{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 171,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017543859649122806,
      "grad_norm": 3.9338779838849725,
      "learning_rate": 0.0,
      "loss": 1.2337,
      "num_tokens": 415561.0,
      "step": 1
    },
    {
      "epoch": 0.03508771929824561,
      "grad_norm": 4.061961880981431,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.2551,
      "num_tokens": 811930.0,
      "step": 2
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 4.126747331618448,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.2772,
      "num_tokens": 1198366.0,
      "step": 3
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 3.410910565563502,
      "learning_rate": 5e-06,
      "loss": 1.1776,
      "num_tokens": 1608933.0,
      "step": 4
    },
    {
      "epoch": 0.08771929824561403,
      "grad_norm": 2.233718748318594,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9942,
      "num_tokens": 2068099.0,
      "step": 5
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 1.470021030676499,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.8448,
      "num_tokens": 2506575.0,
      "step": 6
    },
    {
      "epoch": 0.12280701754385964,
      "grad_norm": 1.4565364805839704,
      "learning_rate": 1e-05,
      "loss": 0.8213,
      "num_tokens": 2932014.0,
      "step": 7
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 2.3516845691230217,
      "learning_rate": 9.999184354855868e-06,
      "loss": 0.6275,
      "num_tokens": 3370254.0,
      "step": 8
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 1.8135724347288096,
      "learning_rate": 9.996737715102133e-06,
      "loss": 0.5931,
      "num_tokens": 3782161.0,
      "step": 9
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 1.4106997469551075,
      "learning_rate": 9.99266096766761e-06,
      "loss": 0.4933,
      "num_tokens": 4211970.0,
      "step": 10
    },
    {
      "epoch": 0.19298245614035087,
      "grad_norm": 0.8131031118317348,
      "learning_rate": 9.98695559040975e-06,
      "loss": 0.4229,
      "num_tokens": 4613367.0,
      "step": 11
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 0.7764134903704204,
      "learning_rate": 9.979623651578881e-06,
      "loss": 0.3888,
      "num_tokens": 5012484.0,
      "step": 12
    },
    {
      "epoch": 0.22807017543859648,
      "grad_norm": 0.37508885186302826,
      "learning_rate": 9.970667809068476e-06,
      "loss": 0.3783,
      "num_tokens": 5420781.0,
      "step": 13
    },
    {
      "epoch": 0.24561403508771928,
      "grad_norm": 0.3209831897748014,
      "learning_rate": 9.960091309451626e-06,
      "loss": 0.3308,
      "num_tokens": 5823019.0,
      "step": 14
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 0.2796386507143645,
      "learning_rate": 9.947897986804131e-06,
      "loss": 0.3456,
      "num_tokens": 6241736.0,
      "step": 15
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 0.25191330470494944,
      "learning_rate": 9.93409226131462e-06,
      "loss": 0.3172,
      "num_tokens": 6680804.0,
      "step": 16
    },
    {
      "epoch": 0.2982456140350877,
      "grad_norm": 0.2415593584049499,
      "learning_rate": 9.91867913768218e-06,
      "loss": 0.3253,
      "num_tokens": 7107939.0,
      "step": 17
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 0.3003770654260172,
      "learning_rate": 9.901664203302126e-06,
      "loss": 0.3213,
      "num_tokens": 7524281.0,
      "step": 18
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.3539478460817712,
      "learning_rate": 9.883053626240503e-06,
      "loss": 0.2966,
      "num_tokens": 7939796.0,
      "step": 19
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 0.2280690675113219,
      "learning_rate": 9.862854152998112e-06,
      "loss": 0.3043,
      "num_tokens": 8366406.0,
      "step": 20
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 0.21683554455375648,
      "learning_rate": 9.841073106064852e-06,
      "loss": 0.3057,
      "num_tokens": 8772526.0,
      "step": 21
    },
    {
      "epoch": 0.38596491228070173,
      "grad_norm": 0.19203110899204243,
      "learning_rate": 9.81771838126524e-06,
      "loss": 0.2899,
      "num_tokens": 9193705.0,
      "step": 22
    },
    {
      "epoch": 0.40350877192982454,
      "grad_norm": 0.21290762042544037,
