{
  "best_global_step": 2500,
  "best_metric": 0.999875,
  "best_model_checkpoint": "./output/checkpoint-2500",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004,
      "grad_norm": 1.8289061784744263,
      "learning_rate": 0.0,
      "loss": 0.7116,
      "step": 1
    },
    {
      "epoch": 0.004,
      "grad_norm": 1.8662855625152588,
      "learning_rate": 1.44e-05,
      "loss": 0.6501,
      "step": 10
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.9602453112602234,
      "learning_rate": 3.04e-05,
      "loss": 0.2983,
      "step": 20
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.017806751653552055,
      "learning_rate": 4.64e-05,
      "loss": 0.0069,
      "step": 30
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.005567730404436588,
      "learning_rate": 6.24e-05,
      "loss": 0.0056,
      "step": 40
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.0028145266696810722,
      "learning_rate": 7.840000000000001e-05,
      "loss": 0.0083,
      "step": 50
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.002064670203253627,
      "learning_rate": 9.44e-05,
      "loss": 0.0001,
      "step": 60
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.0016958912601694465,
      "learning_rate": 0.00011040000000000001,
      "loss": 0.0,
      "step": 70
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.0036686942912638187,
      "learning_rate": 0.0001264,
      "loss": 0.0447,
      "step": 80
    },
    {
      "epoch": 0.036,
      "grad_norm": 8.377481460571289,
      "learning_rate": 0.0001424,
      "loss": 0.0066,
      "step": 90
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.64479923248291,
      "learning_rate": 0.00015840000000000003,
      "loss": 0.0222,
      "step": 100
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.01056673564016819,
      "learning_rate": 0.0001744,
      "loss": 0.0128,
      "step": 110
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.007756778970360756,
      "learning_rate": 0.0001904,
      "loss": 0.0306,
      "step": 120
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.027605902403593063,
      "learning_rate": 0.0002064,
      "loss": 0.0185,
      "step": 130
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.0016576085472479463,
      "learning_rate": 0.00022240000000000004,
      "loss": 0.0001,
      "step": 140
    },
    {
      "epoch": 0.06,
      "grad_norm": 16.3161678314209,
      "learning_rate": 0.0002384,
      "loss": 0.0472,
      "step": 150
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.385207861661911,
      "learning_rate": 0.0002544,
      "loss": 0.0789,
      "step": 160
    },
    {
      "epoch": 0.068,
      "grad_norm": 1.33426833152771,
      "learning_rate": 0.0002704,
      "loss": 0.0577,
      "step": 170
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.7534772753715515,
      "learning_rate": 0.0002864,
      "loss": 0.0337,
      "step": 180
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.4073263108730316,
      "learning_rate": 0.00030240000000000003,
      "loss": 0.0354,
      "step": 190
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5311892628669739,
      "learning_rate": 0.00031840000000000004,
      "loss": 0.0886,
      "step": 200
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.04398029297590256,
      "learning_rate": 0.0003344,
      "loss": 0.0649,
      "step": 210
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.02103251777589321,
      "learning_rate": 0.0003504,
      "loss": 0.0122,
      "step": 220
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.021729949861764908,
      "learning_rate": 0.0003664,
      "loss": 0.0007,
      "step": 230
    },
    {
      "epoch": 0.096,
      "grad_norm": 10.518205642700195,
      "learning_rate": 0.0003824,
      "loss": 0.0542,
      "step": 240
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2691439688205719,
      "learning_rate": 0.00039840000000000003,
      "loss": 0.1148,
      "step": 250
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.006621910724788904,
      "learning_rate": 0.00039840000000000003,
      "loss": 0.0033,
      "step": 260
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.006356918718665838,
      "learning_rate": 0.00039662222222222224,
      "loss": 0.0142,
      "step": 270
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.19780497252941132,
      "learning_rate": 0.0003948444444444445,
      "loss": 0.0532,
      "step": 280
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.022000627592206,
      "learning_rate": 0.0003930666666666667,
      "loss": 0.023,
      "step": 290
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.021432537585496902,
      "learning_rate": 0.0003912888888888889,
      "loss": 0.0132,
      "step": 300
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.45686644315719604,
      "learning_rate": 0.00038951111111111113,
      "loss": 0.0491,
      "step": 310
    },
    {
      "epoch": 0.128,
      "grad_norm": 1.6421903371810913,
