{
  "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": 6.0688,
      "step": 1
    },
    {
      "epoch": 0.001968503937007874,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 7.5576,
      "step": 2
    },
    {
      "epoch": 0.002952755905511811,
      "grad_norm": Infinity,
      "learning_rate": 0.0,
      "loss": 4.6849,
      "step": 3
    },
    {
      "epoch": 0.003937007874015748,
      "grad_norm": Infinity,
      "learning_rate": 0.0,
      "loss": 5.4503,
      "step": 4
    },
    {
      "epoch": 0.004921259842519685,
      "grad_norm": 32.947513580322266,
      "learning_rate": 9.940357852882705e-10,
      "loss": 5.6057,
      "step": 5
    },
    {
      "epoch": 0.005905511811023622,
      "grad_norm": 40.39633560180664,
      "learning_rate": 1.988071570576541e-09,
      "loss": 6.3049,
      "step": 6
    },
    {
      "epoch": 0.006889763779527559,
      "grad_norm": 44.38965606689453,
      "learning_rate": 2.9821073558648116e-09,
      "loss": 6.8336,
      "step": 7
    },
    {
      "epoch": 0.007874015748031496,
      "grad_norm": 28.46092987060547,
      "learning_rate": 3.976143141153082e-09,
      "loss": 5.0777,
      "step": 8
    },
    {
      "epoch": 0.008858267716535433,
      "grad_norm": 25.817373275756836,
      "learning_rate": 4.970178926441353e-09,
      "loss": 4.8358,
      "step": 9
    },
    {
      "epoch": 0.00984251968503937,
      "grad_norm": 27.06571388244629,
      "learning_rate": 5.964214711729623e-09,
      "loss": 4.641,
      "step": 10
    },
    {
      "epoch": 0.010826771653543307,
      "grad_norm": 27.857297897338867,
      "learning_rate": 6.9582504970178946e-09,
      "loss": 4.828,
      "step": 11
    },
    {
      "epoch": 0.011811023622047244,
      "grad_norm": 29.35377311706543,
      "learning_rate": 7.952286282306164e-09,
      "loss": 5.2269,
      "step": 12
    },
    {
      "epoch": 0.012795275590551181,
      "grad_norm": 32.06682586669922,
      "learning_rate": 8.946322067594435e-09,
      "loss": 5.6772,
      "step": 13
    },
    {
      "epoch": 0.013779527559055118,
      "grad_norm": 30.774206161499023,
      "learning_rate": 9.940357852882705e-09,
      "loss": 5.1422,
      "step": 14
    },
    {
      "epoch": 0.014763779527559055,
      "grad_norm": 45.917659759521484,
      "learning_rate": 1.0934393638170978e-08,
      "loss": 6.2469,
      "step": 15
    },
    {
      "epoch": 0.015748031496062992,
      "grad_norm": 30.189363479614258,
      "learning_rate": 1.1928429423459246e-08,
      "loss": 4.6802,
      "step": 16
    },
    {
      "epoch": 0.01673228346456693,
      "grad_norm": 25.023223876953125,
      "learning_rate": 1.2922465208747517e-08,
      "loss": 4.5492,
      "step": 17
    },
    {
      "epoch": 0.017716535433070866,
      "grad_norm": 26.336748123168945,
      "learning_rate": 1.3916500994035789e-08,
      "loss": 4.8062,
      "step": 18
    },
    {
      "epoch": 0.018700787401574805,
      "grad_norm": 51.350032806396484,
      "learning_rate": 1.4910536779324056e-08,
      "loss": 7.5141,
      "step": 19
    },
    {
      "epoch": 0.01968503937007874,
      "grad_norm": 33.953224182128906,
      "learning_rate": 1.590457256461233e-08,
      "loss": 5.5202,
      "step": 20
    },
    {
      "epoch": 0.02066929133858268,
      "grad_norm": 36.828433990478516,
      "learning_rate": 1.68986083499006e-08,
      "loss": 6.5025,
      "step": 21
    },
    {
      "epoch": 0.021653543307086614,
      "grad_norm": 43.68119812011719,
      "learning_rate": 1.789264413518887e-08,
      "loss": 7.318,
      "step": 22
    },
    {
      "epoch": 0.022637795275590553,
      "grad_norm": 23.9952392578125,
      "learning_rate": 1.888667992047714e-08,
      "loss": 4.6458,
      "step": 23
    },
    {
      "epoch": 0.023622047244094488,
      "grad_norm": 19.315523147583008,
      "learning_rate": 1.988071570576541e-08,
      "loss": 4.6191,
      "step": 24
    },
    {
      "epoch": 0.024606299212598427,
      "grad_norm": 22.747207641601562,
      "learning_rate": 2.087475149105368e-08,
      "loss": 4.3159,
      "step": 25
    },
    {
      "epoch": 0.025590551181102362,
      "grad_norm": 32.90699768066406,
      "learning_rate": 2.1868787276341955e-08,
      "loss": 6.3677,
      "step": 26
    },
    {
      "epoch": 0.0265748031496063,
      "grad_norm": 25.224443435668945,
      "learning_rate": 2.2862823061630224e-08,
      "loss": 5.6052,
      "step": 27
    },
    {
      "epoch": 0.027559055118110236,
      "grad_norm": 16.32752799987793,
      "learning_rate": 2.3856858846918493e-08,
      "loss": 4.196,
      "step": 28
    },
    {
      "epoch": 0.028543307086614175,
      "grad_norm": 19.50617218017578,
      "learning_rate": 2.4850894632206765e-08,
      "loss": 4.4802,
      "step": 29
    },
    {
      "epoch": 0.02952755905511811,
      "grad_norm": 21.613880157470703,
      "learning_rate": 2.5844930417495034e-08,
      "loss": 4.9193,
      "step": 30
    },
    {
      "epoch": 0.03051181102362205,
      "grad_norm": 16.059247970581055,
      "learning_rate": 2.6838966202783303e-08,
      "loss": 4.0996,
      "step": 31
    },
    {
      "epoch": 0.031496062992125984,
      "grad_norm": 27.45536994934082,
      "learning_rate": 2.7833001988071578e-08,
      "loss": 5.6307,
      "step": 32
    },
    {
      "epoch": 0.03248031496062992,
      "grad_norm": 19.249103546142578,
      "learning_rate": 2.8827037773359847e-08,
      "loss": 4.5745,
      "step": 33
    },
    {
      "epoch": 0.03346456692913386,
      "grad_norm": 17.516521453857422,
      "learning_rate": 2.982107355864811e-08,
      "loss": 4.4514,
      "step": 34
    },
    {
      "epoch": 0.0344488188976378,
      "grad_norm": 11.684857368469238,
      "learning_rate": 3.081510934393639e-08,
      "loss": 4.0617,
      "step": 35
    },
    {
      "epoch": 0.03543307086614173,
      "grad_norm": 16.813453674316406,
      "learning_rate": 3.180914512922466e-08,
      "loss": 5.0298,
      "step": 36
    },
    {
      "epoch": 0.03641732283464567,
      "grad_norm": 10.92259407043457,
      "learning_rate": 3.280318091451293e-08,
      "loss": 3.9815,
      "step": 37
    },
    {
      "epoch": 0.03740157480314961,
      "grad_norm": 11.449631690979004,
      "learning_rate": 3.37972166998012e-08,
      "loss": 4.0871,
      "step": 38
    },
    {
      "epoch": 0.038385826771653545,
      "grad_norm": 11.223074913024902,
      "learning_rate": 3.479125248508947e-08,
      "loss": 4.2378,
      "step": 39
    },
    {
      "epoch": 0.03937007874015748,
      "grad_norm": 9.864290237426758,
      "learning_rate": 3.578528827037774e-08,
      "loss": 3.8226,
      "step": 40
    },
    {
      "epoch": 0.040354330708661415,
      "grad_norm": 11.084328651428223,
      "learning_rate": 3.6779324055666005e-08,
      "loss": 4.3519,
      "step": 41
    },
    {
      "epoch": 0.04133858267716536,
      "grad_norm": 9.965168952941895,
      "learning_rate": 3.777335984095428e-08,
      "loss": 3.6345,
      "step": 42
    },
    {
      "epoch": 0.04232283464566929,
      "grad_norm": 18.375577926635742,
      "learning_rate": 3.8767395626242556e-08,
