{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9872340425531916,
  "eval_steps": 500,
  "global_step": 234,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01276595744680851,
      "grad_norm": 7.015718466989074,
      "learning_rate": 4.1666666666666667e-07,
      "loss": 1.117,
      "step": 1
    },
    {
      "epoch": 0.02553191489361702,
      "grad_norm": 6.94762834330754,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.1134,
      "step": 2
    },
    {
      "epoch": 0.03829787234042553,
      "grad_norm": 6.801049953756127,
      "learning_rate": 1.25e-06,
      "loss": 1.1129,
      "step": 3
    },
    {
      "epoch": 0.05106382978723404,
      "grad_norm": 6.772438165340198,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.0998,
      "step": 4
    },
    {
      "epoch": 0.06382978723404255,
      "grad_norm": 6.296447609181628,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 1.079,
      "step": 5
    },
    {
      "epoch": 0.07659574468085106,
      "grad_norm": 5.215909882705263,
      "learning_rate": 2.5e-06,
      "loss": 1.0326,
      "step": 6
    },
    {
      "epoch": 0.08936170212765958,
      "grad_norm": 5.0240627402893425,
      "learning_rate": 2.916666666666667e-06,
      "loss": 1.0168,
      "step": 7
    },
    {
      "epoch": 0.10212765957446808,
      "grad_norm": 3.125198769881761,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0226,
      "step": 8
    },
    {
      "epoch": 0.1148936170212766,
      "grad_norm": 2.8348625516600063,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.9845,
      "step": 9
    },
    {
      "epoch": 0.1276595744680851,
      "grad_norm": 2.5539753733700277,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9905,
      "step": 10
    },
    {
      "epoch": 0.14042553191489363,
      "grad_norm": 4.74032301961336,
      "learning_rate": 4.583333333333333e-06,
      "loss": 0.9639,
      "step": 11
    },
    {
      "epoch": 0.15319148936170213,
      "grad_norm": 4.86790394934032,
      "learning_rate": 5e-06,
      "loss": 0.9165,
      "step": 12
    },
    {
      "epoch": 0.16595744680851063,
      "grad_norm": 5.066912804753986,
      "learning_rate": 5.416666666666667e-06,
      "loss": 1.0023,
      "step": 13
    },
    {
      "epoch": 0.17872340425531916,
      "grad_norm": 4.08224163704665,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9128,
      "step": 14
    },
    {
      "epoch": 0.19148936170212766,
      "grad_norm": 4.091250067697973,
      "learning_rate": 6.25e-06,
      "loss": 0.8716,
      "step": 15
    },
    {
      "epoch": 0.20425531914893616,
      "grad_norm": 3.4732433394882944,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.8829,
      "step": 16
    },
    {
      "epoch": 0.2170212765957447,
      "grad_norm": 2.5197296768780895,
      "learning_rate": 7.083333333333335e-06,
      "loss": 0.8787,
      "step": 17
    },
    {
      "epoch": 0.2297872340425532,
      "grad_norm": 2.125408786161704,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.8183,
      "step": 18
    },
    {
      "epoch": 0.2425531914893617,
      "grad_norm": 2.2748157897060506,
      "learning_rate": 7.916666666666667e-06,
      "loss": 0.8937,
      "step": 19
    },
    {
      "epoch": 0.2553191489361702,
      "grad_norm": 2.333974324631764,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.8386,
      "step": 20
    },
    {
      "epoch": 0.2680851063829787,
      "grad_norm": 1.8128182674638877,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.8552,
      "step": 21
    },
    {
      "epoch": 0.28085106382978725,
      "grad_norm": 1.7125731325584599,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8767,
      "step": 22
    },
    {
      "epoch": 0.2936170212765957,
      "grad_norm": 1.683343575929496,
      "learning_rate": 9.583333333333335e-06,
      "loss": 0.8357,
      "step": 23
    },
    {
      "epoch": 0.30638297872340425,
      "grad_norm": 1.373668760486191,
      "learning_rate": 1e-05,
      "loss": 0.7915,
      "step": 24
    },
    {
      "epoch": 0.3191489361702128,
      "grad_norm": 1.0055571221242297,
      "learning_rate": 9.999440509051367e-06,
      "loss": 0.7785,
      "step": 25
    },
    {
      "epoch": 0.33191489361702126,
      "grad_norm": 0.9027056985862393,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.8068,
      "step": 26
    },
    {
      "epoch": 0.3446808510638298,
      "grad_norm": 1.2667053539496995,
      "learning_rate": 9.994965332706574e-06,
      "loss": 0.8507,
      "step": 27
    },
    {
      "epoch": 0.3574468085106383,
      "grad_norm": 0.9917915369326918,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.8097,
      "step": 28
    },
    {
      "epoch": 0.3702127659574468,
      "grad_norm": 0.8527855285775848,
      "learning_rate": 9.986018985905901e-06,
      "loss": 0.795,
      "step": 29
    },
    {
      "epoch": 0.3829787234042553,
