{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.955555555555556,
  "eval_steps": 500,
  "global_step": 560,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017777777777777778,
      "grad_norm": 1.7422537803649902,
      "learning_rate": 1.7857142857142857e-06,
      "loss": 0.7247,
      "step": 1
    },
    {
      "epoch": 0.035555555555555556,
      "grad_norm": 1.5862261056900024,
      "learning_rate": 3.5714285714285714e-06,
      "loss": 0.6432,
      "step": 2
    },
    {
      "epoch": 0.05333333333333334,
      "grad_norm": 1.3740944862365723,
      "learning_rate": 5.357142857142857e-06,
      "loss": 0.6158,
      "step": 3
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 1.5012167692184448,
      "learning_rate": 7.142857142857143e-06,
      "loss": 0.6322,
      "step": 4
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 1.4912959337234497,
      "learning_rate": 8.92857142857143e-06,
      "loss": 0.5955,
      "step": 5
    },
    {
      "epoch": 0.10666666666666667,
      "grad_norm": 1.462948203086853,
      "learning_rate": 1.0714285714285714e-05,
      "loss": 0.6794,
      "step": 6
    },
    {
      "epoch": 0.12444444444444444,
      "grad_norm": 1.226659893989563,
      "learning_rate": 1.25e-05,
      "loss": 0.5716,
      "step": 7
    },
    {
      "epoch": 0.14222222222222222,
      "grad_norm": 1.1903551816940308,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.6785,
      "step": 8
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.4021542072296143,
      "learning_rate": 1.6071428571428572e-05,
      "loss": 0.5768,
      "step": 9
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 1.4458407163619995,
      "learning_rate": 1.785714285714286e-05,
      "loss": 0.653,
      "step": 10
    },
    {
      "epoch": 0.19555555555555557,
      "grad_norm": 1.669067144393921,
      "learning_rate": 1.9642857142857145e-05,
      "loss": 0.608,
      "step": 11
    },
    {
      "epoch": 0.21333333333333335,
      "grad_norm": 1.5338914394378662,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.6273,
      "step": 12
    },
    {
      "epoch": 0.2311111111111111,
      "grad_norm": 1.299443006515503,
      "learning_rate": 2.3214285714285715e-05,
      "loss": 0.5894,
      "step": 13
    },
    {
      "epoch": 0.24888888888888888,
      "grad_norm": 1.3132046461105347,
      "learning_rate": 2.5e-05,
      "loss": 0.6952,
      "step": 14
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 1.780645728111267,
      "learning_rate": 2.6785714285714288e-05,
      "loss": 0.695,
      "step": 15
    },
    {
      "epoch": 0.28444444444444444,
      "grad_norm": 1.391330361366272,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.5933,
      "step": 16
    },
    {
      "epoch": 0.3022222222222222,
      "grad_norm": 1.3170968294143677,
      "learning_rate": 3.0357142857142857e-05,
      "loss": 0.5422,
      "step": 17
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0980342626571655,
      "learning_rate": 3.2142857142857144e-05,
      "loss": 0.5283,
      "step": 18
    },
    {
      "epoch": 0.3377777777777778,
      "grad_norm": 1.122798204421997,
      "learning_rate": 3.392857142857143e-05,
      "loss": 0.5295,
      "step": 19
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 1.741687297821045,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.4891,
      "step": 20
    },
    {
      "epoch": 0.37333333333333335,
      "grad_norm": 1.3390421867370605,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.5169,
      "step": 21
    },
    {
      "epoch": 0.39111111111111113,
      "grad_norm": 1.5962032079696655,
      "learning_rate": 3.928571428571429e-05,
      "loss": 0.5211,
      "step": 22
    },
    {
      "epoch": 0.4088888888888889,
      "grad_norm": 1.4163485765457153,
      "learning_rate": 4.107142857142857e-05,
      "loss": 0.4661,
      "step": 23
    },
    {
      "epoch": 0.4266666666666667,
      "grad_norm": 1.2807841300964355,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.4794,
      "step": 24
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 1.7780894041061401,
      "learning_rate": 4.464285714285715e-05,
      "loss": 0.4965,
      "step": 25
    },
    {
      "epoch": 0.4622222222222222,
      "grad_norm": 3.661797285079956,
      "learning_rate": 4.642857142857143e-05,
      "loss": 0.8358,
      "step": 26
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.6696407794952393,
      "learning_rate": 4.8214285714285716e-05,
      "loss": 0.4957,
      "step": 27
    },
    {
      "epoch": 0.49777777777777776,
      "grad_norm": 1.1838289499282837,
      "learning_rate": 5e-05,
      "loss": 0.5393,
      "step": 28
    },
    {
      "epoch": 0.5155555555555555,
      "grad_norm": 1.532225489616394,
      "learning_rate": 5.1785714285714296e-05,
      "loss": 0.5784,
      "step": 29
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 2.2218728065490723,
      "learning_rate": 5.3571428571428575e-05,
      "loss": 0.4446,
      "step": 30
    },
    {
      "epoch": 0.5511111111111111,
      "grad_norm": 1.4392008781433105,
      "learning_rate": 5.535714285714286e-05,
      "loss": 0.3995,
      "step": 31
    },
    {
      "epoch": 0.5688888888888889,
      "grad_norm": 0.8953332901000977,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.3399,
      "step": 32
    },
    {
      "epoch": 0.5866666666666667,
      "grad_norm": 1.1794021129608154,
      "learning_rate": 5.8928571428571435e-05,
      "loss": 0.3983,
      "step": 33
    },
    {
      "epoch": 0.6044444444444445,
      "grad_norm": 0.9146949052810669,
      "learning_rate": 6.0714285714285715e-05,
      "loss": 0.3831,
      "step": 34
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 0.4634794294834137,
      "learning_rate": 6.25e-05,
      "loss": 0.3359,
      "step": 35
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.5568346977233887,
      "learning_rate": 6.428571428571429e-05,
      "loss": 0.3597,
      "step": 36
    },
    {
      "epoch": 0.6577777777777778,
      "grad_norm": 0.6084932684898376,
      "learning_rate": 6.607142857142857e-05,
      "loss": 0.4061,
      "step": 37
    },
    {
      "epoch": 0.6755555555555556,
      "grad_norm": 0.6706041693687439,
      "learning_rate": 6.785714285714286e-05,
      "loss": 0.3533,
      "step": 38
    },
    {
      "epoch": 0.6933333333333334,
      "grad_norm": 0.9489718079566956,
      "learning_rate": 6.964285714285715e-05,
      "loss": 0.3598,
      "step": 39
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 1.2667860984802246,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.3796,
      "step": 40
    },
    {
      "epoch": 0.7288888888888889,
      "grad_norm": 1.1869858503341675,
      "learning_rate": 7.321428571428571e-05,
      "loss": 0.4386,
      "step": 41
    },
    {
      "epoch": 0.7466666666666667,
      "grad_norm": 0.5942434668540955,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.3265,
      "step": 42
    },
    {
      "epoch": 0.7644444444444445,
      "grad_norm": 0.7414906620979309,
      "learning_rate": 7.67857142857143e-05,
      "loss": 0.3344,
      "step": 43
    },
    {
      "epoch": 0.7822222222222223,
      "grad_norm": 0.6775667071342468,
      "learning_rate": 7.857142857142858e-05,
      "loss": 0.3514,
      "step": 44
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5995815992355347,
      "learning_rate": 8.035714285714287e-05,
      "loss": 0.3258,
      "step": 45
    },
    {
      "epoch": 0.8177777777777778,
      "grad_norm": 0.42830514907836914,
      "learning_rate": 8.214285714285714e-05,
      "loss": 0.3228,
      "step": 46
    },
    {
      "epoch": 0.8355555555555556,
      "grad_norm": 0.5796249508857727,
      "learning_rate": 8.392857142857144e-05,
      "loss": 0.3752,
      "step": 47
    },
    {
      "epoch": 0.8533333333333334,
      "grad_norm": 0.5575869083404541,
      "learning_rate": 8.571428571428571e-05,
      "loss": 0.3656,
      "step": 48
    },
    {
      "epoch": 0.8711111111111111,
      "grad_norm": 0.3702499270439148,
      "learning_rate": 8.75e-05,
      "loss": 0.3254,
      "step": 49
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.3036387860774994,
      "learning_rate": 8.92857142857143e-05,
      "loss": 0.2911,
      "step": 50
    },
    {
      "epoch": 0.9066666666666666,
      "grad_norm": 0.6176546216011047,
      "learning_rate": 9.107142857142857e-05,
      "loss": 0.3641,
      "step": 51
    },
    {
      "epoch": 0.9244444444444444,
      "grad_norm": 0.46132686734199524,
      "learning_rate": 9.285714285714286e-05,
      "loss": 0.3164,
      "step": 52
    },
    {
      "epoch": 0.9422222222222222,
      "grad_norm": 0.42897599935531616,
      "learning_rate": 9.464285714285715e-05,
      "loss": 0.3216,
      "step": 53
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.1321483850479126,
      "learning_rate": 9.642857142857143e-05,
      "loss": 0.4645,
      "step": 54
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 0.345939576625824,
      "learning_rate": 9.821428571428572e-05,
      "loss": 0.341,
      "step": 55
    },
    {
      "epoch": 0.9955555555555555,
      "grad_norm": 0.2705654203891754,
      "learning_rate": 0.0001,
      "loss": 0.3029,
      "step": 56
    },
    {
      "epoch": 1.0133333333333334,
      "grad_norm": 0.38632267713546753,
      "learning_rate": 9.999902864657691e-05,
      "loss": 0.3005,
      "step": 57
    },
    {
      "epoch": 1.031111111111111,
      "grad_norm": 0.3401973247528076,
      "learning_rate": 9.999611462404875e-05,
      "loss": 0.3182,
      "step": 58
    },
    {
      "epoch": 1.048888888888889,
      "grad_norm": 1.9517568349838257,
      "learning_rate": 9.999125804563732e-05,
      "loss": 0.4438,
      "step": 59
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 0.3414149582386017,
      "learning_rate": 9.998445910004082e-05,
      "loss": 0.3115,
      "step": 60
    },
    {
      "epoch": 1.0844444444444445,
      "grad_norm": 0.655548095703125,
      "learning_rate": 9.997571805142639e-05,
      "loss": 0.4094,
      "step": 61
    },
    {
      "epoch": 1.1022222222222222,
      "grad_norm": 0.49635574221611023,
      "learning_rate": 9.996503523941994e-05,
      "loss": 0.3135,
      "step": 62
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.34827786684036255,
      "learning_rate": 9.99524110790929e-05,
      "loss": 0.2901,
      "step": 63
    },
    {
      "epoch": 1.1377777777777778,
      "grad_norm": 0.4313528835773468,
      "learning_rate": 9.993784606094612e-05,
      "loss": 0.3178,
      "step": 64
    },
    {
      "epoch": 1.1555555555555554,
      "grad_norm": 0.5291451215744019,
      "learning_rate": 9.992134075089084e-05,
      "loss": 0.3534,
      "step": 65
    },
    {
      "epoch": 1.1733333333333333,
      "grad_norm": 0.4702828526496887,
      "learning_rate": 9.99028957902266e-05,
      "loss": 0.3155,
      "step": 66
    },
    {
      "epoch": 1.1911111111111112,
      "grad_norm": 0.5662183165550232,
      "learning_rate": 9.988251189561645e-05,
      "loss": 0.3555,
      "step": 67
    },
    {
      "epoch": 1.208888888888889,
      "grad_norm": 0.3226587772369385,
      "learning_rate": 9.986018985905901e-05,
      "loss": 0.33,
      "step": 68
    },
    {
      "epoch": 1.2266666666666666,
      "grad_norm": 0.7809450030326843,
      "learning_rate": 9.983593054785776e-05,
      "loss": 0.3499,
      "step": 69
    },
    {
      "epoch": 1.2444444444444445,
      "grad_norm": 0.6240183115005493,
      "learning_rate": 9.980973490458728e-05,
      "loss": 0.353,
      "step": 70
    },
    {
      "epoch": 1.2622222222222224,
      "grad_norm": 0.34356218576431274,
      "learning_rate": 9.978160394705668e-05,
      "loss": 0.2993,
      "step": 71
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.33020302653312683,
