{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9974102848686646,
  "eval_steps": 100,
  "global_step": 1011,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0029596744358120607,
      "grad_norm": 0.16762526333332062,
      "learning_rate": 1.9607843137254904e-07,
      "loss": 1.0892,
      "mean_token_accuracy": 0.7134666713588034,
      "step": 1
    },
    {
      "epoch": 0.0059193488716241215,
      "grad_norm": 0.15522930026054382,
      "learning_rate": 3.921568627450981e-07,
      "loss": 1.0859,
      "mean_token_accuracy": 0.7130540900903558,
      "step": 2
    },
    {
      "epoch": 0.008879023307436182,
      "grad_norm": 0.1598200649023056,
      "learning_rate": 5.882352941176471e-07,
      "loss": 1.1084,
      "mean_token_accuracy": 0.7057607800943448,
      "step": 3
    },
    {
      "epoch": 0.011838697743248243,
      "grad_norm": 0.16198603808879852,
      "learning_rate": 7.843137254901962e-07,
      "loss": 1.1252,
      "mean_token_accuracy": 0.7020035438412923,
      "step": 4
    },
    {
      "epoch": 0.014798372179060304,
      "grad_norm": 0.16225993633270264,
      "learning_rate": 9.80392156862745e-07,
      "loss": 1.0616,
      "mean_token_accuracy": 0.720360737247177,
      "step": 5
    },
    {
      "epoch": 0.017758046614872364,
      "grad_norm": 0.15833528339862823,
      "learning_rate": 1.1764705882352942e-06,
      "loss": 1.15,
      "mean_token_accuracy": 0.6989465627144322,
      "step": 6
    },
    {
      "epoch": 0.020717721050684423,
      "grad_norm": 0.15328213572502136,
      "learning_rate": 1.3725490196078434e-06,
      "loss": 1.0692,
      "mean_token_accuracy": 0.7184121167108843,
      "step": 7
    },
    {
      "epoch": 0.023677395486496486,
      "grad_norm": 0.1418345421552658,
      "learning_rate": 1.5686274509803923e-06,
      "loss": 1.1028,
      "mean_token_accuracy": 0.7096372641647224,
      "step": 8
    },
    {
      "epoch": 0.026637069922308545,
      "grad_norm": 0.1514817476272583,
      "learning_rate": 1.7647058823529414e-06,
      "loss": 1.1374,
      "mean_token_accuracy": 0.7002159446531399,
      "step": 9
    },
    {
      "epoch": 0.029596744358120607,
      "grad_norm": 0.1426001787185669,
      "learning_rate": 1.96078431372549e-06,
      "loss": 1.0702,
      "mean_token_accuracy": 0.7191991800747065,
      "step": 10
    },
    {
      "epoch": 0.032556418793932666,
      "grad_norm": 0.1281445175409317,
      "learning_rate": 2.1568627450980393e-06,
      "loss": 1.0694,
      "mean_token_accuracy": 0.71691801950529,
      "step": 11
    },
    {
      "epoch": 0.03551609322974473,
      "grad_norm": 0.1249074786901474,
      "learning_rate": 2.3529411764705885e-06,
      "loss": 1.0593,
      "mean_token_accuracy": 0.7178305997624074,
      "step": 12
    },
    {
      "epoch": 0.03847576766555679,
      "grad_norm": 0.12432456761598587,
      "learning_rate": 2.549019607843137e-06,
      "loss": 1.0596,
      "mean_token_accuracy": 0.7183453209613041,
      "step": 13
    },
    {
      "epoch": 0.04143544210136885,
      "grad_norm": 0.11549073457717896,
      "learning_rate": 2.7450980392156867e-06,
      "loss": 1.0325,
      "mean_token_accuracy": 0.7258331228051237,
      "step": 14
    },
    {
      "epoch": 0.04439511653718091,
      "grad_norm": 0.09011338651180267,
      "learning_rate": 2.9411764705882355e-06,
      "loss": 1.0511,
      "mean_token_accuracy": 0.718035698553114,
      "step": 15
    },
    {
      "epoch": 0.04735479097299297,
      "grad_norm": 0.08878307044506073,
      "learning_rate": 3.1372549019607846e-06,
      "loss": 1.0827,
      "mean_token_accuracy": 0.7073935642252629,
      "step": 16
    },
    {
      "epoch": 0.050314465408805034,
      "grad_norm": 0.08482303470373154,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0326,
      "mean_token_accuracy": 0.7214633725074016,
      "step": 17
    },
    {
      "epoch": 0.05327413984461709,
      "grad_norm": 0.08270992338657379,
      "learning_rate": 3.529411764705883e-06,
      "loss": 1.0698,
      "mean_token_accuracy": 0.7109063855863134,
      "step": 18
    },
    {
      "epoch": 0.05623381428042915,
      "grad_norm": 0.08383882790803909,
      "learning_rate": 3.7254901960784316e-06,
      "loss": 1.0637,
      "mean_token_accuracy": 0.7127537261513002,
      "step": 19
    },
    {
      "epoch": 0.059193488716241215,
      "grad_norm": 0.06499592959880829,
      "learning_rate": 3.92156862745098e-06,
      "loss": 1.0019,
      "mean_token_accuracy": 0.7259147749731993,
      "step": 20
    },
    {
      "epoch": 0.06215316315205328,
      "grad_norm": 0.07826568931341171,
      "learning_rate": 4.11764705882353e-06,
      "loss": 1.0081,
      "mean_token_accuracy": 0.7232646130331987,
      "step": 21
    },
    {
      "epoch": 0.06511283758786533,
      "grad_norm": 0.08427096158266068,
      "learning_rate": 4.313725490196079e-06,
      "loss": 0.9901,
      "mean_token_accuracy": 0.7260792816120228,
      "step": 22
    },
    {
      "epoch": 0.0680725120236774,
      "grad_norm": 0.08436324447393417,
      "learning_rate": 4.509803921568628e-06,
      "loss": 0.9666,
      "mean_token_accuracy": 0.7345528010123215,
      "step": 23
    },
    {
      "epoch": 0.07103218645948946,
      "grad_norm": 0.08232813328504562,
      "learning_rate": 4.705882352941177e-06,
      "loss": 0.9673,
      "mean_token_accuracy": 0.7324904968946799,
      "step": 24
    },
    {
      "epoch": 0.07399186089530152,
      "grad_norm": 0.07241534441709518,
      "learning_rate": 4.901960784313726e-06,
      "loss": 0.9912,
      "mean_token_accuracy": 0.7251441637707999,
      "step": 25
    },
    {
      "epoch": 0.07695153533111358,
      "grad_norm": 0.06652858108282089,
      "learning_rate": 5.098039215686274e-06,
      "loss": 1.013,
      "mean_token_accuracy": 0.7185447007427124,
      "step": 26
    },
    {
      "epoch": 0.07991120976692564,
      "grad_norm": 0.05553431063890457,
      "learning_rate": 5.294117647058824e-06,
      "loss": 0.9626,
      "mean_token_accuracy": 0.7310570228170078,
      "step": 27
    },
    {
      "epoch": 0.0828708842027377,
      "grad_norm": 0.05903884023427963,
      "learning_rate": 5.4901960784313735e-06,
      "loss": 0.9626,
      "mean_token_accuracy": 0.7309964084777121,
      "step": 28
    },
    {
      "epoch": 0.08583055863854976,
      "grad_norm": 0.06741348654031754,
      "learning_rate": 5.686274509803922e-06,
      "loss": 0.9659,
      "mean_token_accuracy": 0.7304780944549785,
      "step": 29
    },
    {
      "epoch": 0.08879023307436182,
      "grad_norm": 0.07335620373487473,
      "learning_rate": 5.882352941176471e-06,
      "loss": 0.9528,
      "mean_token_accuracy": 0.735141297162976,
      "step": 30
    },
    {
      "epoch": 0.09174990751017388,
      "grad_norm": 0.0699823796749115,
      "learning_rate": 6.07843137254902e-06,
      "loss": 0.9788,
      "mean_token_accuracy": 0.7230978860942294,
      "step": 31
    },
    {
      "epoch": 0.09470958194598594,
      "grad_norm": 0.057243529707193375,
      "learning_rate": 6.274509803921569e-06,
      "loss": 0.8985,
      "mean_token_accuracy": 0.7457556690066438,
      "step": 32
    },
    {
      "epoch": 0.097669256381798,
      "grad_norm": 0.054736681282520294,
      "learning_rate": 6.470588235294119e-06,
      "loss": 0.9344,
      "mean_token_accuracy": 0.7364832933769531,
      "step": 33
    },
    {
      "epoch": 0.10062893081761007,
      "grad_norm": 0.04590689763426781,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9022,
      "mean_token_accuracy": 0.7468134852529875,
      "step": 34
    },
    {
      "epoch": 0.10358860525342212,
      "grad_norm": 0.04902254790067673,
      "learning_rate": 6.862745098039216e-06,
      "loss": 0.871,
      "mean_token_accuracy": 0.7504876513230005,
      "step": 35
    },
    {
      "epoch": 0.10654827968923418,
      "grad_norm": 0.04902803525328636,
      "learning_rate": 7.058823529411766e-06,
      "loss": 0.8905,
      "mean_token_accuracy": 0.7462919644706687,
      "step": 36
    },
    {
      "epoch": 0.10950795412504624,
      "grad_norm": 0.051301341503858566,
      "learning_rate": 7.2549019607843145e-06,
      "loss": 0.8811,
      "mean_token_accuracy": 0.7487371815251612,
      "step": 37
    },
    {
      "epoch": 0.1124676285608583,
      "grad_norm": 0.04607445001602173,
      "learning_rate": 7.450980392156863e-06,
      "loss": 0.8995,
      "mean_token_accuracy": 0.7445922596551929,
      "step": 38
    },
    {
      "epoch": 0.11542730299667037,
      "grad_norm": 0.04521537572145462,
      "learning_rate": 7.647058823529411e-06,
      "loss": 0.863,
      "mean_token_accuracy": 0.7525979239300858,
      "step": 39
    },
    {
      "epoch": 0.11838697743248243,
      "grad_norm": 0.03848574310541153,
      "learning_rate": 7.84313725490196e-06,
      "loss": 0.8953,
      "mean_token_accuracy": 0.7438669082059773,
      "step": 40
    },
    {
      "epoch": 0.12134665186829449,
      "grad_norm": 0.04125151038169861,
      "learning_rate": 8.03921568627451e-06,
      "loss": 0.8951,
      "mean_token_accuracy": 0.7446916010263452,
      "step": 41
    },
    {
      "epoch": 0.12430632630410655,
      "grad_norm": 0.039782147854566574,
      "learning_rate": 8.23529411764706e-06,
      "loss": 0.898,
      "mean_token_accuracy": 0.7417729737072621,
      "step": 42
    },
    {
      "epoch": 0.12726600073991862,
      "grad_norm": 0.0405898317694664,
      "learning_rate": 8.43137254901961e-06,
      "loss": 0.846,
      "mean_token_accuracy": 0.7583980485632862,
      "step": 43
    },
    {
      "epoch": 0.13022567517573067,
      "grad_norm": 0.03678755462169647,
      "learning_rate": 8.627450980392157e-06,
      "loss": 0.8774,
      "mean_token_accuracy": 0.747737609406417,
      "step": 44
    },
    {
      "epoch": 0.13318534961154274,
      "grad_norm": 0.03588786721229553,
      "learning_rate": 8.823529411764707e-06,
      "loss": 0.892,
      "mean_token_accuracy": 0.7425907739306867,
      "step": 45
    },
    {
      "epoch": 0.1361450240473548,
      "grad_norm": 0.035192012786865234,
      "learning_rate": 9.019607843137256e-06,
      "loss": 0.8359,
      "mean_token_accuracy": 0.7598674536604115,
      "step": 46
    },
    {
      "epoch": 0.13910469848316684,
      "grad_norm": 0.03437209874391556,
      "learning_rate": 9.215686274509804e-06,
      "loss": 0.8408,
      "mean_token_accuracy": 0.7562485513629261,
      "step": 47
    },
    {
      "epoch": 0.14206437291897892,
      "grad_norm": 0.039065975695848465,
      "learning_rate": 9.411764705882354e-06,
      "loss": 0.8563,
      "mean_token_accuracy": 0.7523758470154138,
      "step": 48
    },
    {
      "epoch": 0.14502404735479096,
      "grad_norm": 0.03827141225337982,
      "learning_rate": 9.607843137254903e-06,
      "loss": 0.8831,
      "mean_token_accuracy": 0.7441134515720478,
      "step": 49
    },
    {
      "epoch": 0.14798372179060304,
      "grad_norm": 0.038179270923137665,
      "learning_rate": 9.803921568627451e-06,
      "loss": 0.8679,
      "mean_token_accuracy": 0.7485590506876966,
      "step": 50
    },
    {
      "epoch": 0.1509433962264151,
      "grad_norm": 0.03490907698869705,
      "learning_rate": 1e-05,
      "loss": 0.8911,
      "mean_token_accuracy": 0.7421432801131297,
      "step": 51
    },
    {
      "epoch": 0.15390307066222716,
      "grad_norm": 0.032188620418310165,
      "learning_rate": 1.0196078431372549e-05,
      "loss": 0.8333,
      "mean_token_accuracy": 0.7583880230838789,
      "step": 52
    },
    {
      "epoch": 0.1568627450980392,
      "grad_norm": 0.034065958112478256,
      "learning_rate": 1.03921568627451e-05,
      "loss": 0.8796,
      "mean_token_accuracy": 0.7459146148009512,
      "step": 53
    },
    {
      "epoch": 0.1598224195338513,
      "grad_norm": 0.0375240258872509,
      "learning_rate": 1.0588235294117648e-05,
      "loss": 0.8748,
      "mean_token_accuracy": 0.745663908179616,
      "step": 54
    },
    {
      "epoch": 0.16278209396966334,
      "grad_norm": 0.03509582206606865,
      "learning_rate": 1.0784313725490196e-05,
      "loss": 0.8219,
      "mean_token_accuracy": 0.7617277634792063,
      "step": 55
    },
    {
      "epoch": 0.1657417684054754,
      "grad_norm": 0.031145978718996048,
      "learning_rate": 1.0980392156862747e-05,
      "loss": 0.8281,
      "mean_token_accuracy": 0.7589892108618271,
      "step": 56
    },
    {
      "epoch": 0.16870144284128746,
      "grad_norm": 0.0317564457654953,
      "learning_rate": 1.1176470588235295e-05,
      "loss": 0.8503,
      "mean_token_accuracy": 0.7520432527990282,
      "step": 57
    },
    {
      "epoch": 0.1716611172770995,
      "grad_norm": 0.0326925553381443,
      "learning_rate": 1.1372549019607844e-05,
      "loss": 0.8557,
      "mean_token_accuracy": 0.7526200242465962,
      "step": 58
    },
    {
      "epoch": 0.1746207917129116,
      "grad_norm": 0.03252090886235237,
      "learning_rate": 1.1568627450980394e-05,
      "loss": 0.8228,
      "mean_token_accuracy": 0.7611397936393716,
      "step": 59
    },
    {
      "epoch": 0.17758046614872364,
      "grad_norm": 0.03402877226471901,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.8618,
      "mean_token_accuracy": 0.7489423826424529,
      "step": 60
    },
    {
      "epoch": 0.1805401405845357,
      "grad_norm": 0.031763315200805664,
      "learning_rate": 1.1960784313725491e-05,
      "loss": 0.8232,
      "mean_token_accuracy": 0.7577058865835121,
      "step": 61
    },
    {
      "epoch": 0.18349981502034776,
      "grad_norm": 0.03277522698044777,
      "learning_rate": 1.215686274509804e-05,
      "loss": 0.8468,
      "mean_token_accuracy": 0.751973355587776,
      "step": 62
    },
    {
      "epoch": 0.1864594894561598,
      "grad_norm": 0.03080877847969532,
      "learning_rate": 1.235294117647059e-05,
      "loss": 0.8275,
      "mean_token_accuracy": 0.7575937164187081,
      "step": 63
    },
    {
      "epoch": 0.1894191638919719,
      "grad_norm": 0.03214840590953827,
      "learning_rate": 1.2549019607843138e-05,
      "loss": 0.8512,
      "mean_token_accuracy": 0.7504745296737662,
      "step": 64
    },
    {
      "epoch": 0.19237883832778394,
      "grad_norm": 0.02780843898653984,
      "learning_rate": 1.2745098039215686e-05,
      "loss": 0.8004,
      "mean_token_accuracy": 0.7650133400923254,
      "step": 65
    },
    {
      "epoch": 0.195338512763596,
      "grad_norm": 0.030489617958664894,
      "learning_rate": 1.2941176470588238e-05,
      "loss": 0.8417,
      "mean_token_accuracy": 0.751373090730022,
      "step": 66
    },
    {
      "epoch": 0.19829818719940806,
      "grad_norm": 0.030239971354603767,
      "learning_rate": 1.3137254901960785e-05,
      "loss": 0.801,
      "mean_token_accuracy": 0.7657282535175904,
      "step": 67
    },
    {
      "epoch": 0.20125786163522014,
      "grad_norm": 0.030548961833119392,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.8236,
      "mean_token_accuracy": 0.7588368671361572,
      "step": 68
    },
    {
      "epoch": 0.20421753607103219,
      "grad_norm": 0.03228336572647095,
      "learning_rate": 1.3529411764705885e-05,
      "loss": 0.8389,
      "mean_token_accuracy": 0.75389222307077,
      "step": 69
    },
    {
      "epoch": 0.20717721050684423,
      "grad_norm": 0.031210200861096382,
      "learning_rate": 1.3725490196078432e-05,
      "loss": 0.8427,
      "mean_token_accuracy": 0.7538669036916475,
      "step": 70
    },
    {
      "epoch": 0.2101368849426563,
      "grad_norm": 0.03246346861124039,
      "learning_rate": 1.392156862745098e-05,
      "loss": 0.7855,
      "mean_token_accuracy": 0.7686714193639335,
      "step": 71
    },
    {
      "epoch": 0.21309655937846836,
      "grad_norm": 0.033433880656957626,
      "learning_rate": 1.4117647058823532e-05,
      "loss": 0.8237,
      "mean_token_accuracy": 0.7583954136539371,
      "step": 72
    },
    {
      "epoch": 0.21605623381428044,
      "grad_norm": 0.02908838354051113,
      "learning_rate": 1.431372549019608e-05,
      "loss": 0.7897,
      "mean_token_accuracy": 0.7670688477944143,
      "step": 73
    },
    {
      "epoch": 0.21901590825009248,
      "grad_norm": 0.029152696952223778,
      "learning_rate": 1.4509803921568629e-05,
      "loss": 0.8158,
      "mean_token_accuracy": 0.759069492682175,
      "step": 74
    },
    {
      "epoch": 0.22197558268590456,
      "grad_norm": 0.02914673648774624,
      "learning_rate": 1.4705882352941179e-05,
      "loss": 0.8246,
      "mean_token_accuracy": 0.7565679063863135,
      "step": 75
    },
    {
      "epoch": 0.2249352571217166,
      "grad_norm": 0.031528692692518234,
      "learning_rate": 1.4901960784313726e-05,
      "loss": 0.7983,
      "mean_token_accuracy": 0.7653819572853421,
      "step": 76
    },
    {
      "epoch": 0.22789493155752868,
      "grad_norm": 0.03318175673484802,
      "learning_rate": 1.5098039215686276e-05,
      "loss": 0.8051,
      "mean_token_accuracy": 0.7626766998590977,
      "step": 77
    },
    {
      "epoch": 0.23085460599334073,
      "grad_norm": 0.03181340545415878,
      "learning_rate": 1.5294117647058822e-05,
      "loss": 0.8314,
      "mean_token_accuracy": 0.7562527153037738,
      "step": 78
    },
    {
      "epoch": 0.23381428042915278,
      "grad_norm": 0.030570238828659058,
      "learning_rate": 1.5490196078431373e-05,
      "loss": 0.8098,
      "mean_token_accuracy": 0.7612361942973398,
      "step": 79
    },
    {
      "epoch": 0.23677395486496486,
      "grad_norm": 0.03228567913174629,
      "learning_rate": 1.568627450980392e-05,
      "loss": 0.8313,
      "mean_token_accuracy": 0.7553067383295745,
      "step": 80
    },
    {
      "epoch": 0.2397336293007769,
      "grad_norm": 0.028338870033621788,
      "learning_rate": 1.5882352941176473e-05,
      "loss": 0.7833,
      "mean_token_accuracy": 0.7682124908200851,
      "step": 81
    },
    {
      "epoch": 0.24269330373658898,
      "grad_norm": 0.029070047661662102,
      "learning_rate": 1.607843137254902e-05,
      "loss": 0.8024,
      "mean_token_accuracy": 0.7619718358937104,
      "step": 82
    },
    {
      "epoch": 0.24565297817240103,
      "grad_norm": 0.030931631103157997,
      "learning_rate": 1.627450980392157e-05,
      "loss": 0.8406,
      "mean_token_accuracy": 0.7543044976811706,
      "step": 83
    },
    {
      "epoch": 0.2486126526082131,
      "grad_norm": 0.03312648832798004,
      "learning_rate": 1.647058823529412e-05,
      "loss": 0.8356,
      "mean_token_accuracy": 0.7525359946108627,
      "step": 84
    },
    {
      "epoch": 0.25157232704402516,
      "grad_norm": 0.029619930312037468,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.8198,
      "mean_token_accuracy": 0.7574176097969649,
      "step": 85
    },
    {
      "epoch": 0.25453200147983723,
      "grad_norm": 0.030434174463152885,
      "learning_rate": 1.686274509803922e-05,
      "loss": 0.7725,
      "mean_token_accuracy": 0.7706778885187738,
      "step": 86
    },
    {
      "epoch": 0.25749167591564925,
      "grad_norm": 0.03060806915163994,
      "learning_rate": 1.7058823529411767e-05,
      "loss": 0.7619,
      "mean_token_accuracy": 0.7759168992180178,
      "step": 87
    },
    {
      "epoch": 0.26045135035146133,
      "grad_norm": 0.03023487515747547,
      "learning_rate": 1.7254901960784314e-05,
      "loss": 0.7719,
      "mean_token_accuracy": 0.7711198031069662,
      "step": 88
    },
    {
      "epoch": 0.2634110247872734,
      "grad_norm": 0.03228628635406494,
      "learning_rate": 1.7450980392156866e-05,
      "loss": 0.8326,
      "mean_token_accuracy": 0.7541014395136457,
      "step": 89
    },
    {
      "epoch": 0.2663706992230855,
      "grad_norm": 0.030876586213707924,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 0.809,
      "mean_token_accuracy": 0.7600534063596812,
      "step": 90
    },
    {
      "epoch": 0.2693303736588975,
      "grad_norm": 0.030030572786927223,
      "learning_rate": 1.7843137254901965e-05,
      "loss": 0.8093,
      "mean_token_accuracy": 0.7603780399368074,
      "step": 91
    },
    {
      "epoch": 0.2722900480947096,
      "grad_norm": 0.03146005794405937,
      "learning_rate": 1.8039215686274513e-05,
      "loss": 0.8255,
      "mean_token_accuracy": 0.7566988371936587,
      "step": 92
    },
    {
      "epoch": 0.27524972253052166,
      "grad_norm": 0.031429581344127655,
      "learning_rate": 1.823529411764706e-05,
      "loss": 0.7499,
      "mean_token_accuracy": 0.776043410115711,
      "step": 93
    },
    {
      "epoch": 0.2782093969663337,
      "grad_norm": 0.02945755235850811,
      "learning_rate": 1.843137254901961e-05,
      "loss": 0.8001,
      "mean_token_accuracy": 0.7639983029478665,
      "step": 94
    },
    {
      "epoch": 0.28116907140214575,
      "grad_norm": 0.03111177310347557,
      "learning_rate": 1.862745098039216e-05,
      "loss": 0.7978,
      "mean_token_accuracy": 0.7633804190251812,
      "step": 95
    },
    {
      "epoch": 0.28412874583795783,
      "grad_norm": 0.032041724771261215,
      "learning_rate": 1.8823529411764708e-05,
      "loss": 0.7835,
      "mean_token_accuracy": 0.7653787019038502,
      "step": 96
    },
    {
      "epoch": 0.2870884202737699,
      "grad_norm": 0.030550945550203323,
      "learning_rate": 1.9019607843137255e-05,
      "loss": 0.8153,
      "mean_token_accuracy": 0.7571907561061959,
      "step": 97
    },
    {
      "epoch": 0.2900480947095819,
      "grad_norm": 0.030825288966298103,
      "learning_rate": 1.9215686274509807e-05,
      "loss": 0.7778,
      "mean_token_accuracy": 0.769872313705694,
      "step": 98
    },
    {
      "epoch": 0.293007769145394,
      "grad_norm": 0.03140871599316597,
      "learning_rate": 1.9411764705882355e-05,
      "loss": 0.8065,
      "mean_token_accuracy": 0.7614259731399569,
      "step": 99
    },
    {
      "epoch": 0.2959674435812061,
      "grad_norm": 0.03500783070921898,
      "learning_rate": 1.9607843137254903e-05,
      "loss": 0.8073,
      "mean_token_accuracy": 0.7603054008149568,
      "step": 100
    },
    {
      "epoch": 0.2959674435812061,
      "eval_loss": 0.8200445771217346,
      "eval_mean_token_accuracy": 0.7534407818769884,
      "eval_runtime": 24.7077,
      "eval_samples_per_second": 5.221,
      "eval_steps_per_second": 1.336,
      "step": 100
    },
    {
      "epoch": 0.2989271180170181,
      "grad_norm": 0.0325748473405838,
      "learning_rate": 1.9803921568627454e-05,
      "loss": 0.8204,
      "mean_token_accuracy": 0.7574438522520155,
      "step": 101
    },
    {
      "epoch": 0.3018867924528302,
      "grad_norm": 0.035838086158037186,
      "learning_rate": 2e-05,
      "loss": 0.7885,
      "mean_token_accuracy": 0.7662196280575948,
      "step": 102
    },
    {
      "epoch": 0.30484646688864225,
      "grad_norm": 0.03514498472213745,
      "learning_rate": 1.9999940277008807e-05,
      "loss": 0.8109,
      "mean_token_accuracy": 0.7592109922581233,
      "step": 103
    },
    {
      "epoch": 0.30780614132445433,
      "grad_norm": 0.030624302104115486,
      "learning_rate": 1.99997611087486e-05,
      "loss": 0.7923,
      "mean_token_accuracy": 0.7646137676651172,
      "step": 104
    },
    {
      "epoch": 0.31076581576026635,
      "grad_norm": 0.03328769654035568,
      "learning_rate": 1.9999462497359468e-05,
      "loss": 0.7618,
      "mean_token_accuracy": 0.7724664009420006,
      "step": 105
    },
    {
      "epoch": 0.3137254901960784,
      "grad_norm": 0.032032452523708344,
      "learning_rate": 1.9999044446408203e-05,
      "loss": 0.7794,
      "mean_token_accuracy": 0.7676382888194131,
      "step": 106
    },
    {
      "epoch": 0.3166851646318905,
      "grad_norm": 0.038596246391534805,
      "learning_rate": 1.9998506960888258e-05,
      "loss": 0.8055,
      "mean_token_accuracy": 0.7598404595982178,
      "step": 107
    },
    {
      "epoch": 0.3196448390677026,
      "grad_norm": 0.03659771382808685,
      "learning_rate": 1.999785004721968e-05,
      "loss": 0.7996,
      "mean_token_accuracy": 0.7617447897080775,
      "step": 108
    },
    {
      "epoch": 0.3226045135035146,
      "grad_norm": 0.03398187831044197,
      "learning_rate": 1.999707371324904e-05,
      "loss": 0.8159,
      "mean_token_accuracy": 0.7579143887098169,
      "step": 109
    },
    {
      "epoch": 0.3255641879393267,
      "grad_norm": 0.034746043384075165,
      "learning_rate": 1.9996177968249336e-05,
      "loss": 0.7447,
      "mean_token_accuracy": 0.7767408029730972,
      "step": 110
    },
    {
      "epoch": 0.32852386237513875,
      "grad_norm": 0.034224774688482285,
      "learning_rate": 1.999516282291988e-05,
      "loss": 0.8118,
      "mean_token_accuracy": 0.7602916024979172,
      "step": 111
    },
    {
      "epoch": 0.3314835368109508,
      "grad_norm": 0.03755398839712143,
      "learning_rate": 1.999402828938618e-05,
      "loss": 0.8064,
      "mean_token_accuracy": 0.7601404942437737,
      "step": 112
    },
    {
      "epoch": 0.33444321124676285,
      "grad_norm": 0.03318261727690697,
      "learning_rate": 1.999277438119978e-05,
      "loss": 0.7842,
      "mean_token_accuracy": 0.7667829450005377,
      "step": 113
    },
    {
      "epoch": 0.3374028856825749,
      "grad_norm": 0.03273628279566765,
      "learning_rate": 1.9991401113338103e-05,
      "loss": 0.8084,
      "mean_token_accuracy": 0.7604015155881398,
      "step": 114
    },
    {
      "epoch": 0.340362560118387,
      "grad_norm": 0.03294860199093819,
      "learning_rate": 1.9989908502204295e-05,
      "loss": 0.7853,
      "mean_token_accuracy": 0.7652222709961788,
      "step": 115
    },
    {
      "epoch": 0.343322234554199,
      "grad_norm": 0.03144688904285431,
      "learning_rate": 1.9988296565626988e-05,
      "loss": 0.7633,
      "mean_token_accuracy": 0.771602340876925,
      "step": 116
    },
    {
      "epoch": 0.3462819089900111,
      "grad_norm": 0.03655346855521202,
      "learning_rate": 1.9986565322860117e-05,
      "loss": 0.8287,
      "mean_token_accuracy": 0.7540871834712267,
      "step": 117
    },
    {
      "epoch": 0.3492415834258232,
      "grad_norm": 0.03457004949450493,
      "learning_rate": 1.9984714794582682e-05,
      "loss": 0.8229,
      "mean_token_accuracy": 0.7544258713699461,
      "step": 118
    },
    {
      "epoch": 0.3522012578616352,
      "grad_norm": 0.03338496387004852,
      "learning_rate": 1.99827450028985e-05,
      "loss": 0.7869,
      "mean_token_accuracy": 0.7650590695958719,
      "step": 119
    },
    {
      "epoch": 0.3551609322974473,
      "grad_norm": 0.032108377665281296,
      "learning_rate": 1.9980655971335944e-05,
      "loss": 0.8162,
      "mean_token_accuracy": 0.758311915938832,
      "step": 120
    },
    {
      "epoch": 0.35812060673325935,
      "grad_norm": 0.0332258902490139,
      "learning_rate": 1.9978447724847655e-05,
      "loss": 0.8016,
      "mean_token_accuracy": 0.7604372139446551,
      "step": 121
    },
    {
      "epoch": 0.3610802811690714,
      "grad_norm": 0.03431647643446922,
      "learning_rate": 1.9976120289810247e-05,
      "loss": 0.817,
      "mean_token_accuracy": 0.7565450471641515,
      "step": 122
    },
    {
      "epoch": 0.36403995560488345,
      "grad_norm": 0.03282129764556885,
      "learning_rate": 1.9973673694024002e-05,
      "loss": 0.7913,
      "mean_token_accuracy": 0.7636492958717122,
      "step": 123
    },
    {
      "epoch": 0.3669996300406955,
      "grad_norm": 0.03169738128781319,
      "learning_rate": 1.9971107966712518e-05,
      "loss": 0.7519,
      "mean_token_accuracy": 0.7740747801822188,
      "step": 124
    },
    {
      "epoch": 0.3699593044765076,
      "grad_norm": 0.03179707005620003,
      "learning_rate": 1.9968423138522382e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.7666833966716972,
      "step": 125
    },
    {
      "epoch": 0.3729189789123196,
      "grad_norm": 0.032772284001111984,
      "learning_rate": 1.996561924152278e-05,
      "loss": 0.8061,
      "mean_token_accuracy": 0.7606085219927862,
      "step": 126
    },
    {
      "epoch": 0.3758786533481317,
      "grad_norm": 0.03113570809364319,
      "learning_rate": 1.9962696309205146e-05,
      "loss": 0.7834,
