{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.3001968503937008,
  "eval_steps": 153,
  "global_step": 305,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000984251968503937,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 4.9603,
      "step": 1
    },
    {
      "epoch": 0.001968503937007874,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 28.2529,
      "step": 2
    },
    {
      "epoch": 0.002952755905511811,
      "grad_norm": Infinity,
      "learning_rate": 0.0,
      "loss": 27.6365,
      "step": 3
    },
    {
      "epoch": 0.003937007874015748,
      "grad_norm": 36.58856201171875,
      "learning_rate": 9.940357852882705e-10,
      "loss": 6.1387,
      "step": 4
    },
    {
      "epoch": 0.004921259842519685,
      "grad_norm": 34.671714782714844,
      "learning_rate": 1.988071570576541e-09,
      "loss": 5.5753,
      "step": 5
    },
    {
      "epoch": 0.005905511811023622,
      "grad_norm": 35.53398132324219,
      "learning_rate": 2.9821073558648116e-09,
      "loss": 5.6951,
      "step": 6
    },
    {
      "epoch": 0.006889763779527559,
      "grad_norm": Infinity,
      "learning_rate": 2.9821073558648116e-09,
      "loss": 6.3533,
      "step": 7
    },
    {
      "epoch": 0.007874015748031496,
      "grad_norm": Infinity,
      "learning_rate": 2.9821073558648116e-09,
      "loss": 27.3848,
      "step": 8
    },
    {
      "epoch": 0.008858267716535433,
      "grad_norm": 16.95815086364746,
      "learning_rate": 3.976143141153082e-09,
      "loss": 3.8501,
      "step": 9
    },
    {
      "epoch": 0.00984251968503937,
      "grad_norm": 69.63166046142578,
      "learning_rate": 4.970178926441353e-09,
      "loss": 27.911,
      "step": 10
    },
    {
      "epoch": 0.010826771653543307,
      "grad_norm": 29.516401290893555,
      "learning_rate": 5.964214711729623e-09,
      "loss": 4.9042,
      "step": 11
    },
    {
      "epoch": 0.011811023622047244,
      "grad_norm": 39.30358123779297,
      "learning_rate": 6.9582504970178946e-09,
      "loss": 6.8003,
      "step": 12
    },
    {
      "epoch": 0.012795275590551181,
      "grad_norm": 36.83983612060547,
      "learning_rate": 7.952286282306164e-09,
      "loss": 5.7317,
      "step": 13
    },
    {
      "epoch": 0.013779527559055118,
      "grad_norm": 55.4377326965332,
      "learning_rate": 8.946322067594435e-09,
      "loss": 20.261,
      "step": 14
    },
    {
      "epoch": 0.014763779527559055,
      "grad_norm": 68.62684631347656,
      "learning_rate": 9.940357852882705e-09,
      "loss": 27.9051,
      "step": 15
    },
    {
      "epoch": 0.015748031496062992,
      "grad_norm": 31.27193832397461,
      "learning_rate": 1.0934393638170978e-08,
      "loss": 5.5959,
      "step": 16
    },
    {
      "epoch": 0.01673228346456693,
      "grad_norm": 36.56179428100586,
      "learning_rate": 1.1928429423459246e-08,
      "loss": 5.8052,
      "step": 17
    },
    {
      "epoch": 0.017716535433070866,
      "grad_norm": 23.220964431762695,
      "learning_rate": 1.2922465208747517e-08,
      "loss": 4.5088,
      "step": 18
    },
    {
      "epoch": 0.018700787401574805,
      "grad_norm": 44.375823974609375,
      "learning_rate": 1.3916500994035789e-08,
      "loss": 7.3472,
      "step": 19
    },
    {
      "epoch": 0.01968503937007874,
      "grad_norm": 40.480628967285156,
      "learning_rate": 1.4910536779324056e-08,
      "loss": 5.8668,
      "step": 20
    },
    {
      "epoch": 0.02066929133858268,
      "grad_norm": 45.778358459472656,
      "learning_rate": 1.590457256461233e-08,
      "loss": 6.4083,
      "step": 21
    },
    {
      "epoch": 0.021653543307086614,
      "grad_norm": 41.16820526123047,
      "learning_rate": 1.68986083499006e-08,
      "loss": 6.011,
      "step": 22
    },
    {
      "epoch": 0.022637795275590553,
      "grad_norm": 27.49931526184082,
      "learning_rate": 1.789264413518887e-08,
      "loss": 5.2394,
      "step": 23
    },
    {
      "epoch": 0.023622047244094488,
      "grad_norm": 20.837919235229492,
      "learning_rate": 1.888667992047714e-08,
      "loss": 4.2966,
      "step": 24
    },
    {
      "epoch": 0.024606299212598427,
      "grad_norm": 65.7834243774414,
      "learning_rate": 1.988071570576541e-08,
      "loss": 26.605,
      "step": 25
    },
    {
      "epoch": 0.025590551181102362,
      "grad_norm": 44.794960021972656,
      "learning_rate": 2.087475149105368e-08,
      "loss": 6.2067,
      "step": 26
    },
    {
      "epoch": 0.0265748031496063,
      "grad_norm": 30.213058471679688,
      "learning_rate": 2.1868787276341955e-08,
      "loss": 6.0346,
      "step": 27
    },
    {
      "epoch": 0.027559055118110236,
      "grad_norm": 27.49605941772461,
      "learning_rate": 2.2862823061630224e-08,
      "loss": 5.4676,
      "step": 28
    },
    {
      "epoch": 0.028543307086614175,
      "grad_norm": 35.94675827026367,
      "learning_rate": 2.3856858846918493e-08,
      "loss": 6.4292,
      "step": 29
    },
    {
      "epoch": 0.02952755905511811,
      "grad_norm": 65.74870300292969,
      "learning_rate": 2.4850894632206765e-08,
      "loss": 26.6452,
      "step": 30
    },
    {
      "epoch": 0.03051181102362205,
      "grad_norm": 49.05170440673828,
      "learning_rate": 2.5844930417495034e-08,
      "loss": 18.8401,
      "step": 31
    },
    {
      "epoch": 0.031496062992125984,
      "grad_norm": 49.38396453857422,
      "learning_rate": 2.6838966202783303e-08,
      "loss": 7.4531,
      "step": 32
    },
    {
      "epoch": 0.03248031496062992,
      "grad_norm": 25.29888916015625,
      "learning_rate": 2.7833001988071578e-08,
      "loss": 4.8286,
      "step": 33
    },
    {
      "epoch": 0.03346456692913386,
      "grad_norm": 28.889131546020508,
      "learning_rate": 2.8827037773359847e-08,
      "loss": 5.0078,
      "step": 34
    },
    {
      "epoch": 0.0344488188976378,
      "grad_norm": 33.611812591552734,
      "learning_rate": 2.982107355864811e-08,
      "loss": 5.4115,
      "step": 35
    },
    {
      "epoch": 0.03543307086614173,
      "grad_norm": 31.503385543823242,
      "learning_rate": 3.081510934393639e-08,
      "loss": 5.4196,
      "step": 36
    },
    {
      "epoch": 0.03641732283464567,
      "grad_norm": 23.436307907104492,
      "learning_rate": 3.180914512922466e-08,
      "loss": 4.5023,
      "step": 37
    },
    {
      "epoch": 0.03740157480314961,
      "grad_norm": 35.096893310546875,
      "learning_rate": 3.280318091451293e-08,
      "loss": 5.376,
      "step": 38
    },
    {
      "epoch": 0.038385826771653545,
      "grad_norm": 25.531570434570312,
      "learning_rate": 3.37972166998012e-08,
      "loss": 5.2303,
      "step": 39
    },
    {
      "epoch": 0.03937007874015748,
      "grad_norm": 29.393512725830078,
      "learning_rate": 3.479125248508947e-08,
      "loss": 5.6694,
      "step": 40
    },
    {
      "epoch": 0.040354330708661415,
      "grad_norm": 26.839847564697266,
      "learning_rate": 3.578528827037774e-08,
      "loss": 4.7825,
      "step": 41
    },
    {
      "epoch": 0.04133858267716536,
      "grad_norm": 21.11309814453125,
      "learning_rate": 3.6779324055666005e-08,
      "loss": 4.6507,
