{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 230,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004357298474945534,
      "grad_norm": 0.9746105074882507,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 1.7828,
      "step": 1
    },
    {
      "epoch": 0.008714596949891068,
      "grad_norm": 0.9253537654876709,
      "learning_rate": 4.800000000000001e-06,
      "loss": 1.8214,
      "step": 2
    },
    {
      "epoch": 0.013071895424836602,
      "grad_norm": 1.5627340078353882,
      "learning_rate": 7.2e-06,
      "loss": 1.8438,
      "step": 3
    },
    {
      "epoch": 0.017429193899782137,
      "grad_norm": 0.38599923253059387,
      "learning_rate": 9.600000000000001e-06,
      "loss": 1.7347,
      "step": 4
    },
    {
      "epoch": 0.02178649237472767,
      "grad_norm": 0.5854593515396118,
      "learning_rate": 1.2e-05,
      "loss": 1.5746,
      "step": 5
    },
    {
      "epoch": 0.026143790849673203,
      "grad_norm": 0.5655810236930847,
      "learning_rate": 1.44e-05,
      "loss": 1.6704,
      "step": 6
    },
    {
      "epoch": 0.030501089324618737,
      "grad_norm": 0.5204870700836182,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 1.7218,
      "step": 7
    },
    {
      "epoch": 0.034858387799564274,
      "grad_norm": 0.3172350227832794,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 1.7356,
      "step": 8
    },
    {
      "epoch": 0.0392156862745098,
      "grad_norm": 0.3486512005329132,
      "learning_rate": 2.16e-05,
      "loss": 1.6989,
      "step": 9
    },
    {
      "epoch": 0.04357298474945534,
      "grad_norm": 0.3535781800746918,
      "learning_rate": 2.4e-05,
      "loss": 1.8387,
      "step": 10
    },
    {
      "epoch": 0.04793028322440087,
      "grad_norm": 0.5419184565544128,
      "learning_rate": 2.64e-05,
      "loss": 1.6227,
      "step": 11
    },
    {
      "epoch": 0.05228758169934641,
      "grad_norm": 0.48862531781196594,
      "learning_rate": 2.88e-05,
      "loss": 1.7614,
      "step": 12
    },
    {
      "epoch": 0.05664488017429194,
      "grad_norm": 0.2664097547531128,
      "learning_rate": 3.12e-05,
      "loss": 1.6985,
      "step": 13
    },
    {
      "epoch": 0.06100217864923747,
      "grad_norm": 0.28547024726867676,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 1.6369,
      "step": 14
    },
    {
      "epoch": 0.06535947712418301,
      "grad_norm": 0.7763476371765137,
      "learning_rate": 3.6e-05,
      "loss": 1.6875,
      "step": 15
    },
    {
      "epoch": 0.06971677559912855,
      "grad_norm": 0.40384408831596375,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 1.5499,
      "step": 16
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 0.2866835594177246,
      "learning_rate": 4.08e-05,
      "loss": 1.7804,
      "step": 17
    },
    {
      "epoch": 0.0784313725490196,
      "grad_norm": 0.36669981479644775,
      "learning_rate": 4.32e-05,
      "loss": 1.6886,
      "step": 18
    },
    {
      "epoch": 0.08278867102396514,
      "grad_norm": 0.31070730090141296,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 1.5843,
      "step": 19
    },
    {
      "epoch": 0.08714596949891068,
      "grad_norm": 0.2698839604854584,
      "learning_rate": 4.8e-05,
      "loss": 1.6562,
      "step": 20
    },
    {
      "epoch": 0.0915032679738562,
      "grad_norm": 0.3030396103858948,
      "learning_rate": 5.04e-05,
      "loss": 1.7683,
      "step": 21
    },
    {
      "epoch": 0.09586056644880174,
      "grad_norm": 0.34498634934425354,
      "learning_rate": 5.28e-05,
      "loss": 1.7619,
      "step": 22
    },
    {
      "epoch": 0.10021786492374728,
      "grad_norm": 0.315149188041687,
      "learning_rate": 5.520000000000001e-05,
      "loss": 1.6201,
      "step": 23
    },
    {
      "epoch": 0.10457516339869281,
      "grad_norm": 0.2636476159095764,
      "learning_rate": 5.76e-05,
      "loss": 1.6439,
      "step": 24
    },
    {
      "epoch": 0.10893246187363835,
      "grad_norm": 0.5684914588928223,
      "learning_rate": 6e-05,
      "loss": 1.823,
      "step": 25
    },
    {
      "epoch": 0.11328976034858387,
      "grad_norm": 0.3564894497394562,
      "learning_rate": 5.9999210388787696e-05,
      "loss": 1.6807,
      "step": 26
    },
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 0.3562137186527252,
      "learning_rate": 5.999684159671647e-05,
      "loss": 1.7543,
      "step": 27
    },
    {
      "epoch": 0.12200435729847495,
      "grad_norm": 0.32931092381477356,
      "learning_rate": 5.999289374848134e-05,
      "loss": 1.7032,
      "step": 28
    },
    {
      "epoch": 0.12636165577342048,
      "grad_norm": 0.3593999743461609,
      "learning_rate": 5.998736705189997e-05,
      "loss": 1.5749,
      "step": 29
    },
    {
      "epoch": 0.13071895424836602,
      "grad_norm": 0.31765711307525635,
      "learning_rate": 5.998026179790181e-05,
