{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 7630,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001310615989515072,
      "grad_norm": 5.242395877838135,
      "learning_rate": 9e-06,
      "loss": 10.790785217285157,
      "step": 10
    },
    {
      "epoch": 0.002621231979030144,
      "grad_norm": 2.062730550765991,
      "learning_rate": 1.9e-05,
      "loss": 9.78466339111328,
      "step": 20
    },
    {
      "epoch": 0.003931847968545216,
      "grad_norm": 1.7534589767456055,
      "learning_rate": 2.9e-05,
      "loss": 9.188233947753906,
      "step": 30
    },
    {
      "epoch": 0.005242463958060288,
      "grad_norm": 1.5791245698928833,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 8.873410034179688,
      "step": 40
    },
    {
      "epoch": 0.00655307994757536,
      "grad_norm": 1.2260531187057495,
      "learning_rate": 4.9e-05,
      "loss": 8.46854019165039,
      "step": 50
    },
    {
      "epoch": 0.007863695937090432,
      "grad_norm": 1.790460467338562,
      "learning_rate": 5.9e-05,
      "loss": 8.018362426757813,
      "step": 60
    },
    {
      "epoch": 0.009174311926605505,
      "grad_norm": 0.9511985182762146,
      "learning_rate": 6.9e-05,
      "loss": 7.632910919189453,
      "step": 70
    },
    {
      "epoch": 0.010484927916120577,
      "grad_norm": 0.7320564985275269,
      "learning_rate": 7.900000000000001e-05,
      "loss": 7.3374168395996096,
      "step": 80
    },
    {
      "epoch": 0.011795543905635648,
      "grad_norm": 0.522661566734314,
      "learning_rate": 8.900000000000001e-05,
      "loss": 7.13555908203125,
      "step": 90
    },
    {
      "epoch": 0.01310615989515072,
      "grad_norm": 0.8992032408714294,
      "learning_rate": 9.900000000000001e-05,
      "loss": 7.0307365417480465,
      "step": 100
    },
    {
      "epoch": 0.014416775884665793,
      "grad_norm": 0.855212926864624,
      "learning_rate": 0.000109,
      "loss": 6.860060119628907,
      "step": 110
    },
    {
      "epoch": 0.015727391874180863,
      "grad_norm": 0.7742177248001099,
      "learning_rate": 0.000119,
      "loss": 6.729600524902343,
      "step": 120
    },
    {
      "epoch": 0.01703800786369594,
      "grad_norm": 0.9360373616218567,
      "learning_rate": 0.00012900000000000002,
      "loss": 6.606645965576172,
      "step": 130
    },
    {
      "epoch": 0.01834862385321101,
      "grad_norm": 0.6685154438018799,
      "learning_rate": 0.000139,
      "loss": 6.490238952636719,
      "step": 140
    },
    {
      "epoch": 0.019659239842726082,
      "grad_norm": 0.872371256351471,
      "learning_rate": 0.00014900000000000002,
      "loss": 6.396331024169922,
      "step": 150
    },
    {
      "epoch": 0.020969855832241154,
      "grad_norm": 1.0915614366531372,
      "learning_rate": 0.00015900000000000002,
      "loss": 6.287831497192383,
      "step": 160
    },
    {
      "epoch": 0.022280471821756225,
      "grad_norm": 1.1108781099319458,
      "learning_rate": 0.00016900000000000002,
      "loss": 6.205494689941406,
      "step": 170
    },
    {
      "epoch": 0.023591087811271297,
      "grad_norm": 1.1550058126449585,
      "learning_rate": 0.00017900000000000001,
      "loss": 6.144885635375976,
      "step": 180
    },
    {
      "epoch": 0.02490170380078637,
      "grad_norm": 0.7530354261398315,
      "learning_rate": 0.00018899999999999999,
      "loss": 6.053673553466797,
      "step": 190
    },
    {
      "epoch": 0.02621231979030144,
      "grad_norm": 1.1098612546920776,
      "learning_rate": 0.000199,
      "loss": 5.951513671875,
      "step": 200
    },
    {
      "epoch": 0.027522935779816515,
      "grad_norm": 1.0251952409744263,
      "learning_rate": 0.00019999927593656382,
      "loss": 5.907609176635742,
      "step": 210
    },
    {
      "epoch": 0.028833551769331587,
      "grad_norm": 0.775165319442749,
      "learning_rate": 0.00019999677301469056,
      "loss": 5.847573089599609,
      "step": 220
    },
    {
      "epoch": 0.03014416775884666,
      "grad_norm": 0.8522757291793823,
      "learning_rate": 0.0001999924823400622,
      "loss": 5.793457794189453,
      "step": 230
    },
    {
      "epoch": 0.03145478374836173,
      "grad_norm": 0.7196369171142578,
      "learning_rate": 0.00019998640398938796,
      "loss": 5.722650527954102,
      "step": 240
    },
    {
      "epoch": 0.0327653997378768,
      "grad_norm": 0.6231063604354858,
      "learning_rate": 0.00019997853807133715,
      "loss": 5.683350372314453,
      "step": 250
    },
    {
      "epoch": 0.03407601572739188,
      "grad_norm": 0.8363837599754333,
      "learning_rate": 0.00019996888472653752,
      "loss": 5.643273544311524,
      "step": 260
    },
    {
      "epoch": 0.035386631716906945,
      "grad_norm": 0.9375743865966797,
      "learning_rate": 0.0001999574441275726,
      "loss": 5.602783584594727,
      "step": 270
    },
    {
      "epoch": 0.03669724770642202,
      "grad_norm": 0.7113925218582153,
      "learning_rate": 0.00019994421647897856,
      "loss": 5.550882720947266,
      "step": 280
    },
    {
      "epoch": 0.03800786369593709,
      "grad_norm": 0.7970216870307922,
      "learning_rate": 0.00019992920201724072,
      "loss": 5.500176239013672,
      "step": 290
    },
    {
      "epoch": 0.039318479685452164,
      "grad_norm": 1.1392215490341187,
      "learning_rate": 0.0001999124010107892,
      "loss": 5.474890518188476,
      "step": 300
    },
    {
      "epoch": 0.04062909567496723,
      "grad_norm": 1.0643885135650635,
      "learning_rate": 0.0001998938137599941,
      "loss": 5.416056060791016,
      "step": 310
    },
    {
      "epoch": 0.04193971166448231,
      "grad_norm": 1.3106462955474854,
      "learning_rate": 0.00019987344059716028,
      "loss": 5.401145935058594,
      "step": 320
    },
    {
      "epoch": 0.04325032765399738,
      "grad_norm": 0.5807702541351318,
      "learning_rate": 0.00019985128188652125,
      "loss": 5.387224578857422,
      "step": 330
    },
    {
      "epoch": 0.04456094364351245,
      "grad_norm": 0.6966476440429688,
      "learning_rate": 0.00019982733802423276,
      "loss": 5.356132888793946,
      "step": 340
    },
    {
      "epoch": 0.045871559633027525,
      "grad_norm": 0.690979540348053,
      "learning_rate": 0.00019980160943836568,
      "loss": 5.295703506469726,
      "step": 350
    },
    {
      "epoch": 0.047182175622542594,
      "grad_norm": 0.6240167021751404,
      "learning_rate": 0.0001997740965888984,
      "loss": 5.23779296875,
      "step": 360
    },
    {
      "epoch": 0.04849279161205767,
      "grad_norm": 0.6696836352348328,
      "learning_rate": 0.00019974479996770844,
      "loss": 5.1716056823730465,
      "step": 370
    },
    {
      "epoch": 0.04980340760157274,
      "grad_norm": 0.7365891933441162,
      "learning_rate": 0.00019971372009856388,
      "loss": 5.193817138671875,
      "step": 380
    },
    {
      "epoch": 0.05111402359108781,
      "grad_norm": 0.6361269354820251,
      "learning_rate": 0.0001996808575371139,
      "loss": 5.149017333984375,
      "step": 390
    },
    {
      "epoch": 0.05242463958060288,
      "grad_norm": 0.712479293346405,
      "learning_rate": 0.0001996462128708788,
      "loss": 5.106535339355469,
      "step": 400
    },
    {
      "epoch": 0.053735255570117955,
      "grad_norm": 0.6224299669265747,
      "learning_rate": 0.00019960978671923946,
      "loss": 5.112558746337891,
      "step": 410
    },
    {
      "epoch": 0.05504587155963303,
      "grad_norm": 0.7293247580528259,
      "learning_rate": 0.00019957157973342655,
      "loss": 5.080678939819336,
      "step": 420
    },
    {
      "epoch": 0.0563564875491481,
      "grad_norm": 0.6226844787597656,
      "learning_rate": 0.00019953159259650848,
      "loss": 5.052564239501953,
      "step": 430
    },
    {
      "epoch": 0.057667103538663174,
      "grad_norm": 0.7167950868606567,
      "learning_rate": 0.0001994898260233795,
      "loss": 4.9948577880859375,
      "step": 440
    },
    {
      "epoch": 0.05897771952817824,
      "grad_norm": 0.5642083883285522,
      "learning_rate": 0.00019944628076074672,
      "loss": 4.983065795898438,
      "step": 450
    },
    {
      "epoch": 0.06028833551769332,
      "grad_norm": 0.7142404317855835,
      "learning_rate": 0.0001994009575871169,
      "loss": 4.959147262573242,
      "step": 460
    },
    {
      "epoch": 0.061598951507208385,
      "grad_norm": 0.5722715258598328,
      "learning_rate": 0.00019935385731278255,
      "loss": 4.973221969604492,
      "step": 470
    },
    {
      "epoch": 0.06290956749672345,
      "grad_norm": 0.6105995774269104,
      "learning_rate": 0.00019930498077980722,
      "loss": 4.937568664550781,
      "step": 480
    },
    {
      "epoch": 0.06422018348623854,
      "grad_norm": 0.6237621903419495,
      "learning_rate": 0.00019925432886201068,
      "loss": 4.889268493652343,
      "step": 490
    },
    {
      "epoch": 0.0655307994757536,
      "grad_norm": 0.8787188529968262,
      "learning_rate": 0.00019920190246495322,
      "loss": 4.858717346191407,
      "step": 500
    },
    {
      "epoch": 0.0655307994757536,
      "eval_loss": 4.885393142700195,
      "eval_runtime": 1.0773,
      "eval_samples_per_second": 118.82,
      "eval_steps_per_second": 29.705,
      "step": 500
    },
    {
      "epoch": 0.06684141546526867,
      "grad_norm": 0.6522945165634155,
      "learning_rate": 0.00019914770252591946,
      "loss": 4.920278167724609,
      "step": 510
    },
    {
      "epoch": 0.06815203145478375,
      "grad_norm": 0.5460343956947327,
      "learning_rate": 0.00019909173001390152,
      "loss": 4.868672943115234,
      "step": 520
    },
    {
      "epoch": 0.06946264744429882,
      "grad_norm": 0.5967769622802734,
      "learning_rate": 0.00019903398592958186,
      "loss": 4.8205101013183596,
      "step": 530
    },
    {
      "epoch": 0.07077326343381389,
      "grad_norm": 0.5859412550926208,
      "learning_rate": 0.0001989744713053153,
      "loss": 4.790373229980469,
      "step": 540
    },
    {
      "epoch": 0.07208387942332896,
      "grad_norm": 0.516115665435791,
      "learning_rate": 0.00019891318720511038,
      "loss": 4.786814880371094,
      "step": 550
    },
    {
      "epoch": 0.07339449541284404,
      "grad_norm": 0.5131822228431702,
      "learning_rate": 0.0001988501347246108,
      "loss": 4.762861251831055,
      "step": 560
    },
    {
      "epoch": 0.07470511140235911,
      "grad_norm": 0.6577239036560059,
      "learning_rate": 0.00019878531499107526,
      "loss": 4.764720153808594,
      "step": 570
    },
    {
      "epoch": 0.07601572739187418,
      "grad_norm": 0.48921218514442444,
      "learning_rate": 0.00019871872916335778,
      "loss": 4.729512786865234,
      "step": 580
    },
    {
      "epoch": 0.07732634338138926,
      "grad_norm": 0.6154564619064331,
      "learning_rate": 0.0001986503784318868,
      "loss": 4.702789688110352,
      "step": 590
    },
    {
      "epoch": 0.07863695937090433,
      "grad_norm": 0.5709148049354553,
      "learning_rate": 0.00019858026401864374,
      "loss": 4.705944061279297,
      "step": 600
    },
    {
      "epoch": 0.0799475753604194,
      "grad_norm": 0.5497680306434631,
      "learning_rate": 0.00019850838717714147,
      "loss": 4.667179870605469,
      "step": 610
    },
    {
      "epoch": 0.08125819134993446,
      "grad_norm": 0.5805076360702515,
      "learning_rate": 0.00019843474919240166,
      "loss": 4.635091400146484,
      "step": 620
    },
    {
      "epoch": 0.08256880733944955,
      "grad_norm": 0.6312486529350281,
      "learning_rate": 0.00019835935138093187,
      "loss": 4.6185142517089846,
      "step": 630
    },
    {
      "epoch": 0.08387942332896461,
      "grad_norm": 0.6897740364074707,
      "learning_rate": 0.0001982821950907021,
      "loss": 4.612821960449219,
      "step": 640
    },
    {
      "epoch": 0.08519003931847968,
      "grad_norm": 0.7656643390655518,
      "learning_rate": 0.00019820328170112053,
      "loss": 4.646015167236328,
      "step": 650
    },
    {
      "epoch": 0.08650065530799476,
      "grad_norm": 0.5744809508323669,
      "learning_rate": 0.00019812261262300895,
      "loss": 4.638483047485352,
      "step": 660
    },
    {
      "epoch": 0.08781127129750983,
      "grad_norm": 0.5200093388557434,
      "learning_rate": 0.00019804018929857765,
      "loss": 4.617862701416016,
      "step": 670
    },
    {
      "epoch": 0.0891218872870249,
      "grad_norm": 0.5013971328735352,
      "learning_rate": 0.00019795601320139936,
      "loss": 4.57965087890625,
      "step": 680
    },
    {
      "epoch": 0.09043250327653997,
      "grad_norm": 0.5955597758293152,
      "learning_rate": 0.00019787008583638317,
      "loss": 4.556201934814453,
      "step": 690
    },
    {
      "epoch": 0.09174311926605505,
      "grad_norm": 0.5395663380622864,
      "learning_rate": 0.0001977824087397475,
      "loss": 4.503931045532227,
      "step": 700
    },
    {
      "epoch": 0.09305373525557012,
      "grad_norm": 0.4899281859397888,
      "learning_rate": 0.00019769298347899265,
      "loss": 4.521625518798828,
      "step": 710
    },
    {
      "epoch": 0.09436435124508519,
      "grad_norm": 0.6566157937049866,
      "learning_rate": 0.00019760181165287275,
      "loss": 4.531532287597656,
      "step": 720
    },
    {
      "epoch": 0.09567496723460026,
      "grad_norm": 0.519769012928009,
      "learning_rate": 0.00019750889489136725,
      "loss": 4.497549438476563,
      "step": 730
    },
    {
      "epoch": 0.09698558322411534,
      "grad_norm": 0.4846694767475128,
      "learning_rate": 0.00019741423485565167,
      "loss": 4.458761215209961,
      "step": 740
    },
    {
      "epoch": 0.0982961992136304,
      "grad_norm": 0.4592718780040741,
      "learning_rate": 0.00019731783323806808,
      "loss": 4.491049957275391,
      "step": 750
    },
    {
      "epoch": 0.09960681520314547,
      "grad_norm": 0.5449084639549255,
      "learning_rate": 0.00019721969176209468,
      "loss": 4.426659774780274,
      "step": 760
    },
    {
      "epoch": 0.10091743119266056,
      "grad_norm": 0.6002539396286011,
      "learning_rate": 0.00019711981218231497,
      "loss": 4.436170959472657,
      "step": 770
    },
    {
      "epoch": 0.10222804718217562,
      "grad_norm": 0.6023821830749512,
      "learning_rate": 0.00019701819628438653,
      "loss": 4.420704650878906,
      "step": 780
    },
    {
      "epoch": 0.10353866317169069,
      "grad_norm": 0.5346150398254395,
      "learning_rate": 0.00019691484588500903,
      "loss": 4.404710388183593,
      "step": 790
    },
    {
      "epoch": 0.10484927916120576,
      "grad_norm": 0.5418703556060791,
      "learning_rate": 0.00019680976283189167,
      "loss": 4.380912780761719,
      "step": 800
    },
    {
      "epoch": 0.10615989515072084,
      "grad_norm": 0.6761361360549927,
      "learning_rate": 0.0001967029490037203,
      "loss": 4.386772918701172,
      "step": 810
    },
    {
      "epoch": 0.10747051114023591,
      "grad_norm": 0.5387855768203735,
      "learning_rate": 0.00019659440631012367,
      "loss": 4.348167419433594,
      "step": 820
    },
    {
      "epoch": 0.10878112712975098,
      "grad_norm": 0.6417603492736816,
      "learning_rate": 0.00019648413669163937,
      "loss": 4.338081359863281,
      "step": 830
    },
    {
      "epoch": 0.11009174311926606,
      "grad_norm": 0.4582243859767914,
      "learning_rate": 0.0001963721421196792,
      "loss": 4.333487319946289,
      "step": 840
    },
    {
      "epoch": 0.11140235910878113,
      "grad_norm": 0.6252626180648804,
      "learning_rate": 0.0001962584245964938,
      "loss": 4.334003448486328,
      "step": 850
    },
    {
      "epoch": 0.1127129750982962,
      "grad_norm": 0.4850163757801056,
      "learning_rate": 0.00019614298615513694,
      "loss": 4.3195758819580075,
      "step": 860
    },
    {
      "epoch": 0.11402359108781127,
      "grad_norm": 0.5418576002120972,
      "learning_rate": 0.0001960258288594291,
      "loss": 4.299655914306641,
      "step": 870
    },
    {
      "epoch": 0.11533420707732635,
      "grad_norm": 0.5620196461677551,
      "learning_rate": 0.00019590695480392072,
      "loss": 4.279837036132813,
      "step": 880
    },
    {
      "epoch": 0.11664482306684142,
      "grad_norm": 0.5274762511253357,
      "learning_rate": 0.00019578636611385452,
      "loss": 4.282602310180664,
      "step": 890
    },
    {
      "epoch": 0.11795543905635648,
      "grad_norm": 0.6437764763832092,
      "learning_rate": 0.00019566406494512774,
      "loss": 4.253681945800781,
      "step": 900
    },
    {
      "epoch": 0.11926605504587157,
      "grad_norm": 0.536810040473938,
      "learning_rate": 0.00019554005348425345,
      "loss": 4.289502334594727,
      "step": 910
    },
    {
      "epoch": 0.12057667103538663,
      "grad_norm": 0.4443066120147705,
      "learning_rate": 0.00019541433394832143,
      "loss": 4.243859481811524,
      "step": 920
    },
    {
      "epoch": 0.1218872870249017,
      "grad_norm": 0.44103753566741943,
      "learning_rate": 0.00019528690858495876,
      "loss": 4.224325180053711,
      "step": 930
    },
    {
      "epoch": 0.12319790301441677,
      "grad_norm": 0.43676555156707764,
      "learning_rate": 0.00019515777967228943,
      "loss": 4.204811096191406,
      "step": 940
    },
    {
      "epoch": 0.12450851900393185,
      "grad_norm": 0.4667107164859772,
      "learning_rate": 0.00019502694951889352,
      "loss": 4.207023620605469,
      "step": 950
    },
    {
      "epoch": 0.1258191349934469,
      "grad_norm": 0.4878367483615875,
      "learning_rate": 0.00019489442046376627,
      "loss": 4.198031616210938,
      "step": 960
    },
    {
      "epoch": 0.127129750982962,
      "grad_norm": 0.4191912114620209,
      "learning_rate": 0.00019476019487627594,
      "loss": 4.176810073852539,
      "step": 970
    },
    {
      "epoch": 0.12844036697247707,