      "learning_rate": 0.0003877333333333334,
      "loss": 0.034,
      "step": 320
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.10952561348676682,
      "learning_rate": 0.0003859555555555556,
      "loss": 0.0945,
      "step": 330
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.05203436315059662,
      "learning_rate": 0.0003841777777777778,
      "loss": 0.01,
      "step": 340
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.06594817340373993,
      "learning_rate": 0.0003824,
      "loss": 0.0229,
      "step": 350
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.03412980958819389,
      "learning_rate": 0.00038062222222222223,
      "loss": 0.0086,
      "step": 360
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.022151971235871315,
      "learning_rate": 0.00037884444444444444,
      "loss": 0.0098,
      "step": 370
    },
    {
      "epoch": 0.152,
      "grad_norm": 4.759343147277832,
      "learning_rate": 0.00037706666666666665,
      "loss": 0.0014,
      "step": 380
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.031497757881879807,
      "learning_rate": 0.0003752888888888889,
      "loss": 0.0455,
      "step": 390
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.11943238973617554,
      "learning_rate": 0.0003735111111111111,
      "loss": 0.017,
      "step": 400
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.020562870427966118,
      "learning_rate": 0.00037173333333333333,
      "loss": 0.0036,
      "step": 410
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.00521384971216321,
      "learning_rate": 0.00036995555555555554,
      "loss": 0.0004,
      "step": 420
    },
    {
      "epoch": 0.172,
      "grad_norm": 2.1800899505615234,
      "learning_rate": 0.00036817777777777775,
      "loss": 0.0337,
      "step": 430
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.14933565258979797,
      "learning_rate": 0.0003664,
      "loss": 0.02,
      "step": 440
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.230485439300537,
      "learning_rate": 0.0003646222222222222,
      "loss": 0.0351,
      "step": 450
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.012220942415297031,
      "learning_rate": 0.00036284444444444443,
      "loss": 0.0026,
      "step": 460
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.00523759750649333,
      "learning_rate": 0.00036106666666666664,
      "loss": 0.0002,
      "step": 470
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.19926713407039642,
      "learning_rate": 0.0003592888888888889,
      "loss": 0.0263,
      "step": 480
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.0035884054377675056,
      "learning_rate": 0.0003575111111111111,
      "loss": 0.0267,
      "step": 490
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.30148079991340637,
      "learning_rate": 0.0003557333333333333,
      "loss": 0.0446,
      "step": 500
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.046939630061388016,
      "learning_rate": 0.00035395555555555553,
      "loss": 0.0084,
      "step": 510
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.02836071327328682,
      "learning_rate": 0.0003521777777777778,
      "loss": 0.0161,
      "step": 520
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.06878186017274857,
      "learning_rate": 0.0003504,
      "loss": 0.0082,
      "step": 530
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.14635486900806427,
      "learning_rate": 0.0003486222222222222,
      "loss": 0.033,
      "step": 540
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.02621046081185341,
      "learning_rate": 0.0003468444444444445,
      "loss": 0.0234,
      "step": 550
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.020144036039710045,
      "learning_rate": 0.0003450666666666667,
      "loss": 0.005,
      "step": 560
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.06109397113323212,
      "learning_rate": 0.0003432888888888889,
      "loss": 0.0126,
      "step": 570
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.02218439429998398,
      "learning_rate": 0.0003415111111111111,
      "loss": 0.0011,
      "step": 580
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.0069372705183923244,
      "learning_rate": 0.00033973333333333337,
      "loss": 0.0003,
      "step": 590
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.004556322004646063,
      "learning_rate": 0.0003379555555555556,
      "loss": 0.0001,
      "step": 600
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.003253826405853033,
      "learning_rate": 0.0003361777777777778,
      "loss": 0.0001,
      "step": 610
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.0028496645390987396,
      "learning_rate": 0.0003344,
      "loss": 0.0001,
      "step": 620
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.0023998478427529335,
      "learning_rate": 0.00033262222222222226,
      "loss": 0.0001,
      "step": 630
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.0020449627190828323,