      "learning_rate": 9.975153876827008e-05,
      "loss": 0.3122,
      "step": 72
    },
    {
      "epoch": 1.2977777777777777,
      "grad_norm": 0.4314161241054535,
      "learning_rate": 9.971954053638399e-05,
      "loss": 0.3258,
      "step": 73
    },
    {
      "epoch": 1.3155555555555556,
      "grad_norm": 0.5286882519721985,
      "learning_rate": 9.968561049466214e-05,
      "loss": 0.3835,
      "step": 74
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.2787860929965973,
      "learning_rate": 9.964974996142698e-05,
      "loss": 0.2817,
      "step": 75
    },
    {
      "epoch": 1.3511111111111112,
      "grad_norm": 0.29439663887023926,
      "learning_rate": 9.961196033000861e-05,
      "loss": 0.2828,
      "step": 76
    },
    {
      "epoch": 1.3688888888888888,
      "grad_norm": 0.27987906336784363,
      "learning_rate": 9.957224306869053e-05,
      "loss": 0.3253,
      "step": 77
    },
    {
      "epoch": 1.3866666666666667,
      "grad_norm": 0.3439163863658905,
      "learning_rate": 9.953059972065265e-05,
      "loss": 0.3215,
      "step": 78
    },
    {
      "epoch": 1.4044444444444444,
      "grad_norm": 0.32061126828193665,
      "learning_rate": 9.948703190391131e-05,
      "loss": 0.323,
      "step": 79
    },
    {
      "epoch": 1.4222222222222223,
      "grad_norm": 0.4793407917022705,
      "learning_rate": 9.944154131125642e-05,
      "loss": 0.2962,
      "step": 80
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.2514803111553192,
      "learning_rate": 9.939412971018574e-05,
      "loss": 0.314,
      "step": 81
    },
    {
      "epoch": 1.4577777777777778,
      "grad_norm": 0.4879082143306732,
      "learning_rate": 9.934479894283606e-05,
      "loss": 0.2828,
      "step": 82
    },
    {
      "epoch": 1.4755555555555555,
      "grad_norm": 0.41817083954811096,
      "learning_rate": 9.92935509259118e-05,
      "loss": 0.3124,
      "step": 83
    },
    {
      "epoch": 1.4933333333333334,
      "grad_norm": 0.3143192529678345,
      "learning_rate": 9.924038765061042e-05,
      "loss": 0.2861,
      "step": 84
    },
    {
      "epoch": 1.511111111111111,
      "grad_norm": 0.33786436915397644,
      "learning_rate": 9.918531118254507e-05,
      "loss": 0.286,
      "step": 85
    },
    {
      "epoch": 1.528888888888889,
      "grad_norm": 0.3116290867328644,
      "learning_rate": 9.912832366166442e-05,
      "loss": 0.307,
      "step": 86
    },
    {
      "epoch": 1.5466666666666666,
      "grad_norm": 0.5235384106636047,
      "learning_rate": 9.906942730216939e-05,
      "loss": 0.3053,
      "step": 87
    },
    {
      "epoch": 1.5644444444444443,
      "grad_norm": 0.3879358172416687,
      "learning_rate": 9.900862439242719e-05,
      "loss": 0.3108,
      "step": 88
    },
    {
      "epoch": 1.5822222222222222,
      "grad_norm": 0.4028690755367279,
      "learning_rate": 9.894591729488242e-05,
      "loss": 0.3178,
      "step": 89
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.3827480375766754,
      "learning_rate": 9.888130844596524e-05,
      "loss": 0.303,
      "step": 90
    },
    {
      "epoch": 1.6177777777777778,
      "grad_norm": 0.48203060030937195,
      "learning_rate": 9.881480035599667e-05,
      "loss": 0.3503,
      "step": 91
    },
    {
      "epoch": 1.6355555555555554,
      "grad_norm": 0.30467236042022705,
      "learning_rate": 9.874639560909117e-05,
      "loss": 0.2876,
      "step": 92
    },
    {
      "epoch": 1.6533333333333333,
      "grad_norm": 0.4495410621166229,
      "learning_rate": 9.867609686305617e-05,
      "loss": 0.2937,
      "step": 93
    },
    {
      "epoch": 1.6711111111111112,
      "grad_norm": 0.29854950308799744,
      "learning_rate": 9.860390684928873e-05,
      "loss": 0.3155,
      "step": 94
    },
    {
      "epoch": 1.6888888888888889,
      "grad_norm": 0.4602304995059967,
      "learning_rate": 9.852982837266955e-05,
      "loss": 0.276,
      "step": 95
    },
    {
      "epoch": 1.7066666666666666,
      "grad_norm": 0.34135401248931885,
      "learning_rate": 9.84538643114539e-05,
      "loss": 0.2831,
      "step": 96
    },
    {
      "epoch": 1.7244444444444444,
      "grad_norm": 0.26776087284088135,
      "learning_rate": 9.837601761715983e-05,
      "loss": 0.2964,
      "step": 97
    },
    {
      "epoch": 1.7422222222222223,
      "grad_norm": 0.3476792573928833,
      "learning_rate": 9.829629131445342e-05,
      "loss": 0.2814,
      "step": 98
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.2479284703731537,
      "learning_rate": 9.82146885010314e-05,
      "loss": 0.2806,
      "step": 99
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 0.32680875062942505,
      "learning_rate": 9.81312123475006e-05,
      "loss": 0.3065,
      "step": 100
    },
    {
      "epoch": 1.7955555555555556,
      "grad_norm": 0.28494346141815186,
      "learning_rate": 9.804586609725499e-05,
      "loss": 0.311,
      "step": 101
    },
    {
      "epoch": 1.8133333333333335,
      "grad_norm": 0.444627583026886,
      "learning_rate": 9.79586530663494e-05,
      "loss": 0.3003,
      "step": 102
    },
    {
      "epoch": 1.8311111111111111,
      "grad_norm": 0.35969412326812744,
      "learning_rate": 9.78695766433709e-05,
      "loss": 0.2727,
      "step": 103
    },
    {
      "epoch": 1.8488888888888888,
      "grad_norm": 0.378616064786911,
      "learning_rate": 9.777864028930705e-05,
      "loss": 0.2834,
      "step": 104
    },
    {
      "epoch": 1.8666666666666667,
      "grad_norm": 0.2953629493713379,
      "learning_rate": 9.768584753741134e-05,
      "loss": 0.2806,
      "step": 105
    },
    {
      "epoch": 1.8844444444444446,
      "grad_norm": 0.44163236021995544,
      "learning_rate": 9.759120199306613e-05,
      "loss": 0.2951,
      "step": 106
    },
    {
      "epoch": 1.9022222222222223,
      "grad_norm": 0.26908189058303833,
      "learning_rate": 9.74947073336423e-05,
      "loss": 0.2975,
      "step": 107
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.2417159080505371,
      "learning_rate": 9.73963673083566e-05,
      "loss": 0.2677,
      "step": 108
    },
    {
      "epoch": 1.9377777777777778,
      "grad_norm": 0.3327369689941406,
      "learning_rate": 9.72961857381258e-05,
      "loss": 0.3219,
      "step": 109
    },
    {
      "epoch": 1.9555555555555557,
      "grad_norm": 0.27677884697914124,
      "learning_rate": 9.719416651541839e-05,
      "loss": 0.2631,
      "step": 110
    },
    {
      "epoch": 1.9733333333333334,
      "grad_norm": 0.41239938139915466,
      "learning_rate": 9.709031360410318e-05,
      "loss": 0.2788,
      "step": 111
    },
    {
      "epoch": 1.991111111111111,
      "grad_norm": 0.28369349241256714,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.2723,
      "step": 112
    },
    {
      "epoch": 2.008888888888889,
      "grad_norm": 0.3098125457763672,
      "learning_rate": 9.687712292719997e-05,
      "loss": 0.3279,
      "step": 113
    },
    {
      "epoch": 2.026666666666667,
      "grad_norm": 0.35817790031433105,
      "learning_rate": 9.67677934449517e-05,
      "loss": 0.2708,
      "step": 114
    },
    {
      "epoch": 2.0444444444444443,
      "grad_norm": 0.32206666469573975,
      "learning_rate": 9.665664684045333e-05,
      "loss": 0.2801,
      "step": 115
    },
    {
      "epoch": 2.062222222222222,
      "grad_norm": 0.2292087823152542,
      "learning_rate": 9.654368743221022e-05,
      "loss": 0.218,
      "step": 116
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.30288153886795044,
      "learning_rate": 9.642891960916268e-05,
      "loss": 0.2865,
      "step": 117
    },
    {
      "epoch": 2.097777777777778,
      "grad_norm": 0.33361321687698364,
      "learning_rate": 9.631234783051544e-05,
      "loss": 0.2866,
      "step": 118
    },
    {
      "epoch": 2.1155555555555554,
      "grad_norm": 0.2744784951210022,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.2776,
      "step": 119
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 0.29272395372390747,
      "learning_rate": 9.607381059352038e-05,
      "loss": 0.2655,
      "step": 120
    },
    {
      "epoch": 2.151111111111111,
      "grad_norm": 1.0778076648712158,
      "learning_rate": 9.595185440333103e-05,
      "loss": 0.3282,
      "step": 121
    },
    {
      "epoch": 2.168888888888889,
      "grad_norm": 0.48110413551330566,
      "learning_rate": 9.582811279349882e-05,
      "loss": 0.3087,
      "step": 122
    },
    {
      "epoch": 2.1866666666666665,
      "grad_norm": 0.28637051582336426,
      "learning_rate": 9.570259057189717e-05,
      "loss": 0.2924,
      "step": 123
    },
    {
      "epoch": 2.2044444444444444,
      "grad_norm": 0.27911806106567383,
      "learning_rate": 9.557529261558367e-05,
      "loss": 0.2613,
      "step": 124
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.34397801756858826,
      "learning_rate": 9.544622387061055e-05,
      "loss": 0.2574,
      "step": 125
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.2947402000427246,
      "learning_rate": 9.53153893518325e-05,
      "loss": 0.2883,
      "step": 126
    },
    {
      "epoch": 2.2577777777777777,
      "grad_norm": 0.39530470967292786,
      "learning_rate": 9.518279414271183e-05,
      "loss": 0.3131,
      "step": 127
    },
    {
      "epoch": 2.2755555555555556,
      "grad_norm": 0.24501116573810577,
      "learning_rate": 9.504844339512095e-05,
      "loss": 0.3184,
      "step": 128
    },
    {
      "epoch": 2.2933333333333334,
      "grad_norm": 0.25899410247802734,
      "learning_rate": 9.491234232914221e-05,
      "loss": 0.2794,
      "step": 129
    },
    {
      "epoch": 2.311111111111111,
      "grad_norm": 0.41601303219795227,
      "learning_rate": 9.477449623286505e-05,
      "loss": 0.2884,
      "step": 130
    },
    {
      "epoch": 2.328888888888889,
      "grad_norm": 0.3597092628479004,
      "learning_rate": 9.463491046218058e-05,
      "loss": 0.2655,
      "step": 131
    },
    {
      "epoch": 2.3466666666666667,
      "grad_norm": 0.361904501914978,
      "learning_rate": 9.449359044057345e-05,
      "loss": 0.2611,
      "step": 132
    },
    {
      "epoch": 2.3644444444444446,
      "grad_norm": 0.4677833020687103,
      "learning_rate": 9.435054165891109e-05,
      "loss": 0.2589,
      "step": 133
    },
    {
      "epoch": 2.3822222222222225,
      "grad_norm": 0.5052748918533325,
      "learning_rate": 9.420576967523049e-05,
      "loss": 0.3022,
      "step": 134
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.2593027651309967,
      "learning_rate": 9.405928011452211e-05,
      "loss": 0.2498,
      "step": 135
    },
    {
      "epoch": 2.417777777777778,
      "grad_norm": 0.2840777337551117,
      "learning_rate": 9.391107866851143e-05,
      "loss": 0.2636,
      "step": 136
    },
    {
      "epoch": 2.4355555555555557,
      "grad_norm": 0.3062927722930908,
      "learning_rate": 9.376117109543769e-05,
      "loss": 0.25,
      "step": 137
    },
    {
      "epoch": 2.453333333333333,
      "grad_norm": 0.31112948060035706,
      "learning_rate": 9.360956321983028e-05,
      "loss": 0.2435,
      "step": 138
    },
    {
      "epoch": 2.471111111111111,
      "grad_norm": 0.4025621712207794,
      "learning_rate": 9.345626093228233e-05,
      "loss": 0.2427,
      "step": 139
    },
    {
      "epoch": 2.488888888888889,
      "grad_norm": 0.39693212509155273,
      "learning_rate": 9.330127018922194e-05,
      "loss": 0.2897,
      "step": 140
    },
    {
      "epoch": 2.506666666666667,
      "grad_norm": 0.30706965923309326,
      "learning_rate": 9.314459701268065e-05,
      "loss": 0.2833,
      "step": 141
    },
    {
      "epoch": 2.5244444444444447,
      "grad_norm": 0.45809707045555115,
      "learning_rate": 9.298624749005951e-05,
      "loss": 0.2698,
      "step": 142
    },
    {
