{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 118,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01694915254237288,
      "grad_norm": 13.582984428143623,
      "learning_rate": 8.333333333333333e-07,
      "loss": 4.6947,
      "step": 1
    },
    {
      "epoch": 0.03389830508474576,
      "grad_norm": 14.123407176783754,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 4.7048,
      "step": 2
    },
    {
      "epoch": 0.05084745762711865,
      "grad_norm": 14.538515649925046,
      "learning_rate": 2.5e-06,
      "loss": 4.4774,
      "step": 3
    },
    {
      "epoch": 0.06779661016949153,
      "grad_norm": 12.917201583651316,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 4.5093,
      "step": 4
    },
    {
      "epoch": 0.0847457627118644,
      "grad_norm": 12.878282813705553,
      "learning_rate": 4.166666666666667e-06,
      "loss": 4.5946,
      "step": 5
    },
    {
      "epoch": 0.1016949152542373,
      "grad_norm": 13.424010409794782,
      "learning_rate": 5e-06,
      "loss": 4.6969,
      "step": 6
    },
    {
      "epoch": 0.11864406779661017,
      "grad_norm": 12.247691943280598,
      "learning_rate": 5.833333333333334e-06,
      "loss": 4.4876,
      "step": 7
    },
    {
      "epoch": 0.13559322033898305,
      "grad_norm": 13.219835780776348,
      "learning_rate": 6.666666666666667e-06,
      "loss": 4.551,
      "step": 8
    },
    {
      "epoch": 0.15254237288135594,
      "grad_norm": 13.377987371106677,
      "learning_rate": 7.500000000000001e-06,
      "loss": 4.1263,
      "step": 9
    },
    {
      "epoch": 0.1694915254237288,
      "grad_norm": 12.792383156845725,
      "learning_rate": 8.333333333333334e-06,
      "loss": 4.1093,
      "step": 10
    },
    {
      "epoch": 0.1864406779661017,
      "grad_norm": 11.58927724154859,
      "learning_rate": 9.166666666666666e-06,
      "loss": 4.0903,
      "step": 11
    },
    {
      "epoch": 0.2033898305084746,
      "grad_norm": 12.857260545779804,
      "learning_rate": 1e-05,
      "loss": 3.7907,
      "step": 12
    },
    {
      "epoch": 0.22033898305084745,
      "grad_norm": 11.263428499431882,
      "learning_rate": 9.997804182543973e-06,
      "loss": 3.4157,
      "step": 13
    },
    {
      "epoch": 0.23728813559322035,
      "grad_norm": 9.096821276875527,
      "learning_rate": 9.991218658821609e-06,
      "loss": 3.248,
      "step": 14
    },
    {
      "epoch": 0.2542372881355932,
      "grad_norm": 8.683777340261885,
      "learning_rate": 9.980249213076085e-06,
      "loss": 3.0936,
      "step": 15
    },
    {
      "epoch": 0.2711864406779661,
      "grad_norm": 7.426067753465437,
      "learning_rate": 9.964905480067585e-06,
      "loss": 2.9003,
      "step": 16
    },
    {
      "epoch": 0.288135593220339,
      "grad_norm": 6.801819161049987,
      "learning_rate": 9.945200936610821e-06,
      "loss": 2.7429,
      "step": 17
    },
    {
      "epoch": 0.3050847457627119,
      "grad_norm": 5.517891034162887,
      "learning_rate": 9.921152889737985e-06,
      "loss": 2.4546,
      "step": 18
    },
    {
      "epoch": 0.3220338983050847,
      "grad_norm": 4.607884862007527,
      "learning_rate": 9.892782461497521e-06,
      "loss": 2.2101,
      "step": 19
    },
    {
      "epoch": 0.3389830508474576,
      "grad_norm": 4.198018015086853,
      "learning_rate": 9.860114570402055e-06,
      "loss": 2.002,
      "step": 20