      "learning_rate": 0.00033084444444444447,
      "loss": 0.0,
      "step": 640
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.13762225210666656,
      "learning_rate": 0.0003290666666666667,
      "loss": 0.0216,
      "step": 650
    },
    {
      "epoch": 0.264,
      "grad_norm": 9.54612922668457,
      "learning_rate": 0.0003272888888888889,
      "loss": 0.0634,
      "step": 660
    },
    {
      "epoch": 0.268,
      "grad_norm": 2.0493884086608887,
      "learning_rate": 0.00032551111111111115,
      "loss": 0.116,
      "step": 670
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.08442521095275879,
      "learning_rate": 0.00032373333333333336,
      "loss": 0.0113,
      "step": 680
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.010931574739515781,
      "learning_rate": 0.00032195555555555557,
      "loss": 0.0089,
      "step": 690
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.08980958163738251,
      "learning_rate": 0.0003201777777777778,
      "loss": 0.0173,
      "step": 700
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.0746162161231041,
      "learning_rate": 0.00031840000000000004,
      "loss": 0.009,
      "step": 710
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.022072354331612587,
      "learning_rate": 0.00031662222222222225,
      "loss": 0.0016,
      "step": 720
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.007012135349214077,
      "learning_rate": 0.00031484444444444446,
      "loss": 0.0084,
      "step": 730
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.009781530126929283,
      "learning_rate": 0.00031306666666666667,
      "loss": 0.0291,
      "step": 740
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.03511832281947136,
      "learning_rate": 0.00031128888888888893,
      "loss": 0.0103,
      "step": 750
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.03628341481089592,
      "learning_rate": 0.00030951111111111114,
      "loss": 0.0013,
      "step": 760
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.008655968122184277,
      "learning_rate": 0.00030773333333333335,
      "loss": 0.0005,
      "step": 770
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.015080983750522137,
      "learning_rate": 0.00030595555555555556,
      "loss": 0.0209,
      "step": 780
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.025626640766859055,
      "learning_rate": 0.0003041777777777778,
      "loss": 0.0068,
      "step": 790
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.02122630551457405,
      "learning_rate": 0.00030240000000000003,
      "loss": 0.0018,
      "step": 800
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.019404616206884384,
      "learning_rate": 0.00030062222222222224,
      "loss": 0.0088,
      "step": 810
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.037797220051288605,
      "learning_rate": 0.00029884444444444445,
      "loss": 0.0139,
      "step": 820
    },
    {
      "epoch": 0.332,
      "grad_norm": 1.059258222579956,
      "learning_rate": 0.0002970666666666667,
      "loss": 0.0051,
      "step": 830
    },
    {
      "epoch": 0.336,
      "grad_norm": 2.2363479137420654,
      "learning_rate": 0.0002952888888888889,
      "loss": 0.0125,
      "step": 840
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.0257643461227417,
      "learning_rate": 0.00029351111111111113,
      "loss": 0.0121,
      "step": 850
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.023524092510342598,
      "learning_rate": 0.00029173333333333334,
      "loss": 0.0011,
      "step": 860
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.07415185868740082,
      "learning_rate": 0.0002899555555555556,
      "loss": 0.0008,
      "step": 870
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.003202820662409067,
      "learning_rate": 0.0002881777777777778,
      "loss": 0.0011,
      "step": 880
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.4437563121318817,
      "learning_rate": 0.0002864,
      "loss": 0.0088,
      "step": 890
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.00555839529260993,
      "learning_rate": 0.00028462222222222223,
      "loss": 0.0107,
      "step": 900
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.008060933090746403,
      "learning_rate": 0.00028284444444444444,
      "loss": 0.002,
      "step": 910
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.011799403466284275,
      "learning_rate": 0.0002810666666666667,
      "loss": 0.0248,
      "step": 920
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.042705778032541275,
      "learning_rate": 0.0002792888888888889,
      "loss": 0.0022,
      "step": 930
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.005020072218030691,
      "learning_rate": 0.0002775111111111111,
      "loss": 0.0072,
      "step": 940
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.002978922799229622,
      "learning_rate": 0.00027573333333333333,
      "loss": 0.0154,
      "step": 950
    },
    {
      "epoch": 0.384,