      "mean_token_accuracy": 0.7672297772401319,
      "step": 127
    },
    {
      "epoch": 0.3788383277839438,
      "grad_norm": 0.030488718301057816,
      "learning_rate": 1.995965437648273e-05,
      "loss": 0.7818,
      "mean_token_accuracy": 0.7655031341083235,
      "step": 128
    },
    {
      "epoch": 0.38179800221975585,
      "grad_norm": 0.031420253217220306,
      "learning_rate": 1.995649347969019e-05,
      "loss": 0.7507,
      "mean_token_accuracy": 0.7733494279033952,
      "step": 129
    },
    {
      "epoch": 0.38475767665556787,
      "grad_norm": 0.03062237985432148,
      "learning_rate": 1.995321365658317e-05,
      "loss": 0.8055,
      "mean_token_accuracy": 0.7586643497349167,
      "step": 130
    },
    {
      "epoch": 0.38771735109137995,
      "grad_norm": 0.038481730967760086,
      "learning_rate": 1.994981494633784e-05,
      "loss": 0.8032,
      "mean_token_accuracy": 0.7589717154545749,
      "step": 131
    },
    {
      "epoch": 0.390677025527192,
      "grad_norm": 0.03341325372457504,
      "learning_rate": 1.9946297389550433e-05,
      "loss": 0.8052,
      "mean_token_accuracy": 0.7596351234251352,
      "step": 132
    },
    {
      "epoch": 0.39363669996300404,
      "grad_norm": 0.033091627061367035,
      "learning_rate": 1.9942661028236746e-05,
      "loss": 0.7924,
      "mean_token_accuracy": 0.7640783045123354,
      "step": 133
    },
    {
      "epoch": 0.3965963743988161,
      "grad_norm": 0.03521052747964859,
      "learning_rate": 1.9938905905831657e-05,
      "loss": 0.7896,
      "mean_token_accuracy": 0.7640436688330697,
      "step": 134
    },
    {
      "epoch": 0.3995560488346282,
      "grad_norm": 0.03265683352947235,
      "learning_rate": 1.993503206718859e-05,
      "loss": 0.7752,
      "mean_token_accuracy": 0.7686113466804103,
      "step": 135
    },
    {
      "epoch": 0.4025157232704403,
      "grad_norm": 0.03359968215227127,
      "learning_rate": 1.9931039558578997e-05,
      "loss": 0.7605,
      "mean_token_accuracy": 0.7727853653709049,
      "step": 136
    },
    {
      "epoch": 0.4054753977062523,
      "grad_norm": 0.033013999462127686,
      "learning_rate": 1.9926928427691788e-05,
      "loss": 0.7382,
      "mean_token_accuracy": 0.7783862904202015,
      "step": 137
    },
    {
      "epoch": 0.40843507214206437,
      "grad_norm": 0.03365851938724518,
      "learning_rate": 1.992269872363277e-05,
      "loss": 0.7845,
      "mean_token_accuracy": 0.7661636540016182,
      "step": 138
    },
    {
      "epoch": 0.41139474657787645,
      "grad_norm": 0.03333881497383118,
      "learning_rate": 1.991835049692405e-05,
      "loss": 0.7643,
      "mean_token_accuracy": 0.7699004379989075,
      "step": 139
    },
    {
      "epoch": 0.41435442101368847,
      "grad_norm": 0.04011340066790581,
      "learning_rate": 1.991388379950346e-05,
      "loss": 0.7608,
      "mean_token_accuracy": 0.7714213787682671,
      "step": 140
    },
    {
      "epoch": 0.41731409544950054,
      "grad_norm": 0.030173150822520256,
      "learning_rate": 1.9909298684723905e-05,
      "loss": 0.7748,
      "mean_token_accuracy": 0.7668684509979319,
      "step": 141
    },
    {
      "epoch": 0.4202737698853126,
      "grad_norm": 0.029973339289426804,
      "learning_rate": 1.9904595207352736e-05,
      "loss": 0.7639,
      "mean_token_accuracy": 0.7699766148227036,
      "step": 142
    },
    {
      "epoch": 0.4232334443211247,
      "grad_norm": 0.035035885870456696,
      "learning_rate": 1.9899773423571102e-05,
      "loss": 0.7475,
      "mean_token_accuracy": 0.7755862537612642,
      "step": 143
    },
    {
      "epoch": 0.4261931187569367,
      "grad_norm": 0.033620089292526245,
      "learning_rate": 1.9894833390973266e-05,
      "loss": 0.8153,
      "mean_token_accuracy": 0.7567288748378098,
      "step": 144
    },
    {
      "epoch": 0.4291527931927488,
      "grad_norm": 0.030382465571165085,
      "learning_rate": 1.9889775168565942e-05,
      "loss": 0.7797,
      "mean_token_accuracy": 0.767362073302904,
      "step": 145
    },
    {
      "epoch": 0.43211246762856087,
      "grad_norm": 0.03639741614460945,
      "learning_rate": 1.9884598816767563e-05,
      "loss": 0.8099,
      "mean_token_accuracy": 0.7583598824820752,
      "step": 146
    },
    {
      "epoch": 0.43507214206437295,
      "grad_norm": 0.033540308475494385,
      "learning_rate": 1.987930439740757e-05,
      "loss": 0.7581,
      "mean_token_accuracy": 0.7727086981336495,
      "step": 147
    },
    {
      "epoch": 0.43803181650018497,
      "grad_norm": 0.03171444311738014,
      "learning_rate": 1.9873891973725673e-05,
      "loss": 0.7566,
      "mean_token_accuracy": 0.7747774133710759,
      "step": 148
    },
    {
      "epoch": 0.44099149093599704,
      "grad_norm": 0.030361434444785118,
      "learning_rate": 1.98683616103711e-05,
      "loss": 0.7671,
      "mean_token_accuracy": 0.7702413004460634,
      "step": 149
    },
    {
      "epoch": 0.4439511653718091,
      "grad_norm": 0.030990034341812134,
      "learning_rate": 1.986271337340182e-05,
      "loss": 0.7808,
      "mean_token_accuracy": 0.7653441489524639,
      "step": 150
    },
    {
      "epoch": 0.44691083980762114,
      "grad_norm": 0.0336206890642643,
      "learning_rate": 1.9856947330283752e-05,
      "loss": 0.7466,
      "mean_token_accuracy": 0.7738241553325558,
      "step": 151
    },
    {
      "epoch": 0.4498705142434332,
      "grad_norm": 0.03193899616599083,
      "learning_rate": 1.985106354988997e-05,
      "loss": 0.7565,
      "mean_token_accuracy": 0.770569745061826,
      "step": 152
    },
    {
      "epoch": 0.4528301886792453,
      "grad_norm": 0.03213547170162201,
      "learning_rate": 1.984506210249986e-05,
      "loss": 0.7423,
      "mean_token_accuracy": 0.7773415450825415,
      "step": 153
    },
    {
      "epoch": 0.45578986311505737,
      "grad_norm": 0.029736842960119247,
      "learning_rate": 1.9838943059798305e-05,
      "loss": 0.7675,
      "mean_token_accuracy": 0.7703503542161088,
      "step": 154
    },
    {
      "epoch": 0.4587495375508694,
      "grad_norm": 0.036596570163965225,
      "learning_rate": 1.9832706494874812e-05,
      "loss": 0.7898,
      "mean_token_accuracy": 0.763933543416774,
      "step": 155
    },
    {
      "epoch": 0.46170921198668147,
      "grad_norm": 0.03609054908156395,
      "learning_rate": 1.982635248222264e-05,
      "loss": 0.8192,
      "mean_token_accuracy": 0.753698721940512,
      "step": 156
    },
    {
      "epoch": 0.46466888642249354,
      "grad_norm": 0.03234592452645302,
      "learning_rate": 1.9819881097737917e-05,
      "loss": 0.7801,
      "mean_token_accuracy": 0.7656625837258381,
      "step": 157
    },
    {
      "epoch": 0.46762856085830556,
      "grad_norm": 0.036302950233221054,
      "learning_rate": 1.9813292418718734e-05,
      "loss": 0.8232,
      "mean_token_accuracy": 0.7549797273411322,
      "step": 158
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 0.033823274075984955,
      "learning_rate": 1.9806586523864212e-05,
      "loss": 0.7838,
      "mean_token_accuracy": 0.7643192081109611,
      "step": 159
    },
    {
      "epoch": 0.4735479097299297,
      "grad_norm": 0.03188293054699898,
      "learning_rate": 1.9799763493273572e-05,
      "loss": 0.7497,
      "mean_token_accuracy": 0.7747934222001206,
      "step": 160
    },
    {
      "epoch": 0.4765075841657418,
      "grad_norm": 0.03294404223561287,
      "learning_rate": 1.9792823408445173e-05,
      "loss": 0.7842,
      "mean_token_accuracy": 0.7655804595799393,
      "step": 161
    },
    {
      "epoch": 0.4794672586015538,
      "grad_norm": 0.03698144853115082,
      "learning_rate": 1.978576635227554e-05,
      "loss": 0.7749,
      "mean_token_accuracy": 0.7673428027423348,
      "step": 162
    },
    {
      "epoch": 0.4824269330373659,
      "grad_norm": 0.03282702714204788,
      "learning_rate": 1.9778592409058376e-05,
      "loss": 0.7539,
      "mean_token_accuracy": 0.7742788469344546,
      "step": 163
    },
    {
      "epoch": 0.48538660747317797,
      "grad_norm": 0.033910758793354034,
      "learning_rate": 1.9771301664483548e-05,
      "loss": 0.768,
      "mean_token_accuracy": 0.7694873822487575,
      "step": 164
    },
    {
      "epoch": 0.48834628190899,
      "grad_norm": 0.0356927327811718,
      "learning_rate": 1.976389420563607e-05,
      "loss": 0.7679,
      "mean_token_accuracy": 0.7698343825834856,
      "step": 165
    },
    {
      "epoch": 0.49130595634480206,
      "grad_norm": 0.03199487179517746,
      "learning_rate": 1.975637012099507e-05,
      "loss": 0.7508,
      "mean_token_accuracy": 0.7741949459067823,
      "step": 166
    },
    {
      "epoch": 0.49426563078061414,
      "grad_norm": 0.03319091349840164,
      "learning_rate": 1.97487295004327e-05,
      "loss": 0.7978,
      "mean_token_accuracy": 0.760116249903723,
      "step": 167
    },
    {
      "epoch": 0.4972253052164262,
      "grad_norm": 0.03303578495979309,
      "learning_rate": 1.9740972435213114e-05,
      "loss": 0.797,
      "mean_token_accuracy": 0.7613210071158084,
      "step": 168
    },
    {
      "epoch": 0.5001849796522383,
      "grad_norm": 0.03376881778240204,
      "learning_rate": 1.9733099017991342e-05,
      "loss": 0.7916,
      "mean_token_accuracy": 0.7619543804125304,
      "step": 169
    },
    {
      "epoch": 0.5031446540880503,
      "grad_norm": 0.031745847314596176,
      "learning_rate": 1.972510934281218e-05,
      "loss": 0.7683,
      "mean_token_accuracy": 0.7699950251417537,
      "step": 170
    },
    {
      "epoch": 0.5061043285238623,
      "grad_norm": 0.03418955206871033,
      "learning_rate": 1.9717003505109097e-05,
      "loss": 0.8034,
      "mean_token_accuracy": 0.7578885558201578,
      "step": 171
    },
    {
      "epoch": 0.5090640029596745,
      "grad_norm": 0.0315784327685833,
      "learning_rate": 1.9708781601703066e-05,
      "loss": 0.7673,
      "mean_token_accuracy": 0.768625610334258,
      "step": 172
    },
    {
      "epoch": 0.5120236773954865,
      "grad_norm": 0.030939601361751556,
      "learning_rate": 1.9700443730801412e-05,
      "loss": 0.7694,
      "mean_token_accuracy": 0.7694912459598255,
      "step": 173
    },
    {
      "epoch": 0.5149833518312985,
      "grad_norm": 0.035644952207803726,
      "learning_rate": 1.9691989991996663e-05,
      "loss": 0.7683,
      "mean_token_accuracy": 0.7670837518034129,
      "step": 174
    },
    {
      "epoch": 0.5179430262671106,
      "grad_norm": 0.03552411496639252,
      "learning_rate": 1.9683420486265328e-05,
      "loss": 0.7694,
      "mean_token_accuracy": 0.7701203461477576,
      "step": 175
    },
    {
      "epoch": 0.5209027007029227,
      "grad_norm": 0.0308453980833292,
      "learning_rate": 1.967473531596671e-05,
      "loss": 0.7559,
      "mean_token_accuracy": 0.772178540695049,
      "step": 176
    },
    {
      "epoch": 0.5238623751387348,
      "grad_norm": 0.033064551651477814,
      "learning_rate": 1.966593458484168e-05,
      "loss": 0.7761,
      "mean_token_accuracy": 0.7676401076309817,
      "step": 177
    },
    {
      "epoch": 0.5268220495745468,
      "grad_norm": 0.03514972701668739,
      "learning_rate": 1.9657018398011435e-05,
      "loss": 0.7721,
      "mean_token_accuracy": 0.7678882657093409,
      "step": 178
    },
    {
      "epoch": 0.5297817240103588,
      "grad_norm": 0.03215213119983673,
      "learning_rate": 1.9647986861976246e-05,
      "loss": 0.7774,
      "mean_token_accuracy": 0.7681080094947464,
      "step": 179
    },
    {
      "epoch": 0.532741398446171,
      "grad_norm": 0.034506384283304214,
      "learning_rate": 1.9638840084614182e-05,
      "loss": 0.7249,
      "mean_token_accuracy": 0.7819193656782826,
      "step": 180
    },
    {
      "epoch": 0.535701072881983,
      "grad_norm": 0.035173896700143814,
      "learning_rate": 1.9629578175179823e-05,
      "loss": 0.7629,
      "mean_token_accuracy": 0.7710785158328923,
      "step": 181
    },
    {
      "epoch": 0.538660747317795,
      "grad_norm": 0.03143983334302902,
      "learning_rate": 1.9620201244302952e-05,
      "loss": 0.7523,
      "mean_token_accuracy": 0.7738275868869572,
      "step": 182
    },
    {
      "epoch": 0.5416204217536071,
      "grad_norm": 0.03433418273925781,
      "learning_rate": 1.9610709403987248e-05,
      "loss": 0.7625,
      "mean_token_accuracy": 0.7716964302162167,
      "step": 183
    },
    {
      "epoch": 0.5445800961894192,
      "grad_norm": 0.03088780678808689,
      "learning_rate": 1.9601102767608924e-05,
      "loss": 0.777,
      "mean_token_accuracy": 0.7660960314455441,
      "step": 184
    },
    {
      "epoch": 0.5475397706252312,
      "grad_norm": 0.033923640847206116,
      "learning_rate": 1.95913814499154e-05,
      "loss": 0.78,
      "mean_token_accuracy": 0.7651355540613629,
      "step": 185
    },
    {
      "epoch": 0.5504994450610433,
      "grad_norm": 0.03130833059549332,
      "learning_rate": 1.95815455670239e-05,
      "loss": 0.7845,
      "mean_token_accuracy": 0.7638504411686589,
      "step": 186
    },
    {
      "epoch": 0.5534591194968553,
      "grad_norm": 0.03128202632069588,
      "learning_rate": 1.9571595236420103e-05,
      "loss": 0.7695,
      "mean_token_accuracy": 0.7674976238469747,
      "step": 187
    },
    {
      "epoch": 0.5564187939326674,
      "grad_norm": 0.02871876023709774,
      "learning_rate": 1.9561530576956703e-05,
      "loss": 0.7335,
      "mean_token_accuracy": 0.7774578028670619,
      "step": 188
    },
    {
      "epoch": 0.5593784683684795,
      "grad_norm": 0.03246740996837616,
      "learning_rate": 1.955135170885202e-05,
      "loss": 0.7467,
      "mean_token_accuracy": 0.7755275648723853,
      "step": 189
    },
    {
      "epoch": 0.5623381428042915,
      "grad_norm": 0.033313024789094925,
      "learning_rate": 1.9541058753688538e-05,
      "loss": 0.7516,
      "mean_token_accuracy": 0.7735669161702174,
      "step": 190
    },
    {
      "epoch": 0.5652978172401036,
      "grad_norm": 0.03222542256116867,
      "learning_rate": 1.9530651834411477e-05,
      "loss": 0.7643,
      "mean_token_accuracy": 0.7690212391129012,
      "step": 191
    },
    {
      "epoch": 0.5682574916759157,
      "grad_norm": 0.032517172396183014,
      "learning_rate": 1.95201310753273e-05,
      "loss": 0.7267,
      "mean_token_accuracy": 0.7777439806072167,
      "step": 192
    },
    {
      "epoch": 0.5712171661117277,
      "grad_norm": 0.02986653707921505,
      "learning_rate": 1.9509496602102253e-05,
      "loss": 0.7296,
      "mean_token_accuracy": 0.7795480557007181,
      "step": 193
    },
    {
      "epoch": 0.5741768405475398,
      "grad_norm": 0.030299223959445953,
      "learning_rate": 1.9498748541760845e-05,
      "loss": 0.7443,
      "mean_token_accuracy": 0.7743541396332051,
      "step": 194
    },
    {
      "epoch": 0.5771365149833518,
      "grad_norm": 0.035258956253528595,
      "learning_rate": 1.9487887022684336e-05,
      "loss": 0.7648,
      "mean_token_accuracy": 0.7696968324451735,
      "step": 195
    },
    {
      "epoch": 0.5800961894191639,
      "grad_norm": 0.03274618461728096,
      "learning_rate": 1.947691217460921e-05,
      "loss": 0.7584,
      "mean_token_accuracy": 0.7702343159675701,
      "step": 196
    },
    {
      "epoch": 0.583055863854976,
      "grad_norm": 0.030494721606373787,
      "learning_rate": 1.946582412862562e-05,
      "loss": 0.7704,
      "mean_token_accuracy": 0.7677169512560594,
      "step": 197
    },
    {
      "epoch": 0.586015538290788,
      "grad_norm": 0.03319782018661499,
      "learning_rate": 1.9454623017175814e-05,
      "loss": 0.7463,
      "mean_token_accuracy": 0.7741017223360237,
      "step": 198
    },
    {
      "epoch": 0.5889752127266,
      "grad_norm": 0.030228402465581894,
      "learning_rate": 1.9443308974052574e-05,
      "loss": 0.7533,
      "mean_token_accuracy": 0.7735382499602192,
      "step": 199
    },
    {
      "epoch": 0.5919348871624122,
      "grad_norm": 0.034058261662721634,
      "learning_rate": 1.9431882134397596e-05,
      "loss": 0.7701,
      "mean_token_accuracy": 0.7672398018870049,
      "step": 200
    },
    {
      "epoch": 0.5919348871624122,
      "eval_loss": 0.782730221748352,
      "eval_mean_token_accuracy": 0.760863277387458,
      "eval_runtime": 24.7217,
      "eval_samples_per_second": 5.218,
      "eval_steps_per_second": 1.335,
      "step": 200
    },
    {
      "epoch": 0.5948945615982242,
      "grad_norm": 0.0337730310857296,
      "learning_rate": 1.9420342634699893e-05,
      "loss": 0.7253,
      "mean_token_accuracy": 0.7804546641367628,
      "step": 201
    },
    {
      "epoch": 0.5978542360340362,
      "grad_norm": 0.033327993005514145,
      "learning_rate": 1.9408690612794146e-05,
      "loss": 0.7806,
      "mean_token_accuracy": 0.7645614384278685,
      "step": 202
    },
    {
      "epoch": 0.6008139104698483,
      "grad_norm": 0.03368711844086647,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.7145,
      "mean_token_accuracy": 0.7817845508341903,
      "step": 203
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 0.03376063331961632,
      "learning_rate": 1.9385049560415794e-05,
      "loss": 0.7859,
      "mean_token_accuracy": 0.7638255756811193,
      "step": 204
    },
    {
      "epoch": 0.6067332593414725,
      "grad_norm": 0.0319255031645298,
      "learning_rate": 1.9373060812326053e-05,
      "loss": 0.7403,
      "mean_token_accuracy": 0.7763434787707683,
      "step": 205
    },
    {
      "epoch": 0.6096929337772845,
      "grad_norm": 0.03392772004008293,
      "learning_rate": 1.9360960106790645e-05,
      "loss": 0.7678,
      "mean_token_accuracy": 0.7679852215182981,
      "step": 206
    },
    {
      "epoch": 0.6126526082130965,
      "grad_norm": 0.03226333484053612,
      "learning_rate": 1.9348747588347637e-05,
      "loss": 0.7679,
      "mean_token_accuracy": 0.7704594970034341,
      "step": 207
    },
    {
      "epoch": 0.6156122826489087,
      "grad_norm": 0.03440292552113533,
      "learning_rate": 1.9336423402870655e-05,
      "loss": 0.7673,
      "mean_token_accuracy": 0.7694151930145929,
      "step": 208
    },
    {
      "epoch": 0.6185719570847207,
      "grad_norm": 0.0319603867828846,
      "learning_rate": 1.932398769756714e-05,
      "loss": 0.7386,
      "mean_token_accuracy": 0.7752009209087904,
      "step": 209
    },
    {
      "epoch": 0.6215316315205327,
      "grad_norm": 0.03767745569348335,
      "learning_rate": 1.9311440620976597e-05,
      "loss": 0.7417,
      "mean_token_accuracy": 0.7745681696302529,
      "step": 210
    },
    {
      "epoch": 0.6244913059563448,
      "grad_norm": 0.03788335621356964,
      "learning_rate": 1.9298782322968817e-05,
      "loss": 0.7781,
      "mean_token_accuracy": 0.7633484681572329,
      "step": 211
    },
    {
      "epoch": 0.6274509803921569,
      "grad_norm": 0.034501463174819946,
      "learning_rate": 1.9286012954742078e-05,
      "loss": 0.7465,
      "mean_token_accuracy": 0.7742809201549735,
      "step": 212
    },
    {
      "epoch": 0.6304106548279689,
      "grad_norm": 0.0353686548769474,
      "learning_rate": 1.9273132668821363e-05,
      "loss": 0.7933,
      "mean_token_accuracy": 0.7619696827977732,
      "step": 213
    },
    {
      "epoch": 0.633370329263781,
      "grad_norm": 0.03463656082749367,
      "learning_rate": 1.9260141619056507e-05,
      "loss": 0.8108,
      "mean_token_accuracy": 0.7571709317487088,
      "step": 214
    },
    {
      "epoch": 0.636330003699593,
      "grad_norm": 0.03792795166373253,
      "learning_rate": 1.924703996062038e-05,
      "loss": 0.7872,
      "mean_token_accuracy": 0.7638574171086279,
      "step": 215
    },
    {
      "epoch": 0.6392896781354052,
      "grad_norm": 0.03332408517599106,
      "learning_rate": 1.9233827850007028e-05,
      "loss": 0.746,
      "mean_token_accuracy": 0.7746787184963281,
      "step": 216
    },
    {
      "epoch": 0.6422493525712172,
      "grad_norm": 0.032017435878515244,
      "learning_rate": 1.9220505445029803e-05,
      "loss": 0.7461,
      "mean_token_accuracy": 0.775977442375706,
      "step": 217
    },
    {
      "epoch": 0.6452090270070292,
      "grad_norm": 0.03525840863585472,
      "learning_rate": 1.9207072904819484e-05,
      "loss": 0.7903,
      "mean_token_accuracy": 0.7608929466647594,
      "step": 218
    },
    {
      "epoch": 0.6481687014428413,
      "grad_norm": 0.031097477301955223,
      "learning_rate": 1.9193530389822364e-05,
      "loss": 0.7588,
      "mean_token_accuracy": 0.7714811406133166,
      "step": 219
    },
    {
      "epoch": 0.6511283758786534,
      "grad_norm": 0.03184192255139351,
      "learning_rate": 1.9179878061798347e-05,
      "loss": 0.7459,
      "mean_token_accuracy": 0.774873205605776,
      "step": 220
    },
    {
      "epoch": 0.6540880503144654,
      "grad_norm": 0.03820689022541046,
      "learning_rate": 1.9166116083819002e-05,
      "loss": 0.7776,
      "mean_token_accuracy": 0.7658224949653946,
      "step": 221
    },
    {
      "epoch": 0.6570477247502775,
      "grad_norm": 0.030225373804569244,
      "learning_rate": 1.915224462026563e-05,
      "loss": 0.7729,
      "mean_token_accuracy": 0.7672185036607989,
      "step": 222
    },
    {
      "epoch": 0.6600073991860895,
      "grad_norm": 0.03312043100595474,
      "learning_rate": 1.913826383682729e-05,
      "loss": 0.7818,
      "mean_token_accuracy": 0.7636147231412597,
      "step": 223
    },
    {
      "epoch": 0.6629670736219015,
      "grad_norm": 0.032151564955711365,
      "learning_rate": 1.912417390049882e-05,
      "loss": 0.7613,
      "mean_token_accuracy": 0.7697268726451575,
      "step": 224
    },
    {
      "epoch": 0.6659267480577137,
      "grad_norm": 0.03240545839071274,
      "learning_rate": 1.9109974979578852e-05,
      "loss": 0.7388,
      "mean_token_accuracy": 0.7747339698476239,
      "step": 225
    },
    {
      "epoch": 0.6688864224935257,
      "grad_norm": 0.03305862843990326,
      "learning_rate": 1.909566724366779e-05,
      "loss": 0.766,
      "mean_token_accuracy": 0.7685600886615002,
      "step": 226
    },
    {
      "epoch": 0.6718460969293377,
      "grad_norm": 0.03305797651410103,
      "learning_rate": 1.9081250863665794e-05,
      "loss": 0.7499,
      "mean_token_accuracy": 0.773900004152648,
      "step": 227
    },
    {
      "epoch": 0.6748057713651499,
      "grad_norm": 0.03515514358878136,
      "learning_rate": 1.9066726011770725e-05,
      "loss": 0.7446,
      "mean_token_accuracy": 0.7749729001990995,
      "step": 228
    },
    {
      "epoch": 0.6777654458009619,
      "grad_norm": 0.030068494379520416,
      "learning_rate": 1.905209286147611e-05,
      "loss": 0.7332,
      "mean_token_accuracy": 0.7778847562042903,
      "step": 229
    },
    {
      "epoch": 0.680725120236774,
      "grad_norm": 0.02978881262242794,
      "learning_rate": 1.903735158756905e-05,
      "loss": 0.7307,
      "mean_token_accuracy": 0.7791969915963137,
      "step": 230
    },
    {
      "epoch": 0.683684794672586,
      "grad_norm": 0.03472378849983215,
      "learning_rate": 1.9022502366128136e-05,
      "loss": 0.7668,
      "mean_token_accuracy": 0.7695760948903868,
      "step": 231
    },
    {
      "epoch": 0.686644469108398,
      "grad_norm": 0.0332939587533474,
      "learning_rate": 1.9007545374521354e-05,
      "loss": 0.777,
      "mean_token_accuracy": 0.7659026934900942,
      "step": 232
    },
    {
      "epoch": 0.6896041435442102,
      "grad_norm": 0.03128567710518837,
      "learning_rate": 1.8992480791403957e-05,
      "loss": 0.7298,
      "mean_token_accuracy": 0.7803031759905384,
      "step": 233
    },
    {
      "epoch": 0.6925638179800222,
      "grad_norm": 0.03323891758918762,
      "learning_rate": 1.897730879671634e-05,
      "loss": 0.7491,
      "mean_token_accuracy": 0.7727914915658459,
      "step": 234
    },
    {
      "epoch": 0.6955234924158342,
      "grad_norm": 0.030116857960820198,
      "learning_rate": 1.8962029571681887e-05,
      "loss": 0.7134,
      "mean_token_accuracy": 0.7848846599693088,
      "step": 235
    },
    {
      "epoch": 0.6984831668516464,
      "grad_norm": 0.0338609404861927,
      "learning_rate": 1.8946643298804794e-05,
      "loss": 0.7744,
      "mean_token_accuracy": 0.7647683694177534,
      "step": 236
    },
    {
      "epoch": 0.7014428412874584,
      "grad_norm": 0.03095398098230362,
      "learning_rate": 1.8931150161867917e-05,
      "loss": 0.7327,
      "mean_token_accuracy": 0.7784814340852172,
      "step": 237
    },
    {
      "epoch": 0.7044025157232704,
      "grad_norm": 0.032056860625743866,
      "learning_rate": 1.891555034593055e-05,
      "loss": 0.7504,
      "mean_token_accuracy": 0.7726877638212124,
      "step": 238
    },
    {
      "epoch": 0.7073621901590825,
      "grad_norm": 0.033068981021642685,
      "learning_rate": 1.8899844037326227e-05,
      "loss": 0.723,
      "mean_token_accuracy": 0.7817239369543367,
      "step": 239
    },
    {
      "epoch": 0.7103218645948945,
      "grad_norm": 0.03178001567721367,
      "learning_rate": 1.8884031423660492e-05,
      "loss": 0.7076,
      "mean_token_accuracy": 0.7852642237822066,
      "step": 240
    },
    {
      "epoch": 0.7132815390307066,
      "grad_norm": 0.033711135387420654,
      "learning_rate": 1.8868112693808664e-05,
      "loss": 0.7702,
      "mean_token_accuracy": 0.766820921581672,
      "step": 241
    },
    {
      "epoch": 0.7162412134665187,
      "grad_norm": 0.03204965591430664,
      "learning_rate": 1.8852088037913577e-05,
      "loss": 0.7512,
      "mean_token_accuracy": 0.7735279189652862,
      "step": 242
    },
    {
      "epoch": 0.7192008879023307,
      "grad_norm": 0.03038489818572998,
      "learning_rate": 1.8835957647383304e-05,
      "loss": 0.7058,
      "mean_token_accuracy": 0.7862652574110619,
      "step": 243
    },
    {
      "epoch": 0.7221605623381429,
      "grad_norm": 0.033698491752147675,
      "learning_rate": 1.8819721714888878e-05,
      "loss": 0.7838,
      "mean_token_accuracy": 0.7635040231728452,
      "step": 244
    },
    {
      "epoch": 0.7251202367739549,
      "grad_norm": 0.032471273094415665,
      "learning_rate": 1.8803380434362e-05,
      "loss": 0.7386,
      "mean_token_accuracy": 0.7770512330256151,
      "step": 245
    },
    {
      "epoch": 0.7280799112097669,
      "grad_norm": 0.030981970950961113,
      "learning_rate": 1.878693400099269e-05,
      "loss": 0.7339,
      "mean_token_accuracy": 0.7781330708128527,
      "step": 246
    },
    {
      "epoch": 0.731039585645579,
      "grad_norm": 0.03179538995027542,
      "learning_rate": 1.877038261122699e-05,
      "loss": 0.7633,
      "mean_token_accuracy": 0.7710850820754268,
      "step": 247
    },
    {
      "epoch": 0.733999260081391,
      "grad_norm": 0.03083074279129505,
      "learning_rate": 1.87537264627646e-05,
      "loss": 0.7253,
      "mean_token_accuracy": 0.7805634785668045,
      "step": 248
    },
    {
      "epoch": 0.7369589345172031,
      "grad_norm": 0.03432825207710266,
      "learning_rate": 1.8736965754556527e-05,
      "loss": 0.7664,
      "mean_token_accuracy": 0.7683185981997055,
      "step": 249
    },
    {
      "epoch": 0.7399186089530152,
      "grad_norm": 0.034216973930597305,
      "learning_rate": 1.8720100686802693e-05,
      "loss": 0.7595,
      "mean_token_accuracy": 0.7692014276613891,
      "step": 250
    },
    {
      "epoch": 0.7428782833888272,
      "grad_norm": 0.03181126341223717,
      "learning_rate": 1.8703131460949555e-05,
      "loss": 0.7821,
      "mean_token_accuracy": 0.762776204780229,
      "step": 251
    },
    {
      "epoch": 0.7458379578246392,
      "grad_norm": 0.0322745256125927,
      "learning_rate": 1.86860582796877e-05,
      "loss": 0.7402,
      "mean_token_accuracy": 0.7755791912742929,
      "step": 252
    },
    {
      "epoch": 0.7487976322604514,
      "grad_norm": 0.041702453047037125,
      "learning_rate": 1.866888134694942e-05,
      "loss": 0.7495,
      "mean_token_accuracy": 0.7741103389303773,
      "step": 253
    },
    {
      "epoch": 0.7517573066962634,
      "grad_norm": 0.03554534539580345,
      "learning_rate": 1.865160086790627e-05,