      "grad_norm": 0.838806489476804,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.7381,
      "step": 30
    },
    {
      "epoch": 0.39574468085106385,
      "grad_norm": 1.0269042283041614,
      "learning_rate": 9.972609476841368e-06,
      "loss": 0.8167,
      "step": 31
    },
    {
      "epoch": 0.4085106382978723,
      "grad_norm": 0.9178642201492891,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.7563,
      "step": 32
    },
    {
      "epoch": 0.42127659574468085,
      "grad_norm": 0.7926111862163232,
      "learning_rate": 9.954748808839675e-06,
      "loss": 0.7787,
      "step": 33
    },
    {
      "epoch": 0.4340425531914894,
      "grad_norm": 0.75573330787621,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.78,
      "step": 34
    },
    {
      "epoch": 0.44680851063829785,
      "grad_norm": 0.815145181705661,
      "learning_rate": 9.932452969617607e-06,
      "loss": 0.7871,
      "step": 35
    },
    {
      "epoch": 0.4595744680851064,
      "grad_norm": 0.9752539663188574,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.7976,
      "step": 36
    },
    {
      "epoch": 0.4723404255319149,
      "grad_norm": 0.6959577877909565,
      "learning_rate": 9.905741916970863e-06,
      "loss": 0.7809,
      "step": 37
    },
    {
      "epoch": 0.4851063829787234,
      "grad_norm": 0.8883024852628244,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.7876,
      "step": 38
    },
    {
      "epoch": 0.4978723404255319,
      "grad_norm": 0.773663747644755,
      "learning_rate": 9.874639560909118e-06,
      "loss": 0.7954,
      "step": 39
    },
    {
      "epoch": 0.5106382978723404,
      "grad_norm": 0.6398699968223347,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.766,
      "step": 40
    },
    {
      "epoch": 0.5234042553191489,
      "grad_norm": 0.746186258074948,
      "learning_rate": 9.839173742253334e-06,
      "loss": 0.8023,
      "step": 41
    },
    {
      "epoch": 0.5361702127659574,
      "grad_norm": 0.621705451607282,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.7787,
      "step": 42
    },
    {
      "epoch": 0.548936170212766,
      "grad_norm": 0.6165255600989523,
      "learning_rate": 9.799376207714446e-06,
      "loss": 0.7676,
      "step": 43
    },
    {
      "epoch": 0.5617021276595745,
      "grad_norm": 0.6336316514181005,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.7733,
      "step": 44
    },
    {
      "epoch": 0.574468085106383,
      "grad_norm": 0.5891614649703093,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.7726,
      "step": 45
    },
    {
      "epoch": 0.5872340425531914,
      "grad_norm": 0.5953555831423718,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.7886,
      "step": 46
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.6255442079564976,
      "learning_rate": 9.706932333304518e-06,
      "loss": 0.7776,
      "step": 47
    },
    {
      "epoch": 0.6127659574468085,
      "grad_norm": 0.5944241623820739,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.8059,
      "step": 48
    },
    {
      "epoch": 0.625531914893617,
      "grad_norm": 0.5568557096343427,
      "learning_rate": 9.654368743221022e-06,
      "loss": 0.7354,
      "step": 49
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 0.6038601495314921,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.7841,
      "step": 50
    },
    {
      "epoch": 0.6510638297872341,
      "grad_norm": 0.5270561177136355,
      "learning_rate": 9.597638862757255e-06,
      "loss": 0.7536,
      "step": 51
    },
    {
      "epoch": 0.6638297872340425,
      "grad_norm": 0.5844453802384513,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.7836,
      "step": 52
    },
    {
      "epoch": 0.676595744680851,
      "grad_norm": 0.6102351937764352,
      "learning_rate": 9.536793472839325e-06,
      "loss": 0.7027,
      "step": 53
    },
    {
      "epoch": 0.6893617021276596,
      "grad_norm": 0.6035603241490802,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.7938,
      "step": 54
    },
    {
      "epoch": 0.7021276595744681,
      "grad_norm": 0.6261987250094778,
      "learning_rate": 9.471887038331686e-06,
      "loss": 0.8058,
      "step": 55
    },
    {
      "epoch": 0.7148936170212766,
      "grad_norm": 0.5496448961935777,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.7234,
      "step": 56
    },
    {
      "epoch": 0.7276595744680852,
      "grad_norm": 0.6005800193481035,
      "learning_rate": 9.40297765928369e-06,
      "loss": 0.7323,
      "step": 57
    },
    {
      "epoch": 0.7404255319148936,
      "grad_norm": 0.9150485257772174,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.7737,
      "step": 58
    },
    {
      "epoch": 0.7531914893617021,
      "grad_norm": 0.5971646523259739,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.713,
      "step": 59
    },
    {