      "grad_norm": 0.4245511591434479,
      "learning_rate": 0.0001946242751561216,
      "loss": 4.14481201171875,
      "step": 980
    },
    {
      "epoch": 0.12975098296199214,
      "grad_norm": 0.659008264541626,
      "learning_rate": 0.0001944866637332902,
      "loss": 4.157764053344726,
      "step": 990
    },
    {
      "epoch": 0.1310615989515072,
      "grad_norm": 0.4773654043674469,
      "learning_rate": 0.00019434736306801317,
      "loss": 4.149207305908203,
      "step": 1000
    },
    {
      "epoch": 0.1310615989515072,
      "eval_loss": 4.139154434204102,
      "eval_runtime": 1.078,
      "eval_samples_per_second": 118.738,
      "eval_steps_per_second": 29.685,
      "step": 1000
    },
    {
      "epoch": 0.13237221494102228,
      "grad_norm": 0.5556576251983643,
      "learning_rate": 0.00019420637565072232,
      "loss": 4.14703369140625,
      "step": 1010
    },
    {
      "epoch": 0.13368283093053734,
      "grad_norm": 0.45232802629470825,
      "learning_rate": 0.00019406370400200547,
      "loss": 4.113636016845703,
      "step": 1020
    },
    {
      "epoch": 0.1349934469200524,
      "grad_norm": 0.43646669387817383,
      "learning_rate": 0.00019391935067256122,
      "loss": 4.12705078125,
      "step": 1030
    },
    {
      "epoch": 0.1363040629095675,
      "grad_norm": 0.44690200686454773,
      "learning_rate": 0.00019377331824315349,
      "loss": 4.117148971557617,
      "step": 1040
    },
    {
      "epoch": 0.13761467889908258,
      "grad_norm": 0.4386990964412689,
      "learning_rate": 0.00019362560932456528,
      "loss": 4.091374206542969,
      "step": 1050
    },
    {
      "epoch": 0.13892529488859764,
      "grad_norm": 0.5231616497039795,
      "learning_rate": 0.00019347622655755214,
      "loss": 4.157352447509766,
      "step": 1060
    },
    {
      "epoch": 0.1402359108781127,
      "grad_norm": 0.5323877930641174,
      "learning_rate": 0.00019332517261279471,
      "loss": 4.1090538024902346,
      "step": 1070
    },
    {
      "epoch": 0.14154652686762778,
      "grad_norm": 0.5148825645446777,
      "learning_rate": 0.00019317245019085126,
      "loss": 4.0859214782714846,
      "step": 1080
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 0.45412108302116394,
      "learning_rate": 0.00019301806202210914,
      "loss": 4.056915283203125,
      "step": 1090
    },
    {
      "epoch": 0.14416775884665792,
      "grad_norm": 0.44947361946105957,
      "learning_rate": 0.00019286201086673622,
      "loss": 4.110357284545898,
      "step": 1100
    },
    {
      "epoch": 0.145478374836173,
      "grad_norm": 0.6319316625595093,
      "learning_rate": 0.0001927042995146313,
      "loss": 4.054293823242188,
      "step": 1110
    },
    {
      "epoch": 0.14678899082568808,
      "grad_norm": 0.5883367657661438,
      "learning_rate": 0.00019254493078537448,
      "loss": 4.099805450439453,
      "step": 1120
    },
    {
      "epoch": 0.14809960681520315,
      "grad_norm": 0.4896838366985321,
      "learning_rate": 0.0001923839075281765,
      "loss": 4.092933654785156,
      "step": 1130
    },
    {
      "epoch": 0.14941022280471822,
      "grad_norm": 0.4674643576145172,
      "learning_rate": 0.00019222123262182798,
      "loss": 4.056755828857422,
      "step": 1140
    },
    {
      "epoch": 0.15072083879423329,
      "grad_norm": 0.46807295083999634,
      "learning_rate": 0.00019205690897464793,
      "loss": 4.0553325653076175,
      "step": 1150
    },
    {
      "epoch": 0.15203145478374835,
      "grad_norm": 0.42275893688201904,
      "learning_rate": 0.0001918909395244316,
      "loss": 4.0182861328125,
      "step": 1160
    },
    {
      "epoch": 0.15334207077326342,
      "grad_norm": 0.4824959933757782,
      "learning_rate": 0.00019172332723839825,
      "loss": 4.008975219726563,
      "step": 1170
    },
    {
      "epoch": 0.15465268676277852,
      "grad_norm": 0.39416125416755676,
      "learning_rate": 0.00019155407511313777,
      "loss": 4.009979629516602,
      "step": 1180
    },
    {
      "epoch": 0.1559633027522936,
      "grad_norm": 0.4873785674571991,
      "learning_rate": 0.00019138318617455743,
      "loss": 4.049541854858399,
      "step": 1190
    },
    {
      "epoch": 0.15727391874180865,
      "grad_norm": 0.5093808770179749,
      "learning_rate": 0.0001912106634778274,
      "loss": 4.0228271484375,
      "step": 1200
    },
    {
      "epoch": 0.15858453473132372,
      "grad_norm": 0.4115995764732361,
      "learning_rate": 0.00019103651010732658,
      "loss": 3.9854972839355467,
      "step": 1210
    },
    {
      "epoch": 0.1598951507208388,
      "grad_norm": 0.4606524109840393,
      "learning_rate": 0.00019086072917658706,
      "loss": 3.9953903198242187,
      "step": 1220
    },
    {
      "epoch": 0.16120576671035386,
      "grad_norm": 0.4111475348472595,
      "learning_rate": 0.00019068332382823875,
      "loss": 3.974542999267578,
      "step": 1230
    },
    {
      "epoch": 0.16251638269986893,
      "grad_norm": 0.42832401394844055,
      "learning_rate": 0.00019050429723395292,
      "loss": 3.989231491088867,
      "step": 1240
    },
    {
      "epoch": 0.16382699868938402,
      "grad_norm": 0.40471911430358887,
      "learning_rate": 0.0001903236525943858,
      "loss": 3.9893341064453125,
      "step": 1250
    },
    {
      "epoch": 0.1651376146788991,
      "grad_norm": 0.42500078678131104,
      "learning_rate": 0.00019014139313912112,
      "loss": 3.974134826660156,
      "step": 1260
    },
    {
      "epoch": 0.16644823066841416,
      "grad_norm": 0.3931633234024048,
      "learning_rate": 0.0001899575221266124,
      "loss": 3.942657470703125,
      "step": 1270
    },
    {
      "epoch": 0.16775884665792923,
      "grad_norm": 0.39109498262405396,
      "learning_rate": 0.00018977204284412487,
      "loss": 3.9631961822509765,
      "step": 1280
    },
    {
      "epoch": 0.1690694626474443,
      "grad_norm": 0.3855876624584198,
      "learning_rate": 0.00018958495860767652,
      "loss": 3.9366249084472655,
      "step": 1290
    },
    {
      "epoch": 0.17038007863695936,
      "grad_norm": 0.4450657069683075,
      "learning_rate": 0.00018939627276197882,
      "loss": 3.948311996459961,
      "step": 1300
    },
    {
      "epoch": 0.17169069462647443,
      "grad_norm": 0.48241353034973145,
      "learning_rate": 0.00018920598868037709,
      "loss": 3.931996154785156,
      "step": 1310
    },
    {
      "epoch": 0.17300131061598953,
      "grad_norm": 0.412693589925766,
      "learning_rate": 0.00018901410976479,
      "loss": 3.929762268066406,
      "step": 1320
    },
    {
      "epoch": 0.1743119266055046,
      "grad_norm": 0.4669123589992523,
      "learning_rate": 0.00018882063944564883,
      "loss": 3.919912338256836,
      "step": 1330
    },
    {
      "epoch": 0.17562254259501967,
      "grad_norm": 0.7486332654953003,
      "learning_rate": 0.0001886255811818362,
      "loss": 3.9282821655273437,
      "step": 1340
    },
    {
      "epoch": 0.17693315858453473,
      "grad_norm": 0.6442509293556213,
      "learning_rate": 0.0001884289384606241,
      "loss": 3.916644287109375,
      "step": 1350
    },
    {
      "epoch": 0.1782437745740498,
      "grad_norm": 0.6581181883811951,
      "learning_rate": 0.00018823071479761168,
      "loss": 3.9413246154785155,
      "step": 1360
    },
    {
      "epoch": 0.17955439056356487,
      "grad_norm": 0.41414180397987366,
      "learning_rate": 0.0001880309137366622,
      "loss": 3.9043121337890625,
      "step": 1370
    },
    {
      "epoch": 0.18086500655307994,
      "grad_norm": 0.38383692502975464,
      "learning_rate": 0.00018782953884984003,
      "loss": 3.898161697387695,
      "step": 1380
    },
    {
      "epoch": 0.182175622542595,
      "grad_norm": 0.3519379496574402,
      "learning_rate": 0.00018762659373734635,
      "loss": 3.8826065063476562,
      "step": 1390
    },
    {
      "epoch": 0.1834862385321101,
      "grad_norm": 0.4244498908519745,
      "learning_rate": 0.00018742208202745514,
      "loss": 3.8820899963378905,
      "step": 1400
    },
    {
      "epoch": 0.18479685452162517,
      "grad_norm": 0.4080543518066406,
      "learning_rate": 0.00018721600737644804,
      "loss": 3.8755558013916014,
      "step": 1410
    },
    {
      "epoch": 0.18610747051114024,
      "grad_norm": 0.4284413754940033,
      "learning_rate": 0.00018700837346854923,
      "loss": 3.88403434753418,
      "step": 1420
    },
    {
      "epoch": 0.1874180865006553,
      "grad_norm": 0.4326043426990509,
      "learning_rate": 0.00018679918401585943,
      "loss": 3.867059326171875,
      "step": 1430
    },
    {
      "epoch": 0.18872870249017037,
      "grad_norm": 0.5231949090957642,
      "learning_rate": 0.00018658844275828951,
      "loss": 3.846060943603516,
      "step": 1440
    },
    {
      "epoch": 0.19003931847968544,
      "grad_norm": 0.39231449365615845,
      "learning_rate": 0.0001863761534634937,
      "loss": 3.872735595703125,
      "step": 1450
    },
    {
      "epoch": 0.1913499344692005,
      "grad_norm": 0.3590643107891083,
      "learning_rate": 0.0001861623199268022,
      "loss": 3.8776809692382814,
      "step": 1460
    },
    {
      "epoch": 0.1926605504587156,
      "grad_norm": 0.3989343047142029,
      "learning_rate": 0.00018594694597115328,
      "loss": 3.859600830078125,
      "step": 1470
    },
    {
      "epoch": 0.19397116644823068,
      "grad_norm": 0.41816872358322144,
      "learning_rate": 0.000185730035447025,
      "loss": 3.8512451171875,
      "step": 1480
    },
    {
      "epoch": 0.19528178243774574,
      "grad_norm": 0.3651232123374939,
      "learning_rate": 0.00018551159223236647,
      "loss": 3.848764419555664,
      "step": 1490
    },
    {
      "epoch": 0.1965923984272608,
      "grad_norm": 0.5443365573883057,
      "learning_rate": 0.00018529162023252823,
      "loss": 3.859080505371094,
      "step": 1500
    },
    {
      "epoch": 0.1965923984272608,
      "eval_loss": 3.845022201538086,
      "eval_runtime": 1.0753,
      "eval_samples_per_second": 119.032,
      "eval_steps_per_second": 29.758,
      "step": 1500
    },
    {
      "epoch": 0.19790301441677588,
      "grad_norm": 0.3784181773662567,
      "learning_rate": 0.00018507012338019272,
      "loss": 3.8517822265625,
      "step": 1510
    },
    {
      "epoch": 0.19921363040629095,
      "grad_norm": 0.37363308668136597,
      "learning_rate": 0.00018484710563530373,
      "loss": 3.845747375488281,
      "step": 1520
    },
    {
      "epoch": 0.20052424639580602,
      "grad_norm": 0.39282530546188354,
      "learning_rate": 0.00018462257098499587,
      "loss": 3.848657989501953,
      "step": 1530
    },
    {
      "epoch": 0.2018348623853211,
      "grad_norm": 0.3724451959133148,
      "learning_rate": 0.00018439652344352306,
      "loss": 3.8459407806396486,
      "step": 1540
    },
    {
      "epoch": 0.20314547837483618,
      "grad_norm": 0.40862613916397095,
      "learning_rate": 0.00018416896705218688,
      "loss": 3.8338302612304687,
      "step": 1550
    },
    {
      "epoch": 0.20445609436435125,
      "grad_norm": 0.35828253626823425,
      "learning_rate": 0.00018393990587926422,
      "loss": 3.8235458374023437,
      "step": 1560
    },
    {
      "epoch": 0.20576671035386632,
      "grad_norm": 0.48027393221855164,
      "learning_rate": 0.0001837093440199347,
      "loss": 3.8863067626953125,
      "step": 1570
    },
    {
      "epoch": 0.20707732634338138,
      "grad_norm": 0.40401071310043335,
      "learning_rate": 0.0001834772855962073,
      "loss": 3.8339820861816407,
      "step": 1580
    },
    {
      "epoch": 0.20838794233289645,
      "grad_norm": 0.4017929434776306,
      "learning_rate": 0.0001832437347568468,
      "loss": 3.8291915893554687,
      "step": 1590
    },
    {
      "epoch": 0.20969855832241152,
      "grad_norm": 0.4115326404571533,
      "learning_rate": 0.00018300869567729948,
      "loss": 3.819872283935547,
      "step": 1600
    },
    {
      "epoch": 0.21100917431192662,
      "grad_norm": 0.35475432872772217,
      "learning_rate": 0.00018277217255961858,
      "loss": 3.802693176269531,
      "step": 1610
    },
    {
      "epoch": 0.21231979030144169,
      "grad_norm": 0.3963567614555359,
      "learning_rate": 0.0001825341696323891,
      "loss": 3.821527862548828,
      "step": 1620
    },
    {
      "epoch": 0.21363040629095675,
      "grad_norm": 0.37121492624282837,
      "learning_rate": 0.00018229469115065223,
      "loss": 3.800576019287109,
      "step": 1630
    },
    {
      "epoch": 0.21494102228047182,
      "grad_norm": 0.39280757308006287,
      "learning_rate": 0.00018205374139582928,
      "loss": 3.8270397186279297,
      "step": 1640
    },
    {
      "epoch": 0.2162516382699869,
      "grad_norm": 0.43435242772102356,
      "learning_rate": 0.00018181132467564513,
      "loss": 3.809955596923828,
      "step": 1650
    },
    {
      "epoch": 0.21756225425950196,
      "grad_norm": 0.46009910106658936,
      "learning_rate": 0.00018156744532405128,
      "loss": 3.8037109375,
      "step": 1660
    },
    {
      "epoch": 0.21887287024901703,
      "grad_norm": 0.4487770199775696,
      "learning_rate": 0.00018132210770114822,
      "loss": 3.7775798797607423,
      "step": 1670
    },
    {
      "epoch": 0.22018348623853212,
      "grad_norm": 0.34337013959884644,
      "learning_rate": 0.0001810753161931076,
      "loss": 3.795824432373047,
      "step": 1680
    },
    {
      "epoch": 0.2214941022280472,
      "grad_norm": 0.4121147394180298,
      "learning_rate": 0.0001808270752120938,
      "loss": 3.7986644744873046,
      "step": 1690
    },
    {
      "epoch": 0.22280471821756226,
      "grad_norm": 0.4773680567741394,
      "learning_rate": 0.000180577389196185,
      "loss": 3.803488922119141,
      "step": 1700
    },
    {
      "epoch": 0.22411533420707733,
      "grad_norm": 0.5654820799827576,
      "learning_rate": 0.0001803262626092939,
      "loss": 3.7989051818847654,
      "step": 1710
    },
    {
      "epoch": 0.2254259501965924,
      "grad_norm": 0.3598356544971466,
      "learning_rate": 0.00018007369994108783,
      "loss": 3.806232452392578,
      "step": 1720
    },
    {
      "epoch": 0.22673656618610746,
      "grad_norm": 0.3509199321269989,
      "learning_rate": 0.00017981970570690858,
      "loss": 3.7837455749511717,
      "step": 1730
    },
    {
      "epoch": 0.22804718217562253,
      "grad_norm": 0.39005377888679504,
      "learning_rate": 0.0001795642844476916,
      "loss": 3.7680877685546874,
      "step": 1740
    },
    {
      "epoch": 0.22935779816513763,
      "grad_norm": 0.37989184260368347,
      "learning_rate": 0.00017930744072988482,
      "loss": 3.768391418457031,
      "step": 1750
    },
    {
      "epoch": 0.2306684141546527,
      "grad_norm": 0.4358330965042114,
      "learning_rate": 0.0001790491791453671,
      "loss": 3.8231929779052733,
      "step": 1760
    },
    {
      "epoch": 0.23197903014416776,
      "grad_norm": 0.38792684674263,
      "learning_rate": 0.000178789504311366,
      "loss": 3.7802581787109375,
      "step": 1770
    },
    {
      "epoch": 0.23328964613368283,
      "grad_norm": 0.435153603553772,
      "learning_rate": 0.00017852842087037534,
      "loss": 3.7689659118652346,
      "step": 1780
    },
    {
      "epoch": 0.2346002621231979,
      "grad_norm": 0.3951868414878845,
      "learning_rate": 0.0001782659334900722,
      "loss": 3.791453552246094,
      "step": 1790
    },
    {
      "epoch": 0.23591087811271297,
      "grad_norm": 0.42941734194755554,
      "learning_rate": 0.00017800204686323329,
      "loss": 3.7726814270019533,
      "step": 1800
    },
    {
      "epoch": 0.23722149410222804,
      "grad_norm": 0.3489938974380493,
      "learning_rate": 0.0001777367657076514,
      "loss": 3.7656307220458984,
      "step": 1810
    },
    {
      "epoch": 0.23853211009174313,
      "grad_norm": 0.35292667150497437,
      "learning_rate": 0.00017747009476605069,
      "loss": 3.757402801513672,
      "step": 1820
    },
    {
      "epoch": 0.2398427260812582,
      "grad_norm": 0.39420491456985474,
      "learning_rate": 0.0001772020388060021,
      "loss": 3.7317531585693358,
      "step": 1830
    },
    {
      "epoch": 0.24115334207077327,
      "grad_norm": 0.41197261214256287,
      "learning_rate": 0.0001769326026198381,
      "loss": 3.7472084045410154,
      "step": 1840
    },
    {
      "epoch": 0.24246395806028834,
      "grad_norm": 0.3579365909099579,
      "learning_rate": 0.000176661791024567,
      "loss": 3.749358367919922,
      "step": 1850
    },
    {
      "epoch": 0.2437745740498034,
      "grad_norm": 0.3467675447463989,
      "learning_rate": 0.00017638960886178678,
      "loss": 3.7561313629150392,
      "step": 1860
    },
    {
      "epoch": 0.24508519003931847,
      "grad_norm": 0.33683568239212036,
      "learning_rate": 0.00017611606099759855,
      "loss": 3.718791198730469,
      "step": 1870
    },
    {
      "epoch": 0.24639580602883354,
      "grad_norm": 0.36595526337623596,
      "learning_rate": 0.00017584115232251968,
      "loss": 3.7227977752685546,
      "step": 1880
    },
    {
      "epoch": 0.24770642201834864,
      "grad_norm": 0.37135109305381775,
      "learning_rate": 0.0001755648877513961,
      "loss": 3.699530029296875,
      "step": 1890
    },
    {
      "epoch": 0.2490170380078637,
      "grad_norm": 0.38147202134132385,
      "learning_rate": 0.0001752872722233147,
      "loss": 3.7416641235351564,
      "step": 1900
    },
    {
      "epoch": 0.2503276539973788,
      "grad_norm": 0.35841041803359985,
      "learning_rate": 0.000175008310701515,
      "loss": 3.7277050018310547,
      "step": 1910
    },
    {
      "epoch": 0.2516382699868938,
      "grad_norm": 0.4510025680065155,
      "learning_rate": 0.00017472800817330003,
      "loss": 3.745488739013672,
      "step": 1920
    },
    {
      "epoch": 0.2529488859764089,
      "grad_norm": 0.3533204197883606,
