diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9834 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.7138122673736808,
+  "eval_steps": 500,
+  "global_step": 14000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005098659052669148,
+      "grad_norm": 251.65310853808273,
+      "learning_rate": 7.640067911714771e-08,
+      "loss": 7.2611,
+      "step": 10
+    },
+    {
+      "epoch": 0.0010197318105338296,
+      "grad_norm": 223.4280231492118,
+      "learning_rate": 1.6129032258064518e-07,
+      "loss": 7.5935,
+      "step": 20
+    },
+    {
+      "epoch": 0.0015295977158007444,
+      "grad_norm": 193.4556196678711,
+      "learning_rate": 2.461799660441426e-07,
+      "loss": 7.1911,
+      "step": 30
+    },
+    {
+      "epoch": 0.002039463621067659,
+      "grad_norm": 154.93275414899998,
+      "learning_rate": 3.310696095076401e-07,
+      "loss": 6.6369,
+      "step": 40
+    },
+    {
+      "epoch": 0.002549329526334574,
+      "grad_norm": 179.53329242395242,
+      "learning_rate": 4.1595925297113753e-07,
+      "loss": 5.5123,
+      "step": 50
+    },
+    {
+      "epoch": 0.0030591954316014887,
+      "grad_norm": 141.11077623847112,
+      "learning_rate": 5.00848896434635e-07,
+      "loss": 4.381,
+      "step": 60
+    },
+    {
+      "epoch": 0.0035690613368684035,
+      "grad_norm": 46.48467601369476,
+      "learning_rate": 5.857385398981324e-07,
+      "loss": 3.1309,
+      "step": 70
+    },
+    {
+      "epoch": 0.004078927242135318,
+      "grad_norm": 53.96813671274727,
+      "learning_rate": 6.706281833616299e-07,
+      "loss": 2.5806,
+      "step": 80
+    },
+    {
+      "epoch": 0.0045887931474022335,
+      "grad_norm": 34.935060044045606,
+      "learning_rate": 7.555178268251275e-07,
+      "loss": 1.699,
+      "step": 90
+    },
+    {
+      "epoch": 0.005098659052669148,
+      "grad_norm": 67.7244290501232,
+      "learning_rate": 8.404074702886249e-07,
+      "loss": 1.2787,
+      "step": 100
+    },
+    {
+      "epoch": 0.005608524957936063,
+      "grad_norm": 48.186462961125564,
+      "learning_rate": 9.252971137521223e-07,
+      "loss": 1.104,
+      "step": 110
+    },
+    {
+      "epoch": 0.0061183908632029774,
+      "grad_norm": 71.68206146928328,
+      "learning_rate": 1.0101867572156197e-06,
+      "loss": 0.9308,
+      "step": 120
+    },
+    {
+      "epoch": 0.006628256768469893,
+      "grad_norm": 84.19717289430702,
+      "learning_rate": 1.0950764006791174e-06,
+      "loss": 0.9437,
+      "step": 130
+    },
+    {
+      "epoch": 0.007138122673736807,
+      "grad_norm": 22.01714846636104,
+      "learning_rate": 1.1799660441426147e-06,
+      "loss": 0.9429,
+      "step": 140
+    },
+    {
+      "epoch": 0.007647988579003722,
+      "grad_norm": 35.42757584454689,
+      "learning_rate": 1.2648556876061122e-06,
+      "loss": 0.865,
+      "step": 150
+    },
+    {
+      "epoch": 0.008157854484270637,
+      "grad_norm": 55.96764262360704,
+      "learning_rate": 1.3497453310696096e-06,
+      "loss": 0.9103,
+      "step": 160
+    },
+    {
+      "epoch": 0.008667720389537552,
+      "grad_norm": 29.05124237209439,
+      "learning_rate": 1.434634974533107e-06,
+      "loss": 0.8996,
+      "step": 170
+    },
+    {
+      "epoch": 0.009177586294804467,
+      "grad_norm": 26.211170552849907,
+      "learning_rate": 1.5195246179966044e-06,
+      "loss": 0.9332,
+      "step": 180
+    },
+    {
+      "epoch": 0.00968745220007138,
+      "grad_norm": 18.947652048653687,
+      "learning_rate": 1.604414261460102e-06,
+      "loss": 0.9318,
+      "step": 190
+    },
+    {
+      "epoch": 0.010197318105338296,
+      "grad_norm": 30.129455976188698,
+      "learning_rate": 1.6893039049235995e-06,
+      "loss": 0.8478,
+      "step": 200
+    },
+    {
+      "epoch": 0.010707184010605211,
+      "grad_norm": 82.99133324029893,
+      "learning_rate": 1.774193548387097e-06,
+      "loss": 0.832,
+      "step": 210
+    },
+    {
+      "epoch": 0.011217049915872126,
+      "grad_norm": 26.23660369155024,
+      "learning_rate": 1.8590831918505945e-06,
+      "loss": 0.8242,
+      "step": 220
+    },
+    {
+      "epoch": 0.01172691582113904,
+      "grad_norm": 23.197604257350026,
+      "learning_rate": 1.943972835314092e-06,
+      "loss": 0.8753,
+      "step": 230
+    },
+    {
+      "epoch": 0.012236781726405955,
+      "grad_norm": 19.501104263510726,
+      "learning_rate": 2.028862478777589e-06,
+      "loss": 0.8768,
+      "step": 240
+    },
+    {
+      "epoch": 0.01274664763167287,
+      "grad_norm": 24.897662656911052,
+      "learning_rate": 2.113752122241087e-06,
+      "loss": 0.869,
+      "step": 250
+    },
+    {
+      "epoch": 0.013256513536939785,
+      "grad_norm": 21.333577680411484,
+      "learning_rate": 2.1986417657045842e-06,
+      "loss": 0.9116,
+      "step": 260
+    },
+    {
+      "epoch": 0.013766379442206699,
+      "grad_norm": 62.90240224505213,
+      "learning_rate": 2.2835314091680816e-06,
+      "loss": 0.841,
+      "step": 270
+    },
+    {
+      "epoch": 0.014276245347473614,
+      "grad_norm": 20.59553192116265,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 0.904,
+      "step": 280
+    },
+    {
+      "epoch": 0.01478611125274053,
+      "grad_norm": 20.161810893917714,
+      "learning_rate": 2.4533106960950766e-06,
+      "loss": 0.8194,
+      "step": 290
+    },
+    {
+      "epoch": 0.015295977158007444,
+      "grad_norm": 16.206444109540872,
+      "learning_rate": 2.538200339558574e-06,
+      "loss": 0.85,
+      "step": 300
+    },
+    {
+      "epoch": 0.015805843063274358,
+      "grad_norm": 24.712582437250955,
+      "learning_rate": 2.6230899830220713e-06,
+      "loss": 0.858,
+      "step": 310
+    },
+    {
+      "epoch": 0.016315708968541273,
+      "grad_norm": 11.762759748871197,
+      "learning_rate": 2.707979626485569e-06,
+      "loss": 0.916,
+      "step": 320
+    },
+    {
+      "epoch": 0.01682557487380819,
+      "grad_norm": 12.553238322483976,
+      "learning_rate": 2.7928692699490667e-06,
+      "loss": 0.8499,
+      "step": 330
+    },
+    {
+      "epoch": 0.017335440779075104,
+      "grad_norm": 16.760823082900814,
+      "learning_rate": 2.877758913412564e-06,
+      "loss": 0.9279,
+      "step": 340
+    },
+    {
+      "epoch": 0.01784530668434202,
+      "grad_norm": 28.504528294646803,
+      "learning_rate": 2.9626485568760614e-06,
+      "loss": 0.9199,
+      "step": 350
+    },
+    {
+      "epoch": 0.018355172589608934,
+      "grad_norm": 14.4930273261429,
+      "learning_rate": 3.0475382003395587e-06,
+      "loss": 0.8613,
+      "step": 360
+    },
+    {
+      "epoch": 0.01886503849487585,
+      "grad_norm": 29.69282833969532,
+      "learning_rate": 3.1324278438030564e-06,
+      "loss": 0.9243,
+      "step": 370
+    },
+    {
+      "epoch": 0.01937490440014276,
+      "grad_norm": 11.766751225603937,
+      "learning_rate": 3.2173174872665538e-06,
+      "loss": 0.8861,
+      "step": 380
+    },
+    {
+      "epoch": 0.019884770305409676,
+      "grad_norm": 14.676218010280364,
+      "learning_rate": 3.302207130730051e-06,
+      "loss": 0.7766,
+      "step": 390
+    },
+    {
+      "epoch": 0.02039463621067659,
+      "grad_norm": 24.371393167451448,
+      "learning_rate": 3.3870967741935484e-06,
+      "loss": 0.8223,
+      "step": 400
+    },
+    {
+      "epoch": 0.020904502115943507,
+      "grad_norm": 43.948339877931836,
+      "learning_rate": 3.471986417657046e-06,
+      "loss": 0.8368,
+      "step": 410
+    },
+    {
+      "epoch": 0.021414368021210422,
+      "grad_norm": 33.83725422441187,
+      "learning_rate": 3.556876061120544e-06,
+      "loss": 0.8302,
+      "step": 420
+    },
+    {
+      "epoch": 0.021924233926477337,
+      "grad_norm": 20.093752196073595,
+      "learning_rate": 3.641765704584041e-06,
+      "loss": 0.904,
+      "step": 430
+    },
+    {
+      "epoch": 0.022434099831744252,
+      "grad_norm": 17.199162746926195,
+      "learning_rate": 3.7266553480475385e-06,
+      "loss": 0.8352,
+      "step": 440
+    },
+    {
+      "epoch": 0.022943965737011168,
+      "grad_norm": 16.80175387256325,
+      "learning_rate": 3.811544991511036e-06,
+      "loss": 0.8199,
+      "step": 450
+    },
+    {
+      "epoch": 0.02345383164227808,
+      "grad_norm": 13.110708597973506,
+      "learning_rate": 3.896434634974533e-06,
+      "loss": 0.8551,
+      "step": 460
+    },
+    {
+      "epoch": 0.023963697547544995,
+      "grad_norm": 14.670368305103178,
+      "learning_rate": 3.981324278438031e-06,
+      "loss": 0.8223,
+      "step": 470
+    },
+    {
+      "epoch": 0.02447356345281191,
+      "grad_norm": 19.68234113783656,
+      "learning_rate": 4.066213921901529e-06,
+      "loss": 0.8651,
+      "step": 480
+    },
+    {
+      "epoch": 0.024983429358078825,
+      "grad_norm": 12.331385026856294,
+      "learning_rate": 4.1511035653650255e-06,
+      "loss": 0.8172,
+      "step": 490
+    },
+    {
+      "epoch": 0.02549329526334574,
+      "grad_norm": 20.223300847863285,
+      "learning_rate": 4.235993208828523e-06,
+      "loss": 0.9137,
+      "step": 500
+    },
+    {
+      "epoch": 0.026003161168612655,
+      "grad_norm": 12.887254567357738,
+      "learning_rate": 4.320882852292021e-06,
+      "loss": 0.8477,
+      "step": 510
+    },
+    {
+      "epoch": 0.02651302707387957,
+      "grad_norm": 31.28097745090668,
+      "learning_rate": 4.405772495755518e-06,
+      "loss": 0.8975,
+      "step": 520
+    },
+    {
+      "epoch": 0.027022892979146486,
+      "grad_norm": 14.546358470787888,
+      "learning_rate": 4.490662139219016e-06,
+      "loss": 0.8026,
+      "step": 530
+    },
+    {
+      "epoch": 0.027532758884413398,
+      "grad_norm": 12.893626716420755,
+      "learning_rate": 4.575551782682513e-06,
+      "loss": 0.8576,
+      "step": 540
+    },
+    {
+      "epoch": 0.028042624789680313,
+      "grad_norm": 15.633191864977086,
+      "learning_rate": 4.66044142614601e-06,
+      "loss": 0.8412,
+      "step": 550
+    },
+    {
+      "epoch": 0.028552490694947228,
+      "grad_norm": 17.734319032596773,
+      "learning_rate": 4.745331069609508e-06,
+      "loss": 0.8408,
+      "step": 560
+    },
+    {
+      "epoch": 0.029062356600214143,
+      "grad_norm": 14.072877238199819,
+      "learning_rate": 4.830220713073006e-06,
+      "loss": 0.8334,
+      "step": 570
+    },
+    {
+      "epoch": 0.02957222250548106,
+      "grad_norm": 15.606824188231757,
+      "learning_rate": 4.915110356536503e-06,
+      "loss": 0.827,
+      "step": 580
+    },
+    {
+      "epoch": 0.030082088410747974,
+      "grad_norm": 38.3978369308503,
+      "learning_rate": 5e-06,
+      "loss": 0.8455,
+      "step": 590
+    },
+    {
+      "epoch": 0.03059195431601489,
+      "grad_norm": 12.840980533572772,
+      "learning_rate": 4.999996591164963e-06,
+      "loss": 0.8228,
+      "step": 600
+    },
+    {
+      "epoch": 0.031101820221281804,
+      "grad_norm": 30.338227470007126,
+      "learning_rate": 4.999986364669145e-06,
+      "loss": 0.7974,
+      "step": 610
+    },
+    {
+      "epoch": 0.031611686126548716,
+      "grad_norm": 18.602004650159476,
+      "learning_rate": 4.999969320540435e-06,
+      "loss": 0.8005,
+      "step": 620
+    },
+    {
+      "epoch": 0.03212155203181563,
+      "grad_norm": 55.38008193860322,
+      "learning_rate": 4.999945458825315e-06,
+      "loss": 0.8082,
+      "step": 630
+    },
+    {
+      "epoch": 0.032631417937082546,
+      "grad_norm": 36.39600509819533,
+      "learning_rate": 4.9999147795888545e-06,
+      "loss": 0.7613,
+      "step": 640
+    },
+    {
+      "epoch": 0.03314128384234946,
+      "grad_norm": 11.625306380518289,
+      "learning_rate": 4.999877282914722e-06,
+      "loss": 0.8314,
+      "step": 650
+    },
+    {
+      "epoch": 0.03365114974761638,
+      "grad_norm": 14.287588770540392,
+      "learning_rate": 4.999832968905171e-06,
+      "loss": 0.8271,
+      "step": 660
+    },
+    {
+      "epoch": 0.03416101565288329,
+      "grad_norm": 19.005445741004017,
+      "learning_rate": 4.999781837681048e-06,
+      "loss": 0.8386,
+      "step": 670
+    },
+    {
+      "epoch": 0.03467088155815021,
+      "grad_norm": 26.674618201945194,
+      "learning_rate": 4.999723889381793e-06,
+      "loss": 0.7765,
+      "step": 680
+    },
+    {
+      "epoch": 0.03518074746341712,
+      "grad_norm": 13.637855817556927,
+      "learning_rate": 4.999659124165434e-06,
+      "loss": 0.8344,
+      "step": 690
+    },
+    {
+      "epoch": 0.03569061336868404,
+      "grad_norm": 21.01966042290558,
+      "learning_rate": 4.999587542208591e-06,
+      "loss": 0.7641,
+      "step": 700
+    },
+    {
+      "epoch": 0.03620047927395095,
+      "grad_norm": 30.965125642767354,
+      "learning_rate": 4.999509143706472e-06,
+      "loss": 0.8969,
+      "step": 710
+    },
+    {
+      "epoch": 0.03671034517921787,
+      "grad_norm": 14.067836644004847,
+      "learning_rate": 4.999423928872876e-06,
+      "loss": 0.8912,
+      "step": 720
+    },
+    {
+      "epoch": 0.03722021108448478,
+      "grad_norm": 22.189111598465264,
+      "learning_rate": 4.999331897940189e-06,
+      "loss": 0.755,
+      "step": 730
+    },
+    {
+      "epoch": 0.0377300769897517,
+      "grad_norm": 23.40728571879822,
+      "learning_rate": 4.999233051159385e-06,
+      "loss": 0.8145,
+      "step": 740
+    },
+    {
+      "epoch": 0.03823994289501861,
+      "grad_norm": 11.14841713621582,
+      "learning_rate": 4.999127388800028e-06,
+      "loss": 0.805,
+      "step": 750
+    },
+    {
+      "epoch": 0.03874980880028552,
+      "grad_norm": 23.128314203665813,
+      "learning_rate": 4.999014911150264e-06,
+      "loss": 0.796,
+      "step": 760
+    },
+    {
+      "epoch": 0.03925967470555244,
+      "grad_norm": 26.492358982029966,
+      "learning_rate": 4.998895618516829e-06,
+      "loss": 0.8336,
+      "step": 770
+    },
+    {
+      "epoch": 0.03976954061081935,
+      "grad_norm": 22.55100498684785,
+      "learning_rate": 4.998769511225041e-06,
+      "loss": 0.766,
+      "step": 780
+    },
+    {
+      "epoch": 0.04027940651608627,
+      "grad_norm": 34.760159393788676,
+      "learning_rate": 4.998636589618803e-06,
+      "loss": 0.8454,
+      "step": 790
+    },
+    {
+      "epoch": 0.04078927242135318,
+      "grad_norm": 14.886385472521404,
+      "learning_rate": 4.998496854060603e-06,
+      "loss": 0.7684,
+      "step": 800
+    },
+    {
+      "epoch": 0.0412991383266201,
+      "grad_norm": 20.35215207377879,
+      "learning_rate": 4.998350304931507e-06,
+      "loss": 0.7806,
+      "step": 810
+    },
+    {
+      "epoch": 0.04180900423188701,
+      "grad_norm": 53.835011257055584,
+      "learning_rate": 4.998196942631166e-06,
+      "loss": 0.7874,
+      "step": 820
+    },
+    {
+      "epoch": 0.04231887013715393,
+      "grad_norm": 19.79335731744932,
+      "learning_rate": 4.998036767577809e-06,
+      "loss": 0.7609,
+      "step": 830
+    },
+    {
+      "epoch": 0.042828736042420844,
+      "grad_norm": 41.363363492847824,
+      "learning_rate": 4.997869780208244e-06,
+      "loss": 0.8284,
+      "step": 840
+    },
+    {
+      "epoch": 0.04333860194768776,
+      "grad_norm": 14.006984552833496,
+      "learning_rate": 4.997695980977858e-06,
+      "loss": 0.784,
+      "step": 850
+    },
+    {
+      "epoch": 0.043848467852954674,
+      "grad_norm": 13.221811726431245,
+      "learning_rate": 4.997515370360613e-06,
+      "loss": 0.8128,
+      "step": 860
+    },
+    {
+      "epoch": 0.04435833375822159,
+      "grad_norm": 38.14389153979424,
+      "learning_rate": 4.997327948849044e-06,
+      "loss": 0.7496,
+      "step": 870
+    },
+    {
+      "epoch": 0.044868199663488505,
+      "grad_norm": 16.60716058264655,
+      "learning_rate": 4.997133716954266e-06,
+      "loss": 0.8008,
+      "step": 880
+    },
+    {
+      "epoch": 0.04537806556875542,
+      "grad_norm": 10.131272882306668,
+      "learning_rate": 4.996932675205961e-06,
+      "loss": 0.7418,
+      "step": 890
+    },
+    {
+      "epoch": 0.045887931474022335,
+      "grad_norm": 19.781623800186882,
+      "learning_rate": 4.996724824152382e-06,
+      "loss": 0.7843,
+      "step": 900
+    },
+    {
+      "epoch": 0.04639779737928925,
+      "grad_norm": 11.205179068676344,
+      "learning_rate": 4.996510164360355e-06,
+      "loss": 0.801,
+      "step": 910
+    },
+    {
+      "epoch": 0.04690766328455616,
+      "grad_norm": 7.133132254705291,
+      "learning_rate": 4.996288696415271e-06,
+      "loss": 0.7436,
+      "step": 920
+    },
+    {
+      "epoch": 0.047417529189823074,
+      "grad_norm": 7.361016575983099,
+      "learning_rate": 4.9960604209210885e-06,
+      "loss": 0.7697,
+      "step": 930
+    },
+    {
+      "epoch": 0.04792739509508999,
+      "grad_norm": 14.875145921702854,
+      "learning_rate": 4.99582533850033e-06,
+      "loss": 0.7774,
+      "step": 940
+    },
+    {
+      "epoch": 0.048437261000356904,
+      "grad_norm": 11.641434144165558,
+      "learning_rate": 4.995583449794081e-06,
+      "loss": 0.7916,
+      "step": 950
+    },
+    {
+      "epoch": 0.04894712690562382,
+      "grad_norm": 12.61376024781143,
+      "learning_rate": 4.99533475546199e-06,
+      "loss": 0.721,
+      "step": 960
+    },
+    {
+      "epoch": 0.049456992810890735,
+      "grad_norm": 10.992135613925235,
+      "learning_rate": 4.995079256182261e-06,
+      "loss": 0.8025,
+      "step": 970
+    },
+    {
+      "epoch": 0.04996685871615765,
+      "grad_norm": 23.077584592510522,
+      "learning_rate": 4.99481695265166e-06,
+      "loss": 0.7668,
+      "step": 980
+    },
+    {
+      "epoch": 0.050476724621424565,
+      "grad_norm": 20.805552814430307,
+      "learning_rate": 4.994547845585505e-06,
+      "loss": 0.7682,
+      "step": 990
+    },
+    {
+      "epoch": 0.05098659052669148,
+      "grad_norm": 37.749437539255986,
+      "learning_rate": 4.994271935717671e-06,
+      "loss": 0.8547,
+      "step": 1000
+    },
+    {
+      "epoch": 0.051496456431958396,
+      "grad_norm": 26.52405401012962,
+      "learning_rate": 4.99398922380058e-06,
+      "loss": 0.8675,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05200632233722531,
+      "grad_norm": 11.386281013139724,
+      "learning_rate": 4.99369971060521e-06,
+      "loss": 0.7781,
+      "step": 1020
+    },
+    {
+      "epoch": 0.052516188242492226,
+      "grad_norm": 23.97388675424806,
+      "learning_rate": 4.993403396921082e-06,
+      "loss": 0.7452,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05302605414775914,
+      "grad_norm": 26.25744423502963,
+      "learning_rate": 4.993100283556262e-06,
+      "loss": 0.7255,
+      "step": 1040
+    },
+    {
+      "epoch": 0.053535920053026056,
+      "grad_norm": 30.57573685033861,
+      "learning_rate": 4.992790371337363e-06,
+      "loss": 0.8226,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05404578595829297,
+      "grad_norm": 11.6456271061302,
+      "learning_rate": 4.992473661109535e-06,
+      "loss": 0.7672,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05455565186355989,
+      "grad_norm": 15.587338850002327,
+      "learning_rate": 4.99215015373647e-06,
+      "loss": 0.7913,
+      "step": 1070
+    },
+    {
+      "epoch": 0.055065517768826795,
+      "grad_norm": 8.306218641787943,
+      "learning_rate": 4.991819850100394e-06,
+      "loss": 0.8413,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05557538367409371,
+      "grad_norm": 13.813826007031592,
+      "learning_rate": 4.991482751102066e-06,
+      "loss": 0.7028,
+      "step": 1090
+    },
+    {
+      "epoch": 0.056085249579360626,
+      "grad_norm": 39.13157110924138,
+      "learning_rate": 4.99113885766078e-06,
+      "loss": 0.7273,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05659511548462754,
+      "grad_norm": 14.998429544518908,
+      "learning_rate": 4.990788170714355e-06,
+      "loss": 0.7754,
+      "step": 1110
+    },
+    {
+      "epoch": 0.057104981389894456,
+      "grad_norm": 37.8467854367974,
+      "learning_rate": 4.99043069121914e-06,
+      "loss": 0.7971,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05761484729516137,
+      "grad_norm": 9.645482089818124,
+      "learning_rate": 4.990066420150005e-06,
+      "loss": 0.7079,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05812471320042829,
+      "grad_norm": 36.873667812807035,
+      "learning_rate": 4.989695358500342e-06,
+      "loss": 0.7912,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0586345791056952,
+      "grad_norm": 19.38112557912608,
+      "learning_rate": 4.989317507282061e-06,
+      "loss": 0.809,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05914444501096212,
+      "grad_norm": 14.049014149225295,
+      "learning_rate": 4.988932867525588e-06,
+      "loss": 0.6839,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05965431091622903,
+      "grad_norm": 12.306775901881515,
+      "learning_rate": 4.988541440279862e-06,
+      "loss": 0.8302,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06016417682149595,
+      "grad_norm": 18.42396715041709,
+      "learning_rate": 4.988143226612333e-06,
+      "loss": 0.7578,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06067404272676286,
+      "grad_norm": 33.067925037551674,
+      "learning_rate": 4.987738227608954e-06,
+      "loss": 0.7718,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06118390863202978,
+      "grad_norm": 19.216201729152147,
+      "learning_rate": 4.987326444374189e-06,
+      "loss": 0.7885,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06169377453729669,
+      "grad_norm": 9.210294599142257,
+      "learning_rate": 4.986907878030994e-06,
+      "loss": 0.6944,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06220364044256361,
+      "grad_norm": 8.483381230394919,
+      "learning_rate": 4.986482529720831e-06,
+      "loss": 0.7622,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06271350634783052,
+      "grad_norm": 22.65788410832388,
+      "learning_rate": 4.986050400603653e-06,
+      "loss": 0.7881,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06322337225309743,
+      "grad_norm": 12.272586735224095,
+      "learning_rate": 4.985611491857906e-06,
+      "loss": 0.7109,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06373323815836435,
+      "grad_norm": 24.66263007900279,
+      "learning_rate": 4.9851658046805226e-06,
+      "loss": 0.715,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06424310406363126,
+      "grad_norm": 14.842375084400794,
+      "learning_rate": 4.9847133402869235e-06,
+      "loss": 0.7369,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06475296996889818,
+      "grad_norm": 22.216116344429377,
+      "learning_rate": 4.984254099911009e-06,
+      "loss": 0.6989,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06526283587416509,
+      "grad_norm": 45.589598576972115,
+      "learning_rate": 4.98378808480516e-06,
+      "loss": 0.7822,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06577270177943201,
+      "grad_norm": 86.77969204698395,
+      "learning_rate": 4.98331529624023e-06,
+      "loss": 0.7097,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06628256768469892,
+      "grad_norm": 17.69167242840336,
+      "learning_rate": 4.982835735505545e-06,
+      "loss": 0.8053,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06679243358996584,
+      "grad_norm": 13.284491837077573,
+      "learning_rate": 4.982349403908902e-06,
+      "loss": 0.7353,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06730229949523275,
+      "grad_norm": 8.48209924744741,
+      "learning_rate": 4.98185630277656e-06,
+      "loss": 0.7611,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06781216540049967,
+      "grad_norm": 8.856486864478638,
+      "learning_rate": 4.981356433453238e-06,
+      "loss": 0.6889,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06832203130576658,
+      "grad_norm": 107.59240891987359,
+      "learning_rate": 4.9808497973021146e-06,
+      "loss": 0.8094,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0688318972110335,
+      "grad_norm": 17.91017719005489,
+      "learning_rate": 4.98033639570482e-06,
+      "loss": 0.7585,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06934176311630041,
+      "grad_norm": 10.058369133706213,
+      "learning_rate": 4.979816230061436e-06,
+      "loss": 0.7193,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06985162902156733,
+      "grad_norm": 20.144328160083482,
+      "learning_rate": 4.97928930179049e-06,
+      "loss": 0.7395,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07036149492683424,
+      "grad_norm": 8.706394083724875,
+      "learning_rate": 4.978755612328951e-06,
+      "loss": 0.7464,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07087136083210116,
+      "grad_norm": 15.894421245682398,
+      "learning_rate": 4.978215163132226e-06,
+      "loss": 0.7485,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07138122673736808,
+      "grad_norm": 62.536082721075196,
+      "learning_rate": 4.977667955674158e-06,
+      "loss": 0.7638,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07189109264263499,
+      "grad_norm": 13.081976664557795,
+      "learning_rate": 4.977113991447017e-06,
+      "loss": 0.731,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0724009585479019,
+      "grad_norm": 15.993828063445294,
+      "learning_rate": 4.976553271961503e-06,
+      "loss": 0.843,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07291082445316882,
+      "grad_norm": 18.85801856028921,
+      "learning_rate": 4.975985798746736e-06,
+      "loss": 0.7829,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07342069035843574,
+      "grad_norm": 15.059259870020968,
+      "learning_rate": 4.975411573350252e-06,
+      "loss": 0.7585,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07393055626370265,
+      "grad_norm": 37.30959187853398,
+      "learning_rate": 4.974830597338004e-06,
+      "loss": 0.7719,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07444042216896957,
+      "grad_norm": 16.804542039630977,
+      "learning_rate": 4.974242872294354e-06,
+      "loss": 0.7128,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07495028807423648,
+      "grad_norm": 11.815174351719946,
+      "learning_rate": 4.973648399822068e-06,
+      "loss": 0.7417,
+      "step": 1470
+    },
+    {
+      "epoch": 0.0754601539795034,
+      "grad_norm": 13.040163493202366,
+      "learning_rate": 4.9730471815423124e-06,
+      "loss": 0.8031,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07597001988477031,
+      "grad_norm": 12.047839519102967,
+      "learning_rate": 4.972439219094649e-06,
+      "loss": 0.7246,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07647988579003721,
+      "grad_norm": 17.021489876135128,
+      "learning_rate": 4.971824514137035e-06,
+      "loss": 0.681,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07698975169530413,
+      "grad_norm": 9.685468993709286,
+      "learning_rate": 4.971203068345811e-06,
+      "loss": 0.732,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07749961760057104,
+      "grad_norm": 15.944959721940311,
+      "learning_rate": 4.970574883415704e-06,
+      "loss": 0.7806,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07800948350583796,
+      "grad_norm": 24.268530969836814,
+      "learning_rate": 4.969939961059814e-06,
+      "loss": 0.7435,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07851934941110487,
+      "grad_norm": 13.030992822878428,
+      "learning_rate": 4.969298303009621e-06,
+      "loss": 0.7274,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07902921531637179,
+      "grad_norm": 13.947966567662132,
+      "learning_rate": 4.968649911014967e-06,
+      "loss": 0.6783,
+      "step": 1550
+    },
+    {
+      "epoch": 0.0795390812216387,
+      "grad_norm": 14.570495534169975,
+      "learning_rate": 4.9679947868440625e-06,
+      "loss": 0.7491,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08004894712690562,
+      "grad_norm": 7.132526560379979,
+      "learning_rate": 4.967332932283476e-06,
+      "loss": 0.7223,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08055881303217254,
+      "grad_norm": 7.875957907008374,
+      "learning_rate": 4.966664349138129e-06,
+      "loss": 0.7368,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08106867893743945,
+      "grad_norm": 14.20041503911161,
+      "learning_rate": 4.9659890392312935e-06,
+      "loss": 0.7542,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08157854484270637,
+      "grad_norm": 37.488065773340224,
+      "learning_rate": 4.965307004404586e-06,
+      "loss": 0.7465,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08208841074797328,
+      "grad_norm": 14.65121446955515,
+      "learning_rate": 4.964618246517962e-06,
+      "loss": 0.7168,
+      "step": 1610
+    },
+    {
+      "epoch": 0.0825982766532402,
+      "grad_norm": 35.950774694044625,
+      "learning_rate": 4.96392276744971e-06,
+      "loss": 0.6847,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08310814255850711,
+      "grad_norm": 25.12291831378265,
+      "learning_rate": 4.9632205690964505e-06,
+      "loss": 0.6943,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08361800846377403,
+      "grad_norm": 8.204862149437655,
+      "learning_rate": 4.962511653373124e-06,
+      "loss": 0.699,
+      "step": 1640
+    },
+    {
+      "epoch": 0.08412787436904094,
+      "grad_norm": 7.350705825407557,
+      "learning_rate": 4.961796022212994e-06,
+      "loss": 0.7053,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08463774027430786,
+      "grad_norm": 11.340599409165993,
+      "learning_rate": 4.961073677567634e-06,
+      "loss": 0.6997,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08514760617957477,
+      "grad_norm": 14.015749677589326,
+      "learning_rate": 4.960344621406927e-06,
+      "loss": 0.8495,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08565747208484169,
+      "grad_norm": 18.004132825340353,
+      "learning_rate": 4.959608855719059e-06,
+      "loss": 0.7093,
+      "step": 1680
+    },
+    {
+      "epoch": 0.0861673379901086,
+      "grad_norm": 12.254937804304983,
+      "learning_rate": 4.958866382510515e-06,
+      "loss": 0.758,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08667720389537552,
+      "grad_norm": 16.92387976586447,
+      "learning_rate": 4.958117203806067e-06,
+      "loss": 0.6934,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08718706980064243,
+      "grad_norm": 10.301187293252587,
+      "learning_rate": 4.957361321648777e-06,
+      "loss": 0.6859,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08769693570590935,
+      "grad_norm": 26.036964612219144,
+      "learning_rate": 4.956598738099988e-06,
+      "loss": 0.751,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08820680161117626,
+      "grad_norm": 19.235345452442886,
+      "learning_rate": 4.955829455239316e-06,
+      "loss": 0.7173,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08871666751644318,
+      "grad_norm": 12.24883597729765,
+      "learning_rate": 4.95505347516465e-06,
+      "loss": 0.742,
+      "step": 1740
+    },
+    {
+      "epoch": 0.0892265334217101,
+      "grad_norm": 14.574956559044802,