      "loss": 5.0829,
      "step": 43
    },
    {
      "epoch": 0.04330708661417323,
      "grad_norm": 11.232723236083984,
      "learning_rate": 3.976143141153082e-08,
      "loss": 4.6701,
      "step": 44
    },
    {
      "epoch": 0.04429133858267716,
      "grad_norm": 10.896126747131348,
      "learning_rate": 4.0755467196819094e-08,
      "loss": 4.1371,
      "step": 45
    },
    {
      "epoch": 0.045275590551181105,
      "grad_norm": 10.237566947937012,
      "learning_rate": 4.174950298210736e-08,
      "loss": 4.2418,
      "step": 46
    },
    {
      "epoch": 0.04625984251968504,
      "grad_norm": 13.484391212463379,
      "learning_rate": 4.274353876739563e-08,
      "loss": 4.4766,
      "step": 47
    },
    {
      "epoch": 0.047244094488188976,
      "grad_norm": 9.769365310668945,
      "learning_rate": 4.373757455268391e-08,
      "loss": 4.4797,
      "step": 48
    },
    {
      "epoch": 0.04822834645669291,
      "grad_norm": 9.953036308288574,
      "learning_rate": 4.4731610337972176e-08,
      "loss": 3.8471,
      "step": 49
    },
    {
      "epoch": 0.04921259842519685,
      "grad_norm": 15.378759384155273,
      "learning_rate": 4.572564612326045e-08,
      "loss": 4.3194,
      "step": 50
    },
    {
      "epoch": 0.05019685039370079,
      "grad_norm": 9.474905967712402,
      "learning_rate": 4.6719681908548713e-08,
      "loss": 3.9426,
      "step": 51
    },
    {
      "epoch": 0.051181102362204724,
      "grad_norm": 8.882713317871094,
      "learning_rate": 4.7713717693836986e-08,
      "loss": 3.5333,
      "step": 52
    },
    {
      "epoch": 0.05216535433070866,
      "grad_norm": 14.956167221069336,
      "learning_rate": 4.870775347912525e-08,
      "loss": 4.2426,
      "step": 53
    },
    {
      "epoch": 0.0531496062992126,
      "grad_norm": 9.333856582641602,
      "learning_rate": 4.970178926441353e-08,
      "loss": 3.9816,
      "step": 54
    },
    {
      "epoch": 0.054133858267716536,
      "grad_norm": 9.034503936767578,
      "learning_rate": 5.06958250497018e-08,
      "loss": 3.663,
      "step": 55
    },
    {
      "epoch": 0.05511811023622047,
      "grad_norm": 10.114209175109863,
      "learning_rate": 5.168986083499007e-08,
      "loss": 3.9057,
      "step": 56
    },
    {
      "epoch": 0.05610236220472441,
      "grad_norm": 9.802759170532227,
      "learning_rate": 5.268389662027834e-08,
      "loss": 4.0345,
      "step": 57
    },
    {
      "epoch": 0.05708661417322835,
      "grad_norm": 10.378804206848145,
      "learning_rate": 5.3677932405566605e-08,
      "loss": 3.5233,
      "step": 58
    },
    {
      "epoch": 0.058070866141732284,
      "grad_norm": 10.237386703491211,
      "learning_rate": 5.467196819085488e-08,
      "loss": 3.7999,
      "step": 59
    },
    {
      "epoch": 0.05905511811023622,
      "grad_norm": 11.103493690490723,
      "learning_rate": 5.5666003976143156e-08,
      "loss": 3.1885,
      "step": 60
    },
    {
      "epoch": 0.060039370078740155,
      "grad_norm": 11.967391014099121,
      "learning_rate": 5.666003976143142e-08,
      "loss": 3.6013,
      "step": 61
    },
    {
      "epoch": 0.0610236220472441,
      "grad_norm": 10.79468822479248,
      "learning_rate": 5.7654075546719694e-08,
      "loss": 3.392,
      "step": 62
    },
    {
      "epoch": 0.06200787401574803,
      "grad_norm": 13.18323802947998,
      "learning_rate": 5.864811133200796e-08,
      "loss": 3.3814,
      "step": 63
    },
    {
      "epoch": 0.06299212598425197,
      "grad_norm": 15.541842460632324,
      "learning_rate": 5.964214711729623e-08,
      "loss": 4.0428,
      "step": 64
    },
    {
      "epoch": 0.0639763779527559,
      "grad_norm": 13.166780471801758,
      "learning_rate": 6.06361829025845e-08,
      "loss": 3.7825,
      "step": 65
    },
    {
      "epoch": 0.06496062992125984,
      "grad_norm": 12.48865795135498,
      "learning_rate": 6.163021868787278e-08,
      "loss": 3.4181,
      "step": 66
    },
    {
      "epoch": 0.06594488188976377,
      "grad_norm": 14.15877628326416,
      "learning_rate": 6.262425447316104e-08,
      "loss": 3.7793,
      "step": 67
    },
    {
      "epoch": 0.06692913385826772,
      "grad_norm": 18.16546630859375,
      "learning_rate": 6.361829025844931e-08,
      "loss": 3.8344,
      "step": 68
    },
    {
      "epoch": 0.06791338582677166,
      "grad_norm": 14.758286476135254,
      "learning_rate": 6.461232604373759e-08,
      "loss": 3.2165,
      "step": 69
    },
    {
      "epoch": 0.0688976377952756,
      "grad_norm": 15.924538612365723,
      "learning_rate": 6.560636182902586e-08,
      "loss": 3.3811,
      "step": 70
    },
    {
      "epoch": 0.06988188976377953,
      "grad_norm": 17.181415557861328,
      "learning_rate": 6.660039761431412e-08,
      "loss": 3.5984,
      "step": 71
    },
    {
      "epoch": 0.07086614173228346,
      "grad_norm": 16.691272735595703,
      "learning_rate": 6.75944333996024e-08,
      "loss": 3.8583,
      "step": 72
    },
    {
      "epoch": 0.0718503937007874,
      "grad_norm": 17.478809356689453,
      "learning_rate": 6.858846918489067e-08,
      "loss": 3.296,
      "step": 73
    },
    {
      "epoch": 0.07283464566929133,
      "grad_norm": 15.149248123168945,
      "learning_rate": 6.958250497017893e-08,
      "loss": 2.7661,
      "step": 74
    },
    {
      "epoch": 0.07381889763779527,
      "grad_norm": 15.139701843261719,
      "learning_rate": 7.057654075546721e-08,
      "loss": 2.9805,
      "step": 75
    },
    {
      "epoch": 0.07480314960629922,
      "grad_norm": 14.794472694396973,
      "learning_rate": 7.157057654075548e-08,
      "loss": 2.566,
      "step": 76
    },
    {
      "epoch": 0.07578740157480315,
      "grad_norm": 15.433120727539062,
      "learning_rate": 7.256461232604374e-08,
      "loss": 3.258,
      "step": 77
    },
    {
      "epoch": 0.07677165354330709,
      "grad_norm": 22.558115005493164,
      "learning_rate": 7.355864811133201e-08,
      "loss": 3.3804,
      "step": 78
    },
    {
      "epoch": 0.07775590551181102,
      "grad_norm": 19.630340576171875,
      "learning_rate": 7.455268389662029e-08,
      "loss": 2.8828,
      "step": 79
    },
    {
      "epoch": 0.07874015748031496,
      "grad_norm": 17.1998233795166,
      "learning_rate": 7.554671968190855e-08,
      "loss": 3.1077,
      "step": 80
    },
    {
      "epoch": 0.0797244094488189,
      "grad_norm": 13.535666465759277,
      "learning_rate": 7.654075546719683e-08,
      "loss": 2.9441,
      "step": 81
    },
    {
      "epoch": 0.08070866141732283,
      "grad_norm": 14.031355857849121,
      "learning_rate": 7.753479125248511e-08,
      "loss": 2.9465,
      "step": 82
    },
    {
      "epoch": 0.08169291338582677,
      "grad_norm": 13.701343536376953,
      "learning_rate": 7.852882703777338e-08,
      "loss": 2.7088,
      "step": 83
    },
    {
      "epoch": 0.08267716535433071,
      "grad_norm": 15.102946281433105,
      "learning_rate": 7.952286282306164e-08,
      "loss": 2.9215,
      "step": 84
    },
    {
      "epoch": 0.08366141732283465,
      "grad_norm": 20.15928077697754,