      "loss": 0.7276,
      "mean_token_accuracy": 0.7793432956847204,
      "step": 254
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 0.03583842143416405,
      "learning_rate": 1.8634217048966638e-05,
      "loss": 0.7696,
      "mean_token_accuracy": 0.7683482746075759,
      "step": 255
    },
    {
      "epoch": 0.7576766555678875,
      "grad_norm": 0.03574317321181297,
      "learning_rate": 1.861673009777325e-05,
      "loss": 0.7495,
      "mean_token_accuracy": 0.7721361491342168,
      "step": 256
    },
    {
      "epoch": 0.7606363300036996,
      "grad_norm": 0.03595171496272087,
      "learning_rate": 1.8599140223200716e-05,
      "loss": 0.752,
      "mean_token_accuracy": 0.7720915463273443,
      "step": 257
    },
    {
      "epoch": 0.7635960044395117,
      "grad_norm": 0.0364031083881855,
      "learning_rate": 1.858144763535302e-05,
      "loss": 0.7826,
      "mean_token_accuracy": 0.7632796827613793,
      "step": 258
    },
    {
      "epoch": 0.7665556788753237,
      "grad_norm": 0.03217500075697899,
      "learning_rate": 1.8563652545561014e-05,
      "loss": 0.7373,
      "mean_token_accuracy": 0.776466212518149,
      "step": 259
    },
    {
      "epoch": 0.7695153533111357,
      "grad_norm": 0.03651575371623039,
      "learning_rate": 1.8545755166379898e-05,
      "loss": 0.7225,
      "mean_token_accuracy": 0.7818389519130157,
      "step": 260
    },
    {
      "epoch": 0.7724750277469479,
      "grad_norm": 0.035515498369932175,
      "learning_rate": 1.852775571158668e-05,
      "loss": 0.7754,
      "mean_token_accuracy": 0.7653143060755303,
      "step": 261
    },
    {
      "epoch": 0.7754347021827599,
      "grad_norm": 0.032655250281095505,
      "learning_rate": 1.850965439617761e-05,
      "loss": 0.7441,
      "mean_token_accuracy": 0.7729170520455341,
      "step": 262
    },
    {
      "epoch": 0.7783943766185719,
      "grad_norm": 0.04255848377943039,
      "learning_rate": 1.8491451436365628e-05,
      "loss": 0.7795,
      "mean_token_accuracy": 0.7633152491917201,
      "step": 263
    },
    {
      "epoch": 0.781354051054384,
      "grad_norm": 0.03399297595024109,
      "learning_rate": 1.8473147049577777e-05,
      "loss": 0.7706,
      "mean_token_accuracy": 0.7676074807018852,
      "step": 264
    },
    {
      "epoch": 0.7843137254901961,
      "grad_norm": 0.03587772697210312,
      "learning_rate": 1.8454741454452604e-05,
      "loss": 0.7559,
      "mean_token_accuracy": 0.7714855442630545,
      "step": 265
    },
    {
      "epoch": 0.7872733999260081,
      "grad_norm": 0.034946031868457794,
      "learning_rate": 1.843623487083755e-05,
      "loss": 0.7439,
      "mean_token_accuracy": 0.7756796284499916,
      "step": 266
    },
    {
      "epoch": 0.7902330743618202,
      "grad_norm": 0.033369943499565125,
      "learning_rate": 1.8417627519786317e-05,
      "loss": 0.7628,
      "mean_token_accuracy": 0.7685935411526189,
      "step": 267
    },
    {
      "epoch": 0.7931927487976322,
      "grad_norm": 0.03367667272686958,
      "learning_rate": 1.839891962355624e-05,
      "loss": 0.7201,
      "mean_token_accuracy": 0.781186419549585,
      "step": 268
    },
    {
      "epoch": 0.7961524232334444,
      "grad_norm": 0.03550582006573677,
      "learning_rate": 1.838011140560562e-05,
      "loss": 0.7602,
      "mean_token_accuracy": 0.7696177576821034,
      "step": 269
    },
    {
      "epoch": 0.7991120976692564,
      "grad_norm": 0.03280024975538254,
      "learning_rate": 1.836120309059107e-05,
      "loss": 0.7529,
      "mean_token_accuracy": 0.7721933474820526,
      "step": 270
    },
    {
      "epoch": 0.8020717721050684,
      "grad_norm": 0.03471779450774193,
      "learning_rate": 1.8342194904364815e-05,
      "loss": 0.7175,
      "mean_token_accuracy": 0.7819206631878622,
      "step": 271
    },
    {
      "epoch": 0.8050314465408805,
      "grad_norm": 0.03391272574663162,
      "learning_rate": 1.8323087073971996e-05,
      "loss": 0.7399,
      "mean_token_accuracy": 0.7750955451541647,
      "step": 272
    },
    {
      "epoch": 0.8079911209766926,
      "grad_norm": 0.038367222994565964,
      "learning_rate": 1.8303879827647977e-05,
      "loss": 0.7581,
      "mean_token_accuracy": 0.7710080861085928,
      "step": 273
    },
    {
      "epoch": 0.8109507954125046,
      "grad_norm": 0.037487514317035675,
      "learning_rate": 1.8284573394815596e-05,
      "loss": 0.7487,
      "mean_token_accuracy": 0.7729222710688437,
      "step": 274
    },
    {
      "epoch": 0.8139104698483167,
      "grad_norm": 0.03280162066221237,
      "learning_rate": 1.826516800608244e-05,
      "loss": 0.7663,
      "mean_token_accuracy": 0.7687775106057849,
      "step": 275
    },
    {
      "epoch": 0.8168701442841287,
      "grad_norm": 0.03341701626777649,
      "learning_rate": 1.8245663893238075e-05,
      "loss": 0.7695,
      "mean_token_accuracy": 0.76759030475242,
      "step": 276
    },
    {
      "epoch": 0.8198298187199408,
      "grad_norm": 0.040311548858881,
      "learning_rate": 1.8226061289251297e-05,
      "loss": 0.7672,
      "mean_token_accuracy": 0.76806103155876,
      "step": 277
    },
    {
      "epoch": 0.8227894931557529,
      "grad_norm": 0.03274424374103546,
      "learning_rate": 1.8206360428267332e-05,
      "loss": 0.6876,
      "mean_token_accuracy": 0.790432438931161,
      "step": 278
    },
    {
      "epoch": 0.8257491675915649,
      "grad_norm": 0.037516999989748,
      "learning_rate": 1.8186561545605055e-05,
      "loss": 0.7635,
      "mean_token_accuracy": 0.769564809199736,
      "step": 279
    },
    {
      "epoch": 0.8287088420273769,
      "grad_norm": 0.03393593803048134,
      "learning_rate": 1.816666487775416e-05,
      "loss": 0.7492,
      "mean_token_accuracy": 0.7740442354463241,
      "step": 280
    },
    {
      "epoch": 0.8316685164631891,
      "grad_norm": 0.036721568554639816,
      "learning_rate": 1.8146670662372353e-05,
      "loss": 0.747,
      "mean_token_accuracy": 0.7739746940135478,
      "step": 281
    },
    {
      "epoch": 0.8346281908990011,
      "grad_norm": 0.03595879301428795,
      "learning_rate": 1.8126579138282502e-05,
      "loss": 0.7555,
      "mean_token_accuracy": 0.7710237547523787,
      "step": 282
    },
    {
      "epoch": 0.8375878653348132,
      "grad_norm": 0.03116425685584545,
      "learning_rate": 1.8106390545469797e-05,
      "loss": 0.7645,
      "mean_token_accuracy": 0.769266064492148,
      "step": 283
    },
    {
      "epoch": 0.8405475397706252,
      "grad_norm": 0.03314213454723358,
      "learning_rate": 1.8086105125078858e-05,
      "loss": 0.7368,
      "mean_token_accuracy": 0.7773449894515131,
      "step": 284
    },
    {
      "epoch": 0.8435072142064373,
      "grad_norm": 0.03777587041258812,
      "learning_rate": 1.8065723119410885e-05,
      "loss": 0.734,
      "mean_token_accuracy": 0.7763653275974284,
      "step": 285
    },
    {
      "epoch": 0.8464668886422494,
      "grad_norm": 0.03679841011762619,
      "learning_rate": 1.804524477192075e-05,
      "loss": 0.7372,
      "mean_token_accuracy": 0.775884892394405,
      "step": 286
    },
    {
      "epoch": 0.8494265630780614,
      "grad_norm": 0.031244000419974327,
      "learning_rate": 1.8024670327214084e-05,
      "loss": 0.7297,
      "mean_token_accuracy": 0.7799712358309591,
      "step": 287
    },
    {
      "epoch": 0.8523862375138734,
      "grad_norm": 0.04045666381716728,
      "learning_rate": 1.8004000031044363e-05,
      "loss": 0.7837,
      "mean_token_accuracy": 0.7637096806986703,
      "step": 288
    },
    {
      "epoch": 0.8553459119496856,
      "grad_norm": 0.030689913779497147,
      "learning_rate": 1.798323413030997e-05,
      "loss": 0.7337,
      "mean_token_accuracy": 0.7768989749579454,
      "step": 289
    },
    {
      "epoch": 0.8583055863854976,
      "grad_norm": 0.03591914474964142,
      "learning_rate": 1.796237287305125e-05,
      "loss": 0.7353,
      "mean_token_accuracy": 0.7765039076965347,
      "step": 290
    },
    {
      "epoch": 0.8612652608213096,
      "grad_norm": 0.03441183269023895,
      "learning_rate": 1.7941416508447537e-05,
      "loss": 0.7409,
      "mean_token_accuracy": 0.7749446991304727,
      "step": 291
    },
    {
      "epoch": 0.8642249352571217,
      "grad_norm": 0.03485677018761635,
      "learning_rate": 1.792036528681418e-05,
      "loss": 0.7488,
      "mean_token_accuracy": 0.7733238513849428,
      "step": 292
    },
    {
      "epoch": 0.8671846096929338,
      "grad_norm": 0.03616423159837723,
      "learning_rate": 1.789921945959958e-05,
      "loss": 0.7331,
      "mean_token_accuracy": 0.7764074323111411,
      "step": 293
    },
    {
      "epoch": 0.8701442841287459,
      "grad_norm": 0.030521664768457413,
      "learning_rate": 1.7877979279382135e-05,
      "loss": 0.7237,
      "mean_token_accuracy": 0.7802865303017775,
      "step": 294
    },
    {
      "epoch": 0.8731039585645579,
      "grad_norm": 0.035453200340270996,
      "learning_rate": 1.7856644999867264e-05,
      "loss": 0.7532,
      "mean_token_accuracy": 0.7719881399745001,
      "step": 295
    },
    {
      "epoch": 0.8760636330003699,
      "grad_norm": 0.032341089099645615,
      "learning_rate": 1.783521687588437e-05,
      "loss": 0.7014,
      "mean_token_accuracy": 0.7874739737787696,
      "step": 296
    },
    {
      "epoch": 0.8790233074361821,
      "grad_norm": 0.033419959247112274,
      "learning_rate": 1.781369516338378e-05,
      "loss": 0.7435,
      "mean_token_accuracy": 0.7743909186756861,
      "step": 297
    },
    {
      "epoch": 0.8819829818719941,
      "grad_norm": 0.03388496860861778,
      "learning_rate": 1.779208011943371e-05,
      "loss": 0.7373,
      "mean_token_accuracy": 0.7755390679346732,
      "step": 298
    },
    {
      "epoch": 0.8849426563078061,
      "grad_norm": 0.029631685465574265,
      "learning_rate": 1.777037200221717e-05,
      "loss": 0.7428,
      "mean_token_accuracy": 0.774006184394115,
      "step": 299
    },
    {
      "epoch": 0.8879023307436182,
      "grad_norm": 0.03248713165521622,
      "learning_rate": 1.77485710710289e-05,
      "loss": 0.7327,
      "mean_token_accuracy": 0.7779741478131739,
      "step": 300
    },
    {
      "epoch": 0.8879023307436182,
      "eval_loss": 0.7651689648628235,
      "eval_mean_token_accuracy": 0.7653411134270157,
      "eval_runtime": 24.7317,
      "eval_samples_per_second": 5.216,
      "eval_steps_per_second": 1.334,
      "step": 300
    },
    {
      "epoch": 0.8908620051794303,
      "grad_norm": 0.03294381499290466,
      "learning_rate": 1.7726677586272263e-05,
      "loss": 0.7287,
      "mean_token_accuracy": 0.7794406515697183,
      "step": 301
    },
    {
      "epoch": 0.8938216796152423,
      "grad_norm": 0.03231232240796089,
      "learning_rate": 1.7704691809456142e-05,
      "loss": 0.766,
      "mean_token_accuracy": 0.7679922917966525,
      "step": 302
    },
    {
      "epoch": 0.8967813540510544,
      "grad_norm": 0.03185324743390083,
      "learning_rate": 1.7682614003191807e-05,
      "loss": 0.7227,
      "mean_token_accuracy": 0.7820203432226969,
      "step": 303
    },
    {
      "epoch": 0.8997410284868664,
      "grad_norm": 0.029139429330825806,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.7399,
      "mean_token_accuracy": 0.7751363574324961,
      "step": 304
    },
    {
      "epoch": 0.9027007029226785,
      "grad_norm": 0.03121069073677063,
      "learning_rate": 1.76381833582567e-05,
      "loss": 0.7385,
      "mean_token_accuracy": 0.7743020754632624,
      "step": 305
    },
    {
      "epoch": 0.9056603773584906,
      "grad_norm": 0.03469480201601982,
      "learning_rate": 1.761583105029213e-05,
      "loss": 0.7126,
      "mean_token_accuracy": 0.7825847092542977,
      "step": 306
    },
    {
      "epoch": 0.9086200517943026,
      "grad_norm": 0.03172752633690834,
      "learning_rate": 1.7593387774285412e-05,
      "loss": 0.73,
      "mean_token_accuracy": 0.7784024192768745,
      "step": 307
    },
    {
      "epoch": 0.9115797262301147,
      "grad_norm": 0.029237743467092514,
      "learning_rate": 1.7570853798312462e-05,
      "loss": 0.7267,
      "mean_token_accuracy": 0.7799445593889766,
      "step": 308
    },
    {
      "epoch": 0.9145394006659268,
      "grad_norm": 0.031245123594999313,
      "learning_rate": 1.7548229391532572e-05,
      "loss": 0.6603,
      "mean_token_accuracy": 0.7977340496924726,
      "step": 309
    },
    {
      "epoch": 0.9174990751017388,
      "grad_norm": 0.03181256353855133,
      "learning_rate": 1.7525514824185187e-05,
      "loss": 0.7271,
      "mean_token_accuracy": 0.7796218791896655,
      "step": 310
    },
    {
      "epoch": 0.9204587495375509,
      "grad_norm": 0.029375318437814713,
      "learning_rate": 1.750271036758669e-05,
      "loss": 0.7599,
      "mean_token_accuracy": 0.7702703929768216,
      "step": 311
    },
    {
      "epoch": 0.9234184239733629,
      "grad_norm": 0.03115256503224373,
      "learning_rate": 1.747981629412715e-05,
      "loss": 0.7415,
      "mean_token_accuracy": 0.7748722985282385,
      "step": 312
    },
    {
      "epoch": 0.926378098409175,
      "grad_norm": 0.03140730410814285,
      "learning_rate": 1.7456832877267083e-05,
      "loss": 0.7182,
      "mean_token_accuracy": 0.782194479289576,
      "step": 313
    },
    {
      "epoch": 0.9293377728449871,
      "grad_norm": 0.029602287337183952,
      "learning_rate": 1.7433760391534166e-05,
      "loss": 0.7286,
      "mean_token_accuracy": 0.7793288365526181,
      "step": 314
    },
    {
      "epoch": 0.9322974472807991,
      "grad_norm": 0.033283088356256485,
      "learning_rate": 1.741059911251997e-05,
      "loss": 0.7429,
      "mean_token_accuracy": 0.7746304922297019,
      "step": 315
    },
    {
      "epoch": 0.9352571217166111,
      "grad_norm": 0.03298504278063774,
      "learning_rate": 1.7387349316876668e-05,
      "loss": 0.7249,
      "mean_token_accuracy": 0.7800789377237446,
      "step": 316
    },
    {
      "epoch": 0.9382167961524233,
      "grad_norm": 0.029268553480505943,
      "learning_rate": 1.7364011282313732e-05,
      "loss": 0.7163,
      "mean_token_accuracy": 0.7810958489914521,
      "step": 317
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 0.03102250210940838,
      "learning_rate": 1.7340585287594605e-05,
      "loss": 0.7314,
      "mean_token_accuracy": 0.7777864735018288,
      "step": 318
    },
    {
      "epoch": 0.9441361450240473,
      "grad_norm": 0.03122146986424923,
      "learning_rate": 1.731707161253338e-05,
      "loss": 0.7682,
      "mean_token_accuracy": 0.7676880887984262,
      "step": 319
    },
    {
      "epoch": 0.9470958194598594,
      "grad_norm": 0.030520522966980934,
      "learning_rate": 1.7293470537991463e-05,
      "loss": 0.7329,
      "mean_token_accuracy": 0.7772511521206702,
      "step": 320
    },
    {
      "epoch": 0.9500554938956715,
      "grad_norm": 0.032705482095479965,
      "learning_rate": 1.7269782345874204e-05,
      "loss": 0.7063,
      "mean_token_accuracy": 0.7847350651684459,
      "step": 321
    },
    {
      "epoch": 0.9530151683314836,
      "grad_norm": 0.032125309109687805,
      "learning_rate": 1.7246007319127547e-05,
      "loss": 0.7507,
      "mean_token_accuracy": 0.773278755434117,
      "step": 322
    },
    {
      "epoch": 0.9559748427672956,
      "grad_norm": 0.03482387587428093,
      "learning_rate": 1.7222145741734625e-05,
      "loss": 0.7238,
      "mean_token_accuracy": 0.7801127727300564,
      "step": 323
    },
    {
      "epoch": 0.9589345172031076,
      "grad_norm": 0.03271453455090523,
      "learning_rate": 1.7198197898712402e-05,
      "loss": 0.7312,
      "mean_token_accuracy": 0.7778095028206786,
      "step": 324
    },
    {
      "epoch": 0.9618941916389198,
      "grad_norm": 0.031074879691004753,
      "learning_rate": 1.717416407610824e-05,
      "loss": 0.6925,
      "mean_token_accuracy": 0.7869436601525868,
      "step": 325
    },
    {
      "epoch": 0.9648538660747318,
      "grad_norm": 0.031365539878606796,
      "learning_rate": 1.7150044560996488e-05,
      "loss": 0.7504,
      "mean_token_accuracy": 0.7736639198741512,
      "step": 326
    },
    {
      "epoch": 0.9678135405105438,
      "grad_norm": 0.03309888020157814,
      "learning_rate": 1.7125839641475074e-05,
      "loss": 0.7508,
      "mean_token_accuracy": 0.7741240045337013,
      "step": 327
    },
    {
      "epoch": 0.9707732149463559,
      "grad_norm": 0.028719095513224602,
      "learning_rate": 1.7101549606662025e-05,
      "loss": 0.7629,
      "mean_token_accuracy": 0.7664604053923443,
      "step": 328
    },
    {
      "epoch": 0.973732889382168,
      "grad_norm": 0.030317535623908043,
      "learning_rate": 1.7077174746692054e-05,
      "loss": 0.7125,
      "mean_token_accuracy": 0.7827231553164308,
      "step": 329
    },
    {
      "epoch": 0.97669256381798,
      "grad_norm": 0.03435689955949783,
      "learning_rate": 1.7052715352713076e-05,
      "loss": 0.6956,
      "mean_token_accuracy": 0.7874807466792983,
      "step": 330
    },
    {
      "epoch": 0.9796522382537921,
      "grad_norm": 0.02943146787583828,
      "learning_rate": 1.7028171716882714e-05,
      "loss": 0.7304,
      "mean_token_accuracy": 0.7787544388176159,
      "step": 331
    },
    {
      "epoch": 0.9826119126896041,
      "grad_norm": 0.030871093273162842,
      "learning_rate": 1.7003544132364847e-05,
      "loss": 0.7529,
      "mean_token_accuracy": 0.7713072336847819,
      "step": 332
    },
    {
      "epoch": 0.9855715871254163,
      "grad_norm": 0.0330820232629776,
      "learning_rate": 1.6978832893326074e-05,
      "loss": 0.7311,
      "mean_token_accuracy": 0.7764928547898482,
      "step": 333
    },
    {
      "epoch": 0.9885312615612283,
      "grad_norm": 0.03206097334623337,
      "learning_rate": 1.6954038294932215e-05,
      "loss": 0.7309,
      "mean_token_accuracy": 0.7781804958794002,
      "step": 334
    },
    {
      "epoch": 0.9914909359970403,
      "grad_norm": 0.03135286644101143,
      "learning_rate": 1.692916063334479e-05,
      "loss": 0.7194,
      "mean_token_accuracy": 0.7799673950007542,
      "step": 335
    },
    {
      "epoch": 0.9944506104328524,
      "grad_norm": 0.033186834305524826,
      "learning_rate": 1.690420020571747e-05,
      "loss": 0.7899,
      "mean_token_accuracy": 0.7602483320552219,
      "step": 336
    },
    {
      "epoch": 0.9974102848686645,
      "grad_norm": 0.03052665665745735,
      "learning_rate": 1.6879157310192537e-05,
      "loss": 0.7273,
      "mean_token_accuracy": 0.7788000687099865,
      "step": 337
    },
    {
      "epoch": 1.002959674435812,
      "grad_norm": 0.05862506106495857,
      "learning_rate": 1.685403224589731e-05,
      "loss": 1.4416,
      "mean_token_accuracy": 0.780696022213719,
      "step": 338
    },
    {
      "epoch": 1.005919348871624,
      "grad_norm": 0.029678061604499817,
      "learning_rate": 1.6828825312940594e-05,
      "loss": 0.7204,
      "mean_token_accuracy": 0.7797792510578121,
      "step": 339
    },
    {
      "epoch": 1.0088790233074363,
      "grad_norm": 0.03100208193063736,
      "learning_rate": 1.6803536812409077e-05,
      "loss": 0.6611,
      "mean_token_accuracy": 0.7959624931730475,
      "step": 340
    },
    {
      "epoch": 1.0118386977432483,
      "grad_norm": 0.03420529514551163,
      "learning_rate": 1.6778167046363735e-05,
      "loss": 0.6712,
      "mean_token_accuracy": 0.7932777434121004,
      "step": 341
    },
    {
      "epoch": 1.0147983721790603,
      "grad_norm": 0.03360075131058693,
      "learning_rate": 1.675271631783623e-05,
      "loss": 0.7009,
      "mean_token_accuracy": 0.7850573385535741,
      "step": 342
    },
    {
      "epoch": 1.0177580466148723,
      "grad_norm": 0.033614061772823334,
      "learning_rate": 1.672718493082529e-05,
      "loss": 0.7038,
      "mean_token_accuracy": 0.7845647773208296,
      "step": 343
    },
    {
      "epoch": 1.0207177210506844,
      "grad_norm": 0.034418750554323196,
      "learning_rate": 1.6701573190293076e-05,
      "loss": 0.7169,
      "mean_token_accuracy": 0.7794328807774604,
      "step": 344
    },
    {
      "epoch": 1.0236773954864964,
      "grad_norm": 0.03483947366476059,
      "learning_rate": 1.667588140216154e-05,
      "loss": 0.7262,
      "mean_token_accuracy": 0.7765482334577232,
      "step": 345
    },
    {
      "epoch": 1.0266370699223086,
      "grad_norm": 0.03210083022713661,
      "learning_rate": 1.6650109873308763e-05,
      "loss": 0.6822,
      "mean_token_accuracy": 0.7909700146422477,
      "step": 346
    },
    {
      "epoch": 1.0295967443581207,
      "grad_norm": 0.03218446671962738,
      "learning_rate": 1.6624258911565312e-05,
      "loss": 0.7038,
      "mean_token_accuracy": 0.7832506511626859,
      "step": 347
    },
    {
      "epoch": 1.0325564187939327,
      "grad_norm": 0.035634465515613556,
      "learning_rate": 1.6598328825710536e-05,
      "loss": 0.7513,
      "mean_token_accuracy": 0.7698961666083765,
      "step": 348
    },
    {
      "epoch": 1.0355160932297447,
      "grad_norm": 0.037157826125621796,
      "learning_rate": 1.6572319925468892e-05,
      "loss": 0.6925,
      "mean_token_accuracy": 0.787021142427316,
      "step": 349
    },
    {
      "epoch": 1.0384757676655567,
      "grad_norm": 0.030800631269812584,
      "learning_rate": 1.654623252150624e-05,
      "loss": 0.6957,
      "mean_token_accuracy": 0.7864814152952101,
      "step": 350
    },
    {
      "epoch": 1.0414354421013687,
      "grad_norm": 0.029243871569633484,
      "learning_rate": 1.6520066925426146e-05,
      "loss": 0.6844,
      "mean_token_accuracy": 0.7880249869746129,
      "step": 351
    },
    {
      "epoch": 1.044395116537181,
      "grad_norm": 0.03222661465406418,
      "learning_rate": 1.6493823449766137e-05,
      "loss": 0.708,
      "mean_token_accuracy": 0.7815879649114263,
      "step": 352
    },
    {
      "epoch": 1.047354790972993,
      "grad_norm": 0.03295891731977463,
      "learning_rate": 1.6467502407993995e-05,
      "loss": 0.6869,
      "mean_token_accuracy": 0.7876301328483187,
      "step": 353
    },
    {
      "epoch": 1.050314465408805,
      "grad_norm": 0.03705889359116554,
      "learning_rate": 1.644110411450398e-05,
      "loss": 0.7113,
      "mean_token_accuracy": 0.7823622133534957,
      "step": 354
    },
    {
      "epoch": 1.053274139844617,
      "grad_norm": 0.029595624655485153,
      "learning_rate": 1.6414628884613106e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.785448577457593,
      "step": 355
    },
    {
      "epoch": 1.056233814280429,
      "grad_norm": 0.031605131924152374,
      "learning_rate": 1.6388077034557355e-05,
      "loss": 0.7192,
      "mean_token_accuracy": 0.7785981425945822,
      "step": 356
    },
    {
      "epoch": 1.0591934887162413,
      "grad_norm": 0.030189666897058487,
      "learning_rate": 1.6361448881487913e-05,
      "loss": 0.684,
      "mean_token_accuracy": 0.789518852268955,
      "step": 357
    },
    {
      "epoch": 1.0621531631520533,
      "grad_norm": 0.028912125155329704,
      "learning_rate": 1.6334744743467366e-05,
      "loss": 0.6958,
      "mean_token_accuracy": 0.7862255119868089,
      "step": 358
    },
    {
      "epoch": 1.0651128375878653,
      "grad_norm": 0.029534118250012398,
      "learning_rate": 1.6307964939465914e-05,
      "loss": 0.6914,
      "mean_token_accuracy": 0.7878655276020197,
      "step": 359
    },
    {
      "epoch": 1.0680725120236774,
      "grad_norm": 0.03177528455853462,
      "learning_rate": 1.628110978935756e-05,
      "loss": 0.6985,
      "mean_token_accuracy": 0.7852362445011971,
      "step": 360
    },
    {
      "epoch": 1.0710321864594894,
      "grad_norm": 0.03307664394378662,
      "learning_rate": 1.625417961391628e-05,
      "loss": 0.6551,
      "mean_token_accuracy": 0.7970499807101614,
      "step": 361
    },
    {
      "epoch": 1.0739918608953016,
      "grad_norm": 0.030627764761447906,
      "learning_rate": 1.62271747348122e-05,
      "loss": 0.7017,
      "mean_token_accuracy": 0.78515864040978,
      "step": 362
    },
    {
      "epoch": 1.0769515353311137,
      "grad_norm": 0.034586723893880844,
      "learning_rate": 1.6200095474607753e-05,
      "loss": 0.697,
      "mean_token_accuracy": 0.7843048297400937,
      "step": 363
    },
    {
      "epoch": 1.0799112097669257,
      "grad_norm": 0.029668590053915977,
      "learning_rate": 1.6172942156753822e-05,
      "loss": 0.6819,
      "mean_token_accuracy": 0.7900832488648033,
      "step": 364
    },
    {
      "epoch": 1.0828708842027377,
      "grad_norm": 0.02863898128271103,
      "learning_rate": 1.614571510558588e-05,
      "loss": 0.6814,
      "mean_token_accuracy": 0.7914928366027313,
      "step": 365
    },
    {
      "epoch": 1.0858305586385497,
      "grad_norm": 0.03317256271839142,
      "learning_rate": 1.6118414646320115e-05,
      "loss": 0.6863,
      "mean_token_accuracy": 0.789663594690608,
      "step": 366
    },
    {
      "epoch": 1.0887902330743617,
      "grad_norm": 0.03485291823744774,
      "learning_rate": 1.6091041105049542e-05,
      "loss": 0.7029,
      "mean_token_accuracy": 0.7833869138445729,
      "step": 367
    },
    {
      "epoch": 1.091749907510174,
      "grad_norm": 0.027830787003040314,
      "learning_rate": 1.6063594808740112e-05,
      "loss": 0.649,
      "mean_token_accuracy": 0.7993210376928372,
      "step": 368
    },
    {
      "epoch": 1.094709581945986,
      "grad_norm": 0.03277172893285751,
      "learning_rate": 1.6036076085226813e-05,
      "loss": 0.7406,
      "mean_token_accuracy": 0.7719364426094815,
      "step": 369
    },
    {
      "epoch": 1.097669256381798,
      "grad_norm": 0.03152583912014961,
      "learning_rate": 1.6008485263209742e-05,
      "loss": 0.6595,
      "mean_token_accuracy": 0.7972482679965673,
      "step": 370
    },
    {
      "epoch": 1.10062893081761,
      "grad_norm": 0.030170999467372894,
      "learning_rate": 1.598082267225018e-05,
      "loss": 0.7197,
      "mean_token_accuracy": 0.7789193631242192,
      "step": 371
    },
    {
      "epoch": 1.103588605253422,
      "grad_norm": 0.03111015260219574,
      "learning_rate": 1.595308864276666e-05,
      "loss": 0.7293,
      "mean_token_accuracy": 0.7759533419825515,
      "step": 372
    },
    {
      "epoch": 1.106548279689234,
      "grad_norm": 0.03666509687900543,
      "learning_rate": 1.592528350603103e-05,
      "loss": 0.6997,
      "mean_token_accuracy": 0.783954573590657,
      "step": 373
    },
    {
      "epoch": 1.1095079541250463,
      "grad_norm": 0.030199384316802025,
      "learning_rate": 1.5897407594164468e-05,
      "loss": 0.7077,
      "mean_token_accuracy": 0.7838309524793444,
      "step": 374
    },
    {
      "epoch": 1.1124676285608583,
      "grad_norm": 0.030692126601934433,
      "learning_rate": 1.586946124013354e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.7885178859843318,
      "step": 375
    },
    {
      "epoch": 1.1154273029966704,
      "grad_norm": 0.031568851321935654,
      "learning_rate": 1.5841444777746232e-05,
      "loss": 0.6529,
      "mean_token_accuracy": 0.7981953267564101,
      "step": 376
    },
    {
      "epoch": 1.1183869774324824,
      "grad_norm": 0.034029778093099594,
      "learning_rate": 1.5813358541647915e-05,
      "loss": 0.69,
      "mean_token_accuracy": 0.7882065265685139,
      "step": 377
    },
    {
      "epoch": 1.1213466518682944,
      "grad_norm": 0.030421767383813858,
      "learning_rate": 1.578520286731741e-05,
      "loss": 0.7185,
      "mean_token_accuracy": 0.7780810677041956,
      "step": 378
    },
    {
      "epoch": 1.1243063263041067,
      "grad_norm": 0.032709620893001556,
      "learning_rate": 1.575697809106292e-05,
      "loss": 0.7005,
      "mean_token_accuracy": 0.7837346904672076,
      "step": 379
    },
    {
      "epoch": 1.1272660007399187,
      "grad_norm": 0.03561005741357803,
      "learning_rate": 1.5728684550018066e-05,
      "loss": 0.7017,
      "mean_token_accuracy": 0.7845035737312578,
      "step": 380
    },
    {
      "epoch": 1.1302256751757307,
      "grad_norm": 0.028430243954062462,
