{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.09620780886715305,
  "eval_steps": 500,
  "global_step": 300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0003206926962238435,
      "grad_norm": 2.7483954429626465,
      "learning_rate": 1e-05,
      "loss": 1.9483,
      "step": 1
    },
    {
      "epoch": 0.000641385392447687,
      "grad_norm": 2.690746307373047,
      "learning_rate": 2e-05,
      "loss": 1.7416,
      "step": 2
    },
    {
      "epoch": 0.0009620780886715305,
      "grad_norm": 2.7132160663604736,
      "learning_rate": 3e-05,
      "loss": 1.8029,
      "step": 3
    },
    {
      "epoch": 0.001282770784895374,
      "grad_norm": 2.7139906883239746,
      "learning_rate": 4e-05,
      "loss": 1.9292,
      "step": 4
    },
    {
      "epoch": 0.0016034634811192174,
      "grad_norm": 2.8549463748931885,
      "learning_rate": 5e-05,
      "loss": 1.8055,
      "step": 5
    },
    {
      "epoch": 0.001924156177343061,
      "grad_norm": 2.5766773223876953,
      "learning_rate": 6e-05,
      "loss": 2.042,
      "step": 6
    },
    {
      "epoch": 0.0022448488735669044,
      "grad_norm": 2.1945223808288574,
      "learning_rate": 7e-05,
      "loss": 1.8585,
      "step": 7
    },
    {
      "epoch": 0.002565541569790748,
      "grad_norm": 2.143160581588745,
      "learning_rate": 8e-05,
      "loss": 1.7026,
      "step": 8
    },
    {
      "epoch": 0.0028862342660145915,
      "grad_norm": 2.1353299617767334,
      "learning_rate": 9e-05,
      "loss": 1.7981,
      "step": 9
    },
    {
      "epoch": 0.003206926962238435,
      "grad_norm": 2.080528497695923,
      "learning_rate": 0.0001,
      "loss": 1.7712,
      "step": 10
    },
    {
      "epoch": 0.0035276196584622787,
      "grad_norm": 2.391787052154541,
      "learning_rate": 9.996782496782497e-05,
      "loss": 1.7666,
      "step": 11
    },
    {
      "epoch": 0.003848312354686122,
      "grad_norm": 2.2097723484039307,
      "learning_rate": 9.993564993564995e-05,
      "loss": 1.8056,
      "step": 12
    },
    {
      "epoch": 0.004169005050909965,
      "grad_norm": 1.883785367012024,
      "learning_rate": 9.99034749034749e-05,
      "loss": 1.5309,
      "step": 13
    },
    {
      "epoch": 0.004489697747133809,
      "grad_norm": 1.9296449422836304,
      "learning_rate": 9.987129987129988e-05,
      "loss": 1.5142,
      "step": 14
    },
    {
      "epoch": 0.004810390443357652,
      "grad_norm": 2.2633562088012695,
      "learning_rate": 9.983912483912484e-05,
      "loss": 1.6335,
      "step": 15
    },
    {
      "epoch": 0.005131083139581496,
      "grad_norm": 2.1033310890197754,
      "learning_rate": 9.98069498069498e-05,
      "loss": 1.7974,
      "step": 16
    },
    {
      "epoch": 0.00545177583580534,
      "grad_norm": 2.003077268600464,
      "learning_rate": 9.977477477477478e-05,
      "loss": 1.5864,
      "step": 17
    },
    {
      "epoch": 0.005772468532029183,
      "grad_norm": 2.1461055278778076,
      "learning_rate": 9.974259974259975e-05,
      "loss": 1.5678,
      "step": 18
    },
    {
      "epoch": 0.0060931612282530264,
      "grad_norm": 1.9527181386947632,
      "learning_rate": 9.971042471042471e-05,
      "loss": 1.8429,
      "step": 19
    },
    {
      "epoch": 0.00641385392447687,
      "grad_norm": 2.050208330154419,
      "learning_rate": 9.967824967824968e-05,
      "loss": 1.6551,
      "step": 20
    },
    {
      "epoch": 0.006734546620700713,
      "grad_norm": 1.841902256011963,
      "learning_rate": 9.964607464607466e-05,
      "loss": 1.8851,
      "step": 21
    },
    {
      "epoch": 0.007055239316924557,
      "grad_norm": 1.8649909496307373,
      "learning_rate": 9.961389961389962e-05,
      "loss": 1.7153,
      "step": 22
    },
    {
      "epoch": 0.007375932013148401,
      "grad_norm": 1.9829213619232178,
      "learning_rate": 9.958172458172458e-05,
      "loss": 1.7295,
      "step": 23
    },
    {
      "epoch": 0.007696624709372244,
      "grad_norm": 1.945168137550354,
      "learning_rate": 9.954954954954956e-05,
      "loss": 1.6252,
      "step": 24
    },
    {
      "epoch": 0.008017317405596087,
      "grad_norm": 1.7710901498794556,
      "learning_rate": 9.951737451737451e-05,
      "loss": 1.8558,
      "step": 25
    },
    {
      "epoch": 0.00833801010181993,
      "grad_norm": 1.8060033321380615,
      "learning_rate": 9.948519948519949e-05,
      "loss": 1.6409,
      "step": 26
    },
    {
      "epoch": 0.008658702798043774,
      "grad_norm": 1.737451195716858,
      "learning_rate": 9.945302445302446e-05,
      "loss": 1.6147,
      "step": 27
    },
    {
      "epoch": 0.008979395494267618,
      "grad_norm": 1.7312341928482056,
      "learning_rate": 9.942084942084942e-05,
      "loss": 1.6207,
      "step": 28
    },
    {
      "epoch": 0.009300088190491461,
      "grad_norm": 1.850573182106018,
      "learning_rate": 9.93886743886744e-05,
      "loss": 1.6429,
      "step": 29
    },
    {
      "epoch": 0.009620780886715304,
      "grad_norm": 1.892348051071167,
      "learning_rate": 9.935649935649936e-05,
      "loss": 1.7195,
      "step": 30
    },
    {
      "epoch": 0.00994147358293915,
      "grad_norm": 1.9182771444320679,
      "learning_rate": 9.932432432432433e-05,
      "loss": 1.6534,
      "step": 31
    },
    {
      "epoch": 0.010262166279162993,
      "grad_norm": 1.7795934677124023,
      "learning_rate": 9.929214929214929e-05,
      "loss": 1.5351,
      "step": 32
    },
    {
      "epoch": 0.010582858975386836,
      "grad_norm": 1.794394850730896,
      "learning_rate": 9.925997425997427e-05,
      "loss": 1.6685,
      "step": 33
    },
    {
      "epoch": 0.01090355167161068,
      "grad_norm": 1.784448504447937,
      "learning_rate": 9.922779922779923e-05,
      "loss": 1.6303,
      "step": 34
    },
    {
      "epoch": 0.011224244367834523,
      "grad_norm": 1.6641632318496704,
      "learning_rate": 9.91956241956242e-05,
      "loss": 1.4606,
      "step": 35
    },
    {
      "epoch": 0.011544937064058366,
      "grad_norm": 1.7565549612045288,
      "learning_rate": 9.916344916344918e-05,
      "loss": 1.7965,
      "step": 36
    },
    {
      "epoch": 0.01186562976028221,
      "grad_norm": 1.863232135772705,
      "learning_rate": 9.913127413127413e-05,
      "loss": 1.6193,
      "step": 37
    },
    {
      "epoch": 0.012186322456506053,
      "grad_norm": 1.9475373029708862,
      "learning_rate": 9.90990990990991e-05,