      "learning_rate": 9.792798444896107e-06,
      "loss": 0.3073,
      "num_tokens": 9641419.0,
      "step": 23
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 0.1917960836643891,
      "learning_rate": 9.766322330657499e-06,
      "loss": 0.2921,
      "num_tokens": 10041552.0,
      "step": 24
    },
    {
      "epoch": 0.43859649122807015,
      "grad_norm": 0.18577210434865676,
      "learning_rate": 9.738299636377863e-06,
      "loss": 0.28,
      "num_tokens": 10476244.0,
      "step": 25
    },
    {
      "epoch": 0.45614035087719296,
      "grad_norm": 0.18309063662126532,
      "learning_rate": 9.70874052053476e-06,
      "loss": 0.2785,
      "num_tokens": 10863935.0,
      "step": 26
    },
    {
      "epoch": 0.47368421052631576,
      "grad_norm": 0.18434080604136077,
      "learning_rate": 9.677655698572326e-06,
      "loss": 0.2661,
      "num_tokens": 11259192.0,
      "step": 27
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 0.16956692098933204,
      "learning_rate": 9.645056439016827e-06,
      "loss": 0.273,
      "num_tokens": 11708724.0,
      "step": 28
    },
    {
      "epoch": 0.5087719298245614,
      "grad_norm": 0.16434051233500432,
      "learning_rate": 9.610954559391704e-06,
      "loss": 0.2667,
      "num_tokens": 12137403.0,
      "step": 29
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.17669427582556774,
      "learning_rate": 9.57536242193364e-06,
      "loss": 0.2584,
      "num_tokens": 12543239.0,
      "step": 30
    },
    {
      "epoch": 0.543859649122807,
      "grad_norm": 0.1652310917436867,
      "learning_rate": 9.538292929111114e-06,
      "loss": 0.2569,
      "num_tokens": 12940013.0,
      "step": 31
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 0.15646296852545927,
      "learning_rate": 9.499759518947156e-06,
      "loss": 0.2607,
      "num_tokens": 13409215.0,
      "step": 32
    },
    {
      "epoch": 0.5789473684210527,
      "grad_norm": 0.16249861759922887,
      "learning_rate": 9.459776160147941e-06,
      "loss": 0.2461,
      "num_tokens": 13806115.0,
      "step": 33
    },
    {
      "epoch": 0.5964912280701754,
      "grad_norm": 0.15613005334949157,
      "learning_rate": 9.418357347038999e-06,
      "loss": 0.2493,
      "num_tokens": 14248072.0,
      "step": 34
    },
    {
      "epoch": 0.6140350877192983,
      "grad_norm": 0.16063917924130028,
      "learning_rate": 9.375518094310904e-06,
      "loss": 0.255,
      "num_tokens": 14680852.0,
      "step": 35
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 0.15840497254274827,
      "learning_rate": 9.331273931576306e-06,
      "loss": 0.2459,
      "num_tokens": 15109781.0,
      "step": 36
    },
    {
      "epoch": 0.6491228070175439,
      "grad_norm": 0.1549313386657812,
      "learning_rate": 9.285640897740316e-06,
      "loss": 0.2461,
      "num_tokens": 15556489.0,
      "step": 37
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.15967707273841447,
      "learning_rate": 9.238635535186247e-06,
      "loss": 0.2358,
      "num_tokens": 15975315.0,
      "step": 38
    },
    {
      "epoch": 0.6842105263157895,
      "grad_norm": 0.1664153879515682,
      "learning_rate": 9.19027488377886e-06,
      "loss": 0.2453,
      "num_tokens": 16366263.0,
      "step": 39
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 0.16085937077964602,
      "learning_rate": 9.140576474687263e-06,
      "loss": 0.2339,
      "num_tokens": 16780696.0,
      "step": 40
    },
    {
      "epoch": 0.7192982456140351,
      "grad_norm": 0.1525506357194124,
      "learning_rate": 9.0895583240297e-06,
      "loss": 0.2454,
      "num_tokens": 17226365.0,
      "step": 41
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 0.15414442592489289,
      "learning_rate": 9.037238926342544e-06,
      "loss": 0.2315,
      "num_tokens": 17651462.0,
      "step": 42
    },
    {
      "epoch": 0.7543859649122807,
      "grad_norm": 0.160113618784668,
      "learning_rate": 8.983637247875872e-06,
      "loss": 0.24,
      "num_tokens": 18064676.0,
      "step": 43
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 0.1592020616116995,
      "learning_rate": 8.92877271971802e-06,
      "loss": 0.236,
      "num_tokens": 18474727.0,
      "step": 44
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 0.15245337800135356,