      "loss": 1.5925,
      "step": 30
    },
    {
      "epoch": 0.13507625272331156,
      "grad_norm": 1.1995874643325806,
      "learning_rate": 5.9971578360512726e-05,
      "loss": 1.6874,
      "step": 31
    },
    {
      "epoch": 0.1394335511982571,
      "grad_norm": 0.3488420248031616,
      "learning_rate": 5.996131719683537e-05,
      "loss": 1.8334,
      "step": 32
    },
    {
      "epoch": 0.1437908496732026,
      "grad_norm": 1.2170417308807373,
      "learning_rate": 5.994947884702506e-05,
      "loss": 1.8111,
      "step": 33
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 7.140843868255615,
      "learning_rate": 5.993606393426137e-05,
      "loss": 1.7542,
      "step": 34
    },
    {
      "epoch": 0.15250544662309368,
      "grad_norm": 6.034117221832275,
      "learning_rate": 5.992107316471536e-05,
      "loss": 1.6763,
      "step": 35
    },
    {
      "epoch": 0.1568627450980392,
      "grad_norm": 24.70130729675293,
      "learning_rate": 5.990450732751232e-05,
      "loss": 1.7738,
      "step": 36
    },
    {
      "epoch": 0.16122004357298475,
      "grad_norm": 2.477736234664917,
      "learning_rate": 5.988636729469032e-05,
      "loss": 1.8413,
      "step": 37
    },
    {
      "epoch": 0.1655773420479303,
      "grad_norm": 0.5539727210998535,
      "learning_rate": 5.986665402115423e-05,
      "loss": 1.7711,
      "step": 38
    },
    {
      "epoch": 0.16993464052287582,
      "grad_norm": 0.651229739189148,
      "learning_rate": 5.9845368544625516e-05,
      "loss": 1.809,
      "step": 39
    },
    {
      "epoch": 0.17429193899782136,
      "grad_norm": 0.4992523491382599,
      "learning_rate": 5.9822511985587577e-05,
      "loss": 1.9164,
      "step": 40
    },
    {
      "epoch": 0.1786492374727669,
      "grad_norm": 0.5860515832901001,
      "learning_rate": 5.979808554722675e-05,
      "loss": 1.737,
      "step": 41
    },
    {
      "epoch": 0.1830065359477124,
      "grad_norm": 0.9775006175041199,
      "learning_rate": 5.977209051536903e-05,
      "loss": 1.6852,
      "step": 42
    },
    {
      "epoch": 0.18736383442265794,
      "grad_norm": 0.5079925656318665,
      "learning_rate": 5.974452825841231e-05,
      "loss": 1.6264,
      "step": 43
    },
    {
      "epoch": 0.19172113289760348,
      "grad_norm": 0.4939415752887726,
      "learning_rate": 5.97154002272544e-05,
      "loss": 1.6138,
      "step": 44
    },
    {
      "epoch": 0.19607843137254902,
      "grad_norm": 0.4919627606868744,
      "learning_rate": 5.968470795521663e-05,
      "loss": 1.6991,
      "step": 45
    },
    {
      "epoch": 0.20043572984749455,
      "grad_norm": 0.9349490404129028,
      "learning_rate": 5.965245305796316e-05,
      "loss": 1.7488,
      "step": 46
    },
    {
      "epoch": 0.2047930283224401,
      "grad_norm": 8.618950843811035,
      "learning_rate": 5.9618637233415866e-05,
      "loss": 1.7847,
      "step": 47
    },
    {
      "epoch": 0.20915032679738563,
      "grad_norm": 0.656200110912323,
      "learning_rate": 5.958326226166505e-05,
      "loss": 1.7535,
      "step": 48
    },
    {
      "epoch": 0.21350762527233116,
      "grad_norm": 0.49261513352394104,
      "learning_rate": 5.954633000487565e-05,
      "loss": 1.6509,
      "step": 49
    },
    {
      "epoch": 0.2178649237472767,
      "grad_norm": 0.7735104560852051,
      "learning_rate": 5.950784240718929e-05,
      "loss": 1.7014,
      "step": 50
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.5278734564781189,
      "learning_rate": 5.946780149462187e-05,
      "loss": 1.6646,
      "step": 51
    },
    {
      "epoch": 0.22657952069716775,
      "grad_norm": 10.276693344116211,
      "learning_rate": 5.942620937495696e-05,
      "loss": 1.9713,
      "step": 52
    },
    {
      "epoch": 0.23093681917211328,
      "grad_norm": 0.9025694727897644,
      "learning_rate": 5.938306823763481e-05,
      "loss": 1.7381,
      "step": 53
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 0.6549501419067383,
      "learning_rate": 5.933838035363717e-05,
      "loss": 1.5728,
      "step": 54
    },
    {
      "epoch": 0.23965141612200436,
      "grad_norm": 0.6484580039978027,
      "learning_rate": 5.9292148075367635e-05,
      "loss": 1.7029,
      "step": 55
    },
    {
      "epoch": 0.2440087145969499,
      "grad_norm": 0.5284441113471985,
      "learning_rate": 5.924437383652789e-05,
      "loss": 1.6064,
      "step": 56
    },
    {
      "epoch": 0.24836601307189543,
      "grad_norm": 0.4835830330848694,
      "learning_rate": 5.9195060151989595e-05,
      "loss": 1.6933,
      "step": 57
    },
    {
      "epoch": 0.25272331154684097,
      "grad_norm": 0.4144735336303711,
      "learning_rate": 5.914420961766194e-05,
      "loss": 1.7435,
      "step": 58
    },
    {
      "epoch": 0.2570806100217865,
      "grad_norm": 0.41090619564056396,
      "learning_rate": 5.909182491035509e-05,
