{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9983443708609272,
  "eval_steps": 500,
  "global_step": 201,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004966887417218543,
      "grad_norm": 35.1089973449707,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 5.1044,
      "step": 1
    },
    {
      "epoch": 0.009933774834437087,
      "grad_norm": 35.22032165527344,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 5.0735,
      "step": 2
    },
    {
      "epoch": 0.014900662251655629,
      "grad_norm": 35.83143997192383,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 5.1028,
      "step": 3
    },
    {
      "epoch": 0.019867549668874173,
      "grad_norm": 35.02216339111328,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 5.0132,
      "step": 4
    },
    {
      "epoch": 0.024834437086092714,
      "grad_norm": 35.219600677490234,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 4.9707,
      "step": 5
    },
    {
      "epoch": 0.029801324503311258,
      "grad_norm": 36.146759033203125,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 5.0997,
      "step": 6
    },
    {
      "epoch": 0.0347682119205298,
      "grad_norm": 35.079566955566406,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 5.0102,
      "step": 7
    },
    {
      "epoch": 0.039735099337748346,
      "grad_norm": 33.08433532714844,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 4.7433,
      "step": 8
    },
    {
      "epoch": 0.04470198675496689,
      "grad_norm": 33.799949645996094,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 4.8223,
      "step": 9
    },
    {
      "epoch": 0.04966887417218543,
      "grad_norm": 35.39978790283203,
      "learning_rate": 5.000000000000001e-07,
      "loss": 5.1291,
      "step": 10
    },
    {
      "epoch": 0.054635761589403975,
      "grad_norm": 33.871028900146484,
      "learning_rate": 5.5e-07,
      "loss": 4.8411,
      "step": 11
    },
    {
      "epoch": 0.059602649006622516,
      "grad_norm": 33.94557571411133,
      "learning_rate": 6.000000000000001e-07,
      "loss": 4.8888,
      "step": 12
    },
    {
      "epoch": 0.06456953642384106,
      "grad_norm": 35.7645263671875,
      "learning_rate": 6.5e-07,
      "loss": 4.9706,
      "step": 13
    },
    {
      "epoch": 0.0695364238410596,
      "grad_norm": 33.34524917602539,
      "learning_rate": 7.000000000000001e-07,
      "loss": 4.8249,
      "step": 14
    },
    {
      "epoch": 0.07450331125827815,
      "grad_norm": 31.388317108154297,
      "learning_rate": 7.5e-07,
      "loss": 4.4837,
      "step": 15
    },
    {
      "epoch": 0.07947019867549669,
      "grad_norm": 30.788236618041992,
      "learning_rate": 8.000000000000001e-07,
      "loss": 4.4459,
      "step": 16
    },
    {
      "epoch": 0.08443708609271523,
      "grad_norm": 30.05523109436035,
      "learning_rate": 8.500000000000001e-07,
      "loss": 4.4497,
      "step": 17
    },
    {
      "epoch": 0.08940397350993377,
      "grad_norm": 27.767133712768555,
      "learning_rate": 9.000000000000001e-07,
      "loss": 4.3426,
      "step": 18
    },
    {
      "epoch": 0.09437086092715231,
      "grad_norm": 26.096942901611328,
      "learning_rate": 9.500000000000001e-07,
      "loss": 4.1846,
      "step": 19
    },
    {
      "epoch": 0.09933774834437085,
      "grad_norm": 24.659503936767578,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 4.1404,
      "step": 20
    },
    {
      "epoch": 0.10430463576158941,
      "grad_norm": 22.810558319091797,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 3.9254,
      "step": 21
    },
    {
      "epoch": 0.10927152317880795,
      "grad_norm": 21.360898971557617,
      "learning_rate": 1.1e-06,
      "loss": 3.7266,
      "step": 22
    },
    {
      "epoch": 0.11423841059602649,
      "grad_norm": 22.106414794921875,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 3.6787,
      "step": 23
    },
    {
      "epoch": 0.11920529801324503,
      "grad_norm": 19.86927032470703,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 3.3947,
      "step": 24
    },
    {
      "epoch": 0.12417218543046357,
      "grad_norm": 20.86886215209961,
      "learning_rate": 1.25e-06,
      "loss": 3.3619,
