{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9946524064171123,
  "eval_steps": 500,
  "global_step": 210,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0142602495543672,
      "grad_norm": 5.998478174665459,
      "learning_rate": 4.7619047619047623e-07,
      "loss": 0.7176,
      "step": 1
    },
    {
      "epoch": 0.0285204991087344,
      "grad_norm": 6.157130000522602,
      "learning_rate": 9.523809523809525e-07,
      "loss": 0.7167,
      "step": 2
    },
    {
      "epoch": 0.0427807486631016,
      "grad_norm": 6.037876002899973,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 0.7267,
      "step": 3
    },
    {
      "epoch": 0.0570409982174688,
      "grad_norm": 6.15687961618456,
      "learning_rate": 1.904761904761905e-06,
      "loss": 0.7342,
      "step": 4
    },
    {
      "epoch": 0.07130124777183601,
      "grad_norm": 5.532445239097713,
      "learning_rate": 2.380952380952381e-06,
      "loss": 0.6988,
      "step": 5
    },
    {
      "epoch": 0.0855614973262032,
      "grad_norm": 4.591054025035638,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.6871,
      "step": 6
    },
    {
      "epoch": 0.09982174688057041,
      "grad_norm": 4.163450801982825,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.6551,
      "step": 7
    },
    {
      "epoch": 0.1140819964349376,
      "grad_norm": 3.0450299693083838,
      "learning_rate": 3.80952380952381e-06,
      "loss": 0.5935,
      "step": 8
    },
    {
      "epoch": 0.12834224598930483,
      "grad_norm": 3.0171447511514304,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 0.61,
      "step": 9
    },
    {
      "epoch": 0.14260249554367202,
      "grad_norm": 3.412973255907802,
      "learning_rate": 4.761904761904762e-06,
      "loss": 0.5653,
      "step": 10
    },
    {
      "epoch": 0.1568627450980392,
      "grad_norm": 3.2639215322983093,
      "learning_rate": 5.2380952380952384e-06,
      "loss": 0.5505,
      "step": 11
    },
    {
      "epoch": 0.1711229946524064,
      "grad_norm": 2.7810256770169555,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.5189,
      "step": 12
    },
    {
      "epoch": 0.18538324420677363,
      "grad_norm": 1.855708432778589,
      "learning_rate": 6.1904761904761914e-06,
      "loss": 0.4986,
      "step": 13
    },
    {
      "epoch": 0.19964349376114082,
      "grad_norm": 2.4285867830525225,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.4825,
      "step": 14
    },
    {
      "epoch": 0.21390374331550802,
      "grad_norm": 2.0406805172911455,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.4781,
      "step": 15
    },
    {
      "epoch": 0.2281639928698752,
      "grad_norm": 1.7117644118774973,
      "learning_rate": 7.61904761904762e-06,
      "loss": 0.4717,
      "step": 16
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 1.3431052662117142,
      "learning_rate": 8.095238095238097e-06,
      "loss": 0.4626,
      "step": 17
    },
    {
      "epoch": 0.25668449197860965,
      "grad_norm": 1.2491319368394767,
      "learning_rate": 8.571428571428571e-06,
      "loss": 0.4483,
      "step": 18
    },
    {
      "epoch": 0.2709447415329768,
      "grad_norm": 1.064103742518249,
      "learning_rate": 9.047619047619049e-06,
      "loss": 0.4309,
      "step": 19
    },
    {
      "epoch": 0.28520499108734404,
      "grad_norm": 1.308684310053338,
      "learning_rate": 9.523809523809525e-06,
      "loss": 0.4186,
      "step": 20
    },
    {
      "epoch": 0.2994652406417112,
      "grad_norm": 1.0005997347773574,
      "learning_rate": 1e-05,
      "loss": 0.4289,
      "step": 21
    },
    {
      "epoch": 0.3137254901960784,
      "grad_norm": 1.0855537879905526,
      "learning_rate": 9.99930927345553e-06,
      "loss": 0.4509,
      "step": 22
    },
    {
      "epoch": 0.32798573975044565,
      "grad_norm": 0.9598047755603596,
      "learning_rate": 9.99723728466338e-06,
      "loss": 0.4166,
      "step": 23
    },
    {
      "epoch": 0.3422459893048128,
      "grad_norm": 0.9378891290105057,
      "learning_rate": 9.993784606094612e-06,
      "loss": 0.4124,
      "step": 24
    },
    {
      "epoch": 0.35650623885918004,
      "grad_norm": 0.8238281484919813,
      "learning_rate": 9.988952191691925e-06,
      "loss": 0.4111,
      "step": 25
    },
    {
      "epoch": 0.37076648841354726,
      "grad_norm": 0.8249603588788698,
      "learning_rate": 9.982741376606077e-06,
      "loss": 0.4022,
      "step": 26
    },
    {
