{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.094890510948905,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000364963503649635,
      "grad_norm": 2.6652841567993164,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.1504,
      "step": 1
    },
    {
      "epoch": 0.00072992700729927,
      "grad_norm": 2.3818790912628174,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.0753,
      "step": 2
    },
    {
      "epoch": 0.0010948905109489052,
      "grad_norm": 3.150963544845581,
      "learning_rate": 1.2e-05,
      "loss": 2.0346,
      "step": 3
    },
    {
      "epoch": 0.00145985401459854,
      "grad_norm": 2.5247068405151367,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.0954,
      "step": 4
    },
    {
      "epoch": 0.0018248175182481751,
      "grad_norm": 1.5434119701385498,
      "learning_rate": 2e-05,
      "loss": 1.868,
      "step": 5
    },
    {
      "epoch": 0.0021897810218978104,
      "grad_norm": 0.9381213784217834,
      "learning_rate": 2.4e-05,
      "loss": 1.9279,
      "step": 6
    },
    {
      "epoch": 0.0025547445255474453,
      "grad_norm": 0.983879029750824,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 1.976,
      "step": 7
    },
    {
      "epoch": 0.00291970802919708,
      "grad_norm": 0.5694950222969055,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 1.8964,
      "step": 8
    },
    {
      "epoch": 0.0032846715328467154,
      "grad_norm": 0.44347015023231506,
      "learning_rate": 3.6e-05,
      "loss": 1.8543,
      "step": 9
    },
    {
      "epoch": 0.0036496350364963502,
      "grad_norm": 0.5018998384475708,
      "learning_rate": 4e-05,
      "loss": 1.8235,
      "step": 10
    },
    {
      "epoch": 0.0040145985401459855,
      "grad_norm": 0.4777238667011261,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.7983,
      "step": 11
    },
    {
      "epoch": 0.004379562043795621,
      "grad_norm": 0.4549931585788727,
      "learning_rate": 4.8e-05,
      "loss": 1.7466,
      "step": 12
    },
    {
      "epoch": 0.004744525547445255,
      "grad_norm": 0.5172353386878967,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 1.8362,
      "step": 13
    },
    {
      "epoch": 0.0051094890510948905,
      "grad_norm": 0.38685303926467896,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 1.9337,
      "step": 14
    },
    {
      "epoch": 0.005474452554744526,
      "grad_norm": 0.333118736743927,
      "learning_rate": 6e-05,
      "loss": 1.7544,
      "step": 15
    },
    {
      "epoch": 0.00583941605839416,
      "grad_norm": 0.2205871045589447,
      "learning_rate": 6.400000000000001e-05,
      "loss": 1.8537,
      "step": 16
    },
    {
      "epoch": 0.0062043795620437955,
      "grad_norm": 0.2148354947566986,
      "learning_rate": 6.800000000000001e-05,
      "loss": 1.7724,
      "step": 17
    },
    {
      "epoch": 0.006569343065693431,
      "grad_norm": 0.2315240353345871,
      "learning_rate": 7.2e-05,
      "loss": 1.7859,
      "step": 18
    },
    {
      "epoch": 0.006934306569343066,
      "grad_norm": 0.20595505833625793,
      "learning_rate": 7.6e-05,
      "loss": 1.8268,
      "step": 19
    },
    {
      "epoch": 0.0072992700729927005,
      "grad_norm": 0.23184804618358612,
      "learning_rate": 8e-05,
      "loss": 1.8584,
      "step": 20
    },
    {
      "epoch": 0.007664233576642336,
      "grad_norm": 0.22077114880084991,
      "learning_rate": 8.4e-05,
      "loss": 1.797,
      "step": 21
    },
    {
      "epoch": 0.008029197080291971,
      "grad_norm": 0.22543348371982574,
      "learning_rate": 8.800000000000001e-05,
      "loss": 1.8684,
      "step": 22
    },
    {
      "epoch": 0.008394160583941606,
      "grad_norm": 0.22065207362174988,
      "learning_rate": 9.200000000000001e-05,
      "loss": 1.7857,
      "step": 23
    },
    {
      "epoch": 0.008759124087591242,
      "grad_norm": 0.23953720927238464,
      "learning_rate": 9.6e-05,
      "loss": 1.8335,
      "step": 24
    },
    {
      "epoch": 0.009124087591240875,
      "grad_norm": 0.1894896775484085,
      "learning_rate": 0.0001,
      "loss": 1.7867,
      "step": 25
    },
    {
      "epoch": 0.00948905109489051,
      "grad_norm": 0.169825479388237,
      "learning_rate": 0.00010400000000000001,
      "loss": 1.7816,
      "step": 26
    },
    {
      "epoch": 0.009854014598540146,
      "grad_norm": 0.17420512437820435,
      "learning_rate": 0.00010800000000000001,
      "loss": 1.73,
      "step": 27
    },
    {
      "epoch": 0.010218978102189781,
      "grad_norm": 0.15135613083839417,
      "learning_rate": 0.00011200000000000001,
      "loss": 1.739,
      "step": 28
    },
    {
      "epoch": 0.010583941605839416,
      "grad_norm": 0.16301587224006653,
      "learning_rate": 0.000116,
      "loss": 1.7856,
      "step": 29
    },
    {
      "epoch": 0.010948905109489052,
      "grad_norm": 0.16961249709129333,
      "learning_rate": 0.00012,
      "loss": 1.7099,
      "step": 30
    },
    {
      "epoch": 0.011313868613138687,
      "grad_norm": 0.18324609100818634,
      "learning_rate": 0.000124,
      "loss": 1.7587,
      "step": 31
    },
    {
      "epoch": 0.01167883211678832,
      "grad_norm": 0.15606151521205902,
      "learning_rate": 0.00012800000000000002,
      "loss": 1.7949,
      "step": 32
    },
    {
      "epoch": 0.012043795620437956,
      "grad_norm": 0.14110200107097626,
      "learning_rate": 0.000132,
      "loss": 1.7098,
      "step": 33
    },
    {
      "epoch": 0.012408759124087591,
      "grad_norm": 0.1338261365890503,
      "learning_rate": 0.00013600000000000003,
      "loss": 1.6709,
      "step": 34
    },
    {
      "epoch": 0.012773722627737226,
      "grad_norm": 0.16924400627613068,
      "learning_rate": 0.00014,
      "loss": 1.7945,
      "step": 35
    },
    {
      "epoch": 0.013138686131386862,
      "grad_norm": 0.15808552503585815,
      "learning_rate": 0.000144,
      "loss": 1.7776,
      "step": 36
    },
    {
      "epoch": 0.013503649635036497,
      "grad_norm": 0.15632940828800201,
      "learning_rate": 0.000148,
      "loss": 1.7272,
      "step": 37
    },
    {
      "epoch": 0.013868613138686132,
      "grad_norm": 0.13921982049942017,
      "learning_rate": 0.000152,
      "loss": 1.8103,
      "step": 38
    },
    {
      "epoch": 0.014233576642335766,
      "grad_norm": 0.12642377614974976,
      "learning_rate": 0.00015600000000000002,
      "loss": 1.6595,
      "step": 39
    },
    {
      "epoch": 0.014598540145985401,
      "grad_norm": 0.12675824761390686,
      "learning_rate": 0.00016,
      "loss": 1.7175,
      "step": 40
    },
    {
      "epoch": 0.014963503649635036,
      "grad_norm": 0.12876400351524353,
      "learning_rate": 0.000164,
      "loss": 1.71,
      "step": 41
    },
    {
      "epoch": 0.015328467153284672,
      "grad_norm": 0.12826906144618988,
      "learning_rate": 0.000168,
      "loss": 1.7481,
      "step": 42
    },
    {
      "epoch": 0.015693430656934307,
      "grad_norm": 0.12679742276668549,
      "learning_rate": 0.000172,
      "loss": 1.7828,
      "step": 43
    },
    {
      "epoch": 0.016058394160583942,
      "grad_norm": 0.11728192120790482,
      "learning_rate": 0.00017600000000000002,
      "loss": 1.7076,
      "step": 44
    },
    {
      "epoch": 0.016423357664233577,
      "grad_norm": 0.12272967398166656,
      "learning_rate": 0.00018,
      "loss": 1.6388,
      "step": 45
    },
    {
      "epoch": 0.016788321167883213,
      "grad_norm": 0.12605397403240204,
      "learning_rate": 0.00018400000000000003,
      "loss": 1.7846,
      "step": 46
    },
    {
      "epoch": 0.017153284671532848,
      "grad_norm": 0.13425585627555847,
      "learning_rate": 0.000188,
      "loss": 1.6826,
      "step": 47
    },
    {
      "epoch": 0.017518248175182483,
      "grad_norm": 0.12273597717285156,
      "learning_rate": 0.000192,
      "loss": 1.6744,
      "step": 48
    },
    {
      "epoch": 0.017883211678832115,
      "grad_norm": 0.12498332560062408,
      "learning_rate": 0.000196,
      "loss": 1.6898,
      "step": 49
    },
    {
      "epoch": 0.01824817518248175,
      "grad_norm": 0.11545863747596741,
      "learning_rate": 0.0002,
      "loss": 1.7596,
      "step": 50
    },
    {
      "epoch": 0.018613138686131386,
      "grad_norm": 0.12387639284133911,
      "learning_rate": 0.0002,
      "loss": 1.6486,
      "step": 51
    },
    {
      "epoch": 0.01897810218978102,
      "grad_norm": 0.11782488971948624,
      "learning_rate": 0.0002,
      "loss": 1.7644,
      "step": 52
    },
    {
      "epoch": 0.019343065693430656,
      "grad_norm": 0.11852247267961502,
      "learning_rate": 0.0002,
      "loss": 1.6751,
      "step": 53
    },
    {
      "epoch": 0.01970802919708029,
      "grad_norm": 0.12549245357513428,
      "learning_rate": 0.0002,
      "loss": 1.7141,
      "step": 54
    },
    {
      "epoch": 0.020072992700729927,
      "grad_norm": 0.11000435799360275,
      "learning_rate": 0.0002,
      "loss": 1.6059,
      "step": 55
    },
    {
      "epoch": 0.020437956204379562,
      "grad_norm": 0.12420882284641266,
      "learning_rate": 0.0002,
      "loss": 1.7063,
      "step": 56
    },
    {
      "epoch": 0.020802919708029197,
      "grad_norm": 0.11040220409631729,
      "learning_rate": 0.0002,
      "loss": 1.5744,
      "step": 57
    },
    {
      "epoch": 0.021167883211678833,
      "grad_norm": 0.11722172051668167,
      "learning_rate": 0.0002,
      "loss": 1.7319,
      "step": 58
    },
    {
      "epoch": 0.021532846715328468,
      "grad_norm": 0.11857732385396957,
      "learning_rate": 0.0002,
      "loss": 1.656,
      "step": 59
    },
    {
      "epoch": 0.021897810218978103,
      "grad_norm": 0.13230836391448975,
      "learning_rate": 0.0002,
      "loss": 1.7264,
      "step": 60
    },
    {
      "epoch": 0.02226277372262774,
      "grad_norm": 0.11560218036174774,
      "learning_rate": 0.0002,
      "loss": 1.6478,
      "step": 61
    },
    {
      "epoch": 0.022627737226277374,
      "grad_norm": 0.1175946518778801,
      "learning_rate": 0.0002,
      "loss": 1.6417,
      "step": 62
    },
    {
      "epoch": 0.022992700729927006,
      "grad_norm": 0.11106300354003906,
      "learning_rate": 0.0002,
      "loss": 1.6562,
      "step": 63
    },
    {
      "epoch": 0.02335766423357664,
      "grad_norm": 0.11284781992435455,
      "learning_rate": 0.0002,
      "loss": 1.7173,
      "step": 64
    },
    {
      "epoch": 0.023722627737226276,
      "grad_norm": 0.12277217209339142,
      "learning_rate": 0.0002,
      "loss": 1.7087,
      "step": 65
    },
    {
      "epoch": 0.02408759124087591,
      "grad_norm": 0.12378532439470291,
      "learning_rate": 0.0002,
      "loss": 1.705,
      "step": 66
    },
    {
      "epoch": 0.024452554744525547,
      "grad_norm": 0.11504566669464111,
      "learning_rate": 0.0002,
      "loss": 1.6186,
      "step": 67
    },
    {
      "epoch": 0.024817518248175182,
      "grad_norm": 0.12077129632234573,
      "learning_rate": 0.0002,
      "loss": 1.6381,
      "step": 68
    },
    {
      "epoch": 0.025182481751824817,
      "grad_norm": 0.11701320856809616,
      "learning_rate": 0.0002,
      "loss": 1.7318,
      "step": 69
    },
    {
      "epoch": 0.025547445255474453,
      "grad_norm": 0.12836797535419464,
      "learning_rate": 0.0002,
      "loss": 1.719,
      "step": 70
    },
    {
      "epoch": 0.025912408759124088,
      "grad_norm": 0.12307879328727722,
      "learning_rate": 0.0002,
      "loss": 1.6991,
      "step": 71
    },
    {
      "epoch": 0.026277372262773723,
      "grad_norm": 0.11436320841312408,
      "learning_rate": 0.0002,
      "loss": 1.6712,
      "step": 72
    },
    {
      "epoch": 0.02664233576642336,
      "grad_norm": 0.1211828961968422,
      "learning_rate": 0.0002,
      "loss": 1.7003,
      "step": 73
    },
    {
      "epoch": 0.027007299270072994,
      "grad_norm": 0.11544755101203918,
      "learning_rate": 0.0002,
      "loss": 1.6992,
      "step": 74
    },
    {
      "epoch": 0.02737226277372263,
      "grad_norm": 0.12271855026483536,
      "learning_rate": 0.0002,
      "loss": 1.6341,
      "step": 75
    },
    {
      "epoch": 0.027737226277372264,
      "grad_norm": 0.11788719892501831,
      "learning_rate": 0.0002,
      "loss": 1.7064,
      "step": 76
    },
    {
      "epoch": 0.028102189781021896,
      "grad_norm": 0.11568509042263031,
      "learning_rate": 0.0002,
      "loss": 1.6778,
      "step": 77
    },
    {
      "epoch": 0.02846715328467153,
      "grad_norm": 0.12990358471870422,
      "learning_rate": 0.0002,
      "loss": 1.7488,
      "step": 78
    },
    {
      "epoch": 0.028832116788321167,
      "grad_norm": 0.10779744386672974,
      "learning_rate": 0.0002,
      "loss": 1.545,
      "step": 79
    },
    {
      "epoch": 0.029197080291970802,
      "grad_norm": 0.11127518862485886,
      "learning_rate": 0.0002,
      "loss": 1.6328,
      "step": 80
    },
    {
      "epoch": 0.029562043795620437,
      "grad_norm": 0.1286451816558838,
      "learning_rate": 0.0002,
      "loss": 1.6964,
      "step": 81
    },
    {
      "epoch": 0.029927007299270073,
      "grad_norm": 0.12511080503463745,
      "learning_rate": 0.0002,
      "loss": 1.6018,
      "step": 82
    },
    {
      "epoch": 0.030291970802919708,
      "grad_norm": 0.15127132833003998,
      "learning_rate": 0.0002,
      "loss": 1.6947,
      "step": 83
    },
    {
      "epoch": 0.030656934306569343,
      "grad_norm": 0.11705619096755981,
      "learning_rate": 0.0002,
      "loss": 1.6303,
      "step": 84
    },
    {
      "epoch": 0.03102189781021898,
      "grad_norm": 0.14967235922813416,
      "learning_rate": 0.0002,
      "loss": 1.6805,
      "step": 85
    },
    {
      "epoch": 0.031386861313868614,
      "grad_norm": 0.12219926714897156,
      "learning_rate": 0.0002,
      "loss": 1.6277,
      "step": 86
    },
    {
      "epoch": 0.031751824817518245,
      "grad_norm": 0.14408400654792786,
      "learning_rate": 0.0002,
      "loss": 1.6724,
      "step": 87
    },
    {
      "epoch": 0.032116788321167884,
      "grad_norm": 0.12710924446582794,
      "learning_rate": 0.0002,
      "loss": 1.6975,
      "step": 88
    },
    {
      "epoch": 0.032481751824817516,
      "grad_norm": 0.11335797607898712,
      "learning_rate": 0.0002,
      "loss": 1.6642,
      "step": 89
    },
    {
      "epoch": 0.032846715328467155,
      "grad_norm": 0.12308158725500107,
      "learning_rate": 0.0002,
      "loss": 1.6919,
      "step": 90
    },
    {
      "epoch": 0.03321167883211679,
      "grad_norm": 0.11992021650075912,
      "learning_rate": 0.0002,
      "loss": 1.6693,
      "step": 91
    },
    {
      "epoch": 0.033576642335766425,
      "grad_norm": 0.12926547229290009,
      "learning_rate": 0.0002,
      "loss": 1.5803,
      "step": 92
    },
    {
      "epoch": 0.03394160583941606,
      "grad_norm": 0.12466432899236679,
      "learning_rate": 0.0002,
      "loss": 1.6567,
      "step": 93
    },
    {
      "epoch": 0.034306569343065696,
      "grad_norm": 0.11879820376634598,
      "learning_rate": 0.0002,
      "loss": 1.6812,
      "step": 94
    },
    {
      "epoch": 0.03467153284671533,
      "grad_norm": 0.12993916869163513,
      "learning_rate": 0.0002,
      "loss": 1.6826,
      "step": 95
    },
    {
      "epoch": 0.035036496350364967,
      "grad_norm": 0.11428281664848328,
      "learning_rate": 0.0002,
      "loss": 1.6718,
      "step": 96
    },
    {
      "epoch": 0.0354014598540146,
      "grad_norm": 0.12646667659282684,
      "learning_rate": 0.0002,
      "loss": 1.5986,
      "step": 97
    },
    {
      "epoch": 0.03576642335766423,
      "grad_norm": 0.11796025931835175,
      "learning_rate": 0.0002,
      "loss": 1.6826,
      "step": 98
    },
    {
      "epoch": 0.03613138686131387,
      "grad_norm": 0.13471877574920654,
      "learning_rate": 0.0002,
      "loss": 1.6348,
      "step": 99
    },
    {
      "epoch": 0.0364963503649635,
      "grad_norm": 0.12839972972869873,
      "learning_rate": 0.0002,
      "loss": 1.7218,
      "step": 100
    },
    {
      "epoch": 0.03686131386861314,
      "grad_norm": 0.11712581664323807,
      "learning_rate": 0.0002,
      "loss": 1.6728,
      "step": 101
    },
    {
      "epoch": 0.03722627737226277,
      "grad_norm": 0.1140546202659607,
      "learning_rate": 0.0002,
      "loss": 1.6563,
      "step": 102
    },
    {
      "epoch": 0.03759124087591241,
      "grad_norm": 0.12183567881584167,
      "learning_rate": 0.0002,
      "loss": 1.663,
      "step": 103
    },
    {
      "epoch": 0.03795620437956204,
      "grad_norm": 0.14103741943836212,
      "learning_rate": 0.0002,
      "loss": 1.7295,
      "step": 104
    },
    {
      "epoch": 0.03832116788321168,
      "grad_norm": 0.11011626571416855,
      "learning_rate": 0.0002,
      "loss": 1.5848,
      "step": 105
    },
    {
      "epoch": 0.03868613138686131,
      "grad_norm": 0.16225740313529968,
      "learning_rate": 0.0002,
      "loss": 1.6975,
      "step": 106
    },
    {
      "epoch": 0.03905109489051095,
      "grad_norm": 0.12195593118667603,
      "learning_rate": 0.0002,
      "loss": 1.6313,
      "step": 107
    },
    {
      "epoch": 0.03941605839416058,
      "grad_norm": 0.13149568438529968,