      "learning_rate": 1.570032258213783e-05,
      "loss": 0.7093,
      "mean_token_accuracy": 0.7803112147422935,
      "step": 381
    },
    {
      "epoch": 1.1331853496115427,
      "grad_norm": 0.03315449133515358,
      "learning_rate": 1.5671892526194515e-05,
      "loss": 0.6877,
      "mean_token_accuracy": 0.7902133633539826,
      "step": 382
    },
    {
      "epoch": 1.1361450240473547,
      "grad_norm": 0.03168237954378128,
      "learning_rate": 1.564339472177373e-05,
      "loss": 0.6781,
      "mean_token_accuracy": 0.7911075978390688,
      "step": 383
    },
    {
      "epoch": 1.1391046984831668,
      "grad_norm": 0.032132744789123535,
      "learning_rate": 1.561482950927029e-05,
      "loss": 0.7111,
      "mean_token_accuracy": 0.7805541308517087,
      "step": 384
    },
    {
      "epoch": 1.142064372918979,
      "grad_norm": 0.03324122726917267,
      "learning_rate": 1.5586197229884185e-05,
      "loss": 0.6625,
      "mean_token_accuracy": 0.7956164012918618,
      "step": 385
    },
    {
      "epoch": 1.145024047354791,
      "grad_norm": 0.03547290712594986,
      "learning_rate": 1.5557498225616488e-05,
      "loss": 0.7163,
      "mean_token_accuracy": 0.780174921283727,
      "step": 386
    },
    {
      "epoch": 1.147983721790603,
      "grad_norm": 0.029701631516218185,
      "learning_rate": 1.5528732839265272e-05,
      "loss": 0.7086,
      "mean_token_accuracy": 0.7812849116287365,
      "step": 387
    },
    {
      "epoch": 1.150943396226415,
      "grad_norm": 0.03197464719414711,
      "learning_rate": 1.549990141442153e-05,
      "loss": 0.6898,
      "mean_token_accuracy": 0.788741800444653,
      "step": 388
    },
    {
      "epoch": 1.153903070662227,
      "grad_norm": 0.03248845040798187,
      "learning_rate": 1.5471004295465034e-05,
      "loss": 0.7419,
      "mean_token_accuracy": 0.7732000110125795,
      "step": 389
    },
    {
      "epoch": 1.156862745098039,
      "grad_norm": 0.029760310426354408,
      "learning_rate": 1.5442041827560274e-05,
      "loss": 0.7022,
      "mean_token_accuracy": 0.784173848208745,
      "step": 390
    },
    {
      "epoch": 1.1598224195338513,
      "grad_norm": 0.03191938251256943,
      "learning_rate": 1.5413014356652287e-05,
      "loss": 0.6837,
      "mean_token_accuracy": 0.7880149835350957,
      "step": 391
    },
    {
      "epoch": 1.1627820939696634,
      "grad_norm": 0.03169205039739609,
      "learning_rate": 1.538392222946255e-05,
      "loss": 0.7068,
      "mean_token_accuracy": 0.7833675806723954,
      "step": 392
    },
    {
      "epoch": 1.1657417684054754,
      "grad_norm": 0.027826927602291107,
      "learning_rate": 1.5354765793484834e-05,
      "loss": 0.6859,
      "mean_token_accuracy": 0.7903316417663963,
      "step": 393
    },
    {
      "epoch": 1.1687014428412874,
      "grad_norm": 0.03408754616975784,
      "learning_rate": 1.5325545396981053e-05,
      "loss": 0.702,
      "mean_token_accuracy": 0.784203998343302,
      "step": 394
    },
    {
      "epoch": 1.1716611172770994,
      "grad_norm": 0.030327627435326576,
      "learning_rate": 1.5296261388977107e-05,
      "loss": 0.6366,
      "mean_token_accuracy": 0.8032635644624665,
      "step": 395
    },
    {
      "epoch": 1.1746207917129117,
      "grad_norm": 0.031098052859306335,
      "learning_rate": 1.52669141192587e-05,
      "loss": 0.7303,
      "mean_token_accuracy": 0.7770230512290136,
      "step": 396
    },
    {
      "epoch": 1.1775804661487237,
      "grad_norm": 0.03328563645482063,
      "learning_rate": 1.5237503938367186e-05,
      "loss": 0.7042,
      "mean_token_accuracy": 0.7828977364053202,
      "step": 397
    },
    {
      "epoch": 1.1805401405845357,
      "grad_norm": 0.030965222045779228,
      "learning_rate": 1.5208031197595357e-05,
      "loss": 0.6253,
      "mean_token_accuracy": 0.8065848166559944,
      "step": 398
    },
    {
      "epoch": 1.1834998150203477,
      "grad_norm": 0.02860182337462902,
      "learning_rate": 1.5178496248983254e-05,
      "loss": 0.6519,
      "mean_token_accuracy": 0.7978071202469641,
      "step": 399
    },
    {
      "epoch": 1.1864594894561598,
      "grad_norm": 0.033300597220659256,
      "learning_rate": 1.5148899445313983e-05,
      "loss": 0.6465,
      "mean_token_accuracy": 0.7988422267440451,
      "step": 400
    },
    {
      "epoch": 1.1864594894561598,
      "eval_loss": 0.7561641931533813,
      "eval_mean_token_accuracy": 0.7670792346072827,
      "eval_runtime": 24.707,
      "eval_samples_per_second": 5.221,
      "eval_steps_per_second": 1.336,
      "step": 400
    },
    {
      "epoch": 1.189419163891972,
      "grad_norm": 0.03117380291223526,
      "learning_rate": 1.5119241140109466e-05,
      "loss": 0.6521,
      "mean_token_accuracy": 0.7991485365284977,
      "step": 401
    },
    {
      "epoch": 1.192378838327784,
      "grad_norm": 0.028914451599121094,
      "learning_rate": 1.5089521687626243e-05,
      "loss": 0.6504,
      "mean_token_accuracy": 0.7983551944392973,
      "step": 402
    },
    {
      "epoch": 1.195338512763596,
      "grad_norm": 0.0327957347035408,
      "learning_rate": 1.505974144285124e-05,
      "loss": 0.7229,
      "mean_token_accuracy": 0.7790167661248313,
      "step": 403
    },
    {
      "epoch": 1.198298187199408,
      "grad_norm": 0.03273818641901016,
      "learning_rate": 1.5029900761497507e-05,
      "loss": 0.7549,
      "mean_token_accuracy": 0.7701969901774389,
      "step": 404
    },
    {
      "epoch": 1.20125786163522,
      "grad_norm": 0.03297555446624756,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7825334431876964,
      "step": 405
    },
    {
      "epoch": 1.204217536071032,
      "grad_norm": 0.031928740441799164,
      "learning_rate": 1.4970039515511303e-05,
      "loss": 0.6881,
      "mean_token_accuracy": 0.7880845361352189,
      "step": 406
    },
    {
      "epoch": 1.2071772105068441,
      "grad_norm": 0.0341012105345726,
      "learning_rate": 1.4940019665897363e-05,
      "loss": 0.6969,
      "mean_token_accuracy": 0.785301415759886,
      "step": 407
    },
    {
      "epoch": 1.2101368849426564,
      "grad_norm": 0.03146205097436905,
      "learning_rate": 1.4909940809733223e-05,
      "loss": 0.7439,
      "mean_token_accuracy": 0.7705267279680645,
      "step": 408
    },
    {
      "epoch": 1.2130965593784684,
      "grad_norm": 0.03223543241620064,
      "learning_rate": 1.4879803306298736e-05,
      "loss": 0.7044,
      "mean_token_accuracy": 0.7822857797292601,
      "step": 409
    },
    {
      "epoch": 1.2160562338142804,
      "grad_norm": 0.03370080888271332,
      "learning_rate": 1.4849607515574276e-05,
      "loss": 0.6567,
      "mean_token_accuracy": 0.7980374918045445,
      "step": 410
    },
    {
      "epoch": 1.2190159082500924,
      "grad_norm": 0.02980733849108219,
      "learning_rate": 1.4819353798236427e-05,
      "loss": 0.7073,
      "mean_token_accuracy": 0.781863166451634,
      "step": 411
    },
    {
      "epoch": 1.2219755826859044,
      "grad_norm": 0.03150355443358421,
      "learning_rate": 1.4789042515653687e-05,
      "loss": 0.7022,
      "mean_token_accuracy": 0.7834270768353983,
      "step": 412
    },
    {
      "epoch": 1.2249352571217167,
      "grad_norm": 0.034390807151794434,
      "learning_rate": 1.4758674029882152e-05,
      "loss": 0.6618,
      "mean_token_accuracy": 0.7955240786394708,
      "step": 413
    },
    {
      "epoch": 1.2278949315575287,
      "grad_norm": 0.03246401622891426,
      "learning_rate": 1.4728248703661183e-05,
      "loss": 0.7031,
      "mean_token_accuracy": 0.7822834100695022,
      "step": 414
    },
    {
      "epoch": 1.2308546059933407,
      "grad_norm": 0.029854416847229004,
      "learning_rate": 1.4697766900409076e-05,
      "loss": 0.677,
      "mean_token_accuracy": 0.7911603334828089,
      "step": 415
    },
    {
      "epoch": 1.2338142804291528,
      "grad_norm": 0.030364803969860077,
      "learning_rate": 1.466722898421873e-05,
      "loss": 0.709,
      "mean_token_accuracy": 0.7811259247175466,
      "step": 416
    },
    {
      "epoch": 1.2367739548649648,
      "grad_norm": 0.03201625868678093,
      "learning_rate": 1.4636635319853274e-05,
      "loss": 0.676,
      "mean_token_accuracy": 0.790977511963359,
      "step": 417
    },
    {
      "epoch": 1.239733629300777,
      "grad_norm": 0.02855406329035759,
      "learning_rate": 1.4605986272741748e-05,
      "loss": 0.6981,
      "mean_token_accuracy": 0.785527533021502,
      "step": 418
    },
    {
      "epoch": 1.242693303736589,
      "grad_norm": 0.029217202216386795,
      "learning_rate": 1.4575282208974704e-05,
      "loss": 0.6897,
      "mean_token_accuracy": 0.7877112145346629,
      "step": 419
    },
    {
      "epoch": 1.245652978172401,
      "grad_norm": 0.030246706679463387,
      "learning_rate": 1.4544523495299843e-05,
      "loss": 0.6911,
      "mean_token_accuracy": 0.7860055778545931,
      "step": 420
    },
    {
      "epoch": 1.248612652608213,
      "grad_norm": 0.028721818700432777,
      "learning_rate": 1.4513710499117648e-05,
      "loss": 0.6379,
      "mean_token_accuracy": 0.8035476022671627,
      "step": 421
    },
    {
      "epoch": 1.251572327044025,
      "grad_norm": 0.02835766412317753,
      "learning_rate": 1.4482843588476976e-05,
      "loss": 0.7029,
      "mean_token_accuracy": 0.7822389641279357,
      "step": 422
    },
    {
      "epoch": 1.2545320014798373,
      "grad_norm": 0.030195439234375954,
      "learning_rate": 1.445192313207067e-05,
      "loss": 0.6853,
      "mean_token_accuracy": 0.7896894172789144,
      "step": 423
    },
    {
      "epoch": 1.2574916759156491,
      "grad_norm": 0.029506726190447807,
      "learning_rate": 1.4420949499231172e-05,
      "loss": 0.6886,
      "mean_token_accuracy": 0.7870159629402342,
      "step": 424
    },
    {
      "epoch": 1.2604513503514614,
      "grad_norm": 0.028463300317525864,
      "learning_rate": 1.4389923059926064e-05,
      "loss": 0.6795,
      "mean_token_accuracy": 0.7904310451756502,
      "step": 425
    },
    {
      "epoch": 1.2634110247872734,
      "grad_norm": 0.031092476099729538,
      "learning_rate": 1.4358844184753713e-05,
      "loss": 0.6585,
      "mean_token_accuracy": 0.7961934733751264,
      "step": 426
    },
    {
      "epoch": 1.2663706992230854,
      "grad_norm": 0.03061164729297161,
      "learning_rate": 1.432771324493879e-05,
      "loss": 0.6824,
      "mean_token_accuracy": 0.7889904725449806,
      "step": 427
    },
    {
      "epoch": 1.2693303736588974,
      "grad_norm": 0.03030516766011715,
      "learning_rate": 1.4296530612327864e-05,
      "loss": 0.7121,
      "mean_token_accuracy": 0.7804462426480531,
      "step": 428
    },
    {
      "epoch": 1.2722900480947095,
      "grad_norm": 0.02747941017150879,
      "learning_rate": 1.4265296659384956e-05,
      "loss": 0.7098,
      "mean_token_accuracy": 0.7817476233877318,
      "step": 429
    },
    {
      "epoch": 1.2752497225305217,
      "grad_norm": 0.030901644378900528,
      "learning_rate": 1.4234011759187084e-05,
      "loss": 0.6986,
      "mean_token_accuracy": 0.7867150025833007,
      "step": 430
    },
    {
      "epoch": 1.2782093969663337,
      "grad_norm": 0.027185294777154922,
      "learning_rate": 1.4202676285419811e-05,
      "loss": 0.6511,
      "mean_token_accuracy": 0.7984791356153214,
      "step": 431
    },
    {
      "epoch": 1.2811690714021458,
      "grad_norm": 0.029310813173651695,
      "learning_rate": 1.4171290612372781e-05,
      "loss": 0.6986,
      "mean_token_accuracy": 0.784810926660307,
      "step": 432
    },
    {
      "epoch": 1.2841287458379578,
      "grad_norm": 0.029999373480677605,
      "learning_rate": 1.4139855114935253e-05,
      "loss": 0.6729,
      "mean_token_accuracy": 0.793846290501721,
      "step": 433
    },
    {
      "epoch": 1.2870884202737698,
      "grad_norm": 0.027550166472792625,
      "learning_rate": 1.410837016859161e-05,
      "loss": 0.682,
      "mean_token_accuracy": 0.7892334348310344,
      "step": 434
    },
    {
      "epoch": 1.290048094709582,
      "grad_norm": 0.028653500601649284,
      "learning_rate": 1.4076836149416889e-05,
      "loss": 0.6669,
      "mean_token_accuracy": 0.7949336027322277,
      "step": 435
    },
    {
      "epoch": 1.293007769145394,
      "grad_norm": 0.03023136779665947,
      "learning_rate": 1.4045253434072278e-05,
      "loss": 0.7204,
      "mean_token_accuracy": 0.778514000463678,
      "step": 436
    },
    {
      "epoch": 1.295967443581206,
      "grad_norm": 0.032578323036432266,
      "learning_rate": 1.4013622399800628e-05,
      "loss": 0.7322,
      "mean_token_accuracy": 0.7759616268580662,
      "step": 437
    },
    {
      "epoch": 1.298927118017018,
      "grad_norm": 0.03219889849424362,
      "learning_rate": 1.3981943424421932e-05,
      "loss": 0.7057,
      "mean_token_accuracy": 0.7823459917807589,
      "step": 438
    },
    {
      "epoch": 1.3018867924528301,
      "grad_norm": 0.029690273106098175,
      "learning_rate": 1.3950216886328818e-05,
      "loss": 0.7054,
      "mean_token_accuracy": 0.7824452622468316,
      "step": 439
    },
    {
      "epoch": 1.3048464668886424,
      "grad_norm": 0.0296969935297966,
      "learning_rate": 1.3918443164482048e-05,
      "loss": 0.7041,
      "mean_token_accuracy": 0.7845265919170226,
      "step": 440
    },
    {
      "epoch": 1.3078061413244544,
      "grad_norm": 0.027904395014047623,
      "learning_rate": 1.3886622638405953e-05,
      "loss": 0.703,
      "mean_token_accuracy": 0.7837437940026634,
      "step": 441
    },
    {
      "epoch": 1.3107658157602664,
      "grad_norm": 0.029914019629359245,
      "learning_rate": 1.3854755688183941e-05,
      "loss": 0.73,
      "mean_token_accuracy": 0.7762600806136151,
      "step": 442
    },
    {
      "epoch": 1.3137254901960784,
      "grad_norm": 0.029954560101032257,
      "learning_rate": 1.3822842694453923e-05,
      "loss": 0.6964,
      "mean_token_accuracy": 0.7856698827775562,
      "step": 443
    },
    {
      "epoch": 1.3166851646318904,
      "grad_norm": 0.030304700136184692,
      "learning_rate": 1.3790884038403796e-05,
      "loss": 0.6993,
      "mean_token_accuracy": 0.7846048101626529,
      "step": 444
    },
    {
      "epoch": 1.3196448390677027,
      "grad_norm": 0.031101198866963387,
      "learning_rate": 1.375888010176686e-05,
      "loss": 0.6745,
      "mean_token_accuracy": 0.791563660302581,
      "step": 445
    },
    {
      "epoch": 1.3226045135035145,
      "grad_norm": 0.030776219442486763,
      "learning_rate": 1.3726831266817278e-05,
      "loss": 0.696,
      "mean_token_accuracy": 0.7863943901455982,
      "step": 446
    },
    {
      "epoch": 1.3255641879393267,
      "grad_norm": 0.02914304845035076,
      "learning_rate": 1.3694737916365517e-05,
      "loss": 0.7097,
      "mean_token_accuracy": 0.7815690533016747,
      "step": 447
    },
    {
      "epoch": 1.3285238623751388,
      "grad_norm": 0.02826066128909588,
      "learning_rate": 1.3662600433753746e-05,
      "loss": 0.6974,
      "mean_token_accuracy": 0.7859719689506738,
      "step": 448
    },
    {
      "epoch": 1.3314835368109508,
      "grad_norm": 0.028485247865319252,
      "learning_rate": 1.3630419202851287e-05,
      "loss": 0.7054,
      "mean_token_accuracy": 0.7820680554449749,
      "step": 449
    },
    {
      "epoch": 1.3344432112467628,
      "grad_norm": 0.029415471479296684,
      "learning_rate": 1.3598194608050011e-05,
      "loss": 0.7125,
      "mean_token_accuracy": 0.7813712762181255,
      "step": 450
    },
    {
      "epoch": 1.3374028856825748,
      "grad_norm": 0.030356241390109062,
      "learning_rate": 1.3565927034259757e-05,
      "loss": 0.7038,
      "mean_token_accuracy": 0.784404349262648,
      "step": 451
    },
    {
      "epoch": 1.340362560118387,
      "grad_norm": 0.02904743328690529,
      "learning_rate": 1.3533616866903736e-05,
      "loss": 0.6852,
      "mean_token_accuracy": 0.7884821796631603,
      "step": 452
    },
    {
      "epoch": 1.343322234554199,
      "grad_norm": 0.028087515383958817,
      "learning_rate": 1.3501264491913909e-05,
      "loss": 0.7078,
      "mean_token_accuracy": 0.7810583202817751,
      "step": 453
    },
    {
      "epoch": 1.346281908990011,
      "grad_norm": 0.03136244788765907,
      "learning_rate": 1.3468870295726399e-05,
      "loss": 0.7278,
      "mean_token_accuracy": 0.7758812655015149,
      "step": 454
    },
    {
      "epoch": 1.3492415834258231,
      "grad_norm": 0.027985604479908943,
      "learning_rate": 1.3436434665276865e-05,
      "loss": 0.6818,
      "mean_token_accuracy": 0.7898017303776185,
      "step": 455
    },
    {
      "epoch": 1.3522012578616351,
      "grad_norm": 0.02957519330084324,
      "learning_rate": 1.3403957987995884e-05,
      "loss": 0.6879,
      "mean_token_accuracy": 0.787421315955905,
      "step": 456
    },
    {
      "epoch": 1.3551609322974474,
      "grad_norm": 0.028551606461405754,
      "learning_rate": 1.3371440651804313e-05,
      "loss": 0.7142,
      "mean_token_accuracy": 0.7799032924651638,
      "step": 457
    },
    {
      "epoch": 1.3581206067332594,
      "grad_norm": 0.029369303956627846,
      "learning_rate": 1.3338883045108674e-05,
      "loss": 0.6926,
      "mean_token_accuracy": 0.786089013532344,
      "step": 458
    },
    {
      "epoch": 1.3610802811690714,
      "grad_norm": 0.031690262258052826,
      "learning_rate": 1.3306285556796494e-05,
      "loss": 0.6951,
      "mean_token_accuracy": 0.7869912621818624,
      "step": 459
    },
    {
      "epoch": 1.3640399556048834,
      "grad_norm": 0.027342651039361954,
      "learning_rate": 1.327364857623168e-05,
      "loss": 0.7086,
      "mean_token_accuracy": 0.7826821517369422,
      "step": 460
    },
    {
      "epoch": 1.3669996300406955,
      "grad_norm": 0.030994607135653496,
      "learning_rate": 1.3240972493249846e-05,
      "loss": 0.6984,
      "mean_token_accuracy": 0.7852839444382589,
      "step": 461
    },
    {
      "epoch": 1.3699593044765077,
      "grad_norm": 0.027840347960591316,
      "learning_rate": 1.3208257698153677e-05,
      "loss": 0.7264,
      "mean_token_accuracy": 0.7753326959711423,
      "step": 462
    },
    {
      "epoch": 1.3729189789123195,
      "grad_norm": 0.030181415379047394,
      "learning_rate": 1.3175504581708261e-05,
      "loss": 0.6959,
      "mean_token_accuracy": 0.7858881660692992,
      "step": 463
    },
    {
      "epoch": 1.3758786533481318,
      "grad_norm": 0.02758677490055561,
      "learning_rate": 1.3142713535136413e-05,
      "loss": 0.7041,
      "mean_token_accuracy": 0.7823435129723749,
      "step": 464
    },
    {
      "epoch": 1.3788383277839438,
      "grad_norm": 0.031985413283109665,
      "learning_rate": 1.3109884950114007e-05,
      "loss": 0.7056,
      "mean_token_accuracy": 0.7814022483974902,
      "step": 465
    },
    {
      "epoch": 1.3817980022197558,
      "grad_norm": 0.03063722513616085,
      "learning_rate": 1.3077019218765306e-05,
      "loss": 0.6761,
      "mean_token_accuracy": 0.7908319957754819,
      "step": 466
    },
    {
      "epoch": 1.3847576766555678,
      "grad_norm": 0.02932468056678772,
      "learning_rate": 1.3044116733658261e-05,
      "loss": 0.6818,
      "mean_token_accuracy": 0.7889467099257529,
      "step": 467
    },
    {
      "epoch": 1.3877173510913798,
      "grad_norm": 0.029189975932240486,
      "learning_rate": 1.3011177887799846e-05,
      "loss": 0.6669,
      "mean_token_accuracy": 0.7953069364813705,
      "step": 468
    },
    {
      "epoch": 1.390677025527192,
      "grad_norm": 0.029496444389224052,
      "learning_rate": 1.2978203074631335e-05,
      "loss": 0.6909,
      "mean_token_accuracy": 0.7868717328013188,
      "step": 469
    },
    {
      "epoch": 1.393636699963004,
      "grad_norm": 0.0296261478215456,
      "learning_rate": 1.2945192688023625e-05,
      "loss": 0.7306,
      "mean_token_accuracy": 0.7760671546041664,
      "step": 470
    },
    {
      "epoch": 1.3965963743988161,
      "grad_norm": 0.03375551104545593,
      "learning_rate": 1.2912147122272523e-05,
      "loss": 0.6994,
      "mean_token_accuracy": 0.7834207157496026,
      "step": 471
    },
    {
      "epoch": 1.3995560488346281,
      "grad_norm": 0.03269757702946663,
      "learning_rate": 1.287906677209403e-05,
      "loss": 0.6735,
      "mean_token_accuracy": 0.7919020007052182,
      "step": 472
    },
    {
      "epoch": 1.4025157232704402,
      "grad_norm": 0.0318894125521183,
      "learning_rate": 1.2845952032619651e-05,
      "loss": 0.725,
      "mean_token_accuracy": 0.7789513879429198,
      "step": 473
    },
    {
      "epoch": 1.4054753977062524,
      "grad_norm": 0.030185015872120857,
      "learning_rate": 1.2812803299391629e-05,
      "loss": 0.736,
      "mean_token_accuracy": 0.7739720307998011,
      "step": 474
    },
    {
      "epoch": 1.4084350721420644,
      "grad_norm": 0.03134026750922203,
      "learning_rate": 1.2779620968358276e-05,
      "loss": 0.6657,
      "mean_token_accuracy": 0.79394898639027,
      "step": 475
    },
    {
      "epoch": 1.4113947465778764,
      "grad_norm": 0.03099486045539379,
      "learning_rate": 1.2746405435869198e-05,
      "loss": 0.6746,
      "mean_token_accuracy": 0.7901352687212624,
      "step": 476
    },
    {
      "epoch": 1.4143544210136885,
      "grad_norm": 0.031111029908061028,
      "learning_rate": 1.271315709867059e-05,
      "loss": 0.7108,
      "mean_token_accuracy": 0.7808969929086254,
      "step": 477
    },
    {
      "epoch": 1.4173140954495005,
      "grad_norm": 0.031750526279211044,
      "learning_rate": 1.2679876353900482e-05,
      "loss": 0.7155,
      "mean_token_accuracy": 0.7798662047700707,
      "step": 478
    },
    {
      "epoch": 1.4202737698853127,
      "grad_norm": 0.02914981171488762,
      "learning_rate": 1.2646563599083997e-05,
      "loss": 0.7316,
      "mean_token_accuracy": 0.7748036737839163,
      "step": 479
    },
    {
      "epoch": 1.4232334443211248,
      "grad_norm": 0.03101126290857792,
      "learning_rate": 1.2613219232128608e-05,
      "loss": 0.6703,
      "mean_token_accuracy": 0.7929325497223755,
      "step": 480
    },
    {
      "epoch": 1.4261931187569368,
      "grad_norm": 0.03043644316494465,
      "learning_rate": 1.2579843651319382e-05,
      "loss": 0.7408,
      "mean_token_accuracy": 0.7704327906858951,
      "step": 481
    },
    {
      "epoch": 1.4291527931927488,
      "grad_norm": 0.03125700727105141,
      "learning_rate": 1.2546437255314223e-05,
      "loss": 0.6725,
      "mean_token_accuracy": 0.7928834670777763,
      "step": 482
    },
    {
      "epoch": 1.4321124676285608,
      "grad_norm": 0.02958809956908226,
      "learning_rate": 1.2513000443139112e-05,
      "loss": 0.7004,
      "mean_token_accuracy": 0.7830677966303454,
      "step": 483
    },
    {
      "epoch": 1.435072142064373,
      "grad_norm": 0.028092646971344948,
      "learning_rate": 1.2479533614183334e-05,
      "loss": 0.6859,
      "mean_token_accuracy": 0.788886189737907,
      "step": 484
    },
    {
      "epoch": 1.4380318165001849,
      "grad_norm": 0.02891010046005249,
      "learning_rate": 1.2446037168194716e-05,
      "loss": 0.7025,
      "mean_token_accuracy": 0.7828560085056796,
      "step": 485
    },
    {
      "epoch": 1.440991490935997,
      "grad_norm": 0.030256103724241257,
      "learning_rate": 1.2412511505274845e-05,
      "loss": 0.6677,
      "mean_token_accuracy": 0.7945832908557738,
      "step": 486
    },
    {
      "epoch": 1.4439511653718091,
      "grad_norm": 0.029633628204464912,
      "learning_rate": 1.23789570258743e-05,
      "loss": 0.7028,
      "mean_token_accuracy": 0.7822005274678082,
      "step": 487
    },
    {
      "epoch": 1.4469108398076211,
      "grad_norm": 0.029555251821875572,
      "learning_rate": 1.2345374130787855e-05,
      "loss": 0.7005,
      "mean_token_accuracy": 0.7844486052731944,
      "step": 488
    },
    {
      "epoch": 1.4498705142434332,
      "grad_norm": 0.028284987434744835,
      "learning_rate": 1.23117632211497e-05,
      "loss": 0.6636,
      "mean_token_accuracy": 0.7948520620989358,
      "step": 489
    },
    {
      "epoch": 1.4528301886792452,
      "grad_norm": 0.029071122407913208,
      "learning_rate": 1.2278124698428643e-05,
      "loss": 0.692,
      "mean_token_accuracy": 0.785599726958277,
      "step": 490
    },
    {
      "epoch": 1.4557898631150574,
      "grad_norm": 0.026755662634968758,
      "learning_rate": 1.2244458964423328e-05,
      "loss": 0.7028,
      "mean_token_accuracy": 0.7821296757882308,
      "step": 491
    },
    {
      "epoch": 1.4587495375508694,
      "grad_norm": 0.028367044404149055,
      "learning_rate": 1.221076642125742e-05,
      "loss": 0.6993,
      "mean_token_accuracy": 0.784891485056386,
      "step": 492
    },
    {
      "epoch": 1.4617092119866815,
      "grad_norm": 0.029388481751084328,
      "learning_rate": 1.2177047471374808e-05,
      "loss": 0.6751,
      "mean_token_accuracy": 0.7925045197001166,
      "step": 493
    },
    {
      "epoch": 1.4646688864224935,
      "grad_norm": 0.02947191148996353,
      "learning_rate": 1.214330251753481e-05,
      "loss": 0.6854,
      "mean_token_accuracy": 0.7895612954928821,
      "step": 494
    },
    {
      "epoch": 1.4676285608583055,
      "grad_norm": 0.028657473623752594,
      "learning_rate": 1.2109531962807333e-05,
      "loss": 0.6639,
      "mean_token_accuracy": 0.7951647984782646,
      "step": 495
    },
    {
      "epoch": 1.4705882352941178,
      "grad_norm": 0.029021482914686203,
      "learning_rate": 1.207573621056809e-05,
      "loss": 0.6857,
      "mean_token_accuracy": 0.7899984070146465,
      "step": 496
    },
    {
      "epoch": 1.4735479097299298,
      "grad_norm": 0.029676727950572968,
      "learning_rate": 1.2041915664493763e-05,
      "loss": 0.7187,
      "mean_token_accuracy": 0.7797242570868168,
      "step": 497
    },
    {
      "epoch": 1.4765075841657418,
      "grad_norm": 0.028637081384658813,
      "learning_rate": 1.2008070728557186e-05,
      "loss": 0.7026,
      "mean_token_accuracy": 0.7819584120529711,
      "step": 498
    },
    {
      "epoch": 1.4794672586015538,
      "grad_norm": 0.027969013899564743,
      "learning_rate": 1.1974201807022525e-05,
      "loss": 0.667,
      "mean_token_accuracy": 0.7936783567610219,
      "step": 499
    },
    {
      "epoch": 1.4824269330373658,
      "grad_norm": 0.027376718819141388,
      "learning_rate": 1.1940309304440434e-05,
      "loss": 0.6623,
      "mean_token_accuracy": 0.7948291241642872,
      "step": 500
    },
    {
      "epoch": 1.4824269330373658,
      "eval_loss": 0.7481726408004761,
      "eval_mean_token_accuracy": 0.7689154129322073,
      "eval_runtime": 25.172,
      "eval_samples_per_second": 5.125,
      "eval_steps_per_second": 1.311,
      "step": 500
    },
    {
      "epoch": 1.485386607473178,
      "grad_norm": 0.026877794414758682,
      "learning_rate": 1.1906393625643244e-05,
      "loss": 0.6985,
      "mean_token_accuracy": 0.7859683500641782,
      "step": 501
    },
    {
      "epoch": 1.4883462819089899,
      "grad_norm": 0.02793138287961483,
      "learning_rate": 1.1872455175740111e-05,
      "loss": 0.7117,
      "mean_token_accuracy": 0.7810684257549364,
      "step": 502
    },
    {
      "epoch": 1.4913059563448021,
      "grad_norm": 0.029749806970357895,
      "learning_rate": 1.1838494360112185e-05,
      "loss": 0.6905,
      "mean_token_accuracy": 0.7874104240423798,
      "step": 503
    },
    {
      "epoch": 1.4942656307806141,
      "grad_norm": 0.027249010279774666,
      "learning_rate": 1.1804511584407763e-05,
      "loss": 0.654,
      "mean_token_accuracy": 0.7970481968001146,
      "step": 504
    },
    {
      "epoch": 1.4972253052164262,
      "grad_norm": 0.02847236581146717,
      "learning_rate": 1.1770507254537454e-05,
      "loss": 0.664,
      "mean_token_accuracy": 0.7958982191365208,
      "step": 505
    },
    {
      "epoch": 1.5001849796522384,
      "grad_norm": 0.029989562928676605,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.673,
      "mean_token_accuracy": 0.7924119981004011,
      "step": 506
    },
    {
      "epoch": 1.5031446540880502,