      "epoch": 2.542222222222222,
      "grad_norm": 0.295108824968338,
      "learning_rate": 9.282622777389258e-05,
      "loss": 0.2372,
      "step": 143
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.35757213830947876,
      "learning_rate": 9.266454408160779e-05,
      "loss": 0.2756,
      "step": 144
    },
    {
      "epoch": 2.5777777777777775,
      "grad_norm": 0.2958184778690338,
      "learning_rate": 9.250120269528546e-05,
      "loss": 0.2661,
      "step": 145
    },
    {
      "epoch": 2.5955555555555554,
      "grad_norm": 0.36897847056388855,
      "learning_rate": 9.233620996141421e-05,
      "loss": 0.2945,
      "step": 146
    },
    {
      "epoch": 2.6133333333333333,
      "grad_norm": 0.2986817955970764,
      "learning_rate": 9.21695722906443e-05,
      "loss": 0.2391,
      "step": 147
    },
    {
      "epoch": 2.631111111111111,
      "grad_norm": 0.28577160835266113,
      "learning_rate": 9.200129615753859e-05,
      "loss": 0.2664,
      "step": 148
    },
    {
      "epoch": 2.648888888888889,
      "grad_norm": 0.363800585269928,
      "learning_rate": 9.183138810032099e-05,
      "loss": 0.2371,
      "step": 149
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.34674468636512756,
      "learning_rate": 9.165985472062246e-05,
      "loss": 0.2883,
      "step": 150
    },
    {
      "epoch": 2.6844444444444444,
      "grad_norm": 0.35801827907562256,
      "learning_rate": 9.148670268322438e-05,
      "loss": 0.2572,
      "step": 151
    },
    {
      "epoch": 2.7022222222222223,
      "grad_norm": 0.24792705476284027,
      "learning_rate": 9.131193871579975e-05,
      "loss": 0.2423,
      "step": 152
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 0.31966882944107056,
      "learning_rate": 9.113556960865167e-05,
      "loss": 0.2453,
      "step": 153
    },
    {
      "epoch": 2.7377777777777776,
      "grad_norm": 0.27944958209991455,
      "learning_rate": 9.09576022144496e-05,
      "loss": 0.2297,
      "step": 154
    },
    {
      "epoch": 2.7555555555555555,
      "grad_norm": 0.34474286437034607,
      "learning_rate": 9.077804344796302e-05,
      "loss": 0.2865,
      "step": 155
    },
    {
      "epoch": 2.7733333333333334,
      "grad_norm": 0.36895522475242615,
      "learning_rate": 9.059690028579284e-05,
      "loss": 0.2567,
      "step": 156
    },
    {
      "epoch": 2.7911111111111113,
      "grad_norm": 0.26620030403137207,
      "learning_rate": 9.041417976610027e-05,
      "loss": 0.2537,
      "step": 157
    },
    {
      "epoch": 2.8088888888888888,
      "grad_norm": 0.24772542715072632,
      "learning_rate": 9.022988898833342e-05,
      "loss": 0.2212,
      "step": 158
    },
    {
      "epoch": 2.8266666666666667,
      "grad_norm": 0.3265734016895294,
      "learning_rate": 9.004403511295141e-05,
      "loss": 0.2514,
      "step": 159
    },
    {
      "epoch": 2.8444444444444446,
      "grad_norm": 0.44243156909942627,
      "learning_rate": 8.985662536114613e-05,
      "loss": 0.2857,
      "step": 160
    },
    {
      "epoch": 2.862222222222222,
      "grad_norm": 0.27827122807502747,
      "learning_rate": 8.966766701456177e-05,
      "loss": 0.2507,
      "step": 161
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.47080013155937195,
      "learning_rate": 8.947716741501177e-05,
      "loss": 0.2875,
      "step": 162
    },
    {
      "epoch": 2.897777777777778,
      "grad_norm": 0.37070050835609436,
      "learning_rate": 8.928513396419368e-05,
      "loss": 0.2601,
      "step": 163
    },
    {
      "epoch": 2.9155555555555557,
      "grad_norm": 0.3937135338783264,
      "learning_rate": 8.90915741234015e-05,
      "loss": 0.2624,
      "step": 164
    },
    {
      "epoch": 2.9333333333333336,
      "grad_norm": 0.3232908844947815,
      "learning_rate": 8.889649541323574e-05,
      "loss": 0.2578,
      "step": 165
    },
    {
      "epoch": 2.951111111111111,
      "grad_norm": 0.29915979504585266,
      "learning_rate": 8.869990541331138e-05,
      "loss": 0.2354,
      "step": 166
    },
    {
      "epoch": 2.968888888888889,
      "grad_norm": 0.4688418209552765,
      "learning_rate": 8.850181176196315e-05,
      "loss": 0.3058,
      "step": 167
    },
    {
      "epoch": 2.986666666666667,
      "grad_norm": 0.21794776618480682,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.2424,
      "step": 168
    },
    {
      "epoch": 3.0044444444444443,
      "grad_norm": 0.20859044790267944,
      "learning_rate": 8.810114435015054e-05,
      "loss": 0.2527,
      "step": 169
    },
    {
      "epoch": 3.022222222222222,
      "grad_norm": 0.329470157623291,
      "learning_rate": 8.789858615727265e-05,
      "loss": 0.2644,
      "step": 170
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.2836751341819763,
      "learning_rate": 8.7694555447539e-05,
      "loss": 0.2419,
      "step": 171
    },
    {
      "epoch": 3.057777777777778,
      "grad_norm": 0.2949230968952179,
      "learning_rate": 8.748906014838672e-05,
      "loss": 0.2201,
      "step": 172
    },
    {
      "epoch": 3.0755555555555554,
      "grad_norm": 0.4008142054080963,
      "learning_rate": 8.728210824415827e-05,
      "loss": 0.2479,
      "step": 173
    },
    {
      "epoch": 3.0933333333333333,
      "grad_norm": 0.25974419713020325,
      "learning_rate": 8.707370777579133e-05,
      "loss": 0.2311,
      "step": 174
    },
    {
      "epoch": 3.111111111111111,
      "grad_norm": 0.2546006143093109,
      "learning_rate": 8.68638668405062e-05,
      "loss": 0.2239,
      "step": 175
    },
    {
      "epoch": 3.128888888888889,
      "grad_norm": 0.5272888541221619,
      "learning_rate": 8.665259359149132e-05,
      "loss": 0.2798,
      "step": 176
    },
    {
      "epoch": 3.1466666666666665,
      "grad_norm": 0.4749833047389984,
      "learning_rate": 8.643989623758643e-05,
      "loss": 0.2376,
      "step": 177
    },
    {
      "epoch": 3.1644444444444444,
      "grad_norm": 0.3421812653541565,
      "learning_rate": 8.622578304296364e-05,
      "loss": 0.2672,
      "step": 178
    },
    {
      "epoch": 3.1822222222222223,
      "grad_norm": 0.3668532073497772,
      "learning_rate": 8.601026232680634e-05,
      "loss": 0.2463,
      "step": 179
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.2934477925300598,
      "learning_rate": 8.579334246298593e-05,
      "loss": 0.2367,
      "step": 180
    },
    {
      "epoch": 3.2177777777777776,
      "grad_norm": 0.26318293809890747,
      "learning_rate": 8.557503187973651e-05,
      "loss": 0.2362,
      "step": 181
    },
    {
      "epoch": 3.2355555555555555,
      "grad_norm": 0.3571169972419739,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.2331,
      "step": 182
    },
    {
      "epoch": 3.2533333333333334,
      "grad_norm": 0.39406391978263855,
      "learning_rate": 8.513427253773346e-05,
      "loss": 0.2775,
      "step": 183
    },
    {
      "epoch": 3.2711111111111113,
      "grad_norm": 0.39618298411369324,
      "learning_rate": 8.491184090430364e-05,
      "loss": 0.2239,
      "step": 184
    },
    {
      "epoch": 3.2888888888888888,
      "grad_norm": 0.30980485677719116,
      "learning_rate": 8.468805280142709e-05,
      "loss": 0.2395,
      "step": 185
    },
    {
      "epoch": 3.3066666666666666,
      "grad_norm": 0.3459194004535675,
      "learning_rate": 8.446291692419736e-05,
      "loss": 0.2222,
      "step": 186
    },
    {
      "epoch": 3.3244444444444445,
      "grad_norm": 0.2490205019712448,
      "learning_rate": 8.423644202007467e-05,
      "loss": 0.2151,
      "step": 187
    },
    {
      "epoch": 3.3422222222222224,
      "grad_norm": 0.4106178879737854,
      "learning_rate": 8.400863688854597e-05,
      "loss": 0.2501,
      "step": 188
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.5236695408821106,
      "learning_rate": 8.377951038078302e-05,
      "loss": 0.2801,
      "step": 189
    },
    {
      "epoch": 3.3777777777777778,
      "grad_norm": 0.39424973726272583,
      "learning_rate": 8.354907139929851e-05,
      "loss": 0.2216,
      "step": 190
    },
    {
      "epoch": 3.3955555555555557,
      "grad_norm": 0.38971391320228577,
      "learning_rate": 8.33173288976002e-05,
      "loss": 0.228,
      "step": 191
    },
    {
      "epoch": 3.413333333333333,
      "grad_norm": 0.48786965012550354,
      "learning_rate": 8.308429187984297e-05,
      "loss": 0.2521,
      "step": 192
    },
    {
      "epoch": 3.431111111111111,
      "grad_norm": 0.37168943881988525,
      "learning_rate": 8.284996940047903e-05,
      "loss": 0.2397,
      "step": 193
    },
    {
      "epoch": 3.448888888888889,
      "grad_norm": 0.36831504106521606,
      "learning_rate": 8.261437056390606e-05,
      "loss": 0.2204,
      "step": 194
    },
    {
      "epoch": 3.466666666666667,
      "grad_norm": 0.32543519139289856,
      "learning_rate": 8.237750452411353e-05,
      "loss": 0.2291,
      "step": 195
    },
    {
      "epoch": 3.4844444444444447,
      "grad_norm": 0.3227090835571289,
      "learning_rate": 8.213938048432697e-05,
      "loss": 0.2218,
      "step": 196
    },
    {
      "epoch": 3.502222222222222,
      "grad_norm": 0.623389720916748,
      "learning_rate": 8.190000769665044e-05,
      "loss": 0.249,
      "step": 197
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.35425952076911926,
      "learning_rate": 8.1659395461707e-05,
      "loss": 0.2749,
      "step": 198
    },
    {
      "epoch": 3.537777777777778,
      "grad_norm": 0.33658289909362793,
      "learning_rate": 8.141755312827736e-05,
      "loss": 0.2134,
      "step": 199
    },
    {
      "epoch": 3.5555555555555554,
      "grad_norm": 0.34356260299682617,
      "learning_rate": 8.117449009293668e-05,
      "loss": 0.2216,
      "step": 200
    },
    {
      "epoch": 3.5733333333333333,
      "grad_norm": 0.32880356907844543,
      "learning_rate": 8.093021579968941e-05,
      "loss": 0.2369,
      "step": 201
    },
    {
      "epoch": 3.591111111111111,
      "grad_norm": 0.47223803400993347,
      "learning_rate": 8.068473973960238e-05,
      "loss": 0.2433,
      "step": 202
    },
    {
      "epoch": 3.608888888888889,
      "grad_norm": 0.37927350401878357,
      "learning_rate": 8.043807145043604e-05,
      "loss": 0.2405,
      "step": 203
    },
    {
      "epoch": 3.626666666666667,
      "grad_norm": 0.28793418407440186,
      "learning_rate": 8.019022051627388e-05,
      "loss": 0.2345,
      "step": 204
    },
    {
      "epoch": 3.6444444444444444,
      "grad_norm": 0.3489110767841339,
      "learning_rate": 7.994119656715002e-05,
      "loss": 0.2574,
      "step": 205
    },
    {
      "epoch": 3.6622222222222223,
      "grad_norm": 0.3693983554840088,
      "learning_rate": 7.969100927867507e-05,
      "loss": 0.2445,
      "step": 206
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.3646757900714874,
      "learning_rate": 7.943966837166023e-05,
      "loss": 0.2099,
      "step": 207
    },
    {
      "epoch": 3.6977777777777776,
      "grad_norm": 0.31038224697113037,
      "learning_rate": 7.91871836117395e-05,
      "loss": 0.2051,
      "step": 208
    },
    {
      "epoch": 3.7155555555555555,
      "grad_norm": 0.2764589786529541,
      "learning_rate": 7.89335648089903e-05,
      "loss": 0.2132,
      "step": 209
    },
    {
      "epoch": 3.7333333333333334,
      "grad_norm": 0.3864257037639618,
      "learning_rate": 7.86788218175523e-05,
      "loss": 0.2284,
      "step": 210
    },
    {
      "epoch": 3.7511111111111113,
      "grad_norm": 0.8009142279624939,
      "learning_rate": 7.842296453524463e-05,
      "loss": 0.3221,
      "step": 211
    },
    {
      "epoch": 3.7688888888888887,
      "grad_norm": 0.3626723885536194,
      "learning_rate": 7.81660029031811e-05,
      "loss": 0.2558,
      "step": 212
    },
    {
      "epoch": 3.7866666666666666,