      "step": 25
    },
    {
      "epoch": 0.1291390728476821,
      "grad_norm": 20.39168357849121,
      "learning_rate": 1.3e-06,
      "loss": 3.1309,
      "step": 26
    },
    {
      "epoch": 0.13410596026490065,
      "grad_norm": 20.989742279052734,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 3.0103,
      "step": 27
    },
    {
      "epoch": 0.1390728476821192,
      "grad_norm": 19.428770065307617,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 2.8232,
      "step": 28
    },
    {
      "epoch": 0.14403973509933773,
      "grad_norm": 16.829797744750977,
      "learning_rate": 1.45e-06,
      "loss": 2.6573,
      "step": 29
    },
    {
      "epoch": 0.1490066225165563,
      "grad_norm": 15.379148483276367,
      "learning_rate": 1.5e-06,
      "loss": 2.5938,
      "step": 30
    },
    {
      "epoch": 0.15397350993377484,
      "grad_norm": 14.703971862792969,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 2.4588,
      "step": 31
    },
    {
      "epoch": 0.15894039735099338,
      "grad_norm": 14.567538261413574,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 2.211,
      "step": 32
    },
    {
      "epoch": 0.16390728476821192,
      "grad_norm": 14.908529281616211,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 2.0326,
      "step": 33
    },
    {
      "epoch": 0.16887417218543047,
      "grad_norm": 15.71338939666748,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 1.9728,
      "step": 34
    },
    {
      "epoch": 0.173841059602649,
      "grad_norm": 15.91982364654541,
      "learning_rate": 1.75e-06,
      "loss": 1.8389,
      "step": 35
    },
    {
      "epoch": 0.17880794701986755,
      "grad_norm": 15.001914024353027,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 1.6801,
      "step": 36
    },
    {
      "epoch": 0.1837748344370861,
      "grad_norm": 13.963605880737305,
      "learning_rate": 1.85e-06,
      "loss": 1.4723,
      "step": 37
    },
    {
      "epoch": 0.18874172185430463,
      "grad_norm": 13.326650619506836,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 1.3989,
      "step": 38
    },
    {
      "epoch": 0.19370860927152317,
      "grad_norm": 13.237666130065918,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 1.2334,
      "step": 39
    },
    {
      "epoch": 0.1986754966887417,
      "grad_norm": 12.756895065307617,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.0125,
      "step": 40
    },
    {
      "epoch": 0.20364238410596028,
      "grad_norm": 12.65739917755127,
      "learning_rate": 2.05e-06,
      "loss": 0.9345,
      "step": 41
    },
    {
      "epoch": 0.20860927152317882,
      "grad_norm": 12.268824577331543,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.7775,
      "step": 42
    },
    {
      "epoch": 0.21357615894039736,
      "grad_norm": 11.216493606567383,
      "learning_rate": 2.15e-06,
      "loss": 0.6453,
      "step": 43
    },
    {
      "epoch": 0.2185430463576159,
      "grad_norm": 10.217277526855469,
      "learning_rate": 2.2e-06,
      "loss": 0.5069,
      "step": 44
    },
    {
      "epoch": 0.22350993377483444,
      "grad_norm": 8.963619232177734,
      "learning_rate": 2.25e-06,
      "loss": 0.4146,
      "step": 45
    },
    {
      "epoch": 0.22847682119205298,
      "grad_norm": 7.5903706550598145,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.3154,
      "step": 46
    },
    {
      "epoch": 0.23344370860927152,
      "grad_norm": 5.7549147605896,
      "learning_rate": 2.35e-06,
      "loss": 0.2397,
      "step": 47
    },
    {
      "epoch": 0.23841059602649006,
      "grad_norm": 3.816410779953003,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.1793,
      "step": 48
    },
    {
      "epoch": 0.2433774834437086,
      "grad_norm": 2.867004156112671,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.1624,
      "step": 49
    },
    {
      "epoch": 0.24834437086092714,
      "grad_norm": 2.462583303451538,
      "learning_rate": 2.5e-06,
      "loss": 0.135,
      "step": 50
    },
    {
      "epoch": 0.2533112582781457,
      "grad_norm": 2.168409824371338,
      "learning_rate": 2.55e-06,
      "loss": 0.1176,
      "step": 51
    },
    {
      "epoch": 0.2582781456953642,
