{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7640319258496395,
  "eval_steps": 500,
  "global_step": 5935,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0012873326467559218,
      "grad_norm": 420.831391091495,
      "learning_rate": 2.910112359550562e-06,
      "loss": 1.9312837600708008,
      "step": 10
    },
    {
      "epoch": 0.0025746652935118436,
      "grad_norm": 183.74059849187165,
      "learning_rate": 3.921348314606742e-06,
      "loss": 1.0653209686279297,
      "step": 20
    },
    {
      "epoch": 0.0038619979402677654,
      "grad_norm": 209.9298344604747,
      "learning_rate": 4.932584269662922e-06,
      "loss": 0.9056184768676758,
      "step": 30
    },
    {
      "epoch": 0.005149330587023687,
      "grad_norm": 77.6500325582647,
      "learning_rate": 5.943820224719102e-06,
      "loss": 0.8960558891296386,
      "step": 40
    },
    {
      "epoch": 0.006436663233779609,
      "grad_norm": 284.47230377729784,
      "learning_rate": 6.955056179775282e-06,
      "loss": 0.8537825584411621,
      "step": 50
    },
    {
      "epoch": 0.007723995880535531,
      "grad_norm": 39.79412112786639,
      "learning_rate": 7.966292134831462e-06,
      "loss": 0.9032889366149902,
      "step": 60
    },
    {
      "epoch": 0.009011328527291453,
      "grad_norm": 19.11666869823677,
      "learning_rate": 8.977528089887641e-06,
      "loss": 0.8577792167663574,
      "step": 70
    },
    {
      "epoch": 0.010298661174047374,
      "grad_norm": 194.15492366269652,
      "learning_rate": 9.988764044943822e-06,
      "loss": 0.7952291011810303,
      "step": 80
    },
    {
      "epoch": 0.011585993820803296,
      "grad_norm": 69.50742426880747,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.922086524963379,
      "step": 90
    },
    {
      "epoch": 0.012873326467559218,
      "grad_norm": 14.130976114030805,
      "learning_rate": 1.2011235955056182e-05,
      "loss": 0.8015565872192383,
      "step": 100
    },
    {
      "epoch": 0.01416065911431514,
      "grad_norm": 53.39139843875411,
      "learning_rate": 1.3022471910112362e-05,
      "loss": 0.798184585571289,
      "step": 110
    },
    {
      "epoch": 0.015447991761071062,
      "grad_norm": 9.335386256610237,
      "learning_rate": 1.403370786516854e-05,
      "loss": 0.8048606872558594,
      "step": 120
    },
    {
      "epoch": 0.016735324407826983,
      "grad_norm": 16.579058293598365,
      "learning_rate": 1.504494382022472e-05,
      "loss": 0.7458171844482422,
      "step": 130
    },
    {
      "epoch": 0.018022657054582905,
      "grad_norm": 10.216833945092391,
      "learning_rate": 1.60561797752809e-05,
      "loss": 0.7754960536956788,
      "step": 140
    },
    {
      "epoch": 0.019309989701338827,
      "grad_norm": 4.716209240987936,
      "learning_rate": 1.706741573033708e-05,
      "loss": 0.7410296440124512,
      "step": 150
    },
    {
      "epoch": 0.02059732234809475,
      "grad_norm": 7.373374743890979,
      "learning_rate": 1.8078651685393256e-05,
      "loss": 0.7318727493286132,
      "step": 160
    },
    {
      "epoch": 0.02188465499485067,
      "grad_norm": 2.832195397619456,
      "learning_rate": 1.908988764044944e-05,
      "loss": 0.7140597343444824,
      "step": 170
    },
    {
      "epoch": 0.023171987641606592,
      "grad_norm": 8.407785684478657,
      "learning_rate": 2e-05,
      "loss": 0.7257099151611328,
      "step": 180
    },
    {
      "epoch": 0.024459320288362514,
      "grad_norm": 68.6412229116751,
      "learning_rate": 2e-05,
      "loss": 0.7461828231811524,
      "step": 190
    },
    {
      "epoch": 0.025746652935118436,
      "grad_norm": 6.17895198608428,
      "learning_rate": 2e-05,
      "loss": 0.7821893215179443,
      "step": 200
    },
    {
      "epoch": 0.027033985581874358,
      "grad_norm": 5.194394848131495,
      "learning_rate": 2e-05,
      "loss": 0.7375617027282715,
      "step": 210
    },
    {
      "epoch": 0.02832131822863028,
      "grad_norm": 201.26051157890333,
      "learning_rate": 2e-05,
      "loss": 0.7686418056488037,
      "step": 220
    },
    {
      "epoch": 0.0296086508753862,
      "grad_norm": 28.711764224889286,
      "learning_rate": 2e-05,
      "loss": 0.7339415550231934,
      "step": 230
    },
    {
      "epoch": 0.030895983522142123,
      "grad_norm": 9.11168724332578,
      "learning_rate": 2e-05,
      "loss": 0.7917026519775391,
      "step": 240
    },
    {
      "epoch": 0.032183316168898045,
      "grad_norm": 18.075887930769106,
      "learning_rate": 2e-05,
      "loss": 0.7600293636322022,
      "step": 250
    },
    {
      "epoch": 0.03347064881565397,
      "grad_norm": 6.699833454866563,
      "learning_rate": 2e-05,
      "loss": 0.8107472419738769,
      "step": 260
    },
    {
      "epoch": 0.03475798146240989,
      "grad_norm": 14.699594175403195,
      "learning_rate": 2e-05,
      "loss": 0.8207590103149414,
      "step": 270
    },
    {
      "epoch": 0.03604531410916581,
      "grad_norm": 1.9113896642609345,
      "learning_rate": 2e-05,
      "loss": 0.7948224544525146,
      "step": 280
    },
    {
      "epoch": 0.03733264675592173,
      "grad_norm": 5.196177104534392,
      "learning_rate": 2e-05,
      "loss": 0.7511525630950928,
      "step": 290
    },
    {
      "epoch": 0.038619979402677654,
      "grad_norm": 6.297124236914922,
      "learning_rate": 2e-05,
      "loss": 0.7438372611999512,
      "step": 300
    },
    {
      "epoch": 0.039907312049433576,
      "grad_norm": 11.256600743830386,
      "learning_rate": 2e-05,
      "loss": 0.7340342998504639,
      "step": 310
    },
    {
      "epoch": 0.0411946446961895,
      "grad_norm": 12.03605841268819,
      "learning_rate": 2e-05,
      "loss": 0.7668326377868653,
      "step": 320
    },
    {
      "epoch": 0.04248197734294542,
      "grad_norm": 10.454100982721279,
      "learning_rate": 2e-05,
      "loss": 0.744885778427124,
      "step": 330
    },
    {
      "epoch": 0.04376930998970134,
      "grad_norm": 31.42354342277286,
      "learning_rate": 2e-05,
      "loss": 0.7327960968017578,
      "step": 340
    },
    {
      "epoch": 0.04505664263645726,
      "grad_norm": 8.65794301406385,
      "learning_rate": 2e-05,
      "loss": 0.8318517684936524,
      "step": 350
    },
    {
      "epoch": 0.046343975283213185,
      "grad_norm": 64.03382663360811,
      "learning_rate": 2e-05,
      "loss": 0.7727777004241944,
      "step": 360
    },
    {
      "epoch": 0.047631307929969106,
      "grad_norm": 5.767854062175975,
      "learning_rate": 2e-05,
      "loss": 0.7937121868133545,
      "step": 370
    },
    {
      "epoch": 0.04891864057672503,
      "grad_norm": 34.86438082009496,
      "learning_rate": 2e-05,
      "loss": 0.8124513626098633,
      "step": 380
    },
    {
      "epoch": 0.05020597322348095,
      "grad_norm": 17.861985503962952,
      "learning_rate": 2e-05,
      "loss": 0.782343578338623,
      "step": 390
    },
    {
      "epoch": 0.05149330587023687,
      "grad_norm": 5.273597714761184,
      "learning_rate": 2e-05,
      "loss": 0.7975768089294434,
      "step": 400
    },
    {
      "epoch": 0.052780638516992794,
      "grad_norm": 3.785340696847242,
      "learning_rate": 2e-05,
      "loss": 0.7699785232543945,
      "step": 410
    },
    {
      "epoch": 0.054067971163748715,
      "grad_norm": 2.0191165245283518,
      "learning_rate": 2e-05,
      "loss": 0.7567674160003662,
      "step": 420
    },
    {
      "epoch": 0.05535530381050464,
      "grad_norm": 2.684997153366744,
      "learning_rate": 2e-05,
      "loss": 0.8061090469360351,
      "step": 430
    },
    {
      "epoch": 0.05664263645726056,
      "grad_norm": 4.464002674035226,
      "learning_rate": 2e-05,
      "loss": 0.7761743068695068,
      "step": 440
    },
    {
      "epoch": 0.05792996910401648,
      "grad_norm": 2.113290503315734,
      "learning_rate": 2e-05,
      "loss": 0.7724208354949951,
      "step": 450
    },
    {
      "epoch": 0.0592173017507724,
      "grad_norm": 4.63126727046047,
      "learning_rate": 2e-05,
      "loss": 0.7622805595397949,
      "step": 460
    },
    {
      "epoch": 0.060504634397528324,
      "grad_norm": 9.246708733487706,
      "learning_rate": 2e-05,
      "loss": 0.6900830268859863,
      "step": 470
    },
    {
      "epoch": 0.061791967044284246,
      "grad_norm": 5.015943304904472,
      "learning_rate": 2e-05,
      "loss": 0.7525927543640136,
      "step": 480
    },
    {
      "epoch": 0.06307929969104016,
      "grad_norm": 7.279183898929732,
      "learning_rate": 2e-05,
      "loss": 0.7771492004394531,
      "step": 490
    },
    {
      "epoch": 0.06436663233779609,
      "grad_norm": 5.373066892709014,
      "learning_rate": 2e-05,
      "loss": 0.8580041885375976,
      "step": 500
    },
    {
      "epoch": 0.065653964984552,
      "grad_norm": 15.27286294173738,
      "learning_rate": 2e-05,
      "loss": 0.8757843017578125,
      "step": 510
    },
    {
      "epoch": 0.06694129763130793,
      "grad_norm": 9.932666259805304,
      "learning_rate": 2e-05,
      "loss": 0.8020495414733887,
      "step": 520
    },
    {
      "epoch": 0.06822863027806385,
      "grad_norm": 3.301286188883536,
      "learning_rate": 2e-05,
      "loss": 0.7554676055908203,
      "step": 530
    },
    {
      "epoch": 0.06951596292481978,
      "grad_norm": 5.387035655614053,
      "learning_rate": 2e-05,
      "loss": 0.748227596282959,
      "step": 540
    },
    {
      "epoch": 0.07080329557157569,
      "grad_norm": 51.37416211353129,
      "learning_rate": 2e-05,
      "loss": 0.7636474609375,
      "step": 550
    },
    {
      "epoch": 0.07209062821833162,
      "grad_norm": 33.339739170373484,
      "learning_rate": 2e-05,
      "loss": 0.7783924102783203,
      "step": 560
    },
    {
      "epoch": 0.07337796086508754,
      "grad_norm": 5.732333781925769,
      "learning_rate": 2e-05,
      "loss": 0.793330717086792,
      "step": 570
    },
    {
      "epoch": 0.07466529351184346,
      "grad_norm": 4.100917123304358,
      "learning_rate": 2e-05,
      "loss": 0.7694579601287842,
      "step": 580
    },
    {
      "epoch": 0.07595262615859938,
      "grad_norm": 2.0715078041261332,
      "learning_rate": 2e-05,
      "loss": 0.8294933319091797,
      "step": 590
    },
    {
      "epoch": 0.07723995880535531,
      "grad_norm": 3.184866964539304,
      "learning_rate": 2e-05,
      "loss": 0.7459193706512451,
      "step": 600
    },
    {
      "epoch": 0.07852729145211122,
      "grad_norm": 7.478907825414544,
      "learning_rate": 2e-05,
      "loss": 0.7193670272827148,
      "step": 610
    },
    {
      "epoch": 0.07981462409886715,
      "grad_norm": 8.452718548100744,
      "learning_rate": 2e-05,
      "loss": 0.8021256446838378,
      "step": 620
    },
    {
      "epoch": 0.08110195674562307,
      "grad_norm": 5.097868976449567,
      "learning_rate": 2e-05,
      "loss": 0.7970829963684082,
      "step": 630
    },
    {
      "epoch": 0.082389289392379,
      "grad_norm": 2.4901894680955787,
      "learning_rate": 2e-05,
      "loss": 0.7695016384124755,
      "step": 640
    },
    {
      "epoch": 0.08367662203913491,
      "grad_norm": 6.436792505064742,
      "learning_rate": 2e-05,
      "loss": 0.7511703491210937,
      "step": 650
    },
    {
      "epoch": 0.08496395468589084,
      "grad_norm": 5.76943441019596,
      "learning_rate": 2e-05,
      "loss": 0.7517711639404296,
      "step": 660
    },
    {
      "epoch": 0.08625128733264675,
      "grad_norm": 1.8786628503818215,
      "learning_rate": 2e-05,
      "loss": 0.7410395622253418,
      "step": 670
    },
    {
      "epoch": 0.08753861997940268,
      "grad_norm": 4.587266282984627,
      "learning_rate": 2e-05,
      "loss": 0.771084976196289,
      "step": 680
    },
    {
      "epoch": 0.0888259526261586,
      "grad_norm": 13.971035209507322,
      "learning_rate": 2e-05,
      "loss": 0.7819516658782959,
      "step": 690
    },
    {
      "epoch": 0.09011328527291453,
      "grad_norm": 8.1570226178038,
      "learning_rate": 2e-05,
      "loss": 0.7403954982757568,
      "step": 700
    },
    {
      "epoch": 0.09140061791967044,
      "grad_norm": 4.688221044744754,
      "learning_rate": 2e-05,
      "loss": 0.786048412322998,
      "step": 710
    },
    {
      "epoch": 0.09268795056642637,
      "grad_norm": 2.778648354062325,
      "learning_rate": 2e-05,
      "loss": 0.7478150367736817,
      "step": 720
    },
    {
      "epoch": 0.09397528321318228,
      "grad_norm": 3.652349932924877,
      "learning_rate": 2e-05,
      "loss": 0.7311888694763183,
      "step": 730
    },
    {
