{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0899488926746166,
  "eval_steps": 500,
  "global_step": 1600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0034071550255536627,
      "grad_norm": 9.859827995300293,
      "learning_rate": 4.917142857142858e-06,
      "loss": 0.7937,
      "mean_token_accuracy": 0.8738817512989044,
      "num_tokens": 739751.0,
      "step": 5
    },
    {
      "epoch": 0.0068143100511073255,
      "grad_norm": 1.87760329246521,
      "learning_rate": 1.106357142857143e-05,
      "loss": 0.3667,
      "mean_token_accuracy": 0.9218809515237808,
      "num_tokens": 1481661.0,
      "step": 10
    },
    {
      "epoch": 0.010221465076660987,
      "grad_norm": 1.129870891571045,
      "learning_rate": 1.7210000000000003e-05,
      "loss": 0.1964,
      "mean_token_accuracy": 0.9465200901031494,
      "num_tokens": 2206090.0,
      "step": 15
    },
    {
      "epoch": 0.013628620102214651,
      "grad_norm": 0.831566333770752,
      "learning_rate": 2.3356428571428574e-05,
      "loss": 0.0839,
      "mean_token_accuracy": 0.9683015376329422,
      "num_tokens": 2929568.0,
      "step": 20
    },
    {
      "epoch": 0.017035775127768313,
      "grad_norm": 0.6768407225608826,
      "learning_rate": 2.950285714285715e-05,
      "loss": 0.058,
      "mean_token_accuracy": 0.9755744874477387,
      "num_tokens": 3656469.0,
      "step": 25
    },
    {
      "epoch": 0.020442930153321975,
      "grad_norm": 0.5586706399917603,
      "learning_rate": 3.5649285714285724e-05,
      "loss": 0.0467,
      "mean_token_accuracy": 0.9799701869487762,
      "num_tokens": 4382989.0,
      "step": 30
    },
    {
      "epoch": 0.02385008517887564,
      "grad_norm": 0.6383557319641113,
      "learning_rate": 4.179571428571429e-05,
      "loss": 0.0407,
      "mean_token_accuracy": 0.9826571196317673,
      "num_tokens": 5122605.0,
      "step": 35
    },
    {
      "epoch": 0.027257240204429302,
      "grad_norm": 0.5180884599685669,
      "learning_rate": 4.3024933261338815e-05,
      "loss": 0.042,
      "mean_token_accuracy": 0.9811285108327865,
      "num_tokens": 5850567.0,
      "step": 40
    },
    {
      "epoch": 0.030664395229982964,
      "grad_norm": 0.3146582245826721,
      "learning_rate": 4.302466213647396e-05,
      "loss": 0.0322,
      "mean_token_accuracy": 0.9858334630727768,
      "num_tokens": 6574358.0,
      "step": 45
    },
    {
      "epoch": 0.034071550255536626,
      "grad_norm": 0.504967212677002,
      "learning_rate": 4.3024182457741095e-05,
      "loss": 0.0324,
      "mean_token_accuracy": 0.9867621034383773,
      "num_tokens": 7296052.0,
      "step": 50
    },
    {
      "epoch": 0.03747870528109029,
      "grad_norm": 0.4430710971355438,
      "learning_rate": 4.302349423134068e-05,
      "loss": 0.0257,
      "mean_token_accuracy": 0.9878164857625962,
      "num_tokens": 8040783.0,
      "step": 55
    },
    {
      "epoch": 0.04088586030664395,
      "grad_norm": 0.2613435983657837,
      "learning_rate": 4.3022597466168976e-05,
      "loss": 0.0253,
      "mean_token_accuracy": 0.988895320892334,
      "num_tokens": 8772505.0,
      "step": 60
    },
    {
      "epoch": 0.044293015332197615,
      "grad_norm": 0.28959766030311584,
      "learning_rate": 4.302149217381784e-05,
      "loss": 0.0244,
      "mean_token_accuracy": 0.9890943795442582,
      "num_tokens": 9502629.0,
      "step": 65
    },
    {
      "epoch": 0.04770017035775128,
      "grad_norm": 0.265918105840683,
      "learning_rate": 4.3020178368574676e-05,
      "loss": 0.0198,
      "mean_token_accuracy": 0.9914545267820358,
      "num_tokens": 10233837.0,
      "step": 70
    },
    {
      "epoch": 0.05110732538330494,
      "grad_norm": 0.25154563784599304,
      "learning_rate": 4.301865606742213e-05,
      "loss": 0.0189,
      "mean_token_accuracy": 0.9921916872262955,
      "num_tokens": 10958817.0,
      "step": 75
    },
    {
      "epoch": 0.054514480408858604,
      "grad_norm": 0.31910449266433716,
      "learning_rate": 4.3016925290038e-05,
      "loss": 0.0173,
      "mean_token_accuracy": 0.993075042963028,
      "num_tokens": 11691714.0,
      "step": 80
    },
    {
      "epoch": 0.05792163543441227,
      "grad_norm": 0.4202966094017029,
      "learning_rate": 4.3014986058794856e-05,
      "loss": 0.0171,
      "mean_token_accuracy": 0.99135502576828,
      "num_tokens": 12423624.0,
      "step": 85
    },
    {
      "epoch": 0.06132879045996593,
      "grad_norm": 0.2884162366390228,
      "learning_rate": 4.301283839875985e-05,
      "loss": 0.0129,
      "mean_token_accuracy": 0.9945991188287735,
      "num_tokens": 13164708.0,
      "step": 90
    },
    {
      "epoch": 0.06473594548551959,
      "grad_norm": 0.2969397008419037,
      "learning_rate": 4.301048233769434e-05,
      "loss": 0.0157,
      "mean_token_accuracy": 0.9931935012340546,
      "num_tokens": 13891801.0,
      "step": 95
    },
    {
      "epoch": 0.06814310051107325,
      "grad_norm": 0.20829303562641144,
      "learning_rate": 4.300791790605355e-05,
      "loss": 0.0145,
      "mean_token_accuracy": 0.9934373378753663,
      "num_tokens": 14623442.0,
      "step": 100
    },
    {
      "epoch": 0.07155025553662692,
      "grad_norm": 0.20044466853141785,
      "learning_rate": 4.300514513698616e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.9947379142045975,
      "num_tokens": 15358748.0,
      "step": 105
    },
    {
      "epoch": 0.07495741056218058,
      "grad_norm": 0.19780954718589783,
      "learning_rate": 4.3002164066333896e-05,
      "loss": 0.0146,
      "mean_token_accuracy": 0.9937324762344361,
      "num_tokens": 16083021.0,
      "step": 110
    },
    {
      "epoch": 0.07836456558773425,
      "grad_norm": 0.2107764631509781,
      "learning_rate": 4.299897473263106e-05,
      "loss": 0.0125,
      "mean_token_accuracy": 0.9938120871782303,
      "num_tokens": 16817511.0,
      "step": 115
    },
    {
      "epoch": 0.0817717206132879,
      "grad_norm": 0.19894278049468994,
      "learning_rate": 4.299557717710403e-05,
      "loss": 0.0138,
      "mean_token_accuracy": 0.9935688495635986,
      "num_tokens": 17545152.0,
      "step": 120
    },
    {
      "epoch": 0.08517887563884156,
      "grad_norm": 0.1144740954041481,
      "learning_rate": 4.299197144367072e-05,
      "loss": 0.0109,
      "mean_token_accuracy": 0.9947319984436035,
      "num_tokens": 18276796.0,
      "step": 125
    },
    {
      "epoch": 0.08858603066439523,
      "grad_norm": 0.14482234418392181,
      "learning_rate": 4.298815757894002e-05,
      "loss": 0.0116,
      "mean_token_accuracy": 0.9940285742282867,
      "num_tokens": 19001338.0,
      "step": 130
    },
    {
      "epoch": 0.0919931856899489,
      "grad_norm": 0.103367380797863,
      "learning_rate": 4.298413563221122e-05,
      "loss": 0.0093,
      "mean_token_accuracy": 0.9953823149204254,
      "num_tokens": 19726859.0,
      "step": 135
    },
    {
      "epoch": 0.09540034071550256,
      "grad_norm": 0.44147828221321106,
      "learning_rate": 4.297990565547331e-05,
      "loss": 0.0095,
      "mean_token_accuracy": 0.9961490094661712,
      "num_tokens": 20465377.0,
      "step": 140
    },
    {
      "epoch": 0.09880749574105621,
      "grad_norm": 0.4469468295574188,
      "learning_rate": 4.2975467703404354e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.994527843594551,
      "num_tokens": 21190517.0,
      "step": 145
    },
    {
      "epoch": 0.10221465076660988,
      "grad_norm": 0.31705236434936523,
      "learning_rate": 4.2970821833370796e-05,
      "loss": 0.011,
      "mean_token_accuracy": 0.9949095755815506,
      "num_tokens": 21925938.0,
      "step": 150
    },
    {
      "epoch": 0.10562180579216354,
      "grad_norm": 0.27338773012161255,
      "learning_rate": 4.296596810542667e-05,
      "loss": 0.0084,
      "mean_token_accuracy": 0.9966103434562683,
      "num_tokens": 22665847.0,
      "step": 155
    },
    {
      "epoch": 0.10902896081771721,
      "grad_norm": 0.15007174015045166,
      "learning_rate": 4.296090658231286e-05,
      "loss": 0.0126,
      "mean_token_accuracy": 0.9939430177211761,
      "num_tokens": 23395678.0,
      "step": 160
    },
    {
      "epoch": 0.11243611584327087,
      "grad_norm": 0.1724090427160263,
      "learning_rate": 4.295563732945628e-05,
      "loss": 0.0118,
      "mean_token_accuracy": 0.9948061585426331,
      "num_tokens": 24123619.0,
      "step": 165
    },
    {
      "epoch": 0.11584327086882454,
      "grad_norm": 0.14645731449127197,
      "learning_rate": 4.295016041496902e-05,
      "loss": 0.01,
      "mean_token_accuracy": 0.9955295205116272,
      "num_tokens": 24850815.0,
      "step": 170
    },
    {
      "epoch": 0.11925042589437819,
      "grad_norm": 0.31964659690856934,
      "learning_rate": 4.294447590964748e-05,
      "loss": 0.0112,
      "mean_token_accuracy": 0.9953448295593261,
      "num_tokens": 25565061.0,
      "step": 175
    },
    {
      "epoch": 0.12265758091993186,
      "grad_norm": 0.19131319224834442,
      "learning_rate": 4.293858388697145e-05,
      "loss": 0.0088,
      "mean_token_accuracy": 0.9952926129102707,
      "num_tokens": 26291395.0,
      "step": 180
    },
    {
      "epoch": 0.1260647359454855,
      "grad_norm": 0.0894119068980217,
      "learning_rate": 4.293248442310314e-05,
      "loss": 0.0084,
      "mean_token_accuracy": 0.9962390959262848,
      "num_tokens": 27020673.0,
      "step": 185
    },
    {
      "epoch": 0.12947189097103917,
      "grad_norm": 0.1555798500776291,
      "learning_rate": 4.292617759688624e-05,
      "loss": 0.0083,
      "mean_token_accuracy": 0.9956040650606155,
      "num_tokens": 27751201.0,
      "step": 190
    },
    {
      "epoch": 0.13287904599659284,
      "grad_norm": 0.19205354154109955,
      "learning_rate": 4.2919663489844846e-05,
      "loss": 0.0094,
      "mean_token_accuracy": 0.995469844341278,
      "num_tokens": 28485445.0,
      "step": 195
    },
    {
      "epoch": 0.1362862010221465,
      "grad_norm": 0.10893631726503372,
      "learning_rate": 4.291294218618245e-05,
      "loss": 0.0075,
      "mean_token_accuracy": 0.9966305375099183,
