{
  "best_metric": 0.9536241841291653,
  "best_model_checkpoint": "melanoma-v2\\checkpoint-8535",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 8535,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005858230814294083,
      "grad_norm": 11.994911193847656,
      "learning_rate": 4.098360655737705e-06,
      "loss": 0.6539,
      "step": 10
    },
    {
      "epoch": 0.011716461628588167,
      "grad_norm": 18.7818603515625,
      "learning_rate": 9.36768149882904e-06,
      "loss": 0.6538,
      "step": 20
    },
    {
      "epoch": 0.01757469244288225,
      "grad_norm": 7.247508525848389,
      "learning_rate": 1.5222482435597189e-05,
      "loss": 0.648,
      "step": 30
    },
    {
      "epoch": 0.023432923257176334,
      "grad_norm": 23.158416748046875,
      "learning_rate": 2.107728337236534e-05,
      "loss": 0.6352,
      "step": 40
    },
    {
      "epoch": 0.029291154071470416,
      "grad_norm": 12.200428009033203,
      "learning_rate": 2.693208430913349e-05,
      "loss": 0.609,
      "step": 50
    },
    {
      "epoch": 0.0351493848857645,
      "grad_norm": 10.321218490600586,
      "learning_rate": 3.278688524590164e-05,
      "loss": 0.6022,
      "step": 60
    },
    {
      "epoch": 0.041007615700058585,
      "grad_norm": 35.08419418334961,
      "learning_rate": 3.864168618266979e-05,
      "loss": 0.5575,
      "step": 70
    },
    {
      "epoch": 0.04686584651435267,
      "grad_norm": 27.759618759155273,
      "learning_rate": 4.3911007025761124e-05,
      "loss": 0.514,
      "step": 80
    },
    {
      "epoch": 0.05272407732864675,
      "grad_norm": 16.956911087036133,
      "learning_rate": 4.9180327868852456e-05,
      "loss": 0.5048,
      "step": 90
    },
    {
      "epoch": 0.05858230814294083,
      "grad_norm": 23.36929702758789,
      "learning_rate": 5.5035128805620606e-05,
      "loss": 0.4818,
      "step": 100
    },
    {
      "epoch": 0.06444053895723492,
      "grad_norm": 35.85573196411133,
      "learning_rate": 6.0889929742388756e-05,
      "loss": 0.4162,
      "step": 110
    },
    {
      "epoch": 0.070298769771529,
      "grad_norm": 23.913671493530273,
      "learning_rate": 6.674473067915692e-05,
      "loss": 0.4388,
      "step": 120
    },
    {
      "epoch": 0.07615700058582309,
      "grad_norm": 8.763222694396973,
      "learning_rate": 7.259953161592506e-05,
      "loss": 0.3439,
      "step": 130
    },
    {
      "epoch": 0.08201523140011717,
      "grad_norm": 8.551299095153809,
      "learning_rate": 7.845433255269322e-05,
      "loss": 0.3399,
      "step": 140
    },
    {
      "epoch": 0.08787346221441125,
      "grad_norm": 36.60219955444336,
      "learning_rate": 8.430913348946136e-05,
      "loss": 0.353,
      "step": 150
    },
    {
      "epoch": 0.09373169302870533,
      "grad_norm": 16.17277717590332,
      "learning_rate": 9.016393442622952e-05,
      "loss": 0.3191,
      "step": 160
    },
    {
      "epoch": 0.09958992384299942,
      "grad_norm": 22.695581436157227,
      "learning_rate": 9.601873536299766e-05,
      "loss": 0.3274,
      "step": 170
    },
    {
      "epoch": 0.1054481546572935,
      "grad_norm": 14.668584823608398,
      "learning_rate": 0.00010187353629976582,
      "loss": 0.206,
      "step": 180
    },
    {
      "epoch": 0.11130638547158758,
      "grad_norm": 10.050411224365234,
      "learning_rate": 0.00010772833723653396,
      "loss": 0.2838,
      "step": 190
    },
    {
      "epoch": 0.11716461628588166,
      "grad_norm": 8.341797828674316,
      "learning_rate": 0.0001135831381733021,
      "loss": 0.2025,
      "step": 200
    },
    {
      "epoch": 0.12302284710017575,
      "grad_norm": 12.642868995666504,
      "learning_rate": 0.00011943793911007025,
      "loss": 0.2302,
      "step": 210
    },
    {
      "epoch": 0.12888107791446984,
      "grad_norm": 6.499959945678711,
      "learning_rate": 0.00012529274004683842,
      "loss": 0.1666,
      "step": 220
    },
    {
      "epoch": 0.1347393087287639,
      "grad_norm": 26.433914184570312,
      "learning_rate": 0.00013114754098360657,
      "loss": 0.2277,
      "step": 230
    },
    {
      "epoch": 0.140597539543058,
      "grad_norm": 22.857593536376953,
      "learning_rate": 0.00013700234192037472,
      "loss": 0.2807,
      "step": 240
    },
    {
      "epoch": 0.14645577035735208,
      "grad_norm": 6.599829196929932,
      "learning_rate": 0.00014285714285714284,
      "loss": 0.218,
      "step": 250
    },
    {
      "epoch": 0.15231400117164617,
      "grad_norm": 14.18704605102539,
      "learning_rate": 0.00014871194379391102,
      "loss": 0.2214,
      "step": 260
    },
    {
      "epoch": 0.15817223198594024,
      "grad_norm": 6.61388635635376,
      "learning_rate": 0.00015456674473067917,
      "loss": 0.1327,
      "step": 270
    },
    {
      "epoch": 0.16403046280023434,
      "grad_norm": 7.001180171966553,
      "learning_rate": 0.00016042154566744732,
      "loss": 0.1527,
      "step": 280
    },
    {
      "epoch": 0.1698886936145284,
      "grad_norm": 17.781526565551758,
      "learning_rate": 0.00016627634660421544,
      "loss": 0.2506,
      "step": 290
    },
    {
      "epoch": 0.1757469244288225,
      "grad_norm": 11.289440155029297,
      "learning_rate": 0.00017213114754098362,
      "loss": 0.2424,
      "step": 300
    },
    {
      "epoch": 0.18160515524311657,
      "grad_norm": 14.80530834197998,
      "learning_rate": 0.00017798594847775177,
      "loss": 0.2481,
      "step": 310
    },
    {
      "epoch": 0.18746338605741067,
      "grad_norm": 5.781233310699463,
      "learning_rate": 0.00018384074941451992,
      "loss": 0.203,
      "step": 320
    },
    {
      "epoch": 0.19332161687170474,
      "grad_norm": 4.2908477783203125,
      "learning_rate": 0.00018969555035128804,
      "loss": 0.1923,
      "step": 330
    },
    {
      "epoch": 0.19917984768599883,
      "grad_norm": 3.176772117614746,
      "learning_rate": 0.00019555035128805622,
      "loss": 0.1573,
      "step": 340
    },
    {
      "epoch": 0.2050380785002929,
      "grad_norm": 22.84259605407715,
      "learning_rate": 0.00020140515222482437,
      "loss": 0.2,
      "step": 350
    },
    {
      "epoch": 0.210896309314587,
      "grad_norm": 3.60316801071167,
      "learning_rate": 0.00020725995316159252,
      "loss": 0.2558,
      "step": 360
    },
    {
      "epoch": 0.21675454012888107,
      "grad_norm": 15.83247184753418,
      "learning_rate": 0.00021311475409836064,
      "loss": 0.1953,
      "step": 370
    },
    {
      "epoch": 0.22261277094317516,
      "grad_norm": 3.053990602493286,
      "learning_rate": 0.00021896955503512882,
      "loss": 0.2094,
      "step": 380
    },
    {
      "epoch": 0.22847100175746923,
      "grad_norm": 4.581020355224609,
      "learning_rate": 0.00022482435597189697,
      "loss": 0.166,
      "step": 390
    },
    {
      "epoch": 0.23432923257176333,
      "grad_norm": 4.517653465270996,
      "learning_rate": 0.00023067915690866512,
      "loss": 0.1749,
      "step": 400
    },
    {
      "epoch": 0.2401874633860574,
      "grad_norm": 15.956161499023438,
      "learning_rate": 0.00023653395784543327,
      "loss": 0.2458,
      "step": 410
    },
    {
      "epoch": 0.2460456942003515,
      "grad_norm": 3.738513231277466,
      "learning_rate": 0.0002423887587822014,
      "loss": 0.2622,
      "step": 420
    },
    {
      "epoch": 0.2519039250146456,
      "grad_norm": 4.261665344238281,
      "learning_rate": 0.00024824355971896957,
      "loss": 0.21,
      "step": 430
    },
    {
      "epoch": 0.2577621558289397,
      "grad_norm": 15.049210548400879,
      "learning_rate": 0.0002540983606557377,
      "loss": 0.2058,
      "step": 440
    },
    {
      "epoch": 0.26362038664323373,
      "grad_norm": 2.9831387996673584,
      "learning_rate": 0.00025995316159250587,
      "loss": 0.1936,
      "step": 450
    },
    {
      "epoch": 0.2694786174575278,
      "grad_norm": 7.315589904785156,
      "learning_rate": 0.000265807962529274,
      "loss": 0.2204,
      "step": 460
    },
    {
      "epoch": 0.2753368482718219,
      "grad_norm": 9.33542251586914,
      "learning_rate": 0.00027166276346604217,
      "loss": 0.2543,
      "step": 470
    },
    {
      "epoch": 0.281195079086116,
      "grad_norm": 4.937420845031738,
      "learning_rate": 0.00027751756440281026,
      "loss": 0.2941,
      "step": 480
    },
    {
      "epoch": 0.28705330990041006,
      "grad_norm": 1.6123785972595215,
      "learning_rate": 0.00028337236533957846,
      "loss": 0.2231,
      "step": 490
    },
    {
      "epoch": 0.29291154071470415,
      "grad_norm": 2.655095338821411,
      "learning_rate": 0.0002892271662763466,
      "loss": 0.16,
      "step": 500
    },
    {
      "epoch": 0.29876977152899825,
      "grad_norm": 5.088364601135254,
      "learning_rate": 0.00029508196721311476,
      "loss": 0.2443,
      "step": 510
    },
    {
      "epoch": 0.30462800234329235,
      "grad_norm": 13.077324867248535,
      "learning_rate": 0.0003009367681498829,
      "loss": 0.2526,
      "step": 520
    },
    {
      "epoch": 0.3104862331575864,
      "grad_norm": 3.225377082824707,
      "learning_rate": 0.00030679156908665106,
      "loss": 0.2461,
      "step": 530
    },
    {
      "epoch": 0.3163444639718805,
      "grad_norm": 2.392651319503784,
      "learning_rate": 0.0003126463700234192,
      "loss": 0.2216,
      "step": 540
    },
    {
      "epoch": 0.3222026947861746,
      "grad_norm": 1.9468374252319336,
      "learning_rate": 0.00031850117096018736,
      "loss": 0.1632,
      "step": 550
    },
    {
      "epoch": 0.3280609256004687,
      "grad_norm": 1.429714560508728,
      "learning_rate": 0.0003243559718969555,
      "loss": 0.232,
      "step": 560
    },
    {
      "epoch": 0.3339191564147627,
      "grad_norm": 2.8054699897766113,
      "learning_rate": 0.00033021077283372366,
      "loss": 0.1954,
      "step": 570
    },
    {
      "epoch": 0.3397773872290568,
      "grad_norm": 3.628207206726074,
      "learning_rate": 0.0003360655737704918,
      "loss": 0.1579,
      "step": 580
    },
    {
      "epoch": 0.3456356180433509,
      "grad_norm": 1.0688583850860596,
      "learning_rate": 0.00034192037470725996,
      "loss": 0.2699,
      "step": 590
    },
    {
      "epoch": 0.351493848857645,
      "grad_norm": 5.8670220375061035,
      "learning_rate": 0.0003477751756440281,
      "loss": 0.2075,
      "step": 600
    },
    {
      "epoch": 0.35735207967193905,
      "grad_norm": 2.418936252593994,
      "learning_rate": 0.00035362997658079626,
      "loss": 0.2845,
      "step": 610
    },
    {
      "epoch": 0.36321031048623315,
      "grad_norm": 1.8158751726150513,
      "learning_rate": 0.0003594847775175644,
      "loss": 0.1336,
      "step": 620
    },
    {
      "epoch": 0.36906854130052724,
      "grad_norm": 3.2167751789093018,
      "learning_rate": 0.00036533957845433256,
      "loss": 0.1885,
      "step": 630
    },
    {
      "epoch": 0.37492677211482134,
      "grad_norm": 0.4845765233039856,
      "learning_rate": 0.0003711943793911007,
      "loss": 0.179,
      "step": 640
    },
    {
      "epoch": 0.38078500292911543,
      "grad_norm": 1.4025683403015137,
      "learning_rate": 0.0003770491803278688,
      "loss": 0.1934,
      "step": 650
    },
    {
      "epoch": 0.3866432337434095,
      "grad_norm": 3.7113711833953857,
      "learning_rate": 0.000382903981264637,
      "loss": 0.2392,
      "step": 660
    },
    {
      "epoch": 0.39250146455770357,
      "grad_norm": 0.9925859570503235,
      "learning_rate": 0.00038875878220140516,
      "loss": 0.1762,
      "step": 670
    },
    {
      "epoch": 0.39835969537199767,
      "grad_norm": 2.3232357501983643,
      "learning_rate": 0.0003946135831381733,
      "loss": 0.3168,
      "step": 680
    },
    {
      "epoch": 0.40421792618629176,
      "grad_norm": 2.2389273643493652,
      "learning_rate": 0.00040046838407494146,
      "loss": 0.1604,
      "step": 690
    },
    {
      "epoch": 0.4100761570005858,
      "grad_norm": 3.8855361938476562,
      "learning_rate": 0.0004063231850117096,
      "loss": 0.1886,
      "step": 700
    },
    {
      "epoch": 0.4159343878148799,
      "grad_norm": 3.099923610687256,
      "learning_rate": 0.00041217798594847776,
      "loss": 0.2951,
      "step": 710
    },
    {
      "epoch": 0.421792618629174,
      "grad_norm": 4.365396022796631,
      "learning_rate": 0.0004180327868852459,
      "loss": 0.1872,
      "step": 720
    },
    {
      "epoch": 0.4276508494434681,
      "grad_norm": 1.8989065885543823,
      "learning_rate": 0.00042388758782201406,
      "loss": 0.204,
      "step": 730
    },
    {
      "epoch": 0.43350908025776214,
      "grad_norm": 1.8920644521713257,
      "learning_rate": 0.0004297423887587822,
      "loss": 0.2229,
      "step": 740
    },
    {
      "epoch": 0.43936731107205623,
      "grad_norm": 2.187156915664673,
      "learning_rate": 0.00043559718969555036,
      "loss": 0.1378,
      "step": 750
    },
    {
      "epoch": 0.44522554188635033,
      "grad_norm": 3.7022440433502197,
      "learning_rate": 0.0004414519906323185,
      "loss": 0.2533,
      "step": 760
    },
    {
      "epoch": 0.4510837727006444,
      "grad_norm": 2.7125163078308105,
      "learning_rate": 0.00044730679156908666,
      "loss": 0.1884,
      "step": 770
    },
    {
      "epoch": 0.45694200351493847,
      "grad_norm": 3.3647499084472656,
      "learning_rate": 0.0004531615925058548,
      "loss": 0.2834,
      "step": 780
    },
    {
      "epoch": 0.46280023432923256,
      "grad_norm": 0.8961579203605652,
      "learning_rate": 0.00045901639344262296,
      "loss": 0.228,
      "step": 790
    },
    {
      "epoch": 0.46865846514352666,
      "grad_norm": 1.321513056755066,
      "learning_rate": 0.0004648711943793911,
      "loss": 0.1791,
      "step": 800
    },
    {
      "epoch": 0.47451669595782076,
      "grad_norm": 1.7761132717132568,
      "learning_rate": 0.00047072599531615926,
      "loss": 0.2495,
      "step": 810
    },
    {
      "epoch": 0.4803749267721148,
      "grad_norm": 2.2934768199920654,
      "learning_rate": 0.0004765807962529274,
      "loss": 0.201,
      "step": 820
    },
    {
      "epoch": 0.4862331575864089,
      "grad_norm": 2.120138645172119,
      "learning_rate": 0.00048243559718969556,
      "loss": 0.2572,
      "step": 830
    },
    {
      "epoch": 0.492091388400703,
      "grad_norm": 1.623761534690857,
      "learning_rate": 0.0004882903981264637,
      "loss": 0.2508,
      "step": 840
    },
    {
      "epoch": 0.4979496192149971,
      "grad_norm": 1.8615626096725464,
      "learning_rate": 0.0004941451990632318,
      "loss": 0.2244,
      "step": 850
    },
    {
      "epoch": 0.5038078500292912,
      "grad_norm": 1.1661752462387085,
      "learning_rate": 0.0005,
      "loss": 0.258,
      "step": 860
    },
    {
      "epoch": 0.5096660808435852,
      "grad_norm": 1.3036870956420898,
      "learning_rate": 0.0004993490430933472,
      "loss": 0.2324,
      "step": 870
    },
    {
      "epoch": 0.5155243116578794,
      "grad_norm": 3.418734073638916,
      "learning_rate": 0.0004986980861866944,
      "loss": 0.2483,
      "step": 880
    },
    {
      "epoch": 0.5213825424721734,
      "grad_norm": 1.805756688117981,
      "learning_rate": 0.0004980471292800417,
      "loss": 0.1911,
      "step": 890
    },
    {
      "epoch": 0.5272407732864675,
      "grad_norm": 0.6934542655944824,
      "learning_rate": 0.0004973961723733889,
      "loss": 0.2273,
      "step": 900
    },
    {
      "epoch": 0.5330990041007616,
      "grad_norm": 1.0683369636535645,
      "learning_rate": 0.0004967452154667361,
      "loss": 0.1814,
      "step": 910
    },
    {
      "epoch": 0.5389572349150556,
      "grad_norm": 0.795211672782898,
      "learning_rate": 0.0004960942585600834,
      "loss": 0.2065,
      "step": 920
    },
    {
      "epoch": 0.5448154657293497,
      "grad_norm": 0.7547960877418518,
      "learning_rate": 0.0004954433016534306,
      "loss": 0.147,
      "step": 930
    },
    {
      "epoch": 0.5506736965436438,
      "grad_norm": 0.7984468340873718,
      "learning_rate": 0.0004947923447467778,
      "loss": 0.2025,
      "step": 940
    },
    {
      "epoch": 0.5565319273579379,
      "grad_norm": 0.9543431997299194,
      "learning_rate": 0.000494141387840125,
      "loss": 0.2016,
      "step": 950
    },
    {
      "epoch": 0.562390158172232,
      "grad_norm": 0.5806155800819397,
      "learning_rate": 0.0004934904309334722,
      "loss": 0.1749,
      "step": 960
    },
    {
      "epoch": 0.5682483889865261,
      "grad_norm": 0.5972631573677063,
      "learning_rate": 0.0004928394740268194,
      "loss": 0.176,
      "step": 970
    },
    {
      "epoch": 0.5741066198008201,
      "grad_norm": 0.4694497585296631,
      "learning_rate": 0.0004921885171201666,
      "loss": 0.2667,
      "step": 980
    },
    {
      "epoch": 0.5799648506151143,
      "grad_norm": 0.84099942445755,
      "learning_rate": 0.0004915375602135138,
      "loss": 0.1912,
      "step": 990
    },
    {
      "epoch": 0.5858230814294083,
      "grad_norm": 2.690175771713257,
      "learning_rate": 0.0004908866033068611,
      "loss": 0.1685,
      "step": 1000
    },
    {
      "epoch": 0.5916813122437024,
      "grad_norm": 1.2537461519241333,
      "learning_rate": 0.0004902356464002083,
      "loss": 0.2628,
      "step": 1010
    },
    {
      "epoch": 0.5975395430579965,
      "grad_norm": 3.283203363418579,
      "learning_rate": 0.0004895846894935555,
      "loss": 0.2186,
      "step": 1020
    },
    {
      "epoch": 0.6033977738722905,
      "grad_norm": 1.1899466514587402,
      "learning_rate": 0.0004889337325869028,
      "loss": 0.178,
      "step": 1030
    },
    {
      "epoch": 0.6092560046865847,
      "grad_norm": 2.0233116149902344,
      "learning_rate": 0.00048828277568025,
      "loss": 0.2323,
      "step": 1040
    },
    {
      "epoch": 0.6151142355008787,
      "grad_norm": 1.075803279876709,
      "learning_rate": 0.0004876318187735972,
      "loss": 0.1844,
      "step": 1050
    },
    {
      "epoch": 0.6209724663151728,
      "grad_norm": 1.9664595127105713,
      "learning_rate": 0.0004869808618669444,
      "loss": 0.1697,
      "step": 1060
    },
    {
      "epoch": 0.6268306971294669,
      "grad_norm": 0.497203528881073,
      "learning_rate": 0.0004863299049602916,
      "loss": 0.1803,
      "step": 1070
    },
    {
      "epoch": 0.632688927943761,