      "loss": 1.6691,
      "step": 38
    },
    {
      "epoch": 0.012507015152729896,
      "grad_norm": 1.7293490171432495,
      "learning_rate": 9.906692406692407e-05,
      "loss": 1.6587,
      "step": 39
    },
    {
      "epoch": 0.01282770784895374,
      "grad_norm": 1.7146735191345215,
      "learning_rate": 9.903474903474904e-05,
      "loss": 1.7512,
      "step": 40
    },
    {
      "epoch": 0.013148400545177583,
      "grad_norm": 1.6908844709396362,
      "learning_rate": 9.900257400257401e-05,
      "loss": 1.5004,
      "step": 41
    },
    {
      "epoch": 0.013469093241401426,
      "grad_norm": 2.0432796478271484,
      "learning_rate": 9.897039897039896e-05,
      "loss": 1.5675,
      "step": 42
    },
    {
      "epoch": 0.013789785937625271,
      "grad_norm": 1.9518095254898071,
      "learning_rate": 9.893822393822394e-05,
      "loss": 1.5885,
      "step": 43
    },
    {
      "epoch": 0.014110478633849115,
      "grad_norm": 2.044656753540039,
      "learning_rate": 9.89060489060489e-05,
      "loss": 1.914,
      "step": 44
    },
    {
      "epoch": 0.014431171330072958,
      "grad_norm": 1.8500516414642334,
      "learning_rate": 9.887387387387388e-05,
      "loss": 1.6865,
      "step": 45
    },
    {
      "epoch": 0.014751864026296801,
      "grad_norm": 1.7090437412261963,
      "learning_rate": 9.884169884169885e-05,
      "loss": 1.6582,
      "step": 46
    },
    {
      "epoch": 0.015072556722520645,
      "grad_norm": 1.6555441617965698,
      "learning_rate": 9.880952380952381e-05,
      "loss": 1.4794,
      "step": 47
    },
    {
      "epoch": 0.015393249418744488,
      "grad_norm": 1.6648783683776855,
      "learning_rate": 9.877734877734878e-05,
      "loss": 1.2951,
      "step": 48
    },
    {
      "epoch": 0.01571394211496833,
      "grad_norm": 1.7010596990585327,
      "learning_rate": 9.874517374517374e-05,
      "loss": 1.5657,
      "step": 49
    },
    {
      "epoch": 0.016034634811192175,
      "grad_norm": 1.7453527450561523,
      "learning_rate": 9.871299871299872e-05,
      "loss": 1.4699,
      "step": 50
    },
    {
      "epoch": 0.01635532750741602,
      "grad_norm": 1.7735224962234497,
      "learning_rate": 9.868082368082369e-05,
      "loss": 1.724,
      "step": 51
    },
    {
      "epoch": 0.01667602020363986,
      "grad_norm": 1.5949498414993286,
      "learning_rate": 9.864864864864865e-05,
      "loss": 1.6525,
      "step": 52
    },
    {
      "epoch": 0.016996712899863705,
      "grad_norm": 1.8592952489852905,
      "learning_rate": 9.861647361647363e-05,
      "loss": 1.6122,
      "step": 53
    },
    {
      "epoch": 0.01731740559608755,
      "grad_norm": 1.7868975400924683,
      "learning_rate": 9.858429858429858e-05,
      "loss": 1.5783,
      "step": 54
    },
    {
      "epoch": 0.01763809829231139,
      "grad_norm": 1.6725190877914429,
      "learning_rate": 9.855212355212356e-05,
      "loss": 1.6519,
      "step": 55
    },
    {
      "epoch": 0.017958790988535235,
      "grad_norm": 2.030893564224243,
      "learning_rate": 9.851994851994852e-05,
      "loss": 1.881,
      "step": 56
    },
    {
      "epoch": 0.01827948368475908,
      "grad_norm": 1.79916512966156,
      "learning_rate": 9.84877734877735e-05,
      "loss": 1.6116,
      "step": 57
    },
    {
      "epoch": 0.018600176380982922,
      "grad_norm": 1.7612411975860596,
      "learning_rate": 9.845559845559846e-05,
      "loss": 1.5393,
      "step": 58
    },
    {
      "epoch": 0.018920869077206765,
      "grad_norm": 1.8098998069763184,
      "learning_rate": 9.842342342342343e-05,
      "loss": 1.693,
      "step": 59
    },
    {
      "epoch": 0.01924156177343061,
      "grad_norm": 1.8984328508377075,
      "learning_rate": 9.839124839124839e-05,
      "loss": 1.6939,
      "step": 60
    },
    {
      "epoch": 0.019562254469654452,
      "grad_norm": 1.751279592514038,
      "learning_rate": 9.835907335907336e-05,
      "loss": 1.6109,
      "step": 61
    },
    {
      "epoch": 0.0198829471658783,
      "grad_norm": 1.8087176084518433,
      "learning_rate": 9.832689832689834e-05,
      "loss": 1.5514,
      "step": 62
    },
    {
      "epoch": 0.020203639862102142,
      "grad_norm": 1.7087894678115845,
      "learning_rate": 9.82947232947233e-05,
      "loss": 1.6691,
      "step": 63
    },
    {
      "epoch": 0.020524332558325985,
      "grad_norm": 1.963853120803833,
      "learning_rate": 9.826254826254826e-05,
      "loss": 1.7157,
      "step": 64
    },
    {
      "epoch": 0.02084502525454983,
      "grad_norm": 1.9152112007141113,
      "learning_rate": 9.823037323037324e-05,
      "loss": 1.5779,
      "step": 65
    },
    {
      "epoch": 0.021165717950773672,
      "grad_norm": 1.856195092201233,
      "learning_rate": 9.81981981981982e-05,
      "loss": 1.6279,
      "step": 66
    },
    {
      "epoch": 0.021486410646997516,
      "grad_norm": 1.8876991271972656,
      "learning_rate": 9.816602316602317e-05,
      "loss": 1.7734,
      "step": 67
    },
    {
      "epoch": 0.02180710334322136,
      "grad_norm": 1.8838751316070557,
      "learning_rate": 9.813384813384814e-05,
      "loss": 1.6114,
      "step": 68
    },
    {
      "epoch": 0.022127796039445202,
      "grad_norm": 1.553562045097351,
      "learning_rate": 9.810167310167311e-05,
      "loss": 1.4989,
      "step": 69
    },
    {
      "epoch": 0.022448488735669046,
      "grad_norm": 1.9303158521652222,
      "learning_rate": 9.806949806949808e-05,
      "loss": 1.7525,
      "step": 70
    },
    {
      "epoch": 0.02276918143189289,
      "grad_norm": 1.655938982963562,
      "learning_rate": 9.803732303732304e-05,
      "loss": 1.6089,
      "step": 71
    },
    {
      "epoch": 0.023089874128116732,
      "grad_norm": 2.0096657276153564,
      "learning_rate": 9.800514800514801e-05,
      "loss": 1.6957,
      "step": 72
    },
    {
      "epoch": 0.023410566824340576,
      "grad_norm": 1.8276057243347168,
      "learning_rate": 9.797297297297297e-05,
      "loss": 1.684,
      "step": 73
    },
    {
      "epoch": 0.02373125952056442,
      "grad_norm": 1.6198551654815674,
      "learning_rate": 9.794079794079795e-05,
      "loss": 1.5419,
      "step": 74
    },
    {
      "epoch": 0.024051952216788262,
      "grad_norm": 1.850451111793518,
      "learning_rate": 9.790862290862292e-05,
      "loss": 1.5769,
      "step": 75
    },
    {
      "epoch": 0.024372644913012106,
      "grad_norm": 1.8147132396697998,