      "learning_rate": 8.051689860834992e-08,
      "loss": 3.4698,
      "step": 85
    },
    {
      "epoch": 0.08464566929133858,
      "grad_norm": 14.807561874389648,
      "learning_rate": 8.151093439363819e-08,
      "loss": 2.2414,
      "step": 86
    },
    {
      "epoch": 0.08562992125984252,
      "grad_norm": 18.390995025634766,
      "learning_rate": 8.250497017892645e-08,
      "loss": 3.1601,
      "step": 87
    },
    {
      "epoch": 0.08661417322834646,
      "grad_norm": 17.47426986694336,
      "learning_rate": 8.349900596421472e-08,
      "loss": 2.7714,
      "step": 88
    },
    {
      "epoch": 0.08759842519685039,
      "grad_norm": 19.336660385131836,
      "learning_rate": 8.4493041749503e-08,
      "loss": 3.0311,
      "step": 89
    },
    {
      "epoch": 0.08858267716535433,
      "grad_norm": 22.308195114135742,
      "learning_rate": 8.548707753479126e-08,
      "loss": 3.0336,
      "step": 90
    },
    {
      "epoch": 0.08956692913385826,
      "grad_norm": 16.85112762451172,
      "learning_rate": 8.648111332007953e-08,
      "loss": 1.9358,
      "step": 91
    },
    {
      "epoch": 0.09055118110236221,
      "grad_norm": 18.357120513916016,
      "learning_rate": 8.747514910536782e-08,
      "loss": 2.6031,
      "step": 92
    },
    {
      "epoch": 0.09153543307086615,
      "grad_norm": 21.082101821899414,
      "learning_rate": 8.846918489065609e-08,
      "loss": 2.7515,
      "step": 93
    },
    {
      "epoch": 0.09251968503937008,
      "grad_norm": 21.116485595703125,
      "learning_rate": 8.946322067594435e-08,
      "loss": 2.8496,
      "step": 94
    },
    {
      "epoch": 0.09350393700787402,
      "grad_norm": 15.967782020568848,
      "learning_rate": 9.045725646123262e-08,
      "loss": 1.8015,
      "step": 95
    },
    {
      "epoch": 0.09448818897637795,
      "grad_norm": 20.430376052856445,
      "learning_rate": 9.14512922465209e-08,
      "loss": 2.8138,
      "step": 96
    },
    {
      "epoch": 0.09547244094488189,
      "grad_norm": 18.769359588623047,
      "learning_rate": 9.244532803180916e-08,
      "loss": 2.0597,
      "step": 97
    },
    {
      "epoch": 0.09645669291338582,
      "grad_norm": 19.360021591186523,
      "learning_rate": 9.343936381709743e-08,
      "loss": 2.1053,
      "step": 98
    },
    {
      "epoch": 0.09744094488188976,
      "grad_norm": 22.48013687133789,
      "learning_rate": 9.44333996023857e-08,
      "loss": 2.6785,
      "step": 99
    },
    {
      "epoch": 0.0984251968503937,
      "grad_norm": 21.50940704345703,
      "learning_rate": 9.542743538767397e-08,
      "loss": 2.588,
      "step": 100
    },
    {
      "epoch": 0.09940944881889764,
      "grad_norm": 17.409393310546875,
      "learning_rate": 9.642147117296224e-08,
      "loss": 2.0099,
      "step": 101
    },
    {
      "epoch": 0.10039370078740158,
      "grad_norm": 22.641996383666992,
      "learning_rate": 9.74155069582505e-08,
      "loss": 2.7947,
      "step": 102
    },
    {
      "epoch": 0.10137795275590551,
      "grad_norm": 18.91701889038086,
      "learning_rate": 9.840954274353878e-08,
      "loss": 2.3274,
      "step": 103
    },
    {
      "epoch": 0.10236220472440945,
      "grad_norm": 21.317251205444336,
      "learning_rate": 9.940357852882706e-08,
      "loss": 2.2545,
      "step": 104
    },
    {
      "epoch": 0.10334645669291338,
      "grad_norm": 24.298248291015625,
      "learning_rate": 1.0039761431411533e-07,
      "loss": 2.4575,
      "step": 105
    },
    {
      "epoch": 0.10433070866141732,
      "grad_norm": 19.284252166748047,
      "learning_rate": 1.013916500994036e-07,
      "loss": 2.4413,
      "step": 106
    },
    {
      "epoch": 0.10531496062992125,
      "grad_norm": 19.619426727294922,
      "learning_rate": 1.0238568588469187e-07,
      "loss": 2.3185,
      "step": 107
    },
    {
      "epoch": 0.1062992125984252,
      "grad_norm": 20.323917388916016,
      "learning_rate": 1.0337972166998014e-07,
      "loss": 2.1577,
      "step": 108
    },
    {
      "epoch": 0.10728346456692914,
      "grad_norm": 20.885499954223633,
      "learning_rate": 1.043737574552684e-07,
      "loss": 2.1278,
      "step": 109
    },
    {
      "epoch": 0.10826771653543307,
      "grad_norm": 21.687036514282227,
      "learning_rate": 1.0536779324055668e-07,
      "loss": 2.0967,
      "step": 110
    },
    {
      "epoch": 0.10925196850393701,
      "grad_norm": 23.031898498535156,
      "learning_rate": 1.0636182902584495e-07,
      "loss": 2.6142,
      "step": 111
    },
    {
      "epoch": 0.11023622047244094,
      "grad_norm": 20.64027976989746,
      "learning_rate": 1.0735586481113321e-07,
      "loss": 1.8553,
      "step": 112
    },
    {
      "epoch": 0.11122047244094488,
      "grad_norm": 18.268877029418945,
      "learning_rate": 1.0834990059642149e-07,
      "loss": 2.1523,
      "step": 113
    },
    {
      "epoch": 0.11220472440944881,
      "grad_norm": 22.190549850463867,
      "learning_rate": 1.0934393638170976e-07,
      "loss": 2.1726,
      "step": 114
    },
    {
      "epoch": 0.11318897637795275,
      "grad_norm": 18.314851760864258,
      "learning_rate": 1.1033797216699802e-07,
      "loss": 1.8564,
      "step": 115
    },
    {
      "epoch": 0.1141732283464567,
      "grad_norm": 19.17725944519043,
      "learning_rate": 1.1133200795228631e-07,
      "loss": 1.8413,
      "step": 116
    },
    {
      "epoch": 0.11515748031496063,
      "grad_norm": 24.518449783325195,
      "learning_rate": 1.1232604373757458e-07,
      "loss": 2.0441,
      "step": 117
    },
    {
      "epoch": 0.11614173228346457,
      "grad_norm": 24.78040313720703,
      "learning_rate": 1.1332007952286284e-07,
      "loss": 2.2159,
      "step": 118
    },
    {
      "epoch": 0.1171259842519685,
      "grad_norm": 28.552289962768555,
      "learning_rate": 1.1431411530815111e-07,
      "loss": 2.6779,
      "step": 119
    },
    {
      "epoch": 0.11811023622047244,
      "grad_norm": 21.049219131469727,
      "learning_rate": 1.1530815109343939e-07,
      "loss": 2.2976,
      "step": 120
    },
    {
      "epoch": 0.11909448818897637,
      "grad_norm": 23.962575912475586,
      "learning_rate": 1.1630218687872765e-07,
      "loss": 1.9407,
      "step": 121
    },
    {
      "epoch": 0.12007874015748031,
      "grad_norm": 20.15193748474121,
      "learning_rate": 1.1729622266401592e-07,
      "loss": 1.9019,
      "step": 122
    },
    {
      "epoch": 0.12106299212598425,
      "grad_norm": 22.563459396362305,
      "learning_rate": 1.182902584493042e-07,
      "loss": 2.2149,
      "step": 123
    },
    {
      "epoch": 0.1220472440944882,
      "grad_norm": 16.86638832092285,
      "learning_rate": 1.1928429423459245e-07,
      "loss": 1.6823,
      "step": 124
    },
    {
      "epoch": 0.12303149606299213,
      "grad_norm": 20.249584197998047,
      "learning_rate": 1.2027833001988073e-07,
      "loss": 1.8402,
      "step": 125
    },
    {
      "epoch": 0.12401574803149606,
      "grad_norm": 19.5487117767334,
      "learning_rate": 1.21272365805169e-07,
      "loss": 1.6914,
      "step": 126