      "learning_rate": 0.00017444636964994783,
      "loss": 3.7404788970947265,
      "step": 1930
    },
    {
      "epoch": 0.254259501965924,
      "grad_norm": 0.36999598145484924,
      "learning_rate": 0.0001741634001666213,
      "loss": 3.7083168029785156,
      "step": 1940
    },
    {
      "epoch": 0.25557011795543905,
      "grad_norm": 0.39530521631240845,
      "learning_rate": 0.00017387910478227841,
      "loss": 3.7123844146728517,
      "step": 1950
    },
    {
      "epoch": 0.25688073394495414,
      "grad_norm": 0.3846464455127716,
      "learning_rate": 0.00017359348857958176,
      "loss": 3.7118610382080077,
      "step": 1960
    },
    {
      "epoch": 0.2581913499344692,
      "grad_norm": 0.37135645747184753,
      "learning_rate": 0.00017330655666480767,
      "loss": 3.7165260314941406,
      "step": 1970
    },
    {
      "epoch": 0.2595019659239843,
      "grad_norm": 0.35037359595298767,
      "learning_rate": 0.00017301831416775487,
      "loss": 3.706393814086914,
      "step": 1980
    },
    {
      "epoch": 0.2608125819134993,
      "grad_norm": 0.3349663019180298,
      "learning_rate": 0.00017272876624165284,
      "loss": 3.6959980010986326,
      "step": 1990
    },
    {
      "epoch": 0.2621231979030144,
      "grad_norm": 0.3627758026123047,
      "learning_rate": 0.00017243791806306966,
      "loss": 3.7695777893066404,
      "step": 2000
    },
    {
      "epoch": 0.2621231979030144,
      "eval_loss": 3.678098678588867,
      "eval_runtime": 1.0771,
      "eval_samples_per_second": 118.835,
      "eval_steps_per_second": 29.709,
      "step": 2000
    },
    {
      "epoch": 0.2634338138925295,
      "grad_norm": 0.31398406624794006,
      "learning_rate": 0.0001721457748318194,
      "loss": 3.705975341796875,
      "step": 2010
    },
    {
      "epoch": 0.26474442988204455,
      "grad_norm": 0.33383071422576904,
      "learning_rate": 0.00017185234177086937,
      "loss": 3.715074920654297,
      "step": 2020
    },
    {
      "epoch": 0.26605504587155965,
      "grad_norm": 0.33901411294937134,
      "learning_rate": 0.00017155762412624645,
      "loss": 3.682500457763672,
      "step": 2030
    },
    {
      "epoch": 0.2673656618610747,
      "grad_norm": 0.35773271322250366,
      "learning_rate": 0.00017126162716694346,
      "loss": 3.70422477722168,
      "step": 2040
    },
    {
      "epoch": 0.2686762778505898,
      "grad_norm": 0.37591323256492615,
      "learning_rate": 0.00017096435618482495,
      "loss": 3.6637428283691404,
      "step": 2050
    },
    {
      "epoch": 0.2699868938401048,
      "grad_norm": 0.36375704407691956,
      "learning_rate": 0.00017066581649453265,
      "loss": 3.69287109375,
      "step": 2060
    },
    {
      "epoch": 0.2712975098296199,
      "grad_norm": 0.3371437191963196,
      "learning_rate": 0.00017036601343339025,
      "loss": 3.6953739166259765,
      "step": 2070
    },
    {
      "epoch": 0.272608125819135,
      "grad_norm": 0.4015010893344879,
      "learning_rate": 0.0001700649523613083,
      "loss": 3.684417724609375,
      "step": 2080
    },
    {
      "epoch": 0.27391874180865006,
      "grad_norm": 0.35754695534706116,
      "learning_rate": 0.000169762638660688,
      "loss": 3.6570167541503906,
      "step": 2090
    },
    {
      "epoch": 0.27522935779816515,
      "grad_norm": 0.36966967582702637,
      "learning_rate": 0.00016945907773632536,
      "loss": 3.6873245239257812,
      "step": 2100
    },
    {
      "epoch": 0.2765399737876802,
      "grad_norm": 0.3702266812324524,
      "learning_rate": 0.00016915427501531423,
      "loss": 3.6908008575439455,
      "step": 2110
    },
    {
      "epoch": 0.2778505897771953,
      "grad_norm": 0.3939582407474518,
      "learning_rate": 0.00016884823594694957,
      "loss": 3.6503292083740235,
      "step": 2120
    },
    {
      "epoch": 0.27916120576671033,
      "grad_norm": 0.3116343319416046,
      "learning_rate": 0.00016854096600262976,
      "loss": 3.661838150024414,
      "step": 2130
    },
    {
      "epoch": 0.2804718217562254,
      "grad_norm": 0.3592735826969147,
      "learning_rate": 0.00016823247067575899,
      "loss": 3.6377189636230467,
      "step": 2140
    },
    {
      "epoch": 0.2817824377457405,
      "grad_norm": 0.3495638370513916,
      "learning_rate": 0.00016792275548164893,
      "loss": 3.68336181640625,
      "step": 2150
    },
    {
      "epoch": 0.28309305373525556,
      "grad_norm": 0.3648313879966736,
      "learning_rate": 0.0001676118259574202,
      "loss": 3.6855892181396483,
      "step": 2160
    },
    {
      "epoch": 0.28440366972477066,
      "grad_norm": 0.3525349795818329,
      "learning_rate": 0.00016729968766190332,
      "loss": 3.6899765014648436,
      "step": 2170
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.35812753438949585,
      "learning_rate": 0.00016698634617553931,
      "loss": 3.629052734375,
      "step": 2180
    },
    {
      "epoch": 0.2870249017038008,
      "grad_norm": 0.34052297472953796,
      "learning_rate": 0.00016667180710028005,
      "loss": 3.607305908203125,
      "step": 2190
    },
    {
      "epoch": 0.28833551769331583,
      "grad_norm": 0.3272421061992645,
      "learning_rate": 0.00016635607605948794,
      "loss": 3.661695861816406,
      "step": 2200
    },
    {
      "epoch": 0.28964613368283093,
      "grad_norm": 0.3278498947620392,
      "learning_rate": 0.00016603915869783554,
      "loss": 3.6350292205810546,
      "step": 2210
    },
    {
      "epoch": 0.290956749672346,
      "grad_norm": 0.3760678470134735,
      "learning_rate": 0.00016572106068120452,
      "loss": 3.642927551269531,
      "step": 2220
    },
    {
      "epoch": 0.29226736566186107,
      "grad_norm": 0.3561151921749115,
      "learning_rate": 0.00016540178769658455,
      "loss": 3.6332794189453126,
      "step": 2230
    },
    {
      "epoch": 0.29357798165137616,
      "grad_norm": 0.3576511740684509,
      "learning_rate": 0.00016508134545197126,
      "loss": 3.649431610107422,
      "step": 2240
    },
    {
      "epoch": 0.2948885976408912,
      "grad_norm": 0.3461667001247406,
      "learning_rate": 0.00016475973967626468,
      "loss": 3.62464599609375,
      "step": 2250
    },
    {
      "epoch": 0.2961992136304063,
      "grad_norm": 0.3600212633609772,
      "learning_rate": 0.00016443697611916646,
      "loss": 3.6348335266113283,
      "step": 2260
    },
    {
      "epoch": 0.29750982961992134,
      "grad_norm": 0.37926310300827026,
      "learning_rate": 0.00016411306055107714,
      "loss": 3.6472763061523437,
      "step": 2270
    },
    {
      "epoch": 0.29882044560943644,
      "grad_norm": 0.36611098051071167,
      "learning_rate": 0.00016378799876299316,
      "loss": 3.628095245361328,
      "step": 2280
    },
    {
      "epoch": 0.30013106159895153,
      "grad_norm": 0.32461971044540405,
      "learning_rate": 0.00016346179656640306,
      "loss": 3.647011566162109,
      "step": 2290
    },
    {
      "epoch": 0.30144167758846657,
      "grad_norm": 0.34934699535369873,
      "learning_rate": 0.00016313445979318383,
      "loss": 3.632682037353516,
      "step": 2300
    },
    {
      "epoch": 0.30275229357798167,
      "grad_norm": 0.4701174795627594,
      "learning_rate": 0.00016280599429549647,
      "loss": 3.620447540283203,
      "step": 2310
    },
    {
      "epoch": 0.3040629095674967,
      "grad_norm": 0.3500426709651947,
      "learning_rate": 0.0001624764059456815,
      "loss": 3.6313808441162108,
      "step": 2320
    },
    {
      "epoch": 0.3053735255570118,
      "grad_norm": 0.3685113787651062,
      "learning_rate": 0.0001621457006361539,
      "loss": 3.6274288177490233,
      "step": 2330
    },
    {
      "epoch": 0.30668414154652685,
      "grad_norm": 0.3543446958065033,
      "learning_rate": 0.00016181388427929762,
      "loss": 3.6003509521484376,
      "step": 2340
    },
    {
      "epoch": 0.30799475753604194,
      "grad_norm": 0.3971234858036041,
      "learning_rate": 0.00016148096280736027,
      "loss": 3.610228729248047,
      "step": 2350
    },
    {
      "epoch": 0.30930537352555704,
      "grad_norm": 0.387471079826355,
      "learning_rate": 0.00016114694217234668,
      "loss": 3.6144920349121095,
      "step": 2360
    },
    {
      "epoch": 0.3106159895150721,
      "grad_norm": 0.3746248483657837,
      "learning_rate": 0.00016081182834591263,
      "loss": 3.605546569824219,
      "step": 2370
    },
    {
      "epoch": 0.3119266055045872,
      "grad_norm": 0.3339202404022217,
      "learning_rate": 0.00016047562731925816,
      "loss": 3.6020408630371095,
      "step": 2380
    },
    {
      "epoch": 0.3132372214941022,
      "grad_norm": 0.39812102913856506,
      "learning_rate": 0.00016013834510302032,
      "loss": 3.635127639770508,
      "step": 2390
    },
    {
      "epoch": 0.3145478374836173,
      "grad_norm": 0.3335367739200592,
      "learning_rate": 0.0001597999877271658,
      "loss": 3.6050186157226562,
      "step": 2400
    },
    {
      "epoch": 0.31585845347313235,
      "grad_norm": 0.35353904962539673,
      "learning_rate": 0.00015946056124088316,
      "loss": 3.584572601318359,
      "step": 2410
    },
    {
      "epoch": 0.31716906946264745,
      "grad_norm": 0.34234654903411865,
      "learning_rate": 0.00015912007171247454,
      "loss": 3.606672668457031,
      "step": 2420
    },
    {
      "epoch": 0.31847968545216254,
      "grad_norm": 0.343987375497818,
      "learning_rate": 0.00015877852522924732,
      "loss": 3.586076354980469,
      "step": 2430
    },
    {
      "epoch": 0.3197903014416776,
      "grad_norm": 0.34024274349212646,
      "learning_rate": 0.0001584359278974052,
      "loss": 3.593122100830078,
      "step": 2440
    },
    {
      "epoch": 0.3211009174311927,
      "grad_norm": 0.379170686006546,
      "learning_rate": 0.0001580922858419391,
      "loss": 3.5916927337646483,
      "step": 2450
    },
    {
      "epoch": 0.3224115334207077,
      "grad_norm": 0.4286021590232849,
      "learning_rate": 0.00015774760520651757,
      "loss": 3.6012496948242188,
      "step": 2460
    },
    {
      "epoch": 0.3237221494102228,
      "grad_norm": 0.3811999559402466,
      "learning_rate": 0.000157401892153377,
      "loss": 3.5763626098632812,
      "step": 2470
    },
    {
      "epoch": 0.32503276539973786,
      "grad_norm": 0.33412134647369385,
      "learning_rate": 0.00015705515286321148,
      "loss": 3.582170104980469,
      "step": 2480
    },
    {
      "epoch": 0.32634338138925295,
      "grad_norm": 0.3218776285648346,
      "learning_rate": 0.0001567073935350623,
      "loss": 3.5959442138671873,
      "step": 2490
    },
    {
      "epoch": 0.32765399737876805,
      "grad_norm": 0.3195808231830597,
      "learning_rate": 0.00015635862038620703,
      "loss": 3.5555511474609376,
      "step": 2500
    },
    {
      "epoch": 0.32765399737876805,
      "eval_loss": 3.5740904808044434,
      "eval_runtime": 1.0751,
      "eval_samples_per_second": 119.06,
      "eval_steps_per_second": 29.765,
      "step": 2500
    },
    {
      "epoch": 0.3289646133682831,
      "grad_norm": 0.3520723879337311,
      "learning_rate": 0.00015600883965204845,
      "loss": 3.563397979736328,
      "step": 2510
    },
    {
      "epoch": 0.3302752293577982,
      "grad_norm": 0.33916330337524414,
      "learning_rate": 0.00015565805758600308,
      "loss": 3.5895103454589843,
      "step": 2520
    },
    {
      "epoch": 0.3315858453473132,
      "grad_norm": 0.6582748293876648,
      "learning_rate": 0.0001553062804593894,
      "loss": 3.5820648193359377,
      "step": 2530
    },
    {
      "epoch": 0.3328964613368283,
      "grad_norm": 0.36067548394203186,
      "learning_rate": 0.00015495351456131556,
      "loss": 3.5789104461669923,
      "step": 2540
    },
    {
      "epoch": 0.33420707732634336,
      "grad_norm": 0.36550480127334595,
      "learning_rate": 0.00015459976619856716,
      "loss": 3.551074981689453,
      "step": 2550
    },
    {
      "epoch": 0.33551769331585846,
      "grad_norm": 0.35055071115493774,
      "learning_rate": 0.00015424504169549437,
      "loss": 3.5598285675048826,
      "step": 2560
    },
    {
      "epoch": 0.33682830930537355,
      "grad_norm": 0.3762514889240265,
      "learning_rate": 0.00015388934739389896,
      "loss": 3.590013122558594,
      "step": 2570
    },
    {
      "epoch": 0.3381389252948886,
      "grad_norm": 0.36996179819107056,
      "learning_rate": 0.00015353268965292077,
      "loss": 3.5537513732910155,
      "step": 2580
    },
    {
      "epoch": 0.3394495412844037,
      "grad_norm": 0.3500744700431824,
      "learning_rate": 0.0001531750748489242,
      "loss": 3.560906982421875,
      "step": 2590
    },
    {
      "epoch": 0.34076015727391873,
      "grad_norm": 0.3688618242740631,
      "learning_rate": 0.00015281650937538406,
      "loss": 3.534657287597656,
      "step": 2600
    },
    {
      "epoch": 0.3420707732634338,
      "grad_norm": 0.3809558153152466,
      "learning_rate": 0.00015245699964277132,
      "loss": 3.5656692504882814,
      "step": 2610
    },
    {
      "epoch": 0.34338138925294887,
      "grad_norm": 0.33842557668685913,
      "learning_rate": 0.00015209655207843855,
      "loss": 3.559683990478516,
      "step": 2620
    },
    {
      "epoch": 0.34469200524246396,
      "grad_norm": 0.36254531145095825,
      "learning_rate": 0.000151735173126505,
      "loss": 3.5632965087890627,
      "step": 2630
    },
    {
      "epoch": 0.34600262123197906,
      "grad_norm": 0.31712251901626587,
      "learning_rate": 0.0001513728692477413,
      "loss": 3.529654693603516,
      "step": 2640
    },
    {
      "epoch": 0.3473132372214941,
      "grad_norm": 0.32539448142051697,
      "learning_rate": 0.000151009646919454,
      "loss": 3.518568420410156,
      "step": 2650
    },
    {
      "epoch": 0.3486238532110092,
      "grad_norm": 0.32836371660232544,
      "learning_rate": 0.0001506455126353699,
      "loss": 3.5583621978759767,
      "step": 2660
    },
    {
      "epoch": 0.34993446920052423,
      "grad_norm": 0.32125183939933777,
      "learning_rate": 0.0001502804729055198,
      "loss": 3.5487037658691407,
      "step": 2670
    },
    {
      "epoch": 0.35124508519003933,
      "grad_norm": 0.3169119656085968,
      "learning_rate": 0.000149914534256122,
      "loss": 3.5633468627929688,
      "step": 2680
    },
    {
      "epoch": 0.35255570117955437,
      "grad_norm": 0.3467177450656891,
      "learning_rate": 0.00014954770322946602,
      "loss": 3.541538619995117,
      "step": 2690
    },
    {
      "epoch": 0.35386631716906947,
      "grad_norm": 0.33007171750068665,
      "learning_rate": 0.00014917998638379524,
      "loss": 3.571522521972656,
      "step": 2700
    },
    {
      "epoch": 0.35517693315858456,
      "grad_norm": 0.3987392783164978,
      "learning_rate": 0.00014881139029318985,
      "loss": 3.5742645263671875,
      "step": 2710
    },
    {
      "epoch": 0.3564875491480996,
      "grad_norm": 0.3675360381603241,
      "learning_rate": 0.00014844192154744927,
      "loss": 3.5792030334472655,
      "step": 2720
    },
    {
      "epoch": 0.3577981651376147,
      "grad_norm": 0.3798048496246338,
      "learning_rate": 0.00014807158675197432,
      "loss": 3.552878570556641,
      "step": 2730
    },
    {
      "epoch": 0.35910878112712974,
      "grad_norm": 0.3176024854183197,
      "learning_rate": 0.00014770039252764915,
      "loss": 3.57054443359375,
      "step": 2740
    },
    {
      "epoch": 0.36041939711664484,
      "grad_norm": 0.3383607864379883,
      "learning_rate": 0.00014732834551072296,
      "loss": 3.53015251159668,
      "step": 2750
    },
    {
      "epoch": 0.3617300131061599,
      "grad_norm": 0.31845852732658386,
      "learning_rate": 0.00014695545235269108,
      "loss": 3.546007537841797,
      "step": 2760
    },
    {
      "epoch": 0.36304062909567497,
      "grad_norm": 0.3677367866039276,
      "learning_rate": 0.00014658171972017648,
      "loss": 3.5461830139160155,
      "step": 2770
    },
    {
      "epoch": 0.36435124508519,
      "grad_norm": 0.351630836725235,
      "learning_rate": 0.00014620715429481003,
      "loss": 3.539386749267578,
      "step": 2780
    },
    {
      "epoch": 0.3656618610747051,
      "grad_norm": 0.31386688351631165,
      "learning_rate": 0.0001458317627731117,
      "loss": 3.5225189208984373,
      "step": 2790
    },
    {
      "epoch": 0.3669724770642202,
      "grad_norm": 0.3266359865665436,
      "learning_rate": 0.00014545555186637027,
      "loss": 3.542737579345703,
      "step": 2800
    },
    {
      "epoch": 0.36828309305373524,
      "grad_norm": 0.3688671588897705,
      "learning_rate": 0.00014507852830052362,
      "loss": 3.549066162109375,
      "step": 2810
    },
    {
      "epoch": 0.36959370904325034,
      "grad_norm": 0.33406293392181396,
      "learning_rate": 0.00014470069881603848,
      "loss": 3.519769287109375,
      "step": 2820
    },
    {
      "epoch": 0.3709043250327654,
      "grad_norm": 0.30831819772720337,
      "learning_rate": 0.0001443220701677898,
      "loss": 3.5477012634277343,
      "step": 2830
    },
    {
      "epoch": 0.3722149410222805,
      "grad_norm": 0.3329363763332367,
      "learning_rate": 0.00014394264912494017,
      "loss": 3.5297164916992188,
      "step": 2840
    },
    {
      "epoch": 0.3735255570117955,
      "grad_norm": 0.3543044626712799,
      "learning_rate": 0.00014356244247081863,
      "loss": 3.5188308715820313,
      "step": 2850
    },
    {
      "epoch": 0.3748361730013106,
      "grad_norm": 0.35182806849479675,
      "learning_rate": 0.0001431814570027994,
      "loss": 3.5263519287109375,
      "step": 2860
    },
    {
      "epoch": 0.3761467889908257,
      "grad_norm": 0.3496115505695343,
      "learning_rate": 0.00014279969953218054,
      "loss": 3.558135223388672,
      "step": 2870
    },
    {
      "epoch": 0.37745740498034075,
      "grad_norm": 0.3155076503753662,
