{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.990689013035381,
  "eval_steps": 500,
  "global_step": 335,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0148975791433892,
      "grad_norm": 5.950184932951318,
      "learning_rate": 2.3529411764705885e-06,
      "loss": 0.7975,
      "step": 1
    },
    {
      "epoch": 0.0297951582867784,
      "grad_norm": 5.889999633747208,
      "learning_rate": 4.705882352941177e-06,
      "loss": 0.7918,
      "step": 2
    },
    {
      "epoch": 0.0446927374301676,
      "grad_norm": 5.479080731430823,
      "learning_rate": 7.058823529411766e-06,
      "loss": 0.7842,
      "step": 3
    },
    {
      "epoch": 0.0595903165735568,
      "grad_norm": 4.072102472920601,
      "learning_rate": 9.411764705882354e-06,
      "loss": 0.7305,
      "step": 4
    },
    {
      "epoch": 0.074487895716946,
      "grad_norm": 2.0993469862660445,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.6919,
      "step": 5
    },
    {
      "epoch": 0.0893854748603352,
      "grad_norm": 5.471012079326577,
      "learning_rate": 1.4117647058823532e-05,
      "loss": 0.7321,
      "step": 6
    },
    {
      "epoch": 0.1042830540037244,
      "grad_norm": 7.497778887994976,
      "learning_rate": 1.647058823529412e-05,
      "loss": 0.7089,
      "step": 7
    },
    {
      "epoch": 0.1191806331471136,
      "grad_norm": 7.744362603097892,
      "learning_rate": 1.8823529411764708e-05,
      "loss": 0.7259,
      "step": 8
    },
    {
      "epoch": 0.1340782122905028,
      "grad_norm": 4.336088431882978,
      "learning_rate": 2.1176470588235296e-05,
      "loss": 0.6722,
      "step": 9
    },
    {
      "epoch": 0.148975791433892,
      "grad_norm": 3.1425590071036407,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.6323,
      "step": 10
    },
    {
      "epoch": 0.16387337057728119,
      "grad_norm": 2.293080261987849,
      "learning_rate": 2.5882352941176475e-05,
      "loss": 0.6058,
      "step": 11
    },
    {
      "epoch": 0.1787709497206704,
      "grad_norm": 1.5511769350436764,
      "learning_rate": 2.8235294117647063e-05,
      "loss": 0.5911,
      "step": 12
    },
    {
      "epoch": 0.19366852886405958,
      "grad_norm": 1.43403410737388,
      "learning_rate": 3.0588235294117644e-05,
      "loss": 0.5715,
      "step": 13
    },
    {
      "epoch": 0.2085661080074488,
      "grad_norm": 1.0282624052041764,
      "learning_rate": 3.294117647058824e-05,
      "loss": 0.5482,
      "step": 14
    },
    {
      "epoch": 0.22346368715083798,
      "grad_norm": 22.401057880346002,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.6398,
      "step": 15
    },
    {
      "epoch": 0.2383612662942272,
      "grad_norm": 4.036215605056668,
      "learning_rate": 3.7647058823529415e-05,
      "loss": 0.5932,
      "step": 16
    },
    {
      "epoch": 0.2532588454376164,
      "grad_norm": 2.1089398812874904,
      "learning_rate": 4e-05,
      "loss": 0.5449,
      "step": 17
    },
    {
      "epoch": 0.2681564245810056,
      "grad_norm": 1.8917019656777936,
      "learning_rate": 4.235294117647059e-05,
      "loss": 0.545,
      "step": 18
    },
    {
      "epoch": 0.28305400372439476,
      "grad_norm": 1.9916573775756592,
      "learning_rate": 4.470588235294118e-05,
      "loss": 0.5414,
      "step": 19
    },
    {
      "epoch": 0.297951582867784,
      "grad_norm": 1.8920562076710838,
      "learning_rate": 4.705882352941177e-05,
      "loss": 0.5407,
      "step": 20
    },
    {
      "epoch": 0.3128491620111732,
      "grad_norm": 1.0243631917296254,
      "learning_rate": 4.941176470588236e-05,
      "loss": 0.5197,
      "step": 21
    },
    {
      "epoch": 0.32774674115456237,
      "grad_norm": 1.8353601120734953,
      "learning_rate": 5.176470588235295e-05,
      "loss": 0.5242,
      "step": 22
    },
    {
      "epoch": 0.3426443202979516,
      "grad_norm": 1.0556187539851671,
      "learning_rate": 5.411764705882354e-05,
      "loss": 0.5115,
      "step": 23
    },
    {
      "epoch": 0.3575418994413408,
      "grad_norm": 1.6827786939863305,
      "learning_rate": 5.6470588235294126e-05,
      "loss": 0.5057,
      "step": 24
    },
    {
      "epoch": 0.37243947858473,
      "grad_norm": 1.3016042554320595,
      "learning_rate": 5.8823529411764714e-05,
      "loss": 0.5114,
      "step": 25
    },
    {
      "epoch": 0.38733705772811916,
      "grad_norm": 1.2977216399034888,
      "learning_rate": 6.117647058823529e-05,
      "loss": 0.4991,
      "step": 26
    },
    {
      "epoch": 0.4022346368715084,
      "grad_norm": 1.223225396686375,
      "learning_rate": 6.352941176470589e-05,
      "loss": 0.4941,
      "step": 27
    },
    {
      "epoch": 0.4171322160148976,
      "grad_norm": 0.9631867501185041,
      "learning_rate": 6.588235294117648e-05,
      "loss": 0.4906,
      "step": 28
    },
    {
      "epoch": 0.43202979515828677,
      "grad_norm": 1.2189198129233703,
      "learning_rate": 6.823529411764707e-05,
      "loss": 0.4913,
      "step": 29
    },
    {
      "epoch": 0.44692737430167595,
      "grad_norm": 0.7374132716188752,
      "learning_rate": 7.058823529411765e-05,
      "loss": 0.4785,
      "step": 30
    },
    {
      "epoch": 0.4618249534450652,
      "grad_norm": 4.329279729767168,
      "learning_rate": 7.294117647058824e-05,
      "loss": 0.4855,
      "step": 31
    },
    {
      "epoch": 0.4767225325884544,
      "grad_norm": 1.9935350979555255,
      "learning_rate": 7.529411764705883e-05,
      "loss": 0.5068,
      "step": 32
    },
    {
      "epoch": 0.49162011173184356,
      "grad_norm": 1.0099023827824438,
      "learning_rate": 7.764705882352942e-05,
      "loss": 0.4878,
      "step": 33
    },
    {
      "epoch": 0.5065176908752328,
      "grad_norm": 1.453165427026921,
      "learning_rate": 8e-05,
      "loss": 0.4887,
      "step": 34
    },
    {
      "epoch": 0.521415270018622,
      "grad_norm": 1.0681365388503938,
      "learning_rate": 7.999782132320701e-05,
      "loss": 0.4792,
      "step": 35
    },
    {
      "epoch": 0.5363128491620112,
      "grad_norm": 1.2991278121605818,
      "learning_rate": 7.999128553015966e-05,
      "loss": 0.4746,
      "step": 36
    },
    {
      "epoch": 0.5512104283054003,
      "grad_norm": 0.9705929073068399,
      "learning_rate": 7.998039333282696e-05,
      "loss": 0.47,
      "step": 37
    },
    {
      "epoch": 0.5661080074487895,
      "grad_norm": 1.288262936234187,
      "learning_rate": 7.996514591773782e-05,
      "loss": 0.4758,
      "step": 38
    },
    {
      "epoch": 0.5810055865921788,
      "grad_norm": 0.6816553781708893,
      "learning_rate": 7.99455449458517e-05,
      "loss": 0.4627,
      "step": 39
    },
    {
      "epoch": 0.595903165735568,
      "grad_norm": 0.9231036421804909,
      "learning_rate": 7.992159255237773e-05,
      "loss": 0.4691,
      "step": 40
    },
    {
      "epoch": 0.6108007448789572,
      "grad_norm": 1.1402552499674083,
      "learning_rate": 7.989329134654207e-05,
      "loss": 0.4638,
      "step": 41
    },
    {
      "epoch": 0.6256983240223464,
      "grad_norm": 0.7253677973155603,
      "learning_rate": 7.986064441130378e-05,
      "loss": 0.4593,
      "step": 42
    },
    {