      "grad_norm": 0.4381621778011322,
      "learning_rate": 7.79079469053842e-05,
      "loss": 0.228,
      "step": 213
    },
    {
      "epoch": 3.8044444444444445,
      "grad_norm": 0.38796478509902954,
      "learning_rate": 7.764880656839696e-05,
      "loss": 0.2025,
      "step": 214
    },
    {
      "epoch": 3.822222222222222,
      "grad_norm": 0.27901729941368103,
      "learning_rate": 7.738859196089358e-05,
      "loss": 0.2142,
      "step": 215
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.4677048623561859,
      "learning_rate": 7.712731319328798e-05,
      "loss": 0.2614,
      "step": 216
    },
    {
      "epoch": 3.8577777777777778,
      "grad_norm": 0.2969120740890503,
      "learning_rate": 7.68649804173412e-05,
      "loss": 0.24,
      "step": 217
    },
    {
      "epoch": 3.8755555555555556,
      "grad_norm": 0.42577284574508667,
      "learning_rate": 7.660160382576683e-05,
      "loss": 0.2054,
      "step": 218
    },
    {
      "epoch": 3.8933333333333335,
      "grad_norm": 0.40695130825042725,
      "learning_rate": 7.633719365183504e-05,
      "loss": 0.2464,
      "step": 219
    },
    {
      "epoch": 3.911111111111111,
      "grad_norm": 0.5424155592918396,
      "learning_rate": 7.60717601689749e-05,
      "loss": 0.2583,
      "step": 220
    },
    {
      "epoch": 3.928888888888889,
      "grad_norm": 0.683966338634491,
      "learning_rate": 7.580531369037533e-05,
      "loss": 0.2387,
      "step": 221
    },
    {
      "epoch": 3.9466666666666668,
      "grad_norm": 0.37036415934562683,
      "learning_rate": 7.553786456858429e-05,
      "loss": 0.2333,
      "step": 222
    },
    {
      "epoch": 3.964444444444444,
      "grad_norm": 0.5287876129150391,
      "learning_rate": 7.526942319510655e-05,
      "loss": 0.2353,
      "step": 223
    },
    {
      "epoch": 3.982222222222222,
      "grad_norm": 0.31925442814826965,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.2411,
      "step": 224
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5316102504730225,
      "learning_rate": 7.472960545147038e-05,
      "loss": 0.2472,
      "step": 225
    },
    {
      "epoch": 4.017777777777778,
      "grad_norm": 0.24691040813922882,
      "learning_rate": 7.445825005546448e-05,
      "loss": 0.217,
      "step": 226
    },
    {
      "epoch": 4.035555555555556,
      "grad_norm": 0.3662887215614319,
      "learning_rate": 7.4185944355262e-05,
      "loss": 0.2026,
      "step": 227
    },
    {
      "epoch": 4.053333333333334,
      "grad_norm": 0.36757993698120117,
      "learning_rate": 7.391269893106592e-05,
      "loss": 0.2364,
      "step": 228
    },
    {
      "epoch": 4.071111111111111,
      "grad_norm": 0.29429471492767334,
      "learning_rate": 7.363852439959135e-05,
      "loss": 0.1774,
      "step": 229
    },
    {
      "epoch": 4.088888888888889,
      "grad_norm": 0.48565641045570374,
      "learning_rate": 7.33634314136531e-05,
      "loss": 0.2156,
      "step": 230
    },
    {
      "epoch": 4.1066666666666665,
      "grad_norm": 0.3768898844718933,
      "learning_rate": 7.308743066175172e-05,
      "loss": 0.2006,
      "step": 231
    },
    {
      "epoch": 4.124444444444444,
      "grad_norm": 0.32797563076019287,
      "learning_rate": 7.281053286765815e-05,
      "loss": 0.2195,
      "step": 232
    },
    {
      "epoch": 4.142222222222222,
      "grad_norm": 0.3106316030025482,
      "learning_rate": 7.253274878999727e-05,
      "loss": 0.2268,
      "step": 233
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.43699216842651367,
      "learning_rate": 7.225408922182961e-05,
      "loss": 0.2186,
      "step": 234
    },
    {
      "epoch": 4.177777777777778,
      "grad_norm": 0.569455623626709,
      "learning_rate": 7.197456499023225e-05,
      "loss": 0.2039,
      "step": 235
    },
    {
      "epoch": 4.195555555555556,
      "grad_norm": 0.49764353036880493,
      "learning_rate": 7.169418695587791e-05,
      "loss": 0.2107,
      "step": 236
    },
    {
      "epoch": 4.213333333333333,
      "grad_norm": 0.5107737183570862,
      "learning_rate": 7.141296601261314e-05,
      "loss": 0.2387,
      "step": 237
    },
    {
      "epoch": 4.231111111111111,
      "grad_norm": 0.531856119632721,
      "learning_rate": 7.113091308703498e-05,
      "loss": 0.1914,
      "step": 238
    },
    {
      "epoch": 4.248888888888889,
      "grad_norm": 0.3557555377483368,
      "learning_rate": 7.084803913806641e-05,
      "loss": 0.2192,
      "step": 239
    },
    {
      "epoch": 4.266666666666667,
      "grad_norm": 0.520222544670105,
      "learning_rate": 7.056435515653059e-05,
      "loss": 0.2406,
      "step": 240
    },
    {
      "epoch": 4.2844444444444445,
      "grad_norm": 0.5725359916687012,
      "learning_rate": 7.027987216472377e-05,
      "loss": 0.2102,
      "step": 241
    },
    {
      "epoch": 4.302222222222222,
      "grad_norm": 0.39855217933654785,
      "learning_rate": 6.999460121598704e-05,
      "loss": 0.1943,
      "step": 242
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.47654294967651367,
      "learning_rate": 6.970855339427698e-05,
      "loss": 0.1955,
      "step": 243
    },
    {
      "epoch": 4.337777777777778,
      "grad_norm": 0.6176052093505859,
      "learning_rate": 6.942173981373474e-05,
      "loss": 0.2316,
      "step": 244
    },
    {
      "epoch": 4.355555555555555,
      "grad_norm": 0.446646511554718,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.2038,
      "step": 245
    },
    {
      "epoch": 4.373333333333333,
      "grad_norm": 0.4111061692237854,
      "learning_rate": 6.884585998105026e-05,
      "loss": 0.2137,
      "step": 246
    },
    {
      "epoch": 4.391111111111111,
      "grad_norm": 0.5625485777854919,
      "learning_rate": 6.855681610422189e-05,
      "loss": 0.2059,
      "step": 247
    },
    {
      "epoch": 4.408888888888889,
      "grad_norm": 0.30589085817337036,
      "learning_rate": 6.826705121831976e-05,
      "loss": 0.2213,
      "step": 248
    },
    {
      "epoch": 4.426666666666667,
      "grad_norm": 0.766379177570343,
      "learning_rate": 6.797657658190839e-05,
      "loss": 0.2061,
      "step": 249
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.3027375340461731,
      "learning_rate": 6.768540348112907e-05,
      "loss": 0.2185,
      "step": 250
    },
    {
      "epoch": 4.4622222222222225,
      "grad_norm": 0.49446240067481995,
      "learning_rate": 6.739354322926136e-05,
      "loss": 0.1946,
      "step": 251
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.5550069212913513,
      "learning_rate": 6.710100716628344e-05,
      "loss": 0.2716,
      "step": 252
    },
    {
      "epoch": 4.497777777777777,
      "grad_norm": 0.4387371838092804,
      "learning_rate": 6.680780665843155e-05,
      "loss": 0.1893,
      "step": 253
    },
    {
      "epoch": 4.515555555555555,
      "grad_norm": 0.3026201128959656,
      "learning_rate": 6.651395309775837e-05,
      "loss": 0.216,
      "step": 254
    },
    {
      "epoch": 4.533333333333333,
      "grad_norm": 0.4486079216003418,
      "learning_rate": 6.621945790169036e-05,
      "loss": 0.1849,
      "step": 255
    },
    {
      "epoch": 4.551111111111111,
      "grad_norm": 0.4118870496749878,
      "learning_rate": 6.592433251258423e-05,
      "loss": 0.1977,
      "step": 256
    },
    {
      "epoch": 4.568888888888889,
      "grad_norm": 0.5137110352516174,
      "learning_rate": 6.562858839728223e-05,
      "loss": 0.2167,
      "step": 257
    },
    {
      "epoch": 4.586666666666667,
      "grad_norm": 0.34046244621276855,
      "learning_rate": 6.533223704666672e-05,
      "loss": 0.2128,
      "step": 258
    },
    {
      "epoch": 4.604444444444445,
      "grad_norm": 0.44516146183013916,
      "learning_rate": 6.503528997521366e-05,
      "loss": 0.1932,
      "step": 259
    },
    {
      "epoch": 4.622222222222222,
      "grad_norm": 0.2885585427284241,
      "learning_rate": 6.473775872054521e-05,
      "loss": 0.1991,
      "step": 260
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.37257564067840576,
      "learning_rate": 6.44396548429815e-05,
      "loss": 0.179,
      "step": 261
    },
    {
      "epoch": 4.657777777777778,
      "grad_norm": 0.43277639150619507,
      "learning_rate": 6.414098992509138e-05,
      "loss": 0.2154,
      "step": 262
    },
    {
      "epoch": 4.6755555555555555,
      "grad_norm": 0.4736541509628296,
      "learning_rate": 6.384177557124247e-05,
      "loss": 0.2024,
      "step": 263
    },
    {
      "epoch": 4.693333333333333,
      "grad_norm": 0.3058260381221771,
      "learning_rate": 6.354202340715026e-05,
      "loss": 0.2264,
      "step": 264
    },
    {
      "epoch": 4.711111111111111,
      "grad_norm": 0.33973225951194763,
      "learning_rate": 6.324174507942637e-05,
      "loss": 0.2175,
      "step": 265
    },
    {
      "epoch": 4.728888888888889,
      "grad_norm": 0.370286762714386,
      "learning_rate": 6.294095225512603e-05,
      "loss": 0.1969,
      "step": 266
    },
    {
      "epoch": 4.746666666666667,
      "grad_norm": 0.5065796375274658,
      "learning_rate": 6.263965662129487e-05,
      "loss": 0.2322,
      "step": 267
    },
    {
      "epoch": 4.764444444444445,
      "grad_norm": 0.2974378764629364,
      "learning_rate": 6.233786988451468e-05,
      "loss": 0.1617,
      "step": 268
    },
    {
      "epoch": 4.782222222222222,
      "grad_norm": 0.37753137946128845,
      "learning_rate": 6.203560377044866e-05,
      "loss": 0.2324,
      "step": 269
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.5713666677474976,
      "learning_rate": 6.173287002338577e-05,
      "loss": 0.2213,
      "step": 270
    },
    {
      "epoch": 4.817777777777778,
      "grad_norm": 0.5500757694244385,
      "learning_rate": 6.142968040578449e-05,
      "loss": 0.2163,
      "step": 271
    },
    {
      "epoch": 4.835555555555556,
      "grad_norm": 0.3870888948440552,
      "learning_rate": 6.112604669781572e-05,
      "loss": 0.2014,
      "step": 272
    },
    {
      "epoch": 4.8533333333333335,
      "grad_norm": 0.4181479811668396,
      "learning_rate": 6.0821980696905146e-05,
      "loss": 0.2123,
      "step": 273
    },
    {
      "epoch": 4.871111111111111,
      "grad_norm": 0.3824799656867981,
      "learning_rate": 6.0517494217274794e-05,
      "loss": 0.2008,
      "step": 274
    },
    {
      "epoch": 4.888888888888889,
      "grad_norm": 0.3899323642253876,
      "learning_rate": 6.021259908948402e-05,
      "loss": 0.2011,
      "step": 275
    },
    {
      "epoch": 4.906666666666666,
      "grad_norm": 0.4690476059913635,
      "learning_rate": 5.9907307159969884e-05,
      "loss": 0.2215,
      "step": 276
    },
    {
      "epoch": 4.924444444444444,
      "grad_norm": 0.3852761387825012,
      "learning_rate": 5.960163029058682e-05,
      "loss": 0.1943,
      "step": 277
    },
    {
      "epoch": 4.942222222222222,
      "grad_norm": 0.40581774711608887,
      "learning_rate": 5.9295580358145744e-05,
      "loss": 0.2132,
      "step": 278
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.7496148943901062,
      "learning_rate": 5.898916925395264e-05,
      "loss": 0.2227,
      "step": 279
    },
    {
      "epoch": 4.977777777777778,
      "grad_norm": 0.3892797529697418,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.2165,
      "step": 280
    },
    {
      "epoch": 4.995555555555556,
      "grad_norm": 0.284811794757843,
      "learning_rate": 5.837531116523682e-05,
      "loss": 0.1669,
      "step": 281
    },
    {
      "epoch": 5.013333333333334,
      "grad_norm": 0.416891872882843,
      "learning_rate": 5.806788803164034e-05,
      "loss": 0.1828,
      "step": 282
    },
    {
      "epoch": 5.0311111111111115,
      "grad_norm": 0.4431915283203125,
      "learning_rate": 5.7760151427217576e-05,
      "loss": 0.1957,