    },
    {
      "epoch": 0.125,
      "grad_norm": 19.093505859375,
      "learning_rate": 1.222664015904573e-07,
      "loss": 2.1626,
      "step": 127
    },
    {
      "epoch": 0.12598425196850394,
      "grad_norm": 17.598743438720703,
      "learning_rate": 1.2326043737574557e-07,
      "loss": 1.6414,
      "step": 128
    },
    {
      "epoch": 0.12696850393700787,
      "grad_norm": 22.712181091308594,
      "learning_rate": 1.2425447316103382e-07,
      "loss": 2.2043,
      "step": 129
    },
    {
      "epoch": 0.1279527559055118,
      "grad_norm": 20.516571044921875,
      "learning_rate": 1.2524850894632207e-07,
      "loss": 1.9987,
      "step": 130
    },
    {
      "epoch": 0.12893700787401574,
      "grad_norm": 22.58023452758789,
      "learning_rate": 1.2624254473161035e-07,
      "loss": 1.8868,
      "step": 131
    },
    {
      "epoch": 0.12992125984251968,
      "grad_norm": 17.800390243530273,
      "learning_rate": 1.2723658051689863e-07,
      "loss": 1.8262,
      "step": 132
    },
    {
      "epoch": 0.1309055118110236,
      "grad_norm": 20.7684383392334,
      "learning_rate": 1.282306163021869e-07,
      "loss": 2.0404,
      "step": 133
    },
    {
      "epoch": 0.13188976377952755,
      "grad_norm": 20.88973617553711,
      "learning_rate": 1.2922465208747519e-07,
      "loss": 1.9134,
      "step": 134
    },
    {
      "epoch": 0.1328740157480315,
      "grad_norm": 21.78728485107422,
      "learning_rate": 1.3021868787276344e-07,
      "loss": 2.3725,
      "step": 135
    },
    {
      "epoch": 0.13385826771653545,
      "grad_norm": 16.885374069213867,
      "learning_rate": 1.3121272365805172e-07,
      "loss": 1.4127,
      "step": 136
    },
    {
      "epoch": 0.13484251968503938,
      "grad_norm": 18.944265365600586,
      "learning_rate": 1.3220675944333997e-07,
      "loss": 1.6876,
      "step": 137
    },
    {
      "epoch": 0.13582677165354332,
      "grad_norm": 18.579692840576172,
      "learning_rate": 1.3320079522862825e-07,
      "loss": 1.8376,
      "step": 138
    },
    {
      "epoch": 0.13681102362204725,
      "grad_norm": 19.68845558166504,
      "learning_rate": 1.3419483101391653e-07,
      "loss": 1.6992,
      "step": 139
    },
    {
      "epoch": 0.1377952755905512,
      "grad_norm": 21.27842903137207,
      "learning_rate": 1.351888667992048e-07,
      "loss": 1.5032,
      "step": 140
    },
    {
      "epoch": 0.13877952755905512,
      "grad_norm": 28.506824493408203,
      "learning_rate": 1.3618290258449306e-07,
      "loss": 2.0334,
      "step": 141
    },
    {
      "epoch": 0.13976377952755906,
      "grad_norm": 24.97527313232422,
      "learning_rate": 1.3717693836978134e-07,
      "loss": 2.3581,
      "step": 142
    },
    {
      "epoch": 0.140748031496063,
      "grad_norm": 22.391145706176758,
      "learning_rate": 1.381709741550696e-07,
      "loss": 1.4236,
      "step": 143
    },
    {
      "epoch": 0.14173228346456693,
      "grad_norm": 22.977148056030273,
      "learning_rate": 1.3916500994035787e-07,
      "loss": 2.202,
      "step": 144
    },
    {
      "epoch": 0.14271653543307086,
      "grad_norm": 25.096189498901367,
      "learning_rate": 1.4015904572564615e-07,
      "loss": 1.7654,
      "step": 145
    },
    {
      "epoch": 0.1437007874015748,
      "grad_norm": 28.528640747070312,
      "learning_rate": 1.4115308151093443e-07,
      "loss": 1.5748,
      "step": 146
    },
    {
      "epoch": 0.14468503937007873,
      "grad_norm": 30.80350685119629,
      "learning_rate": 1.421471172962227e-07,
      "loss": 1.7996,
      "step": 147
    },
    {
      "epoch": 0.14566929133858267,
      "grad_norm": 22.726850509643555,
      "learning_rate": 1.4314115308151096e-07,
      "loss": 1.7517,
      "step": 148
    },
    {
      "epoch": 0.1466535433070866,
      "grad_norm": 24.926170349121094,
      "learning_rate": 1.4413518886679924e-07,
      "loss": 1.8933,
      "step": 149
    },
    {
      "epoch": 0.14763779527559054,
      "grad_norm": 20.570375442504883,
      "learning_rate": 1.451292246520875e-07,
      "loss": 1.2836,
      "step": 150
    },
    {
      "epoch": 0.1486220472440945,
      "grad_norm": 21.08650779724121,
      "learning_rate": 1.4612326043737577e-07,
      "loss": 1.7145,
      "step": 151
    },
    {
      "epoch": 0.14960629921259844,
      "grad_norm": 24.37209701538086,
      "learning_rate": 1.4711729622266402e-07,
      "loss": 1.6499,
      "step": 152
    },
    {
      "epoch": 0.15059055118110237,
      "grad_norm": 26.801511764526367,
      "learning_rate": 1.4811133200795232e-07,
      "loss": 1.8273,
      "step": 153
    },
    {
      "epoch": 0.15059055118110237,
      "eval_Qnli-dev_cosine_accuracy": 0.646484375,
      "eval_Qnli-dev_cosine_accuracy_threshold": 0.8057259321212769,
      "eval_Qnli-dev_cosine_ap": 0.6720663622193426,
      "eval_Qnli-dev_cosine_f1": 0.6688102893890675,
      "eval_Qnli-dev_cosine_f1_threshold": 0.7187118530273438,
      "eval_Qnli-dev_cosine_precision": 0.538860103626943,
      "eval_Qnli-dev_cosine_recall": 0.8813559322033898,
      "eval_Qnli-dev_dot_accuracy": 0.646484375,
      "eval_Qnli-dev_dot_accuracy_threshold": 618.8643798828125,
      "eval_Qnli-dev_dot_ap": 0.672083506527328,
      "eval_Qnli-dev_dot_f1": 0.6688102893890675,
      "eval_Qnli-dev_dot_f1_threshold": 552.0260009765625,
      "eval_Qnli-dev_dot_precision": 0.538860103626943,
      "eval_Qnli-dev_dot_recall": 0.8813559322033898,
      "eval_Qnli-dev_euclidean_accuracy": 0.646484375,
      "eval_Qnli-dev_euclidean_accuracy_threshold": 17.27533721923828,
      "eval_Qnli-dev_euclidean_ap": 0.6720591998758361,
      "eval_Qnli-dev_euclidean_f1": 0.6688102893890675,
      "eval_Qnli-dev_euclidean_f1_threshold": 20.787063598632812,
      "eval_Qnli-dev_euclidean_precision": 0.538860103626943,
      "eval_Qnli-dev_euclidean_recall": 0.8813559322033898,
      "eval_Qnli-dev_manhattan_accuracy": 0.6484375,
      "eval_Qnli-dev_manhattan_accuracy_threshold": 386.58905029296875,
      "eval_Qnli-dev_manhattan_ap": 0.6724653688821339,
      "eval_Qnli-dev_manhattan_f1": 0.6645569620253164,
      "eval_Qnli-dev_manhattan_f1_threshold": 462.609130859375,
      "eval_Qnli-dev_manhattan_precision": 0.5303030303030303,
      "eval_Qnli-dev_manhattan_recall": 0.8898305084745762,
      "eval_Qnli-dev_max_accuracy": 0.6484375,
      "eval_Qnli-dev_max_accuracy_threshold": 618.8643798828125,
      "eval_Qnli-dev_max_ap": 0.6724653688821339,
      "eval_Qnli-dev_max_f1": 0.6688102893890675,
      "eval_Qnli-dev_max_f1_threshold": 552.0260009765625,
      "eval_Qnli-dev_max_precision": 0.538860103626943,
      "eval_Qnli-dev_max_recall": 0.8898305084745762,
      "eval_allNLI-dev_cosine_accuracy": 0.67578125,
      "eval_allNLI-dev_cosine_accuracy_threshold": 0.9427558183670044,
      "eval_allNLI-dev_cosine_ap": 0.4368751759846574,
      "eval_allNLI-dev_cosine_f1": 0.5225225225225225,
      "eval_allNLI-dev_cosine_f1_threshold": 0.8046966791152954,
      "eval_allNLI-dev_cosine_precision": 0.3795811518324607,
      "eval_allNLI-dev_cosine_recall": 0.838150289017341,
      "eval_allNLI-dev_dot_accuracy": 0.67578125,