+      "learning_rate": 4.954270799992138e-06,
+      "loss": 0.7574,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08973639932697701,
+      "grad_norm": 34.43135113831238,
+      "learning_rate": 4.953481431856189e-06,
+      "loss": 0.7639,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09024626523224392,
+      "grad_norm": 11.963547537782372,
+      "learning_rate": 4.952685372909465e-06,
+      "loss": 0.697,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09075613113751084,
+      "grad_norm": 14.879992437433783,
+      "learning_rate": 4.951882625322871e-06,
+      "loss": 0.7327,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09126599704277776,
+      "grad_norm": 10.687966849821434,
+      "learning_rate": 4.951073191285555e-06,
+      "loss": 0.7173,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09177586294804467,
+      "grad_norm": 173.65543370635663,
+      "learning_rate": 4.9502570730048995e-06,
+      "loss": 0.7795,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09228572885331159,
+      "grad_norm": 33.83810505997628,
+      "learning_rate": 4.949434272706514e-06,
+      "loss": 0.6531,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0927955947585785,
+      "grad_norm": 15.988283096618108,
+      "learning_rate": 4.9486047926342294e-06,
+      "loss": 0.6574,
+      "step": 1820
+    },
+    {
+      "epoch": 0.0933054606638454,
+      "grad_norm": 14.479744812619012,
+      "learning_rate": 4.947768635050098e-06,
+      "loss": 0.7244,
+      "step": 1830
+    },
+    {
+      "epoch": 0.09381532656911232,
+      "grad_norm": 12.06402074877087,
+      "learning_rate": 4.946925802234373e-06,
+      "loss": 0.7125,
+      "step": 1840
+    },
+    {
+      "epoch": 0.09432519247437923,
+      "grad_norm": 8.125406898256518,
+      "learning_rate": 4.946076296485522e-06,
+      "loss": 0.6822,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09483505837964615,
+      "grad_norm": 20.671526831422415,
+      "learning_rate": 4.945220120120203e-06,
+      "loss": 0.6637,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09534492428491306,
+      "grad_norm": 17.716989185929332,
+      "learning_rate": 4.9443572754732675e-06,
+      "loss": 0.6626,
+      "step": 1870
+    },
+    {
+      "epoch": 0.09585479019017998,
+      "grad_norm": 159.8354306714128,
+      "learning_rate": 4.943487764897749e-06,
+      "loss": 0.7197,
+      "step": 1880
+    },
+    {
+      "epoch": 0.0963646560954469,
+      "grad_norm": 12.61619164444456,
+      "learning_rate": 4.942611590764866e-06,
+      "loss": 0.6883,
+      "step": 1890
+    },
+    {
+      "epoch": 0.09687452200071381,
+      "grad_norm": 11.46693117898443,
+      "learning_rate": 4.941728755464003e-06,
+      "loss": 0.7083,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09738438790598072,
+      "grad_norm": 11.479207415736344,
+      "learning_rate": 4.940839261402711e-06,
+      "loss": 0.7645,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09789425381124764,
+      "grad_norm": 16.330332090944445,
+      "learning_rate": 4.939943111006702e-06,
+      "loss": 0.6721,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09840411971651455,
+      "grad_norm": 24.904337194527546,
+      "learning_rate": 4.93904030671984e-06,
+      "loss": 0.7213,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09891398562178147,
+      "grad_norm": 13.987479654668915,
+      "learning_rate": 4.938130851004131e-06,
+      "loss": 0.8048,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09942385152704838,
+      "grad_norm": 27.075097531744582,
+      "learning_rate": 4.937214746339726e-06,
+      "loss": 0.7305,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0999337174323153,
+      "grad_norm": 8.923073340927486,
+      "learning_rate": 4.936291995224902e-06,
+      "loss": 0.7636,
+      "step": 1960
+    },
+    {
+      "epoch": 0.10044358333758222,
+      "grad_norm": 7.439882170995172,
+      "learning_rate": 4.935362600176064e-06,
+      "loss": 0.7253,
+      "step": 1970
+    },
+    {
+      "epoch": 0.10095344924284913,
+      "grad_norm": 4.931935226467838,
+      "learning_rate": 4.934426563727739e-06,
+      "loss": 0.719,
+      "step": 1980
+    },
+    {
+      "epoch": 0.10146331514811605,
+      "grad_norm": 13.417900875214697,
+      "learning_rate": 4.933483888432558e-06,
+      "loss": 0.7065,
+      "step": 1990
+    },
+    {
+      "epoch": 0.10197318105338296,
+      "grad_norm": 12.076546088417926,
+      "learning_rate": 4.932534576861263e-06,
+      "loss": 0.7119,
+      "step": 2000
+    },
+    {
+      "epoch": 0.10248304695864988,
+      "grad_norm": 9.339538376806033,
+      "learning_rate": 4.931578631602691e-06,
+      "loss": 0.7747,
+      "step": 2010
+    },
+    {
+      "epoch": 0.10299291286391679,
+      "grad_norm": 9.290207456660328,
+      "learning_rate": 4.930616055263768e-06,
+      "loss": 0.7296,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1035027787691837,
+      "grad_norm": 13.938471789842117,
+      "learning_rate": 4.9296468504695075e-06,
+      "loss": 0.716,
+      "step": 2030
+    },
+    {
+      "epoch": 0.10401264467445062,
+      "grad_norm": 7.794410431885056,
+      "learning_rate": 4.928671019862995e-06,
+      "loss": 0.7138,
+      "step": 2040
+    },
+    {
+      "epoch": 0.10452251057971754,
+      "grad_norm": 7.090928909024373,
+      "learning_rate": 4.927688566105388e-06,
+      "loss": 0.7331,
+      "step": 2050
+    },
+    {
+      "epoch": 0.10503237648498445,
+      "grad_norm": 17.67001545795707,
+      "learning_rate": 4.926699491875905e-06,
+      "loss": 0.758,
+      "step": 2060
+    },
+    {
+      "epoch": 0.10554224239025137,
+      "grad_norm": 18.577244712263273,
+      "learning_rate": 4.925703799871818e-06,
+      "loss": 0.7324,
+      "step": 2070
+    },
+    {
+      "epoch": 0.10605210829551828,
+      "grad_norm": 11.69929635491638,
+      "learning_rate": 4.924701492808447e-06,
+      "loss": 0.7171,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1065619742007852,
+      "grad_norm": 7.939657680078957,
+      "learning_rate": 4.923692573419152e-06,
+      "loss": 0.7254,
+      "step": 2090
+    },
+    {
+      "epoch": 0.10707184010605211,
+      "grad_norm": 13.41877483458534,
+      "learning_rate": 4.922677044455324e-06,
+      "loss": 0.8326,
+      "step": 2100
+    },
+    {
+      "epoch": 0.10758170601131903,
+      "grad_norm": 9.545281303803277,
+      "learning_rate": 4.921654908686381e-06,
+      "loss": 0.7448,
+      "step": 2110
+    },
+    {
+      "epoch": 0.10809157191658594,
+      "grad_norm": 7.3352190302911,
+      "learning_rate": 4.920626168899755e-06,
+      "loss": 0.7348,
+      "step": 2120
+    },
+    {
+      "epoch": 0.10860143782185286,
+      "grad_norm": 26.685965241526155,
+      "learning_rate": 4.91959082790089e-06,
+      "loss": 0.7507,
+      "step": 2130
+    },
+    {
+      "epoch": 0.10911130372711977,
+      "grad_norm": 12.909643886403847,
+      "learning_rate": 4.918548888513232e-06,
+      "loss": 0.651,
+      "step": 2140
+    },
+    {
+      "epoch": 0.10962116963238669,
+      "grad_norm": 15.646611507641264,
+      "learning_rate": 4.91750035357822e-06,
+      "loss": 0.6946,
+      "step": 2150
+    },
+    {
+      "epoch": 0.11013103553765359,
+      "grad_norm": 12.998595411207628,
+      "learning_rate": 4.9164452259552805e-06,
+      "loss": 0.7434,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1106409014429205,
+      "grad_norm": 11.660405346251727,
+      "learning_rate": 4.9153835085218175e-06,
+      "loss": 0.7105,
+      "step": 2170
+    },
+    {
+      "epoch": 0.11115076734818742,
+      "grad_norm": 14.143309967401331,
+      "learning_rate": 4.9143152041732074e-06,
+      "loss": 0.7133,
+      "step": 2180
+    },
+    {
+      "epoch": 0.11166063325345434,
+      "grad_norm": 19.03028150637859,
+      "learning_rate": 4.91324031582279e-06,
+      "loss": 0.7891,
+      "step": 2190
+    },
+    {
+      "epoch": 0.11217049915872125,
+      "grad_norm": 16.352679932832064,
+      "learning_rate": 4.9121588464018555e-06,
+      "loss": 0.7343,
+      "step": 2200
+    },
+    {
+      "epoch": 0.11268036506398817,
+      "grad_norm": 13.050194377945338,
+      "learning_rate": 4.911070798859647e-06,
+      "loss": 0.6784,
+      "step": 2210
+    },
+    {
+      "epoch": 0.11319023096925508,
+      "grad_norm": 16.676160812772732,
+      "learning_rate": 4.909976176163345e-06,
+      "loss": 0.7425,
+      "step": 2220
+    },
+    {
+      "epoch": 0.113700096874522,
+      "grad_norm": 24.842180760788374,
+      "learning_rate": 4.908874981298058e-06,
+      "loss": 0.7591,
+      "step": 2230
+    },
+    {
+      "epoch": 0.11420996277978891,
+      "grad_norm": 27.043355467028597,
+      "learning_rate": 4.90776721726682e-06,
+      "loss": 0.7038,
+      "step": 2240
+    },
+    {
+      "epoch": 0.11471982868505583,
+      "grad_norm": 17.602292546113226,
+      "learning_rate": 4.90665288709058e-06,
+      "loss": 0.7011,
+      "step": 2250
+    },
+    {
+      "epoch": 0.11522969459032274,
+      "grad_norm": 9.491025673003486,
+      "learning_rate": 4.905531993808191e-06,
+      "loss": 0.6496,
+      "step": 2260
+    },
+    {
+      "epoch": 0.11573956049558966,
+      "grad_norm": 10.245512707949505,
+      "learning_rate": 4.904404540476405e-06,
+      "loss": 0.7504,
+      "step": 2270
+    },
+    {
+      "epoch": 0.11624942640085657,
+      "grad_norm": 12.898093672630402,
+      "learning_rate": 4.903270530169865e-06,
+      "loss": 0.7097,
+      "step": 2280
+    },
+    {
+      "epoch": 0.11675929230612349,
+      "grad_norm": 6.129617674226962,
+      "learning_rate": 4.902129965981094e-06,
+      "loss": 0.6947,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1172691582113904,
+      "grad_norm": 11.65520326626469,
+      "learning_rate": 4.900982851020487e-06,
+      "loss": 0.7449,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11777902411665732,
+      "grad_norm": 13.834178026043013,
+      "learning_rate": 4.899829188416306e-06,
+      "loss": 0.6752,
+      "step": 2310
+    },
+    {
+      "epoch": 0.11828889002192423,
+      "grad_norm": 12.848364562587328,
+      "learning_rate": 4.898668981314667e-06,
+      "loss": 0.6874,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11879875592719115,
+      "grad_norm": 20.406006817456138,
+      "learning_rate": 4.897502232879533e-06,
+      "loss": 0.7308,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11930862183245806,
+      "grad_norm": 8.308281701307099,
+      "learning_rate": 4.896328946292706e-06,
+      "loss": 0.6662,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11981848773772498,
+      "grad_norm": 34.003846931457545,
+      "learning_rate": 4.895149124753821e-06,
+      "loss": 0.7199,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1203283536429919,
+      "grad_norm": 17.039906833615753,
+      "learning_rate": 4.893962771480329e-06,
+      "loss": 0.7301,
+      "step": 2360
+    },
+    {
+      "epoch": 0.12083821954825881,
+      "grad_norm": 7.361968459417762,
+      "learning_rate": 4.892769889707497e-06,
+      "loss": 0.7211,
+      "step": 2370
+    },
+    {
+      "epoch": 0.12134808545352573,
+      "grad_norm": 11.789122532853575,
+      "learning_rate": 4.891570482688395e-06,
+      "loss": 0.6497,
+      "step": 2380
+    },
+    {
+      "epoch": 0.12185795135879264,
+      "grad_norm": 30.87304515963633,
+      "learning_rate": 4.890364553693886e-06,
+      "loss": 0.7441,
+      "step": 2390
+    },
+    {
+      "epoch": 0.12236781726405956,
+      "grad_norm": 29.669298108318465,
+      "learning_rate": 4.889152106012623e-06,
+      "loss": 0.7157,
+      "step": 2400
+    },
+    {
+      "epoch": 0.12287768316932647,
+      "grad_norm": 11.71514161705899,
+      "learning_rate": 4.88793314295103e-06,
+      "loss": 0.6329,
+      "step": 2410
+    },
+    {
+      "epoch": 0.12338754907459339,
+      "grad_norm": 20.714383976038274,
+      "learning_rate": 4.886707667833306e-06,
+      "loss": 0.7138,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1238974149798603,
+      "grad_norm": 6.276871926336472,
+      "learning_rate": 4.885475684001401e-06,
+      "loss": 0.6407,
+      "step": 2430
+    },
+    {
+      "epoch": 0.12440728088512722,
+      "grad_norm": 4.832640606709299,
+      "learning_rate": 4.884237194815023e-06,
+      "loss": 0.6979,
+      "step": 2440
+    },
+    {
+      "epoch": 0.12491714679039413,
+      "grad_norm": 9.349322693892326,
+      "learning_rate": 4.882992203651613e-06,
+      "loss": 0.7813,
+      "step": 2450
+    },
+    {
+      "epoch": 0.12542701269566103,
+      "grad_norm": 7.413743472965248,
+      "learning_rate": 4.881740713906348e-06,
+      "loss": 0.7791,
+      "step": 2460
+    },
+    {
+      "epoch": 0.12593687860092795,
+      "grad_norm": 5.503594248421507,
+      "learning_rate": 4.880482728992126e-06,
+      "loss": 0.8045,
+      "step": 2470
+    },
+    {
+      "epoch": 0.12644674450619486,
+      "grad_norm": 12.070817085771614,
+      "learning_rate": 4.8792182523395555e-06,
+      "loss": 0.6917,
+      "step": 2480
+    },
+    {
+      "epoch": 0.12695661041146178,
+      "grad_norm": 15.300633215523083,
+      "learning_rate": 4.877947287396952e-06,
+      "loss": 0.6394,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1274664763167287,
+      "grad_norm": 19.470260448454457,
+      "learning_rate": 4.876669837630324e-06,
+      "loss": 0.6972,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1279763422219956,
+      "grad_norm": 17.203397243947155,
+      "learning_rate": 4.875385906523361e-06,
+      "loss": 0.7206,
+      "step": 2510
+    },
+    {
+      "epoch": 0.12848620812726252,
+      "grad_norm": 14.59124108252345,
+      "learning_rate": 4.874095497577434e-06,
+      "loss": 0.6691,
+      "step": 2520
+    },
+    {
+      "epoch": 0.12899607403252944,
+      "grad_norm": 6.680251131013468,
+      "learning_rate": 4.872798614311574e-06,
+      "loss": 0.6992,
+      "step": 2530
+    },
+    {
+      "epoch": 0.12950593993779635,
+      "grad_norm": 9.142831606794571,
+      "learning_rate": 4.87149526026247e-06,
+      "loss": 0.7222,
+      "step": 2540
+    },
+    {
+      "epoch": 0.13001580584306327,
+      "grad_norm": 11.653155354293201,
+      "learning_rate": 4.870185438984458e-06,
+      "loss": 0.7009,
+      "step": 2550
+    },
+    {
+      "epoch": 0.13052567174833019,
+      "grad_norm": 6.406235878833066,
+      "learning_rate": 4.868869154049509e-06,
+      "loss": 0.6904,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1310355376535971,
+      "grad_norm": 8.058447568972042,
+      "learning_rate": 4.867546409047221e-06,
+      "loss": 0.7581,
+      "step": 2570
+    },
+    {
+      "epoch": 0.13154540355886402,
+      "grad_norm": 6.710720353699215,
+      "learning_rate": 4.866217207584811e-06,
+      "loss": 0.6432,
+      "step": 2580
+    },
+    {
+      "epoch": 0.13205526946413093,
+      "grad_norm": 7.407470629761652,
+      "learning_rate": 4.864881553287101e-06,
+      "loss": 0.6199,
+      "step": 2590
+    },
+    {
+      "epoch": 0.13256513536939785,
+      "grad_norm": 11.56809676811302,
+      "learning_rate": 4.863539449796511e-06,
+      "loss": 0.7045,
+      "step": 2600
+    },
+    {
+      "epoch": 0.13307500127466476,
+      "grad_norm": 8.306230590884352,
+      "learning_rate": 4.86219090077305e-06,
+      "loss": 0.671,
+      "step": 2610
+    },
+    {
+      "epoch": 0.13358486717993168,
+      "grad_norm": 13.957448529489147,
+      "learning_rate": 4.8608359098943014e-06,
+      "loss": 0.6508,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1340947330851986,
+      "grad_norm": 7.784662376461,
+      "learning_rate": 4.859474480855417e-06,
+      "loss": 0.6498,
+      "step": 2630
+    },
+    {
+      "epoch": 0.1346045989904655,
+      "grad_norm": 29.944496551875012,
+      "learning_rate": 4.858106617369108e-06,
+      "loss": 0.7051,
+      "step": 2640
+    },
+    {
+      "epoch": 0.13511446489573242,
+      "grad_norm": 10.71685346194294,
+      "learning_rate": 4.85673232316563e-06,
+      "loss": 0.7297,
+      "step": 2650
+    },
+    {
+      "epoch": 0.13562433080099934,
+      "grad_norm": 49.808581866792125,
+      "learning_rate": 4.855351601992777e-06,
+      "loss": 0.6998,
+      "step": 2660
+    },
+    {
+      "epoch": 0.13613419670626625,
+      "grad_norm": 13.335763810361753,
+      "learning_rate": 4.853964457615871e-06,
+      "loss": 0.6777,
+      "step": 2670
+    },
+    {
+      "epoch": 0.13664406261153317,
+      "grad_norm": 9.028936515529828,
+      "learning_rate": 4.852570893817747e-06,
+      "loss": 0.6742,
+      "step": 2680
+    },
+    {
+      "epoch": 0.13715392851680008,
+      "grad_norm": 11.470738059837727,
+      "learning_rate": 4.851170914398749e-06,
+      "loss": 0.671,
+      "step": 2690
+    },
+    {
+      "epoch": 0.137663794422067,
+      "grad_norm": 32.73029360165739,
+      "learning_rate": 4.849764523176716e-06,
+      "loss": 0.7377,
+      "step": 2700
+    },
+    {
+      "epoch": 0.1381736603273339,
+      "grad_norm": 12.819394755469444,
+      "learning_rate": 4.848351723986974e-06,
+      "loss": 0.7482,
+      "step": 2710
+    },
+    {
+      "epoch": 0.13868352623260083,
+      "grad_norm": 7.937408615887619,
+      "learning_rate": 4.84693252068232e-06,
+      "loss": 0.6308,
+      "step": 2720
+    },
+    {
+      "epoch": 0.13919339213786774,
+      "grad_norm": 7.448706904228568,
+      "learning_rate": 4.845506917133021e-06,
+      "loss": 0.6944,
+      "step": 2730
+    },
+    {
+      "epoch": 0.13970325804313466,
+      "grad_norm": 7.180411136542497,
+      "learning_rate": 4.844074917226792e-06,
+      "loss": 0.6722,
+      "step": 2740
+    },
+    {
+      "epoch": 0.14021312394840157,
+      "grad_norm": 11.468772906657533,
+      "learning_rate": 4.842636524868796e-06,
+      "loss": 0.6626,
+      "step": 2750
+    },
+    {
+      "epoch": 0.1407229898536685,
+      "grad_norm": 6.371452808310344,
+      "learning_rate": 4.8411917439816245e-06,
+      "loss": 0.6499,
+      "step": 2760
+    },
+    {
+      "epoch": 0.1412328557589354,
+      "grad_norm": 19.49041066130283,
+      "learning_rate": 4.839740578505297e-06,
+      "loss": 0.7403,
+      "step": 2770
+    },
+    {
+      "epoch": 0.14174272166420232,
+      "grad_norm": 12.99945610718753,
+      "learning_rate": 4.838283032397237e-06,
+      "loss": 0.7243,
+      "step": 2780
+    },
+    {
+      "epoch": 0.14225258756946924,
+      "grad_norm": 17.642109740820253,
+      "learning_rate": 4.8368191096322734e-06,
+      "loss": 0.7556,
+      "step": 2790
+    },
+    {
+      "epoch": 0.14276245347473615,
+      "grad_norm": 3.9040981516739803,
+      "learning_rate": 4.835348814202624e-06,
+      "loss": 0.7256,
+      "step": 2800
+    },
+    {
+      "epoch": 0.14327231938000307,
+      "grad_norm": 71.87809054881508,
+      "learning_rate": 4.833872150117883e-06,
+      "loss": 0.7095,
+      "step": 2810
+    },
+    {
+      "epoch": 0.14378218528526998,
+      "grad_norm": 6.094260739225736,
+      "learning_rate": 4.832389121405013e-06,
+      "loss": 0.7223,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1442920511905369,
+      "grad_norm": 8.473852645006467,
+      "learning_rate": 4.830899732108337e-06,
+      "loss": 0.7042,
+      "step": 2830
+    },
+    {
+      "epoch": 0.1448019170958038,
+      "grad_norm": 7.816652193027694,
+      "learning_rate": 4.829403986289519e-06,
+      "loss": 0.7272,
+      "step": 2840
+    },
+    {
+      "epoch": 0.14531178300107073,
+      "grad_norm": 17.989801416107305,
+      "learning_rate": 4.827901888027561e-06,
+      "loss": 0.6549,
+      "step": 2850
+    },
+    {
+      "epoch": 0.14582164890633764,
+      "grad_norm": 13.395657682163579,
+      "learning_rate": 4.826393441418785e-06,
+      "loss": 0.7444,
+      "step": 2860
+    },
+    {
+      "epoch": 0.14633151481160456,
+      "grad_norm": 8.540968081865213,
+      "learning_rate": 4.824878650576829e-06,
+      "loss": 0.7088,
+      "step": 2870
+    },
+    {
+      "epoch": 0.14684138071687147,
+      "grad_norm": 8.779541567306202,
+      "learning_rate": 4.823357519632631e-06,
+      "loss": 0.6694,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1473512466221384,
+      "grad_norm": 8.84843067718907,
+      "learning_rate": 4.821830052734418e-06,
+      "loss": 0.6761,
+      "step": 2890
+    },
+    {
+      "epoch": 0.1478611125274053,
+      "grad_norm": 8.526698581121803,
+      "learning_rate": 4.820296254047695e-06,
+      "loss": 0.7178,
+      "step": 2900
+    },
+    {
+      "epoch": 0.14837097843267222,
+      "grad_norm": 13.165812013075216,
+      "learning_rate": 4.8187561277552376e-06,
+      "loss": 0.7361,
+      "step": 2910
+    },
+    {
+      "epoch": 0.14888084433793913,
+      "grad_norm": 17.809515537069476,
+      "learning_rate": 4.817209678057073e-06,
+      "loss": 0.7227,
+      "step": 2920
+    },
+    {
+      "epoch": 0.14939071024320605,
+      "grad_norm": 5.14603155282455,
+      "learning_rate": 4.815656909170476e-06,
+      "loss": 0.6763,
+      "step": 2930
+    },
+    {
+      "epoch": 0.14990057614847296,
+      "grad_norm": 11.830391658586644,
+      "learning_rate": 4.814097825329953e-06,
+      "loss": 0.7027,
+      "step": 2940
+    },
+    {
+      "epoch": 0.15041044205373988,
+      "grad_norm": 13.309313102975352,
+      "learning_rate": 4.81253243078723e-06,
+      "loss": 0.6576,
+      "step": 2950
+    },
+    {
+      "epoch": 0.1509203079590068,
+      "grad_norm": 14.121127903581213,
+      "learning_rate": 4.810960729811247e-06,
+      "loss": 0.7173,
+      "step": 2960
+    },
+    {
+      "epoch": 0.1514301738642737,
+      "grad_norm": 12.247498251382844,
+      "learning_rate": 4.8093827266881375e-06,
+      "loss": 0.6516,
+      "step": 2970
+    },
+    {
+      "epoch": 0.15194003976954062,
+      "grad_norm": 11.887753993952547,
+      "learning_rate": 4.807798425721224e-06,
+      "loss": 0.6781,
+      "step": 2980
+    },
+    {
+      "epoch": 0.1524499056748075,
+      "grad_norm": 10.393760480263255,
+      "learning_rate": 4.806207831231004e-06,
+      "loss": 0.6878,
+      "step": 2990
+    },
+    {
+      "epoch": 0.15295977158007443,
+      "grad_norm": 15.180506006866409,
+      "learning_rate": 4.804610947555135e-06,
+      "loss": 0.6866,
+      "step": 3000
+    },
+    {
+      "epoch": 0.15346963748534134,
+      "grad_norm": 8.098256391581655,
+      "learning_rate": 4.803007779048428e-06,
+      "loss": 0.6332,
+      "step": 3010
+    },
+    {
+      "epoch": 0.15397950339060826,
+      "grad_norm": 18.217729986739197,
+      "learning_rate": 4.801398330082834e-06,
+      "loss": 0.7395,
+      "step": 3020
+    },
+    {
+      "epoch": 0.15448936929587517,
+      "grad_norm": 8.190425991582323,
+      "learning_rate": 4.7997826050474284e-06,
+      "loss": 0.6646,
+      "step": 3030
+    },
+    {
+      "epoch": 0.1549992352011421,
+      "grad_norm": 11.396659175805825,
+      "learning_rate": 4.798160608348404e-06,
+      "loss": 0.7352,
+      "step": 3040
+    },
+    {
+      "epoch": 0.155509101106409,
+      "grad_norm": 7.578872856762376,
+      "learning_rate": 4.796532344409055e-06,
+      "loss": 0.6937,
+      "step": 3050
+    },
+    {
+      "epoch": 0.15601896701167592,
+      "grad_norm": 12.65465529170778,
+      "learning_rate": 4.794897817669769e-06,
+      "loss": 0.7079,
+      "step": 3060
+    },
+    {
+      "epoch": 0.15652883291694283,
+      "grad_norm": 8.332077761823149,
+      "learning_rate": 4.7932570325880114e-06,
+      "loss": 0.6553,
+      "step": 3070
+    },
+    {
+      "epoch": 0.15703869882220975,
+      "grad_norm": 7.493874145117778,
+      "learning_rate": 4.791609993638315e-06,
+      "loss": 0.6221,
+      "step": 3080
+    },
+    {
+      "epoch": 0.15754856472747666,
+      "grad_norm": 9.006760792815905,
+      "learning_rate": 4.789956705312266e-06,
+      "loss": 0.7162,
+      "step": 3090
+    },
+    {
+      "epoch": 0.15805843063274358,
+      "grad_norm": 6.626245426845903,
+      "learning_rate": 4.7882971721184955e-06,
+      "loss": 0.6945,
+      "step": 3100
+    },
+    {
+      "epoch": 0.1585682965380105,
+      "grad_norm": 5.003198013650634,
+      "learning_rate": 4.786631398582663e-06,
+      "loss": 0.6242,
+      "step": 3110
+    },
+    {
+      "epoch": 0.1590781624432774,
+      "grad_norm": 70.12439147426733,
+      "learning_rate": 4.784959389247445e-06,
+      "loss": 0.694,
+      "step": 3120
+    },
+    {
+      "epoch": 0.15958802834854433,
+      "grad_norm": 21.090289465260433,
+      "learning_rate": 4.7832811486725275e-06,
+      "loss": 0.7578,
+      "step": 3130
+    },
+    {
+      "epoch": 0.16009789425381124,
+      "grad_norm": 11.217345377559456,
+      "learning_rate": 4.781596681434584e-06,
+      "loss": 0.6675,
+      "step": 3140
+    },
+    {
+      "epoch": 0.16060776015907816,
+      "grad_norm": 27.493496228858692,
+      "learning_rate": 4.779905992127273e-06,
+      "loss": 0.6608,
+      "step": 3150
+    },
+    {
+      "epoch": 0.16111762606434507,
+      "grad_norm": 7.606733875884907,
+      "learning_rate": 4.778209085361217e-06,
+      "loss": 0.759,
+      "step": 3160
+    },
+    {
+      "epoch": 0.16162749196961199,
+      "grad_norm": 6.704916972776797,
+      "learning_rate": 4.776505965763999e-06,
+      "loss": 0.6207,
+      "step": 3170
+    },
+    {
+      "epoch": 0.1621373578748789,
+      "grad_norm": 6.560606605224566,
+      "learning_rate": 4.77479663798014e-06,
+      "loss": 0.7267,
+      "step": 3180
+    },
+    {
+      "epoch": 0.16264722378014582,
+      "grad_norm": 6.665804741801049,
+      "learning_rate": 4.773081106671094e-06,
+      "loss": 0.7096,
+      "step": 3190
+    },
+    {
+      "epoch": 0.16315708968541273,
+      "grad_norm": 19.678768608251875,
+      "learning_rate": 4.771359376515231e-06,
+      "loss": 0.7245,
+      "step": 3200
+    },
+    {
+      "epoch": 0.16366695559067965,
+      "grad_norm": 7.586591364128146,
+      "learning_rate": 4.769631452207828e-06,
+      "loss": 0.6273,
+      "step": 3210
+    },
+    {
+      "epoch": 0.16417682149594656,
+      "grad_norm": 17.479097758456334,
+      "learning_rate": 4.76789733846105e-06,
+      "loss": 0.6805,
+      "step": 3220
+    },
+    {
+      "epoch": 0.16468668740121348,
+      "grad_norm": 6.709811820698195,
+      "learning_rate": 4.766157040003944e-06,
+      "loss": 0.5998,
+      "step": 3230
+    },
+    {
+      "epoch": 0.1651965533064804,
+      "grad_norm": 10.454340440990537,
+      "learning_rate": 4.7644105615824226e-06,
+      "loss": 0.6798,
+      "step": 3240
+    },
+    {
+      "epoch": 0.1657064192117473,
+      "grad_norm": 51.363584768392315,
+      "learning_rate": 4.7626579079592504e-06,
+      "loss": 0.645,
+      "step": 3250
+    },
+    {
+      "epoch": 0.16621628511701422,
+      "grad_norm": 8.002930949956061,
+      "learning_rate": 4.760899083914035e-06,
+      "loss": 0.6527,
+      "step": 3260
+    },
+    {
+      "epoch": 0.16672615102228114,
+      "grad_norm": 9.644972860936063,
+      "learning_rate": 4.759134094243206e-06,
+      "loss": 0.696,
+      "step": 3270
+    },
+    {
+      "epoch": 0.16723601692754805,
+      "grad_norm": 7.7053569212568265,
+      "learning_rate": 4.757362943760013e-06,
+      "loss": 0.6,
+      "step": 3280
+    },
+    {
+      "epoch": 0.16774588283281497,
+      "grad_norm": 18.816000872116042,
+      "learning_rate": 4.755585637294503e-06,
+      "loss": 0.6497,
+      "step": 3290
+    },
+    {
+      "epoch": 0.16825574873808188,
+      "grad_norm": 9.174567143984415,
+      "learning_rate": 4.753802179693512e-06,
+      "loss": 0.6686,
+      "step": 3300
+    },
+    {
+      "epoch": 0.1687656146433488,
+      "grad_norm": 18.969088719334287,
+      "learning_rate": 4.7520125758206495e-06,
+      "loss": 0.6925,
+      "step": 3310
+    },
+    {
+      "epoch": 0.16927548054861571,
+      "grad_norm": 5.900876067555642,
+      "learning_rate": 4.750216830556287e-06,
+      "loss": 0.6485,
+      "step": 3320
+    },
+    {
+      "epoch": 0.16978534645388263,
+      "grad_norm": 6.471858950566868,
+      "learning_rate": 4.748414948797545e-06,
+      "loss": 0.695,
+      "step": 3330
+    },
+    {
+      "epoch": 0.17029521235914954,
+      "grad_norm": 15.140328131097245,
+      "learning_rate": 4.746606935458277e-06,
+      "loss": 0.6557,
+      "step": 3340
+    },
+    {
+      "epoch": 0.17080507826441646,
+      "grad_norm": 17.19040083808326,
+      "learning_rate": 4.744792795469058e-06,
+      "loss": 0.7159,
+      "step": 3350
+    },
+    {
+      "epoch": 0.17131494416968338,
+      "grad_norm": 9.99786870874283,
+      "learning_rate": 4.742972533777172e-06,
+      "loss": 0.6673,
+      "step": 3360
+    },
+    {
+      "epoch": 0.1718248100749503,
+      "grad_norm": 7.203764779303473,
+      "learning_rate": 4.741146155346596e-06,
+      "loss": 0.6452,
+      "step": 3370
+    },
+    {
+      "epoch": 0.1723346759802172,
+      "grad_norm": 6.320161749809006,
+      "learning_rate": 4.739313665157988e-06,
+      "loss": 0.6395,
+      "step": 3380
+    },
+    {
+      "epoch": 0.17284454188548412,
+      "grad_norm": 11.183461388166837,
+      "learning_rate": 4.737475068208673e-06,
+      "loss": 0.6887,
+      "step": 3390
+    },
+    {
+      "epoch": 0.17335440779075104,
+      "grad_norm": 22.535580807371765,
+      "learning_rate": 4.7356303695126315e-06,
+      "loss": 0.6866,
+      "step": 3400
+    },
+    {
+      "epoch": 0.17386427369601795,
+      "grad_norm": 9.11855222839424,
+      "learning_rate": 4.733779574100482e-06,
+      "loss": 0.6396,
+      "step": 3410
+    },
+    {
+      "epoch": 0.17437413960128487,
+      "grad_norm": 16.74772610855082,
+      "learning_rate": 4.7319226870194676e-06,
+      "loss": 0.675,
+      "step": 3420
+    },
+    {
+      "epoch": 0.17488400550655178,
+      "grad_norm": 10.312136704375964,
+      "learning_rate": 4.730059713333448e-06,
+      "loss": 0.7076,
+      "step": 3430
+    },
+    {
+      "epoch": 0.1753938714118187,
+      "grad_norm": 23.686147104032347,
+      "learning_rate": 4.728190658122878e-06,
+      "loss": 0.5949,
+      "step": 3440
+    },
+    {
+      "epoch": 0.1759037373170856,
+      "grad_norm": 12.583921383761359,
+      "learning_rate": 4.726315526484799e-06,
+      "loss": 0.6345,
+      "step": 3450
+    },
+    {
+      "epoch": 0.17641360322235253,
+      "grad_norm": 8.080013852290937,
+      "learning_rate": 4.724434323532822e-06,
+      "loss": 0.6328,
+      "step": 3460
+    },
+    {
+      "epoch": 0.17692346912761944,
+      "grad_norm": 19.18676141068983,
+      "learning_rate": 4.722547054397114e-06,
+      "loss": 0.6573,
+      "step": 3470
+    },
+    {
+      "epoch": 0.17743333503288636,
+      "grad_norm": 13.275647879472974,
+      "learning_rate": 4.720653724224389e-06,
+      "loss": 0.6805,
+      "step": 3480
+    },
+    {
+      "epoch": 0.17794320093815327,
+      "grad_norm": 14.808744880492414,
+      "learning_rate": 4.718754338177887e-06,
+      "loss": 0.611,
+      "step": 3490
+    },
+    {
+      "epoch": 0.1784530668434202,
+      "grad_norm": 8.849696816700126,
+      "learning_rate": 4.716848901437361e-06,