      "grad_norm": 1.5076923370361328,
      "learning_rate": 2.6e-06,
      "loss": 0.1294,
      "step": 52
    },
    {
      "epoch": 0.2632450331125828,
      "grad_norm": 1.422659158706665,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.0929,
      "step": 53
    },
    {
      "epoch": 0.2682119205298013,
      "grad_norm": 1.2882518768310547,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.0872,
      "step": 54
    },
    {
      "epoch": 0.2731788079470199,
      "grad_norm": 0.9768906235694885,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.1038,
      "step": 55
    },
    {
      "epoch": 0.2781456953642384,
      "grad_norm": 0.6768905520439148,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0839,
      "step": 56
    },
    {
      "epoch": 0.28311258278145696,
      "grad_norm": 0.8317804932594299,
      "learning_rate": 2.85e-06,
      "loss": 0.0793,
      "step": 57
    },
    {
      "epoch": 0.28807947019867547,
      "grad_norm": 1.0260225534439087,
      "learning_rate": 2.9e-06,
      "loss": 0.0907,
      "step": 58
    },
    {
      "epoch": 0.29304635761589404,
      "grad_norm": 0.69094318151474,
      "learning_rate": 2.95e-06,
      "loss": 0.0748,
      "step": 59
    },
    {
      "epoch": 0.2980132450331126,
      "grad_norm": 0.6794005036354065,
      "learning_rate": 3e-06,
      "loss": 0.0784,
      "step": 60
    },
    {
      "epoch": 0.3029801324503311,
      "grad_norm": 0.6374951004981995,
      "learning_rate": 3.05e-06,
      "loss": 0.0739,
      "step": 61
    },
    {
      "epoch": 0.3079470198675497,
      "grad_norm": 0.7064375281333923,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.0729,
      "step": 62
    },
    {
      "epoch": 0.3129139072847682,
      "grad_norm": 0.4373936951160431,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.0682,
      "step": 63
    },
    {
      "epoch": 0.31788079470198677,
      "grad_norm": 0.5880410075187683,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0666,
      "step": 64
    },
    {
      "epoch": 0.3228476821192053,
      "grad_norm": 0.4951966106891632,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.0709,
      "step": 65
    },
    {
      "epoch": 0.32781456953642385,
      "grad_norm": 0.46179434657096863,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.0669,
      "step": 66
    },
    {
      "epoch": 0.33278145695364236,
      "grad_norm": 0.543757438659668,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.0716,
      "step": 67
    },
    {
      "epoch": 0.33774834437086093,
      "grad_norm": 0.5536766052246094,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.0751,
      "step": 68
    },
    {
      "epoch": 0.34271523178807944,
      "grad_norm": 0.4959389865398407,
      "learning_rate": 3.45e-06,
      "loss": 0.0725,
      "step": 69
    },
    {
      "epoch": 0.347682119205298,
      "grad_norm": 0.40888911485671997,
      "learning_rate": 3.5e-06,
      "loss": 0.0696,
      "step": 70
    },
    {
      "epoch": 0.3526490066225166,
      "grad_norm": 0.41407281160354614,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.0577,
      "step": 71
    },
    {
      "epoch": 0.3576158940397351,
      "grad_norm": 0.45143064856529236,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.0666,
      "step": 72
    },
    {
      "epoch": 0.36258278145695366,
      "grad_norm": 0.4229239523410797,
      "learning_rate": 3.65e-06,
      "loss": 0.0657,
      "step": 73
    },
    {
      "epoch": 0.3675496688741722,
      "grad_norm": 0.32228904962539673,
      "learning_rate": 3.7e-06,
      "loss": 0.0604,
      "step": 74
    },
    {
      "epoch": 0.37251655629139074,
      "grad_norm": 0.48797911405563354,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.0658,
      "step": 75
    },
    {
      "epoch": 0.37748344370860926,
      "grad_norm": 0.5396533012390137,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.0652,
      "step": 76
    },
    {
      "epoch": 0.3824503311258278,
      "grad_norm": 0.3672020733356476,
      "learning_rate": 3.85e-06,
      "loss": 0.0559,
      "step": 77
    },
    {
      "epoch": 0.38741721854304634,
      "grad_norm": 0.4659889340400696,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.0596,