      "epoch": 0.09526261585993821,
      "grad_norm": 5.5932577832394,
      "learning_rate": 2e-05,
      "loss": 0.7629296779632568,
      "step": 740
    },
    {
      "epoch": 0.09654994850669413,
      "grad_norm": 1.535821909119182,
      "learning_rate": 2e-05,
      "loss": 0.7407299041748047,
      "step": 750
    },
    {
      "epoch": 0.09783728115345006,
      "grad_norm": 1.45905132393004,
      "learning_rate": 2e-05,
      "loss": 0.7430490016937256,
      "step": 760
    },
    {
      "epoch": 0.09912461380020597,
      "grad_norm": 4.619712730990858,
      "learning_rate": 2e-05,
      "loss": 0.7504862785339356,
      "step": 770
    },
    {
      "epoch": 0.1004119464469619,
      "grad_norm": 3.9224749754444566,
      "learning_rate": 2e-05,
      "loss": 0.7686273574829101,
      "step": 780
    },
    {
      "epoch": 0.10169927909371781,
      "grad_norm": 6.098937347148086,
      "learning_rate": 2e-05,
      "loss": 0.7319077491760254,
      "step": 790
    },
    {
      "epoch": 0.10298661174047374,
      "grad_norm": 2.4496087544978336,
      "learning_rate": 2e-05,
      "loss": 0.7697375297546387,
      "step": 800
    },
    {
      "epoch": 0.10427394438722966,
      "grad_norm": 2.8551662537697102,
      "learning_rate": 2e-05,
      "loss": 0.742626953125,
      "step": 810
    },
    {
      "epoch": 0.10556127703398559,
      "grad_norm": 3.205355126068373,
      "learning_rate": 2e-05,
      "loss": 0.7027194023132324,
      "step": 820
    },
    {
      "epoch": 0.1068486096807415,
      "grad_norm": 2.253436553246574,
      "learning_rate": 2e-05,
      "loss": 0.7767892360687256,
      "step": 830
    },
    {
      "epoch": 0.10813594232749743,
      "grad_norm": 4.087718084462775,
      "learning_rate": 2e-05,
      "loss": 0.7374720573425293,
      "step": 840
    },
    {
      "epoch": 0.10942327497425335,
      "grad_norm": 4.050374141792059,
      "learning_rate": 2e-05,
      "loss": 0.7483660697937011,
      "step": 850
    },
    {
      "epoch": 0.11071060762100927,
      "grad_norm": 2.239445345441048,
      "learning_rate": 2e-05,
      "loss": 0.7472774505615234,
      "step": 860
    },
    {
      "epoch": 0.11199794026776519,
      "grad_norm": 3.9463791962767836,
      "learning_rate": 2e-05,
      "loss": 0.7285970687866211,
      "step": 870
    },
    {
      "epoch": 0.11328527291452112,
      "grad_norm": 12.421260769301227,
      "learning_rate": 2e-05,
      "loss": 0.7539933681488037,
      "step": 880
    },
    {
      "epoch": 0.11457260556127703,
      "grad_norm": 4.885002632304067,
      "learning_rate": 2e-05,
      "loss": 0.7439227104187012,
      "step": 890
    },
    {
      "epoch": 0.11585993820803296,
      "grad_norm": 1.5566251726058724,
      "learning_rate": 2e-05,
      "loss": 0.735724401473999,
      "step": 900
    },
    {
      "epoch": 0.11714727085478888,
      "grad_norm": 2.865049116950206,
      "learning_rate": 2e-05,
      "loss": 0.7331388473510743,
      "step": 910
    },
    {
      "epoch": 0.1184346035015448,
      "grad_norm": 2.960680025346854,
      "learning_rate": 2e-05,
      "loss": 0.74105224609375,
      "step": 920
    },
    {
      "epoch": 0.11972193614830072,
      "grad_norm": 2.0043561901231413,
      "learning_rate": 2e-05,
      "loss": 0.6742991447448731,
      "step": 930
    },
    {
      "epoch": 0.12100926879505665,
      "grad_norm": 1.6534160074854731,
      "learning_rate": 2e-05,
      "loss": 0.7274767875671386,
      "step": 940
    },
    {
      "epoch": 0.12229660144181256,
      "grad_norm": 3.0327709147346353,
      "learning_rate": 2e-05,
      "loss": 0.7146361351013184,
      "step": 950
    },
    {
      "epoch": 0.12358393408856849,
      "grad_norm": 3.9665339761676237,
      "learning_rate": 2e-05,
      "loss": 0.7229518413543701,
      "step": 960
    },
    {
      "epoch": 0.12487126673532441,
      "grad_norm": 3.8479184119373193,
      "learning_rate": 2e-05,
      "loss": 0.7719802856445312,
      "step": 970
    },
    {
      "epoch": 0.12615859938208032,
      "grad_norm": 1.2184604438873134,
      "learning_rate": 2e-05,
      "loss": 0.7337092876434326,
      "step": 980
    },
    {
      "epoch": 0.12744593202883625,
      "grad_norm": 5.350440266680789,
      "learning_rate": 2e-05,
      "loss": 0.7010088920593261,
      "step": 990
    },
    {
      "epoch": 0.12873326467559218,
      "grad_norm": 2.7811359489299248,
      "learning_rate": 2e-05,
      "loss": 0.7292609691619873,
      "step": 1000
    },
    {
      "epoch": 0.1300205973223481,
      "grad_norm": 17.912803780366907,
      "learning_rate": 2e-05,
      "loss": 0.7163818359375,
      "step": 1010
    },
    {
      "epoch": 0.131307929969104,
      "grad_norm": 2.11582488712836,
      "learning_rate": 2e-05,
      "loss": 0.7069422721862793,
      "step": 1020
    },
    {
      "epoch": 0.13259526261585994,
      "grad_norm": 11.247133917028588,
      "learning_rate": 2e-05,
      "loss": 0.7331721782684326,
      "step": 1030
    },
    {
      "epoch": 0.13388259526261587,
      "grad_norm": 2.7388313221050575,
      "learning_rate": 2e-05,
      "loss": 0.6841031551361084,
      "step": 1040
    },
    {
      "epoch": 0.1351699279093718,
      "grad_norm": 1.207061425581625,
      "learning_rate": 2e-05,
      "loss": 0.733606481552124,
      "step": 1050
    },
    {
      "epoch": 0.1364572605561277,
      "grad_norm": 47.69246880805316,
      "learning_rate": 2e-05,
      "loss": 0.7302389144897461,
      "step": 1060
    },
    {
      "epoch": 0.13774459320288363,
      "grad_norm": 2.171556278425703,
      "learning_rate": 2e-05,
      "loss": 0.7309787750244141,
      "step": 1070
    },
    {
      "epoch": 0.13903192584963955,
      "grad_norm": 3.360889748897153,
      "learning_rate": 2e-05,
      "loss": 0.7237205028533935,
      "step": 1080
    },
    {
      "epoch": 0.14031925849639545,
      "grad_norm": 1.5969376240757103,
      "learning_rate": 2e-05,
      "loss": 0.7478039741516114,
      "step": 1090
    },
    {
      "epoch": 0.14160659114315138,
      "grad_norm": 1.9534203833516763,
      "learning_rate": 2e-05,
      "loss": 0.7363515377044678,
      "step": 1100
    },
    {
      "epoch": 0.1428939237899073,
      "grad_norm": 2.6654514628849952,
      "learning_rate": 2e-05,
      "loss": 0.7492703437805176,
      "step": 1110
    },
    {
      "epoch": 0.14418125643666324,
      "grad_norm": 1.68379520116963,
      "learning_rate": 2e-05,
      "loss": 0.7549358367919922,
      "step": 1120
    },
    {
      "epoch": 0.14546858908341914,
      "grad_norm": 1.632554767797848,
      "learning_rate": 2e-05,
      "loss": 0.6897923946380615,
      "step": 1130
    },
    {
      "epoch": 0.14675592173017507,
      "grad_norm": 1.653687869959576,
      "learning_rate": 2e-05,
      "loss": 0.7304753303527832,
      "step": 1140
    },
    {
      "epoch": 0.148043254376931,
      "grad_norm": 1.6347977972264258,
      "learning_rate": 2e-05,
      "loss": 0.7604644775390625,
      "step": 1150
    },
    {
      "epoch": 0.14933058702368693,
      "grad_norm": 1.8718744344050506,
      "learning_rate": 2e-05,
      "loss": 0.697459602355957,
      "step": 1160
    },
    {
      "epoch": 0.15061791967044283,
      "grad_norm": 2.532644944339293,
      "learning_rate": 2e-05,
      "loss": 0.7188455104827881,
      "step": 1170
    },
    {
      "epoch": 0.15190525231719876,
      "grad_norm": 1.806201422331298,
      "learning_rate": 2e-05,
      "loss": 0.7081408977508545,
      "step": 1180
    },
    {
      "epoch": 0.1531925849639547,
      "grad_norm": 1.6489310799297887,
      "learning_rate": 2e-05,
      "loss": 0.7228747367858886,
      "step": 1190
    },
    {
      "epoch": 0.15447991761071062,
      "grad_norm": 2.9974385822785408,
      "learning_rate": 2e-05,
      "loss": 0.7244661331176758,
      "step": 1200
    },
    {
      "epoch": 0.15576725025746652,
      "grad_norm": 22.594724424794563,
      "learning_rate": 2e-05,
      "loss": 0.7405924797058105,
      "step": 1210
    },
    {
      "epoch": 0.15705458290422245,
      "grad_norm": 3.7604685572670684,
      "learning_rate": 2e-05,
      "loss": 0.7070909023284913,
      "step": 1220
    },
    {
      "epoch": 0.15834191555097837,
      "grad_norm": 3.3763336904748615,
      "learning_rate": 2e-05,
      "loss": 0.7894343376159668,
      "step": 1230
    },
    {
      "epoch": 0.1596292481977343,
      "grad_norm": 10.941310418547834,
      "learning_rate": 2e-05,
      "loss": 0.7505155086517334,
      "step": 1240
    },
    {
      "epoch": 0.1609165808444902,
      "grad_norm": 4.887708688260815,
      "learning_rate": 2e-05,
      "loss": 0.7505541801452636,
      "step": 1250
    },
    {
      "epoch": 0.16220391349124613,
      "grad_norm": 6.54025957968161,
      "learning_rate": 2e-05,
      "loss": 0.716888427734375,
      "step": 1260
    },
    {
      "epoch": 0.16349124613800206,
      "grad_norm": 3.6877006120763585,
      "learning_rate": 2e-05,
      "loss": 0.7247855186462402,
      "step": 1270
    },
    {
      "epoch": 0.164778578784758,
      "grad_norm": 15.044650338028674,
      "learning_rate": 2e-05,
      "loss": 0.7331586360931397,
      "step": 1280
    },
    {
      "epoch": 0.1660659114315139,
      "grad_norm": 2.538457484617354,
      "learning_rate": 2e-05,
      "loss": 0.7180121421813965,
      "step": 1290
    },
    {
      "epoch": 0.16735324407826982,
      "grad_norm": 1.6984019588574841,
      "learning_rate": 2e-05,
      "loss": 0.7547801494598388,
      "step": 1300
    },
    {
      "epoch": 0.16864057672502575,
      "grad_norm": 4.940123866109371,
      "learning_rate": 2e-05,
      "loss": 0.7324605464935303,
      "step": 1310
    },
    {
      "epoch": 0.16992790937178168,
      "grad_norm": 1.4925558541074413,
      "learning_rate": 2e-05,
      "loss": 0.7033469676971436,
      "step": 1320
    },
    {
      "epoch": 0.17121524201853758,
      "grad_norm": 2.0954880927362307,
      "learning_rate": 2e-05,
      "loss": 0.7216341972351075,
      "step": 1330
    },
    {
      "epoch": 0.1725025746652935,
      "grad_norm": 3.965453670040223,
      "learning_rate": 2e-05,
      "loss": 0.7113677501678467,
      "step": 1340
    },
    {
      "epoch": 0.17378990731204944,
      "grad_norm": 2.62486775378727,
      "learning_rate": 2e-05,
      "loss": 0.6789872646331787,
      "step": 1350
    },
    {
      "epoch": 0.17507723995880536,
      "grad_norm": 1.6392487264123656,
      "learning_rate": 2e-05,
      "loss": 0.6960799217224121,
      "step": 1360
    },
    {
      "epoch": 0.17636457260556127,
      "grad_norm": 1.6246985763924315,
      "learning_rate": 2e-05,
      "loss": 0.7166689872741699,
      "step": 1370
    },
    {
      "epoch": 0.1776519052523172,
      "grad_norm": 3.586582883055029,
      "learning_rate": 2e-05,
      "loss": 0.7102786064147949,
      "step": 1380
    },
    {
      "epoch": 0.17893923789907312,
      "grad_norm": 7.353760235895877,
      "learning_rate": 2e-05,
      "loss": 0.7114831447601319,
      "step": 1390
    },
    {
      "epoch": 0.18022657054582905,
      "grad_norm": 3.600918049571546,
      "learning_rate": 2e-05,
      "loss": 0.722894287109375,
      "step": 1400
    },
    {
      "epoch": 0.18151390319258495,
      "grad_norm": 6.977869760230939,
      "learning_rate": 2e-05,
      "loss": 0.7068233013153076,
      "step": 1410
    },
    {
      "epoch": 0.18280123583934088,
      "grad_norm": 2.6291241838271486,
      "learning_rate": 2e-05,
      "loss": 0.7131928920745849,
      "step": 1420
    },
    {
      "epoch": 0.1840885684860968,
      "grad_norm": 2.47189101333238,
      "learning_rate": 2e-05,
      "loss": 0.7333631038665771,
      "step": 1430
    },
    {
      "epoch": 0.18537590113285274,
      "grad_norm": 2.5920454659070438,
      "learning_rate": 2e-05,
      "loss": 0.6987507820129395,
      "step": 1440
    },
    {
      "epoch": 0.18666323377960864,
      "grad_norm": 1.840628594046838,
      "learning_rate": 2e-05,
      "loss": 0.7048477172851563,
      "step": 1450
    },
    {
      "epoch": 0.18795056642636457,
      "grad_norm": 2.3416098111869093,
      "learning_rate": 2e-05,
      "loss": 0.7118321895599365,
      "step": 1460
    },
    {
      "epoch": 0.1892378990731205,
      "grad_norm": 1.530570157282232,
      "learning_rate": 2e-05,
      "loss": 0.7082484245300293,
      "step": 1470
    },
    {
      "epoch": 0.19052523171987643,
      "grad_norm": 2.2055714727842477,
      "learning_rate": 2e-05,
      "loss": 0.7013579368591308,
      "step": 1480
    },
    {
      "epoch": 0.19181256436663233,
