{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 19.962616822429908,
  "eval_steps": 500,
  "global_step": 2136,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09345794392523364,
      "grad_norm": 4.17994499206543,
      "learning_rate": 1.869158878504673e-05,
      "loss": 0.9424,
      "step": 10
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 2.603377342224121,
      "learning_rate": 3.738317757009346e-05,
      "loss": 0.4804,
      "step": 20
    },
    {
      "epoch": 0.2803738317757009,
      "grad_norm": 1.72787606716156,
      "learning_rate": 5.607476635514019e-05,
      "loss": 0.2668,
      "step": 30
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 0.5973343849182129,
      "learning_rate": 7.476635514018692e-05,
      "loss": 0.1802,
      "step": 40
    },
    {
      "epoch": 0.4672897196261682,
      "grad_norm": 1.0159034729003906,
      "learning_rate": 9.345794392523365e-05,
      "loss": 0.1275,
      "step": 50
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 0.8789228200912476,
      "learning_rate": 0.00011214953271028037,
      "loss": 0.1168,
      "step": 60
    },
    {
      "epoch": 0.6542056074766355,
      "grad_norm": 1.1317589282989502,
      "learning_rate": 0.0001308411214953271,
      "loss": 0.1075,
      "step": 70
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 1.391225814819336,
      "learning_rate": 0.00014953271028037384,
      "loss": 0.1022,
      "step": 80
    },
    {
      "epoch": 0.8411214953271028,
      "grad_norm": 1.0744414329528809,
      "learning_rate": 0.00016822429906542056,
      "loss": 0.0905,
      "step": 90
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 0.6361038684844971,
      "learning_rate": 0.0001869158878504673,
      "loss": 0.0794,
      "step": 100
    },
    {
      "epoch": 1.02803738317757,
      "grad_norm": 0.6731115579605103,
      "learning_rate": 0.00019999892118398673,
      "loss": 0.0823,
      "step": 110
    },
    {
      "epoch": 1.1214953271028036,
      "grad_norm": 0.6741369962692261,
      "learning_rate": 0.0001999797428801708,
      "loss": 0.0762,
      "step": 120
    },
    {
      "epoch": 1.2149532710280373,
      "grad_norm": 0.7907876372337341,
      "learning_rate": 0.00019993659617930244,
      "loss": 0.0669,
      "step": 130
    },
    {
      "epoch": 1.308411214953271,
      "grad_norm": 0.7666932940483093,
      "learning_rate": 0.00019986949142504959,
      "loss": 0.0739,
      "step": 140
    },
    {
      "epoch": 1.4018691588785046,
      "grad_norm": 0.7501648664474487,
      "learning_rate": 0.00019977844470460494,
      "loss": 0.0688,
      "step": 150
    },
    {
      "epoch": 1.4953271028037383,
      "grad_norm": 0.45659929513931274,
      "learning_rate": 0.00019966347784482945,
      "loss": 0.0654,
      "step": 160
    },
    {
      "epoch": 1.588785046728972,
      "grad_norm": 0.6907665729522705,
      "learning_rate": 0.0001995246184070197,
      "loss": 0.0613,
      "step": 170
    },
    {
      "epoch": 1.6822429906542056,
      "grad_norm": 0.47750258445739746,
      "learning_rate": 0.0001993618996803005,
      "loss": 0.0536,
      "step": 180
    },
    {
      "epoch": 1.7757009345794392,
      "grad_norm": 0.7647998332977295,
      "learning_rate": 0.00019917536067364444,
      "loss": 0.0579,
      "step": 190
    },
    {
      "epoch": 1.8691588785046729,
      "grad_norm": 0.5567947030067444,
      "learning_rate": 0.00019896504610652016,
      "loss": 0.0548,
      "step": 200
    },
    {
      "epoch": 1.9626168224299065,
      "grad_norm": 0.46288764476776123,
      "learning_rate": 0.00019873100639817166,
      "loss": 0.0527,
      "step": 210
    },
    {
      "epoch": 2.05607476635514,
      "grad_norm": 0.4530458450317383,
      "learning_rate": 0.00019847329765553115,
      "loss": 0.0465,
      "step": 220
    },
    {
      "epoch": 2.149532710280374,
      "grad_norm": 0.3930870592594147,
      "learning_rate": 0.00019819198165976838,
      "loss": 0.0519,
      "step": 230
    },
    {
      "epoch": 2.2429906542056073,
      "grad_norm": 0.41593921184539795,
      "learning_rate": 0.00019788712585147976,
      "loss": 0.0507,
      "step": 240
    },
    {
      "epoch": 2.336448598130841,
      "grad_norm": 0.47823095321655273,
      "learning_rate": 0.00019755880331452045,
      "loss": 0.043,
      "step": 250
    },
    {
      "epoch": 2.4299065420560746,
      "grad_norm": 0.316738486289978,
      "learning_rate": 0.00019720709275848408,
      "loss": 0.0421,
      "step": 260
    },
    {
      "epoch": 2.5233644859813085,
      "grad_norm": 0.33323222398757935,
      "learning_rate": 0.00019683207849983326,