      "eval_allNLI-dev_dot_accuracy_threshold": 724.1080322265625,
      "eval_allNLI-dev_dot_ap": 0.436842886797982,
      "eval_allNLI-dev_dot_f1": 0.5225225225225225,
      "eval_allNLI-dev_dot_f1_threshold": 618.074951171875,
      "eval_allNLI-dev_dot_precision": 0.3795811518324607,
      "eval_allNLI-dev_dot_recall": 0.838150289017341,
      "eval_allNLI-dev_euclidean_accuracy": 0.67578125,
      "eval_allNLI-dev_euclidean_accuracy_threshold": 9.377331733703613,
      "eval_allNLI-dev_euclidean_ap": 0.4368602200677977,
      "eval_allNLI-dev_euclidean_f1": 0.5225225225225225,
      "eval_allNLI-dev_euclidean_f1_threshold": 17.321048736572266,
      "eval_allNLI-dev_euclidean_precision": 0.3795811518324607,
      "eval_allNLI-dev_euclidean_recall": 0.838150289017341,
      "eval_allNLI-dev_manhattan_accuracy": 0.677734375,
      "eval_allNLI-dev_manhattan_accuracy_threshold": 223.6764373779297,
      "eval_allNLI-dev_manhattan_ap": 0.43892484929307635,
      "eval_allNLI-dev_manhattan_f1": 0.5239852398523985,
      "eval_allNLI-dev_manhattan_f1_threshold": 372.31396484375,
      "eval_allNLI-dev_manhattan_precision": 0.38482384823848237,
      "eval_allNLI-dev_manhattan_recall": 0.8208092485549133,
      "eval_allNLI-dev_max_accuracy": 0.677734375,
      "eval_allNLI-dev_max_accuracy_threshold": 724.1080322265625,
      "eval_allNLI-dev_max_ap": 0.43892484929307635,
      "eval_allNLI-dev_max_f1": 0.5239852398523985,
      "eval_allNLI-dev_max_f1_threshold": 618.074951171875,
      "eval_allNLI-dev_max_precision": 0.38482384823848237,
      "eval_allNLI-dev_max_recall": 0.838150289017341,
      "eval_runtime": 1.2344,
      "eval_samples_per_second": 0.0,
      "eval_sequential_score": 0.6724653688821339,
      "eval_steps_per_second": 0.0,
      "eval_sts-test_pearson_cosine": 0.3774946012125992,
      "eval_sts-test_pearson_dot": 0.3774648453085433,
      "eval_sts-test_pearson_euclidean": 0.38652243004790016,
      "eval_sts-test_pearson_manhattan": 0.3861982631744407,
      "eval_sts-test_pearson_max": 0.38652243004790016,
      "eval_sts-test_spearman_cosine": 0.4056589966976888,
      "eval_sts-test_spearman_dot": 0.40563469676275316,
      "eval_sts-test_spearman_euclidean": 0.4056589966976888,
      "eval_sts-test_spearman_manhattan": 0.4059364545183154,
      "eval_sts-test_spearman_max": 0.4059364545183154,
      "step": 153
    },
    {
      "epoch": 0.1515748031496063,
      "grad_norm": 27.144521713256836,
      "learning_rate": 1.4910536779324058e-07,
      "loss": 2.2859,
      "step": 154
    },
    {
      "epoch": 0.15255905511811024,
      "grad_norm": 16.310321807861328,
      "learning_rate": 1.5009940357852886e-07,
      "loss": 1.0833,
      "step": 155
    },
    {
      "epoch": 0.15354330708661418,
      "grad_norm": 18.788604736328125,
      "learning_rate": 1.510934393638171e-07,
      "loss": 1.6829,
      "step": 156
    },
    {
      "epoch": 0.1545275590551181,
      "grad_norm": 23.34528923034668,
      "learning_rate": 1.5208747514910539e-07,
      "loss": 2.1464,
      "step": 157
    },
    {
      "epoch": 0.15551181102362205,
      "grad_norm": 19.06661033630371,
      "learning_rate": 1.5308151093439367e-07,
      "loss": 1.745,
      "step": 158
    },
    {
      "epoch": 0.15649606299212598,
      "grad_norm": 23.650297164916992,
      "learning_rate": 1.5407554671968192e-07,
      "loss": 1.7319,
      "step": 159
    },
    {
      "epoch": 0.15748031496062992,
      "grad_norm": 22.01628303527832,
      "learning_rate": 1.5506958250497022e-07,
      "loss": 1.6968,
      "step": 160
    },
    {
      "epoch": 0.15846456692913385,
      "grad_norm": 20.905855178833008,
      "learning_rate": 1.5606361829025848e-07,
      "loss": 1.7401,
      "step": 161
    },
    {
      "epoch": 0.1594488188976378,
      "grad_norm": 22.40980339050293,
      "learning_rate": 1.5705765407554675e-07,
      "loss": 1.729,
      "step": 162
    },
    {
      "epoch": 0.16043307086614172,
      "grad_norm": 23.574447631835938,
      "learning_rate": 1.58051689860835e-07,
      "loss": 2.0782,
      "step": 163
    },
    {
      "epoch": 0.16141732283464566,
      "grad_norm": 28.43741798400879,
      "learning_rate": 1.5904572564612329e-07,
      "loss": 2.6545,
      "step": 164
    },
    {
      "epoch": 0.1624015748031496,
      "grad_norm": 18.536827087402344,
      "learning_rate": 1.6003976143141154e-07,
      "loss": 1.4045,
      "step": 165
    },
    {
      "epoch": 0.16338582677165353,
      "grad_norm": 17.013965606689453,
      "learning_rate": 1.6103379721669984e-07,
      "loss": 1.2937,
      "step": 166
    },
    {
      "epoch": 0.1643700787401575,
      "grad_norm": 16.021455764770508,
      "learning_rate": 1.620278330019881e-07,
      "loss": 1.1171,
      "step": 167
    },
    {
      "epoch": 0.16535433070866143,
      "grad_norm": 20.248655319213867,
      "learning_rate": 1.6302186878727637e-07,
      "loss": 1.3537,
      "step": 168
    },
    {
      "epoch": 0.16633858267716536,
      "grad_norm": 21.731477737426758,
      "learning_rate": 1.6401590457256465e-07,
      "loss": 1.7028,
      "step": 169
    },
    {
      "epoch": 0.1673228346456693,
      "grad_norm": 20.372398376464844,
      "learning_rate": 1.650099403578529e-07,
      "loss": 1.4143,
      "step": 170
    },
    {
      "epoch": 0.16830708661417323,
      "grad_norm": 21.89658546447754,
      "learning_rate": 1.6600397614314118e-07,
      "loss": 1.8648,
      "step": 171
    },
    {
      "epoch": 0.16929133858267717,
      "grad_norm": 24.402984619140625,
      "learning_rate": 1.6699801192842944e-07,
      "loss": 1.6768,
      "step": 172
    },
    {
      "epoch": 0.1702755905511811,
      "grad_norm": 28.650039672851562,
      "learning_rate": 1.6799204771371774e-07,
      "loss": 1.9528,
      "step": 173
    },
    {
      "epoch": 0.17125984251968504,
      "grad_norm": 18.586729049682617,
      "learning_rate": 1.68986083499006e-07,
      "loss": 1.1718,
      "step": 174
    },
    {
      "epoch": 0.17224409448818898,
      "grad_norm": 22.66750717163086,
      "learning_rate": 1.6998011928429427e-07,
      "loss": 1.8176,
      "step": 175
    },
    {
      "epoch": 0.1732283464566929,
      "grad_norm": 15.157270431518555,
      "learning_rate": 1.7097415506958253e-07,
      "loss": 0.8439,
      "step": 176
    },
    {
      "epoch": 0.17421259842519685,
      "grad_norm": 23.08325958251953,
      "learning_rate": 1.719681908548708e-07,
      "loss": 1.5092,
      "step": 177
    },
    {
      "epoch": 0.17519685039370078,
      "grad_norm": 19.215696334838867,
      "learning_rate": 1.7296222664015906e-07,
      "loss": 1.1947,
      "step": 178
    },
    {
      "epoch": 0.17618110236220472,
      "grad_norm": 23.268930435180664,
      "learning_rate": 1.7395626242544734e-07,
      "loss": 1.6395,
      "step": 179
    },
    {
      "epoch": 0.17716535433070865,
      "grad_norm": 24.91169548034668,
      "learning_rate": 1.7495029821073564e-07,
      "loss": 1.4394,
      "step": 180
    },
    {
      "epoch": 0.1781496062992126,
      "grad_norm": 26.92424201965332,