      "num_tokens": 29213760.0,
      "step": 200
    },
    {
      "epoch": 0.13969335604770017,
      "grad_norm": 0.15836721658706665,
      "learning_rate": 4.2906013772780834e-05,
      "loss": 0.0087,
      "mean_token_accuracy": 0.9962090939283371,
      "num_tokens": 29955463.0,
      "step": 205
    },
    {
      "epoch": 0.14310051107325383,
      "grad_norm": 0.18292012810707092,
      "learning_rate": 4.289887833919896e-05,
      "loss": 0.0098,
      "mean_token_accuracy": 0.995222756266594,
      "num_tokens": 30692343.0,
      "step": 210
    },
    {
      "epoch": 0.1465076660988075,
      "grad_norm": 0.10357630252838135,
      "learning_rate": 4.289153597767176e-05,
      "loss": 0.009,
      "mean_token_accuracy": 0.9959124267101288,
      "num_tokens": 31413819.0,
      "step": 215
    },
    {
      "epoch": 0.14991482112436116,
      "grad_norm": 0.13664352893829346,
      "learning_rate": 4.288398678310904e-05,
      "loss": 0.0101,
      "mean_token_accuracy": 0.9948742359876632,
      "num_tokens": 32136634.0,
      "step": 220
    },
    {
      "epoch": 0.15332197614991483,
      "grad_norm": 0.08003365248441696,
      "learning_rate": 4.2876230853094156e-05,
      "loss": 0.0098,
      "mean_token_accuracy": 0.9949465572834015,
      "num_tokens": 32848525.0,
      "step": 225
    },
    {
      "epoch": 0.1567291311754685,
      "grad_norm": 0.061436109244823456,
      "learning_rate": 4.286826828788281e-05,
      "loss": 0.0074,
      "mean_token_accuracy": 0.996098268032074,
      "num_tokens": 33582190.0,
      "step": 230
    },
    {
      "epoch": 0.16013628620102216,
      "grad_norm": 0.10085827112197876,
      "learning_rate": 4.286009919040175e-05,
      "loss": 0.0071,
      "mean_token_accuracy": 0.9966446220874786,
      "num_tokens": 34321697.0,
      "step": 235
    },
    {
      "epoch": 0.1635434412265758,
      "grad_norm": 0.04359719902276993,
      "learning_rate": 4.28517236662474e-05,
      "loss": 0.0068,
      "mean_token_accuracy": 0.9967904776334763,
      "num_tokens": 35050174.0,
      "step": 240
    },
    {
      "epoch": 0.16695059625212946,
      "grad_norm": 0.15128719806671143,
      "learning_rate": 4.284314182368455e-05,
      "loss": 0.0074,
      "mean_token_accuracy": 0.9964352756738662,
      "num_tokens": 35784128.0,
      "step": 245
    },
    {
      "epoch": 0.17035775127768313,
      "grad_norm": 0.1763068437576294,
      "learning_rate": 4.2834353773644904e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9973692268133163,
      "num_tokens": 36513412.0,
      "step": 250
    },
    {
      "epoch": 0.1737649063032368,
      "grad_norm": 0.10363134741783142,
      "learning_rate": 4.28253596297257e-05,
      "loss": 0.0101,
      "mean_token_accuracy": 0.9954942435026168,
      "num_tokens": 37229271.0,
      "step": 255
    },
    {
      "epoch": 0.17717206132879046,
      "grad_norm": 0.08556711673736572,
      "learning_rate": 4.28161595081882e-05,
      "loss": 0.0093,
      "mean_token_accuracy": 0.995372086763382,
      "num_tokens": 37954980.0,
      "step": 260
    },
    {
      "epoch": 0.18057921635434412,
      "grad_norm": 0.11393541842699051,
      "learning_rate": 4.2806753527956174e-05,
      "loss": 0.0088,
      "mean_token_accuracy": 0.9960035026073456,
      "num_tokens": 38680301.0,
      "step": 265
    },
    {
      "epoch": 0.1839863713798978,
      "grad_norm": 0.11953490972518921,
      "learning_rate": 4.279714181061445e-05,
      "loss": 0.008,
      "mean_token_accuracy": 0.9957300066947937,
      "num_tokens": 39399422.0,
      "step": 270
    },
    {
      "epoch": 0.18739352640545145,
      "grad_norm": 0.18713217973709106,
      "learning_rate": 4.278732448040723e-05,
      "loss": 0.0081,
      "mean_token_accuracy": 0.9962434619665146,
      "num_tokens": 40128793.0,
      "step": 275
    },
    {
      "epoch": 0.19080068143100512,
      "grad_norm": 0.12800747156143188,
      "learning_rate": 4.2777301664236546e-05,
      "loss": 0.0077,
      "mean_token_accuracy": 0.9958244591951371,
      "num_tokens": 40852378.0,
      "step": 280
    },
    {
      "epoch": 0.19420783645655879,
      "grad_norm": 0.1399906575679779,
      "learning_rate": 4.2767073491660625e-05,
      "loss": 0.0071,
      "mean_token_accuracy": 0.9958741277456283,
      "num_tokens": 41585390.0,
      "step": 285
    },
    {
      "epoch": 0.19761499148211242,
      "grad_norm": 0.14149264991283417,
      "learning_rate": 4.275664009489218e-05,
      "loss": 0.0097,
      "mean_token_accuracy": 0.9950805366039276,
      "num_tokens": 42308088.0,
      "step": 290
    },
    {
      "epoch": 0.2010221465076661,
      "grad_norm": 0.1298685073852539,
      "learning_rate": 4.274600160879674e-05,
      "loss": 0.0071,
      "mean_token_accuracy": 0.9965922892093658,
      "num_tokens": 43032168.0,
      "step": 295
    },
    {
      "epoch": 0.20442930153321975,
      "grad_norm": 0.09813089668750763,
      "learning_rate": 4.2735158170890864e-05,
      "loss": 0.0073,
      "mean_token_accuracy": 0.9965780138969421,
      "num_tokens": 43771887.0,
      "step": 300
    },
    {
      "epoch": 0.20783645655877342,
      "grad_norm": 0.09381706267595291,
      "learning_rate": 4.272410992134039e-05,
      "loss": 0.0079,
      "mean_token_accuracy": 0.9966065973043442,
      "num_tokens": 44511954.0,
      "step": 305
    },
    {
      "epoch": 0.21124361158432708,
      "grad_norm": 0.07301747798919678,
      "learning_rate": 4.271285700295863e-05,
      "loss": 0.0085,
      "mean_token_accuracy": 0.9960614979267121,
      "num_tokens": 45242949.0,
      "step": 310
    },
    {
      "epoch": 0.21465076660988075,
      "grad_norm": 0.1007862314581871,
      "learning_rate": 4.2701399561204506e-05,
      "loss": 0.0077,
      "mean_token_accuracy": 0.9957682400941849,
      "num_tokens": 45965684.0,
      "step": 315
    },
    {
      "epoch": 0.21805792163543442,
      "grad_norm": 0.14022916555404663,
      "learning_rate": 4.2689737744180684e-05,
      "loss": 0.007,
      "mean_token_accuracy": 0.9966099619865417,
      "num_tokens": 46690860.0,
      "step": 320
    },
    {
      "epoch": 0.22146507666098808,
      "grad_norm": 0.08956095576286316,
      "learning_rate": 4.267787170263163e-05,
      "loss": 0.0075,
      "mean_token_accuracy": 0.9969005316495896,
      "num_tokens": 47418842.0,
      "step": 325
    },
    {
      "epoch": 0.22487223168654175,
      "grad_norm": 0.17472460865974426,
      "learning_rate": 4.266580158994171e-05,
      "loss": 0.0085,
      "mean_token_accuracy": 0.9956019967794418,
      "num_tokens": 48141378.0,
      "step": 330
    },
    {
      "epoch": 0.2282793867120954,
      "grad_norm": 0.07788616418838501,
      "learning_rate": 4.2653527562133166e-05,
      "loss": 0.0067,
      "mean_token_accuracy": 0.9971259295940399,
      "num_tokens": 48873966.0,
      "step": 335
    },
    {
      "epoch": 0.23168654173764908,
      "grad_norm": 0.06162365898489952,
      "learning_rate": 4.264104977786411e-05,
      "loss": 0.0068,
      "mean_token_accuracy": 0.9966557651758194,
      "num_tokens": 49605425.0,
      "step": 340
    },
    {
      "epoch": 0.23509369676320271,
      "grad_norm": 0.11083303391933441,
      "learning_rate": 4.262836839842649e-05,
      "loss": 0.0087,
      "mean_token_accuracy": 0.9957881480455398,
      "num_tokens": 50337086.0,
      "step": 345
    },
    {
      "epoch": 0.23850085178875638,
      "grad_norm": 0.07837757468223572,
      "learning_rate": 4.2615483587743984e-05,
      "loss": 0.0071,
      "mean_token_accuracy": 0.9961496263742446,
      "num_tokens": 51064074.0,
      "step": 350
    },
    {
      "epoch": 0.24190800681431004,
      "grad_norm": 0.06718693673610687,
      "learning_rate": 4.260239551236989e-05,
      "loss": 0.007,
      "mean_token_accuracy": 0.9963487684726715,
      "num_tokens": 51798154.0,
      "step": 355
    },
    {
      "epoch": 0.2453151618398637,
      "grad_norm": 0.10025134682655334,
      "learning_rate": 4.258910434148497e-05,
      "loss": 0.0086,
      "mean_token_accuracy": 0.9955656617879868,
      "num_tokens": 52519253.0,
      "step": 360
    },
    {
      "epoch": 0.24872231686541738,
      "grad_norm": 0.08592838793992996,
      "learning_rate": 4.257561024689527e-05,
      "loss": 0.0074,
      "mean_token_accuracy": 0.9956583797931671,
      "num_tokens": 53250055.0,
      "step": 365
    },
    {
      "epoch": 0.252129471890971,
      "grad_norm": 0.0698947086930275,
      "learning_rate": 4.256191340302991e-05,
      "loss": 0.0069,
      "mean_token_accuracy": 0.9965931236743927,
      "num_tokens": 53966965.0,
      "step": 370
    },
    {
      "epoch": 0.2555366269165247,
      "grad_norm": 0.34333905577659607,
      "learning_rate": 4.254801398693877e-05,
      "loss": 0.0065,
      "mean_token_accuracy": 0.9969754308462143,
      "num_tokens": 54702533.0,
      "step": 375
    },
    {
      "epoch": 0.25894378194207834,
      "grad_norm": 0.09627313166856766,
      "learning_rate": 4.253391217829028e-05,
      "loss": 0.0082,
      "mean_token_accuracy": 0.9958543449640274,
      "num_tokens": 55438831.0,
      "step": 380
    },
    {
      "epoch": 0.262350936967632,
      "grad_norm": 0.11762526631355286,
      "learning_rate": 4.251960815936906e-05,
      "loss": 0.0081,
      "mean_token_accuracy": 0.9956094622612,
      "num_tokens": 56166035.0,
      "step": 385
    },
    {
      "epoch": 0.2657580919931857,
      "grad_norm": 0.14286452531814575,
      "learning_rate": 4.250510211507356e-05,
      "loss": 0.0076,
      "mean_token_accuracy": 0.9963731884956359,
      "num_tokens": 56897804.0,
      "step": 390
    },
    {
      "epoch": 0.26916524701873934,
      "grad_norm": 0.17650602757930756,
      "learning_rate": 4.249039423291368e-05,
      "loss": 0.0071,
      "mean_token_accuracy": 0.9962969094514846,
      "num_tokens": 57634989.0,
      "step": 395
    },
    {
      "epoch": 0.272572402044293,
      "grad_norm": 0.0812569335103035,
      "learning_rate": 4.247548470300833e-05,
      "loss": 0.0069,
      "mean_token_accuracy": 0.9965808987617493,
      "num_tokens": 58366231.0,
      "step": 400
    },