      "loss": 1.6487,
      "step": 59
    },
    {
      "epoch": 0.26143790849673204,
      "grad_norm": 0.35188132524490356,
      "learning_rate": 5.9037908787639174e-05,
      "loss": 1.7431,
      "step": 60
    },
    {
      "epoch": 0.2657952069716776,
      "grad_norm": 0.5362684726715088,
      "learning_rate": 5.898246408769921e-05,
      "loss": 1.7848,
      "step": 61
    },
    {
      "epoch": 0.2701525054466231,
      "grad_norm": 0.37840691208839417,
      "learning_rate": 5.892549372918564e-05,
      "loss": 1.807,
      "step": 62
    },
    {
      "epoch": 0.27450980392156865,
      "grad_norm": 0.6966099143028259,
      "learning_rate": 5.8867000711060704e-05,
      "loss": 1.6143,
      "step": 63
    },
    {
      "epoch": 0.2788671023965142,
      "grad_norm": 1.4057034254074097,
      "learning_rate": 5.8806988112440624e-05,
      "loss": 1.5769,
      "step": 64
    },
    {
      "epoch": 0.28322440087145967,
      "grad_norm": 0.38457629084587097,
      "learning_rate": 5.8745459092433436e-05,
      "loss": 1.6125,
      "step": 65
    },
    {
      "epoch": 0.2875816993464052,
      "grad_norm": 0.3964170813560486,
      "learning_rate": 5.868241688997275e-05,
      "loss": 1.9211,
      "step": 66
    },
    {
      "epoch": 0.29193899782135074,
      "grad_norm": 0.4357544183731079,
      "learning_rate": 5.861786482364723e-05,
      "loss": 1.7145,
      "step": 67
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 0.4696052670478821,
      "learning_rate": 5.8551806291525885e-05,
      "loss": 1.5768,
      "step": 68
    },
    {
      "epoch": 0.3006535947712418,
      "grad_norm": 0.5432698130607605,
      "learning_rate": 5.848424477097924e-05,
      "loss": 1.7934,
      "step": 69
    },
    {
      "epoch": 0.30501089324618735,
      "grad_norm": 0.44644972681999207,
      "learning_rate": 5.8415183818496234e-05,
      "loss": 1.7114,
      "step": 70
    },
    {
      "epoch": 0.3093681917211329,
      "grad_norm": 0.5297689437866211,
      "learning_rate": 5.8344627069497025e-05,
      "loss": 1.6654,
      "step": 71
    },
    {
      "epoch": 0.3137254901960784,
      "grad_norm": 0.5777753591537476,
      "learning_rate": 5.827257823814162e-05,
      "loss": 1.7446,
      "step": 72
    },
    {
      "epoch": 0.31808278867102396,
      "grad_norm": 5.469757556915283,
      "learning_rate": 5.819904111713436e-05,
      "loss": 1.7395,
      "step": 73
    },
    {
      "epoch": 0.3224400871459695,
      "grad_norm": 0.5288973450660706,
      "learning_rate": 5.812401957752426e-05,
      "loss": 1.8596,
      "step": 74
    },
    {
      "epoch": 0.32679738562091504,
      "grad_norm": 1.6414694786071777,
      "learning_rate": 5.804751756850124e-05,
      "loss": 1.6706,
      "step": 75
    },
    {
      "epoch": 0.3311546840958606,
      "grad_norm": 0.9940863847732544,
      "learning_rate": 5.7969539117188256e-05,
      "loss": 1.7504,
      "step": 76
    },
    {
      "epoch": 0.3355119825708061,
      "grad_norm": 4.836872577667236,
      "learning_rate": 5.789008832842924e-05,
      "loss": 1.8491,
      "step": 77
    },
    {
      "epoch": 0.33986928104575165,
      "grad_norm": 1.7528856992721558,
      "learning_rate": 5.780916938457314e-05,
      "loss": 1.8635,
      "step": 78
    },
    {
      "epoch": 0.3442265795206972,
      "grad_norm": 1.1119537353515625,
      "learning_rate": 5.772678654525362e-05,
      "loss": 1.7444,
      "step": 79
    },
    {
      "epoch": 0.3485838779956427,
      "grad_norm": 1.9335803985595703,
      "learning_rate": 5.764294414716494e-05,
      "loss": 1.6888,
      "step": 80
    },
    {
      "epoch": 0.35294117647058826,
      "grad_norm": 2.479970932006836,
      "learning_rate": 5.755764660383358e-05,
      "loss": 1.5989,
      "step": 81
    },
    {
      "epoch": 0.3572984749455338,
      "grad_norm": 0.6857799887657166,
      "learning_rate": 5.7470898405386005e-05,
      "loss": 1.7211,
      "step": 82
    },
    {
      "epoch": 0.3616557734204793,
      "grad_norm": 0.5701134204864502,
      "learning_rate": 5.738270411831222e-05,
      "loss": 1.5853,
      "step": 83
    },
    {
      "epoch": 0.3660130718954248,
      "grad_norm": 1.7818999290466309,
      "learning_rate": 5.729306838522541e-05,
      "loss": 1.6756,
      "step": 84
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 0.6445415019989014,
      "learning_rate": 5.720199592461757e-05,
      "loss": 1.7658,
      "step": 85
    },
    {
      "epoch": 0.3747276688453159,
      "grad_norm": 0.7761105298995972,
      "learning_rate": 5.7109491530611106e-05,
      "loss": 1.6389,
      "step": 86
    },
    {
      "epoch": 0.3790849673202614,
      "grad_norm": 0.40940380096435547,
      "learning_rate": 5.701556007270647e-05,
      "loss": 1.6816,
      "step": 87
    },
    {
      "epoch": 0.38344226579520696,
      "grad_norm": 0.640964925289154,