      "epoch": 0.7659574468085106,
      "grad_norm": 0.6498404792169774,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.7276,
      "step": 60
    },
    {
      "epoch": 0.7787234042553192,
      "grad_norm": 0.8683199966188363,
      "learning_rate": 9.253400328436699e-06,
      "loss": 0.7949,
      "step": 61
    },
    {
      "epoch": 0.7914893617021277,
      "grad_norm": 0.6642758991694608,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.8019,
      "step": 62
    },
    {
      "epoch": 0.8042553191489362,
      "grad_norm": 0.6410474363247081,
      "learning_rate": 9.172866268606514e-06,
      "loss": 0.7602,
      "step": 63
    },
    {
      "epoch": 0.8170212765957446,
      "grad_norm": 0.5904684410337478,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.7189,
      "step": 64
    },
    {
      "epoch": 0.8297872340425532,
      "grad_norm": 0.6709679491441708,
      "learning_rate": 9.088596928322158e-06,
      "loss": 0.8073,
      "step": 65
    },
    {
      "epoch": 0.8425531914893617,
      "grad_norm": 0.556019821773472,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.7782,
      "step": 66
    },
    {
      "epoch": 0.8553191489361702,
      "grad_norm": 0.5882607467493168,
      "learning_rate": 9.000667740056033e-06,
      "loss": 0.7552,
      "step": 67
    },
    {
      "epoch": 0.8680851063829788,
      "grad_norm": 0.6128020803055951,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.7859,
      "step": 68
    },
    {
      "epoch": 0.8808510638297873,
      "grad_norm": 0.5441564996823284,
      "learning_rate": 8.90915741234015e-06,
      "loss": 0.8118,
      "step": 69
    },
    {
      "epoch": 0.8936170212765957,
      "grad_norm": 0.5071072554693752,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.7541,
      "step": 70
    },
    {
      "epoch": 0.9063829787234042,
      "grad_norm": 0.526853686583371,
      "learning_rate": 8.814147859311333e-06,
      "loss": 0.7476,
      "step": 71
    },
    {
      "epoch": 0.9191489361702128,
      "grad_norm": 0.565655168554739,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.7464,
      "step": 72
    },
    {
      "epoch": 0.9319148936170213,
      "grad_norm": 0.5679672017867912,
      "learning_rate": 8.715724127386971e-06,
      "loss": 0.8133,
      "step": 73
    },
    {
      "epoch": 0.9446808510638298,
      "grad_norm": 0.5492621036178154,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.8062,
      "step": 74
    },
    {
      "epoch": 0.9574468085106383,
      "grad_norm": 0.5463633975363426,
      "learning_rate": 8.613974319136959e-06,
      "loss": 0.79,
      "step": 75
    },
    {
      "epoch": 0.9702127659574468,
      "grad_norm": 0.5714899949507461,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.7951,
      "step": 76
    },
    {
      "epoch": 0.9829787234042553,
      "grad_norm": 0.5754994749263996,
      "learning_rate": 8.508989514419959e-06,
      "loss": 0.7233,
      "step": 77
    },
    {
      "epoch": 0.9957446808510638,
      "grad_norm": 0.5760269816494827,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.7733,
      "step": 78
    },
    {
      "epoch": 1.0085106382978724,
      "grad_norm": 1.0945888585918737,
      "learning_rate": 8.400863688854598e-06,
      "loss": 1.1355,
      "step": 79
    },
    {
      "epoch": 1.0212765957446808,
      "grad_norm": 0.5489153017138312,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.6646,
      "step": 80
    },
    {
      "epoch": 1.0340425531914894,
      "grad_norm": 0.550042864839408,
      "learning_rate": 8.289693629698564e-06,
      "loss": 0.7469,
      "step": 81
    },
    {
      "epoch": 1.0468085106382978,
      "grad_norm": 0.549804299148006,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.7351,
      "step": 82
    },
    {
      "epoch": 1.0595744680851065,
      "grad_norm": 0.46781275812787854,
      "learning_rate": 8.175578849210894e-06,
      "loss": 0.6464,
      "step": 83
    },
    {
      "epoch": 1.0723404255319149,
      "grad_norm": 0.6076923590979096,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.7995,
      "step": 84
    },
    {
      "epoch": 1.0851063829787233,
      "grad_norm": 0.5641940949871667,
      "learning_rate": 8.058621495575032e-06,
      "loss": 0.7216,
      "step": 85
    },
    {
      "epoch": 1.097872340425532,
      "grad_norm": 0.5993520107937353,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.7149,
      "step": 86
    },
    {
      "epoch": 1.1106382978723404,
      "grad_norm": 0.567214363627709,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.658,
      "step": 87
    },
    {
      "epoch": 1.123404255319149,
      "grad_norm": 0.5349806750291352,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.649,
      "step": 88
    },
    {
      "epoch": 1.1361702127659574,
      "grad_norm": 0.703875833332789,
      "learning_rate": 7.81660029031811e-06,
      "loss": 0.8051,