      "loss": 0.0451,
      "step": 270
    },
    {
      "epoch": 2.616822429906542,
      "grad_norm": 0.43355607986450195,
      "learning_rate": 0.00019643385044168632,
      "loss": 0.0402,
      "step": 280
    },
    {
      "epoch": 2.710280373831776,
      "grad_norm": 0.40903353691101074,
      "learning_rate": 0.0001960125040522645,
      "loss": 0.0433,
      "step": 290
    },
    {
      "epoch": 2.803738317757009,
      "grad_norm": 0.3322107493877411,
      "learning_rate": 0.00019556814034200518,
      "loss": 0.0415,
      "step": 300
    },
    {
      "epoch": 2.897196261682243,
      "grad_norm": 0.3895340859889984,
      "learning_rate": 0.00019510086583934643,
      "loss": 0.0419,
      "step": 310
    },
    {
      "epoch": 2.9906542056074765,
      "grad_norm": 0.39366844296455383,
      "learning_rate": 0.0001946107925651885,
      "loss": 0.0358,
      "step": 320
    },
    {
      "epoch": 3.0841121495327104,
      "grad_norm": 0.3918449282646179,
      "learning_rate": 0.00019409803800603904,
      "loss": 0.041,
      "step": 330
    },
    {
      "epoch": 3.177570093457944,
      "grad_norm": 0.273848295211792,
      "learning_rate": 0.00019356272508584753,
      "loss": 0.0447,
      "step": 340
    },
    {
      "epoch": 3.2710280373831777,
      "grad_norm": 0.4000929296016693,
      "learning_rate": 0.00019300498213653647,
      "loss": 0.0434,
      "step": 350
    },
    {
      "epoch": 3.364485981308411,
      "grad_norm": 0.2600472867488861,
      "learning_rate": 0.00019242494286723612,
      "loss": 0.0429,
      "step": 360
    },
    {
      "epoch": 3.457943925233645,
      "grad_norm": 0.29327982664108276,
      "learning_rate": 0.00019182274633223,
      "loss": 0.0362,
      "step": 370
    },
    {
      "epoch": 3.5514018691588785,
      "grad_norm": 0.35170868039131165,
      "learning_rate": 0.00019119853689761912,
      "loss": 0.0342,
      "step": 380
    },
    {
      "epoch": 3.6448598130841123,
      "grad_norm": 0.46894872188568115,
      "learning_rate": 0.00019055246420671257,
      "loss": 0.0364,
      "step": 390
    },
    {
      "epoch": 3.7383177570093458,
      "grad_norm": 0.4111644923686981,
      "learning_rate": 0.0001898846831441533,
      "loss": 0.0336,
      "step": 400
    },
    {
      "epoch": 3.831775700934579,
      "grad_norm": 0.40484148263931274,
      "learning_rate": 0.0001891953537987873,
      "loss": 0.0406,
      "step": 410
    },
    {
      "epoch": 3.925233644859813,
      "grad_norm": 0.29574188590049744,
      "learning_rate": 0.00018848464142528481,
      "loss": 0.0344,
      "step": 420
    },
    {
      "epoch": 4.018691588785047,
      "grad_norm": 0.30757296085357666,
      "learning_rate": 0.00018775271640452377,
      "loss": 0.0365,
      "step": 430
    },
    {
      "epoch": 4.11214953271028,
      "grad_norm": 0.26074057817459106,
      "learning_rate": 0.00018699975420274358,
      "loss": 0.0356,
      "step": 440
    },
    {
      "epoch": 4.205607476635514,
      "grad_norm": 0.3508523404598236,
      "learning_rate": 0.00018622593532948055,
      "loss": 0.0325,
      "step": 450
    },
    {
      "epoch": 4.299065420560748,
      "grad_norm": 0.2975287437438965,
      "learning_rate": 0.0001854314452942935,
      "loss": 0.0321,
      "step": 460
    },
    {
      "epoch": 4.392523364485982,
      "grad_norm": 0.30913522839546204,
      "learning_rate": 0.0001846164745622914,
      "loss": 0.0329,
      "step": 470
    },
    {
      "epoch": 4.485981308411215,
      "grad_norm": 0.30218392610549927,
      "learning_rate": 0.00018378121850847249,
      "loss": 0.0344,
      "step": 480
    },
    {
      "epoch": 4.579439252336448,
      "grad_norm": 0.4823051691055298,
      "learning_rate": 0.00018292587737088665,
      "loss": 0.0331,
      "step": 490
    },
    {
      "epoch": 4.672897196261682,
      "grad_norm": 0.28159743547439575,
      "learning_rate": 0.00018205065620263168,
      "loss": 0.0374,
      "step": 500
    },
    {
      "epoch": 4.766355140186916,
      "grad_norm": 0.36674001812934875,
      "learning_rate": 0.0001811557648226956,
      "loss": 0.036,
      "step": 510
    },
    {
      "epoch": 4.859813084112149,
      "grad_norm": 0.31007957458496094,
      "learning_rate": 0.00018024141776565606,
      "loss": 0.0321,
      "step": 520
    },
    {
      "epoch": 4.953271028037383,
      "grad_norm": 0.2760983109474182,
      "learning_rate": 0.0001793078342302495,
      "loss": 0.0326,
      "step": 530
    },
    {
      "epoch": 5.046728971962617,
      "grad_norm": 0.23232242465019226,
      "learning_rate": 0.00017835523802682194,