      "learning_rate": 0.00014241717688406197,
      "loss": 3.5189140319824217,
      "step": 2880
    },
    {
      "epoch": 0.37876802096985585,
      "grad_norm": 0.8486918210983276,
      "learning_rate": 0.00014203389589722355,
      "loss": 3.5444580078125,
      "step": 2890
    },
    {
      "epoch": 0.3800786369593709,
      "grad_norm": 0.369697242975235,
      "learning_rate": 0.00014164986342400285,
      "loss": 3.5387516021728516,
      "step": 2900
    },
    {
      "epoch": 0.381389252948886,
      "grad_norm": 0.3202561140060425,
      "learning_rate": 0.0001412650863301725,
      "loss": 3.5395660400390625,
      "step": 2910
    },
    {
      "epoch": 0.382699868938401,
      "grad_norm": 0.3366410434246063,
      "learning_rate": 0.00014087957149481762,
      "loss": 3.555801773071289,
      "step": 2920
    },
    {
      "epoch": 0.3840104849279161,
      "grad_norm": 0.34785890579223633,
      "learning_rate": 0.00014049332581021266,
      "loss": 3.5291671752929688,
      "step": 2930
    },
    {
      "epoch": 0.3853211009174312,
      "grad_norm": 0.3297308385372162,
      "learning_rate": 0.0001401063561816984,
      "loss": 3.5065231323242188,
      "step": 2940
    },
    {
      "epoch": 0.38663171690694625,
      "grad_norm": 0.3157079815864563,
      "learning_rate": 0.00013971866952755817,
      "loss": 3.499795913696289,
      "step": 2950
    },
    {
      "epoch": 0.38794233289646135,
      "grad_norm": 0.3249460458755493,
      "learning_rate": 0.0001393302727788946,
      "loss": 3.535822296142578,
      "step": 2960
    },
    {
      "epoch": 0.3892529488859764,
      "grad_norm": 0.38803529739379883,
      "learning_rate": 0.00013894117287950526,
      "loss": 3.5551456451416015,
      "step": 2970
    },
    {
      "epoch": 0.3905635648754915,
      "grad_norm": 0.3703489899635315,
      "learning_rate": 0.0001385513767857588,
      "loss": 3.5351242065429687,
      "step": 2980
    },
    {
      "epoch": 0.3918741808650065,
      "grad_norm": 0.33586183190345764,
      "learning_rate": 0.00013816089146647062,
      "loss": 3.558055877685547,
      "step": 2990
    },
    {
      "epoch": 0.3931847968545216,
      "grad_norm": 0.36820846796035767,
      "learning_rate": 0.0001377697239027779,
      "loss": 3.5234466552734376,
      "step": 3000
    },
    {
      "epoch": 0.3931847968545216,
      "eval_loss": 3.497260093688965,
      "eval_runtime": 1.0781,
      "eval_samples_per_second": 118.728,
      "eval_steps_per_second": 29.682,
      "step": 3000
    },
    {
      "epoch": 0.3944954128440367,
      "grad_norm": 0.37449905276298523,
      "learning_rate": 0.00013737788108801531,
      "loss": 3.5460636138916017,
      "step": 3010
    },
    {
      "epoch": 0.39580602883355176,
      "grad_norm": 0.32343071699142456,
      "learning_rate": 0.00013698537002758963,
      "loss": 3.5344863891601563,
      "step": 3020
    },
    {
      "epoch": 0.39711664482306686,
      "grad_norm": 0.31588372588157654,
      "learning_rate": 0.00013659219773885453,
      "loss": 3.514875030517578,
      "step": 3030
    },
    {
      "epoch": 0.3984272608125819,
      "grad_norm": 0.35336896777153015,
      "learning_rate": 0.00013619837125098533,
      "loss": 3.5521949768066405,
      "step": 3040
    },
    {
      "epoch": 0.399737876802097,
      "grad_norm": 0.3647097945213318,
      "learning_rate": 0.0001358038976048531,
      "loss": 3.5020431518554687,
      "step": 3050
    },
    {
      "epoch": 0.40104849279161203,
      "grad_norm": 0.34516170620918274,
      "learning_rate": 0.0001354087838528989,
      "loss": 3.5423301696777343,
      "step": 3060
    },
    {
      "epoch": 0.40235910878112713,
      "grad_norm": 0.33818644285202026,
      "learning_rate": 0.00013501303705900768,
      "loss": 3.5412689208984376,
      "step": 3070
    },
    {
      "epoch": 0.4036697247706422,
      "grad_norm": 0.30593037605285645,
      "learning_rate": 0.00013461666429838193,
      "loss": 3.5437835693359374,
      "step": 3080
    },
    {
      "epoch": 0.40498034076015726,
      "grad_norm": 0.3402905762195587,
      "learning_rate": 0.00013421967265741535,
      "loss": 3.5333160400390624,
      "step": 3090
    },
    {
      "epoch": 0.40629095674967236,
      "grad_norm": 0.31519201397895813,
      "learning_rate": 0.0001338220692335659,
      "loss": 3.510814666748047,
      "step": 3100
    },
    {
      "epoch": 0.4076015727391874,
      "grad_norm": 0.32896602153778076,
      "learning_rate": 0.00013342386113522915,
      "loss": 3.5413749694824217,
      "step": 3110
    },
    {
      "epoch": 0.4089121887287025,
      "grad_norm": 0.35383597016334534,
      "learning_rate": 0.0001330250554816111,
      "loss": 3.5362228393554687,
      "step": 3120
    },
    {
      "epoch": 0.41022280471821754,
      "grad_norm": 0.3658854067325592,
      "learning_rate": 0.00013262565940260087,
      "loss": 3.531380462646484,
      "step": 3130
    },
    {
      "epoch": 0.41153342070773263,
      "grad_norm": 0.34737735986709595,
      "learning_rate": 0.00013222568003864332,
      "loss": 3.508496856689453,
      "step": 3140
    },
    {
      "epoch": 0.41284403669724773,
      "grad_norm": 0.3503577709197998,
      "learning_rate": 0.00013182512454061127,
      "loss": 3.5122150421142577,
      "step": 3150
    },
    {
      "epoch": 0.41415465268676277,
      "grad_norm": 0.41411256790161133,
      "learning_rate": 0.00013142400006967776,
      "loss": 3.518244171142578,
      "step": 3160
    },
    {
      "epoch": 0.41546526867627787,
      "grad_norm": 0.3081640303134918,
      "learning_rate": 0.00013102231379718803,
      "loss": 3.502947235107422,
      "step": 3170
    },
    {
      "epoch": 0.4167758846657929,
      "grad_norm": 0.29901912808418274,
      "learning_rate": 0.00013062007290453117,
      "loss": 3.4975021362304686,
      "step": 3180
    },
    {
      "epoch": 0.418086500655308,
      "grad_norm": 0.32307663559913635,
      "learning_rate": 0.00013021728458301196,
      "loss": 3.474725341796875,
      "step": 3190
    },
    {
      "epoch": 0.41939711664482304,
      "grad_norm": 0.3179335296154022,
      "learning_rate": 0.000129813956033722,
      "loss": 3.4851280212402345,
      "step": 3200
    },
    {
      "epoch": 0.42070773263433814,
      "grad_norm": 0.3789829611778259,
      "learning_rate": 0.00012941009446741132,
      "loss": 3.493799591064453,
      "step": 3210
    },
    {
      "epoch": 0.42201834862385323,
      "grad_norm": 0.3218543529510498,
      "learning_rate": 0.0001290057071043591,
      "loss": 3.500946807861328,
      "step": 3220
    },
    {
      "epoch": 0.4233289646133683,
      "grad_norm": 0.467252641916275,
      "learning_rate": 0.00012860080117424496,
      "loss": 3.476967620849609,
      "step": 3230
    },
    {
      "epoch": 0.42463958060288337,
      "grad_norm": 0.32915982604026794,
      "learning_rate": 0.00012819538391601934,
      "loss": 3.499958801269531,
      "step": 3240
    },
    {
      "epoch": 0.4259501965923984,
      "grad_norm": 0.32382312417030334,
      "learning_rate": 0.00012778946257777443,
      "loss": 3.5042747497558593,
      "step": 3250
    },
    {
      "epoch": 0.4272608125819135,
      "grad_norm": 0.4081120789051056,
      "learning_rate": 0.00012738304441661425,
      "loss": 3.4861988067626952,
      "step": 3260
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.3331235945224762,
      "learning_rate": 0.00012697613669852514,
      "loss": 3.542892074584961,
      "step": 3270
    },
    {
      "epoch": 0.42988204456094364,
      "grad_norm": 0.3115365207195282,
      "learning_rate": 0.00012656874669824582,
      "loss": 3.5056705474853516,
      "step": 3280
    },
    {
      "epoch": 0.43119266055045874,
      "grad_norm": 0.31235191226005554,
      "learning_rate": 0.0001261608816991373,
      "loss": 3.4678718566894533,
      "step": 3290
    },
    {
      "epoch": 0.4325032765399738,
      "grad_norm": 0.34477317333221436,
      "learning_rate": 0.00012575254899305262,
      "loss": 3.4818206787109376,
      "step": 3300
    },
    {
      "epoch": 0.4338138925294889,
      "grad_norm": 0.39361926913261414,
      "learning_rate": 0.00012534375588020658,
      "loss": 3.49686279296875,
      "step": 3310
    },
    {
      "epoch": 0.4351245085190039,
      "grad_norm": 0.3232015073299408,
      "learning_rate": 0.0001249345096690451,
      "loss": 3.4711750030517576,
      "step": 3320
    },
    {
      "epoch": 0.436435124508519,
      "grad_norm": 0.32783228158950806,
      "learning_rate": 0.0001245248176761148,
      "loss": 3.450823974609375,
      "step": 3330
    },
    {
      "epoch": 0.43774574049803405,
      "grad_norm": 0.34539034962654114,
      "learning_rate": 0.00012411468722593183,
      "loss": 3.4882408142089845,
      "step": 3340
    },
    {
      "epoch": 0.43905635648754915,
      "grad_norm": 0.3215061128139496,
      "learning_rate": 0.00012370412565085128,
      "loss": 3.4584945678710937,
      "step": 3350
    },
    {
      "epoch": 0.44036697247706424,
      "grad_norm": 0.30352655053138733,
      "learning_rate": 0.0001232931402909359,
      "loss": 3.4660221099853517,
      "step": 3360
    },
    {
      "epoch": 0.4416775884665793,
      "grad_norm": 0.349607914686203,
      "learning_rate": 0.00012288173849382483,
      "loss": 3.4675846099853516,
      "step": 3370
    },
    {
      "epoch": 0.4429882044560944,
      "grad_norm": 0.3552814722061157,
      "learning_rate": 0.0001224699276146024,
      "loss": 3.4534225463867188,
      "step": 3380
    },
    {
      "epoch": 0.4442988204456094,
      "grad_norm": 0.33739328384399414,
      "learning_rate": 0.00012205771501566659,
      "loss": 3.4672878265380858,
      "step": 3390
    },
    {
      "epoch": 0.4456094364351245,
      "grad_norm": 0.4542648196220398,
      "learning_rate": 0.00012164510806659723,
      "loss": 3.4786422729492186,
      "step": 3400
    },
    {
      "epoch": 0.44692005242463956,
      "grad_norm": 0.3146779239177704,
      "learning_rate": 0.00012123211414402451,
      "loss": 3.478990173339844,
      "step": 3410
    },
    {
      "epoch": 0.44823066841415465,
      "grad_norm": 0.3412339389324188,
      "learning_rate": 0.00012081874063149694,
      "loss": 3.4552181243896483,
      "step": 3420
    },
    {
      "epoch": 0.44954128440366975,
      "grad_norm": 0.2967008054256439,
      "learning_rate": 0.0001204049949193493,
      "loss": 3.4537937164306642,
      "step": 3430
    },
    {
      "epoch": 0.4508519003931848,
      "grad_norm": 0.3511996567249298,
      "learning_rate": 0.00011999088440457072,
      "loss": 3.450330352783203,
      "step": 3440
    },
    {
      "epoch": 0.4521625163826999,
      "grad_norm": 0.31556078791618347,
      "learning_rate": 0.00011957641649067217,
      "loss": 3.444178009033203,
      "step": 3450
    },
    {
      "epoch": 0.4534731323722149,
      "grad_norm": 0.4603165090084076,
      "learning_rate": 0.00011916159858755442,
      "loss": 3.412762451171875,
      "step": 3460
    },
    {
      "epoch": 0.45478374836173,
      "grad_norm": 0.35244429111480713,
      "learning_rate": 0.00011874643811137523,
      "loss": 3.4760997772216795,
      "step": 3470
    },
    {
      "epoch": 0.45609436435124506,
      "grad_norm": 0.36484965682029724,
      "learning_rate": 0.00011833094248441694,
      "loss": 3.4620925903320314,
      "step": 3480
    },
    {
      "epoch": 0.45740498034076016,
      "grad_norm": 0.34024161100387573,
      "learning_rate": 0.00011791511913495381,
      "loss": 3.469159698486328,
      "step": 3490
    },
    {
      "epoch": 0.45871559633027525,
      "grad_norm": 0.3085269033908844,
      "learning_rate": 0.00011749897549711912,
      "loss": 3.4188804626464844,
      "step": 3500
    },
    {
      "epoch": 0.45871559633027525,
      "eval_loss": 3.4381401538848877,
      "eval_runtime": 1.077,
      "eval_samples_per_second": 118.85,
      "eval_steps_per_second": 29.712,
      "step": 3500
    },
    {
      "epoch": 0.4600262123197903,
      "grad_norm": 0.3239330053329468,
      "learning_rate": 0.0001170825190107723,
      "loss": 3.476500701904297,
      "step": 3510
    },
    {
      "epoch": 0.4613368283093054,
      "grad_norm": 0.33591270446777344,
      "learning_rate": 0.00011666575712136591,
      "loss": 3.4304973602294924,
      "step": 3520
    },
    {
      "epoch": 0.46264744429882043,
      "grad_norm": 0.44769638776779175,
      "learning_rate": 0.00011624869727981257,
      "loss": 3.45989990234375,
      "step": 3530
    },
    {
      "epoch": 0.4639580602883355,
      "grad_norm": 0.32596758008003235,
      "learning_rate": 0.00011583134694235169,
      "loss": 3.4599456787109375,
      "step": 3540
    },
    {
      "epoch": 0.46526867627785057,
      "grad_norm": 0.3291541635990143,
      "learning_rate": 0.00011541371357041619,
      "loss": 3.4460033416748046,
      "step": 3550
    },
    {
      "epoch": 0.46657929226736566,
      "grad_norm": 0.3279915750026703,
      "learning_rate": 0.00011499580463049917,
      "loss": 3.4722900390625,
      "step": 3560
    },
    {
      "epoch": 0.46788990825688076,
      "grad_norm": 0.34962865710258484,
      "learning_rate": 0.00011457762759402026,
      "loss": 3.4722564697265623,
      "step": 3570
    },
    {
      "epoch": 0.4692005242463958,
      "grad_norm": 0.3223172128200531,
      "learning_rate": 0.00011415918993719228,
      "loss": 3.4083541870117187,
      "step": 3580
    },
    {
      "epoch": 0.4705111402359109,
      "grad_norm": 0.4092220366001129,
      "learning_rate": 0.00011374049914088729,
      "loss": 3.3749027252197266,
      "step": 3590
    },
    {
      "epoch": 0.47182175622542594,
      "grad_norm": 0.3289068639278412,
      "learning_rate": 0.00011332156269050314,
      "loss": 3.4786697387695313,
      "step": 3600
    },
    {
      "epoch": 0.47313237221494103,
      "grad_norm": 0.3178092837333679,
      "learning_rate": 0.00011290238807582944,
      "loss": 3.451682281494141,
      "step": 3610
    },
    {
      "epoch": 0.4744429882044561,
      "grad_norm": 0.31205493211746216,
      "learning_rate": 0.00011248298279091378,
      "loss": 3.415204620361328,
      "step": 3620
    },
    {
      "epoch": 0.47575360419397117,
      "grad_norm": 0.33566388487815857,
      "learning_rate": 0.00011206335433392762,
      "loss": 3.466032791137695,
      "step": 3630
    },
    {
      "epoch": 0.47706422018348627,
      "grad_norm": 0.32014167308807373,
      "learning_rate": 0.00011164351020703236,
      "loss": 3.417876052856445,
      "step": 3640
    },
    {
      "epoch": 0.4783748361730013,
      "grad_norm": 0.30793535709381104,
      "learning_rate": 0.00011122345791624518,
      "loss": 3.4089370727539063,
      "step": 3650
    },
    {
      "epoch": 0.4796854521625164,
      "grad_norm": 0.3043949604034424,
      "learning_rate": 0.00011080320497130482,
      "loss": 3.441085433959961,
      "step": 3660
    },
    {
      "epoch": 0.48099606815203144,
      "grad_norm": 0.3487127423286438,
      "learning_rate": 0.00011038275888553731,
      "loss": 3.426648712158203,
      "step": 3670
    },
    {
      "epoch": 0.48230668414154654,
      "grad_norm": 0.33338889479637146,
      "learning_rate": 0.00010996212717572174,
      "loss": 3.410777282714844,
      "step": 3680
    },
    {
      "epoch": 0.4836173001310616,
      "grad_norm": 0.3135928809642792,
      "learning_rate": 0.00010954131736195567,
      "loss": 3.42159423828125,
      "step": 3690
    },
    {
      "epoch": 0.4849279161205767,
      "grad_norm": 0.3235618770122528,
      "learning_rate": 0.00010912033696752104,
      "loss": 3.4308517456054686,
      "step": 3700
    },
    {
      "epoch": 0.48623853211009177,
      "grad_norm": 0.32941797375679016,
      "learning_rate": 0.00010869919351874924,
      "loss": 3.4475376129150392,
      "step": 3710
    },
    {
      "epoch": 0.4875491480996068,
      "grad_norm": 0.3030867278575897,
      "learning_rate": 0.00010827789454488689,
      "loss": 3.4102359771728517,
      "step": 3720
    },
    {
      "epoch": 0.4888597640891219,
      "grad_norm": 0.343825101852417,
      "learning_rate": 0.00010785644757796103,
      "loss": 3.4191864013671873,
      "step": 3730
    },
    {
      "epoch": 0.49017038007863695,
      "grad_norm": 0.3237217664718628,
      "learning_rate": 0.00010743486015264457,
      "loss": 3.440430450439453,
      "step": 3740
    },
    {
      "epoch": 0.49148099606815204,
      "grad_norm": 0.3112372159957886,
      "learning_rate": 0.00010701313980612155,
      "loss": 3.42187385559082,
      "step": 3750
    },
    {
      "epoch": 0.4927916120576671,
      "grad_norm": 0.3234122097492218,
      "learning_rate": 0.00010659129407795239,
      "loss": 3.4447341918945313,
      "step": 3760
    },
    {
      "epoch": 0.4941022280471822,
      "grad_norm": 0.3233126103878021,
      "learning_rate": 0.00010616933050993908,
      "loss": 3.408727264404297,
      "step": 3770
    },
    {
      "epoch": 0.4954128440366973,
      "grad_norm": 0.3332885205745697,
      "learning_rate": 0.00010574725664599039,
      "loss": 3.419011688232422,
      "step": 3780
    },
    {
      "epoch": 0.4967234600262123,
      "grad_norm": 0.32122355699539185,
      "learning_rate": 0.00010532508003198694,
      "loss": 3.409676742553711,
      "step": 3790
    },
    {
      "epoch": 0.4980340760157274,
      "grad_norm": 0.3335682451725006,
      "learning_rate": 0.00010490280821564628,
      "loss": 3.3936233520507812,
      "step": 3800
    },
    {
      "epoch": 0.49934469200524245,
      "grad_norm": 0.30686959624290466,
      "learning_rate": 0.00010448044874638822,
      "loss": 3.4175350189208986,
      "step": 3810
    },
    {
      "epoch": 0.5006553079947575,
      "grad_norm": 0.31994932889938354,
      "learning_rate": 0.0001040580091751993,
      "loss": 3.3717323303222657,
      "step": 3820
    },
    {
      "epoch": 0.5019659239842726,
      "grad_norm": 0.3275805413722992,