      "grad_norm": 3.051888942718506,
      "learning_rate": 0.00027395555555555554,
      "loss": 0.0065,
      "step": 960
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.10347535461187363,
      "learning_rate": 0.00027217777777777775,
      "loss": 0.0033,
      "step": 970
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.008343281224370003,
      "learning_rate": 0.0002704,
      "loss": 0.0003,
      "step": 980
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.0071385628543794155,
      "learning_rate": 0.0002686222222222222,
      "loss": 0.0003,
      "step": 990
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.003130185417830944,
      "learning_rate": 0.00026684444444444443,
      "loss": 0.0062,
      "step": 1000
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.001496330020017922,
      "learning_rate": 0.00026506666666666664,
      "loss": 0.0005,
      "step": 1010
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.8067265748977661,
      "learning_rate": 0.0002632888888888889,
      "loss": 0.0147,
      "step": 1020
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.0030374648049473763,
      "learning_rate": 0.0002615111111111111,
      "loss": 0.0,
      "step": 1030
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.005266970954835415,
      "learning_rate": 0.0002597333333333333,
      "loss": 0.0018,
      "step": 1040
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.015546767972409725,
      "learning_rate": 0.00025795555555555553,
      "loss": 0.0078,
      "step": 1050
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.009915663860738277,
      "learning_rate": 0.0002561777777777778,
      "loss": 0.0021,
      "step": 1060
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.027595413848757744,
      "learning_rate": 0.0002544,
      "loss": 0.0069,
      "step": 1070
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.033019836992025375,
      "learning_rate": 0.0002526222222222222,
      "loss": 0.0231,
      "step": 1080
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.06228971853852272,
      "learning_rate": 0.0002508444444444444,
      "loss": 0.0162,
      "step": 1090
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.02616875059902668,
      "learning_rate": 0.0002490666666666667,
      "loss": 0.0029,
      "step": 1100
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.01255160290747881,
      "learning_rate": 0.0002472888888888889,
      "loss": 0.0014,
      "step": 1110
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.007698772940784693,
      "learning_rate": 0.0002455111111111111,
      "loss": 0.0003,
      "step": 1120
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.00799358356744051,
      "learning_rate": 0.0002437333333333333,
      "loss": 0.0003,
      "step": 1130
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.09039674699306488,
      "learning_rate": 0.00024195555555555558,
      "loss": 0.0003,
      "step": 1140
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.0030508041381835938,
      "learning_rate": 0.00024017777777777779,
      "loss": 0.0004,
      "step": 1150
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.0018616029992699623,
      "learning_rate": 0.0002384,
      "loss": 0.0053,
      "step": 1160
    },
    {
      "epoch": 0.468,
      "grad_norm": 0.0019433458801358938,
      "learning_rate": 0.0002366222222222222,
      "loss": 0.0,
      "step": 1170
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.0020802582148462534,
      "learning_rate": 0.00023484444444444447,
      "loss": 0.0018,
      "step": 1180
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.2650572657585144,
      "learning_rate": 0.00023306666666666668,
      "loss": 0.0005,
      "step": 1190
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.0014364586677402258,
      "learning_rate": 0.00023128888888888889,
      "loss": 0.0114,
      "step": 1200
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.0016622517723590136,
      "learning_rate": 0.00022951111111111115,
      "loss": 0.0032,
      "step": 1210
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.0018463089363649487,
      "learning_rate": 0.00022773333333333336,
      "loss": 0.0001,
      "step": 1220
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.0017726500518620014,
      "learning_rate": 0.00022595555555555557,
      "loss": 0.0,
      "step": 1230
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.0016171345487236977,
      "learning_rate": 0.00022417777777777778,
      "loss": 0.0,
      "step": 1240
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.001747308298945427,
      "learning_rate": 0.00022240000000000004,
      "loss": 0.0,
      "step": 1250
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.01567615382373333,
      "learning_rate": 0.00022062222222222225,
      "loss": 0.0,
      "step": 1260
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.03938116133213043,
      "learning_rate": 0.00021884444444444446,
      "loss": 0.0001,