      "step": 89
    },
    {
      "epoch": 1.148936170212766,
      "grad_norm": 0.6291964370143157,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.698,
      "step": 90
    },
    {
      "epoch": 1.1617021276595745,
      "grad_norm": 0.4666655295099628,
      "learning_rate": 7.691753080453413e-06,
      "loss": 0.6655,
      "step": 91
    },
    {
      "epoch": 1.174468085106383,
      "grad_norm": 0.5368319006956255,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.7435,
      "step": 92
    },
    {
      "epoch": 1.1872340425531915,
      "grad_norm": 0.5153563571772463,
      "learning_rate": 7.564496387029532e-06,
      "loss": 0.7504,
      "step": 93
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.5466125166103902,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.7341,
      "step": 94
    },
    {
      "epoch": 1.2127659574468086,
      "grad_norm": 0.5224826643535506,
      "learning_rate": 7.434944122021837e-06,
      "loss": 0.7498,
      "step": 95
    },
    {
      "epoch": 1.225531914893617,
      "grad_norm": 0.5708976206267316,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.708,
      "step": 96
    },
    {
      "epoch": 1.2382978723404254,
      "grad_norm": 0.563262494604646,
      "learning_rate": 7.303212252253163e-06,
      "loss": 0.7551,
      "step": 97
    },
    {
      "epoch": 1.251063829787234,
      "grad_norm": 0.5511363211697056,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.7711,
      "step": 98
    },
    {
      "epoch": 1.2638297872340425,
      "grad_norm": 0.4874825499668437,
      "learning_rate": 7.169418695587791e-06,
      "loss": 0.6459,
      "step": 99
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 0.465488823301378,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.6609,
      "step": 100
    },
    {
      "epoch": 1.2893617021276595,
      "grad_norm": 0.5900785967913017,
      "learning_rate": 7.033683215379002e-06,
      "loss": 0.7567,
      "step": 101
    },
    {
      "epoch": 1.302127659574468,
      "grad_norm": 0.5150565984498293,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.6895,
      "step": 102
    },
    {
      "epoch": 1.3148936170212766,
      "grad_norm": 0.5415463201111466,
      "learning_rate": 6.896127313264643e-06,
      "loss": 0.6849,
      "step": 103
    },
    {
      "epoch": 1.327659574468085,
      "grad_norm": 0.5828549026140929,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.7744,
      "step": 104
    },
    {
      "epoch": 1.3404255319148937,
      "grad_norm": 0.5705104446754513,
      "learning_rate": 6.7568741204067145e-06,
      "loss": 0.7876,
      "step": 105
    },
    {
      "epoch": 1.353191489361702,
      "grad_norm": 0.5521243149798348,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.624,
      "step": 106
    },
    {
      "epoch": 1.3659574468085105,
      "grad_norm": 0.6413422766450044,
      "learning_rate": 6.616048287272301e-06,
      "loss": 0.7829,
      "step": 107
    },
    {
      "epoch": 1.3787234042553191,
      "grad_norm": 0.46827552326425964,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.6761,
      "step": 108
    },
    {
      "epoch": 1.3914893617021278,
      "grad_norm": 0.6205358424015903,
      "learning_rate": 6.473775872054522e-06,
      "loss": 0.7192,
      "step": 109
    },
    {
      "epoch": 1.4042553191489362,
      "grad_norm": 0.5005870891529645,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.7401,
      "step": 110
    },
    {
      "epoch": 1.4170212765957446,
      "grad_norm": 0.5009298310214143,
      "learning_rate": 6.330184227833376e-06,
      "loss": 0.6096,
      "step": 111
    },
    {
      "epoch": 1.4297872340425533,
      "grad_norm": 0.600500474185216,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.77,
      "step": 112
    },
    {
      "epoch": 1.4425531914893617,
      "grad_norm": 0.4995998386976486,
      "learning_rate": 6.185401888577488e-06,
      "loss": 0.6739,
      "step": 113
    },
    {
      "epoch": 1.4553191489361703,
      "grad_norm": 0.4798072520496255,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.6769,
      "step": 114
    },
    {
      "epoch": 1.4680851063829787,
      "grad_norm": 0.5176238294523213,
      "learning_rate": 6.039558454088796e-06,
      "loss": 0.8088,
      "step": 115
    },
    {
      "epoch": 1.4808510638297872,
      "grad_norm": 0.4911445924656217,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.6255,
      "step": 116
    },
    {
      "epoch": 1.4936170212765958,
      "grad_norm": 0.5344710529397226,
      "learning_rate": 5.892784473993184e-06,
      "loss": 0.8014,
      "step": 117
    },
    {
      "epoch": 1.5063829787234042,
      "grad_norm": 0.525330620190939,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.6739,
      "step": 118
    },
    {
      "epoch": 1.5191489361702128,
      "grad_norm": 0.5049522069915814,