    {
      "epoch": 0.27597955706984667,
      "grad_norm": 0.09315631538629532,
      "learning_rate": 4.2460373718083e-05,
      "loss": 0.0086,
      "mean_token_accuracy": 0.9956678539514542,
      "num_tokens": 59087658.0,
      "step": 405
    },
    {
      "epoch": 0.27938671209540034,
      "grad_norm": 0.1458846479654312,
      "learning_rate": 4.244506147346725e-05,
      "loss": 0.0072,
      "mean_token_accuracy": 0.9962426245212554,
      "num_tokens": 59828436.0,
      "step": 410
    },
    {
      "epoch": 0.282793867120954,
      "grad_norm": 0.15108242630958557,
      "learning_rate": 4.2429548167092186e-05,
      "loss": 0.0091,
      "mean_token_accuracy": 0.9956044554710388,
      "num_tokens": 60541919.0,
      "step": 415
    },
    {
      "epoch": 0.28620102214650767,
      "grad_norm": 0.12515656650066376,
      "learning_rate": 4.241383399948791e-05,
      "loss": 0.0071,
      "mean_token_accuracy": 0.9964348942041397,
      "num_tokens": 61269079.0,
      "step": 420
    },
    {
      "epoch": 0.28960817717206133,
      "grad_norm": 0.0677802786231041,
      "learning_rate": 4.239791917378092e-05,
      "loss": 0.007,
      "mean_token_accuracy": 0.9968853920698166,
      "num_tokens": 61994345.0,
      "step": 425
    },
    {
      "epoch": 0.293015332197615,
      "grad_norm": 0.1035194844007492,
      "learning_rate": 4.2381803895691476e-05,
      "loss": 0.0084,
      "mean_token_accuracy": 0.996373075246811,
      "num_tokens": 62722716.0,
      "step": 430
    },
    {
      "epoch": 0.29642248722316866,
      "grad_norm": 0.13024494051933289,
      "learning_rate": 4.2365488373530966e-05,
      "loss": 0.0095,
      "mean_token_accuracy": 0.9957513123750686,
      "num_tokens": 63450632.0,
      "step": 435
    },
    {
      "epoch": 0.29982964224872233,
      "grad_norm": 0.08041954040527344,
      "learning_rate": 4.234897281819919e-05,
      "loss": 0.0092,
      "mean_token_accuracy": 0.9949093878269195,
      "num_tokens": 64169205.0,
      "step": 440
    },
    {
      "epoch": 0.303236797274276,
      "grad_norm": 0.08703923970460892,
      "learning_rate": 4.233225744318165e-05,
      "loss": 0.0072,
      "mean_token_accuracy": 0.996146160364151,
      "num_tokens": 64886929.0,
      "step": 445
    },
    {
      "epoch": 0.30664395229982966,
      "grad_norm": 0.08134111016988754,
      "learning_rate": 4.2315342464546776e-05,
      "loss": 0.0059,
      "mean_token_accuracy": 0.9973055094480514,
      "num_tokens": 65606118.0,
      "step": 450
    },
    {
      "epoch": 0.3100511073253833,
      "grad_norm": 0.08342619985342026,
      "learning_rate": 4.229822810094314e-05,
      "loss": 0.0071,
      "mean_token_accuracy": 0.9963917940855026,
      "num_tokens": 66332962.0,
      "step": 455
    },
    {
      "epoch": 0.313458262350937,
      "grad_norm": 0.07163641601800919,
      "learning_rate": 4.2280914573596646e-05,
      "loss": 0.0068,
      "mean_token_accuracy": 0.996507003903389,
      "num_tokens": 67058111.0,
      "step": 460
    },
    {
      "epoch": 0.31686541737649065,
      "grad_norm": 0.07286284118890762,
      "learning_rate": 4.226340210630762e-05,
      "loss": 0.0075,
      "mean_token_accuracy": 0.9964655578136444,
      "num_tokens": 67781600.0,
      "step": 465
    },
    {
      "epoch": 0.3202725724020443,
      "grad_norm": 0.10659992694854736,
      "learning_rate": 4.2245690925447985e-05,
      "loss": 0.0065,
      "mean_token_accuracy": 0.9970693528652191,
      "num_tokens": 68521956.0,
      "step": 470
    },
    {
      "epoch": 0.32367972742759793,
      "grad_norm": 0.0698583796620369,
      "learning_rate": 4.222778125995828e-05,
      "loss": 0.007,
      "mean_token_accuracy": 0.9972410887479782,
      "num_tokens": 69257650.0,
      "step": 475
    },
    {
      "epoch": 0.3270868824531516,
      "grad_norm": 0.07957218587398529,
      "learning_rate": 4.220967334134473e-05,
      "loss": 0.0062,
      "mean_token_accuracy": 0.9966995030641556,
      "num_tokens": 69989942.0,
      "step": 480
    },
    {
      "epoch": 0.33049403747870526,
      "grad_norm": 0.13976924121379852,
      "learning_rate": 4.219136740367625e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9967512339353561,
      "num_tokens": 70710975.0,
      "step": 485
    },
    {
      "epoch": 0.3339011925042589,
      "grad_norm": 0.2637229263782501,
      "learning_rate": 4.21728636835814e-05,
      "loss": 0.0069,
      "mean_token_accuracy": 0.9974818140268326,
      "num_tokens": 71432676.0,
      "step": 490
    },
    {
      "epoch": 0.3373083475298126,
      "grad_norm": 0.0588751882314682,
      "learning_rate": 4.215416242024536e-05,
      "loss": 0.0073,
      "mean_token_accuracy": 0.9966814190149307,
      "num_tokens": 72161382.0,
      "step": 495
    },
    {
      "epoch": 0.34071550255536626,
      "grad_norm": 0.06412996351718903,
      "learning_rate": 4.21352638554068e-05,
      "loss": 0.0062,
      "mean_token_accuracy": 0.9972320377826691,
      "num_tokens": 72893241.0,
      "step": 500
    },
    {
      "epoch": 0.3441226575809199,
      "grad_norm": 0.0671689510345459,
      "learning_rate": 4.211616823335478e-05,
      "loss": 0.0073,
      "mean_token_accuracy": 0.9966926813125611,
      "num_tokens": 73623855.0,
      "step": 505
    },
    {
      "epoch": 0.3475298126064736,
      "grad_norm": 0.07332868129014969,
      "learning_rate": 4.2096875800925595e-05,
      "loss": 0.0073,
      "mean_token_accuracy": 0.9962316125631332,
      "num_tokens": 74362370.0,
      "step": 510
    },
    {
      "epoch": 0.35093696763202725,
      "grad_norm": 0.04477046802639961,
      "learning_rate": 4.207738680749956e-05,
      "loss": 0.0059,
      "mean_token_accuracy": 0.9971085309982299,
      "num_tokens": 75089913.0,
      "step": 515
    },
    {
      "epoch": 0.3543441226575809,
      "grad_norm": 0.05897574499249458,
      "learning_rate": 4.2057701504997816e-05,
      "loss": 0.0065,
      "mean_token_accuracy": 0.9970849037170411,
      "num_tokens": 75820952.0,
      "step": 520
    },
    {
      "epoch": 0.3577512776831346,
      "grad_norm": 0.08671557903289795,
      "learning_rate": 4.203782014787905e-05,
      "loss": 0.0069,
      "mean_token_accuracy": 0.9971589148044586,
      "num_tokens": 76554503.0,
      "step": 525
    },
    {
      "epoch": 0.36115843270868825,
      "grad_norm": 0.11050685495138168,
      "learning_rate": 4.201774299313624e-05,
      "loss": 0.0061,
      "mean_token_accuracy": 0.996543338894844,
      "num_tokens": 77286518.0,
      "step": 530
    },
    {
      "epoch": 0.3645655877342419,
      "grad_norm": 0.18593242764472961,
      "learning_rate": 4.199747030029327e-05,
      "loss": 0.0057,
      "mean_token_accuracy": 0.9971242040395737,
      "num_tokens": 77999789.0,
      "step": 535
    },
    {
      "epoch": 0.3679727427597956,
      "grad_norm": 0.15596704185009003,
      "learning_rate": 4.197700233140163e-05,
      "loss": 0.0073,
      "mean_token_accuracy": 0.9962549567222595,
      "num_tokens": 78738433.0,
      "step": 540
    },
    {
      "epoch": 0.37137989778534924,
      "grad_norm": 0.1119828075170517,
      "learning_rate": 4.195633935103701e-05,
      "loss": 0.0055,
      "mean_token_accuracy": 0.9968603044748306,
      "num_tokens": 79464554.0,
      "step": 545
    },
    {
      "epoch": 0.3747870528109029,
      "grad_norm": 0.0868990421295166,
      "learning_rate": 4.193548162629591e-05,
      "loss": 0.0067,
      "mean_token_accuracy": 0.9966555297374725,
      "num_tokens": 80196885.0,
      "step": 550
    },
    {
      "epoch": 0.3781942078364566,
      "grad_norm": 0.29425886273384094,
      "learning_rate": 4.1914429426792125e-05,
      "loss": 0.007,
      "mean_token_accuracy": 0.9964345067739486,
      "num_tokens": 80922485.0,
      "step": 555
    },
    {
      "epoch": 0.38160136286201024,
      "grad_norm": 0.36037981510162354,
      "learning_rate": 4.18931830246533e-05,
      "loss": 0.0074,
      "mean_token_accuracy": 0.9965087443590164,
      "num_tokens": 81646973.0,
      "step": 560
    },
    {
      "epoch": 0.3850085178875639,
      "grad_norm": 0.28591904044151306,
      "learning_rate": 4.187174269451742e-05,
      "loss": 0.0085,
      "mean_token_accuracy": 0.9958249300718307,
      "num_tokens": 82364390.0,
      "step": 565
    },
    {
      "epoch": 0.38841567291311757,
      "grad_norm": 0.1833687424659729,
      "learning_rate": 4.185010871352924e-05,
      "loss": 0.0076,
      "mean_token_accuracy": 0.9960876494646073,
      "num_tokens": 83093721.0,
      "step": 570
    },
    {
      "epoch": 0.39182282793867124,
      "grad_norm": 0.12492405623197556,
      "learning_rate": 4.1828281361336715e-05,
      "loss": 0.0072,
      "mean_token_accuracy": 0.9967440903186798,
      "num_tokens": 83815727.0,
      "step": 575
    },
    {
      "epoch": 0.39522998296422485,
      "grad_norm": 0.09792337566614151,
      "learning_rate": 4.180626092008737e-05,
      "loss": 0.0077,
      "mean_token_accuracy": 0.9961873531341553,
      "num_tokens": 84533677.0,
      "step": 580
    },
    {
      "epoch": 0.3986371379897785,
      "grad_norm": 0.07072041183710098,
      "learning_rate": 4.178404767442467e-05,
      "loss": 0.0057,
      "mean_token_accuracy": 0.9970175117254257,
      "num_tokens": 85262349.0,
      "step": 585
    },
    {
      "epoch": 0.4020442930153322,
      "grad_norm": 0.7827273607254028,
      "learning_rate": 4.176164191148432e-05,
      "loss": 0.0068,
      "mean_token_accuracy": 0.9970578879117966,
      "num_tokens": 85983657.0,
      "step": 590
    },
    {
      "epoch": 0.40545144804088584,
      "grad_norm": 0.15216010808944702,
      "learning_rate": 4.173904392089059e-05,
      "loss": 0.0099,
      "mean_token_accuracy": 0.9962922394275665,
      "num_tokens": 86704026.0,
      "step": 595
    },
    {
      "epoch": 0.4088586030664395,
      "grad_norm": 0.2555491030216217,
      "learning_rate": 4.171625399475253e-05,
      "loss": 0.0079,
      "mean_token_accuracy": 0.9959048479795456,
      "num_tokens": 87425440.0,
      "step": 600
    },
    {
      "epoch": 0.4122657580919932,
      "grad_norm": 0.09258874505758286,