    },
    {
      "epoch": 0.3559322033898305,
      "grad_norm": 3.8937197111619337,
      "learning_rate": 9.823177909541795e-06,
      "loss": 1.9571,
      "step": 21
    },
    {
      "epoch": 0.3728813559322034,
      "grad_norm": 3.116953631354838,
      "learning_rate": 9.782004921382612e-06,
      "loss": 1.7892,
      "step": 22
    },
    {
      "epoch": 0.3898305084745763,
      "grad_norm": 2.88170444003024,
      "learning_rate": 9.736631769270958e-06,
      "loss": 1.8111,
      "step": 23
    },
    {
      "epoch": 0.4067796610169492,
      "grad_norm": 2.360981852701721,
      "learning_rate": 9.687098305670606e-06,
      "loss": 1.6501,
      "step": 24
    },
    {
      "epoch": 0.423728813559322,
      "grad_norm": 1.9535896880852222,
      "learning_rate": 9.633448037159167e-06,
      "loss": 1.4975,
      "step": 25
    },
    {
      "epoch": 0.4406779661016949,
      "grad_norm": 1.742928903020929,
      "learning_rate": 9.575728086215093e-06,
      "loss": 1.6203,
      "step": 26
    },
    {
      "epoch": 0.4576271186440678,
      "grad_norm": 1.6631917739747308,
      "learning_rate": 9.513989149828718e-06,
      "loss": 1.5796,
      "step": 27
    },
    {
      "epoch": 0.4745762711864407,
      "grad_norm": 1.1808991053273623,
      "learning_rate": 9.448285454973739e-06,
      "loss": 1.4377,
      "step": 28
    },
    {
      "epoch": 0.4915254237288136,
      "grad_norm": 1.0227483418783903,
      "learning_rate": 9.378674710978185e-06,
      "loss": 1.4009,
      "step": 29
    },
    {
      "epoch": 0.5084745762711864,
      "grad_norm": 0.9157571127323277,
      "learning_rate": 9.305218058836778e-06,
      "loss": 1.2478,
      "step": 30
    },
    {
      "epoch": 0.5254237288135594,
      "grad_norm": 0.9338477321471075,
      "learning_rate": 9.22798001750913e-06,
      "loss": 1.3688,
      "step": 31
    },
    {
      "epoch": 0.5423728813559322,
      "grad_norm": 0.8796568592495111,
      "learning_rate": 9.14702842725101e-06,
      "loss": 1.2535,
      "step": 32
    },
    {
      "epoch": 0.559322033898305,
      "grad_norm": 0.9219993618943527,
      "learning_rate": 9.062434390028407e-06,
      "loss": 1.2974,
      "step": 33
    },
    {
      "epoch": 0.576271186440678,
      "grad_norm": 0.9825650845819016,
      "learning_rate": 8.974272207066767e-06,
      "loss": 1.3383,
      "step": 34
    },
    {
      "epoch": 0.5932203389830508,
      "grad_norm": 0.9055827980183389,
      "learning_rate": 8.882619313590212e-06,
      "loss": 1.2601,
      "step": 35
    },
    {
      "epoch": 0.6101694915254238,
      "grad_norm": 0.9167811987216382,
      "learning_rate": 8.787556210808101e-06,
      "loss": 1.226,
      "step": 36
    },
    {
      "epoch": 0.6271186440677966,
      "grad_norm": 0.8488058692801289,
      "learning_rate": 8.689166395208638e-06,
      "loss": 1.1668,
      "step": 37
    },
    {
      "epoch": 0.6440677966101694,
      "grad_norm": 0.8462653312925347,
      "learning_rate": 8.587536285221656e-06,
      "loss": 1.2131,
      "step": 38
    },
    {
      "epoch": 0.6610169491525424,
      "grad_norm": 0.741157564370513,
      "learning_rate": 8.482755145314987e-06,
      "loss": 1.1322,
      "step": 39
    },
    {
      "epoch": 0.6779661016949152,
      "grad_norm": 0.7694007284550851,
      "learning_rate": 8.374915007591053e-06,
      "loss": 1.1526,
      "step": 40
    },
    {