      "learning_rate": 0.0002,
      "loss": 1.6496,
      "step": 108
    },
    {
      "epoch": 0.03978102189781022,
      "grad_norm": 0.11954066157341003,
      "learning_rate": 0.0002,
      "loss": 1.6312,
      "step": 109
    },
    {
      "epoch": 0.040145985401459854,
      "grad_norm": 0.1461978256702423,
      "learning_rate": 0.0002,
      "loss": 1.6639,
      "step": 110
    },
    {
      "epoch": 0.04051094890510949,
      "grad_norm": 0.12941400706768036,
      "learning_rate": 0.0002,
      "loss": 1.675,
      "step": 111
    },
    {
      "epoch": 0.040875912408759124,
      "grad_norm": 0.11342988908290863,
      "learning_rate": 0.0002,
      "loss": 1.593,
      "step": 112
    },
    {
      "epoch": 0.041240875912408756,
      "grad_norm": 0.11693161725997925,
      "learning_rate": 0.0002,
      "loss": 1.7049,
      "step": 113
    },
    {
      "epoch": 0.041605839416058395,
      "grad_norm": 0.1243276372551918,
      "learning_rate": 0.0002,
      "loss": 1.666,
      "step": 114
    },
    {
      "epoch": 0.041970802919708027,
      "grad_norm": 0.12542955577373505,
      "learning_rate": 0.0002,
      "loss": 1.72,
      "step": 115
    },
    {
      "epoch": 0.042335766423357665,
      "grad_norm": 0.13715729117393494,
      "learning_rate": 0.0002,
      "loss": 1.6217,
      "step": 116
    },
    {
      "epoch": 0.0427007299270073,
      "grad_norm": 0.11925441771745682,
      "learning_rate": 0.0002,
      "loss": 1.6519,
      "step": 117
    },
    {
      "epoch": 0.043065693430656936,
      "grad_norm": 0.12592147290706635,
      "learning_rate": 0.0002,
      "loss": 1.6013,
      "step": 118
    },
    {
      "epoch": 0.04343065693430657,
      "grad_norm": 0.11545930802822113,
      "learning_rate": 0.0002,
      "loss": 1.6331,
      "step": 119
    },
    {
      "epoch": 0.043795620437956206,
      "grad_norm": 0.13424816727638245,
      "learning_rate": 0.0002,
      "loss": 1.5907,
      "step": 120
    },
    {
      "epoch": 0.04416058394160584,
      "grad_norm": 0.11689130961894989,
      "learning_rate": 0.0002,
      "loss": 1.6729,
      "step": 121
    },
    {
      "epoch": 0.04452554744525548,
      "grad_norm": 0.1326444000005722,
      "learning_rate": 0.0002,
      "loss": 1.6217,
      "step": 122
    },
    {
      "epoch": 0.04489051094890511,
      "grad_norm": 0.12135671079158783,
      "learning_rate": 0.0002,
      "loss": 1.6552,
      "step": 123
    },
    {
      "epoch": 0.04525547445255475,
      "grad_norm": 0.11764615029096603,
      "learning_rate": 0.0002,
      "loss": 1.6352,
      "step": 124
    },
    {
      "epoch": 0.04562043795620438,
      "grad_norm": 0.13443878293037415,
      "learning_rate": 0.0002,
      "loss": 1.6209,
      "step": 125
    },
    {
      "epoch": 0.04598540145985401,
      "grad_norm": 0.12052415311336517,
      "learning_rate": 0.0002,
      "loss": 1.6961,
      "step": 126
    },
    {
      "epoch": 0.04635036496350365,
      "grad_norm": 0.1330593228340149,
      "learning_rate": 0.0002,
      "loss": 1.6827,
      "step": 127
    },
    {
      "epoch": 0.04671532846715328,
      "grad_norm": 0.12354149669408798,
      "learning_rate": 0.0002,
      "loss": 1.7034,
      "step": 128
    },
    {
      "epoch": 0.04708029197080292,
      "grad_norm": 0.132972851395607,
      "learning_rate": 0.0002,
      "loss": 1.7028,
      "step": 129
    },
    {
      "epoch": 0.04744525547445255,
      "grad_norm": 0.12247125804424286,
      "learning_rate": 0.0002,
      "loss": 1.6168,
      "step": 130
    },
    {
      "epoch": 0.04781021897810219,
      "grad_norm": 0.11995014548301697,
      "learning_rate": 0.0002,
      "loss": 1.5748,
      "step": 131
    },
    {
      "epoch": 0.04817518248175182,
      "grad_norm": 0.11698607355356216,
      "learning_rate": 0.0002,
      "loss": 1.5708,
      "step": 132
    },
    {
      "epoch": 0.04854014598540146,
      "grad_norm": 0.12620186805725098,
      "learning_rate": 0.0002,
      "loss": 1.6245,
      "step": 133
    },
    {
      "epoch": 0.04890510948905109,
      "grad_norm": 0.1219867542386055,
      "learning_rate": 0.0002,
      "loss": 1.6171,
      "step": 134
    },
    {
      "epoch": 0.04927007299270073,
      "grad_norm": 0.11038133502006531,
      "learning_rate": 0.0002,
      "loss": 1.5365,
      "step": 135
    },
    {
      "epoch": 0.049635036496350364,
      "grad_norm": 0.11814630031585693,
      "learning_rate": 0.0002,
      "loss": 1.6574,
      "step": 136
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.12024957686662674,
      "learning_rate": 0.0002,
      "loss": 1.6861,
      "step": 137
    },
    {
      "epoch": 0.050364963503649635,
      "grad_norm": 0.11635082215070724,
      "learning_rate": 0.0002,
      "loss": 1.6249,
      "step": 138
    },
    {
      "epoch": 0.05072992700729927,
      "grad_norm": 0.12423109263181686,
      "learning_rate": 0.0002,
      "loss": 1.689,
      "step": 139
    },
    {
      "epoch": 0.051094890510948905,
      "grad_norm": 0.12072642147541046,
      "learning_rate": 0.0002,
      "loss": 1.6745,
      "step": 140
    },
    {
      "epoch": 0.05145985401459854,
      "grad_norm": 0.12270359694957733,
      "learning_rate": 0.0002,
      "loss": 1.6473,
      "step": 141
    },
    {
      "epoch": 0.051824817518248176,
      "grad_norm": 0.12165810167789459,
      "learning_rate": 0.0002,
      "loss": 1.6911,
      "step": 142
    },
    {
      "epoch": 0.05218978102189781,
      "grad_norm": 0.12689775228500366,
      "learning_rate": 0.0002,
      "loss": 1.629,
      "step": 143
    },
    {
      "epoch": 0.052554744525547446,
      "grad_norm": 0.12048137187957764,
      "learning_rate": 0.0002,
      "loss": 1.6412,
      "step": 144
    },
    {
      "epoch": 0.05291970802919708,
      "grad_norm": 0.12080144137144089,
      "learning_rate": 0.0002,
      "loss": 1.6775,
      "step": 145
    },
    {
      "epoch": 0.05328467153284672,
      "grad_norm": 0.11550788581371307,
      "learning_rate": 0.0002,
      "loss": 1.657,
      "step": 146
    },
    {
      "epoch": 0.05364963503649635,
      "grad_norm": 0.11445775628089905,
      "learning_rate": 0.0002,
      "loss": 1.6275,
      "step": 147
    },
    {
      "epoch": 0.05401459854014599,
      "grad_norm": 0.11275278031826019,
      "learning_rate": 0.0002,
      "loss": 1.5232,
      "step": 148
    },
    {
      "epoch": 0.05437956204379562,
      "grad_norm": 0.11367645114660263,
      "learning_rate": 0.0002,
      "loss": 1.5819,
      "step": 149
    },
    {
      "epoch": 0.05474452554744526,
      "grad_norm": 0.12075960636138916,
      "learning_rate": 0.0002,
      "loss": 1.7352,
      "step": 150
    },
    {
      "epoch": 0.05510948905109489,
      "grad_norm": 0.1164778470993042,
      "learning_rate": 0.0002,
      "loss": 1.6517,
      "step": 151
    },
    {
      "epoch": 0.05547445255474453,
      "grad_norm": 0.11737172305583954,
      "learning_rate": 0.0002,
      "loss": 1.6553,
      "step": 152
    },
    {
      "epoch": 0.05583941605839416,
      "grad_norm": 0.11728131026029587,
      "learning_rate": 0.0002,
      "loss": 1.6393,
      "step": 153
    },
    {
      "epoch": 0.05620437956204379,
      "grad_norm": 0.11871258914470673,
      "learning_rate": 0.0002,
      "loss": 1.6233,
      "step": 154
    },
    {
      "epoch": 0.05656934306569343,
      "grad_norm": 0.12492727488279343,
      "learning_rate": 0.0002,
      "loss": 1.6445,
      "step": 155
    },
    {
      "epoch": 0.05693430656934306,
      "grad_norm": 0.11382921040058136,
      "learning_rate": 0.0002,
      "loss": 1.6398,
      "step": 156
    },
    {
      "epoch": 0.0572992700729927,
      "grad_norm": 0.1339925229549408,
      "learning_rate": 0.0002,
      "loss": 1.7087,
      "step": 157
    },
    {
      "epoch": 0.05766423357664233,
      "grad_norm": 0.11956602334976196,
      "learning_rate": 0.0002,
      "loss": 1.5726,
      "step": 158
    },
    {
      "epoch": 0.05802919708029197,
      "grad_norm": 0.11748247593641281,
      "learning_rate": 0.0002,
      "loss": 1.6286,
      "step": 159
    },
    {
      "epoch": 0.058394160583941604,
      "grad_norm": 0.12338212132453918,
      "learning_rate": 0.0002,
      "loss": 1.6794,
      "step": 160
    },
    {
      "epoch": 0.05875912408759124,
      "grad_norm": 0.11834801733493805,
      "learning_rate": 0.0002,
      "loss": 1.668,
      "step": 161
    },
    {
      "epoch": 0.059124087591240874,
      "grad_norm": 0.1320778876543045,
      "learning_rate": 0.0002,
      "loss": 1.7479,
      "step": 162
    },
    {
      "epoch": 0.05948905109489051,
      "grad_norm": 0.11785445362329483,
      "learning_rate": 0.0002,
      "loss": 1.6111,
      "step": 163
    },
    {
      "epoch": 0.059854014598540145,
      "grad_norm": 0.1434011459350586,
      "learning_rate": 0.0002,
      "loss": 1.6283,
      "step": 164
    },
    {
      "epoch": 0.060218978102189784,
      "grad_norm": 0.11256599426269531,
      "learning_rate": 0.0002,
      "loss": 1.5934,
      "step": 165
    },
    {
      "epoch": 0.060583941605839416,
      "grad_norm": 0.11245762556791306,
      "learning_rate": 0.0002,
      "loss": 1.5825,
      "step": 166
    },
    {
      "epoch": 0.060948905109489054,
      "grad_norm": 0.1373375505208969,
      "learning_rate": 0.0002,
      "loss": 1.7148,
      "step": 167
    },
    {
      "epoch": 0.061313868613138686,
      "grad_norm": 0.11637628823518753,
      "learning_rate": 0.0002,
      "loss": 1.6123,
      "step": 168
    },
    {
      "epoch": 0.06167883211678832,
      "grad_norm": 0.13764281570911407,
      "learning_rate": 0.0002,
      "loss": 1.6396,
      "step": 169
    },
    {
      "epoch": 0.06204379562043796,
      "grad_norm": 0.11316309124231339,
      "learning_rate": 0.0002,
      "loss": 1.5679,
      "step": 170
    },
    {
      "epoch": 0.06240875912408759,
      "grad_norm": 0.1379183530807495,
      "learning_rate": 0.0002,
      "loss": 1.6464,
      "step": 171
    },
    {
      "epoch": 0.06277372262773723,
      "grad_norm": 0.12394940853118896,
      "learning_rate": 0.0002,
      "loss": 1.6212,
      "step": 172
    },
    {
      "epoch": 0.06313868613138686,
      "grad_norm": 0.14400620758533478,
      "learning_rate": 0.0002,
      "loss": 1.6683,
      "step": 173
    },
    {
      "epoch": 0.06350364963503649,
      "grad_norm": 0.12766240537166595,
      "learning_rate": 0.0002,
      "loss": 1.6804,
      "step": 174
    },
    {
      "epoch": 0.06386861313868614,
      "grad_norm": 0.12509246170520782,
      "learning_rate": 0.0002,
      "loss": 1.6617,
      "step": 175
    },
    {
      "epoch": 0.06423357664233577,
      "grad_norm": 0.11990375816822052,
      "learning_rate": 0.0002,
      "loss": 1.6391,
      "step": 176
    },
    {
      "epoch": 0.0645985401459854,
      "grad_norm": 0.11096862703561783,
      "learning_rate": 0.0002,
      "loss": 1.5013,
      "step": 177
    },
    {
      "epoch": 0.06496350364963503,
      "grad_norm": 0.12917746603488922,
      "learning_rate": 0.0002,
      "loss": 1.6549,
      "step": 178
    },
    {
      "epoch": 0.06532846715328468,
      "grad_norm": 0.12188570201396942,
      "learning_rate": 0.0002,
      "loss": 1.5994,
      "step": 179
    },
    {
      "epoch": 0.06569343065693431,
      "grad_norm": 0.1272428333759308,
      "learning_rate": 0.0002,
      "loss": 1.5493,
      "step": 180
    },
    {
      "epoch": 0.06605839416058394,
      "grad_norm": 0.11665729433298111,
      "learning_rate": 0.0002,
      "loss": 1.5915,
      "step": 181
    },
    {
      "epoch": 0.06642335766423357,
      "grad_norm": 0.1397407054901123,
      "learning_rate": 0.0002,
      "loss": 1.6569,
      "step": 182
    },
    {
      "epoch": 0.0667883211678832,
      "grad_norm": 0.13038963079452515,
      "learning_rate": 0.0002,
      "loss": 1.7703,
      "step": 183
    },
    {
      "epoch": 0.06715328467153285,
      "grad_norm": 0.13244028389453888,
      "learning_rate": 0.0002,
      "loss": 1.6495,
      "step": 184
    },
    {
      "epoch": 0.06751824817518248,
      "grad_norm": 0.1212984174489975,
      "learning_rate": 0.0002,
      "loss": 1.5804,
      "step": 185
    },
    {
      "epoch": 0.06788321167883211,
      "grad_norm": 0.14059199392795563,
      "learning_rate": 0.0002,
      "loss": 1.6633,
      "step": 186
    },
    {
      "epoch": 0.06824817518248175,
      "grad_norm": 0.12205739319324493,
      "learning_rate": 0.0002,
      "loss": 1.6186,
      "step": 187
    },
    {
      "epoch": 0.06861313868613139,
      "grad_norm": 0.1352696418762207,
      "learning_rate": 0.0002,
      "loss": 1.6352,
      "step": 188
    },
    {
      "epoch": 0.06897810218978102,
      "grad_norm": 0.11795849353075027,
      "learning_rate": 0.0002,
      "loss": 1.6392,
      "step": 189
    },
    {
      "epoch": 0.06934306569343066,
      "grad_norm": 0.13905206322669983,
      "learning_rate": 0.0002,
      "loss": 1.5881,
      "step": 190
    },
    {
      "epoch": 0.06970802919708029,
      "grad_norm": 0.12037964910268784,
      "learning_rate": 0.0002,
      "loss": 1.7036,
      "step": 191
    },
    {
      "epoch": 0.07007299270072993,
      "grad_norm": 0.1346055120229721,
      "learning_rate": 0.0002,
      "loss": 1.6547,
      "step": 192
    },
    {
      "epoch": 0.07043795620437956,
      "grad_norm": 0.12371893227100372,
      "learning_rate": 0.0002,
      "loss": 1.6899,
      "step": 193
    },
    {
      "epoch": 0.0708029197080292,
      "grad_norm": 0.14225134253501892,
      "learning_rate": 0.0002,
      "loss": 1.6335,
      "step": 194
    },
    {
      "epoch": 0.07116788321167883,
      "grad_norm": 0.11774294078350067,
      "learning_rate": 0.0002,
      "loss": 1.6116,
      "step": 195
    },
    {
      "epoch": 0.07153284671532846,
      "grad_norm": 0.11762867867946625,
      "learning_rate": 0.0002,
      "loss": 1.5792,
      "step": 196
    },
    {
      "epoch": 0.0718978102189781,
      "grad_norm": 0.13361147046089172,
      "learning_rate": 0.0002,
      "loss": 1.6677,
      "step": 197
    },
    {
      "epoch": 0.07226277372262774,
      "grad_norm": 0.11872056871652603,
      "learning_rate": 0.0002,
      "loss": 1.642,
      "step": 198
    },
    {
      "epoch": 0.07262773722627737,
      "grad_norm": 0.12665650248527527,
      "learning_rate": 0.0002,
      "loss": 1.6497,
      "step": 199
    },
    {
      "epoch": 0.072992700729927,
      "grad_norm": 0.12323873490095139,
      "learning_rate": 0.0002,
      "loss": 1.684,
      "step": 200
    },
    {
      "epoch": 0.07335766423357665,
      "grad_norm": 0.1252090185880661,
      "learning_rate": 0.0002,
      "loss": 1.6634,
      "step": 201
    },
    {
      "epoch": 0.07372262773722628,
      "grad_norm": 0.11640435457229614,
      "learning_rate": 0.0002,
      "loss": 1.6255,
      "step": 202
    },
    {
      "epoch": 0.07408759124087591,
      "grad_norm": 0.13391569256782532,
      "learning_rate": 0.0002,
      "loss": 1.6389,
      "step": 203
    },
    {
      "epoch": 0.07445255474452554,
      "grad_norm": 0.11862675100564957,
      "learning_rate": 0.0002,
      "loss": 1.6165,
      "step": 204
    },
    {
      "epoch": 0.07481751824817519,
      "grad_norm": 0.11248727887868881,
      "learning_rate": 0.0002,
      "loss": 1.5304,
      "step": 205
    },
    {
      "epoch": 0.07518248175182482,
      "grad_norm": 0.12208462506532669,
      "learning_rate": 0.0002,
      "loss": 1.6079,
      "step": 206
    },
    {
      "epoch": 0.07554744525547445,
      "grad_norm": 0.12427826225757599,
      "learning_rate": 0.0002,
      "loss": 1.6729,
      "step": 207
    },
    {
      "epoch": 0.07591240875912408,
      "grad_norm": 0.11961633712053299,
      "learning_rate": 0.0002,
      "loss": 1.7584,
      "step": 208
    },
    {
      "epoch": 0.07627737226277373,
      "grad_norm": 0.1133294552564621,
      "learning_rate": 0.0002,
      "loss": 1.6211,
      "step": 209
    },
    {
      "epoch": 0.07664233576642336,
      "grad_norm": 0.11330787092447281,
      "learning_rate": 0.0002,
      "loss": 1.5788,
      "step": 210
    },
    {
      "epoch": 0.07700729927007299,
      "grad_norm": 0.12008119374513626,
      "learning_rate": 0.0002,
      "loss": 1.6359,
      "step": 211
    },
    {
      "epoch": 0.07737226277372262,
      "grad_norm": 0.11228281259536743,
      "learning_rate": 0.0002,
      "loss": 1.5832,
      "step": 212
    },
    {
      "epoch": 0.07773722627737226,
      "grad_norm": 0.12557366490364075,
      "learning_rate": 0.0002,
      "loss": 1.6456,
      "step": 213
    },
    {
      "epoch": 0.0781021897810219,
      "grad_norm": 0.11967014521360397,
      "learning_rate": 0.0002,
      "loss": 1.6521,
      "step": 214
    },
    {
      "epoch": 0.07846715328467153,
      "grad_norm": 0.13786721229553223,
      "learning_rate": 0.0002,
      "loss": 1.6335,
      "step": 215
    },
    {
      "epoch": 0.07883211678832117,
      "grad_norm": 0.11887596547603607,
      "learning_rate": 0.0002,
      "loss": 1.6091,
      "step": 216
    },
    {
      "epoch": 0.0791970802919708,
      "grad_norm": 0.13055916130542755,
      "learning_rate": 0.0002,
      "loss": 1.5713,
      "step": 217
    },