      "grad_norm": 1.3227007389068604,
      "learning_rate": 0.00048567894805363887,
      "loss": 0.2132,
      "step": 1080
    },
    {
      "epoch": 0.6385471587580551,
      "grad_norm": 1.1600663661956787,
      "learning_rate": 0.0004850279911469861,
      "loss": 0.2151,
      "step": 1090
    },
    {
      "epoch": 0.6444053895723492,
      "grad_norm": 1.4555237293243408,
      "learning_rate": 0.0004843770342403333,
      "loss": 0.2224,
      "step": 1100
    },
    {
      "epoch": 0.6502636203866432,
      "grad_norm": 1.7892956733703613,
      "learning_rate": 0.0004837260773336805,
      "loss": 0.2004,
      "step": 1110
    },
    {
      "epoch": 0.6561218512009374,
      "grad_norm": 0.6487146615982056,
      "learning_rate": 0.00048307512042702774,
      "loss": 0.136,
      "step": 1120
    },
    {
      "epoch": 0.6619800820152314,
      "grad_norm": 1.258460521697998,
      "learning_rate": 0.00048242416352037494,
      "loss": 0.1574,
      "step": 1130
    },
    {
      "epoch": 0.6678383128295254,
      "grad_norm": 1.852683424949646,
      "learning_rate": 0.0004817732066137222,
      "loss": 0.1983,
      "step": 1140
    },
    {
      "epoch": 0.6736965436438196,
      "grad_norm": 0.7946725487709045,
      "learning_rate": 0.0004811222497070694,
      "loss": 0.2243,
      "step": 1150
    },
    {
      "epoch": 0.6795547744581136,
      "grad_norm": 2.0183424949645996,
      "learning_rate": 0.0004804712928004166,
      "loss": 0.2063,
      "step": 1160
    },
    {
      "epoch": 0.6854130052724078,
      "grad_norm": 1.2471932172775269,
      "learning_rate": 0.00047982033589376386,
      "loss": 0.1675,
      "step": 1170
    },
    {
      "epoch": 0.6912712360867018,
      "grad_norm": 1.7430349588394165,
      "learning_rate": 0.00047916937898711105,
      "loss": 0.1867,
      "step": 1180
    },
    {
      "epoch": 0.6971294669009959,
      "grad_norm": 0.6978306770324707,
      "learning_rate": 0.00047851842208045825,
      "loss": 0.1756,
      "step": 1190
    },
    {
      "epoch": 0.70298769771529,
      "grad_norm": 1.8298044204711914,
      "learning_rate": 0.00047786746517380555,
      "loss": 0.199,
      "step": 1200
    },
    {
      "epoch": 0.7088459285295841,
      "grad_norm": 1.6273216009140015,
      "learning_rate": 0.00047721650826715274,
      "loss": 0.2137,
      "step": 1210
    },
    {
      "epoch": 0.7147041593438781,
      "grad_norm": 0.8626360893249512,
      "learning_rate": 0.00047656555136049993,
      "loss": 0.1602,
      "step": 1220
    },
    {
      "epoch": 0.7205623901581723,
      "grad_norm": 1.0214688777923584,
      "learning_rate": 0.0004759145944538472,
      "loss": 0.2219,
      "step": 1230
    },
    {
      "epoch": 0.7264206209724663,
      "grad_norm": 1.6713732481002808,
      "learning_rate": 0.00047526363754719437,
      "loss": 0.2086,
      "step": 1240
    },
    {
      "epoch": 0.7322788517867604,
      "grad_norm": 1.6285948753356934,
      "learning_rate": 0.00047461268064054156,
      "loss": 0.1886,
      "step": 1250
    },
    {
      "epoch": 0.7381370826010545,
      "grad_norm": 1.1279932260513306,
      "learning_rate": 0.00047396172373388886,
      "loss": 0.1786,
      "step": 1260
    },
    {
      "epoch": 0.7439953134153485,
      "grad_norm": 1.1999080181121826,
      "learning_rate": 0.00047331076682723605,
      "loss": 0.2515,
      "step": 1270
    },
    {
      "epoch": 0.7498535442296427,
      "grad_norm": 0.6995332837104797,
      "learning_rate": 0.0004726598099205833,
      "loss": 0.193,
      "step": 1280
    },
    {
      "epoch": 0.7557117750439367,
      "grad_norm": 1.1866198778152466,
      "learning_rate": 0.0004720088530139305,
      "loss": 0.1641,
      "step": 1290
    },
    {
      "epoch": 0.7615700058582309,
      "grad_norm": 0.7438950538635254,
      "learning_rate": 0.0004713578961072777,
      "loss": 0.2166,
      "step": 1300
    },
    {
      "epoch": 0.7674282366725249,
      "grad_norm": 1.5525139570236206,
      "learning_rate": 0.0004707069392006249,
      "loss": 0.1733,
      "step": 1310
    },
    {
      "epoch": 0.773286467486819,
      "grad_norm": 1.763756275177002,
      "learning_rate": 0.00047005598229397217,
      "loss": 0.1777,
      "step": 1320
    },
    {
      "epoch": 0.7791446983011131,
      "grad_norm": 1.4447243213653564,
      "learning_rate": 0.00046940502538731936,
      "loss": 0.1423,
      "step": 1330
    },
    {
      "epoch": 0.7850029291154071,
      "grad_norm": 0.3512236177921295,
      "learning_rate": 0.0004687540684806666,
      "loss": 0.2049,
      "step": 1340
    },
    {
      "epoch": 0.7908611599297012,
      "grad_norm": 1.6187587976455688,
      "learning_rate": 0.0004681031115740138,
      "loss": 0.2369,
      "step": 1350
    },
    {
      "epoch": 0.7967193907439953,
      "grad_norm": 1.3473628759384155,
      "learning_rate": 0.00046745215466736104,
      "loss": 0.2218,
      "step": 1360
    },
    {
      "epoch": 0.8025776215582894,
      "grad_norm": 0.4788994789123535,
      "learning_rate": 0.00046680119776070823,
      "loss": 0.1477,
      "step": 1370
    },
    {
      "epoch": 0.8084358523725835,
      "grad_norm": 1.1703928709030151,
      "learning_rate": 0.0004661502408540555,
      "loss": 0.2457,
      "step": 1380
    },
    {
      "epoch": 0.8142940831868776,
      "grad_norm": 1.6234064102172852,
      "learning_rate": 0.0004654992839474027,
      "loss": 0.232,
      "step": 1390
    },
    {
      "epoch": 0.8201523140011716,
      "grad_norm": 0.4102720618247986,
      "learning_rate": 0.0004648483270407499,
      "loss": 0.1856,
      "step": 1400
    },
    {
      "epoch": 0.8260105448154658,
      "grad_norm": 0.49339228868484497,
      "learning_rate": 0.0004641973701340971,
      "loss": 0.1961,
      "step": 1410
    },
    {
      "epoch": 0.8318687756297598,
      "grad_norm": 1.422866702079773,
      "learning_rate": 0.00046354641322744435,
      "loss": 0.2002,
      "step": 1420
    },
    {
      "epoch": 0.8377270064440538,
      "grad_norm": 2.2886526584625244,
      "learning_rate": 0.00046289545632079154,
      "loss": 0.2098,
      "step": 1430
    },
    {
      "epoch": 0.843585237258348,
      "grad_norm": 1.019761085510254,
      "learning_rate": 0.00046224449941413884,
      "loss": 0.1656,
      "step": 1440
    },
    {
      "epoch": 0.849443468072642,
      "grad_norm": 0.9850839972496033,
      "learning_rate": 0.00046159354250748603,
      "loss": 0.1725,
      "step": 1450
    },
    {
      "epoch": 0.8553016988869362,
      "grad_norm": 0.8339518308639526,
      "learning_rate": 0.0004609425856008332,
      "loss": 0.1312,
      "step": 1460
    },
    {
      "epoch": 0.8611599297012302,
      "grad_norm": 2.1276755332946777,
      "learning_rate": 0.00046029162869418047,
      "loss": 0.1281,
      "step": 1470
    },
    {
      "epoch": 0.8670181605155243,
      "grad_norm": 2.526759386062622,
      "learning_rate": 0.00045964067178752766,
      "loss": 0.2324,
      "step": 1480
    },
    {
      "epoch": 0.8728763913298184,
      "grad_norm": 1.4450796842575073,
      "learning_rate": 0.00045898971488087485,
      "loss": 0.191,
      "step": 1490
    },
    {
      "epoch": 0.8787346221441125,
      "grad_norm": 1.1786580085754395,
      "learning_rate": 0.00045833875797422215,
      "loss": 0.2706,
      "step": 1500
    },
    {
      "epoch": 0.8845928529584065,
      "grad_norm": 1.5052456855773926,
      "learning_rate": 0.00045768780106756934,
      "loss": 0.2679,
      "step": 1510
    },
    {
      "epoch": 0.8904510837727007,
      "grad_norm": 0.3879900872707367,
      "learning_rate": 0.0004570368441609166,
      "loss": 0.1693,
      "step": 1520
    },
    {
      "epoch": 0.8963093145869947,
      "grad_norm": 1.613844394683838,
      "learning_rate": 0.0004563858872542638,
      "loss": 0.2008,
      "step": 1530
    },
    {
      "epoch": 0.9021675454012889,
      "grad_norm": 1.3180102109909058,
      "learning_rate": 0.00045573493034761097,
      "loss": 0.2123,
      "step": 1540
    },
    {
      "epoch": 0.9080257762155829,
      "grad_norm": 1.0241358280181885,
      "learning_rate": 0.0004550839734409582,
      "loss": 0.153,
      "step": 1550
    },
    {
      "epoch": 0.9138840070298769,
      "grad_norm": 0.9837117791175842,
      "learning_rate": 0.00045443301653430546,
      "loss": 0.2067,
      "step": 1560
    },
    {
      "epoch": 0.9197422378441711,
      "grad_norm": 0.6446996927261353,
      "learning_rate": 0.00045378205962765266,
      "loss": 0.1446,
      "step": 1570
    },
    {
      "epoch": 0.9256004686584651,
      "grad_norm": 0.733072817325592,
      "learning_rate": 0.0004531311027209999,
      "loss": 0.1972,
      "step": 1580
    },
    {
      "epoch": 0.9314586994727593,
      "grad_norm": 1.2420257329940796,
      "learning_rate": 0.0004524801458143471,
      "loss": 0.1854,
      "step": 1590
    },
    {
      "epoch": 0.9373169302870533,
      "grad_norm": 2.1636710166931152,
      "learning_rate": 0.00045182918890769434,
      "loss": 0.2689,
      "step": 1600
    },
    {
      "epoch": 0.9431751611013474,
      "grad_norm": 0.833899199962616,
      "learning_rate": 0.00045117823200104153,
      "loss": 0.1944,
      "step": 1610
    },
    {
      "epoch": 0.9490333919156415,
      "grad_norm": 0.98338782787323,
      "learning_rate": 0.0004505272750943887,
      "loss": 0.1568,
      "step": 1620
    },
    {
      "epoch": 0.9548916227299356,
      "grad_norm": 0.2556711435317993,
      "learning_rate": 0.000449876318187736,
      "loss": 0.145,
      "step": 1630
    },
    {
      "epoch": 0.9607498535442296,
      "grad_norm": 1.3900483846664429,
      "learning_rate": 0.0004492253612810832,
      "loss": 0.1975,
      "step": 1640
    },
    {
      "epoch": 0.9666080843585237,
      "grad_norm": 0.9346416592597961,
      "learning_rate": 0.0004485744043744304,
      "loss": 0.2289,
      "step": 1650
    },
    {
      "epoch": 0.9724663151728178,
      "grad_norm": 0.9890295267105103,
      "learning_rate": 0.00044792344746777765,
      "loss": 0.1623,
      "step": 1660
    },
    {
      "epoch": 0.9783245459871119,
      "grad_norm": 0.3821526765823364,
      "learning_rate": 0.00044727249056112484,
      "loss": 0.1841,
      "step": 1670
    },
    {
      "epoch": 0.984182776801406,
      "grad_norm": 0.9447836875915527,
      "learning_rate": 0.00044662153365447203,
      "loss": 0.155,
      "step": 1680
    },
    {
      "epoch": 0.9900410076157,
      "grad_norm": 1.4631438255310059,
      "learning_rate": 0.00044597057674781933,
      "loss": 0.1878,
      "step": 1690
    },
    {
      "epoch": 0.9958992384299942,
      "grad_norm": 0.7002029418945312,
      "learning_rate": 0.0004453196198411665,
      "loss": 0.1585,
      "step": 1700
    },
    {
      "epoch": 1.0,
      "eval_f1": 0.8884904983865185,
      "eval_loss": 0.29340454936027527,
      "eval_runtime": 15.1829,
      "eval_samples_per_second": 197.591,
      "eval_steps_per_second": 6.191,
      "step": 1707
    },
    {
      "epoch": 1.0017574692442883,
      "grad_norm": 1.4538880586624146,
      "learning_rate": 0.00044466866293451377,
      "loss": 0.1659,
      "step": 1710
    },
    {
      "epoch": 1.0076157000585824,
      "grad_norm": 1.1730880737304688,
      "learning_rate": 0.00044401770602786096,
      "loss": 0.2061,
      "step": 1720
    },
    {
      "epoch": 1.0134739308728764,
      "grad_norm": 1.1373745203018188,
      "learning_rate": 0.00044336674912120815,
      "loss": 0.1307,
      "step": 1730
    },
    {
      "epoch": 1.0193321616871704,
      "grad_norm": 1.5837347507476807,
      "learning_rate": 0.0004427157922145554,
      "loss": 0.1837,
      "step": 1740
    },
    {
      "epoch": 1.0251903925014645,
      "grad_norm": 1.2740920782089233,
      "learning_rate": 0.00044206483530790264,
      "loss": 0.2228,
      "step": 1750
    },
    {
      "epoch": 1.0310486233157587,
      "grad_norm": 0.7164502739906311,
      "learning_rate": 0.00044141387840124983,
      "loss": 0.1628,
      "step": 1760
    },
    {
      "epoch": 1.0369068541300528,
      "grad_norm": 0.4816601872444153,
      "learning_rate": 0.0004407629214945971,
      "loss": 0.1548,
      "step": 1770
    },
    {
      "epoch": 1.0427650849443468,
      "grad_norm": 2.0663034915924072,
      "learning_rate": 0.00044011196458794427,
      "loss": 0.1997,
      "step": 1780
    },
    {
      "epoch": 1.0486233157586409,
      "grad_norm": 0.48207902908325195,
      "learning_rate": 0.0004394610076812915,
      "loss": 0.1376,
      "step": 1790
    },
    {
      "epoch": 1.054481546572935,
      "grad_norm": 1.8158009052276611,
      "learning_rate": 0.0004388100507746387,
      "loss": 0.1746,
      "step": 1800
    },
    {
      "epoch": 1.060339777387229,
      "grad_norm": 0.8723987936973572,
      "learning_rate": 0.00043815909386798595,
      "loss": 0.182,
      "step": 1810
    },
    {
      "epoch": 1.0661980082015232,
      "grad_norm": 1.8085635900497437,
      "learning_rate": 0.0004375081369613332,
      "loss": 0.1686,
      "step": 1820
    },
    {
      "epoch": 1.0720562390158173,
      "grad_norm": 1.800154447555542,
      "learning_rate": 0.0004368571800546804,
      "loss": 0.1744,
      "step": 1830
    },
    {
      "epoch": 1.0779144698301113,
      "grad_norm": 0.9144867062568665,
      "learning_rate": 0.0004362062231480276,
      "loss": 0.1776,
      "step": 1840
    },
    {
      "epoch": 1.0837727006444053,
      "grad_norm": 1.4381457567214966,
      "learning_rate": 0.0004355552662413748,
      "loss": 0.1944,
      "step": 1850
    },
    {
      "epoch": 1.0896309314586994,
      "grad_norm": 0.435210645198822,
      "learning_rate": 0.000434904309334722,
      "loss": 0.1903,
      "step": 1860
    },
    {
      "epoch": 1.0954891622729936,
      "grad_norm": 1.2425928115844727,
      "learning_rate": 0.0004342533524280693,
      "loss": 0.1455,
      "step": 1870
    },
    {
      "epoch": 1.1013473930872877,
      "grad_norm": 1.3132634162902832,
      "learning_rate": 0.0004336023955214165,
      "loss": 0.1694,
      "step": 1880
    },
    {
      "epoch": 1.1072056239015817,
      "grad_norm": 0.8136222958564758,
      "learning_rate": 0.0004329514386147637,
      "loss": 0.1287,
      "step": 1890
    },
    {
      "epoch": 1.1130638547158758,
      "grad_norm": 1.0159581899642944,
      "learning_rate": 0.00043230048170811095,
      "loss": 0.1716,
      "step": 1900
    },
    {
      "epoch": 1.1189220855301698,
      "grad_norm": 1.047573208808899,
      "learning_rate": 0.00043164952480145814,
      "loss": 0.203,
      "step": 1910
    },
    {
      "epoch": 1.124780316344464,
      "grad_norm": 0.5778195858001709,
      "learning_rate": 0.00043099856789480533,
      "loss": 0.1901,
      "step": 1920
    },
    {
      "epoch": 1.1306385471587581,
      "grad_norm": 0.5892320871353149,
      "learning_rate": 0.00043034761098815263,
      "loss": 0.1872,
      "step": 1930
    },
    {
      "epoch": 1.1364967779730522,
      "grad_norm": 0.8339682817459106,
      "learning_rate": 0.0004296966540814998,
      "loss": 0.1555,
      "step": 1940
    },
    {
      "epoch": 1.1423550087873462,
      "grad_norm": 1.1029021739959717,
      "learning_rate": 0.00042904569717484706,
      "loss": 0.1283,
      "step": 1950
    },
    {
      "epoch": 1.1482132396016402,
      "grad_norm": 0.9936773180961609,
      "learning_rate": 0.00042839474026819426,
      "loss": 0.157,
      "step": 1960
    },
    {
      "epoch": 1.1540714704159343,
      "grad_norm": 1.108543038368225,
      "learning_rate": 0.00042774378336154145,
      "loss": 0.1472,
      "step": 1970
    },
    {
      "epoch": 1.1599297012302285,
      "grad_norm": 0.6184877157211304,
      "learning_rate": 0.0004270928264548887,
      "loss": 0.127,
      "step": 1980
    },
    {
      "epoch": 1.1657879320445226,
      "grad_norm": 0.9069430828094482,
      "learning_rate": 0.00042644186954823594,
      "loss": 0.1197,
      "step": 1990
    },
    {
      "epoch": 1.1716461628588166,
      "grad_norm": 1.8886994123458862,
      "learning_rate": 0.00042579091264158313,
      "loss": 0.1614,
      "step": 2000
    },
    {
      "epoch": 1.1775043936731107,
      "grad_norm": 0.9473972916603088,
      "learning_rate": 0.0004251399557349304,
      "loss": 0.1793,
      "step": 2010
    },
    {
      "epoch": 1.1833626244874047,
      "grad_norm": 0.49200665950775146,
      "learning_rate": 0.00042448899882827757,
      "loss": 0.1558,
      "step": 2020
    },
    {
      "epoch": 1.189220855301699,
      "grad_norm": 1.1897913217544556,
      "learning_rate": 0.00042383804192162476,
      "loss": 0.1287,
      "step": 2030
    },
    {
      "epoch": 1.195079086115993,
      "grad_norm": 0.4664163291454315,
      "learning_rate": 0.000423187085014972,
      "loss": 0.1385,
      "step": 2040
    },
    {
      "epoch": 1.200937316930287,
      "grad_norm": 0.6468362808227539,
      "learning_rate": 0.00042253612810831925,
      "loss": 0.1646,
      "step": 2050
    },
    {
      "epoch": 1.206795547744581,
      "grad_norm": 0.9662730097770691,
      "learning_rate": 0.0004218851712016665,
      "loss": 0.1461,
      "step": 2060
    },
    {
      "epoch": 1.2126537785588751,
      "grad_norm": 1.2626930475234985,
      "learning_rate": 0.0004212342142950137,
      "loss": 0.1373,
      "step": 2070
    },
    {
      "epoch": 1.2185120093731694,
      "grad_norm": 2.0867109298706055,
      "learning_rate": 0.0004205832573883609,
      "loss": 0.2043,
      "step": 2080
    },
    {
      "epoch": 1.2243702401874634,
      "grad_norm": 1.0699323415756226,
      "learning_rate": 0.0004199323004817081,
      "loss": 0.1493,
      "step": 2090
    },
    {
      "epoch": 1.2302284710017575,
      "grad_norm": 0.5708653926849365,
      "learning_rate": 0.0004192813435750553,
      "loss": 0.1596,
      "step": 2100
    },
    {
      "epoch": 1.2360867018160515,
      "grad_norm": 1.6544972658157349,
      "learning_rate": 0.00041863038666840256,
      "loss": 0.1858,
      "step": 2110
    },
    {
      "epoch": 1.2419449326303456,
      "grad_norm": 1.7309073209762573,
      "learning_rate": 0.0004179794297617498,
      "loss": 0.093,
      "step": 2120
    },
    {
      "epoch": 1.2478031634446398,
      "grad_norm": 1.0835295915603638,
      "learning_rate": 0.000417328472855097,
      "loss": 0.1629,
      "step": 2130
    },
    {
      "epoch": 1.2536613942589339,
      "grad_norm": 0.587459146976471,