      "learning_rate": 9.787644787644788e-05,
      "loss": 1.4616,
      "step": 76
    },
    {
      "epoch": 0.02469333760923595,
      "grad_norm": 1.6501047611236572,
      "learning_rate": 9.784427284427284e-05,
      "loss": 1.3938,
      "step": 77
    },
    {
      "epoch": 0.025014030305459792,
      "grad_norm": 1.7198457717895508,
      "learning_rate": 9.781209781209781e-05,
      "loss": 1.4913,
      "step": 78
    },
    {
      "epoch": 0.025334723001683636,
      "grad_norm": 1.8030346632003784,
      "learning_rate": 9.777992277992279e-05,
      "loss": 1.5601,
      "step": 79
    },
    {
      "epoch": 0.02565541569790748,
      "grad_norm": 1.6368941068649292,
      "learning_rate": 9.774774774774775e-05,
      "loss": 1.5337,
      "step": 80
    },
    {
      "epoch": 0.025976108394131323,
      "grad_norm": 1.6790492534637451,
      "learning_rate": 9.771557271557273e-05,
      "loss": 1.5556,
      "step": 81
    },
    {
      "epoch": 0.026296801090355166,
      "grad_norm": 1.7556215524673462,
      "learning_rate": 9.76833976833977e-05,
      "loss": 1.5403,
      "step": 82
    },
    {
      "epoch": 0.02661749378657901,
      "grad_norm": 1.6657801866531372,
      "learning_rate": 9.765122265122264e-05,
      "loss": 1.5023,
      "step": 83
    },
    {
      "epoch": 0.026938186482802853,
      "grad_norm": 1.8705981969833374,
      "learning_rate": 9.761904761904762e-05,
      "loss": 1.5976,
      "step": 84
    },
    {
      "epoch": 0.027258879179026696,
      "grad_norm": 1.8940606117248535,
      "learning_rate": 9.758687258687259e-05,
      "loss": 1.7891,
      "step": 85
    },
    {
      "epoch": 0.027579571875250543,
      "grad_norm": 1.5879243612289429,
      "learning_rate": 9.755469755469757e-05,
      "loss": 1.4251,
      "step": 86
    },
    {
      "epoch": 0.027900264571474386,
      "grad_norm": 1.5877267122268677,
      "learning_rate": 9.752252252252253e-05,
      "loss": 1.6625,
      "step": 87
    },
    {
      "epoch": 0.02822095726769823,
      "grad_norm": 1.6735244989395142,
      "learning_rate": 9.74903474903475e-05,
      "loss": 1.5938,
      "step": 88
    },
    {
      "epoch": 0.028541649963922073,
      "grad_norm": 1.6752715110778809,
      "learning_rate": 9.745817245817246e-05,
      "loss": 1.4759,
      "step": 89
    },
    {
      "epoch": 0.028862342660145916,
      "grad_norm": 1.8093688488006592,
      "learning_rate": 9.742599742599742e-05,
      "loss": 1.8628,
      "step": 90
    },
    {
      "epoch": 0.02918303535636976,
      "grad_norm": 1.822588324546814,
      "learning_rate": 9.73938223938224e-05,
      "loss": 1.6622,
      "step": 91
    },
    {
      "epoch": 0.029503728052593603,
      "grad_norm": 1.7710521221160889,
      "learning_rate": 9.736164736164737e-05,
      "loss": 1.4719,
      "step": 92
    },
    {
      "epoch": 0.029824420748817446,
      "grad_norm": 1.4404919147491455,
      "learning_rate": 9.732947232947234e-05,
      "loss": 1.1332,
      "step": 93
    },
    {
      "epoch": 0.03014511344504129,
      "grad_norm": 1.6102395057678223,
      "learning_rate": 9.729729729729731e-05,
      "loss": 1.4751,
      "step": 94
    },
    {
      "epoch": 0.030465806141265133,
      "grad_norm": 1.706247091293335,
      "learning_rate": 9.726512226512226e-05,
      "loss": 1.5941,
      "step": 95
    },
    {
      "epoch": 0.030786498837488976,
      "grad_norm": 1.6406103372573853,
      "learning_rate": 9.723294723294724e-05,
      "loss": 1.5951,
      "step": 96
    },
    {
      "epoch": 0.03110719153371282,
      "grad_norm": 1.6317566633224487,
      "learning_rate": 9.72007722007722e-05,
      "loss": 1.4737,
      "step": 97
    },
    {
      "epoch": 0.03142788422993666,
      "grad_norm": 1.8702404499053955,
      "learning_rate": 9.716859716859718e-05,
      "loss": 1.5181,
      "step": 98
    },
    {
      "epoch": 0.03174857692616051,
      "grad_norm": 1.725326657295227,
      "learning_rate": 9.713642213642214e-05,
      "loss": 1.6582,
      "step": 99
    },
    {
      "epoch": 0.03206926962238435,
      "grad_norm": 1.7484664916992188,
      "learning_rate": 9.710424710424711e-05,
      "loss": 1.6073,
      "step": 100
    },
    {
      "epoch": 0.0323899623186082,
      "grad_norm": 1.8844904899597168,
      "learning_rate": 9.707207207207207e-05,
      "loss": 1.6232,
      "step": 101
    },
    {
      "epoch": 0.03271065501483204,
      "grad_norm": 1.8485490083694458,
      "learning_rate": 9.703989703989704e-05,
      "loss": 1.5857,
      "step": 102
    },
    {
      "epoch": 0.03303134771105588,
      "grad_norm": 1.7108585834503174,
      "learning_rate": 9.700772200772202e-05,
      "loss": 1.7378,
      "step": 103
    },
    {
      "epoch": 0.03335204040727972,
      "grad_norm": 1.869952917098999,
      "learning_rate": 9.697554697554698e-05,
      "loss": 1.7696,
      "step": 104
    },
    {
      "epoch": 0.03367273310350357,
      "grad_norm": 1.6918983459472656,
      "learning_rate": 9.694337194337196e-05,
      "loss": 1.4724,
      "step": 105
    },
    {
      "epoch": 0.03399342579972741,
      "grad_norm": 1.903573989868164,
      "learning_rate": 9.691119691119691e-05,
      "loss": 1.7818,
      "step": 106
    },
    {
      "epoch": 0.03431411849595126,
      "grad_norm": 1.7081507444381714,
      "learning_rate": 9.687902187902187e-05,
      "loss": 1.5507,
      "step": 107
    },
    {
      "epoch": 0.0346348111921751,
      "grad_norm": 1.8181805610656738,
      "learning_rate": 9.684684684684685e-05,
      "loss": 1.6716,
      "step": 108
    },
    {
      "epoch": 0.034955503888398944,
      "grad_norm": 1.7593623399734497,
      "learning_rate": 9.681467181467182e-05,
      "loss": 1.6878,
      "step": 109
    },
    {
      "epoch": 0.03527619658462278,
      "grad_norm": 1.6997793912887573,
      "learning_rate": 9.67824967824968e-05,
      "loss": 1.5886,
      "step": 110
    },
    {
      "epoch": 0.03559688928084663,
      "grad_norm": 1.5424351692199707,
      "learning_rate": 9.675032175032176e-05,
      "loss": 1.6448,
      "step": 111
    },
    {
      "epoch": 0.03591758197707047,
      "grad_norm": 1.8513628244400024,
      "learning_rate": 9.671814671814672e-05,
      "loss": 1.4854,
      "step": 112
    },
    {
      "epoch": 0.03623827467329432,
      "grad_norm": 1.7762032747268677,
      "learning_rate": 9.668597168597169e-05,
      "loss": 1.5484,
      "step": 113
    },
    {