      "learning_rate": 0.00010363549705449846,
      "loss": 3.4295333862304687,
      "step": 3830
    },
    {
      "epoch": 0.5032765399737876,
      "grad_norm": 0.3137935698032379,
      "learning_rate": 0.00010321291993800141,
      "loss": 3.4044025421142576,
      "step": 3840
    },
    {
      "epoch": 0.5045871559633027,
      "grad_norm": 0.2928585410118103,
      "learning_rate": 0.00010279028538058608,
      "loss": 3.3909950256347656,
      "step": 3850
    },
    {
      "epoch": 0.5058977719528178,
      "grad_norm": 0.4643808603286743,
      "learning_rate": 0.00010236760093815716,
      "loss": 3.4161136627197264,
      "step": 3860
    },
    {
      "epoch": 0.5072083879423329,
      "grad_norm": 0.3479609489440918,
      "learning_rate": 0.00010194487416751133,
      "loss": 3.4109344482421875,
      "step": 3870
    },
    {
      "epoch": 0.508519003931848,
      "grad_norm": 0.3002571165561676,
      "learning_rate": 0.00010152211262620197,
      "loss": 3.4208934783935545,
      "step": 3880
    },
    {
      "epoch": 0.509829619921363,
      "grad_norm": 0.38259080052375793,
      "learning_rate": 0.00010109932387240404,
      "loss": 3.445271682739258,
      "step": 3890
    },
    {
      "epoch": 0.5111402359108781,
      "grad_norm": 0.32081010937690735,
      "learning_rate": 0.00010067651546477905,
      "loss": 3.4343528747558594,
      "step": 3900
    },
    {
      "epoch": 0.5124508519003932,
      "grad_norm": 0.3057716488838196,
      "learning_rate": 0.00010025369496233992,
      "loss": 3.409467315673828,
      "step": 3910
    },
    {
      "epoch": 0.5137614678899083,
      "grad_norm": 0.33491888642311096,
      "learning_rate": 9.983086992431575e-05,
      "loss": 3.3883102416992186,
      "step": 3920
    },
    {
      "epoch": 0.5150720838794234,
      "grad_norm": 0.2997762858867645,
      "learning_rate": 9.940804791001674e-05,
      "loss": 3.4108463287353517,
      "step": 3930
    },
    {
      "epoch": 0.5163826998689384,
      "grad_norm": 0.3269652724266052,
      "learning_rate": 9.898523647869901e-05,
      "loss": 3.3795501708984377,
      "step": 3940
    },
    {
      "epoch": 0.5176933158584535,
      "grad_norm": 0.32590535283088684,
      "learning_rate": 9.856244318942957e-05,
      "loss": 3.4327816009521483,
      "step": 3950
    },
    {
      "epoch": 0.5190039318479686,
      "grad_norm": 0.30181077122688293,
      "learning_rate": 9.813967560095096e-05,
      "loss": 3.4569053649902344,
      "step": 3960
    },
    {
      "epoch": 0.5203145478374837,
      "grad_norm": 0.33557194471359253,
      "learning_rate": 9.771694127154635e-05,
      "loss": 3.3972793579101563,
      "step": 3970
    },
    {
      "epoch": 0.5216251638269986,
      "grad_norm": 0.32555773854255676,
      "learning_rate": 9.729424775890415e-05,
      "loss": 3.414725112915039,
      "step": 3980
    },
    {
      "epoch": 0.5229357798165137,
      "grad_norm": 0.3153853118419647,
      "learning_rate": 9.687160261998323e-05,
      "loss": 3.432520294189453,
      "step": 3990
    },
    {
      "epoch": 0.5242463958060288,
      "grad_norm": 0.30330994725227356,
      "learning_rate": 9.64490134108775e-05,
      "loss": 3.4177536010742187,
      "step": 4000
    },
    {
      "epoch": 0.5242463958060288,
      "eval_loss": 3.3871026039123535,
      "eval_runtime": 1.0791,
      "eval_samples_per_second": 118.615,
      "eval_steps_per_second": 29.654,
      "step": 4000
    },
    {
      "epoch": 0.5255570117955439,
      "grad_norm": 0.326238751411438,
      "learning_rate": 9.602648768668102e-05,
      "loss": 3.4518890380859375,
      "step": 4010
    },
    {
      "epoch": 0.526867627785059,
      "grad_norm": 0.3168793022632599,
      "learning_rate": 9.560403300135276e-05,
      "loss": 3.4074005126953124,
      "step": 4020
    },
    {
      "epoch": 0.528178243774574,
      "grad_norm": 0.33325210213661194,
      "learning_rate": 9.518165690758179e-05,
      "loss": 3.4176284790039064,
      "step": 4030
    },
    {
      "epoch": 0.5294888597640891,
      "grad_norm": 0.3835427165031433,
      "learning_rate": 9.475936695665203e-05,
      "loss": 3.4074501037597655,
      "step": 4040
    },
    {
      "epoch": 0.5307994757536042,
      "grad_norm": 0.32493406534194946,
      "learning_rate": 9.43371706983073e-05,
      "loss": 3.4552375793457033,
      "step": 4050
    },
    {
      "epoch": 0.5321100917431193,
      "grad_norm": 0.32197055220603943,
      "learning_rate": 9.391507568061649e-05,
      "loss": 3.3711593627929686,
      "step": 4060
    },
    {
      "epoch": 0.5334207077326344,
      "grad_norm": 0.32301968336105347,
      "learning_rate": 9.349308944983837e-05,
      "loss": 3.411232757568359,
      "step": 4070
    },
    {
      "epoch": 0.5347313237221494,
      "grad_norm": 0.3350101411342621,
      "learning_rate": 9.307121955028685e-05,
      "loss": 3.4191001892089843,
      "step": 4080
    },
    {
      "epoch": 0.5360419397116645,
      "grad_norm": 0.32984063029289246,
      "learning_rate": 9.264947352419608e-05,
      "loss": 3.429418182373047,
      "step": 4090
    },
    {
      "epoch": 0.5373525557011796,
      "grad_norm": 0.3136809468269348,
      "learning_rate": 9.222785891158561e-05,
      "loss": 3.417095184326172,
      "step": 4100
    },
    {
      "epoch": 0.5386631716906947,
      "grad_norm": 0.31202802062034607,
      "learning_rate": 9.180638325012552e-05,
      "loss": 3.4011585235595705,
      "step": 4110
    },
    {
      "epoch": 0.5399737876802096,
      "grad_norm": 0.32449594140052795,
      "learning_rate": 9.138505407500169e-05,
      "loss": 3.388248825073242,
      "step": 4120
    },
    {
      "epoch": 0.5412844036697247,
      "grad_norm": 0.3140908181667328,
      "learning_rate": 9.096387891878112e-05,
      "loss": 3.3917205810546873,
      "step": 4130
    },
    {
      "epoch": 0.5425950196592398,
      "grad_norm": 0.32939860224723816,
      "learning_rate": 9.05428653112773e-05,
      "loss": 3.415624237060547,
      "step": 4140
    },
    {
      "epoch": 0.5439056356487549,
      "grad_norm": 0.33560019731521606,
      "learning_rate": 9.012202077941546e-05,
      "loss": 3.4133262634277344,
      "step": 4150
    },
    {
      "epoch": 0.54521625163827,
      "grad_norm": 0.34747788310050964,
      "learning_rate": 8.970135284709819e-05,
      "loss": 3.3784561157226562,
      "step": 4160
    },
    {
      "epoch": 0.546526867627785,
      "grad_norm": 0.309505432844162,
      "learning_rate": 8.928086903507059e-05,
      "loss": 3.393187713623047,
      "step": 4170
    },
    {
      "epoch": 0.5478374836173001,
      "grad_norm": 0.3165646493434906,
      "learning_rate": 8.886057686078628e-05,
      "loss": 3.4088680267333986,
      "step": 4180
    },
    {
      "epoch": 0.5491480996068152,
      "grad_norm": 0.30743396282196045,
      "learning_rate": 8.844048383827262e-05,
      "loss": 3.3991775512695312,
      "step": 4190
    },
    {
      "epoch": 0.5504587155963303,
      "grad_norm": 0.3135233223438263,
      "learning_rate": 8.802059747799658e-05,
      "loss": 3.403910827636719,
      "step": 4200
    },
    {
      "epoch": 0.5517693315858454,
      "grad_norm": 0.36344054341316223,
      "learning_rate": 8.760092528673039e-05,
      "loss": 3.428179931640625,
      "step": 4210
    },
    {
      "epoch": 0.5530799475753604,
      "grad_norm": 0.3224182724952698,
      "learning_rate": 8.71814747674173e-05,
      "loss": 3.4004798889160157,
      "step": 4220
    },
    {
      "epoch": 0.5543905635648755,
      "grad_norm": 0.4850703775882721,
      "learning_rate": 8.67622534190375e-05,
      "loss": 3.4311748504638673,
      "step": 4230
    },
    {
      "epoch": 0.5557011795543906,
      "grad_norm": 0.31577715277671814,
      "learning_rate": 8.63432687364741e-05,
      "loss": 3.412723922729492,
      "step": 4240
    },
    {
      "epoch": 0.5570117955439057,
      "grad_norm": 0.31131064891815186,
      "learning_rate": 8.5924528210379e-05,
      "loss": 3.383625793457031,
      "step": 4250
    },
    {
      "epoch": 0.5583224115334207,
      "grad_norm": 0.3247140347957611,
      "learning_rate": 8.550603932703911e-05,
      "loss": 3.3689788818359374,
      "step": 4260
    },
    {
      "epoch": 0.5596330275229358,
      "grad_norm": 0.3325023353099823,
      "learning_rate": 8.508780956824231e-05,
      "loss": 3.3962554931640625,
      "step": 4270
    },
    {
      "epoch": 0.5609436435124509,
      "grad_norm": 0.31043484807014465,
      "learning_rate": 8.466984641114404e-05,
      "loss": 3.387592315673828,
      "step": 4280
    },
    {
      "epoch": 0.562254259501966,
      "grad_norm": 0.30663248896598816,
      "learning_rate": 8.425215732813323e-05,
      "loss": 3.4229507446289062,
      "step": 4290
    },
    {
      "epoch": 0.563564875491481,
      "grad_norm": 0.3366420269012451,
      "learning_rate": 8.383474978669891e-05,
      "loss": 3.379755401611328,
      "step": 4300
    },
    {
      "epoch": 0.564875491480996,
      "grad_norm": 0.3328545093536377,
      "learning_rate": 8.341763124929683e-05,
      "loss": 3.37796630859375,
      "step": 4310
    },
    {
      "epoch": 0.5661861074705111,
      "grad_norm": 0.30118587613105774,
      "learning_rate": 8.300080917321567e-05,
      "loss": 3.4230308532714844,
      "step": 4320
    },
    {
      "epoch": 0.5674967234600262,
      "grad_norm": 0.3139464557170868,
      "learning_rate": 8.258429101044409e-05,
      "loss": 3.399740219116211,
      "step": 4330
    },
    {
      "epoch": 0.5688073394495413,
      "grad_norm": 0.4278660714626312,
      "learning_rate": 8.21680842075373e-05,
      "loss": 3.3967803955078124,
      "step": 4340
    },
    {
      "epoch": 0.5701179554390564,
      "grad_norm": 0.31098660826683044,
      "learning_rate": 8.175219620548404e-05,
      "loss": 3.3812324523925783,
      "step": 4350
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.3112267255783081,
      "learning_rate": 8.133663443957348e-05,
      "loss": 3.4089019775390623,
      "step": 4360
    },
    {
      "epoch": 0.5727391874180865,
      "grad_norm": 0.3149000406265259,
      "learning_rate": 8.092140633926221e-05,
      "loss": 3.3670639038085937,
      "step": 4370
    },
    {
      "epoch": 0.5740498034076016,
      "grad_norm": 0.30315423011779785,
      "learning_rate": 8.050651932804163e-05,
      "loss": 3.3814231872558596,
      "step": 4380
    },
    {
      "epoch": 0.5753604193971167,
      "grad_norm": 0.3087703287601471,
      "learning_rate": 8.009198082330508e-05,
      "loss": 3.370252990722656,
      "step": 4390
    },
    {
      "epoch": 0.5766710353866317,
      "grad_norm": 0.3247099816799164,
      "learning_rate": 7.967779823621522e-05,
      "loss": 3.408329391479492,
      "step": 4400
    },
    {
      "epoch": 0.5779816513761468,
      "grad_norm": 0.30703648924827576,
      "learning_rate": 7.926397897157171e-05,
      "loss": 3.3721439361572267,
      "step": 4410
    },
    {
      "epoch": 0.5792922673656619,
      "grad_norm": 0.3469173312187195,
      "learning_rate": 7.885053042767847e-05,
      "loss": 3.3755874633789062,
      "step": 4420
    },
    {
      "epoch": 0.580602883355177,
      "grad_norm": 0.35509803891181946,
      "learning_rate": 7.843745999621184e-05,
      "loss": 3.3569053649902343,
      "step": 4430
    },
    {
      "epoch": 0.581913499344692,
      "grad_norm": 0.3195863366127014,
      "learning_rate": 7.80247750620881e-05,
      "loss": 3.3575416564941407,
      "step": 4440
    },
    {
      "epoch": 0.583224115334207,
      "grad_norm": 0.3031638264656067,
      "learning_rate": 7.761248300333165e-05,
      "loss": 3.3901832580566404,
      "step": 4450
    },
    {
      "epoch": 0.5845347313237221,
      "grad_norm": 0.30322331190109253,
      "learning_rate": 7.720059119094304e-05,
      "loss": 3.4193626403808595,
      "step": 4460
    },
    {
      "epoch": 0.5858453473132372,
      "grad_norm": 0.32810741662979126,
      "learning_rate": 7.678910698876703e-05,
      "loss": 3.39061279296875,
      "step": 4470
    },
    {
      "epoch": 0.5871559633027523,
      "grad_norm": 0.3103083670139313,
      "learning_rate": 7.637803775336121e-05,
      "loss": 3.3724815368652346,
      "step": 4480
    },
    {
      "epoch": 0.5884665792922673,
      "grad_norm": 0.3157121539115906,
      "learning_rate": 7.596739083386432e-05,
      "loss": 3.363127899169922,
      "step": 4490
    },
    {
      "epoch": 0.5897771952817824,
      "grad_norm": 0.31368428468704224,
      "learning_rate": 7.555717357186491e-05,
      "loss": 3.3873497009277345,
      "step": 4500
    },
    {
      "epoch": 0.5897771952817824,
      "eval_loss": 3.349853515625,
      "eval_runtime": 1.0755,
      "eval_samples_per_second": 119.01,
      "eval_steps_per_second": 29.753,
      "step": 4500
    },
    {
      "epoch": 0.5910878112712975,
      "grad_norm": 0.3187400996685028,
      "learning_rate": 7.514739330127006e-05,
      "loss": 3.38061637878418,
      "step": 4510
    },
    {
      "epoch": 0.5923984272608126,
      "grad_norm": 0.31810012459754944,
      "learning_rate": 7.473805734817416e-05,
      "loss": 3.3528343200683595,
      "step": 4520
    },
    {
      "epoch": 0.5937090432503277,
      "grad_norm": 0.30577442049980164,
      "learning_rate": 7.432917303072827e-05,
      "loss": 3.4044944763183596,
      "step": 4530
    },
    {
      "epoch": 0.5950196592398427,
      "grad_norm": 0.31250351667404175,
      "learning_rate": 7.392074765900885e-05,
      "loss": 3.37048454284668,
      "step": 4540
    },
    {
      "epoch": 0.5963302752293578,
      "grad_norm": 0.30148154497146606,
      "learning_rate": 7.351278853488737e-05,
      "loss": 3.3811817169189453,
      "step": 4550
    },
    {
      "epoch": 0.5976408912188729,
      "grad_norm": 0.32013002038002014,
      "learning_rate": 7.310530295189974e-05,
      "loss": 3.3760913848876952,
      "step": 4560
    },
    {
      "epoch": 0.598951507208388,
      "grad_norm": 0.3121527433395386,
      "learning_rate": 7.269829819511574e-05,
      "loss": 3.3625667572021483,
      "step": 4570
    },
    {
      "epoch": 0.6002621231979031,
      "grad_norm": 0.32702282071113586,
      "learning_rate": 7.22917815410089e-05,
      "loss": 3.3423603057861326,
      "step": 4580
    },
    {
      "epoch": 0.601572739187418,
      "grad_norm": 0.3107105493545532,
      "learning_rate": 7.188576025732647e-05,
      "loss": 3.3608413696289063,
      "step": 4590
    },
    {
      "epoch": 0.6028833551769331,
      "grad_norm": 0.3107401132583618,
      "learning_rate": 7.148024160295941e-05,
      "loss": 3.3815628051757813,
      "step": 4600
    },
    {
      "epoch": 0.6041939711664482,
      "grad_norm": 0.3191450536251068,
      "learning_rate": 7.107523282781254e-05,
      "loss": 3.3593799591064455,
      "step": 4610
    },
    {
      "epoch": 0.6055045871559633,
      "grad_norm": 0.3236934542655945,
      "learning_rate": 7.067074117267505e-05,
      "loss": 3.352891540527344,
      "step": 4620
    },
    {
      "epoch": 0.6068152031454783,
      "grad_norm": 0.37565138936042786,
      "learning_rate": 7.026677386909102e-05,
      "loss": 3.387762451171875,
      "step": 4630
    },
    {
      "epoch": 0.6081258191349934,
      "grad_norm": 0.3366449177265167,
      "learning_rate": 6.986333813923014e-05,
      "loss": 3.3691715240478515,
      "step": 4640
    },
    {
      "epoch": 0.6094364351245085,
      "grad_norm": 0.3368198871612549,
      "learning_rate": 6.946044119575852e-05,
      "loss": 3.361377716064453,
      "step": 4650
    },
    {
      "epoch": 0.6107470511140236,
      "grad_norm": 0.32912445068359375,
      "learning_rate": 6.905809024170979e-05,
      "loss": 3.377477264404297,
      "step": 4660
    },
    {
      "epoch": 0.6120576671035387,
      "grad_norm": 0.3042116165161133,
      "learning_rate": 6.865629247035634e-05,
      "loss": 3.3426773071289064,
      "step": 4670
    },
    {
      "epoch": 0.6133682830930537,
      "grad_norm": 0.3233180642127991,
      "learning_rate": 6.825505506508073e-05,
      "loss": 3.3775726318359376,
      "step": 4680
    },
    {
      "epoch": 0.6146788990825688,
      "grad_norm": 0.29626721143722534,
      "learning_rate": 6.785438519924716e-05,
      "loss": 3.355507659912109,
      "step": 4690
    },
    {
      "epoch": 0.6159895150720839,
      "grad_norm": 0.31352996826171875,
      "learning_rate": 6.745429003607347e-05,
      "loss": 3.3301483154296876,
      "step": 4700
    },
    {
      "epoch": 0.617300131061599,
      "grad_norm": 0.29384997487068176,
      "learning_rate": 6.705477672850264e-05,
      "loss": 3.340058135986328,
      "step": 4710
    },
    {
      "epoch": 0.6186107470511141,
      "grad_norm": 0.297488808631897,
      "learning_rate": 6.665585241907541e-05,
      "loss": 3.356072998046875,
      "step": 4720
    },
    {
      "epoch": 0.6199213630406291,
      "grad_norm": 0.3099939227104187,
      "learning_rate": 6.62575242398022e-05,
      "loss": 3.35523681640625,
      "step": 4730
    },
    {
      "epoch": 0.6212319790301442,
      "grad_norm": 0.3405107259750366,
      "learning_rate": 6.585979931203589e-05,
      "loss": 3.3339542388916015,
      "step": 4740
    },
    {
      "epoch": 0.6225425950196593,
      "grad_norm": 0.29546186327934265,
      "learning_rate": 6.546268474634427e-05,
      "loss": 3.3676658630371095,
      "step": 4750
    },
    {
      "epoch": 0.6238532110091743,
      "grad_norm": 0.31302016973495483,
      "learning_rate": 6.506618764238299e-05,
      "loss": 3.3599761962890624,
      "step": 4760
    },
    {
      "epoch": 0.6251638269986893,
      "grad_norm": 0.30460891127586365,
      "learning_rate": 6.46703150887687e-05,
      "loss": 3.3548095703125,
      "step": 4770
    },
    {
      "epoch": 0.6264744429882044,
      "grad_norm": 0.30694299936294556,
      "learning_rate": 6.427507416295233e-05,
      "loss": 3.343377685546875,
      "step": 4780
    },
    {
      "epoch": 0.6277850589777195,
