{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.273117937291103,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006366385484641095,
      "grad_norm": 28.7295833201599,
      "learning_rate": 0.00011219390703061137,
      "loss": 6.6223,
      "step": 10
    },
    {
      "epoch": 0.01273277096928219,
      "grad_norm": 15.081303427956621,
      "learning_rate": 0.00014596763837756142,
      "loss": 2.7812,
      "step": 20
    },
    {
      "epoch": 0.019099156453923283,
      "grad_norm": 17.081283324526982,
      "learning_rate": 0.00016572400472495783,
      "loss": 2.6574,
      "step": 30
    },
    {
      "epoch": 0.02546554193856438,
      "grad_norm": 15.201762160364256,
      "learning_rate": 0.00017974136972451145,
      "loss": 2.9039,
      "step": 40
    },
    {
      "epoch": 0.031831927423205474,
      "grad_norm": 11.557044296106628,
      "learning_rate": 0.0001906140827142727,
      "loss": 2.9066,
      "step": 50
    },
    {
      "epoch": 0.03819831290784657,
      "grad_norm": 14.915432761425704,
      "learning_rate": 0.00019949773607190786,
      "loss": 2.8914,
      "step": 60
    },
    {
      "epoch": 0.04456469839248767,
      "grad_norm": 8.745902020340656,
      "learning_rate": 0.0002070087579637228,
      "loss": 2.9014,
      "step": 70
    },
    {
      "epoch": 0.05093108387712876,
      "grad_norm": 8.857658675771324,
      "learning_rate": 0.0002135151010714615,
      "loss": 2.9293,
      "step": 80
    },
    {
      "epoch": 0.057297469361769854,
      "grad_norm": 11.48593285000581,
      "learning_rate": 0.0002192541024193043,
      "loss": 2.8625,
      "step": 90
    },
    {
      "epoch": 0.06366385484641095,
      "grad_norm": 7.815960040918993,
      "learning_rate": 0.00022438781406122275,
      "loss": 3.0063,
      "step": 100
    },
    {
      "epoch": 0.07003024033105204,
      "grad_norm": 8.632620514774269,
      "learning_rate": 0.00022903182113161202,
      "loss": 2.9527,
      "step": 110
    },
    {
      "epoch": 0.07639662581569313,
      "grad_norm": 7.955348207734164,
      "learning_rate": 0.00023327146741885792,
      "loss": 3.0082,
      "step": 120
    },
    {
      "epoch": 0.08276301130033424,
      "grad_norm": 7.216941061016277,
      "learning_rate": 0.00023717156393669215,
      "loss": 2.7871,
      "step": 130
    },
    {
      "epoch": 0.08912939678497533,
      "grad_norm": 5.822424121539524,
      "learning_rate": 0.00024078248931067283,
      "loss": 2.7844,
      "step": 140
    },
    {
      "epoch": 0.09549578226961643,
      "grad_norm": 7.906464628684026,
      "learning_rate": 0.00024414418040861915,
      "loss": 2.8514,
      "step": 150
    },
    {
      "epoch": 0.10186216775425752,
      "grad_norm": 7.514192904534888,
      "learning_rate": 0.00024728883241841157,
      "loss": 3.0086,
      "step": 160
    },
    {
      "epoch": 0.10822855323889861,
      "grad_norm": 6.425203024033055,
      "learning_rate": 0.0002502427789216415,
      "loss": 2.8754,
      "step": 170
    },
    {
      "epoch": 0.11459493872353971,
      "grad_norm": 7.088171053123606,
      "learning_rate": 0.00025302783376625435,
      "loss": 2.9094,
      "step": 180
    },
    {
      "epoch": 0.1209613242081808,
      "grad_norm": 6.538392044337595,
      "learning_rate": 0.00025566226965097254,
      "loss": 2.9262,
      "step": 190
    },
    {
      "epoch": 0.1273277096928219,
      "grad_norm": 8.672338929603391,
      "learning_rate": 0.0002581615454081728,
      "loss": 2.8391,
      "step": 200
    },
    {
      "epoch": 0.133694095177463,
      "grad_norm": 7.204443295582082,
      "learning_rate": 0.00026053885565806924,
      "loss": 2.9121,
      "step": 210
    },
    {
      "epoch": 0.14006048066210408,
      "grad_norm": 6.463760927732387,
      "learning_rate": 0.0002628055524785621,
      "loss": 2.9268,
      "step": 220
    },
    {
      "epoch": 0.1464268661467452,
      "grad_norm": 6.869739493367925,
      "learning_rate": 0.0002649714732657648,
      "loss": 2.8965,
      "step": 230
    },
    {
      "epoch": 0.15279325163138627,
      "grad_norm": 5.613452359830457,
      "learning_rate": 0.00026704519876580795,
      "loss": 2.9688,
      "step": 240
    },
    {
      "epoch": 0.15915963711602737,
      "grad_norm": 5.7575322101372866,