      "epoch": 0.6405959031657356,
      "grad_norm": 1.1918043936517428,
      "learning_rate": 7.982365530301885e-05,
      "loss": 0.4624,
      "step": 43
    },
    {
      "epoch": 0.6554934823091247,
      "grad_norm": 1.0659452922693904,
      "learning_rate": 7.978232805105288e-05,
      "loss": 0.4514,
      "step": 44
    },
    {
      "epoch": 0.6703910614525139,
      "grad_norm": 0.9974435758912656,
      "learning_rate": 7.97366671573421e-05,
      "loss": 0.4495,
      "step": 45
    },
    {
      "epoch": 0.6852886405959032,
      "grad_norm": 0.741244345164511,
      "learning_rate": 7.968667759590298e-05,
      "loss": 0.4498,
      "step": 46
    },
    {
      "epoch": 0.7001862197392924,
      "grad_norm": 0.700212224489924,
      "learning_rate": 7.96323648122904e-05,
      "loss": 0.4496,
      "step": 47
    },
    {
      "epoch": 0.7150837988826816,
      "grad_norm": 0.6861544024192645,
      "learning_rate": 7.957373472300442e-05,
      "loss": 0.4425,
      "step": 48
    },
    {
      "epoch": 0.7299813780260708,
      "grad_norm": 0.9097284558328156,
      "learning_rate": 7.951079371484578e-05,
      "loss": 0.4435,
      "step": 49
    },
    {
      "epoch": 0.74487895716946,
      "grad_norm": 0.8086828171768753,
      "learning_rate": 7.944354864422016e-05,
      "loss": 0.4427,
      "step": 50
    },
    {
      "epoch": 0.7597765363128491,
      "grad_norm": 0.7336340975598995,
      "learning_rate": 7.937200683639133e-05,
      "loss": 0.4418,
      "step": 51
    },
    {
      "epoch": 0.7746741154562383,
      "grad_norm": 0.709266378863673,
      "learning_rate": 7.929617608468308e-05,
      "loss": 0.4413,
      "step": 52
    },
    {
      "epoch": 0.7895716945996276,
      "grad_norm": 0.6997957286653358,
      "learning_rate": 7.921606464963037e-05,
      "loss": 0.4343,
      "step": 53
    },
    {
      "epoch": 0.8044692737430168,
      "grad_norm": 0.6312526710744601,
      "learning_rate": 7.913168125807943e-05,
      "loss": 0.4357,
      "step": 54
    },
    {
      "epoch": 0.819366852886406,
      "grad_norm": 0.8682993581299321,
      "learning_rate": 7.90430351022371e-05,
      "loss": 0.4256,
      "step": 55
    },
    {
      "epoch": 0.8342644320297952,
      "grad_norm": 1.0211685354189273,
      "learning_rate": 7.895013583866949e-05,
      "loss": 0.4383,
      "step": 56
    },
    {
      "epoch": 0.8491620111731844,
      "grad_norm": 0.5942703334288042,
      "learning_rate": 7.88529935872501e-05,
      "loss": 0.4266,
      "step": 57
    },
    {
      "epoch": 0.8640595903165735,
      "grad_norm": 0.4905469818523975,
      "learning_rate": 7.875161893005736e-05,
      "loss": 0.4275,
      "step": 58
    },
    {
      "epoch": 0.8789571694599627,
      "grad_norm": 0.713524523381961,
      "learning_rate": 7.864602291022193e-05,
      "loss": 0.425,
      "step": 59
    },
    {
      "epoch": 0.8938547486033519,
      "grad_norm": 0.7215883306108987,
      "learning_rate": 7.85362170307237e-05,
      "loss": 0.4296,
      "step": 60
    },
    {
      "epoch": 0.9087523277467412,
      "grad_norm": 0.5476669418552049,
      "learning_rate": 7.842221325313873e-05,
      "loss": 0.4264,
      "step": 61
    },
    {
      "epoch": 0.9236499068901304,
      "grad_norm": 0.5026608565405364,
      "learning_rate": 7.830402399633624e-05,
      "loss": 0.4279,
      "step": 62
    },
    {
      "epoch": 0.9385474860335196,
      "grad_norm": 0.43642895836746987,
      "learning_rate": 7.818166213512581e-05,
      "loss": 0.4231,
      "step": 63
    },
    {
      "epoch": 0.9534450651769087,
      "grad_norm": 0.5516886415086357,
      "learning_rate": 7.805514099885479e-05,
      "loss": 0.4286,
      "step": 64
    },
    {
      "epoch": 0.9683426443202979,
      "grad_norm": 0.6202502097645083,
      "learning_rate": 7.792447436995634e-05,
      "loss": 0.4238,
      "step": 65
    },
    {
      "epoch": 0.9832402234636871,
      "grad_norm": 0.8483095514593049,
      "learning_rate": 7.778967648244807e-05,
      "loss": 0.4251,
      "step": 66
    },
    {
      "epoch": 0.9981378026070763,
      "grad_norm": 0.9010008832838037,
      "learning_rate": 7.765076202038145e-05,
      "loss": 0.4227,
      "step": 67
    },
    {
      "epoch": 1.0130353817504656,
      "grad_norm": 1.1210899234881495,
      "learning_rate": 7.750774611624222e-05,
      "loss": 0.7678,
      "step": 68
    },
    {
      "epoch": 1.0279329608938548,
      "grad_norm": 1.4294187108260912,
      "learning_rate": 7.736064434930193e-05,
      "loss": 0.4298,
      "step": 69
    },
    {
      "epoch": 1.042830540037244,
      "grad_norm": 0.5162696432741285,
      "learning_rate": 7.720947274392087e-05,
      "loss": 0.4051,
      "step": 70
    },
    {
      "epoch": 1.0577281191806331,
      "grad_norm": 1.2710462444765749,
      "learning_rate": 7.705424776780249e-05,
      "loss": 0.4149,
      "step": 71
    },
    {
      "epoch": 1.0726256983240223,
      "grad_norm": 0.7995934416533542,
      "learning_rate": 7.689498633019941e-05,
      "loss": 0.4007,
      "step": 72
    },
    {
      "epoch": 1.0875232774674115,
      "grad_norm": 0.880390966869298,
      "learning_rate": 7.673170578007157e-05,
      "loss": 0.4116,
      "step": 73
    },
    {
      "epoch": 1.1024208566108007,
      "grad_norm": 0.7628454204466086,
      "learning_rate": 7.656442390419622e-05,
      "loss": 0.4008,
      "step": 74
    },
    {
      "epoch": 1.1173184357541899,
      "grad_norm": 0.5370866632430045,
      "learning_rate": 7.63931589252304e-05,
      "loss": 0.3935,
      "step": 75
    },
    {
      "epoch": 1.132216014897579,
      "grad_norm": 0.6503753033899171,
      "learning_rate": 7.621792949972588e-05,
      "loss": 0.3983,
      "step": 76
    },
    {
      "epoch": 1.1471135940409685,
      "grad_norm": 0.483185187236749,
      "learning_rate": 7.603875471609677e-05,
      "loss": 0.4092,
      "step": 77
    },
    {
      "epoch": 1.1620111731843576,
      "grad_norm": 0.5469246081737635,
      "learning_rate": 7.585565409254025e-05,
      "loss": 0.3988,
      "step": 78
    },
    {
      "epoch": 1.1769087523277468,
      "grad_norm": 0.3809737733256841,
      "learning_rate": 7.566864757491027e-05,
      "loss": 0.4049,
      "step": 79
    },
    {
      "epoch": 1.191806331471136,
      "grad_norm": 0.4978345458094043,
      "learning_rate": 7.547775553454485e-05,
      "loss": 0.3991,
      "step": 80
    },
    {
      "epoch": 1.2067039106145252,
      "grad_norm": 0.3419277603644534,
      "learning_rate": 7.528299876604689e-05,
      "loss": 0.4015,
      "step": 81
    },
    {
      "epoch": 1.2216014897579144,
      "grad_norm": 0.3978591421270835,
      "learning_rate": 7.508439848501899e-05,
      "loss": 0.3948,
      "step": 82
    },
    {
      "epoch": 1.2364990689013036,
      "grad_norm": 0.3215672869140669,
      "learning_rate": 7.488197632575232e-05,
      "loss": 0.3946,
      "step": 83
    },
    {
      "epoch": 1.2513966480446927,
      "grad_norm": 0.2805092237683431,
      "learning_rate": 7.467575433886989e-05,
      "loss": 0.3925,
      "step": 84
    },
    {
      "epoch": 1.266294227188082,
      "grad_norm": 0.29668081700935095,