      "step": 42
    },
    {
      "epoch": 0.04232283464566929,
      "grad_norm": 61.134098052978516,
      "learning_rate": 3.777335984095428e-08,
      "loss": 24.2072,
      "step": 43
    },
    {
      "epoch": 0.04330708661417323,
      "grad_norm": 26.884740829467773,
      "learning_rate": 3.8767395626242556e-08,
      "loss": 4.9285,
      "step": 44
    },
    {
      "epoch": 0.04429133858267716,
      "grad_norm": 33.500144958496094,
      "learning_rate": 3.976143141153082e-08,
      "loss": 6.326,
      "step": 45
    },
    {
      "epoch": 0.045275590551181105,
      "grad_norm": 17.54262924194336,
      "learning_rate": 4.0755467196819094e-08,
      "loss": 4.5724,
      "step": 46
    },
    {
      "epoch": 0.04625984251968504,
      "grad_norm": 23.30596351623535,
      "learning_rate": 4.174950298210736e-08,
      "loss": 4.754,
      "step": 47
    },
    {
      "epoch": 0.047244094488188976,
      "grad_norm": 34.042816162109375,
      "learning_rate": 4.274353876739563e-08,
      "loss": 5.5443,
      "step": 48
    },
    {
      "epoch": 0.04822834645669291,
      "grad_norm": 21.270071029663086,
      "learning_rate": 4.373757455268391e-08,
      "loss": 4.5764,
      "step": 49
    },
    {
      "epoch": 0.04921259842519685,
      "grad_norm": 24.815349578857422,
      "learning_rate": 4.4731610337972176e-08,
      "loss": 5.1434,
      "step": 50
    },
    {
      "epoch": 0.05019685039370079,
      "grad_norm": 55.756900787353516,
      "learning_rate": 4.572564612326045e-08,
      "loss": 22.6991,
      "step": 51
    },
    {
      "epoch": 0.051181102362204724,
      "grad_norm": 23.544273376464844,
      "learning_rate": 4.6719681908548713e-08,
      "loss": 5.4277,
      "step": 52
    },
    {
      "epoch": 0.05216535433070866,
      "grad_norm": 21.845703125,
      "learning_rate": 4.7713717693836986e-08,
      "loss": 5.0178,
      "step": 53
    },
    {
      "epoch": 0.0531496062992126,
      "grad_norm": 16.331026077270508,
      "learning_rate": 4.870775347912525e-08,
      "loss": 4.8779,
      "step": 54
    },
    {
      "epoch": 0.054133858267716536,
      "grad_norm": 16.72958755493164,
      "learning_rate": 4.970178926441353e-08,
      "loss": 4.2884,
      "step": 55
    },
    {
      "epoch": 0.05511811023622047,
      "grad_norm": 41.22899627685547,
      "learning_rate": 5.06958250497018e-08,
      "loss": 16.0994,
      "step": 56
    },
    {
      "epoch": 0.05610236220472441,
      "grad_norm": 52.578941345214844,
      "learning_rate": 5.168986083499007e-08,
      "loss": 21.31,
      "step": 57
    },
    {
      "epoch": 0.05708661417322835,
      "grad_norm": 15.741512298583984,
      "learning_rate": 5.268389662027834e-08,
      "loss": 4.9721,
      "step": 58
    },
    {
      "epoch": 0.058070866141732284,
      "grad_norm": 23.57728385925293,
      "learning_rate": 5.3677932405566605e-08,
      "loss": 5.143,
      "step": 59
    },
    {
      "epoch": 0.05905511811023622,
      "grad_norm": 12.699495315551758,
      "learning_rate": 5.467196819085488e-08,
      "loss": 3.5933,
      "step": 60
    },
    {
      "epoch": 0.060039370078740155,
      "grad_norm": 17.49776840209961,
      "learning_rate": 5.5666003976143156e-08,
      "loss": 5.2559,
      "step": 61
    },
    {
      "epoch": 0.0610236220472441,
      "grad_norm": 13.251837730407715,
      "learning_rate": 5.666003976143142e-08,
      "loss": 4.0757,
      "step": 62
    },
    {
      "epoch": 0.06200787401574803,
      "grad_norm": 11.610112190246582,
      "learning_rate": 5.7654075546719694e-08,
      "loss": 3.6612,
      "step": 63
    },
    {
      "epoch": 0.06299212598425197,
      "grad_norm": 19.652385711669922,
      "learning_rate": 5.864811133200796e-08,
      "loss": 4.7505,
      "step": 64
    },
    {
      "epoch": 0.0639763779527559,
      "grad_norm": 13.930652618408203,
      "learning_rate": 5.964214711729623e-08,
      "loss": 4.1979,
      "step": 65
    },
    {
      "epoch": 0.06496062992125984,
      "grad_norm": 11.817291259765625,
      "learning_rate": 6.06361829025845e-08,
      "loss": 3.9982,
      "step": 66
    },
    {
      "epoch": 0.06594488188976377,
      "grad_norm": 18.6019287109375,
      "learning_rate": 6.163021868787278e-08,
      "loss": 4.7065,
      "step": 67
    },
    {
      "epoch": 0.06692913385826772,
      "grad_norm": 27.056259155273438,
      "learning_rate": 6.262425447316104e-08,
      "loss": 5.3413,
      "step": 68
    },
    {
      "epoch": 0.06791338582677166,
      "grad_norm": 13.010573387145996,
      "learning_rate": 6.361829025844931e-08,
      "loss": 3.6964,
      "step": 69
    },
    {
      "epoch": 0.0688976377952756,
      "grad_norm": 49.04872131347656,
      "learning_rate": 6.461232604373759e-08,
      "loss": 17.8774,
      "step": 70
    },
    {
      "epoch": 0.06988188976377953,
      "grad_norm": 19.028602600097656,
      "learning_rate": 6.560636182902586e-08,
      "loss": 4.8154,
      "step": 71
    },
    {
      "epoch": 0.07086614173228346,
      "grad_norm": 17.006460189819336,
      "learning_rate": 6.660039761431412e-08,
      "loss": 4.8356,
      "step": 72
    },
    {
      "epoch": 0.0718503937007874,
      "grad_norm": 17.15074920654297,
      "learning_rate": 6.75944333996024e-08,
      "loss": 4.568,
      "step": 73
    },
    {
      "epoch": 0.07283464566929133,
      "grad_norm": 14.456765174865723,
      "learning_rate": 6.858846918489067e-08,
      "loss": 4.0898,
      "step": 74
    },
    {
      "epoch": 0.07381889763779527,
      "grad_norm": 9.999987602233887,
      "learning_rate": 6.958250497017893e-08,
      "loss": 3.4502,
      "step": 75
    },
    {
      "epoch": 0.07480314960629922,
      "grad_norm": 13.652220726013184,
      "learning_rate": 7.057654075546721e-08,
      "loss": 3.7733,
      "step": 76
    },
    {
      "epoch": 0.07578740157480315,
      "grad_norm": 17.76757049560547,
      "learning_rate": 7.157057654075548e-08,
      "loss": 4.5204,
      "step": 77
    },
    {
      "epoch": 0.07677165354330709,
      "grad_norm": 11.42149829864502,
      "learning_rate": 7.256461232604374e-08,
      "loss": 4.2526,
      "step": 78
    },
    {
      "epoch": 0.07775590551181102,
      "grad_norm": 16.21160125732422,
      "learning_rate": 7.355864811133201e-08,
      "loss": 4.4398,
      "step": 79
    },
    {
      "epoch": 0.07874015748031496,
      "grad_norm": 12.522687911987305,
      "learning_rate": 7.455268389662029e-08,
      "loss": 4.0988,
      "step": 80
    },
    {
      "epoch": 0.0797244094488189,
      "grad_norm": 12.63741683959961,
      "learning_rate": 7.554671968190855e-08,
      "loss": 3.9704,
      "step": 81
    },
    {
      "epoch": 0.08070866141732283,
      "grad_norm": 11.259520530700684,
      "learning_rate": 7.654075546719683e-08,
      "loss": 4.3343,
      "step": 82
    },
    {
      "epoch": 0.08169291338582677,
      "grad_norm": 14.228102684020996,
      "learning_rate": 7.753479125248511e-08,
      "loss": 4.2587,
      "step": 83
    },
    {
      "epoch": 0.08267716535433071,
      "grad_norm": 45.13947677612305,
      "learning_rate": 7.852882703777338e-08,
      "loss": 15.0149,