      "learning_rate": 5.692020649552581e-05,
      "loss": 1.5405,
      "step": 88
    },
    {
      "epoch": 0.3877995642701525,
      "grad_norm": 0.5489770174026489,
      "learning_rate": 5.6823435818552696e-05,
      "loss": 1.7949,
      "step": 89
    },
    {
      "epoch": 0.39215686274509803,
      "grad_norm": 0.7703227400779724,
      "learning_rate": 5.6725253135867915e-05,
      "loss": 1.7931,
      "step": 90
    },
    {
      "epoch": 0.39651416122004357,
      "grad_norm": 0.5832772254943848,
      "learning_rate": 5.662566361588127e-05,
      "loss": 1.8679,
      "step": 91
    },
    {
      "epoch": 0.4008714596949891,
      "grad_norm": 0.8693870902061462,
      "learning_rate": 5.652467250105952e-05,
      "loss": 1.6827,
      "step": 92
    },
    {
      "epoch": 0.40522875816993464,
      "grad_norm": 2.1690566539764404,
      "learning_rate": 5.642228510765046e-05,
      "loss": 1.7037,
      "step": 93
    },
    {
      "epoch": 0.4095860566448802,
      "grad_norm": 1.1100988388061523,
      "learning_rate": 5.6318506825403015e-05,
      "loss": 1.6876,
      "step": 94
    },
    {
      "epoch": 0.4139433551198257,
      "grad_norm": 10.220698356628418,
      "learning_rate": 5.6213343117283506e-05,
      "loss": 1.8567,
      "step": 95
    },
    {
      "epoch": 0.41830065359477125,
      "grad_norm": 4.02601432800293,
      "learning_rate": 5.610679951918817e-05,
      "loss": 1.7431,
      "step": 96
    },
    {
      "epoch": 0.4226579520697168,
      "grad_norm": 9.665276527404785,
      "learning_rate": 5.5998881639651626e-05,
      "loss": 1.6413,
      "step": 97
    },
    {
      "epoch": 0.42701525054466233,
      "grad_norm": 16.3509578704834,
      "learning_rate": 5.588959515955173e-05,
      "loss": 2.2493,
      "step": 98
    },
    {
      "epoch": 0.43137254901960786,
      "grad_norm": 24.08732795715332,
      "learning_rate": 5.577894583181048e-05,
      "loss": 2.2252,
      "step": 99
    },
    {
      "epoch": 0.4357298474945534,
      "grad_norm": 17.57210350036621,
      "learning_rate": 5.566693948109122e-05,
      "loss": 1.7175,
      "step": 100
    },
    {
      "epoch": 0.4400871459694989,
      "grad_norm": 43.53101348876953,
      "learning_rate": 5.5553582003491944e-05,
      "loss": 1.9579,
      "step": 101
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 27.57504653930664,
      "learning_rate": 5.543887936623503e-05,
      "loss": 1.9186,
      "step": 102
    },
    {
      "epoch": 0.44880174291938996,
      "grad_norm": 18.363325119018555,
      "learning_rate": 5.532283760735302e-05,
      "loss": 1.8096,
      "step": 103
    },
    {
      "epoch": 0.4531590413943355,
      "grad_norm": 27.843307495117188,
      "learning_rate": 5.520546283537086e-05,
      "loss": 1.8707,
      "step": 104
    },
    {
      "epoch": 0.45751633986928103,
      "grad_norm": 11.314703941345215,
      "learning_rate": 5.508676122898427e-05,
      "loss": 1.5912,
      "step": 105
    },
    {
      "epoch": 0.46187363834422657,
      "grad_norm": 9.792496681213379,
      "learning_rate": 5.496673903673455e-05,
      "loss": 1.6825,
      "step": 106
    },
    {
      "epoch": 0.4662309368191721,
      "grad_norm": 11.959925651550293,
      "learning_rate": 5.484540257667961e-05,
      "loss": 1.7789,
      "step": 107
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 2.4276187419891357,
      "learning_rate": 5.47227582360614e-05,
      "loss": 1.6628,
      "step": 108
    },
    {
      "epoch": 0.4749455337690632,
      "grad_norm": 2.457439661026001,
      "learning_rate": 5.459881247096969e-05,
      "loss": 1.6083,
      "step": 109
    },
    {
      "epoch": 0.4793028322440087,
      "grad_norm": 19.251781463623047,
      "learning_rate": 5.447357180600219e-05,
      "loss": 1.6644,
      "step": 110
    },
    {
      "epoch": 0.48366013071895425,
      "grad_norm": 1.5374093055725098,
      "learning_rate": 5.434704283392115e-05,
      "loss": 1.6576,
      "step": 111
    },
    {
      "epoch": 0.4880174291938998,
      "grad_norm": 1.1958122253417969,
      "learning_rate": 5.421923221530622e-05,
      "loss": 1.5076,
      "step": 112
    },
    {
      "epoch": 0.4923747276688453,
      "grad_norm": 1.085802435874939,
      "learning_rate": 5.409014667820389e-05,
      "loss": 1.7292,
      "step": 113
    },
    {
      "epoch": 0.49673202614379086,
      "grad_norm": 1.2091395854949951,
      "learning_rate": 5.395979301777334e-05,
      "loss": 1.5753,
      "step": 114
    },
    {
      "epoch": 0.5010893246187363,
      "grad_norm": 2.773833990097046,
      "learning_rate": 5.3828178095928684e-05,
      "loss": 1.7763,
      "step": 115
    },
    {
      "epoch": 0.5054466230936819,
      "grad_norm": 0.7476174831390381,
      "learning_rate": 5.369530884097779e-05,
      "loss": 1.7815,
      "step": 116
    },
    {
      "epoch": 0.5098039215686274,