      "grad_norm": 1.2321264302309138,
      "learning_rate": 2e-05,
      "loss": 0.720965576171875,
      "step": 1490
    },
    {
      "epoch": 0.19309989701338826,
      "grad_norm": 2.7847826433903067,
      "learning_rate": 2e-05,
      "loss": 0.7270987510681153,
      "step": 1500
    },
    {
      "epoch": 0.19438722966014418,
      "grad_norm": 1.8686276102701087,
      "learning_rate": 2e-05,
      "loss": 0.7201802730560303,
      "step": 1510
    },
    {
      "epoch": 0.1956745623069001,
      "grad_norm": 1.569912150907071,
      "learning_rate": 2e-05,
      "loss": 0.7226099967956543,
      "step": 1520
    },
    {
      "epoch": 0.196961894953656,
      "grad_norm": 2.283386362656244,
      "learning_rate": 2e-05,
      "loss": 0.7405374050140381,
      "step": 1530
    },
    {
      "epoch": 0.19824922760041194,
      "grad_norm": 1.5156912593840797,
      "learning_rate": 2e-05,
      "loss": 0.7121532440185547,
      "step": 1540
    },
    {
      "epoch": 0.19953656024716787,
      "grad_norm": 1.4580219323036991,
      "learning_rate": 2e-05,
      "loss": 0.6971864223480224,
      "step": 1550
    },
    {
      "epoch": 0.2008238928939238,
      "grad_norm": 5.82821608514732,
      "learning_rate": 2e-05,
      "loss": 0.7024627685546875,
      "step": 1560
    },
    {
      "epoch": 0.2021112255406797,
      "grad_norm": 1.1362087794702644,
      "learning_rate": 2e-05,
      "loss": 0.6774281978607177,
      "step": 1570
    },
    {
      "epoch": 0.20339855818743563,
      "grad_norm": 2.443427160363054,
      "learning_rate": 2e-05,
      "loss": 0.6777324676513672,
      "step": 1580
    },
    {
      "epoch": 0.20468589083419156,
      "grad_norm": 3.4197112552113254,
      "learning_rate": 2e-05,
      "loss": 0.6910956859588623,
      "step": 1590
    },
    {
      "epoch": 0.2059732234809475,
      "grad_norm": 2.7142923935215904,
      "learning_rate": 2e-05,
      "loss": 0.7175305366516114,
      "step": 1600
    },
    {
      "epoch": 0.2072605561277034,
      "grad_norm": 2.9446280303316854,
      "learning_rate": 2e-05,
      "loss": 0.6894670009613038,
      "step": 1610
    },
    {
      "epoch": 0.20854788877445932,
      "grad_norm": 1.369637715767581,
      "learning_rate": 2e-05,
      "loss": 0.6922161102294921,
      "step": 1620
    },
    {
      "epoch": 0.20983522142121525,
      "grad_norm": 1.3879222035136711,
      "learning_rate": 2e-05,
      "loss": 0.6919208526611328,
      "step": 1630
    },
    {
      "epoch": 0.21112255406797117,
      "grad_norm": 3.065083029358514,
      "learning_rate": 2e-05,
      "loss": 0.6886746883392334,
      "step": 1640
    },
    {
      "epoch": 0.21240988671472708,
      "grad_norm": 10.325097544156115,
      "learning_rate": 2e-05,
      "loss": 0.6938620567321777,
      "step": 1650
    },
    {
      "epoch": 0.213697219361483,
      "grad_norm": 1.6072671865699064,
      "learning_rate": 2e-05,
      "loss": 0.7277378082275391,
      "step": 1660
    },
    {
      "epoch": 0.21498455200823893,
      "grad_norm": 1.3380594408338742,
      "learning_rate": 2e-05,
      "loss": 0.7332149505615234,
      "step": 1670
    },
    {
      "epoch": 0.21627188465499486,
      "grad_norm": 1.5454219317409892,
      "learning_rate": 2e-05,
      "loss": 0.7071063995361329,
      "step": 1680
    },
    {
      "epoch": 0.21755921730175076,
      "grad_norm": 1.8153666337523178,
      "learning_rate": 2e-05,
      "loss": 0.7288019180297851,
      "step": 1690
    },
    {
      "epoch": 0.2188465499485067,
      "grad_norm": 1.877709344780682,
      "learning_rate": 2e-05,
      "loss": 0.7265975475311279,
      "step": 1700
    },
    {
      "epoch": 0.22013388259526262,
      "grad_norm": 4.043860838359305,
      "learning_rate": 2e-05,
      "loss": 0.7083337783813477,
      "step": 1710
    },
    {
      "epoch": 0.22142121524201855,
      "grad_norm": 4.246037776394193,
      "learning_rate": 2e-05,
      "loss": 0.678376293182373,
      "step": 1720
    },
    {
      "epoch": 0.22270854788877445,
      "grad_norm": 1.4558126421583217,
      "learning_rate": 2e-05,
      "loss": 0.7107050895690918,
      "step": 1730
    },
    {
      "epoch": 0.22399588053553038,
      "grad_norm": 2.1461600575963056,
      "learning_rate": 2e-05,
      "loss": 0.6994074821472168,
      "step": 1740
    },
    {
      "epoch": 0.2252832131822863,
      "grad_norm": 2.138539490137605,
      "learning_rate": 2e-05,
      "loss": 0.6766193389892579,
      "step": 1750
    },
    {
      "epoch": 0.22657054582904224,
      "grad_norm": 1.230257735080685,
      "learning_rate": 2e-05,
      "loss": 0.6743141174316406,
      "step": 1760
    },
    {
      "epoch": 0.22785787847579814,
      "grad_norm": 5.529787951162952,
      "learning_rate": 2e-05,
      "loss": 0.6881375789642334,
      "step": 1770
    },
    {
      "epoch": 0.22914521112255407,
      "grad_norm": 1.6357599532681988,
      "learning_rate": 2e-05,
      "loss": 0.6633195877075195,
      "step": 1780
    },
    {
      "epoch": 0.23043254376931,
      "grad_norm": 6.229625650998896,
      "learning_rate": 2e-05,
      "loss": 0.6983210563659668,
      "step": 1790
    },
    {
      "epoch": 0.23171987641606592,
      "grad_norm": 1.3301077031080362,
      "learning_rate": 2e-05,
      "loss": 0.7201817512512207,
      "step": 1800
    },
    {
      "epoch": 0.23300720906282182,
      "grad_norm": 3.39358729212885,
      "learning_rate": 2e-05,
      "loss": 0.6951064109802246,
      "step": 1810
    },
    {
      "epoch": 0.23429454170957775,
      "grad_norm": 4.181323376087317,
      "learning_rate": 2e-05,
      "loss": 0.6929269790649414,
      "step": 1820
    },
    {
      "epoch": 0.23558187435633368,
      "grad_norm": 1.5545245966089207,
      "learning_rate": 2e-05,
      "loss": 0.6981029033660888,
      "step": 1830
    },
    {
      "epoch": 0.2368692070030896,
      "grad_norm": 2.1015275437163385,
      "learning_rate": 2e-05,
      "loss": 0.6695489883422852,
      "step": 1840
    },
    {
      "epoch": 0.2381565396498455,
      "grad_norm": 1.8856355185752995,
      "learning_rate": 2e-05,
      "loss": 0.7271100521087647,
      "step": 1850
    },
    {
      "epoch": 0.23944387229660144,
      "grad_norm": 3.425995350551422,
      "learning_rate": 2e-05,
      "loss": 0.6916310787200928,
      "step": 1860
    },
    {
      "epoch": 0.24073120494335737,
      "grad_norm": 1.2448080896450548,
      "learning_rate": 2e-05,
      "loss": 0.705747127532959,
      "step": 1870
    },
    {
      "epoch": 0.2420185375901133,
      "grad_norm": 6.4595134842908495,
      "learning_rate": 2e-05,
      "loss": 0.6798825740814209,
      "step": 1880
    },
    {
      "epoch": 0.2433058702368692,
      "grad_norm": 1.2226636920681981,
      "learning_rate": 2e-05,
      "loss": 0.7015285015106201,
      "step": 1890
    },
    {
      "epoch": 0.24459320288362513,
      "grad_norm": 2.3842419146911635,
      "learning_rate": 2e-05,
      "loss": 0.7172482490539551,
      "step": 1900
    },
    {
      "epoch": 0.24588053553038106,
      "grad_norm": 1.7333858862906162,
      "learning_rate": 2e-05,
      "loss": 0.6919423580169678,
      "step": 1910
    },
    {
      "epoch": 0.24716786817713698,
      "grad_norm": 1.2738406384985739,
      "learning_rate": 2e-05,
      "loss": 0.7124980926513672,
      "step": 1920
    },
    {
      "epoch": 0.24845520082389289,
      "grad_norm": 3.089273068519385,
      "learning_rate": 2e-05,
      "loss": 0.7129861354827881,
      "step": 1930
    },
    {
      "epoch": 0.24974253347064881,
      "grad_norm": 5.4254882757315395,
      "learning_rate": 2e-05,
      "loss": 0.7517825126647949,
      "step": 1940
    },
    {
      "epoch": 0.25102986611740474,
      "grad_norm": 1.71140671632292,
      "learning_rate": 2e-05,
      "loss": 0.7149953365325927,
      "step": 1950
    },
    {
      "epoch": 0.25231719876416064,
      "grad_norm": 2.7193201670484886,
      "learning_rate": 2e-05,
      "loss": 0.7189720153808594,
      "step": 1960
    },
    {
      "epoch": 0.2536045314109166,
      "grad_norm": 10.070508073263046,
      "learning_rate": 2e-05,
      "loss": 0.6905816555023193,
      "step": 1970
    },
    {
      "epoch": 0.2548918640576725,
      "grad_norm": 8.024891526041356,
      "learning_rate": 2e-05,
      "loss": 0.6899857521057129,
      "step": 1980
    },
    {
      "epoch": 0.2561791967044284,
      "grad_norm": 14.975571956672278,
      "learning_rate": 2e-05,
      "loss": 0.7188075065612793,
      "step": 1990
    },
    {
      "epoch": 0.25746652935118436,
      "grad_norm": 2.4641111805113103,
      "learning_rate": 2e-05,
      "loss": 0.7120208263397216,
      "step": 2000
    },
    {
      "epoch": 0.25875386199794026,
      "grad_norm": 2.7546066287502833,
      "learning_rate": 2e-05,
      "loss": 0.6824542999267578,
      "step": 2010
    },
    {
      "epoch": 0.2600411946446962,
      "grad_norm": 2.180296993017307,
      "learning_rate": 2e-05,
      "loss": 0.7345854759216308,
      "step": 2020
    },
    {
      "epoch": 0.2613285272914521,
      "grad_norm": 3.1963581482749137,
      "learning_rate": 2e-05,
      "loss": 0.689295482635498,
      "step": 2030
    },
    {
      "epoch": 0.262615859938208,
      "grad_norm": 3.340370740500395,
      "learning_rate": 2e-05,
      "loss": 0.6768050193786621,
      "step": 2040
    },
    {
      "epoch": 0.263903192584964,
      "grad_norm": 2.7181711140892375,
      "learning_rate": 2e-05,
      "loss": 0.755255126953125,
      "step": 2050
    },
    {
      "epoch": 0.2651905252317199,
      "grad_norm": 5.156287372699048,
      "learning_rate": 2e-05,
      "loss": 0.6873680114746094,
      "step": 2060
    },
    {
      "epoch": 0.2664778578784758,
      "grad_norm": 0.9527725260292781,
      "learning_rate": 2e-05,
      "loss": 0.7040456771850586,
      "step": 2070
    },
    {
      "epoch": 0.26776519052523173,
      "grad_norm": 1.1259153607047196,
      "learning_rate": 2e-05,
      "loss": 0.6887578010559082,
      "step": 2080
    },
    {
      "epoch": 0.26905252317198763,
      "grad_norm": 1.4640224316338672,
      "learning_rate": 2e-05,
      "loss": 0.7209594249725342,
      "step": 2090
    },
    {
      "epoch": 0.2703398558187436,
      "grad_norm": 1.6949802858236316,
      "learning_rate": 2e-05,
      "loss": 0.7076781272888184,
      "step": 2100
    },
    {
      "epoch": 0.2716271884654995,
      "grad_norm": 1.9455095056367833,
      "learning_rate": 2e-05,
      "loss": 0.7013013362884521,
      "step": 2110
    },
    {
      "epoch": 0.2729145211122554,
      "grad_norm": 1.885815138529574,
      "learning_rate": 2e-05,
      "loss": 0.70594482421875,
      "step": 2120
    },
    {
      "epoch": 0.27420185375901135,
      "grad_norm": 2.8614315685772205,
      "learning_rate": 2e-05,
      "loss": 0.7034307003021241,
      "step": 2130
    },
    {
      "epoch": 0.27548918640576725,
      "grad_norm": 12.720304202743357,
      "learning_rate": 2e-05,
      "loss": 0.7098933219909668,
      "step": 2140
    },
    {
      "epoch": 0.27677651905252315,
      "grad_norm": 2.1655512205055225,
      "learning_rate": 2e-05,
      "loss": 0.7170520305633545,
      "step": 2150
    },
    {
      "epoch": 0.2780638516992791,
      "grad_norm": 8.691373263950895,
      "learning_rate": 2e-05,
      "loss": 0.7013160228729248,
      "step": 2160
    },
    {
      "epoch": 0.279351184346035,
      "grad_norm": 1.608771090637953,
      "learning_rate": 2e-05,
      "loss": 0.7061017036437989,
      "step": 2170
    },
    {
      "epoch": 0.2806385169927909,
      "grad_norm": 11.833877668015374,
      "learning_rate": 2e-05,
      "loss": 0.6650290489196777,
      "step": 2180
    },
    {
      "epoch": 0.28192584963954687,
      "grad_norm": 1.4006501457101668,
      "learning_rate": 2e-05,
      "loss": 0.7379101753234864,
      "step": 2190
    },
    {
      "epoch": 0.28321318228630277,
      "grad_norm": 1.6610926197943376,
      "learning_rate": 2e-05,
      "loss": 0.6706048011779785,
      "step": 2200
    },
    {
      "epoch": 0.2845005149330587,
      "grad_norm": 1.487648001601594,
      "learning_rate": 2e-05,
      "loss": 0.6972060680389405,
      "step": 2210
    },
    {
      "epoch": 0.2857878475798146,
      "grad_norm": 7.7682006345537715,
      "learning_rate": 2e-05,
      "loss": 0.6770853996276855,
      "step": 2220
    },
    {
      "epoch": 0.2870751802265705,
      "grad_norm": 2.0707865262416623,
      "learning_rate": 2e-05,
      "loss": 0.6925491333007813,
      "step": 2230
    },
    {
      "epoch": 0.2883625128733265,
      "grad_norm": 1.7595244490954383,