      "epoch": 0.3850267379679144,
      "grad_norm": 0.7944536509888138,
      "learning_rate": 9.975153876827008e-06,
      "loss": 0.4096,
      "step": 27
    },
    {
      "epoch": 0.39928698752228164,
      "grad_norm": 0.8877336723504325,
      "learning_rate": 9.966191788709716e-06,
      "loss": 0.4078,
      "step": 28
    },
    {
      "epoch": 0.41354723707664887,
      "grad_norm": 0.9362944528113393,
      "learning_rate": 9.955857588395065e-06,
      "loss": 0.3974,
      "step": 29
    },
    {
      "epoch": 0.42780748663101603,
      "grad_norm": 1.0365875514927283,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.3907,
      "step": 30
    },
    {
      "epoch": 0.44206773618538325,
      "grad_norm": 0.9187628487659182,
      "learning_rate": 9.931084650456892e-06,
      "loss": 0.3788,
      "step": 31
    },
    {
      "epoch": 0.4563279857397504,
      "grad_norm": 0.9346374036471025,
      "learning_rate": 9.916652757363698e-06,
      "loss": 0.3933,
      "step": 32
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 0.8638182734526192,
      "learning_rate": 9.900862439242719e-06,
      "loss": 0.377,
      "step": 33
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 0.8897918969636153,
      "learning_rate": 9.883718058810708e-06,
      "loss": 0.3879,
      "step": 34
    },
    {
      "epoch": 0.49910873440285203,
      "grad_norm": 0.905899013833491,
      "learning_rate": 9.86522435289912e-06,
      "loss": 0.393,
      "step": 35
    },
    {
      "epoch": 0.5133689839572193,
      "grad_norm": 0.8313526898166544,
      "learning_rate": 9.84538643114539e-06,
      "loss": 0.3802,
      "step": 36
    },
    {
      "epoch": 0.5276292335115864,
      "grad_norm": 0.7981624010904073,
      "learning_rate": 9.824209774581176e-06,
      "loss": 0.3798,
      "step": 37
    },
    {
      "epoch": 0.5418894830659536,
      "grad_norm": 0.8107386340463024,
      "learning_rate": 9.801700234118e-06,
      "loss": 0.3776,
      "step": 38
    },
    {
      "epoch": 0.5561497326203209,
      "grad_norm": 0.7619681977203558,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.3815,
      "step": 39
    },
    {
      "epoch": 0.5704099821746881,
      "grad_norm": 0.8454306401600613,
      "learning_rate": 9.752707744739146e-06,
      "loss": 0.3776,
      "step": 40
    },
    {
      "epoch": 0.5846702317290553,
      "grad_norm": 0.8447995805383889,
      "learning_rate": 9.726238331988625e-06,
      "loss": 0.3704,
      "step": 41
    },
    {
      "epoch": 0.5989304812834224,
      "grad_norm": 0.785542071668213,
      "learning_rate": 9.698463103929542e-06,
      "loss": 0.3713,
      "step": 42
    },
    {
      "epoch": 0.6131907308377896,
      "grad_norm": 0.8441633538171278,
      "learning_rate": 9.669389734596819e-06,
      "loss": 0.3521,
      "step": 43
    },
    {
      "epoch": 0.6274509803921569,
      "grad_norm": 0.8104799297356389,
      "learning_rate": 9.639026256689628e-06,
      "loss": 0.379,
      "step": 44
    },
    {
      "epoch": 0.6417112299465241,
      "grad_norm": 0.7777936361439105,
      "learning_rate": 9.60738105935204e-06,
      "loss": 0.3682,
      "step": 45
    },
    {
      "epoch": 0.6559714795008913,
      "grad_norm": 0.7900521212207957,
      "learning_rate": 9.574462885855173e-06,
      "loss": 0.3708,
      "step": 46
    },
    {
      "epoch": 0.6702317290552585,
      "grad_norm": 0.8163359807292335,
      "learning_rate": 9.540280831181525e-06,
      "loss": 0.3605,
      "step": 47
    },
    {
      "epoch": 0.6844919786096256,
      "grad_norm": 0.872836606578105,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.3661,
      "step": 48
    },
    {
      "epoch": 0.6987522281639929,
      "grad_norm": 0.9033570080860079,
      "learning_rate": 9.468163201617063e-06,
      "loss": 0.3672,
      "step": 49
    },
    {
      "epoch": 0.7130124777183601,
      "grad_norm": 0.8384671316276696,
      "learning_rate": 9.430247552150673e-06,
      "loss": 0.3671,
      "step": 50
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.8616090850522994,
      "learning_rate": 9.391107866851143e-06,
      "loss": 0.3765,
      "step": 51
    },
    {
      "epoch": 0.7415329768270945,
      "grad_norm": 0.9146332401497722,
      "learning_rate": 9.350754959646306e-06,
      "loss": 0.3567,
      "step": 52
    },
    {
      "epoch": 0.7557932263814616,
      "grad_norm": 0.7467925012563463,
      "learning_rate": 9.309199979665821e-06,
      "loss": 0.3624,
      "step": 53