      "grad_norm": 0.7140160202980042,
      "learning_rate": 5.356119224725757e-05,
      "loss": 1.7438,
      "step": 117
    },
    {
      "epoch": 0.514161220043573,
      "grad_norm": 0.6035992503166199,
      "learning_rate": 5.3425835374765745e-05,
      "loss": 1.7846,
      "step": 118
    },
    {
      "epoch": 0.5185185185185185,
      "grad_norm": 3.483454465866089,
      "learning_rate": 5.328924534878927e-05,
      "loss": 1.8446,
      "step": 119
    },
    {
      "epoch": 0.5228758169934641,
      "grad_norm": 0.5726216435432434,
      "learning_rate": 5.3151429359529226e-05,
      "loss": 1.692,
      "step": 120
    },
    {
      "epoch": 0.5272331154684096,
      "grad_norm": 0.7607144117355347,
      "learning_rate": 5.3012394661722275e-05,
      "loss": 1.7304,
      "step": 121
    },
    {
      "epoch": 0.5315904139433552,
      "grad_norm": 0.586825966835022,
      "learning_rate": 5.2872148574258864e-05,
      "loss": 1.7993,
      "step": 122
    },
    {
      "epoch": 0.5359477124183006,
      "grad_norm": 0.4952510893344879,
      "learning_rate": 5.273069847979786e-05,
      "loss": 1.6976,
      "step": 123
    },
    {
      "epoch": 0.5403050108932462,
      "grad_norm": 0.5482230186462402,
      "learning_rate": 5.258805182437795e-05,
      "loss": 1.6702,
      "step": 124
    },
    {
      "epoch": 0.5446623093681917,
      "grad_norm": 1.4653637409210205,
      "learning_rate": 5.244421611702573e-05,
      "loss": 1.7064,
      "step": 125
    },
    {
      "epoch": 0.5490196078431373,
      "grad_norm": 0.7224717140197754,
      "learning_rate": 5.229919892936033e-05,
      "loss": 1.6133,
      "step": 126
    },
    {
      "epoch": 0.5533769063180828,
      "grad_norm": 2.9934468269348145,
      "learning_rate": 5.2153007895194915e-05,
      "loss": 1.8177,
      "step": 127
    },
    {
      "epoch": 0.5577342047930284,
      "grad_norm": 0.53426593542099,
      "learning_rate": 5.2005650710134804e-05,
      "loss": 1.775,
      "step": 128
    },
    {
      "epoch": 0.5620915032679739,
      "grad_norm": 0.5842443108558655,
      "learning_rate": 5.185713513117236e-05,
      "loss": 1.7364,
      "step": 129
    },
    {
      "epoch": 0.5664488017429193,
      "grad_norm": 0.7634817361831665,
      "learning_rate": 5.1707468976278674e-05,
      "loss": 1.6766,
      "step": 130
    },
    {
      "epoch": 0.5708061002178649,
      "grad_norm": 2.6095964908599854,
      "learning_rate": 5.1556660123992014e-05,
      "loss": 1.5793,
      "step": 131
    },
    {
      "epoch": 0.5751633986928104,
      "grad_norm": 0.6185954809188843,
      "learning_rate": 5.140471651300309e-05,
      "loss": 1.7295,
      "step": 132
    },
    {
      "epoch": 0.579520697167756,
      "grad_norm": 0.7229522466659546,
      "learning_rate": 5.125164614173716e-05,
      "loss": 1.6844,
      "step": 133
    },
    {
      "epoch": 0.5838779956427015,
      "grad_norm": 1.0191963911056519,
      "learning_rate": 5.109745706793299e-05,
      "loss": 1.6971,
      "step": 134
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 7.9863972663879395,
      "learning_rate": 5.094215740821867e-05,
      "loss": 1.6789,
      "step": 135
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 0.5257764458656311,
      "learning_rate": 5.07857553376844e-05,
      "loss": 1.6463,
      "step": 136
    },
    {
      "epoch": 0.5969498910675382,
      "grad_norm": 0.6847944855690002,
      "learning_rate": 5.062825908945204e-05,
      "loss": 1.7736,
      "step": 137
    },
    {
      "epoch": 0.6013071895424836,
      "grad_norm": 0.46256768703460693,
      "learning_rate": 5.0469676954241857e-05,
      "loss": 1.721,
      "step": 138
    },
    {
      "epoch": 0.6056644880174292,
      "grad_norm": 0.5772983431816101,
      "learning_rate": 5.0310017279935964e-05,
      "loss": 1.8956,
      "step": 139
    },
    {
      "epoch": 0.6100217864923747,
      "grad_norm": 1.1113687753677368,
      "learning_rate": 5.0149288471138975e-05,
      "loss": 1.8724,
      "step": 140
    },
    {
      "epoch": 0.6143790849673203,
      "grad_norm": 1.8465428352355957,
      "learning_rate": 4.9987498988735525e-05,
      "loss": 1.6231,
      "step": 141
    },
    {
      "epoch": 0.6187363834422658,
      "grad_norm": 0.6225913166999817,
      "learning_rate": 4.982465734944489e-05,
      "loss": 1.7372,
      "step": 142
    },
    {
      "epoch": 0.6230936819172114,
      "grad_norm": 0.6797170042991638,
      "learning_rate": 4.9660772125372715e-05,
      "loss": 1.6512,
      "step": 143
    },
    {
      "epoch": 0.6274509803921569,
      "grad_norm": 0.7226108312606812,
      "learning_rate": 4.949585194355966e-05,
      "loss": 1.7288,
      "step": 144
    },
    {
      "epoch": 0.6318082788671024,
      "grad_norm": 2.4343645572662354,
      "learning_rate": 4.9329905485527394e-05,
      "loss": 1.5985,
      "step": 145
    },