+      "loss": 0.6432,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1789629327486871,
+      "grad_norm": 7.567552114191146,
+      "learning_rate": 4.714937419199067e-06,
+      "loss": 0.6534,
+      "step": 3510
+    },
+    {
+      "epoch": 0.17947279865395402,
+      "grad_norm": 13.33853831817063,
+      "learning_rate": 4.713019896675749e-06,
+      "loss": 0.7478,
+      "step": 3520
+    },
+    {
+      "epoch": 0.17998266455922093,
+      "grad_norm": 7.1558035745268365,
+      "learning_rate": 4.711096339096619e-06,
+      "loss": 0.6852,
+      "step": 3530
+    },
+    {
+      "epoch": 0.18049253046448785,
+      "grad_norm": 11.762580657805662,
+      "learning_rate": 4.709166751707351e-06,
+      "loss": 0.6843,
+      "step": 3540
+    },
+    {
+      "epoch": 0.18100239636975476,
+      "grad_norm": 14.327525728374841,
+      "learning_rate": 4.7072311397700605e-06,
+      "loss": 0.6476,
+      "step": 3550
+    },
+    {
+      "epoch": 0.18151226227502168,
+      "grad_norm": 17.965119556017292,
+      "learning_rate": 4.705289508563293e-06,
+      "loss": 0.6633,
+      "step": 3560
+    },
+    {
+      "epoch": 0.1820221281802886,
+      "grad_norm": 11.95934033677814,
+      "learning_rate": 4.703341863382009e-06,
+      "loss": 0.6615,
+      "step": 3570
+    },
+    {
+      "epoch": 0.1825319940855555,
+      "grad_norm": 6.200151009583062,
+      "learning_rate": 4.701388209537569e-06,
+      "loss": 0.651,
+      "step": 3580
+    },
+    {
+      "epoch": 0.18304185999082243,
+      "grad_norm": 10.837831286173747,
+      "learning_rate": 4.69942855235772e-06,
+      "loss": 0.702,
+      "step": 3590
+    },
+    {
+      "epoch": 0.18355172589608934,
+      "grad_norm": 12.987261700429512,
+      "learning_rate": 4.697462897186581e-06,
+      "loss": 0.6598,
+      "step": 3600
+    },
+    {
+      "epoch": 0.18406159180135626,
+      "grad_norm": 1953.3292817352285,
+      "learning_rate": 4.695491249384628e-06,
+      "loss": 0.6659,
+      "step": 3610
+    },
+    {
+      "epoch": 0.18457145770662317,
+      "grad_norm": 6.83565071579364,
+      "learning_rate": 4.693513614328676e-06,
+      "loss": 0.6841,
+      "step": 3620
+    },
+    {
+      "epoch": 0.1850813236118901,
+      "grad_norm": 12.146874651168142,
+      "learning_rate": 4.691529997411873e-06,
+      "loss": 0.738,
+      "step": 3630
+    },
+    {
+      "epoch": 0.185591189517157,
+      "grad_norm": 22.745683956578276,
+      "learning_rate": 4.689540404043677e-06,
+      "loss": 0.6929,
+      "step": 3640
+    },
+    {
+      "epoch": 0.1861010554224239,
+      "grad_norm": 11.50571746380886,
+      "learning_rate": 4.687544839649842e-06,
+      "loss": 0.6619,
+      "step": 3650
+    },
+    {
+      "epoch": 0.1866109213276908,
+      "grad_norm": 13.388716539489481,
+      "learning_rate": 4.685543309672411e-06,
+      "loss": 0.6279,
+      "step": 3660
+    },
+    {
+      "epoch": 0.18712078723295772,
+      "grad_norm": 12.157108250773055,
+      "learning_rate": 4.683535819569691e-06,
+      "loss": 0.705,
+      "step": 3670
+    },
+    {
+      "epoch": 0.18763065313822463,
+      "grad_norm": 12.180479581675126,
+      "learning_rate": 4.681522374816244e-06,
+      "loss": 0.6288,
+      "step": 3680
+    },
+    {
+      "epoch": 0.18814051904349155,
+      "grad_norm": 16.82193055581822,
+      "learning_rate": 4.679502980902871e-06,
+      "loss": 0.6532,
+      "step": 3690
+    },
+    {
+      "epoch": 0.18865038494875846,
+      "grad_norm": 16.649758754627584,
+      "learning_rate": 4.6774776433365965e-06,
+      "loss": 0.5991,
+      "step": 3700
+    },
+    {
+      "epoch": 0.18916025085402538,
+      "grad_norm": 9.18637457024438,
+      "learning_rate": 4.6754463676406545e-06,
+      "loss": 0.6731,
+      "step": 3710
+    },
+    {
+      "epoch": 0.1896701167592923,
+      "grad_norm": 36.51851079033295,
+      "learning_rate": 4.6734091593544705e-06,
+      "loss": 0.6507,
+      "step": 3720
+    },
+    {
+      "epoch": 0.1901799826645592,
+      "grad_norm": 23.088525840764767,
+      "learning_rate": 4.671366024033651e-06,
+      "loss": 0.6136,
+      "step": 3730
+    },
+    {
+      "epoch": 0.19068984856982613,
+      "grad_norm": 11.393600238288462,
+      "learning_rate": 4.669316967249966e-06,
+      "loss": 0.721,
+      "step": 3740
+    },
+    {
+      "epoch": 0.19119971447509304,
+      "grad_norm": 23.092054098850237,
+      "learning_rate": 4.667261994591331e-06,
+      "loss": 0.6668,
+      "step": 3750
+    },
+    {
+      "epoch": 0.19170958038035996,
+      "grad_norm": 9.043611501137145,
+      "learning_rate": 4.665201111661797e-06,
+      "loss": 0.7159,
+      "step": 3760
+    },
+    {
+      "epoch": 0.19221944628562687,
+      "grad_norm": 6.873039056267214,
+      "learning_rate": 4.663134324081533e-06,
+      "loss": 0.6576,
+      "step": 3770
+    },
+    {
+      "epoch": 0.1927293121908938,
+      "grad_norm": 18.365114545851974,
+      "learning_rate": 4.6610616374868066e-06,
+      "loss": 0.7423,
+      "step": 3780
+    },
+    {
+      "epoch": 0.1932391780961607,
+      "grad_norm": 7.330732108249972,
+      "learning_rate": 4.658983057529978e-06,
+      "loss": 0.552,
+      "step": 3790
+    },
+    {
+      "epoch": 0.19374904400142762,
+      "grad_norm": 44.58519185418861,
+      "learning_rate": 4.656898589879475e-06,
+      "loss": 0.6274,
+      "step": 3800
+    },
+    {
+      "epoch": 0.19425890990669453,
+      "grad_norm": 10.959535571320709,
+      "learning_rate": 4.654808240219782e-06,
+      "loss": 0.678,
+      "step": 3810
+    },
+    {
+      "epoch": 0.19476877581196145,
+      "grad_norm": 9.443844486690475,
+      "learning_rate": 4.652712014251426e-06,
+      "loss": 0.6216,
+      "step": 3820
+    },
+    {
+      "epoch": 0.19527864171722836,
+      "grad_norm": 29.311966069946706,
+      "learning_rate": 4.650609917690957e-06,
+      "loss": 0.5818,
+      "step": 3830
+    },
+    {
+      "epoch": 0.19578850762249528,
+      "grad_norm": 10.093424824180262,
+      "learning_rate": 4.648501956270936e-06,
+      "loss": 0.687,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1962983735277622,
+      "grad_norm": 10.636454112575786,
+      "learning_rate": 4.646388135739915e-06,
+      "loss": 0.6343,
+      "step": 3850
+    },
+    {
+      "epoch": 0.1968082394330291,
+      "grad_norm": 23.62252499935202,
+      "learning_rate": 4.64426846186243e-06,
+      "loss": 0.5967,
+      "step": 3860
+    },
+    {
+      "epoch": 0.19731810533829602,
+      "grad_norm": 5.867497909475541,
+      "learning_rate": 4.642142940418973e-06,
+      "loss": 0.7358,
+      "step": 3870
+    },
+    {
+      "epoch": 0.19782797124356294,
+      "grad_norm": 16.58867776820823,
+      "learning_rate": 4.640011577205987e-06,
+      "loss": 0.6752,
+      "step": 3880
+    },
+    {
+      "epoch": 0.19833783714882985,
+      "grad_norm": 10.374371049382303,
+      "learning_rate": 4.637874378035845e-06,
+      "loss": 0.691,
+      "step": 3890
+    },
+    {
+      "epoch": 0.19884770305409677,
+      "grad_norm": 7.507499494510421,
+      "learning_rate": 4.635731348736832e-06,
+      "loss": 0.6807,
+      "step": 3900
+    },
+    {
+      "epoch": 0.19935756895936368,
+      "grad_norm": 52.904799589576946,
+      "learning_rate": 4.633582495153137e-06,
+      "loss": 0.6904,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1998674348646306,
+      "grad_norm": 8.485499924274286,
+      "learning_rate": 4.631427823144829e-06,
+      "loss": 0.6324,
+      "step": 3920
+    },
+    {
+      "epoch": 0.20037730076989752,
+      "grad_norm": 11.64397173810255,
+      "learning_rate": 4.6292673385878466e-06,
+      "loss": 0.6987,
+      "step": 3930
+    },
+    {
+      "epoch": 0.20088716667516443,
+      "grad_norm": 19.05962502430355,
+      "learning_rate": 4.6271010473739754e-06,
+      "loss": 0.766,
+      "step": 3940
+    },
+    {
+      "epoch": 0.20139703258043135,
+      "grad_norm": 12.18391244127119,
+      "learning_rate": 4.624928955410841e-06,
+      "loss": 0.6047,
+      "step": 3950
+    },
+    {
+      "epoch": 0.20190689848569826,
+      "grad_norm": 8.895134290225426,
+      "learning_rate": 4.622751068621886e-06,
+      "loss": 0.6302,
+      "step": 3960
+    },
+    {
+      "epoch": 0.20241676439096518,
+      "grad_norm": 30.39435655960845,
+      "learning_rate": 4.620567392946355e-06,
+      "loss": 0.644,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2029266302962321,
+      "grad_norm": 7.3427483079553175,
+      "learning_rate": 4.618377934339279e-06,
+      "loss": 0.6704,
+      "step": 3980
+    },
+    {
+      "epoch": 0.203436496201499,
+      "grad_norm": 18.877200527683843,
+      "learning_rate": 4.616182698771463e-06,
+      "loss": 0.6056,
+      "step": 3990
+    },
+    {
+      "epoch": 0.20394636210676592,
+      "grad_norm": 11.482887115059379,
+      "learning_rate": 4.613981692229462e-06,
+      "loss": 0.6122,
+      "step": 4000
+    },
+    {
+      "epoch": 0.20445622801203284,
+      "grad_norm": 7.986812774045366,
+      "learning_rate": 4.611774920715572e-06,
+      "loss": 0.6509,
+      "step": 4010
+    },
+    {
+      "epoch": 0.20496609391729975,
+      "grad_norm": 10.669994016142317,
+      "learning_rate": 4.609562390247808e-06,
+      "loss": 0.652,
+      "step": 4020
+    },
+    {
+      "epoch": 0.20547595982256667,
+      "grad_norm": 8.22738123199278,
+      "learning_rate": 4.607344106859891e-06,
+      "loss": 0.6581,
+      "step": 4030
+    },
+    {
+      "epoch": 0.20598582572783358,
+      "grad_norm": 8.962593320055948,
+      "learning_rate": 4.605120076601231e-06,
+      "loss": 0.6999,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2064956916331005,
+      "grad_norm": 14.157563821046,
+      "learning_rate": 4.602890305536911e-06,
+      "loss": 0.6804,
+      "step": 4050
+    },
+    {
+      "epoch": 0.2070055575383674,
+      "grad_norm": 10.129186014658032,
+      "learning_rate": 4.6006547997476666e-06,
+      "loss": 0.706,
+      "step": 4060
+    },
+    {
+      "epoch": 0.20751542344363433,
+      "grad_norm": 7.004517518261975,
+      "learning_rate": 4.598413565329876e-06,
+      "loss": 0.6189,
+      "step": 4070
+    },
+    {
+      "epoch": 0.20802528934890124,
+      "grad_norm": 9.96490109965192,
+      "learning_rate": 4.596166608395535e-06,
+      "loss": 0.6288,
+      "step": 4080
+    },
+    {
+      "epoch": 0.20853515525416816,
+      "grad_norm": 122.00941188724634,
+      "learning_rate": 4.593913935072251e-06,
+      "loss": 0.6688,
+      "step": 4090
+    },
+    {
+      "epoch": 0.20904502115943507,
+      "grad_norm": 5.837271689245049,
+      "learning_rate": 4.591655551503215e-06,
+      "loss": 0.6697,
+      "step": 4100
+    },
+    {
+      "epoch": 0.209554887064702,
+      "grad_norm": 7.591682709402776,
+      "learning_rate": 4.589391463847194e-06,
+      "loss": 0.7342,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2100647529699689,
+      "grad_norm": 8.603671132021212,
+      "learning_rate": 4.58712167827851e-06,
+      "loss": 0.6465,
+      "step": 4120
+    },
+    {
+      "epoch": 0.21057461887523582,
+      "grad_norm": 11.385898364821118,
+      "learning_rate": 4.584846200987022e-06,
+      "loss": 0.669,
+      "step": 4130
+    },
+    {
+      "epoch": 0.21108448478050273,
+      "grad_norm": 13.24419242506655,
+      "learning_rate": 4.582565038178109e-06,
+      "loss": 0.6474,
+      "step": 4140
+    },
+    {
+      "epoch": 0.21159435068576965,
+      "grad_norm": 12.744833317770105,
+      "learning_rate": 4.58027819607266e-06,
+      "loss": 0.665,
+      "step": 4150
+    },
+    {
+      "epoch": 0.21210421659103657,
+      "grad_norm": 8.386511453662118,
+      "learning_rate": 4.577985680907049e-06,
+      "loss": 0.6657,
+      "step": 4160
+    },
+    {
+      "epoch": 0.21261408249630348,
+      "grad_norm": 16.427408003447585,
+      "learning_rate": 4.575687498933119e-06,
+      "loss": 0.6532,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2131239484015704,
+      "grad_norm": 7.603221921574408,
+      "learning_rate": 4.573383656418169e-06,
+      "loss": 0.707,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2136338143068373,
+      "grad_norm": 10.770116825808305,
+      "learning_rate": 4.571074159644936e-06,
+      "loss": 0.6435,
+      "step": 4190
+    },
+    {
+      "epoch": 0.21414368021210423,
+      "grad_norm": 26.871687440222775,
+      "learning_rate": 4.568759014911573e-06,
+      "loss": 0.621,
+      "step": 4200
+    },
+    {
+      "epoch": 0.21465354611737114,
+      "grad_norm": 6.392324082965119,
+      "learning_rate": 4.566438228531638e-06,
+      "loss": 0.6996,
+      "step": 4210
+    },
+    {
+      "epoch": 0.21516341202263806,
+      "grad_norm": 18.486331848454444,
+      "learning_rate": 4.564111806834073e-06,
+      "loss": 0.5856,
+      "step": 4220
+    },
+    {
+      "epoch": 0.21567327792790497,
+      "grad_norm": 7.730881101997795,
+      "learning_rate": 4.5617797561631885e-06,
+      "loss": 0.7293,
+      "step": 4230
+    },
+    {
+      "epoch": 0.2161831438331719,
+      "grad_norm": 13.315083409768397,
+      "learning_rate": 4.559442082878645e-06,
+      "loss": 0.6832,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2166930097384388,
+      "grad_norm": 36.886853297901304,
+      "learning_rate": 4.557098793355436e-06,
+      "loss": 0.5822,
+      "step": 4250
+    },
+    {
+      "epoch": 0.21720287564370572,
+      "grad_norm": 11.837773372638708,
+      "learning_rate": 4.554749893983874e-06,
+      "loss": 0.5926,
+      "step": 4260
+    },
+    {
+      "epoch": 0.21771274154897263,
+      "grad_norm": 17.923474859429835,
+      "learning_rate": 4.552395391169564e-06,
+      "loss": 0.6853,
+      "step": 4270
+    },
+    {
+      "epoch": 0.21822260745423955,
+      "grad_norm": 14.002965124115262,
+      "learning_rate": 4.550035291333398e-06,
+      "loss": 0.5624,
+      "step": 4280
+    },
+    {
+      "epoch": 0.21873247335950646,
+      "grad_norm": 6.554152865278325,
+      "learning_rate": 4.547669600911527e-06,
+      "loss": 0.7063,
+      "step": 4290
+    },
+    {
+      "epoch": 0.21924233926477338,
+      "grad_norm": 14.860982258088095,
+      "learning_rate": 4.545298326355351e-06,
+      "loss": 0.634,
+      "step": 4300
+    },
+    {
+      "epoch": 0.21975220517004027,
+      "grad_norm": 8.1627141188177,
+      "learning_rate": 4.542921474131497e-06,
+      "loss": 0.6895,
+      "step": 4310
+    },
+    {
+      "epoch": 0.22026207107530718,
+      "grad_norm": 6.129164499117784,
+      "learning_rate": 4.540539050721801e-06,
+      "loss": 0.6775,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2207719369805741,
+      "grad_norm": 9.514850934457902,
+      "learning_rate": 4.538151062623296e-06,
+      "loss": 0.5793,
+      "step": 4330
+    },
+    {
+      "epoch": 0.221281802885841,
+      "grad_norm": 6.705922905219092,
+      "learning_rate": 4.535757516348186e-06,
+      "loss": 0.6321,
+      "step": 4340
+    },
+    {
+      "epoch": 0.22179166879110793,
+      "grad_norm": 8.845979013568288,
+      "learning_rate": 4.533358418423837e-06,
+      "loss": 0.6229,
+      "step": 4350
+    },
+    {
+      "epoch": 0.22230153469637484,
+      "grad_norm": 25.37797825124637,
+      "learning_rate": 4.530953775392749e-06,
+      "loss": 0.6514,
+      "step": 4360
+    },
+    {
+      "epoch": 0.22281140060164176,
+      "grad_norm": 14.002012381654973,
+      "learning_rate": 4.52854359381255e-06,
+      "loss": 0.6388,
+      "step": 4370
+    },
+    {
+      "epoch": 0.22332126650690867,
+      "grad_norm": 7.685017242976436,
+      "learning_rate": 4.5261278802559675e-06,
+      "loss": 0.6719,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2238311324121756,
+      "grad_norm": 8.792948491297658,
+      "learning_rate": 4.523706641310817e-06,
+      "loss": 0.629,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2243409983174425,
+      "grad_norm": 10.547436698801492,
+      "learning_rate": 4.521279883579982e-06,
+      "loss": 0.5841,
+      "step": 4400
+    },
+    {
+      "epoch": 0.22485086422270942,
+      "grad_norm": 11.439199546981827,
+      "learning_rate": 4.518847613681397e-06,
+      "loss": 0.7102,
+      "step": 4410
+    },
+    {
+      "epoch": 0.22536073012797633,
+      "grad_norm": 5.762383327340069,
+      "learning_rate": 4.516409838248026e-06,
+      "loss": 0.6329,
+      "step": 4420
+    },
+    {
+      "epoch": 0.22587059603324325,
+      "grad_norm": 7.5671046403384965,
+      "learning_rate": 4.513966563927849e-06,
+      "loss": 0.635,
+      "step": 4430
+    },
+    {
+      "epoch": 0.22638046193851016,
+      "grad_norm": 15.470629027815528,
+      "learning_rate": 4.511517797383841e-06,
+      "loss": 0.6468,
+      "step": 4440
+    },
+    {
+      "epoch": 0.22689032784377708,
+      "grad_norm": 20.905186400076417,
+      "learning_rate": 4.509063545293954e-06,
+      "loss": 0.6701,
+      "step": 4450
+    },
+    {
+      "epoch": 0.227400193749044,
+      "grad_norm": 13.868107533971003,
+      "learning_rate": 4.506603814351103e-06,
+      "loss": 0.5854,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2279100596543109,
+      "grad_norm": 16.533672160629905,
+      "learning_rate": 4.5041386112631394e-06,
+      "loss": 0.6443,
+      "step": 4470
+    },
+    {
+      "epoch": 0.22841992555957782,
+      "grad_norm": 9.198456027284987,
+      "learning_rate": 4.501667942752841e-06,
+      "loss": 0.5938,
+      "step": 4480
+    },
+    {
+      "epoch": 0.22892979146484474,
+      "grad_norm": 9.416687423454851,
+      "learning_rate": 4.499191815557888e-06,
+      "loss": 0.6675,
+      "step": 4490
+    },
+    {
+      "epoch": 0.22943965737011165,
+      "grad_norm": 72.40302034695415,
+      "learning_rate": 4.496710236430848e-06,
+      "loss": 0.6253,
+      "step": 4500
+    },
+    {
+      "epoch": 0.22994952327537857,
+      "grad_norm": 8.499089870026229,
+      "learning_rate": 4.4942232121391565e-06,
+      "loss": 0.6272,
+      "step": 4510
+    },
+    {
+      "epoch": 0.23045938918064549,
+      "grad_norm": 38.011204918994885,
+      "learning_rate": 4.4917307494650975e-06,
+      "loss": 0.5957,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2309692550859124,
+      "grad_norm": 15.172657432433963,
+      "learning_rate": 4.489232855205787e-06,
+      "loss": 0.6854,
+      "step": 4530
+    },
+    {
+      "epoch": 0.23147912099117932,
+      "grad_norm": 19.75927037161284,
+      "learning_rate": 4.4867295361731515e-06,
+      "loss": 0.6695,
+      "step": 4540
+    },
+    {
+      "epoch": 0.23198898689644623,
+      "grad_norm": 13.9619388763187,
+      "learning_rate": 4.484220799193913e-06,
+      "loss": 0.6195,
+      "step": 4550
+    },
+    {
+      "epoch": 0.23249885280171315,
+      "grad_norm": 7.928460498876682,
+      "learning_rate": 4.481706651109567e-06,
+      "loss": 0.5898,
+      "step": 4560
+    },
+    {
+      "epoch": 0.23300871870698006,
+      "grad_norm": 8.027758710964898,
+      "learning_rate": 4.479187098776368e-06,
+      "loss": 0.6324,
+      "step": 4570
+    },
+    {
+      "epoch": 0.23351858461224698,
+      "grad_norm": 7.7963017806518,
+      "learning_rate": 4.476662149065306e-06,
+      "loss": 0.6704,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2340284505175139,
+      "grad_norm": 8.785758250481065,
+      "learning_rate": 4.474131808862089e-06,
+      "loss": 0.7118,
+      "step": 4590
+    },
+    {
+      "epoch": 0.2345383164227808,
+      "grad_norm": 372.06481432872613,
+      "learning_rate": 4.471596085067129e-06,
+      "loss": 0.6486,
+      "step": 4600
+    },
+    {
+      "epoch": 0.23504818232804772,
+      "grad_norm": 5.977610469289572,
+      "learning_rate": 4.469054984595517e-06,
+      "loss": 0.6187,
+      "step": 4610
+    },
+    {
+      "epoch": 0.23555804823331464,
+      "grad_norm": 15.41921778689407,
+      "learning_rate": 4.466508514377006e-06,
+      "loss": 0.637,
+      "step": 4620
+    },
+    {
+      "epoch": 0.23606791413858155,
+      "grad_norm": 9.985726301822796,
+      "learning_rate": 4.463956681355993e-06,
+      "loss": 0.6474,
+      "step": 4630
+    },
+    {
+      "epoch": 0.23657778004384847,
+      "grad_norm": 8.63927014722445,
+      "learning_rate": 4.461399492491502e-06,
+      "loss": 0.6374,
+      "step": 4640
+    },
+    {
+      "epoch": 0.23708764594911538,
+      "grad_norm": 10.214411222798754,
+      "learning_rate": 4.458836954757161e-06,
+      "loss": 0.6459,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2375975118543823,
+      "grad_norm": 11.80289354020228,
+      "learning_rate": 4.456269075141183e-06,
+      "loss": 0.6386,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2381073777596492,
+      "grad_norm": 6.858411227021501,
+      "learning_rate": 4.4536958606463506e-06,
+      "loss": 0.6957,
+      "step": 4670
+    },
+    {
+      "epoch": 0.23861724366491613,
+      "grad_norm": 10.1818081988257,
+      "learning_rate": 4.451117318289996e-06,
+      "loss": 0.663,
+      "step": 4680
+    },
+    {
+      "epoch": 0.23912710957018304,
+      "grad_norm": 4.4783201565409945,
+      "learning_rate": 4.448533455103979e-06,
+      "loss": 0.6272,
+      "step": 4690
+    },
+    {
+      "epoch": 0.23963697547544996,
+      "grad_norm": 13.274713770638163,
+      "learning_rate": 4.445944278134671e-06,
+      "loss": 0.7047,
+      "step": 4700
+    },
+    {
+      "epoch": 0.24014684138071687,
+      "grad_norm": 13.898990696837775,
+      "learning_rate": 4.4433497944429325e-06,
+      "loss": 0.6115,
+      "step": 4710
+    },
+    {
+      "epoch": 0.2406567072859838,
+      "grad_norm": 7.610144075363796,
+      "learning_rate": 4.440750011104098e-06,
+      "loss": 0.5935,
+      "step": 4720
+    },
+    {
+      "epoch": 0.2411665731912507,
+      "grad_norm": 7.152607433595776,
+      "learning_rate": 4.438144935207953e-06,
+      "loss": 0.5946,
+      "step": 4730
+    },
+    {
+      "epoch": 0.24167643909651762,
+      "grad_norm": 10.716962757277734,
+      "learning_rate": 4.435534573858717e-06,
+      "loss": 0.6335,
+      "step": 4740
+    },
+    {
+      "epoch": 0.24218630500178454,
+      "grad_norm": 4.684618729043833,
+      "learning_rate": 4.432918934175023e-06,
+      "loss": 0.6108,
+      "step": 4750
+    },
+    {
+      "epoch": 0.24269617090705145,
+      "grad_norm": 8.294094418125644,
+      "learning_rate": 4.430298023289897e-06,
+      "loss": 0.6291,
+      "step": 4760
+    },
+    {
+      "epoch": 0.24320603681231837,
+      "grad_norm": 17.780227528515248,
+      "learning_rate": 4.427671848350744e-06,
+      "loss": 0.6159,
+      "step": 4770
+    },
+    {
+      "epoch": 0.24371590271758528,
+      "grad_norm": 5.890151944674759,
+      "learning_rate": 4.425040416519319e-06,
+      "loss": 0.5393,
+      "step": 4780
+    },
+    {
+      "epoch": 0.2442257686228522,
+      "grad_norm": 7.1241490526375255,
+      "learning_rate": 4.422403734971718e-06,
+      "loss": 0.5862,
+      "step": 4790
+    },
+    {
+      "epoch": 0.2447356345281191,
+      "grad_norm": 7.856600008931303,
+      "learning_rate": 4.419761810898349e-06,
+      "loss": 0.6156,
+      "step": 4800
+    },
+    {
+      "epoch": 0.24524550043338603,
+      "grad_norm": 13.581267616482041,
+      "learning_rate": 4.4171146515039206e-06,
+      "loss": 0.6246,
+      "step": 4810
+    },
+    {
+      "epoch": 0.24575536633865294,
+      "grad_norm": 5.113231100197785,
+      "learning_rate": 4.414462264007414e-06,
+      "loss": 0.6797,
+      "step": 4820
+    },
+    {
+      "epoch": 0.24626523224391986,
+      "grad_norm": 11.231112402680115,
+      "learning_rate": 4.4118046556420725e-06,
+      "loss": 0.6074,
+      "step": 4830
+    },
+    {
+      "epoch": 0.24677509814918677,
+      "grad_norm": 8.052570264788237,
+      "learning_rate": 4.409141833655375e-06,
+      "loss": 0.5878,
+      "step": 4840
+    },
+    {
+      "epoch": 0.2472849640544537,
+      "grad_norm": 6.660041473213604,
+      "learning_rate": 4.406473805309016e-06,
+      "loss": 0.6666,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2477948299597206,
+      "grad_norm": 7.323309513429638,
+      "learning_rate": 4.403800577878892e-06,
+      "loss": 0.5889,
+      "step": 4860
+    },
+    {
+      "epoch": 0.24830469586498752,
+      "grad_norm": 20.81071641709809,
+      "learning_rate": 4.401122158655076e-06,
+      "loss": 0.6094,
+      "step": 4870
+    },
+    {
+      "epoch": 0.24881456177025443,
+      "grad_norm": 10.74009093392373,
+      "learning_rate": 4.3984385549418e-06,
+      "loss": 0.6416,
+      "step": 4880
+    },
+    {
+      "epoch": 0.24932442767552135,
+      "grad_norm": 6.980032132546093,
+      "learning_rate": 4.395749774057432e-06,
+      "loss": 0.6539,
+      "step": 4890
+    },
+    {
+      "epoch": 0.24983429358078826,
+      "grad_norm": 82.35304909424602,
+      "learning_rate": 4.393055823334461e-06,
+      "loss": 0.6577,
+      "step": 4900
+    },
+    {
+      "epoch": 0.25034415948605515,
+      "grad_norm": 68.9873160694173,
+      "learning_rate": 4.390356710119476e-06,
+      "loss": 0.6415,
+      "step": 4910
+    },
+    {
+      "epoch": 0.25085402539132207,
+      "grad_norm": 8.53782382775481,
+      "learning_rate": 4.38765244177314e-06,
+      "loss": 0.6141,
+      "step": 4920
+    },
+    {
+      "epoch": 0.251363891296589,
+      "grad_norm": 10.124372790395688,
+      "learning_rate": 4.3849430256701765e-06,
+      "loss": 0.601,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2518737572018559,
+      "grad_norm": 7.77404112045145,
+      "learning_rate": 4.38222846919935e-06,
+      "loss": 0.6295,
+      "step": 4940
+    },
+    {
+      "epoch": 0.2523836231071228,
+      "grad_norm": 15.321610414707543,
+      "learning_rate": 4.379508779763438e-06,
+      "loss": 0.719,
+      "step": 4950
+    },
+    {
+      "epoch": 0.2528934890123897,
+      "grad_norm": 8.797795586013368,
+      "learning_rate": 4.376783964779221e-06,
+      "loss": 0.6559,
+      "step": 4960
+    },
+    {
+      "epoch": 0.25340335491765664,
+      "grad_norm": 14.17555125096599,
+      "learning_rate": 4.3740540316774535e-06,
+      "loss": 0.6231,
+      "step": 4970
+    },
+    {
+      "epoch": 0.25391322082292356,
+      "grad_norm": 9.755110695106975,
+      "learning_rate": 4.3713189879028485e-06,
+      "loss": 0.5748,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2544230867281905,
+      "grad_norm": 6.830509512931204,
+      "learning_rate": 4.3685788409140564e-06,
+      "loss": 0.6683,
+      "step": 4990
+    },
+    {
+      "epoch": 0.2549329526334574,
+      "grad_norm": 22.277722816391652,
+      "learning_rate": 4.365833598183645e-06,
+      "loss": 0.6583,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2554428185387243,
+      "grad_norm": 8.773826215610796,
+      "learning_rate": 4.363083267198079e-06,
+      "loss": 0.6008,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2559526844439912,
+      "grad_norm": 7.782050291060473,
+      "learning_rate": 4.360327855457696e-06,
+      "loss": 0.5959,
+      "step": 5020
+    },
+    {
+      "epoch": 0.25646255034925813,
+      "grad_norm": 6.610786664149157,
+      "learning_rate": 4.357567370476693e-06,
+      "loss": 0.5854,
+      "step": 5030
+    },
+    {
+      "epoch": 0.25697241625452505,
+      "grad_norm": 9.898788342889924,
+      "learning_rate": 4.354801819783099e-06,
+      "loss": 0.5716,
+      "step": 5040
+    },
+    {
+      "epoch": 0.25748228215979196,
+      "grad_norm": 7.0022547012783445,
+      "learning_rate": 4.35203121091876e-06,
+      "loss": 0.5536,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2579921480650589,
+      "grad_norm": 8.798700085731559,
+      "learning_rate": 4.349255551439314e-06,
+      "loss": 0.6537,
+      "step": 5060
+    },
+    {
+      "epoch": 0.2585020139703258,
+      "grad_norm": 7.284575104853222,
+      "learning_rate": 4.346474848914174e-06,
+      "loss": 0.6439,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2590118798755927,
+      "grad_norm": 12.366489512249572,
+      "learning_rate": 4.343689110926504e-06,
+      "loss": 0.6505,
+      "step": 5080
+    },
+    {
+      "epoch": 0.2595217457808596,
+      "grad_norm": 18.160906934404036,
+      "learning_rate": 4.340898345073202e-06,
+      "loss": 0.7467,
+      "step": 5090
+    },
+    {
+      "epoch": 0.26003161168612654,
+      "grad_norm": 6.074259845271382,
+      "learning_rate": 4.338102558964876e-06,
+      "loss": 0.683,
+      "step": 5100
+    },
+    {
+      "epoch": 0.26054147759139346,
+      "grad_norm": 20.71328456392585,
+      "learning_rate": 4.335301760225824e-06,
+      "loss": 0.6004,
+      "step": 5110
+    },
+    {
+      "epoch": 0.26105134349666037,
+      "grad_norm": 27.825218691728683,
+      "learning_rate": 4.3324959564940165e-06,
+      "loss": 0.5665,
+      "step": 5120
+    },
+    {
+      "epoch": 0.2615612094019273,
+      "grad_norm": 8.576618170922808,
+      "learning_rate": 4.329685155421069e-06,
+      "loss": 0.6139,
+      "step": 5130
+    },
+    {
+      "epoch": 0.2620710753071942,
+      "grad_norm": 5.191563874692266,
+      "learning_rate": 4.326869364672229e-06,
+      "loss": 0.6224,
+      "step": 5140
+    },
+    {
+      "epoch": 0.2625809412124611,
+      "grad_norm": 7.644239450475295,
+      "learning_rate": 4.324048591926349e-06,
+      "loss": 0.6289,
+      "step": 5150
+    },
+    {
+      "epoch": 0.26309080711772803,
+      "grad_norm": 7.166245372410849,
+      "learning_rate": 4.321222844875869e-06,
+      "loss": 0.6658,
+      "step": 5160
+    },
+    {
+      "epoch": 0.26360067302299495,
+      "grad_norm": 5.350234410317429,
+      "learning_rate": 4.318392131226791e-06,
+      "loss": 0.6947,
+      "step": 5170
+    },
+    {
+      "epoch": 0.26411053892826186,
+      "grad_norm": 5.755572274000322,
+      "learning_rate": 4.315556458698665e-06,
+      "loss": 0.6358,
+      "step": 5180
+    },
+    {
+      "epoch": 0.2646204048335288,
+      "grad_norm": 38.17798848705542,
+      "learning_rate": 4.312715835024565e-06,
+      "loss": 0.6454,
+      "step": 5190
+    },
+    {
+      "epoch": 0.2651302707387957,
+      "grad_norm": 11.241404253833712,
+      "learning_rate": 4.309870267951061e-06,
+      "loss": 0.7519,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2656401366440626,
+      "grad_norm": 22.28328130251413,
+      "learning_rate": 4.30701976523821e-06,
+      "loss": 0.6471,
+      "step": 5210
+    },
+    {
+      "epoch": 0.2661500025493295,
+      "grad_norm": 9.842249709147955,
+      "learning_rate": 4.3041643346595285e-06,
+      "loss": 0.5849,
+      "step": 5220
+    },
+    {
+      "epoch": 0.26665986845459644,
+      "grad_norm": 9.637874026517801,
+      "learning_rate": 4.3013039840019675e-06,
+      "loss": 0.6153,
+      "step": 5230
+    },
+    {
+      "epoch": 0.26716973435986335,
+      "grad_norm": 4.617183170224206,
+      "learning_rate": 4.298438721065899e-06,
+      "loss": 0.6211,
+      "step": 5240
+    },
+    {
+      "epoch": 0.26767960026513027,
+      "grad_norm": 10.65168368583943,
+      "learning_rate": 4.295568553665089e-06,
+      "loss": 0.6364,
+      "step": 5250
+    },
+    {
+      "epoch": 0.2681894661703972,
+      "grad_norm": 10.613247539596093,