      "learning_rate": 1.759443339960239e-07,
      "loss": 1.7548,
      "step": 181
    },
    {
      "epoch": 0.17913385826771652,
      "grad_norm": 22.831315994262695,
      "learning_rate": 1.7693836978131217e-07,
      "loss": 1.1181,
      "step": 182
    },
    {
      "epoch": 0.18011811023622049,
      "grad_norm": 18.39922332763672,
      "learning_rate": 1.7793240556660042e-07,
      "loss": 1.0271,
      "step": 183
    },
    {
      "epoch": 0.18110236220472442,
      "grad_norm": 30.0064697265625,
      "learning_rate": 1.789264413518887e-07,
      "loss": 2.3108,
      "step": 184
    },
    {
      "epoch": 0.18208661417322836,
      "grad_norm": 25.531293869018555,
      "learning_rate": 1.7992047713717695e-07,
      "loss": 2.1242,
      "step": 185
    },
    {
      "epoch": 0.1830708661417323,
      "grad_norm": 25.3813533782959,
      "learning_rate": 1.8091451292246523e-07,
      "loss": 1.9822,
      "step": 186
    },
    {
      "epoch": 0.18405511811023623,
      "grad_norm": 22.24681282043457,
      "learning_rate": 1.819085487077535e-07,
      "loss": 2.3605,
      "step": 187
    },
    {
      "epoch": 0.18503937007874016,
      "grad_norm": 19.291316986083984,
      "learning_rate": 1.829025844930418e-07,
      "loss": 1.5251,
      "step": 188
    },
    {
      "epoch": 0.1860236220472441,
      "grad_norm": 18.280820846557617,
      "learning_rate": 1.8389662027833004e-07,
      "loss": 1.2351,
      "step": 189
    },
    {
      "epoch": 0.18700787401574803,
      "grad_norm": 18.341115951538086,
      "learning_rate": 1.8489065606361832e-07,
      "loss": 1.5859,
      "step": 190
    },
    {
      "epoch": 0.18799212598425197,
      "grad_norm": 21.981475830078125,
      "learning_rate": 1.8588469184890657e-07,
      "loss": 1.8056,
      "step": 191
    },
    {
      "epoch": 0.1889763779527559,
      "grad_norm": 23.482725143432617,
      "learning_rate": 1.8687872763419485e-07,
      "loss": 1.349,
      "step": 192
    },
    {
      "epoch": 0.18996062992125984,
      "grad_norm": 15.451715469360352,
      "learning_rate": 1.8787276341948313e-07,
      "loss": 0.893,
      "step": 193
    },
    {
      "epoch": 0.19094488188976377,
      "grad_norm": 25.24738121032715,
      "learning_rate": 1.888667992047714e-07,
      "loss": 1.5122,
      "step": 194
    },
    {
      "epoch": 0.1919291338582677,
      "grad_norm": 22.348981857299805,
      "learning_rate": 1.898608349900597e-07,
      "loss": 1.3875,
      "step": 195
    },
    {
      "epoch": 0.19291338582677164,
      "grad_norm": 22.770132064819336,
      "learning_rate": 1.9085487077534794e-07,
      "loss": 1.29,
      "step": 196
    },
    {
      "epoch": 0.19389763779527558,
      "grad_norm": 31.219194412231445,
      "learning_rate": 1.9184890656063622e-07,
      "loss": 2.2931,
      "step": 197
    },
    {
      "epoch": 0.19488188976377951,
      "grad_norm": 24.657859802246094,
      "learning_rate": 1.9284294234592447e-07,
      "loss": 1.2663,
      "step": 198
    },
    {
      "epoch": 0.19586614173228348,
      "grad_norm": 26.825185775756836,
      "learning_rate": 1.9383697813121275e-07,
      "loss": 1.9712,
      "step": 199
    },
    {
      "epoch": 0.1968503937007874,
      "grad_norm": 30.97873306274414,
      "learning_rate": 1.94831013916501e-07,
      "loss": 2.3307,
      "step": 200
    },
    {
      "epoch": 0.19783464566929135,
      "grad_norm": 24.922365188598633,
      "learning_rate": 1.958250497017893e-07,
      "loss": 1.6544,
      "step": 201
    },
    {
      "epoch": 0.19881889763779528,
      "grad_norm": 22.842063903808594,
      "learning_rate": 1.9681908548707756e-07,
      "loss": 1.638,
      "step": 202
    },
    {
      "epoch": 0.19980314960629922,
      "grad_norm": 17.376962661743164,
      "learning_rate": 1.9781312127236584e-07,
      "loss": 1.3412,
      "step": 203
    },
    {
      "epoch": 0.20078740157480315,
      "grad_norm": 17.828535079956055,
      "learning_rate": 1.9880715705765412e-07,
      "loss": 1.4454,
      "step": 204
    },
    {
      "epoch": 0.2017716535433071,
      "grad_norm": 18.73072624206543,
      "learning_rate": 1.9980119284294237e-07,
      "loss": 1.5437,
      "step": 205
    },
    {
      "epoch": 0.20275590551181102,
      "grad_norm": 22.05230140686035,
      "learning_rate": 2.0079522862823065e-07,
      "loss": 1.4921,
      "step": 206
    },
    {
      "epoch": 0.20374015748031496,
      "grad_norm": 18.090892791748047,
      "learning_rate": 2.017892644135189e-07,
      "loss": 1.4298,
      "step": 207
    },
    {
      "epoch": 0.2047244094488189,
      "grad_norm": 22.94427490234375,
      "learning_rate": 2.027833001988072e-07,
      "loss": 1.6174,
      "step": 208
    },
    {
      "epoch": 0.20570866141732283,
      "grad_norm": 20.903331756591797,
      "learning_rate": 2.0377733598409546e-07,
      "loss": 1.4137,
      "step": 209
    },
    {
      "epoch": 0.20669291338582677,
      "grad_norm": 22.67809295654297,
      "learning_rate": 2.0477137176938374e-07,
      "loss": 1.5652,
      "step": 210
    },
    {
      "epoch": 0.2076771653543307,
      "grad_norm": 16.824064254760742,
      "learning_rate": 2.05765407554672e-07,
      "loss": 1.1631,
      "step": 211
    },
    {
      "epoch": 0.20866141732283464,
      "grad_norm": 21.46413230895996,
      "learning_rate": 2.0675944333996027e-07,
      "loss": 1.2351,
      "step": 212
    },
    {
      "epoch": 0.20964566929133857,
      "grad_norm": 22.461610794067383,
      "learning_rate": 2.0775347912524852e-07,
      "loss": 1.7537,
      "step": 213
    },
    {
      "epoch": 0.2106299212598425,
      "grad_norm": 22.150938034057617,
      "learning_rate": 2.087475149105368e-07,
      "loss": 1.3186,
      "step": 214
    },
    {
      "epoch": 0.21161417322834647,
      "grad_norm": 22.82622718811035,
      "learning_rate": 2.097415506958251e-07,
      "loss": 1.2258,
      "step": 215
    },
    {
      "epoch": 0.2125984251968504,
      "grad_norm": 13.514169692993164,
      "learning_rate": 2.1073558648111336e-07,
      "loss": 0.7695,
      "step": 216
    },
    {
      "epoch": 0.21358267716535434,
      "grad_norm": 22.254596710205078,
      "learning_rate": 2.1172962226640164e-07,
      "loss": 1.2775,
      "step": 217
    },
    {
      "epoch": 0.21456692913385828,
      "grad_norm": 23.31028938293457,
      "learning_rate": 2.127236580516899e-07,
      "loss": 1.6795,
      "step": 218
    },
    {
      "epoch": 0.2155511811023622,
      "grad_norm": 24.057025909423828,
      "learning_rate": 2.1371769383697817e-07,
      "loss": 1.2862,
      "step": 219
    },
    {
      "epoch": 0.21653543307086615,
      "grad_norm": 17.363821029663086,
      "learning_rate": 2.1471172962226642e-07,
      "loss": 1.1723,
      "step": 220
    },
    {
      "epoch": 0.21751968503937008,
      "grad_norm": 22.69175148010254,
      "learning_rate": 2.1570576540755473e-07,
      "loss": 1.3322,
      "step": 221
    },
    {
      "epoch": 0.21850393700787402,
      "grad_norm": 24.083465576171875,
      "learning_rate": 2.1669980119284298e-07,
      "loss": 1.7564,
      "step": 222
    },
    {