      "learning_rate": 7.44657549889246e-05,
      "loss": 0.3944,
      "step": 85
    },
    {
      "epoch": 1.2811918063314711,
      "grad_norm": 0.32093150433950507,
      "learning_rate": 7.425200115195193e-05,
      "loss": 0.395,
      "step": 86
    },
    {
      "epoch": 1.2960893854748603,
      "grad_norm": 0.3114671727422004,
      "learning_rate": 7.403451611297808e-05,
      "loss": 0.3896,
      "step": 87
    },
    {
      "epoch": 1.3109869646182495,
      "grad_norm": 0.2621069849515514,
      "learning_rate": 7.381332356348343e-05,
      "loss": 0.3972,
      "step": 88
    },
    {
      "epoch": 1.3258845437616387,
      "grad_norm": 0.35041652635951187,
      "learning_rate": 7.358844759882168e-05,
      "loss": 0.3947,
      "step": 89
    },
    {
      "epoch": 1.3407821229050279,
      "grad_norm": 0.35165133417817274,
      "learning_rate": 7.335991271559512e-05,
      "loss": 0.3924,
      "step": 90
    },
    {
      "epoch": 1.355679702048417,
      "grad_norm": 0.3622625923423102,
      "learning_rate": 7.312774380898608e-05,
      "loss": 0.3802,
      "step": 91
    },
    {
      "epoch": 1.3705772811918062,
      "grad_norm": 0.34568371596915276,
      "learning_rate": 7.289196617004499e-05,
      "loss": 0.3921,
      "step": 92
    },
    {
      "epoch": 1.3854748603351954,
      "grad_norm": 0.36106236048997714,
      "learning_rate": 7.265260548293535e-05,
      "loss": 0.3852,
      "step": 93
    },
    {
      "epoch": 1.4003724394785848,
      "grad_norm": 0.3591694116875136,
      "learning_rate": 7.24096878221359e-05,
      "loss": 0.3864,
      "step": 94
    },
    {
      "epoch": 1.415270018621974,
      "grad_norm": 0.4611887631357433,
      "learning_rate": 7.216323964960013e-05,
      "loss": 0.3917,
      "step": 95
    },
    {
      "epoch": 1.4301675977653632,
      "grad_norm": 0.5900177988285087,
      "learning_rate": 7.191328781187374e-05,
      "loss": 0.3889,
      "step": 96
    },
    {
      "epoch": 1.4450651769087524,
      "grad_norm": 0.6476785738427469,
      "learning_rate": 7.165985953717017e-05,
      "loss": 0.3976,
      "step": 97
    },
    {
      "epoch": 1.4599627560521415,
      "grad_norm": 0.5959898548929913,
      "learning_rate": 7.140298243240444e-05,
      "loss": 0.3991,
      "step": 98
    },
    {
      "epoch": 1.4748603351955307,
      "grad_norm": 0.45154508108061203,
      "learning_rate": 7.114268448018589e-05,
      "loss": 0.395,
      "step": 99
    },
    {
      "epoch": 1.48975791433892,
      "grad_norm": 0.41405552449106015,
      "learning_rate": 7.087899403576992e-05,
      "loss": 0.3904,
      "step": 100
    },
    {
      "epoch": 1.504655493482309,
      "grad_norm": 0.550596726311639,
      "learning_rate": 7.06119398239691e-05,
      "loss": 0.3809,
      "step": 101
    },
    {
      "epoch": 1.5195530726256983,
      "grad_norm": 0.6087570876386633,
      "learning_rate": 7.034155093602413e-05,
      "loss": 0.3962,
      "step": 102
    },
    {
      "epoch": 1.5344506517690877,
      "grad_norm": 0.533129391668596,
      "learning_rate": 7.006785682643479e-05,
      "loss": 0.3976,
      "step": 103
    },
    {
      "epoch": 1.5493482309124769,
      "grad_norm": 0.43458513735835697,
      "learning_rate": 6.979088730975128e-05,
      "loss": 0.3917,
      "step": 104
    },
    {
      "epoch": 1.564245810055866,
      "grad_norm": 0.3518341619336429,
      "learning_rate": 6.951067255732655e-05,
      "loss": 0.3858,
      "step": 105
    },
    {
      "epoch": 1.5791433891992552,
      "grad_norm": 0.40015247854184544,
      "learning_rate": 6.92272430940295e-05,
      "loss": 0.3864,
      "step": 106
    },
    {
      "epoch": 1.5940409683426444,
      "grad_norm": 0.3910716089734902,
      "learning_rate": 6.894062979491987e-05,
      "loss": 0.3883,
      "step": 107
    },
    {
      "epoch": 1.6089385474860336,
      "grad_norm": 0.35066296337841696,
      "learning_rate": 6.865086388188476e-05,
      "loss": 0.386,
      "step": 108
    },
    {
      "epoch": 1.6238361266294228,
      "grad_norm": 0.3072713840073755,
      "learning_rate": 6.835797692023774e-05,
      "loss": 0.39,
      "step": 109
    },
    {
      "epoch": 1.638733705772812,
      "grad_norm": 0.3550143349129471,
      "learning_rate": 6.806200081528008e-05,
      "loss": 0.3872,
      "step": 110
    },
    {
      "epoch": 1.6536312849162011,
      "grad_norm": 0.37053547013887106,
      "learning_rate": 6.776296780882537e-05,
      "loss": 0.3813,
      "step": 111
    },
    {
      "epoch": 1.6685288640595903,
      "grad_norm": 0.24071852433891697,
      "learning_rate": 6.746091047568716e-05,
      "loss": 0.3812,
      "step": 112
    },
    {
      "epoch": 1.6834264432029795,
      "grad_norm": 0.3042807960699796,
      "learning_rate": 6.715586172013054e-05,
      "loss": 0.3835,
      "step": 113
    },
    {
      "epoch": 1.6983240223463687,
      "grad_norm": 0.3538152445413312,
      "learning_rate": 6.684785477228777e-05,
      "loss": 0.3813,
      "step": 114
    },
    {
      "epoch": 1.7132216014897579,
      "grad_norm": 0.24086260757195502,
      "learning_rate": 6.653692318453831e-05,
      "loss": 0.3852,
      "step": 115
    },
    {
      "epoch": 1.728119180633147,
      "grad_norm": 0.3025847544952239,
      "learning_rate": 6.622310082785384e-05,
      "loss": 0.3913,
      "step": 116
    },
    {
      "epoch": 1.7430167597765363,
      "grad_norm": 0.31641711838551906,
      "learning_rate": 6.590642188810869e-05,
      "loss": 0.3827,
      "step": 117
    },
    {
      "epoch": 1.7579143389199254,
      "grad_norm": 0.2688168797114581,
      "learning_rate": 6.558692086235565e-05,
      "loss": 0.3744,
      "step": 118
    },
    {
      "epoch": 1.7728119180633146,
      "grad_norm": 0.2984297462672916,
      "learning_rate": 6.526463255506828e-05,
      "loss": 0.3811,
      "step": 119
    },
    {
      "epoch": 1.7877094972067038,
      "grad_norm": 0.30746963837396896,
      "learning_rate": 6.493959207434934e-05,
      "loss": 0.3758,
      "step": 120
    },
    {
      "epoch": 1.802607076350093,
      "grad_norm": 0.3230084067714464,
      "learning_rate": 6.461183482810646e-05,
      "loss": 0.3853,
      "step": 121
    },
    {
      "epoch": 1.8175046554934822,
      "grad_norm": 0.4000243849303593,
      "learning_rate": 6.42813965201949e-05,
      "loss": 0.3882,
      "step": 122
    },
    {
      "epoch": 1.8324022346368714,
      "grad_norm": 0.4288088964712755,
      "learning_rate": 6.394831314652835e-05,
      "loss": 0.3858,
      "step": 123
    },
    {
      "epoch": 1.8472998137802608,
      "grad_norm": 0.5085831972309272,
      "learning_rate": 6.361262099115761e-05,
      "loss": 0.3856,
      "step": 124
    },
    {
      "epoch": 1.86219739292365,
      "grad_norm": 0.6105698024621092,
      "learning_rate": 6.327435662231812e-05,
      "loss": 0.3849,
      "step": 125
    },
    {
      "epoch": 1.8770949720670391,
      "grad_norm": 0.5824359200910605,
      "learning_rate": 6.293355688844637e-05,
      "loss": 0.3866,
      "step": 126
    },
    {
      "epoch": 1.8919925512104283,
      "grad_norm": 0.44326416200929236,
      "learning_rate": 6.259025891416594e-05,
      "loss": 0.3737,