      "learning_rate": 8.872665230751644e-06,
      "loss": 0.2405,
      "num_tokens": 18902046.0,
      "step": 45
    },
    {
      "epoch": 0.8070175438596491,
      "grad_norm": 0.20723449258057428,
      "learning_rate": 8.815335120443822e-06,
      "loss": 0.224,
      "num_tokens": 19319711.0,
      "step": 46
    },
    {
      "epoch": 0.8245614035087719,
      "grad_norm": 0.1474862726874557,
      "learning_rate": 8.756803171472817e-06,
      "loss": 0.2412,
      "num_tokens": 19757131.0,
      "step": 47
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.15052042969864995,
      "learning_rate": 8.69709060219416e-06,
      "loss": 0.2327,
      "num_tokens": 20194869.0,
      "step": 48
    },
    {
      "epoch": 0.8596491228070176,
      "grad_norm": 0.15385476085770403,
      "learning_rate": 8.636219058948823e-06,
      "loss": 0.2266,
      "num_tokens": 20597224.0,
      "step": 49
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 0.15228109362462297,
      "learning_rate": 8.574210608216206e-06,
      "loss": 0.239,
      "num_tokens": 21031895.0,
      "step": 50
    },
    {
      "epoch": 0.8947368421052632,
      "grad_norm": 0.14813164351615135,
      "learning_rate": 8.511087728614863e-06,
      "loss": 0.2271,
      "num_tokens": 21445108.0,
      "step": 51
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 0.14876791381975332,
      "learning_rate": 8.446873302753783e-06,
      "loss": 0.2277,
      "num_tokens": 21886020.0,
      "step": 52
    },
    {
      "epoch": 0.9298245614035088,
      "grad_norm": 0.14949287901477407,
      "learning_rate": 8.381590608937251e-06,
      "loss": 0.2395,
      "num_tokens": 22300331.0,
      "step": 53
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 0.1413875737037552,
      "learning_rate": 8.315263312726248e-06,
      "loss": 0.2321,
      "num_tokens": 22731352.0,
      "step": 54
    },
    {
      "epoch": 0.9649122807017544,
      "grad_norm": 0.14923397563157811,
      "learning_rate": 8.247915458359473e-06,
      "loss": 0.2169,
      "num_tokens": 23159887.0,
      "step": 55
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 0.15363902325932008,
      "learning_rate": 8.179571460037096e-06,
      "loss": 0.2357,
      "num_tokens": 23618660.0,
      "step": 56
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.14687085036662337,
      "learning_rate": 8.110256093070393e-06,
      "loss": 0.2334,
      "num_tokens": 24039673.0,
      "step": 57
    },
    {
      "epoch": 1.0175438596491229,
      "grad_norm": 0.1476041876990848,
      "learning_rate": 8.039994484900463e-06,
      "loss": 0.2011,
      "num_tokens": 24434236.0,
      "step": 58
    },
    {
      "epoch": 1.0350877192982457,
      "grad_norm": 0.1419449695352204,
      "learning_rate": 7.968812105989316e-06,
      "loss": 0.2114,
      "num_tokens": 24853504.0,
      "step": 59
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 0.15337353318491537,
      "learning_rate": 7.896734760586599e-06,
      "loss": 0.2052,
      "num_tokens": 25255898.0,
      "step": 60
    },
    {
      "epoch": 1.0701754385964912,
      "grad_norm": 0.14113455068915454,
      "learning_rate": 7.82378857737533e-06,
      "loss": 0.2098,
      "num_tokens": 25667644.0,
      "step": 61
    },
    {
      "epoch": 1.087719298245614,
      "grad_norm": 0.14898616979585055,
      "learning_rate": 7.75e-06,
      "loss": 0.2041,
      "num_tokens": 26083880.0,
      "step": 62
    },
    {
      "epoch": 1.1052631578947367,
      "grad_norm": 0.15200676955309492,
      "learning_rate": 7.675395777480538e-06,
      "loss": 0.1911,
      "num_tokens": 26465737.0,
      "step": 63
    },
    {
      "epoch": 1.1228070175438596,
      "grad_norm": 0.15217717203165435,
      "learning_rate": 7.600002954515532e-06,
      "loss": 0.2113,
      "num_tokens": 26905038.0,
      "step": 64
    },
    {
      "epoch": 1.1403508771929824,
      "grad_norm": 0.1392806691058708,
      "learning_rate": 7.523848861678297e-06,
      "loss": 0.1981,
      "num_tokens": 27327803.0,
      "step": 65
    },
    {
      "epoch": 1.1578947368421053,
      "grad_norm": 0.14823210297550762,
      "learning_rate": 7.446961105509289e-06,
      "loss": 0.199,
      "num_tokens": 27743615.0,
      "step": 66
    },
    {
      "epoch": 1.1754385964912282,