      "learning_rate": 0.00041667751594844424,
      "loss": 0.1727,
      "step": 2140
    },
    {
      "epoch": 1.259519625073228,
      "grad_norm": 3.7887632846832275,
      "learning_rate": 0.00041602655904179143,
      "loss": 0.1967,
      "step": 2150
    },
    {
      "epoch": 1.265377855887522,
      "grad_norm": 1.1685255765914917,
      "learning_rate": 0.0004153756021351386,
      "loss": 0.1311,
      "step": 2160
    },
    {
      "epoch": 1.271236086701816,
      "grad_norm": 0.7206849455833435,
      "learning_rate": 0.0004147246452284859,
      "loss": 0.1448,
      "step": 2170
    },
    {
      "epoch": 1.2770943175161102,
      "grad_norm": 1.402601718902588,
      "learning_rate": 0.0004140736883218331,
      "loss": 0.1841,
      "step": 2180
    },
    {
      "epoch": 1.2829525483304043,
      "grad_norm": 0.5288113355636597,
      "learning_rate": 0.0004134227314151803,
      "loss": 0.0906,
      "step": 2190
    },
    {
      "epoch": 1.2888107791446983,
      "grad_norm": 1.5257741212844849,
      "learning_rate": 0.00041277177450852755,
      "loss": 0.1277,
      "step": 2200
    },
    {
      "epoch": 1.2946690099589924,
      "grad_norm": 0.6375269889831543,
      "learning_rate": 0.00041212081760187474,
      "loss": 0.1615,
      "step": 2210
    },
    {
      "epoch": 1.3005272407732864,
      "grad_norm": 1.8483799695968628,
      "learning_rate": 0.000411469860695222,
      "loss": 0.1615,
      "step": 2220
    },
    {
      "epoch": 1.3063854715875807,
      "grad_norm": 0.9884814620018005,
      "learning_rate": 0.00041081890378856924,
      "loss": 0.1591,
      "step": 2230
    },
    {
      "epoch": 1.3122437024018747,
      "grad_norm": 1.186396598815918,
      "learning_rate": 0.0004101679468819164,
      "loss": 0.1757,
      "step": 2240
    },
    {
      "epoch": 1.3181019332161688,
      "grad_norm": 0.8369488716125488,
      "learning_rate": 0.00040951698997526367,
      "loss": 0.1475,
      "step": 2250
    },
    {
      "epoch": 1.3239601640304628,
      "grad_norm": 1.4090021848678589,
      "learning_rate": 0.00040886603306861086,
      "loss": 0.1246,
      "step": 2260
    },
    {
      "epoch": 1.3298183948447568,
      "grad_norm": 2.212960958480835,
      "learning_rate": 0.00040821507616195805,
      "loss": 0.2006,
      "step": 2270
    },
    {
      "epoch": 1.335676625659051,
      "grad_norm": 1.3161710500717163,
      "learning_rate": 0.0004075641192553053,
      "loss": 0.1085,
      "step": 2280
    },
    {
      "epoch": 1.341534856473345,
      "grad_norm": 2.699848175048828,
      "learning_rate": 0.00040691316234865255,
      "loss": 0.1698,
      "step": 2290
    },
    {
      "epoch": 1.3473930872876392,
      "grad_norm": 0.8176013827323914,
      "learning_rate": 0.0004062622054419998,
      "loss": 0.0913,
      "step": 2300
    },
    {
      "epoch": 1.3532513181019332,
      "grad_norm": 1.959524393081665,
      "learning_rate": 0.000405611248535347,
      "loss": 0.1192,
      "step": 2310
    },
    {
      "epoch": 1.3591095489162273,
      "grad_norm": 1.9907140731811523,
      "learning_rate": 0.0004049602916286942,
      "loss": 0.1565,
      "step": 2320
    },
    {
      "epoch": 1.3649677797305213,
      "grad_norm": 0.7778204679489136,
      "learning_rate": 0.0004043093347220414,
      "loss": 0.1835,
      "step": 2330
    },
    {
      "epoch": 1.3708260105448153,
      "grad_norm": 0.8291440606117249,
      "learning_rate": 0.0004036583778153886,
      "loss": 0.1627,
      "step": 2340
    },
    {
      "epoch": 1.3766842413591096,
      "grad_norm": 0.6740533709526062,
      "learning_rate": 0.00040300742090873586,
      "loss": 0.1645,
      "step": 2350
    },
    {
      "epoch": 1.3825424721734036,
      "grad_norm": 1.1750842332839966,
      "learning_rate": 0.0004023564640020831,
      "loss": 0.1404,
      "step": 2360
    },
    {
      "epoch": 1.3884007029876977,
      "grad_norm": 0.880556046962738,
      "learning_rate": 0.0004017055070954303,
      "loss": 0.1242,
      "step": 2370
    },
    {
      "epoch": 1.3942589338019917,
      "grad_norm": 0.932043731212616,
      "learning_rate": 0.0004010545501887775,
      "loss": 0.1217,
      "step": 2380
    },
    {
      "epoch": 1.4001171646162858,
      "grad_norm": 0.5808772444725037,
      "learning_rate": 0.00040040359328212473,
      "loss": 0.2381,
      "step": 2390
    },
    {
      "epoch": 1.40597539543058,
      "grad_norm": 0.8846039175987244,
      "learning_rate": 0.0003997526363754719,
      "loss": 0.1702,
      "step": 2400
    },
    {
      "epoch": 1.411833626244874,
      "grad_norm": 2.054563522338867,
      "learning_rate": 0.00039910167946881917,
      "loss": 0.1562,
      "step": 2410
    },
    {
      "epoch": 1.4176918570591681,
      "grad_norm": 1.1283624172210693,
      "learning_rate": 0.0003984507225621664,
      "loss": 0.2018,
      "step": 2420
    },
    {
      "epoch": 1.4235500878734622,
      "grad_norm": 0.28652748465538025,
      "learning_rate": 0.0003977997656555136,
      "loss": 0.1612,
      "step": 2430
    },
    {
      "epoch": 1.4294083186877562,
      "grad_norm": 2.846034288406372,
      "learning_rate": 0.00039714880874886085,
      "loss": 0.1342,
      "step": 2440
    },
    {
      "epoch": 1.4352665495020505,
      "grad_norm": 2.90832257270813,
      "learning_rate": 0.00039649785184220804,
      "loss": 0.1695,
      "step": 2450
    },
    {
      "epoch": 1.4411247803163445,
      "grad_norm": 0.8565723299980164,
      "learning_rate": 0.00039584689493555523,
      "loss": 0.1446,
      "step": 2460
    },
    {
      "epoch": 1.4469830111306385,
      "grad_norm": 0.6863169074058533,
      "learning_rate": 0.0003951959380289025,
      "loss": 0.1666,
      "step": 2470
    },
    {
      "epoch": 1.4528412419449326,
      "grad_norm": 1.357521414756775,
      "learning_rate": 0.0003945449811222497,
      "loss": 0.1354,
      "step": 2480
    },
    {
      "epoch": 1.4586994727592266,
      "grad_norm": 0.3757944703102112,
      "learning_rate": 0.00039389402421559697,
      "loss": 0.1386,
      "step": 2490
    },
    {
      "epoch": 1.4645577035735209,
      "grad_norm": 0.5341873168945312,
      "learning_rate": 0.00039324306730894416,
      "loss": 0.1479,
      "step": 2500
    },
    {
      "epoch": 1.470415934387815,
      "grad_norm": 0.9452828168869019,
      "learning_rate": 0.00039259211040229135,
      "loss": 0.0944,
      "step": 2510
    },
    {
      "epoch": 1.476274165202109,
      "grad_norm": 0.904435932636261,
      "learning_rate": 0.0003919411534956386,
      "loss": 0.0872,
      "step": 2520
    },
    {
      "epoch": 1.482132396016403,
      "grad_norm": 1.0896387100219727,
      "learning_rate": 0.0003912901965889858,
      "loss": 0.1463,
      "step": 2530
    },
    {
      "epoch": 1.487990626830697,
      "grad_norm": 0.9536682367324829,
      "learning_rate": 0.00039063923968233303,
      "loss": 0.1748,
      "step": 2540
    },
    {
      "epoch": 1.4938488576449913,
      "grad_norm": 0.42995181679725647,
      "learning_rate": 0.0003899882827756803,
      "loss": 0.1005,
      "step": 2550
    },
    {
      "epoch": 1.4997070884592854,
      "grad_norm": 1.3272085189819336,
      "learning_rate": 0.00038933732586902747,
      "loss": 0.1755,
      "step": 2560
    },
    {
      "epoch": 1.5055653192735794,
      "grad_norm": 1.5230518579483032,
      "learning_rate": 0.0003886863689623747,
      "loss": 0.137,
      "step": 2570
    },
    {
      "epoch": 1.5114235500878734,
      "grad_norm": 1.9343452453613281,
      "learning_rate": 0.0003880354120557219,
      "loss": 0.0864,
      "step": 2580
    },
    {
      "epoch": 1.5172817809021675,
      "grad_norm": 0.27782219648361206,
      "learning_rate": 0.0003873844551490691,
      "loss": 0.232,
      "step": 2590
    },
    {
      "epoch": 1.5231400117164617,
      "grad_norm": 0.78903728723526,
      "learning_rate": 0.0003867334982424164,
      "loss": 0.1586,
      "step": 2600
    },
    {
      "epoch": 1.5289982425307556,
      "grad_norm": 1.1951757669448853,
      "learning_rate": 0.0003860825413357636,
      "loss": 0.2101,
      "step": 2610
    },
    {
      "epoch": 1.5348564733450498,
      "grad_norm": 0.6688250303268433,
      "learning_rate": 0.0003854315844291108,
      "loss": 0.1457,
      "step": 2620
    },
    {
      "epoch": 1.5407147041593439,
      "grad_norm": 0.8118303418159485,
      "learning_rate": 0.000384780627522458,
      "loss": 0.1941,
      "step": 2630
    },
    {
      "epoch": 1.546572934973638,
      "grad_norm": 0.43793728947639465,
      "learning_rate": 0.0003841296706158052,
      "loss": 0.139,
      "step": 2640
    },
    {
      "epoch": 1.5524311657879322,
      "grad_norm": 0.8637521862983704,
      "learning_rate": 0.00038347871370915246,
      "loss": 0.1452,
      "step": 2650
    },
    {
      "epoch": 1.558289396602226,
      "grad_norm": 1.656689167022705,
      "learning_rate": 0.0003828277568024997,
      "loss": 0.1937,
      "step": 2660
    },
    {
      "epoch": 1.5641476274165202,
      "grad_norm": 1.5395698547363281,
      "learning_rate": 0.0003821767998958469,
      "loss": 0.1465,
      "step": 2670
    },
    {
      "epoch": 1.5700058582308143,
      "grad_norm": 0.8242401480674744,
      "learning_rate": 0.00038152584298919415,
      "loss": 0.1497,
      "step": 2680
    },
    {
      "epoch": 1.5758640890451083,
      "grad_norm": 1.000875473022461,
      "learning_rate": 0.00038087488608254134,
      "loss": 0.1878,
      "step": 2690
    },
    {
      "epoch": 1.5817223198594026,
      "grad_norm": 0.8596873879432678,
      "learning_rate": 0.00038022392917588853,
      "loss": 0.1509,
      "step": 2700
    },
    {
      "epoch": 1.5875805506736964,
      "grad_norm": 2.0601985454559326,
      "learning_rate": 0.0003795729722692358,
      "loss": 0.1647,
      "step": 2710
    },
    {
      "epoch": 1.5934387814879907,
      "grad_norm": 1.012863278388977,
      "learning_rate": 0.000378922015362583,
      "loss": 0.1352,
      "step": 2720
    },
    {
      "epoch": 1.5992970123022847,
      "grad_norm": 1.4908140897750854,
      "learning_rate": 0.00037827105845593027,
      "loss": 0.1741,
      "step": 2730
    },
    {
      "epoch": 1.6051552431165788,
      "grad_norm": 0.9020434021949768,
      "learning_rate": 0.00037762010154927746,
      "loss": 0.1376,
      "step": 2740
    },
    {
      "epoch": 1.611013473930873,
      "grad_norm": 0.7156019806861877,
      "learning_rate": 0.00037696914464262465,
      "loss": 0.1685,
      "step": 2750
    },
    {
      "epoch": 1.6168717047451668,
      "grad_norm": 1.503300428390503,
      "learning_rate": 0.0003763181877359719,
      "loss": 0.1478,
      "step": 2760
    },
    {
      "epoch": 1.622729935559461,
      "grad_norm": 0.546312153339386,
      "learning_rate": 0.0003756672308293191,
      "loss": 0.1061,
      "step": 2770
    },
    {
      "epoch": 1.6285881663737551,
      "grad_norm": 0.4983227550983429,
      "learning_rate": 0.00037501627392266633,
      "loss": 0.1315,
      "step": 2780
    },
    {
      "epoch": 1.6344463971880492,
      "grad_norm": 0.6810446381568909,
      "learning_rate": 0.0003743653170160136,
      "loss": 0.1199,
      "step": 2790
    },
    {
      "epoch": 1.6403046280023434,
      "grad_norm": 1.2596421241760254,
      "learning_rate": 0.00037371436010936077,
      "loss": 0.1201,
      "step": 2800
    },
    {
      "epoch": 1.6461628588166373,
      "grad_norm": 0.9648050665855408,
      "learning_rate": 0.00037306340320270796,
      "loss": 0.1358,
      "step": 2810
    },
    {
      "epoch": 1.6520210896309315,
      "grad_norm": 1.0759117603302002,
      "learning_rate": 0.0003724124462960552,
      "loss": 0.1664,
      "step": 2820
    },
    {
      "epoch": 1.6578793204452256,
      "grad_norm": 1.0369553565979004,
      "learning_rate": 0.0003717614893894024,
      "loss": 0.1737,
      "step": 2830
    },
    {
      "epoch": 1.6637375512595196,
      "grad_norm": 1.0915820598602295,
      "learning_rate": 0.0003711105324827497,
      "loss": 0.1355,
      "step": 2840
    },
    {
      "epoch": 1.6695957820738139,
      "grad_norm": 1.3576059341430664,
      "learning_rate": 0.0003704595755760969,
      "loss": 0.1674,
      "step": 2850
    },
    {
      "epoch": 1.6754540128881077,
      "grad_norm": 0.8195040822029114,
      "learning_rate": 0.0003698086186694441,
      "loss": 0.1284,
      "step": 2860
    },
    {
      "epoch": 1.681312243702402,
      "grad_norm": 1.8879438638687134,
      "learning_rate": 0.0003691576617627913,
      "loss": 0.1336,
      "step": 2870
    },
    {
      "epoch": 1.687170474516696,
      "grad_norm": 0.7765256762504578,
      "learning_rate": 0.0003685067048561385,
      "loss": 0.1347,
      "step": 2880
    },
    {
      "epoch": 1.69302870533099,
      "grad_norm": 1.017264723777771,
      "learning_rate": 0.0003678557479494857,
      "loss": 0.1548,
      "step": 2890
    },
    {
      "epoch": 1.698886936145284,
      "grad_norm": 0.42928430438041687,
      "learning_rate": 0.000367204791042833,
      "loss": 0.1526,
      "step": 2900
    },
    {
      "epoch": 1.7047451669595781,
      "grad_norm": 0.75179123878479,
      "learning_rate": 0.0003665538341361802,
      "loss": 0.1415,
      "step": 2910
    },
    {
      "epoch": 1.7106033977738724,
      "grad_norm": 1.2041255235671997,
      "learning_rate": 0.00036590287722952744,
      "loss": 0.1065,
      "step": 2920
    },
    {
      "epoch": 1.7164616285881664,
      "grad_norm": 1.6273362636566162,
      "learning_rate": 0.00036525192032287463,
      "loss": 0.1683,
      "step": 2930
    },
    {
      "epoch": 1.7223198594024605,
      "grad_norm": 1.5621204376220703,
      "learning_rate": 0.0003646009634162218,
      "loss": 0.1626,
      "step": 2940
    },
    {
      "epoch": 1.7281780902167545,
      "grad_norm": 1.275732398033142,
      "learning_rate": 0.00036395000650956907,
      "loss": 0.1235,
      "step": 2950
    },
    {
      "epoch": 1.7340363210310485,
      "grad_norm": 1.7712377309799194,
      "learning_rate": 0.0003632990496029163,
      "loss": 0.1851,
      "step": 2960
    },
    {
      "epoch": 1.7398945518453428,
      "grad_norm": 0.7929120659828186,
      "learning_rate": 0.0003626480926962635,
      "loss": 0.1422,
      "step": 2970
    },
    {
      "epoch": 1.7457527826596366,
      "grad_norm": 0.8228936791419983,
      "learning_rate": 0.00036199713578961075,
      "loss": 0.1494,
      "step": 2980
    },
    {
      "epoch": 1.751611013473931,
      "grad_norm": 1.4195990562438965,
      "learning_rate": 0.00036134617888295795,
      "loss": 0.1466,
      "step": 2990
    },
    {
      "epoch": 1.757469244288225,
      "grad_norm": 0.2758101522922516,
      "learning_rate": 0.0003606952219763052,
      "loss": 0.1443,
      "step": 3000
    },
    {
      "epoch": 1.763327475102519,
      "grad_norm": 0.9897528886795044,
      "learning_rate": 0.0003600442650696524,
      "loss": 0.1148,
      "step": 3010
    },
    {
      "epoch": 1.7691857059168132,
      "grad_norm": 0.1884101778268814,
      "learning_rate": 0.00035939330816299963,
      "loss": 0.0635,
      "step": 3020
    },
    {
      "epoch": 1.775043936731107,
      "grad_norm": 1.4347259998321533,
      "learning_rate": 0.0003587423512563469,
      "loss": 0.1919,
      "step": 3030
    },
    {
      "epoch": 1.7809021675454013,
      "grad_norm": 0.8466237783432007,
      "learning_rate": 0.00035809139434969406,
      "loss": 0.1725,
      "step": 3040
    },
    {
      "epoch": 1.7867603983596954,
      "grad_norm": 1.203354001045227,
      "learning_rate": 0.00035744043744304126,
      "loss": 0.1351,
      "step": 3050
    },
    {
      "epoch": 1.7926186291739894,
      "grad_norm": 0.47008630633354187,
      "learning_rate": 0.0003567894805363885,
      "loss": 0.1546,
      "step": 3060
    },
    {
      "epoch": 1.7984768599882837,
      "grad_norm": 1.0304838418960571,
      "learning_rate": 0.0003561385236297357,
      "loss": 0.1989,
      "step": 3070
    },
    {
      "epoch": 1.8043350908025775,
      "grad_norm": 0.6836739182472229,
      "learning_rate": 0.000355487566723083,
      "loss": 0.1482,
      "step": 3080
    },
    {
      "epoch": 1.8101933216168717,
      "grad_norm": 1.5532166957855225,
      "learning_rate": 0.0003548366098164302,
      "loss": 0.2138,
      "step": 3090
    },
    {
      "epoch": 1.8160515524311658,
      "grad_norm": 0.4863007962703705,
      "learning_rate": 0.0003541856529097774,
      "loss": 0.1219,
      "step": 3100
    },
    {
      "epoch": 1.8219097832454598,
      "grad_norm": 0.6631212830543518,
      "learning_rate": 0.0003535346960031246,
      "loss": 0.1539,
      "step": 3110
    },
    {
      "epoch": 1.827768014059754,
      "grad_norm": 0.7438946962356567,
      "learning_rate": 0.0003528837390964718,
      "loss": 0.1615,
      "step": 3120
    },
    {
      "epoch": 1.833626244874048,
      "grad_norm": 1.0698827505111694,
      "learning_rate": 0.000352232782189819,
      "loss": 0.2044,
      "step": 3130
    },
    {
      "epoch": 1.8394844756883422,
      "grad_norm": 0.41544845700263977,
      "learning_rate": 0.00035158182528316625,
      "loss": 0.1379,
      "step": 3140
    },
    {
      "epoch": 1.8453427065026362,
      "grad_norm": 1.023944616317749,
      "learning_rate": 0.0003509308683765135,
      "loss": 0.1466,
      "step": 3150
    },
    {
      "epoch": 1.8512009373169303,
      "grad_norm": 0.8214076161384583,
      "learning_rate": 0.0003502799114698607,
      "loss": 0.1181,
      "step": 3160
    },
    {
      "epoch": 1.8570591681312245,
      "grad_norm": 1.027405858039856,
      "learning_rate": 0.00034962895456320793,
      "loss": 0.1083,
      "step": 3170
    },
    {
      "epoch": 1.8629173989455183,
      "grad_norm": 3.0552237033843994,
      "learning_rate": 0.0003489779976565551,
      "loss": 0.1734,
      "step": 3180
    },
    {
      "epoch": 1.8687756297598126,
      "grad_norm": 1.0937753915786743,
      "learning_rate": 0.00034832704074990237,
      "loss": 0.1142,
      "step": 3190
    },
    {
      "epoch": 1.8746338605741066,
      "grad_norm": 1.0819835662841797,
      "learning_rate": 0.00034767608384324956,
      "loss": 0.1135,
      "step": 3200
    },
    {
      "epoch": 1.8804920913884007,
      "grad_norm": 1.1981408596038818,
      "learning_rate": 0.0003470251269365968,
      "loss": 0.1087,
      "step": 3210
    },
    {
      "epoch": 1.886350322202695,