      "step": 127
    },
    {
      "epoch": 1.9068901303538175,
      "grad_norm": 0.3430006009849836,
      "learning_rate": 6.224450009624332e-05,
      "loss": 0.3741,
      "step": 128
    },
    {
      "epoch": 1.9217877094972067,
      "grad_norm": 0.3048045239289814,
      "learning_rate": 6.18963180995141e-05,
      "loss": 0.3767,
      "step": 129
    },
    {
      "epoch": 1.9366852886405959,
      "grad_norm": 0.3664037873157842,
      "learning_rate": 6.154575085278012e-05,
      "loss": 0.3857,
      "step": 130
    },
    {
      "epoch": 1.9515828677839853,
      "grad_norm": 0.3595451501590318,
      "learning_rate": 6.119283654467761e-05,
      "loss": 0.3806,
      "step": 131
    },
    {
      "epoch": 1.9664804469273744,
      "grad_norm": 0.3076574382248462,
      "learning_rate": 6.083761361951722e-05,
      "loss": 0.3802,
      "step": 132
    },
    {
      "epoch": 1.9813780260707636,
      "grad_norm": 0.25896414972948856,
      "learning_rate": 6.048012077309612e-05,
      "loss": 0.3773,
      "step": 133
    },
    {
      "epoch": 1.9962756052141528,
      "grad_norm": 0.2438114619698701,
      "learning_rate": 6.01203969484827e-05,
      "loss": 0.3832,
      "step": 134
    },
    {
      "epoch": 2.011173184357542,
      "grad_norm": 0.4880986651628094,
      "learning_rate": 5.975848133177442e-05,
      "loss": 0.7025,
      "step": 135
    },
    {
      "epoch": 2.026070763500931,
      "grad_norm": 0.7683631863694373,
      "learning_rate": 5.939441334782901e-05,
      "loss": 0.3581,
      "step": 136
    },
    {
      "epoch": 2.0409683426443204,
      "grad_norm": 1.0088180311972472,
      "learning_rate": 5.9028232655969866e-05,
      "loss": 0.3662,
      "step": 137
    },
    {
      "epoch": 2.0558659217877095,
      "grad_norm": 0.8036028379230191,
      "learning_rate": 5.865997914566577e-05,
      "loss": 0.3606,
      "step": 138
    },
    {
      "epoch": 2.0707635009310987,
      "grad_norm": 0.6488117999773151,
      "learning_rate": 5.8289692932185546e-05,
      "loss": 0.3619,
      "step": 139
    },
    {
      "epoch": 2.085661080074488,
      "grad_norm": 0.6603945376014646,
      "learning_rate": 5.791741435222821e-05,
      "loss": 0.3613,
      "step": 140
    },
    {
      "epoch": 2.100558659217877,
      "grad_norm": 0.6137565370101866,
      "learning_rate": 5.7543183959528886e-05,
      "loss": 0.3523,
      "step": 141
    },
    {
      "epoch": 2.1154562383612663,
      "grad_norm": 0.6393236118969376,
      "learning_rate": 5.716704252044116e-05,
      "loss": 0.3543,
      "step": 142
    },
    {
      "epoch": 2.1303538175046555,
      "grad_norm": 0.37602818408299177,
      "learning_rate": 5.678903100949625e-05,
      "loss": 0.3564,
      "step": 143
    },
    {
      "epoch": 2.1452513966480447,
      "grad_norm": 0.4873243740997193,
      "learning_rate": 5.640919060493948e-05,
      "loss": 0.3547,
      "step": 144
    },
    {
      "epoch": 2.160148975791434,
      "grad_norm": 0.399733374633036,
      "learning_rate": 5.602756268424457e-05,
      "loss": 0.3521,
      "step": 145
    },
    {
      "epoch": 2.175046554934823,
      "grad_norm": 0.38098879738078906,
      "learning_rate": 5.564418881960624e-05,
      "loss": 0.3525,
      "step": 146
    },
    {
      "epoch": 2.189944134078212,
      "grad_norm": 0.3112584196059825,
      "learning_rate": 5.5259110773411596e-05,
      "loss": 0.3578,
      "step": 147
    },
    {
      "epoch": 2.2048417132216014,
      "grad_norm": 0.33974576612810525,
      "learning_rate": 5.487237049369075e-05,
      "loss": 0.3485,
      "step": 148
    },
    {
      "epoch": 2.2197392923649906,
      "grad_norm": 0.31599699558663985,
      "learning_rate": 5.448401010954733e-05,
      "loss": 0.3482,
      "step": 149
    },
    {
      "epoch": 2.2346368715083798,
      "grad_norm": 0.23588411232355225,
      "learning_rate": 5.4094071926569146e-05,
      "loss": 0.3465,
      "step": 150
    },
    {
      "epoch": 2.249534450651769,
      "grad_norm": 0.259141065976224,
      "learning_rate": 5.370259842221972e-05,
      "loss": 0.3574,
      "step": 151
    },
    {
      "epoch": 2.264432029795158,
      "grad_norm": 0.2948627119507419,
      "learning_rate": 5.330963224121096e-05,
      "loss": 0.356,
      "step": 152
    },
    {
      "epoch": 2.2793296089385473,
      "grad_norm": 0.2398464878022089,
      "learning_rate": 5.291521619085785e-05,
      "loss": 0.3536,
      "step": 153
    },
    {
      "epoch": 2.294227188081937,
      "grad_norm": 0.2554013267182894,
      "learning_rate": 5.251939323641516e-05,
      "loss": 0.3494,
      "step": 154
    },
    {
      "epoch": 2.3091247672253257,
      "grad_norm": 0.2781655892263444,
      "learning_rate": 5.212220649639715e-05,
      "loss": 0.3511,
      "step": 155
    },
    {
      "epoch": 2.3240223463687153,
      "grad_norm": 0.22065610655196213,
      "learning_rate": 5.172369923788046e-05,
      "loss": 0.3589,
      "step": 156
    },
    {
      "epoch": 2.338919925512104,
      "grad_norm": 0.2632172433601213,
      "learning_rate": 5.132391487179088e-05,
      "loss": 0.3537,
      "step": 157
    },
    {
      "epoch": 2.3538175046554937,
      "grad_norm": 0.18801127961637412,
      "learning_rate": 5.092289694817446e-05,
      "loss": 0.3483,
      "step": 158
    },
    {
      "epoch": 2.368715083798883,
      "grad_norm": 0.2368052254172045,
      "learning_rate": 5.052068915145336e-05,
      "loss": 0.3571,
      "step": 159
    },
    {
      "epoch": 2.383612662942272,
      "grad_norm": 0.23655539210427956,
      "learning_rate": 5.011733529566723e-05,
      "loss": 0.3446,
      "step": 160
    },
    {
      "epoch": 2.398510242085661,
      "grad_norm": 0.2138112386109768,
      "learning_rate": 4.971287931970033e-05,
      "loss": 0.3482,
      "step": 161
    },
    {
      "epoch": 2.4134078212290504,
      "grad_norm": 0.23467262592443583,
      "learning_rate": 4.9307365282495075e-05,
      "loss": 0.3473,
      "step": 162
    },
    {
      "epoch": 2.4283054003724396,
      "grad_norm": 0.1804084648863834,
      "learning_rate": 4.890083735825258e-05,
      "loss": 0.3531,
      "step": 163
    },
    {
      "epoch": 2.4432029795158288,
      "grad_norm": 0.22285709273589918,
      "learning_rate": 4.849333983162056e-05,
      "loss": 0.3516,
      "step": 164
    },
    {
      "epoch": 2.458100558659218,
      "grad_norm": 0.14828204605953096,
      "learning_rate": 4.808491709286921e-05,
      "loss": 0.356,
      "step": 165
    },
    {
      "epoch": 2.472998137802607,
      "grad_norm": 0.1797938362516586,
      "learning_rate": 4.76756136330557e-05,
      "loss": 0.3491,
      "step": 166
    },
    {
      "epoch": 2.4878957169459963,
      "grad_norm": 0.15466564418943565,
      "learning_rate": 4.726547403917746e-05,
      "loss": 0.3435,
      "step": 167
    },
    {
      "epoch": 2.5027932960893855,
      "grad_norm": 0.1866313977972178,
      "learning_rate": 4.685454298931527e-05,
      "loss": 0.3506,
      "step": 168
    },
    {
      "epoch": 2.5176908752327747,
      "grad_norm": 0.15144628774380284,
      "learning_rate": 4.6442865247766203e-05,
      "loss": 0.3547,
      "step": 169
    },
    {