      "grad_norm": 0.15025342094871685,
      "learning_rate": 7.36936755850849e-06,
      "loss": 0.2035,
      "num_tokens": 28159990.0,
      "step": 67
    },
    {
      "epoch": 1.1929824561403508,
      "grad_norm": 0.1424694767142918,
      "learning_rate": 7.2910963490313815e-06,
      "loss": 0.1997,
      "num_tokens": 28556831.0,
      "step": 68
    },
    {
      "epoch": 1.2105263157894737,
      "grad_norm": 0.14322092181407153,
      "learning_rate": 7.212175851092154e-06,
      "loss": 0.1961,
      "num_tokens": 28995194.0,
      "step": 69
    },
    {
      "epoch": 1.2280701754385965,
      "grad_norm": 0.1526511754370368,
      "learning_rate": 7.132634674077884e-06,
      "loss": 0.1921,
      "num_tokens": 29410916.0,
      "step": 70
    },
    {
      "epoch": 1.2456140350877192,
      "grad_norm": 0.139735447301268,
      "learning_rate": 7.052501652377368e-06,
      "loss": 0.2063,
      "num_tokens": 29913120.0,
      "step": 71
    },
    {
      "epoch": 1.263157894736842,
      "grad_norm": 0.14344116097027,
      "learning_rate": 6.971805834928399e-06,
      "loss": 0.2027,
      "num_tokens": 30351179.0,
      "step": 72
    },
    {
      "epoch": 1.280701754385965,
      "grad_norm": 0.1426174151835526,
      "learning_rate": 6.890576474687264e-06,
      "loss": 0.2064,
      "num_tokens": 30791604.0,
      "step": 73
    },
    {
      "epoch": 1.2982456140350878,
      "grad_norm": 0.14994921794042249,
      "learning_rate": 6.808843018024296e-06,
      "loss": 0.2065,
      "num_tokens": 31261315.0,
      "step": 74
    },
    {
      "epoch": 1.3157894736842106,
      "grad_norm": 0.15092392748993957,
      "learning_rate": 6.726635094049291e-06,
      "loss": 0.1908,
      "num_tokens": 31659426.0,
      "step": 75
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.14611148416178402,
      "learning_rate": 6.643982503870693e-06,
      "loss": 0.2051,
      "num_tokens": 32057379.0,
      "step": 76
    },
    {
      "epoch": 1.3508771929824561,
      "grad_norm": 0.1427497694651295,
      "learning_rate": 6.560915209792424e-06,
      "loss": 0.1992,
      "num_tokens": 32476948.0,
      "step": 77
    },
    {
      "epoch": 1.368421052631579,
      "grad_norm": 0.14625754427445514,
      "learning_rate": 6.477463324452286e-06,
      "loss": 0.1966,
      "num_tokens": 32880627.0,
      "step": 78
    },
    {
      "epoch": 1.3859649122807016,
      "grad_norm": 0.14428604329114683,
      "learning_rate": 6.393657099905854e-06,
      "loss": 0.2,
      "num_tokens": 33298268.0,
      "step": 79
    },
    {
      "epoch": 1.4035087719298245,
      "grad_norm": 0.14115165460163148,
      "learning_rate": 6.309526916659843e-06,
      "loss": 0.1961,
      "num_tokens": 33735278.0,
      "step": 80
    },
    {
      "epoch": 1.4210526315789473,
      "grad_norm": 0.14007382911380342,
      "learning_rate": 6.225103272658889e-06,
      "loss": 0.199,
      "num_tokens": 34180174.0,
      "step": 81
    },
    {
      "epoch": 1.4385964912280702,
      "grad_norm": 0.1464893959125568,
      "learning_rate": 6.140416772229785e-06,
      "loss": 0.1996,
      "num_tokens": 34600325.0,
      "step": 82
    },
    {
      "epoch": 1.456140350877193,
      "grad_norm": 0.14708924135818197,
      "learning_rate": 6.0554981149871276e-06,
      "loss": 0.1978,
      "num_tokens": 35028867.0,
      "step": 83
    },
    {
      "epoch": 1.4736842105263157,
      "grad_norm": 0.15137327410210355,
      "learning_rate": 5.970378084704441e-06,
      "loss": 0.1971,
      "num_tokens": 35465932.0,
      "step": 84
    },
    {
      "epoch": 1.4912280701754386,
      "grad_norm": 0.14187565136169733,
      "learning_rate": 5.88508753815478e-06,
      "loss": 0.1922,
      "num_tokens": 35857083.0,
      "step": 85
    },
    {
      "epoch": 1.5087719298245614,
      "grad_norm": 0.14750567085422958,
      "learning_rate": 5.799657393924869e-06,
      "loss": 0.1886,
      "num_tokens": 36243816.0,
      "step": 86
    },
    {
      "epoch": 1.526315789473684,
      "grad_norm": 0.13584876516210329,
      "learning_rate": 5.714118621206843e-06,
      "loss": 0.1949,
      "num_tokens": 36682515.0,
      "step": 87
    },
    {
      "epoch": 1.543859649122807,
      "grad_norm": 0.13138592337104565,
      "learning_rate": 5.6285022285716325e-06,