      "step": 283
    },
    {
      "epoch": 5.0488888888888885,
      "grad_norm": 0.5158615708351135,
      "learning_rate": 5.745211330880872e-05,
      "loss": 0.2331,
      "step": 284
    },
    {
      "epoch": 5.066666666666666,
      "grad_norm": 0.3418184220790863,
      "learning_rate": 5.714378564496901e-05,
      "loss": 0.2065,
      "step": 285
    },
    {
      "epoch": 5.084444444444444,
      "grad_norm": 0.5028918981552124,
      "learning_rate": 5.683518041550368e-05,
      "loss": 0.1657,
      "step": 286
    },
    {
      "epoch": 5.102222222222222,
      "grad_norm": 0.4428432285785675,
      "learning_rate": 5.6526309611002594e-05,
      "loss": 0.1948,
      "step": 287
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.3536894917488098,
      "learning_rate": 5.621718523237427e-05,
      "loss": 0.1709,
      "step": 288
    },
    {
      "epoch": 5.137777777777778,
      "grad_norm": 0.4637707769870758,
      "learning_rate": 5.590781929037965e-05,
      "loss": 0.1662,
      "step": 289
    },
    {
      "epoch": 5.155555555555556,
      "grad_norm": 0.5988882780075073,
      "learning_rate": 5.559822380516539e-05,
      "loss": 0.1477,
      "step": 290
    },
    {
      "epoch": 5.173333333333334,
      "grad_norm": 0.45593711733818054,
      "learning_rate": 5.5288410805796895e-05,
      "loss": 0.1802,
      "step": 291
    },
    {
      "epoch": 5.191111111111111,
      "grad_norm": 0.46239393949508667,
      "learning_rate": 5.497839232979084e-05,
      "loss": 0.178,
      "step": 292
    },
    {
      "epoch": 5.208888888888889,
      "grad_norm": 0.5151268839836121,
      "learning_rate": 5.466818042264753e-05,
      "loss": 0.1453,
      "step": 293
    },
    {
      "epoch": 5.226666666666667,
      "grad_norm": 0.5723825097084045,
      "learning_rate": 5.435778713738292e-05,
      "loss": 0.212,
      "step": 294
    },
    {
      "epoch": 5.2444444444444445,
      "grad_norm": 0.5398524403572083,
      "learning_rate": 5.404722453406017e-05,
      "loss": 0.1773,
      "step": 295
    },
    {
      "epoch": 5.262222222222222,
      "grad_norm": 0.5362580418586731,
      "learning_rate": 5.373650467932122e-05,
      "loss": 0.1816,
      "step": 296
    },
    {
      "epoch": 5.28,
      "grad_norm": 0.5614521503448486,
      "learning_rate": 5.3425639645917834e-05,
      "loss": 0.1544,
      "step": 297
    },
    {
      "epoch": 5.297777777777778,
      "grad_norm": 0.44896313548088074,
      "learning_rate": 5.311464151224261e-05,
      "loss": 0.1381,
      "step": 298
    },
    {
      "epoch": 5.315555555555555,
      "grad_norm": 0.35840409994125366,
      "learning_rate": 5.2803522361859594e-05,
      "loss": 0.1653,
      "step": 299
    },
    {
      "epoch": 5.333333333333333,
      "grad_norm": 0.6827304363250732,
      "learning_rate": 5.249229428303486e-05,
      "loss": 0.167,
      "step": 300
    },
    {
      "epoch": 5.351111111111111,
      "grad_norm": 0.3499818444252014,
      "learning_rate": 5.218096936826681e-05,
      "loss": 0.2221,
      "step": 301
    },
    {
      "epoch": 5.368888888888889,
      "grad_norm": 0.48628711700439453,
      "learning_rate": 5.18695597138163e-05,
      "loss": 0.1932,
      "step": 302
    },
    {
      "epoch": 5.386666666666667,
      "grad_norm": 0.5208820104598999,
      "learning_rate": 5.155807741923666e-05,
      "loss": 0.1818,
      "step": 303
    },
    {
      "epoch": 5.404444444444445,
      "grad_norm": 0.4346122741699219,
      "learning_rate": 5.124653458690365e-05,
      "loss": 0.1924,
      "step": 304
    },
    {
      "epoch": 5.4222222222222225,
      "grad_norm": 0.45820167660713196,
      "learning_rate": 5.0934943321545115e-05,
      "loss": 0.1922,
      "step": 305
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.5644346475601196,
      "learning_rate": 5.062331572977076e-05,
      "loss": 0.1828,
      "step": 306
    },
    {
      "epoch": 5.457777777777777,
      "grad_norm": 0.507893979549408,
      "learning_rate": 5.031166391960168e-05,
      "loss": 0.2173,
      "step": 307
    },
    {
      "epoch": 5.475555555555555,
      "grad_norm": 0.4621982276439667,
      "learning_rate": 5e-05,
      "loss": 0.1703,
      "step": 308
    },
    {
      "epoch": 5.493333333333333,
      "grad_norm": 0.4117422103881836,
      "learning_rate": 4.968833608039832e-05,
      "loss": 0.1751,
      "step": 309
    },
    {
      "epoch": 5.511111111111111,
      "grad_norm": 0.4004817605018616,
      "learning_rate": 4.9376684270229254e-05,
      "loss": 0.1835,
      "step": 310
    },
    {
      "epoch": 5.528888888888889,
      "grad_norm": 0.4063107669353485,
      "learning_rate": 4.9065056678454904e-05,
      "loss": 0.1892,
      "step": 311
    },
    {
      "epoch": 5.546666666666667,
      "grad_norm": 0.45984384417533875,
      "learning_rate": 4.875346541309637e-05,
      "loss": 0.181,
      "step": 312
    },
    {
      "epoch": 5.564444444444445,
      "grad_norm": 0.48440927267074585,
      "learning_rate": 4.844192258076336e-05,
      "loss": 0.1723,
      "step": 313
    },
    {
      "epoch": 5.582222222222223,
      "grad_norm": 0.6084842681884766,
      "learning_rate": 4.813044028618373e-05,
      "loss": 0.1875,
      "step": 314
    },
    {
      "epoch": 5.6,
      "grad_norm": 0.5340481996536255,
      "learning_rate": 4.781903063173321e-05,
      "loss": 0.1903,
      "step": 315
    },
    {
      "epoch": 5.6177777777777775,
      "grad_norm": 0.3680734634399414,
      "learning_rate": 4.750770571696514e-05,
      "loss": 0.1742,
      "step": 316
    },
    {
      "epoch": 5.635555555555555,
      "grad_norm": 0.4389323592185974,
      "learning_rate": 4.7196477638140404e-05,
      "loss": 0.1604,
      "step": 317
    },
    {
      "epoch": 5.653333333333333,
      "grad_norm": 0.5682416558265686,
      "learning_rate": 4.68853584877574e-05,
      "loss": 0.2142,
      "step": 318
    },
    {
      "epoch": 5.671111111111111,
      "grad_norm": 0.5514141321182251,
      "learning_rate": 4.657436035408217e-05,
      "loss": 0.1902,
      "step": 319
    },
    {
      "epoch": 5.688888888888889,
      "grad_norm": 0.5815438032150269,
      "learning_rate": 4.626349532067879e-05,
      "loss": 0.1604,
      "step": 320
    },
    {
      "epoch": 5.706666666666667,
      "grad_norm": 0.3311176598072052,
      "learning_rate": 4.595277546593984e-05,
      "loss": 0.2049,
      "step": 321
    },
    {
      "epoch": 5.724444444444444,
      "grad_norm": 0.4205529987812042,
      "learning_rate": 4.564221286261709e-05,
      "loss": 0.1715,
      "step": 322
    },
    {
      "epoch": 5.742222222222222,
      "grad_norm": 0.6624743342399597,
      "learning_rate": 4.5331819577352474e-05,
      "loss": 0.1601,
      "step": 323
    },
    {
      "epoch": 5.76,
      "grad_norm": 0.5203994512557983,
      "learning_rate": 4.502160767020918e-05,
      "loss": 0.1589,
      "step": 324
    },
    {
      "epoch": 5.777777777777778,
      "grad_norm": 0.4620148241519928,
      "learning_rate": 4.471158919420312e-05,
      "loss": 0.1656,
      "step": 325
    },
    {
      "epoch": 5.795555555555556,
      "grad_norm": 0.5596708655357361,
      "learning_rate": 4.4401776194834613e-05,
      "loss": 0.1821,
      "step": 326
    },
    {
      "epoch": 5.8133333333333335,
      "grad_norm": 0.4360620975494385,
      "learning_rate": 4.409218070962036e-05,
      "loss": 0.1752,
      "step": 327
    },
    {
      "epoch": 5.831111111111111,
      "grad_norm": 0.4674508571624756,
      "learning_rate": 4.378281476762576e-05,
      "loss": 0.204,
      "step": 328
    },
    {
      "epoch": 5.848888888888889,
      "grad_norm": 0.9109951853752136,
      "learning_rate": 4.347369038899744e-05,
      "loss": 0.1562,
      "step": 329
    },
    {
      "epoch": 5.866666666666667,
      "grad_norm": 0.8667967915534973,
      "learning_rate": 4.316481958449634e-05,
      "loss": 0.2376,
      "step": 330
    },
    {
      "epoch": 5.884444444444444,
      "grad_norm": 0.6173791289329529,
      "learning_rate": 4.285621435503101e-05,
      "loss": 0.1699,
      "step": 331
    },
    {
      "epoch": 5.902222222222222,
      "grad_norm": 0.6599533557891846,
      "learning_rate": 4.254788669119127e-05,
      "loss": 0.1652,
      "step": 332
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.39456817507743835,
      "learning_rate": 4.223984857278242e-05,
      "loss": 0.1695,
      "step": 333
    },
    {
      "epoch": 5.937777777777778,
      "grad_norm": 0.7024866342544556,
      "learning_rate": 4.1932111968359664e-05,
      "loss": 0.1673,
      "step": 334
    },
    {
      "epoch": 5.955555555555556,
      "grad_norm": 0.5477570295333862,
      "learning_rate": 4.162468883476319e-05,
      "loss": 0.1852,
      "step": 335
    },
    {
      "epoch": 5.973333333333334,
      "grad_norm": 0.5014763474464417,
      "learning_rate": 4.131759111665349e-05,
      "loss": 0.1847,
      "step": 336
    },
    {
      "epoch": 5.9911111111111115,
      "grad_norm": 0.334262877702713,
      "learning_rate": 4.101083074604737e-05,
      "loss": 0.1709,
      "step": 337
    },
    {
      "epoch": 6.0088888888888885,
      "grad_norm": 0.5552971959114075,
      "learning_rate": 4.0704419641854274e-05,
      "loss": 0.1476,
      "step": 338
    },
    {
      "epoch": 6.026666666666666,
      "grad_norm": 0.35900405049324036,
      "learning_rate": 4.03983697094132e-05,
      "loss": 0.1731,
      "step": 339
    },
    {
      "epoch": 6.044444444444444,
      "grad_norm": 0.3748185634613037,
      "learning_rate": 4.0092692840030134e-05,
      "loss": 0.1625,
      "step": 340
    },
    {
      "epoch": 6.062222222222222,
      "grad_norm": 0.5626404285430908,
      "learning_rate": 3.978740091051599e-05,
      "loss": 0.1718,
      "step": 341
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.5533756017684937,
      "learning_rate": 3.9482505782725224e-05,
      "loss": 0.1793,
      "step": 342
    },
    {
      "epoch": 6.097777777777778,
      "grad_norm": 0.45061609148979187,
      "learning_rate": 3.917801930309486e-05,
      "loss": 0.1675,
      "step": 343
    },
    {
      "epoch": 6.115555555555556,
      "grad_norm": 0.6766936182975769,
      "learning_rate": 3.887395330218429e-05,
      "loss": 0.1528,
      "step": 344
    },
    {
      "epoch": 6.133333333333334,
      "grad_norm": 0.4249931573867798,
      "learning_rate": 3.857031959421553e-05,
      "loss": 0.1437,
      "step": 345
    },
    {
      "epoch": 6.151111111111111,
      "grad_norm": 0.4025160074234009,
      "learning_rate": 3.8267129976614254e-05,
      "loss": 0.1647,
      "step": 346
    },
    {
      "epoch": 6.168888888888889,
      "grad_norm": 0.42490458488464355,
      "learning_rate": 3.7964396229551364e-05,
      "loss": 0.1436,
      "step": 347
    },
    {
      "epoch": 6.1866666666666665,
      "grad_norm": 0.6905245184898376,
      "learning_rate": 3.7662130115485314e-05,
      "loss": 0.148,
      "step": 348
    },
    {
      "epoch": 6.204444444444444,
      "grad_norm": 0.608650267124176,
      "learning_rate": 3.7360343378705124e-05,
      "loss": 0.1413,
      "step": 349
    },
    {
      "epoch": 6.222222222222222,
      "grad_norm": 0.560066819190979,
      "learning_rate": 3.705904774487396e-05,
      "loss": 0.1528,
      "step": 350
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.5189685821533203,
      "learning_rate": 3.675825492057364e-05,
      "loss": 0.1639,
      "step": 351
    },
    {
      "epoch": 6.257777777777778,
      "grad_norm": 0.92945796251297,
      "learning_rate": 3.6457976592849754e-05,
      "loss": 0.2024,
      "step": 352
    },
    {
      "epoch": 6.275555555555556,
      "grad_norm": 0.544842541217804,
      "learning_rate": 3.6158224428757535e-05,
      "loss": 0.1443,
      "step": 353
    },
    {
      "epoch": 6.293333333333333,