    {
      "epoch": 0.07956204379562044,
      "grad_norm": 0.12640148401260376,
      "learning_rate": 0.0002,
      "loss": 1.6282,
      "step": 218
    },
    {
      "epoch": 0.07992700729927008,
      "grad_norm": 0.13849961757659912,
      "learning_rate": 0.0002,
      "loss": 1.5846,
      "step": 219
    },
    {
      "epoch": 0.08029197080291971,
      "grad_norm": 0.12171662598848343,
      "learning_rate": 0.0002,
      "loss": 1.677,
      "step": 220
    },
    {
      "epoch": 0.08065693430656934,
      "grad_norm": 0.1360357701778412,
      "learning_rate": 0.0002,
      "loss": 1.6284,
      "step": 221
    },
    {
      "epoch": 0.08102189781021898,
      "grad_norm": 0.12228655815124512,
      "learning_rate": 0.0002,
      "loss": 1.7018,
      "step": 222
    },
    {
      "epoch": 0.08138686131386862,
      "grad_norm": 0.1370847076177597,
      "learning_rate": 0.0002,
      "loss": 1.5757,
      "step": 223
    },
    {
      "epoch": 0.08175182481751825,
      "grad_norm": 0.12627586722373962,
      "learning_rate": 0.0002,
      "loss": 1.6777,
      "step": 224
    },
    {
      "epoch": 0.08211678832116788,
      "grad_norm": 0.13189387321472168,
      "learning_rate": 0.0002,
      "loss": 1.6399,
      "step": 225
    },
    {
      "epoch": 0.08248175182481751,
      "grad_norm": 0.1369723081588745,
      "learning_rate": 0.0002,
      "loss": 1.6396,
      "step": 226
    },
    {
      "epoch": 0.08284671532846716,
      "grad_norm": 0.1290339082479477,
      "learning_rate": 0.0002,
      "loss": 1.6397,
      "step": 227
    },
    {
      "epoch": 0.08321167883211679,
      "grad_norm": 0.1419927328824997,
      "learning_rate": 0.0002,
      "loss": 1.6799,
      "step": 228
    },
    {
      "epoch": 0.08357664233576642,
      "grad_norm": 0.11959163844585419,
      "learning_rate": 0.0002,
      "loss": 1.6488,
      "step": 229
    },
    {
      "epoch": 0.08394160583941605,
      "grad_norm": 0.12422080338001251,
      "learning_rate": 0.0002,
      "loss": 1.6698,
      "step": 230
    },
    {
      "epoch": 0.0843065693430657,
      "grad_norm": 0.11872342228889465,
      "learning_rate": 0.0002,
      "loss": 1.5616,
      "step": 231
    },
    {
      "epoch": 0.08467153284671533,
      "grad_norm": 0.1194247454404831,
      "learning_rate": 0.0002,
      "loss": 1.6418,
      "step": 232
    },
    {
      "epoch": 0.08503649635036496,
      "grad_norm": 0.11817412823438644,
      "learning_rate": 0.0002,
      "loss": 1.5835,
      "step": 233
    },
    {
      "epoch": 0.0854014598540146,
      "grad_norm": 0.12974007427692413,
      "learning_rate": 0.0002,
      "loss": 1.6791,
      "step": 234
    },
    {
      "epoch": 0.08576642335766424,
      "grad_norm": 0.11961016803979874,
      "learning_rate": 0.0002,
      "loss": 1.6231,
      "step": 235
    },
    {
      "epoch": 0.08613138686131387,
      "grad_norm": 0.12387136369943619,
      "learning_rate": 0.0002,
      "loss": 1.6065,
      "step": 236
    },
    {
      "epoch": 0.0864963503649635,
      "grad_norm": 0.12175223976373672,
      "learning_rate": 0.0002,
      "loss": 1.6052,
      "step": 237
    },
    {
      "epoch": 0.08686131386861314,
      "grad_norm": 0.12783817946910858,
      "learning_rate": 0.0002,
      "loss": 1.63,
      "step": 238
    },
    {
      "epoch": 0.08722627737226277,
      "grad_norm": 0.12543965876102448,
      "learning_rate": 0.0002,
      "loss": 1.7156,
      "step": 239
    },
    {
      "epoch": 0.08759124087591241,
      "grad_norm": 0.12341178953647614,
      "learning_rate": 0.0002,
      "loss": 1.6519,
      "step": 240
    },
    {
      "epoch": 0.08795620437956204,
      "grad_norm": 0.11857528239488602,
      "learning_rate": 0.0002,
      "loss": 1.6229,
      "step": 241
    },
    {
      "epoch": 0.08832116788321168,
      "grad_norm": 0.12251981347799301,
      "learning_rate": 0.0002,
      "loss": 1.7066,
      "step": 242
    },
    {
      "epoch": 0.08868613138686131,
      "grad_norm": 0.11930924654006958,
      "learning_rate": 0.0002,
      "loss": 1.5395,
      "step": 243
    },
    {
      "epoch": 0.08905109489051095,
      "grad_norm": 0.12633389234542847,
      "learning_rate": 0.0002,
      "loss": 1.7253,
      "step": 244
    },
    {
      "epoch": 0.08941605839416059,
      "grad_norm": 0.12682409584522247,
      "learning_rate": 0.0002,
      "loss": 1.5634,
      "step": 245
    },
    {
      "epoch": 0.08978102189781022,
      "grad_norm": 0.11788968741893768,
      "learning_rate": 0.0002,
      "loss": 1.5916,
      "step": 246
    },
    {
      "epoch": 0.09014598540145985,
      "grad_norm": 0.13407382369041443,
      "learning_rate": 0.0002,
      "loss": 1.6277,
      "step": 247
    },
    {
      "epoch": 0.0905109489051095,
      "grad_norm": 0.11324549466371536,
      "learning_rate": 0.0002,
      "loss": 1.5323,
      "step": 248
    },
    {
      "epoch": 0.09087591240875913,
      "grad_norm": 0.13209547102451324,
      "learning_rate": 0.0002,
      "loss": 1.7101,
      "step": 249
    },
    {
      "epoch": 0.09124087591240876,
      "grad_norm": 0.12488308548927307,
      "learning_rate": 0.0002,
      "loss": 1.6505,
      "step": 250
    },
    {
      "epoch": 0.09160583941605839,
      "grad_norm": 0.1370416134595871,
      "learning_rate": 0.0002,
      "loss": 1.6232,
      "step": 251
    },
    {
      "epoch": 0.09197080291970802,
      "grad_norm": 0.12230122834444046,
      "learning_rate": 0.0002,
      "loss": 1.721,
      "step": 252
    },
    {
      "epoch": 0.09233576642335767,
      "grad_norm": 0.12129965424537659,
      "learning_rate": 0.0002,
      "loss": 1.641,
      "step": 253
    },
    {
      "epoch": 0.0927007299270073,
      "grad_norm": 0.12389133870601654,
      "learning_rate": 0.0002,
      "loss": 1.617,
      "step": 254
    },
    {
      "epoch": 0.09306569343065693,
      "grad_norm": 0.12198037654161453,
      "learning_rate": 0.0002,
      "loss": 1.6696,
      "step": 255
    },
    {
      "epoch": 0.09343065693430656,
      "grad_norm": 0.12175638973712921,
      "learning_rate": 0.0002,
      "loss": 1.6385,
      "step": 256
    },
    {
      "epoch": 0.09379562043795621,
      "grad_norm": 0.11863414198160172,
      "learning_rate": 0.0002,
      "loss": 1.6908,
      "step": 257
    },
    {
      "epoch": 0.09416058394160584,
      "grad_norm": 0.12191036343574524,
      "learning_rate": 0.0002,
      "loss": 1.6952,
      "step": 258
    },
    {
      "epoch": 0.09452554744525547,
      "grad_norm": 0.11907047033309937,
      "learning_rate": 0.0002,
      "loss": 1.6363,
      "step": 259
    },
    {
      "epoch": 0.0948905109489051,
      "grad_norm": 0.1218395084142685,
      "learning_rate": 0.0002,
      "loss": 1.6128,
      "step": 260
    },
    {
      "epoch": 0.09525547445255475,
      "grad_norm": 0.11693360656499863,
      "learning_rate": 0.0002,
      "loss": 1.5374,
      "step": 261
    },
    {
      "epoch": 0.09562043795620438,
      "grad_norm": 0.11564788222312927,
      "learning_rate": 0.0002,
      "loss": 1.538,
      "step": 262
    },
    {
      "epoch": 0.09598540145985401,
      "grad_norm": 0.1199735701084137,
      "learning_rate": 0.0002,
      "loss": 1.6431,
      "step": 263
    },
    {
      "epoch": 0.09635036496350365,
      "grad_norm": 0.1216713935136795,
      "learning_rate": 0.0002,
      "loss": 1.6472,
      "step": 264
    },
    {
      "epoch": 0.09671532846715329,
      "grad_norm": 0.11472850292921066,
      "learning_rate": 0.0002,
      "loss": 1.5855,
      "step": 265
    },
    {
      "epoch": 0.09708029197080292,
      "grad_norm": 0.12120837718248367,
      "learning_rate": 0.0002,
      "loss": 1.6294,
      "step": 266
    },
    {
      "epoch": 0.09744525547445256,
      "grad_norm": 0.118916817009449,
      "learning_rate": 0.0002,
      "loss": 1.647,
      "step": 267
    },
    {
      "epoch": 0.09781021897810219,
      "grad_norm": 0.12737038731575012,
      "learning_rate": 0.0002,
      "loss": 1.7492,
      "step": 268
    },
    {
      "epoch": 0.09817518248175182,
      "grad_norm": 0.11635486036539078,
      "learning_rate": 0.0002,
      "loss": 1.6176,
      "step": 269
    },
    {
      "epoch": 0.09854014598540146,
      "grad_norm": 0.11855744570493698,
      "learning_rate": 0.0002,
      "loss": 1.6883,
      "step": 270
    },
    {
      "epoch": 0.0989051094890511,
      "grad_norm": 0.11844778060913086,
      "learning_rate": 0.0002,
      "loss": 1.6546,
      "step": 271
    },
    {
      "epoch": 0.09927007299270073,
      "grad_norm": 0.11695360392332077,
      "learning_rate": 0.0002,
      "loss": 1.5561,
      "step": 272
    },
    {
      "epoch": 0.09963503649635036,
      "grad_norm": 0.12796904146671295,
      "learning_rate": 0.0002,
      "loss": 1.6883,
      "step": 273
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.12001866847276688,
      "learning_rate": 0.0002,
      "loss": 1.6524,
      "step": 274
    },
    {
      "epoch": 0.10036496350364964,
      "grad_norm": 0.11756692826747894,
      "learning_rate": 0.0002,
      "loss": 1.5384,
      "step": 275
    },
    {
      "epoch": 0.10072992700729927,
      "grad_norm": 0.1221008375287056,
      "learning_rate": 0.0002,
      "loss": 1.6073,
      "step": 276
    },
    {
      "epoch": 0.1010948905109489,
      "grad_norm": 0.1234850063920021,
      "learning_rate": 0.0002,
      "loss": 1.5973,
      "step": 277
    },
    {
      "epoch": 0.10145985401459855,
      "grad_norm": 0.11612263321876526,
      "learning_rate": 0.0002,
      "loss": 1.5448,
      "step": 278
    },
    {
      "epoch": 0.10182481751824818,
      "grad_norm": 0.11482261121273041,
      "learning_rate": 0.0002,
      "loss": 1.5362,
      "step": 279
    },
    {
      "epoch": 0.10218978102189781,
      "grad_norm": 0.1227899119257927,
      "learning_rate": 0.0002,
      "loss": 1.5913,
      "step": 280
    },
    {
      "epoch": 0.10255474452554744,
      "grad_norm": 0.11733414232730865,
      "learning_rate": 0.0002,
      "loss": 1.6347,
      "step": 281
    },
    {
      "epoch": 0.10291970802919707,
      "grad_norm": 0.12448490411043167,
      "learning_rate": 0.0002,
      "loss": 1.6459,
      "step": 282
    },
    {
      "epoch": 0.10328467153284672,
      "grad_norm": 0.11743931472301483,
      "learning_rate": 0.0002,
      "loss": 1.6707,
      "step": 283
    },
    {
      "epoch": 0.10364963503649635,
      "grad_norm": 0.12791165709495544,
      "learning_rate": 0.0002,
      "loss": 1.6734,
      "step": 284
    },
    {
      "epoch": 0.10401459854014598,
      "grad_norm": 0.1253037005662918,
      "learning_rate": 0.0002,
      "loss": 1.6721,
      "step": 285
    },
    {
      "epoch": 0.10437956204379562,
      "grad_norm": 0.11901380121707916,
      "learning_rate": 0.0002,
      "loss": 1.5865,
      "step": 286
    },
    {
      "epoch": 0.10474452554744526,
      "grad_norm": 0.12020629644393921,
      "learning_rate": 0.0002,
      "loss": 1.6199,
      "step": 287
    },
    {
      "epoch": 0.10510948905109489,
      "grad_norm": 0.12800563871860504,
      "learning_rate": 0.0002,
      "loss": 1.5972,
      "step": 288
    },
    {
      "epoch": 0.10547445255474452,
      "grad_norm": 0.12339998781681061,
      "learning_rate": 0.0002,
      "loss": 1.5933,
      "step": 289
    },
    {
      "epoch": 0.10583941605839416,
      "grad_norm": 0.11812318861484528,
      "learning_rate": 0.0002,
      "loss": 1.633,
      "step": 290
    },
    {
      "epoch": 0.1062043795620438,
      "grad_norm": 0.11325936019420624,
      "learning_rate": 0.0002,
      "loss": 1.5521,
      "step": 291
    },
    {
      "epoch": 0.10656934306569343,
      "grad_norm": 0.1314803510904312,
      "learning_rate": 0.0002,
      "loss": 1.5925,
      "step": 292
    },
    {
      "epoch": 0.10693430656934307,
      "grad_norm": 0.12405530363321304,
      "learning_rate": 0.0002,
      "loss": 1.6646,
      "step": 293
    },
    {
      "epoch": 0.1072992700729927,
      "grad_norm": 0.12453662604093552,
      "learning_rate": 0.0002,
      "loss": 1.6052,
      "step": 294
    },
    {
      "epoch": 0.10766423357664233,
      "grad_norm": 0.11974222213029861,
      "learning_rate": 0.0002,
      "loss": 1.6117,
      "step": 295
    },
    {
      "epoch": 0.10802919708029197,
      "grad_norm": 0.11925089359283447,
      "learning_rate": 0.0002,
      "loss": 1.61,
      "step": 296
    },
    {
      "epoch": 0.1083941605839416,
      "grad_norm": 0.12501589953899384,
      "learning_rate": 0.0002,
      "loss": 1.6382,
      "step": 297
    },
    {
      "epoch": 0.10875912408759124,
      "grad_norm": 0.12182055413722992,
      "learning_rate": 0.0002,
      "loss": 1.5895,
      "step": 298
    },
    {
      "epoch": 0.10912408759124087,
      "grad_norm": 0.13116660714149475,
      "learning_rate": 0.0002,
      "loss": 1.5923,
      "step": 299
    },
    {
      "epoch": 0.10948905109489052,
      "grad_norm": 0.11703726649284363,
      "learning_rate": 0.0002,
      "loss": 1.5461,
      "step": 300
    },
    {
      "epoch": 0.10985401459854015,
      "grad_norm": 0.1421850323677063,
      "learning_rate": 0.0002,
      "loss": 1.5613,
      "step": 301
    },
    {
      "epoch": 0.11021897810218978,
      "grad_norm": 0.1327492594718933,
      "learning_rate": 0.0002,
      "loss": 1.616,
      "step": 302
    },
    {
      "epoch": 0.11058394160583941,
      "grad_norm": 0.13673241436481476,
      "learning_rate": 0.0002,
      "loss": 1.605,
      "step": 303
    },
    {
      "epoch": 0.11094890510948906,
      "grad_norm": 0.1417289525270462,
      "learning_rate": 0.0002,
      "loss": 1.6445,
      "step": 304
    },
    {
      "epoch": 0.11131386861313869,
      "grad_norm": 0.13772191107273102,
      "learning_rate": 0.0002,
      "loss": 1.6688,
      "step": 305
    },
    {
      "epoch": 0.11167883211678832,
      "grad_norm": 0.11876555532217026,
      "learning_rate": 0.0002,
      "loss": 1.5819,
      "step": 306
    },
    {
      "epoch": 0.11204379562043795,
      "grad_norm": 0.12445658445358276,
      "learning_rate": 0.0002,
      "loss": 1.504,
      "step": 307
    },
    {
      "epoch": 0.11240875912408758,
      "grad_norm": 0.12812760472297668,
      "learning_rate": 0.0002,
      "loss": 1.6091,
      "step": 308
    },
    {
      "epoch": 0.11277372262773723,
      "grad_norm": 0.1273098737001419,
      "learning_rate": 0.0002,
      "loss": 1.6773,
      "step": 309
    },
    {
      "epoch": 0.11313868613138686,
      "grad_norm": 0.1302049309015274,
      "learning_rate": 0.0002,
      "loss": 1.619,
      "step": 310
    },
    {
      "epoch": 0.1135036496350365,
      "grad_norm": 0.13052499294281006,
      "learning_rate": 0.0002,
      "loss": 1.6132,
      "step": 311
    },
    {
      "epoch": 0.11386861313868613,
      "grad_norm": 0.12880726158618927,
      "learning_rate": 0.0002,
      "loss": 1.6362,
      "step": 312
    },
    {
      "epoch": 0.11423357664233577,
      "grad_norm": 0.13387270271778107,
      "learning_rate": 0.0002,
      "loss": 1.6109,
      "step": 313
    },
    {
      "epoch": 0.1145985401459854,
      "grad_norm": 0.1276913434267044,
      "learning_rate": 0.0002,
      "loss": 1.6827,
      "step": 314
    },
    {
      "epoch": 0.11496350364963503,
      "grad_norm": 0.1302744597196579,
      "learning_rate": 0.0002,
      "loss": 1.6233,
      "step": 315
    },
    {
      "epoch": 0.11532846715328467,
      "grad_norm": 0.1209658607840538,
      "learning_rate": 0.0002,
      "loss": 1.688,
      "step": 316
    },
    {
      "epoch": 0.11569343065693431,
      "grad_norm": 0.12607784569263458,
      "learning_rate": 0.0002,
      "loss": 1.6436,
      "step": 317
    },
    {
      "epoch": 0.11605839416058394,
      "grad_norm": 0.1240229457616806,
      "learning_rate": 0.0002,
      "loss": 1.6267,
      "step": 318
    },
    {
      "epoch": 0.11642335766423358,
      "grad_norm": 0.12903936207294464,
      "learning_rate": 0.0002,
      "loss": 1.6097,
      "step": 319
    },
    {
      "epoch": 0.11678832116788321,
      "grad_norm": 0.11992821842432022,
      "learning_rate": 0.0002,
      "loss": 1.5992,
      "step": 320
    },
    {
      "epoch": 0.11715328467153285,
      "grad_norm": 0.1317678540945053,
      "learning_rate": 0.0002,
      "loss": 1.6073,
      "step": 321
    },
    {
      "epoch": 0.11751824817518249,
      "grad_norm": 0.11902587115764618,
      "learning_rate": 0.0002,
      "loss": 1.5888,
      "step": 322
    },
    {
      "epoch": 0.11788321167883212,
      "grad_norm": 0.11648491770029068,
      "learning_rate": 0.0002,
      "loss": 1.5993,
      "step": 323
    },
    {
      "epoch": 0.11824817518248175,
      "grad_norm": 0.12216935306787491,
      "learning_rate": 0.0002,
      "loss": 1.5889,
      "step": 324
    },
    {
      "epoch": 0.11861313868613138,
      "grad_norm": 0.12259191274642944,
      "learning_rate": 0.0002,
      "loss": 1.6071,
      "step": 325
    },
    {
      "epoch": 0.11897810218978103,
      "grad_norm": 0.1317819207906723,
      "learning_rate": 0.0002,
      "loss": 1.6512,
      "step": 326
    },
    {
      "epoch": 0.11934306569343066,
      "grad_norm": 0.12047944217920303,
      "learning_rate": 0.0002,
      "loss": 1.5807,