      "loss": 0.032,
      "step": 540
    },
    {
      "epoch": 5.140186915887851,
      "grad_norm": 0.3053811192512512,
      "learning_rate": 0.0001773838575236747,
      "loss": 0.0302,
      "step": 550
    },
    {
      "epoch": 5.233644859813084,
      "grad_norm": 0.2696075737476349,
      "learning_rate": 0.0001763939255923166,
      "loss": 0.0325,
      "step": 560
    },
    {
      "epoch": 5.327102803738318,
      "grad_norm": 0.28349894285202026,
      "learning_rate": 0.00017538567955163755,
      "loss": 0.0358,
      "step": 570
    },
    {
      "epoch": 5.420560747663552,
      "grad_norm": 0.20442534983158112,
      "learning_rate": 0.00017435936111101518,
      "loss": 0.0293,
      "step": 580
    },
    {
      "epoch": 5.5140186915887845,
      "grad_norm": 0.271004855632782,
      "learning_rate": 0.0001733152163123694,
      "loss": 0.0275,
      "step": 590
    },
    {
      "epoch": 5.607476635514018,
      "grad_norm": 0.14870816469192505,
      "learning_rate": 0.00017225349547117798,
      "loss": 0.0315,
      "step": 600
    },
    {
      "epoch": 5.700934579439252,
      "grad_norm": 0.3340059518814087,
      "learning_rate": 0.00017117445311646802,
      "loss": 0.029,
      "step": 610
    },
    {
      "epoch": 5.794392523364486,
      "grad_norm": 0.32217147946357727,
      "learning_rate": 0.00017007834792979695,
      "loss": 0.0294,
      "step": 620
    },
    {
      "epoch": 5.88785046728972,
      "grad_norm": 0.32910802960395813,
      "learning_rate": 0.00016896544268323792,
      "loss": 0.033,
      "step": 630
    },
    {
      "epoch": 5.981308411214953,
      "grad_norm": 0.5195891261100769,
      "learning_rate": 0.00016783600417638535,
      "loss": 0.0334,
      "step": 640
    },
    {
      "epoch": 6.074766355140187,
      "grad_norm": 0.31266453862190247,
      "learning_rate": 0.0001666903031723937,
      "loss": 0.0311,
      "step": 650
    },
    {
      "epoch": 6.168224299065421,
      "grad_norm": 0.21037207543849945,
      "learning_rate": 0.00016552861433306726,
      "loss": 0.0297,
      "step": 660
    },
    {
      "epoch": 6.261682242990654,
      "grad_norm": 0.20664794743061066,
      "learning_rate": 0.00016435121615301457,
      "loss": 0.0286,
      "step": 670
    },
    {
      "epoch": 6.355140186915888,
      "grad_norm": 0.2566857635974884,
      "learning_rate": 0.00016315839089288411,
      "loss": 0.0229,
      "step": 680
    },
    {
      "epoch": 6.4485981308411215,
      "grad_norm": 0.2402873933315277,
      "learning_rate": 0.00016195042451169753,
      "loss": 0.0264,
      "step": 690
    },
    {
      "epoch": 6.542056074766355,
      "grad_norm": 0.36038416624069214,
      "learning_rate": 0.00016072760659829562,
      "loss": 0.0252,
      "step": 700
    },
    {
      "epoch": 6.635514018691588,
      "grad_norm": 0.4320838749408722,
      "learning_rate": 0.00015949023030191475,
      "loss": 0.0293,
      "step": 710
    },
    {
      "epoch": 6.728971962616822,
      "grad_norm": 0.26101937890052795,
      "learning_rate": 0.0001582385922619092,
      "loss": 0.0286,
      "step": 720
    },
    {
      "epoch": 6.822429906542056,
      "grad_norm": 0.32656329870224,
      "learning_rate": 0.00015697299253663723,
      "loss": 0.0294,
      "step": 730
    },
    {
      "epoch": 6.91588785046729,
      "grad_norm": 0.2876569926738739,
      "learning_rate": 0.00015569373453152728,
      "loss": 0.0309,
      "step": 740
    },
    {
      "epoch": 7.009345794392523,
      "grad_norm": 0.2590012848377228,
      "learning_rate": 0.00015440112492634192,
      "loss": 0.0261,
      "step": 750
    },
    {
      "epoch": 7.102803738317757,
      "grad_norm": 0.37834206223487854,
      "learning_rate": 0.00015309547360165647,
      "loss": 0.03,
      "step": 760
    },
    {
      "epoch": 7.196261682242991,
      "grad_norm": 0.27962639927864075,
      "learning_rate": 0.00015177709356457077,
      "loss": 0.0297,
      "step": 770
    },
    {
      "epoch": 7.289719626168225,
      "grad_norm": 0.3405240476131439,
      "learning_rate": 0.00015044630087367118,
      "loss": 0.0314,
      "step": 780
    },
    {
      "epoch": 7.383177570093458,
      "grad_norm": 0.31650567054748535,
      "learning_rate": 0.0001491034145632609,
      "loss": 0.0249,
      "step": 790
    },
    {
      "epoch": 7.4766355140186915,
      "grad_norm": 0.3112686276435852,
      "learning_rate": 0.00014774875656687727,
      "loss": 0.0257,
      "step": 800
    },
    {
      "epoch": 7.570093457943925,
      "grad_norm": 0.23344942927360535,
      "learning_rate": 0.0001463826516401137,
      "loss": 0.0245,