      "grad_norm": 0.4329567849636078,
      "learning_rate": 3.585901007490863e-05,
      "loss": 0.1314,
      "step": 354
    },
    {
      "epoch": 6.311111111111111,
      "grad_norm": 0.5567479729652405,
      "learning_rate": 3.556034515701852e-05,
      "loss": 0.1495,
      "step": 355
    },
    {
      "epoch": 6.328888888888889,
      "grad_norm": 0.45176929235458374,
      "learning_rate": 3.5262241279454785e-05,
      "loss": 0.1571,
      "step": 356
    },
    {
      "epoch": 6.346666666666667,
      "grad_norm": 0.4304347634315491,
      "learning_rate": 3.4964710024786354e-05,
      "loss": 0.1413,
      "step": 357
    },
    {
      "epoch": 6.364444444444445,
      "grad_norm": 0.5478505492210388,
      "learning_rate": 3.4667762953333295e-05,
      "loss": 0.1643,
      "step": 358
    },
    {
      "epoch": 6.3822222222222225,
      "grad_norm": 0.5922195911407471,
      "learning_rate": 3.4371411602717784e-05,
      "loss": 0.1716,
      "step": 359
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.4263589084148407,
      "learning_rate": 3.4075667487415785e-05,
      "loss": 0.1671,
      "step": 360
    },
    {
      "epoch": 6.417777777777777,
      "grad_norm": 0.44783204793930054,
      "learning_rate": 3.3780542098309654e-05,
      "loss": 0.167,
      "step": 361
    },
    {
      "epoch": 6.435555555555555,
      "grad_norm": 0.5761044025421143,
      "learning_rate": 3.3486046902241664e-05,
      "loss": 0.1518,
      "step": 362
    },
    {
      "epoch": 6.453333333333333,
      "grad_norm": 0.4984096884727478,
      "learning_rate": 3.319219334156847e-05,
      "loss": 0.1888,
      "step": 363
    },
    {
      "epoch": 6.471111111111111,
      "grad_norm": 0.5709213614463806,
      "learning_rate": 3.289899283371657e-05,
      "loss": 0.1374,
      "step": 364
    },
    {
      "epoch": 6.488888888888889,
      "grad_norm": 0.45197322964668274,
      "learning_rate": 3.2606456770738636e-05,
      "loss": 0.1823,
      "step": 365
    },
    {
      "epoch": 6.506666666666667,
      "grad_norm": 0.3853418231010437,
      "learning_rate": 3.231459651887093e-05,
      "loss": 0.1424,
      "step": 366
    },
    {
      "epoch": 6.524444444444445,
      "grad_norm": 0.6892653107643127,
      "learning_rate": 3.2023423418091626e-05,
      "loss": 0.1354,
      "step": 367
    },
    {
      "epoch": 6.542222222222223,
      "grad_norm": 0.8659518957138062,
      "learning_rate": 3.173294878168025e-05,
      "loss": 0.1658,
      "step": 368
    },
    {
      "epoch": 6.5600000000000005,
      "grad_norm": 0.6673440933227539,
      "learning_rate": 3.1443183895778105e-05,
      "loss": 0.1341,
      "step": 369
    },
    {
      "epoch": 6.5777777777777775,
      "grad_norm": 0.49865254759788513,
      "learning_rate": 3.115414001894974e-05,
      "loss": 0.1746,
      "step": 370
    },
    {
      "epoch": 6.595555555555555,
      "grad_norm": 0.5364950299263,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.1655,
      "step": 371
    },
    {
      "epoch": 6.613333333333333,
      "grad_norm": 0.6427706480026245,
      "learning_rate": 3.0578260186265265e-05,
      "loss": 0.1489,
      "step": 372
    },
    {
      "epoch": 6.631111111111111,
      "grad_norm": 0.5947421193122864,
      "learning_rate": 3.029144660572304e-05,
      "loss": 0.1327,
      "step": 373
    },
    {
      "epoch": 6.648888888888889,
      "grad_norm": 0.5604392886161804,
      "learning_rate": 3.000539878401296e-05,
      "loss": 0.1593,
      "step": 374
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.4848502576351166,
      "learning_rate": 2.9720127835276256e-05,
      "loss": 0.1551,
      "step": 375
    },
    {
      "epoch": 6.684444444444445,
      "grad_norm": 0.5442904829978943,
      "learning_rate": 2.9435644843469436e-05,
      "loss": 0.1526,
      "step": 376
    },
    {
      "epoch": 6.702222222222222,
      "grad_norm": 0.4855183959007263,
      "learning_rate": 2.9151960861933614e-05,
      "loss": 0.183,
      "step": 377
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.8890539407730103,
      "learning_rate": 2.886908691296504e-05,
      "loss": 0.1624,
      "step": 378
    },
    {
      "epoch": 6.737777777777778,
      "grad_norm": 0.47631746530532837,
      "learning_rate": 2.858703398738686e-05,
      "loss": 0.1373,
      "step": 379
    },
    {
      "epoch": 6.7555555555555555,
      "grad_norm": 0.46250614523887634,
      "learning_rate": 2.8305813044122097e-05,
      "loss": 0.1786,
      "step": 380
    },
    {
      "epoch": 6.773333333333333,
      "grad_norm": 0.5734685063362122,
      "learning_rate": 2.8025435009767747e-05,
      "loss": 0.1432,
      "step": 381
    },
    {
      "epoch": 6.791111111111111,
      "grad_norm": 0.7250308394432068,
      "learning_rate": 2.774591077817038e-05,
      "loss": 0.1483,
      "step": 382
    },
    {
      "epoch": 6.808888888888889,
      "grad_norm": 0.4598230719566345,
      "learning_rate": 2.746725121000273e-05,
      "loss": 0.1709,
      "step": 383
    },
    {
      "epoch": 6.826666666666666,
      "grad_norm": 0.5403770208358765,
      "learning_rate": 2.718946713234185e-05,
      "loss": 0.1944,
      "step": 384
    },
    {
      "epoch": 6.844444444444444,
      "grad_norm": 0.43435898423194885,
      "learning_rate": 2.6912569338248315e-05,
      "loss": 0.1471,
      "step": 385
    },
    {
      "epoch": 6.862222222222222,
      "grad_norm": 0.7569040656089783,
      "learning_rate": 2.66365685863469e-05,
      "loss": 0.1755,
      "step": 386
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.4488949775695801,
      "learning_rate": 2.636147560040866e-05,
      "loss": 0.1354,
      "step": 387
    },
    {
      "epoch": 6.897777777777778,
      "grad_norm": 0.43409401178359985,
      "learning_rate": 2.6087301068934106e-05,
      "loss": 0.1803,
      "step": 388
    },
    {
      "epoch": 6.915555555555556,
      "grad_norm": 0.38693767786026,
      "learning_rate": 2.581405564473801e-05,
      "loss": 0.1762,
      "step": 389
    },
    {
      "epoch": 6.933333333333334,
      "grad_norm": 0.3770747780799866,
      "learning_rate": 2.5541749944535554e-05,
      "loss": 0.1644,
      "step": 390
    },
    {
      "epoch": 6.9511111111111115,
      "grad_norm": 0.5996742844581604,
      "learning_rate": 2.527039454852963e-05,
      "loss": 0.1719,
      "step": 391
    },
    {
      "epoch": 6.968888888888889,
      "grad_norm": 0.54173743724823,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.1366,
      "step": 392
    },
    {
      "epoch": 6.986666666666666,
      "grad_norm": 0.5965537428855896,
      "learning_rate": 2.473057680489348e-05,
      "loss": 0.155,
      "step": 393
    },
    {
      "epoch": 7.004444444444444,
      "grad_norm": 0.3956046998500824,
      "learning_rate": 2.4462135431415733e-05,
      "loss": 0.1627,
      "step": 394
    },
    {
      "epoch": 7.022222222222222,
      "grad_norm": 0.530720055103302,
      "learning_rate": 2.4194686309624663e-05,
      "loss": 0.1384,
      "step": 395
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.4579175114631653,
      "learning_rate": 2.39282398310251e-05,
      "loss": 0.1731,
      "step": 396
    },
    {
      "epoch": 7.057777777777778,
      "grad_norm": 0.807518720626831,
      "learning_rate": 2.366280634816496e-05,
      "loss": 0.1471,
      "step": 397
    },
    {
      "epoch": 7.075555555555556,
      "grad_norm": 0.5390831232070923,
      "learning_rate": 2.3398396174233178e-05,
      "loss": 0.1066,
      "step": 398
    },
    {
      "epoch": 7.093333333333334,
      "grad_norm": 0.4245457649230957,
      "learning_rate": 2.3135019582658802e-05,
      "loss": 0.1201,
      "step": 399
    },
    {
      "epoch": 7.111111111111111,
      "grad_norm": 0.40093037486076355,
      "learning_rate": 2.2872686806712035e-05,
      "loss": 0.1401,
      "step": 400
    },
    {
      "epoch": 7.128888888888889,
      "grad_norm": 0.8866329789161682,
      "learning_rate": 2.261140803910644e-05,
      "loss": 0.1096,
      "step": 401
    },
    {
      "epoch": 7.1466666666666665,
      "grad_norm": 0.45840370655059814,
      "learning_rate": 2.235119343160303e-05,
      "loss": 0.1555,
      "step": 402
    },
    {
      "epoch": 7.164444444444444,
      "grad_norm": 0.478550523519516,
      "learning_rate": 2.2092053094615813e-05,
      "loss": 0.1301,
      "step": 403
    },
    {
      "epoch": 7.182222222222222,
      "grad_norm": 0.4727082848548889,
      "learning_rate": 2.1833997096818898e-05,
      "loss": 0.1345,
      "step": 404
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.5048395991325378,
      "learning_rate": 2.157703546475539e-05,
      "loss": 0.1733,
      "step": 405
    },
    {
      "epoch": 7.217777777777778,
      "grad_norm": 0.40645188093185425,
      "learning_rate": 2.132117818244771e-05,
      "loss": 0.1491,
      "step": 406
    },
    {
      "epoch": 7.235555555555556,
      "grad_norm": 0.560034990310669,
      "learning_rate": 2.1066435191009715e-05,
      "loss": 0.1226,
      "step": 407
    },
    {
      "epoch": 7.253333333333333,
      "grad_norm": 0.7475597262382507,
      "learning_rate": 2.0812816388260518e-05,
      "loss": 0.1455,
      "step": 408
    },
    {
      "epoch": 7.271111111111111,
      "grad_norm": 0.4967871904373169,
      "learning_rate": 2.056033162833977e-05,
      "loss": 0.1176,
      "step": 409
    },
    {
      "epoch": 7.288888888888889,
      "grad_norm": 1.1265000104904175,
      "learning_rate": 2.0308990721324927e-05,
      "loss": 0.1654,
      "step": 410
    },
    {
      "epoch": 7.306666666666667,
      "grad_norm": 1.0787440538406372,
      "learning_rate": 2.0058803432849987e-05,
      "loss": 0.1741,
      "step": 411
    },
    {
      "epoch": 7.3244444444444445,
      "grad_norm": 0.7019418478012085,
      "learning_rate": 1.980977948372612e-05,
      "loss": 0.1198,
      "step": 412
    },
    {
      "epoch": 7.342222222222222,
      "grad_norm": 0.5232211947441101,
      "learning_rate": 1.9561928549563968e-05,
      "loss": 0.1176,
      "step": 413
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.4455171823501587,
      "learning_rate": 1.931526026039764e-05,
      "loss": 0.1487,
      "step": 414
    },
    {
      "epoch": 7.377777777777778,
      "grad_norm": 0.8757272362709045,
      "learning_rate": 1.906978420031059e-05,
      "loss": 0.1809,
      "step": 415
    },
    {
      "epoch": 7.395555555555555,
      "grad_norm": 0.4900935888290405,
      "learning_rate": 1.8825509907063327e-05,
      "loss": 0.1134,
      "step": 416
    },
    {
      "epoch": 7.413333333333333,
      "grad_norm": 0.4308149218559265,
      "learning_rate": 1.8582446871722636e-05,
      "loss": 0.1467,
      "step": 417
    },
    {
      "epoch": 7.431111111111111,
      "grad_norm": 0.42540425062179565,
      "learning_rate": 1.8340604538293015e-05,
      "loss": 0.1813,
      "step": 418
    },
    {
      "epoch": 7.448888888888889,
      "grad_norm": 0.9641832113265991,
      "learning_rate": 1.8099992303349577e-05,
      "loss": 0.1514,
      "step": 419
    },
    {
      "epoch": 7.466666666666667,
      "grad_norm": 0.47594931721687317,
      "learning_rate": 1.7860619515673033e-05,
      "loss": 0.1348,
      "step": 420
    },
    {
      "epoch": 7.484444444444445,
      "grad_norm": 0.7819913625717163,
      "learning_rate": 1.7622495475886487e-05,
      "loss": 0.127,
      "step": 421
    },
    {
      "epoch": 7.502222222222223,
      "grad_norm": 0.4303249418735504,
      "learning_rate": 1.738562943609396e-05,
      "loss": 0.1081,
      "step": 422
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.8481171131134033,
      "learning_rate": 1.7150030599520984e-05,
      "loss": 0.1875,
      "step": 423
    },
    {
      "epoch": 7.5377777777777775,
      "grad_norm": 0.5441702008247375,