      "grad_norm": 1.8808680772781372,
      "learning_rate": 0.00034637417002994405,
      "loss": 0.1477,
      "step": 3220
    },
    {
      "epoch": 1.8922085530169888,
      "grad_norm": 1.3297356367111206,
      "learning_rate": 0.00034572321312329124,
      "loss": 0.1422,
      "step": 3230
    },
    {
      "epoch": 1.898066783831283,
      "grad_norm": 1.0454509258270264,
      "learning_rate": 0.00034507225621663843,
      "loss": 0.1246,
      "step": 3240
    },
    {
      "epoch": 1.903925014645577,
      "grad_norm": 0.9511450529098511,
      "learning_rate": 0.0003444212993099857,
      "loss": 0.1102,
      "step": 3250
    },
    {
      "epoch": 1.909783245459871,
      "grad_norm": 1.416601300239563,
      "learning_rate": 0.00034377034240333287,
      "loss": 0.1576,
      "step": 3260
    },
    {
      "epoch": 1.9156414762741654,
      "grad_norm": 0.8354918956756592,
      "learning_rate": 0.00034311938549668017,
      "loss": 0.1596,
      "step": 3270
    },
    {
      "epoch": 1.9214997070884592,
      "grad_norm": 1.1035876274108887,
      "learning_rate": 0.00034246842859002736,
      "loss": 0.1418,
      "step": 3280
    },
    {
      "epoch": 1.9273579379027534,
      "grad_norm": 1.3299360275268555,
      "learning_rate": 0.00034181747168337455,
      "loss": 0.1353,
      "step": 3290
    },
    {
      "epoch": 1.9332161687170475,
      "grad_norm": 1.3646624088287354,
      "learning_rate": 0.0003411665147767218,
      "loss": 0.1488,
      "step": 3300
    },
    {
      "epoch": 1.9390743995313415,
      "grad_norm": 1.2904136180877686,
      "learning_rate": 0.000340515557870069,
      "loss": 0.14,
      "step": 3310
    },
    {
      "epoch": 1.9449326303456356,
      "grad_norm": 0.9034157991409302,
      "learning_rate": 0.0003398646009634162,
      "loss": 0.1954,
      "step": 3320
    },
    {
      "epoch": 1.9507908611599296,
      "grad_norm": 0.7134235501289368,
      "learning_rate": 0.0003392136440567635,
      "loss": 0.1842,
      "step": 3330
    },
    {
      "epoch": 1.9566490919742239,
      "grad_norm": 1.0327099561691284,
      "learning_rate": 0.00033856268715011067,
      "loss": 0.1927,
      "step": 3340
    },
    {
      "epoch": 1.962507322788518,
      "grad_norm": 0.6372695565223694,
      "learning_rate": 0.0003379117302434579,
      "loss": 0.1247,
      "step": 3350
    },
    {
      "epoch": 1.968365553602812,
      "grad_norm": 0.30220475792884827,
      "learning_rate": 0.0003372607733368051,
      "loss": 0.142,
      "step": 3360
    },
    {
      "epoch": 1.974223784417106,
      "grad_norm": 0.4559396207332611,
      "learning_rate": 0.0003366098164301523,
      "loss": 0.1475,
      "step": 3370
    },
    {
      "epoch": 1.9800820152314,
      "grad_norm": 0.6560627222061157,
      "learning_rate": 0.00033595885952349955,
      "loss": 0.1602,
      "step": 3380
    },
    {
      "epoch": 1.9859402460456943,
      "grad_norm": 1.2091546058654785,
      "learning_rate": 0.0003353079026168468,
      "loss": 0.1333,
      "step": 3390
    },
    {
      "epoch": 1.9917984768599881,
      "grad_norm": 0.6147709488868713,
      "learning_rate": 0.000334656945710194,
      "loss": 0.1473,
      "step": 3400
    },
    {
      "epoch": 1.9976567076742824,
      "grad_norm": 0.751679539680481,
      "learning_rate": 0.00033400598880354123,
      "loss": 0.1326,
      "step": 3410
    },
    {
      "epoch": 2.0,
      "eval_f1": 0.9051568698160837,
      "eval_loss": 0.2498554289340973,
      "eval_runtime": 14.577,
      "eval_samples_per_second": 205.804,
      "eval_steps_per_second": 6.449,
      "step": 3414
    },
    {
      "epoch": 2.0035149384885766,
      "grad_norm": 0.8273231387138367,
      "learning_rate": 0.0003333550318968884,
      "loss": 0.1969,
      "step": 3420
    },
    {
      "epoch": 2.0093731693028705,
      "grad_norm": 0.7016593217849731,
      "learning_rate": 0.00033270407499023566,
      "loss": 0.1126,
      "step": 3430
    },
    {
      "epoch": 2.0152314001171647,
      "grad_norm": 0.6760423183441162,
      "learning_rate": 0.00033205311808358286,
      "loss": 0.1554,
      "step": 3440
    },
    {
      "epoch": 2.0210896309314585,
      "grad_norm": 0.7149805426597595,
      "learning_rate": 0.0003314021611769301,
      "loss": 0.1065,
      "step": 3450
    },
    {
      "epoch": 2.026947861745753,
      "grad_norm": 1.2349637746810913,
      "learning_rate": 0.00033075120427027735,
      "loss": 0.1916,
      "step": 3460
    },
    {
      "epoch": 2.032806092560047,
      "grad_norm": 0.8604615330696106,
      "learning_rate": 0.00033010024736362454,
      "loss": 0.1068,
      "step": 3470
    },
    {
      "epoch": 2.038664323374341,
      "grad_norm": 0.24382354319095612,
      "learning_rate": 0.00032944929045697173,
      "loss": 0.1375,
      "step": 3480
    },
    {
      "epoch": 2.044522554188635,
      "grad_norm": 0.656825602054596,
      "learning_rate": 0.000328798333550319,
      "loss": 0.1482,
      "step": 3490
    },
    {
      "epoch": 2.050380785002929,
      "grad_norm": 0.9618995785713196,
      "learning_rate": 0.00032814737664366617,
      "loss": 0.1149,
      "step": 3500
    },
    {
      "epoch": 2.0562390158172232,
      "grad_norm": 1.9244462251663208,
      "learning_rate": 0.00032749641973701347,
      "loss": 0.1389,
      "step": 3510
    },
    {
      "epoch": 2.0620972466315175,
      "grad_norm": 1.6664360761642456,
      "learning_rate": 0.00032684546283036066,
      "loss": 0.1432,
      "step": 3520
    },
    {
      "epoch": 2.0679554774458113,
      "grad_norm": 0.4561314284801483,
      "learning_rate": 0.00032619450592370785,
      "loss": 0.1132,
      "step": 3530
    },
    {
      "epoch": 2.0738137082601056,
      "grad_norm": 0.33170029520988464,
      "learning_rate": 0.0003255435490170551,
      "loss": 0.1682,
      "step": 3540
    },
    {
      "epoch": 2.0796719390743994,
      "grad_norm": 0.4367826282978058,
      "learning_rate": 0.0003248925921104023,
      "loss": 0.0934,
      "step": 3550
    },
    {
      "epoch": 2.0855301698886937,
      "grad_norm": 0.7262292504310608,
      "learning_rate": 0.0003242416352037495,
      "loss": 0.1238,
      "step": 3560
    },
    {
      "epoch": 2.0913884007029875,
      "grad_norm": 1.0613982677459717,
      "learning_rate": 0.0003235906782970968,
      "loss": 0.0882,
      "step": 3570
    },
    {
      "epoch": 2.0972466315172817,
      "grad_norm": 2.260653018951416,
      "learning_rate": 0.00032293972139044397,
      "loss": 0.2016,
      "step": 3580
    },
    {
      "epoch": 2.103104862331576,
      "grad_norm": 1.0498729944229126,
      "learning_rate": 0.00032228876448379116,
      "loss": 0.1245,
      "step": 3590
    },
    {
      "epoch": 2.10896309314587,
      "grad_norm": 1.473840594291687,
      "learning_rate": 0.0003216378075771384,
      "loss": 0.1602,
      "step": 3600
    },
    {
      "epoch": 2.114821323960164,
      "grad_norm": 0.40939509868621826,
      "learning_rate": 0.0003209868506704856,
      "loss": 0.1458,
      "step": 3610
    },
    {
      "epoch": 2.120679554774458,
      "grad_norm": 0.44588029384613037,
      "learning_rate": 0.00032033589376383284,
      "loss": 0.1174,
      "step": 3620
    },
    {
      "epoch": 2.126537785588752,
      "grad_norm": 1.196237325668335,
      "learning_rate": 0.0003196849368571801,
      "loss": 0.1936,
      "step": 3630
    },
    {
      "epoch": 2.1323960164030464,
      "grad_norm": 0.3714032769203186,
      "learning_rate": 0.0003190339799505273,
      "loss": 0.1558,
      "step": 3640
    },
    {
      "epoch": 2.1382542472173403,
      "grad_norm": 0.8900133967399597,
      "learning_rate": 0.0003183830230438745,
      "loss": 0.1434,
      "step": 3650
    },
    {
      "epoch": 2.1441124780316345,
      "grad_norm": 0.21238140761852264,
      "learning_rate": 0.0003177320661372217,
      "loss": 0.1501,
      "step": 3660
    },
    {
      "epoch": 2.1499707088459283,
      "grad_norm": 0.45354288816452026,
      "learning_rate": 0.0003170811092305689,
      "loss": 0.1279,
      "step": 3670
    },
    {
      "epoch": 2.1558289396602226,
      "grad_norm": 1.1276099681854248,
      "learning_rate": 0.00031643015232391615,
      "loss": 0.1237,
      "step": 3680
    },
    {
      "epoch": 2.161687170474517,
      "grad_norm": 1.6248077154159546,
      "learning_rate": 0.0003157791954172634,
      "loss": 0.0997,
      "step": 3690
    },
    {
      "epoch": 2.1675454012888107,
      "grad_norm": 0.5289869904518127,
      "learning_rate": 0.00031512823851061064,
      "loss": 0.1384,
      "step": 3700
    },
    {
      "epoch": 2.173403632103105,
      "grad_norm": 0.781470000743866,
      "learning_rate": 0.00031447728160395784,
      "loss": 0.107,
      "step": 3710
    },
    {
      "epoch": 2.1792618629173988,
      "grad_norm": 0.49066823720932007,
      "learning_rate": 0.000313826324697305,
      "loss": 0.1406,
      "step": 3720
    },
    {
      "epoch": 2.185120093731693,
      "grad_norm": 0.9173513054847717,
      "learning_rate": 0.00031317536779065227,
      "loss": 0.1404,
      "step": 3730
    },
    {
      "epoch": 2.1909783245459873,
      "grad_norm": 0.29051652550697327,
      "learning_rate": 0.00031252441088399946,
      "loss": 0.1694,
      "step": 3740
    },
    {
      "epoch": 2.196836555360281,
      "grad_norm": 0.5923140645027161,
      "learning_rate": 0.0003118734539773467,
      "loss": 0.1195,
      "step": 3750
    },
    {
      "epoch": 2.2026947861745754,
      "grad_norm": 0.6858001947402954,
      "learning_rate": 0.00031122249707069395,
      "loss": 0.18,
      "step": 3760
    },
    {
      "epoch": 2.208553016988869,
      "grad_norm": 0.9082098603248596,
      "learning_rate": 0.00031057154016404115,
      "loss": 0.1295,
      "step": 3770
    },
    {
      "epoch": 2.2144112478031635,
      "grad_norm": 0.8940849304199219,
      "learning_rate": 0.0003099205832573884,
      "loss": 0.163,
      "step": 3780
    },
    {
      "epoch": 2.2202694786174577,
      "grad_norm": 0.36999017000198364,
      "learning_rate": 0.0003092696263507356,
      "loss": 0.1384,
      "step": 3790
    },
    {
      "epoch": 2.2261277094317515,
      "grad_norm": 1.1672558784484863,
      "learning_rate": 0.0003086186694440828,
      "loss": 0.138,
      "step": 3800
    },
    {
      "epoch": 2.231985940246046,
      "grad_norm": 0.8477211594581604,
      "learning_rate": 0.0003079677125374301,
      "loss": 0.1636,
      "step": 3810
    },
    {
      "epoch": 2.2378441710603396,
      "grad_norm": 1.2284730672836304,
      "learning_rate": 0.00030731675563077727,
      "loss": 0.1117,
      "step": 3820
    },
    {
      "epoch": 2.243702401874634,
      "grad_norm": 0.745220959186554,
      "learning_rate": 0.00030666579872412446,
      "loss": 0.1254,
      "step": 3830
    },
    {
      "epoch": 2.249560632688928,
      "grad_norm": 1.0539661645889282,
      "learning_rate": 0.0003060148418174717,
      "loss": 0.1172,
      "step": 3840
    },
    {
      "epoch": 2.255418863503222,
      "grad_norm": 0.3214583992958069,
      "learning_rate": 0.0003053638849108189,
      "loss": 0.1075,
      "step": 3850
    },
    {
      "epoch": 2.2612770943175162,
      "grad_norm": 0.11182527989149094,
      "learning_rate": 0.00030471292800416614,
      "loss": 0.1132,
      "step": 3860
    },
    {
      "epoch": 2.26713532513181,
      "grad_norm": 1.5970757007598877,
      "learning_rate": 0.00030406197109751333,
      "loss": 0.1469,
      "step": 3870
    },
    {
      "epoch": 2.2729935559461043,
      "grad_norm": 1.57491135597229,
      "learning_rate": 0.0003034110141908606,
      "loss": 0.1708,
      "step": 3880
    },
    {
      "epoch": 2.2788517867603986,
      "grad_norm": 0.7461884617805481,
      "learning_rate": 0.0003027600572842078,
      "loss": 0.1209,
      "step": 3890
    },
    {
      "epoch": 2.2847100175746924,
      "grad_norm": 1.06315016746521,
      "learning_rate": 0.000302109100377555,
      "loss": 0.1731,
      "step": 3900
    },
    {
      "epoch": 2.2905682483889866,
      "grad_norm": 1.3551595211029053,
      "learning_rate": 0.0003014581434709022,
      "loss": 0.1369,
      "step": 3910
    },
    {
      "epoch": 2.2964264792032805,
      "grad_norm": 0.7530913352966309,
      "learning_rate": 0.00030080718656424945,
      "loss": 0.1821,
      "step": 3920
    },
    {
      "epoch": 2.3022847100175747,
      "grad_norm": 0.5707684755325317,
      "learning_rate": 0.00030015622965759664,
      "loss": 0.1104,
      "step": 3930
    },
    {
      "epoch": 2.3081429408318686,
      "grad_norm": 0.8748314380645752,
      "learning_rate": 0.0002995052727509439,
      "loss": 0.0961,
      "step": 3940
    },
    {
      "epoch": 2.314001171646163,
      "grad_norm": 1.2551450729370117,
      "learning_rate": 0.00029885431584429113,
      "loss": 0.1167,
      "step": 3950
    },
    {
      "epoch": 2.319859402460457,
      "grad_norm": 0.5332285761833191,
      "learning_rate": 0.0002982033589376383,
      "loss": 0.0624,
      "step": 3960
    },
    {
      "epoch": 2.325717633274751,
      "grad_norm": 1.4492905139923096,
      "learning_rate": 0.00029755240203098557,
      "loss": 0.1021,
      "step": 3970
    },
    {
      "epoch": 2.331575864089045,
      "grad_norm": 0.294634610414505,
      "learning_rate": 0.00029690144512433276,
      "loss": 0.118,
      "step": 3980
    },
    {
      "epoch": 2.3374340949033394,
      "grad_norm": 0.9232218861579895,
      "learning_rate": 0.00029625048821767995,
      "loss": 0.1014,
      "step": 3990
    },
    {
      "epoch": 2.3432923257176332,
      "grad_norm": 0.928164005279541,
      "learning_rate": 0.00029559953131102725,
      "loss": 0.0991,
      "step": 4000
    },
    {
      "epoch": 2.3491505565319275,
      "grad_norm": 0.6724839806556702,
      "learning_rate": 0.00029494857440437444,
      "loss": 0.1867,
      "step": 4010
    },
    {
      "epoch": 2.3550087873462213,
      "grad_norm": 1.1483349800109863,
      "learning_rate": 0.00029429761749772163,
      "loss": 0.1504,
      "step": 4020
    },
    {
      "epoch": 2.3608670181605156,
      "grad_norm": 0.6115332841873169,
      "learning_rate": 0.0002936466605910689,
      "loss": 0.0969,
      "step": 4030
    },
    {
      "epoch": 2.3667252489748094,
      "grad_norm": 0.43446749448776245,
      "learning_rate": 0.00029299570368441607,
      "loss": 0.1511,
      "step": 4040
    },
    {
      "epoch": 2.3725834797891037,
      "grad_norm": 1.4291220903396606,
      "learning_rate": 0.0002923447467777633,
      "loss": 0.136,
      "step": 4050
    },
    {
      "epoch": 2.378441710603398,
      "grad_norm": 0.640063464641571,
      "learning_rate": 0.00029169378987111056,
      "loss": 0.104,
      "step": 4060
    },
    {
      "epoch": 2.3842999414176917,
      "grad_norm": 0.8535376191139221,
      "learning_rate": 0.00029104283296445775,
      "loss": 0.1197,
      "step": 4070
    },
    {
      "epoch": 2.390158172231986,
      "grad_norm": 1.2727094888687134,
      "learning_rate": 0.000290391876057805,
      "loss": 0.1475,
      "step": 4080
    },
    {
      "epoch": 2.39601640304628,
      "grad_norm": 1.8613280057907104,
      "learning_rate": 0.0002897409191511522,
      "loss": 0.121,
      "step": 4090
    },
    {
      "epoch": 2.401874633860574,
      "grad_norm": 1.230644702911377,
      "learning_rate": 0.0002890899622444994,
      "loss": 0.0983,
      "step": 4100
    },
    {
      "epoch": 2.4077328646748684,
      "grad_norm": 0.5424458980560303,
      "learning_rate": 0.00028843900533784663,
      "loss": 0.1382,
      "step": 4110
    },
    {
      "epoch": 2.413591095489162,
      "grad_norm": 0.5892564058303833,
      "learning_rate": 0.0002877880484311939,
      "loss": 0.112,
      "step": 4120
    },
    {
      "epoch": 2.4194493263034564,
      "grad_norm": 0.936575710773468,
      "learning_rate": 0.0002871370915245411,
      "loss": 0.1243,
      "step": 4130
    },
    {
      "epoch": 2.4253075571177503,
      "grad_norm": 1.3329449892044067,
      "learning_rate": 0.0002864861346178883,
      "loss": 0.1001,
      "step": 4140
    },
    {
      "epoch": 2.4311657879320445,
      "grad_norm": 0.28441017866134644,
      "learning_rate": 0.0002858351777112355,
      "loss": 0.1086,
      "step": 4150
    },
    {
      "epoch": 2.437024018746339,
      "grad_norm": 0.7536140084266663,
      "learning_rate": 0.00028518422080458275,
      "loss": 0.1712,
      "step": 4160
    },
    {
      "epoch": 2.4428822495606326,
      "grad_norm": 0.4054887592792511,
      "learning_rate": 0.00028453326389792994,
      "loss": 0.091,
      "step": 4170
    },
    {
      "epoch": 2.448740480374927,
      "grad_norm": 0.4561721384525299,
      "learning_rate": 0.0002838823069912772,
      "loss": 0.1377,
      "step": 4180
    },
    {
      "epoch": 2.4545987111892207,
      "grad_norm": 0.9957545399665833,
      "learning_rate": 0.00028323135008462443,
      "loss": 0.1018,
      "step": 4190
    },
    {
      "epoch": 2.460456942003515,
      "grad_norm": 1.5235391855239868,
      "learning_rate": 0.0002825803931779716,
      "loss": 0.1161,
      "step": 4200
    },
    {
      "epoch": 2.466315172817809,
      "grad_norm": 1.5290122032165527,
      "learning_rate": 0.00028192943627131887,
      "loss": 0.1101,
      "step": 4210
    },
    {
      "epoch": 2.472173403632103,
      "grad_norm": 0.8591200113296509,
      "learning_rate": 0.00028127847936466606,
      "loss": 0.1148,
      "step": 4220
    },
    {
      "epoch": 2.4780316344463973,
      "grad_norm": 1.4811750650405884,
      "learning_rate": 0.00028062752245801325,
      "loss": 0.1753,
      "step": 4230
    },
    {
      "epoch": 2.483889865260691,
      "grad_norm": 0.9379531145095825,
      "learning_rate": 0.00027997656555136055,
      "loss": 0.1451,
      "step": 4240
    },
    {
      "epoch": 2.4897480960749854,
      "grad_norm": 1.2333440780639648,
      "learning_rate": 0.00027932560864470774,
      "loss": 0.1101,
      "step": 4250
    },
    {
      "epoch": 2.4956063268892796,
      "grad_norm": 1.016764760017395,
      "learning_rate": 0.00027867465173805493,
      "loss": 0.124,
      "step": 4260
    },
    {
      "epoch": 2.5014645577035735,
      "grad_norm": 1.6167750358581543,
      "learning_rate": 0.0002780236948314022,
      "loss": 0.1398,
      "step": 4270
    },
    {
      "epoch": 2.5073227885178677,
      "grad_norm": 0.5222885012626648,
      "learning_rate": 0.00027737273792474937,