      "step": 810
    },
    {
      "epoch": 7.663551401869158,
      "grad_norm": 0.1889248788356781,
      "learning_rate": 0.00014500542728276518,
      "loss": 0.0229,
      "step": 820
    },
    {
      "epoch": 7.757009345794392,
      "grad_norm": 0.24938005208969116,
      "learning_rate": 0.00014361741366031603,
      "loss": 0.0246,
      "step": 830
    },
    {
      "epoch": 7.850467289719626,
      "grad_norm": 0.22689329087734222,
      "learning_rate": 0.0001422189435247884,
      "loss": 0.0251,
      "step": 840
    },
    {
      "epoch": 7.94392523364486,
      "grad_norm": 0.28100717067718506,
      "learning_rate": 0.00014081035213497082,
      "loss": 0.0277,
      "step": 850
    },
    {
      "epoch": 8.037383177570094,
      "grad_norm": 0.2773335874080658,
      "learning_rate": 0.0001393919771760457,
      "loss": 0.027,
      "step": 860
    },
    {
      "epoch": 8.130841121495328,
      "grad_norm": 0.2691657841205597,
      "learning_rate": 0.00013796415867863528,
      "loss": 0.0239,
      "step": 870
    },
    {
      "epoch": 8.22429906542056,
      "grad_norm": 0.4514601230621338,
      "learning_rate": 0.00013652723893728532,
      "loss": 0.029,
      "step": 880
    },
    {
      "epoch": 8.317757009345794,
      "grad_norm": 0.3693794012069702,
      "learning_rate": 0.00013508156242840566,
      "loss": 0.0251,
      "step": 890
    },
    {
      "epoch": 8.411214953271028,
      "grad_norm": 0.2953411042690277,
      "learning_rate": 0.00013362747572768832,
      "loss": 0.0284,
      "step": 900
    },
    {
      "epoch": 8.504672897196262,
      "grad_norm": 0.25673815608024597,
      "learning_rate": 0.0001321653274270219,
      "loss": 0.0251,
      "step": 910
    },
    {
      "epoch": 8.598130841121495,
      "grad_norm": 0.2374838888645172,
      "learning_rate": 0.00013069546805092276,
      "loss": 0.0211,
      "step": 920
    },
    {
      "epoch": 8.69158878504673,
      "grad_norm": 0.2258038967847824,
      "learning_rate": 0.00012921824997250294,
      "loss": 0.0235,
      "step": 930
    },
    {
      "epoch": 8.785046728971963,
      "grad_norm": 0.35665759444236755,
      "learning_rate": 0.00012773402732899481,
      "loss": 0.0262,
      "step": 940
    },
    {
      "epoch": 8.878504672897197,
      "grad_norm": 0.2746776342391968,
      "learning_rate": 0.00012624315593685282,
      "loss": 0.0228,
      "step": 950
    },
    {
      "epoch": 8.97196261682243,
      "grad_norm": 0.1950758695602417,
      "learning_rate": 0.0001247459932064529,
      "loss": 0.028,
      "step": 960
    },
    {
      "epoch": 9.065420560747663,
      "grad_norm": 0.25761884450912476,
      "learning_rate": 0.00012324289805640934,
      "loss": 0.0256,
      "step": 970
    },
    {
      "epoch": 9.158878504672897,
      "grad_norm": 0.24331067502498627,
      "learning_rate": 0.00012173423082753041,
      "loss": 0.0226,
      "step": 980
    },
    {
      "epoch": 9.25233644859813,
      "grad_norm": 0.2336030751466751,
      "learning_rate": 0.00012022035319643305,
      "loss": 0.024,
      "step": 990
    },
    {
      "epoch": 9.345794392523365,
      "grad_norm": 0.26760831475257874,
      "learning_rate": 0.00011870162808883703,
      "loss": 0.0231,
      "step": 1000
    },
    {
      "epoch": 9.439252336448599,
      "grad_norm": 0.22399088740348816,
      "learning_rate": 0.00011717841959255999,
      "loss": 0.0212,
      "step": 1010
    },
    {
      "epoch": 9.532710280373832,
      "grad_norm": 0.13075126707553864,
      "learning_rate": 0.0001156510928702337,
      "loss": 0.027,
      "step": 1020
    },
    {
      "epoch": 9.626168224299064,
      "grad_norm": 0.19196636974811554,
      "learning_rate": 0.00011412001407176265,
      "loss": 0.0213,
      "step": 1030
    },
    {
      "epoch": 9.719626168224298,
      "grad_norm": 0.24825584888458252,
      "learning_rate": 0.00011258555024654612,
      "loss": 0.02,
      "step": 1040
    },
    {
      "epoch": 9.813084112149532,
      "grad_norm": 0.20732861757278442,
      "learning_rate": 0.00011104806925548435,
      "loss": 0.0209,
      "step": 1050
    },
    {
      "epoch": 9.906542056074766,
      "grad_norm": 0.15755069255828857,
      "learning_rate": 0.00010950793968279045,
      "loss": 0.0198,
      "step": 1060
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.261300265789032,
      "learning_rate": 0.00010796553074762867,
      "loss": 0.0228,
      "step": 1070
    },
    {
      "epoch": 10.093457943925234,
      "grad_norm": 0.20064909756183624,
      "learning_rate": 0.00010642121221560066,
      "loss": 0.0237,
      "step": 1080
    },