      "epoch": 0.03655896736951816,
      "grad_norm": 1.8025050163269043,
      "learning_rate": 9.665379665379665e-05,
      "loss": 1.7242,
      "step": 114
    },
    {
      "epoch": 0.036879660065742004,
      "grad_norm": 1.6463018655776978,
      "learning_rate": 9.662162162162163e-05,
      "loss": 1.5908,
      "step": 115
    },
    {
      "epoch": 0.037200352761965844,
      "grad_norm": 1.640679121017456,
      "learning_rate": 9.65894465894466e-05,
      "loss": 1.4735,
      "step": 116
    },
    {
      "epoch": 0.03752104545818969,
      "grad_norm": 1.7171475887298584,
      "learning_rate": 9.655727155727157e-05,
      "loss": 1.5754,
      "step": 117
    },
    {
      "epoch": 0.03784173815441353,
      "grad_norm": 1.6603971719741821,
      "learning_rate": 9.652509652509652e-05,
      "loss": 1.6748,
      "step": 118
    },
    {
      "epoch": 0.03816243085063738,
      "grad_norm": 1.7294347286224365,
      "learning_rate": 9.649292149292149e-05,
      "loss": 1.6821,
      "step": 119
    },
    {
      "epoch": 0.03848312354686122,
      "grad_norm": 1.8348171710968018,
      "learning_rate": 9.646074646074647e-05,
      "loss": 1.6028,
      "step": 120
    },
    {
      "epoch": 0.038803816243085064,
      "grad_norm": 1.729529857635498,
      "learning_rate": 9.642857142857143e-05,
      "loss": 1.5516,
      "step": 121
    },
    {
      "epoch": 0.039124508939308904,
      "grad_norm": 1.7815181016921997,
      "learning_rate": 9.639639639639641e-05,
      "loss": 1.6472,
      "step": 122
    },
    {
      "epoch": 0.03944520163553275,
      "grad_norm": 1.8458462953567505,
      "learning_rate": 9.636422136422137e-05,
      "loss": 1.5905,
      "step": 123
    },
    {
      "epoch": 0.0397658943317566,
      "grad_norm": 1.6666098833084106,
      "learning_rate": 9.633204633204634e-05,
      "loss": 1.4131,
      "step": 124
    },
    {
      "epoch": 0.04008658702798044,
      "grad_norm": 1.9331727027893066,
      "learning_rate": 9.62998712998713e-05,
      "loss": 1.5866,
      "step": 125
    },
    {
      "epoch": 0.040407279724204284,
      "grad_norm": 1.8131427764892578,
      "learning_rate": 9.626769626769627e-05,
      "loss": 1.6006,
      "step": 126
    },
    {
      "epoch": 0.040727972420428124,
      "grad_norm": 1.6893302202224731,
      "learning_rate": 9.623552123552125e-05,
      "loss": 1.7013,
      "step": 127
    },
    {
      "epoch": 0.04104866511665197,
      "grad_norm": 1.8650181293487549,
      "learning_rate": 9.620334620334621e-05,
      "loss": 1.7522,
      "step": 128
    },
    {
      "epoch": 0.04136935781287581,
      "grad_norm": 1.8159857988357544,
      "learning_rate": 9.617117117117117e-05,
      "loss": 1.4935,
      "step": 129
    },
    {
      "epoch": 0.04169005050909966,
      "grad_norm": 1.919482707977295,
      "learning_rate": 9.613899613899614e-05,
      "loss": 1.6084,
      "step": 130
    },
    {
      "epoch": 0.0420107432053235,
      "grad_norm": 1.721970558166504,
      "learning_rate": 9.61068211068211e-05,
      "loss": 1.5999,
      "step": 131
    },
    {
      "epoch": 0.042331435901547344,
      "grad_norm": 1.6646255254745483,
      "learning_rate": 9.607464607464608e-05,
      "loss": 1.5181,
      "step": 132
    },
    {
      "epoch": 0.042652128597771184,
      "grad_norm": 1.7984895706176758,
      "learning_rate": 9.604247104247105e-05,
      "loss": 1.6474,
      "step": 133
    },
    {
      "epoch": 0.04297282129399503,
      "grad_norm": 1.83111572265625,
      "learning_rate": 9.601029601029602e-05,
      "loss": 1.671,
      "step": 134
    },
    {
      "epoch": 0.04329351399021887,
      "grad_norm": 1.7685164213180542,
      "learning_rate": 9.597812097812098e-05,
      "loss": 1.6629,
      "step": 135
    },
    {
      "epoch": 0.04361420668644272,
      "grad_norm": 1.7368009090423584,
      "learning_rate": 9.594594594594595e-05,
      "loss": 1.5108,
      "step": 136
    },
    {
      "epoch": 0.04393489938266656,
      "grad_norm": 1.9609910249710083,
      "learning_rate": 9.591377091377092e-05,
      "loss": 1.6246,
      "step": 137
    },
    {
      "epoch": 0.044255592078890404,
      "grad_norm": 2.0718841552734375,
      "learning_rate": 9.588159588159588e-05,
      "loss": 1.6929,
      "step": 138
    },
    {
      "epoch": 0.044576284775114244,
      "grad_norm": 1.6211580038070679,
      "learning_rate": 9.584942084942086e-05,
      "loss": 1.4574,
      "step": 139
    },
    {
      "epoch": 0.04489697747133809,
      "grad_norm": 1.6323758363723755,
      "learning_rate": 9.581724581724583e-05,
      "loss": 1.5247,
      "step": 140
    },
    {
      "epoch": 0.04521767016756193,
      "grad_norm": 1.8777216672897339,
      "learning_rate": 9.578507078507079e-05,
      "loss": 1.7277,
      "step": 141
    },
    {
      "epoch": 0.04553836286378578,
      "grad_norm": 1.6735459566116333,
      "learning_rate": 9.575289575289575e-05,
      "loss": 1.5469,
      "step": 142
    },
    {
      "epoch": 0.04585905556000962,
      "grad_norm": 1.5979048013687134,
      "learning_rate": 9.572072072072072e-05,
      "loss": 1.524,
      "step": 143
    },
    {
      "epoch": 0.046179748256233465,
      "grad_norm": 1.7182780504226685,
      "learning_rate": 9.56885456885457e-05,
      "loss": 1.6218,
      "step": 144
    },
    {
      "epoch": 0.046500440952457305,
      "grad_norm": 1.7319525480270386,
      "learning_rate": 9.565637065637066e-05,
      "loss": 1.684,
      "step": 145
    },
    {
      "epoch": 0.04682113364868115,
      "grad_norm": 1.7818727493286133,
      "learning_rate": 9.562419562419564e-05,
      "loss": 1.5768,
      "step": 146
    },
    {
      "epoch": 0.047141826344905,
      "grad_norm": 1.6169071197509766,
      "learning_rate": 9.559202059202059e-05,
      "loss": 1.5124,
      "step": 147
    },
    {
      "epoch": 0.04746251904112884,
      "grad_norm": 1.8243296146392822,
      "learning_rate": 9.555984555984557e-05,
      "loss": 1.7473,
      "step": 148
    },
    {
      "epoch": 0.047783211737352685,
      "grad_norm": 1.7316741943359375,
      "learning_rate": 9.552767052767053e-05,
      "loss": 1.6526,
      "step": 149
    },
    {
      "epoch": 0.048103904433576525,
      "grad_norm": 1.712912678718567,
      "learning_rate": 9.54954954954955e-05,
      "loss": 1.7089,
      "step": 150
    },
    {
      "epoch": 0.04842459712980037,
      "grad_norm": 1.6389209032058716,