      "learning_rate": 5.745211330880872e-06,
      "loss": 0.7689,
      "step": 119
    },
    {
      "epoch": 1.5319148936170213,
      "grad_norm": 0.5368024598915486,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.6955,
      "step": 120
    },
    {
      "epoch": 1.5446808510638297,
      "grad_norm": 0.5586168781796989,
      "learning_rate": 5.596971122701221e-06,
      "loss": 0.6957,
      "step": 121
    },
    {
      "epoch": 1.5574468085106383,
      "grad_norm": 0.4948472638820811,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.6908,
      "step": 122
    },
    {
      "epoch": 1.570212765957447,
      "grad_norm": 0.6059945133709344,
      "learning_rate": 5.448196544517168e-06,
      "loss": 0.8481,
      "step": 123
    },
    {
      "epoch": 1.5829787234042554,
      "grad_norm": 0.4870810475990154,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.6862,
      "step": 124
    },
    {
      "epoch": 1.5957446808510638,
      "grad_norm": 0.6318974529790354,
      "learning_rate": 5.299020769725172e-06,
      "loss": 0.6743,
      "step": 125
    },
    {
      "epoch": 1.6085106382978722,
      "grad_norm": 0.665188557566122,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.7473,
      "step": 126
    },
    {
      "epoch": 1.6212765957446809,
      "grad_norm": 0.4922894454426491,
      "learning_rate": 5.1495773308469935e-06,
      "loss": 0.6891,
      "step": 127
    },
    {
      "epoch": 1.6340425531914895,
      "grad_norm": 0.5207672174829043,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.622,
      "step": 128
    },
    {
      "epoch": 1.646808510638298,
      "grad_norm": 0.5266233826579009,
      "learning_rate": 5e-06,
      "loss": 0.7025,
      "step": 129
    },
    {
      "epoch": 1.6595744680851063,
      "grad_norm": 0.5449941486655594,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.6184,
      "step": 130
    },
    {
      "epoch": 1.6723404255319148,
      "grad_norm": 0.6474022684476997,
      "learning_rate": 4.850422669153009e-06,
      "loss": 0.706,
      "step": 131
    },
    {
      "epoch": 1.6851063829787234,
      "grad_norm": 0.49589919470436855,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.7047,
      "step": 132
    },
    {
      "epoch": 1.697872340425532,
      "grad_norm": 0.5426968909626301,
      "learning_rate": 4.700979230274829e-06,
      "loss": 0.7253,
      "step": 133
    },
    {
      "epoch": 1.7106382978723405,
      "grad_norm": 0.60876343138784,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.7898,
      "step": 134
    },
    {
      "epoch": 1.7234042553191489,
      "grad_norm": 0.5027585714953703,
      "learning_rate": 4.551803455482833e-06,
      "loss": 0.691,
      "step": 135
    },
    {
      "epoch": 1.7361702127659573,
      "grad_norm": 0.5526361540526097,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.7306,
      "step": 136
    },
    {
      "epoch": 1.748936170212766,
      "grad_norm": 0.4564058130594461,
      "learning_rate": 4.4030288772987795e-06,
      "loss": 0.677,
      "step": 137
    },
    {
      "epoch": 1.7617021276595746,
      "grad_norm": 0.4265652793878518,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.6141,
      "step": 138
    },
    {
      "epoch": 1.774468085106383,
      "grad_norm": 0.4451324891275968,
      "learning_rate": 4.254788669119127e-06,
      "loss": 0.6652,
      "step": 139
    },
    {
      "epoch": 1.7872340425531914,
      "grad_norm": 0.531036312152885,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.6635,
      "step": 140
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.522235798644409,
      "learning_rate": 4.107215526006818e-06,
      "loss": 0.7722,
      "step": 141
    },
    {
      "epoch": 1.8127659574468085,
      "grad_norm": 0.446699355693039,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.6111,
      "step": 142
    },
    {
      "epoch": 1.825531914893617,
      "grad_norm": 0.4882147681991008,
      "learning_rate": 3.960441545911205e-06,
      "loss": 0.7362,
      "step": 143
    },
    {
      "epoch": 1.8382978723404255,
      "grad_norm": 0.5436602159483871,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.7662,
      "step": 144
    },
    {
      "epoch": 1.851063829787234,
      "grad_norm": 0.4097709684721711,
      "learning_rate": 3.8145981114225135e-06,
      "loss": 0.627,
      "step": 145
    },
    {
      "epoch": 1.8638297872340426,
      "grad_norm": 0.5774202757015185,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.8162,
      "step": 146
    },
    {
      "epoch": 1.8765957446808512,
      "grad_norm": 0.5449728825694614,
      "learning_rate": 3.669815772166625e-06,
      "loss": 0.7026,
      "step": 147
    },
    {
      "epoch": 1.8893617021276596,
      "grad_norm": 0.4576714982855058,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.7011,
      "step": 148
    },
    {
      "epoch": 1.902127659574468,