      "learning_rate": 0.000269034258397934,
      "loss": 2.9008,
      "step": 250
    },
    {
      "epoch": 0.16552602260066848,
      "grad_norm": 7.337481922102622,
      "learning_rate": 0.00027094529528364224,
      "loss": 2.9398,
      "step": 260
    },
    {
      "epoch": 0.17189240808530956,
      "grad_norm": 6.95242288679829,
      "learning_rate": 0.00027278420011365073,
      "loss": 2.8973,
      "step": 270
    },
    {
      "epoch": 0.17825879356995067,
      "grad_norm": 4.554710482953035,
      "learning_rate": 0.00027455622065762283,
      "loss": 2.9867,
      "step": 280
    },
    {
      "epoch": 0.18462517905459175,
      "grad_norm": 7.670629856220385,
      "learning_rate": 0.00027626605204863905,
      "loss": 2.9191,
      "step": 290
    },
    {
      "epoch": 0.19099156453923286,
      "grad_norm": 6.140158435864239,
      "learning_rate": 0.0002779179117555692,
      "loss": 2.8199,
      "step": 300
    },
    {
      "epoch": 0.19735795002387393,
      "grad_norm": 6.06880330501044,
      "learning_rate": 0.00027951560225766885,
      "loss": 2.959,
      "step": 310
    },
    {
      "epoch": 0.20372433550851504,
      "grad_norm": 5.932458703590213,
      "learning_rate": 0.0002810625637653616,
      "loss": 2.9984,
      "step": 320
    },
    {
      "epoch": 0.21009072099315615,
      "grad_norm": 5.413884214713577,
      "learning_rate": 0.0002825619188259585,
      "loss": 2.8937,
      "step": 330
    },
    {
      "epoch": 0.21645710647779723,
      "grad_norm": 5.6133458231383315,
      "learning_rate": 0.00028401651026859154,
      "loss": 2.8844,
      "step": 340
    },
    {
      "epoch": 0.22282349196243834,
      "grad_norm": 6.157515185075721,
      "learning_rate": 0.00028542893364738413,
      "loss": 2.9137,
      "step": 350
    },
    {
      "epoch": 0.22918987744707942,
      "grad_norm": 6.150435968994946,
      "learning_rate": 0.0002868015651132044,
      "loss": 2.8086,
      "step": 360
    },
    {
      "epoch": 0.23555626293172052,
      "grad_norm": 6.1391865377467205,
      "learning_rate": 0.00028813658546582825,
      "loss": 2.9246,
      "step": 370
    },
    {
      "epoch": 0.2419226484163616,
      "grad_norm": 4.720212170517568,
      "learning_rate": 0.0002894360009979226,
      "loss": 2.7273,
      "step": 380
    },
    {
      "epoch": 0.2482890339010027,
      "grad_norm": 5.221438084968306,
      "learning_rate": 0.0002907016616310386,
      "loss": 2.8332,
      "step": 390
    },
    {
      "epoch": 0.2546554193856438,
      "grad_norm": 5.397389581466363,
      "learning_rate": 0.00029193527675512284,
      "loss": 2.776,
      "step": 400
    },
    {
      "epoch": 0.26102180487028487,
      "grad_norm": 7.525981812223918,
      "learning_rate": 0.0002931384291118962,
      "loss": 2.8727,
      "step": 410
    },
    {
      "epoch": 0.267388190354926,
      "grad_norm": 5.7116155190893805,
      "learning_rate": 0.00029431258700501927,
      "loss": 2.9766,
      "step": 420
    },
    {
      "epoch": 0.2737545758395671,
      "grad_norm": 7.012023587035556,
      "learning_rate": 0.00029545911507334384,
      "loss": 2.7902,
      "step": 430
    },
    {
      "epoch": 0.28012096132420816,
      "grad_norm": 4.8865448559050435,
      "learning_rate": 0.0002965792838255121,
      "loss": 2.6873,
      "step": 440
    },
    {
      "epoch": 0.2864873468088493,
      "grad_norm": 4.682656054807096,
      "learning_rate": 0.0002976742781029656,
      "loss": 2.8498,
      "step": 450
    },
    {
      "epoch": 0.2928537322934904,
      "grad_norm": 4.114956894922335,
      "learning_rate": 0.00029874520461271485,
      "loss": 2.7461,
      "step": 460
    },
    {
      "epoch": 0.29922011777813146,
      "grad_norm": 4.388435500348782,
      "learning_rate": 0.0002997930986499357,
      "loss": 2.9004,
      "step": 470
    },
    {
      "epoch": 0.30558650326277254,
      "grad_norm": 4.06992963685517,
      "learning_rate": 0.0002995048337656213,
      "loss": 2.8203,
      "step": 480
    },
    {
      "epoch": 0.31195288874741367,
      "grad_norm": 5.7122354924160375,
      "learning_rate": 0.0002987974534307946,
      "loss": 2.8477,
      "step": 490
    },
    {
      "epoch": 0.31831927423205475,
      "grad_norm": 4.543743759310822,
      "learning_rate": 0.0002980900730959679,