      "grad_norm": 0.34920060634613037,
      "learning_rate": 6.38804719310924e-05,
      "loss": 3.34407958984375,
      "step": 4790
    },
    {
      "epoch": 0.6290956749672346,
      "grad_norm": 0.30971863865852356,
      "learning_rate": 6.348651544792887e-05,
      "loss": 3.345227813720703,
      "step": 4800
    },
    {
      "epoch": 0.6304062909567497,
      "grad_norm": 0.32231196761131287,
      "learning_rate": 6.309321175665686e-05,
      "loss": 3.355621337890625,
      "step": 4810
    },
    {
      "epoch": 0.6317169069462647,
      "grad_norm": 0.30323100090026855,
      "learning_rate": 6.270056788880088e-05,
      "loss": 3.339982604980469,
      "step": 4820
    },
    {
      "epoch": 0.6330275229357798,
      "grad_norm": 0.3249286413192749,
      "learning_rate": 6.230859086408901e-05,
      "loss": 3.370708465576172,
      "step": 4830
    },
    {
      "epoch": 0.6343381389252949,
      "grad_norm": 0.3215145766735077,
      "learning_rate": 6.191728769032744e-05,
      "loss": 3.328388214111328,
      "step": 4840
    },
    {
      "epoch": 0.63564875491481,
      "grad_norm": 0.3169178366661072,
      "learning_rate": 6.152666536327524e-05,
      "loss": 3.338909149169922,
      "step": 4850
    },
    {
      "epoch": 0.6369593709043251,
      "grad_norm": 0.3184073567390442,
      "learning_rate": 6.113673086651914e-05,
      "loss": 3.353936767578125,
      "step": 4860
    },
    {
      "epoch": 0.6382699868938401,
      "grad_norm": 0.3235366940498352,
      "learning_rate": 6.074749117134881e-05,
      "loss": 3.29986457824707,
      "step": 4870
    },
    {
      "epoch": 0.6395806028833552,
      "grad_norm": 0.3182830512523651,
      "learning_rate": 6.035895323663219e-05,
      "loss": 3.3734458923339843,
      "step": 4880
    },
    {
      "epoch": 0.6408912188728703,
      "grad_norm": 0.3092668056488037,
      "learning_rate": 5.9971124008691085e-05,
      "loss": 3.318250274658203,
      "step": 4890
    },
    {
      "epoch": 0.6422018348623854,
      "grad_norm": 0.28549256920814514,
      "learning_rate": 5.958401042117696e-05,
      "loss": 3.3585628509521483,
      "step": 4900
    },
    {
      "epoch": 0.6435124508519003,
      "grad_norm": 0.4132247269153595,
      "learning_rate": 5.919761939494695e-05,
      "loss": 3.3403884887695314,
      "step": 4910
    },
    {
      "epoch": 0.6448230668414154,
      "grad_norm": 0.31634393334388733,
      "learning_rate": 5.881195783794016e-05,
      "loss": 3.3272781372070312,
      "step": 4920
    },
    {
      "epoch": 0.6461336828309305,
      "grad_norm": 0.311374306678772,
      "learning_rate": 5.842703264505427e-05,
      "loss": 3.3597652435302736,
      "step": 4930
    },
    {
      "epoch": 0.6474442988204456,
      "grad_norm": 0.3237077593803406,
      "learning_rate": 5.804285069802202e-05,
      "loss": 3.3317779541015624,
      "step": 4940
    },
    {
      "epoch": 0.6487549148099607,
      "grad_norm": 0.29717254638671875,
      "learning_rate": 5.765941886528851e-05,
      "loss": 3.323468017578125,
      "step": 4950
    },
    {
      "epoch": 0.6500655307994757,
      "grad_norm": 0.31114426255226135,
      "learning_rate": 5.727674400188798e-05,
      "loss": 3.3352993011474608,
      "step": 4960
    },
    {
      "epoch": 0.6513761467889908,
      "grad_norm": 0.298746794462204,
      "learning_rate": 5.689483294932163e-05,
      "loss": 3.3412445068359373,
      "step": 4970
    },
    {
      "epoch": 0.6526867627785059,
      "grad_norm": 0.3022208511829376,
      "learning_rate": 5.651369253543523e-05,
      "loss": 3.3568382263183594,
      "step": 4980
    },
    {
      "epoch": 0.653997378768021,
      "grad_norm": 0.29490622878074646,
      "learning_rate": 5.613332957429689e-05,
      "loss": 3.3298187255859375,
      "step": 4990
    },
    {
      "epoch": 0.6553079947575361,
      "grad_norm": 0.3072386384010315,
      "learning_rate": 5.5753750866075384e-05,
      "loss": 3.305432891845703,
      "step": 5000
    },
    {
      "epoch": 0.6553079947575361,
      "eval_loss": 3.3193883895874023,
      "eval_runtime": 1.1003,
      "eval_samples_per_second": 116.328,
      "eval_steps_per_second": 29.082,
      "step": 5000
    },
    {
      "epoch": 0.6566186107470511,
      "grad_norm": 0.31927192211151123,
      "learning_rate": 5.537496319691842e-05,
      "loss": 3.3518016815185545,
      "step": 5010
    },
    {
      "epoch": 0.6579292267365662,
      "grad_norm": 0.30445393919944763,
      "learning_rate": 5.4996973338831626e-05,
      "loss": 3.324379730224609,
      "step": 5020
    },
    {
      "epoch": 0.6592398427260813,
      "grad_norm": 0.3192574381828308,
      "learning_rate": 5.461978804955714e-05,
      "loss": 3.286786651611328,
      "step": 5030
    },
    {
      "epoch": 0.6605504587155964,
      "grad_norm": 0.3199535608291626,
      "learning_rate": 5.424341407245298e-05,
      "loss": 3.3442413330078127,
      "step": 5040
    },
    {
      "epoch": 0.6618610747051114,
      "grad_norm": 0.36999377608299255,
      "learning_rate": 5.3867858136372404e-05,
      "loss": 3.3658817291259764,
      "step": 5050
    },
    {
      "epoch": 0.6631716906946264,
      "grad_norm": 0.3546248972415924,
      "learning_rate": 5.3493126955543704e-05,
      "loss": 3.3181053161621095,
      "step": 5060
    },
    {
      "epoch": 0.6644823066841415,
      "grad_norm": 0.3281347155570984,
      "learning_rate": 5.311922722945006e-05,
      "loss": 3.33707275390625,
      "step": 5070
    },
    {
      "epoch": 0.6657929226736566,
      "grad_norm": 0.3314729928970337,
      "learning_rate": 5.274616564270984e-05,
      "loss": 3.341175842285156,
      "step": 5080
    },
    {
      "epoch": 0.6671035386631717,
      "grad_norm": 0.2999439537525177,
      "learning_rate": 5.237394886495707e-05,
      "loss": 3.3121627807617187,
      "step": 5090
    },
    {
      "epoch": 0.6684141546526867,
      "grad_norm": 0.31203365325927734,
      "learning_rate": 5.2002583550722183e-05,
      "loss": 3.3463668823242188,
      "step": 5100
    },
    {
      "epoch": 0.6697247706422018,
      "grad_norm": 0.32296445965766907,
      "learning_rate": 5.163207633931308e-05,
      "loss": 3.3332698822021483,
      "step": 5110
    },
    {
      "epoch": 0.6710353866317169,
      "grad_norm": 0.2978368103504181,
      "learning_rate": 5.126243385469639e-05,
      "loss": 3.3421607971191407,
      "step": 5120
    },
    {
      "epoch": 0.672346002621232,
      "grad_norm": 0.30745768547058105,
      "learning_rate": 5.089366270537907e-05,
      "loss": 3.3254119873046877,
      "step": 5130
    },
    {
      "epoch": 0.6736566186107471,
      "grad_norm": 0.30247536301612854,
      "learning_rate": 5.05257694842902e-05,
      "loss": 3.3183547973632814,
      "step": 5140
    },
    {
      "epoch": 0.6749672346002621,
      "grad_norm": 0.34190618991851807,
      "learning_rate": 5.015876076866336e-05,
      "loss": 3.3317501068115236,
      "step": 5150
    },
    {
      "epoch": 0.6762778505897772,
      "grad_norm": 0.3016245663166046,
      "learning_rate": 4.9792643119918594e-05,
      "loss": 3.338121032714844,
      "step": 5160
    },
    {
      "epoch": 0.6775884665792923,
      "grad_norm": 0.3206377625465393,
      "learning_rate": 4.94274230835455e-05,
      "loss": 3.3291389465332033,
      "step": 5170
    },
    {
      "epoch": 0.6788990825688074,
      "grad_norm": 0.33241593837738037,
      "learning_rate": 4.9063107188986014e-05,
      "loss": 3.300611114501953,
      "step": 5180
    },
    {
      "epoch": 0.6802096985583224,
      "grad_norm": 0.3275558054447174,
      "learning_rate": 4.869970194951784e-05,
      "loss": 3.337323760986328,
      "step": 5190
    },
    {
      "epoch": 0.6815203145478375,
      "grad_norm": 0.3159787356853485,
      "learning_rate": 4.833721386213781e-05,
      "loss": 3.329510498046875,
      "step": 5200
    },
    {
      "epoch": 0.6828309305373526,
      "grad_norm": 0.3475024104118347,
      "learning_rate": 4.797564940744575e-05,
      "loss": 3.323708343505859,
      "step": 5210
    },
    {
      "epoch": 0.6841415465268676,
      "grad_norm": 0.32093843817710876,
      "learning_rate": 4.7615015049528746e-05,
      "loss": 3.3412124633789064,
      "step": 5220
    },
    {
      "epoch": 0.6854521625163827,
      "grad_norm": 0.2981502413749695,
      "learning_rate": 4.7255317235845605e-05,
      "loss": 3.359756851196289,
      "step": 5230
    },
    {
      "epoch": 0.6867627785058977,
      "grad_norm": 0.30133214592933655,
      "learning_rate": 4.6896562397111445e-05,
      "loss": 3.3311183929443358,
      "step": 5240
    },
    {
      "epoch": 0.6880733944954128,
      "grad_norm": 0.33328819274902344,
      "learning_rate": 4.6538756947182635e-05,
      "loss": 3.3565765380859376,
      "step": 5250
    },
    {
      "epoch": 0.6893840104849279,
      "grad_norm": 0.31691062450408936,
      "learning_rate": 4.6181907282942416e-05,
      "loss": 3.299510192871094,
      "step": 5260
    },
    {
      "epoch": 0.690694626474443,
      "grad_norm": 0.3374110460281372,
      "learning_rate": 4.582601978418643e-05,
      "loss": 3.2947620391845702,
      "step": 5270
    },
    {
      "epoch": 0.6920052424639581,
      "grad_norm": 0.31588372588157654,
      "learning_rate": 4.547110081350855e-05,
      "loss": 3.3121360778808593,
      "step": 5280
    },
    {
      "epoch": 0.6933158584534731,
      "grad_norm": 0.295403391122818,
      "learning_rate": 4.5117156716187216e-05,
      "loss": 3.332244110107422,
      "step": 5290
    },
    {
      "epoch": 0.6946264744429882,
      "grad_norm": 0.3001668155193329,
      "learning_rate": 4.476419382007189e-05,
      "loss": 3.3460357666015623,
      "step": 5300
    },
    {
      "epoch": 0.6959370904325033,
      "grad_norm": 0.2911960780620575,
      "learning_rate": 4.44122184354702e-05,
      "loss": 3.310851287841797,
      "step": 5310
    },
    {
      "epoch": 0.6972477064220184,
      "grad_norm": 0.3135843575000763,
      "learning_rate": 4.40612368550348e-05,
      "loss": 3.334528350830078,
      "step": 5320
    },
    {
      "epoch": 0.6985583224115334,
      "grad_norm": 0.3108680844306946,
      "learning_rate": 4.3711255353651095e-05,
      "loss": 3.3180381774902346,
      "step": 5330
    },
    {
      "epoch": 0.6998689384010485,
      "grad_norm": 0.31819745898246765,
      "learning_rate": 4.336228018832489e-05,
      "loss": 3.335260772705078,
      "step": 5340
    },
    {
      "epoch": 0.7011795543905636,
      "grad_norm": 0.2930546700954437,
      "learning_rate": 4.3014317598070664e-05,
      "loss": 3.308660125732422,
      "step": 5350
    },
    {
      "epoch": 0.7024901703800787,
      "grad_norm": 0.32272499799728394,
      "learning_rate": 4.266737380379996e-05,
      "loss": 3.320453643798828,
      "step": 5360
    },
    {
      "epoch": 0.7038007863695938,
      "grad_norm": 0.31959226727485657,
      "learning_rate": 4.232145500821018e-05,
      "loss": 3.342560577392578,
      "step": 5370
    },
    {
      "epoch": 0.7051114023591087,
      "grad_norm": 0.29557809233665466,
      "learning_rate": 4.1976567395673684e-05,
      "loss": 3.2771766662597654,
      "step": 5380
    },
    {
      "epoch": 0.7064220183486238,
      "grad_norm": 0.31558558344841003,
      "learning_rate": 4.163271713212724e-05,
      "loss": 3.3245262145996093,
      "step": 5390
    },
    {
      "epoch": 0.7077326343381389,
      "grad_norm": 0.3526941239833832,
      "learning_rate": 4.1289910364961804e-05,
      "loss": 3.3200958251953123,
      "step": 5400
    },
    {
      "epoch": 0.709043250327654,
      "grad_norm": 0.35238364338874817,
      "learning_rate": 4.094815322291254e-05,
      "loss": 3.3825469970703126,
      "step": 5410
    },
    {
      "epoch": 0.7103538663171691,
      "grad_norm": 0.3272700011730194,
      "learning_rate": 4.060745181594937e-05,
      "loss": 3.3269989013671877,
      "step": 5420
    },
    {
      "epoch": 0.7116644823066841,
      "grad_norm": 0.3352244794368744,
      "learning_rate": 4.026781223516758e-05,
      "loss": 3.324008560180664,
      "step": 5430
    },
    {
      "epoch": 0.7129750982961992,
      "grad_norm": 0.3269336223602295,
      "learning_rate": 3.9929240552679206e-05,
      "loss": 3.324179840087891,
      "step": 5440
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.3039933145046234,
      "learning_rate": 3.959174282150405e-05,
      "loss": 3.3177886962890626,
      "step": 5450
    },
    {
      "epoch": 0.7155963302752294,
      "grad_norm": 0.31648242473602295,
      "learning_rate": 3.925532507546184e-05,
      "loss": 3.3113357543945314,
      "step": 5460
    },
    {
      "epoch": 0.7169069462647444,
      "grad_norm": 0.4337925314903259,
      "learning_rate": 3.891999332906414e-05,
      "loss": 3.3319847106933596,
      "step": 5470
    },
    {
      "epoch": 0.7182175622542595,
      "grad_norm": 0.30732420086860657,
      "learning_rate": 3.8585753577407036e-05,
      "loss": 3.2889732360839843,
      "step": 5480
    },
    {
      "epoch": 0.7195281782437746,
      "grad_norm": 0.29712918400764465,
      "learning_rate": 3.825261179606372e-05,
      "loss": 3.3131568908691404,
      "step": 5490
    },
    {
      "epoch": 0.7208387942332897,
      "grad_norm": 0.4064099192619324,
      "learning_rate": 3.792057394097769e-05,
      "loss": 3.3167797088623048,
      "step": 5500
    },
    {
      "epoch": 0.7208387942332897,
      "eval_loss": 3.2948293685913086,
      "eval_runtime": 1.0775,
      "eval_samples_per_second": 118.791,
      "eval_steps_per_second": 29.698,
      "step": 5500
    },
    {
      "epoch": 0.7221494102228048,
      "grad_norm": 0.2900736629962921,
      "learning_rate": 3.758964594835641e-05,
      "loss": 3.334580993652344,
      "step": 5510
    },
    {
      "epoch": 0.7234600262123198,
      "grad_norm": 0.33125314116477966,
      "learning_rate": 3.725983373456519e-05,
      "loss": 3.319670867919922,
      "step": 5520
    },
    {
      "epoch": 0.7247706422018348,
      "grad_norm": 0.322269082069397,
      "learning_rate": 3.693114319602119e-05,
      "loss": 3.3071502685546874,
      "step": 5530
    },
    {
      "epoch": 0.7260812581913499,
      "grad_norm": 0.2971557378768921,
      "learning_rate": 3.660358020908823e-05,
      "loss": 3.3088722229003906,
      "step": 5540
    },
    {
      "epoch": 0.727391874180865,
      "grad_norm": 0.32254520058631897,
      "learning_rate": 3.6277150629971534e-05,
      "loss": 3.30521240234375,
      "step": 5550
    },
    {
      "epoch": 0.72870249017038,
      "grad_norm": 0.3058118224143982,
      "learning_rate": 3.595186029461331e-05,
      "loss": 3.339150238037109,
      "step": 5560
    },
    {
      "epoch": 0.7300131061598951,
      "grad_norm": 0.30846792459487915,
      "learning_rate": 3.56277150185881e-05,
      "loss": 3.291063690185547,
      "step": 5570
    },
    {
      "epoch": 0.7313237221494102,
      "grad_norm": 0.29057520627975464,
      "learning_rate": 3.5304720596999056e-05,
      "loss": 3.2958980560302735,
      "step": 5580
    },
    {
      "epoch": 0.7326343381389253,
      "grad_norm": 0.2985893189907074,
      "learning_rate": 3.498288280437416e-05,
      "loss": 3.2895973205566404,
      "step": 5590
    },
    {
      "epoch": 0.7339449541284404,
      "grad_norm": 0.29524528980255127,
      "learning_rate": 3.466220739456312e-05,
      "loss": 3.2810997009277343,
      "step": 5600
    },
    {
      "epoch": 0.7352555701179554,
      "grad_norm": 0.2950055003166199,
      "learning_rate": 3.4342700100634384e-05,
      "loss": 3.35821647644043,
      "step": 5610
    },
    {
      "epoch": 0.7365661861074705,
      "grad_norm": 0.3141389489173889,
      "learning_rate": 3.402436663477275e-05,
      "loss": 3.326692581176758,
      "step": 5620
    },
    {
      "epoch": 0.7378768020969856,
      "grad_norm": 0.29024651646614075,
      "learning_rate": 3.370721268817719e-05,
      "loss": 3.328185272216797,
      "step": 5630
    },
    {
      "epoch": 0.7391874180865007,
      "grad_norm": 0.30281588435173035,
      "learning_rate": 3.339124393095907e-05,
      "loss": 3.311102294921875,
      "step": 5640
    },
    {
      "epoch": 0.7404980340760158,
      "grad_norm": 0.3014330565929413,
      "learning_rate": 3.3076466012040866e-05,
      "loss": 3.312317657470703,
      "step": 5650
    },
    {
      "epoch": 0.7418086500655308,
      "grad_norm": 0.29511767625808716,
      "learning_rate": 3.2762884559055075e-05,
      "loss": 3.296949768066406,
      "step": 5660
    },
    {
      "epoch": 0.7431192660550459,
      "grad_norm": 0.2992333173751831,
      "learning_rate": 3.245050517824368e-05,
      "loss": 3.3102745056152343,
      "step": 5670
    },
    {
      "epoch": 0.744429882044561,
      "grad_norm": 0.30958986282348633,
      "learning_rate": 3.213933345435788e-05,
      "loss": 3.302378845214844,
      "step": 5680
    },
    {
      "epoch": 0.745740498034076,
      "grad_norm": 0.30180537700653076,
      "learning_rate": 3.182937495055833e-05,
      "loss": 3.315709686279297,
      "step": 5690
    },
    {
      "epoch": 0.747051114023591,
      "grad_norm": 0.2902790904045105,
      "learning_rate": 3.1520635208315475e-05,
      "loss": 3.308466339111328,
      "step": 5700
    },
    {
      "epoch": 0.7483617300131061,
      "grad_norm": 0.3062317967414856,
      "learning_rate": 3.121311974731068e-05,
      "loss": 3.3105705261230467,
      "step": 5710
    },
    {
      "epoch": 0.7496723460026212,
      "grad_norm": 0.30234888195991516,
      "learning_rate": 3.090683406533748e-05,
      "loss": 3.321080780029297,
      "step": 5720
    },
    {
      "epoch": 0.7509829619921363,
      "grad_norm": 0.28379008173942566,
      "learning_rate": 3.060178363820335e-05,
      "loss": 3.3270851135253907,
      "step": 5730
    },
    {
      "epoch": 0.7522935779816514,
      "grad_norm": 0.28985893726348877,
      "learning_rate": 3.0297973919631694e-05,
      "loss": 3.3171066284179687,