      "learning_rate": 1.691570812015704e-05,
      "loss": 0.1278,
      "step": 424
    },
    {
      "epoch": 7.555555555555555,
      "grad_norm": 0.6104997992515564,
      "learning_rate": 1.6682671102399805e-05,
      "loss": 0.1391,
      "step": 425
    },
    {
      "epoch": 7.573333333333333,
      "grad_norm": 0.4054010212421417,
      "learning_rate": 1.6450928600701504e-05,
      "loss": 0.1568,
      "step": 426
    },
    {
      "epoch": 7.591111111111111,
      "grad_norm": 0.4448254108428955,
      "learning_rate": 1.622048961921699e-05,
      "loss": 0.1352,
      "step": 427
    },
    {
      "epoch": 7.608888888888889,
      "grad_norm": 0.5302364826202393,
      "learning_rate": 1.599136311145402e-05,
      "loss": 0.1635,
      "step": 428
    },
    {
      "epoch": 7.626666666666667,
      "grad_norm": 0.42252257466316223,
      "learning_rate": 1.5763557979925324e-05,
      "loss": 0.1637,
      "step": 429
    },
    {
      "epoch": 7.644444444444445,
      "grad_norm": 0.508930504322052,
      "learning_rate": 1.553708307580265e-05,
      "loss": 0.1537,
      "step": 430
    },
    {
      "epoch": 7.662222222222223,
      "grad_norm": 0.6009820103645325,
      "learning_rate": 1.531194719857292e-05,
      "loss": 0.1205,
      "step": 431
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.4471452236175537,
      "learning_rate": 1.5088159095696363e-05,
      "loss": 0.128,
      "step": 432
    },
    {
      "epoch": 7.697777777777778,
      "grad_norm": 0.6378930807113647,
      "learning_rate": 1.4865727462266543e-05,
      "loss": 0.1254,
      "step": 433
    },
    {
      "epoch": 7.7155555555555555,
      "grad_norm": 0.627171516418457,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.1451,
      "step": 434
    },
    {
      "epoch": 7.733333333333333,
      "grad_norm": 0.4755552411079407,
      "learning_rate": 1.4424968120263504e-05,
      "loss": 0.1285,
      "step": 435
    },
    {
      "epoch": 7.751111111111111,
      "grad_norm": 0.6819527745246887,
      "learning_rate": 1.4206657537014079e-05,
      "loss": 0.1817,
      "step": 436
    },
    {
      "epoch": 7.768888888888889,
      "grad_norm": 0.5436893701553345,
      "learning_rate": 1.398973767319368e-05,
      "loss": 0.1569,
      "step": 437
    },
    {
      "epoch": 7.786666666666667,
      "grad_norm": 0.506430447101593,
      "learning_rate": 1.3774216957036367e-05,
      "loss": 0.1493,
      "step": 438
    },
    {
      "epoch": 7.804444444444444,
      "grad_norm": 0.8908332586288452,
      "learning_rate": 1.3560103762413584e-05,
      "loss": 0.1215,
      "step": 439
    },
    {
      "epoch": 7.822222222222222,
      "grad_norm": 0.38750067353248596,
      "learning_rate": 1.3347406408508695e-05,
      "loss": 0.1389,
      "step": 440
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.44923001527786255,
      "learning_rate": 1.3136133159493802e-05,
      "loss": 0.113,
      "step": 441
    },
    {
      "epoch": 7.857777777777778,
      "grad_norm": 0.4674062132835388,
      "learning_rate": 1.2926292224208664e-05,
      "loss": 0.1344,
      "step": 442
    },
    {
      "epoch": 7.875555555555556,
      "grad_norm": 0.47324058413505554,
      "learning_rate": 1.2717891755841722e-05,
      "loss": 0.1363,
      "step": 443
    },
    {
      "epoch": 7.8933333333333335,
      "grad_norm": 0.6032904386520386,
      "learning_rate": 1.2510939851613285e-05,
      "loss": 0.1293,
      "step": 444
    },
    {
      "epoch": 7.911111111111111,
      "grad_norm": 0.4771234393119812,
      "learning_rate": 1.230544455246101e-05,
      "loss": 0.161,
      "step": 445
    },
    {
      "epoch": 7.928888888888888,
      "grad_norm": 0.48323342204093933,
      "learning_rate": 1.2101413842727345e-05,
      "loss": 0.1336,
      "step": 446
    },
    {
      "epoch": 7.946666666666666,
      "grad_norm": 0.34565791487693787,
      "learning_rate": 1.1898855649849461e-05,
      "loss": 0.177,
      "step": 447
    },
    {
      "epoch": 7.964444444444444,
      "grad_norm": 0.39549532532691956,
      "learning_rate": 1.1697777844051105e-05,
      "loss": 0.1431,
      "step": 448
    },
    {
      "epoch": 7.982222222222222,
      "grad_norm": 0.45379385352134705,
      "learning_rate": 1.1498188238036861e-05,
      "loss": 0.1503,
      "step": 449
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.5633464455604553,
      "learning_rate": 1.130009458668863e-05,
      "loss": 0.1328,
      "step": 450
    },
    {
      "epoch": 8.017777777777777,
      "grad_norm": 0.42473697662353516,
      "learning_rate": 1.1103504586764263e-05,
      "loss": 0.1182,
      "step": 451
    },
    {
      "epoch": 8.035555555555556,
      "grad_norm": 0.34909459948539734,
      "learning_rate": 1.090842587659851e-05,
      "loss": 0.1369,
      "step": 452
    },
    {
      "epoch": 8.053333333333333,
      "grad_norm": 0.49491363763809204,
      "learning_rate": 1.0714866035806326e-05,
      "loss": 0.1346,
      "step": 453
    },
    {
      "epoch": 8.071111111111112,
      "grad_norm": 0.39719170331954956,
      "learning_rate": 1.0522832584988234e-05,
      "loss": 0.144,
      "step": 454
    },
    {
      "epoch": 8.088888888888889,
      "grad_norm": 0.530826210975647,
      "learning_rate": 1.0332332985438248e-05,
      "loss": 0.1522,
      "step": 455
    },
    {
      "epoch": 8.106666666666667,
      "grad_norm": 0.44326362013816833,
      "learning_rate": 1.0143374638853891e-05,
      "loss": 0.1347,
      "step": 456
    },
    {
      "epoch": 8.124444444444444,
      "grad_norm": 0.47626104950904846,
      "learning_rate": 9.955964887048607e-06,
      "loss": 0.1371,
      "step": 457
    },
    {
      "epoch": 8.142222222222221,
      "grad_norm": 0.532647430896759,
      "learning_rate": 9.770111011666583e-06,
      "loss": 0.1247,
      "step": 458
    },
    {
      "epoch": 8.16,
      "grad_norm": 0.6383976340293884,
      "learning_rate": 9.58582023389974e-06,
      "loss": 0.1328,
      "step": 459
    },
    {
      "epoch": 8.177777777777777,
      "grad_norm": 0.6057294011116028,
      "learning_rate": 9.403099714207175e-06,
      "loss": 0.1485,
      "step": 460
    },
    {
      "epoch": 8.195555555555556,
      "grad_norm": 0.3655204474925995,
      "learning_rate": 9.221956552036992e-06,
      "loss": 0.1226,
      "step": 461
    },
    {
      "epoch": 8.213333333333333,
      "grad_norm": 0.3895055055618286,
      "learning_rate": 9.042397785550405e-06,
      "loss": 0.1297,
      "step": 462
    },
    {
      "epoch": 8.231111111111112,
      "grad_norm": 0.42906251549720764,
      "learning_rate": 8.864430391348332e-06,
      "loss": 0.1209,
      "step": 463
    },
    {
      "epoch": 8.248888888888889,
      "grad_norm": 0.6001704335212708,
      "learning_rate": 8.688061284200266e-06,
      "loss": 0.0914,
      "step": 464
    },
    {
      "epoch": 8.266666666666667,
      "grad_norm": 0.5638251304626465,
      "learning_rate": 8.513297316775625e-06,
      "loss": 0.1514,
      "step": 465
    },
    {
      "epoch": 8.284444444444444,
      "grad_norm": 0.6026332378387451,
      "learning_rate": 8.34014527937756e-06,
      "loss": 0.1336,
      "step": 466
    },
    {
      "epoch": 8.302222222222222,
      "grad_norm": 0.4549403190612793,
      "learning_rate": 8.168611899679013e-06,
      "loss": 0.1287,
      "step": 467
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.48877498507499695,
      "learning_rate": 7.998703842461431e-06,
      "loss": 0.1336,
      "step": 468
    },
    {
      "epoch": 8.337777777777777,
      "grad_norm": 0.7382359504699707,
      "learning_rate": 7.830427709355725e-06,
      "loss": 0.1321,
      "step": 469
    },
    {
      "epoch": 8.355555555555556,
      "grad_norm": 0.6301836371421814,
      "learning_rate": 7.663790038585793e-06,
      "loss": 0.1072,
      "step": 470
    },
    {
      "epoch": 8.373333333333333,
      "grad_norm": 0.45454826951026917,
      "learning_rate": 7.498797304714544e-06,
      "loss": 0.1336,
      "step": 471
    },
    {
      "epoch": 8.391111111111112,
      "grad_norm": 0.5664876103401184,
      "learning_rate": 7.33545591839222e-06,
      "loss": 0.1141,
      "step": 472
    },
    {
      "epoch": 8.408888888888889,
      "grad_norm": 0.45692479610443115,
      "learning_rate": 7.173772226107434e-06,
      "loss": 0.1143,
      "step": 473
    },
    {
      "epoch": 8.426666666666666,
      "grad_norm": 0.3842525780200958,
      "learning_rate": 7.013752509940485e-06,
      "loss": 0.1242,
      "step": 474
    },
    {
      "epoch": 8.444444444444445,
      "grad_norm": 0.5034791827201843,
      "learning_rate": 6.855402987319348e-06,
      "loss": 0.1302,
      "step": 475
    },
    {
      "epoch": 8.462222222222222,
      "grad_norm": 1.0054627656936646,
      "learning_rate": 6.698729810778065e-06,
      "loss": 0.1245,
      "step": 476
    },
    {
      "epoch": 8.48,
      "grad_norm": 0.5008876919746399,
      "learning_rate": 6.54373906771768e-06,
      "loss": 0.134,
      "step": 477
    },
    {
      "epoch": 8.497777777777777,
      "grad_norm": 0.6585814356803894,
      "learning_rate": 6.390436780169734e-06,
      "loss": 0.1106,
      "step": 478
    },
    {
      "epoch": 8.515555555555556,
      "grad_norm": 0.40638384222984314,
      "learning_rate": 6.238828904562316e-06,
      "loss": 0.1309,
      "step": 479
    },
    {
      "epoch": 8.533333333333333,
      "grad_norm": 0.6327247619628906,
      "learning_rate": 6.088921331488568e-06,
      "loss": 0.1263,
      "step": 480
    },
    {
      "epoch": 8.551111111111112,
      "grad_norm": 0.4367387592792511,
      "learning_rate": 5.94071988547788e-06,
      "loss": 0.1018,
      "step": 481
    },
    {
      "epoch": 8.568888888888889,
      "grad_norm": 0.5796259641647339,
      "learning_rate": 5.794230324769517e-06,
      "loss": 0.1568,
      "step": 482
    },
    {
      "epoch": 8.586666666666666,
      "grad_norm": 0.485882043838501,
      "learning_rate": 5.649458341088915e-06,
      "loss": 0.1537,
      "step": 483
    },
    {
      "epoch": 8.604444444444445,
      "grad_norm": 0.38468489050865173,
      "learning_rate": 5.506409559426573e-06,
      "loss": 0.1364,
      "step": 484
    },
    {
      "epoch": 8.622222222222222,
      "grad_norm": 2.784595012664795,
      "learning_rate": 5.365089537819434e-06,
      "loss": 0.1214,
      "step": 485
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.40455055236816406,
      "learning_rate": 5.2255037671349535e-06,
      "loss": 0.1029,
      "step": 486
    },
    {
      "epoch": 8.657777777777778,
      "grad_norm": 0.5025741457939148,
      "learning_rate": 5.087657670857798e-06,
      "loss": 0.1279,
      "step": 487
    },
    {
      "epoch": 8.675555555555556,
      "grad_norm": 0.5535502433776855,
      "learning_rate": 4.951556604879048e-06,
      "loss": 0.1055,
      "step": 488
    },
    {
      "epoch": 8.693333333333333,
      "grad_norm": 0.47892096638679504,
      "learning_rate": 4.8172058572881765e-06,
      "loss": 0.1115,
      "step": 489
    },
    {
      "epoch": 8.71111111111111,
      "grad_norm": 0.6920176148414612,
      "learning_rate": 4.684610648167503e-06,
      "loss": 0.1337,
      "step": 490
    },
    {
      "epoch": 8.72888888888889,
      "grad_norm": 0.41214314103126526,
      "learning_rate": 4.5537761293894535e-06,
      "loss": 0.1677,
      "step": 491
    },
    {
      "epoch": 8.746666666666666,
      "grad_norm": 0.4349137544631958,
      "learning_rate": 4.424707384416344e-06,
      "loss": 0.1281,
      "step": 492
    },
    {
      "epoch": 8.764444444444445,
      "grad_norm": 0.5308365225791931,
      "learning_rate": 4.29740942810285e-06,
      "loss": 0.1464,
      "step": 493
    },
    {
      "epoch": 8.782222222222222,
      "grad_norm": 0.5176928043365479,
      "learning_rate": 4.1718872065011904e-06,