      "step": 78
    },
    {
      "epoch": 0.3923841059602649,
      "grad_norm": 0.40139755606651306,
      "learning_rate": 3.95e-06,
      "loss": 0.0594,
      "step": 79
    },
    {
      "epoch": 0.3973509933774834,
      "grad_norm": 0.32445284724235535,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0574,
      "step": 80
    },
    {
      "epoch": 0.402317880794702,
      "grad_norm": 0.5123882293701172,
      "learning_rate": 4.05e-06,
      "loss": 0.0694,
      "step": 81
    },
    {
      "epoch": 0.40728476821192056,
      "grad_norm": 0.3744450509548187,
      "learning_rate": 4.1e-06,
      "loss": 0.0663,
      "step": 82
    },
    {
      "epoch": 0.41225165562913907,
      "grad_norm": 0.4288279414176941,
      "learning_rate": 4.15e-06,
      "loss": 0.0575,
      "step": 83
    },
    {
      "epoch": 0.41721854304635764,
      "grad_norm": 0.3742220997810364,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0643,
      "step": 84
    },
    {
      "epoch": 0.42218543046357615,
      "grad_norm": 0.3413388133049011,
      "learning_rate": 4.25e-06,
      "loss": 0.0599,
      "step": 85
    },
    {
      "epoch": 0.4271523178807947,
      "grad_norm": 0.42702731490135193,
      "learning_rate": 4.3e-06,
      "loss": 0.0607,
      "step": 86
    },
    {
      "epoch": 0.43211920529801323,
      "grad_norm": 0.28603485226631165,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.0539,
      "step": 87
    },
    {
      "epoch": 0.4370860927152318,
      "grad_norm": 0.39420223236083984,
      "learning_rate": 4.4e-06,
      "loss": 0.0536,
      "step": 88
    },
    {
      "epoch": 0.4420529801324503,
      "grad_norm": 0.3824305534362793,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.0545,
      "step": 89
    },
    {
      "epoch": 0.4470198675496689,
      "grad_norm": 0.7128148674964905,
      "learning_rate": 4.5e-06,
      "loss": 0.066,
      "step": 90
    },
    {
      "epoch": 0.4519867549668874,
      "grad_norm": 0.43943652510643005,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.0551,
      "step": 91
    },
    {
      "epoch": 0.45695364238410596,
      "grad_norm": 0.4158875048160553,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.0597,
      "step": 92
    },
    {
      "epoch": 0.46192052980132453,
      "grad_norm": 0.4504539668560028,
      "learning_rate": 4.65e-06,
      "loss": 0.0651,
      "step": 93
    },
    {
      "epoch": 0.46688741721854304,
      "grad_norm": 0.2968044579029083,
      "learning_rate": 4.7e-06,
      "loss": 0.0601,
      "step": 94
    },
    {
      "epoch": 0.4718543046357616,
      "grad_norm": 0.3136419951915741,
      "learning_rate": 4.75e-06,
      "loss": 0.0537,
      "step": 95
    },
    {
      "epoch": 0.4768211920529801,
      "grad_norm": 0.3263697922229767,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0571,
      "step": 96
    },
    {
      "epoch": 0.4817880794701987,
      "grad_norm": 0.4513280391693115,
      "learning_rate": 4.85e-06,
      "loss": 0.0578,
      "step": 97
    },
    {
      "epoch": 0.4867549668874172,
      "grad_norm": 0.32309144735336304,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.0532,
      "step": 98
    },
    {
      "epoch": 0.4917218543046358,
      "grad_norm": 0.3191864788532257,
      "learning_rate": 4.95e-06,
      "loss": 0.0563,
      "step": 99
    },
    {
      "epoch": 0.4966887417218543,
      "grad_norm": 0.3069349229335785,
      "learning_rate": 5e-06,
      "loss": 0.0492,
      "step": 100
    },
    {
      "epoch": 0.5016556291390728,
      "grad_norm": 0.592820942401886,
      "learning_rate": 4.999989914458693e-06,
      "loss": 0.051,
      "step": 101
    },
    {
      "epoch": 0.5066225165562914,
      "grad_norm": 0.6325397491455078,
      "learning_rate": 4.999959657916147e-06,
      "loss": 0.0621,
      "step": 102
    },
    {
      "epoch": 0.5115894039735099,
      "grad_norm": 0.43557360768318176,
      "learning_rate": 4.999909230616483e-06,
      "loss": 0.0579,
      "step": 103
    },
    {
      "epoch": 0.5165562913907285,
      "grad_norm": 0.5436730980873108,
      "learning_rate": 4.999838632966572e-06,
      "loss": 0.0688,