      "learning_rate": 4.1693272427660226e-05,
      "loss": 0.0068,
      "mean_token_accuracy": 0.9973825007677078,
      "num_tokens": 88152625.0,
      "step": 605
    },
    {
      "epoch": 0.41567291311754684,
      "grad_norm": 0.20232026278972626,
      "learning_rate": 4.167009951668096e-05,
      "loss": 0.0092,
      "mean_token_accuracy": 0.9960968136787415,
      "num_tokens": 88892690.0,
      "step": 610
    },
    {
      "epoch": 0.4190800681431005,
      "grad_norm": 0.08462874591350555,
      "learning_rate": 4.16467355613554e-05,
      "loss": 0.0071,
      "mean_token_accuracy": 0.997252830862999,
      "num_tokens": 89630337.0,
      "step": 615
    },
    {
      "epoch": 0.42248722316865417,
      "grad_norm": 0.10808555781841278,
      "learning_rate": 4.162318086369371e-05,
      "loss": 0.0057,
      "mean_token_accuracy": 0.9974013090133667,
      "num_tokens": 90366154.0,
      "step": 620
    },
    {
      "epoch": 0.42589437819420783,
      "grad_norm": 0.07365107536315918,
      "learning_rate": 4.159943572817166e-05,
      "loss": 0.0054,
      "mean_token_accuracy": 0.9974009066820144,
      "num_tokens": 91105141.0,
      "step": 625
    },
    {
      "epoch": 0.4293015332197615,
      "grad_norm": 0.10557448118925095,
      "learning_rate": 4.157550046172669e-05,
      "loss": 0.0077,
      "mean_token_accuracy": 0.9966617226600647,
      "num_tokens": 91822389.0,
      "step": 630
    },
    {
      "epoch": 0.43270868824531517,
      "grad_norm": 0.05330486595630646,
      "learning_rate": 4.15513753737539e-05,
      "loss": 0.007,
      "mean_token_accuracy": 0.996642854809761,
      "num_tokens": 92544524.0,
      "step": 635
    },
    {
      "epoch": 0.43611584327086883,
      "grad_norm": 0.08263322710990906,
      "learning_rate": 4.152706077610211e-05,
      "loss": 0.0075,
      "mean_token_accuracy": 0.9965035200119019,
      "num_tokens": 93280767.0,
      "step": 640
    },
    {
      "epoch": 0.4395229982964225,
      "grad_norm": 0.1939581036567688,
      "learning_rate": 4.150255698306982e-05,
      "loss": 0.0079,
      "mean_token_accuracy": 0.9966277658939362,
      "num_tokens": 94009326.0,
      "step": 645
    },
    {
      "epoch": 0.44293015332197616,
      "grad_norm": 0.11316569894552231,
      "learning_rate": 4.1477864311401105e-05,
      "loss": 0.0076,
      "mean_token_accuracy": 0.9957149237394333,
      "num_tokens": 94733349.0,
      "step": 650
    },
    {
      "epoch": 0.4463373083475298,
      "grad_norm": 0.05097102373838425,
      "learning_rate": 4.145298308028155e-05,
      "loss": 0.0064,
      "mean_token_accuracy": 0.9973096132278443,
      "num_tokens": 95465665.0,
      "step": 655
    },
    {
      "epoch": 0.4497444633730835,
      "grad_norm": 0.06545665860176086,
      "learning_rate": 4.1427913611334136e-05,
      "loss": 0.0077,
      "mean_token_accuracy": 0.9959337741136551,
      "num_tokens": 96189888.0,
      "step": 660
    },
    {
      "epoch": 0.45315161839863716,
      "grad_norm": 0.07116438448429108,
      "learning_rate": 4.140265622861506e-05,
      "loss": 0.0065,
      "mean_token_accuracy": 0.9962837547063828,
      "num_tokens": 96917405.0,
      "step": 665
    },
    {
      "epoch": 0.4565587734241908,
      "grad_norm": 0.14547786116600037,
      "learning_rate": 4.137721125860956e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9975415617227554,
      "num_tokens": 97654511.0,
      "step": 670
    },
    {
      "epoch": 0.4599659284497445,
      "grad_norm": 0.07389840483665466,
      "learning_rate": 4.135157903022766e-05,
      "loss": 0.0062,
      "mean_token_accuracy": 0.9964825630187988,
      "num_tokens": 98387279.0,
      "step": 675
    },
    {
      "epoch": 0.46337308347529815,
      "grad_norm": 0.06067698821425438,
      "learning_rate": 4.13257598748e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9966069042682648,
      "num_tokens": 99107313.0,
      "step": 680
    },
    {
      "epoch": 0.46678023850085176,
      "grad_norm": 0.05793856456875801,
      "learning_rate": 4.1299754126073456e-05,
      "loss": 0.0078,
      "mean_token_accuracy": 0.996084812283516,
      "num_tokens": 99823780.0,
      "step": 685
    },
    {
      "epoch": 0.47018739352640543,
      "grad_norm": 0.09125217795372009,
      "learning_rate": 4.127356212020688e-05,
      "loss": 0.0061,
      "mean_token_accuracy": 0.9968007355928421,
      "num_tokens": 100561551.0,
      "step": 690
    },
    {
      "epoch": 0.4735945485519591,
      "grad_norm": 0.08726269751787186,
      "learning_rate": 4.124718419576675e-05,
      "loss": 0.0057,
      "mean_token_accuracy": 0.9969796687364578,
      "num_tokens": 101279579.0,
      "step": 695
    },
    {
      "epoch": 0.47700170357751276,
      "grad_norm": 0.06593688577413559,
      "learning_rate": 4.1220620693722814e-05,
      "loss": 0.0055,
      "mean_token_accuracy": 0.9972678184509277,
      "num_tokens": 102011160.0,
      "step": 700
    },
    {
      "epoch": 0.4804088586030664,
      "grad_norm": 0.0862775668501854,
      "learning_rate": 4.11938719574436e-05,
      "loss": 0.0052,
      "mean_token_accuracy": 0.9976744949817657,
      "num_tokens": 102741944.0,
      "step": 705
    },
    {
      "epoch": 0.4838160136286201,
      "grad_norm": 0.04647627845406532,
      "learning_rate": 4.116693833269208e-05,
      "loss": 0.0062,
      "mean_token_accuracy": 0.996781411767006,
      "num_tokens": 103470050.0,
      "step": 710
    },
    {
      "epoch": 0.48722316865417375,
      "grad_norm": 0.10203167796134949,
      "learning_rate": 4.1139820167621156e-05,
      "loss": 0.0073,
      "mean_token_accuracy": 0.9963602036237716,
      "num_tokens": 104200482.0,
      "step": 715
    },
    {
      "epoch": 0.4906303236797274,
      "grad_norm": 0.07260779291391373,
      "learning_rate": 4.1112517812769126e-05,
      "loss": 0.0079,
      "mean_token_accuracy": 0.9962616235017776,
      "num_tokens": 104920828.0,
      "step": 720
    },
    {
      "epoch": 0.4940374787052811,
      "grad_norm": 0.07864541560411453,
      "learning_rate": 4.108503162105521e-05,
      "loss": 0.0074,
      "mean_token_accuracy": 0.9965195417404175,
      "num_tokens": 105648158.0,
      "step": 725
    },
    {
      "epoch": 0.49744463373083475,
      "grad_norm": 0.15095138549804688,
      "learning_rate": 4.105736194777495e-05,
      "loss": 0.0076,
      "mean_token_accuracy": 0.9961408793926239,
      "num_tokens": 106377029.0,
      "step": 730
    },
    {
      "epoch": 0.5008517887563884,
      "grad_norm": 0.2626940608024597,
      "learning_rate": 4.102950915059566e-05,
      "loss": 0.0101,
      "mean_token_accuracy": 0.9956704348325729,
      "num_tokens": 107111684.0,
      "step": 735
    },
    {
      "epoch": 0.504258943781942,
      "grad_norm": 0.10384848713874817,
      "learning_rate": 4.1001473589551726e-05,
      "loss": 0.0054,
      "mean_token_accuracy": 0.9972564399242401,
      "num_tokens": 107844889.0,
      "step": 740
    },
    {
      "epoch": 0.5076660988074957,
      "grad_norm": 0.20145000517368317,
      "learning_rate": 4.0973255627040046e-05,
      "loss": 0.0071,
      "mean_token_accuracy": 0.9963594675064087,
      "num_tokens": 108586454.0,
      "step": 745
    },
    {
      "epoch": 0.5110732538330494,
      "grad_norm": 0.11000243574380875,
      "learning_rate": 4.094485562781527e-05,
      "loss": 0.0078,
      "mean_token_accuracy": 0.9965285181999206,
      "num_tokens": 109307025.0,
      "step": 750
    },
    {
      "epoch": 0.514480408858603,
      "grad_norm": 0.060915376991033554,
      "learning_rate": 4.0916273958985135e-05,
      "loss": 0.0052,
      "mean_token_accuracy": 0.9974155962467194,
      "num_tokens": 110041045.0,
      "step": 755
    },
    {
      "epoch": 0.5178875638841567,
      "grad_norm": 0.06734370440244675,
      "learning_rate": 4.088751099000568e-05,
      "loss": 0.0064,
      "mean_token_accuracy": 0.9968310385942459,
      "num_tokens": 110770640.0,
      "step": 760
    },
    {
      "epoch": 0.5212947189097104,
      "grad_norm": 0.046116940677165985,
      "learning_rate": 4.085856709267651e-05,
      "loss": 0.0053,
      "mean_token_accuracy": 0.9971246629953384,
      "num_tokens": 111518848.0,
      "step": 765
    },
    {
      "epoch": 0.524701873935264,
      "grad_norm": 0.16023464500904083,
      "learning_rate": 4.082944264113593e-05,
      "loss": 0.0072,
      "mean_token_accuracy": 0.996779453754425,
      "num_tokens": 112247792.0,
      "step": 770
    },
    {
      "epoch": 0.5281090289608177,
      "grad_norm": 0.14424645900726318,
      "learning_rate": 4.08001380118562e-05,
      "loss": 0.0064,
      "mean_token_accuracy": 0.9965166807174682,
      "num_tokens": 112975212.0,
      "step": 775
    },
    {
      "epoch": 0.5315161839863713,
      "grad_norm": 0.12981460988521576,
      "learning_rate": 4.0770653583638576e-05,
      "loss": 0.0058,
      "mean_token_accuracy": 0.9971657991409302,
      "num_tokens": 113694371.0,
      "step": 780
    },
    {
      "epoch": 0.534923339011925,
      "grad_norm": 0.16971617937088013,
      "learning_rate": 4.074098973760849e-05,
      "loss": 0.0065,
      "mean_token_accuracy": 0.9967111885547638,
      "num_tokens": 114407689.0,
      "step": 785
    },
    {
      "epoch": 0.5383304940374787,
      "grad_norm": 0.10368596017360687,
      "learning_rate": 4.071114685721055e-05,
      "loss": 0.0074,
      "mean_token_accuracy": 0.9963945209980011,
      "num_tokens": 115130620.0,
      "step": 790
    },
    {
      "epoch": 0.5417376490630323,
      "grad_norm": 0.051713522523641586,
      "learning_rate": 4.068112532820364e-05,
      "loss": 0.0058,
      "mean_token_accuracy": 0.9974117010831833,
      "num_tokens": 115864321.0,
      "step": 795
    },
    {
      "epoch": 0.545144804088586,
      "grad_norm": 0.1162973940372467,
      "learning_rate": 4.065092553865592e-05,
      "loss": 0.0072,
      "mean_token_accuracy": 0.9967932134866715,
      "num_tokens": 116598089.0,
      "step": 800
    },
    {
      "epoch": 0.5485519591141397,
      "grad_norm": 0.08385727554559708,
      "learning_rate": 4.06205478789398e-05,