      "loss": 0.097,
      "step": 4280
    },
    {
      "epoch": 2.5131810193321615,
      "grad_norm": 0.763202428817749,
      "learning_rate": 0.00027672178101809656,
      "loss": 0.1379,
      "step": 4290
    },
    {
      "epoch": 2.519039250146456,
      "grad_norm": 0.9483877420425415,
      "learning_rate": 0.00027607082411144386,
      "loss": 0.1466,
      "step": 4300
    },
    {
      "epoch": 2.5248974809607496,
      "grad_norm": 0.28170886635780334,
      "learning_rate": 0.00027541986720479105,
      "loss": 0.0893,
      "step": 4310
    },
    {
      "epoch": 2.530755711775044,
      "grad_norm": 1.6378175020217896,
      "learning_rate": 0.0002747689102981383,
      "loss": 0.1025,
      "step": 4320
    },
    {
      "epoch": 2.536613942589338,
      "grad_norm": 0.8733501434326172,
      "learning_rate": 0.0002741179533914855,
      "loss": 0.1538,
      "step": 4330
    },
    {
      "epoch": 2.542472173403632,
      "grad_norm": 0.4998052716255188,
      "learning_rate": 0.0002734669964848327,
      "loss": 0.12,
      "step": 4340
    },
    {
      "epoch": 2.5483304042179262,
      "grad_norm": 2.2542619705200195,
      "learning_rate": 0.0002728160395781799,
      "loss": 0.1392,
      "step": 4350
    },
    {
      "epoch": 2.5541886350322205,
      "grad_norm": 0.7784902453422546,
      "learning_rate": 0.00027216508267152717,
      "loss": 0.1211,
      "step": 4360
    },
    {
      "epoch": 2.5600468658465143,
      "grad_norm": 1.080384612083435,
      "learning_rate": 0.00027151412576487436,
      "loss": 0.1912,
      "step": 4370
    },
    {
      "epoch": 2.5659050966608086,
      "grad_norm": 0.5921308398246765,
      "learning_rate": 0.0002708631688582216,
      "loss": 0.111,
      "step": 4380
    },
    {
      "epoch": 2.5717633274751024,
      "grad_norm": 1.543337345123291,
      "learning_rate": 0.0002702122119515688,
      "loss": 0.1058,
      "step": 4390
    },
    {
      "epoch": 2.5776215582893967,
      "grad_norm": 0.8708186149597168,
      "learning_rate": 0.00026956125504491604,
      "loss": 0.1268,
      "step": 4400
    },
    {
      "epoch": 2.5834797891036905,
      "grad_norm": 1.394648790359497,
      "learning_rate": 0.00026891029813826323,
      "loss": 0.144,
      "step": 4410
    },
    {
      "epoch": 2.5893380199179847,
      "grad_norm": 0.597211480140686,
      "learning_rate": 0.0002682593412316105,
      "loss": 0.0984,
      "step": 4420
    },
    {
      "epoch": 2.595196250732279,
      "grad_norm": 0.30916476249694824,
      "learning_rate": 0.0002676083843249577,
      "loss": 0.1638,
      "step": 4430
    },
    {
      "epoch": 2.601054481546573,
      "grad_norm": 0.4281676113605499,
      "learning_rate": 0.0002669574274183049,
      "loss": 0.1351,
      "step": 4440
    },
    {
      "epoch": 2.606912712360867,
      "grad_norm": 0.8316947221755981,
      "learning_rate": 0.0002663064705116521,
      "loss": 0.1458,
      "step": 4450
    },
    {
      "epoch": 2.6127709431751613,
      "grad_norm": 0.8049978017807007,
      "learning_rate": 0.00026565551360499935,
      "loss": 0.1022,
      "step": 4460
    },
    {
      "epoch": 2.618629173989455,
      "grad_norm": 0.4352714419364929,
      "learning_rate": 0.00026500455669834655,
      "loss": 0.1185,
      "step": 4470
    },
    {
      "epoch": 2.6244874048037494,
      "grad_norm": 1.2094974517822266,
      "learning_rate": 0.00026435359979169385,
      "loss": 0.1452,
      "step": 4480
    },
    {
      "epoch": 2.6303456356180432,
      "grad_norm": 0.7460260987281799,
      "learning_rate": 0.00026370264288504104,
      "loss": 0.1086,
      "step": 4490
    },
    {
      "epoch": 2.6362038664323375,
      "grad_norm": 1.3077291250228882,
      "learning_rate": 0.00026305168597838823,
      "loss": 0.1692,
      "step": 4500
    },
    {
      "epoch": 2.6420620972466313,
      "grad_norm": 0.5297181010246277,
      "learning_rate": 0.0002624007290717355,
      "loss": 0.1575,
      "step": 4510
    },
    {
      "epoch": 2.6479203280609256,
      "grad_norm": 0.9857646226882935,
      "learning_rate": 0.00026174977216508266,
      "loss": 0.1359,
      "step": 4520
    },
    {
      "epoch": 2.65377855887522,
      "grad_norm": 0.3707817494869232,
      "learning_rate": 0.00026109881525842986,
      "loss": 0.1226,
      "step": 4530
    },
    {
      "epoch": 2.6596367896895137,
      "grad_norm": 1.480391263961792,
      "learning_rate": 0.00026044785835177716,
      "loss": 0.1046,
      "step": 4540
    },
    {
      "epoch": 2.665495020503808,
      "grad_norm": 0.7787607908248901,
      "learning_rate": 0.00025979690144512435,
      "loss": 0.1064,
      "step": 4550
    },
    {
      "epoch": 2.671353251318102,
      "grad_norm": 0.7129953503608704,
      "learning_rate": 0.0002591459445384716,
      "loss": 0.0566,
      "step": 4560
    },
    {
      "epoch": 2.677211482132396,
      "grad_norm": 0.8584160804748535,
      "learning_rate": 0.0002584949876318188,
      "loss": 0.1639,
      "step": 4570
    },
    {
      "epoch": 2.68306971294669,
      "grad_norm": 1.6028215885162354,
      "learning_rate": 0.000257844030725166,
      "loss": 0.1539,
      "step": 4580
    },
    {
      "epoch": 2.688927943760984,
      "grad_norm": 0.49101898074150085,
      "learning_rate": 0.0002571930738185132,
      "loss": 0.105,
      "step": 4590
    },
    {
      "epoch": 2.6947861745752784,
      "grad_norm": 2.8125131130218506,
      "learning_rate": 0.0002565421169118604,
      "loss": 0.1414,
      "step": 4600
    },
    {
      "epoch": 2.700644405389572,
      "grad_norm": 0.6218743324279785,
      "learning_rate": 0.00025589116000520766,
      "loss": 0.123,
      "step": 4610
    },
    {
      "epoch": 2.7065026362038664,
      "grad_norm": 1.4385967254638672,
      "learning_rate": 0.0002552402030985549,
      "loss": 0.142,
      "step": 4620
    },
    {
      "epoch": 2.7123608670181607,
      "grad_norm": 0.36521002650260925,
      "learning_rate": 0.0002545892461919021,
      "loss": 0.0797,
      "step": 4630
    },
    {
      "epoch": 2.7182190978324545,
      "grad_norm": 1.6112006902694702,
      "learning_rate": 0.0002539382892852493,
      "loss": 0.1632,
      "step": 4640
    },
    {
      "epoch": 2.724077328646749,
      "grad_norm": 0.47123655676841736,
      "learning_rate": 0.00025328733237859653,
      "loss": 0.1085,
      "step": 4650
    },
    {
      "epoch": 2.7299355594610426,
      "grad_norm": 0.8812481164932251,
      "learning_rate": 0.0002526363754719437,
      "loss": 0.1075,
      "step": 4660
    },
    {
      "epoch": 2.735793790275337,
      "grad_norm": 1.782219648361206,
      "learning_rate": 0.000251985418565291,
      "loss": 0.1371,
      "step": 4670
    },
    {
      "epoch": 2.7416520210896307,
      "grad_norm": 1.105644941329956,
      "learning_rate": 0.0002513344616586382,
      "loss": 0.1572,
      "step": 4680
    },
    {
      "epoch": 2.747510251903925,
      "grad_norm": 1.425328016281128,
      "learning_rate": 0.0002506835047519854,
      "loss": 0.131,
      "step": 4690
    },
    {
      "epoch": 2.753368482718219,
      "grad_norm": 0.43381068110466003,
      "learning_rate": 0.00025003254784533265,
      "loss": 0.1552,
      "step": 4700
    },
    {
      "epoch": 2.759226713532513,
      "grad_norm": 2.056269645690918,
      "learning_rate": 0.0002493815909386799,
      "loss": 0.1155,
      "step": 4710
    },
    {
      "epoch": 2.7650849443468073,
      "grad_norm": 1.0049457550048828,
      "learning_rate": 0.0002487306340320271,
      "loss": 0.1399,
      "step": 4720
    },
    {
      "epoch": 2.7709431751611016,
      "grad_norm": 2.309018135070801,
      "learning_rate": 0.0002480796771253743,
      "loss": 0.1408,
      "step": 4730
    },
    {
      "epoch": 2.7768014059753954,
      "grad_norm": 1.171257734298706,
      "learning_rate": 0.0002474287202187215,
      "loss": 0.1368,
      "step": 4740
    },
    {
      "epoch": 2.7826596367896896,
      "grad_norm": 1.2910953760147095,
      "learning_rate": 0.00024677776331206877,
      "loss": 0.1101,
      "step": 4750
    },
    {
      "epoch": 2.7885178676039835,
      "grad_norm": 0.9543918967247009,
      "learning_rate": 0.00024612680640541596,
      "loss": 0.1791,
      "step": 4760
    },
    {
      "epoch": 2.7943760984182777,
      "grad_norm": 0.6492213010787964,
      "learning_rate": 0.0002454758494987632,
      "loss": 0.1034,
      "step": 4770
    },
    {
      "epoch": 2.8002343292325715,
      "grad_norm": 1.3879276514053345,
      "learning_rate": 0.0002448248925921104,
      "loss": 0.1126,
      "step": 4780
    },
    {
      "epoch": 2.806092560046866,
      "grad_norm": 0.843720555305481,
      "learning_rate": 0.00024417393568545764,
      "loss": 0.097,
      "step": 4790
    },
    {
      "epoch": 2.81195079086116,
      "grad_norm": 0.4348345994949341,
      "learning_rate": 0.00024352297877880486,
      "loss": 0.1093,
      "step": 4800
    },
    {
      "epoch": 2.817809021675454,
      "grad_norm": 0.7387756705284119,
      "learning_rate": 0.00024287202187215208,
      "loss": 0.1409,
      "step": 4810
    },
    {
      "epoch": 2.823667252489748,
      "grad_norm": 1.1606628894805908,
      "learning_rate": 0.00024222106496549927,
      "loss": 0.1175,
      "step": 4820
    },
    {
      "epoch": 2.8295254833040424,
      "grad_norm": 0.33284834027290344,
      "learning_rate": 0.00024157010805884652,
      "loss": 0.1256,
      "step": 4830
    },
    {
      "epoch": 2.8353837141183362,
      "grad_norm": 1.8327375650405884,
      "learning_rate": 0.00024091915115219374,
      "loss": 0.1376,
      "step": 4840
    },
    {
      "epoch": 2.8412419449326305,
      "grad_norm": 0.3788931667804718,
      "learning_rate": 0.00024026819424554095,
      "loss": 0.1204,
      "step": 4850
    },
    {
      "epoch": 2.8471001757469243,
      "grad_norm": 0.8661250472068787,
      "learning_rate": 0.00023961723733888817,
      "loss": 0.1064,
      "step": 4860
    },
    {
      "epoch": 2.8529584065612186,
      "grad_norm": 0.6688357591629028,
      "learning_rate": 0.0002389662804322354,
      "loss": 0.1455,
      "step": 4870
    },
    {
      "epoch": 2.8588166373755124,
      "grad_norm": 0.8732729554176331,
      "learning_rate": 0.0002383153235255826,
      "loss": 0.0888,
      "step": 4880
    },
    {
      "epoch": 2.8646748681898067,
      "grad_norm": 0.4746171832084656,
      "learning_rate": 0.00023766436661892986,
      "loss": 0.1252,
      "step": 4890
    },
    {
      "epoch": 2.870533099004101,
      "grad_norm": 0.9698277115821838,
      "learning_rate": 0.00023701340971227705,
      "loss": 0.1077,
      "step": 4900
    },
    {
      "epoch": 2.8763913298183947,
      "grad_norm": 0.39930036664009094,
      "learning_rate": 0.00023636245280562427,
      "loss": 0.0795,
      "step": 4910
    },
    {
      "epoch": 2.882249560632689,
      "grad_norm": 0.9420844912528992,
      "learning_rate": 0.0002357114958989715,
      "loss": 0.1143,
      "step": 4920
    },
    {
      "epoch": 2.8881077914469833,
      "grad_norm": 0.8221577405929565,
      "learning_rate": 0.00023506053899231873,
      "loss": 0.1146,
      "step": 4930
    },
    {
      "epoch": 2.893966022261277,
      "grad_norm": 0.9865840673446655,
      "learning_rate": 0.00023440958208566592,
      "loss": 0.1409,
      "step": 4940
    },
    {
      "epoch": 2.899824253075571,
      "grad_norm": 0.7893118262290955,
      "learning_rate": 0.00023375862517901317,
      "loss": 0.1549,
      "step": 4950
    },
    {
      "epoch": 2.905682483889865,
      "grad_norm": 1.0813521146774292,
      "learning_rate": 0.00023310766827236038,
      "loss": 0.1122,
      "step": 4960
    },
    {
      "epoch": 2.9115407147041594,
      "grad_norm": 1.1200429201126099,
      "learning_rate": 0.0002324567113657076,
      "loss": 0.0944,
      "step": 4970
    },
    {
      "epoch": 2.9173989455184532,
      "grad_norm": 1.6647378206253052,
      "learning_rate": 0.0002318057544590548,
      "loss": 0.1129,
      "step": 4980
    },
    {
      "epoch": 2.9232571763327475,
      "grad_norm": 0.7454376220703125,
      "learning_rate": 0.00023115479755240204,
      "loss": 0.1238,
      "step": 4990
    },
    {
      "epoch": 2.9291154071470418,
      "grad_norm": 0.5001448392868042,
      "learning_rate": 0.00023050384064574926,
      "loss": 0.0679,
      "step": 5000
    },
    {
      "epoch": 2.9349736379613356,
      "grad_norm": 0.8018816709518433,
      "learning_rate": 0.00022985288373909645,
      "loss": 0.115,
      "step": 5010
    },
    {
      "epoch": 2.94083186877563,
      "grad_norm": 0.793289840221405,
      "learning_rate": 0.0002292019268324437,
      "loss": 0.144,
      "step": 5020
    },
    {
      "epoch": 2.946690099589924,
      "grad_norm": 1.4793994426727295,
      "learning_rate": 0.00022855096992579091,
      "loss": 0.126,
      "step": 5030
    },
    {
      "epoch": 2.952548330404218,
      "grad_norm": 1.5347057580947876,
      "learning_rate": 0.00022790001301913813,
      "loss": 0.1249,
      "step": 5040
    },
    {
      "epoch": 2.9584065612185118,
      "grad_norm": 0.9479228258132935,
      "learning_rate": 0.00022724905611248535,
      "loss": 0.1251,
      "step": 5050
    },
    {
      "epoch": 2.964264792032806,
      "grad_norm": 0.7511537075042725,
      "learning_rate": 0.00022659809920583257,
      "loss": 0.1311,
      "step": 5060
    },
    {
      "epoch": 2.9701230228471003,
      "grad_norm": 1.184127926826477,
      "learning_rate": 0.0002259471422991798,
      "loss": 0.1497,
      "step": 5070
    },
    {
      "epoch": 2.975981253661394,
      "grad_norm": 0.5232158899307251,
      "learning_rate": 0.00022529618539252703,
      "loss": 0.1539,
      "step": 5080
    },
    {
      "epoch": 2.9818394844756884,
      "grad_norm": 0.8987466096878052,
      "learning_rate": 0.00022464522848587422,
      "loss": 0.0837,
      "step": 5090
    },
    {
      "epoch": 2.9876977152899826,
      "grad_norm": 0.8917015790939331,
      "learning_rate": 0.00022399427157922144,
      "loss": 0.1197,
      "step": 5100
    },
    {
      "epoch": 2.9935559461042764,
      "grad_norm": 0.277315616607666,
      "learning_rate": 0.0002233433146725687,
      "loss": 0.1388,
      "step": 5110
    },
    {
      "epoch": 2.9994141769185707,
      "grad_norm": 0.8018264770507812,
      "learning_rate": 0.0002226923577659159,
      "loss": 0.1113,
      "step": 5120
    },
    {
      "epoch": 3.0,
      "eval_f1": 0.9308659217877095,
      "eval_loss": 0.19312748312950134,
      "eval_runtime": 14.7246,
      "eval_samples_per_second": 203.741,
      "eval_steps_per_second": 6.384,
      "step": 5121
    },
    {
      "epoch": 3.0052724077328645,
      "grad_norm": 0.6647753715515137,
      "learning_rate": 0.0002220414008592631,
      "loss": 0.1585,
      "step": 5130
    },
    {
      "epoch": 3.011130638547159,
      "grad_norm": 1.3105008602142334,
      "learning_rate": 0.00022139044395261034,
      "loss": 0.1092,
      "step": 5140
    },
    {
      "epoch": 3.016988869361453,
      "grad_norm": 0.5599445700645447,
      "learning_rate": 0.00022073948704595756,
      "loss": 0.0809,
      "step": 5150
    },
    {
      "epoch": 3.022847100175747,
      "grad_norm": 2.208543539047241,
      "learning_rate": 0.00022008853013930478,
      "loss": 0.1188,
      "step": 5160
    },
    {
      "epoch": 3.028705330990041,
      "grad_norm": 1.3378957509994507,
      "learning_rate": 0.000219437573232652,
      "loss": 0.0943,
      "step": 5170
    },
    {
      "epoch": 3.034563561804335,
      "grad_norm": 1.3608115911483765,
      "learning_rate": 0.00021878661632599922,
      "loss": 0.0886,
      "step": 5180
    },
    {
      "epoch": 3.040421792618629,
      "grad_norm": 0.6845038533210754,
      "learning_rate": 0.00021813565941934644,
      "loss": 0.0881,
      "step": 5190
    },
    {
      "epoch": 3.0462800234329235,
      "grad_norm": 1.1577521562576294,
      "learning_rate": 0.00021748470251269368,
      "loss": 0.1168,
      "step": 5200
    },
    {
      "epoch": 3.0521382542472173,
      "grad_norm": 0.5686136484146118,
      "learning_rate": 0.00021683374560604087,
      "loss": 0.0874,
      "step": 5210
    },
    {
      "epoch": 3.0579964850615116,
      "grad_norm": 1.3156239986419678,
      "learning_rate": 0.0002161827886993881,
      "loss": 0.1072,
      "step": 5220
    },
    {
      "epoch": 3.0638547158758054,
      "grad_norm": 1.4853498935699463,
      "learning_rate": 0.00021553183179273534,
      "loss": 0.1296,
      "step": 5230
    },
    {
      "epoch": 3.0697129466900996,
      "grad_norm": 0.6757405400276184,
      "learning_rate": 0.00021488087488608256,
      "loss": 0.0878,
      "step": 5240
    },
    {
      "epoch": 3.0755711775043935,
      "grad_norm": 1.7888120412826538,
      "learning_rate": 0.00021422991797942975,
      "loss": 0.1324,
      "step": 5250
    },
    {
      "epoch": 3.0814294083186877,
      "grad_norm": 0.30726903676986694,
      "learning_rate": 0.000213578961072777,
      "loss": 0.0949,
      "step": 5260
    },
    {
      "epoch": 3.087287639132982,
      "grad_norm": 1.2144439220428467,
      "learning_rate": 0.0002129280041661242,
      "loss": 0.0923,
      "step": 5270
    },
    {
      "epoch": 3.093145869947276,
      "grad_norm": 1.7017472982406616,
      "learning_rate": 0.00021227704725947143,
      "loss": 0.1178,
      "step": 5280
    },
    {
      "epoch": 3.09900410076157,
      "grad_norm": 0.6928281784057617,
      "learning_rate": 0.00021162609035281865,
      "loss": 0.1124,
      "step": 5290
    },
    {
      "epoch": 3.104862331575864,
      "grad_norm": 1.007668375968933,
      "learning_rate": 0.00021097513344616587,
      "loss": 0.1586,
      "step": 5300
    },
    {
      "epoch": 3.110720562390158,
      "grad_norm": 1.729608178138733,
      "learning_rate": 0.00021032417653951308,
      "loss": 0.1273,
      "step": 5310
    },
    {
      "epoch": 3.1165787932044524,
      "grad_norm": 0.3571639955043793,
      "learning_rate": 0.00020967321963286033,
      "loss": 0.1102,
      "step": 5320
    },
    {
      "epoch": 3.1224370240187462,
      "grad_norm": 0.6538040637969971,
      "learning_rate": 0.00020902226272620752,
      "loss": 0.0988,
      "step": 5330
    },
    {
      "epoch": 3.1282952548330405,
      "grad_norm": 1.0057750940322876,
      "learning_rate": 0.00020837130581955474,
      "loss": 0.0928,
      "step": 5340
    },
    {