    {
      "epoch": 10.186915887850468,
      "grad_norm": 0.2862524092197418,
      "learning_rate": 0.00010487535431010049,
      "loss": 0.0189,
      "step": 1090
    },
    {
      "epoch": 10.280373831775702,
      "grad_norm": 0.2752748727798462,
      "learning_rate": 0.0001033283276235602,
      "loss": 0.0231,
      "step": 1100
    },
    {
      "epoch": 10.373831775700934,
      "grad_norm": 0.1559881865978241,
      "learning_rate": 0.00010178050302860662,
      "loss": 0.017,
      "step": 1110
    },
    {
      "epoch": 10.467289719626168,
      "grad_norm": 0.16235685348510742,
      "learning_rate": 0.00010023225158915119,
      "loss": 0.0214,
      "step": 1120
    },
    {
      "epoch": 10.560747663551401,
      "grad_norm": 0.26877355575561523,
      "learning_rate": 9.868394447143394e-05,
      "loss": 0.0221,
      "step": 1130
    },
    {
      "epoch": 10.654205607476635,
      "grad_norm": 0.26213476061820984,
      "learning_rate": 9.713595285504266e-05,
      "loss": 0.0218,
      "step": 1140
    },
    {
      "epoch": 10.74766355140187,
      "grad_norm": 0.2717665433883667,
      "learning_rate": 9.558864784392932e-05,
      "loss": 0.0203,
      "step": 1150
    },
    {
      "epoch": 10.841121495327103,
      "grad_norm": 0.29337963461875916,
      "learning_rate": 9.404240037744418e-05,
      "loss": 0.0208,
      "step": 1160
    },
    {
      "epoch": 10.934579439252337,
      "grad_norm": 0.18261757493019104,
      "learning_rate": 9.249758114140977e-05,
      "loss": 0.0237,
      "step": 1170
    },
    {
      "epoch": 11.02803738317757,
      "grad_norm": 0.18363697826862335,
      "learning_rate": 9.095456047925517e-05,
      "loss": 0.0202,
      "step": 1180
    },
    {
      "epoch": 11.121495327102803,
      "grad_norm": 0.147624671459198,
      "learning_rate": 8.941370830323286e-05,
      "loss": 0.0177,
      "step": 1190
    },
    {
      "epoch": 11.214953271028037,
      "grad_norm": 0.19984780251979828,
      "learning_rate": 8.787539400573867e-05,
      "loss": 0.0208,
      "step": 1200
    },
    {
      "epoch": 11.30841121495327,
      "grad_norm": 0.2558623254299164,
      "learning_rate": 8.633998637075634e-05,
      "loss": 0.0187,
      "step": 1210
    },
    {
      "epoch": 11.401869158878505,
      "grad_norm": 0.15088629722595215,
      "learning_rate": 8.480785348544802e-05,
      "loss": 0.0191,
      "step": 1220
    },
    {
      "epoch": 11.495327102803738,
      "grad_norm": 0.15472617745399475,
      "learning_rate": 8.32793626519119e-05,
      "loss": 0.0174,
      "step": 1230
    },
    {
      "epoch": 11.588785046728972,
      "grad_norm": 0.1284412145614624,
      "learning_rate": 8.175488029912783e-05,
      "loss": 0.0183,
      "step": 1240
    },
    {
      "epoch": 11.682242990654206,
      "grad_norm": 0.2205333113670349,
      "learning_rate": 8.02347718951124e-05,
      "loss": 0.0179,
      "step": 1250
    },
    {
      "epoch": 11.77570093457944,
      "grad_norm": 0.2073797583580017,
      "learning_rate": 7.871940185930438e-05,
      "loss": 0.023,
      "step": 1260
    },
    {
      "epoch": 11.869158878504672,
      "grad_norm": 0.22515133023262024,
      "learning_rate": 7.720913347520154e-05,
      "loss": 0.0178,
      "step": 1270
    },
    {
      "epoch": 11.962616822429906,
      "grad_norm": 0.18330000340938568,
      "learning_rate": 7.57043288032698e-05,
      "loss": 0.0206,
      "step": 1280
    },
    {
      "epoch": 12.05607476635514,
      "grad_norm": 0.19882416725158691,
      "learning_rate": 7.420534859414542e-05,
      "loss": 0.0186,
      "step": 1290
    },
    {
      "epoch": 12.149532710280374,
      "grad_norm": 0.4009910821914673,
      "learning_rate": 7.271255220215159e-05,
      "loss": 0.0203,
      "step": 1300
    },
    {
      "epoch": 12.242990654205608,
      "grad_norm": 0.21823973953723907,
      "learning_rate": 7.122629749914917e-05,
      "loss": 0.0183,
      "step": 1310
    },
    {
      "epoch": 12.336448598130842,
      "grad_norm": 0.17465481162071228,
      "learning_rate": 6.974694078874344e-05,
      "loss": 0.0164,
      "step": 1320
    },
    {
      "epoch": 12.429906542056075,
      "grad_norm": 0.23976965248584747,
      "learning_rate": 6.827483672086649e-05,
      "loss": 0.0228,
      "step": 1330
    },
    {
      "epoch": 12.523364485981308,
      "grad_norm": 0.17434823513031006,
      "learning_rate": 6.681033820675596e-05,
      "loss": 0.0185,
      "step": 1340
    },
    {
      "epoch": 12.616822429906541,
      "grad_norm": 0.21936717629432678,
      "learning_rate": 6.535379633435082e-05,
      "loss": 0.019,
      "step": 1350