+      "learning_rate": 4.292693489626681e-06,
+      "loss": 0.6323,
+      "step": 5260
+    },
+    {
+      "epoch": 0.2686993320756641,
+      "grad_norm": 8.49272818687361,
+      "learning_rate": 4.289813536791168e-06,
+      "loss": 0.575,
+      "step": 5270
+    },
+    {
+      "epoch": 0.269209197980931,
+      "grad_norm": 22.6704390447368,
+      "learning_rate": 4.2869287030123786e-06,
+      "loss": 0.6592,
+      "step": 5280
+    },
+    {
+      "epoch": 0.26971906388619793,
+      "grad_norm": 7.121342392959829,
+      "learning_rate": 4.284038996157451e-06,
+      "loss": 0.6357,
+      "step": 5290
+    },
+    {
+      "epoch": 0.27022892979146484,
+      "grad_norm": 24.88598041603986,
+      "learning_rate": 4.2811444241068115e-06,
+      "loss": 0.5871,
+      "step": 5300
+    },
+    {
+      "epoch": 0.27073879569673176,
+      "grad_norm": 5.368482850926287,
+      "learning_rate": 4.278244994754155e-06,
+      "loss": 0.5931,
+      "step": 5310
+    },
+    {
+      "epoch": 0.2712486616019987,
+      "grad_norm": 5.180107118782333,
+      "learning_rate": 4.275340716006424e-06,
+      "loss": 0.566,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2717585275072656,
+      "grad_norm": 9.352302095228602,
+      "learning_rate": 4.272431595783783e-06,
+      "loss": 0.6013,
+      "step": 5330
+    },
+    {
+      "epoch": 0.2722683934125325,
+      "grad_norm": 5.016831399628619,
+      "learning_rate": 4.269517642019601e-06,
+      "loss": 0.7052,
+      "step": 5340
+    },
+    {
+      "epoch": 0.2727782593177994,
+      "grad_norm": 11.577872049295284,
+      "learning_rate": 4.2665988626604285e-06,
+      "loss": 0.623,
+      "step": 5350
+    },
+    {
+      "epoch": 0.27328812522306634,
+      "grad_norm": 12.110900648060454,
+      "learning_rate": 4.2636752656659745e-06,
+      "loss": 0.6656,
+      "step": 5360
+    },
+    {
+      "epoch": 0.27379799112833325,
+      "grad_norm": 6.509298716020258,
+      "learning_rate": 4.260746859009087e-06,
+      "loss": 0.621,
+      "step": 5370
+    },
+    {
+      "epoch": 0.27430785703360017,
+      "grad_norm": 8.948515262505117,
+      "learning_rate": 4.257813650675732e-06,
+      "loss": 0.6315,
+      "step": 5380
+    },
+    {
+      "epoch": 0.2748177229388671,
+      "grad_norm": 6.910625223428903,
+      "learning_rate": 4.254875648664965e-06,
+      "loss": 0.6059,
+      "step": 5390
+    },
+    {
+      "epoch": 0.275327588844134,
+      "grad_norm": 9.995100150449323,
+      "learning_rate": 4.251932860988921e-06,
+      "loss": 0.5617,
+      "step": 5400
+    },
+    {
+      "epoch": 0.2758374547494009,
+      "grad_norm": 8.682632389217074,
+      "learning_rate": 4.24898529567278e-06,
+      "loss": 0.6523,
+      "step": 5410
+    },
+    {
+      "epoch": 0.2763473206546678,
+      "grad_norm": 8.280215194644136,
+      "learning_rate": 4.246032960754753e-06,
+      "loss": 0.6414,
+      "step": 5420
+    },
+    {
+      "epoch": 0.27685718655993474,
+      "grad_norm": 10.10137835136439,
+      "learning_rate": 4.243075864286059e-06,
+      "loss": 0.5732,
+      "step": 5430
+    },
+    {
+      "epoch": 0.27736705246520166,
+      "grad_norm": 10.8290914906575,
+      "learning_rate": 4.2401140143309e-06,
+      "loss": 0.5839,
+      "step": 5440
+    },
+    {
+      "epoch": 0.2778769183704686,
+      "grad_norm": 6.645724769592051,
+      "learning_rate": 4.237147418966444e-06,
+      "loss": 0.5774,
+      "step": 5450
+    },
+    {
+      "epoch": 0.2783867842757355,
+      "grad_norm": 72.81404017838632,
+      "learning_rate": 4.234176086282797e-06,
+      "loss": 0.5864,
+      "step": 5460
+    },
+    {
+      "epoch": 0.2788966501810024,
+      "grad_norm": 5.314994274785593,
+      "learning_rate": 4.231200024382987e-06,
+      "loss": 0.6449,
+      "step": 5470
+    },
+    {
+      "epoch": 0.2794065160862693,
+      "grad_norm": 31.29101037439269,
+      "learning_rate": 4.228219241382936e-06,
+      "loss": 0.5623,
+      "step": 5480
+    },
+    {
+      "epoch": 0.27991638199153623,
+      "grad_norm": 25.923064324099325,
+      "learning_rate": 4.2252337454114426e-06,
+      "loss": 0.6535,
+      "step": 5490
+    },
+    {
+      "epoch": 0.28042624789680315,
+      "grad_norm": 22.96804242136826,
+      "learning_rate": 4.2222435446101555e-06,
+      "loss": 0.6227,
+      "step": 5500
+    },
+    {
+      "epoch": 0.28093611380207006,
+      "grad_norm": 9.39528145131879,
+      "learning_rate": 4.219248647133559e-06,
+      "loss": 0.5898,
+      "step": 5510
+    },
+    {
+      "epoch": 0.281445979707337,
+      "grad_norm": 14.895333434529494,
+      "learning_rate": 4.216249061148939e-06,
+      "loss": 0.6934,
+      "step": 5520
+    },
+    {
+      "epoch": 0.2819558456126039,
+      "grad_norm": 17.44854189744529,
+      "learning_rate": 4.213244794836373e-06,
+      "loss": 0.6571,
+      "step": 5530
+    },
+    {
+      "epoch": 0.2824657115178708,
+      "grad_norm": 6.045811920782836,
+      "learning_rate": 4.210235856388699e-06,
+      "loss": 0.6137,
+      "step": 5540
+    },
+    {
+      "epoch": 0.2829755774231377,
+      "grad_norm": 6.494290828572964,
+      "learning_rate": 4.2072222540114965e-06,
+      "loss": 0.641,
+      "step": 5550
+    },
+    {
+      "epoch": 0.28348544332840464,
+      "grad_norm": 8.506416783659864,
+      "learning_rate": 4.204203995923064e-06,
+      "loss": 0.5892,
+      "step": 5560
+    },
+    {
+      "epoch": 0.28399530923367156,
+      "grad_norm": 10.074725916553763,
+      "learning_rate": 4.201181090354396e-06,
+      "loss": 0.6319,
+      "step": 5570
+    },
+    {
+      "epoch": 0.28450517513893847,
+      "grad_norm": 12.321876819156738,
+      "learning_rate": 4.198153545549164e-06,
+      "loss": 0.5932,
+      "step": 5580
+    },
+    {
+      "epoch": 0.2850150410442054,
+      "grad_norm": 8.281688914683107,
+      "learning_rate": 4.195121369763687e-06,
+      "loss": 0.6197,
+      "step": 5590
+    },
+    {
+      "epoch": 0.2855249069494723,
+      "grad_norm": 10.71683516397999,
+      "learning_rate": 4.192084571266915e-06,
+      "loss": 0.6103,
+      "step": 5600
+    },
+    {
+      "epoch": 0.2860347728547392,
+      "grad_norm": 5.796597428515388,
+      "learning_rate": 4.189043158340403e-06,
+      "loss": 0.6021,
+      "step": 5610
+    },
+    {
+      "epoch": 0.28654463876000613,
+      "grad_norm": 10.228637952287452,
+      "learning_rate": 4.185997139278292e-06,
+      "loss": 0.5835,
+      "step": 5620
+    },
+    {
+      "epoch": 0.28705450466527305,
+      "grad_norm": 12.590399688941602,
+      "learning_rate": 4.182946522387283e-06,
+      "loss": 0.6131,
+      "step": 5630
+    },
+    {
+      "epoch": 0.28756437057053996,
+      "grad_norm": 11.901456061427117,
+      "learning_rate": 4.179891315986617e-06,
+      "loss": 0.5923,
+      "step": 5640
+    },
+    {
+      "epoch": 0.2880742364758069,
+      "grad_norm": 4.729735791734294,
+      "learning_rate": 4.1768315284080475e-06,
+      "loss": 0.5941,
+      "step": 5650
+    },
+    {
+      "epoch": 0.2885841023810738,
+      "grad_norm": 41.878376647383455,
+      "learning_rate": 4.173767167995825e-06,
+      "loss": 0.6108,
+      "step": 5660
+    },
+    {
+      "epoch": 0.2890939682863407,
+      "grad_norm": 8.57145358299838,
+      "learning_rate": 4.170698243106668e-06,
+      "loss": 0.6225,
+      "step": 5670
+    },
+    {
+      "epoch": 0.2896038341916076,
+      "grad_norm": 57.68706708658153,
+      "learning_rate": 4.1676247621097445e-06,
+      "loss": 0.5881,
+      "step": 5680
+    },
+    {
+      "epoch": 0.29011370009687454,
+      "grad_norm": 14.15474838073681,
+      "learning_rate": 4.164546733386644e-06,
+      "loss": 0.6116,
+      "step": 5690
+    },
+    {
+      "epoch": 0.29062356600214145,
+      "grad_norm": 12.984062057200392,
+      "learning_rate": 4.161464165331363e-06,
+      "loss": 0.6717,
+      "step": 5700
+    },
+    {
+      "epoch": 0.29113343190740837,
+      "grad_norm": 10.624920059707007,
+      "learning_rate": 4.158377066350273e-06,
+      "loss": 0.5929,
+      "step": 5710
+    },
+    {
+      "epoch": 0.2916432978126753,
+      "grad_norm": 30.251451662052922,
+      "learning_rate": 4.1552854448621025e-06,
+      "loss": 0.6397,
+      "step": 5720
+    },
+    {
+      "epoch": 0.2921531637179422,
+      "grad_norm": 28.098357274037717,
+      "learning_rate": 4.152189309297914e-06,
+      "loss": 0.5996,
+      "step": 5730
+    },
+    {
+      "epoch": 0.2926630296232091,
+      "grad_norm": 15.49814773098564,
+      "learning_rate": 4.14908866810108e-06,
+      "loss": 0.6222,
+      "step": 5740
+    },
+    {
+      "epoch": 0.29317289552847603,
+      "grad_norm": 10.342825730948135,
+      "learning_rate": 4.14598352972726e-06,
+      "loss": 0.5807,
+      "step": 5750
+    },
+    {
+      "epoch": 0.29368276143374294,
+      "grad_norm": 6.7777361109333825,
+      "learning_rate": 4.142873902644378e-06,
+      "loss": 0.6487,
+      "step": 5760
+    },
+    {
+      "epoch": 0.29419262733900986,
+      "grad_norm": 55.36024537902725,
+      "learning_rate": 4.139759795332597e-06,
+      "loss": 0.6478,
+      "step": 5770
+    },
+    {
+      "epoch": 0.2947024932442768,
+      "grad_norm": 7.415282240745671,
+      "learning_rate": 4.1366412162843015e-06,
+      "loss": 0.5682,
+      "step": 5780
+    },
+    {
+      "epoch": 0.2952123591495437,
+      "grad_norm": 16.320079976157704,
+      "learning_rate": 4.133518174004068e-06,
+      "loss": 0.5993,
+      "step": 5790
+    },
+    {
+      "epoch": 0.2957222250548106,
+      "grad_norm": 14.575740442925616,
+      "learning_rate": 4.130390677008644e-06,
+      "loss": 0.6656,
+      "step": 5800
+    },
+    {
+      "epoch": 0.2962320909600775,
+      "grad_norm": 23.443828943004323,
+      "learning_rate": 4.127258733826929e-06,
+      "loss": 0.5881,
+      "step": 5810
+    },
+    {
+      "epoch": 0.29674195686534444,
+      "grad_norm": 8.338215537290909,
+      "learning_rate": 4.1241223529999425e-06,
+      "loss": 0.6088,
+      "step": 5820
+    },
+    {
+      "epoch": 0.29725182277061135,
+      "grad_norm": 3.9193795434016976,
+      "learning_rate": 4.12098154308081e-06,
+      "loss": 0.6227,
+      "step": 5830
+    },
+    {
+      "epoch": 0.29776168867587827,
+      "grad_norm": 34.505152074166546,
+      "learning_rate": 4.117836312634734e-06,
+      "loss": 0.6023,
+      "step": 5840
+    },
+    {
+      "epoch": 0.2982715545811452,
+      "grad_norm": 10.358330475944964,
+      "learning_rate": 4.114686670238971e-06,
+      "loss": 0.5958,
+      "step": 5850
+    },
+    {
+      "epoch": 0.2987814204864121,
+      "grad_norm": 10.171023349901475,
+      "learning_rate": 4.111532624482811e-06,
+      "loss": 0.6487,
+      "step": 5860
+    },
+    {
+      "epoch": 0.299291286391679,
+      "grad_norm": 14.306133179044457,
+      "learning_rate": 4.108374183967549e-06,
+      "loss": 0.6554,
+      "step": 5870
+    },
+    {
+      "epoch": 0.2998011522969459,
+      "grad_norm": 15.567524939241922,
+      "learning_rate": 4.10521135730647e-06,
+      "loss": 0.6807,
+      "step": 5880
+    },
+    {
+      "epoch": 0.30031101820221284,
+      "grad_norm": 38.31191365953914,
+      "learning_rate": 4.1020441531248165e-06,
+      "loss": 0.5854,
+      "step": 5890
+    },
+    {
+      "epoch": 0.30082088410747976,
+      "grad_norm": 19.86350978038444,
+      "learning_rate": 4.09887258005977e-06,
+      "loss": 0.5814,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3013307500127467,
+      "grad_norm": 15.358100181159678,
+      "learning_rate": 4.095696646760425e-06,
+      "loss": 0.6046,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3018406159180136,
+      "grad_norm": 15.721665747424774,
+      "learning_rate": 4.09251636188777e-06,
+      "loss": 0.656,
+      "step": 5920
+    },
+    {
+      "epoch": 0.3023504818232805,
+      "grad_norm": 4.81030272554478,
+      "learning_rate": 4.0893317341146545e-06,
+      "loss": 0.6183,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3028603477285474,
+      "grad_norm": 22.743414607219858,
+      "learning_rate": 4.086142772125779e-06,
+      "loss": 0.6162,
+      "step": 5940
+    },
+    {
+      "epoch": 0.30337021363381433,
+      "grad_norm": 339.9381590434665,
+      "learning_rate": 4.082949484617656e-06,
+      "loss": 0.5904,
+      "step": 5950
+    },
+    {
+      "epoch": 0.30388007953908125,
+      "grad_norm": 9.861498398640471,
+      "learning_rate": 4.079751880298601e-06,
+      "loss": 0.6336,
+      "step": 5960
+    },
+    {
+      "epoch": 0.3043899454443481,
+      "grad_norm": 11.112861647101575,
+      "learning_rate": 4.076549967888697e-06,
+      "loss": 0.6418,
+      "step": 5970
+    },
+    {
+      "epoch": 0.304899811349615,
+      "grad_norm": 9.57530538775308,
+      "learning_rate": 4.073343756119778e-06,
+      "loss": 0.5854,
+      "step": 5980
+    },
+    {
+      "epoch": 0.30540967725488194,
+      "grad_norm": 27.895928908583112,
+      "learning_rate": 4.070133253735399e-06,
+      "loss": 0.5991,
+      "step": 5990
+    },
+    {
+      "epoch": 0.30591954316014885,
+      "grad_norm": 5.340717821375443,
+      "learning_rate": 4.066918469490822e-06,
+      "loss": 0.6264,
+      "step": 6000
+    },
+    {
+      "epoch": 0.30642940906541577,
+      "grad_norm": 26.112994712108293,
+      "learning_rate": 4.063699412152979e-06,
+      "loss": 0.5982,
+      "step": 6010
+    },
+    {
+      "epoch": 0.3069392749706827,
+      "grad_norm": 14.195337759571771,
+      "learning_rate": 4.060476090500462e-06,
+      "loss": 0.6299,
+      "step": 6020
+    },
+    {
+      "epoch": 0.3074491408759496,
+      "grad_norm": 5.544117867891527,
+      "learning_rate": 4.057248513323484e-06,
+      "loss": 0.6132,
+      "step": 6030
+    },
+    {
+      "epoch": 0.3079590067812165,
+      "grad_norm": 7.769017752910073,
+      "learning_rate": 4.054016689423871e-06,
+      "loss": 0.6122,
+      "step": 6040
+    },
+    {
+      "epoch": 0.30846887268648343,
+      "grad_norm": 5.614632316438232,
+      "learning_rate": 4.050780627615025e-06,
+      "loss": 0.6126,
+      "step": 6050
+    },
+    {
+      "epoch": 0.30897873859175035,
+      "grad_norm": 25.512901060709,
+      "learning_rate": 4.047540336721909e-06,
+      "loss": 0.5766,
+      "step": 6060
+    },
+    {
+      "epoch": 0.30948860449701726,
+      "grad_norm": 20.54637693246705,
+      "learning_rate": 4.044295825581013e-06,
+      "loss": 0.5903,
+      "step": 6070
+    },
+    {
+      "epoch": 0.3099984704022842,
+      "grad_norm": 4.685213134791145,
+      "learning_rate": 4.041047103040343e-06,
+      "loss": 0.59,
+      "step": 6080
+    },
+    {
+      "epoch": 0.3105083363075511,
+      "grad_norm": 14.698272645628181,
+      "learning_rate": 4.0377941779593835e-06,
+      "loss": 0.5881,
+      "step": 6090
+    },
+    {
+      "epoch": 0.311018202212818,
+      "grad_norm": 11.911269514103715,
+      "learning_rate": 4.034537059209085e-06,
+      "loss": 0.5514,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3115280681180849,
+      "grad_norm": 16.715726537650962,
+      "learning_rate": 4.03127575567183e-06,
+      "loss": 0.5983,
+      "step": 6110
+    },
+    {
+      "epoch": 0.31203793402335184,
+      "grad_norm": 4.360422453535343,
+      "learning_rate": 4.028010276241416e-06,
+      "loss": 0.5862,
+      "step": 6120
+    },
+    {
+      "epoch": 0.31254779992861875,
+      "grad_norm": 3.997756637077826,
+      "learning_rate": 4.0247406298230285e-06,
+      "loss": 0.617,
+      "step": 6130
+    },
+    {
+      "epoch": 0.31305766583388567,
+      "grad_norm": 4.76148406920648,
+      "learning_rate": 4.021466825333215e-06,
+      "loss": 0.5919,
+      "step": 6140
+    },
+    {
+      "epoch": 0.3135675317391526,
+      "grad_norm": 13.942277089049702,
+      "learning_rate": 4.018188871699861e-06,
+      "loss": 0.6223,
+      "step": 6150
+    },
+    {
+      "epoch": 0.3140773976444195,
+      "grad_norm": 13.662148648384237,
+      "learning_rate": 4.014906777862172e-06,
+      "loss": 0.6097,
+      "step": 6160
+    },
+    {
+      "epoch": 0.3145872635496864,
+      "grad_norm": 5.804025458086976,
+      "learning_rate": 4.01162055277064e-06,
+      "loss": 0.6503,
+      "step": 6170
+    },
+    {
+      "epoch": 0.31509712945495333,
+      "grad_norm": 16.82069434398167,
+      "learning_rate": 4.008330205387024e-06,
+      "loss": 0.6452,
+      "step": 6180
+    },
+    {
+      "epoch": 0.31560699536022024,
+      "grad_norm": 16.53797565480657,
+      "learning_rate": 4.005035744684325e-06,
+      "loss": 0.5796,
+      "step": 6190
+    },
+    {
+      "epoch": 0.31611686126548716,
+      "grad_norm": 8.015851023261215,
+      "learning_rate": 4.0017371796467635e-06,
+      "loss": 0.6059,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3166267271707541,
+      "grad_norm": 6.7882106521077805,
+      "learning_rate": 3.998434519269749e-06,
+      "loss": 0.5796,
+      "step": 6210
+    },
+    {
+      "epoch": 0.317136593076021,
+      "grad_norm": 10.623587735462868,
+      "learning_rate": 3.9951277725598604e-06,
+      "loss": 0.6933,
+      "step": 6220
+    },
+    {
+      "epoch": 0.3176464589812879,
+      "grad_norm": 7.334096120730939,
+      "learning_rate": 3.991816948534823e-06,
+      "loss": 0.5632,
+      "step": 6230
+    },
+    {
+      "epoch": 0.3181563248865548,
+      "grad_norm": 16.456732933661677,
+      "learning_rate": 3.988502056223477e-06,
+      "loss": 0.5692,
+      "step": 6240
+    },
+    {
+      "epoch": 0.31866619079182174,
+      "grad_norm": 137.52656755217927,
+      "learning_rate": 3.98518310466576e-06,
+      "loss": 0.5884,
+      "step": 6250
+    },
+    {
+      "epoch": 0.31917605669708865,
+      "grad_norm": 25.065009101683582,
+      "learning_rate": 3.98186010291268e-06,
+      "loss": 0.5745,
+      "step": 6260
+    },
+    {
+      "epoch": 0.31968592260235557,
+      "grad_norm": 8.153973542505792,
+      "learning_rate": 3.978533060026288e-06,
+      "loss": 0.6411,
+      "step": 6270
+    },
+    {
+      "epoch": 0.3201957885076225,
+      "grad_norm": 22.665329895940015,
+      "learning_rate": 3.975201985079655e-06,
+      "loss": 0.5821,
+      "step": 6280
+    },
+    {
+      "epoch": 0.3207056544128894,
+      "grad_norm": 8.578356130956463,
+      "learning_rate": 3.971866887156851e-06,
+      "loss": 0.5774,
+      "step": 6290
+    },
+    {
+      "epoch": 0.3212155203181563,
+      "grad_norm": 10.787741255106576,
+      "learning_rate": 3.968527775352914e-06,
+      "loss": 0.6307,
+      "step": 6300
+    },
+    {
+      "epoch": 0.3217253862234232,
+      "grad_norm": 8.963049497956545,
+      "learning_rate": 3.965184658773828e-06,
+      "loss": 0.6402,
+      "step": 6310
+    },
+    {
+      "epoch": 0.32223525212869014,
+      "grad_norm": 8.266683813731634,
+      "learning_rate": 3.961837546536501e-06,
+      "loss": 0.5837,
+      "step": 6320
+    },
+    {
+      "epoch": 0.32274511803395706,
+      "grad_norm": 9.180307846998755,
+      "learning_rate": 3.958486447768736e-06,
+      "loss": 0.5806,
+      "step": 6330
+    },
+    {
+      "epoch": 0.32325498393922397,
+      "grad_norm": 25.651525690537756,
+      "learning_rate": 3.955131371609206e-06,
+      "loss": 0.6398,
+      "step": 6340
+    },
+    {
+      "epoch": 0.3237648498444909,
+      "grad_norm": 7.1700263175610806,
+      "learning_rate": 3.951772327207432e-06,
+      "loss": 0.6487,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3242747157497578,
+      "grad_norm": 6.599676709961129,
+      "learning_rate": 3.948409323723756e-06,
+      "loss": 0.6023,
+      "step": 6360
+    },
+    {
+      "epoch": 0.3247845816550247,
+      "grad_norm": 20.95163227633846,
+      "learning_rate": 3.945042370329319e-06,
+      "loss": 0.5539,
+      "step": 6370
+    },
+    {
+      "epoch": 0.32529444756029163,
+      "grad_norm": 7.088550847545342,
+      "learning_rate": 3.9416714762060325e-06,
+      "loss": 0.6551,
+      "step": 6380
+    },
+    {
+      "epoch": 0.32580431346555855,
+      "grad_norm": 15.30777318161728,
+      "learning_rate": 3.938296650546552e-06,
+      "loss": 0.5995,
+      "step": 6390
+    },
+    {
+      "epoch": 0.32631417937082546,
+      "grad_norm": 13.542125007259056,
+      "learning_rate": 3.934917902554257e-06,
+      "loss": 0.569,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3268240452760924,
+      "grad_norm": 19.839669107198333,
+      "learning_rate": 3.931535241443225e-06,
+      "loss": 0.5796,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3273339111813593,
+      "grad_norm": 9.379304814298935,
+      "learning_rate": 3.9281486764382e-06,
+      "loss": 0.5959,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3278437770866262,
+      "grad_norm": 4.964025222773356,
+      "learning_rate": 3.924758216774579e-06,
+      "loss": 0.6092,
+      "step": 6430
+    },
+    {
+      "epoch": 0.3283536429918931,
+      "grad_norm": 8.647009136839792,
+      "learning_rate": 3.921363871698372e-06,
+      "loss": 0.6594,
+      "step": 6440
+    },
+    {
+      "epoch": 0.32886350889716004,
+      "grad_norm": 7.0288166471256535,
+      "learning_rate": 3.917965650466192e-06,
+      "loss": 0.6147,
+      "step": 6450
+    },
+    {
+      "epoch": 0.32937337480242695,
+      "grad_norm": 13.463141621042343,
+      "learning_rate": 3.914563562345218e-06,
+      "loss": 0.6147,
+      "step": 6460
+    },
+    {
+      "epoch": 0.32988324070769387,
+      "grad_norm": 10.452423068036081,
+      "learning_rate": 3.911157616613176e-06,
+      "loss": 0.6291,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3303931066129608,
+      "grad_norm": 22.28017669172328,
+      "learning_rate": 3.9077478225583115e-06,
+      "loss": 0.58,
+      "step": 6480
+    },
+    {
+      "epoch": 0.3309029725182277,
+      "grad_norm": 7.096911056626139,
+      "learning_rate": 3.904334189479366e-06,
+      "loss": 0.6304,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3314128384234946,
+      "grad_norm": 7.513660732062853,
+      "learning_rate": 3.900916726685547e-06,
+      "loss": 0.5736,
+      "step": 6500
+    },
+    {
+      "epoch": 0.33192270432876153,
+      "grad_norm": 10.702933677647772,
+      "learning_rate": 3.897495443496511e-06,
+      "loss": 0.6478,
+      "step": 6510
+    },
+    {
+      "epoch": 0.33243257023402845,
+      "grad_norm": 7.712509198957455,
+      "learning_rate": 3.894070349242328e-06,
+      "loss": 0.6366,
+      "step": 6520
+    },
+    {
+      "epoch": 0.33294243613929536,
+      "grad_norm": 12.4801385321312,
+      "learning_rate": 3.890641453263463e-06,
+      "loss": 0.6557,
+      "step": 6530
+    },
+    {
+      "epoch": 0.3334523020445623,
+      "grad_norm": 11.838342547088189,
+      "learning_rate": 3.887208764910749e-06,
+      "loss": 0.6099,
+      "step": 6540
+    },
+    {
+      "epoch": 0.3339621679498292,
+      "grad_norm": 9.848289485775439,
+      "learning_rate": 3.8837722935453615e-06,
+      "loss": 0.625,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3344720338550961,
+      "grad_norm": 20.69381486392122,
+      "learning_rate": 3.880332048538789e-06,
+      "loss": 0.5759,
+      "step": 6560
+    },
+    {
+      "epoch": 0.334981899760363,
+      "grad_norm": 6.137002587152166,
+      "learning_rate": 3.876888039272818e-06,
+      "loss": 0.6291,
+      "step": 6570
+    },
+    {
+      "epoch": 0.33549176566562994,
+      "grad_norm": 8.858939897615745,
+      "learning_rate": 3.8734402751394925e-06,
+      "loss": 0.6023,
+      "step": 6580
+    },
+    {
+      "epoch": 0.33600163157089685,
+      "grad_norm": 5.371446280698542,
+      "learning_rate": 3.869988765541101e-06,
+      "loss": 0.5707,
+      "step": 6590
+    },
+    {
+      "epoch": 0.33651149747616377,
+      "grad_norm": 13.449325689930731,
+      "learning_rate": 3.866533519890145e-06,
+      "loss": 0.5676,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3370213633814307,
+      "grad_norm": 7.58335388392582,
+      "learning_rate": 3.863074547609314e-06,
+      "loss": 0.5677,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3375312292866976,
+      "grad_norm": 6.048763372455439,
+      "learning_rate": 3.859611858131461e-06,
+      "loss": 0.6223,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3380410951919645,
+      "grad_norm": 8.43049059804318,
+      "learning_rate": 3.8561454608995765e-06,
+      "loss": 0.6523,
+      "step": 6630
+    },
+    {
+      "epoch": 0.33855096109723143,
+      "grad_norm": 14.456986021929183,
+      "learning_rate": 3.852675365366762e-06,
+      "loss": 0.5357,
+      "step": 6640
+    },
+    {
+      "epoch": 0.33906082700249834,
+      "grad_norm": 8.698736924699832,
+      "learning_rate": 3.849201580996201e-06,
+      "loss": 0.6119,
+      "step": 6650
+    },
+    {
+      "epoch": 0.33957069290776526,
+      "grad_norm": 12.227061026125032,
+      "learning_rate": 3.845724117261142e-06,
+      "loss": 0.6333,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3400805588130322,
+      "grad_norm": 7.964324332305248,
+      "learning_rate": 3.8422429836448665e-06,
+      "loss": 0.632,
+      "step": 6670
+    },
+    {
+      "epoch": 0.3405904247182991,
+      "grad_norm": 9.02347624995148,
+      "learning_rate": 3.83875818964066e-06,
+      "loss": 0.5645,
+      "step": 6680
+    },
+    {
+      "epoch": 0.341100290623566,
+      "grad_norm": 11.481259913748172,
+      "learning_rate": 3.835269744751796e-06,
+      "loss": 0.6003,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3416101565288329,
+      "grad_norm": 13.058847885580379,
+      "learning_rate": 3.831777658491497e-06,
+      "loss": 0.6493,
+      "step": 6700
+    },
+    {
+      "epoch": 0.34212002243409984,
+      "grad_norm": 36.57172275862477,
+      "learning_rate": 3.828281940382923e-06,
+      "loss": 0.6713,
+      "step": 6710
+    },
+    {
+      "epoch": 0.34262988833936675,
+      "grad_norm": 17.121028748445738,
+      "learning_rate": 3.824782599959134e-06,
+      "loss": 0.6546,
+      "step": 6720
+    },
+    {
+      "epoch": 0.34313975424463367,
+      "grad_norm": 4.667001744714278,
+      "learning_rate": 3.8212796467630685e-06,
+      "loss": 0.5635,
+      "step": 6730
+    },
+    {
+      "epoch": 0.3436496201499006,
+      "grad_norm": 6.6063555790808595,
+      "learning_rate": 3.8177730903475195e-06,
+      "loss": 0.5727,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3441594860551675,
+      "grad_norm": 9.254337862184316,
+      "learning_rate": 3.8142629402751047e-06,
+      "loss": 0.6035,
+      "step": 6750
+    },
+    {
+      "epoch": 0.3446693519604344,
+      "grad_norm": 7.310130343220906,
+      "learning_rate": 3.8107492061182418e-06,
+      "loss": 0.6365,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3451792178657013,
+      "grad_norm": 7.347731976740374,
+      "learning_rate": 3.8072318974591233e-06,
+      "loss": 0.5842,
+      "step": 6770
+    },
+    {
+      "epoch": 0.34568908377096824,
+      "grad_norm": 5.516178192770456,
+      "learning_rate": 3.803711023889688e-06,
+      "loss": 0.5975,
+      "step": 6780
+    },
+    {
+      "epoch": 0.34619894967623516,
+      "grad_norm": 9.877992028377761,
+      "learning_rate": 3.800186595011599e-06,
+      "loss": 0.6027,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3467088155815021,
+      "grad_norm": 8.263217660967262,
+      "learning_rate": 3.7966586204362137e-06,
+      "loss": 0.5963,
+      "step": 6800
+    },
+    {
+      "epoch": 0.347218681486769,
+      "grad_norm": 22.684240498659832,
+      "learning_rate": 3.793127109784558e-06,
+      "loss": 0.6034,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3477285473920359,
+      "grad_norm": 11.932025654487864,
+      "learning_rate": 3.789592072687302e-06,
+      "loss": 0.6842,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3482384132973028,
+      "grad_norm": 5.99767425388713,
+      "learning_rate": 3.7860535187847326e-06,
+      "loss": 0.5786,
+      "step": 6830
+    },
+    {
+      "epoch": 0.34874827920256973,
+      "grad_norm": 11.542318476320958,
+      "learning_rate": 3.782511457726725e-06,
+      "loss": 0.6377,
+      "step": 6840
+    },
+    {
+      "epoch": 0.34925814510783665,
+      "grad_norm": 8.06618580589015,
+      "learning_rate": 3.7789658991727242e-06,
+      "loss": 0.5589,
+      "step": 6850
+    },
+    {
+      "epoch": 0.34976801101310356,
+      "grad_norm": 6.172507472614161,
+      "learning_rate": 3.775416852791707e-06,
+      "loss": 0.5888,
+      "step": 6860
+    },
+    {
+      "epoch": 0.3502778769183705,
+      "grad_norm": 13.480236409716568,
+      "learning_rate": 3.7718643282621648e-06,
+      "loss": 0.5631,
+      "step": 6870
+    },
+    {
+      "epoch": 0.3507877428236374,
+      "grad_norm": 8.245968547125749,
+      "learning_rate": 3.768308335272075e-06,
+      "loss": 0.5725,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3512976087289043,
+      "grad_norm": 5.113284315645673,
+      "learning_rate": 3.7647488835188705e-06,
+      "loss": 0.5578,
+      "step": 6890
+    },
+    {
+      "epoch": 0.3518074746341712,
+      "grad_norm": 21.845139952027658,
+      "learning_rate": 3.76118598270942e-06,
+      "loss": 0.6442,
+      "step": 6900
+    },
+    {
+      "epoch": 0.35231734053943814,
+      "grad_norm": 5.808958140251627,
+      "learning_rate": 3.7576196425599944e-06,
+      "loss": 0.5623,
+      "step": 6910
+    },
+    {
+      "epoch": 0.35282720644470505,
+      "grad_norm": 12.8925779686756,
+      "learning_rate": 3.7540498727962483e-06,
+      "loss": 0.617,
+      "step": 6920
+    },
+    {
+      "epoch": 0.35333707234997197,
+      "grad_norm": 6.351102780495482,
+      "learning_rate": 3.7504766831531848e-06,
+      "loss": 0.6059,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3538469382552389,
+      "grad_norm": 15.781230743083418,
+      "learning_rate": 3.746900083375137e-06,
+      "loss": 0.602,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3543568041605058,
+      "grad_norm": 21.198227388450018,
+      "learning_rate": 3.7433200832157333e-06,
+      "loss": 0.7074,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3548666700657727,
+      "grad_norm": 7.172725885534685,
+      "learning_rate": 3.7397366924378797e-06,
+      "loss": 0.6029,
+      "step": 6960
+    },
+    {
+      "epoch": 0.35537653597103963,
+      "grad_norm": 7.408637085264887,
+      "learning_rate": 3.736149920813726e-06,
+      "loss": 0.6019,
+      "step": 6970
+    },
+    {
+      "epoch": 0.35588640187630655,
+      "grad_norm": 7.798488548214566,
+      "learning_rate": 3.7325597781246426e-06,
+      "loss": 0.583,
+      "step": 6980
+    },
+    {
+      "epoch": 0.35639626778157346,
+      "grad_norm": 9.20496366906781,
+      "learning_rate": 3.728966274161192e-06,
+      "loss": 0.554,
+      "step": 6990
+    },
+    {
+      "epoch": 0.3569061336868404,
+      "grad_norm": 10.6846708940764,
+      "learning_rate": 3.725369418723106e-06,
+      "loss": 0.5573,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3574159995921073,
+      "grad_norm": 6.499329020438048,
+      "learning_rate": 3.721769221619252e-06,
+      "loss": 0.5126,
+      "step": 7010
+    },
+    {