      "loss": 2.8168,
      "step": 500
    },
    {
      "epoch": 0.32468565971669583,
      "grad_norm": 5.2449822235737384,
      "learning_rate": 0.0002973826927611412,
      "loss": 2.7021,
      "step": 510
    },
    {
      "epoch": 0.33105204520133696,
      "grad_norm": 6.42302117474529,
      "learning_rate": 0.0002966753124263145,
      "loss": 2.8545,
      "step": 520
    },
    {
      "epoch": 0.33741843068597804,
      "grad_norm": 6.4116258183908945,
      "learning_rate": 0.0002959679320914878,
      "loss": 2.8436,
      "step": 530
    },
    {
      "epoch": 0.3437848161706191,
      "grad_norm": 4.982523320381435,
      "learning_rate": 0.00029526055175666113,
      "loss": 2.7607,
      "step": 540
    },
    {
      "epoch": 0.3501512016552602,
      "grad_norm": 4.815920229901925,
      "learning_rate": 0.00029455317142183443,
      "loss": 2.734,
      "step": 550
    },
    {
      "epoch": 0.35651758713990134,
      "grad_norm": 9.300484207985622,
      "learning_rate": 0.00029384579108700774,
      "loss": 2.7889,
      "step": 560
    },
    {
      "epoch": 0.3628839726245424,
      "grad_norm": 4.5559522343853285,
      "learning_rate": 0.00029313841075218105,
      "loss": 2.7531,
      "step": 570
    },
    {
      "epoch": 0.3692503581091835,
      "grad_norm": 5.351641807589207,
      "learning_rate": 0.0002924310304173544,
      "loss": 2.7451,
      "step": 580
    },
    {
      "epoch": 0.37561674359382463,
      "grad_norm": 5.696588702079196,
      "learning_rate": 0.00029172365008252766,
      "loss": 2.7189,
      "step": 590
    },
    {
      "epoch": 0.3819831290784657,
      "grad_norm": 4.694068773418385,
      "learning_rate": 0.00029101626974770096,
      "loss": 2.6096,
      "step": 600
    },
    {
      "epoch": 0.3883495145631068,
      "grad_norm": 9.013555425033168,
      "learning_rate": 0.0002903088894128743,
      "loss": 2.6605,
      "step": 610
    },
    {
      "epoch": 0.39471590004774787,
      "grad_norm": 10.964569620769009,
      "learning_rate": 0.00028960150907804763,
      "loss": 2.7258,
      "step": 620
    },
    {
      "epoch": 0.401082285532389,
      "grad_norm": 8.94321766980447,
      "learning_rate": 0.00028889412874322093,
      "loss": 2.6688,
      "step": 630
    },
    {
      "epoch": 0.4074486710170301,
      "grad_norm": 10.742420289850541,
      "learning_rate": 0.0002881867484083942,
      "loss": 2.573,
      "step": 640
    },
    {
      "epoch": 0.41381505650167116,
      "grad_norm": 8.890111021830462,
      "learning_rate": 0.00028747936807356755,
      "loss": 2.5846,
      "step": 650
    },
    {
      "epoch": 0.4201814419863123,
      "grad_norm": 27.744833393643688,
      "learning_rate": 0.00028677198773874085,
      "loss": 2.5406,
      "step": 660
    },
    {
      "epoch": 0.4265478274709534,
      "grad_norm": 8.283239589638123,
      "learning_rate": 0.00028606460740391416,
      "loss": 2.602,
      "step": 670
    },
    {
      "epoch": 0.43291421295559446,
      "grad_norm": 7.738262298672388,
      "learning_rate": 0.00028535722706908746,
      "loss": 2.6947,
      "step": 680
    },
    {
      "epoch": 0.43928059844023554,
      "grad_norm": 11.536106786052837,
      "learning_rate": 0.00028464984673426077,
      "loss": 2.6422,
      "step": 690
    },
    {
      "epoch": 0.44564698392487667,
      "grad_norm": 7.93060742031869,
      "learning_rate": 0.0002839424663994341,
      "loss": 2.5859,
      "step": 700
    },
    {
      "epoch": 0.45201336940951775,
      "grad_norm": 4.031662354294752,
      "learning_rate": 0.0002832350860646074,
      "loss": 2.5947,
      "step": 710
    },
    {
      "epoch": 0.45837975489415883,
      "grad_norm": 5.258278074052536,
      "learning_rate": 0.0002825277057297807,
      "loss": 2.64,
      "step": 720
    },
    {
      "epoch": 0.4647461403787999,
      "grad_norm": 4.323834597980534,
      "learning_rate": 0.000281820325394954,
      "loss": 2.6949,
      "step": 730
    },
    {
      "epoch": 0.47111252586344105,
      "grad_norm": 5.0167173121086135,
      "learning_rate": 0.0002811129450601273,
      "loss": 2.6221,
      "step": 740
    },
    {
      "epoch": 0.4774789113480821,
      "grad_norm": 5.933856548166221,
      "learning_rate": 0.0002804055647253006,
      "loss": 2.632,
      "step": 750