      "step": 327
    },
    {
      "epoch": 0.11970802919708029,
      "grad_norm": 0.12732666730880737,
      "learning_rate": 0.0002,
      "loss": 1.6079,
      "step": 328
    },
    {
      "epoch": 0.12007299270072992,
      "grad_norm": 0.11920438706874847,
      "learning_rate": 0.0002,
      "loss": 1.6526,
      "step": 329
    },
    {
      "epoch": 0.12043795620437957,
      "grad_norm": 0.12505806982517242,
      "learning_rate": 0.0002,
      "loss": 1.7459,
      "step": 330
    },
    {
      "epoch": 0.1208029197080292,
      "grad_norm": 0.12838099896907806,
      "learning_rate": 0.0002,
      "loss": 1.6387,
      "step": 331
    },
    {
      "epoch": 0.12116788321167883,
      "grad_norm": 0.12129881232976913,
      "learning_rate": 0.0002,
      "loss": 1.6692,
      "step": 332
    },
    {
      "epoch": 0.12153284671532846,
      "grad_norm": 0.12050154060125351,
      "learning_rate": 0.0002,
      "loss": 1.6236,
      "step": 333
    },
    {
      "epoch": 0.12189781021897811,
      "grad_norm": 0.12132157385349274,
      "learning_rate": 0.0002,
      "loss": 1.6201,
      "step": 334
    },
    {
      "epoch": 0.12226277372262774,
      "grad_norm": 0.11917468160390854,
      "learning_rate": 0.0002,
      "loss": 1.5816,
      "step": 335
    },
    {
      "epoch": 0.12262773722627737,
      "grad_norm": 0.12340982258319855,
      "learning_rate": 0.0002,
      "loss": 1.6922,
      "step": 336
    },
    {
      "epoch": 0.122992700729927,
      "grad_norm": 0.12510235607624054,
      "learning_rate": 0.0002,
      "loss": 1.6354,
      "step": 337
    },
    {
      "epoch": 0.12335766423357664,
      "grad_norm": 0.12469889968633652,
      "learning_rate": 0.0002,
      "loss": 1.6216,
      "step": 338
    },
    {
      "epoch": 0.12372262773722628,
      "grad_norm": 0.12034323066473007,
      "learning_rate": 0.0002,
      "loss": 1.6245,
      "step": 339
    },
    {
      "epoch": 0.12408759124087591,
      "grad_norm": 0.12086813896894455,
      "learning_rate": 0.0002,
      "loss": 1.6281,
      "step": 340
    },
    {
      "epoch": 0.12445255474452555,
      "grad_norm": 0.11712285876274109,
      "learning_rate": 0.0002,
      "loss": 1.5722,
      "step": 341
    },
    {
      "epoch": 0.12481751824817518,
      "grad_norm": 0.12091507762670517,
      "learning_rate": 0.0002,
      "loss": 1.6108,
      "step": 342
    },
    {
      "epoch": 0.1251824817518248,
      "grad_norm": 0.12395089119672775,
      "learning_rate": 0.0002,
      "loss": 1.6944,
      "step": 343
    },
    {
      "epoch": 0.12554744525547445,
      "grad_norm": 0.11870311945676804,
      "learning_rate": 0.0002,
      "loss": 1.6058,
      "step": 344
    },
    {
      "epoch": 0.1259124087591241,
      "grad_norm": 0.11780521273612976,
      "learning_rate": 0.0002,
      "loss": 1.5643,
      "step": 345
    },
    {
      "epoch": 0.12627737226277372,
      "grad_norm": 0.11802096664905548,
      "learning_rate": 0.0002,
      "loss": 1.5703,
      "step": 346
    },
    {
      "epoch": 0.12664233576642336,
      "grad_norm": 0.11512330919504166,
      "learning_rate": 0.0002,
      "loss": 1.5437,
      "step": 347
    },
    {
      "epoch": 0.12700729927007298,
      "grad_norm": 0.12215406447649002,
      "learning_rate": 0.0002,
      "loss": 1.6841,
      "step": 348
    },
    {
      "epoch": 0.12737226277372263,
      "grad_norm": 0.12097284942865372,
      "learning_rate": 0.0002,
      "loss": 1.6529,
      "step": 349
    },
    {
      "epoch": 0.12773722627737227,
      "grad_norm": 0.11645885556936264,
      "learning_rate": 0.0002,
      "loss": 1.5663,
      "step": 350
    },
    {
      "epoch": 0.1281021897810219,
      "grad_norm": 0.1289779245853424,
      "learning_rate": 0.0002,
      "loss": 1.6553,
      "step": 351
    },
    {
      "epoch": 0.12846715328467154,
      "grad_norm": 0.12239154428243637,
      "learning_rate": 0.0002,
      "loss": 1.6308,
      "step": 352
    },
    {
      "epoch": 0.12883211678832115,
      "grad_norm": 0.13066482543945312,
      "learning_rate": 0.0002,
      "loss": 1.6972,
      "step": 353
    },
    {
      "epoch": 0.1291970802919708,
      "grad_norm": 0.12593978643417358,
      "learning_rate": 0.0002,
      "loss": 1.6843,
      "step": 354
    },
    {
      "epoch": 0.12956204379562045,
      "grad_norm": 0.12491950392723083,
      "learning_rate": 0.0002,
      "loss": 1.6381,
      "step": 355
    },
    {
      "epoch": 0.12992700729927006,
      "grad_norm": 0.13698363304138184,
      "learning_rate": 0.0002,
      "loss": 1.6531,
      "step": 356
    },
    {
      "epoch": 0.1302919708029197,
      "grad_norm": 0.12753282487392426,
      "learning_rate": 0.0002,
      "loss": 1.6197,
      "step": 357
    },
    {
      "epoch": 0.13065693430656936,
      "grad_norm": 0.13995888829231262,
      "learning_rate": 0.0002,
      "loss": 1.6636,
      "step": 358
    },
    {
      "epoch": 0.13102189781021897,
      "grad_norm": 0.11318392306566238,
      "learning_rate": 0.0002,
      "loss": 1.5466,
      "step": 359
    },
    {
      "epoch": 0.13138686131386862,
      "grad_norm": 0.12684592604637146,
      "learning_rate": 0.0002,
      "loss": 1.6467,
      "step": 360
    },
    {
      "epoch": 0.13175182481751824,
      "grad_norm": 0.11644352227449417,
      "learning_rate": 0.0002,
      "loss": 1.5828,
      "step": 361
    },
    {
      "epoch": 0.13211678832116788,
      "grad_norm": 0.12557831406593323,
      "learning_rate": 0.0002,
      "loss": 1.5818,
      "step": 362
    },
    {
      "epoch": 0.13248175182481753,
      "grad_norm": 0.12232189625501633,
      "learning_rate": 0.0002,
      "loss": 1.6495,
      "step": 363
    },
    {
      "epoch": 0.13284671532846715,
      "grad_norm": 0.12774783372879028,
      "learning_rate": 0.0002,
      "loss": 1.6494,
      "step": 364
    },
    {
      "epoch": 0.1332116788321168,
      "grad_norm": 0.12806107103824615,
      "learning_rate": 0.0002,
      "loss": 1.6503,
      "step": 365
    },
    {
      "epoch": 0.1335766423357664,
      "grad_norm": 0.1262645572423935,
      "learning_rate": 0.0002,
      "loss": 1.6018,
      "step": 366
    },
    {
      "epoch": 0.13394160583941606,
      "grad_norm": 0.13683992624282837,
      "learning_rate": 0.0002,
      "loss": 1.5951,
      "step": 367
    },
    {
      "epoch": 0.1343065693430657,
      "grad_norm": 0.11792343109846115,
      "learning_rate": 0.0002,
      "loss": 1.5344,
      "step": 368
    },
    {
      "epoch": 0.13467153284671532,
      "grad_norm": 0.14527153968811035,
      "learning_rate": 0.0002,
      "loss": 1.5948,
      "step": 369
    },
    {
      "epoch": 0.13503649635036497,
      "grad_norm": 0.12317585200071335,
      "learning_rate": 0.0002,
      "loss": 1.5974,
      "step": 370
    },
    {
      "epoch": 0.1354014598540146,
      "grad_norm": 0.12036976963281631,
      "learning_rate": 0.0002,
      "loss": 1.5667,
      "step": 371
    },
    {
      "epoch": 0.13576642335766423,
      "grad_norm": 0.12371034175157547,
      "learning_rate": 0.0002,
      "loss": 1.5849,
      "step": 372
    },
    {
      "epoch": 0.13613138686131387,
      "grad_norm": 0.12332010269165039,
      "learning_rate": 0.0002,
      "loss": 1.6077,
      "step": 373
    },
    {
      "epoch": 0.1364963503649635,
      "grad_norm": 0.12457305938005447,
      "learning_rate": 0.0002,
      "loss": 1.5828,
      "step": 374
    },
    {
      "epoch": 0.13686131386861314,
      "grad_norm": 0.12718334794044495,
      "learning_rate": 0.0002,
      "loss": 1.6866,
      "step": 375
    },
    {
      "epoch": 0.13722627737226278,
      "grad_norm": 0.12438328564167023,
      "learning_rate": 0.0002,
      "loss": 1.632,
      "step": 376
    },
    {
      "epoch": 0.1375912408759124,
      "grad_norm": 0.13365688920021057,
      "learning_rate": 0.0002,
      "loss": 1.6558,
      "step": 377
    },
    {
      "epoch": 0.13795620437956205,
      "grad_norm": 0.12237947434186935,
      "learning_rate": 0.0002,
      "loss": 1.581,
      "step": 378
    },
    {
      "epoch": 0.13832116788321167,
      "grad_norm": 0.1396224945783615,
      "learning_rate": 0.0002,
      "loss": 1.6992,
      "step": 379
    },
    {
      "epoch": 0.1386861313868613,
      "grad_norm": 0.12061601132154465,
      "learning_rate": 0.0002,
      "loss": 1.6394,
      "step": 380
    },
    {
      "epoch": 0.13905109489051096,
      "grad_norm": 0.13178183138370514,
      "learning_rate": 0.0002,
      "loss": 1.7124,
      "step": 381
    },
    {
      "epoch": 0.13941605839416057,
      "grad_norm": 0.12406151741743088,
      "learning_rate": 0.0002,
      "loss": 1.638,
      "step": 382
    },
    {
      "epoch": 0.13978102189781022,
      "grad_norm": 0.12082075327634811,
      "learning_rate": 0.0002,
      "loss": 1.5781,
      "step": 383
    },
    {
      "epoch": 0.14014598540145987,
      "grad_norm": 0.13526132702827454,
      "learning_rate": 0.0002,
      "loss": 1.6054,
      "step": 384
    },
    {
      "epoch": 0.14051094890510948,
      "grad_norm": 0.1378253996372223,
      "learning_rate": 0.0002,
      "loss": 1.6284,
      "step": 385
    },
    {
      "epoch": 0.14087591240875913,
      "grad_norm": 0.13242249190807343,
      "learning_rate": 0.0002,
      "loss": 1.6762,
      "step": 386
    },
    {
      "epoch": 0.14124087591240875,
      "grad_norm": 0.1238362044095993,
      "learning_rate": 0.0002,
      "loss": 1.6889,
      "step": 387
    },
    {
      "epoch": 0.1416058394160584,
      "grad_norm": 0.11768706142902374,
      "learning_rate": 0.0002,
      "loss": 1.5647,
      "step": 388
    },
    {
      "epoch": 0.14197080291970804,
      "grad_norm": 0.12092480808496475,
      "learning_rate": 0.0002,
      "loss": 1.5401,
      "step": 389
    },
    {
      "epoch": 0.14233576642335766,
      "grad_norm": 0.1247277557849884,
      "learning_rate": 0.0002,
      "loss": 1.6637,
      "step": 390
    },
    {
      "epoch": 0.1427007299270073,
      "grad_norm": 0.1271093487739563,
      "learning_rate": 0.0002,
      "loss": 1.6808,
      "step": 391
    },
    {
      "epoch": 0.14306569343065692,
      "grad_norm": 0.13181185722351074,
      "learning_rate": 0.0002,
      "loss": 1.5857,
      "step": 392
    },
    {
      "epoch": 0.14343065693430657,
      "grad_norm": 0.1323055475950241,
      "learning_rate": 0.0002,
      "loss": 1.5976,
      "step": 393
    },
    {
      "epoch": 0.1437956204379562,
      "grad_norm": 0.12518663704395294,
      "learning_rate": 0.0002,
      "loss": 1.6547,
      "step": 394
    },
    {
      "epoch": 0.14416058394160583,
      "grad_norm": 0.12295523285865784,
      "learning_rate": 0.0002,
      "loss": 1.5714,
      "step": 395
    },
    {
      "epoch": 0.14452554744525548,
      "grad_norm": 0.13045430183410645,
      "learning_rate": 0.0002,
      "loss": 1.6287,
      "step": 396
    },
    {
      "epoch": 0.14489051094890512,
      "grad_norm": 0.12984852492809296,
      "learning_rate": 0.0002,
      "loss": 1.6506,
      "step": 397
    },
    {
      "epoch": 0.14525547445255474,
      "grad_norm": 0.12663596868515015,
      "learning_rate": 0.0002,
      "loss": 1.6459,
      "step": 398
    },
    {
      "epoch": 0.14562043795620438,
      "grad_norm": 0.15723878145217896,
      "learning_rate": 0.0002,
      "loss": 1.7081,
      "step": 399
    },
    {
      "epoch": 0.145985401459854,
      "grad_norm": 0.12263409793376923,
      "learning_rate": 0.0002,
      "loss": 1.6117,
      "step": 400
    },
    {
      "epoch": 0.14635036496350365,
      "grad_norm": 0.154331237077713,
      "learning_rate": 0.0002,
      "loss": 1.6825,
      "step": 401
    },
    {
      "epoch": 0.1467153284671533,
      "grad_norm": 0.13326396048069,
      "learning_rate": 0.0002,
      "loss": 1.6244,
      "step": 402
    },
    {
      "epoch": 0.1470802919708029,
      "grad_norm": 0.13876895606517792,
      "learning_rate": 0.0002,
      "loss": 1.6109,
      "step": 403
    },
    {
      "epoch": 0.14744525547445256,
      "grad_norm": 0.12743458151817322,
      "learning_rate": 0.0002,
      "loss": 1.5952,
      "step": 404
    },
    {
      "epoch": 0.1478102189781022,
      "grad_norm": 0.15646688640117645,
      "learning_rate": 0.0002,
      "loss": 1.5812,
      "step": 405
    },
    {
      "epoch": 0.14817518248175182,
      "grad_norm": 0.12506787478923798,
      "learning_rate": 0.0002,
      "loss": 1.5713,
      "step": 406
    },
    {
      "epoch": 0.14854014598540147,
      "grad_norm": 0.13674940168857574,
      "learning_rate": 0.0002,
      "loss": 1.633,
      "step": 407
    },
    {
      "epoch": 0.14890510948905109,
      "grad_norm": 0.12333187460899353,
      "learning_rate": 0.0002,
      "loss": 1.6207,
      "step": 408
    },
    {
      "epoch": 0.14927007299270073,
      "grad_norm": 0.13657552003860474,
      "learning_rate": 0.0002,
      "loss": 1.638,
      "step": 409
    },
    {
      "epoch": 0.14963503649635038,
      "grad_norm": 0.1294127106666565,
      "learning_rate": 0.0002,
      "loss": 1.6221,
      "step": 410
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.126207634806633,
      "learning_rate": 0.0002,
      "loss": 1.6235,
      "step": 411
    },
    {
      "epoch": 0.15036496350364964,
      "grad_norm": 0.1317881941795349,
      "learning_rate": 0.0002,
      "loss": 1.5788,
      "step": 412
    },
    {
      "epoch": 0.15072992700729926,
      "grad_norm": 0.12524400651454926,
      "learning_rate": 0.0002,
      "loss": 1.6088,
      "step": 413
    },
    {
      "epoch": 0.1510948905109489,
      "grad_norm": 0.14089442789554596,
      "learning_rate": 0.0002,
      "loss": 1.5292,
      "step": 414
    },
    {
      "epoch": 0.15145985401459855,
      "grad_norm": 0.11884158849716187,
      "learning_rate": 0.0002,
      "loss": 1.5471,
      "step": 415
    },
    {
      "epoch": 0.15182481751824817,
      "grad_norm": 0.12964744865894318,
      "learning_rate": 0.0002,
      "loss": 1.5829,
      "step": 416
    },
    {
      "epoch": 0.1521897810218978,
      "grad_norm": 0.12144120037555695,
      "learning_rate": 0.0002,
      "loss": 1.5805,
      "step": 417
    },
    {
      "epoch": 0.15255474452554746,
      "grad_norm": 0.13578933477401733,
      "learning_rate": 0.0002,
      "loss": 1.6015,
      "step": 418
    },
    {
      "epoch": 0.15291970802919708,
      "grad_norm": 0.12237313389778137,
      "learning_rate": 0.0002,
      "loss": 1.6,
      "step": 419
    },
    {
      "epoch": 0.15328467153284672,
      "grad_norm": 0.12510788440704346,
      "learning_rate": 0.0002,
      "loss": 1.5866,
      "step": 420
    },
    {
      "epoch": 0.15364963503649634,
      "grad_norm": 0.14027303457260132,
      "learning_rate": 0.0002,
      "loss": 1.5648,
      "step": 421
    },
    {
      "epoch": 0.15401459854014599,
      "grad_norm": 0.1318199187517166,
      "learning_rate": 0.0002,
      "loss": 1.6238,
      "step": 422
    },
    {
      "epoch": 0.15437956204379563,
      "grad_norm": 0.12672249972820282,
      "learning_rate": 0.0002,
      "loss": 1.5874,
      "step": 423
    },
    {
      "epoch": 0.15474452554744525,
      "grad_norm": 0.12423631548881531,
      "learning_rate": 0.0002,
      "loss": 1.6558,
      "step": 424
    },
    {
      "epoch": 0.1551094890510949,
      "grad_norm": 0.13504047691822052,
      "learning_rate": 0.0002,
      "loss": 1.6674,
      "step": 425
    },
    {
      "epoch": 0.1554744525547445,
      "grad_norm": 0.11847669631242752,
      "learning_rate": 0.0002,
      "loss": 1.5164,
      "step": 426
    },
    {
      "epoch": 0.15583941605839416,
      "grad_norm": 0.149508997797966,
      "learning_rate": 0.0002,
      "loss": 1.7467,
      "step": 427
    },
    {
      "epoch": 0.1562043795620438,
      "grad_norm": 0.12399760633707047,
      "learning_rate": 0.0002,
      "loss": 1.556,
      "step": 428
    },
    {
      "epoch": 0.15656934306569342,
      "grad_norm": 0.14854802191257477,
      "learning_rate": 0.0002,
      "loss": 1.7004,
      "step": 429
    },
    {
      "epoch": 0.15693430656934307,
      "grad_norm": 0.12963946163654327,
      "learning_rate": 0.0002,
      "loss": 1.6211,
      "step": 430
    },
    {
      "epoch": 0.15729927007299271,
      "grad_norm": 0.1308518499135971,
      "learning_rate": 0.0002,
      "loss": 1.5677,
      "step": 431
    },
    {
      "epoch": 0.15766423357664233,
      "grad_norm": 0.11560183763504028,
      "learning_rate": 0.0002,
      "loss": 1.5944,
      "step": 432
    },
    {
      "epoch": 0.15802919708029198,
      "grad_norm": 0.1248599961400032,
      "learning_rate": 0.0002,
      "loss": 1.6219,
      "step": 433
    },
    {
      "epoch": 0.1583941605839416,
      "grad_norm": 0.13016974925994873,
      "learning_rate": 0.0002,
      "loss": 1.6203,
      "step": 434
    },
    {
      "epoch": 0.15875912408759124,
      "grad_norm": 0.12397754192352295,
      "learning_rate": 0.0002,
      "loss": 1.6428,
      "step": 435
    },
    {
      "epoch": 0.1591240875912409,
      "grad_norm": 0.12669001519680023,
      "learning_rate": 0.0002,
      "loss": 1.6122,
      "step": 436
    },
    {
      "epoch": 0.1594890510948905,
      "grad_norm": 0.1233924850821495,
      "learning_rate": 0.0002,