      "step": 1270
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.001731475000269711,
      "learning_rate": 0.00021706666666666667,
      "loss": 0.0,
      "step": 1280
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.0011080361437052488,
      "learning_rate": 0.00021528888888888893,
      "loss": 0.0,
      "step": 1290
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.001253710244782269,
      "learning_rate": 0.00021351111111111114,
      "loss": 0.0,
      "step": 1300
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.0009362171404063702,
      "learning_rate": 0.00021173333333333335,
      "loss": 0.0002,
      "step": 1310
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.0010680261766538024,
      "learning_rate": 0.00020995555555555556,
      "loss": 0.0,
      "step": 1320
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.0009266381966881454,
      "learning_rate": 0.0002081777777777778,
      "loss": 0.0,
      "step": 1330
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.0008506786543875933,
      "learning_rate": 0.0002064,
      "loss": 0.0,
      "step": 1340
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.000809083750937134,
      "learning_rate": 0.00020462222222222224,
      "loss": 0.0,
      "step": 1350
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.0008981467690318823,
      "learning_rate": 0.00020284444444444445,
      "loss": 0.0,
      "step": 1360
    },
    {
      "epoch": 0.548,
      "grad_norm": 0.006928270682692528,
      "learning_rate": 0.00020106666666666668,
      "loss": 0.0157,
      "step": 1370
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.019026894122362137,
      "learning_rate": 0.0001992888888888889,
      "loss": 0.0003,
      "step": 1380
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.018439382314682007,
      "learning_rate": 0.0001975111111111111,
      "loss": 0.0019,
      "step": 1390
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.00941744539886713,
      "learning_rate": 0.00019573333333333334,
      "loss": 0.0007,
      "step": 1400
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.00815660785883665,
      "learning_rate": 0.00019395555555555555,
      "loss": 0.0002,
      "step": 1410
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.039481475949287415,
      "learning_rate": 0.00019217777777777778,
      "loss": 0.0096,
      "step": 1420
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.005729417782276869,
      "learning_rate": 0.0001904,
      "loss": 0.0112,
      "step": 1430
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.007418328896164894,
      "learning_rate": 0.00018862222222222223,
      "loss": 0.0077,
      "step": 1440
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.012174281291663647,
      "learning_rate": 0.00018684444444444444,
      "loss": 0.0101,
      "step": 1450
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.015086760744452477,
      "learning_rate": 0.00018506666666666667,
      "loss": 0.0003,
      "step": 1460
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.01587488315999508,
      "learning_rate": 0.00018328888888888888,
      "loss": 0.0004,
      "step": 1470
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.03278527408838272,
      "learning_rate": 0.00018151111111111112,
      "loss": 0.014,
      "step": 1480
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.10622735321521759,
      "learning_rate": 0.00017973333333333333,
      "loss": 0.0012,
      "step": 1490
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.014216107316315174,
      "learning_rate": 0.00017795555555555557,
      "loss": 0.0006,
      "step": 1500
    },
    {
      "epoch": 0.604,
      "grad_norm": 0.008536259643733501,
      "learning_rate": 0.0001761777777777778,
      "loss": 0.0004,
      "step": 1510
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.005456692539155483,
      "learning_rate": 0.0001744,
      "loss": 0.003,
      "step": 1520
    },
    {
      "epoch": 0.612,
      "grad_norm": 0.005484251771122217,
      "learning_rate": 0.00017262222222222225,
      "loss": 0.0002,
      "step": 1530
    },
    {
      "epoch": 0.616,
      "grad_norm": 0.006969142239540815,
      "learning_rate": 0.00017084444444444446,
      "loss": 0.0152,
      "step": 1540
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.01335048582404852,
      "learning_rate": 0.0001690666666666667,
      "loss": 0.0003,
      "step": 1550
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.012201100587844849,
      "learning_rate": 0.0001672888888888889,
      "loss": 0.0026,
      "step": 1560
    },
    {
      "epoch": 0.628,
      "grad_norm": 0.01430495735257864,
      "learning_rate": 0.00016551111111111114,
      "loss": 0.0006,
      "step": 1570
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.03964800387620926,
      "learning_rate": 0.00016373333333333335,
      "loss": 0.0052,
      "step": 1580
    },
    {
      "epoch": 0.636,
      "grad_norm": 0.007560222875326872,