      "learning_rate": 2e-05,
      "loss": 0.6970232009887696,
      "step": 2240
    },
    {
      "epoch": 0.2896498455200824,
      "grad_norm": 1.092700120083694,
      "learning_rate": 2e-05,
      "loss": 0.6953269958496093,
      "step": 2250
    },
    {
      "epoch": 0.2909371781668383,
      "grad_norm": 9.748987611796897,
      "learning_rate": 2e-05,
      "loss": 0.6644187927246094,
      "step": 2260
    },
    {
      "epoch": 0.29222451081359424,
      "grad_norm": 2.492722399232875,
      "learning_rate": 2e-05,
      "loss": 0.6756973266601562,
      "step": 2270
    },
    {
      "epoch": 0.29351184346035014,
      "grad_norm": 1.568787070446361,
      "learning_rate": 2e-05,
      "loss": 0.7013367652893067,
      "step": 2280
    },
    {
      "epoch": 0.2947991761071061,
      "grad_norm": 2.47976302361641,
      "learning_rate": 2e-05,
      "loss": 0.6758729457855225,
      "step": 2290
    },
    {
      "epoch": 0.296086508753862,
      "grad_norm": 1.8280130594138808,
      "learning_rate": 2e-05,
      "loss": 0.718410587310791,
      "step": 2300
    },
    {
      "epoch": 0.2973738414006179,
      "grad_norm": 1.330373350200014,
      "learning_rate": 2e-05,
      "loss": 0.6956700325012207,
      "step": 2310
    },
    {
      "epoch": 0.29866117404737386,
      "grad_norm": 2.1289014923647347,
      "learning_rate": 2e-05,
      "loss": 0.6391764640808105,
      "step": 2320
    },
    {
      "epoch": 0.29994850669412976,
      "grad_norm": 1.4812494641456213,
      "learning_rate": 2e-05,
      "loss": 0.6836237907409668,
      "step": 2330
    },
    {
      "epoch": 0.30123583934088566,
      "grad_norm": 2.182997522790617,
      "learning_rate": 2e-05,
      "loss": 0.6847892284393311,
      "step": 2340
    },
    {
      "epoch": 0.3025231719876416,
      "grad_norm": 1.4120255579731853,
      "learning_rate": 2e-05,
      "loss": 0.6991443634033203,
      "step": 2350
    },
    {
      "epoch": 0.3038105046343975,
      "grad_norm": 1.7715960231377,
      "learning_rate": 2e-05,
      "loss": 0.6520809173583985,
      "step": 2360
    },
    {
      "epoch": 0.30509783728115347,
      "grad_norm": 1.344504528912978,
      "learning_rate": 2e-05,
      "loss": 0.6915669441223145,
      "step": 2370
    },
    {
      "epoch": 0.3063851699279094,
      "grad_norm": 1.418372939618548,
      "learning_rate": 2e-05,
      "loss": 0.6714213371276856,
      "step": 2380
    },
    {
      "epoch": 0.3076725025746653,
      "grad_norm": 2.3106904123208403,
      "learning_rate": 2e-05,
      "loss": 0.6664355278015137,
      "step": 2390
    },
    {
      "epoch": 0.30895983522142123,
      "grad_norm": 1.3253046019523815,
      "learning_rate": 2e-05,
      "loss": 0.6971554756164551,
      "step": 2400
    },
    {
      "epoch": 0.31024716786817713,
      "grad_norm": 1.2161954394286878,
      "learning_rate": 2e-05,
      "loss": 0.644910192489624,
      "step": 2410
    },
    {
      "epoch": 0.31153450051493303,
      "grad_norm": 1.5284008850353576,
      "learning_rate": 2e-05,
      "loss": 0.671488904953003,
      "step": 2420
    },
    {
      "epoch": 0.312821833161689,
      "grad_norm": 2.3280051785201517,
      "learning_rate": 2e-05,
      "loss": 0.6644371032714844,
      "step": 2430
    },
    {
      "epoch": 0.3141091658084449,
      "grad_norm": 2.4846074890678547,
      "learning_rate": 2e-05,
      "loss": 0.6947656631469726,
      "step": 2440
    },
    {
      "epoch": 0.31539649845520085,
      "grad_norm": 2.342854536473009,
      "learning_rate": 2e-05,
      "loss": 0.6945387840270996,
      "step": 2450
    },
    {
      "epoch": 0.31668383110195675,
      "grad_norm": 1.3707661574249077,
      "learning_rate": 2e-05,
      "loss": 0.7021138668060303,
      "step": 2460
    },
    {
      "epoch": 0.31797116374871265,
      "grad_norm": 1.011087502453139,
      "learning_rate": 2e-05,
      "loss": 0.711193037033081,
      "step": 2470
    },
    {
      "epoch": 0.3192584963954686,
      "grad_norm": 1.3124044269321697,
      "learning_rate": 2e-05,
      "loss": 0.7033077239990234,
      "step": 2480
    },
    {
      "epoch": 0.3205458290422245,
      "grad_norm": 0.8810548907751976,
      "learning_rate": 2e-05,
      "loss": 0.6761528015136719,
      "step": 2490
    },
    {
      "epoch": 0.3218331616889804,
      "grad_norm": 1.244342002235635,
      "learning_rate": 2e-05,
      "loss": 0.6616430282592773,
      "step": 2500
    },
    {
      "epoch": 0.32312049433573636,
      "grad_norm": 1.2138059133962842,
      "learning_rate": 2e-05,
      "loss": 0.6823375701904297,
      "step": 2510
    },
    {
      "epoch": 0.32440782698249226,
      "grad_norm": 1.5411359954112853,
      "learning_rate": 2e-05,
      "loss": 0.6461188316345214,
      "step": 2520
    },
    {
      "epoch": 0.3256951596292482,
      "grad_norm": 4.067446254024166,
      "learning_rate": 2e-05,
      "loss": 0.6632846832275391,
      "step": 2530
    },
    {
      "epoch": 0.3269824922760041,
      "grad_norm": 0.9930530111354352,
      "learning_rate": 2e-05,
      "loss": 0.6521314144134521,
      "step": 2540
    },
    {
      "epoch": 0.32826982492276,
      "grad_norm": 1.2183079247100232,
      "learning_rate": 2e-05,
      "loss": 0.6647382259368897,
      "step": 2550
    },
    {
      "epoch": 0.329557157569516,
      "grad_norm": 1.3056406095156692,
      "learning_rate": 2e-05,
      "loss": 0.7056656837463379,
      "step": 2560
    },
    {
      "epoch": 0.3308444902162719,
      "grad_norm": 1.1446095028508332,
      "learning_rate": 2e-05,
      "loss": 0.6724899291992188,
      "step": 2570
    },
    {
      "epoch": 0.3321318228630278,
      "grad_norm": 2.034841488345913,
      "learning_rate": 2e-05,
      "loss": 0.7128400325775146,
      "step": 2580
    },
    {
      "epoch": 0.33341915550978374,
      "grad_norm": 1.1398931102101963,
      "learning_rate": 2e-05,
      "loss": 0.694240665435791,
      "step": 2590
    },
    {
      "epoch": 0.33470648815653964,
      "grad_norm": 1.1474928486855198,
      "learning_rate": 2e-05,
      "loss": 0.6734938621520996,
      "step": 2600
    },
    {
      "epoch": 0.3359938208032956,
      "grad_norm": 1.3857880198125372,
      "learning_rate": 2e-05,
      "loss": 0.6751894950866699,
      "step": 2610
    },
    {
      "epoch": 0.3372811534500515,
      "grad_norm": 1.785399986240084,
      "learning_rate": 2e-05,
      "loss": 0.6587059020996093,
      "step": 2620
    },
    {
      "epoch": 0.3385684860968074,
      "grad_norm": 1.7238102251604501,
      "learning_rate": 2e-05,
      "loss": 0.7021533966064453,
      "step": 2630
    },
    {
      "epoch": 0.33985581874356335,
      "grad_norm": 1.065710140961934,
      "learning_rate": 2e-05,
      "loss": 0.690271520614624,
      "step": 2640
    },
    {
      "epoch": 0.34114315139031925,
      "grad_norm": 4.21226441159709,
      "learning_rate": 2e-05,
      "loss": 0.6952508926391602,
      "step": 2650
    },
    {
      "epoch": 0.34243048403707516,
      "grad_norm": 1.146287127572526,
      "learning_rate": 2e-05,
      "loss": 0.6579763412475585,
      "step": 2660
    },
    {
      "epoch": 0.3437178166838311,
      "grad_norm": 1.1891929638258938,
      "learning_rate": 2e-05,
      "loss": 0.6524415016174316,
      "step": 2670
    },
    {
      "epoch": 0.345005149330587,
      "grad_norm": 2.3605719453162757,
      "learning_rate": 2e-05,
      "loss": 0.6747819900512695,
      "step": 2680
    },
    {
      "epoch": 0.34629248197734297,
      "grad_norm": 2.470411316918508,
      "learning_rate": 2e-05,
      "loss": 0.6213415145874024,
      "step": 2690
    },
    {
      "epoch": 0.34757981462409887,
      "grad_norm": 2.0404873221926123,
      "learning_rate": 2e-05,
      "loss": 0.6706666469573974,
      "step": 2700
    },
    {
      "epoch": 0.34886714727085477,
      "grad_norm": 1.4692042495535325,
      "learning_rate": 2e-05,
      "loss": 0.6797372817993164,
      "step": 2710
    },
    {
      "epoch": 0.35015447991761073,
      "grad_norm": 2.589397458312476,
      "learning_rate": 2e-05,
      "loss": 0.6521578788757324,
      "step": 2720
    },
    {
      "epoch": 0.35144181256436663,
      "grad_norm": 1.1277270836479611,
      "learning_rate": 2e-05,
      "loss": 0.6732677459716797,
      "step": 2730
    },
    {
      "epoch": 0.35272914521112253,
      "grad_norm": 1.5051167160166194,
      "learning_rate": 2e-05,
      "loss": 0.658482837677002,
      "step": 2740
    },
    {
      "epoch": 0.3540164778578785,
      "grad_norm": 2.2443513526369627,
      "learning_rate": 2e-05,
      "loss": 0.6609281063079834,
      "step": 2750
    },
    {
      "epoch": 0.3553038105046344,
      "grad_norm": 2.8898723653557146,
      "learning_rate": 2e-05,
      "loss": 0.6852851867675781,
      "step": 2760
    },
    {
      "epoch": 0.35659114315139034,
      "grad_norm": 1.9403402362804802,
      "learning_rate": 2e-05,
      "loss": 0.629300594329834,
      "step": 2770
    },
    {
      "epoch": 0.35787847579814624,
      "grad_norm": 1.4399739483600749,
      "learning_rate": 2e-05,
      "loss": 0.6836570739746094,
      "step": 2780
    },
    {
      "epoch": 0.35916580844490215,
      "grad_norm": 1.394028020897877,
      "learning_rate": 2e-05,
      "loss": 0.6523145198822021,
      "step": 2790
    },
    {
      "epoch": 0.3604531410916581,
      "grad_norm": 1.519666297429449,
      "learning_rate": 2e-05,
      "loss": 0.6626304626464844,
      "step": 2800
    },
    {
      "epoch": 0.361740473738414,
      "grad_norm": 6.45118774048948,
      "learning_rate": 2e-05,
      "loss": 0.6728658199310302,
      "step": 2810
    },
    {
      "epoch": 0.3630278063851699,
      "grad_norm": 5.278190679624564,
      "learning_rate": 2e-05,
      "loss": 0.6950824737548829,
      "step": 2820
    },
    {
      "epoch": 0.36431513903192586,
      "grad_norm": 3.379627099450112,
      "learning_rate": 2e-05,
      "loss": 0.7113072872161865,
      "step": 2830
    },
    {
      "epoch": 0.36560247167868176,
      "grad_norm": 9.928058665678439,
      "learning_rate": 2e-05,
      "loss": 0.6998546123504639,
      "step": 2840
    },
    {
      "epoch": 0.3668898043254377,
      "grad_norm": 2.4261849530745616,
      "learning_rate": 2e-05,
      "loss": 0.7137026786804199,
      "step": 2850
    },
    {
      "epoch": 0.3681771369721936,
      "grad_norm": 1.705339489405855,
      "learning_rate": 2e-05,
      "loss": 0.6986976623535156,
      "step": 2860
    },
    {
      "epoch": 0.3694644696189495,
      "grad_norm": 1.5901594848354943,
      "learning_rate": 2e-05,
      "loss": 0.6721778869628906,
      "step": 2870
    },
    {
      "epoch": 0.3707518022657055,
      "grad_norm": 2.0976255961116155,
      "learning_rate": 2e-05,
      "loss": 0.6972661972045898,
      "step": 2880
    },
    {
      "epoch": 0.3720391349124614,
      "grad_norm": 1.737453259234397,
      "learning_rate": 2e-05,
      "loss": 0.6896710395812988,
      "step": 2890
    },
    {
      "epoch": 0.3733264675592173,
      "grad_norm": 6.687644336375675,
      "learning_rate": 2e-05,
      "loss": 0.6870223045349121,
      "step": 2900
    },
    {
      "epoch": 0.37461380020597324,
      "grad_norm": 1.7526243999408582,
      "learning_rate": 2e-05,
      "loss": 0.6775806427001954,
      "step": 2910
    },
    {
      "epoch": 0.37590113285272914,
      "grad_norm": 1.449069006898657,
      "learning_rate": 2e-05,
      "loss": 0.6732563018798828,
      "step": 2920
    },
    {
      "epoch": 0.3771884654994851,
      "grad_norm": 1.5495788900320941,
      "learning_rate": 2e-05,
      "loss": 0.6752627849578857,
      "step": 2930
    },
    {
      "epoch": 0.378475798146241,
      "grad_norm": 1.4928682148793067,
      "learning_rate": 2e-05,
      "loss": 0.6585468292236328,
      "step": 2940
    },
    {
      "epoch": 0.3797631307929969,
      "grad_norm": 3.570159299542741,
      "learning_rate": 2e-05,
      "loss": 0.6992852687835693,
      "step": 2950
    },
    {
      "epoch": 0.38105046343975285,
      "grad_norm": 1.745416978112621,
      "learning_rate": 2e-05,
      "loss": 0.6408042430877685,
      "step": 2960
    },
    {
      "epoch": 0.38233779608650875,
      "grad_norm": 1.3878562358445177,
      "learning_rate": 2e-05,
      "loss": 0.652539587020874,
      "step": 2970
    },
    {
      "epoch": 0.38362512873326465,
      "grad_norm": 1.7373358136561254,
      "learning_rate": 2e-05,
      "loss": 0.686555814743042,
      "step": 2980
    },
    {
      "epoch": 0.3849124613800206,
      "grad_norm": 1.6110411266569502,
      "learning_rate": 2e-05,