      "loss": 1.6619,
      "step": 437
    },
    {
      "epoch": 0.15985401459854015,
      "grad_norm": 0.12964662909507751,
      "learning_rate": 0.0002,
      "loss": 1.6966,
      "step": 438
    },
    {
      "epoch": 0.16021897810218977,
      "grad_norm": 0.1253441572189331,
      "learning_rate": 0.0002,
      "loss": 1.5866,
      "step": 439
    },
    {
      "epoch": 0.16058394160583941,
      "grad_norm": 0.12670013308525085,
      "learning_rate": 0.0002,
      "loss": 1.5553,
      "step": 440
    },
    {
      "epoch": 0.16094890510948906,
      "grad_norm": 0.1331230252981186,
      "learning_rate": 0.0002,
      "loss": 1.5895,
      "step": 441
    },
    {
      "epoch": 0.16131386861313868,
      "grad_norm": 0.12280143052339554,
      "learning_rate": 0.0002,
      "loss": 1.5499,
      "step": 442
    },
    {
      "epoch": 0.16167883211678832,
      "grad_norm": 0.13086159527301788,
      "learning_rate": 0.0002,
      "loss": 1.6711,
      "step": 443
    },
    {
      "epoch": 0.16204379562043797,
      "grad_norm": 0.1222786009311676,
      "learning_rate": 0.0002,
      "loss": 1.6114,
      "step": 444
    },
    {
      "epoch": 0.1624087591240876,
      "grad_norm": 0.125692680478096,
      "learning_rate": 0.0002,
      "loss": 1.6215,
      "step": 445
    },
    {
      "epoch": 0.16277372262773723,
      "grad_norm": 0.12965956330299377,
      "learning_rate": 0.0002,
      "loss": 1.6547,
      "step": 446
    },
    {
      "epoch": 0.16313868613138685,
      "grad_norm": 0.1272580623626709,
      "learning_rate": 0.0002,
      "loss": 1.6528,
      "step": 447
    },
    {
      "epoch": 0.1635036496350365,
      "grad_norm": 0.12501932680606842,
      "learning_rate": 0.0002,
      "loss": 1.6243,
      "step": 448
    },
    {
      "epoch": 0.16386861313868614,
      "grad_norm": 0.12629689276218414,
      "learning_rate": 0.0002,
      "loss": 1.6011,
      "step": 449
    },
    {
      "epoch": 0.16423357664233576,
      "grad_norm": 0.11944343894720078,
      "learning_rate": 0.0002,
      "loss": 1.5884,
      "step": 450
    },
    {
      "epoch": 0.1645985401459854,
      "grad_norm": 0.13024847209453583,
      "learning_rate": 0.0002,
      "loss": 1.567,
      "step": 451
    },
    {
      "epoch": 0.16496350364963502,
      "grad_norm": 0.12284905463457108,
      "learning_rate": 0.0002,
      "loss": 1.6044,
      "step": 452
    },
    {
      "epoch": 0.16532846715328467,
      "grad_norm": 0.12351765483617783,
      "learning_rate": 0.0002,
      "loss": 1.6419,
      "step": 453
    },
    {
      "epoch": 0.16569343065693432,
      "grad_norm": 0.1260727494955063,
      "learning_rate": 0.0002,
      "loss": 1.6406,
      "step": 454
    },
    {
      "epoch": 0.16605839416058393,
      "grad_norm": 0.12076357752084732,
      "learning_rate": 0.0002,
      "loss": 1.58,
      "step": 455
    },
    {
      "epoch": 0.16642335766423358,
      "grad_norm": 0.125578373670578,
      "learning_rate": 0.0002,
      "loss": 1.6217,
      "step": 456
    },
    {
      "epoch": 0.16678832116788322,
      "grad_norm": 0.1212897002696991,
      "learning_rate": 0.0002,
      "loss": 1.5974,
      "step": 457
    },
    {
      "epoch": 0.16715328467153284,
      "grad_norm": 0.12076254934072495,
      "learning_rate": 0.0002,
      "loss": 1.6163,
      "step": 458
    },
    {
      "epoch": 0.1675182481751825,
      "grad_norm": 0.1262521594762802,
      "learning_rate": 0.0002,
      "loss": 1.6372,
      "step": 459
    },
    {
      "epoch": 0.1678832116788321,
      "grad_norm": 0.12309537082910538,
      "learning_rate": 0.0002,
      "loss": 1.6165,
      "step": 460
    },
    {
      "epoch": 0.16824817518248175,
      "grad_norm": 0.1166430413722992,
      "learning_rate": 0.0002,
      "loss": 1.5147,
      "step": 461
    },
    {
      "epoch": 0.1686131386861314,
      "grad_norm": 0.11966414004564285,
      "learning_rate": 0.0002,
      "loss": 1.6144,
      "step": 462
    },
    {
      "epoch": 0.16897810218978102,
      "grad_norm": 0.12412495911121368,
      "learning_rate": 0.0002,
      "loss": 1.5889,
      "step": 463
    },
    {
      "epoch": 0.16934306569343066,
      "grad_norm": 0.13024382293224335,
      "learning_rate": 0.0002,
      "loss": 1.705,
      "step": 464
    },
    {
      "epoch": 0.16970802919708028,
      "grad_norm": 0.12120583653450012,
      "learning_rate": 0.0002,
      "loss": 1.6076,
      "step": 465
    },
    {
      "epoch": 0.17007299270072992,
      "grad_norm": 0.1278480887413025,
      "learning_rate": 0.0002,
      "loss": 1.6393,
      "step": 466
    },
    {
      "epoch": 0.17043795620437957,
      "grad_norm": 0.12380988150835037,
      "learning_rate": 0.0002,
      "loss": 1.5887,
      "step": 467
    },
    {
      "epoch": 0.1708029197080292,
      "grad_norm": 0.12113308906555176,
      "learning_rate": 0.0002,
      "loss": 1.607,
      "step": 468
    },
    {
      "epoch": 0.17116788321167883,
      "grad_norm": 0.12327751517295837,
      "learning_rate": 0.0002,
      "loss": 1.5775,
      "step": 469
    },
    {
      "epoch": 0.17153284671532848,
      "grad_norm": 0.12188578397035599,
      "learning_rate": 0.0002,
      "loss": 1.6374,
      "step": 470
    },
    {
      "epoch": 0.1718978102189781,
      "grad_norm": 0.12450867146253586,
      "learning_rate": 0.0002,
      "loss": 1.6115,
      "step": 471
    },
    {
      "epoch": 0.17226277372262774,
      "grad_norm": 0.1270359605550766,
      "learning_rate": 0.0002,
      "loss": 1.5992,
      "step": 472
    },
    {
      "epoch": 0.17262773722627736,
      "grad_norm": 0.12298370897769928,
      "learning_rate": 0.0002,
      "loss": 1.5418,
      "step": 473
    },
    {
      "epoch": 0.172992700729927,
      "grad_norm": 0.12523198127746582,
      "learning_rate": 0.0002,
      "loss": 1.6303,
      "step": 474
    },
    {
      "epoch": 0.17335766423357665,
      "grad_norm": 0.12099114805459976,
      "learning_rate": 0.0002,
      "loss": 1.5323,
      "step": 475
    },
    {
      "epoch": 0.17372262773722627,
      "grad_norm": 0.12272965908050537,
      "learning_rate": 0.0002,
      "loss": 1.5921,
      "step": 476
    },
    {
      "epoch": 0.17408759124087592,
      "grad_norm": 0.1289604753255844,
      "learning_rate": 0.0002,
      "loss": 1.6889,
      "step": 477
    },
    {
      "epoch": 0.17445255474452553,
      "grad_norm": 0.12361368536949158,
      "learning_rate": 0.0002,
      "loss": 1.6201,
      "step": 478
    },
    {
      "epoch": 0.17481751824817518,
      "grad_norm": 0.12392832338809967,
      "learning_rate": 0.0002,
      "loss": 1.5876,
      "step": 479
    },
    {
      "epoch": 0.17518248175182483,
      "grad_norm": 0.12632419168949127,
      "learning_rate": 0.0002,
      "loss": 1.7009,
      "step": 480
    },
    {
      "epoch": 0.17554744525547444,
      "grad_norm": 0.12409470975399017,
      "learning_rate": 0.0002,
      "loss": 1.5598,
      "step": 481
    },
    {
      "epoch": 0.1759124087591241,
      "grad_norm": 0.12016226351261139,
      "learning_rate": 0.0002,
      "loss": 1.5568,
      "step": 482
    },
    {
      "epoch": 0.17627737226277373,
      "grad_norm": 0.12752088904380798,
      "learning_rate": 0.0002,
      "loss": 1.6217,
      "step": 483
    },
    {
      "epoch": 0.17664233576642335,
      "grad_norm": 0.12866921722888947,
      "learning_rate": 0.0002,
      "loss": 1.5857,
      "step": 484
    },
    {
      "epoch": 0.177007299270073,
      "grad_norm": 0.12729863822460175,
      "learning_rate": 0.0002,
      "loss": 1.5651,
      "step": 485
    },
    {
      "epoch": 0.17737226277372262,
      "grad_norm": 0.12759459018707275,
      "learning_rate": 0.0002,
      "loss": 1.536,
      "step": 486
    },
    {
      "epoch": 0.17773722627737226,
      "grad_norm": 0.12400937080383301,
      "learning_rate": 0.0002,
      "loss": 1.5636,
      "step": 487
    },
    {
      "epoch": 0.1781021897810219,
      "grad_norm": 0.1250094324350357,
      "learning_rate": 0.0002,
      "loss": 1.5593,
      "step": 488
    },
    {
      "epoch": 0.17846715328467153,
      "grad_norm": 0.12380323559045792,
      "learning_rate": 0.0002,
      "loss": 1.5767,
      "step": 489
    },
    {
      "epoch": 0.17883211678832117,
      "grad_norm": 0.12458058446645737,
      "learning_rate": 0.0002,
      "loss": 1.6406,
      "step": 490
    },
    {
      "epoch": 0.1791970802919708,
      "grad_norm": 0.12422765791416168,
      "learning_rate": 0.0002,
      "loss": 1.6176,
      "step": 491
    },
    {
      "epoch": 0.17956204379562044,
      "grad_norm": 0.11989676207304001,
      "learning_rate": 0.0002,
      "loss": 1.5694,
      "step": 492
    },
    {
      "epoch": 0.17992700729927008,
      "grad_norm": 0.12428463995456696,
      "learning_rate": 0.0002,
      "loss": 1.6068,
      "step": 493
    },
    {
      "epoch": 0.1802919708029197,
      "grad_norm": 0.12298950552940369,
      "learning_rate": 0.0002,
      "loss": 1.5994,
      "step": 494
    },
    {
      "epoch": 0.18065693430656934,
      "grad_norm": 0.12665534019470215,
      "learning_rate": 0.0002,
      "loss": 1.6717,
      "step": 495
    },
    {
      "epoch": 0.181021897810219,
      "grad_norm": 0.12159518152475357,
      "learning_rate": 0.0002,
      "loss": 1.5706,
      "step": 496
    },
    {
      "epoch": 0.1813868613138686,
      "grad_norm": 0.13005489110946655,
      "learning_rate": 0.0002,
      "loss": 1.5666,
      "step": 497
    },
    {
      "epoch": 0.18175182481751825,
      "grad_norm": 0.11993636935949326,
      "learning_rate": 0.0002,
      "loss": 1.604,
      "step": 498
    },
    {
      "epoch": 0.18211678832116787,
      "grad_norm": 0.12842729687690735,
      "learning_rate": 0.0002,
      "loss": 1.6092,
      "step": 499
    },
    {
      "epoch": 0.18248175182481752,
      "grad_norm": 0.12514735758304596,
      "learning_rate": 0.0002,
      "loss": 1.5833,
      "step": 500
    },
    {
      "epoch": 0.18284671532846716,
      "grad_norm": 0.12309946864843369,
      "learning_rate": 0.0002,
      "loss": 1.572,
      "step": 501
    },
    {
      "epoch": 0.18321167883211678,
      "grad_norm": 0.14705084264278412,
      "learning_rate": 0.0002,
      "loss": 1.5838,
      "step": 502
    },
    {
      "epoch": 0.18357664233576643,
      "grad_norm": 0.1225958913564682,
      "learning_rate": 0.0002,
      "loss": 1.5515,
      "step": 503
    },
    {
      "epoch": 0.18394160583941604,
      "grad_norm": 0.13522018492221832,
      "learning_rate": 0.0002,
      "loss": 1.5135,
      "step": 504
    },
    {
      "epoch": 0.1843065693430657,
      "grad_norm": 0.12841694056987762,
      "learning_rate": 0.0002,
      "loss": 1.5291,
      "step": 505
    },
    {
      "epoch": 0.18467153284671534,
      "grad_norm": 0.13343186676502228,
      "learning_rate": 0.0002,
      "loss": 1.6011,
      "step": 506
    },
    {
      "epoch": 0.18503649635036495,
      "grad_norm": 0.12451590597629547,
      "learning_rate": 0.0002,
      "loss": 1.5332,
      "step": 507
    },
    {
      "epoch": 0.1854014598540146,
      "grad_norm": 0.12955798208713531,
      "learning_rate": 0.0002,
      "loss": 1.5344,
      "step": 508
    },
    {
      "epoch": 0.18576642335766425,
      "grad_norm": 0.12271616607904434,
      "learning_rate": 0.0002,
      "loss": 1.6276,
      "step": 509
    },
    {
      "epoch": 0.18613138686131386,
      "grad_norm": 0.12159144878387451,
      "learning_rate": 0.0002,
      "loss": 1.5768,
      "step": 510
    },
    {
      "epoch": 0.1864963503649635,
      "grad_norm": 0.1294833868741989,
      "learning_rate": 0.0002,
      "loss": 1.6131,
      "step": 511
    },
    {
      "epoch": 0.18686131386861313,
      "grad_norm": 0.12434911727905273,
      "learning_rate": 0.0002,
      "loss": 1.5728,
      "step": 512
    },
    {
      "epoch": 0.18722627737226277,
      "grad_norm": 0.11856069415807724,
      "learning_rate": 0.0002,
      "loss": 1.4826,
      "step": 513
    },
    {
      "epoch": 0.18759124087591242,
      "grad_norm": 0.12816767394542694,
      "learning_rate": 0.0002,
      "loss": 1.6111,
      "step": 514
    },
    {
      "epoch": 0.18795620437956204,
      "grad_norm": 0.12867654860019684,
      "learning_rate": 0.0002,
      "loss": 1.6076,
      "step": 515
    },
    {
      "epoch": 0.18832116788321168,
      "grad_norm": 0.12612290680408478,
      "learning_rate": 0.0002,
      "loss": 1.73,
      "step": 516
    },
    {
      "epoch": 0.18868613138686133,
      "grad_norm": 0.12409399449825287,
      "learning_rate": 0.0002,
      "loss": 1.5821,
      "step": 517
    },
    {
      "epoch": 0.18905109489051095,
      "grad_norm": 0.12502148747444153,
      "learning_rate": 0.0002,
      "loss": 1.5834,
      "step": 518
    },
    {
      "epoch": 0.1894160583941606,
      "grad_norm": 0.11640866845846176,
      "learning_rate": 0.0002,
      "loss": 1.5148,
      "step": 519
    },
    {
      "epoch": 0.1897810218978102,
      "grad_norm": 0.12292591482400894,
      "learning_rate": 0.0002,
      "loss": 1.6643,
      "step": 520
    },
    {
      "epoch": 0.19014598540145985,
      "grad_norm": 0.12456263601779938,
      "learning_rate": 0.0002,
      "loss": 1.5245,
      "step": 521
    },
    {
      "epoch": 0.1905109489051095,
      "grad_norm": 0.12060603499412537,
      "learning_rate": 0.0002,
      "loss": 1.5681,
      "step": 522
    },
    {
      "epoch": 0.19087591240875912,
      "grad_norm": 0.13056325912475586,
      "learning_rate": 0.0002,
      "loss": 1.6876,
      "step": 523
    },
    {
      "epoch": 0.19124087591240876,
      "grad_norm": 0.13116861879825592,
      "learning_rate": 0.0002,
      "loss": 1.5679,
      "step": 524
    },
    {
      "epoch": 0.19160583941605838,
      "grad_norm": 0.12464328855276108,
      "learning_rate": 0.0002,
      "loss": 1.5778,
      "step": 525
    },
    {
      "epoch": 0.19197080291970803,
      "grad_norm": 0.12195053696632385,
      "learning_rate": 0.0002,
      "loss": 1.5499,
      "step": 526
    },
    {
      "epoch": 0.19233576642335767,
      "grad_norm": 0.13058671355247498,
      "learning_rate": 0.0002,
      "loss": 1.6021,
      "step": 527
    },
    {
      "epoch": 0.1927007299270073,
      "grad_norm": 0.12415425479412079,
      "learning_rate": 0.0002,
      "loss": 1.61,
      "step": 528
    },
    {
      "epoch": 0.19306569343065694,
      "grad_norm": 0.12547588348388672,
      "learning_rate": 0.0002,
      "loss": 1.6547,
      "step": 529
    },
    {
      "epoch": 0.19343065693430658,
      "grad_norm": 0.12227174639701843,
      "learning_rate": 0.0002,
      "loss": 1.5307,
      "step": 530
    },
    {
      "epoch": 0.1937956204379562,
      "grad_norm": 0.13205865025520325,
      "learning_rate": 0.0002,
      "loss": 1.6734,
      "step": 531
    },
    {
      "epoch": 0.19416058394160585,
      "grad_norm": 0.11733454465866089,
      "learning_rate": 0.0002,
      "loss": 1.4786,
      "step": 532
    },
    {
      "epoch": 0.19452554744525546,
      "grad_norm": 0.1192682534456253,
      "learning_rate": 0.0002,
      "loss": 1.551,
      "step": 533
    },
    {
      "epoch": 0.1948905109489051,
      "grad_norm": 0.12986639142036438,
      "learning_rate": 0.0002,
      "loss": 1.639,
      "step": 534
    },
    {
      "epoch": 0.19525547445255476,
      "grad_norm": 0.12769947946071625,
      "learning_rate": 0.0002,
      "loss": 1.626,
      "step": 535
    },
    {
      "epoch": 0.19562043795620437,
      "grad_norm": 0.12299970537424088,
      "learning_rate": 0.0002,
      "loss": 1.5738,
      "step": 536
    },
    {
      "epoch": 0.19598540145985402,
      "grad_norm": 0.11823706328868866,
      "learning_rate": 0.0002,
      "loss": 1.4888,
      "step": 537
    },
    {
      "epoch": 0.19635036496350364,
      "grad_norm": 0.12024753540754318,
      "learning_rate": 0.0002,
      "loss": 1.5904,
      "step": 538
    },
    {
      "epoch": 0.19671532846715328,
      "grad_norm": 0.119456946849823,
      "learning_rate": 0.0002,
      "loss": 1.5293,
      "step": 539
    },
    {
      "epoch": 0.19708029197080293,
      "grad_norm": 0.12368359416723251,
      "learning_rate": 0.0002,
      "loss": 1.5584,
      "step": 540
    },
    {
      "epoch": 0.19744525547445255,
      "grad_norm": 0.12554669380187988,
      "learning_rate": 0.0002,
      "loss": 1.6024,
      "step": 541
    },
    {
      "epoch": 0.1978102189781022,
      "grad_norm": 0.1263635903596878,
      "learning_rate": 0.0002,
      "loss": 1.646,
      "step": 542
    },
    {
      "epoch": 0.19817518248175184,
      "grad_norm": 0.12540508806705475,
      "learning_rate": 0.0002,
      "loss": 1.641,
      "step": 543
    },
    {
      "epoch": 0.19854014598540146,
      "grad_norm": 0.12134627997875214,
      "learning_rate": 0.0002,
      "loss": 1.5513,
      "step": 544
    },
    {
      "epoch": 0.1989051094890511,
      "grad_norm": 0.12325450778007507,
      "learning_rate": 0.0002,
      "loss": 1.5899,
      "step": 545
    },
    {
      "epoch": 0.19927007299270072,
      "grad_norm": 0.12610523402690887,
      "learning_rate": 0.0002,
      "loss": 1.596,
      "step": 546
    },
    {
      "epoch": 0.19963503649635037,
      "grad_norm": 0.12681403756141663,