      "loss": 0.0065,
      "mean_token_accuracy": 0.9968057453632355,
      "num_tokens": 117334465.0,
      "step": 805
    },
    {
      "epoch": 0.5519591141396933,
      "grad_norm": 0.11100180447101593,
      "learning_rate": 4.058999274172689e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9965154767036438,
      "num_tokens": 118061885.0,
      "step": 810
    },
    {
      "epoch": 0.555366269165247,
      "grad_norm": 0.09268222004175186,
      "learning_rate": 4.0559260521982936e-05,
      "loss": 0.0061,
      "mean_token_accuracy": 0.9969838321208954,
      "num_tokens": 118797692.0,
      "step": 815
    },
    {
      "epoch": 0.5587734241908007,
      "grad_norm": 0.06716163456439972,
      "learning_rate": 4.052835161696272e-05,
      "loss": 0.0057,
      "mean_token_accuracy": 0.9971336632966995,
      "num_tokens": 119525204.0,
      "step": 820
    },
    {
      "epoch": 0.5621805792163543,
      "grad_norm": 0.0793263241648674,
      "learning_rate": 4.0497266426204905e-05,
      "loss": 0.0058,
      "mean_token_accuracy": 0.9971887588500976,
      "num_tokens": 120257290.0,
      "step": 825
    },
    {
      "epoch": 0.565587734241908,
      "grad_norm": 0.08180940896272659,
      "learning_rate": 4.046600535152689e-05,
      "loss": 0.0064,
      "mean_token_accuracy": 0.9970155090093613,
      "num_tokens": 120983665.0,
      "step": 830
    },
    {
      "epoch": 0.5689948892674617,
      "grad_norm": 0.10042106360197067,
      "learning_rate": 4.04345687970196e-05,
      "loss": 0.0065,
      "mean_token_accuracy": 0.996467211842537,
      "num_tokens": 121702658.0,
      "step": 835
    },
    {
      "epoch": 0.5724020442930153,
      "grad_norm": 0.07629165798425674,
      "learning_rate": 4.040295716904227e-05,
      "loss": 0.0075,
      "mean_token_accuracy": 0.9962219029664994,
      "num_tokens": 122434260.0,
      "step": 840
    },
    {
      "epoch": 0.575809199318569,
      "grad_norm": 0.035190436989068985,
      "learning_rate": 4.037117087621721e-05,
      "loss": 0.0056,
      "mean_token_accuracy": 0.9970576763153076,
      "num_tokens": 123163368.0,
      "step": 845
    },
    {
      "epoch": 0.5792163543441227,
      "grad_norm": 0.038449522107839584,
      "learning_rate": 4.033921032942447e-05,
      "loss": 0.0058,
      "mean_token_accuracy": 0.9972682446241379,
      "num_tokens": 123893557.0,
      "step": 850
    },
    {
      "epoch": 0.5826235093696763,
      "grad_norm": 0.08639010041952133,
      "learning_rate": 4.03070759417966e-05,
      "loss": 0.005,
      "mean_token_accuracy": 0.9972423642873764,
      "num_tokens": 124632464.0,
      "step": 855
    },
    {
      "epoch": 0.58603066439523,
      "grad_norm": 0.0529312863945961,
      "learning_rate": 4.027476812871325e-05,
      "loss": 0.004,
      "mean_token_accuracy": 0.9978580147027969,
      "num_tokens": 125370251.0,
      "step": 860
    },
    {
      "epoch": 0.5894378194207837,
      "grad_norm": 0.09036380797624588,
      "learning_rate": 4.0242287307795846e-05,
      "loss": 0.0074,
      "mean_token_accuracy": 0.996507415175438,
      "num_tokens": 126097603.0,
      "step": 865
    },
    {
      "epoch": 0.5928449744463373,
      "grad_norm": 0.06620994955301285,
      "learning_rate": 4.020963389890214e-05,
      "loss": 0.0061,
      "mean_token_accuracy": 0.9969142705202103,
      "num_tokens": 126831012.0,
      "step": 870
    },
    {
      "epoch": 0.596252129471891,
      "grad_norm": 0.07760175317525864,
      "learning_rate": 4.017680832412084e-05,
      "loss": 0.0059,
      "mean_token_accuracy": 0.9967296481132507,
      "num_tokens": 127552100.0,
      "step": 875
    },
    {
      "epoch": 0.5996592844974447,
      "grad_norm": 0.0371219739317894,
      "learning_rate": 4.014381100776611e-05,
      "loss": 0.0062,
      "mean_token_accuracy": 0.9972570985555649,
      "num_tokens": 128280102.0,
      "step": 880
    },
    {
      "epoch": 0.6030664395229983,
      "grad_norm": 0.04301096498966217,
      "learning_rate": 4.0110642376372114e-05,
      "loss": 0.0052,
      "mean_token_accuracy": 0.9975686013698578,
      "num_tokens": 129011858.0,
      "step": 885
    },
    {
      "epoch": 0.606473594548552,
      "grad_norm": 0.07766426354646683,
      "learning_rate": 4.0077302858687486e-05,
      "loss": 0.0065,
      "mean_token_accuracy": 0.9969834476709366,
      "num_tokens": 129746342.0,
      "step": 890
    },
    {
      "epoch": 0.6098807495741057,
      "grad_norm": 0.13928790390491486,
      "learning_rate": 4.0043792885669795e-05,
      "loss": 0.0064,
      "mean_token_accuracy": 0.996373924612999,
      "num_tokens": 130461868.0,
      "step": 895
    },
    {
      "epoch": 0.6132879045996593,
      "grad_norm": 0.13130903244018555,
      "learning_rate": 4.001011289047996e-05,
      "loss": 0.0067,
      "mean_token_accuracy": 0.9967787474393844,
      "num_tokens": 131187468.0,
      "step": 900
    },
    {
      "epoch": 0.616695059625213,
      "grad_norm": 0.03960848227143288,
      "learning_rate": 3.997626330847668e-05,
      "loss": 0.0071,
      "mean_token_accuracy": 0.9963781535625458,
      "num_tokens": 131912809.0,
      "step": 905
    },
    {
      "epoch": 0.6201022146507666,
      "grad_norm": 0.06704245507717133,
      "learning_rate": 3.994224457721077e-05,
      "loss": 0.0072,
      "mean_token_accuracy": 0.9960544794797898,
      "num_tokens": 132641020.0,
      "step": 910
    },
    {
      "epoch": 0.6235093696763203,
      "grad_norm": 0.07505130767822266,
      "learning_rate": 3.9908057136419536e-05,
      "loss": 0.0072,
      "mean_token_accuracy": 0.9964820086956024,
      "num_tokens": 133364923.0,
      "step": 915
    },
    {
      "epoch": 0.626916524701874,
      "grad_norm": 0.10181338340044022,
      "learning_rate": 3.987370142802109e-05,
      "loss": 0.0082,
      "mean_token_accuracy": 0.9957343071699143,
      "num_tokens": 134085259.0,
      "step": 920
    },
    {
      "epoch": 0.6303236797274276,
      "grad_norm": 0.04167706519365311,
      "learning_rate": 3.983917789610859e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9967679977416992,
      "num_tokens": 134808955.0,
      "step": 925
    },
    {
      "epoch": 0.6337308347529813,
      "grad_norm": 0.05645306780934334,
      "learning_rate": 3.980448698694457e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9970156490802765,
      "num_tokens": 135537312.0,
      "step": 930
    },
    {
      "epoch": 0.637137989778535,
      "grad_norm": 0.06422238796949387,
      "learning_rate": 3.976962914895515e-05,
      "loss": 0.0066,
      "mean_token_accuracy": 0.9968817889690399,
      "num_tokens": 136268523.0,
      "step": 935
    },
    {
      "epoch": 0.6405451448040886,
      "grad_norm": 0.04842214658856392,
      "learning_rate": 3.9734604832724164e-05,
      "loss": 0.0054,
      "mean_token_accuracy": 0.997061750292778,
      "num_tokens": 137002879.0,
      "step": 940
    },
    {
      "epoch": 0.6439522998296422,
      "grad_norm": 0.04557936266064644,
      "learning_rate": 3.969941449098746e-05,
      "loss": 0.0062,
      "mean_token_accuracy": 0.9966436445713043,
      "num_tokens": 137740393.0,
      "step": 945
    },
    {
      "epoch": 0.6473594548551959,
      "grad_norm": 0.04948601499199867,
      "learning_rate": 3.966405857862695e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9971250772476197,
      "num_tokens": 138470335.0,
      "step": 950
    },
    {
      "epoch": 0.6507666098807495,
      "grad_norm": 0.041047368198633194,
      "learning_rate": 3.96285375526648e-05,
      "loss": 0.0059,
      "mean_token_accuracy": 0.996590542793274,
      "num_tokens": 139224387.0,
      "step": 955
    },
    {
      "epoch": 0.6541737649063032,
      "grad_norm": 0.05498833581805229,
      "learning_rate": 3.959285187225746e-05,
      "loss": 0.0052,
      "mean_token_accuracy": 0.9973418712615967,
      "num_tokens": 139952971.0,
      "step": 960
    },
    {
      "epoch": 0.6575809199318569,
      "grad_norm": 0.08102978765964508,
      "learning_rate": 3.9557001998689755e-05,
      "loss": 0.0064,
      "mean_token_accuracy": 0.9969434916973114,
      "num_tokens": 140677739.0,
      "step": 965
    },
    {
      "epoch": 0.6609880749574105,
      "grad_norm": 0.05934149771928787,
      "learning_rate": 3.952098839536895e-05,
      "loss": 0.0057,
      "mean_token_accuracy": 0.9971406131982803,
      "num_tokens": 141400836.0,
      "step": 970
    },
    {
      "epoch": 0.6643952299829642,
      "grad_norm": 0.046910397708415985,
      "learning_rate": 3.948481152781871e-05,
      "loss": 0.0054,
      "mean_token_accuracy": 0.9975356936454773,
      "num_tokens": 142130528.0,
      "step": 975
    },
    {
      "epoch": 0.6678023850085179,
      "grad_norm": 0.05630866438150406,
      "learning_rate": 3.9448471863673134e-05,
      "loss": 0.0054,
      "mean_token_accuracy": 0.9971927255392075,
      "num_tokens": 142865622.0,
      "step": 980
    },
    {
      "epoch": 0.6712095400340715,
      "grad_norm": 0.08571012318134308,
      "learning_rate": 3.941196987267065e-05,
      "loss": 0.0051,
      "mean_token_accuracy": 0.997862359881401,
      "num_tokens": 143598407.0,
      "step": 985
    },
    {
      "epoch": 0.6746166950596252,
      "grad_norm": 0.049263473600149155,
      "learning_rate": 3.9375306026648014e-05,
      "loss": 0.007,
      "mean_token_accuracy": 0.9967271357774734,
      "num_tokens": 144331581.0,
      "step": 990
    },
    {
      "epoch": 0.6780238500851788,
      "grad_norm": 0.05715254694223404,
      "learning_rate": 3.933848079953413e-05,
      "loss": 0.0067,
      "mean_token_accuracy": 0.9966390043497085,
      "num_tokens": 145054612.0,
      "step": 995
    },
    {
      "epoch": 0.6814310051107325,
      "grad_norm": 0.032691799104213715,
      "learning_rate": 3.9301494667344e-05,
      "loss": 0.0062,
      "mean_token_accuracy": 0.9964557588100433,
      "num_tokens": 145777346.0,
      "step": 1000
    },
    {
      "epoch": 0.6848381601362862,
      "grad_norm": 0.055642470717430115,
      "learning_rate": 3.926434810817253e-05,
      "loss": 0.006,