      "loss": 0.6672499656677247,
      "step": 2990
    },
    {
      "epoch": 0.3861997940267765,
      "grad_norm": 8.591842569044532,
      "learning_rate": 2e-05,
      "loss": 0.7332715511322021,
      "step": 3000
    },
    {
      "epoch": 0.38748712667353247,
      "grad_norm": 4.770480208347062,
      "learning_rate": 2e-05,
      "loss": 0.6914670467376709,
      "step": 3010
    },
    {
      "epoch": 0.38877445932028837,
      "grad_norm": 2.5309185366901246,
      "learning_rate": 2e-05,
      "loss": 0.681065845489502,
      "step": 3020
    },
    {
      "epoch": 0.39006179196704427,
      "grad_norm": 1.6387711955152355,
      "learning_rate": 2e-05,
      "loss": 0.6919746398925781,
      "step": 3030
    },
    {
      "epoch": 0.3913491246138002,
      "grad_norm": 1.3566915307344818,
      "learning_rate": 2e-05,
      "loss": 0.6504603385925293,
      "step": 3040
    },
    {
      "epoch": 0.3926364572605561,
      "grad_norm": 2.786785682313342,
      "learning_rate": 2e-05,
      "loss": 0.672648572921753,
      "step": 3050
    },
    {
      "epoch": 0.393923789907312,
      "grad_norm": 1.7948386684878221,
      "learning_rate": 2e-05,
      "loss": 0.6768685817718506,
      "step": 3060
    },
    {
      "epoch": 0.395211122554068,
      "grad_norm": 1.3992759220970516,
      "learning_rate": 2e-05,
      "loss": 0.673085880279541,
      "step": 3070
    },
    {
      "epoch": 0.3964984552008239,
      "grad_norm": 2.2397090140048146,
      "learning_rate": 2e-05,
      "loss": 0.6780692100524902,
      "step": 3080
    },
    {
      "epoch": 0.39778578784757984,
      "grad_norm": 1.825394663917336,
      "learning_rate": 2e-05,
      "loss": 0.6607382774353028,
      "step": 3090
    },
    {
      "epoch": 0.39907312049433574,
      "grad_norm": 4.389762594014447,
      "learning_rate": 2e-05,
      "loss": 0.6736061096191406,
      "step": 3100
    },
    {
      "epoch": 0.40036045314109164,
      "grad_norm": 1.9317885442360987,
      "learning_rate": 2e-05,
      "loss": 0.6794852256774903,
      "step": 3110
    },
    {
      "epoch": 0.4016477857878476,
      "grad_norm": 8.9709361728038,
      "learning_rate": 2e-05,
      "loss": 0.6856914043426514,
      "step": 3120
    },
    {
      "epoch": 0.4029351184346035,
      "grad_norm": 10.096913313964608,
      "learning_rate": 2e-05,
      "loss": 0.6717545986175537,
      "step": 3130
    },
    {
      "epoch": 0.4042224510813594,
      "grad_norm": 4.1135761380680025,
      "learning_rate": 2e-05,
      "loss": 0.6613590240478515,
      "step": 3140
    },
    {
      "epoch": 0.40550978372811536,
      "grad_norm": 5.458974686593683,
      "learning_rate": 2e-05,
      "loss": 0.6676340103149414,
      "step": 3150
    },
    {
      "epoch": 0.40679711637487126,
      "grad_norm": 1.5353876859514601,
      "learning_rate": 2e-05,
      "loss": 0.6625430107116699,
      "step": 3160
    },
    {
      "epoch": 0.4080844490216272,
      "grad_norm": 2.617512976186076,
      "learning_rate": 2e-05,
      "loss": 0.7098607540130615,
      "step": 3170
    },
    {
      "epoch": 0.4093717816683831,
      "grad_norm": 24.853744974348846,
      "learning_rate": 2e-05,
      "loss": 0.6653516292572021,
      "step": 3180
    },
    {
      "epoch": 0.410659114315139,
      "grad_norm": 1.8938125815091902,
      "learning_rate": 2e-05,
      "loss": 0.6713570594787598,
      "step": 3190
    },
    {
      "epoch": 0.411946446961895,
      "grad_norm": 5.279762512651014,
      "learning_rate": 2e-05,
      "loss": 0.6681087493896485,
      "step": 3200
    },
    {
      "epoch": 0.4132337796086509,
      "grad_norm": 8.141366763733092,
      "learning_rate": 2e-05,
      "loss": 0.6675487041473389,
      "step": 3210
    },
    {
      "epoch": 0.4145211122554068,
      "grad_norm": 1.4196281538096058,
      "learning_rate": 2e-05,
      "loss": 0.6300604820251465,
      "step": 3220
    },
    {
      "epoch": 0.41580844490216273,
      "grad_norm": 1.5189135563521712,
      "learning_rate": 2e-05,
      "loss": 0.6708904266357422,
      "step": 3230
    },
    {
      "epoch": 0.41709577754891863,
      "grad_norm": 1.316328436422404,
      "learning_rate": 2e-05,
      "loss": 0.6688165187835693,
      "step": 3240
    },
    {
      "epoch": 0.41838311019567453,
      "grad_norm": 1.490385997513553,
      "learning_rate": 2e-05,
      "loss": 0.6932008266448975,
      "step": 3250
    },
    {
      "epoch": 0.4196704428424305,
      "grad_norm": 6.699064455485067,
      "learning_rate": 2e-05,
      "loss": 0.6734150886535645,
      "step": 3260
    },
    {
      "epoch": 0.4209577754891864,
      "grad_norm": 1.9902209572292322,
      "learning_rate": 2e-05,
      "loss": 0.621324110031128,
      "step": 3270
    },
    {
      "epoch": 0.42224510813594235,
      "grad_norm": 2.3929984697771323,
      "learning_rate": 2e-05,
      "loss": 0.6553589820861816,
      "step": 3280
    },
    {
      "epoch": 0.42353244078269825,
      "grad_norm": 1.4323973020118568,
      "learning_rate": 2e-05,
      "loss": 0.6547521114349365,
      "step": 3290
    },
    {
      "epoch": 0.42481977342945415,
      "grad_norm": 1.722224287548998,
      "learning_rate": 2e-05,
      "loss": 0.7304846763610839,
      "step": 3300
    },
    {
      "epoch": 0.4261071060762101,
      "grad_norm": 2.1934329471769147,
      "learning_rate": 2e-05,
      "loss": 0.6440581321716309,
      "step": 3310
    },
    {
      "epoch": 0.427394438722966,
      "grad_norm": 6.457762244550262,
      "learning_rate": 2e-05,
      "loss": 0.6681999206542969,
      "step": 3320
    },
    {
      "epoch": 0.4286817713697219,
      "grad_norm": 4.0785650101698305,
      "learning_rate": 2e-05,
      "loss": 0.6417197704315185,
      "step": 3330
    },
    {
      "epoch": 0.42996910401647787,
      "grad_norm": 3.671425616461221,
      "learning_rate": 2e-05,
      "loss": 0.6589700222015381,
      "step": 3340
    },
    {
      "epoch": 0.43125643666323377,
      "grad_norm": 2.3178859428756904,
      "learning_rate": 2e-05,
      "loss": 0.661466121673584,
      "step": 3350
    },
    {
      "epoch": 0.4325437693099897,
      "grad_norm": 1.7739984241122377,
      "learning_rate": 2e-05,
      "loss": 0.6633042812347412,
      "step": 3360
    },
    {
      "epoch": 0.4338311019567456,
      "grad_norm": 1.4750459481455314,
      "learning_rate": 2e-05,
      "loss": 0.6860431671142578,
      "step": 3370
    },
    {
      "epoch": 0.4351184346035015,
      "grad_norm": 1.024462615445388,
      "learning_rate": 2e-05,
      "loss": 0.6539207458496094,
      "step": 3380
    },
    {
      "epoch": 0.4364057672502575,
      "grad_norm": 1.0028907091376442,
      "learning_rate": 2e-05,
      "loss": 0.6821427822113038,
      "step": 3390
    },
    {
      "epoch": 0.4376930998970134,
      "grad_norm": 20.32370151247624,
      "learning_rate": 2e-05,
      "loss": 0.6655908584594726,
      "step": 3400
    },
    {
      "epoch": 0.4389804325437693,
      "grad_norm": 1.5298304681583943,
      "learning_rate": 2e-05,
      "loss": 0.6839134216308593,
      "step": 3410
    },
    {
      "epoch": 0.44026776519052524,
      "grad_norm": 2.0967857109698524,
      "learning_rate": 2e-05,
      "loss": 0.6819107055664062,
      "step": 3420
    },
    {
      "epoch": 0.44155509783728114,
      "grad_norm": 2.027706958669317,
      "learning_rate": 2e-05,
      "loss": 0.6640716552734375,
      "step": 3430
    },
    {
      "epoch": 0.4428424304840371,
      "grad_norm": 1.1304241862727915,
      "learning_rate": 2e-05,
      "loss": 0.6362229347229004,
      "step": 3440
    },
    {
      "epoch": 0.444129763130793,
      "grad_norm": 1.0563233431055887,
      "learning_rate": 2e-05,
      "loss": 0.6823592185974121,
      "step": 3450
    },
    {
      "epoch": 0.4454170957775489,
      "grad_norm": 6.906802092336299,
      "learning_rate": 2e-05,
      "loss": 0.723987340927124,
      "step": 3460
    },
    {
      "epoch": 0.44670442842430486,
      "grad_norm": 1.247112091006261,
      "learning_rate": 2e-05,
      "loss": 0.6849922180175781,
      "step": 3470
    },
    {
      "epoch": 0.44799176107106076,
      "grad_norm": 26.02341861091827,
      "learning_rate": 2e-05,
      "loss": 0.630082654953003,
      "step": 3480
    },
    {
      "epoch": 0.44927909371781666,
      "grad_norm": 3.428639375524443,
      "learning_rate": 2e-05,
      "loss": 0.6845290184020996,
      "step": 3490
    },
    {
      "epoch": 0.4505664263645726,
      "grad_norm": 1.166177002019518,
      "learning_rate": 2e-05,
      "loss": 0.6864035606384278,
      "step": 3500
    },
    {
      "epoch": 0.4518537590113285,
      "grad_norm": 1.8605023565035361,
      "learning_rate": 2e-05,
      "loss": 0.6661740779876709,
      "step": 3510
    },
    {
      "epoch": 0.45314109165808447,
      "grad_norm": 0.971442813441688,
      "learning_rate": 2e-05,
      "loss": 0.66938796043396,
      "step": 3520
    },
    {
      "epoch": 0.4544284243048404,
      "grad_norm": 1.3538855308769568,
      "learning_rate": 2e-05,
      "loss": 0.663056468963623,
      "step": 3530
    },
    {
      "epoch": 0.4557157569515963,
      "grad_norm": 1.3794284739098457,
      "learning_rate": 2e-05,
      "loss": 0.6419530391693116,
      "step": 3540
    },
    {
      "epoch": 0.45700308959835223,
      "grad_norm": 2.2926898326260194,
      "learning_rate": 2e-05,
      "loss": 0.6693635940551758,
      "step": 3550
    },
    {
      "epoch": 0.45829042224510813,
      "grad_norm": 5.296737682431592,
      "learning_rate": 2e-05,
      "loss": 0.693416690826416,
      "step": 3560
    },
    {
      "epoch": 0.45957775489186403,
      "grad_norm": 1.4313451910795734,
      "learning_rate": 2e-05,
      "loss": 0.7027761459350585,
      "step": 3570
    },
    {
      "epoch": 0.46086508753862,
      "grad_norm": 8.00538666124274,
      "learning_rate": 2e-05,
      "loss": 0.6691301345825196,
      "step": 3580
    },
    {
      "epoch": 0.4621524201853759,
      "grad_norm": 3.1943754145255605,
      "learning_rate": 2e-05,
      "loss": 0.6514184474945068,
      "step": 3590
    },
    {
      "epoch": 0.46343975283213185,
      "grad_norm": 6.573149792400552,
      "learning_rate": 2e-05,
      "loss": 0.6754954338073731,
      "step": 3600
    },
    {
      "epoch": 0.46472708547888775,
      "grad_norm": 2.154765907874636,
      "learning_rate": 2e-05,
      "loss": 0.6678764343261718,
      "step": 3610
    },
    {
      "epoch": 0.46601441812564365,
      "grad_norm": 3.6800679121985946,
      "learning_rate": 2e-05,
      "loss": 0.692064619064331,
      "step": 3620
    },
    {
      "epoch": 0.4673017507723996,
      "grad_norm": 1.8650670848048954,
      "learning_rate": 2e-05,
      "loss": 0.6698975563049316,
      "step": 3630
    },
    {
      "epoch": 0.4685890834191555,
      "grad_norm": 1.8829809424851136,
      "learning_rate": 2e-05,
      "loss": 0.6646499633789062,
      "step": 3640
    },
    {
      "epoch": 0.4698764160659114,
      "grad_norm": 1.1146778872216527,
      "learning_rate": 2e-05,
      "loss": 0.6541554927825928,
      "step": 3650
    },
    {
      "epoch": 0.47116374871266736,
      "grad_norm": 1.3884161087356786,
      "learning_rate": 2e-05,
      "loss": 0.7008742332458496,
      "step": 3660
    },
    {
      "epoch": 0.47245108135942326,
      "grad_norm": 2.488001552019463,
      "learning_rate": 2e-05,
      "loss": 0.7059492111206055,
      "step": 3670
    },
    {
      "epoch": 0.4737384140061792,
      "grad_norm": 2.1436795964440347,
      "learning_rate": 2e-05,
      "loss": 0.6514601707458496,
      "step": 3680
    },
    {
      "epoch": 0.4750257466529351,
      "grad_norm": 1.6540498436206854,
      "learning_rate": 2e-05,
      "loss": 0.684079933166504,
      "step": 3690
    },
    {
      "epoch": 0.476313079299691,
      "grad_norm": 1.7862339002922256,
      "learning_rate": 2e-05,
      "loss": 0.7529743194580079,
      "step": 3700
    },
    {
      "epoch": 0.477600411946447,
      "grad_norm": 0.9956859259704859,
      "learning_rate": 2e-05,
      "loss": 0.663523817062378,
      "step": 3710
    },
    {
      "epoch": 0.4788877445932029,
      "grad_norm": 1.1151185742888514,
      "learning_rate": 2e-05,
      "loss": 0.6718929290771485,
      "step": 3720
    },
    {
      "epoch": 0.4801750772399588,
      "grad_norm": 1.7975567259722014,
      "learning_rate": 2e-05,
      "loss": 0.6369192123413085,
      "step": 3730
    },
    {
      "epoch": 0.48146240988671474,
      "grad_norm": 18.533022605588275,
      "learning_rate": 2e-05,
      "loss": 0.6432384490966797,
      "step": 3740