    },
    {
      "epoch": 0.4838452968327232,
      "grad_norm": 5.090852518324657,
      "learning_rate": 0.0002796981843904739,
      "loss": 2.8242,
      "step": 760
    },
    {
      "epoch": 0.49021168231736434,
      "grad_norm": 4.60199924467046,
      "learning_rate": 0.0002789908040556472,
      "loss": 2.6357,
      "step": 770
    },
    {
      "epoch": 0.4965780678020054,
      "grad_norm": 4.62334978934682,
      "learning_rate": 0.0002782834237208206,
      "loss": 2.6455,
      "step": 780
    },
    {
      "epoch": 0.5029444532866465,
      "grad_norm": 5.7098868621282834,
      "learning_rate": 0.0002775760433859938,
      "loss": 2.7822,
      "step": 790
    },
    {
      "epoch": 0.5093108387712876,
      "grad_norm": 5.492909395839608,
      "learning_rate": 0.00027686866305116713,
      "loss": 2.5965,
      "step": 800
    },
    {
      "epoch": 0.5156772242559287,
      "grad_norm": 5.007751033932614,
      "learning_rate": 0.00027616128271634044,
      "loss": 2.6355,
      "step": 810
    },
    {
      "epoch": 0.5220436097405697,
      "grad_norm": 6.170648429353342,
      "learning_rate": 0.0002754539023815138,
      "loss": 2.6123,
      "step": 820
    },
    {
      "epoch": 0.5284099952252109,
      "grad_norm": 4.235113061064731,
      "learning_rate": 0.0002747465220466871,
      "loss": 2.6453,
      "step": 830
    },
    {
      "epoch": 0.534776380709852,
      "grad_norm": 4.538569485793573,
      "learning_rate": 0.00027403914171186036,
      "loss": 2.6713,
      "step": 840
    },
    {
      "epoch": 0.5411427661944931,
      "grad_norm": 4.019460059963649,
      "learning_rate": 0.0002733317613770337,
      "loss": 2.5943,
      "step": 850
    },
    {
      "epoch": 0.5475091516791342,
      "grad_norm": 6.254490760667432,
      "learning_rate": 0.000272624381042207,
      "loss": 2.5418,
      "step": 860
    },
    {
      "epoch": 0.5538755371637752,
      "grad_norm": 4.520259916472461,
      "learning_rate": 0.0002719170007073803,
      "loss": 2.584,
      "step": 870
    },
    {
      "epoch": 0.5602419226484163,
      "grad_norm": 5.065731135931019,
      "learning_rate": 0.0002712096203725536,
      "loss": 2.6723,
      "step": 880
    },
    {
      "epoch": 0.5666083081330574,
      "grad_norm": 3.833075034277501,
      "learning_rate": 0.00027050224003772694,
      "loss": 2.5248,
      "step": 890
    },
    {
      "epoch": 0.5729746936176986,
      "grad_norm": 5.507080157861899,
      "learning_rate": 0.00026979485970290024,
      "loss": 2.6379,
      "step": 900
    },
    {
      "epoch": 0.5793410791023397,
      "grad_norm": 14.128248882331409,
      "learning_rate": 0.00026908747936807355,
      "loss": 2.5904,
      "step": 910
    },
    {
      "epoch": 0.5857074645869808,
      "grad_norm": 6.1778278241357265,
      "learning_rate": 0.00026838009903324686,
      "loss": 2.5844,
      "step": 920
    },
    {
      "epoch": 0.5920738500716218,
      "grad_norm": 8.064848357512728,
      "learning_rate": 0.00026767271869842016,
      "loss": 2.6055,
      "step": 930
    },
    {
      "epoch": 0.5984402355562629,
      "grad_norm": 6.16438539022089,
      "learning_rate": 0.00026696533836359347,
      "loss": 2.5506,
      "step": 940
    },
    {
      "epoch": 0.604806621040904,
      "grad_norm": 5.882124357483614,
      "learning_rate": 0.0002662579580287668,
      "loss": 2.4973,
      "step": 950
    },
    {
      "epoch": 0.6111730065255451,
      "grad_norm": 8.373991467730166,
      "learning_rate": 0.0002655505776939401,
      "loss": 2.5428,
      "step": 960
    },
    {
      "epoch": 0.6175393920101863,
      "grad_norm": 6.831051520016273,
      "learning_rate": 0.0002648431973591134,
      "loss": 2.4982,
      "step": 970
    },
    {
      "epoch": 0.6239057774948273,
      "grad_norm": 5.074869694709693,
      "learning_rate": 0.0002641358170242867,
      "loss": 2.5465,
      "step": 980
    },
    {
      "epoch": 0.6302721629794684,
      "grad_norm": 5.503484162830985,
      "learning_rate": 0.00026342843668946,
      "loss": 2.324,
      "step": 990
    },
    {
      "epoch": 0.6366385484641095,
      "grad_norm": 4.385405886090266,
      "learning_rate": 0.0002627210563546333,
      "loss": 2.417,
      "step": 1000
    },
    {
      "epoch": 0.6430049339487506,