      "grad_norm": 0.029826542362570763,
      "learning_rate": 1.1702435557223988e-05,
      "loss": 0.7413,
      "mean_token_accuracy": 0.7711602345991981,
      "step": 507
    },
    {
      "epoch": 1.5061043285238624,
      "grad_norm": 0.02937629073858261,
      "learning_rate": 1.1668369002869912e-05,
      "loss": 0.7031,
      "mean_token_accuracy": 0.7816943042937375,
      "step": 508
    },
    {
      "epoch": 1.5090640029596745,
      "grad_norm": 0.030177390202879906,
      "learning_rate": 1.1634282520518382e-05,
      "loss": 0.692,
      "mean_token_accuracy": 0.7859999122949272,
      "step": 509
    },
    {
      "epoch": 1.5120236773954865,
      "grad_norm": 0.027903199195861816,
      "learning_rate": 1.1600176517318742e-05,
      "loss": 0.7095,
      "mean_token_accuracy": 0.781658248832549,
      "step": 510
    },
    {
      "epoch": 1.5149833518312985,
      "grad_norm": 0.026245184242725372,
      "learning_rate": 1.1566051400653486e-05,
      "loss": 0.6969,
      "mean_token_accuracy": 0.7867325435023516,
      "step": 511
    },
    {
      "epoch": 1.5179430262671105,
      "grad_norm": 0.028284091502428055,
      "learning_rate": 1.153190757813343e-05,
      "loss": 0.6737,
      "mean_token_accuracy": 0.7917562171923704,
      "step": 512
    },
    {
      "epoch": 1.5209027007029228,
      "grad_norm": 0.02920420840382576,
      "learning_rate": 1.1497745457592817e-05,
      "loss": 0.7015,
      "mean_token_accuracy": 0.7843454800190588,
      "step": 513
    },
    {
      "epoch": 1.5238623751387348,
      "grad_norm": 0.029416844248771667,
      "learning_rate": 1.1463565447084446e-05,
      "loss": 0.6782,
      "mean_token_accuracy": 0.7905477218838749,
      "step": 514
    },
    {
      "epoch": 1.5268220495745468,
      "grad_norm": 0.02845824882388115,
      "learning_rate": 1.142936795487482e-05,
      "loss": 0.7105,
      "mean_token_accuracy": 0.7825743294112035,
      "step": 515
    },
    {
      "epoch": 1.5297817240103588,
      "grad_norm": 0.02964814379811287,
      "learning_rate": 1.1395153389439232e-05,
      "loss": 0.6871,
      "mean_token_accuracy": 0.7873465689406989,
      "step": 516
    },
    {
      "epoch": 1.5327413984461709,
      "grad_norm": 0.030287781730294228,
      "learning_rate": 1.1360922159456929e-05,
      "loss": 0.6588,
      "mean_token_accuracy": 0.7957205791765152,
      "step": 517
    },
    {
      "epoch": 1.535701072881983,
      "grad_norm": 0.027695264667272568,
      "learning_rate": 1.1326674673806195e-05,
      "loss": 0.653,
      "mean_token_accuracy": 0.7976352575238422,
      "step": 518
    },
    {
      "epoch": 1.538660747317795,
      "grad_norm": 0.02871226891875267,
      "learning_rate": 1.129241134155949e-05,
      "loss": 0.73,
      "mean_token_accuracy": 0.7754833343000982,
      "step": 519
    },
    {
      "epoch": 1.5416204217536071,
      "grad_norm": 0.031651414930820465,
      "learning_rate": 1.1258132571978555e-05,
      "loss": 0.6941,
      "mean_token_accuracy": 0.7848089037330863,
      "step": 520
    },
    {
      "epoch": 1.5445800961894192,
      "grad_norm": 0.028931137174367905,
      "learning_rate": 1.1223838774509515e-05,
      "loss": 0.6867,
      "mean_token_accuracy": 0.7880591204844969,
      "step": 521
    },
    {
      "epoch": 1.5475397706252312,
      "grad_norm": 0.028498172760009766,
      "learning_rate": 1.1189530358778005e-05,
      "loss": 0.6781,
      "mean_token_accuracy": 0.7904013973143751,
      "step": 522
    },
    {
      "epoch": 1.5504994450610434,
      "grad_norm": 0.02972579561173916,
      "learning_rate": 1.1155207734584264e-05,
      "loss": 0.662,
      "mean_token_accuracy": 0.7951280948805273,
      "step": 523
    },
    {
      "epoch": 1.5534591194968552,
      "grad_norm": 0.02952287159860134,
      "learning_rate": 1.1120871311898254e-05,
      "loss": 0.6693,
      "mean_token_accuracy": 0.7932455871418521,
      "step": 524
    },
    {
      "epoch": 1.5564187939326675,
      "grad_norm": 0.027867093682289124,
      "learning_rate": 1.1086521500854746e-05,
      "loss": 0.6814,
      "mean_token_accuracy": 0.7888936509239802,
      "step": 525
    },
    {
      "epoch": 1.5593784683684795,
      "grad_norm": 0.02665465883910656,
      "learning_rate": 1.1052158711748435e-05,
      "loss": 0.6499,
      "mean_token_accuracy": 0.7985216192089608,
      "step": 526
    },
    {
      "epoch": 1.5623381428042915,
      "grad_norm": 0.02734951861202717,
      "learning_rate": 1.1017783355029027e-05,
      "loss": 0.7045,
      "mean_token_accuracy": 0.7839218834760857,
      "step": 527
    },
    {
      "epoch": 1.5652978172401038,
      "grad_norm": 0.028345216065645218,
      "learning_rate": 1.0983395841296349e-05,
      "loss": 0.7102,
      "mean_token_accuracy": 0.7821431227738681,
      "step": 528
    },
    {
      "epoch": 1.5682574916759155,
      "grad_norm": 0.0284163486212492,
      "learning_rate": 1.0948996581295437e-05,
      "loss": 0.6787,
      "mean_token_accuracy": 0.790229323637791,
      "step": 529
    },
    {
      "epoch": 1.5712171661117278,
      "grad_norm": 0.0328470915555954,
      "learning_rate": 1.0914585985911632e-05,
      "loss": 0.7276,
      "mean_token_accuracy": 0.778051425264871,
      "step": 530
    },
    {
      "epoch": 1.5741768405475398,
      "grad_norm": 0.02768307365477085,
      "learning_rate": 1.0880164466165675e-05,
      "loss": 0.6878,
      "mean_token_accuracy": 0.7872839620098567,
      "step": 531
    },
    {
      "epoch": 1.5771365149833518,
      "grad_norm": 0.028808310627937317,
      "learning_rate": 1.084573243320878e-05,
      "loss": 0.7076,
      "mean_token_accuracy": 0.7827423762289634,
      "step": 532
    },
    {
      "epoch": 1.5800961894191639,
      "grad_norm": 0.029864028096199036,
      "learning_rate": 1.0811290298317755e-05,
      "loss": 0.704,
      "mean_token_accuracy": 0.7836907614704345,
      "step": 533
    },
    {
      "epoch": 1.5830558638549759,
      "grad_norm": 0.027310620993375778,
      "learning_rate": 1.0776838472890065e-05,
      "loss": 0.6688,
      "mean_token_accuracy": 0.7929893034238038,
      "step": 534
    },
    {
      "epoch": 1.5860155382907881,
      "grad_norm": 0.028434861451387405,
      "learning_rate": 1.0742377368438915e-05,
      "loss": 0.6728,
      "mean_token_accuracy": 0.7923653247941337,
      "step": 535
    },
    {
      "epoch": 1.5889752127266,
      "grad_norm": 0.02753947488963604,
      "learning_rate": 1.0707907396588362e-05,
      "loss": 0.6824,
      "mean_token_accuracy": 0.789288914123542,
      "step": 536
    },
    {
      "epoch": 1.5919348871624122,
      "grad_norm": 0.028986243531107903,
      "learning_rate": 1.0673428969068365e-05,
      "loss": 0.6764,
      "mean_token_accuracy": 0.7918605149099156,
      "step": 537
    },
    {
      "epoch": 1.5948945615982242,
      "grad_norm": 0.028384177014231682,
      "learning_rate": 1.063894249770989e-05,
      "loss": 0.7226,
      "mean_token_accuracy": 0.7771391941220901,
      "step": 538
    },
    {
      "epoch": 1.5978542360340362,
      "grad_norm": 0.03014584816992283,
      "learning_rate": 1.0604448394439983e-05,
      "loss": 0.6958,
      "mean_token_accuracy": 0.7866888127818252,
      "step": 539
    },
    {
      "epoch": 1.6008139104698484,
      "grad_norm": 0.027700098231434822,
      "learning_rate": 1.0569947071276847e-05,
      "loss": 0.6854,
      "mean_token_accuracy": 0.7884833991707515,
      "step": 540
    },
    {
      "epoch": 1.6037735849056602,
      "grad_norm": 0.030186597257852554,
      "learning_rate": 1.053543894032493e-05,
      "loss": 0.6555,
      "mean_token_accuracy": 0.7966581896868427,
      "step": 541
    },
    {
      "epoch": 1.6067332593414725,
      "grad_norm": 0.028836503624916077,
      "learning_rate": 1.0500924413769988e-05,
      "loss": 0.7105,
      "mean_token_accuracy": 0.7802640782055931,
      "step": 542
    },
    {
      "epoch": 1.6096929337772845,
      "grad_norm": 0.027741001918911934,
      "learning_rate": 1.0466403903874176e-05,
      "loss": 0.6766,
      "mean_token_accuracy": 0.7904056510280758,
      "step": 543
    },
    {
      "epoch": 1.6126526082130965,
      "grad_norm": 0.02981194108724594,
      "learning_rate": 1.0431877822971118e-05,
      "loss": 0.734,
      "mean_token_accuracy": 0.7746203477446459,
      "step": 544
    },
    {
      "epoch": 1.6156122826489088,
      "grad_norm": 0.028873400762677193,
      "learning_rate": 1.0397346583460972e-05,
      "loss": 0.6822,
      "mean_token_accuracy": 0.7885805178222033,
      "step": 545
    },
    {
      "epoch": 1.6185719570847206,
      "grad_norm": 0.027645166963338852,
      "learning_rate": 1.0362810597805526e-05,
      "loss": 0.7259,
      "mean_token_accuracy": 0.7786934059112492,
      "step": 546
    },
    {
      "epoch": 1.6215316315205328,
      "grad_norm": 0.02878463827073574,
      "learning_rate": 1.0328270278523256e-05,
      "loss": 0.7,
      "mean_token_accuracy": 0.7848884117116051,
      "step": 547
    },
    {
      "epoch": 1.6244913059563448,
      "grad_norm": 0.028918178752064705,
      "learning_rate": 1.0293726038184393e-05,
      "loss": 0.6734,
      "mean_token_accuracy": 0.791649593823967,
      "step": 548
    },
    {
      "epoch": 1.6274509803921569,
      "grad_norm": 0.02728886716067791,
      "learning_rate": 1.0259178289406011e-05,
      "loss": 0.6907,
      "mean_token_accuracy": 0.7856968583915338,
      "step": 549
    },
    {
      "epoch": 1.6304106548279689,
      "grad_norm": 0.03078787960112095,
      "learning_rate": 1.022462744484709e-05,
      "loss": 0.6828,
      "mean_token_accuracy": 0.7884847328062977,
      "step": 550
    },
    {
      "epoch": 1.633370329263781,
      "grad_norm": 0.027883974835276604,
      "learning_rate": 1.019007391720359e-05,
      "loss": 0.6498,
      "mean_token_accuracy": 0.7989072483330585,
      "step": 551
    },
    {
      "epoch": 1.6363300036995931,
      "grad_norm": 0.02902034856379032,
      "learning_rate": 1.0155518119203511e-05,
      "loss": 0.6557,
      "mean_token_accuracy": 0.7977721855590214,
      "step": 552
    },
    {
      "epoch": 1.6392896781354052,
      "grad_norm": 0.027060616761446,
      "learning_rate": 1.0120960463601977e-05,
      "loss": 0.696,
      "mean_token_accuracy": 0.7850648769503469,
      "step": 553
    },
    {
      "epoch": 1.6422493525712172,
      "grad_norm": 0.02980673313140869,
      "learning_rate": 1.0086401363176306e-05,
      "loss": 0.7094,
      "mean_token_accuracy": 0.7815210626510076,
      "step": 554
    },
    {
      "epoch": 1.6452090270070292,
      "grad_norm": 0.02940969541668892,
      "learning_rate": 1.0051841230721065e-05,
      "loss": 0.7099,
      "mean_token_accuracy": 0.7816078970457939,
      "step": 555
    },
    {
      "epoch": 1.6481687014428412,
      "grad_norm": 0.030654923990368843,
      "learning_rate": 1.0017280479043148e-05,
      "loss": 0.6906,
      "mean_token_accuracy": 0.7858616875207767,
      "step": 556
    },
    {
      "epoch": 1.6511283758786535,
      "grad_norm": 0.02817557565867901,
      "learning_rate": 9.982719520956856e-06,
      "loss": 0.7012,
      "mean_token_accuracy": 0.7839192902734954,
      "step": 557
    },
    {
      "epoch": 1.6540880503144653,
      "grad_norm": 0.028098734095692635,
      "learning_rate": 9.948158769278939e-06,
      "loss": 0.6575,
      "mean_token_accuracy": 0.7954805082471507,
      "step": 558
    },
    {
      "epoch": 1.6570477247502775,
      "grad_norm": 0.02810637094080448,
      "learning_rate": 9.913598636823694e-06,
      "loss": 0.6788,
      "mean_token_accuracy": 0.7905485621548396,
      "step": 559
    },
    {
      "epoch": 1.6600073991860895,
      "grad_norm": 0.030547816306352615,
      "learning_rate": 9.879039536398023e-06,
      "loss": 0.6736,
      "mean_token_accuracy": 0.7909516185064236,
      "step": 560
    },
    {
      "epoch": 1.6629670736219015,
      "grad_norm": 0.028826886788010597,
      "learning_rate": 9.844481880796492e-06,
      "loss": 0.6765,
      "mean_token_accuracy": 0.7915564204411224,
      "step": 561
    },
    {
      "epoch": 1.6659267480577138,
      "grad_norm": 0.02765066921710968,
      "learning_rate": 9.809926082796415e-06,
      "loss": 0.6753,
      "mean_token_accuracy": 0.790575449165788,
      "step": 562
    },
    {
      "epoch": 1.6688864224935256,
      "grad_norm": 0.02932964451611042,
      "learning_rate": 9.775372555152912e-06,
      "loss": 0.7171,
      "mean_token_accuracy": 0.7802580640067767,
      "step": 563
    },
    {
      "epoch": 1.6718460969293378,
      "grad_norm": 0.02844865433871746,
      "learning_rate": 9.740821710593989e-06,
      "loss": 0.6797,
      "mean_token_accuracy": 0.7912443602052889,
      "step": 564
    },
    {
      "epoch": 1.6748057713651499,
      "grad_norm": 0.029067976400256157,
      "learning_rate": 9.70627396181561e-06,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7832342601212231,
      "step": 565
    },
    {
      "epoch": 1.6777654458009619,
      "grad_norm": 0.030374307185411453,
      "learning_rate": 9.671729721476747e-06,
      "loss": 0.6853,
      "mean_token_accuracy": 0.7890330391980426,
      "step": 566
    },
    {
      "epoch": 1.6807251202367741,
      "grad_norm": 0.029422570019960403,
      "learning_rate": 9.637189402194477e-06,
      "loss": 0.6705,
      "mean_token_accuracy": 0.7933131421236023,
      "step": 567
    },
    {
      "epoch": 1.683684794672586,
      "grad_norm": 0.029863137751817703,
      "learning_rate": 9.602653416539031e-06,
      "loss": 0.6632,
      "mean_token_accuracy": 0.794249199272412,
      "step": 568
    },
    {
      "epoch": 1.6866444691083982,
      "grad_norm": 0.029888629913330078,
      "learning_rate": 9.568122177028884e-06,
      "loss": 0.6868,
      "mean_token_accuracy": 0.7878144873510193,
      "step": 569
    },
    {
      "epoch": 1.6896041435442102,
      "grad_norm": 0.028478743508458138,
      "learning_rate": 9.533596096125826e-06,
      "loss": 0.7052,
      "mean_token_accuracy": 0.7823080476653739,
      "step": 570
    },
    {
      "epoch": 1.6925638179800222,
      "grad_norm": 0.0327795036137104,
      "learning_rate": 9.499075586230014e-06,
      "loss": 0.7358,
      "mean_token_accuracy": 0.7743240716435916,
      "step": 571
    },
    {
      "epoch": 1.6955234924158342,
      "grad_norm": 0.032194510102272034,
      "learning_rate": 9.464561059675073e-06,
      "loss": 0.689,
      "mean_token_accuracy": 0.7881108192901182,
      "step": 572
    },
    {
      "epoch": 1.6984831668516462,
      "grad_norm": 0.03138433396816254,
      "learning_rate": 9.430052928723153e-06,
      "loss": 0.6766,
      "mean_token_accuracy": 0.7899914931355358,
      "step": 573
    },
    {
      "epoch": 1.7014428412874585,
      "grad_norm": 0.02856632135808468,
      "learning_rate": 9.395551605560018e-06,
      "loss": 0.6615,
      "mean_token_accuracy": 0.792983564239965,
      "step": 574
    },
    {
      "epoch": 1.7044025157232703,
      "grad_norm": 0.027933062985539436,
      "learning_rate": 9.361057502290112e-06,
      "loss": 0.6764,
      "mean_token_accuracy": 0.791124741675713,
      "step": 575
    },
    {
      "epoch": 1.7073621901590825,
      "grad_norm": 0.03196121007204056,
      "learning_rate": 9.326571030931636e-06,
      "loss": 0.6867,
      "mean_token_accuracy": 0.7881935438167171,
      "step": 576
    },
    {
      "epoch": 1.7103218645948945,
      "grad_norm": 0.03020172193646431,
      "learning_rate": 9.292092603411642e-06,
      "loss": 0.6928,
      "mean_token_accuracy": 0.7857717916724626,
      "step": 577
    },
    {
      "epoch": 1.7132815390307066,
      "grad_norm": 0.028756381943821907,
      "learning_rate": 9.257622631561085e-06,
      "loss": 0.6871,
      "mean_token_accuracy": 0.7877761469177995,
      "step": 578
    },
    {
      "epoch": 1.7162412134665188,
      "grad_norm": 0.029508501291275024,
      "learning_rate": 9.223161527109938e-06,
      "loss": 0.7323,
      "mean_token_accuracy": 0.7732244809453903,
      "step": 579
    },
    {
      "epoch": 1.7192008879023306,
      "grad_norm": 0.02920532040297985,
      "learning_rate": 9.188709701682246e-06,
      "loss": 0.687,
      "mean_token_accuracy": 0.7871930277731295,
      "step": 580
    },
    {
      "epoch": 1.7221605623381429,
      "grad_norm": 0.0323316864669323,
      "learning_rate": 9.154267566791224e-06,
      "loss": 0.6822,
      "mean_token_accuracy": 0.7898471637565551,
      "step": 581
    },
    {
      "epoch": 1.7251202367739549,
      "grad_norm": 0.028480004519224167,
      "learning_rate": 9.119835533834332e-06,
      "loss": 0.6517,
      "mean_token_accuracy": 0.7984012253146054,
      "step": 582
    },
    {
      "epoch": 1.728079911209767,
      "grad_norm": 0.026408640667796135,
      "learning_rate": 9.085414014088368e-06,
      "loss": 0.7107,
      "mean_token_accuracy": 0.782095449813971,
      "step": 583
    },
    {
      "epoch": 1.7310395856455791,
      "grad_norm": 0.030046703293919563,
      "learning_rate": 9.051003418704566e-06,
      "loss": 0.687,
      "mean_token_accuracy": 0.7877174303211925,
      "step": 584
    },
    {
      "epoch": 1.733999260081391,
      "grad_norm": 0.029806992039084435,
      "learning_rate": 9.016604158703654e-06,
      "loss": 0.6111,
      "mean_token_accuracy": 0.810001799041272,
      "step": 585
    },
    {
      "epoch": 1.7369589345172032,
      "grad_norm": 0.02893899567425251,
      "learning_rate": 8.982216644970978e-06,
      "loss": 0.7149,
      "mean_token_accuracy": 0.779467916838572,
      "step": 586
    },
    {
      "epoch": 1.7399186089530152,
      "grad_norm": 0.0278693325817585,
      "learning_rate": 8.947841288251568e-06,
      "loss": 0.6845,
      "mean_token_accuracy": 0.788350437808927,
      "step": 587
    },
    {
      "epoch": 1.7428782833888272,
      "grad_norm": 0.0282698143273592,
      "learning_rate": 8.913478499145255e-06,
      "loss": 0.7068,
      "mean_token_accuracy": 0.7829055504378428,
      "step": 588
    },
    {
      "epoch": 1.7458379578246392,
      "grad_norm": 0.0311178807169199,
      "learning_rate": 8.879128688101749e-06,
      "loss": 0.7376,
      "mean_token_accuracy": 0.7729856949958135,
      "step": 589
    },
    {
      "epoch": 1.7487976322604513,
      "grad_norm": 0.026961086317896843,
      "learning_rate": 8.844792265415738e-06,
      "loss": 0.6769,
      "mean_token_accuracy": 0.7915280513720281,
      "step": 590
    },
    {
      "epoch": 1.7517573066962635,
      "grad_norm": 0.026299206539988518,
      "learning_rate": 8.810469641222001e-06,
      "loss": 0.6866,
      "mean_token_accuracy": 0.7877360667145736,
      "step": 591
    },
    {
      "epoch": 1.7547169811320755,
      "grad_norm": 0.027501925826072693,
      "learning_rate": 8.776161225490488e-06,
      "loss": 0.6845,
      "mean_token_accuracy": 0.7873916181335086,
      "step": 592
    },
    {
      "epoch": 1.7576766555678875,
      "grad_norm": 0.02949231117963791,
      "learning_rate": 8.741867428021447e-06,
      "loss": 0.7108,
      "mean_token_accuracy": 0.7801981470561626,
      "step": 593
    },
    {
      "epoch": 1.7606363300036996,
      "grad_norm": 0.025973103940486908,
      "learning_rate": 8.707588658440511e-06,
      "loss": 0.6748,
      "mean_token_accuracy": 0.7900679877899008,
      "step": 594
    },
    {
      "epoch": 1.7635960044395116,
      "grad_norm": 0.027610933408141136,
      "learning_rate": 8.673325326193806e-06,
      "loss": 0.6876,
      "mean_token_accuracy": 0.7891318628138966,
      "step": 595
    },
    {
      "epoch": 1.7665556788753238,
      "grad_norm": 0.03027682937681675,
      "learning_rate": 8.639077840543078e-06,
      "loss": 0.7013,
      "mean_token_accuracy": 0.7830476847803781,
      "step": 596
    },
    {
      "epoch": 1.7695153533111356,
      "grad_norm": 0.02684742584824562,
      "learning_rate": 8.604846610560771e-06,
      "loss": 0.6892,
      "mean_token_accuracy": 0.7863805338109771,
      "step": 597
    },
    {
      "epoch": 1.7724750277469479,
      "grad_norm": 0.0272175595164299,
      "learning_rate": 8.570632045125185e-06,
      "loss": 0.6799,
      "mean_token_accuracy": 0.7892807023909862,
      "step": 598
    },
    {
      "epoch": 1.77543470218276,
      "grad_norm": 0.02956371381878853,
      "learning_rate": 8.536434552915555e-06,
      "loss": 0.6828,
      "mean_token_accuracy": 0.7898981732005309,
      "step": 599
    },
    {
      "epoch": 1.778394376618572,
      "grad_norm": 0.030007705092430115,
      "learning_rate": 8.502254542407186e-06,
      "loss": 0.7057,
      "mean_token_accuracy": 0.7801339855377124,
      "step": 600
    },
    {
      "epoch": 1.778394376618572,
      "eval_loss": 0.7422189712524414,
      "eval_mean_token_accuracy": 0.7706628372210431,
      "eval_runtime": 25.948,
      "eval_samples_per_second": 4.971,
      "eval_steps_per_second": 1.272,
      "step": 600
    },
    {
      "epoch": 1.7813540510543842,
      "grad_norm": 0.03125801309943199,
      "learning_rate": 8.468092421866575e-06,
      "loss": 0.703,
      "mean_token_accuracy": 0.7840786141699138,
      "step": 601
    },
    {
      "epoch": 1.784313725490196,
      "grad_norm": 0.02820964902639389,
      "learning_rate": 8.433948599346516e-06,
      "loss": 0.6794,
      "mean_token_accuracy": 0.7900923655847418,
      "step": 602
    },
    {
      "epoch": 1.7872733999260082,
      "grad_norm": 0.025561511516571045,
      "learning_rate": 8.399823482681263e-06,
      "loss": 0.6725,
      "mean_token_accuracy": 0.7907421479577115,
      "step": 603
    },
    {
      "epoch": 1.7902330743618202,
      "grad_norm": 0.028765974566340446,
      "learning_rate": 8.36571747948162e-06,
      "loss": 0.6585,
      "mean_token_accuracy": 0.795219423137688,
      "step": 604
    },
    {
      "epoch": 1.7931927487976322,
      "grad_norm": 0.02895234152674675,
      "learning_rate": 8.331630997130091e-06,
      "loss": 0.646,
      "mean_token_accuracy": 0.7996773441669457,
      "step": 605
    },
    {
      "epoch": 1.7961524232334445,
      "grad_norm": 0.02965598739683628,
      "learning_rate": 8.297564442776014e-06,
      "loss": 0.7081,
      "mean_token_accuracy": 0.7796690311023957,
      "step": 606
    },
    {
      "epoch": 1.7991120976692563,
      "grad_norm": 0.02854854427278042,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.6727,
      "mean_token_accuracy": 0.7921263343069486,
      "step": 607
    },
    {
      "epoch": 1.8020717721050685,
      "grad_norm": 0.03143621236085892,
      "learning_rate": 8.229492745462551e-06,
      "loss": 0.6808,
      "mean_token_accuracy": 0.7895817190183476,
      "step": 608
    },
    {
      "epoch": 1.8050314465408805,
      "grad_norm": 0.028171831741929054,
      "learning_rate": 8.195488415592238e-06,
      "loss": 0.69,
      "mean_token_accuracy": 0.7871315718886494,
      "step": 609
    },
    {
      "epoch": 1.8079911209766926,
      "grad_norm": 0.03055482916533947,
      "learning_rate": 8.161505639887818e-06,
      "loss": 0.6942,
      "mean_token_accuracy": 0.7849820991609298,
      "step": 610
    },
    {
      "epoch": 1.8109507954125046,
      "grad_norm": 0.028566721826791763,
      "learning_rate": 8.12754482425989e-06,
      "loss": 0.6603,
      "mean_token_accuracy": 0.7962026834361321,
      "step": 611
    },
    {
      "epoch": 1.8139104698483166,
      "grad_norm": 0.027515290305018425,
      "learning_rate": 8.09360637435676e-06,
      "loss": 0.6834,
      "mean_token_accuracy": 0.7882978183658348,
      "step": 612
    },
    {
      "epoch": 1.8168701442841289,
      "grad_norm": 0.028337392956018448,
      "learning_rate": 8.05969069555957e-06,
      "loss": 0.7146,
      "mean_token_accuracy": 0.78017177586895,
      "step": 613
    },
    {
      "epoch": 1.8198298187199407,
      "grad_norm": 0.030293980613350868,
      "learning_rate": 8.025798192977482e-06,
      "loss": 0.6798,
      "mean_token_accuracy": 0.7885316206049379,
      "step": 614
    },
    {
      "epoch": 1.822789493155753,
      "grad_norm": 0.026661008596420288,
      "learning_rate": 7.991929271442817e-06,
      "loss": 0.7014,
      "mean_token_accuracy": 0.7823754400642968,
      "step": 615
    },
    {
      "epoch": 1.825749167591565,
      "grad_norm": 0.026802418753504753,
      "learning_rate": 7.958084335506239e-06,
      "loss": 0.6706,
      "mean_token_accuracy": 0.7922082107064662,
      "step": 616
    },
    {
      "epoch": 1.828708842027377,
      "grad_norm": 0.030152181163430214,
      "learning_rate": 7.924263789431913e-06,
      "loss": 0.7251,
      "mean_token_accuracy": 0.7754623144148004,
      "step": 617
    },
    {
      "epoch": 1.8316685164631892,
      "grad_norm": 0.02956436574459076,
      "learning_rate": 7.89046803719267e-06,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8009759216569763,
      "step": 618
    },
    {
      "epoch": 1.834628190899001,
      "grad_norm": 0.028118127956986427,
      "learning_rate": 7.856697482465195e-06,
      "loss": 0.7131,
      "mean_token_accuracy": 0.779968793402341,
      "step": 619
    },
    {
      "epoch": 1.8375878653348132,
      "grad_norm": 0.02660803310573101,
      "learning_rate": 7.822952528625192e-06,
      "loss": 0.6778,
      "mean_token_accuracy": 0.7895780695711778,
      "step": 620
    },
    {
      "epoch": 1.8405475397706252,
      "grad_norm": 0.026824964210391045,
      "learning_rate": 7.789233578742583e-06,
      "loss": 0.6938,
      "mean_token_accuracy": 0.7869364956239278,
      "step": 621
    },
    {
      "epoch": 1.8435072142064373,
      "grad_norm": 0.029271762818098068,
      "learning_rate": 7.755541035576677e-06,
      "loss": 0.7041,
      "mean_token_accuracy": 0.782565409998345,
      "step": 622
    },
    {
      "epoch": 1.8464668886422495,
      "grad_norm": 0.03059292770922184,
      "learning_rate": 7.721875301571359e-06,
      "loss": 0.6925,
      "mean_token_accuracy": 0.7881261135725389,
      "step": 623
    },
    {
      "epoch": 1.8494265630780613,
      "grad_norm": 0.0281398706138134,
      "learning_rate": 7.688236778850307e-06,
      "loss": 0.6771,
      "mean_token_accuracy": 0.7893142680513425,
      "step": 624
    },
    {
      "epoch": 1.8523862375138735,
      "grad_norm": 0.026483546942472458,
      "learning_rate": 7.654625869212147e-06,
      "loss": 0.6586,
      "mean_token_accuracy": 0.7959044381594015,
      "step": 625
    },
    {
      "epoch": 1.8553459119496856,
      "grad_norm": 0.02909860573709011,
      "learning_rate": 7.621042974125701e-06,
      "loss": 0.7121,
      "mean_token_accuracy": 0.7793595533292483,
      "step": 626
    },
    {
      "epoch": 1.8583055863854976,
      "grad_norm": 0.027886470779776573,
      "learning_rate": 7.587488494725157e-06,
      "loss": 0.7205,
      "mean_token_accuracy": 0.777051614643814,
      "step": 627
    },
    {
      "epoch": 1.8612652608213096,
      "grad_norm": 0.02818593755364418,
      "learning_rate": 7.553962831805291e-06,
      "loss": 0.7001,
      "mean_token_accuracy": 0.7829354410609054,
      "step": 628
    },
    {
      "epoch": 1.8642249352571216,
      "grad_norm": 0.028772005811333656,
      "learning_rate": 7.520466385816672e-06,
      "loss": 0.6921,
      "mean_token_accuracy": 0.7862072112820369,
      "step": 629
    },
    {
      "epoch": 1.8671846096929339,
      "grad_norm": 0.02801499515771866,
      "learning_rate": 7.48699955686089e-06,
      "loss": 0.7116,
      "mean_token_accuracy": 0.7797623820486961,
      "step": 630
    },
    {
      "epoch": 1.870144284128746,
      "grad_norm": 0.03020596317946911,
      "learning_rate": 7.453562744685779e-06,
      "loss": 0.6562,
      "mean_token_accuracy": 0.7967758148394831,
      "step": 631
    },
    {
      "epoch": 1.873103958564558,
      "grad_norm": 0.026859350502490997,
      "learning_rate": 7.420156348680621e-06,
      "loss": 0.7093,
      "mean_token_accuracy": 0.7821791973948278,
      "step": 632
    },
    {
      "epoch": 1.87606363300037,
      "grad_norm": 0.028368927538394928,
      "learning_rate": 7.3867807678713965e-06,
      "loss": 0.6772,
      "mean_token_accuracy": 0.7909775119897093,
      "step": 633
    },
    {