    },
    {
      "epoch": 0.48274974253347064,
      "grad_norm": 2.7878600747183073,
      "learning_rate": 2e-05,
      "loss": 0.6795418739318848,
      "step": 3750
    },
    {
      "epoch": 0.4840370751802266,
      "grad_norm": 1.0890450077296256,
      "learning_rate": 2e-05,
      "loss": 0.6459294319152832,
      "step": 3760
    },
    {
      "epoch": 0.4853244078269825,
      "grad_norm": 6.4422192763167665,
      "learning_rate": 2e-05,
      "loss": 0.674666166305542,
      "step": 3770
    },
    {
      "epoch": 0.4866117404737384,
      "grad_norm": 1.3127188153680827,
      "learning_rate": 2e-05,
      "loss": 0.6758196830749512,
      "step": 3780
    },
    {
      "epoch": 0.48789907312049435,
      "grad_norm": 1.2704832482153694,
      "learning_rate": 2e-05,
      "loss": 0.6724513530731201,
      "step": 3790
    },
    {
      "epoch": 0.48918640576725025,
      "grad_norm": 5.946873165430182,
      "learning_rate": 2e-05,
      "loss": 0.6656732559204102,
      "step": 3800
    },
    {
      "epoch": 0.49047373841400616,
      "grad_norm": 2.1508874984563615,
      "learning_rate": 2e-05,
      "loss": 0.6731793880462646,
      "step": 3810
    },
    {
      "epoch": 0.4917610710607621,
      "grad_norm": 11.013002873008867,
      "learning_rate": 2e-05,
      "loss": 0.6558897972106934,
      "step": 3820
    },
    {
      "epoch": 0.493048403707518,
      "grad_norm": 1.1369603803159063,
      "learning_rate": 2e-05,
      "loss": 0.6341354370117187,
      "step": 3830
    },
    {
      "epoch": 0.49433573635427397,
      "grad_norm": 1.2123188189670462,
      "learning_rate": 2e-05,
      "loss": 0.6724311828613281,
      "step": 3840
    },
    {
      "epoch": 0.49562306900102987,
      "grad_norm": 5.71968755913974,
      "learning_rate": 2e-05,
      "loss": 0.6578932762145996,
      "step": 3850
    },
    {
      "epoch": 0.49691040164778577,
      "grad_norm": 1.667301339852815,
      "learning_rate": 2e-05,
      "loss": 0.6530393600463867,
      "step": 3860
    },
    {
      "epoch": 0.4981977342945417,
      "grad_norm": 2.2052424082783006,
      "learning_rate": 2e-05,
      "loss": 0.6489648342132568,
      "step": 3870
    },
    {
      "epoch": 0.49948506694129763,
      "grad_norm": 1.182655371071617,
      "learning_rate": 2e-05,
      "loss": 0.7217607498168945,
      "step": 3880
    },
    {
      "epoch": 0.5007723995880535,
      "grad_norm": 1.2358308522133377,
      "learning_rate": 2e-05,
      "loss": 0.6366837978363037,
      "step": 3890
    },
    {
      "epoch": 0.5020597322348095,
      "grad_norm": 2.2520605480700366,
      "learning_rate": 2e-05,
      "loss": 0.6712183952331543,
      "step": 3900
    },
    {
      "epoch": 0.5033470648815654,
      "grad_norm": 2.151802936233694,
      "learning_rate": 2e-05,
      "loss": 0.6303033828735352,
      "step": 3910
    },
    {
      "epoch": 0.5046343975283213,
      "grad_norm": 2.4362255545668843,
      "learning_rate": 2e-05,
      "loss": 0.6984471321105957,
      "step": 3920
    },
    {
      "epoch": 0.5059217301750772,
      "grad_norm": 1.3273430757286633,
      "learning_rate": 2e-05,
      "loss": 0.6668586254119873,
      "step": 3930
    },
    {
      "epoch": 0.5072090628218332,
      "grad_norm": 1.4830951454416714,
      "learning_rate": 2e-05,
      "loss": 0.6583553314208984,
      "step": 3940
    },
    {
      "epoch": 0.508496395468589,
      "grad_norm": 2.5789130838170666,
      "learning_rate": 2e-05,
      "loss": 0.6548712730407715,
      "step": 3950
    },
    {
      "epoch": 0.509783728115345,
      "grad_norm": 1.3557484975654484,
      "learning_rate": 2e-05,
      "loss": 0.6732818603515625,
      "step": 3960
    },
    {
      "epoch": 0.511071060762101,
      "grad_norm": 1.5388518917022493,
      "learning_rate": 2e-05,
      "loss": 0.630622386932373,
      "step": 3970
    },
    {
      "epoch": 0.5123583934088568,
      "grad_norm": 3.3880763823383995,
      "learning_rate": 2e-05,
      "loss": 0.6452838897705078,
      "step": 3980
    },
    {
      "epoch": 0.5136457260556128,
      "grad_norm": 1.9510210816666544,
      "learning_rate": 2e-05,
      "loss": 0.6481393814086914,
      "step": 3990
    },
    {
      "epoch": 0.5149330587023687,
      "grad_norm": 3.3445035332864053,
      "learning_rate": 2e-05,
      "loss": 0.6437893867492676,
      "step": 4000
    },
    {
      "epoch": 0.5162203913491246,
      "grad_norm": 1.4430061642931242,
      "learning_rate": 2e-05,
      "loss": 0.6935883522033691,
      "step": 4010
    },
    {
      "epoch": 0.5175077239958805,
      "grad_norm": 1.678957944027874,
      "learning_rate": 2e-05,
      "loss": 0.6531391143798828,
      "step": 4020
    },
    {
      "epoch": 0.5187950566426365,
      "grad_norm": 2.1307079559087994,
      "learning_rate": 2e-05,
      "loss": 0.7015140533447266,
      "step": 4030
    },
    {
      "epoch": 0.5200823892893924,
      "grad_norm": 2.1707573190704275,
      "learning_rate": 2e-05,
      "loss": 0.6406472682952881,
      "step": 4040
    },
    {
      "epoch": 0.5213697219361483,
      "grad_norm": 1.381923517046361,
      "learning_rate": 2e-05,
      "loss": 0.689560842514038,
      "step": 4050
    },
    {
      "epoch": 0.5226570545829042,
      "grad_norm": 2.171024005846642,
      "learning_rate": 2e-05,
      "loss": 0.7308357238769532,
      "step": 4060
    },
    {
      "epoch": 0.5239443872296602,
      "grad_norm": 3.6614263601181762,
      "learning_rate": 2e-05,
      "loss": 0.6891562938690186,
      "step": 4070
    },
    {
      "epoch": 0.525231719876416,
      "grad_norm": 1.8607140040470387,
      "learning_rate": 2e-05,
      "loss": 0.6409766197204589,
      "step": 4080
    },
    {
      "epoch": 0.526519052523172,
      "grad_norm": 1.0660315330714643,
      "learning_rate": 2e-05,
      "loss": 0.6235333919525147,
      "step": 4090
    },
    {
      "epoch": 0.527806385169928,
      "grad_norm": 1.851561848616255,
      "learning_rate": 2e-05,
      "loss": 0.6496839046478271,
      "step": 4100
    },
    {
      "epoch": 0.5290937178166838,
      "grad_norm": 3.2189727986218397,
      "learning_rate": 2e-05,
      "loss": 0.6600148677825928,
      "step": 4110
    },
    {
      "epoch": 0.5303810504634398,
      "grad_norm": 0.994640856729955,
      "learning_rate": 2e-05,
      "loss": 0.6826591491699219,
      "step": 4120
    },
    {
      "epoch": 0.5316683831101957,
      "grad_norm": 1.021253439822631,
      "learning_rate": 2e-05,
      "loss": 0.6925134658813477,
      "step": 4130
    },
    {
      "epoch": 0.5329557157569516,
      "grad_norm": 1.3527319778524327,
      "learning_rate": 2e-05,
      "loss": 0.5992767333984375,
      "step": 4140
    },
    {
      "epoch": 0.5342430484037075,
      "grad_norm": 0.941528852003794,
      "learning_rate": 2e-05,
      "loss": 0.6514298439025878,
      "step": 4150
    },
    {
      "epoch": 0.5355303810504635,
      "grad_norm": 1.1068912732324785,
      "learning_rate": 2e-05,
      "loss": 0.622230052947998,
      "step": 4160
    },
    {
      "epoch": 0.5368177136972193,
      "grad_norm": 1.6844891158297295,
      "learning_rate": 2e-05,
      "loss": 0.7005961418151856,
      "step": 4170
    },
    {
      "epoch": 0.5381050463439753,
      "grad_norm": 1.477298874463471,
      "learning_rate": 2e-05,
      "loss": 0.6360055923461914,
      "step": 4180
    },
    {
      "epoch": 0.5393923789907312,
      "grad_norm": 1.5164764079530957,
      "learning_rate": 2e-05,
      "loss": 0.6254413604736329,
      "step": 4190
    },
    {
      "epoch": 0.5406797116374872,
      "grad_norm": 1.8706816046379673,
      "learning_rate": 2e-05,
      "loss": 0.6307280540466309,
      "step": 4200
    },
    {
      "epoch": 0.541967044284243,
      "grad_norm": 5.012114859267674,
      "learning_rate": 2e-05,
      "loss": 0.6400323390960694,
      "step": 4210
    },
    {
      "epoch": 0.543254376930999,
      "grad_norm": 1.9790677956658003,
      "learning_rate": 2e-05,
      "loss": 0.6628574371337891,
      "step": 4220
    },
    {
      "epoch": 0.5445417095777549,
      "grad_norm": 1.3470501401524795,
      "learning_rate": 2e-05,
      "loss": 0.6545948028564453,
      "step": 4230
    },
    {
      "epoch": 0.5458290422245108,
      "grad_norm": 2.0655432822140667,
      "learning_rate": 2e-05,
      "loss": 0.6282138347625732,
      "step": 4240
    },
    {
      "epoch": 0.5471163748712667,
      "grad_norm": 1.9543886543895945,
      "learning_rate": 2e-05,
      "loss": 0.6307715892791748,
      "step": 4250
    },
    {
      "epoch": 0.5484037075180227,
      "grad_norm": 0.9789158685887359,
      "learning_rate": 2e-05,
      "loss": 0.6398684501647949,
      "step": 4260
    },
    {
      "epoch": 0.5496910401647785,
      "grad_norm": 1.2336567954340274,
      "learning_rate": 2e-05,
      "loss": 0.695476484298706,
      "step": 4270
    },
    {
      "epoch": 0.5509783728115345,
      "grad_norm": 1.446989848485389,
      "learning_rate": 2e-05,
      "loss": 0.6444989204406738,
      "step": 4280
    },
    {
      "epoch": 0.5522657054582905,
      "grad_norm": 1.6899514779115894,
      "learning_rate": 2e-05,
      "loss": 0.6552947998046875,
      "step": 4290
    },
    {
      "epoch": 0.5535530381050463,
      "grad_norm": 4.300141089330589,
      "learning_rate": 2e-05,
      "loss": 0.6904422760009765,
      "step": 4300
    },
    {
      "epoch": 0.5548403707518023,
      "grad_norm": 1.1915809959722747,
      "learning_rate": 2e-05,
      "loss": 0.6382978439331055,
      "step": 4310
    },
    {
      "epoch": 0.5561277033985582,
      "grad_norm": 1.2448701596113807,
      "learning_rate": 2e-05,
      "loss": 0.6655734062194825,
      "step": 4320
    },
    {
      "epoch": 0.5574150360453141,
      "grad_norm": 1.6480835867933197,
      "learning_rate": 2e-05,
      "loss": 0.6510987758636475,
      "step": 4330
    },
    {
      "epoch": 0.55870236869207,
      "grad_norm": 1.1177884658067496,
      "learning_rate": 2e-05,
      "loss": 0.6373340129852295,
      "step": 4340
    },
    {
      "epoch": 0.559989701338826,
      "grad_norm": 1.227905462506924,
      "learning_rate": 2e-05,
      "loss": 0.6498836517333985,
      "step": 4350
    },
    {
      "epoch": 0.5612770339855818,
      "grad_norm": 1.0561761623259625,
      "learning_rate": 2e-05,
      "loss": 0.6252914428710937,
      "step": 4360
    },
    {
      "epoch": 0.5625643666323378,
      "grad_norm": 1.4718178904519381,
      "learning_rate": 2e-05,
      "loss": 0.6598007202148437,
      "step": 4370
    },
    {
      "epoch": 0.5638516992790937,
      "grad_norm": 5.076921050599247,
      "learning_rate": 2e-05,
      "loss": 0.6385498046875,
      "step": 4380
    },
    {
      "epoch": 0.5651390319258497,
      "grad_norm": 2.801980221216199,
      "learning_rate": 2e-05,
      "loss": 0.6354559898376465,
      "step": 4390
    },
    {
      "epoch": 0.5664263645726055,
      "grad_norm": 1.2249997259071304,
      "learning_rate": 2e-05,
      "loss": 0.6390414237976074,
      "step": 4400
    },
    {
      "epoch": 0.5677136972193615,
      "grad_norm": 1.3738188332331804,
      "learning_rate": 2e-05,
      "loss": 0.6940718650817871,
      "step": 4410
    },
    {
      "epoch": 0.5690010298661174,
      "grad_norm": 1.163672143894105,
      "learning_rate": 2e-05,
      "loss": 0.6161147594451905,
      "step": 4420
    },
    {
      "epoch": 0.5702883625128733,
      "grad_norm": 1.7472663262939392,
      "learning_rate": 2e-05,
      "loss": 0.6192094802856445,
      "step": 4430
    },
    {
      "epoch": 0.5715756951596292,
      "grad_norm": 1.0951357239038013,
      "learning_rate": 2e-05,
      "loss": 0.6478008270263672,
      "step": 4440
    },
    {
      "epoch": 0.5728630278063852,
      "grad_norm": 1.4523866646734536,
      "learning_rate": 2e-05,
      "loss": 0.6371118545532226,
      "step": 4450
    },
    {
      "epoch": 0.574150360453141,
      "grad_norm": 0.8879298974135746,
      "learning_rate": 2e-05,
      "loss": 0.6452761650085449,
      "step": 4460
    },
    {
      "epoch": 0.575437693099897,
      "grad_norm": 1.0416755052904407,
      "learning_rate": 2e-05,
      "loss": 0.690460205078125,
      "step": 4470
    },
    {
      "epoch": 0.576725025746653,
      "grad_norm": 1.3544236804335335,
      "learning_rate": 2e-05,
      "loss": 0.6621108055114746,
      "step": 4480
    },
    {
      "epoch": 0.5780123583934088,
      "grad_norm": 1.057843122408065,
      "learning_rate": 2e-05,
      "loss": 0.631985092163086,
      "step": 4490
    },
    {
      "epoch": 0.5792996910401648,