      "epoch": 2.532588454376164,
      "grad_norm": 0.17455463171543423,
      "learning_rate": 4.603048566016735e-05,
      "loss": 0.3486,
      "step": 170
    },
    {
      "epoch": 2.547486033519553,
      "grad_norm": 0.17359950511666877,
      "learning_rate": 4.5617449148610584e-05,
      "loss": 0.3483,
      "step": 171
    },
    {
      "epoch": 2.5623836126629422,
      "grad_norm": 0.16169017429437474,
      "learning_rate": 4.520380070674902e-05,
      "loss": 0.3477,
      "step": 172
    },
    {
      "epoch": 2.5772811918063314,
      "grad_norm": 0.17061911155312887,
      "learning_rate": 4.478958539489569e-05,
      "loss": 0.3571,
      "step": 173
    },
    {
      "epoch": 2.5921787709497206,
      "grad_norm": 0.16695037154209072,
      "learning_rate": 4.437484833511499e-05,
      "loss": 0.3522,
      "step": 174
    },
    {
      "epoch": 2.60707635009311,
      "grad_norm": 0.15955082753766311,
      "learning_rate": 4.395963470630723e-05,
      "loss": 0.3413,
      "step": 175
    },
    {
      "epoch": 2.621973929236499,
      "grad_norm": 0.1573403347105783,
      "learning_rate": 4.3543989739287326e-05,
      "loss": 0.3477,
      "step": 176
    },
    {
      "epoch": 2.636871508379888,
      "grad_norm": 0.14729071549065167,
      "learning_rate": 4.312795871185742e-05,
      "loss": 0.3483,
      "step": 177
    },
    {
      "epoch": 2.6517690875232773,
      "grad_norm": 0.1625891228202278,
      "learning_rate": 4.2711586943874774e-05,
      "loss": 0.3452,
      "step": 178
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.15062427849577625,
      "learning_rate": 4.2294919792314794e-05,
      "loss": 0.3488,
      "step": 179
    },
    {
      "epoch": 2.6815642458100557,
      "grad_norm": 0.14387393723407899,
      "learning_rate": 4.1878002646330144e-05,
      "loss": 0.3524,
      "step": 180
    },
    {
      "epoch": 2.6964618249534453,
      "grad_norm": 0.14777699560280422,
      "learning_rate": 4.1460880922306367e-05,
      "loss": 0.3458,
      "step": 181
    },
    {
      "epoch": 2.711359404096834,
      "grad_norm": 0.1590574368277688,
      "learning_rate": 4.1043600058914436e-05,
      "loss": 0.3479,
      "step": 182
    },
    {
      "epoch": 2.7262569832402237,
      "grad_norm": 0.1285624178683975,
      "learning_rate": 4.0626205512161034e-05,
      "loss": 0.3503,
      "step": 183
    },
    {
      "epoch": 2.7411545623836124,
      "grad_norm": 0.14369454205895854,
      "learning_rate": 4.020874275043679e-05,
      "loss": 0.346,
      "step": 184
    },
    {
      "epoch": 2.756052141527002,
      "grad_norm": 0.12193167492676163,
      "learning_rate": 3.979125724956324e-05,
      "loss": 0.3506,
      "step": 185
    },
    {
      "epoch": 2.770949720670391,
      "grad_norm": 0.1326021753158276,
      "learning_rate": 3.937379448783898e-05,
      "loss": 0.3502,
      "step": 186
    },
    {
      "epoch": 2.7858472998137804,
      "grad_norm": 0.12810706794198765,
      "learning_rate": 3.895639994108558e-05,
      "loss": 0.3541,
      "step": 187
    },
    {
      "epoch": 2.8007448789571696,
      "grad_norm": 0.12656208043359207,
      "learning_rate": 3.853911907769365e-05,
      "loss": 0.3497,
      "step": 188
    },
    {
      "epoch": 2.815642458100559,
      "grad_norm": 0.11404167595412021,
      "learning_rate": 3.812199735366986e-05,
      "loss": 0.3473,
      "step": 189
    },
    {
      "epoch": 2.830540037243948,
      "grad_norm": 0.13878407632584827,
      "learning_rate": 3.770508020768522e-05,
      "loss": 0.3489,
      "step": 190
    },
    {
      "epoch": 2.845437616387337,
      "grad_norm": 0.12311309579828897,
      "learning_rate": 3.728841305612524e-05,
      "loss": 0.3464,
      "step": 191
    },
    {
      "epoch": 2.8603351955307263,
      "grad_norm": 0.13332912284616857,
      "learning_rate": 3.687204128814259e-05,
      "loss": 0.3428,
      "step": 192
    },
    {
      "epoch": 2.8752327746741155,
      "grad_norm": 0.1384361081760211,
      "learning_rate": 3.645601026071269e-05,
      "loss": 0.3505,
      "step": 193
    },
    {
      "epoch": 2.8901303538175047,
      "grad_norm": 0.13736333082695462,
      "learning_rate": 3.604036529369277e-05,
      "loss": 0.347,
      "step": 194
    },
    {
      "epoch": 2.905027932960894,
      "grad_norm": 0.11583463546593212,
      "learning_rate": 3.5625151664885036e-05,
      "loss": 0.3461,
      "step": 195
    },
    {
      "epoch": 2.919925512104283,
      "grad_norm": 0.12889980341532484,
      "learning_rate": 3.5210414605104314e-05,
      "loss": 0.3428,
      "step": 196
    },
    {
      "epoch": 2.9348230912476723,
      "grad_norm": 0.11316852370701413,
      "learning_rate": 3.4796199293250987e-05,
      "loss": 0.3479,
      "step": 197
    },
    {
      "epoch": 2.9497206703910615,
      "grad_norm": 0.11454972761265435,
      "learning_rate": 3.438255085138943e-05,
      "loss": 0.3464,
      "step": 198
    },
    {
      "epoch": 2.9646182495344506,
      "grad_norm": 0.13041453651138907,
      "learning_rate": 3.396951433983266e-05,
      "loss": 0.3458,
      "step": 199
    },
    {
      "epoch": 2.97951582867784,
      "grad_norm": 0.12148037255351198,
      "learning_rate": 3.355713475223382e-05,
      "loss": 0.3463,
      "step": 200
    },
    {
      "epoch": 2.994413407821229,
      "grad_norm": 0.1255719642211306,
      "learning_rate": 3.314545701068475e-05,
      "loss": 0.3491,
      "step": 201
    },
    {
      "epoch": 3.009310986964618,
      "grad_norm": 0.3022359318173022,
      "learning_rate": 3.2734525960822545e-05,
      "loss": 0.6352,
      "step": 202
    },
    {
      "epoch": 3.0242085661080074,
      "grad_norm": 0.24062134731666432,
      "learning_rate": 3.232438636694431e-05,
      "loss": 0.3224,
      "step": 203
    },
    {
      "epoch": 3.0391061452513966,
      "grad_norm": 0.18412193702307067,
      "learning_rate": 3.191508290713079e-05,
      "loss": 0.3206,
      "step": 204
    },
    {
      "epoch": 3.0540037243947857,
      "grad_norm": 0.20860348558828723,
      "learning_rate": 3.150666016837947e-05,
      "loss": 0.3264,
      "step": 205
    },
    {
      "epoch": 3.068901303538175,
      "grad_norm": 0.20845238316157091,
      "learning_rate": 3.109916264174743e-05,
      "loss": 0.327,
      "step": 206
    },
    {
      "epoch": 3.083798882681564,
      "grad_norm": 0.20010239080813352,
      "learning_rate": 3.069263471750493e-05,
      "loss": 0.3223,
      "step": 207
    },
    {
      "epoch": 3.0986964618249533,
      "grad_norm": 0.18760823228422985,
      "learning_rate": 3.0287120680299677e-05,
      "loss": 0.3238,
      "step": 208
    },
    {
      "epoch": 3.1135940409683425,
      "grad_norm": 0.21319162553294677,
      "learning_rate": 2.988266470433277e-05,
      "loss": 0.3249,
      "step": 209
    },
    {
      "epoch": 3.1284916201117317,
      "grad_norm": 0.17329474811121787,
      "learning_rate": 2.9479310848546644e-05,
      "loss": 0.3243,
      "step": 210
    },
    {
      "epoch": 3.143389199255121,
      "grad_norm": 0.20404750548823786,
      "learning_rate": 2.9077103051825567e-05,
      "loss": 0.3203,
      "step": 211
    },
    {
      "epoch": 3.1582867783985105,