      "step": 5740
    },
    {
      "epoch": 0.7536041939711664,
      "grad_norm": 0.3170120418071747,
      "learning_rate": 2.9995410341164344e-05,
      "loss": 3.329694366455078,
      "step": 5750
    },
    {
      "epoch": 0.7549148099606815,
      "grad_norm": 0.30880630016326904,
      "learning_rate": 2.9694098312064555e-05,
      "loss": 3.294874572753906,
      "step": 5760
    },
    {
      "epoch": 0.7562254259501966,
      "grad_norm": 0.294792503118515,
      "learning_rate": 2.9394043219220334e-05,
      "loss": 3.3365081787109374,
      "step": 5770
    },
    {
      "epoch": 0.7575360419397117,
      "grad_norm": 0.3062818944454193,
      "learning_rate": 2.909525042704796e-05,
      "loss": 3.3283241271972654,
      "step": 5780
    },
    {
      "epoch": 0.7588466579292268,
      "grad_norm": 0.29637277126312256,
      "learning_rate": 2.879772527739619e-05,
      "loss": 3.3241039276123048,
      "step": 5790
    },
    {
      "epoch": 0.7601572739187418,
      "grad_norm": 0.3072783052921295,
      "learning_rate": 2.850147308945067e-05,
      "loss": 3.3240474700927733,
      "step": 5800
    },
    {
      "epoch": 0.7614678899082569,
      "grad_norm": 0.302944540977478,
      "learning_rate": 2.820649915963903e-05,
      "loss": 3.289739990234375,
      "step": 5810
    },
    {
      "epoch": 0.762778505897772,
      "grad_norm": 0.29482176899909973,
      "learning_rate": 2.7912808761536003e-05,
      "loss": 3.316808319091797,
      "step": 5820
    },
    {
      "epoch": 0.7640891218872871,
      "grad_norm": 0.2962639033794403,
      "learning_rate": 2.762040714576919e-05,
      "loss": 3.313655471801758,
      "step": 5830
    },
    {
      "epoch": 0.765399737876802,
      "grad_norm": 0.29771360754966736,
      "learning_rate": 2.7329299539925225e-05,
      "loss": 3.298612594604492,
      "step": 5840
    },
    {
      "epoch": 0.7667103538663171,
      "grad_norm": 0.32306742668151855,
      "learning_rate": 2.703949114845631e-05,
      "loss": 3.2987388610839843,
      "step": 5850
    },
    {
      "epoch": 0.7680209698558322,
      "grad_norm": 0.30088114738464355,
      "learning_rate": 2.6750987152587137e-05,
      "loss": 3.2908279418945314,
      "step": 5860
    },
    {
      "epoch": 0.7693315858453473,
      "grad_norm": 0.29777640104293823,
      "learning_rate": 2.646379271022229e-05,
      "loss": 3.3093299865722656,
      "step": 5870
    },
    {
      "epoch": 0.7706422018348624,
      "grad_norm": 0.32324397563934326,
      "learning_rate": 2.6177912955854013e-05,
      "loss": 3.2842891693115233,
      "step": 5880
    },
    {
      "epoch": 0.7719528178243774,
      "grad_norm": 0.3037058711051941,
      "learning_rate": 2.589335300047042e-05,
      "loss": 3.3165321350097656,
      "step": 5890
    },
    {
      "epoch": 0.7732634338138925,
      "grad_norm": 0.29135647416114807,
      "learning_rate": 2.5610117931464128e-05,
      "loss": 3.301674652099609,
      "step": 5900
    },
    {
      "epoch": 0.7745740498034076,
      "grad_norm": 0.2896539866924286,
      "learning_rate": 2.5328212812541287e-05,
      "loss": 3.292359161376953,
      "step": 5910
    },
    {
      "epoch": 0.7758846657929227,
      "grad_norm": 0.29473626613616943,
      "learning_rate": 2.5047642683631058e-05,
      "loss": 3.314501190185547,
      "step": 5920
    },
    {
      "epoch": 0.7771952817824378,
      "grad_norm": 0.30136606097221375,
      "learning_rate": 2.4768412560795528e-05,
      "loss": 3.2981803894042967,
      "step": 5930
    },
    {
      "epoch": 0.7785058977719528,
      "grad_norm": 0.291029691696167,
      "learning_rate": 2.4490527436139998e-05,
      "loss": 3.273956298828125,
      "step": 5940
    },
    {
      "epoch": 0.7798165137614679,
      "grad_norm": 0.28917545080184937,
      "learning_rate": 2.4213992277723764e-05,
      "loss": 3.2718521118164063,
      "step": 5950
    },
    {
      "epoch": 0.781127129750983,
      "grad_norm": 0.294291228055954,
      "learning_rate": 2.393881202947128e-05,
      "loss": 3.2932113647460937,
      "step": 5960
    },
    {
      "epoch": 0.7824377457404981,
      "grad_norm": 0.3048882484436035,
      "learning_rate": 2.3664991611083754e-05,
      "loss": 3.3311233520507812,
      "step": 5970
    },
    {
      "epoch": 0.783748361730013,
      "grad_norm": 0.296068400144577,
      "learning_rate": 2.3392535917951298e-05,
      "loss": 3.279602813720703,
      "step": 5980
    },
    {
      "epoch": 0.7850589777195282,
      "grad_norm": 0.28586626052856445,
      "learning_rate": 2.3121449821065176e-05,
      "loss": 3.2574268341064454,
      "step": 5990
    },
    {
      "epoch": 0.7863695937090432,
      "grad_norm": 0.2919919490814209,
      "learning_rate": 2.285173816693099e-05,
      "loss": 3.2871192932128905,
      "step": 6000
    },
    {
      "epoch": 0.7863695937090432,
      "eval_loss": 3.278139591217041,
      "eval_runtime": 1.0952,
      "eval_samples_per_second": 116.87,
      "eval_steps_per_second": 29.217,
      "step": 6000
    },
    {
      "epoch": 0.7876802096985583,
      "grad_norm": 0.294953316450119,
      "learning_rate": 2.2583405777481837e-05,
      "loss": 3.2963531494140623,
      "step": 6010
    },
    {
      "epoch": 0.7889908256880734,
      "grad_norm": 0.28695785999298096,
      "learning_rate": 2.231645744999227e-05,
      "loss": 3.2852462768554687,
      "step": 6020
    },
    {
      "epoch": 0.7903014416775884,
      "grad_norm": 0.28605547547340393,
      "learning_rate": 2.2050897956992367e-05,
      "loss": 3.2965301513671874,
      "step": 6030
    },
    {
      "epoch": 0.7916120576671035,
      "grad_norm": 0.3020041882991791,
      "learning_rate": 2.1786732046182435e-05,
      "loss": 3.302587127685547,
      "step": 6040
    },
    {
      "epoch": 0.7929226736566186,
      "grad_norm": 0.3116467297077179,
      "learning_rate": 2.1523964440348232e-05,
      "loss": 3.2722827911376955,
      "step": 6050
    },
    {
      "epoch": 0.7942332896461337,
      "grad_norm": 0.2922550141811371,
      "learning_rate": 2.126259983727652e-05,
      "loss": 3.2659141540527346,
      "step": 6060
    },
    {
      "epoch": 0.7955439056356488,
      "grad_norm": 0.3238109350204468,
      "learning_rate": 2.1002642909670955e-05,
      "loss": 3.2884429931640624,
      "step": 6070
    },
    {
      "epoch": 0.7968545216251638,
      "grad_norm": 0.29941412806510925,
      "learning_rate": 2.0744098305068672e-05,
      "loss": 3.2531352996826173,
      "step": 6080
    },
    {
      "epoch": 0.7981651376146789,
      "grad_norm": 0.3028809428215027,
      "learning_rate": 2.048697064575703e-05,
      "loss": 3.277893829345703,
      "step": 6090
    },
    {
      "epoch": 0.799475753604194,
      "grad_norm": 0.2963281571865082,
      "learning_rate": 2.0231264528691286e-05,
      "loss": 3.269486999511719,
      "step": 6100
    },
    {
      "epoch": 0.8007863695937091,
      "grad_norm": 0.2926466763019562,
      "learning_rate": 1.997698452541208e-05,
      "loss": 3.2932498931884764,
      "step": 6110
    },
    {
      "epoch": 0.8020969855832241,
      "grad_norm": 0.2948893904685974,
      "learning_rate": 1.9724135181963865e-05,
      "loss": 3.2725826263427735,
      "step": 6120
    },
    {
      "epoch": 0.8034076015727392,
      "grad_norm": 0.3165523111820221,
      "learning_rate": 1.9472721018813643e-05,
      "loss": 3.2574291229248047,
      "step": 6130
    },
    {
      "epoch": 0.8047182175622543,
      "grad_norm": 0.29546403884887695,
      "learning_rate": 1.9222746530770076e-05,
      "loss": 3.277521514892578,
      "step": 6140
    },
    {
      "epoch": 0.8060288335517694,
      "grad_norm": 0.2877141833305359,
      "learning_rate": 1.8974216186903172e-05,
      "loss": 3.2794921875,
      "step": 6150
    },
    {
      "epoch": 0.8073394495412844,
      "grad_norm": 0.30668383836746216,
      "learning_rate": 1.872713443046439e-05,
      "loss": 3.2715736389160157,
      "step": 6160
    },
    {
      "epoch": 0.8086500655307994,
      "grad_norm": 0.28699082136154175,
      "learning_rate": 1.8481505678807186e-05,
      "loss": 3.275502395629883,
      "step": 6170
    },
    {
      "epoch": 0.8099606815203145,
      "grad_norm": 0.32899871468544006,
      "learning_rate": 1.8237334323308043e-05,
      "loss": 3.255423355102539,
      "step": 6180
    },
    {
      "epoch": 0.8112712975098296,
      "grad_norm": 0.29219210147857666,
      "learning_rate": 1.7994624729287955e-05,
      "loss": 3.2549812316894533,
      "step": 6190
    },
    {
      "epoch": 0.8125819134993447,
      "grad_norm": 0.3040613830089569,
      "learning_rate": 1.7753381235934417e-05,
      "loss": 3.295735168457031,
      "step": 6200
    },
    {
      "epoch": 0.8138925294888598,
      "grad_norm": 0.2960207462310791,
      "learning_rate": 1.7513608156223805e-05,
      "loss": 3.3118080139160155,
      "step": 6210
    },
    {
      "epoch": 0.8152031454783748,
      "grad_norm": 0.29487344622612,
      "learning_rate": 1.727530977684425e-05,
      "loss": 3.282395935058594,
      "step": 6220
    },
    {
      "epoch": 0.8165137614678899,
      "grad_norm": 0.2857603132724762,
      "learning_rate": 1.7038490358119163e-05,
      "loss": 3.295166015625,
      "step": 6230
    },
    {
      "epoch": 0.817824377457405,
      "grad_norm": 0.29180893301963806,
      "learning_rate": 1.6803154133930765e-05,
      "loss": 3.2311470031738283,
      "step": 6240
    },
    {
      "epoch": 0.8191349934469201,
      "grad_norm": 0.2868487536907196,
      "learning_rate": 1.65693053116447e-05,
      "loss": 3.2951675415039063,
      "step": 6250
    },
    {
      "epoch": 0.8204456094364351,
      "grad_norm": 0.30312997102737427,
      "learning_rate": 1.6336948072034607e-05,
      "loss": 3.249333953857422,
      "step": 6260
    },
    {
      "epoch": 0.8217562254259502,
      "grad_norm": 0.2823442220687866,
      "learning_rate": 1.6106086569207547e-05,
      "loss": 3.2559219360351563,
      "step": 6270
    },
    {
      "epoch": 0.8230668414154653,
      "grad_norm": 0.2953377366065979,
      "learning_rate": 1.5876724930529585e-05,
      "loss": 3.2690780639648436,
      "step": 6280
    },
    {
      "epoch": 0.8243774574049804,
      "grad_norm": 0.3080589771270752,
      "learning_rate": 1.5648867256552004e-05,
      "loss": 3.2661956787109374,
      "step": 6290
    },
    {
      "epoch": 0.8256880733944955,
      "grad_norm": 0.2989797294139862,
      "learning_rate": 1.5422517620938115e-05,
      "loss": 3.288446807861328,
      "step": 6300
    },
    {
      "epoch": 0.8269986893840104,
      "grad_norm": 0.3008802831172943,
      "learning_rate": 1.5197680070390397e-05,
      "loss": 3.3186111450195312,
      "step": 6310
    },
    {
      "epoch": 0.8283093053735255,
      "grad_norm": 0.3004642426967621,
      "learning_rate": 1.4974358624578033e-05,
      "loss": 3.2551506042480467,
      "step": 6320
    },
    {
      "epoch": 0.8296199213630406,
      "grad_norm": 0.28708234429359436,
      "learning_rate": 1.4752557276065204e-05,
      "loss": 3.264942169189453,
      "step": 6330
    },
    {
      "epoch": 0.8309305373525557,
      "grad_norm": 0.30744442343711853,
      "learning_rate": 1.4532279990239506e-05,
      "loss": 3.2820091247558594,
      "step": 6340
    },
    {
      "epoch": 0.8322411533420708,
      "grad_norm": 0.28989365696907043,
      "learning_rate": 1.4313530705241351e-05,
      "loss": 3.274460220336914,
      "step": 6350
    },
    {
      "epoch": 0.8335517693315858,
      "grad_norm": 0.2954351007938385,
      "learning_rate": 1.4096313331893274e-05,
      "loss": 3.272370147705078,
      "step": 6360
    },
    {
      "epoch": 0.8348623853211009,
      "grad_norm": 0.2940017282962799,
      "learning_rate": 1.388063175363018e-05,
      "loss": 3.3080726623535157,
      "step": 6370
    },
    {
      "epoch": 0.836173001310616,
      "grad_norm": 0.2876715362071991,
      "learning_rate": 1.3666489826429852e-05,
      "loss": 3.249498748779297,
      "step": 6380
    },
    {
      "epoch": 0.8374836173001311,
      "grad_norm": 0.3198702037334442,
      "learning_rate": 1.3453891378744044e-05,
      "loss": 3.2790260314941406,
      "step": 6390
    },
    {
      "epoch": 0.8387942332896461,
      "grad_norm": 0.31174004077911377,
      "learning_rate": 1.3242840211430007e-05,
      "loss": 3.291934585571289,
      "step": 6400
    },
    {
      "epoch": 0.8401048492791612,
      "grad_norm": 0.32010769844055176,
      "learning_rate": 1.3033340097682556e-05,
      "loss": 3.3114997863769533,
      "step": 6410
    },
    {
      "epoch": 0.8414154652686763,
      "grad_norm": 0.2811467945575714,
      "learning_rate": 1.2825394782966627e-05,
      "loss": 3.262708282470703,
      "step": 6420
    },
    {
      "epoch": 0.8427260812581914,
      "grad_norm": 0.31091687083244324,
      "learning_rate": 1.261900798495027e-05,
      "loss": 3.260384368896484,
      "step": 6430
    },
    {
      "epoch": 0.8440366972477065,
      "grad_norm": 0.2870129942893982,
      "learning_rate": 1.2414183393438228e-05,
      "loss": 3.2830638885498047,
      "step": 6440
    },
    {
      "epoch": 0.8453473132372215,
      "grad_norm": 0.30433380603790283,
      "learning_rate": 1.2210924670305934e-05,
      "loss": 3.304608154296875,
      "step": 6450
    },
    {
      "epoch": 0.8466579292267365,
      "grad_norm": 0.3117028772830963,
      "learning_rate": 1.2009235449434087e-05,
      "loss": 3.28111572265625,
      "step": 6460
    },
    {
      "epoch": 0.8479685452162516,
      "grad_norm": 0.2849375903606415,
      "learning_rate": 1.1809119336643659e-05,
      "loss": 3.278071975708008,
      "step": 6470
    },
    {
      "epoch": 0.8492791612057667,
      "grad_norm": 0.28063881397247314,
      "learning_rate": 1.1610579909631402e-05,
      "loss": 3.2682064056396483,
      "step": 6480
    },
    {
      "epoch": 0.8505897771952818,
      "grad_norm": 0.301556795835495,
      "learning_rate": 1.1413620717905937e-05,
      "loss": 3.2868423461914062,
      "step": 6490
    },
    {
      "epoch": 0.8519003931847968,
      "grad_norm": 0.3041786551475525,
      "learning_rate": 1.1218245282724282e-05,
      "loss": 3.2398521423339846,
      "step": 6500
    },
    {
      "epoch": 0.8519003931847968,
      "eval_loss": 3.26597261428833,
      "eval_runtime": 1.0883,
      "eval_samples_per_second": 117.612,
      "eval_steps_per_second": 29.403,
      "step": 6500
    },
    {
      "epoch": 0.8532110091743119,
      "grad_norm": 0.28452515602111816,
      "learning_rate": 1.102445709702885e-05,
      "loss": 3.296611785888672,
      "step": 6510
    },
    {
      "epoch": 0.854521625163827,
      "grad_norm": 0.32691529393196106,
      "learning_rate": 1.0832259625385132e-05,
      "loss": 3.263344955444336,
      "step": 6520
    },
    {
      "epoch": 0.8558322411533421,
      "grad_norm": 0.2974032163619995,
      "learning_rate": 1.0641656303919556e-05,
      "loss": 3.267037200927734,
      "step": 6530
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.2949809432029724,
      "learning_rate": 1.0452650540258213e-05,
      "loss": 3.318084716796875,
      "step": 6540
    },
    {
      "epoch": 0.8584534731323722,
      "grad_norm": 0.29877355694770813,
      "learning_rate": 1.0265245713465898e-05,
      "loss": 3.2653877258300783,
      "step": 6550
    },
    {
      "epoch": 0.8597640891218873,
      "grad_norm": 0.2992921769618988,
      "learning_rate": 1.0079445173985712e-05,
      "loss": 3.302996063232422,
      "step": 6560
    },
    {
      "epoch": 0.8610747051114024,
      "grad_norm": 0.30970993638038635,
      "learning_rate": 9.895252243579112e-06,
      "loss": 3.318895721435547,
      "step": 6570
    },
    {
      "epoch": 0.8623853211009175,
      "grad_norm": 0.28350308537483215,
      "learning_rate": 9.712670215266506e-06,
      "loss": 3.313739013671875,
      "step": 6580
    },
    {
      "epoch": 0.8636959370904325,
      "grad_norm": 0.28543195128440857,
      "learning_rate": 9.531702353268478e-06,
      "loss": 3.292449188232422,
      "step": 6590
    },
    {
      "epoch": 0.8650065530799476,
      "grad_norm": 0.31342577934265137,
      "learning_rate": 9.352351892947408e-06,
      "loss": 3.300048065185547,
      "step": 6600
    },
    {
      "epoch": 0.8663171690694627,
      "grad_norm": 0.3093763589859009,
      "learning_rate": 9.174622040749536e-06,
      "loss": 3.2877223968505858,
      "step": 6610
    },
    {
      "epoch": 0.8676277850589778,
      "grad_norm": 0.2900940179824829,
      "learning_rate": 8.99851597414777e-06,
      "loss": 3.263954925537109,
      "step": 6620
    },
    {
      "epoch": 0.8689384010484927,
      "grad_norm": 0.29298532009124756,
      "learning_rate": 8.824036841584681e-06,
      "loss": 3.2986801147460936,
      "step": 6630
    },
    {
      "epoch": 0.8702490170380078,
      "grad_norm": 0.29480549693107605,
      "learning_rate": 8.651187762416513e-06,
      "loss": 3.3016666412353515,
      "step": 6640
    },
    {
      "epoch": 0.8715596330275229,
      "grad_norm": 0.31557756662368774,
      "learning_rate": 8.479971826857125e-06,
      "loss": 3.3054836273193358,
      "step": 6650
    },
    {
      "epoch": 0.872870249017038,
      "grad_norm": 0.29097551107406616,
      "learning_rate": 8.310392095922902e-06,
      "loss": 3.3295261383056642,
      "step": 6660
    },
    {
      "epoch": 0.8741808650065531,
      "grad_norm": 0.3085312843322754,
      "learning_rate": 8.142451601377998e-06,
      "loss": 3.299213409423828,
      "step": 6670
    },
    {
      "epoch": 0.8754914809960681,
      "grad_norm": 0.276650071144104,
      "learning_rate": 7.976153345680115e-06,
      "loss": 3.2783363342285154,
      "step": 6680
    },
    {
      "epoch": 0.8768020969855832,
      "grad_norm": 0.3195405900478363,
      "learning_rate": 7.811500301926844e-06,
      "loss": 3.294453430175781,
      "step": 6690
    },
    {
      "epoch": 0.8781127129750983,