      "step": 84
    },
    {
      "epoch": 0.08366141732283465,
      "grad_norm": 45.17081069946289,
      "learning_rate": 7.952286282306164e-08,
      "loss": 14.6599,
      "step": 85
    },
    {
      "epoch": 0.08464566929133858,
      "grad_norm": 15.967412948608398,
      "learning_rate": 8.051689860834992e-08,
      "loss": 4.0623,
      "step": 86
    },
    {
      "epoch": 0.08562992125984252,
      "grad_norm": 10.085712432861328,
      "learning_rate": 8.151093439363819e-08,
      "loss": 3.7597,
      "step": 87
    },
    {
      "epoch": 0.08661417322834646,
      "grad_norm": 13.406641960144043,
      "learning_rate": 8.250497017892645e-08,
      "loss": 4.3433,
      "step": 88
    },
    {
      "epoch": 0.08759842519685039,
      "grad_norm": 9.052105903625488,
      "learning_rate": 8.349900596421472e-08,
      "loss": 4.0287,
      "step": 89
    },
    {
      "epoch": 0.08858267716535433,
      "grad_norm": 12.489309310913086,
      "learning_rate": 8.4493041749503e-08,
      "loss": 4.6257,
      "step": 90
    },
    {
      "epoch": 0.08956692913385826,
      "grad_norm": 40.478675842285156,
      "learning_rate": 8.548707753479126e-08,
      "loss": 13.4689,
      "step": 91
    },
    {
      "epoch": 0.09055118110236221,
      "grad_norm": 14.329568862915039,
      "learning_rate": 8.648111332007953e-08,
      "loss": 4.6583,
      "step": 92
    },
    {
      "epoch": 0.09153543307086615,
      "grad_norm": 10.07358455657959,
      "learning_rate": 8.747514910536782e-08,
      "loss": 4.2682,
      "step": 93
    },
    {
      "epoch": 0.09251968503937008,
      "grad_norm": 12.861531257629395,
      "learning_rate": 8.846918489065609e-08,
      "loss": 4.468,
      "step": 94
    },
    {
      "epoch": 0.09350393700787402,
      "grad_norm": 12.121801376342773,
      "learning_rate": 8.946322067594435e-08,
      "loss": 3.4333,
      "step": 95
    },
    {
      "epoch": 0.09448818897637795,
      "grad_norm": 40.14929962158203,
      "learning_rate": 9.045725646123262e-08,
      "loss": 12.7654,
      "step": 96
    },
    {
      "epoch": 0.09547244094488189,
      "grad_norm": 11.191559791564941,
      "learning_rate": 9.14512922465209e-08,
      "loss": 3.5577,
      "step": 97
    },
    {
      "epoch": 0.09645669291338582,
      "grad_norm": 40.13950729370117,
      "learning_rate": 9.244532803180916e-08,
      "loss": 12.5875,
      "step": 98
    },
    {
      "epoch": 0.09744094488188976,
      "grad_norm": 12.063894271850586,
      "learning_rate": 9.343936381709743e-08,
      "loss": 4.2206,
      "step": 99
    },
    {
      "epoch": 0.0984251968503937,
      "grad_norm": 10.066577911376953,
      "learning_rate": 9.44333996023857e-08,
      "loss": 3.5981,
      "step": 100
    },
    {
      "epoch": 0.09940944881889764,
      "grad_norm": 10.917841911315918,
      "learning_rate": 9.542743538767397e-08,
      "loss": 3.5575,
      "step": 101
    },
    {
      "epoch": 0.10039370078740158,
      "grad_norm": 11.512818336486816,
      "learning_rate": 9.642147117296224e-08,
      "loss": 4.0271,
      "step": 102
    },
    {
      "epoch": 0.10137795275590551,
      "grad_norm": 13.737354278564453,
      "learning_rate": 9.74155069582505e-08,
      "loss": 4.0803,
      "step": 103
    },
    {
      "epoch": 0.10236220472440945,
      "grad_norm": 12.92113208770752,
      "learning_rate": 9.840954274353878e-08,
      "loss": 4.0886,
      "step": 104
    },
    {
      "epoch": 0.10334645669291338,
      "grad_norm": 16.23849868774414,
      "learning_rate": 9.940357852882706e-08,
      "loss": 4.176,
      "step": 105
    },
    {
      "epoch": 0.10433070866141732,
      "grad_norm": 13.244183540344238,
      "learning_rate": 1.0039761431411533e-07,
      "loss": 4.6653,
      "step": 106
    },
    {
      "epoch": 0.10531496062992125,
      "grad_norm": 12.089069366455078,
      "learning_rate": 1.013916500994036e-07,
      "loss": 4.3076,
      "step": 107
    },
    {
      "epoch": 0.1062992125984252,
      "grad_norm": 28.261154174804688,
      "learning_rate": 1.0238568588469187e-07,
      "loss": 8.7282,
      "step": 108
    },
    {
      "epoch": 0.10728346456692914,
      "grad_norm": 10.686351776123047,
      "learning_rate": 1.0337972166998014e-07,
      "loss": 3.4192,
      "step": 109
    },
    {
      "epoch": 0.10826771653543307,
      "grad_norm": 41.12674331665039,
      "learning_rate": 1.043737574552684e-07,
      "loss": 10.6027,
      "step": 110
    },
    {
      "epoch": 0.10925196850393701,
      "grad_norm": 13.403799057006836,
      "learning_rate": 1.0536779324055668e-07,
      "loss": 4.0959,
      "step": 111
    },
    {
      "epoch": 0.11023622047244094,
      "grad_norm": 11.321606636047363,
      "learning_rate": 1.0636182902584495e-07,
      "loss": 4.2785,
      "step": 112
    },
    {
      "epoch": 0.11122047244094488,
      "grad_norm": 14.717891693115234,
      "learning_rate": 1.0735586481113321e-07,
      "loss": 3.9945,
      "step": 113
    },
    {
      "epoch": 0.11220472440944881,
      "grad_norm": 45.00100326538086,
      "learning_rate": 1.0834990059642149e-07,
      "loss": 10.0652,
      "step": 114
    },
    {
      "epoch": 0.11318897637795275,
      "grad_norm": 12.017743110656738,
      "learning_rate": 1.0934393638170976e-07,
      "loss": 3.8621,
      "step": 115
    },
    {
      "epoch": 0.1141732283464567,
      "grad_norm": 14.086198806762695,
      "learning_rate": 1.1033797216699802e-07,
      "loss": 4.3975,
      "step": 116
    },
    {
      "epoch": 0.11515748031496063,
      "grad_norm": 43.2061767578125,
      "learning_rate": 1.1133200795228631e-07,
      "loss": 9.7899,
      "step": 117
    },
    {
      "epoch": 0.11614173228346457,
      "grad_norm": 12.647043228149414,
      "learning_rate": 1.1232604373757458e-07,
      "loss": 4.3812,
      "step": 118
    },
    {
      "epoch": 0.1171259842519685,
      "grad_norm": 11.732013702392578,
      "learning_rate": 1.1332007952286284e-07,
      "loss": 3.8715,
      "step": 119
    },
    {
      "epoch": 0.11811023622047244,
      "grad_norm": 9.888449668884277,
      "learning_rate": 1.1431411530815111e-07,
      "loss": 3.8327,
      "step": 120
    },
    {
      "epoch": 0.11909448818897637,
      "grad_norm": 9.061322212219238,
      "learning_rate": 1.1530815109343939e-07,
      "loss": 3.5103,
      "step": 121
    },
    {
      "epoch": 0.12007874015748031,
      "grad_norm": 38.643943786621094,
      "learning_rate": 1.1630218687872765e-07,
      "loss": 9.3158,
      "step": 122
    },
    {
      "epoch": 0.12106299212598425,
      "grad_norm": 11.240921974182129,
      "learning_rate": 1.1729622266401592e-07,
      "loss": 3.7201,
      "step": 123
    },
    {
      "epoch": 0.1220472440944882,
      "grad_norm": 11.231223106384277,
      "learning_rate": 1.182902584493042e-07,
      "loss": 3.4311,
      "step": 124
    },
    {
      "epoch": 0.12303149606299213,
      "grad_norm": 11.026339530944824,
      "learning_rate": 1.1928429423459245e-07,
      "loss": 3.7946,
      "step": 125
    },
    {
      "epoch": 0.12401574803149606,
      "grad_norm": 11.620814323425293,