    },
    {
      "epoch": 0.7700534759358288,
      "grad_norm": 0.8106016103660585,
      "learning_rate": 9.266454408160779e-06,
      "loss": 0.3647,
      "step": 54
    },
    {
      "epoch": 0.7843137254901961,
      "grad_norm": 0.8134090391372545,
      "learning_rate": 9.22253005533154e-06,
      "loss": 0.3808,
      "step": 55
    },
    {
      "epoch": 0.7985739750445633,
      "grad_norm": 0.7921832122229222,
      "learning_rate": 9.177439057064684e-06,
      "loss": 0.3596,
      "step": 56
    },
    {
      "epoch": 0.8128342245989305,
      "grad_norm": 0.7303835443395563,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.349,
      "step": 57
    },
    {
      "epoch": 0.8270944741532977,
      "grad_norm": 0.7871828583047863,
      "learning_rate": 9.083807275988285e-06,
      "loss": 0.35,
      "step": 58
    },
    {
      "epoch": 0.8413547237076648,
      "grad_norm": 0.881224423943151,
      "learning_rate": 9.035292362761382e-06,
      "loss": 0.3666,
      "step": 59
    },
    {
      "epoch": 0.8556149732620321,
      "grad_norm": 0.6948998923174846,
      "learning_rate": 8.985662536114614e-06,
      "loss": 0.3581,
      "step": 60
    },
    {
      "epoch": 0.8698752228163993,
      "grad_norm": 0.7996766809552465,
      "learning_rate": 8.934931508303446e-06,
      "loss": 0.3618,
      "step": 61
    },
    {
      "epoch": 0.8841354723707665,
      "grad_norm": 0.7926297708544718,
      "learning_rate": 8.883113295834893e-06,
      "loss": 0.3654,
      "step": 62
    },
    {
      "epoch": 0.8983957219251337,
      "grad_norm": 0.8076359165867002,
      "learning_rate": 8.83022221559489e-06,
      "loss": 0.3628,
      "step": 63
    },
    {
      "epoch": 0.9126559714795008,
      "grad_norm": 0.8704725987761389,
      "learning_rate": 8.776272880892675e-06,
      "loss": 0.3615,
      "step": 64
    },
    {
      "epoch": 0.9269162210338681,
      "grad_norm": 0.7259385563641405,
      "learning_rate": 8.721280197423259e-06,
      "loss": 0.3391,
      "step": 65
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 0.7641083584826656,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.3584,
      "step": 66
    },
    {
      "epoch": 0.9554367201426025,
      "grad_norm": 0.8482726911049152,
      "learning_rate": 8.608225844102312e-06,
      "loss": 0.369,
      "step": 67
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.8036946709490872,
      "learning_rate": 8.550195410107903e-06,
      "loss": 0.3736,
      "step": 68
    },
    {
      "epoch": 0.983957219251337,
      "grad_norm": 0.7653533058452437,
      "learning_rate": 8.491184090430365e-06,
      "loss": 0.3401,
      "step": 69
    },
    {
      "epoch": 0.9982174688057041,
      "grad_norm": 0.8073514579287697,
      "learning_rate": 8.43120818934367e-06,
      "loss": 0.3493,
      "step": 70
    },
    {
      "epoch": 1.0124777183600713,
      "grad_norm": 0.6775703542593738,
      "learning_rate": 8.370284277626576e-06,
      "loss": 0.3162,
      "step": 71
    },
    {
      "epoch": 1.0267379679144386,
      "grad_norm": 0.6849387062750477,
      "learning_rate": 8.308429187984298e-06,
      "loss": 0.3132,
      "step": 72
    },
    {
      "epoch": 1.0409982174688057,
      "grad_norm": 0.7501115742708528,
      "learning_rate": 8.24566001039776e-06,
      "loss": 0.3107,
      "step": 73
    },
    {
      "epoch": 1.0552584670231728,
      "grad_norm": 0.7419796127086248,
      "learning_rate": 8.181994087401819e-06,
      "loss": 0.3114,
      "step": 74
    },
    {
      "epoch": 1.0695187165775402,
      "grad_norm": 0.7534872326532702,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.3228,
      "step": 75
    },
    {
      "epoch": 1.0837789661319073,
      "grad_norm": 0.8094312306607193,
      "learning_rate": 8.052042609272817e-06,
      "loss": 0.3022,
      "step": 76
    },
    {
      "epoch": 1.0980392156862746,
      "grad_norm": 0.7273023913777745,
      "learning_rate": 7.985792958513932e-06,
      "loss": 0.3077,
      "step": 77
    },
    {
      "epoch": 1.1122994652406417,
      "grad_norm": 0.6547411987494043,
      "learning_rate": 7.918718361173951e-06,
      "loss": 0.3137,
      "step": 78
    },
    {
      "epoch": 1.1265597147950088,
      "grad_norm": 0.7118409719814686,
      "learning_rate": 7.85083734933481e-06,
      "loss": 0.2952,
      "step": 79
    },
    {
      "epoch": 1.1408199643493762,
      "grad_norm": 0.7635016933875783,
      "learning_rate": 7.782168677883206e-06,