      "epoch": 0.6949152542372882,
      "grad_norm": 0.7658488631154642,
      "learning_rate": 8.264110590952609e-06,
      "loss": 1.1856,
      "step": 41
    },
    {
      "epoch": 0.711864406779661,
      "grad_norm": 0.6405702241168747,
      "learning_rate": 8.150439217908557e-06,
      "loss": 1.0037,
      "step": 42
    },
    {
      "epoch": 0.7288135593220338,
      "grad_norm": 0.616014862055841,
      "learning_rate": 8.034000729092967e-06,
      "loss": 1.0329,
      "step": 43
    },
    {
      "epoch": 0.7457627118644068,
      "grad_norm": 0.6578920077905241,
      "learning_rate": 7.914897395572362e-06,
      "loss": 1.0895,
      "step": 44
    },
    {
      "epoch": 0.7627118644067796,
      "grad_norm": 0.5912450960157642,
      "learning_rate": 7.793233829018263e-06,
      "loss": 1.0218,
      "step": 45
    },
    {
      "epoch": 0.7796610169491526,
      "grad_norm": 0.5582880461976706,
      "learning_rate": 7.669116889823955e-06,
      "loss": 1.013,
      "step": 46
    },
    {
      "epoch": 0.7966101694915254,
      "grad_norm": 0.524466989613803,
      "learning_rate": 7.542655593246103e-06,
      "loss": 0.9566,
      "step": 47
    },
    {
      "epoch": 0.8135593220338984,
      "grad_norm": 0.48335542595907816,
      "learning_rate": 7.413961013653725e-06,
      "loss": 0.8977,
      "step": 48
    },
    {
      "epoch": 0.8305084745762712,
      "grad_norm": 0.5365637758122486,
      "learning_rate": 7.283146186968566e-06,
      "loss": 1.0325,
      "step": 49
    },
    {
      "epoch": 0.847457627118644,
      "grad_norm": 0.5099489526642094,
      "learning_rate": 7.1503260113826035e-06,
      "loss": 1.0182,
      "step": 50
    },
    {
      "epoch": 0.864406779661017,
      "grad_norm": 0.4657708339456314,
      "learning_rate": 7.015617146439863e-06,
      "loss": 0.9675,
      "step": 51
    },
    {
      "epoch": 0.8813559322033898,
      "grad_norm": 0.48914070170734636,
      "learning_rate": 6.879137910571191e-06,
      "loss": 0.9954,
      "step": 52
    },
    {
      "epoch": 0.8983050847457628,
      "grad_norm": 0.46882412163559467,
      "learning_rate": 6.741008177171995e-06,
      "loss": 0.9976,
      "step": 53
    },
    {
      "epoch": 0.9152542372881356,
      "grad_norm": 0.3936331113504765,
      "learning_rate": 6.601349269314188e-06,
      "loss": 0.8422,
      "step": 54
    },
    {
      "epoch": 0.9322033898305084,
      "grad_norm": 0.437572210845862,
      "learning_rate": 6.46028385318488e-06,
      "loss": 0.9785,
      "step": 55
    },
    {
      "epoch": 0.9491525423728814,
      "grad_norm": 0.42239632796928905,
      "learning_rate": 6.3179358303453386e-06,
      "loss": 0.9773,
      "step": 56
    },
    {
      "epoch": 0.9661016949152542,
      "grad_norm": 0.3849602768372098,
      "learning_rate": 6.17443022890492e-06,
      "loss": 0.9132,
      "step": 57
    },
    {
      "epoch": 0.9830508474576272,
      "grad_norm": 0.35887608129604537,
      "learning_rate": 6.029893093705492e-06,
      "loss": 0.8932,
      "step": 58
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3775382679783028,
      "learning_rate": 5.884451375612865e-06,
      "loss": 0.9605,
      "step": 59
    },
    {
      "epoch": 1.0169491525423728,
      "grad_norm": 0.35775761239990117,
      "learning_rate": 5.738232820012407e-06,
      "loss": 0.9254,
      "step": 60
    },
    {
      "epoch": 1.0338983050847457,