      "epoch": 3.1341534856473343,
      "grad_norm": 0.6291601061820984,
      "learning_rate": 0.00020772034891290199,
      "loss": 0.0826,
      "step": 5350
    },
    {
      "epoch": 3.1400117164616286,
      "grad_norm": 1.195389986038208,
      "learning_rate": 0.0002070693920062492,
      "loss": 0.127,
      "step": 5360
    },
    {
      "epoch": 3.145869947275923,
      "grad_norm": 1.0781468152999878,
      "learning_rate": 0.0002064184350995964,
      "loss": 0.1017,
      "step": 5370
    },
    {
      "epoch": 3.1517281780902167,
      "grad_norm": 0.3608550727367401,
      "learning_rate": 0.00020576747819294364,
      "loss": 0.1006,
      "step": 5380
    },
    {
      "epoch": 3.157586408904511,
      "grad_norm": 0.2512178122997284,
      "learning_rate": 0.00020511652128629086,
      "loss": 0.0945,
      "step": 5390
    },
    {
      "epoch": 3.1634446397188047,
      "grad_norm": 1.0521883964538574,
      "learning_rate": 0.00020446556437963805,
      "loss": 0.0758,
      "step": 5400
    },
    {
      "epoch": 3.169302870533099,
      "grad_norm": 0.8204757571220398,
      "learning_rate": 0.0002038146074729853,
      "loss": 0.1369,
      "step": 5410
    },
    {
      "epoch": 3.1751611013473933,
      "grad_norm": 0.2992514669895172,
      "learning_rate": 0.00020316365056633251,
      "loss": 0.0798,
      "step": 5420
    },
    {
      "epoch": 3.181019332161687,
      "grad_norm": 2.212993621826172,
      "learning_rate": 0.00020251269365967973,
      "loss": 0.0935,
      "step": 5430
    },
    {
      "epoch": 3.1868775629759813,
      "grad_norm": 0.2897680401802063,
      "learning_rate": 0.00020186173675302695,
      "loss": 0.0947,
      "step": 5440
    },
    {
      "epoch": 3.192735793790275,
      "grad_norm": 2.02703857421875,
      "learning_rate": 0.00020121077984637417,
      "loss": 0.1168,
      "step": 5450
    },
    {
      "epoch": 3.1985940246045694,
      "grad_norm": 0.859622597694397,
      "learning_rate": 0.0002005598229397214,
      "loss": 0.131,
      "step": 5460
    },
    {
      "epoch": 3.2044522554188637,
      "grad_norm": 1.419415831565857,
      "learning_rate": 0.00019990886603306863,
      "loss": 0.0949,
      "step": 5470
    },
    {
      "epoch": 3.2103104862331575,
      "grad_norm": 0.9570263028144836,
      "learning_rate": 0.00019925790912641583,
      "loss": 0.0848,
      "step": 5480
    },
    {
      "epoch": 3.2161687170474518,
      "grad_norm": 0.33368825912475586,
      "learning_rate": 0.00019860695221976304,
      "loss": 0.1125,
      "step": 5490
    },
    {
      "epoch": 3.2220269478617456,
      "grad_norm": 0.3054542541503906,
      "learning_rate": 0.0001979559953131103,
      "loss": 0.0947,
      "step": 5500
    },
    {
      "epoch": 3.22788517867604,
      "grad_norm": 0.5232771635055542,
      "learning_rate": 0.0001973050384064575,
      "loss": 0.1145,
      "step": 5510
    },
    {
      "epoch": 3.233743409490334,
      "grad_norm": 1.5356754064559937,
      "learning_rate": 0.0001966540814998047,
      "loss": 0.0928,
      "step": 5520
    },
    {
      "epoch": 3.239601640304628,
      "grad_norm": 1.7630113363265991,
      "learning_rate": 0.00019600312459315194,
      "loss": 0.0976,
      "step": 5530
    },
    {
      "epoch": 3.245459871118922,
      "grad_norm": 0.34040358662605286,
      "learning_rate": 0.00019535216768649916,
      "loss": 0.0716,
      "step": 5540
    },
    {
      "epoch": 3.251318101933216,
      "grad_norm": 0.37931734323501587,
      "learning_rate": 0.00019470121077984638,
      "loss": 0.127,
      "step": 5550
    },
    {
      "epoch": 3.2571763327475103,
      "grad_norm": 1.5484894514083862,
      "learning_rate": 0.0001940502538731936,
      "loss": 0.1451,
      "step": 5560
    },
    {
      "epoch": 3.2630345635618045,
      "grad_norm": 1.216162919998169,
      "learning_rate": 0.00019339929696654082,
      "loss": 0.0996,
      "step": 5570
    },
    {
      "epoch": 3.2688927943760984,
      "grad_norm": 0.9533673524856567,
      "learning_rate": 0.00019274834005988804,
      "loss": 0.1204,
      "step": 5580
    },
    {
      "epoch": 3.2747510251903926,
      "grad_norm": 0.6540274024009705,
      "learning_rate": 0.00019209738315323528,
      "loss": 0.1175,
      "step": 5590
    },
    {
      "epoch": 3.2806092560046864,
      "grad_norm": 0.3820449113845825,
      "learning_rate": 0.00019144642624658247,
      "loss": 0.1616,
      "step": 5600
    },
    {
      "epoch": 3.2864674868189807,
      "grad_norm": 0.7934489846229553,
      "learning_rate": 0.0001907954693399297,
      "loss": 0.1217,
      "step": 5610
    },
    {
      "epoch": 3.2923257176332745,
      "grad_norm": 0.5252192616462708,
      "learning_rate": 0.00019014451243327694,
      "loss": 0.1411,
      "step": 5620
    },
    {
      "epoch": 3.298183948447569,
      "grad_norm": 0.40958625078201294,
      "learning_rate": 0.00018949355552662416,
      "loss": 0.1228,
      "step": 5630
    },
    {
      "epoch": 3.304042179261863,
      "grad_norm": 0.5728265047073364,
      "learning_rate": 0.00018884259861997135,
      "loss": 0.0719,
      "step": 5640
    },
    {
      "epoch": 3.309900410076157,
      "grad_norm": 0.24659055471420288,
      "learning_rate": 0.0001881916417133186,
      "loss": 0.1438,
      "step": 5650
    },
    {
      "epoch": 3.315758640890451,
      "grad_norm": 0.8120065927505493,
      "learning_rate": 0.0001875406848066658,
      "loss": 0.163,
      "step": 5660
    },
    {
      "epoch": 3.3216168717047454,
      "grad_norm": 0.25870636105537415,
      "learning_rate": 0.00018688972790001303,
      "loss": 0.0843,
      "step": 5670
    },
    {
      "epoch": 3.327475102519039,
      "grad_norm": 1.5063252449035645,
      "learning_rate": 0.00018623877099336025,
      "loss": 0.1041,
      "step": 5680
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.8163599371910095,
      "learning_rate": 0.00018558781408670747,
      "loss": 0.1018,
      "step": 5690
    },
    {
      "epoch": 3.3391915641476273,
      "grad_norm": 1.1529045104980469,
      "learning_rate": 0.00018493685718005468,
      "loss": 0.1135,
      "step": 5700
    },
    {
      "epoch": 3.3450497949619216,
      "grad_norm": 0.39233940839767456,
      "learning_rate": 0.0001842859002734019,
      "loss": 0.096,
      "step": 5710
    },
    {
      "epoch": 3.3509080257762154,
      "grad_norm": 0.9688923358917236,
      "learning_rate": 0.00018363494336674912,
      "loss": 0.1441,
      "step": 5720
    },
    {
      "epoch": 3.3567662565905096,
      "grad_norm": 1.1118249893188477,
      "learning_rate": 0.00018298398646009634,
      "loss": 0.1334,
      "step": 5730
    },
    {
      "epoch": 3.362624487404804,
      "grad_norm": 0.6611572504043579,
      "learning_rate": 0.00018233302955344356,
      "loss": 0.0873,
      "step": 5740
    },
    {
      "epoch": 3.3684827182190977,
      "grad_norm": 1.4174485206604004,
      "learning_rate": 0.0001816820726467908,
      "loss": 0.138,
      "step": 5750
    },
    {
      "epoch": 3.374340949033392,
      "grad_norm": 0.5151809453964233,
      "learning_rate": 0.000181031115740138,
      "loss": 0.121,
      "step": 5760
    },
    {
      "epoch": 3.380199179847686,
      "grad_norm": 0.7787996530532837,
      "learning_rate": 0.00018038015883348521,
      "loss": 0.0834,
      "step": 5770
    },
    {
      "epoch": 3.38605741066198,
      "grad_norm": 0.3302878737449646,
      "learning_rate": 0.00017972920192683246,
      "loss": 0.107,
      "step": 5780
    },
    {
      "epoch": 3.3919156414762743,
      "grad_norm": 0.5792067646980286,
      "learning_rate": 0.00017907824502017965,
      "loss": 0.1134,
      "step": 5790
    },
    {
      "epoch": 3.397773872290568,
      "grad_norm": 0.9854441285133362,
      "learning_rate": 0.00017842728811352687,
      "loss": 0.1388,
      "step": 5800
    },
    {
      "epoch": 3.4036321031048624,
      "grad_norm": 0.3148748278617859,
      "learning_rate": 0.00017777633120687411,
      "loss": 0.1036,
      "step": 5810
    },
    {
      "epoch": 3.4094903339191562,
      "grad_norm": 0.8012061715126038,
      "learning_rate": 0.00017712537430022133,
      "loss": 0.0934,
      "step": 5820
    },
    {
      "epoch": 3.4153485647334505,
      "grad_norm": 1.5753650665283203,
      "learning_rate": 0.00017647441739356852,
      "loss": 0.087,
      "step": 5830
    },
    {
      "epoch": 3.4212067955477448,
      "grad_norm": 0.9657014608383179,
      "learning_rate": 0.00017582346048691577,
      "loss": 0.1111,
      "step": 5840
    },
    {
      "epoch": 3.4270650263620386,
      "grad_norm": 2.233928918838501,
      "learning_rate": 0.000175172503580263,
      "loss": 0.1123,
      "step": 5850
    },
    {
      "epoch": 3.432923257176333,
      "grad_norm": 0.3911862373352051,
      "learning_rate": 0.0001745215466736102,
      "loss": 0.0758,
      "step": 5860
    },
    {
      "epoch": 3.4387814879906267,
      "grad_norm": 0.20068423449993134,
      "learning_rate": 0.00017387058976695743,
      "loss": 0.1157,
      "step": 5870
    },
    {
      "epoch": 3.444639718804921,
      "grad_norm": 1.0492706298828125,
      "learning_rate": 0.00017321963286030464,
      "loss": 0.0744,
      "step": 5880
    },
    {
      "epoch": 3.450497949619215,
      "grad_norm": 0.5000963807106018,
      "learning_rate": 0.00017256867595365186,
      "loss": 0.1346,
      "step": 5890
    },
    {
      "epoch": 3.456356180433509,
      "grad_norm": 0.9173428416252136,
      "learning_rate": 0.0001719177190469991,
      "loss": 0.0918,
      "step": 5900
    },
    {
      "epoch": 3.4622144112478033,
      "grad_norm": 0.624330461025238,
      "learning_rate": 0.0001712667621403463,
      "loss": 0.1151,
      "step": 5910
    },
    {
      "epoch": 3.468072642062097,
      "grad_norm": 0.414789080619812,
      "learning_rate": 0.00017061580523369352,
      "loss": 0.0942,
      "step": 5920
    },
    {
      "epoch": 3.4739308728763914,
      "grad_norm": 1.336959719657898,
      "learning_rate": 0.00016996484832704076,
      "loss": 0.1209,
      "step": 5930
    },
    {
      "epoch": 3.4797891036906856,
      "grad_norm": 0.5462775826454163,
      "learning_rate": 0.00016931389142038798,
      "loss": 0.0892,
      "step": 5940
    },
    {
      "epoch": 3.4856473345049794,
      "grad_norm": 0.468053936958313,
      "learning_rate": 0.00016866293451373517,
      "loss": 0.1291,
      "step": 5950
    },
    {
      "epoch": 3.4915055653192737,
      "grad_norm": 0.46274715662002563,
      "learning_rate": 0.00016801197760708242,
      "loss": 0.1401,
      "step": 5960
    },
    {
      "epoch": 3.4973637961335675,
      "grad_norm": 0.5142794847488403,
      "learning_rate": 0.00016736102070042964,
      "loss": 0.1012,
      "step": 5970
    },
    {
      "epoch": 3.503222026947862,
      "grad_norm": 0.14458006620407104,
      "learning_rate": 0.00016671006379377686,
      "loss": 0.0723,
      "step": 5980
    },
    {
      "epoch": 3.5090802577621556,
      "grad_norm": 1.0561031103134155,
      "learning_rate": 0.00016605910688712407,
      "loss": 0.1175,
      "step": 5990
    },
    {
      "epoch": 3.51493848857645,
      "grad_norm": 0.40694454312324524,
      "learning_rate": 0.0001654081499804713,
      "loss": 0.092,
      "step": 6000
    },
    {
      "epoch": 3.520796719390744,
      "grad_norm": 0.6344606280326843,
      "learning_rate": 0.0001647571930738185,
      "loss": 0.0944,
      "step": 6010
    },
    {
      "epoch": 3.526654950205038,
      "grad_norm": 1.8507148027420044,
      "learning_rate": 0.00016410623616716576,
      "loss": 0.1041,
      "step": 6020
    },
    {
      "epoch": 3.532513181019332,
      "grad_norm": 0.6685912013053894,
      "learning_rate": 0.00016345527926051295,
      "loss": 0.132,
      "step": 6030
    },
    {
      "epoch": 3.5383714118336265,
      "grad_norm": 1.3463377952575684,
      "learning_rate": 0.00016280432235386017,
      "loss": 0.1003,
      "step": 6040
    },
    {
      "epoch": 3.5442296426479203,
      "grad_norm": 2.4183425903320312,
      "learning_rate": 0.0001621533654472074,
      "loss": 0.1197,
      "step": 6050
    },
    {
      "epoch": 3.5500878734622145,
      "grad_norm": 0.753265917301178,
      "learning_rate": 0.00016150240854055463,
      "loss": 0.1071,
      "step": 6060
    },
    {
      "epoch": 3.5559461042765084,
      "grad_norm": 0.4786781966686249,
      "learning_rate": 0.00016085145163390182,
      "loss": 0.1032,
      "step": 6070
    },
    {
      "epoch": 3.5618043350908026,
      "grad_norm": 0.960931658744812,
      "learning_rate": 0.00016020049472724907,
      "loss": 0.0984,
      "step": 6080
    },
    {
      "epoch": 3.5676625659050965,
      "grad_norm": 0.32585757970809937,
      "learning_rate": 0.00015954953782059629,
      "loss": 0.1358,
      "step": 6090
    },
    {
      "epoch": 3.5735207967193907,
      "grad_norm": 1.226641297340393,
      "learning_rate": 0.0001588985809139435,
      "loss": 0.1098,
      "step": 6100
    },
    {
      "epoch": 3.579379027533685,
      "grad_norm": 2.271745204925537,
      "learning_rate": 0.00015824762400729072,
      "loss": 0.1069,
      "step": 6110
    },
    {
      "epoch": 3.585237258347979,
      "grad_norm": 0.6549329161643982,
      "learning_rate": 0.00015759666710063794,
      "loss": 0.0986,
      "step": 6120
    },
    {
      "epoch": 3.591095489162273,
      "grad_norm": 1.0768980979919434,
      "learning_rate": 0.00015694571019398516,
      "loss": 0.0883,
      "step": 6130
    },
    {
      "epoch": 3.5969537199765673,
      "grad_norm": 0.5941785573959351,
      "learning_rate": 0.0001562947532873324,
      "loss": 0.1091,
      "step": 6140
    },
    {
      "epoch": 3.602811950790861,
      "grad_norm": 1.535934567451477,
      "learning_rate": 0.0001556437963806796,
      "loss": 0.0819,
      "step": 6150
    },
    {
      "epoch": 3.6086701816051554,
      "grad_norm": 0.905866801738739,
      "learning_rate": 0.00015499283947402681,
      "loss": 0.1222,
      "step": 6160
    },
    {
      "epoch": 3.614528412419449,
      "grad_norm": 2.089005470275879,
      "learning_rate": 0.00015434188256737406,
      "loss": 0.1122,
      "step": 6170
    },
    {
      "epoch": 3.6203866432337435,
      "grad_norm": 1.4815788269042969,
      "learning_rate": 0.00015369092566072125,
      "loss": 0.1478,
      "step": 6180
    },
    {
      "epoch": 3.6262448740480373,
      "grad_norm": 1.1967289447784424,
      "learning_rate": 0.00015303996875406847,
      "loss": 0.0809,
      "step": 6190
    },
    {
      "epoch": 3.6321031048623316,
      "grad_norm": 0.13463656604290009,
      "learning_rate": 0.00015238901184741572,
      "loss": 0.0899,
      "step": 6200
    },
    {
      "epoch": 3.637961335676626,
      "grad_norm": 0.9791207909584045,
      "learning_rate": 0.00015173805494076293,
      "loss": 0.1219,
      "step": 6210
    },
    {
      "epoch": 3.6438195664909196,
      "grad_norm": 1.346298336982727,
      "learning_rate": 0.00015108709803411013,
      "loss": 0.1201,
      "step": 6220
    },
    {
      "epoch": 3.649677797305214,
      "grad_norm": 0.5890740752220154,
      "learning_rate": 0.00015043614112745737,
      "loss": 0.0562,
      "step": 6230
    },
    {
      "epoch": 3.655536028119508,
      "grad_norm": 1.8662189245224,
      "learning_rate": 0.0001497851842208046,
      "loss": 0.1372,
      "step": 6240
    },
    {
      "epoch": 3.661394258933802,
      "grad_norm": 0.24533328413963318,
      "learning_rate": 0.0001491342273141518,
      "loss": 0.1056,
      "step": 6250
    },
    {
      "epoch": 3.667252489748096,
      "grad_norm": 1.1817823648452759,
      "learning_rate": 0.00014848327040749903,
      "loss": 0.079,
      "step": 6260
    },
    {
      "epoch": 3.67311072056239,
      "grad_norm": 1.736426591873169,
      "learning_rate": 0.00014783231350084624,
      "loss": 0.1301,
      "step": 6270
    },
    {
      "epoch": 3.6789689513766843,
      "grad_norm": 1.0973541736602783,
      "learning_rate": 0.00014718135659419346,
      "loss": 0.0838,
      "step": 6280
    },
    {
      "epoch": 3.684827182190978,
      "grad_norm": 0.12063330411911011,
      "learning_rate": 0.0001465303996875407,
      "loss": 0.1164,
      "step": 6290
    },
    {
      "epoch": 3.6906854130052724,
      "grad_norm": 0.8762880563735962,
      "learning_rate": 0.0001458794427808879,
      "loss": 0.1077,
      "step": 6300
    },
    {
      "epoch": 3.6965436438195667,
      "grad_norm": 1.3317028284072876,
      "learning_rate": 0.00014522848587423512,
      "loss": 0.1124,
      "step": 6310
    },
    {
      "epoch": 3.7024018746338605,
      "grad_norm": 1.8302267789840698,
      "learning_rate": 0.00014457752896758236,
      "loss": 0.1027,
      "step": 6320
    },
    {
      "epoch": 3.7082601054481548,
      "grad_norm": 0.6655034422874451,
      "learning_rate": 0.00014392657206092958,
      "loss": 0.1061,
      "step": 6330
    },
    {
      "epoch": 3.7141183362624486,
      "grad_norm": 1.7261115312576294,
      "learning_rate": 0.00014327561515427677,
      "loss": 0.0996,
      "step": 6340
    },
    {
      "epoch": 3.719976567076743,
      "grad_norm": 1.1165367364883423,
      "learning_rate": 0.00014262465824762402,
      "loss": 0.0936,
      "step": 6350
    },
    {
      "epoch": 3.7258347978910367,
      "grad_norm": 2.128887414932251,
      "learning_rate": 0.00014197370134097124,
      "loss": 0.1273,
      "step": 6360
    },
    {
      "epoch": 3.731693028705331,
      "grad_norm": 1.7992889881134033,
      "learning_rate": 0.00014132274443431846,
      "loss": 0.1361,
      "step": 6370
    },
    {
      "epoch": 3.737551259519625,
      "grad_norm": 1.342489242553711,
      "learning_rate": 0.00014067178752766567,
      "loss": 0.1468,
      "step": 6380
    },
    {
      "epoch": 3.743409490333919,
      "grad_norm": 0.9803612232208252,
      "learning_rate": 0.0001400208306210129,
      "loss": 0.0943,
      "step": 6390
    },
    {
      "epoch": 3.7492677211482133,
      "grad_norm": 0.8913155198097229,
      "learning_rate": 0.0001393698737143601,
      "loss": 0.1049,
      "step": 6400
    },
    {
      "epoch": 3.7551259519625075,
      "grad_norm": 1.2912554740905762,
      "learning_rate": 0.00013871891680770736,
      "loss": 0.0962,
      "step": 6410
    },
    {
      "epoch": 3.7609841827768014,
      "grad_norm": 0.7567240595817566,
      "learning_rate": 0.00013806795990105455,
      "loss": 0.102,