    },
    {
      "epoch": 12.710280373831775,
      "grad_norm": 0.20729029178619385,
      "learning_rate": 6.390556028412421e-05,
      "loss": 0.0187,
      "step": 1360
    },
    {
      "epoch": 12.80373831775701,
      "grad_norm": 0.20753750205039978,
      "learning_rate": 6.246597724537342e-05,
      "loss": 0.0182,
      "step": 1370
    },
    {
      "epoch": 12.897196261682243,
      "grad_norm": 0.17846998572349548,
      "learning_rate": 6.10353923329875e-05,
      "loss": 0.0158,
      "step": 1380
    },
    {
      "epoch": 12.990654205607477,
      "grad_norm": 0.21448595821857452,
      "learning_rate": 5.961414850471172e-05,
      "loss": 0.0159,
      "step": 1390
    },
    {
      "epoch": 13.08411214953271,
      "grad_norm": 0.21095584332942963,
      "learning_rate": 5.8202586478929646e-05,
      "loss": 0.0169,
      "step": 1400
    },
    {
      "epoch": 13.177570093457945,
      "grad_norm": 0.2717122435569763,
      "learning_rate": 5.680104465298186e-05,
      "loss": 0.0163,
      "step": 1410
    },
    {
      "epoch": 13.271028037383177,
      "grad_norm": 0.15730348229408264,
      "learning_rate": 5.540985902204099e-05,
      "loss": 0.0154,
      "step": 1420
    },
    {
      "epoch": 13.36448598130841,
      "grad_norm": 0.2034991979598999,
      "learning_rate": 5.4029363098562794e-05,
      "loss": 0.0181,
      "step": 1430
    },
    {
      "epoch": 13.457943925233645,
      "grad_norm": 0.1877489984035492,
      "learning_rate": 5.2659887832332424e-05,
      "loss": 0.0149,
      "step": 1440
    },
    {
      "epoch": 13.551401869158878,
      "grad_norm": 0.25242817401885986,
      "learning_rate": 5.130176153112485e-05,
      "loss": 0.0162,
      "step": 1450
    },
    {
      "epoch": 13.644859813084112,
      "grad_norm": 0.3759627342224121,
      "learning_rate": 4.9955309781999096e-05,
      "loss": 0.0205,
      "step": 1460
    },
    {
      "epoch": 13.738317757009346,
      "grad_norm": 0.1744040697813034,
      "learning_rate": 4.8620855373244056e-05,
      "loss": 0.0158,
      "step": 1470
    },
    {
      "epoch": 13.83177570093458,
      "grad_norm": 0.2127053588628769,
      "learning_rate": 4.729871821699587e-05,
      "loss": 0.0181,
      "step": 1480
    },
    {
      "epoch": 13.925233644859812,
      "grad_norm": 0.16114765405654907,
      "learning_rate": 4.5989215272544605e-05,
      "loss": 0.0166,
      "step": 1490
    },
    {
      "epoch": 14.018691588785046,
      "grad_norm": 0.16290584206581116,
      "learning_rate": 4.469266047034861e-05,
      "loss": 0.0157,
      "step": 1500
    },
    {
      "epoch": 14.11214953271028,
      "grad_norm": 0.29187917709350586,
      "learning_rate": 4.3409364636775415e-05,
      "loss": 0.0172,
      "step": 1510
    },
    {
      "epoch": 14.205607476635514,
      "grad_norm": 0.2521297335624695,
      "learning_rate": 4.213963541958631e-05,
      "loss": 0.0147,
      "step": 1520
    },
    {
      "epoch": 14.299065420560748,
      "grad_norm": 0.2301267385482788,
      "learning_rate": 4.0883777214183306e-05,
      "loss": 0.0172,
      "step": 1530
    },
    {
      "epoch": 14.392523364485982,
      "grad_norm": 0.20023106038570404,
      "learning_rate": 3.96420910906356e-05,
      "loss": 0.0178,
      "step": 1540
    },
    {
      "epoch": 14.485981308411215,
      "grad_norm": 0.20669516921043396,
      "learning_rate": 3.8414874721503294e-05,
      "loss": 0.016,
      "step": 1550
    },
    {
      "epoch": 14.57943925233645,
      "grad_norm": 0.22793950140476227,
      "learning_rate": 3.720242231047568e-05,
      "loss": 0.0168,
      "step": 1560
    },
    {
      "epoch": 14.672897196261681,
      "grad_norm": 0.21381746232509613,
      "learning_rate": 3.6005024521840916e-05,
      "loss": 0.0156,
      "step": 1570
    },
    {
      "epoch": 14.766355140186915,
      "grad_norm": 0.20497943460941315,
      "learning_rate": 3.482296841080433e-05,
      "loss": 0.0152,
      "step": 1580
    },
    {
      "epoch": 14.85981308411215,
      "grad_norm": 0.2997715473175049,
      "learning_rate": 3.365653735467217e-05,
      "loss": 0.0159,
      "step": 1590
    },
    {
      "epoch": 14.953271028037383,
      "grad_norm": 0.1698296070098877,
      "learning_rate": 3.2506010984916505e-05,
      "loss": 0.0153,
      "step": 1600
    },
    {
      "epoch": 15.046728971962617,
      "grad_norm": 0.13616669178009033,
      "learning_rate": 3.137166512013882e-05,
      "loss": 0.0155,
      "step": 1610
    },
    {
      "epoch": 15.14018691588785,
      "grad_norm": 0.2119993269443512,
      "learning_rate": 3.025377169994702e-05,
      "loss": 0.013,