      "grad_norm": 0.34923814072726866,
      "learning_rate": 5.591365854606829e-06,
      "loss": 0.9174,
      "step": 61
    },
    {
      "epoch": 1.0508474576271187,
      "grad_norm": 0.34017872341374605,
      "learning_rate": 5.443979476614674e-06,
      "loss": 0.9484,
      "step": 62
    },
    {
      "epoch": 1.0677966101694916,
      "grad_norm": 0.3089243172711035,
      "learning_rate": 5.296203139468572e-06,
      "loss": 0.8374,
      "step": 63
    },
    {
      "epoch": 1.0847457627118644,
      "grad_norm": 0.2871261805872519,
      "learning_rate": 5.148166639112799e-06,
      "loss": 0.8462,
      "step": 64
    },
    {
      "epoch": 1.1016949152542372,
      "grad_norm": 0.30706922907791745,
      "learning_rate": 5e-06,
      "loss": 0.8865,
      "step": 65
    },
    {
      "epoch": 1.11864406779661,
      "grad_norm": 0.2938750275036386,
      "learning_rate": 4.8518333608872015e-06,
      "loss": 0.8858,
      "step": 66
    },
    {
      "epoch": 1.1355932203389831,
      "grad_norm": 0.2549040934942167,
      "learning_rate": 4.703796860531429e-06,
      "loss": 0.7952,
      "step": 67
    },
    {
      "epoch": 1.152542372881356,
      "grad_norm": 0.28685443691965085,
      "learning_rate": 4.556020523385326e-06,
      "loss": 0.8437,
      "step": 68
    },
    {
      "epoch": 1.1694915254237288,
      "grad_norm": 0.2867782597037466,
      "learning_rate": 4.408634145393172e-06,
      "loss": 0.8934,
      "step": 69
    },
    {
      "epoch": 1.1864406779661016,
      "grad_norm": 0.31047696033713723,
      "learning_rate": 4.261767179987595e-06,
      "loss": 0.9716,
      "step": 70
    },
    {
      "epoch": 1.2033898305084745,
      "grad_norm": 0.25941316846282275,
      "learning_rate": 4.115548624387136e-06,
      "loss": 0.8287,
      "step": 71
    },
    {
      "epoch": 1.2203389830508475,
      "grad_norm": 0.2699492319649961,
      "learning_rate": 3.970106906294509e-06,
      "loss": 0.8472,
      "step": 72
    },
    {
      "epoch": 1.2372881355932204,
      "grad_norm": 0.25234340024796287,
      "learning_rate": 3.825569771095082e-06,
      "loss": 0.8287,
      "step": 73
    },
    {
      "epoch": 1.2542372881355932,
      "grad_norm": 0.2660601540051665,
      "learning_rate": 3.682064169654663e-06,
      "loss": 0.8793,
      "step": 74
    },
    {
      "epoch": 1.271186440677966,
      "grad_norm": 0.242634531727012,
      "learning_rate": 3.539716146815122e-06,
      "loss": 0.7988,
      "step": 75
    },
    {
      "epoch": 1.288135593220339,
      "grad_norm": 0.2465357214488806,
      "learning_rate": 3.398650730685813e-06,
      "loss": 0.8259,
      "step": 76
    },
    {
      "epoch": 1.305084745762712,
      "grad_norm": 0.22575413510974532,
      "learning_rate": 3.258991822828007e-06,
      "loss": 0.7525,
      "step": 77
    },
    {
      "epoch": 1.3220338983050848,
      "grad_norm": 0.23693526616000352,
      "learning_rate": 3.1208620894288105e-06,
      "loss": 0.8184,
      "step": 78
    },
    {
      "epoch": 1.3389830508474576,
      "grad_norm": 0.23000459251399064,
      "learning_rate": 2.98438285356014e-06,
      "loss": 0.7968,
      "step": 79
    },
    {
      "epoch": 1.3559322033898304,
      "grad_norm": 0.1958693615856507,
      "learning_rate": 2.8496739886173994e-06,
      "loss": 0.6473,
      "step": 80
    },
    {
      "epoch": 1.3728813559322033,
      "grad_norm": 0.24742346042002725,