      "loss": 0.3138,
      "step": 80
    },
    {
      "epoch": 1.1550802139037433,
      "grad_norm": 0.6817008368615943,
      "learning_rate": 7.712731319328798e-06,
      "loss": 0.3035,
      "step": 81
    },
    {
      "epoch": 1.1693404634581106,
      "grad_norm": 0.8174647964374863,
      "learning_rate": 7.642544458562278e-06,
      "loss": 0.3284,
      "step": 82
    },
    {
      "epoch": 1.1836007130124777,
      "grad_norm": 0.6637911284519812,
      "learning_rate": 7.571627487554769e-06,
      "loss": 0.3174,
      "step": 83
    },
    {
      "epoch": 1.1978609625668448,
      "grad_norm": 0.7495098318592474,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3125,
      "step": 84
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.6420213528206444,
      "learning_rate": 7.4276817859007615e-06,
      "loss": 0.2909,
      "step": 85
    },
    {
      "epoch": 1.2263814616755793,
      "grad_norm": 0.7007340675237256,
      "learning_rate": 7.354692826101102e-06,
      "loss": 0.3114,
      "step": 86
    },
    {
      "epoch": 1.2406417112299466,
      "grad_norm": 0.706271617054947,
      "learning_rate": 7.281053286765816e-06,
      "loss": 0.2988,
      "step": 87
    },
    {
      "epoch": 1.2549019607843137,
      "grad_norm": 0.6255542997192427,
      "learning_rate": 7.206783513808721e-06,
      "loss": 0.3161,
      "step": 88
    },
    {
      "epoch": 1.2691622103386808,
      "grad_norm": 0.6771293525704715,
      "learning_rate": 7.1319040272712705e-06,
      "loss": 0.3164,
      "step": 89
    },
    {
      "epoch": 1.2834224598930482,
      "grad_norm": 0.7070886877032927,
      "learning_rate": 7.056435515653059e-06,
      "loss": 0.3312,
      "step": 90
    },
    {
      "epoch": 1.2976827094474153,
      "grad_norm": 0.6693214606442824,
      "learning_rate": 6.980398830195785e-06,
      "loss": 0.3101,
      "step": 91
    },
    {
      "epoch": 1.3119429590017826,
      "grad_norm": 0.7279758172512709,
      "learning_rate": 6.903814979122249e-06,
      "loss": 0.3197,
      "step": 92
    },
    {
      "epoch": 1.3262032085561497,
      "grad_norm": 0.6520276206041827,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.2975,
      "step": 93
    },
    {
      "epoch": 1.3404634581105168,
      "grad_norm": 0.6107316374770684,
      "learning_rate": 6.749090563055075e-06,
      "loss": 0.2969,
      "step": 94
    },
    {
      "epoch": 1.3547237076648841,
      "grad_norm": 0.6997488359926258,
      "learning_rate": 6.6709927469659385e-06,
      "loss": 0.2991,
      "step": 95
    },
    {
      "epoch": 1.3689839572192513,
      "grad_norm": 0.6541906454481444,
      "learning_rate": 6.592433251258423e-06,
      "loss": 0.3153,
      "step": 96
    },
    {
      "epoch": 1.3832442067736186,
      "grad_norm": 0.6674786478671763,
      "learning_rate": 6.513433781184131e-06,
      "loss": 0.2928,
      "step": 97
    },
    {
      "epoch": 1.3975044563279857,
      "grad_norm": 0.6623084412543156,
      "learning_rate": 6.434016163555452e-06,
      "loss": 0.3043,
      "step": 98
    },
    {
      "epoch": 1.4117647058823528,
      "grad_norm": 0.689312711915529,
      "learning_rate": 6.354202340715027e-06,
      "loss": 0.3344,
      "step": 99
    },
    {
      "epoch": 1.4260249554367201,
      "grad_norm": 0.6311516943939547,
      "learning_rate": 6.274014364473274e-06,
      "loss": 0.2968,
      "step": 100
    },
    {
      "epoch": 1.4402852049910875,
      "grad_norm": 0.6961729490571823,
      "learning_rate": 6.19347439001569e-06,
      "loss": 0.2925,
      "step": 101
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 0.6707000018979263,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.3126,
      "step": 102
    },
    {
      "epoch": 1.4688057040998217,
      "grad_norm": 0.6027031632940896,
      "learning_rate": 6.031427547315889e-06,
      "loss": 0.3023,
      "step": 103
    },
    {
      "epoch": 1.483065953654189,
      "grad_norm": 0.7096258454181166,
      "learning_rate": 5.949965451095952e-06,
      "loss": 0.3083,
      "step": 104
    },
    {
      "epoch": 1.4973262032085561,
      "grad_norm": 0.627103846901669,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.3129,
      "step": 105
    },
    {
      "epoch": 1.5115864527629235,
      "grad_norm": 0.6885664921641609,
      "learning_rate": 5.786276438761928e-06,
      "loss": 0.3134,
      "step": 106
    },
    {
      "epoch": 1.5258467023172906,
      "grad_norm": 0.6047321023835506,