+      "epoch": 0.3579258654973742,
+      "grad_norm": 4.232004126510205,
+      "learning_rate": 3.718165692667613e-06,
+      "loss": 0.5866,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3584357314026411,
+      "grad_norm": 6.522779753834059,
+      "learning_rate": 3.714558841695258e-06,
+      "loss": 0.6243,
+      "step": 7030
+    },
+    {
+      "epoch": 0.35894559730790804,
+      "grad_norm": 7.842308577918714,
+      "learning_rate": 3.710948678538314e-06,
+      "loss": 0.5961,
+      "step": 7040
+    },
+    {
+      "epoch": 0.35945546321317495,
+      "grad_norm": 4.155939637670406,
+      "learning_rate": 3.7073352130419436e-06,
+      "loss": 0.6546,
+      "step": 7050
+    },
+    {
+      "epoch": 0.35996532911844187,
+      "grad_norm": 15.06151436559477,
+      "learning_rate": 3.703718455060311e-06,
+      "loss": 0.6156,
+      "step": 7060
+    },
+    {
+      "epoch": 0.3604751950237088,
+      "grad_norm": 10.46655056093518,
+      "learning_rate": 3.7000984144565617e-06,
+      "loss": 0.6271,
+      "step": 7070
+    },
+    {
+      "epoch": 0.3609850609289757,
+      "grad_norm": 10.530339938427321,
+      "learning_rate": 3.6964751011027933e-06,
+      "loss": 0.5611,
+      "step": 7080
+    },
+    {
+      "epoch": 0.3614949268342426,
+      "grad_norm": 11.596468152577025,
+      "learning_rate": 3.6928485248800273e-06,
+      "loss": 0.5413,
+      "step": 7090
+    },
+    {
+      "epoch": 0.36200479273950953,
+      "grad_norm": 7.438067886146539,
+      "learning_rate": 3.689218695678184e-06,
+      "loss": 0.5483,
+      "step": 7100
+    },
+    {
+      "epoch": 0.36251465864477644,
+      "grad_norm": 5.0417594250753055,
+      "learning_rate": 3.685585623396055e-06,
+      "loss": 0.5861,
+      "step": 7110
+    },
+    {
+      "epoch": 0.36302452455004336,
+      "grad_norm": 6.086227491760784,
+      "learning_rate": 3.681949317941275e-06,
+      "loss": 0.5121,
+      "step": 7120
+    },
+    {
+      "epoch": 0.3635343904553103,
+      "grad_norm": 8.564698930366653,
+      "learning_rate": 3.6783097892302967e-06,
+      "loss": 0.6275,
+      "step": 7130
+    },
+    {
+      "epoch": 0.3640442563605772,
+      "grad_norm": 6.506669188127755,
+      "learning_rate": 3.6746670471883616e-06,
+      "loss": 0.5806,
+      "step": 7140
+    },
+    {
+      "epoch": 0.3645541222658441,
+      "grad_norm": 7.390307218413792,
+      "learning_rate": 3.671021101749476e-06,
+      "loss": 0.5798,
+      "step": 7150
+    },
+    {
+      "epoch": 0.365063988171111,
+      "grad_norm": 10.553770888956274,
+      "learning_rate": 3.6673719628563808e-06,
+      "loss": 0.5747,
+      "step": 7160
+    },
+    {
+      "epoch": 0.36557385407637794,
+      "grad_norm": 24.14877566048891,
+      "learning_rate": 3.6637196404605257e-06,
+      "loss": 0.5573,
+      "step": 7170
+    },
+    {
+      "epoch": 0.36608371998164485,
+      "grad_norm": 13.172173489692812,
+      "learning_rate": 3.660064144522043e-06,
+      "loss": 0.5897,
+      "step": 7180
+    },
+    {
+      "epoch": 0.36659358588691177,
+      "grad_norm": 6.2257304828288955,
+      "learning_rate": 3.656405485009719e-06,
+      "loss": 0.5682,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3671034517921787,
+      "grad_norm": 12.945880033725963,
+      "learning_rate": 3.652743671900967e-06,
+      "loss": 0.5714,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3676133176974456,
+      "grad_norm": 7.236111657688771,
+      "learning_rate": 3.6490787151817986e-06,
+      "loss": 0.5455,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3681231836027125,
+      "grad_norm": 8.565857394457222,
+      "learning_rate": 3.6454106248468024e-06,
+      "loss": 0.6042,
+      "step": 7220
+    },
+    {
+      "epoch": 0.3686330495079794,
+      "grad_norm": 5.938790795920517,
+      "learning_rate": 3.6417394108991096e-06,
+      "loss": 0.5757,
+      "step": 7230
+    },
+    {
+      "epoch": 0.36914291541324634,
+      "grad_norm": 16.58193027443637,
+      "learning_rate": 3.6380650833503705e-06,
+      "loss": 0.5971,
+      "step": 7240
+    },
+    {
+      "epoch": 0.36965278131851326,
+      "grad_norm": 11.450948063607477,
+      "learning_rate": 3.6343876522207253e-06,
+      "loss": 0.6397,
+      "step": 7250
+    },
+    {
+      "epoch": 0.3701626472237802,
+      "grad_norm": 9.41451555929652,
+      "learning_rate": 3.6307071275387807e-06,
+      "loss": 0.5914,
+      "step": 7260
+    },
+    {
+      "epoch": 0.3706725131290471,
+      "grad_norm": 12.429290516145963,
+      "learning_rate": 3.6270235193415754e-06,
+      "loss": 0.5679,
+      "step": 7270
+    },
+    {
+      "epoch": 0.371182379034314,
+      "grad_norm": 13.72537185817539,
+      "learning_rate": 3.6233368376745616e-06,
+      "loss": 0.6502,
+      "step": 7280
+    },
+    {
+      "epoch": 0.37169224493958086,
+      "grad_norm": 9.211876210956733,
+      "learning_rate": 3.6196470925915705e-06,
+      "loss": 0.6053,
+      "step": 7290
+    },
+    {
+      "epoch": 0.3722021108448478,
+      "grad_norm": 6.674342793228114,
+      "learning_rate": 3.6159542941547883e-06,
+      "loss": 0.57,
+      "step": 7300
+    },
+    {
+      "epoch": 0.3727119767501147,
+      "grad_norm": 5.06011137522497,
+      "learning_rate": 3.6122584524347267e-06,
+      "loss": 0.6127,
+      "step": 7310
+    },
+    {
+      "epoch": 0.3732218426553816,
+      "grad_norm": 5.50419488018807,
+      "learning_rate": 3.608559577510198e-06,
+      "loss": 0.5693,
+      "step": 7320
+    },
+    {
+      "epoch": 0.3737317085606485,
+      "grad_norm": 11.98832355243555,
+      "learning_rate": 3.604857679468285e-06,
+      "loss": 0.5676,
+      "step": 7330
+    },
+    {
+      "epoch": 0.37424157446591544,
+      "grad_norm": 25.368564304510052,
+      "learning_rate": 3.601152768404317e-06,
+      "loss": 0.5902,
+      "step": 7340
+    },
+    {
+      "epoch": 0.37475144037118235,
+      "grad_norm": 13.241496574133155,
+      "learning_rate": 3.597444854421837e-06,
+      "loss": 0.6013,
+      "step": 7350
+    },
+    {
+      "epoch": 0.37526130627644927,
+      "grad_norm": 9.605414734958668,
+      "learning_rate": 3.59373394763258e-06,
+      "loss": 0.5892,
+      "step": 7360
+    },
+    {
+      "epoch": 0.3757711721817162,
+      "grad_norm": 10.50506861161194,
+      "learning_rate": 3.5900200581564403e-06,
+      "loss": 0.634,
+      "step": 7370
+    },
+    {
+      "epoch": 0.3762810380869831,
+      "grad_norm": 6.366109630402361,
+      "learning_rate": 3.586303196121447e-06,
+      "loss": 0.5825,
+      "step": 7380
+    },
+    {
+      "epoch": 0.37679090399225,
+      "grad_norm": 6.938308539769015,
+      "learning_rate": 3.5825833716637364e-06,
+      "loss": 0.6521,
+      "step": 7390
+    },
+    {
+      "epoch": 0.37730076989751693,
+      "grad_norm": 26.03998674798817,
+      "learning_rate": 3.5788605949275237e-06,
+      "loss": 0.5358,
+      "step": 7400
+    },
+    {
+      "epoch": 0.37781063580278385,
+      "grad_norm": 11.548585198622813,
+      "learning_rate": 3.5751348760650722e-06,
+      "loss": 0.6379,
+      "step": 7410
+    },
+    {
+      "epoch": 0.37832050170805076,
+      "grad_norm": 5.318555598784323,
+      "learning_rate": 3.5714062252366723e-06,
+      "loss": 0.6551,
+      "step": 7420
+    },
+    {
+      "epoch": 0.3788303676133177,
+      "grad_norm": 5.478810952612762,
+      "learning_rate": 3.5676746526106084e-06,
+      "loss": 0.6603,
+      "step": 7430
+    },
+    {
+      "epoch": 0.3793402335185846,
+      "grad_norm": 9.875012211939511,
+      "learning_rate": 3.5639401683631314e-06,
+      "loss": 0.647,
+      "step": 7440
+    },
+    {
+      "epoch": 0.3798500994238515,
+      "grad_norm": 13.66351116178773,
+      "learning_rate": 3.5602027826784356e-06,
+      "loss": 0.6437,
+      "step": 7450
+    },
+    {
+      "epoch": 0.3803599653291184,
+      "grad_norm": 3.3576915666147977,
+      "learning_rate": 3.556462505748625e-06,
+      "loss": 0.5802,
+      "step": 7460
+    },
+    {
+      "epoch": 0.38086983123438534,
+      "grad_norm": 6.045967456335718,
+      "learning_rate": 3.5527193477736903e-06,
+      "loss": 0.5692,
+      "step": 7470
+    },
+    {
+      "epoch": 0.38137969713965225,
+      "grad_norm": 5.119077954410416,
+      "learning_rate": 3.548973318961477e-06,
+      "loss": 0.5656,
+      "step": 7480
+    },
+    {
+      "epoch": 0.38188956304491917,
+      "grad_norm": 6.179648411829707,
+      "learning_rate": 3.5452244295276604e-06,
+      "loss": 0.6491,
+      "step": 7490
+    },
+    {
+      "epoch": 0.3823994289501861,
+      "grad_norm": 16.80155654981562,
+      "learning_rate": 3.541472689695718e-06,
+      "loss": 0.588,
+      "step": 7500
+    },
+    {
+      "epoch": 0.382909294855453,
+      "grad_norm": 14.227034706280053,
+      "learning_rate": 3.537718109696899e-06,
+      "loss": 0.5874,
+      "step": 7510
+    },
+    {
+      "epoch": 0.3834191607607199,
+      "grad_norm": 6.082648099096069,
+      "learning_rate": 3.5339606997701975e-06,
+      "loss": 0.5422,
+      "step": 7520
+    },
+    {
+      "epoch": 0.38392902666598683,
+      "grad_norm": 22.201688715390624,
+      "learning_rate": 3.530200470162328e-06,
+      "loss": 0.6293,
+      "step": 7530
+    },
+    {
+      "epoch": 0.38443889257125374,
+      "grad_norm": 4.117638443340211,
+      "learning_rate": 3.5264374311276904e-06,
+      "loss": 0.5174,
+      "step": 7540
+    },
+    {
+      "epoch": 0.38494875847652066,
+      "grad_norm": 18.482761872963568,
+      "learning_rate": 3.5226715929283507e-06,
+      "loss": 0.618,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3854586243817876,
+      "grad_norm": 6.392047812297191,
+      "learning_rate": 3.518902965834003e-06,
+      "loss": 0.594,
+      "step": 7560
+    },
+    {
+      "epoch": 0.3859684902870545,
+      "grad_norm": 8.55473811171792,
+      "learning_rate": 3.5151315601219517e-06,
+      "loss": 0.5521,
+      "step": 7570
+    },
+    {
+      "epoch": 0.3864783561923214,
+      "grad_norm": 9.050740419704756,
+      "learning_rate": 3.5113573860770755e-06,
+      "loss": 0.582,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3869882220975883,
+      "grad_norm": 25.688592583626388,
+      "learning_rate": 3.5075804539918047e-06,
+      "loss": 0.7061,
+      "step": 7590
+    },
+    {
+      "epoch": 0.38749808800285523,
+      "grad_norm": 7.353214023143235,
+      "learning_rate": 3.5038007741660895e-06,
+      "loss": 0.5931,
+      "step": 7600
+    },
+    {
+      "epoch": 0.38800795390812215,
+      "grad_norm": 14.032425976466067,
+      "learning_rate": 3.5000183569073743e-06,
+      "loss": 0.6448,
+      "step": 7610
+    },
+    {
+      "epoch": 0.38851781981338906,
+      "grad_norm": 4.8340970822759575,
+      "learning_rate": 3.4962332125305686e-06,
+      "loss": 0.5511,
+      "step": 7620
+    },
+    {
+      "epoch": 0.389027685718656,
+      "grad_norm": 10.049457628084676,
+      "learning_rate": 3.492445351358018e-06,
+      "loss": 0.5893,
+      "step": 7630
+    },
+    {
+      "epoch": 0.3895375516239229,
+      "grad_norm": 6.418594454021383,
+      "learning_rate": 3.488654783719477e-06,
+      "loss": 0.5619,
+      "step": 7640
+    },
+    {
+      "epoch": 0.3900474175291898,
+      "grad_norm": 9.276784267515474,
+      "learning_rate": 3.484861519952083e-06,
+      "loss": 0.5815,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3905572834344567,
+      "grad_norm": 16.507690557273907,
+      "learning_rate": 3.4810655704003237e-06,
+      "loss": 0.5685,
+      "step": 7660
+    },
+    {
+      "epoch": 0.39106714933972364,
+      "grad_norm": 9.18372974531889,
+      "learning_rate": 3.4772669454160113e-06,
+      "loss": 0.5539,
+      "step": 7670
+    },
+    {
+      "epoch": 0.39157701524499056,
+      "grad_norm": 9.95648456172559,
+      "learning_rate": 3.473465655358255e-06,
+      "loss": 0.5408,
+      "step": 7680
+    },
+    {
+      "epoch": 0.39208688115025747,
+      "grad_norm": 7.154378316140561,
+      "learning_rate": 3.469661710593431e-06,
+      "loss": 0.5849,
+      "step": 7690
+    },
+    {
+      "epoch": 0.3925967470555244,
+      "grad_norm": 4.296446896204838,
+      "learning_rate": 3.465855121495156e-06,
+      "loss": 0.4575,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3931066129607913,
+      "grad_norm": 11.971727859126235,
+      "learning_rate": 3.4620458984442564e-06,
+      "loss": 0.553,
+      "step": 7710
+    },
+    {
+      "epoch": 0.3936164788660582,
+      "grad_norm": 8.056036290422684,
+      "learning_rate": 3.4582340518287444e-06,
+      "loss": 0.5746,
+      "step": 7720
+    },
+    {
+      "epoch": 0.39412634477132513,
+      "grad_norm": 3.9615637901197744,
+      "learning_rate": 3.4544195920437834e-06,
+      "loss": 0.5716,
+      "step": 7730
+    },
+    {
+      "epoch": 0.39463621067659205,
+      "grad_norm": 10.049166058251805,
+      "learning_rate": 3.4506025294916656e-06,
+      "loss": 0.585,
+      "step": 7740
+    },
+    {
+      "epoch": 0.39514607658185896,
+      "grad_norm": 10.449434153164734,
+      "learning_rate": 3.44678287458178e-06,
+      "loss": 0.6304,
+      "step": 7750
+    },
+    {
+      "epoch": 0.3956559424871259,
+      "grad_norm": 6.927454525401204,
+      "learning_rate": 3.4429606377305847e-06,
+      "loss": 0.6219,
+      "step": 7760
+    },
+    {
+      "epoch": 0.3961658083923928,
+      "grad_norm": 4.870437160417047,
+      "learning_rate": 3.439135829361581e-06,
+      "loss": 0.5676,
+      "step": 7770
+    },
+    {
+      "epoch": 0.3966756742976597,
+      "grad_norm": 17.89544614300908,
+      "learning_rate": 3.435308459905281e-06,
+      "loss": 0.6855,
+      "step": 7780
+    },
+    {
+      "epoch": 0.3971855402029266,
+      "grad_norm": 12.771844660858063,
+      "learning_rate": 3.4314785397991814e-06,
+      "loss": 0.5657,
+      "step": 7790
+    },
+    {
+      "epoch": 0.39769540610819354,
+      "grad_norm": 7.7960650699904805,
+      "learning_rate": 3.4276460794877343e-06,
+      "loss": 0.5664,
+      "step": 7800
+    },
+    {
+      "epoch": 0.39820527201346045,
+      "grad_norm": 7.228542637831989,
+      "learning_rate": 3.4238110894223205e-06,
+      "loss": 0.6863,
+      "step": 7810
+    },
+    {
+      "epoch": 0.39871513791872737,
+      "grad_norm": 7.760845660983971,
+      "learning_rate": 3.419973580061218e-06,
+      "loss": 0.5669,
+      "step": 7820
+    },
+    {
+      "epoch": 0.3992250038239943,
+      "grad_norm": 12.392495438781795,
+      "learning_rate": 3.4161335618695774e-06,
+      "loss": 0.6402,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3997348697292612,
+      "grad_norm": 26.611025449613667,
+      "learning_rate": 3.4122910453193885e-06,
+      "loss": 0.5825,
+      "step": 7840
+    },
+    {
+      "epoch": 0.4002447356345281,
+      "grad_norm": 18.091448185422944,
+      "learning_rate": 3.4084460408894553e-06,
+      "loss": 0.6152,
+      "step": 7850
+    },
+    {
+      "epoch": 0.40075460153979503,
+      "grad_norm": 15.72425633199789,
+      "learning_rate": 3.4045985590653667e-06,
+      "loss": 0.611,
+      "step": 7860
+    },
+    {
+      "epoch": 0.40126446744506195,
+      "grad_norm": 10.798561704841136,
+      "learning_rate": 3.4007486103394678e-06,
+      "loss": 0.6159,
+      "step": 7870
+    },
+    {
+      "epoch": 0.40177433335032886,
+      "grad_norm": 3.307220226809657,
+      "learning_rate": 3.3968962052108288e-06,
+      "loss": 0.5643,
+      "step": 7880
+    },
+    {
+      "epoch": 0.4022841992555958,
+      "grad_norm": 13.995794219992154,
+      "learning_rate": 3.3930413541852235e-06,
+      "loss": 0.6081,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4027940651608627,
+      "grad_norm": 7.517626739286846,
+      "learning_rate": 3.389184067775091e-06,
+      "loss": 0.5687,
+      "step": 7900
+    },
+    {
+      "epoch": 0.4033039310661296,
+      "grad_norm": 8.606794051353527,
+      "learning_rate": 3.385324356499513e-06,
+      "loss": 0.5603,
+      "step": 7910
+    },
+    {
+      "epoch": 0.4038137969713965,
+      "grad_norm": 9.684832646327461,
+      "learning_rate": 3.381462230884187e-06,
+      "loss": 0.5539,
+      "step": 7920
+    },
+    {
+      "epoch": 0.40432366287666344,
+      "grad_norm": 4.1687295402302444,
+      "learning_rate": 3.377597701461391e-06,
+      "loss": 0.538,
+      "step": 7930
+    },
+    {
+      "epoch": 0.40483352878193035,
+      "grad_norm": 27.225147980543973,
+      "learning_rate": 3.3737307787699593e-06,
+      "loss": 0.5857,
+      "step": 7940
+    },
+    {
+      "epoch": 0.40534339468719727,
+      "grad_norm": 4.638533674818879,
+      "learning_rate": 3.3698614733552537e-06,
+      "loss": 0.5602,
+      "step": 7950
+    },
+    {
+      "epoch": 0.4058532605924642,
+      "grad_norm": 10.400432513347486,
+      "learning_rate": 3.3659897957691334e-06,
+      "loss": 0.6512,
+      "step": 7960
+    },
+    {
+      "epoch": 0.4063631264977311,
+      "grad_norm": 5.130756363065307,
+      "learning_rate": 3.3621157565699265e-06,
+      "loss": 0.5803,
+      "step": 7970
+    },
+    {
+      "epoch": 0.406872992402998,
+      "grad_norm": 7.271116073711001,
+      "learning_rate": 3.3582393663224012e-06,
+      "loss": 0.5683,
+      "step": 7980
+    },
+    {
+      "epoch": 0.40738285830826493,
+      "grad_norm": 7.359012484464728,
+      "learning_rate": 3.3543606355977377e-06,
+      "loss": 0.5638,
+      "step": 7990
+    },
+    {
+      "epoch": 0.40789272421353184,
+      "grad_norm": 7.415959535609534,
+      "learning_rate": 3.350479574973498e-06,
+      "loss": 0.6123,
+      "step": 8000
+    },
+    {
+      "epoch": 0.40840259011879876,
+      "grad_norm": 7.811269462367596,
+      "learning_rate": 3.3465961950335994e-06,
+      "loss": 0.5595,
+      "step": 8010
+    },
+    {
+      "epoch": 0.4089124560240657,
+      "grad_norm": 7.989979255664991,
+      "learning_rate": 3.342710506368282e-06,
+      "loss": 0.5159,
+      "step": 8020
+    },
+    {
+      "epoch": 0.4094223219293326,
+      "grad_norm": 7.869848589658554,
+      "learning_rate": 3.3388225195740844e-06,
+      "loss": 0.5517,
+      "step": 8030
+    },
+    {
+      "epoch": 0.4099321878345995,
+      "grad_norm": 13.414951394387845,
+      "learning_rate": 3.33493224525381e-06,
+      "loss": 0.5044,
+      "step": 8040
+    },
+    {
+      "epoch": 0.4104420537398664,
+      "grad_norm": 7.250013552370468,
+      "learning_rate": 3.3310396940165025e-06,
+      "loss": 0.6043,
+      "step": 8050
+    },
+    {
+      "epoch": 0.41095191964513333,
+      "grad_norm": 29.849356149503027,
+      "learning_rate": 3.327144876477413e-06,
+      "loss": 0.6256,
+      "step": 8060
+    },
+    {
+      "epoch": 0.41146178555040025,
+      "grad_norm": 6.196410170936335,
+      "learning_rate": 3.3232478032579746e-06,
+      "loss": 0.5695,
+      "step": 8070
+    },
+    {
+      "epoch": 0.41197165145566716,
+      "grad_norm": 15.64304736056121,
+      "learning_rate": 3.319348484985771e-06,
+      "loss": 0.5448,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4124815173609341,
+      "grad_norm": 5.370956738872842,
+      "learning_rate": 3.3154469322945083e-06,
+      "loss": 0.5642,
+      "step": 8090
+    },
+    {
+      "epoch": 0.412991383266201,
+      "grad_norm": 10.279572349290994,
+      "learning_rate": 3.311543155823985e-06,
+      "loss": 0.5895,
+      "step": 8100
+    },
+    {
+      "epoch": 0.4135012491714679,
+      "grad_norm": 10.29244692765525,
+      "learning_rate": 3.3076371662200666e-06,
+      "loss": 0.5625,
+      "step": 8110
+    },
+    {
+      "epoch": 0.4140111150767348,
+      "grad_norm": 10.44654118728628,
+      "learning_rate": 3.303728974134653e-06,
+      "loss": 0.5469,
+      "step": 8120
+    },
+    {
+      "epoch": 0.41452098098200174,
+      "grad_norm": 4.83706374349343,
+      "learning_rate": 3.2998185902256475e-06,
+      "loss": 0.5821,
+      "step": 8130
+    },
+    {
+      "epoch": 0.41503084688726866,
+      "grad_norm": 11.49003309549299,
+      "learning_rate": 3.295906025156935e-06,
+      "loss": 0.6202,
+      "step": 8140
+    },
+    {
+      "epoch": 0.41554071279253557,
+      "grad_norm": 10.702272759661504,
+      "learning_rate": 3.291991289598347e-06,
+      "loss": 0.5714,
+      "step": 8150
+    },
+    {
+      "epoch": 0.4160505786978025,
+      "grad_norm": 45.2248551956823,
+      "learning_rate": 3.288074394225632e-06,
+      "loss": 0.6211,
+      "step": 8160
+    },
+    {
+      "epoch": 0.4165604446030694,
+      "grad_norm": 7.4598902579327335,
+      "learning_rate": 3.284155349720431e-06,
+      "loss": 0.6028,
+      "step": 8170
+    },
+    {
+      "epoch": 0.4170703105083363,
+      "grad_norm": 6.274399449869339,
+      "learning_rate": 3.2802341667702448e-06,
+      "loss": 0.6018,
+      "step": 8180
+    },
+    {
+      "epoch": 0.41758017641360323,
+      "grad_norm": 3.9008731665441356,
+      "learning_rate": 3.276310856068406e-06,
+      "loss": 0.6355,
+      "step": 8190
+    },
+    {
+      "epoch": 0.41809004231887015,
+      "grad_norm": 13.84695131835591,
+      "learning_rate": 3.27238542831405e-06,
+      "loss": 0.5286,
+      "step": 8200
+    },
+    {
+      "epoch": 0.41859990822413706,
+      "grad_norm": 9.644587842783794,
+      "learning_rate": 3.2684578942120853e-06,
+      "loss": 0.5411,
+      "step": 8210
+    },
+    {
+      "epoch": 0.419109774129404,
+      "grad_norm": 9.161992322868674,
+      "learning_rate": 3.2645282644731648e-06,
+      "loss": 0.6141,
+      "step": 8220
+    },
+    {
+      "epoch": 0.4196196400346709,
+      "grad_norm": 6.903980241843122,
+      "learning_rate": 3.2605965498136554e-06,
+      "loss": 0.5841,
+      "step": 8230
+    },
+    {
+      "epoch": 0.4201295059399378,
+      "grad_norm": 6.035862909552343,
+      "learning_rate": 3.2566627609556117e-06,
+      "loss": 0.5653,
+      "step": 8240
+    },
+    {
+      "epoch": 0.4206393718452047,
+      "grad_norm": 8.105812505508382,
+      "learning_rate": 3.252726908626742e-06,
+      "loss": 0.6354,
+      "step": 8250
+    },
+    {
+      "epoch": 0.42114923775047164,
+      "grad_norm": 11.937584745156768,
+      "learning_rate": 3.248789003560385e-06,
+      "loss": 0.5163,
+      "step": 8260
+    },
+    {
+      "epoch": 0.42165910365573855,
+      "grad_norm": 21.76832746956224,
+      "learning_rate": 3.2448490564954744e-06,
+      "loss": 0.5452,
+      "step": 8270
+    },
+    {
+      "epoch": 0.42216896956100547,
+      "grad_norm": 18.051485655686268,
+      "learning_rate": 3.2409070781765147e-06,
+      "loss": 0.5664,
+      "step": 8280
+    },
+    {
+      "epoch": 0.4226788354662724,
+      "grad_norm": 8.423685743223801,
+      "learning_rate": 3.236963079353548e-06,
+      "loss": 0.561,
+      "step": 8290
+    },
+    {
+      "epoch": 0.4231887013715393,
+      "grad_norm": 8.624624976972216,
+      "learning_rate": 3.2330170707821283e-06,
+      "loss": 0.5843,
+      "step": 8300
+    },
+    {
+      "epoch": 0.4236985672768062,
+      "grad_norm": 6.933716611383655,
+      "learning_rate": 3.229069063223289e-06,
+      "loss": 0.5982,
+      "step": 8310
+    },
+    {
+      "epoch": 0.42420843318207313,
+      "grad_norm": 6.308662920804664,
+      "learning_rate": 3.225119067443515e-06,
+      "loss": 0.5909,
+      "step": 8320
+    },
+    {
+      "epoch": 0.42471829908734005,
+      "grad_norm": 6.275184288671406,
+      "learning_rate": 3.2211670942147144e-06,
+      "loss": 0.5669,
+      "step": 8330
+    },
+    {
+      "epoch": 0.42522816499260696,
+      "grad_norm": 8.569157246885606,
+      "learning_rate": 3.2172131543141865e-06,
+      "loss": 0.5746,
+      "step": 8340
+    },
+    {
+      "epoch": 0.4257380308978739,
+      "grad_norm": 6.307071145921682,
+      "learning_rate": 3.2132572585245946e-06,
+      "loss": 0.6255,
+      "step": 8350
+    },
+    {
+      "epoch": 0.4262478968031408,
+      "grad_norm": 9.047786484848844,
+      "learning_rate": 3.209299417633936e-06,
+      "loss": 0.608,
+      "step": 8360
+    },
+    {
+      "epoch": 0.4267577627084077,
+      "grad_norm": 7.183817587841396,
+      "learning_rate": 3.2053396424355105e-06,
+      "loss": 0.6324,
+      "step": 8370
+    },
+    {
+      "epoch": 0.4272676286136746,
+      "grad_norm": 4.004382030972596,
+      "learning_rate": 3.201377943727896e-06,
+      "loss": 0.5876,
+      "step": 8380
+    },
+    {
+      "epoch": 0.42777749451894154,
+      "grad_norm": 24.284714914627546,
+      "learning_rate": 3.197414332314914e-06,
+      "loss": 0.4996,
+      "step": 8390
+    },
+    {
+      "epoch": 0.42828736042420845,
+      "grad_norm": 14.820952554168857,
+      "learning_rate": 3.193448819005603e-06,
+      "loss": 0.6185,
+      "step": 8400
+    },
+    {
+      "epoch": 0.42879722632947537,
+      "grad_norm": 6.093851744759424,
+      "learning_rate": 3.189481414614186e-06,
+      "loss": 0.5609,
+      "step": 8410
+    },
+    {
+      "epoch": 0.4293070922347423,
+      "grad_norm": 7.259173190794586,
+      "learning_rate": 3.1855121299600454e-06,
+      "loss": 0.6312,
+      "step": 8420
+    },
+    {
+      "epoch": 0.4298169581400092,
+      "grad_norm": 8.653922058289407,
+      "learning_rate": 3.1815409758676917e-06,
+      "loss": 0.6608,
+      "step": 8430
+    },
+    {
+      "epoch": 0.4303268240452761,
+      "grad_norm": 5.552195512575827,
+      "learning_rate": 3.1775679631667306e-06,
+      "loss": 0.584,
+      "step": 8440
+    },
+    {
+      "epoch": 0.43083668995054303,
+      "grad_norm": 3.743477278951455,
+      "learning_rate": 3.1735931026918393e-06,
+      "loss": 0.6055,
+      "step": 8450
+    },
+    {
+      "epoch": 0.43134655585580994,
+      "grad_norm": 17.16382374151614,
+      "learning_rate": 3.1696164052827318e-06,
+      "loss": 0.5615,
+      "step": 8460
+    },
+    {
+      "epoch": 0.43185642176107686,
+      "grad_norm": 7.525184001694853,
+      "learning_rate": 3.165637881784133e-06,
+      "loss": 0.5091,
+      "step": 8470
+    },
+    {
+      "epoch": 0.4323662876663438,
+      "grad_norm": 7.806074822426972,
+      "learning_rate": 3.161657543045747e-06,
+      "loss": 0.6274,
+      "step": 8480
+    },
+    {
+      "epoch": 0.4328761535716107,
+      "grad_norm": 10.128093552583556,
+      "learning_rate": 3.1576753999222275e-06,
+      "loss": 0.5862,
+      "step": 8490
+    },
+    {
+      "epoch": 0.4333860194768776,
+      "grad_norm": 10.241501484791744,
+      "learning_rate": 3.1536914632731512e-06,
+      "loss": 0.5975,
+      "step": 8500
+    },
+    {
+      "epoch": 0.4338958853821445,
+      "grad_norm": 6.190804768770443,
+      "learning_rate": 3.1497057439629836e-06,
+      "loss": 0.5558,
+      "step": 8510
+    },
+    {
+      "epoch": 0.43440575128741143,
+      "grad_norm": 16.180652163795123,
+      "learning_rate": 3.1457182528610526e-06,
+      "loss": 0.6185,
+      "step": 8520
+    },
+    {
+      "epoch": 0.43491561719267835,
+      "grad_norm": 6.5378974213106416,
+      "learning_rate": 3.1417290008415167e-06,
+      "loss": 0.5623,
+      "step": 8530
+    },
+    {
+      "epoch": 0.43542548309794527,
+      "grad_norm": 17.793095921414608,
+      "learning_rate": 3.1377379987833395e-06,
+      "loss": 0.5711,
+      "step": 8540
+    },
+    {
+      "epoch": 0.4359353490032122,
+      "grad_norm": 5.698719713245101,
+      "learning_rate": 3.133745257570253e-06,
+      "loss": 0.6352,
+      "step": 8550
+    },
+    {
+      "epoch": 0.4364452149084791,
+      "grad_norm": 12.893025105305172,
+      "learning_rate": 3.1297507880907357e-06,
+      "loss": 0.4856,
+      "step": 8560
+    },
+    {
+      "epoch": 0.436955080813746,
+      "grad_norm": 8.399321065682123,
+      "learning_rate": 3.1257546012379775e-06,
+      "loss": 0.5586,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4374649467190129,
+      "grad_norm": 4.831675625103787,
+      "learning_rate": 3.121756707909851e-06,
+      "loss": 0.5308,
+      "step": 8580
+    },
+    {
+      "epoch": 0.43797481262427984,
+      "grad_norm": 5.955633182788592,
+      "learning_rate": 3.1177571190088833e-06,
+      "loss": 0.606,
+      "step": 8590
+    },
+    {
+      "epoch": 0.43848467852954676,
+      "grad_norm": 13.753186960996212,
+      "learning_rate": 3.1137558454422266e-06,
+      "loss": 0.5694,
+      "step": 8600
+    },
+    {
+      "epoch": 0.43899454443481367,
+      "grad_norm": 10.973002877580813,
+      "learning_rate": 3.1097528981216245e-06,
+      "loss": 0.5468,
+      "step": 8610
+    },
+    {
+      "epoch": 0.43950441034008053,
+      "grad_norm": 6.520010650488079,
+      "learning_rate": 3.105748287963388e-06,
+      "loss": 0.6218,
+      "step": 8620
+    },
+    {
+      "epoch": 0.44001427624534745,
+      "grad_norm": 4.663602488972492,
+      "learning_rate": 3.1017420258883607e-06,
+      "loss": 0.6092,
+      "step": 8630
+    },
+    {
+      "epoch": 0.44052414215061436,
+      "grad_norm": 6.0674522796006665,
+      "learning_rate": 3.0977341228218916e-06,
+      "loss": 0.5371,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4410340080558813,
+      "grad_norm": 6.258508973891013,
+      "learning_rate": 3.0937245896938062e-06,
+      "loss": 0.4981,
+      "step": 8650
+    },
+    {
+      "epoch": 0.4415438739611482,
+      "grad_norm": 7.511397618456565,
+      "learning_rate": 3.089713437438373e-06,
+      "loss": 0.556,
+      "step": 8660
+    },
+    {
+      "epoch": 0.4420537398664151,
+      "grad_norm": 10.901260436805492,
+      "learning_rate": 3.085700676994277e-06,
+      "loss": 0.5178,
+      "step": 8670
+    },
+    {
+      "epoch": 0.442563605771682,
+      "grad_norm": 4.032271770877195,
+      "learning_rate": 3.0816863193045898e-06,
+      "loss": 0.5299,
+      "step": 8680
+    },
+    {
+      "epoch": 0.44307347167694894,
+      "grad_norm": 4.682645234637967,
+      "learning_rate": 3.0776703753167382e-06,
+      "loss": 0.5409,
+      "step": 8690
+    },
+    {
+      "epoch": 0.44358333758221585,
+      "grad_norm": 6.869333096460406,
+      "learning_rate": 3.0736528559824736e-06,
+      "loss": 0.5774,
+      "step": 8700
+    },
+    {
+      "epoch": 0.44409320348748277,
+      "grad_norm": 5.646834228989253,
+      "learning_rate": 3.0696337722578444e-06,
+      "loss": 0.5627,
+      "step": 8710
+    },
+    {
+      "epoch": 0.4446030693927497,
+      "grad_norm": 5.011538749911393,
+      "learning_rate": 3.0656131351031663e-06,
+      "loss": 0.5484,
+      "step": 8720
+    },
+    {
+      "epoch": 0.4451129352980166,
+      "grad_norm": 39.470423402922066,
+      "learning_rate": 3.06159095548299e-06,
+      "loss": 0.5335,
+      "step": 8730
+    },
+    {
+      "epoch": 0.4456228012032835,
+      "grad_norm": 5.114310793982849,
+      "learning_rate": 3.057567244366072e-06,
+      "loss": 0.5413,
+      "step": 8740
+    },
+    {
+      "epoch": 0.44613266710855043,
+      "grad_norm": 13.21974253835841,
+      "learning_rate": 3.053542012725347e-06,
+      "loss": 0.6133,
+      "step": 8750
+    },
+    {
+      "epoch": 0.44664253301381734,
+      "grad_norm": 6.883499098740761,
+      "learning_rate": 3.049515271537896e-06,
+      "loss": 0.5629,
+      "step": 8760
+    },
+    {
+      "epoch": 0.44715239891908426,