      "grad_norm": 0.5195319976448147,
      "learning_rate": 3.526224127945479e-06,
      "loss": 0.7526,
      "step": 149
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 0.4707171874496708,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.6694,
      "step": 150
    },
    {
      "epoch": 1.9276595744680851,
      "grad_norm": 0.456671543427713,
      "learning_rate": 3.383951712727701e-06,
      "loss": 0.6953,
      "step": 151
    },
    {
      "epoch": 1.9404255319148938,
      "grad_norm": 0.43257527089711645,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.7002,
      "step": 152
    },
    {
      "epoch": 1.9531914893617022,
      "grad_norm": 0.48731268058311134,
      "learning_rate": 3.2431258795932863e-06,
      "loss": 0.6944,
      "step": 153
    },
    {
      "epoch": 1.9659574468085106,
      "grad_norm": 0.4776960174537403,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.7994,
      "step": 154
    },
    {
      "epoch": 1.978723404255319,
      "grad_norm": 0.45202915072462296,
      "learning_rate": 3.1038726867353587e-06,
      "loss": 0.6511,
      "step": 155
    },
    {
      "epoch": 1.9914893617021276,
      "grad_norm": 0.4585132837327996,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.647,
      "step": 156
    },
    {
      "epoch": 2.0042553191489363,
      "grad_norm": 0.849043915814881,
      "learning_rate": 2.966316784621e-06,
      "loss": 0.9898,
      "step": 157
    },
    {
      "epoch": 2.0170212765957447,
      "grad_norm": 0.47049388201053804,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.706,
      "step": 158
    },
    {
      "epoch": 2.029787234042553,
      "grad_norm": 0.4233596382881546,
      "learning_rate": 2.83058130441221e-06,
      "loss": 0.6692,
      "step": 159
    },
    {
      "epoch": 2.0425531914893615,
      "grad_norm": 0.47661574060831124,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.6987,
      "step": 160
    },
    {
      "epoch": 2.0553191489361704,
      "grad_norm": 0.48301144689234443,
      "learning_rate": 2.6967877477468394e-06,
      "loss": 0.6053,
      "step": 161
    },
    {
      "epoch": 2.068085106382979,
      "grad_norm": 0.4777060565567049,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.6766,
      "step": 162
    },
    {
      "epoch": 2.0808510638297872,
      "grad_norm": 0.38995578479059767,
      "learning_rate": 2.5650558779781635e-06,
      "loss": 0.6486,
      "step": 163
    },
    {
      "epoch": 2.0936170212765957,
      "grad_norm": 0.4469045207268816,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.6803,
      "step": 164
    },
    {
      "epoch": 2.106382978723404,
      "grad_norm": 0.428314133399734,
      "learning_rate": 2.43550361297047e-06,
      "loss": 0.6573,
      "step": 165
    },
    {
      "epoch": 2.119148936170213,
      "grad_norm": 0.4531099548598637,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.6132,
      "step": 166
    },
    {
      "epoch": 2.1319148936170214,
      "grad_norm": 0.5460835483467924,
      "learning_rate": 2.3082469195465893e-06,
      "loss": 0.6436,
      "step": 167
    },
    {
      "epoch": 2.1446808510638298,
      "grad_norm": 0.4634335271948671,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.7057,
      "step": 168
    },
    {
      "epoch": 2.157446808510638,
      "grad_norm": 0.4433923033605567,
      "learning_rate": 2.1833997096818897e-06,
      "loss": 0.6898,
      "step": 169
    },
    {
      "epoch": 2.1702127659574466,
      "grad_norm": 0.4026516337596407,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.6066,
      "step": 170
    },
    {
      "epoch": 2.1829787234042555,
      "grad_norm": 0.4792701280174098,
      "learning_rate": 2.061073738537635e-06,
      "loss": 0.7451,
      "step": 171
    },
    {
      "epoch": 2.195744680851064,
      "grad_norm": 0.5497478648874768,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.6688,
      "step": 172
    },
    {
      "epoch": 2.2085106382978723,
      "grad_norm": 0.4613200283285023,
      "learning_rate": 1.941378504424968e-06,
      "loss": 0.7059,
      "step": 173
    },
    {
      "epoch": 2.2212765957446807,
      "grad_norm": 0.445448575146405,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.6934,
      "step": 174
    },
    {
      "epoch": 2.2340425531914896,
      "grad_norm": 0.3791538615491654,
      "learning_rate": 1.8244211507891064e-06,
      "loss": 0.6081,
      "step": 175
    },
    {
      "epoch": 2.246808510638298,
      "grad_norm": 0.46053396727842905,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.7652,
      "step": 176
    },
    {
      "epoch": 2.2595744680851064,
      "grad_norm": 0.39910064726710237,
      "learning_rate": 1.7103063703014372e-06,
      "loss": 0.7186,
      "step": 177
    },
    {
      "epoch": 2.272340425531915,
      "grad_norm": 0.4019715799995289,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.6669,