      "epoch": 0.21948818897637795,
      "grad_norm": 17.775938034057617,
      "learning_rate": 2.1769383697813126e-07,
      "loss": 1.1071,
      "step": 223
    },
    {
      "epoch": 0.2204724409448819,
      "grad_norm": 18.539440155029297,
      "learning_rate": 2.186878727634195e-07,
      "loss": 1.2011,
      "step": 224
    },
    {
      "epoch": 0.22145669291338582,
      "grad_norm": 18.234455108642578,
      "learning_rate": 2.196819085487078e-07,
      "loss": 1.2303,
      "step": 225
    },
    {
      "epoch": 0.22244094488188976,
      "grad_norm": 20.291522979736328,
      "learning_rate": 2.2067594433399604e-07,
      "loss": 1.212,
      "step": 226
    },
    {
      "epoch": 0.2234251968503937,
      "grad_norm": 15.60973072052002,
      "learning_rate": 2.2166998011928432e-07,
      "loss": 1.0117,
      "step": 227
    },
    {
      "epoch": 0.22440944881889763,
      "grad_norm": 20.482995986938477,
      "learning_rate": 2.2266401590457263e-07,
      "loss": 1.1907,
      "step": 228
    },
    {
      "epoch": 0.22539370078740156,
      "grad_norm": 24.126340866088867,
      "learning_rate": 2.2365805168986088e-07,
      "loss": 2.1293,
      "step": 229
    },
    {
      "epoch": 0.2263779527559055,
      "grad_norm": 21.279150009155273,
      "learning_rate": 2.2465208747514916e-07,
      "loss": 1.3063,
      "step": 230
    },
    {
      "epoch": 0.22736220472440946,
      "grad_norm": 20.62582778930664,
      "learning_rate": 2.256461232604374e-07,
      "loss": 1.2841,
      "step": 231
    },
    {
      "epoch": 0.2283464566929134,
      "grad_norm": 20.63606834411621,
      "learning_rate": 2.266401590457257e-07,
      "loss": 1.3778,
      "step": 232
    },
    {
      "epoch": 0.22933070866141733,
      "grad_norm": 18.099950790405273,
      "learning_rate": 2.2763419483101394e-07,
      "loss": 1.2242,
      "step": 233
    },
    {
      "epoch": 0.23031496062992127,
      "grad_norm": 18.96858024597168,
      "learning_rate": 2.2862823061630222e-07,
      "loss": 0.9227,
      "step": 234
    },
    {
      "epoch": 0.2312992125984252,
      "grad_norm": 22.201169967651367,
      "learning_rate": 2.296222664015905e-07,
      "loss": 1.2221,
      "step": 235
    },
    {
      "epoch": 0.23228346456692914,
      "grad_norm": 26.657480239868164,
      "learning_rate": 2.3061630218687878e-07,
      "loss": 2.1041,
      "step": 236
    },
    {
      "epoch": 0.23326771653543307,
      "grad_norm": 25.558757781982422,
      "learning_rate": 2.3161033797216703e-07,
      "loss": 1.3341,
      "step": 237
    },
    {
      "epoch": 0.234251968503937,
      "grad_norm": 22.0543212890625,
      "learning_rate": 2.326043737574553e-07,
      "loss": 1.0876,
      "step": 238
    },
    {
      "epoch": 0.23523622047244094,
      "grad_norm": 22.258333206176758,
      "learning_rate": 2.335984095427436e-07,
      "loss": 1.3328,
      "step": 239
    },
    {
      "epoch": 0.23622047244094488,
      "grad_norm": 21.519908905029297,
      "learning_rate": 2.3459244532803184e-07,
      "loss": 1.2958,
      "step": 240
    },
    {
      "epoch": 0.2372047244094488,
      "grad_norm": 23.345388412475586,
      "learning_rate": 2.3558648111332012e-07,
      "loss": 1.1522,
      "step": 241
    },
    {
      "epoch": 0.23818897637795275,
      "grad_norm": 25.306140899658203,
      "learning_rate": 2.365805168986084e-07,
      "loss": 1.7942,
      "step": 242
    },
    {
      "epoch": 0.23917322834645668,
      "grad_norm": 21.840364456176758,
      "learning_rate": 2.3757455268389668e-07,
      "loss": 1.1325,
      "step": 243
    },
    {
      "epoch": 0.24015748031496062,
      "grad_norm": 26.5495662689209,
      "learning_rate": 2.385685884691849e-07,
      "loss": 1.6466,
      "step": 244
    },
    {
      "epoch": 0.24114173228346455,
      "grad_norm": 21.928937911987305,
      "learning_rate": 2.395626242544732e-07,
      "loss": 1.4608,
      "step": 245
    },
    {
      "epoch": 0.2421259842519685,
      "grad_norm": 16.23383903503418,
      "learning_rate": 2.4055666003976146e-07,
      "loss": 0.6375,
      "step": 246
    },
    {
      "epoch": 0.24311023622047245,
      "grad_norm": 26.922739028930664,
      "learning_rate": 2.4155069582504976e-07,
      "loss": 2.0177,
      "step": 247
    },
    {
      "epoch": 0.2440944881889764,
      "grad_norm": 23.482742309570312,
      "learning_rate": 2.42544731610338e-07,
      "loss": 1.2069,
      "step": 248
    },
    {
      "epoch": 0.24507874015748032,
      "grad_norm": 14.740228652954102,
      "learning_rate": 2.4353876739562627e-07,
      "loss": 0.7639,
      "step": 249
    },
    {
      "epoch": 0.24606299212598426,
      "grad_norm": 20.238948822021484,
      "learning_rate": 2.445328031809146e-07,
      "loss": 1.3465,
      "step": 250
    },
    {
      "epoch": 0.2470472440944882,
      "grad_norm": 16.91303253173828,
      "learning_rate": 2.4552683896620283e-07,
      "loss": 1.064,
      "step": 251
    },
    {
      "epoch": 0.24803149606299213,
      "grad_norm": 22.081308364868164,
      "learning_rate": 2.4652087475149113e-07,
      "loss": 1.3757,
      "step": 252
    },
    {
      "epoch": 0.24901574803149606,
      "grad_norm": 21.756267547607422,
      "learning_rate": 2.475149105367794e-07,
      "loss": 1.612,
      "step": 253
    },
    {
      "epoch": 0.25,
      "grad_norm": 16.87566566467285,
      "learning_rate": 2.4850894632206764e-07,
      "loss": 0.7917,
      "step": 254
    },
    {
      "epoch": 0.25098425196850394,
      "grad_norm": 22.13374900817871,
      "learning_rate": 2.495029821073559e-07,
      "loss": 1.5515,
      "step": 255
    },
    {
      "epoch": 0.25196850393700787,
      "grad_norm": 18.96138572692871,
      "learning_rate": 2.5049701789264414e-07,
      "loss": 0.799,
      "step": 256
    },
    {
      "epoch": 0.2529527559055118,
      "grad_norm": 17.10991859436035,
      "learning_rate": 2.5149105367793245e-07,
      "loss": 0.9882,
      "step": 257
    },
    {
      "epoch": 0.25393700787401574,
      "grad_norm": 15.347769737243652,
      "learning_rate": 2.524850894632207e-07,
      "loss": 1.1814,
      "step": 258
    },
    {
      "epoch": 0.2549212598425197,
      "grad_norm": 11.857462882995605,
      "learning_rate": 2.53479125248509e-07,
      "loss": 0.6394,
      "step": 259
    },
    {
      "epoch": 0.2559055118110236,
      "grad_norm": 20.00450325012207,
      "learning_rate": 2.5447316103379726e-07,
      "loss": 1.4756,
      "step": 260
    },
    {
      "epoch": 0.25688976377952755,
      "grad_norm": 12.062335968017578,
      "learning_rate": 2.554671968190855e-07,
      "loss": 0.5338,
      "step": 261
    },
    {
      "epoch": 0.2578740157480315,
      "grad_norm": 16.912168502807617,
      "learning_rate": 2.564612326043738e-07,
      "loss": 0.9779,
      "step": 262
    },
    {
      "epoch": 0.2588582677165354,
      "grad_norm": 22.713825225830078,
      "learning_rate": 2.5745526838966207e-07,
      "loss": 1.5307,
      "step": 263
    },
    {
      "epoch": 0.25984251968503935,
      "grad_norm": 22.88834571838379,
      "learning_rate": 2.5844930417495037e-07,