      "learning_rate": 0.00016195555555555558,
      "loss": 0.0002,
      "step": 1590
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.00928121991455555,
      "learning_rate": 0.0001601777777777778,
      "loss": 0.0002,
      "step": 1600
    },
    {
      "epoch": 0.644,
      "grad_norm": 0.016387827694416046,
      "learning_rate": 0.00015840000000000003,
      "loss": 0.0002,
      "step": 1610
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.005653001833707094,
      "learning_rate": 0.00015662222222222224,
      "loss": 0.0002,
      "step": 1620
    },
    {
      "epoch": 0.652,
      "grad_norm": 0.005010788328945637,
      "learning_rate": 0.00015484444444444445,
      "loss": 0.0026,
      "step": 1630
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.00448651984333992,
      "learning_rate": 0.00015306666666666666,
      "loss": 0.0001,
      "step": 1640
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.011779513210058212,
      "learning_rate": 0.0001512888888888889,
      "loss": 0.0001,
      "step": 1650
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.005999232176691294,
      "learning_rate": 0.0001495111111111111,
      "loss": 0.0026,
      "step": 1660
    },
    {
      "epoch": 0.668,
      "grad_norm": 0.005314210895448923,
      "learning_rate": 0.00014773333333333334,
      "loss": 0.0001,
      "step": 1670
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.00803409144282341,
      "learning_rate": 0.00014595555555555555,
      "loss": 0.0008,
      "step": 1680
    },
    {
      "epoch": 0.676,
      "grad_norm": 1.069173812866211,
      "learning_rate": 0.00014417777777777778,
      "loss": 0.0009,
      "step": 1690
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.013021737337112427,
      "learning_rate": 0.0001424,
      "loss": 0.0002,
      "step": 1700
    },
    {
      "epoch": 0.684,
      "grad_norm": 0.0024978595320135355,
      "learning_rate": 0.00014062222222222223,
      "loss": 0.0089,
      "step": 1710
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.003698146902024746,
      "learning_rate": 0.00013884444444444444,
      "loss": 0.0138,
      "step": 1720
    },
    {
      "epoch": 0.692,
      "grad_norm": 0.03162381052970886,
      "learning_rate": 0.00013706666666666667,
      "loss": 0.0047,
      "step": 1730
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.022371523082256317,
      "learning_rate": 0.00013528888888888888,
      "loss": 0.0021,
      "step": 1740
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.008335133083164692,
      "learning_rate": 0.00013351111111111112,
      "loss": 0.004,
      "step": 1750
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.01031657587736845,
      "learning_rate": 0.00013173333333333333,
      "loss": 0.0057,
      "step": 1760
    },
    {
      "epoch": 0.708,
      "grad_norm": 0.008161673322319984,
      "learning_rate": 0.00012995555555555556,
      "loss": 0.0062,
      "step": 1770
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.007272409275174141,
      "learning_rate": 0.00012817777777777777,
      "loss": 0.0002,
      "step": 1780
    },
    {
      "epoch": 0.716,
      "grad_norm": 0.006052941549569368,
      "learning_rate": 0.0001264,
      "loss": 0.0004,
      "step": 1790
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.00472453935071826,
      "learning_rate": 0.00012462222222222222,
      "loss": 0.0015,
      "step": 1800
    },
    {
      "epoch": 0.724,
      "grad_norm": 0.004381476901471615,
      "learning_rate": 0.00012284444444444445,
      "loss": 0.0001,
      "step": 1810
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.004298892803490162,
      "learning_rate": 0.00012106666666666666,
      "loss": 0.0001,
      "step": 1820
    },
    {
      "epoch": 0.732,
      "grad_norm": 0.003755762241780758,
      "learning_rate": 0.0001192888888888889,
      "loss": 0.0069,
      "step": 1830
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.002545682480558753,
      "learning_rate": 0.00011751111111111112,
      "loss": 0.0001,
      "step": 1840
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.0083426209166646,
      "learning_rate": 0.00011573333333333333,
      "loss": 0.0145,
      "step": 1850
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.03931976854801178,
      "learning_rate": 0.00011395555555555557,
      "loss": 0.0011,
      "step": 1860
    },
    {
      "epoch": 0.748,
      "grad_norm": 0.012124504894018173,
      "learning_rate": 0.00011217777777777778,
      "loss": 0.0008,
      "step": 1870
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.00980482716113329,
      "learning_rate": 0.00011040000000000001,
      "loss": 0.0037,
      "step": 1880
    },
    {
      "epoch": 0.756,
      "grad_norm": 2.7182815074920654,
      "learning_rate": 0.00010862222222222222,
      "loss": 0.0013,
      "step": 1890
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.00282659032382071,
      "learning_rate": 0.00010684444444444446,