      "learning_rate": 9.546332046332048e-05,
      "loss": 1.4369,
      "step": 151
    },
    {
      "epoch": 0.04874528982602421,
      "grad_norm": 1.9212989807128906,
      "learning_rate": 9.543114543114544e-05,
      "loss": 1.8143,
      "step": 152
    },
    {
      "epoch": 0.04906598252224806,
      "grad_norm": 1.804646611213684,
      "learning_rate": 9.53989703989704e-05,
      "loss": 1.5463,
      "step": 153
    },
    {
      "epoch": 0.0493866752184719,
      "grad_norm": 1.7606157064437866,
      "learning_rate": 9.536679536679537e-05,
      "loss": 1.6029,
      "step": 154
    },
    {
      "epoch": 0.049707367914695745,
      "grad_norm": 1.673525094985962,
      "learning_rate": 9.533462033462033e-05,
      "loss": 1.4489,
      "step": 155
    },
    {
      "epoch": 0.050028060610919585,
      "grad_norm": 1.677157998085022,
      "learning_rate": 9.530244530244531e-05,
      "loss": 1.5199,
      "step": 156
    },
    {
      "epoch": 0.05034875330714343,
      "grad_norm": 1.804135799407959,
      "learning_rate": 9.527027027027028e-05,
      "loss": 1.5992,
      "step": 157
    },
    {
      "epoch": 0.05066944600336727,
      "grad_norm": 1.651515007019043,
      "learning_rate": 9.523809523809524e-05,
      "loss": 1.5613,
      "step": 158
    },
    {
      "epoch": 0.05099013869959112,
      "grad_norm": 1.7162939310073853,
      "learning_rate": 9.52059202059202e-05,
      "loss": 1.6594,
      "step": 159
    },
    {
      "epoch": 0.05131083139581496,
      "grad_norm": 1.8460568189620972,
      "learning_rate": 9.517374517374518e-05,
      "loss": 1.7585,
      "step": 160
    },
    {
      "epoch": 0.051631524092038805,
      "grad_norm": 1.9340801239013672,
      "learning_rate": 9.514157014157015e-05,
      "loss": 1.656,
      "step": 161
    },
    {
      "epoch": 0.051952216788262645,
      "grad_norm": 1.7593783140182495,
      "learning_rate": 9.510939510939511e-05,
      "loss": 1.4838,
      "step": 162
    },
    {
      "epoch": 0.05227290948448649,
      "grad_norm": 1.7070125341415405,
      "learning_rate": 9.507722007722009e-05,
      "loss": 1.4695,
      "step": 163
    },
    {
      "epoch": 0.05259360218071033,
      "grad_norm": 1.6149088144302368,
      "learning_rate": 9.504504504504504e-05,
      "loss": 1.5294,
      "step": 164
    },
    {
      "epoch": 0.05291429487693418,
      "grad_norm": 1.7407358884811401,
      "learning_rate": 9.501287001287002e-05,
      "loss": 1.5296,
      "step": 165
    },
    {
      "epoch": 0.05323498757315802,
      "grad_norm": 1.8265562057495117,
      "learning_rate": 9.498069498069498e-05,
      "loss": 1.7319,
      "step": 166
    },
    {
      "epoch": 0.053555680269381865,
      "grad_norm": 1.7947865724563599,
      "learning_rate": 9.494851994851995e-05,
      "loss": 1.5892,
      "step": 167
    },
    {
      "epoch": 0.053876372965605705,
      "grad_norm": 1.9553256034851074,
      "learning_rate": 9.491634491634493e-05,
      "loss": 1.8113,
      "step": 168
    },
    {
      "epoch": 0.05419706566182955,
      "grad_norm": 1.8548988103866577,
      "learning_rate": 9.488416988416989e-05,
      "loss": 1.6531,
      "step": 169
    },
    {
      "epoch": 0.05451775835805339,
      "grad_norm": 1.9492192268371582,
      "learning_rate": 9.485199485199486e-05,
      "loss": 1.6863,
      "step": 170
    },
    {
      "epoch": 0.05483845105427724,
      "grad_norm": 1.9482530355453491,
      "learning_rate": 9.481981981981982e-05,
      "loss": 1.7185,
      "step": 171
    },
    {
      "epoch": 0.055159143750501086,
      "grad_norm": 1.7255852222442627,
      "learning_rate": 9.47876447876448e-05,
      "loss": 1.5779,
      "step": 172
    },
    {
      "epoch": 0.055479836446724926,
      "grad_norm": 1.706764578819275,
      "learning_rate": 9.475546975546976e-05,
      "loss": 1.664,
      "step": 173
    },
    {
      "epoch": 0.05580052914294877,
      "grad_norm": 1.8826390504837036,
      "learning_rate": 9.472329472329473e-05,
      "loss": 1.7556,
      "step": 174
    },
    {
      "epoch": 0.05612122183917261,
      "grad_norm": 1.7620404958724976,
      "learning_rate": 9.46911196911197e-05,
      "loss": 1.4872,
      "step": 175
    },
    {
      "epoch": 0.05644191453539646,
      "grad_norm": 1.6721909046173096,
      "learning_rate": 9.465894465894466e-05,
      "loss": 1.4319,
      "step": 176
    },
    {
      "epoch": 0.0567626072316203,
      "grad_norm": 1.5903970003128052,
      "learning_rate": 9.462676962676963e-05,
      "loss": 1.3132,
      "step": 177
    },
    {
      "epoch": 0.057083299927844146,
      "grad_norm": 1.77998685836792,
      "learning_rate": 9.45945945945946e-05,
      "loss": 1.6568,
      "step": 178
    },
    {
      "epoch": 0.057403992624067986,
      "grad_norm": 1.7730937004089355,
      "learning_rate": 9.456241956241956e-05,
      "loss": 1.7011,
      "step": 179
    },
    {
      "epoch": 0.05772468532029183,
      "grad_norm": 1.7409788370132446,
      "learning_rate": 9.453024453024454e-05,
      "loss": 1.4079,
      "step": 180
    },
    {
      "epoch": 0.05804537801651567,
      "grad_norm": 1.6435108184814453,
      "learning_rate": 9.44980694980695e-05,
      "loss": 1.3067,
      "step": 181
    },
    {
      "epoch": 0.05836607071273952,
      "grad_norm": 1.6248952150344849,
      "learning_rate": 9.446589446589447e-05,
      "loss": 1.4392,
      "step": 182
    },
    {
      "epoch": 0.05868676340896336,
      "grad_norm": 1.7086244821548462,
      "learning_rate": 9.443371943371943e-05,
      "loss": 1.5775,
      "step": 183
    },
    {
      "epoch": 0.059007456105187206,
      "grad_norm": 1.8702465295791626,
      "learning_rate": 9.44015444015444e-05,
      "loss": 1.5322,
      "step": 184
    },
    {
      "epoch": 0.059328148801411046,
      "grad_norm": 1.7721097469329834,
      "learning_rate": 9.436936936936938e-05,
      "loss": 1.4723,
      "step": 185
    },
    {
      "epoch": 0.05964884149763489,
      "grad_norm": 1.7834652662277222,
      "learning_rate": 9.433719433719434e-05,
      "loss": 1.5833,
      "step": 186
    },
    {
      "epoch": 0.05996953419385873,
      "grad_norm": 1.840971827507019,
      "learning_rate": 9.43050193050193e-05,
      "loss": 1.6228,
      "step": 187
    },
    {
      "epoch": 0.06029022689008258,
      "grad_norm": 1.7249364852905273,
      "learning_rate": 9.427284427284427e-05,
      "loss": 1.4276,
      "step": 188
    },
    {