    {
      "epoch": 0.6361655773420479,
      "grad_norm": 0.8588735461235046,
      "learning_rate": 4.916294148682151e-05,
      "loss": 1.7933,
      "step": 146
    },
    {
      "epoch": 0.6405228758169934,
      "grad_norm": 0.9162130951881409,
      "learning_rate": 4.899496873655169e-05,
      "loss": 1.7511,
      "step": 147
    },
    {
      "epoch": 0.644880174291939,
      "grad_norm": 1.1275595426559448,
      "learning_rate": 4.882599607692908e-05,
      "loss": 1.7821,
      "step": 148
    },
    {
      "epoch": 0.6492374727668845,
      "grad_norm": 0.596320390701294,
      "learning_rate": 4.865603240280079e-05,
      "loss": 1.4973,
      "step": 149
    },
    {
      "epoch": 0.6535947712418301,
      "grad_norm": 2.232231855392456,
      "learning_rate": 4.8485086661181656e-05,
      "loss": 1.6625,
      "step": 150
    },
    {
      "epoch": 0.6579520697167756,
      "grad_norm": 3.6543140411376953,
      "learning_rate": 4.8313167850783304e-05,
      "loss": 1.7756,
      "step": 151
    },
    {
      "epoch": 0.6623093681917211,
      "grad_norm": 32.78754425048828,
      "learning_rate": 4.8140285021540423e-05,
      "loss": 1.7519,
      "step": 152
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 10.278705596923828,
      "learning_rate": 4.796644727413438e-05,
      "loss": 1.7621,
      "step": 153
    },
    {
      "epoch": 0.6710239651416122,
      "grad_norm": 6.011966705322266,
      "learning_rate": 4.779166375951412e-05,
      "loss": 1.7506,
      "step": 154
    },
    {
      "epoch": 0.6753812636165577,
      "grad_norm": 8.796587944030762,
      "learning_rate": 4.7615943678414515e-05,
      "loss": 1.8206,
      "step": 155
    },
    {
      "epoch": 0.6797385620915033,
      "grad_norm": 9.715092658996582,
      "learning_rate": 4.743929628087197e-05,
      "loss": 1.736,
      "step": 156
    },
    {
      "epoch": 0.6840958605664488,
      "grad_norm": 20.890012741088867,
      "learning_rate": 4.726173086573755e-05,
      "loss": 1.792,
      "step": 157
    },
    {
      "epoch": 0.6884531590413944,
      "grad_norm": 6.689200401306152,
      "learning_rate": 4.708325678018744e-05,
      "loss": 1.8754,
      "step": 158
    },
    {
      "epoch": 0.6928104575163399,
      "grad_norm": 10.521119117736816,
      "learning_rate": 4.6903883419230883e-05,
      "loss": 1.6421,
      "step": 159
    },
    {
      "epoch": 0.6971677559912854,
      "grad_norm": 2.2947447299957275,
      "learning_rate": 4.67236202252157e-05,
      "loss": 1.7782,
      "step": 160
    },
    {
      "epoch": 0.7015250544662309,
      "grad_norm": 2.0073139667510986,
      "learning_rate": 4.6542476687331165e-05,
      "loss": 1.7282,
      "step": 161
    },
    {
      "epoch": 0.7058823529411765,
      "grad_norm": 2.2821898460388184,
      "learning_rate": 4.636046234110852e-05,
      "loss": 1.7421,
      "step": 162
    },
    {
      "epoch": 0.710239651416122,
      "grad_norm": 0.8621974587440491,
      "learning_rate": 4.6177586767918984e-05,
      "loss": 1.7407,
      "step": 163
    },
    {
      "epoch": 0.7145969498910676,
      "grad_norm": 3.2683534622192383,
      "learning_rate": 4.599385959446945e-05,
      "loss": 1.8106,
      "step": 164
    },
    {
      "epoch": 0.7189542483660131,
      "grad_norm": 2.4297866821289062,
      "learning_rate": 4.5809290492295645e-05,
      "loss": 1.5172,
      "step": 165
    },
    {
      "epoch": 0.7233115468409586,
      "grad_norm": 3.033637046813965,
      "learning_rate": 4.5623889177253074e-05,
      "loss": 1.7526,
      "step": 166
    },
    {
      "epoch": 0.7276688453159041,
      "grad_norm": 1.2864795923233032,
      "learning_rate": 4.5437665409005564e-05,
      "loss": 1.7886,
      "step": 167
    },
    {
      "epoch": 0.7320261437908496,
      "grad_norm": 2.133450984954834,
      "learning_rate": 4.525062899051145e-05,
      "loss": 1.7397,
      "step": 168
    },
    {
      "epoch": 0.7363834422657952,
      "grad_norm": 0.7829226851463318,
      "learning_rate": 4.506278976750763e-05,
      "loss": 1.7315,
      "step": 169
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.8621643781661987,
      "learning_rate": 4.487415762799118e-05,
      "loss": 1.7408,
      "step": 170
    },
    {
      "epoch": 0.7450980392156863,
      "grad_norm": 0.6500158905982971,
      "learning_rate": 4.468474250169896e-05,
      "loss": 1.7465,
      "step": 171
    },
    {
      "epoch": 0.7494553376906318,
      "grad_norm": 0.5809956192970276,
      "learning_rate": 4.4494554359584785e-05,
      "loss": 1.7293,
      "step": 172
    },
    {
      "epoch": 0.7538126361655774,
      "grad_norm": 0.7333821058273315,
      "learning_rate": 4.430360321329463e-05,
      "loss": 1.727,
      "step": 173
    },
    {
      "epoch": 0.7581699346405228,
      "grad_norm": 0.5068430304527283,
      "learning_rate": 4.411189911463955e-05,
      "loss": 1.7167,