      "mean_token_accuracy": 0.9971215546131134,
      "num_tokens": 146503175.0,
      "step": 1005
    },
    {
      "epoch": 0.6882453151618398,
      "grad_norm": 0.03214164078235626,
      "learning_rate": 3.922704160218836e-05,
      "loss": 0.0042,
      "mean_token_accuracy": 0.9973462909460068,
      "num_tokens": 147220985.0,
      "step": 1010
    },
    {
      "epoch": 0.6916524701873935,
      "grad_norm": 0.030192743986845016,
      "learning_rate": 3.918957563162764e-05,
      "loss": 0.0054,
      "mean_token_accuracy": 0.9974823087453842,
      "num_tokens": 147964108.0,
      "step": 1015
    },
    {
      "epoch": 0.6950596252129472,
      "grad_norm": 0.06393156200647354,
      "learning_rate": 3.915195068078782e-05,
      "loss": 0.0054,
      "mean_token_accuracy": 0.9972763538360596,
      "num_tokens": 148704779.0,
      "step": 1020
    },
    {
      "epoch": 0.6984667802385008,
      "grad_norm": 0.072189100086689,
      "learning_rate": 3.9114167236021396e-05,
      "loss": 0.0056,
      "mean_token_accuracy": 0.9972026377916337,
      "num_tokens": 149426673.0,
      "step": 1025
    },
    {
      "epoch": 0.7018739352640545,
      "grad_norm": 0.06066165864467621,
      "learning_rate": 3.907622578572959e-05,
      "loss": 0.0061,
      "mean_token_accuracy": 0.9965492695569992,
      "num_tokens": 150145480.0,
      "step": 1030
    },
    {
      "epoch": 0.7052810902896082,
      "grad_norm": 0.17103807628154755,
      "learning_rate": 3.903812682035606e-05,
      "loss": 0.0055,
      "mean_token_accuracy": 0.9974225103855133,
      "num_tokens": 150875830.0,
      "step": 1035
    },
    {
      "epoch": 0.7086882453151618,
      "grad_norm": 0.08268236368894577,
      "learning_rate": 3.899987083238056e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.9977911680936813,
      "num_tokens": 151608707.0,
      "step": 1040
    },
    {
      "epoch": 0.7120954003407155,
      "grad_norm": 0.02316095493733883,
      "learning_rate": 3.896145831631256e-05,
      "loss": 0.0056,
      "mean_token_accuracy": 0.9969914197921753,
      "num_tokens": 152337555.0,
      "step": 1045
    },
    {
      "epoch": 0.7155025553662692,
      "grad_norm": 0.029455585405230522,
      "learning_rate": 3.892288976868488e-05,
      "loss": 0.0046,
      "mean_token_accuracy": 0.9974544763565063,
      "num_tokens": 153076673.0,
      "step": 1050
    },
    {
      "epoch": 0.7189097103918228,
      "grad_norm": 0.0412331223487854,
      "learning_rate": 3.888416568804725e-05,
      "loss": 0.0044,
      "mean_token_accuracy": 0.9975585997104645,
      "num_tokens": 153807260.0,
      "step": 1055
    },
    {
      "epoch": 0.7223168654173765,
      "grad_norm": 0.07391174137592316,
      "learning_rate": 3.884528657495985e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.9975113719701767,
      "num_tokens": 154531559.0,
      "step": 1060
    },
    {
      "epoch": 0.7257240204429302,
      "grad_norm": 0.06137605383992195,
      "learning_rate": 3.880625293198689e-05,
      "loss": 0.0051,
      "mean_token_accuracy": 0.9971971452236176,
      "num_tokens": 155266082.0,
      "step": 1065
    },
    {
      "epoch": 0.7291311754684838,
      "grad_norm": 0.10558829456567764,
      "learning_rate": 3.876706526369007e-05,
      "loss": 0.0051,
      "mean_token_accuracy": 0.9973689585924148,
      "num_tokens": 155990556.0,
      "step": 1070
    },
    {
      "epoch": 0.7325383304940375,
      "grad_norm": 0.03859245777130127,
      "learning_rate": 3.872772407662207e-05,
      "loss": 0.0054,
      "mean_token_accuracy": 0.9970339775085449,
      "num_tokens": 156716466.0,
      "step": 1075
    },
    {
      "epoch": 0.7359454855195912,
      "grad_norm": 0.07865729928016663,
      "learning_rate": 3.8688229879320006e-05,
      "loss": 0.0061,
      "mean_token_accuracy": 0.9968847244977951,
      "num_tokens": 157448670.0,
      "step": 1080
    },
    {
      "epoch": 0.7393526405451448,
      "grad_norm": 0.059538476169109344,
      "learning_rate": 3.864858318229886e-05,
      "loss": 0.006,
      "mean_token_accuracy": 0.9969178944826126,
      "num_tokens": 158172408.0,
      "step": 1085
    },
    {
      "epoch": 0.7427597955706985,
      "grad_norm": 0.17625872790813446,
      "learning_rate": 3.8608784498044875e-05,
      "loss": 0.0073,
      "mean_token_accuracy": 0.9964688360691071,
      "num_tokens": 158890068.0,
      "step": 1090
    },
    {
      "epoch": 0.7461669505962522,
      "grad_norm": 0.05374780297279358,
      "learning_rate": 3.856883434100891e-05,
      "loss": 0.0059,
      "mean_token_accuracy": 0.9972751200199127,
      "num_tokens": 159615299.0,
      "step": 1095
    },
    {
      "epoch": 0.7495741056218058,
      "grad_norm": 0.041930485516786575,
      "learning_rate": 3.8528733227599846e-05,
      "loss": 0.0054,
      "mean_token_accuracy": 0.9975884467363357,
      "num_tokens": 160342037.0,
      "step": 1100
    },
    {
      "epoch": 0.7529812606473595,
      "grad_norm": 0.114886075258255,
      "learning_rate": 3.848848167617785e-05,
      "loss": 0.0055,
      "mean_token_accuracy": 0.997040542960167,
      "num_tokens": 161071038.0,
      "step": 1105
    },
    {
      "epoch": 0.7563884156729132,
      "grad_norm": 0.07053513079881668,
      "learning_rate": 3.844808020704771e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9966237008571625,
      "num_tokens": 161792302.0,
      "step": 1110
    },
    {
      "epoch": 0.7597955706984668,
      "grad_norm": 0.0769307017326355,
      "learning_rate": 3.84075293424521e-05,
      "loss": 0.0045,
      "mean_token_accuracy": 0.9976485043764114,
      "num_tokens": 162513971.0,
      "step": 1115
    },
    {
      "epoch": 0.7632027257240205,
      "grad_norm": 0.04618866741657257,
      "learning_rate": 3.8366829606564826e-05,
      "loss": 0.007,
      "mean_token_accuracy": 0.9960559844970703,
      "num_tokens": 163248314.0,
      "step": 1120
    },
    {
      "epoch": 0.7666098807495741,
      "grad_norm": 0.049940288066864014,
      "learning_rate": 3.832598152548405e-05,
      "loss": 0.0056,
      "mean_token_accuracy": 0.9970351666212082,
      "num_tokens": 163970237.0,
      "step": 1125
    },
    {
      "epoch": 0.7700170357751278,
      "grad_norm": 0.07792378216981888,
      "learning_rate": 3.82849856272255e-05,
      "loss": 0.0064,
      "mean_token_accuracy": 0.997080248594284,
      "num_tokens": 164695025.0,
      "step": 1130
    },
    {
      "epoch": 0.7734241908006815,
      "grad_norm": 0.04957449808716774,
      "learning_rate": 3.824384244171564e-05,
      "loss": 0.0058,
      "mean_token_accuracy": 0.9968085557222366,
      "num_tokens": 165423409.0,
      "step": 1135
    },
    {
      "epoch": 0.7768313458262351,
      "grad_norm": 0.07470090687274933,
      "learning_rate": 3.82025525007848e-05,
      "loss": 0.0057,
      "mean_token_accuracy": 0.9974205553531647,
      "num_tokens": 166147294.0,
      "step": 1140
    },
    {
      "epoch": 0.7802385008517888,
      "grad_norm": 0.038353510200977325,
      "learning_rate": 3.816111633816033e-05,
      "loss": 0.0046,
      "mean_token_accuracy": 0.9978825747966766,
      "num_tokens": 166888187.0,
      "step": 1145
    },
    {
      "epoch": 0.7836456558773425,
      "grad_norm": 0.09847115725278854,
      "learning_rate": 3.81195344894597e-05,
      "loss": 0.005,
      "mean_token_accuracy": 0.9971278429031372,
      "num_tokens": 167627595.0,
      "step": 1150
    },
    {
      "epoch": 0.787052810902896,
      "grad_norm": 0.04033125564455986,
      "learning_rate": 3.8077807492183546e-05,
      "loss": 0.0056,
      "mean_token_accuracy": 0.9969434618949891,
      "num_tokens": 168333907.0,
      "step": 1155
    },
    {
      "epoch": 0.7904599659284497,
      "grad_norm": 0.03185822069644928,
      "learning_rate": 3.803593588570877e-05,
      "loss": 0.0051,
      "mean_token_accuracy": 0.9973024189472198,
      "num_tokens": 169068003.0,
      "step": 1160
    },
    {
      "epoch": 0.7938671209540034,
      "grad_norm": 0.06258576363325119,
      "learning_rate": 3.79939202112815e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.9976115971803665,
      "num_tokens": 169788134.0,
      "step": 1165
    },
    {
      "epoch": 0.797274275979557,
      "grad_norm": 0.04610341414809227,
      "learning_rate": 3.795176101201018e-05,
      "loss": 0.0055,
      "mean_token_accuracy": 0.9972164928913116,
      "num_tokens": 170519826.0,
      "step": 1170
    },
    {
      "epoch": 0.8006814310051107,
      "grad_norm": 0.07702849060297012,
      "learning_rate": 3.7909458832858446e-05,
      "loss": 0.0061,
      "mean_token_accuracy": 0.9968161284923553,
      "num_tokens": 171253972.0,
      "step": 1175
    },
    {
      "epoch": 0.8040885860306644,
      "grad_norm": 0.021018194034695625,
      "learning_rate": 3.786701422063821e-05,
      "loss": 0.0047,
      "mean_token_accuracy": 0.9977647632360458,
      "num_tokens": 171991985.0,
      "step": 1180
    },
    {
      "epoch": 0.807495741056218,
      "grad_norm": 0.11837545782327652,
      "learning_rate": 3.782442772400245e-05,
      "loss": 0.006,
      "mean_token_accuracy": 0.9972640424966812,
      "num_tokens": 172709104.0,
      "step": 1185
    },
    {
      "epoch": 0.8109028960817717,
      "grad_norm": 0.05167927220463753,
      "learning_rate": 3.778169989343825e-05,
      "loss": 0.006,
      "mean_token_accuracy": 0.9971501886844635,
      "num_tokens": 173431766.0,
      "step": 1190
    },
    {
      "epoch": 0.8143100511073254,
      "grad_norm": 0.0540495403110981,
      "learning_rate": 3.773883128125958e-05,
      "loss": 0.0066,
      "mean_token_accuracy": 0.9964712619781494,
      "num_tokens": 174175609.0,
      "step": 1195
    },
    {
      "epoch": 0.817717206132879,
      "grad_norm": 0.044089268893003464,
      "learning_rate": 3.769582244160021e-05,
      "loss": 0.0058,
      "mean_token_accuracy": 0.9972156107425689,
      "num_tokens": 174896740.0,
      "step": 1200
    },
    {
      "epoch": 0.8211243611584327,
      "grad_norm": 0.041980892419815063,
      "learning_rate": 3.765267393040655e-05,
      "loss": 0.0048,
      "mean_token_accuracy": 0.9973050326108932,