      "grad_norm": 5.034207848822367,
      "learning_rate": 0.0002620136760198066,
      "loss": 2.6066,
      "step": 1010
    },
    {
      "epoch": 0.6493713194333917,
      "grad_norm": 3.696840611162144,
      "learning_rate": 0.00026130629568497997,
      "loss": 2.5441,
      "step": 1020
    },
    {
      "epoch": 0.6557377049180327,
      "grad_norm": 4.6314444028768875,
      "learning_rate": 0.0002605989153501533,
      "loss": 2.4752,
      "step": 1030
    },
    {
      "epoch": 0.6621040904026739,
      "grad_norm": 3.4663340841035093,
      "learning_rate": 0.0002598915350153265,
      "loss": 2.4637,
      "step": 1040
    },
    {
      "epoch": 0.668470475887315,
      "grad_norm": 8.334602958229153,
      "learning_rate": 0.00025918415468049983,
      "loss": 2.6082,
      "step": 1050
    },
    {
      "epoch": 0.6748368613719561,
      "grad_norm": 4.575304861734687,
      "learning_rate": 0.0002584767743456732,
      "loss": 2.4842,
      "step": 1060
    },
    {
      "epoch": 0.6812032468565972,
      "grad_norm": 3.618603659881246,
      "learning_rate": 0.0002577693940108465,
      "loss": 2.4982,
      "step": 1070
    },
    {
      "epoch": 0.6875696323412382,
      "grad_norm": 4.4740750234847475,
      "learning_rate": 0.00025706201367601975,
      "loss": 2.4029,
      "step": 1080
    },
    {
      "epoch": 0.6939360178258793,
      "grad_norm": 4.208151138305904,
      "learning_rate": 0.0002563546333411931,
      "loss": 2.4836,
      "step": 1090
    },
    {
      "epoch": 0.7003024033105204,
      "grad_norm": 5.736418133529195,
      "learning_rate": 0.0002556472530063664,
      "loss": 2.6402,
      "step": 1100
    },
    {
      "epoch": 0.7066687887951616,
      "grad_norm": 3.6852138552936395,
      "learning_rate": 0.0002549398726715397,
      "loss": 2.4941,
      "step": 1110
    },
    {
      "epoch": 0.7130351742798027,
      "grad_norm": 4.11339792490071,
      "learning_rate": 0.000254232492336713,
      "loss": 2.4641,
      "step": 1120
    },
    {
      "epoch": 0.7194015597644438,
      "grad_norm": 3.7487169212790477,
      "learning_rate": 0.00025352511200188633,
      "loss": 2.3414,
      "step": 1130
    },
    {
      "epoch": 0.7257679452490848,
      "grad_norm": 3.442251781584789,
      "learning_rate": 0.00025281773166705964,
      "loss": 2.3932,
      "step": 1140
    },
    {
      "epoch": 0.7321343307337259,
      "grad_norm": 3.8470452186807846,
      "learning_rate": 0.00025211035133223294,
      "loss": 2.4334,
      "step": 1150
    },
    {
      "epoch": 0.738500716218367,
      "grad_norm": 4.506628903442333,
      "learning_rate": 0.00025140297099740625,
      "loss": 2.3195,
      "step": 1160
    },
    {
      "epoch": 0.7448671017030081,
      "grad_norm": 4.325389792461109,
      "learning_rate": 0.00025069559066257955,
      "loss": 2.4027,
      "step": 1170
    },
    {
      "epoch": 0.7512334871876493,
      "grad_norm": 3.8669860452410445,
      "learning_rate": 0.00024998821032775286,
      "loss": 2.4686,
      "step": 1180
    },
    {
      "epoch": 0.7575998726722903,
      "grad_norm": 3.561694549590622,
      "learning_rate": 0.00024928082999292617,
      "loss": 2.5395,
      "step": 1190
    },
    {
      "epoch": 0.7639662581569314,
      "grad_norm": 4.156976510059248,
      "learning_rate": 0.00024857344965809947,
      "loss": 2.4416,
      "step": 1200
    },
    {
      "epoch": 0.7703326436415725,
      "grad_norm": 5.616798309964706,
      "learning_rate": 0.0002478660693232728,
      "loss": 2.4055,
      "step": 1210
    },
    {
      "epoch": 0.7766990291262136,
      "grad_norm": 3.149306910194074,
      "learning_rate": 0.0002471586889884461,
      "loss": 2.4488,
      "step": 1220
    },
    {
      "epoch": 0.7830654146108547,
      "grad_norm": 3.0254085146949974,
      "learning_rate": 0.00024645130865361944,
      "loss": 2.44,
      "step": 1230
    },
    {
      "epoch": 0.7894318000954957,
      "grad_norm": 3.3553234269170815,
      "learning_rate": 0.0002457439283187927,
      "loss": 2.4689,
      "step": 1240
    },
    {
      "epoch": 0.7957981855801369,
      "grad_norm": 3.560968004340506,
      "learning_rate": 0.000245036547983966,
      "loss": 2.4328,
      "step": 1250
    },
    {
      "epoch": 0.802164571064778,