      "loss": 0.1848,
      "num_tokens": 37108482.0,
      "step": 88
    },
    {
      "epoch": 1.5614035087719298,
      "grad_norm": 0.14022318239136972,
      "learning_rate": 5.542839252728096e-06,
      "loss": 0.1986,
      "num_tokens": 37546779.0,
      "step": 89
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 0.14886854943483496,
      "learning_rate": 5.457160747271906e-06,
      "loss": 0.2025,
      "num_tokens": 37965357.0,
      "step": 90
    },
    {
      "epoch": 1.5964912280701755,
      "grad_norm": 0.14038303393485826,
      "learning_rate": 5.371497771428368e-06,
      "loss": 0.1975,
      "num_tokens": 38406113.0,
      "step": 91
    },
    {
      "epoch": 1.6140350877192984,
      "grad_norm": 0.14458491708947513,
      "learning_rate": 5.2858813787931605e-06,
      "loss": 0.1872,
      "num_tokens": 38800413.0,
      "step": 92
    },
    {
      "epoch": 1.631578947368421,
      "grad_norm": 0.1480350806263145,
      "learning_rate": 5.2003426060751324e-06,
      "loss": 0.1968,
      "num_tokens": 39200782.0,
      "step": 93
    },
    {
      "epoch": 1.6491228070175439,
      "grad_norm": 0.13730868120439885,
      "learning_rate": 5.114912461845223e-06,
      "loss": 0.2098,
      "num_tokens": 39666940.0,
      "step": 94
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.13357583490843425,
      "learning_rate": 5.02962191529556e-06,
      "loss": 0.2005,
      "num_tokens": 40103554.0,
      "step": 95
    },
    {
      "epoch": 1.6842105263157894,
      "grad_norm": 0.13277577033293314,
      "learning_rate": 4.944501885012875e-06,
      "loss": 0.1894,
      "num_tokens": 40523764.0,
      "step": 96
    },
    {
      "epoch": 1.7017543859649122,
      "grad_norm": 0.1445531104714593,
      "learning_rate": 4.859583227770218e-06,
      "loss": 0.1966,
      "num_tokens": 40923093.0,
      "step": 97
    },
    {
      "epoch": 1.719298245614035,
      "grad_norm": 0.1346892995124459,
      "learning_rate": 4.774896727341113e-06,
      "loss": 0.2085,
      "num_tokens": 41388977.0,
      "step": 98
    },
    {
      "epoch": 1.736842105263158,
      "grad_norm": 0.13950660376280916,
      "learning_rate": 4.6904730833401575e-06,
      "loss": 0.1993,
      "num_tokens": 41804049.0,
      "step": 99
    },
    {
      "epoch": 1.7543859649122808,
      "grad_norm": 0.13419027046897863,
      "learning_rate": 4.606342900094147e-06,
      "loss": 0.197,
      "num_tokens": 42212384.0,
      "step": 100
    },
    {
      "epoch": 1.7719298245614035,
      "grad_norm": 0.13471704895191902,
      "learning_rate": 4.5225366755477165e-06,
      "loss": 0.1991,
      "num_tokens": 42626890.0,
      "step": 101
    },
    {
      "epoch": 1.7894736842105263,
      "grad_norm": 0.1370431857295371,
      "learning_rate": 4.439084790207577e-06,
      "loss": 0.181,
      "num_tokens": 43045185.0,
      "step": 102
    },
    {
      "epoch": 1.807017543859649,
      "grad_norm": 0.13585245130540036,
      "learning_rate": 4.35601749612931e-06,
      "loss": 0.1926,
      "num_tokens": 43469318.0,
      "step": 103
    },
    {
      "epoch": 1.8245614035087718,
      "grad_norm": 0.14177193142263728,
      "learning_rate": 4.273364905950711e-06,
      "loss": 0.1883,
      "num_tokens": 43860735.0,
      "step": 104
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 0.13934124481189936,
      "learning_rate": 4.191156981975704e-06,
      "loss": 0.186,
      "num_tokens": 44259177.0,
      "step": 105
    },
    {
      "epoch": 1.8596491228070176,
      "grad_norm": 0.13757581593440663,
      "learning_rate": 4.109423525312738e-06,
      "loss": 0.1813,
      "num_tokens": 44652826.0,
      "step": 106
    },
    {
      "epoch": 1.8771929824561404,
      "grad_norm": 0.12994085331633362,
      "learning_rate": 4.028194165071603e-06,
      "loss": 0.2007,
      "num_tokens": 45110456.0,
      "step": 107
    },
    {
      "epoch": 1.8947368421052633,
      "grad_norm": 0.1315821088883057,
      "learning_rate": 3.9474983476226335e-06,
      "loss": 0.1984,
      "num_tokens": 45561128.0,
      "step": 108
    },
    {
      "epoch": 1.912280701754386,
      "grad_norm": 0.13466926193354203,
      "learning_rate": 3.867365325922116e-06,
      "loss": 0.195,
      "num_tokens": 45999442.0,
      "step": 109
    },
    {
      "epoch": 1.9298245614035088,