      "epoch": 0.06061091958630642,
      "grad_norm": 1.8125596046447754,
      "learning_rate": 9.424066924066925e-05,
      "loss": 1.4192,
      "step": 189
    },
    {
      "epoch": 0.060931612282530266,
      "grad_norm": 1.6701613664627075,
      "learning_rate": 9.420849420849421e-05,
      "loss": 1.7908,
      "step": 190
    },
    {
      "epoch": 0.061252304978754106,
      "grad_norm": 1.7857857942581177,
      "learning_rate": 9.417631917631918e-05,
      "loss": 1.7204,
      "step": 191
    },
    {
      "epoch": 0.06157299767497795,
      "grad_norm": 1.917222499847412,
      "learning_rate": 9.414414414414416e-05,
      "loss": 1.7897,
      "step": 192
    },
    {
      "epoch": 0.06189369037120179,
      "grad_norm": 1.814931035041809,
      "learning_rate": 9.411196911196911e-05,
      "loss": 1.6911,
      "step": 193
    },
    {
      "epoch": 0.06221438306742564,
      "grad_norm": 1.600833535194397,
      "learning_rate": 9.407979407979409e-05,
      "loss": 1.4422,
      "step": 194
    },
    {
      "epoch": 0.06253507576364949,
      "grad_norm": 1.750791072845459,
      "learning_rate": 9.404761904761905e-05,
      "loss": 1.602,
      "step": 195
    },
    {
      "epoch": 0.06285576845987333,
      "grad_norm": 1.8130338191986084,
      "learning_rate": 9.401544401544401e-05,
      "loss": 1.7479,
      "step": 196
    },
    {
      "epoch": 0.06317646115609717,
      "grad_norm": 1.8416740894317627,
      "learning_rate": 9.398326898326899e-05,
      "loss": 1.7607,
      "step": 197
    },
    {
      "epoch": 0.06349715385232102,
      "grad_norm": 1.7460497617721558,
      "learning_rate": 9.395109395109396e-05,
      "loss": 1.796,
      "step": 198
    },
    {
      "epoch": 0.06381784654854486,
      "grad_norm": 1.8751194477081299,
      "learning_rate": 9.391891891891892e-05,
      "loss": 1.7973,
      "step": 199
    },
    {
      "epoch": 0.0641385392447687,
      "grad_norm": 1.6479636430740356,
      "learning_rate": 9.388674388674389e-05,
      "loss": 1.4562,
      "step": 200
    },
    {
      "epoch": 0.06445923194099254,
      "grad_norm": 1.813751220703125,
      "learning_rate": 9.385456885456886e-05,
      "loss": 1.4338,
      "step": 201
    },
    {
      "epoch": 0.0647799246372164,
      "grad_norm": 1.991184115409851,
      "learning_rate": 9.382239382239383e-05,
      "loss": 1.6387,
      "step": 202
    },
    {
      "epoch": 0.06510061733344023,
      "grad_norm": 1.622376561164856,
      "learning_rate": 9.379021879021879e-05,
      "loss": 1.5389,
      "step": 203
    },
    {
      "epoch": 0.06542131002966407,
      "grad_norm": 1.4437798261642456,
      "learning_rate": 9.375804375804377e-05,
      "loss": 1.1904,
      "step": 204
    },
    {
      "epoch": 0.06574200272588791,
      "grad_norm": 1.7311660051345825,
      "learning_rate": 9.372586872586872e-05,
      "loss": 1.692,
      "step": 205
    },
    {
      "epoch": 0.06606269542211177,
      "grad_norm": 1.5912631750106812,
      "learning_rate": 9.36936936936937e-05,
      "loss": 1.3408,
      "step": 206
    },
    {
      "epoch": 0.0663833881183356,
      "grad_norm": 1.7208282947540283,
      "learning_rate": 9.366151866151866e-05,
      "loss": 1.5267,
      "step": 207
    },
    {
      "epoch": 0.06670408081455945,
      "grad_norm": 1.7791844606399536,
      "learning_rate": 9.362934362934363e-05,
      "loss": 1.6352,
      "step": 208
    },
    {
      "epoch": 0.06702477351078329,
      "grad_norm": 1.7080252170562744,
      "learning_rate": 9.359716859716861e-05,
      "loss": 1.6333,
      "step": 209
    },
    {
      "epoch": 0.06734546620700714,
      "grad_norm": 1.6710137128829956,
      "learning_rate": 9.356499356499357e-05,
      "loss": 1.6988,
      "step": 210
    },
    {
      "epoch": 0.06766615890323098,
      "grad_norm": 1.8755836486816406,
      "learning_rate": 9.353281853281854e-05,
      "loss": 1.7126,
      "step": 211
    },
    {
      "epoch": 0.06798685159945482,
      "grad_norm": 1.7504806518554688,
      "learning_rate": 9.35006435006435e-05,
      "loss": 1.561,
      "step": 212
    },
    {
      "epoch": 0.06830754429567866,
      "grad_norm": 1.6312227249145508,
      "learning_rate": 9.346846846846848e-05,
      "loss": 1.5506,
      "step": 213
    },
    {
      "epoch": 0.06862823699190251,
      "grad_norm": 1.7328168153762817,
      "learning_rate": 9.343629343629344e-05,
      "loss": 1.4784,
      "step": 214
    },
    {
      "epoch": 0.06894892968812635,
      "grad_norm": 1.6513800621032715,
      "learning_rate": 9.340411840411841e-05,
      "loss": 1.5849,
      "step": 215
    },
    {
      "epoch": 0.0692696223843502,
      "grad_norm": 1.6930090188980103,
      "learning_rate": 9.337194337194337e-05,
      "loss": 1.6313,
      "step": 216
    },
    {
      "epoch": 0.06959031508057403,
      "grad_norm": 1.7721333503723145,
      "learning_rate": 9.333976833976834e-05,
      "loss": 1.5716,
      "step": 217
    },
    {
      "epoch": 0.06991100777679789,
      "grad_norm": 1.7768151760101318,
      "learning_rate": 9.330759330759331e-05,
      "loss": 1.5215,
      "step": 218
    },
    {
      "epoch": 0.07023170047302173,
      "grad_norm": 1.6654014587402344,
      "learning_rate": 9.327541827541828e-05,
      "loss": 1.6394,
      "step": 219
    },
    {
      "epoch": 0.07055239316924557,
      "grad_norm": 1.8215490579605103,
      "learning_rate": 9.324324324324324e-05,
      "loss": 1.552,
      "step": 220
    },
    {
      "epoch": 0.07087308586546942,
      "grad_norm": 1.7659235000610352,
      "learning_rate": 9.321106821106822e-05,
      "loss": 1.7689,
      "step": 221
    },
    {
      "epoch": 0.07119377856169326,
      "grad_norm": 1.8315811157226562,
      "learning_rate": 9.317889317889317e-05,
      "loss": 1.518,
      "step": 222
    },
    {
      "epoch": 0.0715144712579171,
      "grad_norm": 1.6116633415222168,
      "learning_rate": 9.314671814671815e-05,
      "loss": 1.5717,
      "step": 223
    },
    {
      "epoch": 0.07183516395414094,
      "grad_norm": 1.6100307703018188,
      "learning_rate": 9.311454311454312e-05,
      "loss": 1.6901,
      "step": 224
    },
    {
      "epoch": 0.0721558566503648,
      "grad_norm": 1.7291269302368164,
      "learning_rate": 9.30823680823681e-05,
      "loss": 1.6414,
      "step": 225
    },
    {
      "epoch": 0.07247654934658863,
      "grad_norm": 1.8281351327896118,
      "learning_rate": 9.305019305019306e-05,