      "epoch": 1.879023307436182,
      "grad_norm": 0.027821959927678108,
      "learning_rate": 7.353436400916006e-06,
      "loss": 0.7309,
      "mean_token_accuracy": 0.774006989047497,
      "step": 634
    },
    {
      "epoch": 1.8819829818719942,
      "grad_norm": 0.028787191957235336,
      "learning_rate": 7.32012364609952e-06,
      "loss": 0.6964,
      "mean_token_accuracy": 0.7859444008537074,
      "step": 635
    },
    {
      "epoch": 1.884942656307806,
      "grad_norm": 0.02959715761244297,
      "learning_rate": 7.286842901329413e-06,
      "loss": 0.6809,
      "mean_token_accuracy": 0.7886240354703831,
      "step": 636
    },
    {
      "epoch": 1.8879023307436182,
      "grad_norm": 0.028586136177182198,
      "learning_rate": 7.253594564130804e-06,
      "loss": 0.638,
      "mean_token_accuracy": 0.8010729061146105,
      "step": 637
    },
    {
      "epoch": 1.8908620051794303,
      "grad_norm": 0.026945089921355247,
      "learning_rate": 7.22037903164173e-06,
      "loss": 0.6723,
      "mean_token_accuracy": 0.7924590728501775,
      "step": 638
    },
    {
      "epoch": 1.8938216796152423,
      "grad_norm": 0.028277359902858734,
      "learning_rate": 7.187196700608373e-06,
      "loss": 0.7127,
      "mean_token_accuracy": 0.7799887033647283,
      "step": 639
    },
    {
      "epoch": 1.8967813540510545,
      "grad_norm": 0.030940737575292587,
      "learning_rate": 7.154047967380353e-06,
      "loss": 0.6867,
      "mean_token_accuracy": 0.7873340124813961,
      "step": 640
    },
    {
      "epoch": 1.8997410284868663,
      "grad_norm": 0.026180449873209,
      "learning_rate": 7.120933227905971e-06,
      "loss": 0.6893,
      "mean_token_accuracy": 0.7868485530143366,
      "step": 641
    },
    {
      "epoch": 1.9027007029226786,
      "grad_norm": 0.02866005152463913,
      "learning_rate": 7.0878528777274814e-06,
      "loss": 0.7183,
      "mean_token_accuracy": 0.777992491800555,
      "step": 642
    },
    {
      "epoch": 1.9056603773584906,
      "grad_norm": 0.02938344143331051,
      "learning_rate": 7.05480731197638e-06,
      "loss": 0.6707,
      "mean_token_accuracy": 0.7938739114456241,
      "step": 643
    },
    {
      "epoch": 1.9086200517943026,
      "grad_norm": 0.028996018692851067,
      "learning_rate": 7.021796925368667e-06,
      "loss": 0.697,
      "mean_token_accuracy": 0.7842826106648617,
      "step": 644
    },
    {
      "epoch": 1.9115797262301149,
      "grad_norm": 0.028850911185145378,
      "learning_rate": 6.988822112200157e-06,
      "loss": 0.7066,
      "mean_token_accuracy": 0.7816915276239196,
      "step": 645
    },
    {
      "epoch": 1.9145394006659266,
      "grad_norm": 0.02863176353275776,
      "learning_rate": 6.955883266341741e-06,
      "loss": 0.6984,
      "mean_token_accuracy": 0.7840305323470096,
      "step": 646
    },
    {
      "epoch": 1.917499075101739,
      "grad_norm": 0.029000280424952507,
      "learning_rate": 6.9229807812346985e-06,
      "loss": 0.7016,
      "mean_token_accuracy": 0.783485416787975,
      "step": 647
    },
    {
      "epoch": 1.920458749537551,
      "grad_norm": 0.02904784306883812,
      "learning_rate": 6.890115049885995e-06,
      "loss": 0.6946,
      "mean_token_accuracy": 0.7850121695593568,
      "step": 648
    },
    {
      "epoch": 1.923418423973363,
      "grad_norm": 0.02796006202697754,
      "learning_rate": 6.85728646486359e-06,
      "loss": 0.6872,
      "mean_token_accuracy": 0.7858824792428694,
      "step": 649
    },
    {
      "epoch": 1.926378098409175,
      "grad_norm": 0.026994947344064713,
      "learning_rate": 6.824495418291741e-06,
      "loss": 0.6696,
      "mean_token_accuracy": 0.7913752429351146,
      "step": 650
    },
    {
      "epoch": 1.929337772844987,
      "grad_norm": 0.029123296961188316,
      "learning_rate": 6.791742301846325e-06,
      "loss": 0.7019,
      "mean_token_accuracy": 0.7835777427919317,
      "step": 651
    },
    {
      "epoch": 1.9322974472807992,
      "grad_norm": 0.02968716248869896,
      "learning_rate": 6.759027506750159e-06,
      "loss": 0.7044,
      "mean_token_accuracy": 0.7811305565589783,
      "step": 652
    },
    {
      "epoch": 1.935257121716611,
      "grad_norm": 0.028995733708143234,
      "learning_rate": 6.726351423768323e-06,
      "loss": 0.7125,
      "mean_token_accuracy": 0.781838803607727,
      "step": 653
    },
    {
      "epoch": 1.9382167961524233,
      "grad_norm": 0.0271761491894722,
      "learning_rate": 6.693714443203507e-06,
      "loss": 0.681,
      "mean_token_accuracy": 0.7889858193285975,
      "step": 654
    },
    {
      "epoch": 1.9411764705882353,
      "grad_norm": 0.027883367612957954,
      "learning_rate": 6.661116954891329e-06,
      "loss": 0.6965,
      "mean_token_accuracy": 0.7858269003851349,
      "step": 655
    },
    {
      "epoch": 1.9441361450240473,
      "grad_norm": 0.029907861724495888,
      "learning_rate": 6.62855934819569e-06,
      "loss": 0.7146,
      "mean_token_accuracy": 0.7782502288395488,
      "step": 656
    },
    {
      "epoch": 1.9470958194598595,
      "grad_norm": 0.027081694453954697,
      "learning_rate": 6.59604201200412e-06,
      "loss": 0.6844,
      "mean_token_accuracy": 0.7883412748666474,
      "step": 657
    },
    {
      "epoch": 1.9500554938956713,
      "grad_norm": 0.030427632853388786,
      "learning_rate": 6.563565334723134e-06,
      "loss": 0.7051,
      "mean_token_accuracy": 0.7835524095761552,
      "step": 658
    },
    {
      "epoch": 1.9530151683314836,
      "grad_norm": 0.027951078489422798,
      "learning_rate": 6.5311297042736046e-06,
      "loss": 0.7245,
      "mean_token_accuracy": 0.7774474913400713,
      "step": 659
    },
    {
      "epoch": 1.9559748427672956,
      "grad_norm": 0.027215953916311264,
      "learning_rate": 6.498735508086094e-06,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8015327084427364,
      "step": 660
    },
    {
      "epoch": 1.9589345172031076,
      "grad_norm": 0.028915971517562866,
      "learning_rate": 6.466383133096268e-06,
      "loss": 0.7031,
      "mean_token_accuracy": 0.7815104672655345,
      "step": 661
    },
    {
      "epoch": 1.9618941916389199,
      "grad_norm": 0.029375633224844933,
      "learning_rate": 6.4340729657402424e-06,
      "loss": 0.662,
      "mean_token_accuracy": 0.7939480151549306,
      "step": 662
    },
    {
      "epoch": 1.9648538660747317,
      "grad_norm": 0.029343878850340843,
      "learning_rate": 6.40180539194999e-06,
      "loss": 0.7084,
      "mean_token_accuracy": 0.7819531152756548,
      "step": 663
    },
    {
      "epoch": 1.967813540510544,
      "grad_norm": 0.028881965205073357,
      "learning_rate": 6.3695807971487175e-06,
      "loss": 0.6812,
      "mean_token_accuracy": 0.7891491178334081,
      "step": 664
    },
    {
      "epoch": 1.970773214946356,
      "grad_norm": 0.02754000574350357,
      "learning_rate": 6.337399566246257e-06,
      "loss": 0.6959,
      "mean_token_accuracy": 0.7853549181325891,
      "step": 665
    },
    {
      "epoch": 1.973732889382168,
      "grad_norm": 0.027574459090828896,
      "learning_rate": 6.305262083634488e-06,
      "loss": 0.7023,
      "mean_token_accuracy": 0.782962098058342,
      "step": 666
    },
    {
      "epoch": 1.97669256381798,
      "grad_norm": 0.03027881495654583,
      "learning_rate": 6.2731687331827214e-06,
      "loss": 0.6993,
      "mean_token_accuracy": 0.7838028336236199,
      "step": 667
    },
    {
      "epoch": 1.979652238253792,
      "grad_norm": 0.03279561921954155,
      "learning_rate": 6.2411198982331435e-06,
      "loss": 0.6652,
      "mean_token_accuracy": 0.7931794557485324,
      "step": 668
    },
    {
      "epoch": 1.9826119126896042,
      "grad_norm": 0.028421377763152122,
      "learning_rate": 6.209115961596208e-06,
      "loss": 0.7039,
      "mean_token_accuracy": 0.7828308947085652,
      "step": 669
    },
    {
      "epoch": 1.9855715871254163,
      "grad_norm": 0.027962755411863327,
      "learning_rate": 6.177157305546077e-06,
      "loss": 0.6986,
      "mean_token_accuracy": 0.7847478913327196,
      "step": 670
    },
    {
      "epoch": 1.9885312615612283,
      "grad_norm": 0.02687322534620762,
      "learning_rate": 6.145244311816063e-06,
      "loss": 0.6887,
      "mean_token_accuracy": 0.7861106503857088,
      "step": 671
    },
    {
      "epoch": 1.9914909359970403,
      "grad_norm": 0.02775583788752556,
      "learning_rate": 6.113377361594048e-06,
      "loss": 0.683,
      "mean_token_accuracy": 0.7885030224129054,
      "step": 672
    },
    {
      "epoch": 1.9944506104328523,
      "grad_norm": 0.02843259647488594,
      "learning_rate": 6.081556835517955e-06,
      "loss": 0.6895,
      "mean_token_accuracy": 0.7873784909684312,
      "step": 673
    },
    {
      "epoch": 1.9974102848686646,
      "grad_norm": 0.029535317793488503,
      "learning_rate": 6.049783113671184e-06,
      "loss": 0.6828,
      "mean_token_accuracy": 0.7881313178490122,
      "step": 674
    },
    {
      "epoch": 2.0029596744358122,
      "grad_norm": 0.045787323266267776,
      "learning_rate": 6.018056575578075e-06,
      "loss": 1.395,
      "mean_token_accuracy": 0.7882934187103929,
      "step": 675
    },
    {
      "epoch": 2.005919348871624,
      "grad_norm": 0.02875291369855404,
      "learning_rate": 5.986377600199371e-06,
      "loss": 0.6926,
      "mean_token_accuracy": 0.783694086228782,
      "step": 676
    },
    {
      "epoch": 2.0088790233074363,
      "grad_norm": 0.02869325503706932,
      "learning_rate": 5.9547465659277215e-06,
      "loss": 0.67,
      "mean_token_accuracy": 0.7912807693801105,
      "step": 677
    },
    {
      "epoch": 2.011838697743248,
      "grad_norm": 0.02715614065527916,
      "learning_rate": 5.923163850583114e-06,
      "loss": 0.6265,
      "mean_token_accuracy": 0.8029483091653056,
      "step": 678
    },
    {
      "epoch": 2.0147983721790603,
      "grad_norm": 0.026809340342879295,
      "learning_rate": 5.891629831408392e-06,
      "loss": 0.6769,
      "mean_token_accuracy": 0.7901606767031425,
      "step": 679
    },
    {
      "epoch": 2.0177580466148726,
      "grad_norm": 0.02626258134841919,
      "learning_rate": 5.8601448850647515e-06,
      "loss": 0.7242,
      "mean_token_accuracy": 0.7743892889344857,
      "step": 680
    },
    {
      "epoch": 2.0207177210506844,
      "grad_norm": 0.027618641033768654,
      "learning_rate": 5.828709387627219e-06,
      "loss": 0.6397,
      "mean_token_accuracy": 0.8006681555060222,
      "step": 681
    },
    {
      "epoch": 2.0236773954864966,
      "grad_norm": 0.028962858021259308,
      "learning_rate": 5.797323714580192e-06,
      "loss": 0.6502,
      "mean_token_accuracy": 0.7962358444644367,
      "step": 682
    },
    {
      "epoch": 2.0266370699223084,
      "grad_norm": 0.029240624979138374,
      "learning_rate": 5.7659882408129204e-06,
      "loss": 0.6625,
      "mean_token_accuracy": 0.7923286909305776,
      "step": 683
    },
    {
      "epoch": 2.0295967443581207,
      "grad_norm": 0.028116624802350998,
      "learning_rate": 5.7347033406150494e-06,
      "loss": 0.6833,
      "mean_token_accuracy": 0.7865030310525323,
      "step": 684
    },
    {
      "epoch": 2.0325564187939325,
      "grad_norm": 0.028533287346363068,
      "learning_rate": 5.703469387672138e-06,
      "loss": 0.6156,
      "mean_token_accuracy": 0.8083995635977411,
      "step": 685
    },
    {
      "epoch": 2.0355160932297447,
      "grad_norm": 0.026627851650118828,
      "learning_rate": 5.672286755061212e-06,
      "loss": 0.6475,
      "mean_token_accuracy": 0.7970526695062603,
      "step": 686
    },
    {
      "epoch": 2.038475767665557,
      "grad_norm": 0.02711411751806736,
      "learning_rate": 5.64115581524629e-06,
      "loss": 0.6562,
      "mean_token_accuracy": 0.7952441052730128,
      "step": 687
    },
    {
      "epoch": 2.0414354421013687,
      "grad_norm": 0.026333443820476532,
      "learning_rate": 5.610076940073939e-06,
      "loss": 0.6505,
      "mean_token_accuracy": 0.7972208890277822,
      "step": 688
    },
    {
      "epoch": 2.044395116537181,
      "grad_norm": 0.027702175080776215,
      "learning_rate": 5.579050500768837e-06,
      "loss": 0.6347,
      "mean_token_accuracy": 0.8013627765726284,
      "step": 689
    },
    {
      "epoch": 2.0473547909729928,
      "grad_norm": 0.027392296120524406,
      "learning_rate": 5.548076867929331e-06,
      "loss": 0.6601,
      "mean_token_accuracy": 0.7935603702263234,
      "step": 690
    },
    {
      "epoch": 2.050314465408805,
      "grad_norm": 0.026571672409772873,
      "learning_rate": 5.517156411523026e-06,
      "loss": 0.6308,
      "mean_token_accuracy": 0.8038993308545019,
      "step": 691
    },
    {
      "epoch": 2.0532741398446173,
      "grad_norm": 0.028205694630742073,
      "learning_rate": 5.486289500882355e-06,
      "loss": 0.6535,
      "mean_token_accuracy": 0.7952194817552366,
      "step": 692
    },
    {
      "epoch": 2.056233814280429,
      "grad_norm": 0.030303819105029106,
      "learning_rate": 5.455476504700161e-06,
      "loss": 0.645,
      "mean_token_accuracy": 0.7977919872763917,
      "step": 693
    },
    {
      "epoch": 2.0591934887162413,
      "grad_norm": 0.027428416535258293,
      "learning_rate": 5.424717791025302e-06,
      "loss": 0.6189,
      "mean_token_accuracy": 0.8051012175713522,
      "step": 694
    },
    {
      "epoch": 2.062153163152053,
      "grad_norm": 0.026880696415901184,
      "learning_rate": 5.3940137272582534e-06,
      "loss": 0.6725,
      "mean_token_accuracy": 0.7919366129740589,
      "step": 695
    },
    {
      "epoch": 2.0651128375878653,
      "grad_norm": 0.026649234816432,
      "learning_rate": 5.3633646801467255e-06,
      "loss": 0.637,
      "mean_token_accuracy": 0.8012966006828176,
      "step": 696
    },
    {
      "epoch": 2.0680725120236776,
      "grad_norm": 0.025737503543496132,
      "learning_rate": 5.332771015781275e-06,
      "loss": 0.6416,
      "mean_token_accuracy": 0.7998217304969483,
      "step": 697
    },
    {
      "epoch": 2.0710321864594894,
      "grad_norm": 0.028061147779226303,
      "learning_rate": 5.302233099590928e-06,
      "loss": 0.6304,
      "mean_token_accuracy": 0.8029078651422608,
      "step": 698
    },
    {
      "epoch": 2.0739918608953016,
      "grad_norm": 0.02635873854160309,
      "learning_rate": 5.271751296338823e-06,
      "loss": 0.6558,
      "mean_token_accuracy": 0.7949186439112171,
      "step": 699
    },
    {
      "epoch": 2.0769515353311134,
      "grad_norm": 0.027407187968492508,
      "learning_rate": 5.241325970117851e-06,
      "loss": 0.6401,
      "mean_token_accuracy": 0.8013597311952657,
      "step": 700
    },
    {
      "epoch": 2.0769515353311134,
      "eval_loss": 0.7419947981834412,
      "eval_mean_token_accuracy": 0.7710200209515181,
      "eval_runtime": 24.7166,
      "eval_samples_per_second": 5.219,
      "eval_steps_per_second": 1.335,
      "step": 700
    },
    {
      "epoch": 2.0799112097669257,
      "grad_norm": 0.02673179842531681,
      "learning_rate": 5.210957484346314e-06,
      "loss": 0.5888,
      "mean_token_accuracy": 0.8146958734098403,
      "step": 701
    },
    {
      "epoch": 2.0828708842027375,
      "grad_norm": 0.026034710928797722,
      "learning_rate": 5.1806462017635775e-06,
      "loss": 0.6544,
      "mean_token_accuracy": 0.7953925705337829,
      "step": 702
    },
    {
      "epoch": 2.0858305586385497,
      "grad_norm": 0.029159262776374817,
      "learning_rate": 5.150392484425728e-06,
      "loss": 0.6534,
      "mean_token_accuracy": 0.7972241240298339,
      "step": 703
    },
    {
      "epoch": 2.088790233074362,
      "grad_norm": 0.027097195386886597,
      "learning_rate": 5.120196693701267e-06,
      "loss": 0.6552,
      "mean_token_accuracy": 0.7952015942605395,
      "step": 704
    },
    {
      "epoch": 2.0917499075101738,
      "grad_norm": 0.026992561295628548,
      "learning_rate": 5.090059190266779e-06,
      "loss": 0.6807,
      "mean_token_accuracy": 0.7870320512938295,
      "step": 705
    },
    {
      "epoch": 2.094709581945986,
      "grad_norm": 0.026859944686293602,
      "learning_rate": 5.059980334102637e-06,
      "loss": 0.6503,
      "mean_token_accuracy": 0.7984751395459512,
      "step": 706
    },
    {
      "epoch": 2.097669256381798,
      "grad_norm": 0.027902353554964066,
      "learning_rate": 5.0299604844886985e-06,
      "loss": 0.6541,
      "mean_token_accuracy": 0.795808366406312,
      "step": 707
    },
    {
      "epoch": 2.10062893081761,
      "grad_norm": 0.0258940476924181,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.6665,
      "mean_token_accuracy": 0.791814400602288,
      "step": 708
    },
    {
      "epoch": 2.1035886052534223,
      "grad_norm": 0.025506624951958656,
      "learning_rate": 4.970099238502494e-06,
      "loss": 0.6619,
      "mean_token_accuracy": 0.7939234290402679,
      "step": 709
    },
    {
      "epoch": 2.106548279689234,
      "grad_norm": 0.026071209460496902,
      "learning_rate": 4.940258557148765e-06,
      "loss": 0.6813,
      "mean_token_accuracy": 0.7877323169707022,
      "step": 710
    },
    {
      "epoch": 2.1095079541250463,
      "grad_norm": 0.026423268020153046,
      "learning_rate": 4.910478312373757e-06,
      "loss": 0.6277,
      "mean_token_accuracy": 0.8043923389854782,
      "step": 711
    },
    {
      "epoch": 2.112467628560858,
      "grad_norm": 0.025508983060717583,
      "learning_rate": 4.8807588598905364e-06,
      "loss": 0.6585,
      "mean_token_accuracy": 0.7954374363543862,
      "step": 712
    },
    {
      "epoch": 2.1154273029966704,
      "grad_norm": 0.027017951011657715,
      "learning_rate": 4.8511005546860214e-06,
      "loss": 0.6596,
      "mean_token_accuracy": 0.7942605528684004,
      "step": 713
    },
    {
      "epoch": 2.1183869774324826,
      "grad_norm": 0.025915304198861122,
      "learning_rate": 4.821503751016746e-06,
      "loss": 0.6327,
      "mean_token_accuracy": 0.8014850499624679,
      "step": 714
    },
    {
      "epoch": 2.1213466518682944,
      "grad_norm": 0.02531328611075878,
      "learning_rate": 4.791968802404648e-06,
      "loss": 0.6489,
      "mean_token_accuracy": 0.7980775534029787,
      "step": 715
    },
    {
      "epoch": 2.1243063263041067,
      "grad_norm": 0.02628396824002266,
      "learning_rate": 4.762496061632814e-06,
      "loss": 0.6058,
      "mean_token_accuracy": 0.8083745324569607,
      "step": 716
    },
    {
      "epoch": 2.1272660007399185,
      "grad_norm": 0.02541985549032688,
      "learning_rate": 4.733085880741301e-06,
      "loss": 0.694,
      "mean_token_accuracy": 0.7830943305243893,
      "step": 717
    },
    {
      "epoch": 2.1302256751757307,
      "grad_norm": 0.026374945417046547,
      "learning_rate": 4.703738611022899e-06,
      "loss": 0.6665,
      "mean_token_accuracy": 0.7928101370241055,
      "step": 718
    },
    {
      "epoch": 2.133185349611543,
      "grad_norm": 0.02606821246445179,
      "learning_rate": 4.674454603018949e-06,
      "loss": 0.6248,
      "mean_token_accuracy": 0.8057719779843474,
      "step": 719
    },
    {
      "epoch": 2.1361450240473547,
      "grad_norm": 0.02639526128768921,
      "learning_rate": 4.645234206515171e-06,
      "loss": 0.6485,
      "mean_token_accuracy": 0.7988039981486942,
      "step": 720
    },
    {
      "epoch": 2.139104698483167,
      "grad_norm": 0.026219215244054794,
      "learning_rate": 4.616077770537453e-06,
      "loss": 0.6334,
      "mean_token_accuracy": 0.801784327189846,
      "step": 721
    },
    {
      "epoch": 2.1420643729189788,
      "grad_norm": 0.025508172810077667,
      "learning_rate": 4.586985643347716e-06,
      "loss": 0.6453,
      "mean_token_accuracy": 0.7974435768625385,
      "step": 722
    },
    {
      "epoch": 2.145024047354791,
      "grad_norm": 0.026285525411367416,
      "learning_rate": 4.557958172439726e-06,
      "loss": 0.662,
      "mean_token_accuracy": 0.7931471630073025,
      "step": 723
    },
    {
      "epoch": 2.1479837217906033,
      "grad_norm": 0.025700561702251434,
      "learning_rate": 4.5289957045349655e-06,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8027245724114495,
      "step": 724
    },
    {
      "epoch": 2.150943396226415,
      "grad_norm": 0.02564854547381401,
      "learning_rate": 4.500098585578475e-06,
      "loss": 0.6299,
      "mean_token_accuracy": 0.8018347141086171,
      "step": 725
    },
    {
      "epoch": 2.1539030706622273,
      "grad_norm": 0.027371376752853394,
      "learning_rate": 4.471267160734731e-06,
      "loss": 0.664,
      "mean_token_accuracy": 0.7915234935011575,
      "step": 726
    },
    {
      "epoch": 2.156862745098039,
      "grad_norm": 0.02621809020638466,
      "learning_rate": 4.4425017743835155e-06,
      "loss": 0.6473,
      "mean_token_accuracy": 0.797941866204852,
      "step": 727
    },
    {
      "epoch": 2.1598224195338513,
      "grad_norm": 0.02413308061659336,
      "learning_rate": 4.413802770115816e-06,
      "loss": 0.6631,
      "mean_token_accuracy": 0.7934499240969237,
      "step": 728
    },
    {
      "epoch": 2.162782093969663,
      "grad_norm": 0.025956308469176292,
      "learning_rate": 4.385170490729712e-06,
      "loss": 0.6521,
      "mean_token_accuracy": 0.7956886134135189,
      "step": 729
    },
    {
      "epoch": 2.1657417684054754,
      "grad_norm": 0.02783017046749592,
      "learning_rate": 4.356605278226274e-06,
      "loss": 0.6742,
      "mean_token_accuracy": 0.7895592407412447,
      "step": 730
    },
    {
      "epoch": 2.1687014428412876,
      "grad_norm": 0.0272431168705225,
      "learning_rate": 4.328107473805487e-06,
      "loss": 0.6487,
      "mean_token_accuracy": 0.7979920405939707,
      "step": 731
    },
    {
      "epoch": 2.1716611172770994,
      "grad_norm": 0.025318261235952377,
      "learning_rate": 4.299677417862174e-06,
      "loss": 0.6657,
      "mean_token_accuracy": 0.791217591423855,
      "step": 732
    },
    {
      "epoch": 2.1746207917129117,
      "grad_norm": 0.027191517874598503,
      "learning_rate": 4.2713154499819345e-06,
      "loss": 0.6744,
      "mean_token_accuracy": 0.789432093853135,
      "step": 733
    },
    {
      "epoch": 2.1775804661487235,
      "grad_norm": 0.026165476068854332,
      "learning_rate": 4.243021908937083e-06,
      "loss": 0.6238,
      "mean_token_accuracy": 0.8054286287972181,
      "step": 734
    },
    {
      "epoch": 2.1805401405845357,
      "grad_norm": 0.0251636765897274,
      "learning_rate": 4.214797132682597e-06,
      "loss": 0.6114,
      "mean_token_accuracy": 0.8081800352561124,
      "step": 735
    },
    {
      "epoch": 2.183499815020348,
      "grad_norm": 0.02719036489725113,
      "learning_rate": 4.186641458352088e-06,
      "loss": 0.6817,
      "mean_token_accuracy": 0.787719062593448,
      "step": 736
    },
    {
      "epoch": 2.1864594894561598,
      "grad_norm": 0.02855796180665493,
      "learning_rate": 4.158555222253772e-06,
      "loss": 0.6847,
      "mean_token_accuracy": 0.7874981314371368,
      "step": 737
    },
    {
      "epoch": 2.189419163891972,
      "grad_norm": 0.026408439502120018,
      "learning_rate": 4.130538759866457e-06,
      "loss": 0.6623,
      "mean_token_accuracy": 0.7937424245147677,
      "step": 738
    },
    {
      "epoch": 2.192378838327784,
      "grad_norm": 0.028109323233366013,
      "learning_rate": 4.102592405835536e-06,
      "loss": 0.6466,
      "mean_token_accuracy": 0.798414769024551,
      "step": 739
    },
    {
      "epoch": 2.195338512763596,
      "grad_norm": 0.026109393686056137,
      "learning_rate": 4.074716493968976e-06,
      "loss": 0.6291,
      "mean_token_accuracy": 0.8041202182781496,
      "step": 740
    },
    {
      "epoch": 2.1982981871994083,
      "grad_norm": 0.026727063581347466,
      "learning_rate": 4.046911357233343e-06,
      "loss": 0.5991,
      "mean_token_accuracy": 0.8107393870634693,
      "step": 741
    },
    {
      "epoch": 2.20125786163522,
      "grad_norm": 0.025831934064626694,
      "learning_rate": 4.019177327749822e-06,
      "loss": 0.6262,
      "mean_token_accuracy": 0.8041446268176082,
      "step": 742
    },
    {
      "epoch": 2.2042175360710323,
      "grad_norm": 0.026398099958896637,
      "learning_rate": 3.991514736790259e-06,
      "loss": 0.6674,
      "mean_token_accuracy": 0.7917718646909211,
      "step": 743
    },
    {
      "epoch": 2.207177210506844,
      "grad_norm": 0.02614120952785015,
      "learning_rate": 3.9639239147731865e-06,
      "loss": 0.6206,
      "mean_token_accuracy": 0.8058362294253705,
      "step": 744
    },
    {
      "epoch": 2.2101368849426564,
      "grad_norm": 0.027411239221692085,
      "learning_rate": 3.936405191259891e-06,
      "loss": 0.6558,
      "mean_token_accuracy": 0.7954637043353722,
      "step": 745
    },
    {
      "epoch": 2.213096559378468,
      "grad_norm": 0.026757799088954926,
      "learning_rate": 3.908958894950465e-06,
      "loss": 0.6712,
      "mean_token_accuracy": 0.7914555146421036,
      "step": 746
    },
    {
      "epoch": 2.2160562338142804,
      "grad_norm": 0.027935413643717766,
      "learning_rate": 3.881585353679891e-06,
      "loss": 0.6752,
      "mean_token_accuracy": 0.7890529538858062,
      "step": 747
    },
    {
      "epoch": 2.2190159082500927,
      "grad_norm": 0.028880253434181213,
      "learning_rate": 3.854284894414122e-06,
      "loss": 0.6391,
      "mean_token_accuracy": 0.7998401837148577,
      "step": 748
    },
    {
      "epoch": 2.2219755826859044,
      "grad_norm": 0.02599172107875347,
      "learning_rate": 3.827057843246181e-06,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8027289918227506,
      "step": 749
    },
    {
      "epoch": 2.2249352571217167,
      "grad_norm": 0.026221485808491707,
      "learning_rate": 3.799904525392251e-06,
      "loss": 0.6174,
      "mean_token_accuracy": 0.8059152410024348,
      "step": 750
    },
    {
      "epoch": 2.2278949315575285,
      "grad_norm": 0.025390662252902985,
      "learning_rate": 3.7728252651878018e-06,
      "loss": 0.6682,
      "mean_token_accuracy": 0.7900244605067172,
      "step": 751
    },
    {
      "epoch": 2.2308546059933407,
      "grad_norm": 0.0265947375446558,
      "learning_rate": 3.745820386083724e-06,
      "loss": 0.6778,
      "mean_token_accuracy": 0.7873313246962839,
      "step": 752
    },
    {
      "epoch": 2.233814280429153,
      "grad_norm": 0.027348464354872704,
      "learning_rate": 3.718890210642442e-06,
      "loss": 0.6708,
      "mean_token_accuracy": 0.7910270915644594,
      "step": 753
    },
    {
      "epoch": 2.2367739548649648,
      "grad_norm": 0.029645796865224838,
      "learning_rate": 3.6920350605340883e-06,
      "loss": 0.6367,
      "mean_token_accuracy": 0.8016885587934341,
      "step": 754
    },
    {
      "epoch": 2.239733629300777,
      "grad_norm": 0.02541356347501278,
      "learning_rate": 3.6652552565326382e-06,
      "loss": 0.6784,
      "mean_token_accuracy": 0.7879687111880507,
      "step": 755
    },
    {
      "epoch": 2.242693303736589,
      "grad_norm": 0.026525545865297318,
      "learning_rate": 3.638551118512089e-06,
      "loss": 0.6665,
      "mean_token_accuracy": 0.7921577833942067,
      "step": 756
    },
    {
      "epoch": 2.245652978172401,
      "grad_norm": 0.029526982456445694,
      "learning_rate": 3.611922965442648e-06,
      "loss": 0.7059,
      "mean_token_accuracy": 0.7816915517325618,
      "step": 757
    },
    {
      "epoch": 2.2486126526082133,
      "grad_norm": 0.02809961326420307,
      "learning_rate": 3.5853711153868962e-06,
      "loss": 0.6544,
      "mean_token_accuracy": 0.7950629710473827,
      "step": 758
    },
    {
      "epoch": 2.251572327044025,
      "grad_norm": 0.028096938505768776,
      "learning_rate": 3.558895885496023e-06,
      "loss": 0.6653,
      "mean_token_accuracy": 0.7915877654144932,
      "step": 759
    },
    {
      "epoch": 2.2545320014798373,
      "grad_norm": 0.026522498577833176,
      "learning_rate": 3.53249759200601e-06,
      "loss": 0.6319,
      "mean_token_accuracy": 0.8026526280769568,
      "step": 760
    },
    {
      "epoch": 2.257491675915649,