      "grad_norm": 3.670143268942363,
      "learning_rate": 0.00024432916764913936,
      "loss": 2.357,
      "step": 1260
    },
    {
      "epoch": 0.8085309565494191,
      "grad_norm": 4.603180595509743,
      "learning_rate": 0.00024362178731431264,
      "loss": 2.4336,
      "step": 1270
    },
    {
      "epoch": 0.8148973420340602,
      "grad_norm": 3.2173958923931845,
      "learning_rate": 0.00024291440697948595,
      "loss": 2.3859,
      "step": 1280
    },
    {
      "epoch": 0.8212637275187012,
      "grad_norm": 4.238015371838477,
      "learning_rate": 0.00024220702664465925,
      "loss": 2.4234,
      "step": 1290
    },
    {
      "epoch": 0.8276301130033423,
      "grad_norm": 3.350821041839247,
      "learning_rate": 0.00024149964630983258,
      "loss": 2.34,
      "step": 1300
    },
    {
      "epoch": 0.8339964984879834,
      "grad_norm": 3.956822835287818,
      "learning_rate": 0.00024079226597500586,
      "loss": 2.409,
      "step": 1310
    },
    {
      "epoch": 0.8403628839726246,
      "grad_norm": 3.206745954274851,
      "learning_rate": 0.00024008488564017917,
      "loss": 2.4766,
      "step": 1320
    },
    {
      "epoch": 0.8467292694572657,
      "grad_norm": 3.16798874537864,
      "learning_rate": 0.0002393775053053525,
      "loss": 2.4918,
      "step": 1330
    },
    {
      "epoch": 0.8530956549419068,
      "grad_norm": 3.2084119932227613,
      "learning_rate": 0.0002386701249705258,
      "loss": 2.3279,
      "step": 1340
    },
    {
      "epoch": 0.8594620404265478,
      "grad_norm": 2.6626163554970694,
      "learning_rate": 0.00023796274463569909,
      "loss": 2.3762,
      "step": 1350
    },
    {
      "epoch": 0.8658284259111889,
      "grad_norm": 3.543568104802714,
      "learning_rate": 0.00023725536430087242,
      "loss": 2.4406,
      "step": 1360
    },
    {
      "epoch": 0.87219481139583,
      "grad_norm": 3.0950338366546832,
      "learning_rate": 0.00023654798396604572,
      "loss": 2.4135,
      "step": 1370
    },
    {
      "epoch": 0.8785611968804711,
      "grad_norm": 3.4459117956268583,
      "learning_rate": 0.00023584060363121903,
      "loss": 2.3371,
      "step": 1380
    },
    {
      "epoch": 0.8849275823651122,
      "grad_norm": 3.3339357175581026,
      "learning_rate": 0.00023513322329639234,
      "loss": 2.3568,
      "step": 1390
    },
    {
      "epoch": 0.8912939678497533,
      "grad_norm": 2.7770602679788836,
      "learning_rate": 0.00023442584296156567,
      "loss": 2.4703,
      "step": 1400
    },
    {
      "epoch": 0.8976603533343944,
      "grad_norm": 3.1587359306963925,
      "learning_rate": 0.00023371846262673895,
      "loss": 2.3687,
      "step": 1410
    },
    {
      "epoch": 0.9040267388190355,
      "grad_norm": 3.6463373041615057,
      "learning_rate": 0.00023301108229191225,
      "loss": 2.5037,
      "step": 1420
    },
    {
      "epoch": 0.9103931243036766,
      "grad_norm": 3.8799007179484066,
      "learning_rate": 0.00023230370195708559,
      "loss": 2.3883,
      "step": 1430
    },
    {
      "epoch": 0.9167595097883177,
      "grad_norm": 2.650501985473451,
      "learning_rate": 0.0002315963216222589,
      "loss": 2.3818,
      "step": 1440
    },
    {
      "epoch": 0.9231258952729587,
      "grad_norm": 3.3575457214628353,
      "learning_rate": 0.00023088894128743217,
      "loss": 2.2014,
      "step": 1450
    },
    {
      "epoch": 0.9294922807575998,
      "grad_norm": 4.160903578127555,
      "learning_rate": 0.00023018156095260548,
      "loss": 2.3586,
      "step": 1460
    },
    {
      "epoch": 0.935858666242241,
      "grad_norm": 3.7275531046661805,
      "learning_rate": 0.0002294741806177788,
      "loss": 2.4684,
      "step": 1470
    },
    {
      "epoch": 0.9422250517268821,
      "grad_norm": 3.5804971221330515,
      "learning_rate": 0.00022876680028295211,
      "loss": 2.3984,
      "step": 1480
    },
    {
      "epoch": 0.9485914372115232,
      "grad_norm": 2.903402028010888,
      "learning_rate": 0.00022805941994812542,
      "loss": 2.3508,
      "step": 1490
    },
    {
      "epoch": 0.9549578226961642,
      "grad_norm": 3.2883720153819804,
      "learning_rate": 0.00022735203961329875,
      "loss": 2.468,
      "step": 1500
    },
    {
      "epoch": 0.9613242081808053,
      "grad_norm": 3.165068282067837,