      "learning_rate": 5.7040947483861845e-06,
      "loss": 0.3037,
      "step": 107
    },
    {
      "epoch": 1.5401069518716577,
      "grad_norm": 0.6014431063746195,
      "learning_rate": 5.621718523237427e-06,
      "loss": 0.3066,
      "step": 108
    },
    {
      "epoch": 1.5543672014260248,
      "grad_norm": 0.6759797800635752,
      "learning_rate": 5.539170523093794e-06,
      "loss": 0.2972,
      "step": 109
    },
    {
      "epoch": 1.5686274509803921,
      "grad_norm": 0.6459776120064094,
      "learning_rate": 5.456473555193242e-06,
      "loss": 0.2868,
      "step": 110
    },
    {
      "epoch": 1.5828877005347595,
      "grad_norm": 0.5990226927063328,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.3098,
      "step": 111
    },
    {
      "epoch": 1.5971479500891266,
      "grad_norm": 0.6724529376774372,
      "learning_rate": 5.290724144552379e-06,
      "loss": 0.296,
      "step": 112
    },
    {
      "epoch": 1.6114081996434937,
      "grad_norm": 0.6674952293711535,
      "learning_rate": 5.207717496819134e-06,
      "loss": 0.3061,
      "step": 113
    },
    {
      "epoch": 1.6256684491978608,
      "grad_norm": 0.6574349669336045,
      "learning_rate": 5.1246534586903655e-06,
      "loss": 0.3028,
      "step": 114
    },
    {
      "epoch": 1.6399286987522281,
      "grad_norm": 0.5996511909752443,
      "learning_rate": 5.041554979980487e-06,
      "loss": 0.306,
      "step": 115
    },
    {
      "epoch": 1.6541889483065955,
      "grad_norm": 0.6324851444349666,
      "learning_rate": 4.958445020019516e-06,
      "loss": 0.3113,
      "step": 116
    },
    {
      "epoch": 1.6684491978609626,
      "grad_norm": 0.5804606484309103,
      "learning_rate": 4.875346541309637e-06,
      "loss": 0.298,
      "step": 117
    },
    {
      "epoch": 1.6827094474153297,
      "grad_norm": 0.6297463707956045,
      "learning_rate": 4.792282503180867e-06,
      "loss": 0.2915,
      "step": 118
    },
    {
      "epoch": 1.696969696969697,
      "grad_norm": 0.59037299101606,
      "learning_rate": 4.7092758554476215e-06,
      "loss": 0.308,
      "step": 119
    },
    {
      "epoch": 1.7112299465240641,
      "grad_norm": 0.6386954686267667,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.2963,
      "step": 120
    },
    {
      "epoch": 1.7254901960784315,
      "grad_norm": 0.6145804027948849,
      "learning_rate": 4.5435264448067595e-06,
      "loss": 0.3089,
      "step": 121
    },
    {
      "epoch": 1.7397504456327986,
      "grad_norm": 0.6053923239121506,
      "learning_rate": 4.460829476906208e-06,
      "loss": 0.307,
      "step": 122
    },
    {
      "epoch": 1.7540106951871657,
      "grad_norm": 0.6502354241272797,
      "learning_rate": 4.3782814767625755e-06,
      "loss": 0.2997,
      "step": 123
    },
    {
      "epoch": 1.768270944741533,
      "grad_norm": 0.6047031298299372,
      "learning_rate": 4.295905251613817e-06,
      "loss": 0.2988,
      "step": 124
    },
    {
      "epoch": 1.7825311942959001,
      "grad_norm": 0.5997702616278708,
      "learning_rate": 4.213723561238074e-06,
      "loss": 0.3123,
      "step": 125
    },
    {
      "epoch": 1.7967914438502675,
      "grad_norm": 0.6254870806600614,
      "learning_rate": 4.131759111665349e-06,
      "loss": 0.3026,
      "step": 126
    },
    {
      "epoch": 1.8110516934046346,
      "grad_norm": 0.633357389799123,
      "learning_rate": 4.0500345489040515e-06,
      "loss": 0.2951,
      "step": 127
    },
    {
      "epoch": 1.8253119429590017,
      "grad_norm": 0.6404830360753402,
      "learning_rate": 3.968572452684113e-06,
      "loss": 0.2999,
      "step": 128
    },
    {
      "epoch": 1.839572192513369,
      "grad_norm": 0.6329066927776358,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.3035,
      "step": 129
    },
    {
      "epoch": 1.8538324420677363,
      "grad_norm": 0.589200866233192,
      "learning_rate": 3.806525609984312e-06,
      "loss": 0.3124,
      "step": 130
    },
    {
      "epoch": 1.8680926916221035,
      "grad_norm": 0.6082320573613086,
      "learning_rate": 3.7259856355267275e-06,
      "loss": 0.2956,
      "step": 131
    },
    {
      "epoch": 1.8823529411764706,
      "grad_norm": 0.6241929627393371,
      "learning_rate": 3.6457976592849753e-06,
      "loss": 0.2843,
      "step": 132
    },
    {
      "epoch": 1.8966131907308377,
      "grad_norm": 0.5731217918233906,
      "learning_rate": 3.5659838364445505e-06,
      "loss": 0.2962,