      "grad_norm": 0.1407044279595099,
      "learning_rate": 3.7878241489078473e-06,
      "loss": 0.1957,
      "num_tokens": 46405471.0,
      "step": 110
    },
    {
      "epoch": 1.9473684210526314,
      "grad_norm": 0.13869286010098864,
      "learning_rate": 3.7089036509686216e-06,
      "loss": 0.2096,
      "num_tokens": 46837446.0,
      "step": 111
    },
    {
      "epoch": 1.9649122807017543,
      "grad_norm": 0.13744222256799704,
      "learning_rate": 3.630632441491512e-06,
      "loss": 0.1874,
      "num_tokens": 47238683.0,
      "step": 112
    },
    {
      "epoch": 1.9824561403508771,
      "grad_norm": 0.13584540109090137,
      "learning_rate": 3.5530388944907124e-06,
      "loss": 0.1944,
      "num_tokens": 47661310.0,
      "step": 113
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.14078927259649465,
      "learning_rate": 3.476151138321705e-06,
      "loss": 0.1893,
      "num_tokens": 48079346.0,
      "step": 114
    },
    {
      "epoch": 2.017543859649123,
      "grad_norm": 0.14801647762626233,
      "learning_rate": 3.3999970454844688e-06,
      "loss": 0.1833,
      "num_tokens": 48481444.0,
      "step": 115
    },
    {
      "epoch": 2.0350877192982457,
      "grad_norm": 0.14207211024797958,
      "learning_rate": 3.3246042225194626e-06,
      "loss": 0.1972,
      "num_tokens": 48904379.0,
      "step": 116
    },
    {
      "epoch": 2.0526315789473686,
      "grad_norm": 0.13459996651406178,
      "learning_rate": 3.2500000000000015e-06,
      "loss": 0.1734,
      "num_tokens": 49314280.0,
      "step": 117
    },
    {
      "epoch": 2.0701754385964914,
      "grad_norm": 0.13420885166445595,
      "learning_rate": 3.176211422624672e-06,
      "loss": 0.1748,
      "num_tokens": 49720807.0,
      "step": 118
    },
    {
      "epoch": 2.087719298245614,
      "grad_norm": 0.1396854400821284,
      "learning_rate": 3.103265239413401e-06,
      "loss": 0.1781,
      "num_tokens": 50151950.0,
      "step": 119
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 0.14321980578287785,
      "learning_rate": 3.0311878940106864e-06,
      "loss": 0.182,
      "num_tokens": 50574817.0,
      "step": 120
    },
    {
      "epoch": 2.1228070175438596,
      "grad_norm": 0.1402669344348115,
      "learning_rate": 2.9600055150995397e-06,
      "loss": 0.1804,
      "num_tokens": 50991178.0,
      "step": 121
    },
    {
      "epoch": 2.1403508771929824,
      "grad_norm": 0.14268249286817555,
      "learning_rate": 2.889743906929609e-06,
      "loss": 0.1701,
      "num_tokens": 51370437.0,
      "step": 122
    },
    {
      "epoch": 2.1578947368421053,
      "grad_norm": 0.13769468594260603,
      "learning_rate": 2.820428539962905e-06,
      "loss": 0.1803,
      "num_tokens": 51807382.0,
      "step": 123
    },
    {
      "epoch": 2.175438596491228,
      "grad_norm": 0.1418977810735795,
      "learning_rate": 2.7520845416405285e-06,
      "loss": 0.1867,
      "num_tokens": 52214420.0,
      "step": 124
    },
    {
      "epoch": 2.192982456140351,
      "grad_norm": 0.1357482680971467,
      "learning_rate": 2.6847366872737535e-06,
      "loss": 0.1855,
      "num_tokens": 52648228.0,
      "step": 125
    },
    {
      "epoch": 2.2105263157894735,
      "grad_norm": 0.13758984714319883,
      "learning_rate": 2.618409391062751e-06,
      "loss": 0.1928,
      "num_tokens": 53085345.0,
      "step": 126
    },
    {
      "epoch": 2.2280701754385963,
      "grad_norm": 0.13835249455110674,
      "learning_rate": 2.5531266972462176e-06,
      "loss": 0.1753,
      "num_tokens": 53493140.0,
      "step": 127
    },
    {
      "epoch": 2.245614035087719,
      "grad_norm": 0.13785872746102823,
      "learning_rate": 2.4889122713851397e-06,
      "loss": 0.1734,
      "num_tokens": 53922832.0,
      "step": 128
    },
    {
      "epoch": 2.263157894736842,
      "grad_norm": 0.1369141853241696,
      "learning_rate": 2.425789391783796e-06,
      "loss": 0.1771,
      "num_tokens": 54319035.0,
      "step": 129
    },
    {
      "epoch": 2.280701754385965,
      "grad_norm": 0.13274261857238875,
      "learning_rate": 2.36378094105118e-06,
      "loss": 0.1759,
      "num_tokens": 54754161.0,
      "step": 130
    },
    {
      "epoch": 2.2982456140350878,
      "grad_norm": 0.13374187150373806,