      "learning_rate": 0.00022664465927847203,
      "loss": 2.3387,
      "step": 1510
    },
    {
      "epoch": 0.9676905936654464,
      "grad_norm": 2.8318847580527895,
      "learning_rate": 0.00022593727894364534,
      "loss": 2.235,
      "step": 1520
    },
    {
      "epoch": 0.9740569791500875,
      "grad_norm": 3.0627223359274556,
      "learning_rate": 0.00022522989860881867,
      "loss": 2.4135,
      "step": 1530
    },
    {
      "epoch": 0.9804233646347287,
      "grad_norm": 4.63667156924246,
      "learning_rate": 0.00022452251827399198,
      "loss": 2.3891,
      "step": 1540
    },
    {
      "epoch": 0.9867897501193698,
      "grad_norm": 2.7155003830350735,
      "learning_rate": 0.00022381513793916526,
      "loss": 2.2814,
      "step": 1550
    },
    {
      "epoch": 0.9931561356040108,
      "grad_norm": 3.743790594304019,
      "learning_rate": 0.00022310775760433856,
      "loss": 2.1729,
      "step": 1560
    },
    {
      "epoch": 0.9995225210886519,
      "grad_norm": 3.0964479659673336,
      "learning_rate": 0.0002224003772695119,
      "loss": 2.3365,
      "step": 1570
    },
    {
      "epoch": 1.005729746936177,
      "grad_norm": 3.9150118918426307,
      "learning_rate": 0.0002216929969346852,
      "loss": 1.9852,
      "step": 1580
    },
    {
      "epoch": 1.012096132420818,
      "grad_norm": 3.5380613834183614,
      "learning_rate": 0.0002209856165998585,
      "loss": 2.0525,
      "step": 1590
    },
    {
      "epoch": 1.0184625179054592,
      "grad_norm": 3.1487817928894613,
      "learning_rate": 0.00022027823626503184,
      "loss": 1.9064,
      "step": 1600
    },
    {
      "epoch": 1.0248289033901004,
      "grad_norm": 3.266369156342291,
      "learning_rate": 0.00021957085593020512,
      "loss": 1.8943,
      "step": 1610
    },
    {
      "epoch": 1.0311952888747413,
      "grad_norm": 3.4336556947723307,
      "learning_rate": 0.00021886347559537842,
      "loss": 1.7861,
      "step": 1620
    },
    {
      "epoch": 1.0375616743593825,
      "grad_norm": 3.832267113304745,
      "learning_rate": 0.00021815609526055173,
      "loss": 2.0424,
      "step": 1630
    },
    {
      "epoch": 1.0439280598440235,
      "grad_norm": 4.130087016761162,
      "learning_rate": 0.00021744871492572506,
      "loss": 1.9139,
      "step": 1640
    },
    {
      "epoch": 1.0502944453286647,
      "grad_norm": 2.7335284776977886,
      "learning_rate": 0.00021674133459089834,
      "loss": 2.0488,
      "step": 1650
    },
    {
      "epoch": 1.0566608308133056,
      "grad_norm": 3.329599622048459,
      "learning_rate": 0.00021603395425607165,
      "loss": 1.9137,
      "step": 1660
    },
    {
      "epoch": 1.0630272162979468,
      "grad_norm": 3.6314777543543313,
      "learning_rate": 0.00021532657392124498,
      "loss": 1.9768,
      "step": 1670
    },
    {
      "epoch": 1.069393601782588,
      "grad_norm": 4.218111606262662,
      "learning_rate": 0.00021461919358641828,
      "loss": 2.0141,
      "step": 1680
    },
    {
      "epoch": 1.075759987267229,
      "grad_norm": 2.895306844768361,
      "learning_rate": 0.0002139118132515916,
      "loss": 1.9783,
      "step": 1690
    },
    {
      "epoch": 1.0821263727518702,
      "grad_norm": 3.753588342038039,
      "learning_rate": 0.00021320443291676492,
      "loss": 2.0227,
      "step": 1700
    },
    {
      "epoch": 1.0884927582365111,
      "grad_norm": 3.250518576003523,
      "learning_rate": 0.0002124970525819382,
      "loss": 1.8848,
      "step": 1710
    },
    {
      "epoch": 1.0948591437211523,
      "grad_norm": 2.980988290019129,
      "learning_rate": 0.0002117896722471115,
      "loss": 1.867,
      "step": 1720
    },
    {
      "epoch": 1.1012255292057933,
      "grad_norm": 3.3311589155605263,
      "learning_rate": 0.0002110822919122848,
      "loss": 1.8766,
      "step": 1730
    },
    {
      "epoch": 1.1075919146904345,
      "grad_norm": 3.2929109952092523,
      "learning_rate": 0.00021037491157745815,
      "loss": 2.0685,
      "step": 1740
    },
    {
      "epoch": 1.1139583001750757,
      "grad_norm": 3.8495434375573243,
      "learning_rate": 0.00020966753124263143,
      "loss": 1.8662,
      "step": 1750
    },
    {
      "epoch": 1.1203246856597167,
      "grad_norm": 3.223669203555661,