      "step": 133
    },
    {
      "epoch": 1.910873440285205,
      "grad_norm": 0.7203946076740332,
      "learning_rate": 3.4865662188158713e-06,
      "loss": 0.3019,
      "step": 134
    },
    {
      "epoch": 1.9251336898395723,
      "grad_norm": 0.5801797087810499,
      "learning_rate": 3.4075667487415785e-06,
      "loss": 0.2919,
      "step": 135
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.6017940447903612,
      "learning_rate": 3.3290072530340628e-06,
      "loss": 0.289,
      "step": 136
    },
    {
      "epoch": 1.9536541889483066,
      "grad_norm": 0.6783329890821536,
      "learning_rate": 3.250909436944928e-06,
      "loss": 0.3004,
      "step": 137
    },
    {
      "epoch": 1.9679144385026737,
      "grad_norm": 0.6181255692540811,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.2907,
      "step": 138
    },
    {
      "epoch": 1.982174688057041,
      "grad_norm": 0.5527014431122037,
      "learning_rate": 3.0961850208777527e-06,
      "loss": 0.3023,
      "step": 139
    },
    {
      "epoch": 1.9964349376114083,
      "grad_norm": 0.6138241158100955,
      "learning_rate": 3.019601169804216e-06,
      "loss": 0.2956,
      "step": 140
    },
    {
      "epoch": 2.0106951871657754,
      "grad_norm": 0.5795269162344017,
      "learning_rate": 2.9435644843469434e-06,
      "loss": 0.2645,
      "step": 141
    },
    {
      "epoch": 2.0249554367201426,
      "grad_norm": 0.6133201918050623,
      "learning_rate": 2.8680959727287316e-06,
      "loss": 0.2519,
      "step": 142
    },
    {
      "epoch": 2.0392156862745097,
      "grad_norm": 0.6132844962849042,
      "learning_rate": 2.7932164861912805e-06,
      "loss": 0.2813,
      "step": 143
    },
    {
      "epoch": 2.053475935828877,
      "grad_norm": 0.6064305890385137,
      "learning_rate": 2.718946713234185e-06,
      "loss": 0.2678,
      "step": 144
    },
    {
      "epoch": 2.0677361853832443,
      "grad_norm": 0.5071570057035093,
      "learning_rate": 2.645307173898901e-06,
      "loss": 0.2569,
      "step": 145
    },
    {
      "epoch": 2.0819964349376114,
      "grad_norm": 0.5593533988862088,
      "learning_rate": 2.5723182140992385e-06,
      "loss": 0.2672,
      "step": 146
    },
    {
      "epoch": 2.0962566844919786,
      "grad_norm": 0.5681953675884334,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.2509,
      "step": 147
    },
    {
      "epoch": 2.1105169340463457,
      "grad_norm": 0.5238572661786828,
      "learning_rate": 2.428372512445233e-06,
      "loss": 0.2654,
      "step": 148
    },
    {
      "epoch": 2.124777183600713,
      "grad_norm": 0.5212764514198992,
      "learning_rate": 2.357455541437723e-06,
      "loss": 0.2593,
      "step": 149
    },
    {
      "epoch": 2.1390374331550803,
      "grad_norm": 0.5009569961577103,
      "learning_rate": 2.2872686806712037e-06,
      "loss": 0.2651,
      "step": 150
    },
    {
      "epoch": 2.1532976827094474,
      "grad_norm": 0.5016236354852869,
      "learning_rate": 2.217831322116797e-06,
      "loss": 0.253,
      "step": 151
    },
    {
      "epoch": 2.1675579322638145,
      "grad_norm": 0.5385410771208106,
      "learning_rate": 2.1491626506651914e-06,
      "loss": 0.2567,
      "step": 152
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 0.52490098008563,
      "learning_rate": 2.081281638826052e-06,
      "loss": 0.2653,
      "step": 153
    },
    {
      "epoch": 2.196078431372549,
      "grad_norm": 0.5547010556928098,
      "learning_rate": 2.0142070414860704e-06,
      "loss": 0.2498,
      "step": 154
    },
    {
      "epoch": 2.2103386809269163,
      "grad_norm": 0.5326534701282144,
      "learning_rate": 1.947957390727185e-06,
      "loss": 0.2677,
      "step": 155
    },
    {
      "epoch": 2.2245989304812834,
      "grad_norm": 0.5676341178655185,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.262,
      "step": 156
    },
    {
      "epoch": 2.2388591800356505,
      "grad_norm": 0.5075550544962953,
      "learning_rate": 1.8180059125981826e-06,
      "loss": 0.2609,
      "step": 157
    },
    {
      "epoch": 2.2531194295900177,
      "grad_norm": 0.5378049836987813,
      "learning_rate": 1.7543399896022406e-06,
      "loss": 0.2401,
      "step": 158
    },
    {
      "epoch": 2.267379679144385,
      "grad_norm": 0.5313735495324294,
      "learning_rate": 1.6915708120157042e-06,
      "loss": 0.2478,
      "step": 159
    },
    {
      "epoch": 2.2816399286987523,