      "grad_norm": 1.1450465434830548,
      "learning_rate": 2e-05,
      "loss": 0.6567739009857178,
      "step": 4500
    },
    {
      "epoch": 0.5805870236869207,
      "grad_norm": 1.5384094917202418,
      "learning_rate": 2e-05,
      "loss": 0.6385008811950683,
      "step": 4510
    },
    {
      "epoch": 0.5818743563336766,
      "grad_norm": 1.7502133918813758,
      "learning_rate": 2e-05,
      "loss": 0.6220414161682128,
      "step": 4520
    },
    {
      "epoch": 0.5831616889804325,
      "grad_norm": 1.4936899817423515,
      "learning_rate": 2e-05,
      "loss": 0.6436448574066163,
      "step": 4530
    },
    {
      "epoch": 0.5844490216271885,
      "grad_norm": 1.450196944000251,
      "learning_rate": 2e-05,
      "loss": 0.637021541595459,
      "step": 4540
    },
    {
      "epoch": 0.5857363542739444,
      "grad_norm": 1.0566363253694624,
      "learning_rate": 2e-05,
      "loss": 0.6220635890960693,
      "step": 4550
    },
    {
      "epoch": 0.5870236869207003,
      "grad_norm": 0.985212887220168,
      "learning_rate": 2e-05,
      "loss": 0.6490059852600097,
      "step": 4560
    },
    {
      "epoch": 0.5883110195674562,
      "grad_norm": 1.9799701522660655,
      "learning_rate": 2e-05,
      "loss": 0.6891635894775391,
      "step": 4570
    },
    {
      "epoch": 0.5895983522142122,
      "grad_norm": 0.7700228407593017,
      "learning_rate": 2e-05,
      "loss": 0.649361515045166,
      "step": 4580
    },
    {
      "epoch": 0.590885684860968,
      "grad_norm": 1.0096645286973764,
      "learning_rate": 2e-05,
      "loss": 0.6665196418762207,
      "step": 4590
    },
    {
      "epoch": 0.592173017507724,
      "grad_norm": 0.9780863034536382,
      "learning_rate": 2e-05,
      "loss": 0.6462597846984863,
      "step": 4600
    },
    {
      "epoch": 0.59346035015448,
      "grad_norm": 5.406540309008322,
      "learning_rate": 2e-05,
      "loss": 0.634643840789795,
      "step": 4610
    },
    {
      "epoch": 0.5947476828012358,
      "grad_norm": 9.491651278813794,
      "learning_rate": 2e-05,
      "loss": 0.6160319328308106,
      "step": 4620
    },
    {
      "epoch": 0.5960350154479918,
      "grad_norm": 1.1718697514540304,
      "learning_rate": 2e-05,
      "loss": 0.6167381286621094,
      "step": 4630
    },
    {
      "epoch": 0.5973223480947477,
      "grad_norm": 1.6490618390346732,
      "learning_rate": 2e-05,
      "loss": 0.6317207336425781,
      "step": 4640
    },
    {
      "epoch": 0.5986096807415036,
      "grad_norm": 0.9903567166113278,
      "learning_rate": 2e-05,
      "loss": 0.6460660457611084,
      "step": 4650
    },
    {
      "epoch": 0.5998970133882595,
      "grad_norm": 1.0849060619579463,
      "learning_rate": 2e-05,
      "loss": 0.6712134838104248,
      "step": 4660
    },
    {
      "epoch": 0.6011843460350155,
      "grad_norm": 1.0279316617591696,
      "learning_rate": 2e-05,
      "loss": 0.6309604167938232,
      "step": 4670
    },
    {
      "epoch": 0.6024716786817713,
      "grad_norm": 1.0068797509794678,
      "learning_rate": 2e-05,
      "loss": 0.6249179363250732,
      "step": 4680
    },
    {
      "epoch": 0.6037590113285273,
      "grad_norm": 0.8781383820091182,
      "learning_rate": 2e-05,
      "loss": 0.662699031829834,
      "step": 4690
    },
    {
      "epoch": 0.6050463439752832,
      "grad_norm": 0.9145210451878394,
      "learning_rate": 2e-05,
      "loss": 0.6563605308532715,
      "step": 4700
    },
    {
      "epoch": 0.6063336766220392,
      "grad_norm": 1.040284357804239,
      "learning_rate": 2e-05,
      "loss": 0.6210678100585938,
      "step": 4710
    },
    {
      "epoch": 0.607621009268795,
      "grad_norm": 0.9415283242197504,
      "learning_rate": 2e-05,
      "loss": 0.6166461944580078,
      "step": 4720
    },
    {
      "epoch": 0.608908341915551,
      "grad_norm": 0.9727403069264652,
      "learning_rate": 2e-05,
      "loss": 0.6694095611572266,
      "step": 4730
    },
    {
      "epoch": 0.6101956745623069,
      "grad_norm": 0.8614268093374182,
      "learning_rate": 2e-05,
      "loss": 0.6036763668060303,
      "step": 4740
    },
    {
      "epoch": 0.6114830072090628,
      "grad_norm": 0.9773333066734369,
      "learning_rate": 2e-05,
      "loss": 0.6359670162200928,
      "step": 4750
    },
    {
      "epoch": 0.6127703398558187,
      "grad_norm": 1.112490918051523,
      "learning_rate": 2e-05,
      "loss": 0.6377384185791015,
      "step": 4760
    },
    {
      "epoch": 0.6140576725025747,
      "grad_norm": 0.7966671243978947,
      "learning_rate": 2e-05,
      "loss": 0.6769555568695068,
      "step": 4770
    },
    {
      "epoch": 0.6153450051493305,
      "grad_norm": 0.9030761946697352,
      "learning_rate": 2e-05,
      "loss": 0.6484799861907959,
      "step": 4780
    },
    {
      "epoch": 0.6166323377960865,
      "grad_norm": 1.0005212814326052,
      "learning_rate": 2e-05,
      "loss": 0.6718368053436279,
      "step": 4790
    },
    {
      "epoch": 0.6179196704428425,
      "grad_norm": 1.077742892482781,
      "learning_rate": 2e-05,
      "loss": 0.6295736312866211,
      "step": 4800
    },
    {
      "epoch": 0.6192070030895983,
      "grad_norm": 28.626120100894322,
      "learning_rate": 2e-05,
      "loss": 0.6363849639892578,
      "step": 4810
    },
    {
      "epoch": 0.6204943357363543,
      "grad_norm": 0.9724802777965487,
      "learning_rate": 2e-05,
      "loss": 0.6536874294281005,
      "step": 4820
    },
    {
      "epoch": 0.6217816683831102,
      "grad_norm": 1.1101268342122879,
      "learning_rate": 2e-05,
      "loss": 0.6011521339416503,
      "step": 4830
    },
    {
      "epoch": 0.6230690010298661,
      "grad_norm": 1.210018604456637,
      "learning_rate": 2e-05,
      "loss": 0.68607177734375,
      "step": 4840
    },
    {
      "epoch": 0.624356333676622,
      "grad_norm": 1.3320515065330811,
      "learning_rate": 2e-05,
      "loss": 0.6384735107421875,
      "step": 4850
    },
    {
      "epoch": 0.625643666323378,
      "grad_norm": 4.1440314155514795,
      "learning_rate": 2e-05,
      "loss": 0.6874375343322754,
      "step": 4860
    },
    {
      "epoch": 0.6269309989701339,
      "grad_norm": 0.9164636348056602,
      "learning_rate": 2e-05,
      "loss": 0.6379546165466309,
      "step": 4870
    },
    {
      "epoch": 0.6282183316168898,
      "grad_norm": 1.6654629652251853,
      "learning_rate": 2e-05,
      "loss": 0.6865750789642334,
      "step": 4880
    },
    {
      "epoch": 0.6295056642636457,
      "grad_norm": 0.9566181268893309,
      "learning_rate": 2e-05,
      "loss": 0.6759776115417481,
      "step": 4890
    },
    {
      "epoch": 0.6307929969104017,
      "grad_norm": 0.9916137026414327,
      "learning_rate": 2e-05,
      "loss": 0.6295784950256348,
      "step": 4900
    },
    {
      "epoch": 0.6320803295571575,
      "grad_norm": 0.9886654124917329,
      "learning_rate": 2e-05,
      "loss": 0.6312230110168457,
      "step": 4910
    },
    {
      "epoch": 0.6333676622039135,
      "grad_norm": 1.1265554071584722,
      "learning_rate": 2e-05,
      "loss": 0.6896268844604492,
      "step": 4920
    },
    {
      "epoch": 0.6346549948506695,
      "grad_norm": 1.5551307141827106,
      "learning_rate": 2e-05,
      "loss": 0.664113712310791,
      "step": 4930
    },
    {
      "epoch": 0.6359423274974253,
      "grad_norm": 1.1244929385209503,
      "learning_rate": 2e-05,
      "loss": 0.6766931056976319,
      "step": 4940
    },
    {
      "epoch": 0.6372296601441813,
      "grad_norm": 1.5539032133273178,
      "learning_rate": 2e-05,
      "loss": 0.583741283416748,
      "step": 4950
    },
    {
      "epoch": 0.6385169927909372,
      "grad_norm": 0.8059692241959814,
      "learning_rate": 2e-05,
      "loss": 0.6013468742370606,
      "step": 4960
    },
    {
      "epoch": 0.6398043254376931,
      "grad_norm": 0.8263809854582325,
      "learning_rate": 2e-05,
      "loss": 0.6536582946777344,
      "step": 4970
    },
    {
      "epoch": 0.641091658084449,
      "grad_norm": 1.2170950033809171,
      "learning_rate": 2e-05,
      "loss": 0.6847995758056641,
      "step": 4980
    },
    {
      "epoch": 0.642378990731205,
      "grad_norm": 1.4508710873711133,
      "learning_rate": 2e-05,
      "loss": 0.610681676864624,
      "step": 4990
    },
    {
      "epoch": 0.6436663233779608,
      "grad_norm": 1.5085502448451786,
      "learning_rate": 2e-05,
      "loss": 0.6235250473022461,
      "step": 5000
    },
    {
      "epoch": 0.6449536560247168,
      "grad_norm": 1.4152370947961994,
      "learning_rate": 2e-05,
      "loss": 0.6345526695251464,
      "step": 5010
    },
    {
      "epoch": 0.6462409886714727,
      "grad_norm": 0.9275258112992915,
      "learning_rate": 2e-05,
      "loss": 0.6097187042236328,
      "step": 5020
    },
    {
      "epoch": 0.6475283213182287,
      "grad_norm": 1.2790296235237106,
      "learning_rate": 2e-05,
      "loss": 0.6325679302215577,
      "step": 5030
    },
    {
      "epoch": 0.6488156539649845,
      "grad_norm": 1.418156164510018,
      "learning_rate": 2e-05,
      "loss": 0.6181277275085449,
      "step": 5040
    },
    {
      "epoch": 0.6501029866117405,
      "grad_norm": 1.112987928286208,
      "learning_rate": 2e-05,
      "loss": 0.6346569061279297,
      "step": 5050
    },
    {
      "epoch": 0.6513903192584964,
      "grad_norm": 3.0749314862416206,
      "learning_rate": 2e-05,
      "loss": 0.6290814876556396,
      "step": 5060
    },
    {
      "epoch": 0.6526776519052523,
      "grad_norm": 0.9323183551240596,
      "learning_rate": 2e-05,
      "loss": 0.6290140151977539,
      "step": 5070
    },
    {
      "epoch": 0.6539649845520082,
      "grad_norm": 2.781319107749208,
      "learning_rate": 2e-05,
      "loss": 0.6491301536560059,
      "step": 5080
    },
    {
      "epoch": 0.6552523171987642,
      "grad_norm": 1.1205853122844738,
      "learning_rate": 2e-05,
      "loss": 0.628472900390625,
      "step": 5090
    },
    {
      "epoch": 0.65653964984552,
      "grad_norm": 1.2741520326100562,
      "learning_rate": 2e-05,
      "loss": 0.6397724151611328,
      "step": 5100
    },
    {
      "epoch": 0.657826982492276,
      "grad_norm": 1.3869228069462751,
      "learning_rate": 2e-05,
      "loss": 0.6083261013031006,
      "step": 5110
    },
    {
      "epoch": 0.659114315139032,
      "grad_norm": 1.2589764737757365,
      "learning_rate": 2e-05,
      "loss": 0.6427864074707031,
      "step": 5120
    },
    {
      "epoch": 0.6604016477857878,
      "grad_norm": 0.997441043420146,
      "learning_rate": 2e-05,
      "loss": 0.6458511352539062,
      "step": 5130
    },
    {
      "epoch": 0.6616889804325438,
      "grad_norm": 1.8159403922513209,
      "learning_rate": 2e-05,
      "loss": 0.6361720085144043,
      "step": 5140
    },
    {
      "epoch": 0.6629763130792997,
      "grad_norm": 1.523814290998718,
      "learning_rate": 2e-05,
      "loss": 0.6253744125366211,
      "step": 5150
    },
    {
      "epoch": 0.6642636457260556,
      "grad_norm": 1.671461072048835,
      "learning_rate": 2e-05,
      "loss": 0.6089911937713623,
      "step": 5160
    },
    {
      "epoch": 0.6655509783728115,
      "grad_norm": 0.8608516899622911,
      "learning_rate": 2e-05,
      "loss": 0.658145809173584,
      "step": 5170
    },
    {
      "epoch": 0.6668383110195675,
      "grad_norm": 0.900994698832759,
      "learning_rate": 2e-05,
      "loss": 0.6144911766052246,
      "step": 5180
    },
    {
      "epoch": 0.6681256436663234,
      "grad_norm": 1.0955431580160213,
      "learning_rate": 2e-05,
      "loss": 0.6374094009399414,
      "step": 5190
    },
    {
      "epoch": 0.6694129763130793,
      "grad_norm": 1.556480675892863,
      "learning_rate": 2e-05,
      "loss": 0.6380134582519531,
      "step": 5200
    },
    {
      "epoch": 0.6707003089598352,
      "grad_norm": 0.8019943589977299,
      "learning_rate": 2e-05,
      "loss": 0.6089535713195801,
      "step": 5210
    },
    {
      "epoch": 0.6719876416065912,
      "grad_norm": 2.20231873430845,
      "learning_rate": 2e-05,
      "loss": 0.6290367126464844,
      "step": 5220
    },
    {
      "epoch": 0.673274974253347,
      "grad_norm": 1.2274276724050663,
      "learning_rate": 2e-05,
      "loss": 0.6053452968597413,
      "step": 5230
    },
    {
      "epoch": 0.674562306900103,
      "grad_norm": 1.1491653075055754,
      "learning_rate": 2e-05,
      "loss": 0.6174365997314453,
      "step": 5240
    },
    {
      "epoch": 0.675849639546859,
      "grad_norm": 1.519214436040967,
      "learning_rate": 2e-05,