      "step": 174
    },
    {
      "epoch": 0.7625272331154684,
      "grad_norm": 0.6892985701560974,
      "learning_rate": 4.3919452155066614e-05,
      "loss": 1.5829,
      "step": 175
    },
    {
      "epoch": 0.7668845315904139,
      "grad_norm": 1.0960310697555542,
      "learning_rate": 4.372627246512761e-05,
      "loss": 1.7752,
      "step": 176
    },
    {
      "epoch": 0.7712418300653595,
      "grad_norm": 0.9808456897735596,
      "learning_rate": 4.3532370213945826e-05,
      "loss": 1.6206,
      "step": 177
    },
    {
      "epoch": 0.775599128540305,
      "grad_norm": 0.9465530514717102,
      "learning_rate": 4.333775560868071e-05,
      "loss": 1.7664,
      "step": 178
    },
    {
      "epoch": 0.7799564270152506,
      "grad_norm": 2.3225557804107666,
      "learning_rate": 4.3142438893990536e-05,
      "loss": 1.7564,
      "step": 179
    },
    {
      "epoch": 0.7843137254901961,
      "grad_norm": 0.833987295627594,
      "learning_rate": 4.294643035149318e-05,
      "loss": 1.7599,
      "step": 180
    },
    {
      "epoch": 0.7886710239651417,
      "grad_norm": 7.097892761230469,
      "learning_rate": 4.274974029922482e-05,
      "loss": 1.7297,
      "step": 181
    },
    {
      "epoch": 0.7930283224400871,
      "grad_norm": 1.9574705362319946,
      "learning_rate": 4.255237909109685e-05,
      "loss": 1.6733,
      "step": 182
    },
    {
      "epoch": 0.7973856209150327,
      "grad_norm": 5.98021125793457,
      "learning_rate": 4.235435711635076e-05,
      "loss": 1.7801,
      "step": 183
    },
    {
      "epoch": 0.8017429193899782,
      "grad_norm": 1.4386473894119263,
      "learning_rate": 4.2155684799011346e-05,
      "loss": 1.7036,
      "step": 184
    },
    {
      "epoch": 0.8061002178649237,
      "grad_norm": 1.5464740991592407,
      "learning_rate": 4.195637259733789e-05,
      "loss": 1.7301,
      "step": 185
    },
    {
      "epoch": 0.8104575163398693,
      "grad_norm": 2.3751494884490967,
      "learning_rate": 4.1756431003273664e-05,
      "loss": 1.7411,
      "step": 186
    },
    {
      "epoch": 0.8148148148148148,
      "grad_norm": 0.7723326086997986,
      "learning_rate": 4.155587054189365e-05,
      "loss": 1.6975,
      "step": 187
    },
    {
      "epoch": 0.8191721132897604,
      "grad_norm": 0.8427643775939941,
      "learning_rate": 4.135470177085042e-05,
      "loss": 1.7452,
      "step": 188
    },
    {
      "epoch": 0.8235294117647058,
      "grad_norm": 0.6999509930610657,
      "learning_rate": 4.11529352798185e-05,
      "loss": 1.8235,
      "step": 189
    },
    {
      "epoch": 0.8278867102396514,
      "grad_norm": 0.6350328326225281,
      "learning_rate": 4.0950581689936744e-05,
      "loss": 1.8013,
      "step": 190
    },
    {
      "epoch": 0.8322440087145969,
      "grad_norm": 0.5478256344795227,
      "learning_rate": 4.074765165324942e-05,
      "loss": 1.8239,
      "step": 191
    },
    {
      "epoch": 0.8366013071895425,
      "grad_norm": 0.7619976997375488,
      "learning_rate": 4.0544155852145334e-05,
      "loss": 1.6571,
      "step": 192
    },
    {
      "epoch": 0.840958605664488,
      "grad_norm": 0.6814029216766357,
      "learning_rate": 4.034010499879557e-05,
      "loss": 1.8015,
      "step": 193
    },
    {
      "epoch": 0.8453159041394336,
      "grad_norm": 1.5520319938659668,
      "learning_rate": 4.0135509834589566e-05,
      "loss": 1.7239,
      "step": 194
    },
    {
      "epoch": 0.8496732026143791,
      "grad_norm": 0.7724117040634155,
      "learning_rate": 3.9930381129569696e-05,
      "loss": 1.6609,
      "step": 195
    },
    {
      "epoch": 0.8540305010893247,
      "grad_norm": 1.1484628915786743,
      "learning_rate": 3.972472968186434e-05,
      "loss": 1.7557,
      "step": 196
    },
    {
      "epoch": 0.8583877995642701,
      "grad_norm": 1.2805622816085815,
      "learning_rate": 3.95185663171194e-05,
      "loss": 1.7035,
      "step": 197
    },
    {
      "epoch": 0.8627450980392157,
      "grad_norm": 0.890356183052063,
      "learning_rate": 3.931190188792853e-05,
      "loss": 1.6805,
      "step": 198
    },
    {
      "epoch": 0.8671023965141612,
      "grad_norm": 1.8623896837234497,
      "learning_rate": 3.910474727326175e-05,
      "loss": 1.6048,
      "step": 199
    },
    {
      "epoch": 0.8714596949891068,
      "grad_norm": 1.6572378873825073,
      "learning_rate": 3.8897113377892805e-05,
      "loss": 1.6146,
      "step": 200
    },
    {
      "epoch": 0.8758169934640523,
      "grad_norm": 0.8090646862983704,
      "learning_rate": 3.8689011131825186e-05,
      "loss": 1.6052,
      "step": 201
    },
    {
      "epoch": 0.8801742919389978,
      "grad_norm": 1.0492998361587524,
      "learning_rate": 3.8480451489716636e-05,
      "loss": 1.8196,
      "step": 202
    },
    {
      "epoch": 0.8845315904139434,
      "grad_norm": 0.5649470686912537,