      "grad_norm": 0.4997207720980407,
      "learning_rate": 1.6297157223734228e-06,
      "loss": 0.2608,
      "step": 160
    },
    {
      "epoch": 2.2959001782531194,
      "grad_norm": 0.5354164171295689,
      "learning_rate": 1.5687918106563326e-06,
      "loss": 0.2521,
      "step": 161
    },
    {
      "epoch": 2.3101604278074865,
      "grad_norm": 0.5521694352110376,
      "learning_rate": 1.5088159095696365e-06,
      "loss": 0.2618,
      "step": 162
    },
    {
      "epoch": 2.3244206773618536,
      "grad_norm": 0.5419705137185197,
      "learning_rate": 1.4498045898920988e-06,
      "loss": 0.2415,
      "step": 163
    },
    {
      "epoch": 2.338680926916221,
      "grad_norm": 0.5289333038282386,
      "learning_rate": 1.3917741558976894e-06,
      "loss": 0.2594,
      "step": 164
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 0.4958046194398315,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.2547,
      "step": 165
    },
    {
      "epoch": 2.3672014260249554,
      "grad_norm": 0.5275471951202156,
      "learning_rate": 1.2787198025767417e-06,
      "loss": 0.2498,
      "step": 166
    },
    {
      "epoch": 2.3814616755793225,
      "grad_norm": 0.5439621033914177,
      "learning_rate": 1.223727119107327e-06,
      "loss": 0.2569,
      "step": 167
    },
    {
      "epoch": 2.3957219251336896,
      "grad_norm": 0.5020641846014723,
      "learning_rate": 1.1697777844051105e-06,
      "loss": 0.2682,
      "step": 168
    },
    {
      "epoch": 2.409982174688057,
      "grad_norm": 0.5332076852350075,
      "learning_rate": 1.1168867041651082e-06,
      "loss": 0.2522,
      "step": 169
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.5479842841748694,
      "learning_rate": 1.065068491696556e-06,
      "loss": 0.242,
      "step": 170
    },
    {
      "epoch": 2.4385026737967914,
      "grad_norm": 0.5337676381332789,
      "learning_rate": 1.0143374638853892e-06,
      "loss": 0.255,
      "step": 171
    },
    {
      "epoch": 2.4527629233511585,
      "grad_norm": 0.4899023747726101,
      "learning_rate": 9.647076372386195e-07,
      "loss": 0.2552,
      "step": 172
    },
    {
      "epoch": 2.4670231729055256,
      "grad_norm": 0.542407864218085,
      "learning_rate": 9.161927240117174e-07,
      "loss": 0.2662,
      "step": 173
    },
    {
      "epoch": 2.481283422459893,
      "grad_norm": 0.47566063751571563,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.2581,
      "step": 174
    },
    {
      "epoch": 2.4955436720142603,
      "grad_norm": 0.47291898298798923,
      "learning_rate": 8.225609429353187e-07,
      "loss": 0.2679,
      "step": 175
    },
    {
      "epoch": 2.5098039215686274,
      "grad_norm": 0.4856493909985723,
      "learning_rate": 7.774699446684608e-07,
      "loss": 0.2558,
      "step": 176
    },
    {
      "epoch": 2.5240641711229945,
      "grad_norm": 0.5468886611950984,
      "learning_rate": 7.33545591839222e-07,
      "loss": 0.2575,
      "step": 177
    },
    {
      "epoch": 2.5383244206773616,
      "grad_norm": 0.5227642129317599,
      "learning_rate": 6.908000203341802e-07,
      "loss": 0.2726,
      "step": 178
    },
    {
      "epoch": 2.552584670231729,
      "grad_norm": 0.5243649599139041,
      "learning_rate": 6.492450403536959e-07,
      "loss": 0.2691,
      "step": 179
    },
    {
      "epoch": 2.5668449197860963,
      "grad_norm": 0.5038584609402851,
      "learning_rate": 6.088921331488568e-07,
      "loss": 0.2717,
      "step": 180
    },
    {
      "epoch": 2.5811051693404634,
      "grad_norm": 0.48239876397554576,
      "learning_rate": 5.697524478493288e-07,
      "loss": 0.2598,
      "step": 181
    },
    {
      "epoch": 2.5953654188948305,
      "grad_norm": 0.5437994297898453,
      "learning_rate": 5.318367983829393e-07,
      "loss": 0.259,
      "step": 182
    },
    {
      "epoch": 2.6096256684491976,
      "grad_norm": 0.49097378469526687,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.262,
      "step": 183
    },
    {
      "epoch": 2.623885918003565,
      "grad_norm": 0.5189871183074696,
      "learning_rate": 4.5971916881847543e-07,
      "loss": 0.2473,
      "step": 184
    },
    {
      "epoch": 2.6381461675579323,
      "grad_norm": 0.4963009680587836,
      "learning_rate": 4.255371141448272e-07,
      "loss": 0.2621,
      "step": 185
    },
    {
      "epoch": 2.6524064171122994,
      "grad_norm": 0.4845160599159596,
      "learning_rate": 3.9261894064796136e-07,
      "loss": 0.2496,