      "learning_rate": 2.716853813031435e-06,
      "loss": 0.8747,
      "step": 81
    },
    {
      "epoch": 1.3898305084745763,
      "grad_norm": 0.2506984767340435,
      "learning_rate": 2.5860389863462765e-06,
      "loss": 0.8807,
      "step": 82
    },
    {
      "epoch": 1.4067796610169492,
      "grad_norm": 0.22963412267254538,
      "learning_rate": 2.457344406753899e-06,
      "loss": 0.8101,
      "step": 83
    },
    {
      "epoch": 1.423728813559322,
      "grad_norm": 0.2317036951827151,
      "learning_rate": 2.330883110176049e-06,
      "loss": 0.8256,
      "step": 84
    },
    {
      "epoch": 1.4406779661016949,
      "grad_norm": 0.19606132470172136,
      "learning_rate": 2.2067661709817384e-06,
      "loss": 0.6938,
      "step": 85
    },
    {
      "epoch": 1.457627118644068,
      "grad_norm": 0.219441511162039,
      "learning_rate": 2.0851026044276405e-06,
      "loss": 0.7687,
      "step": 86
    },
    {
      "epoch": 1.4745762711864407,
      "grad_norm": 0.20310562062320556,
      "learning_rate": 1.9659992709070346e-06,
      "loss": 0.76,
      "step": 87
    },
    {
      "epoch": 1.4915254237288136,
      "grad_norm": 0.21286522814139364,
      "learning_rate": 1.8495607820914451e-06,
      "loss": 0.7625,
      "step": 88
    },
    {
      "epoch": 1.5084745762711864,
      "grad_norm": 0.22107972808860454,
      "learning_rate": 1.7358894090473928e-06,
      "loss": 0.8307,
      "step": 89
    },
    {
      "epoch": 1.5254237288135593,
      "grad_norm": 0.21471387542490836,
      "learning_rate": 1.6250849924089485e-06,
      "loss": 0.7501,
      "step": 90
    },
    {
      "epoch": 1.542372881355932,
      "grad_norm": 0.22047778238568697,
      "learning_rate": 1.5172448546850166e-06,
      "loss": 0.8155,
      "step": 91
    },
    {
      "epoch": 1.559322033898305,
      "grad_norm": 0.21411524032874296,
      "learning_rate": 1.4124637147783431e-06,
      "loss": 0.7937,
      "step": 92
    },
    {
      "epoch": 1.576271186440678,
      "grad_norm": 0.21327217337137594,
      "learning_rate": 1.3108336047913633e-06,
      "loss": 0.7728,
      "step": 93
    },
    {
      "epoch": 1.5932203389830508,
      "grad_norm": 0.1920350890527698,
      "learning_rate": 1.2124437891918995e-06,
      "loss": 0.6718,
      "step": 94
    },
    {
      "epoch": 1.6101694915254239,
      "grad_norm": 0.21902635675798737,
      "learning_rate": 1.1173806864097885e-06,
      "loss": 0.804,
      "step": 95
    },
    {
      "epoch": 1.6271186440677967,
      "grad_norm": 0.22207901313261283,
      "learning_rate": 1.0257277929332332e-06,
      "loss": 0.8154,
      "step": 96
    },
    {
      "epoch": 1.6440677966101696,
      "grad_norm": 0.21571048693428027,
      "learning_rate": 9.375656099715935e-07,
      "loss": 0.7889,
      "step": 97
    },
    {
      "epoch": 1.6610169491525424,
      "grad_norm": 0.21357881995728437,
      "learning_rate": 8.529715727489912e-07,
      "loss": 0.8078,
      "step": 98
    },
    {
      "epoch": 1.6779661016949152,
      "grad_norm": 0.2235192978949996,
      "learning_rate": 7.720199824908692e-07,
      "loss": 0.8096,
      "step": 99
    },
    {
      "epoch": 1.694915254237288,
      "grad_norm": 0.22622640413844738,
      "learning_rate": 6.947819411632223e-07,
      "loss": 0.8223,
      "step": 100
    },
    {
      "epoch": 1.711864406779661,