      "loss": 1.8009,
      "step": 226
    },
    {
      "epoch": 0.07279724204281247,
      "grad_norm": 1.675513744354248,
      "learning_rate": 9.301801801801802e-05,
      "loss": 1.5188,
      "step": 227
    },
    {
      "epoch": 0.07311793473903631,
      "grad_norm": 1.684611201286316,
      "learning_rate": 9.298584298584299e-05,
      "loss": 1.6918,
      "step": 228
    },
    {
      "epoch": 0.07343862743526017,
      "grad_norm": 1.9136089086532593,
      "learning_rate": 9.295366795366795e-05,
      "loss": 1.6034,
      "step": 229
    },
    {
      "epoch": 0.07375932013148401,
      "grad_norm": 1.8346682786941528,
      "learning_rate": 9.292149292149293e-05,
      "loss": 1.618,
      "step": 230
    },
    {
      "epoch": 0.07408001282770785,
      "grad_norm": 1.842098355293274,
      "learning_rate": 9.28893178893179e-05,
      "loss": 1.4825,
      "step": 231
    },
    {
      "epoch": 0.07440070552393169,
      "grad_norm": 1.723360538482666,
      "learning_rate": 9.285714285714286e-05,
      "loss": 1.3566,
      "step": 232
    },
    {
      "epoch": 0.07472139822015554,
      "grad_norm": 1.6950525045394897,
      "learning_rate": 9.282496782496784e-05,
      "loss": 1.6979,
      "step": 233
    },
    {
      "epoch": 0.07504209091637938,
      "grad_norm": 1.7014894485473633,
      "learning_rate": 9.279279279279279e-05,
      "loss": 1.5959,
      "step": 234
    },
    {
      "epoch": 0.07536278361260322,
      "grad_norm": 1.7120134830474854,
      "learning_rate": 9.276061776061777e-05,
      "loss": 1.6729,
      "step": 235
    },
    {
      "epoch": 0.07568347630882706,
      "grad_norm": 1.7131479978561401,
      "learning_rate": 9.272844272844273e-05,
      "loss": 1.6015,
      "step": 236
    },
    {
      "epoch": 0.07600416900505091,
      "grad_norm": 1.7372465133666992,
      "learning_rate": 9.269626769626771e-05,
      "loss": 1.675,
      "step": 237
    },
    {
      "epoch": 0.07632486170127475,
      "grad_norm": 1.7408344745635986,
      "learning_rate": 9.266409266409267e-05,
      "loss": 1.5983,
      "step": 238
    },
    {
      "epoch": 0.0766455543974986,
      "grad_norm": 1.8079317808151245,
      "learning_rate": 9.263191763191764e-05,
      "loss": 1.5834,
      "step": 239
    },
    {
      "epoch": 0.07696624709372243,
      "grad_norm": 1.7480193376541138,
      "learning_rate": 9.25997425997426e-05,
      "loss": 1.6799,
      "step": 240
    },
    {
      "epoch": 0.07728693978994629,
      "grad_norm": 1.6279219388961792,
      "learning_rate": 9.256756756756757e-05,
      "loss": 1.6193,
      "step": 241
    },
    {
      "epoch": 0.07760763248617013,
      "grad_norm": 1.877221941947937,
      "learning_rate": 9.253539253539254e-05,
      "loss": 1.5518,
      "step": 242
    },
    {
      "epoch": 0.07792832518239397,
      "grad_norm": 1.7862673997879028,
      "learning_rate": 9.250321750321751e-05,
      "loss": 1.6225,
      "step": 243
    },
    {
      "epoch": 0.07824901787861781,
      "grad_norm": 1.7293739318847656,
      "learning_rate": 9.247104247104247e-05,
      "loss": 1.6664,
      "step": 244
    },
    {
      "epoch": 0.07856971057484166,
      "grad_norm": 1.6670184135437012,
      "learning_rate": 9.243886743886744e-05,
      "loss": 1.4997,
      "step": 245
    },
    {
      "epoch": 0.0788904032710655,
      "grad_norm": 1.6994450092315674,
      "learning_rate": 9.24066924066924e-05,
      "loss": 1.4323,
      "step": 246
    },
    {
      "epoch": 0.07921109596728934,
      "grad_norm": 1.8251434564590454,
      "learning_rate": 9.237451737451738e-05,
      "loss": 1.7074,
      "step": 247
    },
    {
      "epoch": 0.0795317886635132,
      "grad_norm": 1.6368730068206787,
      "learning_rate": 9.234234234234235e-05,
      "loss": 1.5412,
      "step": 248
    },
    {
      "epoch": 0.07985248135973703,
      "grad_norm": 1.4857044219970703,
      "learning_rate": 9.231016731016732e-05,
      "loss": 1.3139,
      "step": 249
    },
    {
      "epoch": 0.08017317405596087,
      "grad_norm": 1.7797576189041138,
      "learning_rate": 9.227799227799229e-05,
      "loss": 1.6985,
      "step": 250
    },
    {
      "epoch": 0.08049386675218471,
      "grad_norm": 1.8514691591262817,
      "learning_rate": 9.224581724581724e-05,
      "loss": 1.7173,
      "step": 251
    },
    {
      "epoch": 0.08081455944840857,
      "grad_norm": 1.7061352729797363,
      "learning_rate": 9.221364221364222e-05,
      "loss": 1.4749,
      "step": 252
    },
    {
      "epoch": 0.08113525214463241,
      "grad_norm": 1.8911951780319214,
      "learning_rate": 9.218146718146718e-05,
      "loss": 1.612,
      "step": 253
    },
    {
      "epoch": 0.08145594484085625,
      "grad_norm": 1.8928183317184448,
      "learning_rate": 9.214929214929216e-05,
      "loss": 1.7248,
      "step": 254
    },
    {
      "epoch": 0.08177663753708009,
      "grad_norm": 1.69159996509552,
      "learning_rate": 9.211711711711712e-05,
      "loss": 1.5939,
      "step": 255
    },
    {
      "epoch": 0.08209733023330394,
      "grad_norm": 1.7285486459732056,
      "learning_rate": 9.208494208494209e-05,
      "loss": 1.4037,
      "step": 256
    },
    {
      "epoch": 0.08241802292952778,
      "grad_norm": 1.5664756298065186,
      "learning_rate": 9.205276705276705e-05,
      "loss": 1.544,
      "step": 257
    },
    {
      "epoch": 0.08273871562575162,
      "grad_norm": 1.8683611154556274,
      "learning_rate": 9.202059202059202e-05,
      "loss": 1.4889,
      "step": 258
    },
    {
      "epoch": 0.08305940832197546,
      "grad_norm": 2.022183656692505,
      "learning_rate": 9.1988416988417e-05,
      "loss": 1.6101,
      "step": 259
    },
    {
      "epoch": 0.08338010101819932,
      "grad_norm": 1.7333014011383057,
      "learning_rate": 9.195624195624196e-05,
      "loss": 1.4097,
      "step": 260
    },
    {
      "epoch": 0.08370079371442316,
      "grad_norm": 1.693903923034668,
      "learning_rate": 9.192406692406694e-05,
      "loss": 1.6963,
      "step": 261
    },
    {
      "epoch": 0.084021486410647,
      "grad_norm": 1.8906500339508057,
      "learning_rate": 9.18918918918919e-05,
      "loss": 1.7168,
      "step": 262
    },
    {
      "epoch": 0.08434217910687083,
      "grad_norm": 1.8799378871917725,
      "learning_rate": 9.185971685971685e-05,
      "loss": 1.6934,
      "step": 263
    },
    {
      "epoch": 0.08466287180309469,