      "step": 1620
    },
    {
      "epoch": 15.233644859813085,
      "grad_norm": 0.18577918410301208,
      "learning_rate": 2.9152598719762846e-05,
      "loss": 0.0135,
      "step": 1630
    },
    {
      "epoch": 15.327102803738319,
      "grad_norm": 0.16220198571681976,
      "learning_rate": 2.8068410166574566e-05,
      "loss": 0.015,
      "step": 1640
    },
    {
      "epoch": 15.42056074766355,
      "grad_norm": 0.3275705575942993,
      "learning_rate": 2.7001465955650774e-05,
      "loss": 0.0136,
      "step": 1650
    },
    {
      "epoch": 15.514018691588785,
      "grad_norm": 0.14544951915740967,
      "learning_rate": 2.5952021868230135e-05,
      "loss": 0.013,
      "step": 1660
    },
    {
      "epoch": 15.607476635514018,
      "grad_norm": 0.1696663796901703,
      "learning_rate": 2.492032949020249e-05,
      "loss": 0.0135,
      "step": 1670
    },
    {
      "epoch": 15.700934579439252,
      "grad_norm": 0.22859309613704681,
      "learning_rate": 2.390663615179528e-05,
      "loss": 0.0145,
      "step": 1680
    },
    {
      "epoch": 15.794392523364486,
      "grad_norm": 0.16665244102478027,
      "learning_rate": 2.291118486828071e-05,
      "loss": 0.0188,
      "step": 1690
    },
    {
      "epoch": 15.88785046728972,
      "grad_norm": 0.2959059774875641,
      "learning_rate": 2.193421428171685e-05,
      "loss": 0.0176,
      "step": 1700
    },
    {
      "epoch": 15.981308411214954,
      "grad_norm": 0.2710016071796417,
      "learning_rate": 2.0975958603737477e-05,
      "loss": 0.0159,
      "step": 1710
    },
    {
      "epoch": 16.074766355140188,
      "grad_norm": 0.18647173047065735,
      "learning_rate": 2.0036647559403997e-05,
      "loss": 0.0137,
      "step": 1720
    },
    {
      "epoch": 16.16822429906542,
      "grad_norm": 0.17502279579639435,
      "learning_rate": 1.9116506332132734e-05,
      "loss": 0.0149,
      "step": 1730
    },
    {
      "epoch": 16.261682242990656,
      "grad_norm": 0.24166816473007202,
      "learning_rate": 1.8215755509711196e-05,
      "loss": 0.0152,
      "step": 1740
    },
    {
      "epoch": 16.35514018691589,
      "grad_norm": 0.1839790791273117,
      "learning_rate": 1.7334611031416047e-05,
      "loss": 0.0144,
      "step": 1750
    },
    {
      "epoch": 16.44859813084112,
      "grad_norm": 0.20211663842201233,
      "learning_rate": 1.647328413624537e-05,
      "loss": 0.0149,
      "step": 1760
    },
    {
      "epoch": 16.542056074766354,
      "grad_norm": 0.14310133457183838,
      "learning_rate": 1.5631981312277886e-05,
      "loss": 0.0138,
      "step": 1770
    },
    {
      "epoch": 16.635514018691588,
      "grad_norm": 0.1992831528186798,
      "learning_rate": 1.4810904247170953e-05,
      "loss": 0.0145,
      "step": 1780
    },
    {
      "epoch": 16.72897196261682,
      "grad_norm": 0.20767998695373535,
      "learning_rate": 1.4010249779809503e-05,
      "loss": 0.0142,
      "step": 1790
    },
    {
      "epoch": 16.822429906542055,
      "grad_norm": 0.11333203315734863,
      "learning_rate": 1.3230209853117448e-05,
      "loss": 0.0144,
      "step": 1800
    },
    {
      "epoch": 16.91588785046729,
      "grad_norm": 0.10632842034101486,
      "learning_rate": 1.2470971468042558e-05,
      "loss": 0.0139,
      "step": 1810
    },
    {
      "epoch": 17.009345794392523,
      "grad_norm": 0.18203343451023102,
      "learning_rate": 1.1732716638726382e-05,
      "loss": 0.014,
      "step": 1820
    },
    {
      "epoch": 17.102803738317757,
      "grad_norm": 0.21806679666042328,
      "learning_rate": 1.1015622348869491e-05,
      "loss": 0.0129,
      "step": 1830
    },
    {
      "epoch": 17.19626168224299,
      "grad_norm": 0.19995306432247162,
      "learning_rate": 1.0319860509302637e-05,
      "loss": 0.0131,
      "step": 1840
    },
    {
      "epoch": 17.289719626168225,
      "grad_norm": 0.0872287005186081,
      "learning_rate": 9.64559791677434e-06,
      "loss": 0.0153,
      "step": 1850
    },
    {
      "epoch": 17.38317757009346,
      "grad_norm": 0.13513228297233582,
      "learning_rate": 8.992996213964067e-06,
      "loss": 0.0158,
      "step": 1860
    },
    {
      "epoch": 17.476635514018692,
      "grad_norm": 0.1807233691215515,
      "learning_rate": 8.362211850731382e-06,
      "loss": 0.0113,
      "step": 1870
    },
    {
      "epoch": 17.570093457943926,
      "grad_norm": 0.19154734909534454,
      "learning_rate": 7.75339604660973e-06,
      "loss": 0.0161,
      "step": 1880
    },
    {
      "epoch": 17.66355140186916,
      "grad_norm": 0.2096325159072876,
      "learning_rate": 7.166694754554193e-06,