      "grad_norm": 0.026230841875076294,
      "learning_rate": 3.506176550233863e-06,
      "loss": 0.6675,
      "mean_token_accuracy": 0.7919771831874818,
      "step": 761
    },
    {
      "epoch": 2.2604513503514614,
      "grad_norm": 0.02647150494158268,
      "learning_rate": 3.479933074573858e-06,
      "loss": 0.696,
      "mean_token_accuracy": 0.7853976150666319,
      "step": 762
    },
    {
      "epoch": 2.263411024787273,
      "grad_norm": 0.02635926753282547,
      "learning_rate": 3.453767478493761e-06,
      "loss": 0.6546,
      "mean_token_accuracy": 0.7957674214463809,
      "step": 763
    },
    {
      "epoch": 2.2663706992230854,
      "grad_norm": 0.026394182816147804,
      "learning_rate": 3.4276800745311135e-06,
      "loss": 0.6291,
      "mean_token_accuracy": 0.8024884508139499,
      "step": 764
    },
    {
      "epoch": 2.2693303736588977,
      "grad_norm": 0.024270912632346153,
      "learning_rate": 3.401671174289469e-06,
      "loss": 0.6617,
      "mean_token_accuracy": 0.7935239959477985,
      "step": 765
    },
    {
      "epoch": 2.2722900480947095,
      "grad_norm": 0.025384657084941864,
      "learning_rate": 3.37574108843469e-06,
      "loss": 0.6884,
      "mean_token_accuracy": 0.7868102749420524,
      "step": 766
    },
    {
      "epoch": 2.2752497225305217,
      "grad_norm": 0.02624906599521637,
      "learning_rate": 3.3498901266912397e-06,
      "loss": 0.6177,
      "mean_token_accuracy": 0.8059171051001157,
      "step": 767
    },
    {
      "epoch": 2.2782093969663335,
      "grad_norm": 0.026656541973352432,
      "learning_rate": 3.3241185978384636e-06,
      "loss": 0.6797,
      "mean_token_accuracy": 0.7886742381142204,
      "step": 768
    },
    {
      "epoch": 2.2811690714021458,
      "grad_norm": 0.0267485361546278,
      "learning_rate": 3.2984268097069284e-06,
      "loss": 0.6439,
      "mean_token_accuracy": 0.7990039021278321,
      "step": 769
    },
    {
      "epoch": 2.284128745837958,
      "grad_norm": 0.02498735673725605,
      "learning_rate": 3.2728150691747117e-06,
      "loss": 0.6513,
      "mean_token_accuracy": 0.7959641698923955,
      "step": 770
    },
    {
      "epoch": 2.28708842027377,
      "grad_norm": 0.0256348829716444,
      "learning_rate": 3.2472836821637744e-06,
      "loss": 0.6379,
      "mean_token_accuracy": 0.7999466653578364,
      "step": 771
    },
    {
      "epoch": 2.290048094709582,
      "grad_norm": 0.025716498494148254,
      "learning_rate": 3.22183295363627e-06,
      "loss": 0.6366,
      "mean_token_accuracy": 0.7998937718325038,
      "step": 772
    },
    {
      "epoch": 2.293007769145394,
      "grad_norm": 0.02601573057472706,
      "learning_rate": 3.196463187590929e-06,
      "loss": 0.6591,
      "mean_token_accuracy": 0.7943536743268278,
      "step": 773
    },
    {
      "epoch": 2.295967443581206,
      "grad_norm": 0.025225449353456497,
      "learning_rate": 3.1711746870594083e-06,
      "loss": 0.6389,
      "mean_token_accuracy": 0.8002055204444862,
      "step": 774
    },
    {
      "epoch": 2.2989271180170183,
      "grad_norm": 0.026426415890455246,
      "learning_rate": 3.145967754102691e-06,
      "loss": 0.6475,
      "mean_token_accuracy": 0.7981097278800483,
      "step": 775
    },
    {
      "epoch": 2.30188679245283,
      "grad_norm": 0.02602315880358219,
      "learning_rate": 3.1208426898074685e-06,
      "loss": 0.6807,
      "mean_token_accuracy": 0.7877973274370385,
      "step": 776
    },
    {
      "epoch": 2.3048464668886424,
      "grad_norm": 0.025725608691573143,
      "learning_rate": 3.0957997942825337e-06,
      "loss": 0.6469,
      "mean_token_accuracy": 0.7970667319183613,
      "step": 777
    },
    {
      "epoch": 2.307806141324454,
      "grad_norm": 0.026288101449608803,
      "learning_rate": 3.070839366655215e-06,
      "loss": 0.6568,
      "mean_token_accuracy": 0.793979566233237,
      "step": 778
    },
    {
      "epoch": 2.3107658157602664,
      "grad_norm": 0.02703455463051796,
      "learning_rate": 3.045961705067787e-06,
      "loss": 0.6695,
      "mean_token_accuracy": 0.7928558495837194,
      "step": 779
    },
    {
      "epoch": 2.313725490196078,
      "grad_norm": 0.025975678116083145,
      "learning_rate": 3.021167106673928e-06,
      "loss": 0.6525,
      "mean_token_accuracy": 0.7960860984487496,
      "step": 780
    },
    {
      "epoch": 2.3166851646318904,
      "grad_norm": 0.025307759642601013,
      "learning_rate": 2.996455867635155e-06,
      "loss": 0.6714,
      "mean_token_accuracy": 0.7890198580360704,
      "step": 781
    },
    {
      "epoch": 2.3196448390677027,
      "grad_norm": 0.02632983587682247,
      "learning_rate": 2.9718282831172885e-06,
      "loss": 0.6599,
      "mean_token_accuracy": 0.7941125642567792,
      "step": 782
    },
    {
      "epoch": 2.3226045135035145,
      "grad_norm": 0.0252531711012125,
      "learning_rate": 2.94728464728693e-06,
      "loss": 0.6109,
      "mean_token_accuracy": 0.8088270176688649,
      "step": 783
    },
    {
      "epoch": 2.3255641879393267,
      "grad_norm": 0.024317273870110512,
      "learning_rate": 2.922825253307947e-06,
      "loss": 0.6677,
      "mean_token_accuracy": 0.7906171672087918,
      "step": 784
    },
    {
      "epoch": 2.328523862375139,
      "grad_norm": 0.025564562529325485,
      "learning_rate": 2.898450393337977e-06,
      "loss": 0.6851,
      "mean_token_accuracy": 0.7865240457048697,
      "step": 785
    },
    {
      "epoch": 2.3314835368109508,
      "grad_norm": 0.026198633015155792,
      "learning_rate": 2.8741603585249312e-06,
      "loss": 0.6508,
      "mean_token_accuracy": 0.7969678323155496,
      "step": 786
    },
    {
      "epoch": 2.334443211246763,
      "grad_norm": 0.025766175240278244,
      "learning_rate": 2.8499554390035144e-06,
      "loss": 0.6579,
      "mean_token_accuracy": 0.7939616857426376,
      "step": 787
    },
    {
      "epoch": 2.337402885682575,
      "grad_norm": 0.024921085685491562,
      "learning_rate": 2.8258359238917665e-06,
      "loss": 0.6344,
      "mean_token_accuracy": 0.8027122551770116,
      "step": 788
    },
    {
      "epoch": 2.340362560118387,
      "grad_norm": 0.025290053337812424,
      "learning_rate": 2.8018021012875994e-06,
      "loss": 0.6116,
      "mean_token_accuracy": 0.8097985750753174,
      "step": 789
    },
    {
      "epoch": 2.343322234554199,
      "grad_norm": 0.025850266218185425,
      "learning_rate": 2.7778542582653746e-06,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8076698787987452,
      "step": 790
    },
    {
      "epoch": 2.346281908990011,
      "grad_norm": 0.024583814665675163,
      "learning_rate": 2.753992680872457e-06,
      "loss": 0.6157,
      "mean_token_accuracy": 0.8082738707215518,
      "step": 791
    },
    {
      "epoch": 2.3492415834258233,
      "grad_norm": 0.024625372141599655,
      "learning_rate": 2.7302176541257984e-06,
      "loss": 0.6433,
      "mean_token_accuracy": 0.7987133448759101,
      "step": 792
    },
    {
      "epoch": 2.352201257861635,
      "grad_norm": 0.02514871209859848,
      "learning_rate": 2.7065294620085425e-06,
      "loss": 0.6571,
      "mean_token_accuracy": 0.7949285878751765,
      "step": 793
    },
    {
      "epoch": 2.3551609322974474,
      "grad_norm": 0.02804665081202984,
      "learning_rate": 2.6829283874666236e-06,
      "loss": 0.6545,
      "mean_token_accuracy": 0.7965882492826242,
      "step": 794
    },
    {
      "epoch": 2.358120606733259,
      "grad_norm": 0.02476775459945202,
      "learning_rate": 2.6594147124053983e-06,
      "loss": 0.6613,
      "mean_token_accuracy": 0.7936106971507497,
      "step": 795
    },
    {
      "epoch": 2.3610802811690714,
      "grad_norm": 0.02643917314708233,
      "learning_rate": 2.635988717686272e-06,
      "loss": 0.6343,
      "mean_token_accuracy": 0.8011573896262686,
      "step": 796
    },
    {
      "epoch": 2.3640399556048832,
      "grad_norm": 0.025359464809298515,
      "learning_rate": 2.6126506831233343e-06,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8014174196404714,
      "step": 797
    },
    {
      "epoch": 2.3669996300406955,
      "grad_norm": 0.02469559945166111,
      "learning_rate": 2.5894008874800323e-06,
      "loss": 0.6159,
      "mean_token_accuracy": 0.8075957509556305,
      "step": 798
    },
    {
      "epoch": 2.3699593044765077,
      "grad_norm": 0.024650422856211662,
      "learning_rate": 2.5662396084658383e-06,
      "loss": 0.6997,
      "mean_token_accuracy": 0.7822622510937396,
      "step": 799
    },
    {
      "epoch": 2.3729189789123195,
      "grad_norm": 0.026352785527706146,
      "learning_rate": 2.543167122732918e-06,
      "loss": 0.6406,
      "mean_token_accuracy": 0.7996489018271652,
      "step": 800
    },
    {
      "epoch": 2.3729189789123195,
      "eval_loss": 0.7398498058319092,
      "eval_mean_token_accuracy": 0.771194051709573,
      "eval_runtime": 24.6871,
      "eval_samples_per_second": 5.225,
      "eval_steps_per_second": 1.337,
      "step": 800
    },
    {
      "epoch": 2.3758786533481318,
      "grad_norm": 0.026504702866077423,
      "learning_rate": 2.5201837058728506e-06,
      "loss": 0.6522,
      "mean_token_accuracy": 0.7958938862408909,
      "step": 801
    },
    {
      "epoch": 2.378838327783944,
      "grad_norm": 0.02537173219025135,
      "learning_rate": 2.4972896324133143e-06,
      "loss": 0.6505,
      "mean_token_accuracy": 0.7978752996956417,
      "step": 802
    },
    {
      "epoch": 2.381798002219756,
      "grad_norm": 0.02501768432557583,
      "learning_rate": 2.474485175814816e-06,
      "loss": 0.6523,
      "mean_token_accuracy": 0.7982308430662025,
      "step": 803
    },
    {
      "epoch": 2.384757676655568,
      "grad_norm": 0.02492235228419304,
      "learning_rate": 2.451770608467432e-06,
      "loss": 0.6654,
      "mean_token_accuracy": 0.7911750861310474,
      "step": 804
    },
    {
      "epoch": 2.38771735109138,
      "grad_norm": 0.025797097012400627,
      "learning_rate": 2.429146201687538e-06,
      "loss": 0.665,
      "mean_token_accuracy": 0.7931877729009794,
      "step": 805
    },
    {
      "epoch": 2.390677025527192,
      "grad_norm": 0.026369279250502586,
      "learning_rate": 2.4066122257145898e-06,
      "loss": 0.6291,
      "mean_token_accuracy": 0.8019230086243498,
      "step": 806
    },
    {
      "epoch": 2.393636699963004,
      "grad_norm": 0.02507649175822735,
      "learning_rate": 2.3841689497078746e-06,
      "loss": 0.6715,
      "mean_token_accuracy": 0.7905159410520408,
      "step": 807
    },
    {
      "epoch": 2.396596374398816,
      "grad_norm": 0.025938082486391068,
      "learning_rate": 2.361816641743303e-06,
      "loss": 0.6697,
      "mean_token_accuracy": 0.7906795986977561,
      "step": 808
    },
    {
      "epoch": 2.3995560488346284,
      "grad_norm": 0.02559804357588291,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.6755,
      "mean_token_accuracy": 0.7901660477410315,
      "step": 809
    },
    {
      "epoch": 2.40251572327044,
      "grad_norm": 0.026432987302541733,
      "learning_rate": 2.317385996808195e-06,
      "loss": 0.65,
      "mean_token_accuracy": 0.7955027076635426,
      "step": 810
    },
    {
      "epoch": 2.4054753977062524,
      "grad_norm": 0.025431308895349503,
      "learning_rate": 2.295308190543859e-06,
      "loss": 0.666,
      "mean_token_accuracy": 0.7908521754062707,
      "step": 811
    },
    {
      "epoch": 2.408435072142064,
      "grad_norm": 0.026080302894115448,
      "learning_rate": 2.2733224137277366e-06,
      "loss": 0.6724,
      "mean_token_accuracy": 0.7897629836847755,
      "step": 812
    },
    {
      "epoch": 2.4113947465778764,
      "grad_norm": 0.026199713349342346,
      "learning_rate": 2.251428928971102e-06,
      "loss": 0.6517,
      "mean_token_accuracy": 0.7965027056025703,
      "step": 813
    },
    {
      "epoch": 2.4143544210136882,
      "grad_norm": 0.024652190506458282,
      "learning_rate": 2.229627997782834e-06,
      "loss": 0.6629,
      "mean_token_accuracy": 0.7944626450235899,
      "step": 814
    },
    {
      "epoch": 2.4173140954495005,
      "grad_norm": 0.02574991062283516,
      "learning_rate": 2.2079198805662917e-06,
      "loss": 0.6461,
      "mean_token_accuracy": 0.7981591035907828,
      "step": 815
    },
    {
      "epoch": 2.4202737698853127,
      "grad_norm": 0.025130966678261757,
      "learning_rate": 2.186304836616221e-06,
      "loss": 0.6544,
      "mean_token_accuracy": 0.7952903168075519,
      "step": 816
    },
    {
      "epoch": 2.4232334443211245,
      "grad_norm": 0.025584751740098,
      "learning_rate": 2.1647831241156304e-06,
      "loss": 0.66,
      "mean_token_accuracy": 0.7932808343983979,
      "step": 817
    },
    {
      "epoch": 2.4261931187569368,
      "grad_norm": 0.02560119703412056,
      "learning_rate": 2.1433550001327376e-06,
      "loss": 0.6742,
      "mean_token_accuracy": 0.7901575113722394,
      "step": 818
    },
    {
      "epoch": 2.429152793192749,
      "grad_norm": 0.025043940171599388,
      "learning_rate": 2.122020720617869e-06,
      "loss": 0.6708,
      "mean_token_accuracy": 0.7900073733942257,
      "step": 819
    },
    {
      "epoch": 2.432112467628561,
      "grad_norm": 0.025217004120349884,
      "learning_rate": 2.1007805404004247e-06,
      "loss": 0.6221,
      "mean_token_accuracy": 0.8049443290030673,
      "step": 820
    },
    {
      "epoch": 2.435072142064373,
      "grad_norm": 0.025651799514889717,
      "learning_rate": 2.0796347131858187e-06,
      "loss": 0.6015,
      "mean_token_accuracy": 0.8114904410331883,
      "step": 821
    },
    {
      "epoch": 2.438031816500185,
      "grad_norm": 0.0252471175044775,
      "learning_rate": 2.058583491552465e-06,
      "loss": 0.6547,
      "mean_token_accuracy": 0.7951690334151519,
      "step": 822
    },
    {
      "epoch": 2.440991490935997,
      "grad_norm": 0.025844132527709007,
      "learning_rate": 2.037627126948751e-06,
      "loss": 0.6586,
      "mean_token_accuracy": 0.7938731111672259,
      "step": 823
    },
    {
      "epoch": 2.443951165371809,
      "grad_norm": 0.025531049817800522,
      "learning_rate": 2.0167658696900317e-06,
      "loss": 0.66,
      "mean_token_accuracy": 0.7952939311729473,
      "step": 824
    },
    {
      "epoch": 2.446910839807621,
      "grad_norm": 0.025648940354585648,
      "learning_rate": 1.9959999689556407e-06,
      "loss": 0.6613,
      "mean_token_accuracy": 0.794396505438024,
      "step": 825
    },
    {
      "epoch": 2.4498705142434334,
      "grad_norm": 0.024981150403618813,
      "learning_rate": 1.9753296727859195e-06,
      "loss": 0.6527,
      "mean_token_accuracy": 0.795100480026308,
      "step": 826
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.024588439613580704,
      "learning_rate": 1.9547552280792528e-06,
      "loss": 0.6749,
      "mean_token_accuracy": 0.7899342245774533,
      "step": 827
    },
    {
      "epoch": 2.4557898631150574,
      "grad_norm": 0.025872139260172844,
      "learning_rate": 1.9342768805891176e-06,
      "loss": 0.6283,
      "mean_token_accuracy": 0.8026095194197804,
      "step": 828
    },
    {
      "epoch": 2.4587495375508692,
      "grad_norm": 0.025196785107254982,
      "learning_rate": 1.9138948749211473e-06,
      "loss": 0.699,
      "mean_token_accuracy": 0.7819362157832077,
      "step": 829
    },
    {
      "epoch": 2.4617092119866815,
      "grad_norm": 0.025980552658438683,
      "learning_rate": 1.8936094545302098e-06,
      "loss": 0.6428,
      "mean_token_accuracy": 0.7996687751486988,
      "step": 830
    },
    {
      "epoch": 2.4646688864224937,
      "grad_norm": 0.025457216426730156,
      "learning_rate": 1.8734208617174986e-06,
      "loss": 0.6335,
      "mean_token_accuracy": 0.8007493445345365,
      "step": 831
    },
    {
      "epoch": 2.4676285608583055,
      "grad_norm": 0.02512306161224842,
      "learning_rate": 1.8533293376276473e-06,
      "loss": 0.6749,
      "mean_token_accuracy": 0.7885208997148254,
      "step": 832
    },
    {
      "epoch": 2.4705882352941178,
      "grad_norm": 0.024954231455922127,
      "learning_rate": 1.8333351222458407e-06,
      "loss": 0.6552,
      "mean_token_accuracy": 0.7952360154085174,
      "step": 833
    },
    {
      "epoch": 2.4735479097299296,
      "grad_norm": 0.024739854037761688,
      "learning_rate": 1.813438454394948e-06,
      "loss": 0.6765,
      "mean_token_accuracy": 0.7873318838490457,
      "step": 834
    },
    {
      "epoch": 2.476507584165742,
      "grad_norm": 0.026090454310178757,
      "learning_rate": 1.7936395717326705e-06,
      "loss": 0.6511,
      "mean_token_accuracy": 0.796652434689478,
      "step": 835
    },
    {
      "epoch": 2.479467258601554,
      "grad_norm": 0.025344999507069588,
      "learning_rate": 1.773938710748706e-06,
      "loss": 0.672,
      "mean_token_accuracy": 0.7899628771252464,
      "step": 836
    },
    {
      "epoch": 2.482426933037366,
      "grad_norm": 0.02536536566913128,
      "learning_rate": 1.7543361067619269e-06,
      "loss": 0.6249,
      "mean_token_accuracy": 0.803964330032249,
      "step": 837
    },
    {
      "epoch": 2.485386607473178,
      "grad_norm": 0.025098180398344994,
      "learning_rate": 1.734831993917564e-06,
      "loss": 0.6429,
      "mean_token_accuracy": 0.8000461673864955,
      "step": 838
    },
    {
      "epoch": 2.48834628190899,
      "grad_norm": 0.024952976033091545,
      "learning_rate": 1.715426605184407e-06,
      "loss": 0.625,
      "mean_token_accuracy": 0.8038834820447933,
      "step": 839
    },
    {
      "epoch": 2.491305956344802,
      "grad_norm": 0.025531768798828125,
      "learning_rate": 1.6961201723520248e-06,
      "loss": 0.6251,
      "mean_token_accuracy": 0.803729285397345,
      "step": 840
    },
    {
      "epoch": 2.494265630780614,
      "grad_norm": 0.025577286258339882,
      "learning_rate": 1.676912926028007e-06,
      "loss": 0.6149,
      "mean_token_accuracy": 0.8069646247057471,
      "step": 841
    },
    {
      "epoch": 2.497225305216426,
      "grad_norm": 0.024793023243546486,
      "learning_rate": 1.6578050956351887e-06,
      "loss": 0.6349,
      "mean_token_accuracy": 0.8026428810792391,
      "step": 842
    },
    {
      "epoch": 2.5001849796522384,
      "grad_norm": 0.029118113219738007,
      "learning_rate": 1.6387969094089318e-06,
      "loss": 0.6273,
      "mean_token_accuracy": 0.8042415209642944,
      "step": 843
    },
    {
      "epoch": 2.50314465408805,
      "grad_norm": 0.025995168834924698,
      "learning_rate": 1.619888594394382e-06,
      "loss": 0.657,
      "mean_token_accuracy": 0.7943599704287323,
      "step": 844
    },
    {
      "epoch": 2.5061043285238624,
      "grad_norm": 0.025791039690375328,
      "learning_rate": 1.6010803764437633e-06,
      "loss": 0.6386,
      "mean_token_accuracy": 0.8004295628882935,
      "step": 845
    },
    {
      "epoch": 2.5090640029596747,
      "grad_norm": 0.02399960719048977,
      "learning_rate": 1.5823724802136863e-06,
      "loss": 0.6757,
      "mean_token_accuracy": 0.789956235597643,
      "step": 846
    },
    {
      "epoch": 2.5120236773954865,
      "grad_norm": 0.024559002369642258,
      "learning_rate": 1.5637651291624522e-06,
      "loss": 0.6645,
      "mean_token_accuracy": 0.7933182314138616,
      "step": 847
    },
    {
      "epoch": 2.5149833518312983,
      "grad_norm": 0.0261478703469038,
      "learning_rate": 1.545258545547398e-06,
      "loss": 0.6907,
      "mean_token_accuracy": 0.7836512836592634,
      "step": 848
    },
    {
      "epoch": 2.5179430262671105,
      "grad_norm": 0.025508267804980278,
      "learning_rate": 1.5268529504222262e-06,
      "loss": 0.6577,
      "mean_token_accuracy": 0.7934076806152242,
      "step": 849
    },
    {
      "epoch": 2.5209027007029228,
      "grad_norm": 0.025865565985441208,
      "learning_rate": 1.5085485636343755e-06,
      "loss": 0.6407,
      "mean_token_accuracy": 0.7992867330334431,
      "step": 850
    },
    {
      "epoch": 2.5238623751387346,
      "grad_norm": 0.02500130422413349,
      "learning_rate": 1.4903456038223941e-06,
      "loss": 0.6475,
      "mean_token_accuracy": 0.7969763878400493,
      "step": 851
    },
    {
      "epoch": 2.526822049574547,
      "grad_norm": 0.025099368765950203,
      "learning_rate": 1.4722442884133214e-06,
      "loss": 0.5893,
      "mean_token_accuracy": 0.8151317747406919,
      "step": 852
    },
    {
      "epoch": 2.529781724010359,
      "grad_norm": 0.024691209197044373,
      "learning_rate": 1.4542448336201021e-06,
      "loss": 0.6562,
      "mean_token_accuracy": 0.7938335996227099,
      "step": 853
    },
    {
      "epoch": 2.532741398446171,
      "grad_norm": 0.025977561250329018,
      "learning_rate": 1.4363474544389876e-06,
      "loss": 0.6381,
      "mean_token_accuracy": 0.8001415100681745,
      "step": 854
    },
    {
      "epoch": 2.535701072881983,
      "grad_norm": 0.024935878813266754,
      "learning_rate": 1.4185523646469822e-06,
      "loss": 0.6559,
      "mean_token_accuracy": 0.7947584592152319,
      "step": 855
    },
    {
      "epoch": 2.538660747317795,
      "grad_norm": 0.026971381157636642,
      "learning_rate": 1.4008597767992872e-06,
      "loss": 0.6215,
      "mean_token_accuracy": 0.8067008412803375,
      "step": 856
    },
    {
      "epoch": 2.541620421753607,
      "grad_norm": 0.02452334761619568,
      "learning_rate": 1.3832699022267516e-06,
      "loss": 0.6498,
      "mean_token_accuracy": 0.7959469951559066,
      "step": 857
    },
    {
      "epoch": 2.544580096189419,
      "grad_norm": 0.02592734433710575,
      "learning_rate": 1.3657829510333653e-06,
      "loss": 0.673,
      "mean_token_accuracy": 0.7911240420756235,
      "step": 858
    },
    {
      "epoch": 2.547539770625231,
      "grad_norm": 0.025702424347400665,
      "learning_rate": 1.3483991320937307e-06,
      "loss": 0.6471,
      "mean_token_accuracy": 0.7999465837736335,
      "step": 859
    },
    {
      "epoch": 2.5504994450610434,
      "grad_norm": 0.025870032608509064,
      "learning_rate": 1.3311186530505838e-06,
      "loss": 0.6283,
      "mean_token_accuracy": 0.8037564919551831,
      "step": 860
    },
    {
      "epoch": 2.5534591194968552,
      "grad_norm": 0.02403545379638672,
      "learning_rate": 1.313941720312303e-06,
      "loss": 0.6357,
      "mean_token_accuracy": 0.8001821853160078,
      "step": 861
    },
    {
      "epoch": 2.5564187939326675,
      "grad_norm": 0.025059884414076805,
      "learning_rate": 1.2968685390504465e-06,
      "loss": 0.6325,
      "mean_token_accuracy": 0.8013825494175792,
      "step": 862
    },
    {
      "epoch": 2.5593784683684797,
      "grad_norm": 0.02489965409040451,
      "learning_rate": 1.2798993131973093e-06,
      "loss": 0.6842,
      "mean_token_accuracy": 0.7870550003070884,
      "step": 863
    },
    {
      "epoch": 2.5623381428042915,
      "grad_norm": 0.025446029379963875,
      "learning_rate": 1.263034245443473e-06,
      "loss": 0.6662,
      "mean_token_accuracy": 0.7925474939720628,
      "step": 864
    },
    {
      "epoch": 2.5652978172401038,
      "grad_norm": 0.02568916790187359,
      "learning_rate": 1.2462735372353996e-06,
      "loss": 0.6323,
      "mean_token_accuracy": 0.8022879124114145,
      "step": 865
    },
    {
      "epoch": 2.5682574916759155,
      "grad_norm": 0.02477600984275341,
      "learning_rate": 1.2296173887730122e-06,
      "loss": 0.6451,
      "mean_token_accuracy": 0.7997669385622551,
      "step": 866
    },
    {
      "epoch": 2.571217166111728,
      "grad_norm": 0.02490902878344059,
      "learning_rate": 1.2130659990073146e-06,
      "loss": 0.6456,
      "mean_token_accuracy": 0.7973325867475748,
      "step": 867
    },
    {
      "epoch": 2.5741768405475396,
      "grad_norm": 0.024430235847830772,
      "learning_rate": 1.196619565638003e-06,
      "loss": 0.6382,
      "mean_token_accuracy": 0.7996565693111616,
      "step": 868
    },
    {
      "epoch": 2.577136514983352,
      "grad_norm": 0.025014687329530716,
      "learning_rate": 1.1802782851111206e-06,
      "loss": 0.6518,
      "mean_token_accuracy": 0.7963992914489949,
      "step": 869
    },
    {
      "epoch": 2.580096189419164,
      "grad_norm": 0.028634632006287575,
      "learning_rate": 1.1640423526166987e-06,
      "loss": 0.6549,
      "mean_token_accuracy": 0.7952881516518816,
      "step": 870
    },
    {
      "epoch": 2.583055863854976,
      "grad_norm": 0.024887174367904663,
      "learning_rate": 1.1479119620864277e-06,
      "loss": 0.6674,
      "mean_token_accuracy": 0.7934547992415648,
      "step": 871
    },
    {
      "epoch": 2.586015538290788,
      "grad_norm": 0.026073411107063293,
      "learning_rate": 1.1318873061913405e-06,
      "loss": 0.5978,
      "mean_token_accuracy": 0.8119518202925882,
      "step": 872
    },
    {
      "epoch": 2.5889752127266,
      "grad_norm": 0.02333136461675167,
      "learning_rate": 1.1159685763395113e-06,
      "loss": 0.6497,
      "mean_token_accuracy": 0.796374926581197,
      "step": 873
    },
    {
      "epoch": 2.591934887162412,
      "grad_norm": 0.02479637786746025,
      "learning_rate": 1.1001559626737757e-06,
      "loss": 0.6516,
      "mean_token_accuracy": 0.796101681709061,
      "step": 874
    },
    {
      "epoch": 2.594894561598224,
      "grad_norm": 0.02537384442985058,
      "learning_rate": 1.0844496540694515e-06,
      "loss": 0.6365,
      "mean_token_accuracy": 0.8016051534534484,
      "step": 875
    },
    {
      "epoch": 2.597854236034036,
      "grad_norm": 0.025600122287869453,
      "learning_rate": 1.0688498381320855e-06,
      "loss": 0.6421,
      "mean_token_accuracy": 0.7988657540129854,
      "step": 876
    },
    {
      "epoch": 2.6008139104698484,
      "grad_norm": 0.025285426527261734,
      "learning_rate": 1.0533567011952094e-06,
      "loss": 0.6557,
      "mean_token_accuracy": 0.7942064586783343,
      "step": 877
    },
    {
      "epoch": 2.6037735849056602,
      "grad_norm": 0.02693473920226097,
      "learning_rate": 1.037970428318118e-06,
      "loss": 0.6943,
      "mean_token_accuracy": 0.7812896157512945,
      "step": 878
    },
    {
      "epoch": 2.6067332593414725,
      "grad_norm": 0.025598160922527313,
      "learning_rate": 1.022691203283661e-06,
      "loss": 0.6608,
      "mean_token_accuracy": 0.7923492517580604,
      "step": 879
    },
    {
      "epoch": 2.6096929337772847,
      "grad_norm": 0.025789178907871246,
      "learning_rate": 1.0075192085960451e-06,
      "loss": 0.6779,
      "mean_token_accuracy": 0.7885042642671458,
      "step": 880
    },
    {
      "epoch": 2.6126526082130965,
      "grad_norm": 0.024856319651007652,
      "learning_rate": 9.924546254786493e-07,
      "loss": 0.6407,
      "mean_token_accuracy": 0.7998727243366125,
      "step": 881
    },
    {
      "epoch": 2.6156122826489088,
      "grad_norm": 0.02484741620719433,
      "learning_rate": 9.77497633871868e-07,
      "loss": 0.6154,
      "mean_token_accuracy": 0.8069910581994415,
      "step": 882
    },
    {
      "epoch": 2.6185719570847206,
      "grad_norm": 0.024206671863794327,
      "learning_rate": 9.62648412430951e-07,
      "loss": 0.6892,
      "mean_token_accuracy": 0.7860742805534731,
      "step": 883
    },
    {
      "epoch": 2.621531631520533,
      "grad_norm": 0.02509419061243534,
      "learning_rate": 9.479071385238892e-07,
      "loss": 0.6314,
      "mean_token_accuracy": 0.8018883117100286,
      "step": 884
    },
    {
      "epoch": 2.6244913059563446,
      "grad_norm": 0.02489032968878746,
      "learning_rate": 9.332739882292752e-07,
      "loss": 0.6395,
      "mean_token_accuracy": 0.799507930802213,
      "step": 885
    },
    {
      "epoch": 2.627450980392157,
      "grad_norm": 0.024056807160377502,
      "learning_rate": 9.187491363342094e-07,
      "loss": 0.6019,
      "mean_token_accuracy": 0.8121379855585256,
      "step": 886
    },
    {
      "epoch": 2.630410654827969,
      "grad_norm": 0.024712828919291496,
      "learning_rate": 9.043327563322113e-07,
      "loss": 0.648,
      "mean_token_accuracy": 0.7977265157316199,
      "step": 887
    },
    {