      "loss": 0.0015,
      "step": 1900
    },
    {
      "epoch": 0.764,
      "grad_norm": 0.0028170158620923758,
      "learning_rate": 0.00010506666666666667,
      "loss": 0.0002,
      "step": 1910
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.0013493013102561235,
      "learning_rate": 0.0001032888888888889,
      "loss": 0.0011,
      "step": 1920
    },
    {
      "epoch": 0.772,
      "grad_norm": 0.0013236267259344459,
      "learning_rate": 0.00010151111111111111,
      "loss": 0.0113,
      "step": 1930
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.0017112682107836008,
      "learning_rate": 9.973333333333334e-05,
      "loss": 0.0,
      "step": 1940
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.0022746031172573566,
      "learning_rate": 9.795555555555556e-05,
      "loss": 0.0002,
      "step": 1950
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.0015560939209535718,
      "learning_rate": 9.617777777777778e-05,
      "loss": 0.0003,
      "step": 1960
    },
    {
      "epoch": 0.788,
      "grad_norm": 0.001704063848592341,
      "learning_rate": 9.44e-05,
      "loss": 0.0001,
      "step": 1970
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.13681542873382568,
      "learning_rate": 9.262222222222223e-05,
      "loss": 0.0001,
      "step": 1980
    },
    {
      "epoch": 0.796,
      "grad_norm": 0.0013939260970801115,
      "learning_rate": 9.084444444444445e-05,
      "loss": 0.0,
      "step": 1990
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.0016302280128002167,
      "learning_rate": 8.906666666666667e-05,
      "loss": 0.0,
      "step": 2000
    },
    {
      "epoch": 0.804,
      "grad_norm": 0.0014311623526737094,
      "learning_rate": 8.72888888888889e-05,
      "loss": 0.0001,
      "step": 2010
    },
    {
      "epoch": 0.808,
      "grad_norm": 0.0013674832880496979,
      "learning_rate": 8.551111111111112e-05,
      "loss": 0.0,
      "step": 2020
    },
    {
      "epoch": 0.812,
      "grad_norm": 0.0013156217755749822,
      "learning_rate": 8.373333333333334e-05,
      "loss": 0.0,
      "step": 2030
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.7331569790840149,
      "learning_rate": 8.195555555555556e-05,
      "loss": 0.0131,
      "step": 2040
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.006600831635296345,
      "learning_rate": 8.017777777777779e-05,
      "loss": 0.0168,
      "step": 2050
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.012540962547063828,
      "learning_rate": 7.840000000000001e-05,
      "loss": 0.0005,
      "step": 2060
    },
    {
      "epoch": 0.828,
      "grad_norm": 0.023732764646410942,
      "learning_rate": 7.662222222222223e-05,
      "loss": 0.0004,
      "step": 2070
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.01323721744120121,
      "learning_rate": 7.484444444444445e-05,
      "loss": 0.0009,
      "step": 2080
    },
    {
      "epoch": 0.836,
      "grad_norm": 0.009665262885391712,
      "learning_rate": 7.306666666666668e-05,
      "loss": 0.0003,
      "step": 2090
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.008318779990077019,
      "learning_rate": 7.12888888888889e-05,
      "loss": 0.0002,
      "step": 2100
    },
    {
      "epoch": 0.844,
      "grad_norm": 0.0076893046498298645,
      "learning_rate": 6.951111111111112e-05,
      "loss": 0.0002,
      "step": 2110
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.007633762434124947,
      "learning_rate": 6.773333333333333e-05,
      "loss": 0.0002,
      "step": 2120
    },
    {
      "epoch": 0.852,
      "grad_norm": 0.0073342034593224525,
      "learning_rate": 6.595555555555555e-05,
      "loss": 0.0002,
      "step": 2130
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.005839989520609379,
      "learning_rate": 6.417777777777778e-05,
      "loss": 0.0001,
      "step": 2140
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.005278122611343861,
      "learning_rate": 6.24e-05,
      "loss": 0.0001,
      "step": 2150
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.006889206822961569,
      "learning_rate": 6.062222222222222e-05,
      "loss": 0.0001,
      "step": 2160
    },
    {
      "epoch": 0.868,
      "grad_norm": 0.7518675923347473,
      "learning_rate": 5.8844444444444444e-05,
      "loss": 0.0011,
      "step": 2170
    },
    {
      "epoch": 0.872,
      "grad_norm": 0.009561669081449509,
      "learning_rate": 5.706666666666667e-05,
      "loss": 0.0015,
      "step": 2180
    },
    {
      "epoch": 0.876,
      "grad_norm": 0.003935567103326321,
      "learning_rate": 5.528888888888889e-05,
      "loss": 0.0013,
      "step": 2190
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.003710120217874646,
      "learning_rate": 5.351111111111111e-05,
      "loss": 0.0001,
      "step": 2200
    },
    {
      "epoch": 0.884,
      "grad_norm": 0.002907069865614176,
      "learning_rate": 5.1733333333333335e-05,
      "loss": 0.0002,
      "step": 2210
    },
    {
      "epoch": 0.888,