+      "grad_norm": 8.128398135255848,
+      "learning_rate": 3.045487031784916e-06,
+      "loss": 0.5635,
+      "step": 8770
+    },
+    {
+      "epoch": 0.4476622648243512,
+      "grad_norm": 14.008364726787159,
+      "learning_rate": 3.041457304451691e-06,
+      "loss": 0.569,
+      "step": 8780
+    },
+    {
+      "epoch": 0.4481721307296181,
+      "grad_norm": 11.22813055032139,
+      "learning_rate": 3.0374261005275606e-06,
+      "loss": 0.5666,
+      "step": 8790
+    },
+    {
+      "epoch": 0.448681996634885,
+      "grad_norm": 6.139480470965199,
+      "learning_rate": 3.033393431005893e-06,
+      "loss": 0.5756,
+      "step": 8800
+    },
+    {
+      "epoch": 0.4491918625401519,
+      "grad_norm": 8.549581288367204,
+      "learning_rate": 3.0293593068840514e-06,
+      "loss": 0.5939,
+      "step": 8810
+    },
+    {
+      "epoch": 0.44970172844541884,
+      "grad_norm": 12.60039243357793,
+      "learning_rate": 3.0253237391633684e-06,
+      "loss": 0.5794,
+      "step": 8820
+    },
+    {
+      "epoch": 0.45021159435068575,
+      "grad_norm": 6.017339635112227,
+      "learning_rate": 3.02128673884911e-06,
+      "loss": 0.5692,
+      "step": 8830
+    },
+    {
+      "epoch": 0.45072146025595267,
+      "grad_norm": 10.959547011184434,
+      "learning_rate": 3.017248316950452e-06,
+      "loss": 0.5534,
+      "step": 8840
+    },
+    {
+      "epoch": 0.4512313261612196,
+      "grad_norm": 9.79220916497208,
+      "learning_rate": 3.0132084844804444e-06,
+      "loss": 0.5757,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4517411920664865,
+      "grad_norm": 6.681505188215735,
+      "learning_rate": 3.0091672524559855e-06,
+      "loss": 0.5547,
+      "step": 8860
+    },
+    {
+      "epoch": 0.4522510579717534,
+      "grad_norm": 20.40660151851915,
+      "learning_rate": 3.0051246318977913e-06,
+      "loss": 0.577,
+      "step": 8870
+    },
+    {
+      "epoch": 0.4527609238770203,
+      "grad_norm": 10.223009988923437,
+      "learning_rate": 3.0010806338303615e-06,
+      "loss": 0.6965,
+      "step": 8880
+    },
+    {
+      "epoch": 0.45327078978228724,
+      "grad_norm": 7.19787970543309,
+      "learning_rate": 2.997035269281955e-06,
+      "loss": 0.6269,
+      "step": 8890
+    },
+    {
+      "epoch": 0.45378065568755416,
+      "grad_norm": 9.996599464517434,
+      "learning_rate": 2.9929885492845555e-06,
+      "loss": 0.5932,
+      "step": 8900
+    },
+    {
+      "epoch": 0.4542905215928211,
+      "grad_norm": 6.990713058959951,
+      "learning_rate": 2.9889404848738434e-06,
+      "loss": 0.5632,
+      "step": 8910
+    },
+    {
+      "epoch": 0.454800387498088,
+      "grad_norm": 39.445505834540185,
+      "learning_rate": 2.9848910870891667e-06,
+      "loss": 0.5992,
+      "step": 8920
+    },
+    {
+      "epoch": 0.4553102534033549,
+      "grad_norm": 9.930318139540027,
+      "learning_rate": 2.980840366973508e-06,
+      "loss": 0.6362,
+      "step": 8930
+    },
+    {
+      "epoch": 0.4558201193086218,
+      "grad_norm": 6.1911156271130565,
+      "learning_rate": 2.9767883355734567e-06,
+      "loss": 0.5946,
+      "step": 8940
+    },
+    {
+      "epoch": 0.45632998521388873,
+      "grad_norm": 18.274373612330432,
+      "learning_rate": 2.9727350039391782e-06,
+      "loss": 0.6217,
+      "step": 8950
+    },
+    {
+      "epoch": 0.45683985111915565,
+      "grad_norm": 5.521799462230959,
+      "learning_rate": 2.968680383124384e-06,
+      "loss": 0.5643,
+      "step": 8960
+    },
+    {
+      "epoch": 0.45734971702442256,
+      "grad_norm": 4.417246498865753,
+      "learning_rate": 2.9646244841862996e-06,
+      "loss": 0.5954,
+      "step": 8970
+    },
+    {
+      "epoch": 0.4578595829296895,
+      "grad_norm": 7.1357660764881095,
+      "learning_rate": 2.9605673181856386e-06,
+      "loss": 0.5944,
+      "step": 8980
+    },
+    {
+      "epoch": 0.4583694488349564,
+      "grad_norm": 5.689317447969378,
+      "learning_rate": 2.956508896186569e-06,
+      "loss": 0.5749,
+      "step": 8990
+    },
+    {
+      "epoch": 0.4588793147402233,
+      "grad_norm": 15.300454136222234,
+      "learning_rate": 2.9524492292566824e-06,
+      "loss": 0.5954,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4593891806454902,
+      "grad_norm": 4.717762951430906,
+      "learning_rate": 2.948388328466968e-06,
+      "loss": 0.5539,
+      "step": 9010
+    },
+    {
+      "epoch": 0.45989904655075714,
+      "grad_norm": 5.159362880082,
+      "learning_rate": 2.944326204891777e-06,
+      "loss": 0.5472,
+      "step": 9020
+    },
+    {
+      "epoch": 0.46040891245602406,
+      "grad_norm": 7.114476434079598,
+      "learning_rate": 2.940262869608798e-06,
+      "loss": 0.6114,
+      "step": 9030
+    },
+    {
+      "epoch": 0.46091877836129097,
+      "grad_norm": 19.8649933302076,
+      "learning_rate": 2.9361983336990217e-06,
+      "loss": 0.5137,
+      "step": 9040
+    },
+    {
+      "epoch": 0.4614286442665579,
+      "grad_norm": 5.383648367474294,
+      "learning_rate": 2.932132608246715e-06,
+      "loss": 0.5745,
+      "step": 9050
+    },
+    {
+      "epoch": 0.4619385101718248,
+      "grad_norm": 4.573473908276397,
+      "learning_rate": 2.928065704339388e-06,
+      "loss": 0.5508,
+      "step": 9060
+    },
+    {
+      "epoch": 0.4624483760770917,
+      "grad_norm": 7.809278030311417,
+      "learning_rate": 2.9239976330677627e-06,
+      "loss": 0.6065,
+      "step": 9070
+    },
+    {
+      "epoch": 0.46295824198235863,
+      "grad_norm": 7.3170190230299745,
+      "learning_rate": 2.919928405525748e-06,
+      "loss": 0.6124,
+      "step": 9080
+    },
+    {
+      "epoch": 0.46346810788762555,
+      "grad_norm": 9.676008788916786,
+      "learning_rate": 2.9158580328104036e-06,
+      "loss": 0.5445,
+      "step": 9090
+    },
+    {
+      "epoch": 0.46397797379289246,
+      "grad_norm": 6.891950872153487,
+      "learning_rate": 2.9117865260219117e-06,
+      "loss": 0.6456,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4644878396981594,
+      "grad_norm": 5.263756940922855,
+      "learning_rate": 2.907713896263551e-06,
+      "loss": 0.5612,
+      "step": 9110
+    },
+    {
+      "epoch": 0.4649977056034263,
+      "grad_norm": 5.733645589519304,
+      "learning_rate": 2.903640154641657e-06,
+      "loss": 0.6166,
+      "step": 9120
+    },
+    {
+      "epoch": 0.4655075715086932,
+      "grad_norm": 5.622571999492135,
+      "learning_rate": 2.899565312265602e-06,
+      "loss": 0.5459,
+      "step": 9130
+    },
+    {
+      "epoch": 0.4660174374139601,
+      "grad_norm": 7.365458300779295,
+      "learning_rate": 2.895489380247758e-06,
+      "loss": 0.5388,
+      "step": 9140
+    },
+    {
+      "epoch": 0.46652730331922704,
+      "grad_norm": 8.659176612642387,
+      "learning_rate": 2.891412369703469e-06,
+      "loss": 0.5395,
+      "step": 9150
+    },
+    {
+      "epoch": 0.46703716922449395,
+      "grad_norm": 7.9795925132888685,
+      "learning_rate": 2.887334291751019e-06,
+      "loss": 0.5713,
+      "step": 9160
+    },
+    {
+      "epoch": 0.46754703512976087,
+      "grad_norm": 11.811189784886455,
+      "learning_rate": 2.883255157511605e-06,
+      "loss": 0.5767,
+      "step": 9170
+    },
+    {
+      "epoch": 0.4680569010350278,
+      "grad_norm": 7.117984988872768,
+      "learning_rate": 2.8791749781093036e-06,
+      "loss": 0.5649,
+      "step": 9180
+    },
+    {
+      "epoch": 0.4685667669402947,
+      "grad_norm": 4.290368892982683,
+      "learning_rate": 2.8750937646710416e-06,
+      "loss": 0.5301,
+      "step": 9190
+    },
+    {
+      "epoch": 0.4690766328455616,
+      "grad_norm": 17.98069195716884,
+      "learning_rate": 2.8710115283265655e-06,
+      "loss": 0.6331,
+      "step": 9200
+    },
+    {
+      "epoch": 0.46958649875082853,
+      "grad_norm": 8.243633625731393,
+      "learning_rate": 2.866928280208412e-06,
+      "loss": 0.5107,
+      "step": 9210
+    },
+    {
+      "epoch": 0.47009636465609544,
+      "grad_norm": 6.677199266110991,
+      "learning_rate": 2.8628440314518752e-06,
+      "loss": 0.5939,
+      "step": 9220
+    },
+    {
+      "epoch": 0.47060623056136236,
+      "grad_norm": 13.017880763698141,
+      "learning_rate": 2.8587587931949806e-06,
+      "loss": 0.5262,
+      "step": 9230
+    },
+    {
+      "epoch": 0.4711160964666293,
+      "grad_norm": 7.3415953210088105,
+      "learning_rate": 2.854672576578451e-06,
+      "loss": 0.5276,
+      "step": 9240
+    },
+    {
+      "epoch": 0.4716259623718962,
+      "grad_norm": 10.96227824447763,
+      "learning_rate": 2.850585392745676e-06,
+      "loss": 0.5208,
+      "step": 9250
+    },
+    {
+      "epoch": 0.4721358282771631,
+      "grad_norm": 13.046761791051441,
+      "learning_rate": 2.8464972528426847e-06,
+      "loss": 0.5542,
+      "step": 9260
+    },
+    {
+      "epoch": 0.47264569418243,
+      "grad_norm": 9.203867781595244,
+      "learning_rate": 2.842408168018112e-06,
+      "loss": 0.6764,
+      "step": 9270
+    },
+    {
+      "epoch": 0.47315556008769694,
+      "grad_norm": 8.204302843434839,
+      "learning_rate": 2.8383181494231714e-06,
+      "loss": 0.6044,
+      "step": 9280
+    },
+    {
+      "epoch": 0.47366542599296385,
+      "grad_norm": 17.257229733535443,
+      "learning_rate": 2.8342272082116214e-06,
+      "loss": 0.5799,
+      "step": 9290
+    },
+    {
+      "epoch": 0.47417529189823077,
+      "grad_norm": 48.134097471584894,
+      "learning_rate": 2.830135355539737e-06,
+      "loss": 0.5445,
+      "step": 9300
+    },
+    {
+      "epoch": 0.4746851578034977,
+      "grad_norm": 19.275589909403493,
+      "learning_rate": 2.826042602566279e-06,
+      "loss": 0.5852,
+      "step": 9310
+    },
+    {
+      "epoch": 0.4751950237087646,
+      "grad_norm": 11.641780387376368,
+      "learning_rate": 2.821948960452463e-06,
+      "loss": 0.5544,
+      "step": 9320
+    },
+    {
+      "epoch": 0.4757048896140315,
+      "grad_norm": 26.514172353901436,
+      "learning_rate": 2.817854440361929e-06,
+      "loss": 0.5607,
+      "step": 9330
+    },
+    {
+      "epoch": 0.4762147555192984,
+      "grad_norm": 8.912788994240413,
+      "learning_rate": 2.8137590534607123e-06,
+      "loss": 0.599,
+      "step": 9340
+    },
+    {
+      "epoch": 0.47672462142456534,
+      "grad_norm": 4.7043228206878105,
+      "learning_rate": 2.8096628109172125e-06,
+      "loss": 0.5597,
+      "step": 9350
+    },
+    {
+      "epoch": 0.47723448732983226,
+      "grad_norm": 10.056552729096786,
+      "learning_rate": 2.8055657239021605e-06,
+      "loss": 0.5544,
+      "step": 9360
+    },
+    {
+      "epoch": 0.4777443532350992,
+      "grad_norm": 11.358614652667871,
+      "learning_rate": 2.8014678035885913e-06,
+      "loss": 0.5654,
+      "step": 9370
+    },
+    {
+      "epoch": 0.4782542191403661,
+      "grad_norm": 11.285505664189369,
+      "learning_rate": 2.7973690611518124e-06,
+      "loss": 0.6263,
+      "step": 9380
+    },
+    {
+      "epoch": 0.478764085045633,
+      "grad_norm": 14.490746002092285,
+      "learning_rate": 2.7932695077693745e-06,
+      "loss": 0.5964,
+      "step": 9390
+    },
+    {
+      "epoch": 0.4792739509508999,
+      "grad_norm": 16.56839590898341,
+      "learning_rate": 2.7891691546210374e-06,
+      "loss": 0.5001,
+      "step": 9400
+    },
+    {
+      "epoch": 0.47978381685616683,
+      "grad_norm": 6.355268739093313,
+      "learning_rate": 2.7850680128887424e-06,
+      "loss": 0.5557,
+      "step": 9410
+    },
+    {
+      "epoch": 0.48029368276143375,
+      "grad_norm": 4.861294379529152,
+      "learning_rate": 2.780966093756584e-06,
+      "loss": 0.5566,
+      "step": 9420
+    },
+    {
+      "epoch": 0.48080354866670066,
+      "grad_norm": 12.792971199282182,
+      "learning_rate": 2.7768634084107736e-06,
+      "loss": 0.5651,
+      "step": 9430
+    },
+    {
+      "epoch": 0.4813134145719676,
+      "grad_norm": 6.746606350811939,
+      "learning_rate": 2.772759968039614e-06,
+      "loss": 0.5201,
+      "step": 9440
+    },
+    {
+      "epoch": 0.4818232804772345,
+      "grad_norm": 29.540322079081925,
+      "learning_rate": 2.7686557838334644e-06,
+      "loss": 0.6562,
+      "step": 9450
+    },
+    {
+      "epoch": 0.4823331463825014,
+      "grad_norm": 9.069525088118954,
+      "learning_rate": 2.764550866984716e-06,
+      "loss": 0.5443,
+      "step": 9460
+    },
+    {
+      "epoch": 0.4828430122877683,
+      "grad_norm": 9.551367204892276,
+      "learning_rate": 2.760445228687755e-06,
+      "loss": 0.5846,
+      "step": 9470
+    },
+    {
+      "epoch": 0.48335287819303524,
+      "grad_norm": 9.76185721527148,
+      "learning_rate": 2.7563388801389386e-06,
+      "loss": 0.5927,
+      "step": 9480
+    },
+    {
+      "epoch": 0.48386274409830216,
+      "grad_norm": 21.04589411137331,
+      "learning_rate": 2.752231832536556e-06,
+      "loss": 0.5547,
+      "step": 9490
+    },
+    {
+      "epoch": 0.48437261000356907,
+      "grad_norm": 9.066899550488897,
+      "learning_rate": 2.7481240970808074e-06,
+      "loss": 0.572,
+      "step": 9500
+    },
+    {
+      "epoch": 0.484882475908836,
+      "grad_norm": 5.241832407340177,
+      "learning_rate": 2.744015684973766e-06,
+      "loss": 0.6231,
+      "step": 9510
+    },
+    {
+      "epoch": 0.4853923418141029,
+      "grad_norm": 8.390016323543929,
+      "learning_rate": 2.739906607419351e-06,
+      "loss": 0.6537,
+      "step": 9520
+    },
+    {
+      "epoch": 0.4859022077193698,
+      "grad_norm": 4.679399687028415,
+      "learning_rate": 2.7357968756232963e-06,
+      "loss": 0.5742,
+      "step": 9530
+    },
+    {
+      "epoch": 0.48641207362463673,
+      "grad_norm": 12.2527696919582,
+      "learning_rate": 2.7316865007931208e-06,
+      "loss": 0.5758,
+      "step": 9540
+    },
+    {
+      "epoch": 0.48692193952990365,
+      "grad_norm": 11.787110553359827,
+      "learning_rate": 2.727575494138096e-06,
+      "loss": 0.5877,
+      "step": 9550
+    },
+    {
+      "epoch": 0.48743180543517056,
+      "grad_norm": 6.5067870783524,
+      "learning_rate": 2.7234638668692166e-06,
+      "loss": 0.5406,
+      "step": 9560
+    },
+    {
+      "epoch": 0.4879416713404375,
+      "grad_norm": 22.58855874717976,
+      "learning_rate": 2.7193516301991703e-06,
+      "loss": 0.5888,
+      "step": 9570
+    },
+    {
+      "epoch": 0.4884515372457044,
+      "grad_norm": 5.338386745561283,
+      "learning_rate": 2.7152387953423047e-06,
+      "loss": 0.5745,
+      "step": 9580
+    },
+    {
+      "epoch": 0.4889614031509713,
+      "grad_norm": 3.459193769024109,
+      "learning_rate": 2.711125373514602e-06,
+      "loss": 0.4953,
+      "step": 9590
+    },
+    {
+      "epoch": 0.4894712690562382,
+      "grad_norm": 6.446548850190915,
+      "learning_rate": 2.7070113759336424e-06,
+      "loss": 0.5437,
+      "step": 9600
+    },
+    {
+      "epoch": 0.48998113496150514,
+      "grad_norm": 17.79215951866651,
+      "learning_rate": 2.7028968138185783e-06,
+      "loss": 0.5767,
+      "step": 9610
+    },
+    {
+      "epoch": 0.49049100086677205,
+      "grad_norm": 2.9805861974926042,
+      "learning_rate": 2.6987816983900995e-06,
+      "loss": 0.5674,
+      "step": 9620
+    },
+    {
+      "epoch": 0.49100086677203897,
+      "grad_norm": 6.129586543666673,
+      "learning_rate": 2.6946660408704062e-06,
+      "loss": 0.5525,
+      "step": 9630
+    },
+    {
+      "epoch": 0.4915107326773059,
+      "grad_norm": 7.83623462595385,
+      "learning_rate": 2.6905498524831763e-06,
+      "loss": 0.6184,
+      "step": 9640
+    },
+    {
+      "epoch": 0.4920205985825728,
+      "grad_norm": 58.25334222306575,
+      "learning_rate": 2.6864331444535347e-06,
+      "loss": 0.558,
+      "step": 9650
+    },
+    {
+      "epoch": 0.4925304644878397,
+      "grad_norm": 6.3215473826681965,
+      "learning_rate": 2.682315928008026e-06,
+      "loss": 0.5026,
+      "step": 9660
+    },
+    {
+      "epoch": 0.49304033039310663,
+      "grad_norm": 8.52558919954231,
+      "learning_rate": 2.6781982143745776e-06,
+      "loss": 0.5422,
+      "step": 9670
+    },
+    {
+      "epoch": 0.49355019629837354,
+      "grad_norm": 20.2309368399216,
+      "learning_rate": 2.6740800147824764e-06,
+      "loss": 0.5206,
+      "step": 9680
+    },
+    {
+      "epoch": 0.49406006220364046,
+      "grad_norm": 9.45042168411588,
+      "learning_rate": 2.669961340462332e-06,
+      "loss": 0.6294,
+      "step": 9690
+    },
+    {
+      "epoch": 0.4945699281089074,
+      "grad_norm": 4.1312591938210685,
+      "learning_rate": 2.6658422026460493e-06,
+      "loss": 0.5323,
+      "step": 9700
+    },
+    {
+      "epoch": 0.4950797940141743,
+      "grad_norm": 12.26862074644364,
+      "learning_rate": 2.6617226125667977e-06,
+      "loss": 0.6065,
+      "step": 9710
+    },
+    {
+      "epoch": 0.4955896599194412,
+      "grad_norm": 28.97413747857921,
+      "learning_rate": 2.65760258145898e-06,
+      "loss": 0.5238,
+      "step": 9720
+    },
+    {
+      "epoch": 0.4960995258247081,
+      "grad_norm": 46.611911684006444,
+      "learning_rate": 2.653482120558201e-06,
+      "loss": 0.6087,
+      "step": 9730
+    },
+    {
+      "epoch": 0.49660939172997504,
+      "grad_norm": 24.265978020472545,
+      "learning_rate": 2.6493612411012377e-06,
+      "loss": 0.4993,
+      "step": 9740
+    },
+    {
+      "epoch": 0.49711925763524195,
+      "grad_norm": 61.9664610260044,
+      "learning_rate": 2.645239954326009e-06,
+      "loss": 0.583,
+      "step": 9750
+    },
+    {
+      "epoch": 0.49762912354050887,
+      "grad_norm": 11.421201901607155,
+      "learning_rate": 2.641118271471543e-06,
+      "loss": 0.5282,
+      "step": 9760
+    },
+    {
+      "epoch": 0.4981389894457758,
+      "grad_norm": 8.469805395004254,
+      "learning_rate": 2.6369962037779513e-06,
+      "loss": 0.5381,
+      "step": 9770
+    },
+    {
+      "epoch": 0.4986488553510427,
+      "grad_norm": 6.825974444005084,
+      "learning_rate": 2.632873762486392e-06,
+      "loss": 0.5972,
+      "step": 9780
+    },
+    {
+      "epoch": 0.4991587212563096,
+      "grad_norm": 4.502610701034782,
+      "learning_rate": 2.6287509588390424e-06,
+      "loss": 0.5137,
+      "step": 9790
+    },
+    {
+      "epoch": 0.4996685871615765,
+      "grad_norm": 10.021546011417461,
+      "learning_rate": 2.6246278040790696e-06,
+      "loss": 0.5858,
+      "step": 9800
+    },
+    {
+      "epoch": 0.5001784530668434,
+      "grad_norm": 10.603678720209555,
+      "learning_rate": 2.620504309450596e-06,
+      "loss": 0.5989,
+      "step": 9810
+    },
+    {
+      "epoch": 0.5006883189721103,
+      "grad_norm": 31.57403616528839,
+      "learning_rate": 2.6163804861986735e-06,
+      "loss": 0.5516,
+      "step": 9820
+    },
+    {
+      "epoch": 0.5011981848773772,
+      "grad_norm": 11.014820450677353,
+      "learning_rate": 2.6122563455692467e-06,
+      "loss": 0.5738,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5017080507826441,
+      "grad_norm": 13.980415539710211,
+      "learning_rate": 2.608131898809129e-06,
+      "loss": 0.6012,
+      "step": 9840
+    },
+    {
+      "epoch": 0.502217916687911,
+      "grad_norm": 8.212468333614668,
+      "learning_rate": 2.6040071571659676e-06,
+      "loss": 0.5443,
+      "step": 9850
+    },
+    {
+      "epoch": 0.502727782593178,
+      "grad_norm": 14.081381389861937,
+      "learning_rate": 2.5998821318882127e-06,
+      "loss": 0.5794,
+      "step": 9860
+    },
+    {
+      "epoch": 0.5032376484984449,
+      "grad_norm": 9.109868034850592,
+      "learning_rate": 2.595756834225089e-06,
+      "loss": 0.5533,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5037475144037118,
+      "grad_norm": 5.125784400942763,
+      "learning_rate": 2.5916312754265636e-06,
+      "loss": 0.5506,
+      "step": 9880
+    },
+    {
+      "epoch": 0.5042573803089787,
+      "grad_norm": 9.57153158237963,
+      "learning_rate": 2.587505466743317e-06,
+      "loss": 0.5881,
+      "step": 9890
+    },
+    {
+      "epoch": 0.5047672462142456,
+      "grad_norm": 8.591015024768001,
+      "learning_rate": 2.583379419426709e-06,
+      "loss": 0.5838,
+      "step": 9900
+    },
+    {
+      "epoch": 0.5052771121195125,
+      "grad_norm": 4.456326119029988,
+      "learning_rate": 2.5792531447287515e-06,
+      "loss": 0.5543,
+      "step": 9910
+    },
+    {
+      "epoch": 0.5057869780247795,
+      "grad_norm": 5.4644311436455535,
+      "learning_rate": 2.575126653902078e-06,
+      "loss": 0.4771,
+      "step": 9920
+    },
+    {
+      "epoch": 0.5062968439300464,
+      "grad_norm": 7.427951980380359,
+      "learning_rate": 2.570999958199908e-06,
+      "loss": 0.565,
+      "step": 9930
+    },
+    {
+      "epoch": 0.5068067098353133,
+      "grad_norm": 7.106568576861007,
+      "learning_rate": 2.566873068876022e-06,
+      "loss": 0.5849,
+      "step": 9940
+    },
+    {
+      "epoch": 0.5073165757405802,
+      "grad_norm": 9.110513479717502,
+      "learning_rate": 2.5627459971847264e-06,
+      "loss": 0.5208,
+      "step": 9950
+    },
+    {
+      "epoch": 0.5078264416458471,
+      "grad_norm": 10.449121049647742,
+      "learning_rate": 2.558618754380829e-06,
+      "loss": 0.5625,
+      "step": 9960
+    },
+    {
+      "epoch": 0.508336307551114,
+      "grad_norm": 5.528616833916277,
+      "learning_rate": 2.5544913517196006e-06,
+      "loss": 0.6019,
+      "step": 9970
+    },
+    {
+      "epoch": 0.508846173456381,
+      "grad_norm": 5.062026973579031,
+      "learning_rate": 2.5503638004567487e-06,
+      "loss": 0.5975,
+      "step": 9980
+    },
+    {
+      "epoch": 0.5093560393616479,
+      "grad_norm": 5.98683821910533,
+      "learning_rate": 2.546236111848387e-06,
+      "loss": 0.5156,
+      "step": 9990
+    },
+    {
+      "epoch": 0.5098659052669148,
+      "grad_norm": 11.616172837788515,
+      "learning_rate": 2.5421082971510024e-06,
+      "loss": 0.6184,
+      "step": 10000
+    },
+    {
+      "epoch": 0.5103757711721817,
+      "grad_norm": 21.09149624644589,
+      "learning_rate": 2.537980367621427e-06,
+      "loss": 0.5898,
+      "step": 10010
+    },
+    {
+      "epoch": 0.5108856370774486,
+      "grad_norm": 4.710577802867508,
+      "learning_rate": 2.533852334516805e-06,
+      "loss": 0.5942,
+      "step": 10020
+    },
+    {
+      "epoch": 0.5113955029827155,
+      "grad_norm": 5.51514736515005,
+      "learning_rate": 2.5297242090945638e-06,
+      "loss": 0.5705,
+      "step": 10030
+    },
+    {
+      "epoch": 0.5119053688879824,
+      "grad_norm": 5.627220384979693,
+      "learning_rate": 2.5255960026123825e-06,
+      "loss": 0.5193,
+      "step": 10040
+    },
+    {
+      "epoch": 0.5124152347932494,
+      "grad_norm": 69.00517336322274,
+      "learning_rate": 2.5214677263281613e-06,
+      "loss": 0.609,
+      "step": 10050
+    },
+    {
+      "epoch": 0.5129251006985163,
+      "grad_norm": 7.2529107552093555,
+      "learning_rate": 2.5173393914999894e-06,
+      "loss": 0.5935,
+      "step": 10060
+    },
+    {
+      "epoch": 0.5134349666037832,
+      "grad_norm": 10.759484594224745,
+      "learning_rate": 2.5132110093861174e-06,
+      "loss": 0.5138,
+      "step": 10070
+    },
+    {
+      "epoch": 0.5139448325090501,
+      "grad_norm": 3.920463168617169,
+      "learning_rate": 2.5090825912449233e-06,
+      "loss": 0.5137,
+      "step": 10080
+    },
+    {
+      "epoch": 0.514454698414317,
+      "grad_norm": 3.913288130629234,
+      "learning_rate": 2.504954148334886e-06,
+      "loss": 0.5385,
+      "step": 10090
+    },
+    {
+      "epoch": 0.5149645643195839,
+      "grad_norm": 15.125679483487403,
+      "learning_rate": 2.500825691914549e-06,
+      "loss": 0.5652,
+      "step": 10100
+    },
+    {
+      "epoch": 0.5154744302248508,
+      "grad_norm": 27.642563009181146,
+      "learning_rate": 2.4966972332424934e-06,
+      "loss": 0.6017,
+      "step": 10110
+    },
+    {
+      "epoch": 0.5159842961301178,
+      "grad_norm": 9.924339952473618,
+      "learning_rate": 2.492568783577308e-06,
+      "loss": 0.5479,
+      "step": 10120
+    },
+    {
+      "epoch": 0.5164941620353847,
+      "grad_norm": 8.10524709991328,
+      "learning_rate": 2.4884403541775553e-06,
+      "loss": 0.5492,
+      "step": 10130
+    },
+    {
+      "epoch": 0.5170040279406516,
+      "grad_norm": 197.39062095068655,
+      "learning_rate": 2.4843119563017426e-06,
+      "loss": 0.5094,
+      "step": 10140
+    },
+    {
+      "epoch": 0.5175138938459185,
+      "grad_norm": 10.76373884142598,
+      "learning_rate": 2.480183601208293e-06,
+      "loss": 0.6102,
+      "step": 10150
+    },
+    {
+      "epoch": 0.5180237597511854,
+      "grad_norm": 4.331181284635521,
+      "learning_rate": 2.4760553001555105e-06,
+      "loss": 0.5191,
+      "step": 10160
+    },
+    {
+      "epoch": 0.5185336256564523,
+      "grad_norm": 5.8217667885042985,
+      "learning_rate": 2.4719270644015533e-06,
+      "loss": 0.5496,
+      "step": 10170
+    },
+    {
+      "epoch": 0.5190434915617193,
+      "grad_norm": 19.08094034857763,
+      "learning_rate": 2.4677989052044022e-06,
+      "loss": 0.6322,
+      "step": 10180
+    },
+    {
+      "epoch": 0.5195533574669862,
+      "grad_norm": 5.2461850193376485,
+      "learning_rate": 2.4636708338218267e-06,
+      "loss": 0.5394,
+      "step": 10190
+    },
+    {
+      "epoch": 0.5200632233722531,
+      "grad_norm": 20.81830527042792,
+      "learning_rate": 2.4595428615113596e-06,
+      "loss": 0.5523,
+      "step": 10200
+    },
+    {
+      "epoch": 0.52057308927752,
+      "grad_norm": 19.96864700282037,
+      "learning_rate": 2.4554149995302605e-06,
+      "loss": 0.5756,
+      "step": 10210
+    },
+    {
+      "epoch": 0.5210829551827869,
+      "grad_norm": 7.162960743105341,
+      "learning_rate": 2.451287259135491e-06,
+      "loss": 0.5989,
+      "step": 10220
+    },
+    {
+      "epoch": 0.5215928210880538,
+      "grad_norm": 9.571220402814106,
+      "learning_rate": 2.4471596515836797e-06,
+      "loss": 0.5182,
+      "step": 10230
+    },
+    {
+      "epoch": 0.5221026869933207,
+      "grad_norm": 15.182064264052597,
+      "learning_rate": 2.4430321881310928e-06,
+      "loss": 0.6155,
+      "step": 10240
+    },
+    {
+      "epoch": 0.5226125528985877,
+      "grad_norm": 12.09544309621859,
+      "learning_rate": 2.4389048800336056e-06,
+      "loss": 0.4815,
+      "step": 10250
+    },
+    {
+      "epoch": 0.5231224188038546,
+      "grad_norm": 12.673582745109133,
+      "learning_rate": 2.434777738546666e-06,
+      "loss": 0.5715,
+      "step": 10260
+    },
+    {
+      "epoch": 0.5236322847091215,
+      "grad_norm": 4.490508455233823,
+      "learning_rate": 2.4306507749252715e-06,
+      "loss": 0.5596,
+      "step": 10270
+    },
+    {
+      "epoch": 0.5241421506143884,
+      "grad_norm": 23.00387868404976,
+      "learning_rate": 2.426524000423931e-06,
+      "loss": 0.5911,
+      "step": 10280
+    },
+    {
+      "epoch": 0.5246520165196553,
+      "grad_norm": 9.313144992336463,
+      "learning_rate": 2.4223974262966395e-06,
+      "loss": 0.5439,
+      "step": 10290
+    },
+    {
+      "epoch": 0.5251618824249222,
+      "grad_norm": 7.04600299886626,
+      "learning_rate": 2.4182710637968466e-06,
+      "loss": 0.541,
+      "step": 10300
+    },
+    {
+      "epoch": 0.5256717483301891,
+      "grad_norm": 28.8589127862858,
+      "learning_rate": 2.414144924177422e-06,
+      "loss": 0.5036,
+      "step": 10310
+    },
+    {
+      "epoch": 0.5261816142354561,
+      "grad_norm": 13.107283050966476,
+      "learning_rate": 2.4100190186906304e-06,
+      "loss": 0.5215,
+      "step": 10320
+    },
+    {
+      "epoch": 0.526691480140723,
+      "grad_norm": 7.642985447128399,
+      "learning_rate": 2.4058933585880958e-06,
+      "loss": 0.4943,
+      "step": 10330
+    },
+    {
+      "epoch": 0.5272013460459899,
+      "grad_norm": 10.878127942510975,
+      "learning_rate": 2.401767955120776e-06,
+      "loss": 0.573,
+      "step": 10340
+    },
+    {
+      "epoch": 0.5277112119512568,
+      "grad_norm": 7.41098358575074,
+      "learning_rate": 2.3976428195389236e-06,
+      "loss": 0.5247,
+      "step": 10350
+    },
+    {
+      "epoch": 0.5282210778565237,
+      "grad_norm": 5.924014795722462,
+      "learning_rate": 2.393517963092066e-06,
+      "loss": 0.5586,
+      "step": 10360
+    },
+    {
+      "epoch": 0.5287309437617906,
+      "grad_norm": 4.518669209117697,
+      "learning_rate": 2.3893933970289677e-06,
+      "loss": 0.515,
+      "step": 10370
+    },
+    {
+      "epoch": 0.5292408096670576,
+      "grad_norm": 6.223131311678184,
+      "learning_rate": 2.3852691325975996e-06,
+      "loss": 0.6371,
+      "step": 10380
+    },
+    {
+      "epoch": 0.5297506755723245,
+      "grad_norm": 8.75630273349942,
+      "learning_rate": 2.3811451810451132e-06,
+      "loss": 0.5759,
+      "step": 10390
+    },
+    {
+      "epoch": 0.5302605414775914,
+      "grad_norm": 8.600063322715169,
+      "learning_rate": 2.377021553617803e-06,
+      "loss": 0.5489,
+      "step": 10400
+    },
+    {
+      "epoch": 0.5307704073828583,
+      "grad_norm": 4.174747365549438,
+      "learning_rate": 2.372898261561082e-06,
+      "loss": 0.5865,
+      "step": 10410
+    },
+    {
+      "epoch": 0.5312802732881252,
+      "grad_norm": 4.916435449099151,
+      "learning_rate": 2.3687753161194498e-06,
+      "loss": 0.5744,
+      "step": 10420
+    },
+    {
+      "epoch": 0.5317901391933921,
+      "grad_norm": 5.548881930046863,
+      "learning_rate": 2.3646527285364565e-06,
+      "loss": 0.5161,
+      "step": 10430
+    },
+    {
+      "epoch": 0.532300005098659,
+      "grad_norm": 4.378987795858091,
+      "learning_rate": 2.3605305100546807e-06,
+      "loss": 0.5833,
+      "step": 10440
+    },
+    {
+      "epoch": 0.532809871003926,
+      "grad_norm": 18.165930076107532,
+      "learning_rate": 2.356408671915692e-06,
+      "loss": 0.5813,
+      "step": 10450
+    },
+    {
+      "epoch": 0.5333197369091929,
+      "grad_norm": 15.010025469079643,
+      "learning_rate": 2.352287225360024e-06,
+      "loss": 0.5461,
+      "step": 10460
+    },
+    {
+      "epoch": 0.5338296028144598,
+      "grad_norm": 4.72191705604572,
+      "learning_rate": 2.3481661816271413e-06,
+      "loss": 0.577,
+      "step": 10470
+    },
+    {
+      "epoch": 0.5343394687197267,
+      "grad_norm": 7.029303864129285,
+      "learning_rate": 2.3440455519554096e-06,
+      "loss": 0.5542,
+      "step": 10480
+    },
+    {
+      "epoch": 0.5348493346249936,
+      "grad_norm": 5.617285824429663,
+      "learning_rate": 2.339925347582069e-06,
+      "loss": 0.4906,
+      "step": 10490
+    },
+    {
+      "epoch": 0.5353592005302605,
+      "grad_norm": 5.705927749325273,
+      "learning_rate": 2.3358055797431945e-06,
+      "loss": 0.6121,
+      "step": 10500
+    },
+    {
+      "epoch": 0.5358690664355275,
+      "grad_norm": 6.538622967992609,
+      "learning_rate": 2.3316862596736752e-06,
+      "loss": 0.5855,
+      "step": 10510
+    },
+    {
+      "epoch": 0.5363789323407944,
+      "grad_norm": 17.530699610788833,