      "loss": 1.1213,
      "step": 264
    },
    {
      "epoch": 0.2608267716535433,
      "grad_norm": 18.133310317993164,
      "learning_rate": 2.5944333996023857e-07,
      "loss": 0.9482,
      "step": 265
    },
    {
      "epoch": 0.2618110236220472,
      "grad_norm": 19.092018127441406,
      "learning_rate": 2.604373757455269e-07,
      "loss": 0.9599,
      "step": 266
    },
    {
      "epoch": 0.26279527559055116,
      "grad_norm": 25.283647537231445,
      "learning_rate": 2.614314115308152e-07,
      "loss": 1.4455,
      "step": 267
    },
    {
      "epoch": 0.2637795275590551,
      "grad_norm": 28.519742965698242,
      "learning_rate": 2.6242544731610343e-07,
      "loss": 1.6496,
      "step": 268
    },
    {
      "epoch": 0.26476377952755903,
      "grad_norm": 15.56734561920166,
      "learning_rate": 2.634194831013917e-07,
      "loss": 0.7402,
      "step": 269
    },
    {
      "epoch": 0.265748031496063,
      "grad_norm": 16.450769424438477,
      "learning_rate": 2.6441351888667994e-07,
      "loss": 0.7835,
      "step": 270
    },
    {
      "epoch": 0.26673228346456695,
      "grad_norm": 18.447402954101562,
      "learning_rate": 2.6540755467196824e-07,
      "loss": 0.7821,
      "step": 271
    },
    {
      "epoch": 0.2677165354330709,
      "grad_norm": 22.634441375732422,
      "learning_rate": 2.664015904572565e-07,
      "loss": 1.5422,
      "step": 272
    },
    {
      "epoch": 0.2687007874015748,
      "grad_norm": 20.091428756713867,
      "learning_rate": 2.6739562624254475e-07,
      "loss": 1.0995,
      "step": 273
    },
    {
      "epoch": 0.26968503937007876,
      "grad_norm": 22.89579200744629,
      "learning_rate": 2.6838966202783305e-07,
      "loss": 1.378,
      "step": 274
    },
    {
      "epoch": 0.2706692913385827,
      "grad_norm": 25.36764144897461,
      "learning_rate": 2.693836978131213e-07,
      "loss": 1.3562,
      "step": 275
    },
    {
      "epoch": 0.27165354330708663,
      "grad_norm": 15.717765808105469,
      "learning_rate": 2.703777335984096e-07,
      "loss": 0.7376,
      "step": 276
    },
    {
      "epoch": 0.27263779527559057,
      "grad_norm": 21.739999771118164,
      "learning_rate": 2.7137176938369786e-07,
      "loss": 1.1678,
      "step": 277
    },
    {
      "epoch": 0.2736220472440945,
      "grad_norm": 21.12672233581543,
      "learning_rate": 2.723658051689861e-07,
      "loss": 1.2989,
      "step": 278
    },
    {
      "epoch": 0.27460629921259844,
      "grad_norm": 22.614158630371094,
      "learning_rate": 2.7335984095427437e-07,
      "loss": 1.9559,
      "step": 279
    },
    {
      "epoch": 0.2755905511811024,
      "grad_norm": 18.742107391357422,
      "learning_rate": 2.743538767395627e-07,
      "loss": 1.1237,
      "step": 280
    },
    {
      "epoch": 0.2765748031496063,
      "grad_norm": 17.969554901123047,
      "learning_rate": 2.75347912524851e-07,
      "loss": 0.952,
      "step": 281
    },
    {
      "epoch": 0.27755905511811024,
      "grad_norm": 24.01414680480957,
      "learning_rate": 2.763419483101392e-07,
      "loss": 1.6629,
      "step": 282
    },
    {
      "epoch": 0.2785433070866142,
      "grad_norm": 21.362796783447266,
      "learning_rate": 2.773359840954275e-07,
      "loss": 1.871,
      "step": 283
    },
    {
      "epoch": 0.2795275590551181,
      "grad_norm": 19.05644989013672,
      "learning_rate": 2.7833001988071574e-07,
      "loss": 1.5946,
      "step": 284
    },
    {
      "epoch": 0.28051181102362205,
      "grad_norm": 18.40645408630371,
      "learning_rate": 2.7932405566600404e-07,
      "loss": 1.4456,
      "step": 285
    },
    {
      "epoch": 0.281496062992126,
      "grad_norm": 18.53467559814453,
      "learning_rate": 2.803180914512923e-07,
      "loss": 1.4085,
      "step": 286
    },
    {
      "epoch": 0.2824803149606299,
      "grad_norm": 18.1446533203125,
      "learning_rate": 2.8131212723658055e-07,
      "loss": 1.1394,
      "step": 287
    },
    {
      "epoch": 0.28346456692913385,
      "grad_norm": 15.109747886657715,
      "learning_rate": 2.8230616302186885e-07,
      "loss": 1.0315,
      "step": 288
    },
    {
      "epoch": 0.2844488188976378,
      "grad_norm": 22.63469696044922,
      "learning_rate": 2.833001988071571e-07,
      "loss": 1.488,
      "step": 289
    },
    {
      "epoch": 0.2854330708661417,
      "grad_norm": 18.841909408569336,
      "learning_rate": 2.842942345924454e-07,
      "loss": 1.4006,
      "step": 290
    },
    {
      "epoch": 0.28641732283464566,
      "grad_norm": 20.073745727539062,
      "learning_rate": 2.852882703777336e-07,
      "loss": 0.9237,
      "step": 291
    },
    {
      "epoch": 0.2874015748031496,
      "grad_norm": 18.586259841918945,
      "learning_rate": 2.862823061630219e-07,
      "loss": 1.163,
      "step": 292
    },
    {
      "epoch": 0.28838582677165353,
      "grad_norm": 23.599292755126953,
      "learning_rate": 2.8727634194831017e-07,
      "loss": 1.7037,
      "step": 293
    },
    {
      "epoch": 0.28937007874015747,
      "grad_norm": 16.384347915649414,
      "learning_rate": 2.8827037773359847e-07,
      "loss": 0.8715,
      "step": 294
    },
    {
      "epoch": 0.2903543307086614,
      "grad_norm": 20.26093101501465,
      "learning_rate": 2.892644135188867e-07,
      "loss": 1.2101,
      "step": 295
    },
    {
      "epoch": 0.29133858267716534,
      "grad_norm": 19.598831176757812,
      "learning_rate": 2.90258449304175e-07,
      "loss": 1.1179,
      "step": 296
    },
    {
      "epoch": 0.29232283464566927,
      "grad_norm": 22.65572738647461,
      "learning_rate": 2.912524850894633e-07,
      "loss": 1.3986,
      "step": 297
    },
    {
      "epoch": 0.2933070866141732,
      "grad_norm": 22.658931732177734,
      "learning_rate": 2.9224652087475153e-07,
      "loss": 1.7068,
      "step": 298
    },
    {
      "epoch": 0.29429133858267714,
      "grad_norm": 16.94671058654785,
      "learning_rate": 2.9324055666003984e-07,
      "loss": 0.8695,
      "step": 299
    },
    {
      "epoch": 0.2952755905511811,
      "grad_norm": 20.13766860961914,
      "learning_rate": 2.9423459244532804e-07,
      "loss": 1.3778,
      "step": 300
    },
    {
      "epoch": 0.296259842519685,
      "grad_norm": 16.939062118530273,
      "learning_rate": 2.9522862823061634e-07,
      "loss": 1.2834,
      "step": 301
    },
    {
      "epoch": 0.297244094488189,
      "grad_norm": 14.732866287231445,
      "learning_rate": 2.9622266401590465e-07,
      "loss": 0.8123,
      "step": 302
    },
    {
      "epoch": 0.29822834645669294,
      "grad_norm": 23.4341983795166,
      "learning_rate": 2.972166998011929e-07,
      "loss": 1.6521,
      "step": 303
    },
    {
      "epoch": 0.2992125984251969,
      "grad_norm": 15.455023765563965,
      "learning_rate": 2.9821073558648115e-07,
      "loss": 1.1064,
      "step": 304
    },
    {
      "epoch": 0.3001968503937008,
      "grad_norm": 15.683412551879883,
      "learning_rate": 2.992047713717694e-07,
      "loss": 0.9578,
      "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
}