      "grad_norm": 0.17940098271660299,
      "learning_rate": 2.8676085128209133e-05,
      "loss": 0.3237,
      "step": 212
    },
    {
      "epoch": 3.1731843575418996,
      "grad_norm": 0.16237217554172287,
      "learning_rate": 2.8276300762119553e-05,
      "loss": 0.3196,
      "step": 213
    },
    {
      "epoch": 3.188081936685289,
      "grad_norm": 0.15801986044567953,
      "learning_rate": 2.787779350360286e-05,
      "loss": 0.3196,
      "step": 214
    },
    {
      "epoch": 3.202979515828678,
      "grad_norm": 0.1485056929542617,
      "learning_rate": 2.748060676358484e-05,
      "loss": 0.3192,
      "step": 215
    },
    {
      "epoch": 3.217877094972067,
      "grad_norm": 0.15223329100287147,
      "learning_rate": 2.7084783809142164e-05,
      "loss": 0.3192,
      "step": 216
    },
    {
      "epoch": 3.2327746741154564,
      "grad_norm": 0.13836533584306443,
      "learning_rate": 2.6690367758789046e-05,
      "loss": 0.328,
      "step": 217
    },
    {
      "epoch": 3.2476722532588456,
      "grad_norm": 0.1351886559378493,
      "learning_rate": 2.6297401577780295e-05,
      "loss": 0.322,
      "step": 218
    },
    {
      "epoch": 3.2625698324022347,
      "grad_norm": 0.12857547848905795,
      "learning_rate": 2.5905928073430854e-05,
      "loss": 0.319,
      "step": 219
    },
    {
      "epoch": 3.277467411545624,
      "grad_norm": 0.1274395295470198,
      "learning_rate": 2.5515989890452674e-05,
      "loss": 0.3237,
      "step": 220
    },
    {
      "epoch": 3.292364990689013,
      "grad_norm": 0.11899376687857942,
      "learning_rate": 2.5127629506309264e-05,
      "loss": 0.3205,
      "step": 221
    },
    {
      "epoch": 3.3072625698324023,
      "grad_norm": 0.1304655308172865,
      "learning_rate": 2.474088922658842e-05,
      "loss": 0.3188,
      "step": 222
    },
    {
      "epoch": 3.3221601489757915,
      "grad_norm": 0.11611090203712811,
      "learning_rate": 2.4355811180393767e-05,
      "loss": 0.3195,
      "step": 223
    },
    {
      "epoch": 3.3370577281191807,
      "grad_norm": 0.11604416368224107,
      "learning_rate": 2.397243731575543e-05,
      "loss": 0.3201,
      "step": 224
    },
    {
      "epoch": 3.35195530726257,
      "grad_norm": 0.11134670769841407,
      "learning_rate": 2.359080939506052e-05,
      "loss": 0.32,
      "step": 225
    },
    {
      "epoch": 3.366852886405959,
      "grad_norm": 0.10870133266703798,
      "learning_rate": 2.3210968990503755e-05,
      "loss": 0.3208,
      "step": 226
    },
    {
      "epoch": 3.381750465549348,
      "grad_norm": 0.1110502941686079,
      "learning_rate": 2.2832957479558866e-05,
      "loss": 0.317,
      "step": 227
    },
    {
      "epoch": 3.3966480446927374,
      "grad_norm": 0.10438158313371027,
      "learning_rate": 2.245681604047114e-05,
      "loss": 0.3179,
      "step": 228
    },
    {
      "epoch": 3.4115456238361266,
      "grad_norm": 0.10282665897445756,
      "learning_rate": 2.2082585647771807e-05,
      "loss": 0.3228,
      "step": 229
    },
    {
      "epoch": 3.4264432029795158,
      "grad_norm": 0.10672800838711503,
      "learning_rate": 2.171030706781446e-05,
      "loss": 0.3222,
      "step": 230
    },
    {
      "epoch": 3.441340782122905,
      "grad_norm": 0.09786842543478856,
      "learning_rate": 2.1340020854334246e-05,
      "loss": 0.3181,
      "step": 231
    },
    {
      "epoch": 3.456238361266294,
      "grad_norm": 0.10404230481342372,
      "learning_rate": 2.0971767344030144e-05,
      "loss": 0.3225,
      "step": 232
    },
    {
      "epoch": 3.4711359404096833,
      "grad_norm": 0.10712648472255477,
      "learning_rate": 2.0605586652170998e-05,
      "loss": 0.321,
      "step": 233
    },
    {
      "epoch": 3.4860335195530725,
      "grad_norm": 0.1044874161797147,
      "learning_rate": 2.0241518668225595e-05,
      "loss": 0.3252,
      "step": 234
    },
    {
      "epoch": 3.5009310986964617,
      "grad_norm": 0.10451587153618398,
      "learning_rate": 1.98796030515173e-05,
      "loss": 0.3173,
      "step": 235
    },
    {
      "epoch": 3.515828677839851,
      "grad_norm": 0.09831156843059215,
      "learning_rate": 1.9519879226903903e-05,
      "loss": 0.3188,
      "step": 236
    },
    {
      "epoch": 3.5307262569832405,
      "grad_norm": 0.09438833971964583,
      "learning_rate": 1.9162386380482795e-05,
      "loss": 0.3174,
      "step": 237
    },
    {
      "epoch": 3.5456238361266292,
      "grad_norm": 0.10433270782878293,
      "learning_rate": 1.88071634553224e-05,
      "loss": 0.3219,
      "step": 238
    },
    {
      "epoch": 3.560521415270019,
      "grad_norm": 0.09292286071725575,
      "learning_rate": 1.845424914721988e-05,
      "loss": 0.3252,
      "step": 239
    },
    {
      "epoch": 3.5754189944134076,
      "grad_norm": 0.09413410282284697,
      "learning_rate": 1.81036819004859e-05,
      "loss": 0.3163,
      "step": 240
    },
    {
      "epoch": 3.5903165735567972,
      "grad_norm": 0.10158989698055416,
      "learning_rate": 1.7755499903756704e-05,
      "loss": 0.3259,
      "step": 241
    },
    {
      "epoch": 3.605214152700186,
      "grad_norm": 0.09464740249041373,
      "learning_rate": 1.7409741085834066e-05,
      "loss": 0.3191,
      "step": 242
    },
    {
      "epoch": 3.6201117318435756,
      "grad_norm": 0.10726107957817572,
      "learning_rate": 1.7066443111553627e-05,
      "loss": 0.3211,
      "step": 243
    },
    {
      "epoch": 3.635009310986965,
      "grad_norm": 0.08917870235234628,
      "learning_rate": 1.6725643377681893e-05,
      "loss": 0.3225,
      "step": 244
    },
    {
      "epoch": 3.649906890130354,
      "grad_norm": 0.10677910900640168,
      "learning_rate": 1.638737900884239e-05,
      "loss": 0.3214,
      "step": 245
    },
    {
      "epoch": 3.664804469273743,
      "grad_norm": 0.09774898119089985,
      "learning_rate": 1.6051686853471667e-05,
      "loss": 0.3252,
      "step": 246
    },
    {
      "epoch": 3.6797020484171323,
      "grad_norm": 0.08773810545171451,
      "learning_rate": 1.5718603479805113e-05,
      "loss": 0.3235,
      "step": 247
    },
    {
      "epoch": 3.6945996275605215,
      "grad_norm": 0.10111349625699705,
      "learning_rate": 1.538816517189356e-05,
      "loss": 0.3228,
      "step": 248
    },
    {
      "epoch": 3.7094972067039107,
      "grad_norm": 0.09592458935558433,
      "learning_rate": 1.5060407925650662e-05,
      "loss": 0.3234,
      "step": 249
    },
    {
      "epoch": 3.7243947858473,
      "grad_norm": 0.08715032888076826,
      "learning_rate": 1.4735367444931722e-05,
      "loss": 0.3192,
      "step": 250
    },
    {
      "epoch": 3.739292364990689,
      "grad_norm": 0.09053851330147095,
      "learning_rate": 1.4413079137644358e-05,
      "loss": 0.3162,
      "step": 251
    },
    {
      "epoch": 3.7541899441340782,
      "grad_norm": 0.10377676484499614,
      "learning_rate": 1.4093578111891333e-05,
      "loss": 0.3189,
      "step": 252
    },
    {
      "epoch": 3.7690875232774674,
      "grad_norm": 0.08949080189520761,
      "learning_rate": 1.377689917214617e-05,
      "loss": 0.3258,
      "step": 253
    },
    {
      "epoch": 3.7839851024208566,