      "grad_norm": 0.22601175493797776,
      "learning_rate": 6.213252890218163e-07,
      "loss": 0.8135,
      "step": 101
    },
    {
      "epoch": 1.7288135593220337,
      "grad_norm": 0.1911211920002579,
      "learning_rate": 5.517145450262639e-07,
      "loss": 0.7068,
      "step": 102
    },
    {
      "epoch": 1.7457627118644068,
      "grad_norm": 0.21175848010983564,
      "learning_rate": 4.860108501712824e-07,
      "loss": 0.7783,
      "step": 103
    },
    {
      "epoch": 1.7627118644067796,
      "grad_norm": 0.21053330888859195,
      "learning_rate": 4.242719137849077e-07,
      "loss": 0.8187,
      "step": 104
    },
    {
      "epoch": 1.7796610169491527,
      "grad_norm": 0.2130541935136037,
      "learning_rate": 3.665519628408332e-07,
      "loss": 0.7656,
      "step": 105
    },
    {
      "epoch": 1.7966101694915255,
      "grad_norm": 0.22526303577323026,
      "learning_rate": 3.1290169432939556e-07,
      "loss": 0.8245,
      "step": 106
    },
    {
      "epoch": 1.8135593220338984,
      "grad_norm": 0.2072005309322467,
      "learning_rate": 2.6336823072904305e-07,
      "loss": 0.743,
      "step": 107
    },
    {
      "epoch": 1.8305084745762712,
      "grad_norm": 0.22448423254493008,
      "learning_rate": 2.179950786173879e-07,
      "loss": 0.7898,
      "step": 108
    },
    {
      "epoch": 1.847457627118644,
      "grad_norm": 0.2138555012447717,
      "learning_rate": 1.7682209045820687e-07,
      "loss": 0.7782,
      "step": 109
    },
    {
      "epoch": 1.8644067796610169,
      "grad_norm": 0.1993249828378952,
      "learning_rate": 1.3988542959794627e-07,
      "loss": 0.7393,
      "step": 110
    },
    {
      "epoch": 1.8813559322033897,
      "grad_norm": 0.21708860295387616,
      "learning_rate": 1.0721753850247984e-07,
      "loss": 0.7906,
      "step": 111
    },
    {
      "epoch": 1.8983050847457628,
      "grad_norm": 0.18874363435593658,
      "learning_rate": 7.884711026201586e-08,
      "loss": 0.6812,
      "step": 112
    },
    {
      "epoch": 1.9152542372881356,
      "grad_norm": 0.25545899012578377,
      "learning_rate": 5.479906338917984e-08,
      "loss": 0.9486,
      "step": 113
    },
    {
      "epoch": 1.9322033898305084,
      "grad_norm": 0.1936498628431566,
      "learning_rate": 3.5094519932415417e-08,
      "loss": 0.745,
      "step": 114
    },
    {
      "epoch": 1.9491525423728815,
      "grad_norm": 0.20802679119960885,
      "learning_rate": 1.975078692391552e-08,
      "loss": 0.7714,
      "step": 115
    },
    {
      "epoch": 1.9661016949152543,
      "grad_norm": 0.22619971076796636,
      "learning_rate": 8.781341178393244e-09,
      "loss": 0.8194,
      "step": 116
    },
    {
      "epoch": 1.9830508474576272,
      "grad_norm": 0.2211171658614822,
      "learning_rate": 2.19581745602826e-09,
      "loss": 0.7564,
      "step": 117
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.22426334671251008,
      "learning_rate": 0.0,
      "loss": 0.8302,
      "step": 118
    },
    {
      "epoch": 2.0,
      "step": 118,
      "total_flos": 1854342635192320.0,
      "train_loss": 1.4390717199293233,
      "train_runtime": 1287.2643,
      "train_samples_per_second": 5.867,
      "train_steps_per_second": 0.092
    }
  ],
  "logging_steps": 1,
  "max_steps": 118,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 1854342635192320.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}