      "grad_norm": 0.2772200405597687,
      "learning_rate": 7.648495413802503e-06,
      "loss": 3.300670623779297,
      "step": 6700
    },
    {
      "epoch": 0.8794233289646134,
      "grad_norm": 0.29121139645576477,
      "learning_rate": 7.487141595525515e-06,
      "loss": 3.3137420654296874,
      "step": 6710
    },
    {
      "epoch": 0.8807339449541285,
      "grad_norm": 0.288507878780365,
      "learning_rate": 7.327441731796292e-06,
      "loss": 3.2971775054931642,
      "step": 6720
    },
    {
      "epoch": 0.8820445609436435,
      "grad_norm": 0.28590911626815796,
      "learning_rate": 7.169398677745698e-06,
      "loss": 3.3028072357177733,
      "step": 6730
    },
    {
      "epoch": 0.8833551769331586,
      "grad_norm": 0.2808607518672943,
      "learning_rate": 7.013015258883937e-06,
      "loss": 3.3124561309814453,
      "step": 6740
    },
    {
      "epoch": 0.8846657929226737,
      "grad_norm": 0.275604784488678,
      "learning_rate": 6.858294271050136e-06,
      "loss": 3.299256134033203,
      "step": 6750
    },
    {
      "epoch": 0.8859764089121888,
      "grad_norm": 0.3126138150691986,
      "learning_rate": 6.705238480362253e-06,
      "loss": 3.3072059631347654,
      "step": 6760
    },
    {
      "epoch": 0.8872870249017037,
      "grad_norm": 0.2812648117542267,
      "learning_rate": 6.553850623167757e-06,
      "loss": 3.284593200683594,
      "step": 6770
    },
    {
      "epoch": 0.8885976408912188,
      "grad_norm": 0.30022668838500977,
      "learning_rate": 6.404133405994517e-06,
      "loss": 3.358733367919922,
      "step": 6780
    },
    {
      "epoch": 0.8899082568807339,
      "grad_norm": 0.3897671103477478,
      "learning_rate": 6.2560895055026044e-06,
      "loss": 3.2805694580078124,
      "step": 6790
    },
    {
      "epoch": 0.891218872870249,
      "grad_norm": 0.2814595699310303,
      "learning_rate": 6.1097215684363325e-06,
      "loss": 3.289775085449219,
      "step": 6800
    },
    {
      "epoch": 0.8925294888597641,
      "grad_norm": 0.2786250412464142,
      "learning_rate": 5.965032211576993e-06,
      "loss": 3.2834228515625,
      "step": 6810
    },
    {
      "epoch": 0.8938401048492791,
      "grad_norm": 0.2777429223060608,
      "learning_rate": 5.8220240216960266e-06,
      "loss": 3.280866241455078,
      "step": 6820
    },
    {
      "epoch": 0.8951507208387942,
      "grad_norm": 0.2776261270046234,
      "learning_rate": 5.680699555508762e-06,
      "loss": 3.2776443481445314,
      "step": 6830
    },
    {
      "epoch": 0.8964613368283093,
      "grad_norm": 0.2950192093849182,
      "learning_rate": 5.541061339628773e-06,
      "loss": 3.3044166564941406,
      "step": 6840
    },
    {
      "epoch": 0.8977719528178244,
      "grad_norm": 0.2830997407436371,
      "learning_rate": 5.403111870522692e-06,
      "loss": 3.275865936279297,
      "step": 6850
    },
    {
      "epoch": 0.8990825688073395,
      "grad_norm": 0.2779957354068756,
      "learning_rate": 5.266853614465539e-06,
      "loss": 3.296843338012695,
      "step": 6860
    },
    {
      "epoch": 0.9003931847968545,
      "grad_norm": 0.27545085549354553,
      "learning_rate": 5.132289007496649e-06,
      "loss": 3.2827804565429686,
      "step": 6870
    },
    {
      "epoch": 0.9017038007863696,
      "grad_norm": 0.28516528010368347,
      "learning_rate": 4.999420455376102e-06,
      "loss": 3.27435302734375,
      "step": 6880
    },
    {
      "epoch": 0.9030144167758847,
      "grad_norm": 0.2794021964073181,
      "learning_rate": 4.86825033354178e-06,
      "loss": 3.3022735595703123,
      "step": 6890
    },
    {
      "epoch": 0.9043250327653998,
      "grad_norm": 0.27546775341033936,
      "learning_rate": 4.738780987066816e-06,
      "loss": 3.297052001953125,
      "step": 6900
    },
    {
      "epoch": 0.9056356487549148,
      "grad_norm": 0.29891765117645264,
      "learning_rate": 4.6110147306177134e-06,
      "loss": 3.269060516357422,
      "step": 6910
    },
    {
      "epoch": 0.9069462647444299,
      "grad_norm": 0.2922978401184082,
      "learning_rate": 4.484953848412954e-06,
      "loss": 3.276897430419922,
      "step": 6920
    },
    {
      "epoch": 0.908256880733945,
      "grad_norm": 0.284485787153244,
      "learning_rate": 4.3606005941821625e-06,
      "loss": 3.29921875,
      "step": 6930
    },
    {
      "epoch": 0.90956749672346,
      "grad_norm": 0.28652673959732056,
      "learning_rate": 4.237957191125796e-06,
      "loss": 3.2882545471191404,
      "step": 6940
    },
    {
      "epoch": 0.9108781127129751,
      "grad_norm": 0.2855699360370636,
      "learning_rate": 4.117025831875443e-06,
      "loss": 3.291054916381836,
      "step": 6950
    },
    {
      "epoch": 0.9121887287024901,
      "grad_norm": 0.29297029972076416,
      "learning_rate": 3.997808678454573e-06,
      "loss": 3.307445526123047,
      "step": 6960
    },
    {
      "epoch": 0.9134993446920052,
      "grad_norm": 0.28819260001182556,
      "learning_rate": 3.880307862239918e-06,
      "loss": 3.2897270202636717,
      "step": 6970
    },
    {
      "epoch": 0.9148099606815203,
      "grad_norm": 0.28534117341041565,
      "learning_rate": 3.764525483923331e-06,
      "loss": 3.2816349029541017,
      "step": 6980
    },
    {
      "epoch": 0.9161205766710354,
      "grad_norm": 0.2977021634578705,
      "learning_rate": 3.6504636134742644e-06,
      "loss": 3.3019195556640626,
      "step": 6990
    },
    {
      "epoch": 0.9174311926605505,
      "grad_norm": 0.30420050024986267,
      "learning_rate": 3.5381242901027843e-06,
      "loss": 3.248387908935547,
      "step": 7000
    },
    {
      "epoch": 0.9174311926605505,
      "eval_loss": 3.2593278884887695,
      "eval_runtime": 1.077,
      "eval_samples_per_second": 118.848,
      "eval_steps_per_second": 29.712,
      "step": 7000
    },
    {
      "epoch": 0.9187418086500655,
      "grad_norm": 0.2798181474208832,
      "learning_rate": 3.4275095222230045e-06,
      "loss": 3.2874984741210938,
      "step": 7010
    },
    {
      "epoch": 0.9200524246395806,
      "grad_norm": 0.3176014721393585,
      "learning_rate": 3.318621287417356e-06,
      "loss": 3.2528312683105467,
      "step": 7020
    },
    {
      "epoch": 0.9213630406290957,
      "grad_norm": 0.2832520604133606,
      "learning_rate": 3.211461532401039e-06,
      "loss": 3.2792755126953126,
      "step": 7030
    },
    {
      "epoch": 0.9226736566186108,
      "grad_norm": 0.2840670943260193,
      "learning_rate": 3.106032172987361e-06,
      "loss": 3.3148155212402344,
      "step": 7040
    },
    {
      "epoch": 0.9239842726081258,
      "grad_norm": 0.2739748954772949,
      "learning_rate": 3.00233509405341e-06,
      "loss": 3.2669998168945313,
      "step": 7050
    },
    {
      "epoch": 0.9252948885976409,
      "grad_norm": 0.29281947016716003,
      "learning_rate": 2.900372149506414e-06,
      "loss": 3.231282043457031,
      "step": 7060
    },
    {
      "epoch": 0.926605504587156,
      "grad_norm": 0.298840194940567,
      "learning_rate": 2.800145162250567e-06,
      "loss": 3.343215560913086,
      "step": 7070
    },
    {
      "epoch": 0.927916120576671,
      "grad_norm": 0.2800920009613037,
      "learning_rate": 2.701655924154367e-06,
      "loss": 3.288318634033203,
      "step": 7080
    },
    {
      "epoch": 0.9292267365661862,
      "grad_norm": 0.28348249197006226,
      "learning_rate": 2.6049061960187303e-06,
      "loss": 3.284392547607422,
      "step": 7090
    },
    {
      "epoch": 0.9305373525557011,
      "grad_norm": 0.27869218587875366,
      "learning_rate": 2.5098977075454165e-06,
      "loss": 3.26202507019043,
      "step": 7100
    },
    {
      "epoch": 0.9318479685452162,
      "grad_norm": 0.2775638699531555,
      "learning_rate": 2.4166321573060977e-06,
      "loss": 3.2691986083984377,
      "step": 7110
    },
    {
      "epoch": 0.9331585845347313,
      "grad_norm": 0.28189051151275635,
      "learning_rate": 2.3251112127120167e-06,
      "loss": 3.28306884765625,
      "step": 7120
    },
    {
      "epoch": 0.9344692005242464,
      "grad_norm": 0.27239105105400085,
      "learning_rate": 2.2353365099841873e-06,
      "loss": 3.251131057739258,
      "step": 7130
    },
    {
      "epoch": 0.9357798165137615,
      "grad_norm": 0.2699277997016907,
      "learning_rate": 2.14730965412413e-06,
      "loss": 3.2617202758789063,
      "step": 7140
    },
    {
      "epoch": 0.9370904325032765,
      "grad_norm": 0.2918347418308258,
      "learning_rate": 2.0610322188851617e-06,
      "loss": 3.2662094116210936,
      "step": 7150
    },
    {
      "epoch": 0.9384010484927916,
      "grad_norm": 0.29079312086105347,
      "learning_rate": 1.976505746744295e-06,
      "loss": 3.3002155303955076,
      "step": 7160
    },
    {
      "epoch": 0.9397116644823067,
      "grad_norm": 0.2938360273838043,
      "learning_rate": 1.8937317488746165e-06,
      "loss": 3.248193359375,
      "step": 7170
    },
    {
      "epoch": 0.9410222804718218,
      "grad_norm": 0.2767062783241272,
      "learning_rate": 1.8127117051183085e-06,
      "loss": 3.2747344970703125,
      "step": 7180
    },
    {
      "epoch": 0.9423328964613368,
      "grad_norm": 0.297127902507782,
      "learning_rate": 1.7334470639601918e-06,
      "loss": 3.244709014892578,
      "step": 7190
    },
    {
      "epoch": 0.9436435124508519,
      "grad_norm": 0.27425581216812134,
      "learning_rate": 1.6559392425017916e-06,
      "loss": 3.249822998046875,
      "step": 7200
    },
    {
      "epoch": 0.944954128440367,
      "grad_norm": 0.28988516330718994,
      "learning_rate": 1.580189626436046e-06,
      "loss": 3.2722457885742187,
      "step": 7210
    },
    {
      "epoch": 0.9462647444298821,
      "grad_norm": 0.2844417095184326,
      "learning_rate": 1.506199570022493e-06,
      "loss": 3.2284812927246094,
      "step": 7220
    },
    {
      "epoch": 0.9475753604193972,
      "grad_norm": 0.2929475009441376,
      "learning_rate": 1.433970396063089e-06,
      "loss": 3.250102996826172,
      "step": 7230
    },
    {
      "epoch": 0.9488859764089121,
      "grad_norm": 0.28254833817481995,
      "learning_rate": 1.3635033958785626e-06,
      "loss": 3.2649932861328126,
      "step": 7240
    },
    {
      "epoch": 0.9501965923984272,
      "grad_norm": 0.29200100898742676,
      "learning_rate": 1.2947998292853093e-06,
      "loss": 3.2766223907470704,
      "step": 7250
    },
    {
      "epoch": 0.9515072083879423,
      "grad_norm": 0.2859213650226593,
      "learning_rate": 1.2278609245728656e-06,
      "loss": 3.2988685607910155,
      "step": 7260
    },
    {
      "epoch": 0.9528178243774574,
      "grad_norm": 0.299523264169693,
      "learning_rate": 1.1626878784819494e-06,
      "loss": 3.253802490234375,
      "step": 7270
    },
    {
      "epoch": 0.9541284403669725,
      "grad_norm": 0.28574061393737793,
      "learning_rate": 1.0992818561831097e-06,
      "loss": 3.2580291748046877,
      "step": 7280
    },
    {
      "epoch": 0.9554390563564875,
      "grad_norm": 0.294808954000473,
      "learning_rate": 1.0376439912558211e-06,
      "loss": 3.2454341888427733,
      "step": 7290
    },
    {
      "epoch": 0.9567496723460026,
      "grad_norm": 0.2927926778793335,
      "learning_rate": 9.77775385668267e-07,
      "loss": 3.248301696777344,
      "step": 7300
    },
    {
      "epoch": 0.9580602883355177,
      "grad_norm": 0.275357723236084,
      "learning_rate": 9.196771097576661e-07,
      "loss": 3.267353820800781,
      "step": 7310
    },
    {
      "epoch": 0.9593709043250328,
      "grad_norm": 0.28868117928504944,
      "learning_rate": 8.633502022110329e-07,
      "loss": 3.244837188720703,
      "step": 7320
    },
    {
      "epoch": 0.9606815203145478,
      "grad_norm": 0.2750817537307739,
      "learning_rate": 8.087956700467136e-07,
      "loss": 3.2568435668945312,
      "step": 7330
    },
    {
      "epoch": 0.9619921363040629,
      "grad_norm": 0.2710508108139038,
      "learning_rate": 7.56014488596335e-07,
      "loss": 3.253221893310547,
      "step": 7340
    },
    {
      "epoch": 0.963302752293578,
      "grad_norm": 0.2738211154937744,
      "learning_rate": 7.05007601487373e-07,
      "loss": 3.2773902893066404,
      "step": 7350
    },
    {
      "epoch": 0.9646133682830931,
      "grad_norm": 0.27187711000442505,
      "learning_rate": 6.557759206262892e-07,
      "loss": 3.2607303619384767,
      "step": 7360
    },
    {
      "epoch": 0.9659239842726082,
      "grad_norm": 0.2756361961364746,
      "learning_rate": 6.083203261822212e-07,
      "loss": 3.2541645050048826,
      "step": 7370
    },
    {
      "epoch": 0.9672346002621232,
      "grad_norm": 0.2759033739566803,
      "learning_rate": 5.626416665712175e-07,
      "loss": 3.247817611694336,
      "step": 7380
    },
    {
      "epoch": 0.9685452162516383,
      "grad_norm": 0.28787118196487427,
      "learning_rate": 5.187407584411496e-07,
      "loss": 3.260750579833984,
      "step": 7390
    },
    {
      "epoch": 0.9698558322411533,
      "grad_norm": 0.2827833592891693,
      "learning_rate": 4.7661838665704617e-07,
      "loss": 3.2661361694335938,
      "step": 7400
    },
    {
      "epoch": 0.9711664482306684,
      "grad_norm": 0.31101733446121216,
      "learning_rate": 4.362753042870593e-07,
      "loss": 3.2435897827148437,
      "step": 7410
    },
    {
      "epoch": 0.9724770642201835,
      "grad_norm": 0.32739442586898804,
      "learning_rate": 3.9771223258902033e-07,
      "loss": 3.2514717102050783,
      "step": 7420
    },
    {
      "epoch": 0.9737876802096985,
      "grad_norm": 0.29380154609680176,
      "learning_rate": 3.60929860997572e-07,
      "loss": 3.246078872680664,
      "step": 7430
    },
    {
      "epoch": 0.9750982961992136,
      "grad_norm": 0.27706289291381836,
      "learning_rate": 3.259288471117894e-07,
      "loss": 3.2216529846191406,
      "step": 7440
    },
    {
      "epoch": 0.9764089121887287,
      "grad_norm": 0.27122655510902405,
      "learning_rate": 2.927098166834563e-07,
      "loss": 3.2784027099609374,
      "step": 7450
    },
    {
      "epoch": 0.9777195281782438,
      "grad_norm": 0.2749541103839874,
      "learning_rate": 2.6127336360587396e-07,
      "loss": 3.2723121643066406,
      "step": 7460
    },
    {
      "epoch": 0.9790301441677588,
      "grad_norm": 0.27366903424263,
      "learning_rate": 2.3162004990322506e-07,
      "loss": 3.2601028442382813,
      "step": 7470
    },
    {
      "epoch": 0.9803407601572739,
      "grad_norm": 0.2898007333278656,
      "learning_rate": 2.037504057205486e-07,
      "loss": 3.2964656829833983,
      "step": 7480
    },
    {
      "epoch": 0.981651376146789,
      "grad_norm": 0.2722117304801941,
      "learning_rate": 1.776649293142585e-07,
      "loss": 3.328302764892578,
      "step": 7490
    },
    {
      "epoch": 0.9829619921363041,
      "grad_norm": 0.2788948714733124,
      "learning_rate": 1.5336408704323957e-07,
      "loss": 3.26646728515625,
      "step": 7500
    },
    {
      "epoch": 0.9829619921363041,
      "eval_loss": 3.2572457790374756,
      "eval_runtime": 1.0762,
      "eval_samples_per_second": 118.935,
      "eval_steps_per_second": 29.734,
      "step": 7500
    },
    {
      "epoch": 0.9842726081258192,
      "grad_norm": 0.2735145390033722,
      "learning_rate": 1.3084831336047653e-07,
      "loss": 3.320809173583984,
      "step": 7510
    },
    {
      "epoch": 0.9855832241153342,
      "grad_norm": 0.290691077709198,
      "learning_rate": 1.1011801080531569e-07,
      "loss": 3.251786804199219,
      "step": 7520
    },
    {
      "epoch": 0.9868938401048493,
      "grad_norm": 0.28158876299858093,
      "learning_rate": 9.117354999629291e-08,
      "loss": 3.2572391510009764,
      "step": 7530
    },
    {
      "epoch": 0.9882044560943644,
      "grad_norm": 0.3086004853248596,
      "learning_rate": 7.401526962443894e-08,
      "loss": 3.2797309875488283,
      "step": 7540
    },
    {
      "epoch": 0.9895150720838795,
      "grad_norm": 0.280847430229187,
      "learning_rate": 5.86434764472954e-08,
      "loss": 3.3035411834716797,
      "step": 7550
    },
    {
      "epoch": 0.9908256880733946,
      "grad_norm": 0.28159886598587036,
      "learning_rate": 4.505844528337466e-08,
      "loss": 3.2849308013916017,
      "step": 7560
    },
    {
      "epoch": 0.9921363040629095,
      "grad_norm": 0.2907842695713043,
      "learning_rate": 3.326041900729715e-08,
      "loss": 3.3056507110595703,
      "step": 7570
    },
    {
      "epoch": 0.9934469200524246,
      "grad_norm": 0.28100988268852234,
      "learning_rate": 2.324960854540592e-08,
      "loss": 3.28037109375,
      "step": 7580
    },
    {
      "epoch": 0.9947575360419397,
      "grad_norm": 0.27997565269470215,
      "learning_rate": 1.502619287200302e-08,
      "loss": 3.3047534942626955,
      "step": 7590
    },
    {
      "epoch": 0.9960681520314548,
      "grad_norm": 0.28468552231788635,
      "learning_rate": 8.590319006185343e-09,
      "loss": 3.305406951904297,
      "step": 7600
    },
    {
      "epoch": 0.9973787680209698,
      "grad_norm": 0.2723033130168915,
      "learning_rate": 3.942102009180104e-09,
      "loss": 3.257395553588867,
      "step": 7610
    },
    {
      "epoch": 0.9986893840104849,
      "grad_norm": 0.27879369258880615,
      "learning_rate": 1.0816249822909186e-09,
      "loss": 3.2955612182617187,
      "step": 7620
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.28259631991386414,
      "learning_rate": 8.939065421209591e-12,
      "loss": 3.297560119628906,
      "step": 7630
    },
    {
      "epoch": 1.0,
      "eval_loss": 3.257199287414551,
      "eval_runtime": 1.075,
      "eval_samples_per_second": 119.064,
      "eval_steps_per_second": 29.766,
      "step": 7630
    }
  ],
  "logging_steps": 10,
  "max_steps": 7630,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.9198727907167437e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}