      "learning_rate": 0.0002,
      "loss": 1.6101,
      "step": 547
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.1371295154094696,
      "learning_rate": 0.0002,
      "loss": 1.6646,
      "step": 548
    },
    {
      "epoch": 0.20036496350364963,
      "grad_norm": 0.12604184448719025,
      "learning_rate": 0.0002,
      "loss": 1.5634,
      "step": 549
    },
    {
      "epoch": 0.20072992700729927,
      "grad_norm": 0.12495389580726624,
      "learning_rate": 0.0002,
      "loss": 1.6122,
      "step": 550
    },
    {
      "epoch": 0.2010948905109489,
      "grad_norm": 0.1262616366147995,
      "learning_rate": 0.0002,
      "loss": 1.6134,
      "step": 551
    },
    {
      "epoch": 0.20145985401459854,
      "grad_norm": 0.12361679971218109,
      "learning_rate": 0.0002,
      "loss": 1.5422,
      "step": 552
    },
    {
      "epoch": 0.20182481751824818,
      "grad_norm": 0.13471703231334686,
      "learning_rate": 0.0002,
      "loss": 1.6508,
      "step": 553
    },
    {
      "epoch": 0.2021897810218978,
      "grad_norm": 0.12616637349128723,
      "learning_rate": 0.0002,
      "loss": 1.4846,
      "step": 554
    },
    {
      "epoch": 0.20255474452554745,
      "grad_norm": 0.13487878441810608,
      "learning_rate": 0.0002,
      "loss": 1.641,
      "step": 555
    },
    {
      "epoch": 0.2029197080291971,
      "grad_norm": 0.13742412626743317,
      "learning_rate": 0.0002,
      "loss": 1.6499,
      "step": 556
    },
    {
      "epoch": 0.2032846715328467,
      "grad_norm": 0.1438058465719223,
      "learning_rate": 0.0002,
      "loss": 1.6358,
      "step": 557
    },
    {
      "epoch": 0.20364963503649636,
      "grad_norm": 0.12672267854213715,
      "learning_rate": 0.0002,
      "loss": 1.5424,
      "step": 558
    },
    {
      "epoch": 0.20401459854014597,
      "grad_norm": 0.13570484519004822,
      "learning_rate": 0.0002,
      "loss": 1.662,
      "step": 559
    },
    {
      "epoch": 0.20437956204379562,
      "grad_norm": 0.1322007030248642,
      "learning_rate": 0.0002,
      "loss": 1.6108,
      "step": 560
    },
    {
      "epoch": 0.20474452554744527,
      "grad_norm": 0.12482073903083801,
      "learning_rate": 0.0002,
      "loss": 1.5554,
      "step": 561
    },
    {
      "epoch": 0.20510948905109488,
      "grad_norm": 0.12600703537464142,
      "learning_rate": 0.0002,
      "loss": 1.6091,
      "step": 562
    },
    {
      "epoch": 0.20547445255474453,
      "grad_norm": 0.12867365777492523,
      "learning_rate": 0.0002,
      "loss": 1.5661,
      "step": 563
    },
    {
      "epoch": 0.20583941605839415,
      "grad_norm": 0.12389802187681198,
      "learning_rate": 0.0002,
      "loss": 1.5933,
      "step": 564
    },
    {
      "epoch": 0.2062043795620438,
      "grad_norm": 0.12294901907444,
      "learning_rate": 0.0002,
      "loss": 1.5966,
      "step": 565
    },
    {
      "epoch": 0.20656934306569344,
      "grad_norm": 0.12663765251636505,
      "learning_rate": 0.0002,
      "loss": 1.6,
      "step": 566
    },
    {
      "epoch": 0.20693430656934306,
      "grad_norm": 0.13137564063072205,
      "learning_rate": 0.0002,
      "loss": 1.6041,
      "step": 567
    },
    {
      "epoch": 0.2072992700729927,
      "grad_norm": 0.12941476702690125,
      "learning_rate": 0.0002,
      "loss": 1.5802,
      "step": 568
    },
    {
      "epoch": 0.20766423357664235,
      "grad_norm": 0.13174748420715332,
      "learning_rate": 0.0002,
      "loss": 1.5899,
      "step": 569
    },
    {
      "epoch": 0.20802919708029197,
      "grad_norm": 0.12933413684368134,
      "learning_rate": 0.0002,
      "loss": 1.6523,
      "step": 570
    },
    {
      "epoch": 0.2083941605839416,
      "grad_norm": 0.12375111877918243,
      "learning_rate": 0.0002,
      "loss": 1.5936,
      "step": 571
    },
    {
      "epoch": 0.20875912408759123,
      "grad_norm": 0.12206665426492691,
      "learning_rate": 0.0002,
      "loss": 1.5521,
      "step": 572
    },
    {
      "epoch": 0.20912408759124088,
      "grad_norm": 0.12940704822540283,
      "learning_rate": 0.0002,
      "loss": 1.5832,
      "step": 573
    },
    {
      "epoch": 0.20948905109489052,
      "grad_norm": 0.1234569400548935,
      "learning_rate": 0.0002,
      "loss": 1.588,
      "step": 574
    },
    {
      "epoch": 0.20985401459854014,
      "grad_norm": 0.1330309808254242,
      "learning_rate": 0.0002,
      "loss": 1.627,
      "step": 575
    },
    {
      "epoch": 0.21021897810218979,
      "grad_norm": 0.13321800529956818,
      "learning_rate": 0.0002,
      "loss": 1.6522,
      "step": 576
    },
    {
      "epoch": 0.2105839416058394,
      "grad_norm": 0.1326562464237213,
      "learning_rate": 0.0002,
      "loss": 1.6395,
      "step": 577
    },
    {
      "epoch": 0.21094890510948905,
      "grad_norm": 0.12630027532577515,
      "learning_rate": 0.0002,
      "loss": 1.5718,
      "step": 578
    },
    {
      "epoch": 0.2113138686131387,
      "grad_norm": 0.13319642841815948,
      "learning_rate": 0.0002,
      "loss": 1.6591,
      "step": 579
    },
    {
      "epoch": 0.2116788321167883,
      "grad_norm": 0.14198234677314758,
      "learning_rate": 0.0002,
      "loss": 1.6378,
      "step": 580
    },
    {
      "epoch": 0.21204379562043796,
      "grad_norm": 0.12817272543907166,
      "learning_rate": 0.0002,
      "loss": 1.6728,
      "step": 581
    },
    {
      "epoch": 0.2124087591240876,
      "grad_norm": 0.1363651156425476,
      "learning_rate": 0.0002,
      "loss": 1.5868,
      "step": 582
    },
    {
      "epoch": 0.21277372262773722,
      "grad_norm": 0.12817902863025665,
      "learning_rate": 0.0002,
      "loss": 1.5982,
      "step": 583
    },
    {
      "epoch": 0.21313868613138687,
      "grad_norm": 0.1262945979833603,
      "learning_rate": 0.0002,
      "loss": 1.6569,
      "step": 584
    },
    {
      "epoch": 0.21350364963503649,
      "grad_norm": 0.12703324854373932,
      "learning_rate": 0.0002,
      "loss": 1.6063,
      "step": 585
    },
    {
      "epoch": 0.21386861313868613,
      "grad_norm": 0.1290171593427658,
      "learning_rate": 0.0002,
      "loss": 1.6301,
      "step": 586
    },
    {
      "epoch": 0.21423357664233578,
      "grad_norm": 0.12325242906808853,
      "learning_rate": 0.0002,
      "loss": 1.5506,
      "step": 587
    },
    {
      "epoch": 0.2145985401459854,
      "grad_norm": 0.12850339710712433,
      "learning_rate": 0.0002,
      "loss": 1.6141,
      "step": 588
    },
    {
      "epoch": 0.21496350364963504,
      "grad_norm": 0.12642017006874084,
      "learning_rate": 0.0002,
      "loss": 1.5583,
      "step": 589
    },
    {
      "epoch": 0.21532846715328466,
      "grad_norm": 0.1287481039762497,
      "learning_rate": 0.0002,
      "loss": 1.6637,
      "step": 590
    },
    {
      "epoch": 0.2156934306569343,
      "grad_norm": 0.12406617403030396,
      "learning_rate": 0.0002,
      "loss": 1.5082,
      "step": 591
    },
    {
      "epoch": 0.21605839416058395,
      "grad_norm": 0.1366252303123474,
      "learning_rate": 0.0002,
      "loss": 1.7043,
      "step": 592
    },
    {
      "epoch": 0.21642335766423357,
      "grad_norm": 0.13122934103012085,
      "learning_rate": 0.0002,
      "loss": 1.6326,
      "step": 593
    },
    {
      "epoch": 0.2167883211678832,
      "grad_norm": 0.12096287310123444,
      "learning_rate": 0.0002,
      "loss": 1.5145,
      "step": 594
    },
    {
      "epoch": 0.21715328467153286,
      "grad_norm": 0.12783284485340118,
      "learning_rate": 0.0002,
      "loss": 1.5322,
      "step": 595
    },
    {
      "epoch": 0.21751824817518248,
      "grad_norm": 0.12616991996765137,
      "learning_rate": 0.0002,
      "loss": 1.6353,
      "step": 596
    },
    {
      "epoch": 0.21788321167883212,
      "grad_norm": 0.12628446519374847,
      "learning_rate": 0.0002,
      "loss": 1.6123,
      "step": 597
    },
    {
      "epoch": 0.21824817518248174,
      "grad_norm": 0.12966489791870117,
      "learning_rate": 0.0002,
      "loss": 1.611,
      "step": 598
    },
    {
      "epoch": 0.2186131386861314,
      "grad_norm": 0.12685440480709076,
      "learning_rate": 0.0002,
      "loss": 1.6001,
      "step": 599
    },
    {
      "epoch": 0.21897810218978103,
      "grad_norm": 0.13016311824321747,
      "learning_rate": 0.0002,
      "loss": 1.5942,
      "step": 600
    },
    {
      "epoch": 0.21934306569343065,
      "grad_norm": 0.1282406449317932,
      "learning_rate": 0.0002,
      "loss": 1.5933,
      "step": 601
    },
    {
      "epoch": 0.2197080291970803,
      "grad_norm": 0.12475615739822388,
      "learning_rate": 0.0002,
      "loss": 1.5341,
      "step": 602
    },
    {
      "epoch": 0.2200729927007299,
      "grad_norm": 0.13007867336273193,
      "learning_rate": 0.0002,
      "loss": 1.623,
      "step": 603
    },
    {
      "epoch": 0.22043795620437956,
      "grad_norm": 0.12241235375404358,
      "learning_rate": 0.0002,
      "loss": 1.561,
      "step": 604
    },
    {
      "epoch": 0.2208029197080292,
      "grad_norm": 0.1352016180753708,
      "learning_rate": 0.0002,
      "loss": 1.5757,
      "step": 605
    },
    {
      "epoch": 0.22116788321167882,
      "grad_norm": 0.12300044298171997,
      "learning_rate": 0.0002,
      "loss": 1.5032,
      "step": 606
    },
    {
      "epoch": 0.22153284671532847,
      "grad_norm": 0.13144831359386444,
      "learning_rate": 0.0002,
      "loss": 1.5676,
      "step": 607
    },
    {
      "epoch": 0.22189781021897811,
      "grad_norm": 0.13430549204349518,
      "learning_rate": 0.0002,
      "loss": 1.5704,
      "step": 608
    },
    {
      "epoch": 0.22226277372262773,
      "grad_norm": 0.12394355982542038,
      "learning_rate": 0.0002,
      "loss": 1.6019,
      "step": 609
    },
    {
      "epoch": 0.22262773722627738,
      "grad_norm": 0.12226184457540512,
      "learning_rate": 0.0002,
      "loss": 1.5759,
      "step": 610
    },
    {
      "epoch": 0.222992700729927,
      "grad_norm": 0.13529735803604126,
      "learning_rate": 0.0002,
      "loss": 1.6626,
      "step": 611
    },
    {
      "epoch": 0.22335766423357664,
      "grad_norm": 0.1277168244123459,
      "learning_rate": 0.0002,
      "loss": 1.6995,
      "step": 612
    },
    {
      "epoch": 0.2237226277372263,
      "grad_norm": 0.12239306420087814,
      "learning_rate": 0.0002,
      "loss": 1.5152,
      "step": 613
    },
    {
      "epoch": 0.2240875912408759,
      "grad_norm": 0.12527568638324738,
      "learning_rate": 0.0002,
      "loss": 1.6283,
      "step": 614
    },
    {
      "epoch": 0.22445255474452555,
      "grad_norm": 0.12858916819095612,
      "learning_rate": 0.0002,
      "loss": 1.6087,
      "step": 615
    },
    {
      "epoch": 0.22481751824817517,
      "grad_norm": 0.1269419640302658,
      "learning_rate": 0.0002,
      "loss": 1.5659,
      "step": 616
    },
    {
      "epoch": 0.22518248175182481,
      "grad_norm": 0.12576986849308014,
      "learning_rate": 0.0002,
      "loss": 1.5774,
      "step": 617
    },
    {
      "epoch": 0.22554744525547446,
      "grad_norm": 0.12508419156074524,
      "learning_rate": 0.0002,
      "loss": 1.5754,
      "step": 618
    },
    {
      "epoch": 0.22591240875912408,
      "grad_norm": 0.1288011223077774,
      "learning_rate": 0.0002,
      "loss": 1.6216,
      "step": 619
    },
    {
      "epoch": 0.22627737226277372,
      "grad_norm": 0.12680262327194214,
      "learning_rate": 0.0002,
      "loss": 1.6121,
      "step": 620
    },
    {
      "epoch": 0.22664233576642337,
      "grad_norm": 0.1337292641401291,
      "learning_rate": 0.0002,
      "loss": 1.6533,
      "step": 621
    },
    {
      "epoch": 0.227007299270073,
      "grad_norm": 0.1294473260641098,
      "learning_rate": 0.0002,
      "loss": 1.6524,
      "step": 622
    },
    {
      "epoch": 0.22737226277372263,
      "grad_norm": 0.12878432869911194,
      "learning_rate": 0.0002,
      "loss": 1.5529,
      "step": 623
    },
    {
      "epoch": 0.22773722627737225,
      "grad_norm": 0.13039539754390717,
      "learning_rate": 0.0002,
      "loss": 1.6236,
      "step": 624
    },
    {
      "epoch": 0.2281021897810219,
      "grad_norm": 0.13648881018161774,
      "learning_rate": 0.0002,
      "loss": 1.6413,
      "step": 625
    },
    {
      "epoch": 0.22846715328467154,
      "grad_norm": 0.12794329226016998,
      "learning_rate": 0.0002,
      "loss": 1.6038,
      "step": 626
    },
    {
      "epoch": 0.22883211678832116,
      "grad_norm": 0.12740661203861237,
      "learning_rate": 0.0002,
      "loss": 1.5797,
      "step": 627
    },
    {
      "epoch": 0.2291970802919708,
      "grad_norm": 0.12367057055234909,
      "learning_rate": 0.0002,
      "loss": 1.5602,
      "step": 628
    },
    {
      "epoch": 0.22956204379562045,
      "grad_norm": 0.13924606144428253,
      "learning_rate": 0.0002,
      "loss": 1.5937,
      "step": 629
    },
    {
      "epoch": 0.22992700729927007,
      "grad_norm": 0.13208065927028656,
      "learning_rate": 0.0002,
      "loss": 1.5593,
      "step": 630
    },
    {
      "epoch": 0.23029197080291972,
      "grad_norm": 0.12866170704364777,
      "learning_rate": 0.0002,
      "loss": 1.5838,
      "step": 631
    },
    {
      "epoch": 0.23065693430656933,
      "grad_norm": 0.12462903559207916,
      "learning_rate": 0.0002,
      "loss": 1.5887,
      "step": 632
    },
    {
      "epoch": 0.23102189781021898,
      "grad_norm": 0.14653056859970093,
      "learning_rate": 0.0002,
      "loss": 1.6537,
      "step": 633
    },
    {
      "epoch": 0.23138686131386862,
      "grad_norm": 0.12716948986053467,
      "learning_rate": 0.0002,
      "loss": 1.5078,
      "step": 634
    },
    {
      "epoch": 0.23175182481751824,
      "grad_norm": 0.14617744088172913,
      "learning_rate": 0.0002,
      "loss": 1.5916,
      "step": 635
    },
    {
      "epoch": 0.2321167883211679,
      "grad_norm": 0.12871889770030975,
      "learning_rate": 0.0002,
      "loss": 1.5051,
      "step": 636
    },
    {
      "epoch": 0.2324817518248175,
      "grad_norm": 0.14056125283241272,
      "learning_rate": 0.0002,
      "loss": 1.5766,
      "step": 637
    },
    {
      "epoch": 0.23284671532846715,
      "grad_norm": 0.12758252024650574,
      "learning_rate": 0.0002,
      "loss": 1.5652,
      "step": 638
    },
    {
      "epoch": 0.2332116788321168,
      "grad_norm": 0.13819703459739685,
      "learning_rate": 0.0002,
      "loss": 1.5764,
      "step": 639
    },
    {
      "epoch": 0.23357664233576642,
      "grad_norm": 0.12751556932926178,
      "learning_rate": 0.0002,
      "loss": 1.5233,
      "step": 640
    },
    {
      "epoch": 0.23394160583941606,
      "grad_norm": 0.14769837260246277,
      "learning_rate": 0.0002,
      "loss": 1.6019,
      "step": 641
    },
    {
      "epoch": 0.2343065693430657,
      "grad_norm": 0.13473080098628998,
      "learning_rate": 0.0002,
      "loss": 1.6334,
      "step": 642
    },
    {
      "epoch": 0.23467153284671532,
      "grad_norm": 0.13933910429477692,
      "learning_rate": 0.0002,
      "loss": 1.5779,
      "step": 643
    },
    {
      "epoch": 0.23503649635036497,
      "grad_norm": 0.1292448490858078,
      "learning_rate": 0.0002,
      "loss": 1.6493,
      "step": 644
    },
    {
      "epoch": 0.2354014598540146,
      "grad_norm": 0.1376861184835434,
      "learning_rate": 0.0002,
      "loss": 1.6159,
      "step": 645
    },
    {
      "epoch": 0.23576642335766423,
      "grad_norm": 0.1310352087020874,
      "learning_rate": 0.0002,
      "loss": 1.6002,
      "step": 646
    },
    {
      "epoch": 0.23613138686131388,
      "grad_norm": 0.13202151656150818,
      "learning_rate": 0.0002,
      "loss": 1.6023,
      "step": 647
    },
    {
      "epoch": 0.2364963503649635,
      "grad_norm": 0.13219547271728516,
      "learning_rate": 0.0002,
      "loss": 1.6607,
      "step": 648
    },
    {
      "epoch": 0.23686131386861314,
      "grad_norm": 0.12737205624580383,
      "learning_rate": 0.0002,
      "loss": 1.5969,
      "step": 649
    },
    {
      "epoch": 0.23722627737226276,
      "grad_norm": 0.12716510891914368,
      "learning_rate": 0.0002,
      "loss": 1.6018,
      "step": 650
    },
    {
      "epoch": 0.2375912408759124,
      "grad_norm": 0.13227479159832,
      "learning_rate": 0.0002,
      "loss": 1.6411,
      "step": 651
    },
    {
      "epoch": 0.23795620437956205,
      "grad_norm": 0.13447023928165436,
      "learning_rate": 0.0002,
      "loss": 1.5636,
      "step": 652
    },
    {
      "epoch": 0.23832116788321167,
      "grad_norm": 0.13219791650772095,
      "learning_rate": 0.0002,
      "loss": 1.616,
      "step": 653
    },
    {
      "epoch": 0.23868613138686132,
      "grad_norm": 0.14518432319164276,
      "learning_rate": 0.0002,
      "loss": 1.6025,
      "step": 654
    },
    {
      "epoch": 0.23905109489051096,
      "grad_norm": 0.13520745933055878,
      "learning_rate": 0.0002,
      "loss": 1.6248,
      "step": 655
    },
    {
      "epoch": 0.23941605839416058,
      "grad_norm": 0.13061542809009552,
      "learning_rate": 0.0002,
      "loss": 1.5589,
      "step": 656
    },
    {
      "epoch": 0.23978102189781023,