+      "learning_rate": 2.327567398607175e-06,
+      "loss": 0.5381,
+      "step": 10520
+    },
+    {
+      "epoch": 0.5368887982460613,
+      "grad_norm": 5.89952620967862,
+      "learning_rate": 2.3234490077761097e-06,
+      "loss": 0.5461,
+      "step": 10530
+    },
+    {
+      "epoch": 0.5373986641513282,
+      "grad_norm": 5.255170494186333,
+      "learning_rate": 2.3193310984116113e-06,
+      "loss": 0.5675,
+      "step": 10540
+    },
+    {
+      "epoch": 0.5379085300565951,
+      "grad_norm": 5.275971519411612,
+      "learning_rate": 2.3152136817434983e-06,
+      "loss": 0.572,
+      "step": 10550
+    },
+    {
+      "epoch": 0.538418395961862,
+      "grad_norm": 14.685487630845353,
+      "learning_rate": 2.311096769000247e-06,
+      "loss": 0.5969,
+      "step": 10560
+    },
+    {
+      "epoch": 0.5389282618671289,
+      "grad_norm": 4.955813987538444,
+      "learning_rate": 2.306980371408957e-06,
+      "loss": 0.5584,
+      "step": 10570
+    },
+    {
+      "epoch": 0.5394381277723959,
+      "grad_norm": 7.380894299801214,
+      "learning_rate": 2.302864500195326e-06,
+      "loss": 0.4963,
+      "step": 10580
+    },
+    {
+      "epoch": 0.5399479936776628,
+      "grad_norm": 4.890069171450678,
+      "learning_rate": 2.2987491665836137e-06,
+      "loss": 0.525,
+      "step": 10590
+    },
+    {
+      "epoch": 0.5404578595829297,
+      "grad_norm": 11.506877076778487,
+      "learning_rate": 2.2946343817966147e-06,
+      "loss": 0.5496,
+      "step": 10600
+    },
+    {
+      "epoch": 0.5409677254881966,
+      "grad_norm": 7.716243227450342,
+      "learning_rate": 2.2905201570556294e-06,
+      "loss": 0.5583,
+      "step": 10610
+    },
+    {
+      "epoch": 0.5414775913934635,
+      "grad_norm": 5.818018263936738,
+      "learning_rate": 2.2864065035804253e-06,
+      "loss": 0.5293,
+      "step": 10620
+    },
+    {
+      "epoch": 0.5419874572987304,
+      "grad_norm": 9.928565196052132,
+      "learning_rate": 2.2822934325892178e-06,
+      "loss": 0.5897,
+      "step": 10630
+    },
+    {
+      "epoch": 0.5424973232039974,
+      "grad_norm": 11.33679136308585,
+      "learning_rate": 2.2781809552986296e-06,
+      "loss": 0.5679,
+      "step": 10640
+    },
+    {
+      "epoch": 0.5430071891092643,
+      "grad_norm": 7.70962478429198,
+      "learning_rate": 2.2740690829236672e-06,
+      "loss": 0.5523,
+      "step": 10650
+    },
+    {
+      "epoch": 0.5435170550145312,
+      "grad_norm": 7.917410798743483,
+      "learning_rate": 2.269957826677685e-06,
+      "loss": 0.5672,
+      "step": 10660
+    },
+    {
+      "epoch": 0.5440269209197981,
+      "grad_norm": 6.7827283040354205,
+      "learning_rate": 2.2658471977723593e-06,
+      "loss": 0.5437,
+      "step": 10670
+    },
+    {
+      "epoch": 0.544536786825065,
+      "grad_norm": 4.342938474458968,
+      "learning_rate": 2.2617372074176565e-06,
+      "loss": 0.5387,
+      "step": 10680
+    },
+    {
+      "epoch": 0.5450466527303319,
+      "grad_norm": 9.43212832651628,
+      "learning_rate": 2.2576278668217967e-06,
+      "loss": 0.5197,
+      "step": 10690
+    },
+    {
+      "epoch": 0.5455565186355988,
+      "grad_norm": 8.004745774649368,
+      "learning_rate": 2.2535191871912337e-06,
+      "loss": 0.5429,
+      "step": 10700
+    },
+    {
+      "epoch": 0.5460663845408658,
+      "grad_norm": 6.872851460413676,
+      "learning_rate": 2.2494111797306146e-06,
+      "loss": 0.5419,
+      "step": 10710
+    },
+    {
+      "epoch": 0.5465762504461327,
+      "grad_norm": 7.289325384183292,
+      "learning_rate": 2.2453038556427557e-06,
+      "loss": 0.6115,
+      "step": 10720
+    },
+    {
+      "epoch": 0.5470861163513996,
+      "grad_norm": 13.306562120222383,
+      "learning_rate": 2.2411972261286107e-06,
+      "loss": 0.5227,
+      "step": 10730
+    },
+    {
+      "epoch": 0.5475959822566665,
+      "grad_norm": 8.891170592948066,
+      "learning_rate": 2.2370913023872357e-06,
+      "loss": 0.6406,
+      "step": 10740
+    },
+    {
+      "epoch": 0.5481058481619334,
+      "grad_norm": 8.565050748681404,
+      "learning_rate": 2.2329860956157655e-06,
+      "loss": 0.563,
+      "step": 10750
+    },
+    {
+      "epoch": 0.5486157140672003,
+      "grad_norm": 10.560003193382766,
+      "learning_rate": 2.2288816170093767e-06,
+      "loss": 0.5297,
+      "step": 10760
+    },
+    {
+      "epoch": 0.5491255799724672,
+      "grad_norm": 21.790652278804174,
+      "learning_rate": 2.224777877761264e-06,
+      "loss": 0.5178,
+      "step": 10770
+    },
+    {
+      "epoch": 0.5496354458777342,
+      "grad_norm": 4.762684973385444,
+      "learning_rate": 2.2206748890626004e-06,
+      "loss": 0.6076,
+      "step": 10780
+    },
+    {
+      "epoch": 0.5501453117830011,
+      "grad_norm": 11.67968286241827,
+      "learning_rate": 2.216572662102518e-06,
+      "loss": 0.636,
+      "step": 10790
+    },
+    {
+      "epoch": 0.550655177688268,
+      "grad_norm": 32.813770920064506,
+      "learning_rate": 2.212471208068068e-06,
+      "loss": 0.5275,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5511650435935349,
+      "grad_norm": 15.057043570306027,
+      "learning_rate": 2.2083705381441933e-06,
+      "loss": 0.5409,
+      "step": 10810
+    },
+    {
+      "epoch": 0.5516749094988018,
+      "grad_norm": 5.890492850664067,
+      "learning_rate": 2.204270663513702e-06,
+      "loss": 0.5254,
+      "step": 10820
+    },
+    {
+      "epoch": 0.5521847754040687,
+      "grad_norm": 7.960866990199008,
+      "learning_rate": 2.200171595357229e-06,
+      "loss": 0.529,
+      "step": 10830
+    },
+    {
+      "epoch": 0.5526946413093357,
+      "grad_norm": 12.87627099794233,
+      "learning_rate": 2.196073344853213e-06,
+      "loss": 0.493,
+      "step": 10840
+    },
+    {
+      "epoch": 0.5532045072146026,
+      "grad_norm": 13.807869879260597,
+      "learning_rate": 2.1919759231778616e-06,
+      "loss": 0.5958,
+      "step": 10850
+    },
+    {
+      "epoch": 0.5537143731198695,
+      "grad_norm": 8.732910844238024,
+      "learning_rate": 2.187879341505123e-06,
+      "loss": 0.5315,
+      "step": 10860
+    },
+    {
+      "epoch": 0.5542242390251364,
+      "grad_norm": 10.627749896245922,
+      "learning_rate": 2.1837836110066544e-06,
+      "loss": 0.6201,
+      "step": 10870
+    },
+    {
+      "epoch": 0.5547341049304033,
+      "grad_norm": 12.410053602064153,
+      "learning_rate": 2.17968874285179e-06,
+      "loss": 0.661,
+      "step": 10880
+    },
+    {
+      "epoch": 0.5552439708356702,
+      "grad_norm": 9.702490596647241,
+      "learning_rate": 2.175594748207516e-06,
+      "loss": 0.5263,
+      "step": 10890
+    },
+    {
+      "epoch": 0.5557538367409371,
+      "grad_norm": 10.737235352322436,
+      "learning_rate": 2.1715016382384314e-06,
+      "loss": 0.5643,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5562637026462041,
+      "grad_norm": 5.205601904965876,
+      "learning_rate": 2.1674094241067275e-06,
+      "loss": 0.4741,
+      "step": 10910
+    },
+    {
+      "epoch": 0.556773568551471,
+      "grad_norm": 9.54609904611359,
+      "learning_rate": 2.1633181169721518e-06,
+      "loss": 0.5626,
+      "step": 10920
+    },
+    {
+      "epoch": 0.5572834344567379,
+      "grad_norm": 17.84606744633947,
+      "learning_rate": 2.159227727991974e-06,
+      "loss": 0.5571,
+      "step": 10930
+    },
+    {
+      "epoch": 0.5577933003620048,
+      "grad_norm": 21.19914314000095,
+      "learning_rate": 2.155138268320966e-06,
+      "loss": 0.5913,
+      "step": 10940
+    },
+    {
+      "epoch": 0.5583031662672717,
+      "grad_norm": 11.748809523593113,
+      "learning_rate": 2.151049749111361e-06,
+      "loss": 0.6247,
+      "step": 10950
+    },
+    {
+      "epoch": 0.5588130321725386,
+      "grad_norm": 9.896026330514129,
+      "learning_rate": 2.146962181512829e-06,
+      "loss": 0.5989,
+      "step": 10960
+    },
+    {
+      "epoch": 0.5593228980778056,
+      "grad_norm": 7.035140414909101,
+      "learning_rate": 2.142875576672446e-06,
+      "loss": 0.5488,
+      "step": 10970
+    },
+    {
+      "epoch": 0.5598327639830725,
+      "grad_norm": 6.072140050707628,
+      "learning_rate": 2.13878994573466e-06,
+      "loss": 0.5407,
+      "step": 10980
+    },
+    {
+      "epoch": 0.5603426298883394,
+      "grad_norm": 7.288317100033059,
+      "learning_rate": 2.1347052998412667e-06,
+      "loss": 0.5561,
+      "step": 10990
+    },
+    {
+      "epoch": 0.5608524957936063,
+      "grad_norm": 5.819210689950646,
+      "learning_rate": 2.1306216501313705e-06,
+      "loss": 0.5428,
+      "step": 11000
+    },
+    {
+      "epoch": 0.5613623616988732,
+      "grad_norm": 12.776136561779468,
+      "learning_rate": 2.1265390077413643e-06,
+      "loss": 0.5003,
+      "step": 11010
+    },
+    {
+      "epoch": 0.5618722276041401,
+      "grad_norm": 18.983352626504544,
+      "learning_rate": 2.1224573838048894e-06,
+      "loss": 0.5445,
+      "step": 11020
+    },
+    {
+      "epoch": 0.562382093509407,
+      "grad_norm": 15.491629082640607,
+      "learning_rate": 2.1183767894528135e-06,
+      "loss": 0.5391,
+      "step": 11030
+    },
+    {
+      "epoch": 0.562891959414674,
+      "grad_norm": 9.763365375568759,
+      "learning_rate": 2.114297235813196e-06,
+      "loss": 0.5338,
+      "step": 11040
+    },
+    {
+      "epoch": 0.5634018253199409,
+      "grad_norm": 6.44544801800945,
+      "learning_rate": 2.110218734011255e-06,
+      "loss": 0.5031,
+      "step": 11050
+    },
+    {
+      "epoch": 0.5639116912252078,
+      "grad_norm": 4.539572048783182,
+      "learning_rate": 2.106141295169344e-06,
+      "loss": 0.539,
+      "step": 11060
+    },
+    {
+      "epoch": 0.5644215571304747,
+      "grad_norm": 7.702392114659815,
+      "learning_rate": 2.1020649304069144e-06,
+      "loss": 0.5167,
+      "step": 11070
+    },
+    {
+      "epoch": 0.5649314230357416,
+      "grad_norm": 6.263165760093371,
+      "learning_rate": 2.0979896508404917e-06,
+      "loss": 0.524,
+      "step": 11080
+    },
+    {
+      "epoch": 0.5654412889410085,
+      "grad_norm": 8.742193198862491,
+      "learning_rate": 2.0939154675836407e-06,
+      "loss": 0.5162,
+      "step": 11090
+    },
+    {
+      "epoch": 0.5659511548462755,
+      "grad_norm": 7.658537397707099,
+      "learning_rate": 2.0898423917469344e-06,
+      "loss": 0.5552,
+      "step": 11100
+    },
+    {
+      "epoch": 0.5664610207515424,
+      "grad_norm": 7.432799918637242,
+      "learning_rate": 2.085770434437931e-06,
+      "loss": 0.4958,
+      "step": 11110
+    },
+    {
+      "epoch": 0.5669708866568093,
+      "grad_norm": 8.525440793917197,
+      "learning_rate": 2.0816996067611315e-06,
+      "loss": 0.5474,
+      "step": 11120
+    },
+    {
+      "epoch": 0.5674807525620762,
+      "grad_norm": 12.292033213542174,
+      "learning_rate": 2.0776299198179624e-06,
+      "loss": 0.6679,
+      "step": 11130
+    },
+    {
+      "epoch": 0.5679906184673431,
+      "grad_norm": 11.936598755773637,
+      "learning_rate": 2.0735613847067355e-06,
+      "loss": 0.5626,
+      "step": 11140
+    },
+    {
+      "epoch": 0.56850048437261,
+      "grad_norm": 15.717441916004525,
+      "learning_rate": 2.0694940125226224e-06,
+      "loss": 0.5813,
+      "step": 11150
+    },
+    {
+      "epoch": 0.5690103502778769,
+      "grad_norm": 4.026163345202539,
+      "learning_rate": 2.0654278143576263e-06,
+      "loss": 0.5576,
+      "step": 11160
+    },
+    {
+      "epoch": 0.5695202161831439,
+      "grad_norm": 6.045823319943218,
+      "learning_rate": 2.0613628013005437e-06,
+      "loss": 0.595,
+      "step": 11170
+    },
+    {
+      "epoch": 0.5700300820884108,
+      "grad_norm": 22.41401539076487,
+      "learning_rate": 2.0572989844369427e-06,
+      "loss": 0.5276,
+      "step": 11180
+    },
+    {
+      "epoch": 0.5705399479936777,
+      "grad_norm": 5.04830021515491,
+      "learning_rate": 2.053236374849128e-06,
+      "loss": 0.5411,
+      "step": 11190
+    },
+    {
+      "epoch": 0.5710498138989446,
+      "grad_norm": 8.227143620060716,
+      "learning_rate": 2.049174983616113e-06,
+      "loss": 0.5622,
+      "step": 11200
+    },
+    {
+      "epoch": 0.5715596798042115,
+      "grad_norm": 4.747071324681422,
+      "learning_rate": 2.045114821813588e-06,
+      "loss": 0.5133,
+      "step": 11210
+    },
+    {
+      "epoch": 0.5720695457094784,
+      "grad_norm": 4.548869894399742,
+      "learning_rate": 2.0410559005138893e-06,
+      "loss": 0.5411,
+      "step": 11220
+    },
+    {
+      "epoch": 0.5725794116147453,
+      "grad_norm": 4.694113174753521,
+      "learning_rate": 2.0369982307859728e-06,
+      "loss": 0.5564,
+      "step": 11230
+    },
+    {
+      "epoch": 0.5730892775200123,
+      "grad_norm": 7.208504892989224,
+      "learning_rate": 2.032941823695378e-06,
+      "loss": 0.4987,
+      "step": 11240
+    },
+    {
+      "epoch": 0.5735991434252792,
+      "grad_norm": 6.550131307780764,
+      "learning_rate": 2.0288866903042055e-06,
+      "loss": 0.5533,
+      "step": 11250
+    },
+    {
+      "epoch": 0.5741090093305461,
+      "grad_norm": 14.482125263295131,
+      "learning_rate": 2.024832841671077e-06,
+      "loss": 0.5762,
+      "step": 11260
+    },
+    {
+      "epoch": 0.574618875235813,
+      "grad_norm": 5.509057904355494,
+      "learning_rate": 2.0207802888511155e-06,
+      "loss": 0.564,
+      "step": 11270
+    },
+    {
+      "epoch": 0.5751287411410799,
+      "grad_norm": 11.023848829252348,
+      "learning_rate": 2.0167290428959082e-06,
+      "loss": 0.5716,
+      "step": 11280
+    },
+    {
+      "epoch": 0.5756386070463468,
+      "grad_norm": 12.049873053442631,
+      "learning_rate": 2.0126791148534777e-06,
+      "loss": 0.5457,
+      "step": 11290
+    },
+    {
+      "epoch": 0.5761484729516138,
+      "grad_norm": 4.3956366771791515,
+      "learning_rate": 2.0086305157682546e-06,
+      "loss": 0.4937,
+      "step": 11300
+    },
+    {
+      "epoch": 0.5766583388568807,
+      "grad_norm": 5.635414557925949,
+      "learning_rate": 2.004583256681042e-06,
+      "loss": 0.5618,
+      "step": 11310
+    },
+    {
+      "epoch": 0.5771682047621476,
+      "grad_norm": 4.7330346651271755,
+      "learning_rate": 2.0005373486289932e-06,
+      "loss": 0.5553,
+      "step": 11320
+    },
+    {
+      "epoch": 0.5776780706674145,
+      "grad_norm": 5.922877822891385,
+      "learning_rate": 1.9964928026455715e-06,
+      "loss": 0.4771,
+      "step": 11330
+    },
+    {
+      "epoch": 0.5781879365726814,
+      "grad_norm": 5.231113774110676,
+      "learning_rate": 1.9924496297605315e-06,
+      "loss": 0.5578,
+      "step": 11340
+    },
+    {
+      "epoch": 0.5786978024779483,
+      "grad_norm": 5.08380380468843,
+      "learning_rate": 1.988407840999881e-06,
+      "loss": 0.4714,
+      "step": 11350
+    },
+    {
+      "epoch": 0.5792076683832152,
+      "grad_norm": 6.004819919545154,
+      "learning_rate": 1.984367447385851e-06,
+      "loss": 0.5764,
+      "step": 11360
+    },
+    {
+      "epoch": 0.5797175342884822,
+      "grad_norm": 18.64442586475368,
+      "learning_rate": 1.9803284599368704e-06,
+      "loss": 0.5513,
+      "step": 11370
+    },
+    {
+      "epoch": 0.5802274001937491,
+      "grad_norm": 11.25732697341979,
+      "learning_rate": 1.976290889667533e-06,
+      "loss": 0.5753,
+      "step": 11380
+    },
+    {
+      "epoch": 0.580737266099016,
+      "grad_norm": 13.524467497890853,
+      "learning_rate": 1.9722547475885685e-06,
+      "loss": 0.5732,
+      "step": 11390
+    },
+    {
+      "epoch": 0.5812471320042829,
+      "grad_norm": 11.958430536712548,
+      "learning_rate": 1.9682200447068095e-06,
+      "loss": 0.5608,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5817569979095498,
+      "grad_norm": 4.929859531783645,
+      "learning_rate": 1.9641867920251655e-06,
+      "loss": 0.5411,
+      "step": 11410
+    },
+    {
+      "epoch": 0.5822668638148167,
+      "grad_norm": 6.115864848385566,
+      "learning_rate": 1.9601550005425925e-06,
+      "loss": 0.5141,
+      "step": 11420
+    },
+    {
+      "epoch": 0.5827767297200837,
+      "grad_norm": 3.8065996292234425,
+      "learning_rate": 1.9561246812540572e-06,
+      "loss": 0.5874,
+      "step": 11430
+    },
+    {
+      "epoch": 0.5832865956253506,
+      "grad_norm": 8.306882889332991,
+      "learning_rate": 1.952095845150518e-06,
+      "loss": 0.5053,
+      "step": 11440
+    },
+    {
+      "epoch": 0.5837964615306175,
+      "grad_norm": 10.01015622593652,
+      "learning_rate": 1.9480685032188816e-06,
+      "loss": 0.5359,
+      "step": 11450
+    },
+    {
+      "epoch": 0.5843063274358844,
+      "grad_norm": 10.699493523122232,
+      "learning_rate": 1.9440426664419855e-06,
+      "loss": 0.5721,
+      "step": 11460
+    },
+    {
+      "epoch": 0.5848161933411513,
+      "grad_norm": 3.2992193035520776,
+      "learning_rate": 1.940018345798561e-06,
+      "loss": 0.4852,
+      "step": 11470
+    },
+    {
+      "epoch": 0.5853260592464182,
+      "grad_norm": 42.64445686580835,
+      "learning_rate": 1.935995552263202e-06,
+      "loss": 0.5427,
+      "step": 11480
+    },
+    {
+      "epoch": 0.5858359251516851,
+      "grad_norm": 4.99320155932582,
+      "learning_rate": 1.931974296806342e-06,
+      "loss": 0.5109,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5863457910569521,
+      "grad_norm": 11.084567601573044,
+      "learning_rate": 1.9279545903942174e-06,
+      "loss": 0.5371,
+      "step": 11500
+    },
+    {
+      "epoch": 0.586855656962219,
+      "grad_norm": 10.31727590216634,
+      "learning_rate": 1.923936443988841e-06,
+      "loss": 0.5611,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5873655228674859,
+      "grad_norm": 12.632248744614495,
+      "learning_rate": 1.9199198685479732e-06,
+      "loss": 0.5298,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5878753887727528,
+      "grad_norm": 7.57910307013232,
+      "learning_rate": 1.9159048750250855e-06,
+      "loss": 0.5541,
+      "step": 11530
+    },
+    {
+      "epoch": 0.5883852546780197,
+      "grad_norm": 7.842350564520856,
+      "learning_rate": 1.9118914743693407e-06,
+      "loss": 0.5822,
+      "step": 11540
+    },
+    {
+      "epoch": 0.5888951205832866,
+      "grad_norm": 14.98917864774212,
+      "learning_rate": 1.907879677525554e-06,
+      "loss": 0.5409,
+      "step": 11550
+    },
+    {
+      "epoch": 0.5894049864885536,
+      "grad_norm": 14.362958611737447,
+      "learning_rate": 1.9038694954341697e-06,
+      "loss": 0.4998,
+      "step": 11560
+    },
+    {
+      "epoch": 0.5899148523938205,
+      "grad_norm": 5.82283027332385,
+      "learning_rate": 1.8998609390312251e-06,
+      "loss": 0.4923,
+      "step": 11570
+    },
+    {
+      "epoch": 0.5904247182990874,
+      "grad_norm": 10.272349364887027,
+      "learning_rate": 1.8958540192483277e-06,
+      "loss": 0.5519,
+      "step": 11580
+    },
+    {
+      "epoch": 0.5909345842043543,
+      "grad_norm": 12.109395292903596,
+      "learning_rate": 1.8918487470126207e-06,
+      "loss": 0.5581,
+      "step": 11590
+    },
+    {
+      "epoch": 0.5914444501096212,
+      "grad_norm": 7.524591927450444,
+      "learning_rate": 1.8878451332467529e-06,
+      "loss": 0.5133,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5919543160148881,
+      "grad_norm": 16.959598592402315,
+      "learning_rate": 1.8838431888688528e-06,
+      "loss": 0.5905,
+      "step": 11610
+    },
+    {
+      "epoch": 0.592464181920155,
+      "grad_norm": 6.178426939909682,
+      "learning_rate": 1.879842924792493e-06,
+      "loss": 0.5712,
+      "step": 11620
+    },
+    {
+      "epoch": 0.592974047825422,
+      "grad_norm": 6.552338716939495,
+      "learning_rate": 1.8758443519266667e-06,
+      "loss": 0.5146,
+      "step": 11630
+    },
+    {
+      "epoch": 0.5934839137306889,
+      "grad_norm": 5.554779498177808,
+      "learning_rate": 1.8718474811757553e-06,
+      "loss": 0.5211,
+      "step": 11640
+    },
+    {
+      "epoch": 0.5939937796359558,
+      "grad_norm": 14.887120778210514,
+      "learning_rate": 1.8678523234394956e-06,
+      "loss": 0.5599,
+      "step": 11650
+    },
+    {
+      "epoch": 0.5945036455412227,
+      "grad_norm": 3.108872975693166,
+      "learning_rate": 1.863858889612956e-06,
+      "loss": 0.4967,
+      "step": 11660
+    },
+    {
+      "epoch": 0.5950135114464896,
+      "grad_norm": 4.95066312376142,
+      "learning_rate": 1.8598671905865002e-06,
+      "loss": 0.5689,
+      "step": 11670
+    },
+    {
+      "epoch": 0.5955233773517565,
+      "grad_norm": 3.8170420467015793,
+      "learning_rate": 1.8558772372457647e-06,
+      "loss": 0.5409,
+      "step": 11680
+    },
+    {
+      "epoch": 0.5960332432570234,
+      "grad_norm": 9.995945349423792,
+      "learning_rate": 1.8518890404716227e-06,
+      "loss": 0.5085,
+      "step": 11690
+    },
+    {
+      "epoch": 0.5965431091622904,
+      "grad_norm": 7.49096178498184,
+      "learning_rate": 1.8479026111401594e-06,
+      "loss": 0.5287,
+      "step": 11700
+    },
+    {
+      "epoch": 0.5970529750675573,
+      "grad_norm": 5.778438270552736,
+      "learning_rate": 1.8439179601226376e-06,
+      "loss": 0.5124,
+      "step": 11710
+    },
+    {
+      "epoch": 0.5975628409728242,
+      "grad_norm": 4.567825321093821,
+      "learning_rate": 1.8399350982854717e-06,
+      "loss": 0.4978,
+      "step": 11720
+    },
+    {
+      "epoch": 0.5980727068780911,
+      "grad_norm": 11.475875662289322,
+      "learning_rate": 1.835954036490198e-06,
+      "loss": 0.5369,
+      "step": 11730
+    },
+    {
+      "epoch": 0.598582572783358,
+      "grad_norm": 5.961118047678783,
+      "learning_rate": 1.8319747855934416e-06,
+      "loss": 0.5744,
+      "step": 11740
+    },
+    {
+      "epoch": 0.5990924386886249,
+      "grad_norm": 6.930421840026854,
+      "learning_rate": 1.8279973564468906e-06,
+      "loss": 0.5138,
+      "step": 11750
+    },
+    {
+      "epoch": 0.5996023045938919,
+      "grad_norm": 6.394493063704984,
+      "learning_rate": 1.8240217598972665e-06,
+      "loss": 0.6055,
+      "step": 11760
+    },
+    {
+      "epoch": 0.6001121704991588,
+      "grad_norm": 6.464271673435689,
+      "learning_rate": 1.8200480067862888e-06,
+      "loss": 0.5635,
+      "step": 11770
+    },
+    {
+      "epoch": 0.6006220364044257,
+      "grad_norm": 6.33418337132712,
+      "learning_rate": 1.8160761079506553e-06,
+      "loss": 0.546,
+      "step": 11780
+    },
+    {
+      "epoch": 0.6011319023096926,
+      "grad_norm": 10.636087327366742,
+      "learning_rate": 1.812106074222002e-06,
+      "loss": 0.5233,
+      "step": 11790
+    },
+    {
+      "epoch": 0.6016417682149595,
+      "grad_norm": 8.78934513556952,
+      "learning_rate": 1.8081379164268826e-06,
+      "loss": 0.574,
+      "step": 11800
+    },
+    {
+      "epoch": 0.6021516341202264,
+      "grad_norm": 32.344538634188154,
+      "learning_rate": 1.804171645386732e-06,
+      "loss": 0.5374,
+      "step": 11810
+    },
+    {
+      "epoch": 0.6026615000254933,
+      "grad_norm": 6.569258991721041,
+      "learning_rate": 1.800207271917842e-06,
+      "loss": 0.549,
+      "step": 11820
+    },
+    {
+      "epoch": 0.6031713659307603,
+      "grad_norm": 7.410474971488895,
+      "learning_rate": 1.7962448068313298e-06,
+      "loss": 0.5449,
+      "step": 11830
+    },
+    {
+      "epoch": 0.6036812318360272,
+      "grad_norm": 13.567886275688778,
+      "learning_rate": 1.7922842609331053e-06,
+      "loss": 0.5348,
+      "step": 11840
+    },
+    {
+      "epoch": 0.6041910977412941,
+      "grad_norm": 22.176157368753856,
+      "learning_rate": 1.788325645023848e-06,
+      "loss": 0.5003,
+      "step": 11850
+    },
+    {
+      "epoch": 0.604700963646561,
+      "grad_norm": 5.704122016801913,
+      "learning_rate": 1.7843689698989715e-06,
+      "loss": 0.4972,
+      "step": 11860
+    },
+    {
+      "epoch": 0.6052108295518279,
+      "grad_norm": 10.267907792805214,
+      "learning_rate": 1.7804142463486e-06,
+      "loss": 0.51,
+      "step": 11870
+    },
+    {
+      "epoch": 0.6057206954570948,
+      "grad_norm": 6.572061790414582,
+      "learning_rate": 1.776461485157531e-06,
+      "loss": 0.5938,
+      "step": 11880
+    },
+    {
+      "epoch": 0.6062305613623618,
+      "grad_norm": 6.2343504746004,
+      "learning_rate": 1.7725106971052147e-06,
+      "loss": 0.5394,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6067404272676287,
+      "grad_norm": 12.18768143965922,
+      "learning_rate": 1.7685618929657193e-06,
+      "loss": 0.5751,
+      "step": 11900
+    },
+    {
+      "epoch": 0.6072502931728956,
+      "grad_norm": 7.201216852229459,
+      "learning_rate": 1.7646150835077014e-06,
+      "loss": 0.5611,
+      "step": 11910
+    },
+    {
+      "epoch": 0.6077601590781625,
+      "grad_norm": 6.707647871833833,
+      "learning_rate": 1.7606702794943803e-06,
+      "loss": 0.539,
+      "step": 11920
+    },
+    {
+      "epoch": 0.6082700249834293,
+      "grad_norm": 16.03213196409904,
+      "learning_rate": 1.756727491683503e-06,
+      "loss": 0.5695,
+      "step": 11930
+    },
+    {
+      "epoch": 0.6087798908886962,
+      "grad_norm": 7.653859604284126,
+      "learning_rate": 1.7527867308273211e-06,
+      "loss": 0.5502,
+      "step": 11940
+    },
+    {
+      "epoch": 0.6092897567939631,
+      "grad_norm": 18.450018356779395,
+      "learning_rate": 1.7488480076725584e-06,
+      "loss": 0.5902,
+      "step": 11950
+    },
+    {
+      "epoch": 0.60979962269923,
+      "grad_norm": 8.81328827050954,
+      "learning_rate": 1.7449113329603787e-06,
+      "loss": 0.5659,
+      "step": 11960
+    },
+    {
+      "epoch": 0.610309488604497,
+      "grad_norm": 16.666299004092103,
+      "learning_rate": 1.7409767174263643e-06,
+      "loss": 0.5547,
+      "step": 11970
+    },
+    {
+      "epoch": 0.6108193545097639,
+      "grad_norm": 8.923550983437284,
+      "learning_rate": 1.7370441718004771e-06,
+      "loss": 0.6135,
+      "step": 11980
+    },
+    {
+      "epoch": 0.6113292204150308,
+      "grad_norm": 6.030797934178173,
+      "learning_rate": 1.733113706807038e-06,
+      "loss": 0.563,
+      "step": 11990
+    },
+    {
+      "epoch": 0.6118390863202977,
+      "grad_norm": 6.60626139740419,
+      "learning_rate": 1.7291853331646917e-06,
+      "loss": 0.5256,
+      "step": 12000
+    },
+    {
+      "epoch": 0.6123489522255646,
+      "grad_norm": 7.553040351819789,
+      "learning_rate": 1.7252590615863809e-06,
+      "loss": 0.5619,
+      "step": 12010
+    },
+    {
+      "epoch": 0.6128588181308315,
+      "grad_norm": 12.276519778238537,
+      "learning_rate": 1.7213349027793153e-06,
+      "loss": 0.5427,
+      "step": 12020
+    },
+    {
+      "epoch": 0.6133686840360985,
+      "grad_norm": 6.7982241497229134,
+      "learning_rate": 1.7174128674449422e-06,
+      "loss": 0.4984,
+      "step": 12030
+    },
+    {
+      "epoch": 0.6138785499413654,
+      "grad_norm": 34.987861391211595,
+      "learning_rate": 1.7134929662789204e-06,
+      "loss": 0.535,
+      "step": 12040
+    },
+    {
+      "epoch": 0.6143884158466323,
+      "grad_norm": 6.391665999406623,
+      "learning_rate": 1.709575209971085e-06,
+      "loss": 0.4856,
+      "step": 12050
+    },
+    {
+      "epoch": 0.6148982817518992,
+      "grad_norm": 10.174572831746715,
+      "learning_rate": 1.7056596092054245e-06,
+      "loss": 0.5267,
+      "step": 12060
+    },
+    {
+      "epoch": 0.6154081476571661,
+      "grad_norm": 3.8316603741601885,
+      "learning_rate": 1.7017461746600506e-06,
+      "loss": 0.56,
+      "step": 12070
+    },
+    {
+      "epoch": 0.615918013562433,
+      "grad_norm": 4.2956097081697315,
+      "learning_rate": 1.697834917007163e-06,
+      "loss": 0.5564,
+      "step": 12080
+    },
+    {
+      "epoch": 0.6164278794677,
+      "grad_norm": 12.793184931761772,
+      "learning_rate": 1.6939258469130288e-06,
+      "loss": 0.5483,
+      "step": 12090
+    },
+    {
+      "epoch": 0.6169377453729669,
+      "grad_norm": 11.38804280431213,
+      "learning_rate": 1.6900189750379469e-06,
+      "loss": 0.5347,
+      "step": 12100
+    },
+    {
+      "epoch": 0.6174476112782338,
+      "grad_norm": 8.594475973751882,
+      "learning_rate": 1.6861143120362239e-06,
+      "loss": 0.5411,
+      "step": 12110
+    },
+    {
+      "epoch": 0.6179574771835007,
+      "grad_norm": 7.3766297715792115,
+      "learning_rate": 1.6822118685561403e-06,
+      "loss": 0.5243,
+      "step": 12120
+    },
+    {
+      "epoch": 0.6184673430887676,
+      "grad_norm": 13.124098897523416,
+      "learning_rate": 1.6783116552399258e-06,
+      "loss": 0.5176,
+      "step": 12130
+    },
+    {
+      "epoch": 0.6189772089940345,
+      "grad_norm": 5.497804698316182,
+      "learning_rate": 1.6744136827237283e-06,
+      "loss": 0.5455,
+      "step": 12140
+    },
+    {
+      "epoch": 0.6194870748993014,
+      "grad_norm": 7.212115337101796,
+      "learning_rate": 1.670517961637582e-06,
+      "loss": 0.5101,
+      "step": 12150
+    },
+    {
+      "epoch": 0.6199969408045684,
+      "grad_norm": 8.604412741667838,
+      "learning_rate": 1.666624502605385e-06,
+      "loss": 0.5701,
+      "step": 12160
+    },
+    {
+      "epoch": 0.6205068067098353,
+      "grad_norm": 19.511180309160327,
+      "learning_rate": 1.6627333162448638e-06,
+      "loss": 0.5202,
+      "step": 12170
+    },
+    {
+      "epoch": 0.6210166726151022,
+      "grad_norm": 8.939814436037937,
+      "learning_rate": 1.6588444131675486e-06,
+      "loss": 0.5488,
+      "step": 12180
+    },
+    {
+      "epoch": 0.6215265385203691,
+      "grad_norm": 8.046046108782505,
+      "learning_rate": 1.6549578039787436e-06,
+      "loss": 0.5954,
+      "step": 12190
+    },
+    {
+      "epoch": 0.622036404425636,
+      "grad_norm": 15.08482992814336,
+      "learning_rate": 1.6510734992774953e-06,
+      "loss": 0.5213,
+      "step": 12200
+    },
+    {
+      "epoch": 0.6225462703309029,
+      "grad_norm": 8.290379527495224,
+      "learning_rate": 1.647191509656567e-06,
+      "loss": 0.5287,
+      "step": 12210
+    },
+    {
+      "epoch": 0.6230561362361698,
+      "grad_norm": 5.2768775686017495,
+      "learning_rate": 1.6433118457024094e-06,
+      "loss": 0.5833,
+      "step": 12220
+    },
+    {
+      "epoch": 0.6235660021414368,
+      "grad_norm": 7.166817676607632,
+      "learning_rate": 1.6394345179951293e-06,
+      "loss": 0.464,
+      "step": 12230
+    },
+    {
+      "epoch": 0.6240758680467037,
+      "grad_norm": 7.212770160890743,
+      "learning_rate": 1.6355595371084627e-06,
+      "loss": 0.5278,
+      "step": 12240
+    },
+    {
+      "epoch": 0.6245857339519706,
+      "grad_norm": 10.940639389400307,
+      "learning_rate": 1.6316869136097467e-06,
+      "loss": 0.541,
+      "step": 12250
+    },
+    {
+      "epoch": 0.6250955998572375,
+      "grad_norm": 7.45936713616669,
+      "learning_rate": 1.6278166580598897e-06,
+      "loss": 0.4824,
+      "step": 12260
+    },
+    {
+      "epoch": 0.6256054657625044,