      "learning_rate": 3.827144543030266e-05,
      "loss": 1.544,
      "step": 203
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.9283618330955505,
      "learning_rate": 3.806200395581841e-05,
      "loss": 1.6018,
      "step": 204
    },
    {
      "epoch": 0.8932461873638344,
      "grad_norm": 1.2240904569625854,
      "learning_rate": 3.785213809141969e-05,
      "loss": 1.8455,
      "step": 205
    },
    {
      "epoch": 0.8976034858387799,
      "grad_norm": 1.7425321340560913,
      "learning_rate": 3.764185888460246e-05,
      "loss": 1.9087,
      "step": 206
    },
    {
      "epoch": 0.9019607843137255,
      "grad_norm": 6.965508460998535,
      "learning_rate": 3.743117740462135e-05,
      "loss": 1.8282,
      "step": 207
    },
    {
      "epoch": 0.906318082788671,
      "grad_norm": 1.315513253211975,
      "learning_rate": 3.722010474190695e-05,
      "loss": 1.8563,
      "step": 208
    },
    {
      "epoch": 0.9106753812636166,
      "grad_norm": 0.5694398283958435,
      "learning_rate": 3.700865200748199e-05,
      "loss": 1.7312,
      "step": 209
    },
    {
      "epoch": 0.9150326797385621,
      "grad_norm": 1.3388159275054932,
      "learning_rate": 3.679683033237648e-05,
      "loss": 1.7612,
      "step": 210
    },
    {
      "epoch": 0.9193899782135077,
      "grad_norm": 1.2813358306884766,
      "learning_rate": 3.658465086704174e-05,
      "loss": 1.6871,
      "step": 211
    },
    {
      "epoch": 0.9237472766884531,
      "grad_norm": 0.9567870497703552,
      "learning_rate": 3.6372124780763396e-05,
      "loss": 1.731,
      "step": 212
    },
    {
      "epoch": 0.9281045751633987,
      "grad_norm": 1.2562271356582642,
      "learning_rate": 3.6159263261073515e-05,
      "loss": 1.6309,
      "step": 213
    },
    {
      "epoch": 0.9324618736383442,
      "grad_norm": 1.0945274829864502,
      "learning_rate": 3.59460775131616e-05,
      "loss": 1.8533,
      "step": 214
    },
    {
      "epoch": 0.9368191721132898,
      "grad_norm": 0.546955406665802,
      "learning_rate": 3.5732578759284776e-05,
      "loss": 1.6818,
      "step": 215
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 0.6648678183555603,
      "learning_rate": 3.551877823817702e-05,
      "loss": 1.8012,
      "step": 216
    },
    {
      "epoch": 0.9455337690631809,
      "grad_norm": 0.9706017374992371,
      "learning_rate": 3.53046872044576e-05,
      "loss": 1.766,
      "step": 217
    },
    {
      "epoch": 0.9498910675381264,
      "grad_norm": 0.685961127281189,
      "learning_rate": 3.5090316928038546e-05,
      "loss": 1.735,
      "step": 218
    },
    {
      "epoch": 0.954248366013072,
      "grad_norm": 1.0843380689620972,
      "learning_rate": 3.487567869353144e-05,
      "loss": 1.5208,
      "step": 219
    },
    {
      "epoch": 0.9586056644880174,
      "grad_norm": 0.7091177701950073,
      "learning_rate": 3.466078379965341e-05,
      "loss": 1.8162,
      "step": 220
    },
    {
      "epoch": 0.9629629629629629,
      "grad_norm": 0.535933792591095,
      "learning_rate": 3.4445643558632295e-05,
      "loss": 1.7281,
      "step": 221
    },
    {
      "epoch": 0.9673202614379085,
      "grad_norm": 0.6523289680480957,
      "learning_rate": 3.423026929561117e-05,
      "loss": 1.7759,
      "step": 222
    },
    {
      "epoch": 0.971677559912854,
      "grad_norm": 0.6564977765083313,
      "learning_rate": 3.401467234805227e-05,
      "loss": 1.5996,
      "step": 223
    },
    {
      "epoch": 0.9760348583877996,
      "grad_norm": 0.7078843712806702,
      "learning_rate": 3.379886406514004e-05,
      "loss": 1.6532,
      "step": 224
    },
    {
      "epoch": 0.9803921568627451,
      "grad_norm": 0.7516049742698669,
      "learning_rate": 3.358285580718383e-05,
      "loss": 1.5464,
      "step": 225
    },
    {
      "epoch": 0.9847494553376906,
      "grad_norm": 0.7365531325340271,
      "learning_rate": 3.336665894501977e-05,
      "loss": 1.7292,
      "step": 226
    },
    {
      "epoch": 0.9891067538126361,
      "grad_norm": 0.9807084202766418,
      "learning_rate": 3.315028485941233e-05,
      "loss": 1.4769,
      "step": 227
    },
    {
      "epoch": 0.9934640522875817,
      "grad_norm": 0.7513118386268616,
      "learning_rate": 3.293374494045507e-05,
      "loss": 1.8098,
      "step": 228
    },
    {
      "epoch": 0.9978213507625272,
      "grad_norm": 0.7991335988044739,
      "learning_rate": 3.271705058697122e-05,
      "loss": 1.7686,
      "step": 229
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.0507432222366333,
      "learning_rate": 3.2500213205913506e-05,
      "loss": 0.9588,
      "step": 230
    }
  ],
  "logging_steps": 1,
  "max_steps": 458,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 23,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.9962116695030497e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}