      "step": 6420
    },
    {
      "epoch": 3.7668424135910956,
      "grad_norm": 1.3401840925216675,
      "learning_rate": 0.00013741700299440177,
      "loss": 0.1328,
      "step": 6430
    },
    {
      "epoch": 3.7727006444053894,
      "grad_norm": 0.6872816681861877,
      "learning_rate": 0.000136766046087749,
      "loss": 0.1094,
      "step": 6440
    },
    {
      "epoch": 3.7785588752196837,
      "grad_norm": 0.5628686547279358,
      "learning_rate": 0.00013611508918109623,
      "loss": 0.078,
      "step": 6450
    },
    {
      "epoch": 3.7844171060339775,
      "grad_norm": 0.6823394298553467,
      "learning_rate": 0.00013546413227444342,
      "loss": 0.0825,
      "step": 6460
    },
    {
      "epoch": 3.790275336848272,
      "grad_norm": 1.2937862873077393,
      "learning_rate": 0.00013481317536779064,
      "loss": 0.1136,
      "step": 6470
    },
    {
      "epoch": 3.796133567662566,
      "grad_norm": 1.6106140613555908,
      "learning_rate": 0.00013416221846113789,
      "loss": 0.0787,
      "step": 6480
    },
    {
      "epoch": 3.80199179847686,
      "grad_norm": 0.094044528901577,
      "learning_rate": 0.0001335112615544851,
      "loss": 0.0737,
      "step": 6490
    },
    {
      "epoch": 3.807850029291154,
      "grad_norm": 1.3219674825668335,
      "learning_rate": 0.0001328603046478323,
      "loss": 0.1171,
      "step": 6500
    },
    {
      "epoch": 3.8137082601054484,
      "grad_norm": 0.3359241783618927,
      "learning_rate": 0.00013220934774117954,
      "loss": 0.0874,
      "step": 6510
    },
    {
      "epoch": 3.819566490919742,
      "grad_norm": 1.6946830749511719,
      "learning_rate": 0.00013155839083452676,
      "loss": 0.1243,
      "step": 6520
    },
    {
      "epoch": 3.8254247217340365,
      "grad_norm": 0.7258983850479126,
      "learning_rate": 0.00013090743392787395,
      "loss": 0.106,
      "step": 6530
    },
    {
      "epoch": 3.8312829525483303,
      "grad_norm": 1.6327470541000366,
      "learning_rate": 0.0001302564770212212,
      "loss": 0.0981,
      "step": 6540
    },
    {
      "epoch": 3.8371411833626246,
      "grad_norm": 0.45188337564468384,
      "learning_rate": 0.00012960552011456842,
      "loss": 0.049,
      "step": 6550
    },
    {
      "epoch": 3.8429994141769184,
      "grad_norm": 1.3675637245178223,
      "learning_rate": 0.00012895456320791563,
      "loss": 0.0854,
      "step": 6560
    },
    {
      "epoch": 3.8488576449912126,
      "grad_norm": 0.9713359475135803,
      "learning_rate": 0.00012830360630126285,
      "loss": 0.0827,
      "step": 6570
    },
    {
      "epoch": 3.854715875805507,
      "grad_norm": 0.08092973381280899,
      "learning_rate": 0.00012765264939461007,
      "loss": 0.1022,
      "step": 6580
    },
    {
      "epoch": 3.8605741066198007,
      "grad_norm": 1.1305333375930786,
      "learning_rate": 0.0001270016924879573,
      "loss": 0.0997,
      "step": 6590
    },
    {
      "epoch": 3.866432337434095,
      "grad_norm": 2.724644184112549,
      "learning_rate": 0.00012635073558130453,
      "loss": 0.0637,
      "step": 6600
    },
    {
      "epoch": 3.8722905682483892,
      "grad_norm": 1.3331657648086548,
      "learning_rate": 0.00012569977867465173,
      "loss": 0.1093,
      "step": 6610
    },
    {
      "epoch": 3.878148799062683,
      "grad_norm": 1.6242687702178955,
      "learning_rate": 0.00012504882176799894,
      "loss": 0.0791,
      "step": 6620
    },
    {
      "epoch": 3.884007029876977,
      "grad_norm": 0.24345877766609192,
      "learning_rate": 0.0001243978648613462,
      "loss": 0.0932,
      "step": 6630
    },
    {
      "epoch": 3.889865260691271,
      "grad_norm": 1.968440294265747,
      "learning_rate": 0.0001237469079546934,
      "loss": 0.1046,
      "step": 6640
    },
    {
      "epoch": 3.8957234915055654,
      "grad_norm": 1.3497897386550903,
      "learning_rate": 0.00012309595104804063,
      "loss": 0.1171,
      "step": 6650
    },
    {
      "epoch": 3.9015817223198592,
      "grad_norm": 0.794064462184906,
      "learning_rate": 0.00012244499414138785,
      "loss": 0.0912,
      "step": 6660
    },
    {
      "epoch": 3.9074399531341535,
      "grad_norm": 1.1040220260620117,
      "learning_rate": 0.00012179403723473506,
      "loss": 0.0932,
      "step": 6670
    },
    {
      "epoch": 3.9132981839484478,
      "grad_norm": 0.6346971988677979,
      "learning_rate": 0.00012114308032808228,
      "loss": 0.0786,
      "step": 6680
    },
    {
      "epoch": 3.9191564147627416,
      "grad_norm": 1.220487117767334,
      "learning_rate": 0.0001204921234214295,
      "loss": 0.1191,
      "step": 6690
    },
    {
      "epoch": 3.925014645577036,
      "grad_norm": 1.4723790884017944,
      "learning_rate": 0.00011984116651477672,
      "loss": 0.0768,
      "step": 6700
    },
    {
      "epoch": 3.93087287639133,
      "grad_norm": 0.9130443334579468,
      "learning_rate": 0.00011919020960812395,
      "loss": 0.0828,
      "step": 6710
    },
    {
      "epoch": 3.936731107205624,
      "grad_norm": 1.0067200660705566,
      "learning_rate": 0.00011853925270147116,
      "loss": 0.0703,
      "step": 6720
    },
    {
      "epoch": 3.9425893380199177,
      "grad_norm": 0.931067168712616,
      "learning_rate": 0.00011788829579481839,
      "loss": 0.145,
      "step": 6730
    },
    {
      "epoch": 3.948447568834212,
      "grad_norm": 1.0656850337982178,
      "learning_rate": 0.0001172373388881656,
      "loss": 0.0963,
      "step": 6740
    },
    {
      "epoch": 3.9543057996485063,
      "grad_norm": 1.115067958831787,
      "learning_rate": 0.00011658638198151282,
      "loss": 0.1641,
      "step": 6750
    },
    {
      "epoch": 3.9601640304628,
      "grad_norm": 0.4096965491771698,
      "learning_rate": 0.00011593542507486004,
      "loss": 0.0915,
      "step": 6760
    },
    {
      "epoch": 3.9660222612770943,
      "grad_norm": 0.781139612197876,
      "learning_rate": 0.00011528446816820728,
      "loss": 0.1007,
      "step": 6770
    },
    {
      "epoch": 3.9718804920913886,
      "grad_norm": 0.6638227701187134,
      "learning_rate": 0.00011463351126155448,
      "loss": 0.0853,
      "step": 6780
    },
    {
      "epoch": 3.9777387229056824,
      "grad_norm": 2.11378812789917,
      "learning_rate": 0.00011398255435490171,
      "loss": 0.0866,
      "step": 6790
    },
    {
      "epoch": 3.9835969537199767,
      "grad_norm": 0.9695467948913574,
      "learning_rate": 0.00011333159744824893,
      "loss": 0.0919,
      "step": 6800
    },
    {
      "epoch": 3.9894551845342705,
      "grad_norm": 1.504145622253418,
      "learning_rate": 0.00011268064054159615,
      "loss": 0.1168,
      "step": 6810
    },
    {
      "epoch": 3.9953134153485648,
      "grad_norm": 2.174940586090088,
      "learning_rate": 0.00011202968363494337,
      "loss": 0.17,
      "step": 6820
    },
    {
      "epoch": 4.0,
      "eval_f1": 0.9495159059474412,
      "eval_loss": 0.14605550467967987,
      "eval_runtime": 15.2718,
      "eval_samples_per_second": 196.44,
      "eval_steps_per_second": 6.155,
      "step": 6828
    },
    {
      "epoch": 4.001171646162859,
      "grad_norm": 0.45618754625320435,
      "learning_rate": 0.0001113787267282906,
      "loss": 0.0885,
      "step": 6830
    },
    {
      "epoch": 4.007029876977153,
      "grad_norm": 0.5365521907806396,
      "learning_rate": 0.0001107277698216378,
      "loss": 0.1259,
      "step": 6840
    },
    {
      "epoch": 4.012888107791447,
      "grad_norm": 0.4301340878009796,
      "learning_rate": 0.00011007681291498504,
      "loss": 0.0706,
      "step": 6850
    },
    {
      "epoch": 4.018746338605741,
      "grad_norm": 0.29090505838394165,
      "learning_rate": 0.00010942585600833225,
      "loss": 0.0633,
      "step": 6860
    },
    {
      "epoch": 4.024604569420035,
      "grad_norm": 0.9259321093559265,
      "learning_rate": 0.00010877489910167947,
      "loss": 0.0879,
      "step": 6870
    },
    {
      "epoch": 4.0304628002343295,
      "grad_norm": 0.9292414784431458,
      "learning_rate": 0.00010812394219502669,
      "loss": 0.0803,
      "step": 6880
    },
    {
      "epoch": 4.036321031048623,
      "grad_norm": 0.9546783566474915,
      "learning_rate": 0.00010747298528837392,
      "loss": 0.0461,
      "step": 6890
    },
    {
      "epoch": 4.042179261862917,
      "grad_norm": 1.73061203956604,
      "learning_rate": 0.00010682202838172113,
      "loss": 0.0864,
      "step": 6900
    },
    {
      "epoch": 4.048037492677212,
      "grad_norm": 1.5470154285430908,
      "learning_rate": 0.00010617107147506836,
      "loss": 0.0595,
      "step": 6910
    },
    {
      "epoch": 4.053895723491506,
      "grad_norm": 0.4693801999092102,
      "learning_rate": 0.00010552011456841558,
      "loss": 0.122,
      "step": 6920
    },
    {
      "epoch": 4.059753954305799,
      "grad_norm": 0.8462125658988953,
      "learning_rate": 0.00010486915766176278,
      "loss": 0.1058,
      "step": 6930
    },
    {
      "epoch": 4.065612185120094,
      "grad_norm": 0.8536199927330017,
      "learning_rate": 0.00010421820075511002,
      "loss": 0.0799,
      "step": 6940
    },
    {
      "epoch": 4.071470415934388,
      "grad_norm": 0.09983082860708237,
      "learning_rate": 0.00010356724384845723,
      "loss": 0.0425,
      "step": 6950
    },
    {
      "epoch": 4.077328646748682,
      "grad_norm": 2.1790366172790527,
      "learning_rate": 0.00010291628694180445,
      "loss": 0.0846,
      "step": 6960
    },
    {
      "epoch": 4.083186877562976,
      "grad_norm": 1.6550222635269165,
      "learning_rate": 0.00010226533003515167,
      "loss": 0.0812,
      "step": 6970
    },
    {
      "epoch": 4.08904510837727,
      "grad_norm": 1.3716167211532593,
      "learning_rate": 0.0001016143731284989,
      "loss": 0.0949,
      "step": 6980
    },
    {
      "epoch": 4.094903339191564,
      "grad_norm": 0.5519310832023621,
      "learning_rate": 0.00010096341622184611,
      "loss": 0.0448,
      "step": 6990
    },
    {
      "epoch": 4.100761570005858,
      "grad_norm": 0.1261308640241623,
      "learning_rate": 0.00010031245931519334,
      "loss": 0.0681,
      "step": 7000
    },
    {
      "epoch": 4.106619800820153,
      "grad_norm": 3.2982888221740723,
      "learning_rate": 9.966150240854056e-05,
      "loss": 0.0542,
      "step": 7010
    },
    {
      "epoch": 4.1124780316344465,
      "grad_norm": 1.6753637790679932,
      "learning_rate": 9.901054550188778e-05,
      "loss": 0.0618,
      "step": 7020
    },
    {
      "epoch": 4.11833626244874,
      "grad_norm": 1.3481342792510986,
      "learning_rate": 9.8359588595235e-05,
      "loss": 0.0991,
      "step": 7030
    },
    {
      "epoch": 4.124194493263035,
      "grad_norm": 1.577588677406311,
      "learning_rate": 9.770863168858221e-05,
      "loss": 0.0559,
      "step": 7040
    },
    {
      "epoch": 4.130052724077329,
      "grad_norm": 0.1360217034816742,
      "learning_rate": 9.705767478192943e-05,
      "loss": 0.0527,
      "step": 7050
    },
    {
      "epoch": 4.135910954891623,
      "grad_norm": 0.4467345178127289,
      "learning_rate": 9.640671787527666e-05,
      "loss": 0.0903,
      "step": 7060
    },
    {
      "epoch": 4.1417691857059165,
      "grad_norm": 0.2530747354030609,
      "learning_rate": 9.575576096862387e-05,
      "loss": 0.0697,
      "step": 7070
    },
    {
      "epoch": 4.147627416520211,
      "grad_norm": 0.8840377926826477,
      "learning_rate": 9.51048040619711e-05,
      "loss": 0.0959,
      "step": 7080
    },
    {
      "epoch": 4.153485647334505,
      "grad_norm": 1.2627636194229126,
      "learning_rate": 9.445384715531832e-05,
      "loss": 0.0849,
      "step": 7090
    },
    {
      "epoch": 4.159343878148799,
      "grad_norm": 2.104389190673828,
      "learning_rate": 9.380289024866554e-05,
      "loss": 0.0869,
      "step": 7100
    },
    {
      "epoch": 4.1652021089630935,
      "grad_norm": 0.8246046304702759,
      "learning_rate": 9.315193334201276e-05,
      "loss": 0.0991,
      "step": 7110
    },
    {
      "epoch": 4.171060339777387,
      "grad_norm": 1.2681169509887695,
      "learning_rate": 9.250097643535999e-05,
      "loss": 0.127,
      "step": 7120
    },
    {
      "epoch": 4.176918570591681,
      "grad_norm": 0.6696609854698181,
      "learning_rate": 9.18500195287072e-05,
      "loss": 0.0894,
      "step": 7130
    },
    {
      "epoch": 4.182776801405975,
      "grad_norm": 1.066689133644104,
      "learning_rate": 9.119906262205443e-05,
      "loss": 0.0874,
      "step": 7140
    },
    {
      "epoch": 4.18863503222027,
      "grad_norm": 0.2604907155036926,
      "learning_rate": 9.054810571540164e-05,
      "loss": 0.0812,
      "step": 7150
    },
    {
      "epoch": 4.1944932630345635,
      "grad_norm": 0.5059824585914612,
      "learning_rate": 8.989714880874886e-05,
      "loss": 0.0683,
      "step": 7160
    },
    {
      "epoch": 4.200351493848857,
      "grad_norm": 0.43476417660713196,
      "learning_rate": 8.924619190209608e-05,
      "loss": 0.0648,
      "step": 7170
    },
    {
      "epoch": 4.206209724663152,
      "grad_norm": 0.4302019476890564,
      "learning_rate": 8.859523499544331e-05,
      "loss": 0.0873,
      "step": 7180
    },
    {
      "epoch": 4.212067955477446,
      "grad_norm": 0.7343120574951172,
      "learning_rate": 8.794427808879052e-05,
      "loss": 0.0909,
      "step": 7190
    },
    {
      "epoch": 4.21792618629174,
      "grad_norm": 1.6828593015670776,
      "learning_rate": 8.729332118213775e-05,
      "loss": 0.0659,
      "step": 7200
    },
    {
      "epoch": 4.223784417106034,
      "grad_norm": 1.096401333808899,
      "learning_rate": 8.664236427548497e-05,
      "loss": 0.0789,
      "step": 7210
    },
    {
      "epoch": 4.229642647920328,
      "grad_norm": 0.8949251770973206,
      "learning_rate": 8.599140736883219e-05,
      "loss": 0.0659,
      "step": 7220
    },
    {
      "epoch": 4.235500878734622,
      "grad_norm": 2.401662588119507,
      "learning_rate": 8.53404504621794e-05,
      "loss": 0.0947,
      "step": 7230
    },
    {
      "epoch": 4.241359109548916,
      "grad_norm": 0.17847134172916412,
      "learning_rate": 8.468949355552664e-05,
      "loss": 0.0576,
      "step": 7240
    },
    {
      "epoch": 4.2472173403632105,
      "grad_norm": 1.1923648118972778,
      "learning_rate": 8.403853664887384e-05,
      "loss": 0.0913,
      "step": 7250
    },
    {
      "epoch": 4.253075571177504,
      "grad_norm": 1.0581272840499878,
      "learning_rate": 8.338757974222107e-05,
      "loss": 0.0991,
      "step": 7260
    },
    {
      "epoch": 4.258933801991798,
      "grad_norm": 0.5614556670188904,
      "learning_rate": 8.273662283556829e-05,
      "loss": 0.0954,
      "step": 7270
    },
    {
      "epoch": 4.264792032806093,
      "grad_norm": 0.4855095148086548,
      "learning_rate": 8.208566592891551e-05,
      "loss": 0.1072,
      "step": 7280
    },
    {
      "epoch": 4.270650263620387,
      "grad_norm": 0.4003661274909973,
      "learning_rate": 8.143470902226273e-05,
      "loss": 0.0616,
      "step": 7290
    },
    {
      "epoch": 4.2765084944346805,
      "grad_norm": 1.0927107334136963,
      "learning_rate": 8.078375211560996e-05,
      "loss": 0.03,
      "step": 7300
    },
    {
      "epoch": 4.282366725248975,
      "grad_norm": 1.1179447174072266,
      "learning_rate": 8.013279520895717e-05,
      "loss": 0.1378,
      "step": 7310
    },
    {
      "epoch": 4.288224956063269,
      "grad_norm": 0.48047196865081787,
      "learning_rate": 7.948183830230438e-05,
      "loss": 0.074,
      "step": 7320
    },
    {
      "epoch": 4.294083186877563,
      "grad_norm": 0.5955802202224731,
      "learning_rate": 7.883088139565162e-05,
      "loss": 0.0636,
      "step": 7330
    },
    {
      "epoch": 4.299941417691857,
      "grad_norm": 0.6766892075538635,
      "learning_rate": 7.817992448899882e-05,
      "loss": 0.0752,
      "step": 7340
    },
    {
      "epoch": 4.305799648506151,
      "grad_norm": 0.9996503591537476,
      "learning_rate": 7.752896758234605e-05,
      "loss": 0.047,
      "step": 7350
    },
    {
      "epoch": 4.311657879320445,
      "grad_norm": 1.4891853332519531,
      "learning_rate": 7.687801067569327e-05,
      "loss": 0.0873,
      "step": 7360
    },
    {
      "epoch": 4.317516110134739,
      "grad_norm": 0.449779212474823,
      "learning_rate": 7.622705376904049e-05,
      "loss": 0.0792,
      "step": 7370
    },
    {
      "epoch": 4.323374340949034,
      "grad_norm": 1.9570748805999756,
      "learning_rate": 7.557609686238771e-05,
      "loss": 0.0836,
      "step": 7380
    },
    {
      "epoch": 4.3292325717633275,
      "grad_norm": 0.9605783224105835,
      "learning_rate": 7.492513995573493e-05,
      "loss": 0.1052,
      "step": 7390
    },
    {
      "epoch": 4.335090802577621,
      "grad_norm": 2.792238712310791,
      "learning_rate": 7.427418304908215e-05,
      "loss": 0.0913,
      "step": 7400
    },
    {
      "epoch": 4.340949033391915,
      "grad_norm": 0.2096194326877594,
      "learning_rate": 7.362322614242938e-05,
      "loss": 0.063,
      "step": 7410
    },
    {
      "epoch": 4.34680726420621,
      "grad_norm": 0.6416425108909607,
      "learning_rate": 7.297226923577658e-05,
      "loss": 0.0934,
      "step": 7420
    },
    {
      "epoch": 4.352665495020504,
      "grad_norm": 0.5825894474983215,
      "learning_rate": 7.232131232912381e-05,
      "loss": 0.086,
      "step": 7430
    },
    {
      "epoch": 4.3585237258347975,
      "grad_norm": 1.1459928750991821,
      "learning_rate": 7.167035542247103e-05,
      "loss": 0.122,
      "step": 7440
    },
    {
      "epoch": 4.364381956649092,
      "grad_norm": 1.567478895187378,
      "learning_rate": 7.101939851581825e-05,
      "loss": 0.0627,
      "step": 7450
    },
    {
      "epoch": 4.370240187463386,
      "grad_norm": 1.6177845001220703,
      "learning_rate": 7.036844160916547e-05,
      "loss": 0.0993,
      "step": 7460
    },
    {
      "epoch": 4.37609841827768,
      "grad_norm": 2.1598875522613525,
      "learning_rate": 6.97174847025127e-05,
      "loss": 0.0849,
      "step": 7470
    },
    {
      "epoch": 4.381956649091975,
      "grad_norm": 0.6902780532836914,
      "learning_rate": 6.90665277958599e-05,
      "loss": 0.111,
      "step": 7480
    },
    {
      "epoch": 4.387814879906268,
      "grad_norm": 2.869980573654175,
      "learning_rate": 6.841557088920714e-05,