      "learning_rate": 1.2027833001988073e-07,
      "loss": 4.0456,
      "step": 126
    },
    {
      "epoch": 0.125,
      "grad_norm": 9.652909278869629,
      "learning_rate": 1.21272365805169e-07,
      "loss": 3.482,
      "step": 127
    },
    {
      "epoch": 0.12598425196850394,
      "grad_norm": 9.82579231262207,
      "learning_rate": 1.222664015904573e-07,
      "loss": 3.1901,
      "step": 128
    },
    {
      "epoch": 0.12696850393700787,
      "grad_norm": 10.219281196594238,
      "learning_rate": 1.2326043737574557e-07,
      "loss": 3.414,
      "step": 129
    },
    {
      "epoch": 0.1279527559055118,
      "grad_norm": 9.734150886535645,
      "learning_rate": 1.2425447316103382e-07,
      "loss": 3.4967,
      "step": 130
    },
    {
      "epoch": 0.12893700787401574,
      "grad_norm": 10.505714416503906,
      "learning_rate": 1.2524850894632207e-07,
      "loss": 3.6594,
      "step": 131
    },
    {
      "epoch": 0.12992125984251968,
      "grad_norm": 39.98779296875,
      "learning_rate": 1.2624254473161035e-07,
      "loss": 8.066,
      "step": 132
    },
    {
      "epoch": 0.1309055118110236,
      "grad_norm": 9.36937427520752,
      "learning_rate": 1.2723658051689863e-07,
      "loss": 3.7872,
      "step": 133
    },
    {
      "epoch": 0.13188976377952755,
      "grad_norm": 10.953369140625,
      "learning_rate": 1.282306163021869e-07,
      "loss": 4.0023,
      "step": 134
    },
    {
      "epoch": 0.1328740157480315,
      "grad_norm": 11.828601837158203,
      "learning_rate": 1.2922465208747519e-07,
      "loss": 3.7728,
      "step": 135
    },
    {
      "epoch": 0.13385826771653545,
      "grad_norm": 9.795777320861816,
      "learning_rate": 1.3021868787276344e-07,
      "loss": 3.1893,
      "step": 136
    },
    {
      "epoch": 0.13484251968503938,
      "grad_norm": 10.846212387084961,
      "learning_rate": 1.3121272365805172e-07,
      "loss": 3.3635,
      "step": 137
    },
    {
      "epoch": 0.13582677165354332,
      "grad_norm": 13.001614570617676,
      "learning_rate": 1.3220675944333997e-07,
      "loss": 4.0195,
      "step": 138
    },
    {
      "epoch": 0.13681102362204725,
      "grad_norm": 11.749286651611328,
      "learning_rate": 1.3320079522862825e-07,
      "loss": 4.1097,
      "step": 139
    },
    {
      "epoch": 0.1377952755905512,
      "grad_norm": 12.08881664276123,
      "learning_rate": 1.3419483101391653e-07,
      "loss": 3.7903,
      "step": 140
    },
    {
      "epoch": 0.13877952755905512,
      "grad_norm": 10.898722648620605,
      "learning_rate": 1.351888667992048e-07,
      "loss": 3.5748,
      "step": 141
    },
    {
      "epoch": 0.13976377952755906,
      "grad_norm": 11.456235885620117,
      "learning_rate": 1.3618290258449306e-07,
      "loss": 3.8104,
      "step": 142
    },
    {
      "epoch": 0.140748031496063,
      "grad_norm": 47.697425842285156,
      "learning_rate": 1.3717693836978134e-07,
      "loss": 8.0411,
      "step": 143
    },
    {
      "epoch": 0.14173228346456693,
      "grad_norm": 10.220965385437012,
      "learning_rate": 1.381709741550696e-07,
      "loss": 3.4819,
      "step": 144
    },
    {
      "epoch": 0.14271653543307086,
      "grad_norm": 13.412939071655273,
      "learning_rate": 1.3916500994035787e-07,
      "loss": 3.452,
      "step": 145
    },
    {
      "epoch": 0.1437007874015748,
      "grad_norm": 11.866227149963379,
      "learning_rate": 1.4015904572564615e-07,
      "loss": 3.5861,
      "step": 146
    },
    {
      "epoch": 0.14468503937007873,
      "grad_norm": 10.724785804748535,
      "learning_rate": 1.4115308151093443e-07,
      "loss": 3.4324,
      "step": 147
    },
    {
      "epoch": 0.14566929133858267,
      "grad_norm": 11.023091316223145,
      "learning_rate": 1.421471172962227e-07,
      "loss": 3.521,
      "step": 148
    },
    {
      "epoch": 0.1466535433070866,
      "grad_norm": 12.216788291931152,
      "learning_rate": 1.4314115308151096e-07,
      "loss": 3.8868,
      "step": 149
    },
    {
      "epoch": 0.14763779527559054,
      "grad_norm": 71.6353530883789,
      "learning_rate": 1.4413518886679924e-07,
      "loss": 8.1191,
      "step": 150
    },
    {
      "epoch": 0.1486220472440945,
      "grad_norm": 10.8847074508667,
      "learning_rate": 1.451292246520875e-07,
      "loss": 3.6447,
      "step": 151
    },
    {
      "epoch": 0.14960629921259844,
      "grad_norm": 11.547443389892578,
      "learning_rate": 1.4612326043737577e-07,
      "loss": 2.9436,
      "step": 152
    },
    {
      "epoch": 0.15059055118110237,
      "grad_norm": 42.24600601196289,
      "learning_rate": 1.4711729622266402e-07,
      "loss": 8.1535,
      "step": 153
    },
    {
      "epoch": 0.15059055118110237,
      "eval_Qnli-dev_cosine_accuracy": 0.591796875,
      "eval_Qnli-dev_cosine_accuracy_threshold": 0.9479926824569702,
      "eval_Qnli-dev_cosine_ap": 0.5658036772817674,
      "eval_Qnli-dev_cosine_f1": 0.6291834002677376,
      "eval_Qnli-dev_cosine_f1_threshold": 0.7761930823326111,
      "eval_Qnli-dev_cosine_precision": 0.4598825831702544,
      "eval_Qnli-dev_cosine_recall": 0.9957627118644068,
      "eval_Qnli-dev_dot_accuracy": 0.59375,
      "eval_Qnli-dev_dot_accuracy_threshold": 724.091064453125,
      "eval_Qnli-dev_dot_ap": 0.5657459555147606,
      "eval_Qnli-dev_dot_f1": 0.6291834002677376,
      "eval_Qnli-dev_dot_f1_threshold": 596.2498779296875,
      "eval_Qnli-dev_dot_precision": 0.4598825831702544,
      "eval_Qnli-dev_dot_recall": 0.9957627118644068,
      "eval_Qnli-dev_euclidean_accuracy": 0.591796875,
      "eval_Qnli-dev_euclidean_accuracy_threshold": 8.938886642456055,
      "eval_Qnli-dev_euclidean_ap": 0.5658036772817674,
      "eval_Qnli-dev_euclidean_f1": 0.6291834002677376,
      "eval_Qnli-dev_euclidean_f1_threshold": 18.542938232421875,
      "eval_Qnli-dev_euclidean_precision": 0.4598825831702544,
      "eval_Qnli-dev_euclidean_recall": 0.9957627118644068,
      "eval_Qnli-dev_manhattan_accuracy": 0.6171875,
      "eval_Qnli-dev_manhattan_accuracy_threshold": 202.07958984375,
      "eval_Qnli-dev_manhattan_ap": 0.5891966424964378,
      "eval_Qnli-dev_manhattan_f1": 0.6291834002677376,
      "eval_Qnli-dev_manhattan_f1_threshold": 307.9236145019531,
      "eval_Qnli-dev_manhattan_precision": 0.4598825831702544,
      "eval_Qnli-dev_manhattan_recall": 0.9957627118644068,
      "eval_Qnli-dev_max_accuracy": 0.6171875,
      "eval_Qnli-dev_max_accuracy_threshold": 724.091064453125,
      "eval_Qnli-dev_max_ap": 0.5891966424964378,
      "eval_Qnli-dev_max_f1": 0.6291834002677376,
      "eval_Qnli-dev_max_f1_threshold": 596.2498779296875,
      "eval_Qnli-dev_max_precision": 0.4598825831702544,
      "eval_Qnli-dev_max_recall": 0.9957627118644068,
      "eval_allNLI-dev_cosine_accuracy": 0.666015625,
      "eval_allNLI-dev_cosine_accuracy_threshold": 0.9797871112823486,
      "eval_allNLI-dev_cosine_ap": 0.4008449937025217,
      "eval_allNLI-dev_cosine_f1": 0.504258943781942,
      "eval_allNLI-dev_cosine_f1_threshold": 0.8929213285446167,
      "eval_allNLI-dev_cosine_precision": 0.357487922705314,
      "eval_allNLI-dev_cosine_recall": 0.8554913294797688,
      "eval_allNLI-dev_dot_accuracy": 0.666015625,