      "loss": 0.6282214164733887,
      "step": 5250
    },
    {
      "epoch": 0.6771369721936148,
      "grad_norm": 1.9681939249837503,
      "learning_rate": 2e-05,
      "loss": 0.6702754974365235,
      "step": 5260
    },
    {
      "epoch": 0.6784243048403708,
      "grad_norm": 2.0663516464135046,
      "learning_rate": 2e-05,
      "loss": 0.6178231716156006,
      "step": 5270
    },
    {
      "epoch": 0.6797116374871267,
      "grad_norm": 3.5457906264943797,
      "learning_rate": 2e-05,
      "loss": 0.6104377269744873,
      "step": 5280
    },
    {
      "epoch": 0.6809989701338826,
      "grad_norm": 2.592039635377849,
      "learning_rate": 2e-05,
      "loss": 0.5997319698333741,
      "step": 5290
    },
    {
      "epoch": 0.6822863027806385,
      "grad_norm": 2.8201001964415813,
      "learning_rate": 2e-05,
      "loss": 0.6083297729492188,
      "step": 5300
    },
    {
      "epoch": 0.6835736354273945,
      "grad_norm": 1.4176264977179571,
      "learning_rate": 2e-05,
      "loss": 0.6262206554412841,
      "step": 5310
    },
    {
      "epoch": 0.6848609680741503,
      "grad_norm": 1.5094906394644259,
      "learning_rate": 2e-05,
      "loss": 0.6512961387634277,
      "step": 5320
    },
    {
      "epoch": 0.6861483007209063,
      "grad_norm": 3.30433650831491,
      "learning_rate": 2e-05,
      "loss": 0.6577296257019043,
      "step": 5330
    },
    {
      "epoch": 0.6874356333676622,
      "grad_norm": 1.069367234110156,
      "learning_rate": 2e-05,
      "loss": 0.708523178100586,
      "step": 5340
    },
    {
      "epoch": 0.6887229660144182,
      "grad_norm": 1.012712029253865,
      "learning_rate": 1.999194520388485e-05,
      "loss": 0.6392881393432617,
      "step": 5350
    },
    {
      "epoch": 0.690010298661174,
      "grad_norm": 3.7475226126382415,
      "learning_rate": 1.9959247303157763e-05,
      "loss": 0.6981637001037597,
      "step": 5360
    },
    {
      "epoch": 0.69129763130793,
      "grad_norm": 0.8725197668588418,
      "learning_rate": 1.990149424459817e-05,
      "loss": 0.6451784133911133,
      "step": 5370
    },
    {
      "epoch": 0.6925849639546859,
      "grad_norm": 2.40668585541682,
      "learning_rate": 1.9818847538648688e-05,
      "loss": 0.6496396064758301,
      "step": 5380
    },
    {
      "epoch": 0.6938722966014418,
      "grad_norm": 1.323407320819958,
      "learning_rate": 1.971153831256721e-05,
      "loss": 0.6111641883850097,
      "step": 5390
    },
    {
      "epoch": 0.6951596292481977,
      "grad_norm": 1.1235618200786057,
      "learning_rate": 1.9579866664063435e-05,
      "loss": 0.6316841125488282,
      "step": 5400
    },
    {
      "epoch": 0.6964469618949537,
      "grad_norm": 0.9306275131637043,
      "learning_rate": 1.9424200822054782e-05,
      "loss": 0.6398882865905762,
      "step": 5410
    },
    {
      "epoch": 0.6977342945417095,
      "grad_norm": 1.3577148815741154,
      "learning_rate": 1.9244976116888626e-05,
      "loss": 0.6030726432800293,
      "step": 5420
    },
    {
      "epoch": 0.6990216271884655,
      "grad_norm": 0.9945542500574273,
      "learning_rate": 1.904269376291071e-05,
      "loss": 0.6424736976623535,
      "step": 5430
    },
    {
      "epoch": 0.7003089598352215,
      "grad_norm": 1.7083784972949956,
      "learning_rate": 1.88179194567844e-05,
      "loss": 0.6232508659362793,
      "step": 5440
    },
    {
      "epoch": 0.7015962924819773,
      "grad_norm": 2.3906782874052777,
      "learning_rate": 1.8571281795480632e-05,
      "loss": 0.6836617946624756,
      "step": 5450
    },
    {
      "epoch": 0.7028836251287333,
      "grad_norm": 0.8983829870244988,
      "learning_rate": 1.8303470518362808e-05,
      "loss": 0.638803243637085,
      "step": 5460
    },
    {
      "epoch": 0.7041709577754892,
      "grad_norm": 1.0702565367827406,
      "learning_rate": 1.801523457828271e-05,
      "loss": 0.5981147766113282,
      "step": 5470
    },
    {
      "epoch": 0.7054582904222451,
      "grad_norm": 0.9245970931900683,
      "learning_rate": 1.7707380047081796e-05,
      "loss": 0.6440218925476074,
      "step": 5480
    },
    {
      "epoch": 0.706745623069001,
      "grad_norm": 3.4329881126173256,
      "learning_rate": 1.7380767861355252e-05,
      "loss": 0.6338693618774414,
      "step": 5490
    },
    {
      "epoch": 0.708032955715757,
      "grad_norm": 0.9662219584364334,
      "learning_rate": 1.703631141478299e-05,
      "loss": 0.6338390350341797,
      "step": 5500
    },
    {
      "epoch": 0.7093202883625128,
      "grad_norm": 0.9456759078552397,
      "learning_rate": 1.6674974003760705e-05,
      "loss": 0.5741812705993652,
      "step": 5510
    },
    {
      "epoch": 0.7106076210092688,
      "grad_norm": 2.323823810675979,
      "learning_rate": 1.6297766133474556e-05,
      "loss": 0.6422665596008301,
      "step": 5520
    },
    {
      "epoch": 0.7118949536560247,
      "grad_norm": 1.2569646315616894,
      "learning_rate": 1.5905742691953186e-05,
      "loss": 0.6417381286621093,
      "step": 5530
    },
    {
      "epoch": 0.7131822863027807,
      "grad_norm": 2.3953140629883984,
      "learning_rate": 1.55e-05,
      "loss": 0.6150378227233887,
      "step": 5540
    },
    {
      "epoch": 0.7144696189495365,
      "grad_norm": 5.844811952352083,
      "learning_rate": 1.5081672745255862e-05,
      "loss": 0.626163387298584,
      "step": 5550
    },
    {
      "epoch": 0.7157569515962925,
      "grad_norm": 0.9024296666383642,
      "learning_rate": 1.4651930808966206e-05,
      "loss": 0.6572855949401856,
      "step": 5560
    },
    {
      "epoch": 0.7170442842430484,
      "grad_norm": 5.050844694270378,
      "learning_rate": 1.4211975994326848e-05,
      "loss": 0.6286805629730224,
      "step": 5570
    },
    {
      "epoch": 0.7183316168898043,
      "grad_norm": 3.4410157275467093,
      "learning_rate": 1.376303866555779e-05,
      "loss": 0.6618189811706543,
      "step": 5580
    },
    {
      "epoch": 0.7196189495365602,
      "grad_norm": 1.1191996896183056,
      "learning_rate": 1.3306374307104164e-05,
      "loss": 0.6437629699707031,
      "step": 5590
    },
    {
      "epoch": 0.7209062821833162,
      "grad_norm": 1.1087845244732786,
      "learning_rate": 1.2843260012586718e-05,
      "loss": 0.6390609741210938,
      "step": 5600
    },
    {
      "epoch": 0.722193614830072,
      "grad_norm": 0.9658111944229392,
      "learning_rate": 1.2374990913320726e-05,
      "loss": 0.5985804557800293,
      "step": 5610
    },
    {
      "epoch": 0.723480947476828,
      "grad_norm": 1.0785028436789506,
      "learning_rate": 1.1902876556391247e-05,
      "loss": 0.6057968139648438,
      "step": 5620
    },
    {
      "epoch": 0.724768280123584,
      "grad_norm": 1.044842463367286,
      "learning_rate": 1.1428237242413685e-05,
      "loss": 0.6438541412353516,
      "step": 5630
    },
    {
      "epoch": 0.7260556127703398,
      "grad_norm": 1.9904090816906963,
      "learning_rate": 1.095240033322132e-05,
      "loss": 0.6080128669738769,
      "step": 5640
    },
    {
      "epoch": 0.7273429454170958,
      "grad_norm": 1.9734716987593224,
      "learning_rate": 1.047669653980572e-05,
      "loss": 0.6000909805297852,
      "step": 5650
    },
    {
      "epoch": 0.7286302780638517,
      "grad_norm": 1.3687742658460382,
      "learning_rate": 1.0002456200890902e-05,
      "loss": 0.6080599784851074,
      "step": 5660
    },
    {
      "epoch": 0.7299176107106076,
      "grad_norm": 1.3540728505304707,
      "learning_rate": 9.531005562548668e-06,
      "loss": 0.6132723808288574,
      "step": 5670
    },
    {
      "epoch": 0.7312049433573635,
      "grad_norm": 1.7082112510990102,
      "learning_rate": 9.06366306925925e-06,
      "loss": 0.605108642578125,
      "step": 5680
    },
    {
      "epoch": 0.7324922760041195,
      "grad_norm": 1.011376961087355,
      "learning_rate": 8.601735676789686e-06,
      "loss": 0.5989008903503418,
      "step": 5690
    },
    {
      "epoch": 0.7337796086508754,
      "grad_norm": 1.5444064496459442,
      "learning_rate": 8.14651519720118e-06,
      "loss": 0.6501237392425537,
      "step": 5700
    },
    {
      "epoch": 0.7350669412976313,
      "grad_norm": 1.3479539557101632,
      "learning_rate": 7.699274686206866e-06,
      "loss": 0.5925134181976318,
      "step": 5710
    },
    {
      "epoch": 0.7363542739443872,
      "grad_norm": 2.9468145741798595,
      "learning_rate": 7.261264882983024e-06,
      "loss": 0.618309497833252,
      "step": 5720
    },
    {
      "epoch": 0.7376416065911432,
      "grad_norm": 1.0247834544205183,
      "learning_rate": 6.8337107123900386e-06,
      "loss": 0.608586311340332,
      "step": 5730
    },
    {
      "epoch": 0.738928939237899,
      "grad_norm": 1.076733528232416,
      "learning_rate": 6.417807859384881e-06,
      "loss": 0.5897451400756836,
      "step": 5740
    },
    {
      "epoch": 0.740216271884655,
      "grad_norm": 1.0892502726130167,
      "learning_rate": 6.014719425205009e-06,
      "loss": 0.604542350769043,
      "step": 5750
    },
    {
      "epoch": 0.741503604531411,
      "grad_norm": 1.1117046845148453,
      "learning_rate": 5.6255726746749275e-06,
      "loss": 0.5969659805297851,
      "step": 5760
    },
    {
      "epoch": 0.7427909371781668,
      "grad_norm": 0.7747936959123887,
      "learning_rate": 5.251455883731728e-06,
      "loss": 0.5851496696472168,
      "step": 5770
    },
    {
      "epoch": 0.7440782698249228,
      "grad_norm": 1.047238462137781,
      "learning_rate": 4.893415295985812e-06,
      "loss": 0.5985543251037597,
      "step": 5780
    },
    {
      "epoch": 0.7453656024716787,
      "grad_norm": 1.276691391352667,
      "learning_rate": 4.552452196827885e-06,
      "loss": 0.6084946632385254,
      "step": 5790
    },
    {
      "epoch": 0.7466529351184346,
      "grad_norm": 1.216415546013822,
      "learning_rate": 4.229520113264785e-06,
      "loss": 0.5801431179046631,
      "step": 5800
    },
    {
      "epoch": 0.7479402677651905,
      "grad_norm": 3.072712328804095,
      "learning_rate": 3.925522147314915e-06,
      "loss": 0.5908690452575683,
      "step": 5810
    },
    {
      "epoch": 0.7492276004119465,
      "grad_norm": 1.468346657504093,
      "learning_rate": 3.641308450420675e-06,
      "loss": 0.6156291961669922,
      "step": 5820
    },
    {
      "epoch": 0.7505149330587023,
      "grad_norm": 1.3309978937821394,
      "learning_rate": 3.37767384594087e-06,
      "loss": 0.5735548973083496,
      "step": 5830
    },
    {
      "epoch": 0.7518022657054583,
      "grad_norm": 2.0653604665768674,
      "learning_rate": 3.1353556063719357e-06,
      "loss": 0.5759501934051514,
      "step": 5840
    },
    {
      "epoch": 0.7530895983522142,
      "grad_norm": 1.0813775445288938,
      "learning_rate": 2.9150313915141474e-06,
      "loss": 0.5929423332214355,
      "step": 5850
    },
    {
      "epoch": 0.7543769309989702,
      "grad_norm": 1.0945791102466997,
      "learning_rate": 2.717317353348891e-06,
      "loss": 0.5981351375579834,
      "step": 5860
    },
    {
      "epoch": 0.755664263645726,
      "grad_norm": 1.3903738685744231,
      "learning_rate": 2.5427664129268253e-06,
      "loss": 0.6201614379882813,
      "step": 5870
    },
    {
      "epoch": 0.756951596292482,
      "grad_norm": 0.858621863265503,
      "learning_rate": 2.391866714085717e-06,
      "loss": 0.5964079856872558,
      "step": 5880
    },
    {
      "epoch": 0.7582389289392379,
      "grad_norm": 0.9432664290459118,
      "learning_rate": 2.2650402583222753e-06,
      "loss": 0.613310718536377,
      "step": 5890
    },
    {
      "epoch": 0.7595262615859938,
      "grad_norm": 7.581701100617174,
      "learning_rate": 2.1626417246356398e-06,
      "loss": 0.5768440246582032,
      "step": 5900
    },
    {
      "epoch": 0.7608135942327497,
      "grad_norm": 1.0994291779942686,
      "learning_rate": 2.084957477642894e-06,
      "loss": 0.5689279079437256,
      "step": 5910
    },
    {
      "epoch": 0.7621009268795057,
      "grad_norm": 0.8746143858458795,
      "learning_rate": 2.0322047667405147e-06,
      "loss": 0.5903809547424317,
      "step": 5920
    },
    {
      "epoch": 0.7633882595262615,
      "grad_norm": 1.0230205196138777,
      "learning_rate": 2.0045311185513342e-06,
      "loss": 0.5622176647186279,
      "step": 5930
    }
  ],
  "logging_steps": 10,
  "max_steps": 5935,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2035,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2810342056919040.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}