      "epoch": 2.633370329263781,
      "grad_norm": 0.024342121556401253,
      "learning_rate": 8.900250204211513e-07,
      "loss": 0.6364,
      "mean_token_accuracy": 0.8029139209078922,
      "step": 888
    },
    {
      "epoch": 2.636330003699593,
      "grad_norm": 0.024350209161639214,
      "learning_rate": 8.758260995011825e-07,
      "loss": 0.6348,
      "mean_token_accuracy": 0.8018293453891258,
      "step": 889
    },
    {
      "epoch": 2.6392896781354054,
      "grad_norm": 0.024455633014440536,
      "learning_rate": 8.617361631727139e-07,
      "loss": 0.6471,
      "mean_token_accuracy": 0.7972542185423029,
      "step": 890
    },
    {
      "epoch": 2.642249352571217,
      "grad_norm": 0.02435615472495556,
      "learning_rate": 8.477553797343729e-07,
      "loss": 0.6031,
      "mean_token_accuracy": 0.8102795388579896,
      "step": 891
    },
    {
      "epoch": 2.645209027007029,
      "grad_norm": 0.024372348561882973,
      "learning_rate": 8.338839161809997e-07,
      "loss": 0.6358,
      "mean_token_accuracy": 0.8012500910642769,
      "step": 892
    },
    {
      "epoch": 2.648168701442841,
      "grad_norm": 0.02625070884823799,
      "learning_rate": 8.201219382016556e-07,
      "loss": 0.6527,
      "mean_token_accuracy": 0.7961821396905197,
      "step": 893
    },
    {
      "epoch": 2.6511283758786535,
      "grad_norm": 0.024691645056009293,
      "learning_rate": 8.06469610177636e-07,
      "loss": 0.6462,
      "mean_token_accuracy": 0.7976159496492162,
      "step": 894
    },
    {
      "epoch": 2.6540880503144653,
      "grad_norm": 0.02393011562526226,
      "learning_rate": 7.92927095180518e-07,
      "loss": 0.6596,
      "mean_token_accuracy": 0.7939745206822114,
      "step": 895
    },
    {
      "epoch": 2.6570477247502775,
      "grad_norm": 0.02625460736453533,
      "learning_rate": 7.794945549701993e-07,
      "loss": 0.6185,
      "mean_token_accuracy": 0.8061236282474503,
      "step": 896
    },
    {
      "epoch": 2.6600073991860897,
      "grad_norm": 0.023810414597392082,
      "learning_rate": 7.661721499929753e-07,
      "loss": 0.6173,
      "mean_token_accuracy": 0.8056948960382845,
      "step": 897
    },
    {
      "epoch": 2.6629670736219015,
      "grad_norm": 0.025793075561523438,
      "learning_rate": 7.529600393796232e-07,
      "loss": 0.6445,
      "mean_token_accuracy": 0.7986423215004738,
      "step": 898
    },
    {
      "epoch": 2.665926748057714,
      "grad_norm": 0.024996111169457436,
      "learning_rate": 7.398583809434944e-07,
      "loss": 0.6292,
      "mean_token_accuracy": 0.8045569361071508,
      "step": 899
    },
    {
      "epoch": 2.6688864224935256,
      "grad_norm": 0.024181660264730453,
      "learning_rate": 7.268673311786378e-07,
      "loss": 0.6645,
      "mean_token_accuracy": 0.7917808356429421,
      "step": 900
    },
    {
      "epoch": 2.6688864224935256,
      "eval_loss": 0.738807737827301,
      "eval_mean_token_accuracy": 0.771508710465834,
      "eval_runtime": 25.0666,
      "eval_samples_per_second": 5.146,
      "eval_steps_per_second": 1.316,
      "step": 900
    },
    {
      "epoch": 2.671846096929338,
      "grad_norm": 0.02503887563943863,
      "learning_rate": 7.1398704525792e-07,
      "loss": 0.6766,
      "mean_token_accuracy": 0.7880289900957871,
      "step": 901
    },
    {
      "epoch": 2.6748057713651496,
      "grad_norm": 0.026051614433526993,
      "learning_rate": 7.012176770311863e-07,
      "loss": 0.6725,
      "mean_token_accuracy": 0.7896488006356566,
      "step": 902
    },
    {
      "epoch": 2.677765445800962,
      "grad_norm": 0.02513834275305271,
      "learning_rate": 6.885593790234057e-07,
      "loss": 0.6472,
      "mean_token_accuracy": 0.7970934905924975,
      "step": 903
    },
    {
      "epoch": 2.680725120236774,
      "grad_norm": 0.024156536906957626,
      "learning_rate": 6.760123024328624e-07,
      "loss": 0.6237,
      "mean_token_accuracy": 0.8055969734199404,
      "step": 904
    },
    {
      "epoch": 2.683684794672586,
      "grad_norm": 0.024541374295949936,
      "learning_rate": 6.635765971293484e-07,
      "loss": 0.6665,
      "mean_token_accuracy": 0.7917083397566683,
      "step": 905
    },
    {
      "epoch": 2.686644469108398,
      "grad_norm": 0.0244311410933733,
      "learning_rate": 6.512524116523633e-07,
      "loss": 0.6375,
      "mean_token_accuracy": 0.8013061395180803,
      "step": 906
    },
    {
      "epoch": 2.6896041435442104,
      "grad_norm": 0.02466966025531292,
      "learning_rate": 6.390398932093555e-07,
      "loss": 0.6176,
      "mean_token_accuracy": 0.8057496292276121,
      "step": 907
    },
    {
      "epoch": 2.692563817980022,
      "grad_norm": 0.023878371343016624,
      "learning_rate": 6.269391876739494e-07,
      "loss": 0.6399,
      "mean_token_accuracy": 0.8016521868224971,
      "step": 908
    },
    {
      "epoch": 2.695523492415834,
      "grad_norm": 0.0244602020829916,
      "learning_rate": 6.149504395842087e-07,
      "loss": 0.6252,
      "mean_token_accuracy": 0.805927353486149,
      "step": 909
    },
    {
      "epoch": 2.6984831668516462,
      "grad_norm": 0.0248147863894701,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.6685,
      "mean_token_accuracy": 0.7912146618011965,
      "step": 910
    },
    {
      "epoch": 2.7014428412874585,
      "grad_norm": 0.02523919567465782,
      "learning_rate": 5.913093872058528e-07,
      "loss": 0.6709,
      "mean_token_accuracy": 0.7915945180644234,
      "step": 911
    },
    {
      "epoch": 2.7044025157232703,
      "grad_norm": 0.0254255048930645,
      "learning_rate": 5.796573653001091e-07,
      "loss": 0.6438,
      "mean_token_accuracy": 0.7992025463389314,
      "step": 912
    },
    {
      "epoch": 2.7073621901590825,
      "grad_norm": 0.025221869349479675,
      "learning_rate": 5.681178656024055e-07,
      "loss": 0.6724,
      "mean_token_accuracy": 0.7908819503712674,
      "step": 913
    },
    {
      "epoch": 2.7103218645948948,
      "grad_norm": 0.02520192414522171,
      "learning_rate": 5.56691025947429e-07,
      "loss": 0.6479,
      "mean_token_accuracy": 0.797362315578991,
      "step": 914
    },
    {
      "epoch": 2.7132815390307066,
      "grad_norm": 0.023382479324936867,
      "learning_rate": 5.453769828241872e-07,
      "loss": 0.6362,
      "mean_token_accuracy": 0.8003035239148327,
      "step": 915
    },
    {
      "epoch": 2.716241213466519,
      "grad_norm": 0.024721728637814522,
      "learning_rate": 5.341758713743828e-07,
      "loss": 0.6693,
      "mean_token_accuracy": 0.7907774298266615,
      "step": 916
    },
    {
      "epoch": 2.7192008879023306,
      "grad_norm": 0.024578237906098366,
      "learning_rate": 5.230878253907911e-07,
      "loss": 0.6514,
      "mean_token_accuracy": 0.7970180769866232,
      "step": 917
    },
    {
      "epoch": 2.722160562338143,
      "grad_norm": 0.02482464909553528,
      "learning_rate": 5.121129773156663e-07,
      "loss": 0.6867,
      "mean_token_accuracy": 0.7856522327001212,
      "step": 918
    },
    {
      "epoch": 2.7251202367739547,
      "grad_norm": 0.025747794657945633,
      "learning_rate": 5.012514582391592e-07,
      "loss": 0.6296,
      "mean_token_accuracy": 0.8027714730751373,
      "step": 919
    },
    {
      "epoch": 2.728079911209767,
      "grad_norm": 0.0240219347178936,
      "learning_rate": 4.905033978977492e-07,
      "loss": 0.6383,
      "mean_token_accuracy": 0.8011803596174307,
      "step": 920
    },
    {
      "epoch": 2.731039585645579,
      "grad_norm": 0.02373625710606575,
      "learning_rate": 4.798689246727006e-07,
      "loss": 0.6243,
      "mean_token_accuracy": 0.8047336419880804,
      "step": 921
    },
    {
      "epoch": 2.733999260081391,
      "grad_norm": 0.024487396702170372,
      "learning_rate": 4.693481655885257e-07,
      "loss": 0.6798,
      "mean_token_accuracy": 0.7893838597760202,
      "step": 922
    },
    {
      "epoch": 2.736958934517203,
      "grad_norm": 0.02495640330016613,
      "learning_rate": 4.58941246311464e-07,
      "loss": 0.64,
      "mean_token_accuracy": 0.8001018380042697,
      "step": 923
    },
    {
      "epoch": 2.7399186089530154,
      "grad_norm": 0.024247821420431137,
      "learning_rate": 4.4864829114798394e-07,
      "loss": 0.6467,
      "mean_token_accuracy": 0.7979800032324957,
      "step": 924
    },
    {
      "epoch": 2.742878283388827,
      "grad_norm": 0.02434772253036499,
      "learning_rate": 4.384694230432984e-07,
      "loss": 0.6045,
      "mean_token_accuracy": 0.8115274818481392,
      "step": 925
    },
    {
      "epoch": 2.745837957824639,
      "grad_norm": 0.024380499497056007,
      "learning_rate": 4.2840476357989825e-07,
      "loss": 0.6208,
      "mean_token_accuracy": 0.8054650293308171,
      "step": 926
    },
    {
      "epoch": 2.7487976322604513,
      "grad_norm": 0.025606529787182808,
      "learning_rate": 4.184544329761009e-07,
      "loss": 0.6308,
      "mean_token_accuracy": 0.8033098980649387,
      "step": 927
    },
    {
      "epoch": 2.7517573066962635,
      "grad_norm": 0.024429449811577797,
      "learning_rate": 4.0861855008460403e-07,
      "loss": 0.6428,
      "mean_token_accuracy": 0.7993372766101048,
      "step": 928
    },
    {
      "epoch": 2.7547169811320753,
      "grad_norm": 0.02489388920366764,
      "learning_rate": 3.988972323910778e-07,
      "loss": 0.6278,
      "mean_token_accuracy": 0.8029963477440261,
      "step": 929
    },
    {
      "epoch": 2.7576766555678875,
      "grad_norm": 0.023982353508472443,
      "learning_rate": 3.8929059601275463e-07,
      "loss": 0.6359,
      "mean_token_accuracy": 0.8001376427714603,
      "step": 930
    },
    {
      "epoch": 2.7606363300037,
      "grad_norm": 0.026325182989239693,
      "learning_rate": 3.797987556970495e-07,
      "loss": 0.6821,
      "mean_token_accuracy": 0.7880444376553832,
      "step": 931
    },
    {
      "epoch": 2.7635960044395116,
      "grad_norm": 0.025254786014556885,
      "learning_rate": 3.7042182482018074e-07,
      "loss": 0.6372,
      "mean_token_accuracy": 0.8017578644051205,
      "step": 932
    },
    {
      "epoch": 2.766555678875324,
      "grad_norm": 0.02403276413679123,
      "learning_rate": 3.611599153858214e-07,
      "loss": 0.6875,
      "mean_token_accuracy": 0.78475783245207,
      "step": 933
    },
    {
      "epoch": 2.7695153533111356,
      "grad_norm": 0.025127604603767395,
      "learning_rate": 3.520131380237546e-07,
      "loss": 0.675,
      "mean_token_accuracy": 0.7890153834639706,
      "step": 934
    },
    {
      "epoch": 2.772475027746948,
      "grad_norm": 0.025778232142329216,
      "learning_rate": 3.429816019885657e-07,
      "loss": 0.6916,
      "mean_token_accuracy": 0.7848815835819323,
      "step": 935
    },
    {
      "epoch": 2.7754347021827597,
      "grad_norm": 0.025138361379504204,
      "learning_rate": 3.3406541515832e-07,
      "loss": 0.6889,
      "mean_token_accuracy": 0.7836224085566339,
      "step": 936
    },
    {
      "epoch": 2.778394376618572,
      "grad_norm": 0.02500680834054947,
      "learning_rate": 3.252646840332918e-07,
      "loss": 0.657,
      "mean_token_accuracy": 0.7943578327625707,
      "step": 937
    },
    {
      "epoch": 2.781354051054384,
      "grad_norm": 0.023927554488182068,
      "learning_rate": 3.16579513734675e-07,
      "loss": 0.6356,
      "mean_token_accuracy": 0.8012900648956005,
      "step": 938
    },
    {
      "epoch": 2.784313725490196,
      "grad_norm": 0.024294385686516762,
      "learning_rate": 3.080100080033388e-07,
      "loss": 0.632,
      "mean_token_accuracy": 0.8028005606873495,
      "step": 939
    },
    {
      "epoch": 2.787273399926008,
      "grad_norm": 0.023370586335659027,
      "learning_rate": 2.995562691985898e-07,
      "loss": 0.6376,
      "mean_token_accuracy": 0.8005631755973619,
      "step": 940
    },
    {
      "epoch": 2.7902330743618204,
      "grad_norm": 0.025261515751481056,
      "learning_rate": 2.9121839829693857e-07,
      "loss": 0.629,
      "mean_token_accuracy": 0.8027061171973076,
      "step": 941
    },
    {
      "epoch": 2.7931927487976322,
      "grad_norm": 0.025184744969010353,
      "learning_rate": 2.829964948909048e-07,
      "loss": 0.6332,
      "mean_token_accuracy": 0.8013635705262054,
      "step": 942
    },
    {
      "epoch": 2.7961524232334445,
      "grad_norm": 0.02480281889438629,
      "learning_rate": 2.748906571878207e-07,
      "loss": 0.6706,
      "mean_token_accuracy": 0.7922778381845065,
      "step": 943
    },
    {
      "epoch": 2.7991120976692563,
      "grad_norm": 0.024982471019029617,
      "learning_rate": 2.6690098200866097e-07,
      "loss": 0.6518,
      "mean_token_accuracy": 0.7972574162482862,
      "step": 944
    },
    {
      "epoch": 2.8020717721050685,
      "grad_norm": 0.025391558185219765,
      "learning_rate": 2.5902756478688674e-07,
      "loss": 0.6534,
      "mean_token_accuracy": 0.7962430394707019,
      "step": 945
    },
    {
      "epoch": 2.8050314465408803,
      "grad_norm": 0.0246841199696064,
      "learning_rate": 2.5127049956730207e-07,
      "loss": 0.6526,
      "mean_token_accuracy": 0.7955391888444348,
      "step": 946
    },
    {
      "epoch": 2.8079911209766926,
      "grad_norm": 0.024317042902112007,
      "learning_rate": 2.436298790049363e-07,
      "loss": 0.6755,
      "mean_token_accuracy": 0.7898710311844659,
      "step": 947
    },
    {
      "epoch": 2.810950795412505,
      "grad_norm": 0.024903837591409683,
      "learning_rate": 2.3610579436392999e-07,
      "loss": 0.6556,
      "mean_token_accuracy": 0.7956978937757702,
      "step": 948
    },
    {
      "epoch": 2.8139104698483166,
      "grad_norm": 0.02431805431842804,
      "learning_rate": 2.2869833551645293e-07,
      "loss": 0.6561,
      "mean_token_accuracy": 0.7947831758698982,
      "step": 949
    },
    {
      "epoch": 2.816870144284129,
      "grad_norm": 0.024643970653414726,
      "learning_rate": 2.2140759094162468e-07,
      "loss": 0.6542,
      "mean_token_accuracy": 0.7953192952935193,
      "step": 950
    },
    {
      "epoch": 2.8198298187199407,
      "grad_norm": 0.024292340502142906,
      "learning_rate": 2.1423364772445886e-07,
      "loss": 0.6326,
      "mean_token_accuracy": 0.801863821030254,
      "step": 951
    },
    {
      "epoch": 2.822789493155753,
      "grad_norm": 0.024100741371512413,
      "learning_rate": 2.071765915548274e-07,
      "loss": 0.6983,
      "mean_token_accuracy": 0.7829299157268246,
      "step": 952
    },
    {
      "epoch": 2.8257491675915647,
      "grad_norm": 0.025073310360312462,
      "learning_rate": 2.002365067264289e-07,
      "loss": 0.6835,
      "mean_token_accuracy": 0.7867106025298752,
      "step": 953
    },
    {
      "epoch": 2.828708842027377,
      "grad_norm": 0.025796640664339066,
      "learning_rate": 1.9341347613579086e-07,
      "loss": 0.6284,
      "mean_token_accuracy": 0.8044256973718155,
      "step": 954
    },
    {
      "epoch": 2.831668516463189,
      "grad_norm": 0.025663964450359344,
      "learning_rate": 1.867075812812691e-07,
      "loss": 0.6486,
      "mean_token_accuracy": 0.7982518960756937,
      "step": 955
    },
    {
      "epoch": 2.834628190899001,
      "grad_norm": 0.024320784956216812,
      "learning_rate": 1.8011890226208527e-07,
      "loss": 0.6229,
      "mean_token_accuracy": 0.8051208299906005,
      "step": 956
    },
    {
      "epoch": 2.837587865334813,
      "grad_norm": 0.023920755833387375,
      "learning_rate": 1.7364751777736334e-07,
      "loss": 0.6608,
      "mean_token_accuracy": 0.7913435589780335,
      "step": 957
    },
    {
      "epoch": 2.8405475397706255,
      "grad_norm": 0.024578483775258064,
      "learning_rate": 1.6729350512519006e-07,
      "loss": 0.6488,
      "mean_token_accuracy": 0.7976782731768994,
      "step": 958
    },
    {
      "epoch": 2.8435072142064373,
      "grad_norm": 0.023927923291921616,
      "learning_rate": 1.6105694020169594e-07,
      "loss": 0.6429,
      "mean_token_accuracy": 0.7981146937230165,
      "step": 959
    },
    {
      "epoch": 2.8464668886422495,
      "grad_norm": 0.024035179987549782,
      "learning_rate": 1.5493789750014032e-07,
      "loss": 0.6283,
      "mean_token_accuracy": 0.8015194859849463,
      "step": 960
    },
    {
      "epoch": 2.8494265630780613,
      "grad_norm": 0.02438296563923359,
      "learning_rate": 1.489364501100332e-07,
      "loss": 0.6657,
      "mean_token_accuracy": 0.7931666858315013,
      "step": 961
    },
    {
      "epoch": 2.8523862375138735,
      "grad_norm": 0.024643562734127045,
      "learning_rate": 1.430526697162482e-07,
      "loss": 0.6548,
      "mean_token_accuracy": 0.7951223486333518,
      "step": 962
    },
    {
      "epoch": 2.8553459119496853,
      "grad_norm": 0.02407108247280121,
      "learning_rate": 1.3728662659818205e-07,
      "loss": 0.6135,
      "mean_token_accuracy": 0.808081339789838,
      "step": 963
    },
    {
      "epoch": 2.8583055863854976,
      "grad_norm": 0.023972326889634132,
      "learning_rate": 1.3163838962890196e-07,
      "loss": 0.6703,
      "mean_token_accuracy": 0.7895250771918654,
      "step": 964
    },
    {
      "epoch": 2.86126526082131,
      "grad_norm": 0.025045253336429596,
      "learning_rate": 1.2610802627432972e-07,
      "loss": 0.6525,
      "mean_token_accuracy": 0.7964547026980748,
      "step": 965
    },
    {
      "epoch": 2.8642249352571216,
      "grad_norm": 0.025660628452897072,
      "learning_rate": 1.206956025924333e-07,
      "loss": 0.6367,
      "mean_token_accuracy": 0.7998873486537863,
      "step": 966
    },
    {
      "epoch": 2.867184609692934,
      "grad_norm": 0.023965923115611076,
      "learning_rate": 1.1540118323243866e-07,
      "loss": 0.5799,
      "mean_token_accuracy": 0.817906123790644,
      "step": 967
    },
    {
      "epoch": 2.870144284128746,
      "grad_norm": 0.02407275140285492,
      "learning_rate": 1.1022483143405705e-07,
      "loss": 0.6173,
      "mean_token_accuracy": 0.8078425425436084,
      "step": 968
    },
    {
      "epoch": 2.873103958564558,
      "grad_norm": 0.024154920130968094,
      "learning_rate": 1.0516660902673448e-07,
      "loss": 0.6571,
      "mean_token_accuracy": 0.7956497116361156,
      "step": 969
    },
    {
      "epoch": 2.8760636330003697,
      "grad_norm": 0.023997806012630463,
      "learning_rate": 1.0022657642890232e-07,
      "loss": 0.6329,
      "mean_token_accuracy": 0.8014162111918257,
      "step": 970
    },
    {
      "epoch": 2.879023307436182,
      "grad_norm": 0.02487342245876789,
      "learning_rate": 9.540479264726676e-08,
      "loss": 0.6611,
      "mean_token_accuracy": 0.7919628230949652,
      "step": 971
    },
    {
      "epoch": 2.881982981871994,
      "grad_norm": 0.02392645925283432,
      "learning_rate": 9.070131527609604e-08,
      "loss": 0.6176,
      "mean_token_accuracy": 0.8069237870660606,
      "step": 972
    },
    {
      "epoch": 2.884942656307806,
      "grad_norm": 0.02404465526342392,
      "learning_rate": 8.61162004965388e-08,
      "loss": 0.6447,
      "mean_token_accuracy": 0.7989509125985043,
      "step": 973
    },
    {
      "epoch": 2.8879023307436182,
      "grad_norm": 0.025083765387535095,
      "learning_rate": 8.16495030759501e-08,
      "loss": 0.6804,
      "mean_token_accuracy": 0.7879754814988429,
      "step": 974
    },
    {
      "epoch": 2.8908620051794305,
      "grad_norm": 0.024577250704169273,
      "learning_rate": 7.730127636723539e-08,
      "loss": 0.6101,
      "mean_token_accuracy": 0.809471025901658,
      "step": 975
    },
    {
      "epoch": 2.8938216796152423,
      "grad_norm": 0.024379462003707886,
      "learning_rate": 7.307157230821426e-08,
      "loss": 0.6557,
      "mean_token_accuracy": 0.7952257879623418,
      "step": 976
    },
    {
      "epoch": 2.8967813540510545,
      "grad_norm": 0.02482825517654419,
      "learning_rate": 6.896044142100433e-08,
      "loss": 0.6676,
      "mean_token_accuracy": 0.791864596358556,
      "step": 977
    },
    {
      "epoch": 2.8997410284868663,
      "grad_norm": 0.025561396032571793,
      "learning_rate": 6.496793281141056e-08,
      "loss": 0.6873,
      "mean_token_accuracy": 0.7856712287579546,
      "step": 978
    },
    {
      "epoch": 2.9027007029226786,
      "grad_norm": 0.023899441584944725,
      "learning_rate": 6.109409416834689e-08,
      "loss": 0.6511,
      "mean_token_accuracy": 0.7969160513759264,
      "step": 979
    },
    {
      "epoch": 2.9056603773584904,
      "grad_norm": 0.02500973455607891,
      "learning_rate": 5.7338971763256646e-08,
      "loss": 0.6322,
      "mean_token_accuracy": 0.8027010748620524,
      "step": 980
    },
    {
      "epoch": 2.9086200517943026,
      "grad_norm": 0.023463763296604156,
      "learning_rate": 5.37026104495697e-08,
      "loss": 0.6408,
      "mean_token_accuracy": 0.8007576973535182,
      "step": 981
    },
    {
      "epoch": 2.911579726230115,
      "grad_norm": 0.025442739948630333,
      "learning_rate": 5.0185053662161756e-08,
      "loss": 0.6422,
      "mean_token_accuracy": 0.7980975950167215,
      "step": 982
    },
    {
      "epoch": 2.9145394006659266,
      "grad_norm": 0.024705683812499046,
      "learning_rate": 4.678634341683252e-08,
      "loss": 0.6315,
      "mean_token_accuracy": 0.8026462818772715,
      "step": 983
    },
    {
      "epoch": 2.917499075101739,
      "grad_norm": 0.02321893535554409,
      "learning_rate": 4.350652030981395e-08,
      "loss": 0.6542,
      "mean_token_accuracy": 0.7955604152394864,
      "step": 984
    },
    {
      "epoch": 2.920458749537551,
      "grad_norm": 0.024503175169229507,
      "learning_rate": 4.0345623517273894e-08,
      "loss": 0.636,
      "mean_token_accuracy": 0.8011414448121752,
      "step": 985
    },
    {
      "epoch": 2.923418423973363,
      "grad_norm": 0.024266136810183525,
      "learning_rate": 3.7303690794854296e-08,
      "loss": 0.7167,
      "mean_token_accuracy": 0.7765485087523469,
      "step": 986
    },
    {
      "epoch": 2.9263780984091747,
      "grad_norm": 0.025390757247805595,
      "learning_rate": 3.438075847721933e-08,
      "loss": 0.594,
      "mean_token_accuracy": 0.8128088417434304,
      "step": 987
    },
    {
      "epoch": 2.929337772844987,
      "grad_norm": 0.02399209700524807,
      "learning_rate": 3.157686147762129e-08,
      "loss": 0.6577,
      "mean_token_accuracy": 0.7952499631209614,
      "step": 988
    },
    {
      "epoch": 2.932297447280799,
      "grad_norm": 0.024280982092022896,
      "learning_rate": 2.8892033287484245e-08,
      "loss": 0.6746,
      "mean_token_accuracy": 0.7908002646119627,
      "step": 989
    },
    {
      "epoch": 2.935257121716611,
      "grad_norm": 0.024518564343452454,
      "learning_rate": 2.6326305976001054e-08,
      "loss": 0.6113,
      "mean_token_accuracy": 0.808108495541965,
      "step": 990
    },
    {
      "epoch": 2.9382167961524233,
      "grad_norm": 0.02435346506536007,
      "learning_rate": 2.3879710189753657e-08,
      "loss": 0.6491,
      "mean_token_accuracy": 0.7966765281876945,
      "step": 991
    },
    {
      "epoch": 2.9411764705882355,
      "grad_norm": 0.023688076063990593,
      "learning_rate": 2.1552275152346702e-08,
      "loss": 0.647,
      "mean_token_accuracy": 0.7983097586711366,
      "step": 992
    },
    {
      "epoch": 2.9441361450240473,
      "grad_norm": 0.024345990270376205,
      "learning_rate": 1.9344028664056715e-08,
      "loss": 0.6127,
      "mean_token_accuracy": 0.8075949132288531,
      "step": 993
    },
    {
      "epoch": 2.9470958194598595,
      "grad_norm": 0.02496313489973545,
      "learning_rate": 1.7254997101500137e-08,
      "loss": 0.6302,
      "mean_token_accuracy": 0.8037297708156844,
      "step": 994
    },
    {
      "epoch": 2.9500554938956713,
      "grad_norm": 0.024208253249526024,
      "learning_rate": 1.528520541731915e-08,
      "loss": 0.6179,
      "mean_token_accuracy": 0.8066043471491108,
      "step": 995
    },
    {
      "epoch": 2.9530151683314836,
      "grad_norm": 0.02446601912379265,
      "learning_rate": 1.3434677139885222e-08,
      "loss": 0.6637,
      "mean_token_accuracy": 0.7923625658161302,
      "step": 996
    },
    {
      "epoch": 2.9559748427672954,
      "grad_norm": 0.02337535284459591,
      "learning_rate": 1.170343437301491e-08,
      "loss": 0.6558,
      "mean_token_accuracy": 0.7938551843926827,
      "step": 997
    },
    {
      "epoch": 2.9589345172031076,
      "grad_norm": 0.02501841075718403,
      "learning_rate": 1.0091497795706728e-08,
      "loss": 0.6497,
      "mean_token_accuracy": 0.7968393039357107,
      "step": 998
    },
    {
      "epoch": 2.96189419163892,
      "grad_norm": 0.025258854031562805,
      "learning_rate": 8.59888666189579e-09,
      "loss": 0.686,
      "mean_token_accuracy": 0.7872833543596387,
      "step": 999
    },
    {
      "epoch": 2.9648538660747317,
      "grad_norm": 0.025254277512431145,
      "learning_rate": 7.225618800222878e-09,
      "loss": 0.6709,
      "mean_token_accuracy": 0.7897079843617865,
      "step": 1000
    },
    {
      "epoch": 2.9648538660747317,
      "eval_loss": 0.7385186553001404,
      "eval_mean_token_accuracy": 0.7715846062541177,
      "eval_runtime": 24.7087,
      "eval_samples_per_second": 5.221,
      "eval_steps_per_second": 1.336,
      "step": 1000
    },
    {
      "epoch": 2.967813540510544,
      "grad_norm": 0.024483192712068558,
      "learning_rate": 5.971710613821291e-09,
      "loss": 0.6833,
      "mean_token_accuracy": 0.7869484667394264,
      "step": 1001
    },
    {
      "epoch": 2.970773214946356,
      "grad_norm": 0.0248421598225832,
      "learning_rate": 4.837177080119215e-09,
      "loss": 0.6321,
      "mean_token_accuracy": 0.8007169446084274,
      "step": 1002
    },
    {
      "epoch": 2.973732889382168,
      "grad_norm": 0.024752933531999588,
      "learning_rate": 3.8220317506654226e-09,
      "loss": 0.6638,
      "mean_token_accuracy": 0.7926716064057001,
      "step": 1003
    },
    {
      "epoch": 2.9766925638179798,
      "grad_norm": 0.02514069154858589,
      "learning_rate": 2.9262867509605164e-09,
      "loss": 0.6494,
      "mean_token_accuracy": 0.7968540201847713,
      "step": 1004
    },
    {
      "epoch": 2.979652238253792,
      "grad_norm": 0.025022801011800766,
      "learning_rate": 2.149952780321485e-09,
      "loss": 0.6744,
      "mean_token_accuracy": 0.7884714897121342,
      "step": 1005
    },
    {
      "epoch": 2.9826119126896042,
      "grad_norm": 0.02561136893928051,
      "learning_rate": 1.4930391117451427e-09,
      "loss": 0.6456,
      "mean_token_accuracy": 0.7974667041991319,
      "step": 1006
    },
    {
      "epoch": 2.985571587125416,
      "grad_norm": 0.026112880557775497,
      "learning_rate": 9.555535917993297e-10,
      "loss": 0.7062,
      "mean_token_accuracy": 0.7800898238870632,
      "step": 1007
    },
    {
      "epoch": 2.9885312615612283,
      "grad_norm": 0.0258809681981802,
      "learning_rate": 5.375026405352035e-10,
      "loss": 0.6348,
      "mean_token_accuracy": 0.801181635576265,
      "step": 1008
    },
    {
      "epoch": 2.9914909359970405,
      "grad_norm": 0.02392594888806343,
      "learning_rate": 2.388912514017516e-10,
      "loss": 0.6731,
      "mean_token_accuracy": 0.7911797656677796,
      "step": 1009
    },
    {
      "epoch": 2.9944506104328523,
      "grad_norm": 0.024932818487286568,
      "learning_rate": 5.972299119250124e-11,
      "loss": 0.5987,
      "mean_token_accuracy": 0.8113517558307672,
      "step": 1010
    },
    {
      "epoch": 2.9974102848686646,
      "grad_norm": 0.025314847007393837,
      "learning_rate": 0.0,
      "loss": 0.673,
      "mean_token_accuracy": 0.7901783390368937,
      "step": 1011
    },
    {
      "epoch": 2.9974102848686646,
      "step": 1011,
      "total_flos": 230593791000576.0,
      "train_loss": 0.7156181658528796,
      "train_runtime": 41815.0838,
      "train_samples_per_second": 1.551,
      "train_steps_per_second": 0.024
    }
  ],
  "logging_steps": 1,
  "max_steps": 1011,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 230593791000576.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}