      "step": 104
    },
    {
      "epoch": 0.5215231788079471,
      "grad_norm": 0.4149836301803589,
      "learning_rate": 4.999747865536025e-06,
      "loss": 0.0504,
      "step": 105
    },
    {
      "epoch": 0.5264900662251656,
      "grad_norm": 0.3056333661079407,
      "learning_rate": 4.999636929057196e-06,
      "loss": 0.0582,
      "step": 106
    },
    {
      "epoch": 0.5314569536423841,
      "grad_norm": 0.3057827651500702,
      "learning_rate": 4.999505824425164e-06,
      "loss": 0.0583,
      "step": 107
    },
    {
      "epoch": 0.5364238410596026,
      "grad_norm": 0.29118627309799194,
      "learning_rate": 4.999354552697742e-06,
      "loss": 0.0545,
      "step": 108
    },
    {
      "epoch": 0.5413907284768212,
      "grad_norm": 0.4180416762828827,
      "learning_rate": 4.999183115095453e-06,
      "loss": 0.0635,
      "step": 109
    },
    {
      "epoch": 0.5463576158940397,
      "grad_norm": 0.36227947473526,
      "learning_rate": 4.998991513001532e-06,
      "loss": 0.0581,
      "step": 110
    },
    {
      "epoch": 0.5513245033112583,
      "grad_norm": 0.4391280710697174,
      "learning_rate": 4.998779747961906e-06,
      "loss": 0.0569,
      "step": 111
    },
    {
      "epoch": 0.5562913907284768,
      "grad_norm": 0.40176400542259216,
      "learning_rate": 4.998547821685188e-06,
      "loss": 0.0479,
      "step": 112
    },
    {
      "epoch": 0.5612582781456954,
      "grad_norm": 0.28267550468444824,
      "learning_rate": 4.998295736042659e-06,
      "loss": 0.0562,
      "step": 113
    },
    {
      "epoch": 0.5662251655629139,
      "grad_norm": 0.29833951592445374,
      "learning_rate": 4.998023493068255e-06,
      "loss": 0.0551,
      "step": 114
    },
    {
      "epoch": 0.5711920529801324,
      "grad_norm": 0.4476202130317688,
      "learning_rate": 4.997731094958551e-06,
      "loss": 0.0493,
      "step": 115
    },
    {
      "epoch": 0.5761589403973509,
      "grad_norm": 0.3935531973838806,
      "learning_rate": 4.997418544072742e-06,
      "loss": 0.0602,
      "step": 116
    },
    {
      "epoch": 0.5811258278145696,
      "grad_norm": 0.3693472743034363,
      "learning_rate": 4.9970858429326215e-06,
      "loss": 0.0567,
      "step": 117
    },
    {
      "epoch": 0.5860927152317881,
      "grad_norm": 0.5814310312271118,
      "learning_rate": 4.99673299422257e-06,
      "loss": 0.0696,
      "step": 118
    },
    {
      "epoch": 0.5910596026490066,
      "grad_norm": 0.32046371698379517,
      "learning_rate": 4.996360000789519e-06,
      "loss": 0.0498,
      "step": 119
    },
    {
      "epoch": 0.5960264900662252,
      "grad_norm": 0.45766863226890564,
      "learning_rate": 4.995966865642946e-06,
      "loss": 0.0501,
      "step": 120
    },
    {
      "epoch": 0.6009933774834437,
      "grad_norm": 0.2858210504055023,
      "learning_rate": 4.995553591954832e-06,
      "loss": 0.0516,
      "step": 121
    },
    {
      "epoch": 0.6059602649006622,
      "grad_norm": 0.2883959412574768,
      "learning_rate": 4.9951201830596505e-06,
      "loss": 0.0565,
      "step": 122
    },
    {
      "epoch": 0.6109271523178808,
      "grad_norm": 0.2570095360279083,
      "learning_rate": 4.994666642454331e-06,
      "loss": 0.0517,
      "step": 123
    },
    {
      "epoch": 0.6158940397350994,
      "grad_norm": 0.34462130069732666,
      "learning_rate": 4.994192973798236e-06,
      "loss": 0.0589,
      "step": 124
    },
    {
      "epoch": 0.6208609271523179,
      "grad_norm": 0.2416890263557434,
      "learning_rate": 4.993699180913127e-06,
      "loss": 0.0544,
      "step": 125
    },
    {
      "epoch": 0.6258278145695364,
      "grad_norm": 0.2676381468772888,
      "learning_rate": 4.993185267783142e-06,
      "loss": 0.0506,
      "step": 126
    },
    {
      "epoch": 0.6307947019867549,
      "grad_norm": 0.2828262150287628,
      "learning_rate": 4.992651238554753e-06,
      "loss": 0.0504,
      "step": 127
    },
    {
      "epoch": 0.6357615894039735,
      "grad_norm": 0.2980761229991913,
      "learning_rate": 4.99209709753674e-06,
      "loss": 0.053,
      "step": 128
    },
    {
      "epoch": 0.640728476821192,
      "grad_norm": 0.30427825450897217,