      "loss": 0.141,
      "step": 7490
    },
    {
      "epoch": 4.393673110720562,
      "grad_norm": 0.31494468450546265,
      "learning_rate": 6.776461398255436e-05,
      "loss": 0.0818,
      "step": 7500
    },
    {
      "epoch": 4.399531341534857,
      "grad_norm": 0.4729668200016022,
      "learning_rate": 6.711365707590158e-05,
      "loss": 0.0808,
      "step": 7510
    },
    {
      "epoch": 4.405389572349151,
      "grad_norm": 1.1855378150939941,
      "learning_rate": 6.64627001692488e-05,
      "loss": 0.0634,
      "step": 7520
    },
    {
      "epoch": 4.411247803163445,
      "grad_norm": 1.2448687553405762,
      "learning_rate": 6.581174326259603e-05,
      "loss": 0.0801,
      "step": 7530
    },
    {
      "epoch": 4.417106033977738,
      "grad_norm": 2.6101768016815186,
      "learning_rate": 6.516078635594323e-05,
      "loss": 0.0942,
      "step": 7540
    },
    {
      "epoch": 4.422964264792033,
      "grad_norm": 0.29874828457832336,
      "learning_rate": 6.450982944929046e-05,
      "loss": 0.0317,
      "step": 7550
    },
    {
      "epoch": 4.428822495606327,
      "grad_norm": 0.7493123412132263,
      "learning_rate": 6.385887254263768e-05,
      "loss": 0.1051,
      "step": 7560
    },
    {
      "epoch": 4.434680726420621,
      "grad_norm": 2.9272103309631348,
      "learning_rate": 6.32079156359849e-05,
      "loss": 0.0689,
      "step": 7570
    },
    {
      "epoch": 4.440538957234915,
      "grad_norm": 0.1127224862575531,
      "learning_rate": 6.255695872933212e-05,
      "loss": 0.0628,
      "step": 7580
    },
    {
      "epoch": 4.446397188049209,
      "grad_norm": 0.74460768699646,
      "learning_rate": 6.190600182267934e-05,
      "loss": 0.0694,
      "step": 7590
    },
    {
      "epoch": 4.452255418863503,
      "grad_norm": 1.3040841817855835,
      "learning_rate": 6.125504491602657e-05,
      "loss": 0.0642,
      "step": 7600
    },
    {
      "epoch": 4.458113649677797,
      "grad_norm": 0.5990577936172485,
      "learning_rate": 6.060408800937378e-05,
      "loss": 0.0823,
      "step": 7610
    },
    {
      "epoch": 4.463971880492092,
      "grad_norm": 1.0939172506332397,
      "learning_rate": 5.9953131102721e-05,
      "loss": 0.0664,
      "step": 7620
    },
    {
      "epoch": 4.469830111306385,
      "grad_norm": 0.4238429367542267,
      "learning_rate": 5.9302174196068224e-05,
      "loss": 0.0908,
      "step": 7630
    },
    {
      "epoch": 4.475688342120679,
      "grad_norm": 0.3552704155445099,
      "learning_rate": 5.865121728941544e-05,
      "loss": 0.0996,
      "step": 7640
    },
    {
      "epoch": 4.481546572934974,
      "grad_norm": 0.44584277272224426,
      "learning_rate": 5.800026038276266e-05,
      "loss": 0.0522,
      "step": 7650
    },
    {
      "epoch": 4.487404803749268,
      "grad_norm": 1.605562448501587,
      "learning_rate": 5.7349303476109886e-05,
      "loss": 0.0935,
      "step": 7660
    },
    {
      "epoch": 4.493263034563562,
      "grad_norm": 1.6356998682022095,
      "learning_rate": 5.6698346569457104e-05,
      "loss": 0.0821,
      "step": 7670
    },
    {
      "epoch": 4.499121265377856,
      "grad_norm": 1.8082013130187988,
      "learning_rate": 5.604738966280432e-05,
      "loss": 0.0748,
      "step": 7680
    },
    {
      "epoch": 4.50497949619215,
      "grad_norm": 2.335904359817505,
      "learning_rate": 5.539643275615154e-05,
      "loss": 0.1107,
      "step": 7690
    },
    {
      "epoch": 4.510837727006444,
      "grad_norm": 2.8229901790618896,
      "learning_rate": 5.4745475849498766e-05,
      "loss": 0.0908,
      "step": 7700
    },
    {
      "epoch": 4.516695957820739,
      "grad_norm": 0.3684004247188568,
      "learning_rate": 5.4094518942845985e-05,
      "loss": 0.0578,
      "step": 7710
    },
    {
      "epoch": 4.5225541886350324,
      "grad_norm": 0.6283183693885803,
      "learning_rate": 5.34435620361932e-05,
      "loss": 0.0323,
      "step": 7720
    },
    {
      "epoch": 4.528412419449326,
      "grad_norm": 2.3907487392425537,
      "learning_rate": 5.279260512954043e-05,
      "loss": 0.09,
      "step": 7730
    },
    {
      "epoch": 4.53427065026362,
      "grad_norm": 1.1042436361312866,
      "learning_rate": 5.214164822288765e-05,
      "loss": 0.0775,
      "step": 7740
    },
    {
      "epoch": 4.540128881077915,
      "grad_norm": 1.1733710765838623,
      "learning_rate": 5.1490691316234865e-05,
      "loss": 0.0773,
      "step": 7750
    },
    {
      "epoch": 4.545987111892209,
      "grad_norm": 1.506988525390625,
      "learning_rate": 5.083973440958209e-05,
      "loss": 0.0922,
      "step": 7760
    },
    {
      "epoch": 4.551845342706502,
      "grad_norm": 0.33986783027648926,
      "learning_rate": 5.018877750292931e-05,
      "loss": 0.1085,
      "step": 7770
    },
    {
      "epoch": 4.557703573520797,
      "grad_norm": 2.0390729904174805,
      "learning_rate": 4.953782059627653e-05,
      "loss": 0.0657,
      "step": 7780
    },
    {
      "epoch": 4.563561804335091,
      "grad_norm": 0.731065571308136,
      "learning_rate": 4.888686368962375e-05,
      "loss": 0.0725,
      "step": 7790
    },
    {
      "epoch": 4.569420035149385,
      "grad_norm": 1.5336552858352661,
      "learning_rate": 4.823590678297097e-05,
      "loss": 0.1152,
      "step": 7800
    },
    {
      "epoch": 4.575278265963679,
      "grad_norm": 1.6048200130462646,
      "learning_rate": 4.758494987631819e-05,
      "loss": 0.0727,
      "step": 7810
    },
    {
      "epoch": 4.581136496777973,
      "grad_norm": 1.8822131156921387,
      "learning_rate": 4.6933992969665415e-05,
      "loss": 0.0926,
      "step": 7820
    },
    {
      "epoch": 4.586994727592267,
      "grad_norm": 1.853079915046692,
      "learning_rate": 4.628303606301263e-05,
      "loss": 0.0745,
      "step": 7830
    },
    {
      "epoch": 4.592852958406561,
      "grad_norm": 0.5452135801315308,
      "learning_rate": 4.563207915635985e-05,
      "loss": 0.0898,
      "step": 7840
    },
    {
      "epoch": 4.598711189220856,
      "grad_norm": 0.16546402871608734,
      "learning_rate": 4.498112224970708e-05,
      "loss": 0.0756,
      "step": 7850
    },
    {
      "epoch": 4.6045694200351495,
      "grad_norm": 1.2575308084487915,
      "learning_rate": 4.4330165343054295e-05,
      "loss": 0.0458,
      "step": 7860
    },
    {
      "epoch": 4.610427650849443,
      "grad_norm": 2.0601093769073486,
      "learning_rate": 4.367920843640151e-05,
      "loss": 0.047,
      "step": 7870
    },
    {
      "epoch": 4.616285881663737,
      "grad_norm": 1.2107033729553223,
      "learning_rate": 4.3028251529748725e-05,
      "loss": 0.0702,
      "step": 7880
    },
    {
      "epoch": 4.622144112478032,
      "grad_norm": 1.5647666454315186,
      "learning_rate": 4.237729462309595e-05,
      "loss": 0.0559,
      "step": 7890
    },
    {
      "epoch": 4.628002343292326,
      "grad_norm": 0.9503174424171448,
      "learning_rate": 4.172633771644317e-05,
      "loss": 0.0579,
      "step": 7900
    },
    {
      "epoch": 4.6338605741066194,
      "grad_norm": 1.5842559337615967,
      "learning_rate": 4.107538080979039e-05,
      "loss": 0.0941,
      "step": 7910
    },
    {
      "epoch": 4.639718804920914,
      "grad_norm": 0.5423364639282227,
      "learning_rate": 4.042442390313761e-05,
      "loss": 0.0358,
      "step": 7920
    },
    {
      "epoch": 4.645577035735208,
      "grad_norm": 0.35448887944221497,
      "learning_rate": 3.977346699648483e-05,
      "loss": 0.0342,
      "step": 7930
    },
    {
      "epoch": 4.651435266549502,
      "grad_norm": 0.5956054925918579,
      "learning_rate": 3.912251008983205e-05,
      "loss": 0.0918,
      "step": 7940
    },
    {
      "epoch": 4.6572934973637965,
      "grad_norm": 0.8710196018218994,
      "learning_rate": 3.8471553183179275e-05,
      "loss": 0.0689,
      "step": 7950
    },
    {
      "epoch": 4.66315172817809,
      "grad_norm": 1.2711126804351807,
      "learning_rate": 3.782059627652649e-05,
      "loss": 0.0706,
      "step": 7960
    },
    {
      "epoch": 4.669009958992384,
      "grad_norm": 1.3305224180221558,
      "learning_rate": 3.716963936987371e-05,
      "loss": 0.0902,
      "step": 7970
    },
    {
      "epoch": 4.674868189806679,
      "grad_norm": 2.224329948425293,
      "learning_rate": 3.651868246322094e-05,
      "loss": 0.084,
      "step": 7980
    },
    {
      "epoch": 4.680726420620973,
      "grad_norm": 2.547774076461792,
      "learning_rate": 3.5867725556568155e-05,
      "loss": 0.0492,
      "step": 7990
    },
    {
      "epoch": 4.6865846514352665,
      "grad_norm": 0.06184905394911766,
      "learning_rate": 3.5216768649915374e-05,
      "loss": 0.0577,
      "step": 8000
    },
    {
      "epoch": 4.69244288224956,
      "grad_norm": 1.6255323886871338,
      "learning_rate": 3.45658117432626e-05,
      "loss": 0.0673,
      "step": 8010
    },
    {
      "epoch": 4.698301113063855,
      "grad_norm": 1.4682468175888062,
      "learning_rate": 3.391485483660982e-05,
      "loss": 0.0767,
      "step": 8020
    },
    {
      "epoch": 4.704159343878149,
      "grad_norm": 2.154526472091675,
      "learning_rate": 3.3263897929957036e-05,
      "loss": 0.0806,
      "step": 8030
    },
    {
      "epoch": 4.710017574692443,
      "grad_norm": 1.4415132999420166,
      "learning_rate": 3.2612941023304254e-05,
      "loss": 0.0657,
      "step": 8040
    },
    {
      "epoch": 4.715875805506737,
      "grad_norm": 2.2767250537872314,
      "learning_rate": 3.196198411665148e-05,
      "loss": 0.0645,
      "step": 8050
    },
    {
      "epoch": 4.721734036321031,
      "grad_norm": 1.7800607681274414,
      "learning_rate": 3.13110272099987e-05,
      "loss": 0.0997,
      "step": 8060
    },
    {
      "epoch": 4.727592267135325,
      "grad_norm": 0.4202064871788025,
      "learning_rate": 3.0660070303345917e-05,
      "loss": 0.0715,
      "step": 8070
    },
    {
      "epoch": 4.733450497949619,
      "grad_norm": 1.201169490814209,
      "learning_rate": 3.0009113396693142e-05,
      "loss": 0.0883,
      "step": 8080
    },
    {
      "epoch": 4.7393087287639135,
      "grad_norm": 1.4814995527267456,
      "learning_rate": 2.935815649004036e-05,
      "loss": 0.0464,
      "step": 8090
    },
    {
      "epoch": 4.745166959578207,
      "grad_norm": 1.5350182056427002,
      "learning_rate": 2.8707199583387582e-05,
      "loss": 0.0605,
      "step": 8100
    },
    {
      "epoch": 4.751025190392501,
      "grad_norm": 0.17659243941307068,
      "learning_rate": 2.80562426767348e-05,
      "loss": 0.0636,
      "step": 8110
    },
    {
      "epoch": 4.756883421206796,
      "grad_norm": 0.8797257542610168,
      "learning_rate": 2.7405285770082022e-05,
      "loss": 0.043,
      "step": 8120
    },
    {
      "epoch": 4.76274165202109,
      "grad_norm": 1.0583323240280151,
      "learning_rate": 2.6754328863429244e-05,
      "loss": 0.0305,
      "step": 8130
    },
    {
      "epoch": 4.7685998828353835,
      "grad_norm": 1.5315287113189697,
      "learning_rate": 2.6103371956776463e-05,
      "loss": 0.0446,
      "step": 8140
    },
    {
      "epoch": 4.774458113649677,
      "grad_norm": 0.2287391722202301,
      "learning_rate": 2.545241505012368e-05,
      "loss": 0.0947,
      "step": 8150
    },
    {
      "epoch": 4.780316344463972,
      "grad_norm": 0.09398193657398224,
      "learning_rate": 2.4801458143470903e-05,
      "loss": 0.0783,
      "step": 8160
    },
    {
      "epoch": 4.786174575278266,
      "grad_norm": 1.1325687170028687,
      "learning_rate": 2.415050123681812e-05,
      "loss": 0.081,
      "step": 8170
    },
    {
      "epoch": 4.79203280609256,
      "grad_norm": 0.5710904598236084,
      "learning_rate": 2.3499544330165343e-05,
      "loss": 0.0852,
      "step": 8180
    },
    {
      "epoch": 4.797891036906854,
      "grad_norm": 0.3212810754776001,
      "learning_rate": 2.284858742351256e-05,
      "loss": 0.0315,
      "step": 8190
    },
    {
      "epoch": 4.803749267721148,
      "grad_norm": 0.5738294124603271,
      "learning_rate": 2.2197630516859783e-05,
      "loss": 0.062,
      "step": 8200
    },
    {
      "epoch": 4.809607498535442,
      "grad_norm": 3.216890811920166,
      "learning_rate": 2.1546673610207005e-05,
      "loss": 0.0752,
      "step": 8210
    },
    {
      "epoch": 4.815465729349737,
      "grad_norm": 0.953929603099823,
      "learning_rate": 2.0895716703554224e-05,
      "loss": 0.0529,
      "step": 8220
    },
    {
      "epoch": 4.8213239601640305,
      "grad_norm": 0.4356798827648163,
      "learning_rate": 2.0244759796901445e-05,
      "loss": 0.1011,
      "step": 8230
    },
    {
      "epoch": 4.827182190978324,
      "grad_norm": 1.009763479232788,
      "learning_rate": 1.9593802890248667e-05,
      "loss": 0.0795,
      "step": 8240
    },
    {
      "epoch": 4.833040421792619,
      "grad_norm": 1.7814884185791016,
      "learning_rate": 1.8942845983595886e-05,
      "loss": 0.0712,
      "step": 8250
    },
    {
      "epoch": 4.838898652606913,
      "grad_norm": 0.352679580450058,
      "learning_rate": 1.8291889076943108e-05,
      "loss": 0.0627,
      "step": 8260
    },
    {
      "epoch": 4.844756883421207,
      "grad_norm": 2.5139098167419434,
      "learning_rate": 1.764093217029033e-05,
      "loss": 0.0543,
      "step": 8270
    },
    {
      "epoch": 4.8506151142355005,
      "grad_norm": 1.804341435432434,
      "learning_rate": 1.6989975263637548e-05,
      "loss": 0.0854,
      "step": 8280
    },
    {
      "epoch": 4.856473345049795,
      "grad_norm": 0.4135284721851349,
      "learning_rate": 1.633901835698477e-05,
      "loss": 0.0646,
      "step": 8290
    },
    {
      "epoch": 4.862331575864089,
      "grad_norm": 2.144594192504883,
      "learning_rate": 1.5688061450331988e-05,
      "loss": 0.0888,
      "step": 8300
    },
    {
      "epoch": 4.868189806678383,
      "grad_norm": 0.48692771792411804,
      "learning_rate": 1.5037104543679208e-05,
      "loss": 0.0783,
      "step": 8310
    },
    {
      "epoch": 4.874048037492678,
      "grad_norm": 0.4761165976524353,
      "learning_rate": 1.4386147637026428e-05,
      "loss": 0.0787,
      "step": 8320
    },
    {
      "epoch": 4.879906268306971,
      "grad_norm": 2.453779697418213,
      "learning_rate": 1.373519073037365e-05,
      "loss": 0.085,
      "step": 8330
    },
    {
      "epoch": 4.885764499121265,
      "grad_norm": 0.2788262665271759,
      "learning_rate": 1.308423382372087e-05,
      "loss": 0.0797,
      "step": 8340
    },
    {
      "epoch": 4.891622729935559,
      "grad_norm": 1.32106614112854,
      "learning_rate": 1.243327691706809e-05,
      "loss": 0.0453,
      "step": 8350
    },
    {
      "epoch": 4.897480960749854,
      "grad_norm": 0.27709731459617615,
      "learning_rate": 1.178232001041531e-05,
      "loss": 0.0665,
      "step": 8360
    },
    {
      "epoch": 4.9033391915641475,
      "grad_norm": 0.5980609655380249,
      "learning_rate": 1.1131363103762532e-05,
      "loss": 0.0666,
      "step": 8370
    },
    {
      "epoch": 4.909197422378441,
      "grad_norm": 0.24671287834644318,
      "learning_rate": 1.0480406197109753e-05,
      "loss": 0.052,
      "step": 8380
    },
    {
      "epoch": 4.915055653192736,
      "grad_norm": 0.6081691980361938,
      "learning_rate": 9.829449290456971e-06,
      "loss": 0.0812,
      "step": 8390
    },
    {
      "epoch": 4.92091388400703,
      "grad_norm": 0.37573495507240295,
      "learning_rate": 9.178492383804191e-06,
      "loss": 0.0421,
      "step": 8400
    },
    {
      "epoch": 4.926772114821324,
      "grad_norm": 1.504637598991394,
      "learning_rate": 8.527535477151413e-06,
      "loss": 0.0747,
      "step": 8410
    },
    {
      "epoch": 4.932630345635618,
      "grad_norm": 0.8133247494697571,
      "learning_rate": 7.876578570498633e-06,
      "loss": 0.1044,
      "step": 8420
    },
    {
      "epoch": 4.938488576449912,
      "grad_norm": 1.2968809604644775,
      "learning_rate": 7.225621663845853e-06,
      "loss": 0.0999,
      "step": 8430
    },
    {
      "epoch": 4.944346807264206,
      "grad_norm": 1.6766571998596191,
      "learning_rate": 6.574664757193074e-06,
      "loss": 0.0636,
      "step": 8440
    },
    {
      "epoch": 4.950205038078501,
      "grad_norm": 1.0582013130187988,
      "learning_rate": 5.923707850540294e-06,
      "loss": 0.0686,
      "step": 8450
    },
    {
      "epoch": 4.956063268892795,
      "grad_norm": 1.32669198513031,
      "learning_rate": 5.2727509438875145e-06,
      "loss": 0.0632,
      "step": 8460
    },
    {
      "epoch": 4.961921499707088,
      "grad_norm": 0.4580432176589966,
      "learning_rate": 4.621794037234735e-06,
      "loss": 0.0913,
      "step": 8470
    },
    {
      "epoch": 4.967779730521382,
      "grad_norm": 0.21711011230945587,
      "learning_rate": 3.970837130581956e-06,
      "loss": 0.0618,
      "step": 8480
    },
    {
      "epoch": 4.973637961335677,
      "grad_norm": 1.316470742225647,
      "learning_rate": 3.3198802239291758e-06,
      "loss": 0.0528,
      "step": 8490
    },
    {
      "epoch": 4.979496192149971,
      "grad_norm": 0.8495000004768372,
      "learning_rate": 2.6689233172763963e-06,
      "loss": 0.0685,
      "step": 8500
    },
    {
      "epoch": 4.985354422964265,
      "grad_norm": 0.20742426812648773,
      "learning_rate": 2.017966410623617e-06,
      "loss": 0.0892,
      "step": 8510
    },
    {
      "epoch": 4.991212653778559,
      "grad_norm": 0.988993227481842,
      "learning_rate": 1.367009503970837e-06,
      "loss": 0.0613,
      "step": 8520
    },
    {
      "epoch": 4.997070884592853,
      "grad_norm": 1.7386090755462646,
      "learning_rate": 7.160525973180576e-07,
      "loss": 0.0695,
      "step": 8530
    },
    {
      "epoch": 5.0,
      "eval_f1": 0.9536241841291653,
      "eval_loss": 0.12523502111434937,
      "eval_runtime": 14.5928,
      "eval_samples_per_second": 205.581,
      "eval_steps_per_second": 6.442,
      "step": 8535
    },
    {
      "epoch": 5.0,
      "step": 8535,
      "total_flos": 1.4339856401173094e+19,
      "train_loss": 0.13944385742400917,
      "train_runtime": 1629.9678,
      "train_samples_per_second": 167.519,
      "train_steps_per_second": 5.236
    }
  ],
  "logging_steps": 10,
  "max_steps": 8535,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 1.4339856401173094e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}