      "eval_allNLI-dev_dot_accuracy_threshold": 752.6634521484375,
      "eval_allNLI-dev_dot_ap": 0.40071344979441287,
      "eval_allNLI-dev_dot_f1": 0.504258943781942,
      "eval_allNLI-dev_dot_f1_threshold": 685.9220581054688,
      "eval_allNLI-dev_dot_precision": 0.357487922705314,
      "eval_allNLI-dev_dot_recall": 0.8554913294797688,
      "eval_allNLI-dev_euclidean_accuracy": 0.666015625,
      "eval_allNLI-dev_euclidean_accuracy_threshold": 5.572628974914551,
      "eval_allNLI-dev_euclidean_ap": 0.40083962142052487,
      "eval_allNLI-dev_euclidean_f1": 0.504258943781942,
      "eval_allNLI-dev_euclidean_f1_threshold": 12.826179504394531,
      "eval_allNLI-dev_euclidean_precision": 0.357487922705314,
      "eval_allNLI-dev_euclidean_recall": 0.8554913294797688,
      "eval_allNLI-dev_manhattan_accuracy": 0.66796875,
      "eval_allNLI-dev_manhattan_accuracy_threshold": 144.52613830566406,
      "eval_allNLI-dev_manhattan_ap": 0.4008700157620745,
      "eval_allNLI-dev_manhattan_f1": 0.5075987841945289,
      "eval_allNLI-dev_manhattan_f1_threshold": 267.046875,
      "eval_allNLI-dev_manhattan_precision": 0.3443298969072165,
      "eval_allNLI-dev_manhattan_recall": 0.9653179190751445,
      "eval_allNLI-dev_max_accuracy": 0.66796875,
      "eval_allNLI-dev_max_accuracy_threshold": 752.6634521484375,
      "eval_allNLI-dev_max_ap": 0.4008700157620745,
      "eval_allNLI-dev_max_f1": 0.5075987841945289,
      "eval_allNLI-dev_max_f1_threshold": 685.9220581054688,
      "eval_allNLI-dev_max_precision": 0.357487922705314,
      "eval_allNLI-dev_max_recall": 0.9653179190751445,
      "eval_loss": 2.203232526779175,
      "eval_runtime": 50.3824,
      "eval_samples_per_second": 33.027,
      "eval_sequential_score": 0.5891966424964378,
      "eval_steps_per_second": 0.139,
      "eval_sts-test_pearson_cosine": 0.1561600438268545,
      "eval_sts-test_pearson_dot": 0.15588248423807516,
      "eval_sts-test_pearson_euclidean": 0.1908690981304929,
      "eval_sts-test_pearson_manhattan": 0.2216924674035587,
      "eval_sts-test_pearson_max": 0.2216924674035587,
      "eval_sts-test_spearman_cosine": 0.22356441354815124,
      "eval_sts-test_spearman_dot": 0.22337189362164545,
      "eval_sts-test_spearman_euclidean": 0.22363767136304896,
      "eval_sts-test_spearman_manhattan": 0.24997065610359018,
      "eval_sts-test_spearman_max": 0.24997065610359018,
      "step": 153
    },
    {
      "epoch": 0.1515748031496063,
      "grad_norm": 13.303340911865234,
      "learning_rate": 1.4811133200795232e-07,
      "loss": 3.9619,
      "step": 154
    },
    {
      "epoch": 0.15255905511811024,
      "grad_norm": 10.895444869995117,
      "learning_rate": 1.4910536779324058e-07,
      "loss": 3.1301,
      "step": 155
    },
    {
      "epoch": 0.15354330708661418,
      "grad_norm": 9.847289085388184,
      "learning_rate": 1.5009940357852886e-07,
      "loss": 3.0478,
      "step": 156
    },
    {
      "epoch": 0.1545275590551181,
      "grad_norm": 9.858047485351562,
      "learning_rate": 1.510934393638171e-07,
      "loss": 3.2986,
      "step": 157
    },
    {
      "epoch": 0.15551181102362205,
      "grad_norm": 11.554423332214355,
      "learning_rate": 1.5208747514910539e-07,
      "loss": 3.2847,
      "step": 158
    },
    {
      "epoch": 0.15649606299212598,
      "grad_norm": 11.145759582519531,
      "learning_rate": 1.5308151093439367e-07,
      "loss": 3.6599,
      "step": 159
    },
    {
      "epoch": 0.15748031496062992,
      "grad_norm": 9.598219871520996,
      "learning_rate": 1.5407554671968192e-07,
      "loss": 3.2238,
      "step": 160
    },
    {
      "epoch": 0.15846456692913385,
      "grad_norm": 11.533960342407227,
      "learning_rate": 1.5506958250497022e-07,
      "loss": 2.8897,
      "step": 161
    },
    {
      "epoch": 0.1594488188976378,
      "grad_norm": 12.17159652709961,
      "learning_rate": 1.5606361829025848e-07,
      "loss": 3.9443,
      "step": 162
    },
    {
      "epoch": 0.16043307086614172,
      "grad_norm": 10.684307098388672,
      "learning_rate": 1.5705765407554675e-07,
      "loss": 3.3733,
      "step": 163
    },
    {
      "epoch": 0.16141732283464566,
      "grad_norm": 12.093358039855957,
      "learning_rate": 1.58051689860835e-07,
      "loss": 3.7444,
      "step": 164
    },
    {
      "epoch": 0.1624015748031496,
      "grad_norm": 12.204547882080078,
      "learning_rate": 1.5904572564612329e-07,
      "loss": 3.4813,
      "step": 165
    },
    {
      "epoch": 0.16338582677165353,
      "grad_norm": 11.32477855682373,
      "learning_rate": 1.6003976143141154e-07,
      "loss": 2.6865,
      "step": 166
    },
    {
      "epoch": 0.1643700787401575,
      "grad_norm": 10.9214506149292,
      "learning_rate": 1.6103379721669984e-07,
      "loss": 2.7587,
      "step": 167
    },
    {
      "epoch": 0.16535433070866143,
      "grad_norm": 10.949960708618164,
      "learning_rate": 1.620278330019881e-07,
      "loss": 3.3628,
      "step": 168
    },
    {
      "epoch": 0.16633858267716536,
      "grad_norm": 12.229423522949219,
      "learning_rate": 1.6302186878727637e-07,
      "loss": 3.0035,
      "step": 169
    },
    {
      "epoch": 0.1673228346456693,
      "grad_norm": 77.52136993408203,
      "learning_rate": 1.6401590457256465e-07,
      "loss": 10.1591,
      "step": 170
    },
    {
      "epoch": 0.16830708661417323,
      "grad_norm": 13.680435180664062,
      "learning_rate": 1.650099403578529e-07,
      "loss": 3.5366,
      "step": 171
    },
    {
      "epoch": 0.16929133858267717,
      "grad_norm": 51.19025421142578,
      "learning_rate": 1.6600397614314118e-07,
      "loss": 8.4047,
      "step": 172
    },
    {
      "epoch": 0.1702755905511811,
      "grad_norm": 12.889047622680664,
      "learning_rate": 1.6699801192842944e-07,
      "loss": 3.8643,
      "step": 173
    },
    {
      "epoch": 0.17125984251968504,
      "grad_norm": 12.575840950012207,
      "learning_rate": 1.6799204771371774e-07,
      "loss": 3.3529,
      "step": 174
    },
    {
      "epoch": 0.17224409448818898,
      "grad_norm": 11.404970169067383,
      "learning_rate": 1.68986083499006e-07,
      "loss": 3.7143,
      "step": 175
    },
    {
      "epoch": 0.1732283464566929,
      "grad_norm": 11.78778076171875,
      "learning_rate": 1.6998011928429427e-07,
      "loss": 3.3323,
      "step": 176
    },
    {
      "epoch": 0.17421259842519685,
      "grad_norm": 11.07532787322998,
      "learning_rate": 1.7097415506958253e-07,
      "loss": 3.1206,
      "step": 177
    },
    {
      "epoch": 0.17519685039370078,
      "grad_norm": 13.040700912475586,
      "learning_rate": 1.719681908548708e-07,
      "loss": 3.1348,
      "step": 178
    },
    {
      "epoch": 0.17618110236220472,
      "grad_norm": 49.289466857910156,
      "learning_rate": 1.7296222664015906e-07,
      "loss": 7.6011,
      "step": 179
    },
    {
      "epoch": 0.17716535433070865,
      "grad_norm": 14.549922943115234,
      "learning_rate": 1.7395626242544734e-07,
      "loss": 3.7025,
      "step": 180
    },
    {
      "epoch": 0.1781496062992126,
      "grad_norm": 104.96332550048828,