      "learning_rate": 4.991522849200152e-06,
      "loss": 0.0517,
      "step": 129
    },
    {
      "epoch": 0.6456953642384106,
      "grad_norm": 0.32718971371650696,
      "learning_rate": 4.990928498178274e-06,
      "loss": 0.0543,
      "step": 130
    },
    {
      "epoch": 0.6506622516556292,
      "grad_norm": 0.26971936225891113,
      "learning_rate": 4.990314049266586e-06,
      "loss": 0.0498,
      "step": 131
    },
    {
      "epoch": 0.6556291390728477,
      "grad_norm": 0.2683403193950653,
      "learning_rate": 4.989679507422728e-06,
      "loss": 0.0456,
      "step": 132
    },
    {
      "epoch": 0.6605960264900662,
      "grad_norm": 0.3088124990463257,
      "learning_rate": 4.989024877766461e-06,
      "loss": 0.0549,
      "step": 133
    },
    {
      "epoch": 0.6655629139072847,
      "grad_norm": 0.2979089021682739,
      "learning_rate": 4.988350165579618e-06,
      "loss": 0.0493,
      "step": 134
    },
    {
      "epoch": 0.6705298013245033,
      "grad_norm": 0.3119763731956482,
      "learning_rate": 4.987655376306069e-06,
      "loss": 0.0494,
      "step": 135
    },
    {
      "epoch": 0.6754966887417219,
      "grad_norm": 0.26640820503234863,
      "learning_rate": 4.986940515551676e-06,
      "loss": 0.0466,
      "step": 136
    },
    {
      "epoch": 0.6804635761589404,
      "grad_norm": 0.34190040826797485,
      "learning_rate": 4.9862055890842455e-06,
      "loss": 0.052,
      "step": 137
    },
    {
      "epoch": 0.6854304635761589,
      "grad_norm": 0.3251188099384308,
      "learning_rate": 4.9854506028334805e-06,
      "loss": 0.0475,
      "step": 138
    },
    {
      "epoch": 0.6903973509933775,
      "grad_norm": 0.31759873032569885,
      "learning_rate": 4.984675562890939e-06,
      "loss": 0.0473,
      "step": 139
    },
    {
      "epoch": 0.695364238410596,
      "grad_norm": 0.382521390914917,
      "learning_rate": 4.983880475509978e-06,
      "loss": 0.0569,
      "step": 140
    },
    {
      "epoch": 0.7003311258278145,
      "grad_norm": 0.2814515233039856,
      "learning_rate": 4.983065347105707e-06,
      "loss": 0.0512,
      "step": 141
    },
    {
      "epoch": 0.7052980132450332,
      "grad_norm": 0.3223007321357727,
      "learning_rate": 4.982230184254934e-06,
      "loss": 0.0563,
      "step": 142
    },
    {
      "epoch": 0.7102649006622517,
      "grad_norm": 0.30451443791389465,
      "learning_rate": 4.981374993696116e-06,
      "loss": 0.0412,
      "step": 143
    },
    {
      "epoch": 0.7152317880794702,
      "grad_norm": 0.39687198400497437,
      "learning_rate": 4.9804997823292996e-06,
      "loss": 0.0507,
      "step": 144
    },
    {
      "epoch": 0.7201986754966887,
      "grad_norm": 0.311040997505188,
      "learning_rate": 4.97960455721607e-06,
      "loss": 0.0483,
      "step": 145
    },
    {
      "epoch": 0.7251655629139073,
      "grad_norm": 0.33765530586242676,
      "learning_rate": 4.978689325579491e-06,
      "loss": 0.0431,
      "step": 146
    },
    {
      "epoch": 0.7301324503311258,
      "grad_norm": 0.321216344833374,
      "learning_rate": 4.9777540948040474e-06,
      "loss": 0.0537,
      "step": 147
    },
    {
      "epoch": 0.7350993377483444,
      "grad_norm": 0.38110417127609253,
      "learning_rate": 4.976798872435586e-06,
      "loss": 0.0535,
      "step": 148
    },
    {
      "epoch": 0.7400662251655629,
      "grad_norm": 0.35567203164100647,
      "learning_rate": 4.975823666181256e-06,
      "loss": 0.0481,
      "step": 149
    },
    {
      "epoch": 0.7450331125827815,
      "grad_norm": 0.27548110485076904,
      "learning_rate": 4.974828483909441e-06,
      "loss": 0.0467,
      "step": 150
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.33838632702827454,
      "learning_rate": 4.9738133336497045e-06,
      "loss": 0.0551,
      "step": 151
    },
    {
      "epoch": 0.7549668874172185,
      "grad_norm": 0.29228779673576355,
      "learning_rate": 4.972778223592717e-06,
      "loss": 0.0483,
      "step": 152
    },
    {
      "epoch": 0.7599337748344371,
      "grad_norm": 0.28962090611457825,
      "learning_rate": 4.9717231620901964e-06,
      "loss": 0.0509,
      "step": 153
    },
    {