      "learning_rate": 2.302909397805841e-06,
      "loss": 0.1757,
      "num_tokens": 55177972.0,
      "step": 131
    },
    {
      "epoch": 2.3157894736842106,
      "grad_norm": 0.14056010545232225,
      "learning_rate": 2.2431968285271843e-06,
      "loss": 0.1762,
      "num_tokens": 55567532.0,
      "step": 132
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.14440093167538778,
      "learning_rate": 2.1846648795561777e-06,
      "loss": 0.1789,
      "num_tokens": 55997257.0,
      "step": 133
    },
    {
      "epoch": 2.3508771929824563,
      "grad_norm": 0.13209921359488944,
      "learning_rate": 2.1273347692483574e-06,
      "loss": 0.1683,
      "num_tokens": 56435780.0,
      "step": 134
    },
    {
      "epoch": 2.3684210526315788,
      "grad_norm": 0.13075669290172312,
      "learning_rate": 2.071227280281982e-06,
      "loss": 0.1766,
      "num_tokens": 56880449.0,
      "step": 135
    },
    {
      "epoch": 2.3859649122807016,
      "grad_norm": 0.15520159951824955,
      "learning_rate": 2.016362752124129e-06,
      "loss": 0.1766,
      "num_tokens": 57298925.0,
      "step": 136
    },
    {
      "epoch": 2.4035087719298245,
      "grad_norm": 0.13674994090087955,
      "learning_rate": 1.9627610736574575e-06,
      "loss": 0.1689,
      "num_tokens": 57702294.0,
      "step": 137
    },
    {
      "epoch": 2.4210526315789473,
      "grad_norm": 0.13321588166669115,
      "learning_rate": 1.9104416759703017e-06,
      "loss": 0.1758,
      "num_tokens": 58154418.0,
      "step": 138
    },
    {
      "epoch": 2.43859649122807,
      "grad_norm": 0.13632315461262803,
      "learning_rate": 1.8594235253127373e-06,
      "loss": 0.1767,
      "num_tokens": 58579683.0,
      "step": 139
    },
    {
      "epoch": 2.456140350877193,
      "grad_norm": 0.1288153252214618,
      "learning_rate": 1.8097251162211405e-06,
      "loss": 0.1811,
      "num_tokens": 59051562.0,
      "step": 140
    },
    {
      "epoch": 2.473684210526316,
      "grad_norm": 0.12987465966750822,
      "learning_rate": 1.7613644648137543e-06,
      "loss": 0.1774,
      "num_tokens": 59501698.0,
      "step": 141
    },
    {
      "epoch": 2.4912280701754383,
      "grad_norm": 0.1374425426257984,
      "learning_rate": 1.7143591022596846e-06,
      "loss": 0.1944,
      "num_tokens": 59955592.0,
      "step": 142
    },
    {
      "epoch": 2.5087719298245617,
      "grad_norm": 0.13531966438989285,
      "learning_rate": 1.6687260684236943e-06,
      "loss": 0.1805,
      "num_tokens": 60381274.0,
      "step": 143
    },
    {
      "epoch": 2.526315789473684,
      "grad_norm": 0.134925096600037,
      "learning_rate": 1.6244819056890975e-06,
      "loss": 0.171,
      "num_tokens": 60788438.0,
      "step": 144
    },
    {
      "epoch": 2.543859649122807,
      "grad_norm": 0.1407719819848176,
      "learning_rate": 1.5816426529610035e-06,
      "loss": 0.1714,
      "num_tokens": 61184551.0,
      "step": 145
    },
    {
      "epoch": 2.56140350877193,
      "grad_norm": 0.1359063010745857,
      "learning_rate": 1.5402238398520614e-06,
      "loss": 0.1758,
      "num_tokens": 61619839.0,
      "step": 146
    },
    {
      "epoch": 2.5789473684210527,
      "grad_norm": 0.13246835060504009,
      "learning_rate": 1.5002404810528452e-06,
      "loss": 0.1739,
      "num_tokens": 62045889.0,
      "step": 147
    },
    {
      "epoch": 2.5964912280701755,
      "grad_norm": 0.13964950522945552,
      "learning_rate": 1.4617070708888882e-06,
      "loss": 0.1718,
      "num_tokens": 62437900.0,
      "step": 148
    },
    {
      "epoch": 2.6140350877192984,
      "grad_norm": 0.13372294299780066,
      "learning_rate": 1.4246375780663613e-06,
      "loss": 0.1661,
      "num_tokens": 62843119.0,
      "step": 149
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 0.13805400712254826,
      "learning_rate": 1.389045440608296e-06,
      "loss": 0.1828,
      "num_tokens": 63262284.0,
      "step": 150
    },
    {
      "epoch": 2.6491228070175437,
      "grad_norm": 0.13808339180910903,
      "learning_rate": 1.354943560983175e-06,
      "loss": 0.1803,
      "num_tokens": 63701914.0,
      "step": 151
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.13165183261262675,
      "learning_rate": 1.3223443014276738e-06,
      "loss": 0.1774,
      "num_tokens": 64149759.0,