      "step": 178
    },
    {
      "epoch": 2.2851063829787233,
      "grad_norm": 0.430365457540682,
      "learning_rate": 1.5991363111454023e-06,
      "loss": 0.6817,
      "step": 179
    },
    {
      "epoch": 2.297872340425532,
      "grad_norm": 0.4951142354660635,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.6174,
      "step": 180
    },
    {
      "epoch": 2.3106382978723405,
      "grad_norm": 0.4703634969823223,
      "learning_rate": 1.4910104855800429e-06,
      "loss": 0.6884,
      "step": 181
    },
    {
      "epoch": 2.323404255319149,
      "grad_norm": 0.4383048478891779,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.6935,
      "step": 182
    },
    {
      "epoch": 2.3361702127659574,
      "grad_norm": 0.40015717483297514,
      "learning_rate": 1.3860256808630429e-06,
      "loss": 0.6116,
      "step": 183
    },
    {
      "epoch": 2.348936170212766,
      "grad_norm": 0.44199347748639095,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.6779,
      "step": 184
    },
    {
      "epoch": 2.3617021276595747,
      "grad_norm": 0.4325390527297763,
      "learning_rate": 1.2842758726130283e-06,
      "loss": 0.733,
      "step": 185
    },
    {
      "epoch": 2.374468085106383,
      "grad_norm": 0.41116195093472185,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.6138,
      "step": 186
    },
    {
      "epoch": 2.3872340425531915,
      "grad_norm": 0.43169027661540094,
      "learning_rate": 1.1858521406886674e-06,
      "loss": 0.7082,
      "step": 187
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.41678457027002164,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.7132,
      "step": 188
    },
    {
      "epoch": 2.4127659574468083,
      "grad_norm": 0.4464999225599113,
      "learning_rate": 1.0908425876598512e-06,
      "loss": 0.595,
      "step": 189
    },
    {
      "epoch": 2.425531914893617,
      "grad_norm": 0.3737938370502847,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.5945,
      "step": 190
    },
    {
      "epoch": 2.4382978723404256,
      "grad_norm": 0.4346593864871973,
      "learning_rate": 9.993322599439692e-07,
      "loss": 0.7383,
      "step": 191
    },
    {
      "epoch": 2.451063829787234,
      "grad_norm": 0.3744054578474546,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.6384,
      "step": 192
    },
    {
      "epoch": 2.4638297872340424,
      "grad_norm": 0.46293489409698124,
      "learning_rate": 9.114030716778433e-07,
      "loss": 0.7709,
      "step": 193
    },
    {
      "epoch": 2.476595744680851,
      "grad_norm": 0.47676213159992253,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.67,
      "step": 194
    },
    {
      "epoch": 2.4893617021276597,
      "grad_norm": 0.4110883133835421,
      "learning_rate": 8.271337313934869e-07,
      "loss": 0.8132,
      "step": 195
    },
    {
      "epoch": 2.502127659574468,
      "grad_norm": 0.3498732547252578,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.609,
      "step": 196
    },
    {
      "epoch": 2.5148936170212766,
      "grad_norm": 0.4399096601042779,
      "learning_rate": 7.465996715633028e-07,
      "loss": 0.846,
      "step": 197
    },
    {
      "epoch": 2.527659574468085,
      "grad_norm": 0.40810583196696365,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.7215,
      "step": 198
    },
    {
      "epoch": 2.5404255319148934,
      "grad_norm": 0.3562003037576008,
      "learning_rate": 6.698729810778065e-07,
      "loss": 0.5401,
      "step": 199
    },
    {
      "epoch": 2.5531914893617023,
      "grad_norm": 0.41794386747962137,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.7005,
      "step": 200
    },
    {
      "epoch": 2.5659574468085107,
      "grad_norm": 0.395478659932515,
      "learning_rate": 5.9702234071631e-07,
      "loss": 0.6182,
      "step": 201
    },
    {
      "epoch": 2.578723404255319,
      "grad_norm": 0.4022640858856306,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.7059,
      "step": 202
    },
    {
      "epoch": 2.5914893617021275,
      "grad_norm": 0.37429538477036206,
      "learning_rate": 5.281129616683167e-07,
      "loss": 0.625,
      "step": 203
    },
    {
      "epoch": 2.604255319148936,
      "grad_norm": 0.4065643629196546,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.7309,
      "step": 204
    },
    {
      "epoch": 2.617021276595745,
      "grad_norm": 0.37314680627544133,
      "learning_rate": 4.632065271606756e-07,
      "loss": 0.6416,
      "step": 205
    },
    {
      "epoch": 2.629787234042553,
      "grad_norm": 0.37232602082470584,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.6622,
      "step": 206
    },
    {
      "epoch": 2.6425531914893616,
      "grad_norm": 0.39090758391988906,
      "learning_rate": 4.0236113724274716e-07,
      "loss": 0.6591,
      "step": 207
    },
    {