      "epoch": 0.7649006622516556,
      "grad_norm": 0.2481975555419922,
      "learning_rate": 4.970648157654836e-06,
      "loss": 0.0473,
      "step": 154
    },
    {
      "epoch": 0.7698675496688742,
      "grad_norm": 0.2947401702404022,
      "learning_rate": 4.969553218960235e-06,
      "loss": 0.0449,
      "step": 155
    },
    {
      "epoch": 0.7748344370860927,
      "grad_norm": 0.291072815656662,
      "learning_rate": 4.968438354840834e-06,
      "loss": 0.0547,
      "step": 156
    },
    {
      "epoch": 0.7798013245033113,
      "grad_norm": 0.2935287356376648,
      "learning_rate": 4.96730357429184e-06,
      "loss": 0.0517,
      "step": 157
    },
    {
      "epoch": 0.7847682119205298,
      "grad_norm": 0.3540295958518982,
      "learning_rate": 4.966148886469153e-06,
      "loss": 0.0468,
      "step": 158
    },
    {
      "epoch": 0.7897350993377483,
      "grad_norm": 0.2809925973415375,
      "learning_rate": 4.964974300689295e-06,
      "loss": 0.0474,
      "step": 159
    },
    {
      "epoch": 0.7947019867549668,
      "grad_norm": 0.3492342531681061,
      "learning_rate": 4.963779826429333e-06,
      "loss": 0.056,
      "step": 160
    },
    {
      "epoch": 0.7996688741721855,
      "grad_norm": 0.39458170533180237,
      "learning_rate": 4.9625654733268016e-06,
      "loss": 0.0496,
      "step": 161
    },
    {
      "epoch": 0.804635761589404,
      "grad_norm": 0.30106887221336365,
      "learning_rate": 4.961331251179629e-06,
      "loss": 0.0401,
      "step": 162
    },
    {
      "epoch": 0.8096026490066225,
      "grad_norm": 0.297493577003479,
      "learning_rate": 4.960077169946052e-06,
      "loss": 0.0488,
      "step": 163
    },
    {
      "epoch": 0.8145695364238411,
      "grad_norm": 0.30854183435440063,
      "learning_rate": 4.958803239744542e-06,
      "loss": 0.0452,
      "step": 164
    },
    {
      "epoch": 0.8195364238410596,
      "grad_norm": 0.3040302097797394,
      "learning_rate": 4.9575094708537205e-06,
      "loss": 0.0497,
      "step": 165
    },
    {
      "epoch": 0.8245033112582781,
      "grad_norm": 0.3189365267753601,
      "learning_rate": 4.956195873712274e-06,
      "loss": 0.0404,
      "step": 166
    },
    {
      "epoch": 0.8294701986754967,
      "grad_norm": 0.27114391326904297,
      "learning_rate": 4.954862458918873e-06,
      "loss": 0.0376,
      "step": 167
    },
    {
      "epoch": 0.8344370860927153,
      "grad_norm": 0.37876346707344055,
      "learning_rate": 4.953509237232086e-06,
      "loss": 0.0536,
      "step": 168
    },
    {
      "epoch": 0.8394039735099338,
      "grad_norm": 0.3651692867279053,
      "learning_rate": 4.952136219570292e-06,
      "loss": 0.0513,
      "step": 169
    },
    {
      "epoch": 0.8443708609271523,
      "grad_norm": 0.2958744764328003,
      "learning_rate": 4.950743417011591e-06,
      "loss": 0.0448,
      "step": 170
    },
    {
      "epoch": 0.8493377483443708,
      "grad_norm": 0.2732123136520386,
      "learning_rate": 4.9493308407937176e-06,
      "loss": 0.0441,
      "step": 171
    },
    {
      "epoch": 0.8543046357615894,
      "grad_norm": 0.2853662073612213,
      "learning_rate": 4.947898502313949e-06,
      "loss": 0.039,
      "step": 172
    },
    {
      "epoch": 0.859271523178808,
      "grad_norm": 0.3872867822647095,
      "learning_rate": 4.946446413129011e-06,
      "loss": 0.0561,
      "step": 173
    },
    {
      "epoch": 0.8642384105960265,
      "grad_norm": 0.33438095450401306,
      "learning_rate": 4.944974584954989e-06,
      "loss": 0.0427,
      "step": 174
    },
    {
      "epoch": 0.8692052980132451,
      "grad_norm": 0.3008469045162201,
      "learning_rate": 4.943483029667231e-06,
      "loss": 0.0417,
      "step": 175
    },
    {
      "epoch": 0.8741721854304636,
      "grad_norm": 0.298586368560791,
      "learning_rate": 4.941971759300249e-06,
      "loss": 0.0508,
      "step": 176
    },
    {
      "epoch": 0.8791390728476821,
      "grad_norm": 0.2677428126335144,
      "learning_rate": 4.9404407860476275e-06,
      "loss": 0.0468,
      "step": 177
    },
    {
      "epoch": 0.8841059602649006,
      "grad_norm": 0.2635156214237213,
      "learning_rate": 4.938890122261923e-06,