      "grad_norm": 0.09258472675428549,
      "learning_rate": 1.3463076815461703e-05,
      "loss": 0.3195,
      "step": 254
    },
    {
      "epoch": 3.798882681564246,
      "grad_norm": 0.0951200280500826,
      "learning_rate": 1.3152145227712221e-05,
      "loss": 0.314,
      "step": 255
    },
    {
      "epoch": 3.813780260707635,
      "grad_norm": 0.09226678982549726,
      "learning_rate": 1.284413827986946e-05,
      "loss": 0.3255,
      "step": 256
    },
    {
      "epoch": 3.828677839851024,
      "grad_norm": 0.08915216421489361,
      "learning_rate": 1.2539089524312855e-05,
      "loss": 0.3261,
      "step": 257
    },
    {
      "epoch": 3.8435754189944134,
      "grad_norm": 0.0925569123285621,
      "learning_rate": 1.2237032191174642e-05,
      "loss": 0.3202,
      "step": 258
    },
    {
      "epoch": 3.8584729981378025,
      "grad_norm": 0.08684065982720417,
      "learning_rate": 1.1937999184719926e-05,
      "loss": 0.3212,
      "step": 259
    },
    {
      "epoch": 3.8733705772811917,
      "grad_norm": 0.08765050211323365,
      "learning_rate": 1.1642023079762281e-05,
      "loss": 0.319,
      "step": 260
    },
    {
      "epoch": 3.888268156424581,
      "grad_norm": 0.08724442004118642,
      "learning_rate": 1.1349136118115242e-05,
      "loss": 0.3154,
      "step": 261
    },
    {
      "epoch": 3.90316573556797,
      "grad_norm": 0.08586178309454286,
      "learning_rate": 1.1059370205080157e-05,
      "loss": 0.3226,
      "step": 262
    },
    {
      "epoch": 3.9180633147113593,
      "grad_norm": 0.08501873030376662,
      "learning_rate": 1.07727569059705e-05,
      "loss": 0.3229,
      "step": 263
    },
    {
      "epoch": 3.9329608938547485,
      "grad_norm": 0.09206134247645327,
      "learning_rate": 1.0489327442673459e-05,
      "loss": 0.3227,
      "step": 264
    },
    {
      "epoch": 3.9478584729981376,
      "grad_norm": 0.08410101453326498,
      "learning_rate": 1.0209112690248726e-05,
      "loss": 0.3177,
      "step": 265
    },
    {
      "epoch": 3.9627560521415273,
      "grad_norm": 0.08821111302237976,
      "learning_rate": 9.932143173565225e-06,
      "loss": 0.3227,
      "step": 266
    },
    {
      "epoch": 3.977653631284916,
      "grad_norm": 0.08616646529797505,
      "learning_rate": 9.658449063975875e-06,
      "loss": 0.3216,
      "step": 267
    },
    {
      "epoch": 3.9925512104283056,
      "grad_norm": 0.0885838583519875,
      "learning_rate": 9.388060176030907e-06,
      "loss": 0.3226,
      "step": 268
    },
    {
      "epoch": 4.007448789571694,
      "grad_norm": 0.22436110155064673,
      "learning_rate": 9.12100596423009e-06,
      "loss": 0.577,
      "step": 269
    },
    {
      "epoch": 4.022346368715084,
      "grad_norm": 0.11607630428391699,
      "learning_rate": 8.857315519814111e-06,
      "loss": 0.306,
      "step": 270
    },
    {
      "epoch": 4.037243947858473,
      "grad_norm": 0.11774451222490284,
      "learning_rate": 8.597017567595562e-06,
      "loss": 0.3064,
      "step": 271
    },
    {
      "epoch": 4.052141527001862,
      "grad_norm": 0.09907950082186576,
      "learning_rate": 8.34014046282984e-06,
      "loss": 0.302,
      "step": 272
    },
    {
      "epoch": 4.067039106145251,
      "grad_norm": 0.10570787971335635,
      "learning_rate": 8.086712188126263e-06,
      "loss": 0.3021,
      "step": 273
    },
    {
      "epoch": 4.081936685288641,
      "grad_norm": 0.10877019752281963,
      "learning_rate": 7.836760350399881e-06,
      "loss": 0.2992,
      "step": 274
    },
    {
      "epoch": 4.0968342644320295,
      "grad_norm": 0.1173773952444213,
      "learning_rate": 7.5903121778641096e-06,
      "loss": 0.3077,
      "step": 275
    },
    {
      "epoch": 4.111731843575419,
      "grad_norm": 0.10898666220477518,
      "learning_rate": 7.347394517064663e-06,
      "loss": 0.3068,
      "step": 276
    },
    {
      "epoch": 4.126629422718808,
      "grad_norm": 0.09803810087456757,
      "learning_rate": 7.108033829955028e-06,
      "loss": 0.3077,
      "step": 277
    },
    {
      "epoch": 4.1415270018621975,
      "grad_norm": 0.09446749801813671,
      "learning_rate": 6.87225619101394e-06,
      "loss": 0.3061,
      "step": 278
    },
    {
      "epoch": 4.156424581005586,
      "grad_norm": 0.0951410747387855,
      "learning_rate": 6.640087284404888e-06,
      "loss": 0.3054,
      "step": 279
    },
    {
      "epoch": 4.171322160148976,
      "grad_norm": 0.10491597814630026,
      "learning_rate": 6.411552401178327e-06,
      "loss": 0.3098,
      "step": 280
    },
    {
      "epoch": 4.186219739292365,
      "grad_norm": 0.09593380058395089,
      "learning_rate": 6.186676436516581e-06,
      "loss": 0.3008,
      "step": 281
    },
    {
      "epoch": 4.201117318435754,
      "grad_norm": 0.09551371288826128,
      "learning_rate": 5.965483887021934e-06,
      "loss": 0.3087,
      "step": 282
    },
    {
      "epoch": 4.216014897579143,
      "grad_norm": 0.0895153508605015,
      "learning_rate": 5.747998848048091e-06,
      "loss": 0.3091,
      "step": 283
    },
    {
      "epoch": 4.230912476722533,
      "grad_norm": 0.09197880309307534,
      "learning_rate": 5.534245011075414e-06,
      "loss": 0.298,
      "step": 284
    },
    {
      "epoch": 4.245810055865922,
      "grad_norm": 0.08575532851271551,
      "learning_rate": 5.3242456611301095e-06,
      "loss": 0.3,
      "step": 285
    },
    {
      "epoch": 4.260707635009311,
      "grad_norm": 0.08276481700085347,
      "learning_rate": 5.118023674247692e-06,
      "loss": 0.2984,
      "step": 286
    },
    {
      "epoch": 4.275605214152701,
      "grad_norm": 0.08713205903583655,
      "learning_rate": 4.915601514981024e-06,
      "loss": 0.3078,
      "step": 287
    },
    {
      "epoch": 4.290502793296089,
      "grad_norm": 0.08363731970724225,
      "learning_rate": 4.717001233953116e-06,
      "loss": 0.2987,
      "step": 288
    },
    {
      "epoch": 4.305400372439479,
      "grad_norm": 0.08529607146039025,
      "learning_rate": 4.522244465455154e-06,
      "loss": 0.3038,
      "step": 289
    },
    {
      "epoch": 4.320297951582868,
      "grad_norm": 0.08215569802598276,
      "learning_rate": 4.33135242508973e-06,
      "loss": 0.3043,
      "step": 290
    },
    {
      "epoch": 4.335195530726257,
      "grad_norm": 0.08149200923150887,
      "learning_rate": 4.144345907459753e-06,
      "loss": 0.3013,
      "step": 291
    },
    {
      "epoch": 4.350093109869646,
      "grad_norm": 0.08086624632566386,
      "learning_rate": 3.961245283903239e-06,
      "loss": 0.308,
      "step": 292
    },
    {
      "epoch": 4.364990689013036,
      "grad_norm": 0.07896402407099593,
      "learning_rate": 3.7820705002741353e-06,
      "loss": 0.3052,
      "step": 293
    },
    {
      "epoch": 4.379888268156424,
      "grad_norm": 0.08231952485165063,
      "learning_rate": 3.6068410747696112e-06,
      "loss": 0.3002,
      "step": 294
    },
    {
      "epoch": 4.394785847299814,
      "grad_norm": 0.07789166320140198,
      "learning_rate": 3.435576095803792e-06,
      "loss": 0.3021,
      "step": 295
    },
    {
      "epoch": 4.409683426443203,