      "grad_norm": 0.0030114708933979273,
      "learning_rate": 4.995555555555556e-05,
      "loss": 0.0023,
      "step": 2220
    },
    {
      "epoch": 0.892,
      "grad_norm": 0.0034878470469266176,
      "learning_rate": 4.817777777777778e-05,
      "loss": 0.0001,
      "step": 2230
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.003132345387712121,
      "learning_rate": 4.64e-05,
      "loss": 0.0004,
      "step": 2240
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.0030635548755526543,
      "learning_rate": 4.4622222222222226e-05,
      "loss": 0.0001,
      "step": 2250
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.0031918687745928764,
      "learning_rate": 4.284444444444445e-05,
      "loss": 0.0031,
      "step": 2260
    },
    {
      "epoch": 0.908,
      "grad_norm": 0.006079728715121746,
      "learning_rate": 4.106666666666667e-05,
      "loss": 0.0001,
      "step": 2270
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.0049600861966609955,
      "learning_rate": 3.9288888888888894e-05,
      "loss": 0.0001,
      "step": 2280
    },
    {
      "epoch": 0.916,
      "grad_norm": 0.006550566293299198,
      "learning_rate": 3.7511111111111116e-05,
      "loss": 0.0001,
      "step": 2290
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.0037377977278083563,
      "learning_rate": 3.573333333333333e-05,
      "loss": 0.0043,
      "step": 2300
    },
    {
      "epoch": 0.924,
      "grad_norm": 0.0033655581064522266,
      "learning_rate": 3.3955555555555555e-05,
      "loss": 0.0001,
      "step": 2310
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.0035526896826922894,
      "learning_rate": 3.217777777777778e-05,
      "loss": 0.0001,
      "step": 2320
    },
    {
      "epoch": 0.932,
      "grad_norm": 0.0034034913405776024,
      "learning_rate": 3.04e-05,
      "loss": 0.0001,
      "step": 2330
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.006957338657230139,
      "learning_rate": 2.8622222222222223e-05,
      "loss": 0.0156,
      "step": 2340
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.007531840819865465,
      "learning_rate": 2.6844444444444446e-05,
      "loss": 0.0001,
      "step": 2350
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.00489840330556035,
      "learning_rate": 2.5066666666666665e-05,
      "loss": 0.0001,
      "step": 2360
    },
    {
      "epoch": 0.948,
      "grad_norm": 0.005002748221158981,
      "learning_rate": 2.328888888888889e-05,
      "loss": 0.0001,
      "step": 2370
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.0048461793921887875,
      "learning_rate": 2.1511111111111114e-05,
      "loss": 0.0001,
      "step": 2380
    },
    {
      "epoch": 0.956,
      "grad_norm": 0.0058443606831133366,
      "learning_rate": 1.9733333333333333e-05,
      "loss": 0.0004,
      "step": 2390
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.0049603781662881374,
      "learning_rate": 1.7955555555555556e-05,
      "loss": 0.0002,
      "step": 2400
    },
    {
      "epoch": 0.964,
      "grad_norm": 0.005183044355362654,
      "learning_rate": 1.617777777777778e-05,
      "loss": 0.0035,
      "step": 2410
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.0050833080895245075,
      "learning_rate": 1.44e-05,
      "loss": 0.0012,
      "step": 2420
    },
    {
      "epoch": 0.972,
      "grad_norm": 0.004844759125262499,
      "learning_rate": 1.2622222222222224e-05,
      "loss": 0.0001,
      "step": 2430
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.004592160694301128,
      "learning_rate": 1.0844444444444445e-05,
      "loss": 0.0012,
      "step": 2440
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.004524698480963707,
      "learning_rate": 9.066666666666667e-06,
      "loss": 0.0001,
      "step": 2450
    },
    {
      "epoch": 0.984,
      "grad_norm": 0.005091739818453789,
      "learning_rate": 7.288888888888889e-06,
      "loss": 0.0002,
      "step": 2460
    },
    {
      "epoch": 0.988,
      "grad_norm": 0.004614758305251598,
      "learning_rate": 5.511111111111111e-06,
      "loss": 0.0001,
      "step": 2470
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.010195998474955559,
      "learning_rate": 3.7333333333333337e-06,
      "loss": 0.0003,
      "step": 2480
    },
    {
      "epoch": 0.996,
      "grad_norm": 0.004264904651790857,
      "learning_rate": 1.9555555555555556e-06,
      "loss": 0.0002,
      "step": 2490
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.004516505636274815,
      "learning_rate": 1.777777777777778e-07,
      "loss": 0.0002,
      "step": 2500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.999875,
      "eval_loss": 0.0004970150184817612,
      "eval_runtime": 2540.0992,
      "eval_samples_per_second": 15.747,
      "eval_steps_per_second": 0.984,
      "step": 2500
    }
  ],
  "logging_steps": 10,
  "max_steps": 2500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 2.0090762496e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}