      "loss": 0.0467,
      "step": 178
    },
    {
      "epoch": 0.8890728476821192,
      "grad_norm": 0.29530155658721924,
      "learning_rate": 4.937319780454559e-06,
      "loss": 0.0408,
      "step": 179
    },
    {
      "epoch": 0.8940397350993378,
      "grad_norm": 0.2985950708389282,
      "learning_rate": 4.935729773295738e-06,
      "loss": 0.0397,
      "step": 180
    },
    {
      "epoch": 0.8990066225165563,
      "grad_norm": 0.3108009696006775,
      "learning_rate": 4.934120113614322e-06,
      "loss": 0.0405,
      "step": 181
    },
    {
      "epoch": 0.9039735099337748,
      "grad_norm": 0.32336103916168213,
      "learning_rate": 4.932490814397744e-06,
      "loss": 0.0435,
      "step": 182
    },
    {
      "epoch": 0.9089403973509934,
      "grad_norm": 0.3786524534225464,
      "learning_rate": 4.930841888791898e-06,
      "loss": 0.0455,
      "step": 183
    },
    {
      "epoch": 0.9139072847682119,
      "grad_norm": 0.32045990228652954,
      "learning_rate": 4.929173350101025e-06,
      "loss": 0.0424,
      "step": 184
    },
    {
      "epoch": 0.9188741721854304,
      "grad_norm": 0.4801805913448334,
      "learning_rate": 4.927485211787622e-06,
      "loss": 0.0558,
      "step": 185
    },
    {
      "epoch": 0.9238410596026491,
      "grad_norm": 0.28597143292427063,
      "learning_rate": 4.925777487472318e-06,
      "loss": 0.0378,
      "step": 186
    },
    {
      "epoch": 0.9288079470198676,
      "grad_norm": 0.2773878872394562,
      "learning_rate": 4.9240501909337725e-06,
      "loss": 0.0439,
      "step": 187
    },
    {
      "epoch": 0.9337748344370861,
      "grad_norm": 0.2605089545249939,
      "learning_rate": 4.922303336108562e-06,
      "loss": 0.0361,
      "step": 188
    },
    {
      "epoch": 0.9387417218543046,
      "grad_norm": 0.2805444598197937,
      "learning_rate": 4.920536937091068e-06,
      "loss": 0.0415,
      "step": 189
    },
    {
      "epoch": 0.9437086092715232,
      "grad_norm": 0.33180198073387146,
      "learning_rate": 4.918751008133362e-06,
      "loss": 0.0427,
      "step": 190
    },
    {
      "epoch": 0.9486754966887417,
      "grad_norm": 0.28023046255111694,
      "learning_rate": 4.916945563645093e-06,
      "loss": 0.0419,
      "step": 191
    },
    {
      "epoch": 0.9536423841059603,
      "grad_norm": 0.32969868183135986,
      "learning_rate": 4.915120618193369e-06,
      "loss": 0.0486,
      "step": 192
    },
    {
      "epoch": 0.9586092715231788,
      "grad_norm": 0.3446713984012604,
      "learning_rate": 4.913276186502639e-06,
      "loss": 0.0479,
      "step": 193
    },
    {
      "epoch": 0.9635761589403974,
      "grad_norm": 0.2720557749271393,
      "learning_rate": 4.911412283454579e-06,
      "loss": 0.0388,
      "step": 194
    },
    {
      "epoch": 0.9685430463576159,
      "grad_norm": 0.28475743532180786,
      "learning_rate": 4.909528924087963e-06,
      "loss": 0.0402,
      "step": 195
    },
    {
      "epoch": 0.9735099337748344,
      "grad_norm": 0.3087317943572998,
      "learning_rate": 4.907626123598552e-06,
      "loss": 0.0421,
      "step": 196
    },
    {
      "epoch": 0.9784768211920529,
      "grad_norm": 0.3794298470020294,
      "learning_rate": 4.9057038973389635e-06,
      "loss": 0.0489,
      "step": 197
    },
    {
      "epoch": 0.9834437086092715,
      "grad_norm": 0.31470081210136414,
      "learning_rate": 4.903762260818552e-06,
      "loss": 0.0419,
      "step": 198
    },
    {
      "epoch": 0.9884105960264901,
      "grad_norm": 0.3073600232601166,
      "learning_rate": 4.90180122970328e-06,
      "loss": 0.0373,
      "step": 199
    },
    {
      "epoch": 0.9933774834437086,
      "grad_norm": 0.35700473189353943,
      "learning_rate": 4.899820819815598e-06,
      "loss": 0.0382,
      "step": 200
    },
    {
      "epoch": 0.9983443708609272,
      "grad_norm": 0.38011306524276733,
      "learning_rate": 4.89782104713431e-06,
      "loss": 0.0437,
      "step": 201
    }
  ],
  "logging_steps": 1,
  "max_steps": 1206,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 201,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.5026874380569805e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}