      "grad_norm": 0.07892696916845467,
      "learning_rate": 3.268294219928434e-06,
      "loss": 0.3046,
      "step": 296
    },
    {
      "epoch": 4.424581005586592,
      "grad_norm": 0.07646811095500806,
      "learning_rate": 3.1050136698005963e-06,
      "loss": 0.3025,
      "step": 297
    },
    {
      "epoch": 4.439478584729981,
      "grad_norm": 0.07957607014767461,
      "learning_rate": 2.9457522321975253e-06,
      "loss": 0.3079,
      "step": 298
    },
    {
      "epoch": 4.454376163873371,
      "grad_norm": 0.08390752993938617,
      "learning_rate": 2.790527256079134e-06,
      "loss": 0.3066,
      "step": 299
    },
    {
      "epoch": 4.4692737430167595,
      "grad_norm": 0.07541814294925331,
      "learning_rate": 2.6393556506980834e-06,
      "loss": 0.3071,
      "step": 300
    },
    {
      "epoch": 4.484171322160149,
      "grad_norm": 0.0714738276840467,
      "learning_rate": 2.4922538837577916e-06,
      "loss": 0.3078,
      "step": 301
    },
    {
      "epoch": 4.499068901303538,
      "grad_norm": 0.0729695749304457,
      "learning_rate": 2.349237979618555e-06,
      "loss": 0.3112,
      "step": 302
    },
    {
      "epoch": 4.5139664804469275,
      "grad_norm": 0.10108195097380965,
      "learning_rate": 2.2103235175519355e-06,
      "loss": 0.3052,
      "step": 303
    },
    {
      "epoch": 4.528864059590316,
      "grad_norm": 0.07138906578544997,
      "learning_rate": 2.0755256300436687e-06,
      "loss": 0.3014,
      "step": 304
    },
    {
      "epoch": 4.543761638733706,
      "grad_norm": 0.07356056598380585,
      "learning_rate": 1.944859001145223e-06,
      "loss": 0.3065,
      "step": 305
    },
    {
      "epoch": 4.558659217877095,
      "grad_norm": 0.07516469607423958,
      "learning_rate": 1.8183378648741979e-06,
      "loss": 0.3049,
      "step": 306
    },
    {
      "epoch": 4.573556797020484,
      "grad_norm": 0.07371825538928153,
      "learning_rate": 1.6959760036637662e-06,
      "loss": 0.304,
      "step": 307
    },
    {
      "epoch": 4.588454376163874,
      "grad_norm": 0.07502909475594224,
      "learning_rate": 1.5777867468612874e-06,
      "loss": 0.3011,
      "step": 308
    },
    {
      "epoch": 4.603351955307263,
      "grad_norm": 0.07352157905066296,
      "learning_rate": 1.4637829692763128e-06,
      "loss": 0.3029,
      "step": 309
    },
    {
      "epoch": 4.618249534450651,
      "grad_norm": 0.07644054766848495,
      "learning_rate": 1.353977089778078e-06,
      "loss": 0.3046,
      "step": 310
    },
    {
      "epoch": 4.633147113594041,
      "grad_norm": 0.07171131988386027,
      "learning_rate": 1.2483810699426458e-06,
      "loss": 0.3068,
      "step": 311
    },
    {
      "epoch": 4.648044692737431,
      "grad_norm": 0.07057850167398541,
      "learning_rate": 1.1470064127499091e-06,
      "loss": 0.3064,
      "step": 312
    },
    {
      "epoch": 4.662942271880819,
      "grad_norm": 0.07626745016001804,
      "learning_rate": 1.0498641613305182e-06,
      "loss": 0.3007,
      "step": 313
    },
    {
      "epoch": 4.677839851024208,
      "grad_norm": 0.0693829644961785,
      "learning_rate": 9.569648977629176e-07,
      "loss": 0.3049,
      "step": 314
    },
    {
      "epoch": 4.692737430167598,
      "grad_norm": 0.0674761214613311,
      "learning_rate": 8.683187419205797e-07,
      "loss": 0.3013,
      "step": 315
    },
    {
      "epoch": 4.707635009310987,
      "grad_norm": 0.06823222699165288,
      "learning_rate": 7.839353503696379e-07,
      "loss": 0.3071,
      "step": 316
    },
    {
      "epoch": 4.722532588454376,
      "grad_norm": 0.06859479836699696,
      "learning_rate": 7.038239153169324e-07,
      "loss": 0.301,
      "step": 317
    },
    {
      "epoch": 4.737430167597766,
      "grad_norm": 0.07031736775350605,
      "learning_rate": 6.279931636086912e-07,
      "loss": 0.3086,
      "step": 318
    },
    {
      "epoch": 4.752327746741154,
      "grad_norm": 0.06915744071480374,
      "learning_rate": 5.564513557798501e-07,
      "loss": 0.3048,
      "step": 319
    },
    {
      "epoch": 4.767225325884544,
      "grad_norm": 0.06815891873681693,
      "learning_rate": 4.892062851542356e-07,
      "loss": 0.309,
      "step": 320
    },
    {
      "epoch": 4.782122905027933,
      "grad_norm": 0.06935050248754245,
      "learning_rate": 4.2626527699558996e-07,
      "loss": 0.3018,
      "step": 321
    },
    {
      "epoch": 4.797020484171322,
      "grad_norm": 0.06784753003557113,
      "learning_rate": 3.6763518770960517e-07,
      "loss": 0.3051,
      "step": 322
    },
    {
      "epoch": 4.811918063314711,
      "grad_norm": 0.06808652827602746,
      "learning_rate": 3.133224040970273e-07,
      "loss": 0.3071,
      "step": 323
    },
    {
      "epoch": 4.826815642458101,
      "grad_norm": 0.07121784851648261,
      "learning_rate": 2.6333284265790627e-07,
      "loss": 0.3047,
      "step": 324
    },
    {
      "epoch": 4.8417132216014895,
      "grad_norm": 0.06700344502725675,
      "learning_rate": 2.1767194894712462e-07,
      "loss": 0.305,
      "step": 325
    },
    {
      "epoch": 4.856610800744879,
      "grad_norm": 0.06927419764551407,
      "learning_rate": 1.763446969811522e-07,
      "loss": 0.306,
      "step": 326
    },
    {
      "epoch": 4.871508379888268,
      "grad_norm": 0.06947638763948212,
      "learning_rate": 1.3935558869622168e-07,
      "loss": 0.3042,
      "step": 327
    },
    {
      "epoch": 4.8864059590316575,
      "grad_norm": 0.06866693792434822,
      "learning_rate": 1.0670865345793425e-07,
      "loss": 0.3094,
      "step": 328
    },
    {
      "epoch": 4.901303538175046,
      "grad_norm": 0.06859589228192586,
      "learning_rate": 7.840744762229069e-08,
      "loss": 0.3054,
      "step": 329
    },
    {
      "epoch": 4.916201117318436,
      "grad_norm": 0.06927380611470749,
      "learning_rate": 5.445505414831242e-08,
      "loss": 0.3054,
      "step": 330
    },
    {
      "epoch": 4.931098696461825,
      "grad_norm": 0.06680927663742126,
      "learning_rate": 3.485408226218567e-08,
      "loss": 0.2998,
      "step": 331
    },
    {
      "epoch": 4.945996275605214,
      "grad_norm": 0.06773777244031369,
      "learning_rate": 1.960666717304438e-08,
      "loss": 0.3058,
      "step": 332
    },
    {
      "epoch": 4.960893854748603,
      "grad_norm": 0.06864556914107234,
      "learning_rate": 8.714469840351848e-09,
      "loss": 0.2984,
      "step": 333
    },
    {
      "epoch": 4.975791433891993,
      "grad_norm": 0.0678647360547641,
      "learning_rate": 2.178676792996548e-09,
      "loss": 0.3051,
      "step": 334
    },
    {
      "epoch": 4.990689013035381,
      "grad_norm": 0.06662103871858636,
      "learning_rate": 0.0,
      "loss": 0.2973,
      "step": 335
    },
    {
      "epoch": 4.990689013035381,
      "step": 335,
      "total_flos": 8.620546826447094e+18,
      "train_loss": 0.380745083004681,
      "train_runtime": 51129.1279,
      "train_samples_per_second": 3.36,
      "train_steps_per_second": 0.007
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 335,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.620546826447094e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}