      "learning_rate": 1.7495029821073564e-07,
      "loss": 10.5662,
      "step": 181
    },
    {
      "epoch": 0.17913385826771652,
      "grad_norm": 61.00530242919922,
      "learning_rate": 1.759443339960239e-07,
      "loss": 8.966,
      "step": 182
    },
    {
      "epoch": 0.18011811023622049,
      "grad_norm": 68.94571685791016,
      "learning_rate": 1.7693836978131217e-07,
      "loss": 9.426,
      "step": 183
    },
    {
      "epoch": 0.18110236220472442,
      "grad_norm": 11.119109153747559,
      "learning_rate": 1.7793240556660042e-07,
      "loss": 3.0025,
      "step": 184
    },
    {
      "epoch": 0.18208661417322836,
      "grad_norm": 34.247840881347656,
      "learning_rate": 1.789264413518887e-07,
      "loss": 7.0984,
      "step": 185
    },
    {
      "epoch": 0.1830708661417323,
      "grad_norm": 47.956634521484375,
      "learning_rate": 1.7992047713717695e-07,
      "loss": 7.3808,
      "step": 186
    },
    {
      "epoch": 0.18405511811023623,
      "grad_norm": 11.798696517944336,
      "learning_rate": 1.8091451292246523e-07,
      "loss": 2.8657,
      "step": 187
    },
    {
      "epoch": 0.18503937007874016,
      "grad_norm": 38.751102447509766,
      "learning_rate": 1.819085487077535e-07,
      "loss": 6.5636,
      "step": 188
    },
    {
      "epoch": 0.1860236220472441,
      "grad_norm": 14.265003204345703,
      "learning_rate": 1.829025844930418e-07,
      "loss": 3.4702,
      "step": 189
    },
    {
      "epoch": 0.18700787401574803,
      "grad_norm": 35.365360260009766,
      "learning_rate": 1.8389662027833004e-07,
      "loss": 5.9302,
      "step": 190
    },
    {
      "epoch": 0.18799212598425197,
      "grad_norm": 10.978341102600098,
      "learning_rate": 1.8489065606361832e-07,
      "loss": 3.2406,
      "step": 191
    },
    {
      "epoch": 0.1889763779527559,
      "grad_norm": 12.03227710723877,
      "learning_rate": 1.8588469184890657e-07,
      "loss": 3.4459,
      "step": 192
    },
    {
      "epoch": 0.18996062992125984,
      "grad_norm": 21.640823364257812,
      "learning_rate": 1.8687872763419485e-07,
      "loss": 5.269,
      "step": 193
    },
    {
      "epoch": 0.19094488188976377,
      "grad_norm": 21.88094139099121,
      "learning_rate": 1.8787276341948313e-07,
      "loss": 4.8605,
      "step": 194
    },
    {
      "epoch": 0.1919291338582677,
      "grad_norm": 10.063691139221191,
      "learning_rate": 1.888667992047714e-07,
      "loss": 2.9891,
      "step": 195
    },
    {
      "epoch": 0.19291338582677164,
      "grad_norm": 10.375823974609375,
      "learning_rate": 1.898608349900597e-07,
      "loss": 3.6681,
      "step": 196
    },
    {
      "epoch": 0.19389763779527558,
      "grad_norm": 10.918342590332031,
      "learning_rate": 1.9085487077534794e-07,
      "loss": 3.1589,
      "step": 197
    },
    {
      "epoch": 0.19488188976377951,
      "grad_norm": 12.367669105529785,
      "learning_rate": 1.9184890656063622e-07,
      "loss": 3.1835,
      "step": 198
    },
    {
      "epoch": 0.19586614173228348,
      "grad_norm": 12.073822021484375,
      "learning_rate": 1.9284294234592447e-07,
      "loss": 3.7561,
      "step": 199
    },
    {
      "epoch": 0.1968503937007874,
      "grad_norm": 12.768759727478027,
      "learning_rate": 1.9383697813121275e-07,
      "loss": 4.0891,
      "step": 200
    },
    {
      "epoch": 0.19783464566929135,
      "grad_norm": 12.075078964233398,
      "learning_rate": 1.94831013916501e-07,
      "loss": 3.563,
      "step": 201
    },
    {
      "epoch": 0.19881889763779528,
      "grad_norm": 13.486553192138672,
      "learning_rate": 1.958250497017893e-07,
      "loss": 3.7433,
      "step": 202
    },
    {
      "epoch": 0.19980314960629922,
      "grad_norm": 12.1553955078125,
      "learning_rate": 1.9681908548707756e-07,
      "loss": 3.3813,
      "step": 203
    },
    {
      "epoch": 0.20078740157480315,
      "grad_norm": 23.534427642822266,
      "learning_rate": 1.9781312127236584e-07,
      "loss": 5.2311,
      "step": 204
    },
    {
      "epoch": 0.2017716535433071,
      "grad_norm": 11.403944969177246,
      "learning_rate": 1.9880715705765412e-07,
      "loss": 3.3494,
      "step": 205
    },
    {
      "epoch": 0.20275590551181102,
      "grad_norm": 10.039087295532227,
      "learning_rate": 1.9980119284294237e-07,
      "loss": 3.3533,
      "step": 206
    },
    {
      "epoch": 0.20374015748031496,
      "grad_norm": 11.510744094848633,
      "learning_rate": 2.0079522862823065e-07,
      "loss": 3.688,
      "step": 207
    },
    {
      "epoch": 0.2047244094488189,
      "grad_norm": 10.009934425354004,
      "learning_rate": 2.017892644135189e-07,
      "loss": 3.5342,
      "step": 208
    },
    {
      "epoch": 0.20570866141732283,
      "grad_norm": 21.598899841308594,
      "learning_rate": 2.027833001988072e-07,
      "loss": 4.9381,
      "step": 209
    },
    {
      "epoch": 0.20669291338582677,
      "grad_norm": 10.349525451660156,
      "learning_rate": 2.0377733598409546e-07,
      "loss": 3.1839,
      "step": 210
    },
    {
      "epoch": 0.2076771653543307,
      "grad_norm": 10.374922752380371,
      "learning_rate": 2.0477137176938374e-07,
      "loss": 3.0465,
      "step": 211
    },
    {
      "epoch": 0.20866141732283464,
      "grad_norm": 12.201885223388672,
      "learning_rate": 2.05765407554672e-07,
      "loss": 3.1232,
      "step": 212
    },
    {
      "epoch": 0.20964566929133857,
      "grad_norm": 21.732847213745117,
      "learning_rate": 2.0675944333996027e-07,
      "loss": 4.6297,
      "step": 213
    },
    {
      "epoch": 0.2106299212598425,
      "grad_norm": 9.907038688659668,
      "learning_rate": 2.0775347912524852e-07,
      "loss": 2.9834,
      "step": 214
    },
    {
      "epoch": 0.21161417322834647,
      "grad_norm": 17.139137268066406,
      "learning_rate": 2.087475149105368e-07,
      "loss": 4.2231,
      "step": 215
    },
    {
      "epoch": 0.2125984251968504,
      "grad_norm": 12.830426216125488,
      "learning_rate": 2.097415506958251e-07,
      "loss": 3.1458,
      "step": 216
    },
    {
      "epoch": 0.21358267716535434,
      "grad_norm": 10.250080108642578,
      "learning_rate": 2.1073558648111336e-07,
      "loss": 3.2525,
      "step": 217
    },
    {
      "epoch": 0.21456692913385828,
      "grad_norm": 11.067953109741211,
      "learning_rate": 2.1172962226640164e-07,
      "loss": 3.5971,
      "step": 218
    },
    {
      "epoch": 0.2155511811023622,
      "grad_norm": 10.0042724609375,
      "learning_rate": 2.127236580516899e-07,
      "loss": 3.5616,
      "step": 219
    },
    {
      "epoch": 0.21653543307086615,
      "grad_norm": 10.706488609313965,
      "learning_rate": 2.1371769383697817e-07,
      "loss": 3.2378,
      "step": 220
    },
    {
      "epoch": 0.21751968503937008,
      "grad_norm": 12.0921630859375,
      "learning_rate": 2.1471172962226642e-07,
      "loss": 2.9075,
      "step": 221
    },
    {
      "epoch": 0.21850393700787402,
      "grad_norm": 11.22988510131836,
      "learning_rate": 2.1570576540755473e-07,
      "loss": 3.0391,
      "step": 222
    },
    {
      "epoch": 0.21948818897637795,