      "num_tokens": 175633023.0,
      "step": 1205
    },
    {
      "epoch": 0.8245315161839863,
      "grad_norm": 0.04354306310415268,
      "learning_rate": 3.760938630543044e-05,
      "loss": 0.0073,
      "mean_token_accuracy": 0.9960598409175873,
      "num_tokens": 176347248.0,
      "step": 1210
    },
    {
      "epoch": 0.82793867120954,
      "grad_norm": 0.04586596414446831,
      "learning_rate": 3.756596012622194e-05,
      "loss": 0.006,
      "mean_token_accuracy": 0.997346556186676,
      "num_tokens": 177066031.0,
      "step": 1215
    },
    {
      "epoch": 0.8313458262350937,
      "grad_norm": 0.038669832050800323,
      "learning_rate": 3.752239595412212e-05,
      "loss": 0.0052,
      "mean_token_accuracy": 0.9973910361528396,
      "num_tokens": 177802910.0,
      "step": 1220
    },
    {
      "epoch": 0.8347529812606473,
      "grad_norm": 0.05740887671709061,
      "learning_rate": 3.747869435225581e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.997280952334404,
      "num_tokens": 178541629.0,
      "step": 1225
    },
    {
      "epoch": 0.838160136286201,
      "grad_norm": 0.03755781799554825,
      "learning_rate": 3.743485588552426e-05,
      "loss": 0.0048,
      "mean_token_accuracy": 0.9979491472244263,
      "num_tokens": 179261844.0,
      "step": 1230
    },
    {
      "epoch": 0.8415672913117547,
      "grad_norm": 0.027860237285494804,
      "learning_rate": 3.739088112059792e-05,
      "loss": 0.0052,
      "mean_token_accuracy": 0.9978325009346009,
      "num_tokens": 180007470.0,
      "step": 1235
    },
    {
      "epoch": 0.8449744463373083,
      "grad_norm": 0.0710931047797203,
      "learning_rate": 3.734677062590904e-05,
      "loss": 0.0051,
      "mean_token_accuracy": 0.996997407078743,
      "num_tokens": 180727845.0,
      "step": 1240
    },
    {
      "epoch": 0.848381601362862,
      "grad_norm": 0.0614597462117672,
      "learning_rate": 3.7302524971644385e-05,
      "loss": 0.0057,
      "mean_token_accuracy": 0.9972827166318894,
      "num_tokens": 181449186.0,
      "step": 1245
    },
    {
      "epoch": 0.8517887563884157,
      "grad_norm": 0.0660260021686554,
      "learning_rate": 3.725814472973784e-05,
      "loss": 0.0051,
      "mean_token_accuracy": 0.9970994561910629,
      "num_tokens": 182181254.0,
      "step": 1250
    },
    {
      "epoch": 0.8551959114139693,
      "grad_norm": 0.08717916905879974,
      "learning_rate": 3.7213630473862975e-05,
      "loss": 0.0067,
      "mean_token_accuracy": 0.9969981014728546,
      "num_tokens": 182906609.0,
      "step": 1255
    },
    {
      "epoch": 0.858603066439523,
      "grad_norm": 0.07495622336864471,
      "learning_rate": 3.716898277942571e-05,
      "loss": 0.0053,
      "mean_token_accuracy": 0.9972448527812958,
      "num_tokens": 183635377.0,
      "step": 1260
    },
    {
      "epoch": 0.8620102214650767,
      "grad_norm": 0.05844198912382126,
      "learning_rate": 3.712420222355683e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9966474562883377,
      "num_tokens": 184362912.0,
      "step": 1265
    },
    {
      "epoch": 0.8654173764906303,
      "grad_norm": 0.041810475289821625,
      "learning_rate": 3.7079289385104504e-05,
      "loss": 0.004,
      "mean_token_accuracy": 0.9977830231189728,
      "num_tokens": 185088332.0,
      "step": 1270
    },
    {
      "epoch": 0.868824531516184,
      "grad_norm": 0.04615911468863487,
      "learning_rate": 3.703424484462686e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.997277969121933,
      "num_tokens": 185818265.0,
      "step": 1275
    },
    {
      "epoch": 0.8722316865417377,
      "grad_norm": 0.11199091374874115,
      "learning_rate": 3.698906918438443e-05,
      "loss": 0.0055,
      "mean_token_accuracy": 0.9965817183256149,
      "num_tokens": 186552002.0,
      "step": 1280
    },
    {
      "epoch": 0.8756388415672913,
      "grad_norm": 0.08411674946546555,
      "learning_rate": 3.694376298833265e-05,
      "loss": 0.006,
      "mean_token_accuracy": 0.9970198780298233,
      "num_tokens": 187272311.0,
      "step": 1285
    },
    {
      "epoch": 0.879045996592845,
      "grad_norm": 0.04271921515464783,
      "learning_rate": 3.6898326842114305e-05,
      "loss": 0.0065,
      "mean_token_accuracy": 0.996277117729187,
      "num_tokens": 188000922.0,
      "step": 1290
    },
    {
      "epoch": 0.8824531516183987,
      "grad_norm": 0.05022366717457771,
      "learning_rate": 3.685276133305196e-05,
      "loss": 0.0052,
      "mean_token_accuracy": 0.9972202897071838,
      "num_tokens": 188724304.0,
      "step": 1295
    },
    {
      "epoch": 0.8858603066439523,
      "grad_norm": 0.06338661909103394,
      "learning_rate": 3.680706705014036e-05,
      "loss": 0.0052,
      "mean_token_accuracy": 0.9970058828592301,
      "num_tokens": 189467176.0,
      "step": 1300
    },
    {
      "epoch": 0.889267461669506,
      "grad_norm": 0.04164168983697891,
      "learning_rate": 3.6761244584038836e-05,
      "loss": 0.0055,
      "mean_token_accuracy": 0.9972519338130951,
      "num_tokens": 190208885.0,
      "step": 1305
    },
    {
      "epoch": 0.8926746166950597,
      "grad_norm": 0.04781107231974602,
      "learning_rate": 3.671529452706364e-05,
      "loss": 0.0059,
      "mean_token_accuracy": 0.9969022154808045,
      "num_tokens": 190933373.0,
      "step": 1310
    },
    {
      "epoch": 0.8960817717206133,
      "grad_norm": 0.05542631074786186,
      "learning_rate": 3.666921747318031e-05,
      "loss": 0.0056,
      "mean_token_accuracy": 0.9969410121440887,
      "num_tokens": 191656704.0,
      "step": 1315
    },
    {
      "epoch": 0.899488926746167,
      "grad_norm": 0.0386984758079052,
      "learning_rate": 3.662301401799601e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9969573467969894,
      "num_tokens": 192379867.0,
      "step": 1320
    },
    {
      "epoch": 0.9028960817717206,
      "grad_norm": 0.03840241581201553,
      "learning_rate": 3.657668475875177e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.9974152386188507,
      "num_tokens": 193094134.0,
      "step": 1325
    },
    {
      "epoch": 0.9063032367972743,
      "grad_norm": 0.06262340396642685,
      "learning_rate": 3.6530230294314845e-05,
      "loss": 0.0071,
      "mean_token_accuracy": 0.9961749941110611,
      "num_tokens": 193816011.0,
      "step": 1330
    },
    {
      "epoch": 0.909710391822828,
      "grad_norm": 0.05942515283823013,
      "learning_rate": 3.6483651225170906e-05,
      "loss": 0.0059,
      "mean_token_accuracy": 0.9972467362880707,
      "num_tokens": 194543017.0,
      "step": 1335
    },
    {
      "epoch": 0.9131175468483816,
      "grad_norm": 0.03937466815114021,
      "learning_rate": 3.6436948153416316e-05,
      "loss": 0.0048,
      "mean_token_accuracy": 0.9978303790092469,
      "num_tokens": 195275759.0,
      "step": 1340
    },
    {
      "epoch": 0.9165247018739353,
      "grad_norm": 0.05984418839216232,
      "learning_rate": 3.6390121682750324e-05,
      "loss": 0.0051,
      "mean_token_accuracy": 0.9976150929927826,
      "num_tokens": 196013305.0,
      "step": 1345
    },
    {
      "epoch": 0.919931856899489,
      "grad_norm": 0.0860673189163208,
      "learning_rate": 3.634317241846729e-05,
      "loss": 0.0045,
      "mean_token_accuracy": 0.9978837996721268,
      "num_tokens": 196740094.0,
      "step": 1350
    },
    {
      "epoch": 0.9233390119250426,
      "grad_norm": 0.059223659336566925,
      "learning_rate": 3.6296100967448845e-05,
      "loss": 0.0066,
      "mean_token_accuracy": 0.9974291831254959,
      "num_tokens": 197465304.0,
      "step": 1355
    },
    {
      "epoch": 0.9267461669505963,
      "grad_norm": 0.04743910953402519,
      "learning_rate": 3.6248907938156024e-05,
      "loss": 0.0058,
      "mean_token_accuracy": 0.9968450695276261,
      "num_tokens": 198181028.0,
      "step": 1360
    },
    {
      "epoch": 0.9301533219761499,
      "grad_norm": 0.04319678619503975,
      "learning_rate": 3.620159394062145e-05,
      "loss": 0.005,
      "mean_token_accuracy": 0.9973400145769119,
      "num_tokens": 198916749.0,
      "step": 1365
    },
    {
      "epoch": 0.9335604770017035,
      "grad_norm": 0.053205423057079315,
      "learning_rate": 3.615415958644141e-05,
      "loss": 0.0057,
      "mean_token_accuracy": 0.9972663849592209,
      "num_tokens": 199650032.0,
      "step": 1370
    },
    {
      "epoch": 0.9369676320272572,
      "grad_norm": 0.029109084978699684,
      "learning_rate": 3.610660548876795e-05,
      "loss": 0.0061,
      "mean_token_accuracy": 0.9969484329223632,
      "num_tokens": 200378622.0,
      "step": 1375
    },
    {
      "epoch": 0.9403747870528109,
      "grad_norm": 0.06647433340549469,
      "learning_rate": 3.605893226230099e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.9973088294267655,
      "num_tokens": 201104269.0,
      "step": 1380
    },
    {
      "epoch": 0.9437819420783645,
      "grad_norm": 0.0842168927192688,
      "learning_rate": 3.6011140523280306e-05,
      "loss": 0.0056,
      "mean_token_accuracy": 0.9967386871576309,
      "num_tokens": 201843493.0,
      "step": 1385
    },
    {
      "epoch": 0.9471890971039182,
      "grad_norm": 0.06159935146570206,
      "learning_rate": 3.5963230889477636e-05,
      "loss": 0.005,
      "mean_token_accuracy": 0.997000876069069,
      "num_tokens": 202552662.0,
      "step": 1390
    },
    {
      "epoch": 0.9505962521294719,
      "grad_norm": 0.05091740936040878,
      "learning_rate": 3.591520398018867e-05,
      "loss": 0.0062,
      "mean_token_accuracy": 0.9968457281589508,
      "num_tokens": 203274960.0,
      "step": 1395
    },
    {
      "epoch": 0.9540034071550255,
      "grad_norm": 0.07609888911247253,
      "learning_rate": 3.5867060416224995e-05,
      "loss": 0.0061,
      "mean_token_accuracy": 0.9963427126407624,
      "num_tokens": 203985778.0,
      "step": 1400
    },
    {
      "epoch": 0.9574105621805792,
      "grad_norm": 0.030158163979649544,
      "learning_rate": 3.581880081990617e-05,
      "loss": 0.005,
      "mean_token_accuracy": 0.9973898887634277,