      "loss": 0.016,
      "step": 1890
    },
    {
      "epoch": 17.757009345794394,
      "grad_norm": 0.16648273169994354,
      "learning_rate": 6.602248625951934e-06,
      "loss": 0.013,
      "step": 1900
    },
    {
      "epoch": 17.850467289719624,
      "grad_norm": 0.19170570373535156,
      "learning_rate": 6.0601929769033115e-06,
      "loss": 0.0116,
      "step": 1910
    },
    {
      "epoch": 17.94392523364486,
      "grad_norm": 0.14835980534553528,
      "learning_rate": 5.540657755782341e-06,
      "loss": 0.0136,
      "step": 1920
    },
    {
      "epoch": 18.037383177570092,
      "grad_norm": 0.10613177716732025,
      "learning_rate": 5.043767512083797e-06,
      "loss": 0.012,
      "step": 1930
    },
    {
      "epoch": 18.130841121495326,
      "grad_norm": 0.18721149861812592,
      "learning_rate": 4.569641366564614e-06,
      "loss": 0.0153,
      "step": 1940
    },
    {
      "epoch": 18.22429906542056,
      "grad_norm": 0.17616799473762512,
      "learning_rate": 4.1183929826868476e-06,
      "loss": 0.0115,
      "step": 1950
    },
    {
      "epoch": 18.317757009345794,
      "grad_norm": 0.2676970064640045,
      "learning_rate": 3.6901305393687634e-06,
      "loss": 0.0113,
      "step": 1960
    },
    {
      "epoch": 18.411214953271028,
      "grad_norm": 0.10376067459583282,
      "learning_rate": 3.284956705050868e-06,
      "loss": 0.0122,
      "step": 1970
    },
    {
      "epoch": 18.50467289719626,
      "grad_norm": 0.12165194004774094,
      "learning_rate": 2.902968613082957e-06,
      "loss": 0.0134,
      "step": 1980
    },
    {
      "epoch": 18.598130841121495,
      "grad_norm": 0.15639284253120422,
      "learning_rate": 2.5442578384380245e-06,
      "loss": 0.0115,
      "step": 1990
    },
    {
      "epoch": 18.69158878504673,
      "grad_norm": 0.14054954051971436,
      "learning_rate": 2.208910375758888e-06,
      "loss": 0.0111,
      "step": 2000
    },
    {
      "epoch": 18.785046728971963,
      "grad_norm": 0.13985030353069305,
      "learning_rate": 1.8970066187423364e-06,
      "loss": 0.0115,
      "step": 2010
    },
    {
      "epoch": 18.878504672897197,
      "grad_norm": 0.12971541285514832,
      "learning_rate": 1.6086213408662232e-06,
      "loss": 0.0117,
      "step": 2020
    },
    {
      "epoch": 18.97196261682243,
      "grad_norm": 0.17401276528835297,
      "learning_rate": 1.343823677463818e-06,
      "loss": 0.0139,
      "step": 2030
    },
    {
      "epoch": 19.065420560747665,
      "grad_norm": 0.24669478833675385,
      "learning_rate": 1.1026771091498079e-06,
      "loss": 0.0143,
      "step": 2040
    },
    {
      "epoch": 19.1588785046729,
      "grad_norm": 0.1944689005613327,
      "learning_rate": 8.852394466019487e-07,
      "loss": 0.0116,
      "step": 2050
    },
    {
      "epoch": 19.252336448598133,
      "grad_norm": 0.15337271988391876,
      "learning_rate": 6.915628167019628e-07,
      "loss": 0.0126,
      "step": 2060
    },
    {
      "epoch": 19.345794392523363,
      "grad_norm": 0.22086739540100098,
      "learning_rate": 5.216936500389679e-07,
      "loss": 0.0123,
      "step": 2070
    },
    {
      "epoch": 19.439252336448597,
      "grad_norm": 0.10731634497642517,
      "learning_rate": 3.7567266977868165e-07,
      "loss": 0.0117,
      "step": 2080
    },
    {
      "epoch": 19.53271028037383,
      "grad_norm": 0.16525092720985413,
      "learning_rate": 2.535348819006078e-07,
      "loss": 0.0142,
      "step": 2090
    },
    {
      "epoch": 19.626168224299064,
      "grad_norm": 0.22534893453121185,
      "learning_rate": 1.5530956680606114e-07,
      "loss": 0.0136,
      "step": 2100
    },
    {
      "epoch": 19.7196261682243,
      "grad_norm": 0.14636439085006714,
      "learning_rate": 8.102027229868192e-08,
      "loss": 0.0098,
      "step": 2110
    },
    {
      "epoch": 19.813084112149532,
      "grad_norm": 0.1717785745859146,
      "learning_rate": 3.0684807939262985e-08,
      "loss": 0.0105,
      "step": 2120
    },
    {
      "epoch": 19.906542056074766,
      "grad_norm": 0.15287938714027405,
      "learning_rate": 4.315240776175955e-09,
      "loss": 0.0135,
      "step": 2130
    },
    {
      "epoch": 19.962616822429908,
      "step": 2136,
      "total_flos": 3.092595209923816e+17,
      "train_loss": 0.03558622600950217,
      "train_runtime": 2455.8794,
      "train_samples_per_second": 55.664,
      "train_steps_per_second": 0.87
    }
  ],
  "logging_steps": 10,
  "max_steps": 2136,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.092595209923816e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}