      "grad_norm": 0.13301809132099152,
      "learning_rate": 0.0002,
      "loss": 1.5978,
      "step": 657
    },
    {
      "epoch": 0.24014598540145984,
      "grad_norm": 0.13062575459480286,
      "learning_rate": 0.0002,
      "loss": 1.6017,
      "step": 658
    },
    {
      "epoch": 0.2405109489051095,
      "grad_norm": 0.13420599699020386,
      "learning_rate": 0.0002,
      "loss": 1.5381,
      "step": 659
    },
    {
      "epoch": 0.24087591240875914,
      "grad_norm": 0.13332746922969818,
      "learning_rate": 0.0002,
      "loss": 1.5302,
      "step": 660
    },
    {
      "epoch": 0.24124087591240875,
      "grad_norm": 0.13045744597911835,
      "learning_rate": 0.0002,
      "loss": 1.5933,
      "step": 661
    },
    {
      "epoch": 0.2416058394160584,
      "grad_norm": 0.12723106145858765,
      "learning_rate": 0.0002,
      "loss": 1.5477,
      "step": 662
    },
    {
      "epoch": 0.24197080291970802,
      "grad_norm": 0.13549931347370148,
      "learning_rate": 0.0002,
      "loss": 1.6687,
      "step": 663
    },
    {
      "epoch": 0.24233576642335766,
      "grad_norm": 0.12326513230800629,
      "learning_rate": 0.0002,
      "loss": 1.5009,
      "step": 664
    },
    {
      "epoch": 0.2427007299270073,
      "grad_norm": 0.1319713294506073,
      "learning_rate": 0.0002,
      "loss": 1.5681,
      "step": 665
    },
    {
      "epoch": 0.24306569343065693,
      "grad_norm": 0.1251668930053711,
      "learning_rate": 0.0002,
      "loss": 1.5842,
      "step": 666
    },
    {
      "epoch": 0.24343065693430657,
      "grad_norm": 0.1359686702489853,
      "learning_rate": 0.0002,
      "loss": 1.6112,
      "step": 667
    },
    {
      "epoch": 0.24379562043795622,
      "grad_norm": 0.1299322545528412,
      "learning_rate": 0.0002,
      "loss": 1.6008,
      "step": 668
    },
    {
      "epoch": 0.24416058394160584,
      "grad_norm": 0.13247615098953247,
      "learning_rate": 0.0002,
      "loss": 1.5851,
      "step": 669
    },
    {
      "epoch": 0.24452554744525548,
      "grad_norm": 0.13385839760303497,
      "learning_rate": 0.0002,
      "loss": 1.6284,
      "step": 670
    },
    {
      "epoch": 0.2448905109489051,
      "grad_norm": 0.13745443522930145,
      "learning_rate": 0.0002,
      "loss": 1.5944,
      "step": 671
    },
    {
      "epoch": 0.24525547445255474,
      "grad_norm": 0.13360241055488586,
      "learning_rate": 0.0002,
      "loss": 1.6416,
      "step": 672
    },
    {
      "epoch": 0.2456204379562044,
      "grad_norm": 0.14015933871269226,
      "learning_rate": 0.0002,
      "loss": 1.6114,
      "step": 673
    },
    {
      "epoch": 0.245985401459854,
      "grad_norm": 0.1356603056192398,
      "learning_rate": 0.0002,
      "loss": 1.5749,
      "step": 674
    },
    {
      "epoch": 0.24635036496350365,
      "grad_norm": 0.1314651370048523,
      "learning_rate": 0.0002,
      "loss": 1.6001,
      "step": 675
    },
    {
      "epoch": 0.24671532846715327,
      "grad_norm": 0.128329336643219,
      "learning_rate": 0.0002,
      "loss": 1.6295,
      "step": 676
    },
    {
      "epoch": 0.24708029197080292,
      "grad_norm": 0.130244642496109,
      "learning_rate": 0.0002,
      "loss": 1.6309,
      "step": 677
    },
    {
      "epoch": 0.24744525547445256,
      "grad_norm": 0.12354037165641785,
      "learning_rate": 0.0002,
      "loss": 1.4648,
      "step": 678
    },
    {
      "epoch": 0.24781021897810218,
      "grad_norm": 0.12949790060520172,
      "learning_rate": 0.0002,
      "loss": 1.5658,
      "step": 679
    },
    {
      "epoch": 0.24817518248175183,
      "grad_norm": 0.13146783411502838,
      "learning_rate": 0.0002,
      "loss": 1.5882,
      "step": 680
    },
    {
      "epoch": 0.24854014598540147,
      "grad_norm": 0.12736576795578003,
      "learning_rate": 0.0002,
      "loss": 1.5917,
      "step": 681
    },
    {
      "epoch": 0.2489051094890511,
      "grad_norm": 0.13470464944839478,
      "learning_rate": 0.0002,
      "loss": 1.5868,
      "step": 682
    },
    {
      "epoch": 0.24927007299270074,
      "grad_norm": 0.12716783583164215,
      "learning_rate": 0.0002,
      "loss": 1.5981,
      "step": 683
    },
    {
      "epoch": 0.24963503649635035,
      "grad_norm": 0.13455115258693695,
      "learning_rate": 0.0002,
      "loss": 1.6147,
      "step": 684
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.12866565585136414,
      "learning_rate": 0.0002,
      "loss": 1.6132,
      "step": 685
    },
    {
      "epoch": 0.2503649635036496,
      "grad_norm": 0.13592790067195892,
      "learning_rate": 0.0002,
      "loss": 1.6316,
      "step": 686
    },
    {
      "epoch": 0.2507299270072993,
      "grad_norm": 0.13539980351924896,
      "learning_rate": 0.0002,
      "loss": 1.6839,
      "step": 687
    },
    {
      "epoch": 0.2510948905109489,
      "grad_norm": 0.13164173066616058,
      "learning_rate": 0.0002,
      "loss": 1.6008,
      "step": 688
    },
    {
      "epoch": 0.2514598540145985,
      "grad_norm": 0.13888196647167206,
      "learning_rate": 0.0002,
      "loss": 1.5962,
      "step": 689
    },
    {
      "epoch": 0.2518248175182482,
      "grad_norm": 0.1348728984594345,
      "learning_rate": 0.0002,
      "loss": 1.5399,
      "step": 690
    },
    {
      "epoch": 0.2521897810218978,
      "grad_norm": 0.14682956039905548,
      "learning_rate": 0.0002,
      "loss": 1.6207,
      "step": 691
    },
    {
      "epoch": 0.25255474452554744,
      "grad_norm": 0.14165958762168884,
      "learning_rate": 0.0002,
      "loss": 1.5923,
      "step": 692
    },
    {
      "epoch": 0.25291970802919705,
      "grad_norm": 0.13476519286632538,
      "learning_rate": 0.0002,
      "loss": 1.613,
      "step": 693
    },
    {
      "epoch": 0.25328467153284673,
      "grad_norm": 0.13607962429523468,
      "learning_rate": 0.0002,
      "loss": 1.5471,
      "step": 694
    },
    {
      "epoch": 0.25364963503649635,
      "grad_norm": 0.14517605304718018,
      "learning_rate": 0.0002,
      "loss": 1.6525,
      "step": 695
    },
    {
      "epoch": 0.25401459854014596,
      "grad_norm": 0.14711929857730865,
      "learning_rate": 0.0002,
      "loss": 1.5862,
      "step": 696
    },
    {
      "epoch": 0.25437956204379564,
      "grad_norm": 0.14860352873802185,
      "learning_rate": 0.0002,
      "loss": 1.6253,
      "step": 697
    },
    {
      "epoch": 0.25474452554744526,
      "grad_norm": 0.13540425896644592,
      "learning_rate": 0.0002,
      "loss": 1.5181,
      "step": 698
    },
    {
      "epoch": 0.2551094890510949,
      "grad_norm": 0.13812172412872314,
      "learning_rate": 0.0002,
      "loss": 1.6303,
      "step": 699
    },
    {
      "epoch": 0.25547445255474455,
      "grad_norm": 0.1302783191204071,
      "learning_rate": 0.0002,
      "loss": 1.6667,
      "step": 700
    },
    {
      "epoch": 0.25583941605839416,
      "grad_norm": 0.138635516166687,
      "learning_rate": 0.0002,
      "loss": 1.6034,
      "step": 701
    },
    {
      "epoch": 0.2562043795620438,
      "grad_norm": 0.13259904086589813,
      "learning_rate": 0.0002,
      "loss": 1.6734,
      "step": 702
    },
    {
      "epoch": 0.25656934306569346,
      "grad_norm": 0.13747696578502655,
      "learning_rate": 0.0002,
      "loss": 1.5746,
      "step": 703
    },
    {
      "epoch": 0.2569343065693431,
      "grad_norm": 0.12199462950229645,
      "learning_rate": 0.0002,
      "loss": 1.4516,
      "step": 704
    },
    {
      "epoch": 0.2572992700729927,
      "grad_norm": 0.14551463723182678,
      "learning_rate": 0.0002,
      "loss": 1.5941,
      "step": 705
    },
    {
      "epoch": 0.2576642335766423,
      "grad_norm": 0.13090601563453674,
      "learning_rate": 0.0002,
      "loss": 1.5723,
      "step": 706
    },
    {
      "epoch": 0.258029197080292,
      "grad_norm": 0.14511674642562866,
      "learning_rate": 0.0002,
      "loss": 1.6544,
      "step": 707
    },
    {
      "epoch": 0.2583941605839416,
      "grad_norm": 0.12983490526676178,
      "learning_rate": 0.0002,
      "loss": 1.5604,
      "step": 708
    },
    {
      "epoch": 0.2587591240875912,
      "grad_norm": 0.13798363506793976,
      "learning_rate": 0.0002,
      "loss": 1.6188,
      "step": 709
    },
    {
      "epoch": 0.2591240875912409,
      "grad_norm": 0.12761889398097992,
      "learning_rate": 0.0002,
      "loss": 1.5055,
      "step": 710
    },
    {
      "epoch": 0.2594890510948905,
      "grad_norm": 0.13984474539756775,
      "learning_rate": 0.0002,
      "loss": 1.6852,
      "step": 711
    },
    {
      "epoch": 0.25985401459854013,
      "grad_norm": 0.12671226263046265,
      "learning_rate": 0.0002,
      "loss": 1.6068,
      "step": 712
    },
    {
      "epoch": 0.2602189781021898,
      "grad_norm": 0.13424624502658844,
      "learning_rate": 0.0002,
      "loss": 1.5933,
      "step": 713
    },
    {
      "epoch": 0.2605839416058394,
      "grad_norm": 0.13040973246097565,
      "learning_rate": 0.0002,
      "loss": 1.5776,
      "step": 714
    },
    {
      "epoch": 0.26094890510948904,
      "grad_norm": 0.13343258202075958,
      "learning_rate": 0.0002,
      "loss": 1.5629,
      "step": 715
    },
    {
      "epoch": 0.2613138686131387,
      "grad_norm": 0.12715087831020355,
      "learning_rate": 0.0002,
      "loss": 1.5663,
      "step": 716
    },
    {
      "epoch": 0.26167883211678833,
      "grad_norm": 0.138636514544487,
      "learning_rate": 0.0002,
      "loss": 1.5638,
      "step": 717
    },
    {
      "epoch": 0.26204379562043795,
      "grad_norm": 0.13081076741218567,
      "learning_rate": 0.0002,
      "loss": 1.589,
      "step": 718
    },
    {
      "epoch": 0.26240875912408756,
      "grad_norm": 0.1312263458967209,
      "learning_rate": 0.0002,
      "loss": 1.6319,
      "step": 719
    },
    {
      "epoch": 0.26277372262773724,
      "grad_norm": 0.12871064245700836,
      "learning_rate": 0.0002,
      "loss": 1.5942,
      "step": 720
    },
    {
      "epoch": 0.26313868613138686,
      "grad_norm": 0.12614786624908447,
      "learning_rate": 0.0002,
      "loss": 1.5886,
      "step": 721
    },
    {
      "epoch": 0.2635036496350365,
      "grad_norm": 0.1270677000284195,
      "learning_rate": 0.0002,
      "loss": 1.5898,
      "step": 722
    },
    {
      "epoch": 0.26386861313868615,
      "grad_norm": 0.14529967308044434,
      "learning_rate": 0.0002,
      "loss": 1.6611,
      "step": 723
    },
    {
      "epoch": 0.26423357664233577,
      "grad_norm": 0.13351039588451385,
      "learning_rate": 0.0002,
      "loss": 1.5706,
      "step": 724
    },
    {
      "epoch": 0.2645985401459854,
      "grad_norm": 0.12604960799217224,
      "learning_rate": 0.0002,
      "loss": 1.4876,
      "step": 725
    },
    {
      "epoch": 0.26496350364963506,
      "grad_norm": 0.14161822199821472,
      "learning_rate": 0.0002,
      "loss": 1.6221,
      "step": 726
    },
    {
      "epoch": 0.2653284671532847,
      "grad_norm": 0.13583475351333618,
      "learning_rate": 0.0002,
      "loss": 1.5831,
      "step": 727
    },
    {
      "epoch": 0.2656934306569343,
      "grad_norm": 0.13326473534107208,
      "learning_rate": 0.0002,
      "loss": 1.5724,
      "step": 728
    },
    {
      "epoch": 0.26605839416058397,
      "grad_norm": 0.13122308254241943,
      "learning_rate": 0.0002,
      "loss": 1.6325,
      "step": 729
    },
    {
      "epoch": 0.2664233576642336,
      "grad_norm": 0.1377830058336258,
      "learning_rate": 0.0002,
      "loss": 1.6684,
      "step": 730
    },
    {
      "epoch": 0.2667883211678832,
      "grad_norm": 0.1253751665353775,
      "learning_rate": 0.0002,
      "loss": 1.5781,
      "step": 731
    },
    {
      "epoch": 0.2671532846715328,
      "grad_norm": 0.13373394310474396,
      "learning_rate": 0.0002,
      "loss": 1.6753,
      "step": 732
    },
    {
      "epoch": 0.2675182481751825,
      "grad_norm": 0.12496975809335709,
      "learning_rate": 0.0002,
      "loss": 1.5032,
      "step": 733
    },
    {
      "epoch": 0.2678832116788321,
      "grad_norm": 0.13128703832626343,
      "learning_rate": 0.0002,
      "loss": 1.5721,
      "step": 734
    },
    {
      "epoch": 0.26824817518248173,
      "grad_norm": 0.13457854092121124,
      "learning_rate": 0.0002,
      "loss": 1.6085,
      "step": 735
    },
    {
      "epoch": 0.2686131386861314,
      "grad_norm": 0.12277831137180328,
      "learning_rate": 0.0002,
      "loss": 1.5689,
      "step": 736
    },
    {
      "epoch": 0.268978102189781,
      "grad_norm": 0.13485166430473328,
      "learning_rate": 0.0002,
      "loss": 1.5571,
      "step": 737
    },
    {
      "epoch": 0.26934306569343064,
      "grad_norm": 0.13914477825164795,
      "learning_rate": 0.0002,
      "loss": 1.59,
      "step": 738
    },
    {
      "epoch": 0.2697080291970803,
      "grad_norm": 0.1311093121767044,
      "learning_rate": 0.0002,
      "loss": 1.5579,
      "step": 739
    },
    {
      "epoch": 0.27007299270072993,
      "grad_norm": 0.14288941025733948,
      "learning_rate": 0.0002,
      "loss": 1.5415,
      "step": 740
    },
    {
      "epoch": 0.27043795620437955,
      "grad_norm": 0.12793388962745667,
      "learning_rate": 0.0002,
      "loss": 1.4666,
      "step": 741
    },
    {
      "epoch": 0.2708029197080292,
      "grad_norm": 0.12849663197994232,
      "learning_rate": 0.0002,
      "loss": 1.5371,
      "step": 742
    },
    {
      "epoch": 0.27116788321167884,
      "grad_norm": 0.13036863505840302,
      "learning_rate": 0.0002,
      "loss": 1.6457,
      "step": 743
    },
    {
      "epoch": 0.27153284671532846,
      "grad_norm": 0.1418592631816864,
      "learning_rate": 0.0002,
      "loss": 1.6412,
      "step": 744
    },
    {
      "epoch": 0.2718978102189781,
      "grad_norm": 0.13614927232265472,
      "learning_rate": 0.0002,
      "loss": 1.6403,
      "step": 745
    },
    {
      "epoch": 0.27226277372262775,
      "grad_norm": 0.1381651908159256,
      "learning_rate": 0.0002,
      "loss": 1.5877,
      "step": 746
    },
    {
      "epoch": 0.27262773722627737,
      "grad_norm": 0.12716172635555267,
      "learning_rate": 0.0002,
      "loss": 1.5043,
      "step": 747
    },
    {
      "epoch": 0.272992700729927,
      "grad_norm": 0.14366447925567627,
      "learning_rate": 0.0002,
      "loss": 1.6122,
      "step": 748
    },
    {
      "epoch": 0.27335766423357666,
      "grad_norm": 0.13839493691921234,
      "learning_rate": 0.0002,
      "loss": 1.6253,
      "step": 749
    },
    {
      "epoch": 0.2737226277372263,
      "grad_norm": 0.13882477581501007,
      "learning_rate": 0.0002,
      "loss": 1.5798,
      "step": 750
    },
    {
      "epoch": 0.2740875912408759,
      "grad_norm": 0.1327245533466339,
      "learning_rate": 0.0002,
      "loss": 1.5956,
      "step": 751
    },
    {
      "epoch": 0.27445255474452557,
      "grad_norm": 0.14007292687892914,
      "learning_rate": 0.0002,
      "loss": 1.6677,
      "step": 752
    },
    {
      "epoch": 0.2748175182481752,
      "grad_norm": 0.1254786103963852,
      "learning_rate": 0.0002,
      "loss": 1.5567,
      "step": 753
    },
    {
      "epoch": 0.2751824817518248,
      "grad_norm": 0.12549081444740295,
      "learning_rate": 0.0002,
      "loss": 1.5124,
      "step": 754
    },
    {
      "epoch": 0.2755474452554745,
      "grad_norm": 0.12922461330890656,
      "learning_rate": 0.0002,
      "loss": 1.6371,
      "step": 755
    },
    {
      "epoch": 0.2759124087591241,
      "grad_norm": 0.12711787223815918,
      "learning_rate": 0.0002,
      "loss": 1.5724,
      "step": 756
    },
    {
      "epoch": 0.2762773722627737,
      "grad_norm": 0.12085840851068497,
      "learning_rate": 0.0002,
      "loss": 1.4923,
      "step": 757
    },
    {
      "epoch": 0.27664233576642333,
      "grad_norm": 0.1347634345293045,
      "learning_rate": 0.0002,
      "loss": 1.6171,
      "step": 758
    },
    {
      "epoch": 0.277007299270073,
      "grad_norm": 0.12539051473140717,
      "learning_rate": 0.0002,
      "loss": 1.5199,
      "step": 759
    },
    {
      "epoch": 0.2773722627737226,
      "grad_norm": 0.13493546843528748,
      "learning_rate": 0.0002,
      "loss": 1.6056,
      "step": 760
    },
    {
      "epoch": 0.27773722627737224,
      "grad_norm": 0.1304313689470291,
      "learning_rate": 0.0002,
      "loss": 1.5663,
      "step": 761
    },
    {
      "epoch": 0.2781021897810219,
      "grad_norm": 0.14230117201805115,
      "learning_rate": 0.0002,
      "loss": 1.6184,
      "step": 762
    },
    {
      "epoch": 0.27846715328467153,
      "grad_norm": 0.1290244311094284,
      "learning_rate": 0.0002,
      "loss": 1.6021,
      "step": 763
    },
    {
      "epoch": 0.27883211678832115,
      "grad_norm": 0.12766176462173462,
      "learning_rate": 0.0002,
      "loss": 1.5578,
      "step": 764
    },
    {
      "epoch": 0.2791970802919708,
      "grad_norm": 0.13934561610221863,
      "learning_rate": 0.0002,
      "loss": 1.5962,
      "step": 765
    },
    {
      "epoch": 0.27956204379562044,
      "grad_norm": 0.12840186059474945,
      "learning_rate": 0.0002,
      "loss": 1.5372,
      "step": 766
    },
    {
      "epoch": 0.27992700729927006,
      "grad_norm": 0.12687906622886658,