      "num_tokens": 204724083.0,
      "step": 1405
    },
    {
      "epoch": 0.9608177172061328,
      "grad_norm": 0.09108585864305496,
      "learning_rate": 3.577042581505159e-05,
      "loss": 0.006,
      "mean_token_accuracy": 0.9973899811506272,
      "num_tokens": 205451123.0,
      "step": 1410
    },
    {
      "epoch": 0.9642248722316865,
      "grad_norm": 0.01543778833001852,
      "learning_rate": 3.572193602697249e-05,
      "loss": 0.0053,
      "mean_token_accuracy": 0.9976825684309005,
      "num_tokens": 206198692.0,
      "step": 1415
    },
    {
      "epoch": 0.9676320272572402,
      "grad_norm": 0.04157339781522751,
      "learning_rate": 3.567333208246379e-05,
      "loss": 0.0053,
      "mean_token_accuracy": 0.9976014912128448,
      "num_tokens": 206934417.0,
      "step": 1420
    },
    {
      "epoch": 0.9710391822827938,
      "grad_norm": 0.026260005310177803,
      "learning_rate": 3.562461460979607e-05,
      "loss": 0.0055,
      "mean_token_accuracy": 0.9967951595783233,
      "num_tokens": 207651564.0,
      "step": 1425
    },
    {
      "epoch": 0.9744463373083475,
      "grad_norm": 0.05099419876933098,
      "learning_rate": 3.5575784238707385e-05,
      "loss": 0.0062,
      "mean_token_accuracy": 0.9968614369630814,
      "num_tokens": 208372673.0,
      "step": 1430
    },
    {
      "epoch": 0.9778534923339012,
      "grad_norm": 0.054835014045238495,
      "learning_rate": 3.5526841600395167e-05,
      "loss": 0.0056,
      "mean_token_accuracy": 0.9975211530923843,
      "num_tokens": 209103059.0,
      "step": 1435
    },
    {
      "epoch": 0.9812606473594548,
      "grad_norm": 0.05773494765162468,
      "learning_rate": 3.547778732750806e-05,
      "loss": 0.0056,
      "mean_token_accuracy": 0.9972155272960663,
      "num_tokens": 209838434.0,
      "step": 1440
    },
    {
      "epoch": 0.9846678023850085,
      "grad_norm": 0.025797275826334953,
      "learning_rate": 3.5428622054137706e-05,
      "loss": 0.0056,
      "mean_token_accuracy": 0.9971807062625885,
      "num_tokens": 210570859.0,
      "step": 1445
    },
    {
      "epoch": 0.9880749574105622,
      "grad_norm": 0.05218036100268364,
      "learning_rate": 3.5379346415810596e-05,
      "loss": 0.0063,
      "mean_token_accuracy": 0.9963378429412841,
      "num_tokens": 211294347.0,
      "step": 1450
    },
    {
      "epoch": 0.9914821124361158,
      "grad_norm": 0.038821663707494736,
      "learning_rate": 3.532996104947985e-05,
      "loss": 0.0058,
      "mean_token_accuracy": 0.9975549668073654,
      "num_tokens": 212020280.0,
      "step": 1455
    },
    {
      "epoch": 0.9948892674616695,
      "grad_norm": 0.05104764550924301,
      "learning_rate": 3.528046659351692e-05,
      "loss": 0.0067,
      "mean_token_accuracy": 0.9965088933706283,
      "num_tokens": 212735703.0,
      "step": 1460
    },
    {
      "epoch": 0.9982964224872232,
      "grad_norm": 0.07035092264413834,
      "learning_rate": 3.523086368770342e-05,
      "loss": 0.0047,
      "mean_token_accuracy": 0.9978703796863556,
      "num_tokens": 213468511.0,
      "step": 1465
    },
    {
      "epoch": 1.0013628620102215,
      "grad_norm": 0.09284400939941406,
      "learning_rate": 3.5181152973222835e-05,
      "loss": 0.0065,
      "mean_token_accuracy": 0.9972891211509705,
      "num_tokens": 214122368.0,
      "step": 1470
    },
    {
      "epoch": 1.0047700170357752,
      "grad_norm": 0.09326088428497314,
      "learning_rate": 3.513133509265219e-05,
      "loss": 0.0059,
      "mean_token_accuracy": 0.9976144641637802,
      "num_tokens": 214844511.0,
      "step": 1475
    },
    {
      "epoch": 1.0081771720613288,
      "grad_norm": 0.07183246314525604,
      "learning_rate": 3.5081410689953784e-05,
      "loss": 0.0056,
      "mean_token_accuracy": 0.9978006035089493,
      "num_tokens": 215565405.0,
      "step": 1480
    },
    {
      "epoch": 1.0115843270868825,
      "grad_norm": 0.04884856566786766,
      "learning_rate": 3.503138041046686e-05,
      "loss": 0.0068,
      "mean_token_accuracy": 0.9968368977308273,
      "num_tokens": 216295640.0,
      "step": 1485
    },
    {
      "epoch": 1.0149914821124362,
      "grad_norm": 0.0549909770488739,
      "learning_rate": 3.4981244900899266e-05,
      "loss": 0.0045,
      "mean_token_accuracy": 0.997943839430809,
      "num_tokens": 217023400.0,
      "step": 1490
    },
    {
      "epoch": 1.0183986371379898,
      "grad_norm": 0.03975891321897507,
      "learning_rate": 3.4931004809319075e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.9976144343614578,
      "num_tokens": 217749356.0,
      "step": 1495
    },
    {
      "epoch": 1.0218057921635435,
      "grad_norm": 0.017692692577838898,
      "learning_rate": 3.488066078514622e-05,
      "loss": 0.0051,
      "mean_token_accuracy": 0.9976278364658355,
      "num_tokens": 218480277.0,
      "step": 1500
    },
    {
      "epoch": 1.0252129471890972,
      "grad_norm": 0.039218395948410034,
      "learning_rate": 3.483021347914413e-05,
      "loss": 0.0047,
      "mean_token_accuracy": 0.9976595848798752,
      "num_tokens": 219216195.0,
      "step": 1505
    },
    {
      "epoch": 1.0286201022146508,
      "grad_norm": 0.08224453032016754,
      "learning_rate": 3.477966354341124e-05,
      "loss": 0.0046,
      "mean_token_accuracy": 0.9971513509750366,
      "num_tokens": 219950008.0,
      "step": 1510
    },
    {
      "epoch": 1.0320272572402045,
      "grad_norm": 0.040846455842256546,
      "learning_rate": 3.472901163137264e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.9978175193071366,
      "num_tokens": 220678308.0,
      "step": 1515
    },
    {
      "epoch": 1.0354344122657582,
      "grad_norm": 0.03683318942785263,
      "learning_rate": 3.467825839777163e-05,
      "loss": 0.0044,
      "mean_token_accuracy": 0.9976907223463058,
      "num_tokens": 221392685.0,
      "step": 1520
    },
    {
      "epoch": 1.0388415672913118,
      "grad_norm": 0.026718707755208015,
      "learning_rate": 3.462740449866115e-05,
      "loss": 0.0044,
      "mean_token_accuracy": 0.9973954141139985,
      "num_tokens": 222122241.0,
      "step": 1525
    },
    {
      "epoch": 1.0422487223168655,
      "grad_norm": 0.060078807175159454,
      "learning_rate": 3.4576450591395436e-05,
      "loss": 0.0046,
      "mean_token_accuracy": 0.9973968118429184,
      "num_tokens": 222856120.0,
      "step": 1530
    },
    {
      "epoch": 1.0456558773424192,
      "grad_norm": 0.03545363247394562,
      "learning_rate": 3.452539733462145e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.9975829154253006,
      "num_tokens": 223582950.0,
      "step": 1535
    },
    {
      "epoch": 1.0490630323679728,
      "grad_norm": 0.023809373378753662,
      "learning_rate": 3.447424538827037e-05,
      "loss": 0.0037,
      "mean_token_accuracy": 0.9979015290737152,
      "num_tokens": 224304965.0,
      "step": 1540
    },
    {
      "epoch": 1.0524701873935265,
      "grad_norm": 0.03015109896659851,
      "learning_rate": 3.442299541354906e-05,
      "loss": 0.0053,
      "mean_token_accuracy": 0.9978000223636627,
      "num_tokens": 225031852.0,
      "step": 1545
    },
    {
      "epoch": 1.0558773424190802,
      "grad_norm": 0.09108318388462067,
      "learning_rate": 3.4371648072931555e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.9980493158102035,
      "num_tokens": 225766001.0,
      "step": 1550
    },
    {
      "epoch": 1.0592844974446338,
      "grad_norm": 0.041708365082740784,
      "learning_rate": 3.432020403015045e-05,
      "loss": 0.0055,
      "mean_token_accuracy": 0.9969852536916732,
      "num_tokens": 226486873.0,
      "step": 1555
    },
    {
      "epoch": 1.0626916524701875,
      "grad_norm": 0.05419030785560608,
      "learning_rate": 3.426866395018836e-05,
      "loss": 0.0057,
      "mean_token_accuracy": 0.9971535056829453,
      "num_tokens": 227226782.0,
      "step": 1560
    },
    {
      "epoch": 1.0660988074957412,
      "grad_norm": 0.020873723551630974,
      "learning_rate": 3.4217028499269304e-05,
      "loss": 0.0045,
      "mean_token_accuracy": 0.9975880146026611,
      "num_tokens": 227962015.0,
      "step": 1565
    },
    {
      "epoch": 1.0695059625212948,
      "grad_norm": 0.03190511837601662,
      "learning_rate": 3.416529834485008e-05,
      "loss": 0.0055,
      "mean_token_accuracy": 0.997125095129013,
      "num_tokens": 228689495.0,
      "step": 1570
    },
    {
      "epoch": 1.0729131175468485,
      "grad_norm": 0.038219112902879715,
      "learning_rate": 3.411347415561169e-05,
      "loss": 0.0052,
      "mean_token_accuracy": 0.9972829192876815,
      "num_tokens": 229430137.0,
      "step": 1575
    },
    {
      "epoch": 1.0763202725724021,
      "grad_norm": 0.05260247364640236,
      "learning_rate": 3.406155660145063e-05,
      "loss": 0.0051,
      "mean_token_accuracy": 0.9974386841058731,
      "num_tokens": 230157030.0,
      "step": 1580
    },
    {
      "epoch": 1.0797274275979558,
      "grad_norm": 0.05825897306203842,
      "learning_rate": 3.400954635347028e-05,
      "loss": 0.0049,
      "mean_token_accuracy": 0.997426375746727,
      "num_tokens": 230884465.0,
      "step": 1585
    },
    {
      "epoch": 1.0831345826235093,
      "grad_norm": 0.034084614366292953,
      "learning_rate": 3.395744408397219e-05,
      "loss": 0.0047,
      "mean_token_accuracy": 0.9975792735815048,
      "num_tokens": 231603606.0,
      "step": 1590
    },
    {
      "epoch": 1.0865417376490631,
      "grad_norm": 0.053616102784872055,
      "learning_rate": 3.3905250466447436e-05,
      "loss": 0.0061,
      "mean_token_accuracy": 0.9969032853841782,
      "num_tokens": 232330755.0,
      "step": 1595
    },
    {
      "epoch": 1.0899488926746166,
      "grad_norm": 0.03791968896985054,
      "learning_rate": 3.385296617556788e-05,
      "loss": 0.0036,
      "mean_token_accuracy": 0.9983588516712188,
      "num_tokens": 233060107.0,
      "step": 1600
    }
  ],
  "logging_steps": 5,
  "max_steps": 4404,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3449394411137925e+19,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}