      "learning_rate": 0.00020896015090780473,
      "loss": 1.9401,
      "step": 1760
    },
    {
      "epoch": 1.1266910711443578,
      "grad_norm": 3.068725890159831,
      "learning_rate": 0.00020825277057297806,
      "loss": 1.9732,
      "step": 1770
    },
    {
      "epoch": 1.1330574566289988,
      "grad_norm": 3.7652954958158724,
      "learning_rate": 0.00020754539023815137,
      "loss": 1.9264,
      "step": 1780
    },
    {
      "epoch": 1.13942384211364,
      "grad_norm": 3.283263110696873,
      "learning_rate": 0.00020683800990332468,
      "loss": 1.8805,
      "step": 1790
    },
    {
      "epoch": 1.145790227598281,
      "grad_norm": 3.578786877708257,
      "learning_rate": 0.00020613062956849795,
      "loss": 1.8764,
      "step": 1800
    },
    {
      "epoch": 1.1521566130829222,
      "grad_norm": 3.2073175602979402,
      "learning_rate": 0.0002054232492336713,
      "loss": 1.9123,
      "step": 1810
    },
    {
      "epoch": 1.1585229985675634,
      "grad_norm": 3.960679432169151,
      "learning_rate": 0.0002047158688988446,
      "loss": 1.8938,
      "step": 1820
    },
    {
      "epoch": 1.1648893840522043,
      "grad_norm": 3.026836467606075,
      "learning_rate": 0.0002040084885640179,
      "loss": 1.9291,
      "step": 1830
    },
    {
      "epoch": 1.1712557695368455,
      "grad_norm": 3.368777784279614,
      "learning_rate": 0.00020330110822919123,
      "loss": 1.9459,
      "step": 1840
    },
    {
      "epoch": 1.1776221550214865,
      "grad_norm": 2.843063910159576,
      "learning_rate": 0.0002025937278943645,
      "loss": 1.9449,
      "step": 1850
    },
    {
      "epoch": 1.1839885405061277,
      "grad_norm": 2.9510127362455876,
      "learning_rate": 0.00020188634755953782,
      "loss": 1.9309,
      "step": 1860
    },
    {
      "epoch": 1.1903549259907686,
      "grad_norm": 4.545581163313047,
      "learning_rate": 0.00020117896722471115,
      "loss": 1.8869,
      "step": 1870
    },
    {
      "epoch": 1.1967213114754098,
      "grad_norm": 3.301260365676958,
      "learning_rate": 0.00020047158688988445,
      "loss": 1.9613,
      "step": 1880
    },
    {
      "epoch": 1.203087696960051,
      "grad_norm": 3.299718394999854,
      "learning_rate": 0.00019976420655505776,
      "loss": 2.0072,
      "step": 1890
    },
    {
      "epoch": 1.209454082444692,
      "grad_norm": 3.5617690109472444,
      "learning_rate": 0.00019905682622023104,
      "loss": 1.9713,
      "step": 1900
    },
    {
      "epoch": 1.2158204679293332,
      "grad_norm": 3.1359097580279767,
      "learning_rate": 0.00019834944588540437,
      "loss": 1.8154,
      "step": 1910
    },
    {
      "epoch": 1.2221868534139742,
      "grad_norm": 3.0984711664236206,
      "learning_rate": 0.00019764206555057768,
      "loss": 1.8568,
      "step": 1920
    },
    {
      "epoch": 1.2285532388986153,
      "grad_norm": 2.842733095279462,
      "learning_rate": 0.00019693468521575098,
      "loss": 1.851,
      "step": 1930
    },
    {
      "epoch": 1.2349196243832563,
      "grad_norm": 2.935936669675825,
      "learning_rate": 0.00019622730488092432,
      "loss": 1.9604,
      "step": 1940
    },
    {
      "epoch": 1.2412860098678975,
      "grad_norm": 3.3716657028276096,
      "learning_rate": 0.0001955199245460976,
      "loss": 1.9171,
      "step": 1950
    },
    {
      "epoch": 1.2476523953525387,
      "grad_norm": 2.9604032685493133,
      "learning_rate": 0.0001948125442112709,
      "loss": 1.8604,
      "step": 1960
    },
    {
      "epoch": 1.2540187808371797,
      "grad_norm": 2.7269826434015427,
      "learning_rate": 0.0001941051638764442,
      "loss": 1.9076,
      "step": 1970
    },
    {
      "epoch": 1.2603851663218208,
      "grad_norm": 4.046544087868519,
      "learning_rate": 0.00019339778354161754,
      "loss": 1.8299,
      "step": 1980
    },
    {
      "epoch": 1.2667515518064618,
      "grad_norm": 3.12579708206849,
      "learning_rate": 0.00019269040320679084,
      "loss": 1.7545,
      "step": 1990
    },
    {
      "epoch": 1.273117937291103,
      "grad_norm": 2.7391520471543225,
      "learning_rate": 0.00019198302287196412,
      "loss": 1.8973,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 4713,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}