      "step": 186
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.5085689793959361,
      "learning_rate": 3.6097374331037326e-07,
      "loss": 0.2422,
      "step": 187
    },
    {
      "epoch": 2.6809269162210336,
      "grad_norm": 0.4910420456640142,
      "learning_rate": 3.306102654031823e-07,
      "loss": 0.2691,
      "step": 188
    },
    {
      "epoch": 2.695187165775401,
      "grad_norm": 0.5080309292352433,
      "learning_rate": 3.015368960704584e-07,
      "loss": 0.2769,
      "step": 189
    },
    {
      "epoch": 2.7094474153297683,
      "grad_norm": 0.4486963045863036,
      "learning_rate": 2.737616680113758e-07,
      "loss": 0.2595,
      "step": 190
    },
    {
      "epoch": 2.7237076648841354,
      "grad_norm": 0.5134768408162061,
      "learning_rate": 2.472922552608559e-07,
      "loss": 0.2466,
      "step": 191
    },
    {
      "epoch": 2.7379679144385025,
      "grad_norm": 0.4729499265055225,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.2438,
      "step": 192
    },
    {
      "epoch": 2.7522281639928696,
      "grad_norm": 0.4756332341553558,
      "learning_rate": 1.982997658820013e-07,
      "loss": 0.2619,
      "step": 193
    },
    {
      "epoch": 2.766488413547237,
      "grad_norm": 0.5111999088028844,
      "learning_rate": 1.757902254188254e-07,
      "loss": 0.242,
      "step": 194
    },
    {
      "epoch": 2.7807486631016043,
      "grad_norm": 0.5009830915446124,
      "learning_rate": 1.5461356885461077e-07,
      "loss": 0.2463,
      "step": 195
    },
    {
      "epoch": 2.7950089126559714,
      "grad_norm": 0.5063176331660432,
      "learning_rate": 1.3477564710088097e-07,
      "loss": 0.2629,
      "step": 196
    },
    {
      "epoch": 2.809269162210339,
      "grad_norm": 0.5389279146565609,
      "learning_rate": 1.1628194118929403e-07,
      "loss": 0.2654,
      "step": 197
    },
    {
      "epoch": 2.8235294117647056,
      "grad_norm": 0.5002474822940156,
      "learning_rate": 9.913756075728088e-08,
      "loss": 0.256,
      "step": 198
    },
    {
      "epoch": 2.837789661319073,
      "grad_norm": 0.4856526207889643,
      "learning_rate": 8.334724263630301e-08,
      "loss": 0.2672,
      "step": 199
    },
    {
      "epoch": 2.8520499108734403,
      "grad_norm": 0.4773574713690703,
      "learning_rate": 6.891534954310886e-08,
      "loss": 0.2569,
      "step": 200
    },
    {
      "epoch": 2.8663101604278074,
      "grad_norm": 0.4725556890683267,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.2484,
      "step": 201
    },
    {
      "epoch": 2.880570409982175,
      "grad_norm": 0.5061174548555849,
      "learning_rate": 4.41424116049366e-08,
      "loss": 0.2564,
      "step": 202
    },
    {
      "epoch": 2.8948306595365416,
      "grad_norm": 0.4978103192012945,
      "learning_rate": 3.3808211290284886e-08,
      "loss": 0.2626,
      "step": 203
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.4646163044086365,
      "learning_rate": 2.4846123172992953e-08,
      "loss": 0.2605,
      "step": 204
    },
    {
      "epoch": 2.9233511586452763,
      "grad_norm": 0.4968912078361725,
      "learning_rate": 1.725862339392259e-08,
      "loss": 0.2543,
      "step": 205
    },
    {
      "epoch": 2.9376114081996434,
      "grad_norm": 0.47130252532014427,
      "learning_rate": 1.1047808308075059e-08,
      "loss": 0.2533,
      "step": 206
    },
    {
      "epoch": 2.951871657754011,
      "grad_norm": 0.5105837310688802,
      "learning_rate": 6.215393905388278e-09,
      "loss": 0.2753,
      "step": 207
    },
    {
      "epoch": 2.966131907308378,
      "grad_norm": 0.5097896921787389,
      "learning_rate": 2.7627153366222014e-09,
      "loss": 0.2676,
      "step": 208
    },
    {
      "epoch": 2.980392156862745,
      "grad_norm": 0.5475905566370015,
      "learning_rate": 6.907265444716649e-10,
      "loss": 0.2399,
      "step": 209
    },
    {
      "epoch": 2.9946524064171123,
      "grad_norm": 0.4933114168605027,
      "learning_rate": 0.0,
      "loss": 0.2595,
      "step": 210
    },
    {
      "epoch": 2.9946524064171123,
      "step": 210,
      "total_flos": 95562719805440.0,
      "train_loss": 0.3320285594179517,
      "train_runtime": 9818.8898,
      "train_samples_per_second": 1.371,
      "train_steps_per_second": 0.021
    }
  ],
  "logging_steps": 1,
  "max_steps": 210,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 95562719805440.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}