+      "grad_norm": 5.728147171292839,
+      "learning_rate": 1.6239487810133404e-06,
+      "loss": 0.5447,
+      "step": 12270
+    },
+    {
+      "epoch": 0.6261153316677713,
+      "grad_norm": 3.2085803535811,
+      "learning_rate": 1.6200832930180643e-06,
+      "loss": 0.489,
+      "step": 12280
+    },
+    {
+      "epoch": 0.6266251975730383,
+      "grad_norm": 11.435384897927873,
+      "learning_rate": 1.6162202046155085e-06,
+      "loss": 0.5655,
+      "step": 12290
+    },
+    {
+      "epoch": 0.6271350634783052,
+      "grad_norm": 5.80944804251215,
+      "learning_rate": 1.6123595263405783e-06,
+      "loss": 0.5259,
+      "step": 12300
+    },
+    {
+      "epoch": 0.6276449293835721,
+      "grad_norm": 8.788085075044139,
+      "learning_rate": 1.6085012687216078e-06,
+      "loss": 0.4623,
+      "step": 12310
+    },
+    {
+      "epoch": 0.628154795288839,
+      "grad_norm": 8.084105701558203,
+      "learning_rate": 1.6046454422803253e-06,
+      "loss": 0.5811,
+      "step": 12320
+    },
+    {
+      "epoch": 0.6286646611941059,
+      "grad_norm": 6.232198726308961,
+      "learning_rate": 1.6007920575318334e-06,
+      "loss": 0.5777,
+      "step": 12330
+    },
+    {
+      "epoch": 0.6291745270993728,
+      "grad_norm": 11.130383595276275,
+      "learning_rate": 1.5969411249845737e-06,
+      "loss": 0.5393,
+      "step": 12340
+    },
+    {
+      "epoch": 0.6296843930046397,
+      "grad_norm": 5.992969945305373,
+      "learning_rate": 1.5930926551403025e-06,
+      "loss": 0.4707,
+      "step": 12350
+    },
+    {
+      "epoch": 0.6301942589099067,
+      "grad_norm": 5.463490867539239,
+      "learning_rate": 1.5892466584940574e-06,
+      "loss": 0.564,
+      "step": 12360
+    },
+    {
+      "epoch": 0.6307041248151736,
+      "grad_norm": 5.145508622723425,
+      "learning_rate": 1.5854031455341332e-06,
+      "loss": 0.5216,
+      "step": 12370
+    },
+    {
+      "epoch": 0.6312139907204405,
+      "grad_norm": 6.430541005929227,
+      "learning_rate": 1.5815621267420526e-06,
+      "loss": 0.5294,
+      "step": 12380
+    },
+    {
+      "epoch": 0.6317238566257074,
+      "grad_norm": 15.519128813172943,
+      "learning_rate": 1.5777236125925333e-06,
+      "loss": 0.4695,
+      "step": 12390
+    },
+    {
+      "epoch": 0.6322337225309743,
+      "grad_norm": 16.999741625919526,
+      "learning_rate": 1.5738876135534659e-06,
+      "loss": 0.5651,
+      "step": 12400
+    },
+    {
+      "epoch": 0.6327435884362412,
+      "grad_norm": 16.449734306702062,
+      "learning_rate": 1.5700541400858793e-06,
+      "loss": 0.5825,
+      "step": 12410
+    },
+    {
+      "epoch": 0.6332534543415081,
+      "grad_norm": 8.678524232749545,
+      "learning_rate": 1.5662232026439172e-06,
+      "loss": 0.5247,
+      "step": 12420
+    },
+    {
+      "epoch": 0.6337633202467751,
+      "grad_norm": 5.544135324516816,
+      "learning_rate": 1.5623948116748074e-06,
+      "loss": 0.4806,
+      "step": 12430
+    },
+    {
+      "epoch": 0.634273186152042,
+      "grad_norm": 8.135829474616635,
+      "learning_rate": 1.5585689776188321e-06,
+      "loss": 0.5846,
+      "step": 12440
+    },
+    {
+      "epoch": 0.6347830520573089,
+      "grad_norm": 4.438831891658201,
+      "learning_rate": 1.5547457109093004e-06,
+      "loss": 0.445,
+      "step": 12450
+    },
+    {
+      "epoch": 0.6352929179625758,
+      "grad_norm": 5.302529484772549,
+      "learning_rate": 1.5509250219725207e-06,
+      "loss": 0.4947,
+      "step": 12460
+    },
+    {
+      "epoch": 0.6358027838678427,
+      "grad_norm": 7.124606505211159,
+      "learning_rate": 1.5471069212277729e-06,
+      "loss": 0.5015,
+      "step": 12470
+    },
+    {
+      "epoch": 0.6363126497731096,
+      "grad_norm": 18.000002055899103,
+      "learning_rate": 1.5432914190872757e-06,
+      "loss": 0.5355,
+      "step": 12480
+    },
+    {
+      "epoch": 0.6368225156783766,
+      "grad_norm": 10.878967475009329,
+      "learning_rate": 1.539478525956164e-06,
+      "loss": 0.5542,
+      "step": 12490
+    },
+    {
+      "epoch": 0.6373323815836435,
+      "grad_norm": 12.088166395329571,
+      "learning_rate": 1.5356682522324578e-06,
+      "loss": 0.4835,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6378422474889104,
+      "grad_norm": 35.076987710234555,
+      "learning_rate": 1.5318606083070305e-06,
+      "loss": 0.5917,
+      "step": 12510
+    },
+    {
+      "epoch": 0.6383521133941773,
+      "grad_norm": 13.112714878589523,
+      "learning_rate": 1.5280556045635881e-06,
+      "loss": 0.5474,
+      "step": 12520
+    },
+    {
+      "epoch": 0.6388619792994442,
+      "grad_norm": 6.40859233326716,
+      "learning_rate": 1.5242532513786334e-06,
+      "loss": 0.5124,
+      "step": 12530
+    },
+    {
+      "epoch": 0.6393718452047111,
+      "grad_norm": 8.147985062950728,
+      "learning_rate": 1.5204535591214428e-06,
+      "loss": 0.5748,
+      "step": 12540
+    },
+    {
+      "epoch": 0.639881711109978,
+      "grad_norm": 5.978197749841948,
+      "learning_rate": 1.516656538154035e-06,
+      "loss": 0.5753,
+      "step": 12550
+    },
+    {
+      "epoch": 0.640391577015245,
+      "grad_norm": 5.3901524765660085,
+      "learning_rate": 1.5128621988311448e-06,
+      "loss": 0.4921,
+      "step": 12560
+    },
+    {
+      "epoch": 0.6409014429205119,
+      "grad_norm": 8.021974547296901,
+      "learning_rate": 1.5090705515001949e-06,
+      "loss": 0.5782,
+      "step": 12570
+    },
+    {
+      "epoch": 0.6414113088257788,
+      "grad_norm": 4.898265970748594,
+      "learning_rate": 1.5052816065012635e-06,
+      "loss": 0.5026,
+      "step": 12580
+    },
+    {
+      "epoch": 0.6419211747310457,
+      "grad_norm": 7.58277321463348,
+      "learning_rate": 1.501495374167063e-06,
+      "loss": 0.5916,
+      "step": 12590
+    },
+    {
+      "epoch": 0.6424310406363126,
+      "grad_norm": 5.643643219733153,
+      "learning_rate": 1.497711864822905e-06,
+      "loss": 0.5112,
+      "step": 12600
+    },
+    {
+      "epoch": 0.6429409065415795,
+      "grad_norm": 25.832450519018945,
+      "learning_rate": 1.4939310887866775e-06,
+      "loss": 0.5532,
+      "step": 12610
+    },
+    {
+      "epoch": 0.6434507724468465,
+      "grad_norm": 11.500383214893693,
+      "learning_rate": 1.4901530563688154e-06,
+      "loss": 0.5108,
+      "step": 12620
+    },
+    {
+      "epoch": 0.6439606383521134,
+      "grad_norm": 13.803826215250618,
+      "learning_rate": 1.4863777778722682e-06,
+      "loss": 0.5415,
+      "step": 12630
+    },
+    {
+      "epoch": 0.6444705042573803,
+      "grad_norm": 23.185859827096497,
+      "learning_rate": 1.482605263592478e-06,
+      "loss": 0.5105,
+      "step": 12640
+    },
+    {
+      "epoch": 0.6449803701626472,
+      "grad_norm": 3.1925340110943874,
+      "learning_rate": 1.4788355238173473e-06,
+      "loss": 0.5143,
+      "step": 12650
+    },
+    {
+      "epoch": 0.6454902360679141,
+      "grad_norm": 8.434282532315683,
+      "learning_rate": 1.4750685688272143e-06,
+      "loss": 0.5254,
+      "step": 12660
+    },
+    {
+      "epoch": 0.646000101973181,
+      "grad_norm": 5.696693313005259,
+      "learning_rate": 1.4713044088948197e-06,
+      "loss": 0.4906,
+      "step": 12670
+    },
+    {
+      "epoch": 0.6465099678784479,
+      "grad_norm": 4.446483290889001,
+      "learning_rate": 1.4675430542852848e-06,
+      "loss": 0.5744,
+      "step": 12680
+    },
+    {
+      "epoch": 0.6470198337837149,
+      "grad_norm": 13.981158431091856,
+      "learning_rate": 1.4637845152560804e-06,
+      "loss": 0.5317,
+      "step": 12690
+    },
+    {
+      "epoch": 0.6475296996889818,
+      "grad_norm": 9.908338337938533,
+      "learning_rate": 1.4600288020569959e-06,
+      "loss": 0.5559,
+      "step": 12700
+    },
+    {
+      "epoch": 0.6480395655942487,
+      "grad_norm": 10.346360874512909,
+      "learning_rate": 1.4562759249301185e-06,
+      "loss": 0.6283,
+      "step": 12710
+    },
+    {
+      "epoch": 0.6485494314995156,
+      "grad_norm": 5.455903851149757,
+      "learning_rate": 1.4525258941097985e-06,
+      "loss": 0.5419,
+      "step": 12720
+    },
+    {
+      "epoch": 0.6490592974047825,
+      "grad_norm": 6.911006067636452,
+      "learning_rate": 1.4487787198226244e-06,
+      "loss": 0.5488,
+      "step": 12730
+    },
+    {
+      "epoch": 0.6495691633100494,
+      "grad_norm": 5.265599029986363,
+      "learning_rate": 1.4450344122873985e-06,
+      "loss": 0.4896,
+      "step": 12740
+    },
+    {
+      "epoch": 0.6500790292153164,
+      "grad_norm": 11.776370208991526,
+      "learning_rate": 1.4412929817150982e-06,
+      "loss": 0.5506,
+      "step": 12750
+    },
+    {
+      "epoch": 0.6505888951205833,
+      "grad_norm": 18.5005112032705,
+      "learning_rate": 1.437554438308863e-06,
+      "loss": 0.4875,
+      "step": 12760
+    },
+    {
+      "epoch": 0.6510987610258502,
+      "grad_norm": 8.771512470871057,
+      "learning_rate": 1.4338187922639506e-06,
+      "loss": 0.5322,
+      "step": 12770
+    },
+    {
+      "epoch": 0.6516086269311171,
+      "grad_norm": 27.262326684678758,
+      "learning_rate": 1.430086053767726e-06,
+      "loss": 0.5692,
+      "step": 12780
+    },
+    {
+      "epoch": 0.652118492836384,
+      "grad_norm": 8.73658779249368,
+      "learning_rate": 1.4263562329996194e-06,
+      "loss": 0.5229,
+      "step": 12790
+    },
+    {
+      "epoch": 0.6526283587416509,
+      "grad_norm": 4.696525831770103,
+      "learning_rate": 1.422629340131106e-06,
+      "loss": 0.5722,
+      "step": 12800
+    },
+    {
+      "epoch": 0.6531382246469178,
+      "grad_norm": 13.98465385717505,
+      "learning_rate": 1.4189053853256757e-06,
+      "loss": 0.5333,
+      "step": 12810
+    },
+    {
+      "epoch": 0.6536480905521848,
+      "grad_norm": 44.135847073923244,
+      "learning_rate": 1.4151843787388062e-06,
+      "loss": 0.584,
+      "step": 12820
+    },
+    {
+      "epoch": 0.6541579564574517,
+      "grad_norm": 17.496468258150834,
+      "learning_rate": 1.4114663305179382e-06,
+      "loss": 0.5256,
+      "step": 12830
+    },
+    {
+      "epoch": 0.6546678223627186,
+      "grad_norm": 3.5798846929402095,
+      "learning_rate": 1.4077512508024382e-06,
+      "loss": 0.4815,
+      "step": 12840
+    },
+    {
+      "epoch": 0.6551776882679855,
+      "grad_norm": 27.319618611904193,
+      "learning_rate": 1.4040391497235845e-06,
+      "loss": 0.5426,
+      "step": 12850
+    },
+    {
+      "epoch": 0.6556875541732524,
+      "grad_norm": 3.9535275012824753,
+      "learning_rate": 1.4003300374045283e-06,
+      "loss": 0.5097,
+      "step": 12860
+    },
+    {
+      "epoch": 0.6561974200785193,
+      "grad_norm": 2.874465898696785,
+      "learning_rate": 1.396623923960271e-06,
+      "loss": 0.5031,
+      "step": 12870
+    },
+    {
+      "epoch": 0.6567072859837862,
+      "grad_norm": 5.723509444766761,
+      "learning_rate": 1.3929208194976362e-06,
+      "loss": 0.5727,
+      "step": 12880
+    },
+    {
+      "epoch": 0.6572171518890532,
+      "grad_norm": 10.234880244026453,
+      "learning_rate": 1.3892207341152416e-06,
+      "loss": 0.5734,
+      "step": 12890
+    },
+    {
+      "epoch": 0.6577270177943201,
+      "grad_norm": 11.590413083459318,
+      "learning_rate": 1.385523677903472e-06,
+      "loss": 0.5068,
+      "step": 12900
+    },
+    {
+      "epoch": 0.658236883699587,
+      "grad_norm": 20.26688405326,
+      "learning_rate": 1.38182966094445e-06,
+      "loss": 0.601,
+      "step": 12910
+    },
+    {
+      "epoch": 0.6587467496048539,
+      "grad_norm": 5.873427482586327,
+      "learning_rate": 1.3781386933120133e-06,
+      "loss": 0.4823,
+      "step": 12920
+    },
+    {
+      "epoch": 0.6592566155101208,
+      "grad_norm": 7.732319317412305,
+      "learning_rate": 1.3744507850716804e-06,
+      "loss": 0.517,
+      "step": 12930
+    },
+    {
+      "epoch": 0.6597664814153877,
+      "grad_norm": 5.717087164447723,
+      "learning_rate": 1.3707659462806284e-06,
+      "loss": 0.5071,
+      "step": 12940
+    },
+    {
+      "epoch": 0.6602763473206547,
+      "grad_norm": 12.546369227920732,
+      "learning_rate": 1.367084186987663e-06,
+      "loss": 0.5283,
+      "step": 12950
+    },
+    {
+      "epoch": 0.6607862132259216,
+      "grad_norm": 9.522985256100982,
+      "learning_rate": 1.3634055172331926e-06,
+      "loss": 0.5291,
+      "step": 12960
+    },
+    {
+      "epoch": 0.6612960791311885,
+      "grad_norm": 6.0967021531973025,
+      "learning_rate": 1.3597299470491986e-06,
+      "loss": 0.6057,
+      "step": 12970
+    },
+    {
+      "epoch": 0.6618059450364554,
+      "grad_norm": 15.313091190736756,
+      "learning_rate": 1.356057486459214e-06,
+      "loss": 0.5561,
+      "step": 12980
+    },
+    {
+      "epoch": 0.6623158109417223,
+      "grad_norm": 7.740759385070269,
+      "learning_rate": 1.352388145478285e-06,
+      "loss": 0.5671,
+      "step": 12990
+    },
+    {
+      "epoch": 0.6628256768469892,
+      "grad_norm": 11.418720457485895,
+      "learning_rate": 1.3487219341129566e-06,
+      "loss": 0.4986,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6633355427522561,
+      "grad_norm": 16.280698683867406,
+      "learning_rate": 1.3450588623612353e-06,
+      "loss": 0.5557,
+      "step": 13010
+    },
+    {
+      "epoch": 0.6638454086575231,
+      "grad_norm": 5.81925433785274,
+      "learning_rate": 1.3413989402125682e-06,
+      "loss": 0.5501,
+      "step": 13020
+    },
+    {
+      "epoch": 0.66435527456279,
+      "grad_norm": 5.939874542788059,
+      "learning_rate": 1.3377421776478111e-06,
+      "loss": 0.5353,
+      "step": 13030
+    },
+    {
+      "epoch": 0.6648651404680569,
+      "grad_norm": 8.305569539285225,
+      "learning_rate": 1.3340885846392032e-06,
+      "loss": 0.5852,
+      "step": 13040
+    },
+    {
+      "epoch": 0.6653750063733238,
+      "grad_norm": 5.767160133744185,
+      "learning_rate": 1.3304381711503444e-06,
+      "loss": 0.5129,
+      "step": 13050
+    },
+    {
+      "epoch": 0.6658848722785907,
+      "grad_norm": 6.875680453072486,
+      "learning_rate": 1.3267909471361574e-06,
+      "loss": 0.5727,
+      "step": 13060
+    },
+    {
+      "epoch": 0.6663947381838576,
+      "grad_norm": 5.03774252438953,
+      "learning_rate": 1.3231469225428726e-06,
+      "loss": 0.5317,
+      "step": 13070
+    },
+    {
+      "epoch": 0.6669046040891246,
+      "grad_norm": 5.1407923174582395,
+      "learning_rate": 1.3195061073079901e-06,
+      "loss": 0.5798,
+      "step": 13080
+    },
+    {
+      "epoch": 0.6674144699943915,
+      "grad_norm": 9.799807279239122,
+      "learning_rate": 1.3158685113602636e-06,
+      "loss": 0.4992,
+      "step": 13090
+    },
+    {
+      "epoch": 0.6679243358996584,
+      "grad_norm": 12.17057925211001,
+      "learning_rate": 1.3122341446196636e-06,
+      "loss": 0.5889,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6684342018049253,
+      "grad_norm": 4.2258115828431535,
+      "learning_rate": 1.3086030169973552e-06,
+      "loss": 0.5236,
+      "step": 13110
+    },
+    {
+      "epoch": 0.6689440677101922,
+      "grad_norm": 13.144950859146647,
+      "learning_rate": 1.3049751383956707e-06,
+      "loss": 0.5002,
+      "step": 13120
+    },
+    {
+      "epoch": 0.6694539336154591,
+      "grad_norm": 9.859556462555908,
+      "learning_rate": 1.301350518708081e-06,
+      "loss": 0.5379,
+      "step": 13130
+    },
+    {
+      "epoch": 0.669963799520726,
+      "grad_norm": 8.997523392336818,
+      "learning_rate": 1.2977291678191733e-06,
+      "loss": 0.5159,
+      "step": 13140
+    },
+    {
+      "epoch": 0.670473665425993,
+      "grad_norm": 11.962543453038034,
+      "learning_rate": 1.2941110956046142e-06,
+      "loss": 0.5201,
+      "step": 13150
+    },
+    {
+      "epoch": 0.6709835313312599,
+      "grad_norm": 9.211901098231362,
+      "learning_rate": 1.290496311931135e-06,
+      "loss": 0.5275,
+      "step": 13160
+    },
+    {
+      "epoch": 0.6714933972365268,
+      "grad_norm": 5.802536276229398,
+      "learning_rate": 1.2868848266564964e-06,
+      "loss": 0.4901,
+      "step": 13170
+    },
+    {
+      "epoch": 0.6720032631417937,
+      "grad_norm": 9.337922159510477,
+      "learning_rate": 1.2832766496294647e-06,
+      "loss": 0.5323,
+      "step": 13180
+    },
+    {
+      "epoch": 0.6725131290470606,
+      "grad_norm": 7.120905849429728,
+      "learning_rate": 1.2796717906897831e-06,
+      "loss": 0.5164,
+      "step": 13190
+    },
+    {
+      "epoch": 0.6730229949523275,
+      "grad_norm": 9.733236048311111,
+      "learning_rate": 1.2760702596681478e-06,
+      "loss": 0.5552,
+      "step": 13200
+    },
+    {
+      "epoch": 0.6735328608575945,
+      "grad_norm": 3.3631007408034677,
+      "learning_rate": 1.2724720663861792e-06,
+      "loss": 0.4785,
+      "step": 13210
+    },
+    {
+      "epoch": 0.6740427267628614,
+      "grad_norm": 13.547952544012343,
+      "learning_rate": 1.2688772206563938e-06,
+      "loss": 0.5631,
+      "step": 13220
+    },
+    {
+      "epoch": 0.6745525926681283,
+      "grad_norm": 7.191362755646714,
+      "learning_rate": 1.2652857322821821e-06,
+      "loss": 0.4977,
+      "step": 13230
+    },
+    {
+      "epoch": 0.6750624585733952,
+      "grad_norm": 5.156968959766061,
+      "learning_rate": 1.2616976110577766e-06,
+      "loss": 0.5453,
+      "step": 13240
+    },
+    {
+      "epoch": 0.6755723244786621,
+      "grad_norm": 11.097231281107701,
+      "learning_rate": 1.2581128667682286e-06,
+      "loss": 0.547,
+      "step": 13250
+    },
+    {
+      "epoch": 0.676082190383929,
+      "grad_norm": 19.165171233840805,
+      "learning_rate": 1.2545315091893784e-06,
+      "loss": 0.5206,
+      "step": 13260
+    },
+    {
+      "epoch": 0.6765920562891959,
+      "grad_norm": 35.51519567896334,
+      "learning_rate": 1.2509535480878325e-06,
+      "loss": 0.4864,
+      "step": 13270
+    },
+    {
+      "epoch": 0.6771019221944629,
+      "grad_norm": 5.511431626465132,
+      "learning_rate": 1.2473789932209331e-06,
+      "loss": 0.4802,
+      "step": 13280
+    },
+    {
+      "epoch": 0.6776117880997298,
+      "grad_norm": 11.049237655524086,
+      "learning_rate": 1.2438078543367377e-06,
+      "loss": 0.6008,
+      "step": 13290
+    },
+    {
+      "epoch": 0.6781216540049967,
+      "grad_norm": 12.081808269170782,
+      "learning_rate": 1.2402401411739806e-06,
+      "loss": 0.5358,
+      "step": 13300
+    },
+    {
+      "epoch": 0.6786315199102636,
+      "grad_norm": 6.161026900799108,
+      "learning_rate": 1.2366758634620615e-06,
+      "loss": 0.5713,
+      "step": 13310
+    },
+    {
+      "epoch": 0.6791413858155305,
+      "grad_norm": 6.293422338189026,
+      "learning_rate": 1.2331150309210075e-06,
+      "loss": 0.556,
+      "step": 13320
+    },
+    {
+      "epoch": 0.6796512517207974,
+      "grad_norm": 4.283687942029315,
+      "learning_rate": 1.2295576532614506e-06,
+      "loss": 0.5029,
+      "step": 13330
+    },
+    {
+      "epoch": 0.6801611176260643,
+      "grad_norm": 7.352841148937369,
+      "learning_rate": 1.226003740184602e-06,
+      "loss": 0.5361,
+      "step": 13340
+    },
+    {
+      "epoch": 0.6806709835313313,
+      "grad_norm": 175.73598698371708,
+      "learning_rate": 1.2224533013822237e-06,
+      "loss": 0.5526,
+      "step": 13350
+    },
+    {
+      "epoch": 0.6811808494365982,
+      "grad_norm": 12.027309626664836,
+      "learning_rate": 1.2189063465366064e-06,
+      "loss": 0.6395,
+      "step": 13360
+    },
+    {
+      "epoch": 0.6816907153418651,
+      "grad_norm": 12.040883913595133,
+      "learning_rate": 1.2153628853205336e-06,
+      "loss": 0.5229,
+      "step": 13370
+    },
+    {
+      "epoch": 0.682200581247132,
+      "grad_norm": 8.000181612684882,
+      "learning_rate": 1.2118229273972684e-06,
+      "loss": 0.5347,
+      "step": 13380
+    },
+    {
+      "epoch": 0.6827104471523989,
+      "grad_norm": 12.028638552739197,
+      "learning_rate": 1.2082864824205138e-06,
+      "loss": 0.5195,
+      "step": 13390
+    },
+    {
+      "epoch": 0.6832203130576658,
+      "grad_norm": 20.371683560997308,
+      "learning_rate": 1.2047535600343984e-06,
+      "loss": 0.5081,
+      "step": 13400
+    },
+    {
+      "epoch": 0.6837301789629328,
+      "grad_norm": 8.136173607931632,
+      "learning_rate": 1.2012241698734408e-06,
+      "loss": 0.5893,
+      "step": 13410
+    },
+    {
+      "epoch": 0.6842400448681997,
+      "grad_norm": 9.232546953455795,
+      "learning_rate": 1.1976983215625285e-06,
+      "loss": 0.5348,
+      "step": 13420
+    },
+    {
+      "epoch": 0.6847499107734666,
+      "grad_norm": 4.377944461206342,
+      "learning_rate": 1.1941760247168893e-06,
+      "loss": 0.5529,
+      "step": 13430
+    },
+    {
+      "epoch": 0.6852597766787335,
+      "grad_norm": 3.459532633536289,
+      "learning_rate": 1.1906572889420655e-06,
+      "loss": 0.4904,
+      "step": 13440
+    },
+    {
+      "epoch": 0.6857696425840004,
+      "grad_norm": 6.8068993996494775,
+      "learning_rate": 1.1871421238338917e-06,
+      "loss": 0.5179,
+      "step": 13450
+    },
+    {
+      "epoch": 0.6862795084892673,
+      "grad_norm": 9.388231515206153,
+      "learning_rate": 1.1836305389784588e-06,
+      "loss": 0.5772,
+      "step": 13460
+    },
+    {
+      "epoch": 0.6867893743945342,
+      "grad_norm": 3.8333730829756174,
+      "learning_rate": 1.1801225439521003e-06,
+      "loss": 0.4779,
+      "step": 13470
+    },
+    {
+      "epoch": 0.6872992402998012,
+      "grad_norm": 6.9070129435643,
+      "learning_rate": 1.176618148321356e-06,
+      "loss": 0.4894,
+      "step": 13480
+    },
+    {
+      "epoch": 0.6878091062050681,
+      "grad_norm": 7.058960081890223,
+      "learning_rate": 1.1731173616429514e-06,
+      "loss": 0.4723,
+      "step": 13490
+    },
+    {
+      "epoch": 0.688318972110335,
+      "grad_norm": 7.419454694113433,
+      "learning_rate": 1.16962019346377e-06,
+      "loss": 0.4849,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6888288380156019,
+      "grad_norm": 6.911766062790672,
+      "learning_rate": 1.1661266533208274e-06,
+      "loss": 0.5458,
+      "step": 13510
+    },
+    {
+      "epoch": 0.6893387039208688,
+      "grad_norm": 25.73500382065693,
+      "learning_rate": 1.1626367507412443e-06,
+      "loss": 0.5025,
+      "step": 13520
+    },
+    {
+      "epoch": 0.6898485698261357,
+      "grad_norm": 9.057037998210602,
+      "learning_rate": 1.1591504952422243e-06,
+      "loss": 0.5203,
+      "step": 13530
+    },
+    {
+      "epoch": 0.6903584357314027,
+      "grad_norm": 16.41052757265655,
+      "learning_rate": 1.1556678963310222e-06,
+      "loss": 0.5844,
+      "step": 13540
+    },
+    {
+      "epoch": 0.6908683016366696,
+      "grad_norm": 5.732859564218691,
+      "learning_rate": 1.152188963504922e-06,
+      "loss": 0.5078,
+      "step": 13550
+    },
+    {
+      "epoch": 0.6913781675419365,
+      "grad_norm": 5.929076421646337,
+      "learning_rate": 1.148713706251211e-06,
+      "loss": 0.5164,
+      "step": 13560
+    },
+    {
+      "epoch": 0.6918880334472034,
+      "grad_norm": 9.301935237585717,
+      "learning_rate": 1.1452421340471514e-06,
+      "loss": 0.55,
+      "step": 13570
+    },
+    {
+      "epoch": 0.6923978993524703,
+      "grad_norm": 8.655543019346618,
+      "learning_rate": 1.1417742563599568e-06,
+      "loss": 0.5696,
+      "step": 13580
+    },
+    {
+      "epoch": 0.6929077652577372,
+      "grad_norm": 10.889091625170066,
+      "learning_rate": 1.1383100826467653e-06,
+      "loss": 0.5868,
+      "step": 13590
+    },
+    {
+      "epoch": 0.6934176311630041,
+      "grad_norm": 5.809789419158467,
+      "learning_rate": 1.1348496223546162e-06,
+      "loss": 0.5087,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6939274970682711,
+      "grad_norm": 21.7701581078474,
+      "learning_rate": 1.131392884920417e-06,
+      "loss": 0.4572,
+      "step": 13610
+    },
+    {
+      "epoch": 0.694437362973538,
+      "grad_norm": 36.72901867309476,
+      "learning_rate": 1.1279398797709293e-06,
+      "loss": 0.5561,
+      "step": 13620
+    },
+    {
+      "epoch": 0.6949472288788049,
+      "grad_norm": 7.015828755810641,
+      "learning_rate": 1.1244906163227295e-06,
+      "loss": 0.5569,
+      "step": 13630
+    },
+    {
+      "epoch": 0.6954570947840718,
+      "grad_norm": 5.717141506625831,
+      "learning_rate": 1.1210451039821965e-06,
+      "loss": 0.5606,
+      "step": 13640
+    },
+    {
+      "epoch": 0.6959669606893387,
+      "grad_norm": 200.18780369162442,
+      "learning_rate": 1.1176033521454758e-06,
+      "loss": 0.5456,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6964768265946056,
+      "grad_norm": 8.019380290741653,
+      "learning_rate": 1.114165370198458e-06,
+      "loss": 0.5384,
+      "step": 13660
+    },
+    {
+      "epoch": 0.6969866924998726,
+      "grad_norm": 15.058488690780067,
+      "learning_rate": 1.1107311675167558e-06,
+      "loss": 0.5397,
+      "step": 13670
+    },
+    {
+      "epoch": 0.6974965584051395,
+      "grad_norm": 8.87322964539086,
+      "learning_rate": 1.1073007534656712e-06,
+      "loss": 0.5328,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6980064243104064,
+      "grad_norm": 7.857307785773062,
+      "learning_rate": 1.1038741374001793e-06,
+      "loss": 0.5493,
+      "step": 13690
+    },
+    {
+      "epoch": 0.6985162902156733,
+      "grad_norm": 48.793916370591866,
+      "learning_rate": 1.1004513286648922e-06,
+      "loss": 0.5117,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6990261561209402,
+      "grad_norm": 3.4690544397426195,
+      "learning_rate": 1.0970323365940443e-06,
+      "loss": 0.5075,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6995360220262071,
+      "grad_norm": 5.720335109430535,
+      "learning_rate": 1.093617170511459e-06,
+      "loss": 0.5554,
+      "step": 13720
+    },
+    {
+      "epoch": 0.700045887931474,
+      "grad_norm": 3.592681308182018,
+      "learning_rate": 1.0902058397305268e-06,
+      "loss": 0.5351,
+      "step": 13730
+    },
+    {
+      "epoch": 0.700555753836741,
+      "grad_norm": 9.773333278685,
+      "learning_rate": 1.0867983535541785e-06,
+      "loss": 0.5818,
+      "step": 13740
+    },
+    {
+      "epoch": 0.7010656197420079,
+      "grad_norm": 10.27186569790017,
+      "learning_rate": 1.0833947212748597e-06,
+      "loss": 0.5552,
+      "step": 13750
+    },
+    {
+      "epoch": 0.7015754856472748,
+      "grad_norm": 13.36103278733716,
+      "learning_rate": 1.07999495217451e-06,
+      "loss": 0.616,
+      "step": 13760
+    },
+    {
+      "epoch": 0.7020853515525417,
+      "grad_norm": 5.392163411689741,
+      "learning_rate": 1.0765990555245275e-06,
+      "loss": 0.5264,
+      "step": 13770
+    },
+    {
+      "epoch": 0.7025952174578086,
+      "grad_norm": 10.72345270938434,
+      "learning_rate": 1.0732070405857562e-06,
+      "loss": 0.5634,
+      "step": 13780
+    },
+    {
+      "epoch": 0.7031050833630755,
+      "grad_norm": 15.80049388002083,
+      "learning_rate": 1.0698189166084501e-06,
+      "loss": 0.466,
+      "step": 13790
+    },
+    {
+      "epoch": 0.7036149492683424,
+      "grad_norm": 4.82132484505896,
+      "learning_rate": 1.0664346928322547e-06,
+      "loss": 0.5186,
+      "step": 13800
+    },
+    {
+      "epoch": 0.7041248151736094,
+      "grad_norm": 6.069159826575058,
+      "learning_rate": 1.063054378486178e-06,
+      "loss": 0.634,
+      "step": 13810
+    },
+    {
+      "epoch": 0.7046346810788763,
+      "grad_norm": 11.614673101846604,
+      "learning_rate": 1.059677982788567e-06,
+      "loss": 0.4856,
+      "step": 13820
+    },
+    {
+      "epoch": 0.7051445469841432,
+      "grad_norm": 14.11494362911992,
+      "learning_rate": 1.056305514947082e-06,
+      "loss": 0.4981,
+      "step": 13830
+    },
+    {
+      "epoch": 0.7056544128894101,
+      "grad_norm": 8.007261173840975,
+      "learning_rate": 1.0529369841586743e-06,
+      "loss": 0.4831,
+      "step": 13840
+    },
+    {
+      "epoch": 0.706164278794677,
+      "grad_norm": 17.304524008597603,
+      "learning_rate": 1.0495723996095533e-06,
+      "loss": 0.4787,
+      "step": 13850
+    },
+    {
+      "epoch": 0.7066741446999439,
+      "grad_norm": 10.05312890548322,
+      "learning_rate": 1.046211770475173e-06,
+      "loss": 0.4995,
+      "step": 13860
+    },
+    {
+      "epoch": 0.7071840106052109,
+      "grad_norm": 7.738138897636104,
+      "learning_rate": 1.0428551059201964e-06,
+      "loss": 0.4952,
+      "step": 13870
+    },
+    {
+      "epoch": 0.7076938765104778,
+      "grad_norm": 3.979757475584907,
+      "learning_rate": 1.039502415098476e-06,
+      "loss": 0.5682,
+      "step": 13880
+    },
+    {
+      "epoch": 0.7082037424157447,
+      "grad_norm": 16.9521365709318,
+      "learning_rate": 1.0361537071530277e-06,
+      "loss": 0.5504,
+      "step": 13890
+    },
+    {
+      "epoch": 0.7087136083210116,
+      "grad_norm": 13.478011647892208,
+      "learning_rate": 1.0328089912160055e-06,
+      "loss": 0.5764,
+      "step": 13900
+    },
+    {
+      "epoch": 0.7092234742262785,
+      "grad_norm": 14.12969618493761,
+      "learning_rate": 1.0294682764086794e-06,
+      "loss": 0.5445,
+      "step": 13910
+    },
+    {
+      "epoch": 0.7097333401315454,
+      "grad_norm": 5.795649950376214,
+      "learning_rate": 1.0261315718414028e-06,
+      "loss": 0.5645,
+      "step": 13920
+    },
+    {
+      "epoch": 0.7102432060368123,
+      "grad_norm": 11.710592013823014,
+      "learning_rate": 1.0227988866135995e-06,
+      "loss": 0.5083,
+      "step": 13930
+    },
+    {
+      "epoch": 0.7107530719420793,
+      "grad_norm": 12.43499461710959,
+      "learning_rate": 1.0194702298137251e-06,
+      "loss": 0.4863,
+      "step": 13940
+    },
+    {
+      "epoch": 0.7112629378473462,
+      "grad_norm": 8.863719068408987,
+      "learning_rate": 1.016145610519256e-06,
+      "loss": 0.5543,
+      "step": 13950
+    },
+    {
+      "epoch": 0.7117728037526131,
+      "grad_norm": 5.42342484070064,
+      "learning_rate": 1.0128250377966545e-06,
+      "loss": 0.559,
+      "step": 13960
+    },
+    {
+      "epoch": 0.71228266965788,
+      "grad_norm": 5.137324371851054,
+      "learning_rate": 1.009508520701347e-06,
+      "loss": 0.5158,
+      "step": 13970
+    },
+    {
+      "epoch": 0.7127925355631469,
+      "grad_norm": 6.755524106797482,
+      "learning_rate": 1.006196068277704e-06,
+      "loss": 0.4776,
+      "step": 13980
+    },
+    {
+      "epoch": 0.7133024014684138,
+      "grad_norm": 5.199631700415942,
+      "learning_rate": 1.002887689559005e-06,
+      "loss": 0.5302,
+      "step": 13990
+    },
+    {
+      "epoch": 0.7138122673736808,
+      "grad_norm": 5.152070108304142,
+      "learning_rate": 9.99583393567428e-07,
+      "loss": 0.4551,
+      "step": 14000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 19613,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9501248720994304.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}