      "epoch": 2.65531914893617,
      "grad_norm": 0.3760763848394524,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.6527,
      "step": 208
    },
    {
      "epoch": 2.6680851063829785,
      "grad_norm": 0.3546421103370596,
      "learning_rate": 3.4563125677897936e-07,
      "loss": 0.6087,
      "step": 209
    },
    {
      "epoch": 2.6808510638297873,
      "grad_norm": 0.4558722715324015,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.7705,
      "step": 210
    },
    {
      "epoch": 2.6936170212765957,
      "grad_norm": 0.3838011747274437,
      "learning_rate": 2.930676666954846e-07,
      "loss": 0.6485,
      "step": 211
    },
    {
      "epoch": 2.706382978723404,
      "grad_norm": 0.3744192449633902,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.5956,
      "step": 212
    },
    {
      "epoch": 2.719148936170213,
      "grad_norm": 0.38821467045177294,
      "learning_rate": 2.447174185242324e-07,
      "loss": 0.6515,
      "step": 213
    },
    {
      "epoch": 2.731914893617021,
      "grad_norm": 0.4320094550489657,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.7521,
      "step": 214
    },
    {
      "epoch": 2.74468085106383,
      "grad_norm": 0.38626349645535607,
      "learning_rate": 2.006237922855553e-07,
      "loss": 0.6371,
      "step": 215
    },
    {
      "epoch": 2.7574468085106383,
      "grad_norm": 0.4022594095078819,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.7641,
      "step": 216
    },
    {
      "epoch": 2.7702127659574467,
      "grad_norm": 0.3752751624701117,
      "learning_rate": 1.6082625774666793e-07,
      "loss": 0.6856,
      "step": 217
    },
    {
      "epoch": 2.7829787234042556,
      "grad_norm": 0.39713460273316725,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.661,
      "step": 218
    },
    {
      "epoch": 2.795744680851064,
      "grad_norm": 0.3951904567722507,
      "learning_rate": 1.253604390908819e-07,
      "loss": 0.7534,
      "step": 219
    },
    {
      "epoch": 2.8085106382978724,
      "grad_norm": 0.3912160449369003,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.678,
      "step": 220
    },
    {
      "epoch": 2.821276595744681,
      "grad_norm": 0.37799117554814204,
      "learning_rate": 9.42580830291373e-08,
      "loss": 0.7268,
      "step": 221
    },
    {
      "epoch": 2.8340425531914892,
      "grad_norm": 0.37825741134583873,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.6707,
      "step": 222
    },
    {
      "epoch": 2.846808510638298,
      "grad_norm": 0.4431058847999343,
      "learning_rate": 6.75470303823933e-08,
      "loss": 0.7161,
      "step": 223
    },
    {
      "epoch": 2.8595744680851065,
      "grad_norm": 0.4156239531283083,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.6633,
      "step": 224
    },
    {
      "epoch": 2.872340425531915,
      "grad_norm": 0.41505028162528623,
      "learning_rate": 4.52511911603265e-08,
      "loss": 0.6042,
      "step": 225
    },
    {
      "epoch": 2.8851063829787233,
      "grad_norm": 0.38641333321427823,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.7831,
      "step": 226
    },
    {
      "epoch": 2.8978723404255318,
      "grad_norm": 0.38474595541773154,
      "learning_rate": 2.7390523158633552e-08,
      "loss": 0.6996,
      "step": 227
    },
    {
      "epoch": 2.9106382978723406,
      "grad_norm": 0.40452129313178115,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.6679,
      "step": 228
    },
    {
      "epoch": 2.923404255319149,
      "grad_norm": 0.37427821508754977,
      "learning_rate": 1.3981014094099354e-08,
      "loss": 0.6368,
      "step": 229
    },
    {
      "epoch": 2.9361702127659575,
      "grad_norm": 0.3995889111864553,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.7147,
      "step": 230
    },
    {
      "epoch": 2.948936170212766,
      "grad_norm": 0.38436649596851297,
      "learning_rate": 5.034667293427053e-09,
      "loss": 0.6326,
      "step": 231
    },
    {
      "epoch": 2.9617021276595743,
      "grad_norm": 0.3717333771442753,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.6752,
      "step": 232
    },
    {
      "epoch": 2.974468085106383,
      "grad_norm": 0.37070096497845584,
      "learning_rate": 5.594909486328348e-10,
      "loss": 0.6547,
      "step": 233
    },
    {
      "epoch": 2.9872340425531916,
      "grad_norm": 0.3783628778531463,
      "learning_rate": 0.0,
      "loss": 0.6771,
      "step": 234
    },
    {
      "epoch": 2.9872340425531916,
      "step": 234,
      "total_flos": 226130136268800.0,
      "train_loss": 0.7424698599383363,
      "train_runtime": 3709.8986,
      "train_samples_per_second": 6.065,
      "train_steps_per_second": 0.063
    }
  ],
  "logging_steps": 1,
  "max_steps": 234,
  "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": 226130136268800.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}