      "grad_norm": 11.993961334228516,
      "learning_rate": 2.1669980119284298e-07,
      "loss": 3.5573,
      "step": 223
    },
    {
      "epoch": 0.2204724409448819,
      "grad_norm": 11.548995971679688,
      "learning_rate": 2.1769383697813126e-07,
      "loss": 3.2092,
      "step": 224
    },
    {
      "epoch": 0.22145669291338582,
      "grad_norm": 9.982552528381348,
      "learning_rate": 2.186878727634195e-07,
      "loss": 3.2646,
      "step": 225
    },
    {
      "epoch": 0.22244094488188976,
      "grad_norm": 9.191473007202148,
      "learning_rate": 2.196819085487078e-07,
      "loss": 3.0886,
      "step": 226
    },
    {
      "epoch": 0.2234251968503937,
      "grad_norm": 14.892210006713867,
      "learning_rate": 2.2067594433399604e-07,
      "loss": 3.5241,
      "step": 227
    },
    {
      "epoch": 0.22440944881889763,
      "grad_norm": 12.9964599609375,
      "learning_rate": 2.2166998011928432e-07,
      "loss": 3.0111,
      "step": 228
    },
    {
      "epoch": 0.22539370078740156,
      "grad_norm": 14.988781929016113,
      "learning_rate": 2.2266401590457263e-07,
      "loss": 3.707,
      "step": 229
    },
    {
      "epoch": 0.2263779527559055,
      "grad_norm": 35.95463180541992,
      "learning_rate": 2.2365805168986088e-07,
      "loss": 5.3822,
      "step": 230
    },
    {
      "epoch": 0.22736220472440946,
      "grad_norm": 14.787928581237793,
      "learning_rate": 2.2465208747514916e-07,
      "loss": 3.2646,
      "step": 231
    },
    {
      "epoch": 0.2283464566929134,
      "grad_norm": 10.399752616882324,
      "learning_rate": 2.256461232604374e-07,
      "loss": 2.7021,
      "step": 232
    },
    {
      "epoch": 0.22933070866141733,
      "grad_norm": 12.82553482055664,
      "learning_rate": 2.266401590457257e-07,
      "loss": 3.5131,
      "step": 233
    },
    {
      "epoch": 0.23031496062992127,
      "grad_norm": 13.701425552368164,
      "learning_rate": 2.2763419483101394e-07,
      "loss": 3.103,
      "step": 234
    },
    {
      "epoch": 0.2312992125984252,
      "grad_norm": 11.487199783325195,
      "learning_rate": 2.2862823061630222e-07,
      "loss": 2.9535,
      "step": 235
    },
    {
      "epoch": 0.23228346456692914,
      "grad_norm": 11.957090377807617,
      "learning_rate": 2.296222664015905e-07,
      "loss": 2.9631,
      "step": 236
    },
    {
      "epoch": 0.23326771653543307,
      "grad_norm": 11.277992248535156,
      "learning_rate": 2.3061630218687878e-07,
      "loss": 2.8068,
      "step": 237
    },
    {
      "epoch": 0.234251968503937,
      "grad_norm": 14.584067344665527,
      "learning_rate": 2.3161033797216703e-07,
      "loss": 3.4251,
      "step": 238
    },
    {
      "epoch": 0.23523622047244094,
      "grad_norm": 12.476365089416504,
      "learning_rate": 2.326043737574553e-07,
      "loss": 2.8495,
      "step": 239
    },
    {
      "epoch": 0.23622047244094488,
      "grad_norm": 10.975320816040039,
      "learning_rate": 2.335984095427436e-07,
      "loss": 2.9972,
      "step": 240
    },
    {
      "epoch": 0.2372047244094488,
      "grad_norm": 15.740063667297363,
      "learning_rate": 2.3459244532803184e-07,
      "loss": 3.3509,
      "step": 241
    },
    {
      "epoch": 0.23818897637795275,
      "grad_norm": 11.939135551452637,
      "learning_rate": 2.3558648111332012e-07,
      "loss": 2.9234,
      "step": 242
    },
    {
      "epoch": 0.23917322834645668,
      "grad_norm": 11.509920120239258,
      "learning_rate": 2.365805168986084e-07,
      "loss": 2.4086,
      "step": 243
    },
    {
      "epoch": 0.24015748031496062,
      "grad_norm": 12.131606101989746,
      "learning_rate": 2.3757455268389668e-07,
      "loss": 3.1282,
      "step": 244
    },
    {
      "epoch": 0.24114173228346455,
      "grad_norm": 10.977266311645508,
      "learning_rate": 2.385685884691849e-07,
      "loss": 2.3352,
      "step": 245
    },
    {
      "epoch": 0.2421259842519685,
      "grad_norm": 13.836033821105957,
      "learning_rate": 2.395626242544732e-07,
      "loss": 2.4706,
      "step": 246
    },
    {
      "epoch": 0.24311023622047245,
      "grad_norm": 14.244443893432617,
      "learning_rate": 2.4055666003976146e-07,
      "loss": 3.5449,
      "step": 247
    },
    {
      "epoch": 0.2440944881889764,
      "grad_norm": 13.271723747253418,
      "learning_rate": 2.4155069582504976e-07,
      "loss": 2.8963,
      "step": 248
    },
    {
      "epoch": 0.24507874015748032,
      "grad_norm": 11.74501895904541,
      "learning_rate": 2.42544731610338e-07,
      "loss": 2.773,
      "step": 249
    },
    {
      "epoch": 0.24606299212598426,
      "grad_norm": 11.362481117248535,
      "learning_rate": 2.4353876739562627e-07,
      "loss": 2.355,
      "step": 250
    },
    {
      "epoch": 0.2470472440944882,
      "grad_norm": 14.688889503479004,
      "learning_rate": 2.445328031809146e-07,
      "loss": 2.656,
      "step": 251
    },
    {
      "epoch": 0.24803149606299213,
      "grad_norm": 10.654073715209961,
      "learning_rate": 2.4552683896620283e-07,
      "loss": 2.6221,
      "step": 252
    },
    {
      "epoch": 0.24901574803149606,
      "grad_norm": 54.549888610839844,
      "learning_rate": 2.4652087475149113e-07,
      "loss": 8.6739,
      "step": 253
    },
    {
      "epoch": 0.25,
      "grad_norm": 96.43404388427734,
      "learning_rate": 2.475149105367794e-07,
      "loss": 10.8242,
      "step": 254
    },
    {
      "epoch": 0.25098425196850394,
      "grad_norm": 12.798384666442871,
      "learning_rate": 2.4850894632206764e-07,
      "loss": 2.3408,
      "step": 255
    },
    {
      "epoch": 0.25196850393700787,
      "grad_norm": 11.494848251342773,
      "learning_rate": 2.495029821073559e-07,
      "loss": 2.1221,
      "step": 256
    },
    {
      "epoch": 0.2529527559055118,
      "grad_norm": 18.67085075378418,
      "learning_rate": 2.5049701789264414e-07,
      "loss": 3.295,
      "step": 257
    },
    {
      "epoch": 0.25393700787401574,
      "grad_norm": 12.794750213623047,
      "learning_rate": 2.5149105367793245e-07,
      "loss": 2.5896,
      "step": 258
    },
    {
      "epoch": 0.2549212598425197,
      "grad_norm": 13.11160659790039,
      "learning_rate": 2.524850894632207e-07,
      "loss": 2.1215,
      "step": 259
    },
    {
      "epoch": 0.2559055118110236,
      "grad_norm": 82.75247192382812,
      "learning_rate": 2.53479125248509e-07,
      "loss": 9.4851,
      "step": 260
    },
    {
      "epoch": 0.25688976377952755,
      "grad_norm": 10.959670066833496,
      "learning_rate": 2.5447316103379726e-07,
      "loss": 2.1982,
      "step": 261
    },
    {
      "epoch": 0.2578740157480315,
      "grad_norm": 35.51136016845703,
      "learning_rate": 2.554671968190855e-07,
      "loss": 3.0568,
      "step": 262
    },
    {
      "epoch": 0.2588582677165354,
      "grad_norm": 11.970784187316895,
      "learning_rate": 2.564612326043738e-07,
      "loss": 2.6269,
      "step": 263
    },
    {
      "epoch": 0.25984251968503935,
      "grad_norm": 13.664376258850098,
      "learning_rate": 2.5745526838966207e-07,
      "loss": 2.4792,
      "step": 264
    },
    {