      "loss": 0.1266,
      "step": 494
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.4264501631259918,
      "learning_rate": 4.048145596668967e-06,
      "loss": 0.1355,
      "step": 495
    },
    {
      "epoch": 8.817777777777778,
      "grad_norm": 0.5254343152046204,
      "learning_rate": 3.9261894064796135e-06,
      "loss": 0.144,
      "step": 496
    },
    {
      "epoch": 8.835555555555555,
      "grad_norm": 0.400968998670578,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.1422,
      "step": 497
    },
    {
      "epoch": 8.853333333333333,
      "grad_norm": 0.7004261016845703,
      "learning_rate": 3.687652169484568e-06,
      "loss": 0.1472,
      "step": 498
    },
    {
      "epoch": 8.87111111111111,
      "grad_norm": 0.42059558629989624,
      "learning_rate": 3.5710803908373224e-06,
      "loss": 0.1145,
      "step": 499
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 0.5794507265090942,
      "learning_rate": 3.4563125677897932e-06,
      "loss": 0.1145,
      "step": 500
    },
    {
      "epoch": 8.88888888888889,
      "eval_loss": 0.3872971832752228,
      "eval_runtime": 44.5374,
      "eval_samples_per_second": 2.245,
      "eval_steps_per_second": 2.245,
      "step": 500
    },
    {
      "epoch": 8.906666666666666,
      "grad_norm": 0.4100549817085266,
      "learning_rate": 3.343353159546675e-06,
      "loss": 0.1311,
      "step": 501
    },
    {
      "epoch": 8.924444444444445,
      "grad_norm": 0.5439295172691345,
      "learning_rate": 3.2322065550483007e-06,
      "loss": 0.1346,
      "step": 502
    },
    {
      "epoch": 8.942222222222222,
      "grad_norm": 0.5619916319847107,
      "learning_rate": 3.1228770728000455e-06,
      "loss": 0.1145,
      "step": 503
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.6183579564094543,
      "learning_rate": 3.0153689607045845e-06,
      "loss": 0.1122,
      "step": 504
    },
    {
      "epoch": 8.977777777777778,
      "grad_norm": 0.39990657567977905,
      "learning_rate": 2.9096863958968268e-06,
      "loss": 0.155,
      "step": 505
    },
    {
      "epoch": 8.995555555555555,
      "grad_norm": 0.3883807063102722,
      "learning_rate": 2.8058334845816213e-06,
      "loss": 0.1568,
      "step": 506
    },
    {
      "epoch": 9.013333333333334,
      "grad_norm": 0.4355883300304413,
      "learning_rate": 2.7038142618741992e-06,
      "loss": 0.129,
      "step": 507
    },
    {
      "epoch": 9.03111111111111,
      "grad_norm": 0.505295991897583,
      "learning_rate": 2.603632691643415e-06,
      "loss": 0.1233,
      "step": 508
    },
    {
      "epoch": 9.04888888888889,
      "grad_norm": 0.5392528176307678,
      "learning_rate": 2.5052926663577e-06,
      "loss": 0.1235,
      "step": 509
    },
    {
      "epoch": 9.066666666666666,
      "grad_norm": 0.4447910785675049,
      "learning_rate": 2.408798006933882e-06,
      "loss": 0.1068,
      "step": 510
    },
    {
      "epoch": 9.084444444444445,
      "grad_norm": 0.5352734923362732,
      "learning_rate": 2.314152462588659e-06,
      "loss": 0.1364,
      "step": 511
    },
    {
      "epoch": 9.102222222222222,
      "grad_norm": 0.5564258694648743,
      "learning_rate": 2.221359710692961e-06,
      "loss": 0.1492,
      "step": 512
    },
    {
      "epoch": 9.12,
      "grad_norm": 0.3572767972946167,
      "learning_rate": 2.1304233566290964e-06,
      "loss": 0.1079,
      "step": 513
    },
    {
      "epoch": 9.137777777777778,
      "grad_norm": 0.5538711547851562,
      "learning_rate": 2.041346933650612e-06,
      "loss": 0.1298,
      "step": 514
    },
    {
      "epoch": 9.155555555555555,
      "grad_norm": 0.5351611375808716,
      "learning_rate": 1.9541339027450256e-06,
      "loss": 0.1046,
      "step": 515
    },
    {
      "epoch": 9.173333333333334,
      "grad_norm": 0.449429988861084,
      "learning_rate": 1.8687876524993987e-06,
      "loss": 0.1129,
      "step": 516
    },
    {
      "epoch": 9.19111111111111,
      "grad_norm": 0.4684661328792572,
      "learning_rate": 1.785311498968617e-06,
      "loss": 0.1036,
      "step": 517
    },
    {
      "epoch": 9.20888888888889,
      "grad_norm": 0.46259376406669617,
      "learning_rate": 1.70370868554659e-06,
      "loss": 0.1266,
      "step": 518
    },
    {
      "epoch": 9.226666666666667,
      "grad_norm": 0.5819494128227234,
      "learning_rate": 1.6239823828401945e-06,
      "loss": 0.1162,
      "step": 519
    },
    {
      "epoch": 9.244444444444444,
      "grad_norm": 0.3985779583454132,
      "learning_rate": 1.5461356885461075e-06,
      "loss": 0.1179,
      "step": 520
    },
    {
      "epoch": 9.262222222222222,
      "grad_norm": 0.4249963164329529,
      "learning_rate": 1.4701716273304521e-06,
      "loss": 0.1342,
      "step": 521
    },
    {
      "epoch": 9.28,
      "grad_norm": 0.3580509424209595,
      "learning_rate": 1.3960931507112752e-06,
      "loss": 0.1196,
      "step": 522
    },
    {
      "epoch": 9.297777777777778,
      "grad_norm": 0.6348000764846802,
      "learning_rate": 1.3239031369438326e-06,
      "loss": 0.1205,
      "step": 523
    },
    {
      "epoch": 9.315555555555555,
      "grad_norm": 0.4758366644382477,
      "learning_rate": 1.2536043909088191e-06,
      "loss": 0.1159,
      "step": 524
    },
    {
      "epoch": 9.333333333333334,
      "grad_norm": 0.39275023341178894,
      "learning_rate": 1.1851996440033319e-06,
      "loss": 0.1529,
      "step": 525
    },
    {
      "epoch": 9.351111111111111,
      "grad_norm": 0.40188562870025635,
      "learning_rate": 1.118691554034773e-06,
      "loss": 0.1398,
      "step": 526
    },
    {
      "epoch": 9.36888888888889,
      "grad_norm": 0.4341106712818146,
      "learning_rate": 1.0540827051175818e-06,
      "loss": 0.1054,
      "step": 527
    },
    {
      "epoch": 9.386666666666667,
      "grad_norm": 0.4394477307796478,
      "learning_rate": 9.913756075728087e-07,
      "loss": 0.1361,
      "step": 528
    },
    {
      "epoch": 9.404444444444444,
      "grad_norm": 0.42253348231315613,
      "learning_rate": 9.305726978306173e-07,
      "loss": 0.0994,
      "step": 529
    },
    {
      "epoch": 9.422222222222222,
      "grad_norm": 0.5073143243789673,
      "learning_rate": 8.716763383355864e-07,
      "loss": 0.1426,
      "step": 530
    },
    {
      "epoch": 9.44,
      "grad_norm": 0.5548211336135864,
      "learning_rate": 8.146888174549339e-07,
      "loss": 0.1543,
      "step": 531
    },
    {
      "epoch": 9.457777777777778,
      "grad_norm": 0.4399406909942627,
      "learning_rate": 7.596123493895991e-07,
      "loss": 0.124,
      "step": 532
    },
    {
      "epoch": 9.475555555555555,
      "grad_norm": 0.46776387095451355,
      "learning_rate": 7.064490740882057e-07,
      "loss": 0.1438,
      "step": 533
    },
    {
      "epoch": 9.493333333333334,
      "grad_norm": 0.466458261013031,
      "learning_rate": 6.552010571639456e-07,
      "loss": 0.1099,
      "step": 534
    },
    {
      "epoch": 9.511111111111111,
      "grad_norm": 0.3930005133152008,
      "learning_rate": 6.058702898142643e-07,
      "loss": 0.1444,
      "step": 535
    },
    {
      "epoch": 9.528888888888888,
      "grad_norm": 0.36770400404930115,
      "learning_rate": 5.584586887435739e-07,
      "loss": 0.1354,
      "step": 536
    },
    {
      "epoch": 9.546666666666667,
      "grad_norm": 0.49141502380371094,
      "learning_rate": 5.129680960887007e-07,
      "loss": 0.1367,
      "step": 537
    },
    {
      "epoch": 9.564444444444444,
      "grad_norm": 0.47367358207702637,
      "learning_rate": 4.6940027934735954e-07,
      "loss": 0.1215,
      "step": 538
    },
    {
      "epoch": 9.582222222222223,
      "grad_norm": 0.4035596251487732,
      "learning_rate": 4.277569313094809e-07,
      "loss": 0.0906,
      "step": 539
    },
    {
      "epoch": 9.6,
      "grad_norm": 0.43266692757606506,
      "learning_rate": 3.8803966999139684e-07,
      "loss": 0.0995,
      "step": 540
    },
    {
      "epoch": 9.617777777777778,
      "grad_norm": 0.4476134181022644,
      "learning_rate": 3.50250038573019e-07,
      "loss": 0.1209,
      "step": 541
    },
    {
      "epoch": 9.635555555555555,
      "grad_norm": 0.3680909276008606,
      "learning_rate": 3.143895053378698e-07,
      "loss": 0.1201,
      "step": 542
    },
    {
      "epoch": 9.653333333333332,
      "grad_norm": 0.3905773162841797,
      "learning_rate": 2.8045946361601183e-07,
      "loss": 0.1151,
      "step": 543
    },
    {
      "epoch": 9.671111111111111,
      "grad_norm": 0.4270147979259491,
      "learning_rate": 2.4846123172992954e-07,
      "loss": 0.1356,
      "step": 544
    },
    {
      "epoch": 9.688888888888888,
      "grad_norm": 0.47167471051216125,
      "learning_rate": 2.1839605294330933e-07,
      "loss": 0.1461,
      "step": 545
    },
    {
      "epoch": 9.706666666666667,
      "grad_norm": 0.3621659576892853,
      "learning_rate": 1.9026509541272275e-07,
      "loss": 0.1065,
      "step": 546
    },
    {
      "epoch": 9.724444444444444,
      "grad_norm": 0.5602201819419861,
      "learning_rate": 1.640694521422459e-07,
      "loss": 0.1043,
      "step": 547
    },
    {
      "epoch": 9.742222222222223,
      "grad_norm": 0.5394431948661804,
      "learning_rate": 1.3981014094099353e-07,
      "loss": 0.1214,
      "step": 548
    },
    {
      "epoch": 9.76,
      "grad_norm": 0.509139895439148,
      "learning_rate": 1.1748810438355628e-07,
      "loss": 0.1368,
      "step": 549
    },
    {
      "epoch": 9.777777777777779,
      "grad_norm": 0.48587632179260254,
      "learning_rate": 9.710420977340762e-08,
      "loss": 0.136,
      "step": 550
    },
    {
      "epoch": 9.795555555555556,
      "grad_norm": 0.38901224732398987,
      "learning_rate": 7.865924910916977e-08,
      "loss": 0.126,
      "step": 551
    },
    {
      "epoch": 9.813333333333333,
      "grad_norm": 0.4409494996070862,
      "learning_rate": 6.215393905388278e-08,
      "loss": 0.1457,
      "step": 552
    },
    {
      "epoch": 9.831111111111111,
      "grad_norm": 0.5387219190597534,
      "learning_rate": 4.7588920907110094e-08,
      "loss": 0.1103,
      "step": 553
    },
    {
      "epoch": 9.848888888888888,
      "grad_norm": 0.48696452379226685,
      "learning_rate": 3.496476058006959e-08,
      "loss": 0.1487,
      "step": 554
    },
    {
      "epoch": 9.866666666666667,
      "grad_norm": 0.4282010495662689,
      "learning_rate": 2.4281948573617874e-08,
      "loss": 0.1124,
      "step": 555
    },
    {
      "epoch": 9.884444444444444,
      "grad_norm": 0.449815034866333,
      "learning_rate": 1.5540899959187727e-08,
      "loss": 0.1004,
      "step": 556
    },
    {
      "epoch": 9.902222222222223,
      "grad_norm": 0.5384798049926758,
      "learning_rate": 8.741954362678772e-09,
      "loss": 0.1479,
      "step": 557
    },
    {
      "epoch": 9.92,
      "grad_norm": 0.47130730748176575,
      "learning_rate": 3.885375951256931e-09,
      "loss": 0.1138,
      "step": 558
    },
    {
      "epoch": 9.937777777777779,
      "grad_norm": 0.6720632910728455,
      "learning_rate": 9.713534230904041e-10,
      "loss": 0.1207,
      "step": 559
    },
    {
      "epoch": 9.955555555555556,
      "grad_norm": 0.4333189129829407,
      "learning_rate": 0.0,
      "loss": 0.1334,
      "step": 560
    },
    {
      "epoch": 9.955555555555556,
      "step": 560,
      "total_flos": 9.116617776876749e+17,
      "train_loss": 0.2247792588014688,
      "train_runtime": 17934.4708,
      "train_samples_per_second": 0.502,
      "train_steps_per_second": 0.031
    }
  ],
  "logging_steps": 1,
  "max_steps": 560,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "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": 9.116617776876749e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}