      "step": 152
    },
    {
      "epoch": 2.6842105263157894,
      "grad_norm": 0.1319243140393186,
      "learning_rate": 1.2912594794652406e-06,
      "loss": 0.1799,
      "num_tokens": 64602514.0,
      "step": 153
    },
    {
      "epoch": 2.7017543859649122,
      "grad_norm": 0.13538494698388223,
      "learning_rate": 1.2617003636221394e-06,
      "loss": 0.1694,
      "num_tokens": 64992832.0,
      "step": 154
    },
    {
      "epoch": 2.719298245614035,
      "grad_norm": 0.13498616412912304,
      "learning_rate": 1.2336776693425028e-06,
      "loss": 0.1707,
      "num_tokens": 65389001.0,
      "step": 155
    },
    {
      "epoch": 2.736842105263158,
      "grad_norm": 0.12924445948277097,
      "learning_rate": 1.2072015551038933e-06,
      "loss": 0.1692,
      "num_tokens": 65842297.0,
      "step": 156
    },
    {
      "epoch": 2.754385964912281,
      "grad_norm": 0.13206490698518994,
      "learning_rate": 1.1822816187347625e-06,
      "loss": 0.1719,
      "num_tokens": 66254756.0,
      "step": 157
    },
    {
      "epoch": 2.7719298245614032,
      "grad_norm": 0.13946311479316087,
      "learning_rate": 1.1589268939351499e-06,
      "loss": 0.1824,
      "num_tokens": 66653349.0,
      "step": 158
    },
    {
      "epoch": 2.7894736842105265,
      "grad_norm": 0.12958606648427337,
      "learning_rate": 1.1371458470018896e-06,
      "loss": 0.1758,
      "num_tokens": 67089072.0,
      "step": 159
    },
    {
      "epoch": 2.807017543859649,
      "grad_norm": 0.12836201287391627,
      "learning_rate": 1.1169463737594995e-06,
      "loss": 0.1725,
      "num_tokens": 67530318.0,
      "step": 160
    },
    {
      "epoch": 2.824561403508772,
      "grad_norm": 0.1312877766550337,
      "learning_rate": 1.0983357966978747e-06,
      "loss": 0.17,
      "num_tokens": 67943670.0,
      "step": 161
    },
    {
      "epoch": 2.8421052631578947,
      "grad_norm": 0.1292736067143446,
      "learning_rate": 1.0813208623178199e-06,
      "loss": 0.1759,
      "num_tokens": 68380170.0,
      "step": 162
    },
    {
      "epoch": 2.8596491228070176,
      "grad_norm": 0.13227684289793154,
      "learning_rate": 1.0659077386853817e-06,
      "loss": 0.1719,
      "num_tokens": 68808114.0,
      "step": 163
    },
    {
      "epoch": 2.8771929824561404,
      "grad_norm": 0.13694616326909123,
      "learning_rate": 1.0521020131958692e-06,
      "loss": 0.168,
      "num_tokens": 69191512.0,
      "step": 164
    },
    {
      "epoch": 2.8947368421052633,
      "grad_norm": 0.135480809490847,
      "learning_rate": 1.0399086905483752e-06,
      "loss": 0.1659,
      "num_tokens": 69582710.0,
      "step": 165
    },
    {
      "epoch": 2.912280701754386,
      "grad_norm": 0.13656650923592858,
      "learning_rate": 1.0293321909315242e-06,
      "loss": 0.1764,
      "num_tokens": 69995390.0,
      "step": 166
    },
    {
      "epoch": 2.9298245614035086,
      "grad_norm": 0.13037307085567856,
      "learning_rate": 1.0203763484211196e-06,
      "loss": 0.1737,
      "num_tokens": 70418613.0,
      "step": 167
    },
    {
      "epoch": 2.9473684210526314,
      "grad_norm": 0.13386161162103719,
      "learning_rate": 1.0130444095902514e-06,
      "loss": 0.1731,
      "num_tokens": 70848952.0,
      "step": 168
    },
    {
      "epoch": 2.9649122807017543,
      "grad_norm": 0.13477603098195323,
      "learning_rate": 1.0073390323323897e-06,
      "loss": 0.1859,
      "num_tokens": 71284648.0,
      "step": 169
    },
    {
      "epoch": 2.982456140350877,
      "grad_norm": 0.13335006132129937,
      "learning_rate": 1.0032622848978689e-06,
      "loss": 0.1727,
      "num_tokens": 71712021.0,
      "step": 170
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.13952787638079295,
      "learning_rate": 1.000815645144134e-06,
      "loss": 0.1803,
      "num_tokens": 72119019.0,
      "step": 171
    },
    {
      "epoch": 3.0,
      "step": 171,
      "total_flos": 2.3168881532705178e+17,
      "train_loss": 0.25350178002614027,
      "train_runtime": 2712.8956,
      "train_samples_per_second": 8.068,
      "train_steps_per_second": 0.063
    }
  ],
  "logging_steps": 1,
  "max_steps": 171,
  "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": 2.3168881532705178e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}