      "epoch": 0.2608267716535433,
      "grad_norm": 10.64925479888916,
      "learning_rate": 2.5844930417495037e-07,
      "loss": 1.9445,
      "step": 265
    },
    {
      "epoch": 0.2618110236220472,
      "grad_norm": 13.321276664733887,
      "learning_rate": 2.5944333996023857e-07,
      "loss": 2.4061,
      "step": 266
    },
    {
      "epoch": 0.26279527559055116,
      "grad_norm": 77.70325469970703,
      "learning_rate": 2.604373757455269e-07,
      "loss": 8.3116,
      "step": 267
    },
    {
      "epoch": 0.2637795275590551,
      "grad_norm": 53.12438201904297,
      "learning_rate": 2.614314115308152e-07,
      "loss": 8.0804,
      "step": 268
    },
    {
      "epoch": 0.26476377952755903,
      "grad_norm": 10.435575485229492,
      "learning_rate": 2.6242544731610343e-07,
      "loss": 2.1674,
      "step": 269
    },
    {
      "epoch": 0.265748031496063,
      "grad_norm": 51.96613311767578,
      "learning_rate": 2.634194831013917e-07,
      "loss": 7.1975,
      "step": 270
    },
    {
      "epoch": 0.26673228346456695,
      "grad_norm": 46.066497802734375,
      "learning_rate": 2.6441351888667994e-07,
      "loss": 5.9104,
      "step": 271
    },
    {
      "epoch": 0.2677165354330709,
      "grad_norm": 11.553542137145996,
      "learning_rate": 2.6540755467196824e-07,
      "loss": 2.498,
      "step": 272
    },
    {
      "epoch": 0.2687007874015748,
      "grad_norm": 12.689590454101562,
      "learning_rate": 2.664015904572565e-07,
      "loss": 2.5249,
      "step": 273
    },
    {
      "epoch": 0.26968503937007876,
      "grad_norm": 11.272449493408203,
      "learning_rate": 2.6739562624254475e-07,
      "loss": 2.7152,
      "step": 274
    },
    {
      "epoch": 0.2706692913385827,
      "grad_norm": 11.380216598510742,
      "learning_rate": 2.6838966202783305e-07,
      "loss": 2.7904,
      "step": 275
    },
    {
      "epoch": 0.27165354330708663,
      "grad_norm": 13.75843334197998,
      "learning_rate": 2.693836978131213e-07,
      "loss": 2.7745,
      "step": 276
    },
    {
      "epoch": 0.27263779527559057,
      "grad_norm": 13.37204647064209,
      "learning_rate": 2.703777335984096e-07,
      "loss": 2.9741,
      "step": 277
    },
    {
      "epoch": 0.2736220472440945,
      "grad_norm": 11.115955352783203,
      "learning_rate": 2.7137176938369786e-07,
      "loss": 1.8215,
      "step": 278
    },
    {
      "epoch": 0.27460629921259844,
      "grad_norm": 26.816680908203125,
      "learning_rate": 2.723658051689861e-07,
      "loss": 4.6844,
      "step": 279
    },
    {
      "epoch": 0.2755905511811024,
      "grad_norm": 12.469667434692383,
      "learning_rate": 2.7335984095427437e-07,
      "loss": 2.8613,
      "step": 280
    },
    {
      "epoch": 0.2765748031496063,
      "grad_norm": 12.972196578979492,
      "learning_rate": 2.743538767395627e-07,
      "loss": 2.7147,
      "step": 281
    },
    {
      "epoch": 0.27755905511811024,
      "grad_norm": 12.858671188354492,
      "learning_rate": 2.75347912524851e-07,
      "loss": 2.814,
      "step": 282
    },
    {
      "epoch": 0.2785433070866142,
      "grad_norm": 10.301759719848633,
      "learning_rate": 2.763419483101392e-07,
      "loss": 2.3569,
      "step": 283
    },
    {
      "epoch": 0.2795275590551181,
      "grad_norm": 12.948614120483398,
      "learning_rate": 2.773359840954275e-07,
      "loss": 2.672,
      "step": 284
    },
    {
      "epoch": 0.28051181102362205,
      "grad_norm": 16.839580535888672,
      "learning_rate": 2.7833001988071574e-07,
      "loss": 3.2052,
      "step": 285
    },
    {
      "epoch": 0.281496062992126,
      "grad_norm": 12.991905212402344,
      "learning_rate": 2.7932405566600404e-07,
      "loss": 2.8056,
      "step": 286
    },
    {
      "epoch": 0.2824803149606299,
      "grad_norm": 12.984047889709473,
      "learning_rate": 2.803180914512923e-07,
      "loss": 2.6268,
      "step": 287
    },
    {
      "epoch": 0.28346456692913385,
      "grad_norm": 12.771299362182617,
      "learning_rate": 2.8131212723658055e-07,
      "loss": 2.5641,
      "step": 288
    },
    {
      "epoch": 0.2844488188976378,
      "grad_norm": 12.763790130615234,
      "learning_rate": 2.8230616302186885e-07,
      "loss": 2.4475,
      "step": 289
    },
    {
      "epoch": 0.2854330708661417,
      "grad_norm": 12.817102432250977,
      "learning_rate": 2.833001988071571e-07,
      "loss": 2.7377,
      "step": 290
    },
    {
      "epoch": 0.28641732283464566,
      "grad_norm": 11.62403678894043,
      "learning_rate": 2.842942345924454e-07,
      "loss": 2.3831,
      "step": 291
    },
    {
      "epoch": 0.2874015748031496,
      "grad_norm": 88.97967529296875,
      "learning_rate": 2.852882703777336e-07,
      "loss": 8.8069,
      "step": 292
    },
    {
      "epoch": 0.28838582677165353,
      "grad_norm": 12.380749702453613,
      "learning_rate": 2.862823061630219e-07,
      "loss": 2.186,
      "step": 293
    },
    {
      "epoch": 0.28937007874015747,
      "grad_norm": 12.181745529174805,
      "learning_rate": 2.8727634194831017e-07,
      "loss": 2.3389,
      "step": 294
    },
    {
      "epoch": 0.2903543307086614,
      "grad_norm": 11.23538875579834,
      "learning_rate": 2.8827037773359847e-07,
      "loss": 1.9744,
      "step": 295
    },
    {
      "epoch": 0.29133858267716534,
      "grad_norm": 13.454959869384766,
      "learning_rate": 2.892644135188867e-07,
      "loss": 2.4491,
      "step": 296
    },
    {
      "epoch": 0.29232283464566927,
      "grad_norm": 12.226387977600098,
      "learning_rate": 2.90258449304175e-07,
      "loss": 2.5668,
      "step": 297
    },
    {
      "epoch": 0.2933070866141732,
      "grad_norm": 13.08324146270752,
      "learning_rate": 2.912524850894633e-07,
      "loss": 2.1939,
      "step": 298
    },
    {
      "epoch": 0.29429133858267714,
      "grad_norm": 12.1226224899292,
      "learning_rate": 2.9224652087475153e-07,
      "loss": 2.2832,
      "step": 299
    },
    {
      "epoch": 0.2952755905511811,
      "grad_norm": 12.738725662231445,
      "learning_rate": 2.9324055666003984e-07,
      "loss": 2.7508,
      "step": 300
    },
    {
      "epoch": 0.296259842519685,
      "grad_norm": 13.919729232788086,
      "learning_rate": 2.9423459244532804e-07,
      "loss": 2.5206,
      "step": 301
    },
    {
      "epoch": 0.297244094488189,
      "grad_norm": 13.623347282409668,
      "learning_rate": 2.9522862823061634e-07,
      "loss": 2.3522,
      "step": 302
    },
    {
      "epoch": 0.29822834645669294,
      "grad_norm": 15.347495079040527,
      "learning_rate": 2.9622266401590465e-07,
      "loss": 2.7186,
      "step": 303
    },
    {
      "epoch": 0.2992125984251969,
      "grad_norm": 13.009486198425293,
      "learning_rate": 2.972166998011929e-07,
      "loss": 2.1369,
      "step": 304
    },
    {
      "epoch": 0.3001968503937008,
      "grad_norm": 77.14462280273438,
      "learning_rate": 2.9821073558648115e-07,
      "loss": 9.7972,
      "step": 305
    }
  ],
  "logging_steps": 1,
  "max_steps": 3048,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 305,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}