      "grad_norm": 1.8366758823394775,
      "learning_rate": 9.182754182754183e-05,
      "loss": 1.7188,
      "step": 264
    },
    {
      "epoch": 0.08498356449931853,
      "grad_norm": 1.7556151151657104,
      "learning_rate": 9.17953667953668e-05,
      "loss": 1.6177,
      "step": 265
    },
    {
      "epoch": 0.08530425719554237,
      "grad_norm": 1.7707295417785645,
      "learning_rate": 9.176319176319177e-05,
      "loss": 1.5682,
      "step": 266
    },
    {
      "epoch": 0.08562494989176621,
      "grad_norm": 1.636765480041504,
      "learning_rate": 9.173101673101674e-05,
      "loss": 1.4409,
      "step": 267
    },
    {
      "epoch": 0.08594564258799006,
      "grad_norm": 1.7112761735916138,
      "learning_rate": 9.16988416988417e-05,
      "loss": 1.6157,
      "step": 268
    },
    {
      "epoch": 0.0862663352842139,
      "grad_norm": 1.9961086511611938,
      "learning_rate": 9.166666666666667e-05,
      "loss": 1.5517,
      "step": 269
    },
    {
      "epoch": 0.08658702798043774,
      "grad_norm": 1.5996118783950806,
      "learning_rate": 9.163449163449163e-05,
      "loss": 1.4812,
      "step": 270
    },
    {
      "epoch": 0.0869077206766616,
      "grad_norm": 1.658401608467102,
      "learning_rate": 9.160231660231661e-05,
      "loss": 1.4842,
      "step": 271
    },
    {
      "epoch": 0.08722841337288544,
      "grad_norm": 1.7346595525741577,
      "learning_rate": 9.157014157014157e-05,
      "loss": 1.55,
      "step": 272
    },
    {
      "epoch": 0.08754910606910928,
      "grad_norm": 1.7823814153671265,
      "learning_rate": 9.153796653796655e-05,
      "loss": 1.6065,
      "step": 273
    },
    {
      "epoch": 0.08786979876533312,
      "grad_norm": 1.7370473146438599,
      "learning_rate": 9.15057915057915e-05,
      "loss": 1.5657,
      "step": 274
    },
    {
      "epoch": 0.08819049146155697,
      "grad_norm": 1.5054192543029785,
      "learning_rate": 9.147361647361647e-05,
      "loss": 1.5523,
      "step": 275
    },
    {
      "epoch": 0.08851118415778081,
      "grad_norm": 1.7042500972747803,
      "learning_rate": 9.144144144144145e-05,
      "loss": 1.6344,
      "step": 276
    },
    {
      "epoch": 0.08883187685400465,
      "grad_norm": 1.6680580377578735,
      "learning_rate": 9.140926640926641e-05,
      "loss": 1.4047,
      "step": 277
    },
    {
      "epoch": 0.08915256955022849,
      "grad_norm": 1.7721407413482666,
      "learning_rate": 9.137709137709139e-05,
      "loss": 1.5141,
      "step": 278
    },
    {
      "epoch": 0.08947326224645234,
      "grad_norm": 1.6130516529083252,
      "learning_rate": 9.134491634491635e-05,
      "loss": 1.3017,
      "step": 279
    },
    {
      "epoch": 0.08979395494267618,
      "grad_norm": 1.6346606016159058,
      "learning_rate": 9.131274131274132e-05,
      "loss": 1.3774,
      "step": 280
    },
    {
      "epoch": 0.09011464763890002,
      "grad_norm": 1.720962643623352,
      "learning_rate": 9.128056628056628e-05,
      "loss": 1.4936,
      "step": 281
    },
    {
      "epoch": 0.09043534033512386,
      "grad_norm": 1.7229537963867188,
      "learning_rate": 9.124839124839125e-05,
      "loss": 1.6381,
      "step": 282
    },
    {
      "epoch": 0.09075603303134772,
      "grad_norm": 1.9374829530715942,
      "learning_rate": 9.121621621621623e-05,
      "loss": 1.5829,
      "step": 283
    },
    {
      "epoch": 0.09107672572757156,
      "grad_norm": 1.6721988916397095,
      "learning_rate": 9.118404118404119e-05,
      "loss": 1.5449,
      "step": 284
    },
    {
      "epoch": 0.0913974184237954,
      "grad_norm": 1.5761111974716187,
      "learning_rate": 9.115186615186617e-05,
      "loss": 1.388,
      "step": 285
    },
    {
      "epoch": 0.09171811112001924,
      "grad_norm": 1.8343489170074463,
      "learning_rate": 9.111969111969112e-05,
      "loss": 1.5724,
      "step": 286
    },
    {
      "epoch": 0.09203880381624309,
      "grad_norm": 1.564759373664856,
      "learning_rate": 9.108751608751608e-05,
      "loss": 1.4356,
      "step": 287
    },
    {
      "epoch": 0.09235949651246693,
      "grad_norm": 1.6440503597259521,
      "learning_rate": 9.105534105534106e-05,
      "loss": 1.4281,
      "step": 288
    },
    {
      "epoch": 0.09268018920869077,
      "grad_norm": 1.7821102142333984,
      "learning_rate": 9.102316602316603e-05,
      "loss": 1.5695,
      "step": 289
    },
    {
      "epoch": 0.09300088190491461,
      "grad_norm": 1.7555899620056152,
      "learning_rate": 9.0990990990991e-05,
      "loss": 1.7338,
      "step": 290
    },
    {
      "epoch": 0.09332157460113846,
      "grad_norm": 1.800379753112793,
      "learning_rate": 9.095881595881597e-05,
      "loss": 1.5684,
      "step": 291
    },
    {
      "epoch": 0.0936422672973623,
      "grad_norm": 1.6488431692123413,
      "learning_rate": 9.092664092664093e-05,
      "loss": 1.4512,
      "step": 292
    },
    {
      "epoch": 0.09396295999358614,
      "grad_norm": 1.72849440574646,
      "learning_rate": 9.08944658944659e-05,
      "loss": 1.6229,
      "step": 293
    },
    {
      "epoch": 0.09428365268981,
      "grad_norm": 1.7736647129058838,
      "learning_rate": 9.086229086229086e-05,
      "loss": 1.4737,
      "step": 294
    },
    {
      "epoch": 0.09460434538603384,
      "grad_norm": 1.7936447858810425,
      "learning_rate": 9.083011583011584e-05,
      "loss": 1.3228,
      "step": 295
    },
    {
      "epoch": 0.09492503808225768,
      "grad_norm": 1.8301600217819214,
      "learning_rate": 9.07979407979408e-05,
      "loss": 1.6702,
      "step": 296
    },
    {
      "epoch": 0.09524573077848152,
      "grad_norm": 1.5292707681655884,
      "learning_rate": 9.076576576576577e-05,
      "loss": 1.5324,
      "step": 297
    },
    {
      "epoch": 0.09556642347470537,
      "grad_norm": 1.7899250984191895,
      "learning_rate": 9.073359073359073e-05,
      "loss": 1.6492,
      "step": 298
    },
    {
      "epoch": 0.09588711617092921,
      "grad_norm": 1.6854242086410522,
      "learning_rate": 9.07014157014157e-05,
      "loss": 1.5048,
      "step": 299
    },
    {
      "epoch": 0.09620780886715305,
      "grad_norm": 1.819797396659851,
      "learning_rate": 9.066924066924068e-05,
      "loss": 1.7865,
      "step": 300
    }
  ],
  "logging_steps": 1,
  "max_steps": 3118,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.366563833469788e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}