      "learning_rate": 0.0002,
      "loss": 1.6095,
      "step": 767
    },
    {
      "epoch": 0.28029197080291973,
      "grad_norm": 0.15552760660648346,
      "learning_rate": 0.0002,
      "loss": 1.6215,
      "step": 768
    },
    {
      "epoch": 0.28065693430656935,
      "grad_norm": 0.14655078947544098,
      "learning_rate": 0.0002,
      "loss": 1.4961,
      "step": 769
    },
    {
      "epoch": 0.28102189781021897,
      "grad_norm": 0.14549200236797333,
      "learning_rate": 0.0002,
      "loss": 1.5839,
      "step": 770
    },
    {
      "epoch": 0.2813868613138686,
      "grad_norm": 0.14957942068576813,
      "learning_rate": 0.0002,
      "loss": 1.577,
      "step": 771
    },
    {
      "epoch": 0.28175182481751826,
      "grad_norm": 0.14191530644893646,
      "learning_rate": 0.0002,
      "loss": 1.559,
      "step": 772
    },
    {
      "epoch": 0.2821167883211679,
      "grad_norm": 0.13750514388084412,
      "learning_rate": 0.0002,
      "loss": 1.6098,
      "step": 773
    },
    {
      "epoch": 0.2824817518248175,
      "grad_norm": 0.13268022239208221,
      "learning_rate": 0.0002,
      "loss": 1.5157,
      "step": 774
    },
    {
      "epoch": 0.28284671532846717,
      "grad_norm": 0.13821713626384735,
      "learning_rate": 0.0002,
      "loss": 1.5786,
      "step": 775
    },
    {
      "epoch": 0.2832116788321168,
      "grad_norm": 0.145687997341156,
      "learning_rate": 0.0002,
      "loss": 1.6512,
      "step": 776
    },
    {
      "epoch": 0.2835766423357664,
      "grad_norm": 0.1283990889787674,
      "learning_rate": 0.0002,
      "loss": 1.5739,
      "step": 777
    },
    {
      "epoch": 0.2839416058394161,
      "grad_norm": 0.14265255630016327,
      "learning_rate": 0.0002,
      "loss": 1.5912,
      "step": 778
    },
    {
      "epoch": 0.2843065693430657,
      "grad_norm": 0.13169783353805542,
      "learning_rate": 0.0002,
      "loss": 1.5433,
      "step": 779
    },
    {
      "epoch": 0.2846715328467153,
      "grad_norm": 0.1440877467393875,
      "learning_rate": 0.0002,
      "loss": 1.6064,
      "step": 780
    },
    {
      "epoch": 0.285036496350365,
      "grad_norm": 0.12635816633701324,
      "learning_rate": 0.0002,
      "loss": 1.5394,
      "step": 781
    },
    {
      "epoch": 0.2854014598540146,
      "grad_norm": 0.13703329861164093,
      "learning_rate": 0.0002,
      "loss": 1.5219,
      "step": 782
    },
    {
      "epoch": 0.2857664233576642,
      "grad_norm": 0.13887226581573486,
      "learning_rate": 0.0002,
      "loss": 1.652,
      "step": 783
    },
    {
      "epoch": 0.28613138686131384,
      "grad_norm": 0.14092965424060822,
      "learning_rate": 0.0002,
      "loss": 1.6154,
      "step": 784
    },
    {
      "epoch": 0.2864963503649635,
      "grad_norm": 0.13508780300617218,
      "learning_rate": 0.0002,
      "loss": 1.6724,
      "step": 785
    },
    {
      "epoch": 0.28686131386861313,
      "grad_norm": 0.14089901745319366,
      "learning_rate": 0.0002,
      "loss": 1.6516,
      "step": 786
    },
    {
      "epoch": 0.28722627737226275,
      "grad_norm": 0.13314324617385864,
      "learning_rate": 0.0002,
      "loss": 1.5446,
      "step": 787
    },
    {
      "epoch": 0.2875912408759124,
      "grad_norm": 0.13465933501720428,
      "learning_rate": 0.0002,
      "loss": 1.6175,
      "step": 788
    },
    {
      "epoch": 0.28795620437956204,
      "grad_norm": 0.12874193489551544,
      "learning_rate": 0.0002,
      "loss": 1.496,
      "step": 789
    },
    {
      "epoch": 0.28832116788321166,
      "grad_norm": 0.13117295503616333,
      "learning_rate": 0.0002,
      "loss": 1.5768,
      "step": 790
    },
    {
      "epoch": 0.28868613138686133,
      "grad_norm": 0.13257136940956116,
      "learning_rate": 0.0002,
      "loss": 1.6393,
      "step": 791
    },
    {
      "epoch": 0.28905109489051095,
      "grad_norm": 0.1290997862815857,
      "learning_rate": 0.0002,
      "loss": 1.6058,
      "step": 792
    },
    {
      "epoch": 0.28941605839416057,
      "grad_norm": 0.13231584429740906,
      "learning_rate": 0.0002,
      "loss": 1.5955,
      "step": 793
    },
    {
      "epoch": 0.28978102189781024,
      "grad_norm": 0.1354554146528244,
      "learning_rate": 0.0002,
      "loss": 1.5689,
      "step": 794
    },
    {
      "epoch": 0.29014598540145986,
      "grad_norm": 0.12847068905830383,
      "learning_rate": 0.0002,
      "loss": 1.5119,
      "step": 795
    },
    {
      "epoch": 0.2905109489051095,
      "grad_norm": 0.1308169811964035,
      "learning_rate": 0.0002,
      "loss": 1.5676,
      "step": 796
    },
    {
      "epoch": 0.2908759124087591,
      "grad_norm": 0.12915584444999695,
      "learning_rate": 0.0002,
      "loss": 1.5956,
      "step": 797
    },
    {
      "epoch": 0.29124087591240877,
      "grad_norm": 0.13203272223472595,
      "learning_rate": 0.0002,
      "loss": 1.5599,
      "step": 798
    },
    {
      "epoch": 0.2916058394160584,
      "grad_norm": 0.12974904477596283,
      "learning_rate": 0.0002,
      "loss": 1.5735,
      "step": 799
    },
    {
      "epoch": 0.291970802919708,
      "grad_norm": 0.12983448803424835,
      "learning_rate": 0.0002,
      "loss": 1.5669,
      "step": 800
    },
    {
      "epoch": 0.2923357664233577,
      "grad_norm": 0.13448183238506317,
      "learning_rate": 0.0002,
      "loss": 1.6201,
      "step": 801
    },
    {
      "epoch": 0.2927007299270073,
      "grad_norm": 0.1279594749212265,
      "learning_rate": 0.0002,
      "loss": 1.5756,
      "step": 802
    },
    {
      "epoch": 0.2930656934306569,
      "grad_norm": 0.13957776129245758,
      "learning_rate": 0.0002,
      "loss": 1.6559,
      "step": 803
    },
    {
      "epoch": 0.2934306569343066,
      "grad_norm": 0.13170307874679565,
      "learning_rate": 0.0002,
      "loss": 1.6318,
      "step": 804
    },
    {
      "epoch": 0.2937956204379562,
      "grad_norm": 0.1365734487771988,
      "learning_rate": 0.0002,
      "loss": 1.5722,
      "step": 805
    },
    {
      "epoch": 0.2941605839416058,
      "grad_norm": 0.12677350640296936,
      "learning_rate": 0.0002,
      "loss": 1.4925,
      "step": 806
    },
    {
      "epoch": 0.2945255474452555,
      "grad_norm": 0.13183702528476715,
      "learning_rate": 0.0002,
      "loss": 1.5682,
      "step": 807
    },
    {
      "epoch": 0.2948905109489051,
      "grad_norm": 0.13015897572040558,
      "learning_rate": 0.0002,
      "loss": 1.5947,
      "step": 808
    },
    {
      "epoch": 0.29525547445255473,
      "grad_norm": 0.1299537569284439,
      "learning_rate": 0.0002,
      "loss": 1.5931,
      "step": 809
    },
    {
      "epoch": 0.2956204379562044,
      "grad_norm": 0.12830056250095367,
      "learning_rate": 0.0002,
      "loss": 1.5553,
      "step": 810
    },
    {
      "epoch": 0.295985401459854,
      "grad_norm": 0.12348118424415588,
      "learning_rate": 0.0002,
      "loss": 1.5778,
      "step": 811
    },
    {
      "epoch": 0.29635036496350364,
      "grad_norm": 0.1324588507413864,
      "learning_rate": 0.0002,
      "loss": 1.5674,
      "step": 812
    },
    {
      "epoch": 0.29671532846715326,
      "grad_norm": 0.13185621798038483,
      "learning_rate": 0.0002,
      "loss": 1.5791,
      "step": 813
    },
    {
      "epoch": 0.29708029197080293,
      "grad_norm": 0.13802723586559296,
      "learning_rate": 0.0002,
      "loss": 1.5674,
      "step": 814
    },
    {
      "epoch": 0.29744525547445255,
      "grad_norm": 0.12998327612876892,
      "learning_rate": 0.0002,
      "loss": 1.6587,
      "step": 815
    },
    {
      "epoch": 0.29781021897810217,
      "grad_norm": 0.1299857795238495,
      "learning_rate": 0.0002,
      "loss": 1.5258,
      "step": 816
    },
    {
      "epoch": 0.29817518248175184,
      "grad_norm": 0.1316968947649002,
      "learning_rate": 0.0002,
      "loss": 1.6334,
      "step": 817
    },
    {
      "epoch": 0.29854014598540146,
      "grad_norm": 0.13051728904247284,
      "learning_rate": 0.0002,
      "loss": 1.5729,
      "step": 818
    },
    {
      "epoch": 0.2989051094890511,
      "grad_norm": 0.13103622198104858,
      "learning_rate": 0.0002,
      "loss": 1.5759,
      "step": 819
    },
    {
      "epoch": 0.29927007299270075,
      "grad_norm": 0.12478207051753998,
      "learning_rate": 0.0002,
      "loss": 1.5169,
      "step": 820
    },
    {
      "epoch": 0.29963503649635037,
      "grad_norm": 0.1357707977294922,
      "learning_rate": 0.0002,
      "loss": 1.5566,
      "step": 821
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.13705214858055115,
      "learning_rate": 0.0002,
      "loss": 1.5948,
      "step": 822
    },
    {
      "epoch": 0.30036496350364966,
      "grad_norm": 0.12958920001983643,
      "learning_rate": 0.0002,
      "loss": 1.5942,
      "step": 823
    },
    {
      "epoch": 0.3007299270072993,
      "grad_norm": 0.1347607970237732,
      "learning_rate": 0.0002,
      "loss": 1.5715,
      "step": 824
    },
    {
      "epoch": 0.3010948905109489,
      "grad_norm": 0.1407051533460617,
      "learning_rate": 0.0002,
      "loss": 1.6463,
      "step": 825
    },
    {
      "epoch": 0.3014598540145985,
      "grad_norm": 0.12976939976215363,
      "learning_rate": 0.0002,
      "loss": 1.5416,
      "step": 826
    },
    {
      "epoch": 0.3018248175182482,
      "grad_norm": 0.13119088113307953,
      "learning_rate": 0.0002,
      "loss": 1.5227,
      "step": 827
    },
    {
      "epoch": 0.3021897810218978,
      "grad_norm": 0.1285337209701538,
      "learning_rate": 0.0002,
      "loss": 1.5152,
      "step": 828
    },
    {
      "epoch": 0.3025547445255474,
      "grad_norm": 0.13610713183879852,
      "learning_rate": 0.0002,
      "loss": 1.6193,
      "step": 829
    },
    {
      "epoch": 0.3029197080291971,
      "grad_norm": 0.1260436624288559,
      "learning_rate": 0.0002,
      "loss": 1.4895,
      "step": 830
    },
    {
      "epoch": 0.3032846715328467,
      "grad_norm": 0.1312754601240158,
      "learning_rate": 0.0002,
      "loss": 1.539,
      "step": 831
    },
    {
      "epoch": 0.30364963503649633,
      "grad_norm": 0.13419954478740692,
      "learning_rate": 0.0002,
      "loss": 1.5538,
      "step": 832
    },
    {
      "epoch": 0.304014598540146,
      "grad_norm": 0.1309952735900879,
      "learning_rate": 0.0002,
      "loss": 1.5525,
      "step": 833
    },
    {
      "epoch": 0.3043795620437956,
      "grad_norm": 0.12890778481960297,
      "learning_rate": 0.0002,
      "loss": 1.5162,
      "step": 834
    },
    {
      "epoch": 0.30474452554744524,
      "grad_norm": 0.1361568719148636,
      "learning_rate": 0.0002,
      "loss": 1.6644,
      "step": 835
    },
    {
      "epoch": 0.3051094890510949,
      "grad_norm": 0.13489140570163727,
      "learning_rate": 0.0002,
      "loss": 1.5152,
      "step": 836
    },
    {
      "epoch": 0.30547445255474454,
      "grad_norm": 0.12985143065452576,
      "learning_rate": 0.0002,
      "loss": 1.563,
      "step": 837
    },
    {
      "epoch": 0.30583941605839415,
      "grad_norm": 0.13642746210098267,
      "learning_rate": 0.0002,
      "loss": 1.5834,
      "step": 838
    },
    {
      "epoch": 0.30620437956204377,
      "grad_norm": 0.13566000759601593,
      "learning_rate": 0.0002,
      "loss": 1.5511,
      "step": 839
    },
    {
      "epoch": 0.30656934306569344,
      "grad_norm": 0.1378335803747177,
      "learning_rate": 0.0002,
      "loss": 1.6435,
      "step": 840
    },
    {
      "epoch": 0.30693430656934306,
      "grad_norm": 0.147769957780838,
      "learning_rate": 0.0002,
      "loss": 1.5508,
      "step": 841
    },
    {
      "epoch": 0.3072992700729927,
      "grad_norm": 0.133962944149971,
      "learning_rate": 0.0002,
      "loss": 1.6109,
      "step": 842
    },
    {
      "epoch": 0.30766423357664235,
      "grad_norm": 0.14057865738868713,
      "learning_rate": 0.0002,
      "loss": 1.6224,
      "step": 843
    },
    {
      "epoch": 0.30802919708029197,
      "grad_norm": 0.13173595070838928,
      "learning_rate": 0.0002,
      "loss": 1.5423,
      "step": 844
    },
    {
      "epoch": 0.3083941605839416,
      "grad_norm": 0.13364249467849731,
      "learning_rate": 0.0002,
      "loss": 1.5926,
      "step": 845
    },
    {
      "epoch": 0.30875912408759126,
      "grad_norm": 0.12767203152179718,
      "learning_rate": 0.0002,
      "loss": 1.5423,
      "step": 846
    },
    {
      "epoch": 0.3091240875912409,
      "grad_norm": 0.13070882856845856,
      "learning_rate": 0.0002,
      "loss": 1.4935,
      "step": 847
    },
    {
      "epoch": 0.3094890510948905,
      "grad_norm": 0.13640448451042175,
      "learning_rate": 0.0002,
      "loss": 1.5441,
      "step": 848
    },
    {
      "epoch": 0.3098540145985402,
      "grad_norm": 0.1349080204963684,
      "learning_rate": 0.0002,
      "loss": 1.6022,
      "step": 849
    },
    {
      "epoch": 0.3102189781021898,
      "grad_norm": 0.13471956551074982,
      "learning_rate": 0.0002,
      "loss": 1.5906,
      "step": 850
    },
    {
      "epoch": 0.3105839416058394,
      "grad_norm": 0.12835094332695007,
      "learning_rate": 0.0002,
      "loss": 1.5984,
      "step": 851
    },
    {
      "epoch": 0.310948905109489,
      "grad_norm": 0.13445474207401276,
      "learning_rate": 0.0002,
      "loss": 1.5774,
      "step": 852
    },
    {
      "epoch": 0.3113138686131387,
      "grad_norm": 0.13178583979606628,
      "learning_rate": 0.0002,
      "loss": 1.5792,
      "step": 853
    },
    {
      "epoch": 0.3116788321167883,
      "grad_norm": 0.1321147084236145,
      "learning_rate": 0.0002,
      "loss": 1.58,
      "step": 854
    },
    {
      "epoch": 0.31204379562043794,
      "grad_norm": 0.13158607482910156,
      "learning_rate": 0.0002,
      "loss": 1.6081,
      "step": 855
    },
    {
      "epoch": 0.3124087591240876,
      "grad_norm": 0.13135871291160583,
      "learning_rate": 0.0002,
      "loss": 1.6051,
      "step": 856
    },
    {
      "epoch": 0.3127737226277372,
      "grad_norm": 0.12961553037166595,
      "learning_rate": 0.0002,
      "loss": 1.5567,
      "step": 857
    },
    {
      "epoch": 0.31313868613138685,
      "grad_norm": 0.13231207430362701,
      "learning_rate": 0.0002,
      "loss": 1.5961,
      "step": 858
    },
    {
      "epoch": 0.3135036496350365,
      "grad_norm": 0.12859250605106354,
      "learning_rate": 0.0002,
      "loss": 1.5713,
      "step": 859
    },
    {
      "epoch": 0.31386861313868614,
      "grad_norm": 0.12963315844535828,
      "learning_rate": 0.0002,
      "loss": 1.5814,
      "step": 860
    },
    {
      "epoch": 0.31423357664233575,
      "grad_norm": 0.1410791575908661,
      "learning_rate": 0.0002,
      "loss": 1.645,
      "step": 861
    },
    {
      "epoch": 0.31459854014598543,
      "grad_norm": 0.13418173789978027,
      "learning_rate": 0.0002,
      "loss": 1.5985,
      "step": 862
    },
    {
      "epoch": 0.31496350364963505,
      "grad_norm": 0.12979480624198914,
      "learning_rate": 0.0002,
      "loss": 1.5725,
      "step": 863
    },
    {
      "epoch": 0.31532846715328466,
      "grad_norm": 0.1312413513660431,
      "learning_rate": 0.0002,
      "loss": 1.5503,
      "step": 864
    },
    {
      "epoch": 0.3156934306569343,
      "grad_norm": 0.1337752491235733,
      "learning_rate": 0.0002,
      "loss": 1.5882,
      "step": 865
    },
    {
      "epoch": 0.31605839416058396,
      "grad_norm": 0.13359582424163818,
      "learning_rate": 0.0002,
      "loss": 1.5893,
      "step": 866
    },
    {
      "epoch": 0.3164233576642336,
      "grad_norm": 0.13061659038066864,
      "learning_rate": 0.0002,
      "loss": 1.4909,
      "step": 867
    },
    {
      "epoch": 0.3167883211678832,
      "grad_norm": 0.13096104562282562,
      "learning_rate": 0.0002,
      "loss": 1.5793,
      "step": 868
    },
    {
      "epoch": 0.31715328467153286,
      "grad_norm": 0.1310334950685501,
      "learning_rate": 0.0002,
      "loss": 1.5398,
      "step": 869
    },
    {
      "epoch": 0.3175182481751825,
      "grad_norm": 0.13355690240859985,
      "learning_rate": 0.0002,
      "loss": 1.5818,
      "step": 870
    },
    {
      "epoch": 0.3178832116788321,
      "grad_norm": 0.13097548484802246,
      "learning_rate": 0.0002,
      "loss": 1.5801,
      "step": 871
    },
    {
      "epoch": 0.3182481751824818,
      "grad_norm": 0.1365012675523758,
      "learning_rate": 0.0002,
      "loss": 1.6612,
      "step": 872
    },
    {
      "epoch": 0.3186131386861314,
      "grad_norm": 0.1399461328983307,
      "learning_rate": 0.0002,
      "loss": 1.5981,
      "step": 873
    },
    {
      "epoch": 0.318978102189781,
      "grad_norm": 0.13965493440628052,
      "learning_rate": 0.0002,
      "loss": 1.6111,
      "step": 874
    },
    {
      "epoch": 0.3193430656934307,
      "grad_norm": 0.13661295175552368,
      "learning_rate": 0.0002,
      "loss": 1.5989,
      "step": 875
    },
    {
      "epoch": 0.3197080291970803,
      "grad_norm": 0.13345752656459808,
      "learning_rate": 0.0002,
      "loss": 1.5211,
      "step": 876
    },
    {
      "epoch": 0.3200729927007299,
      "grad_norm": 0.13443291187286377,
      "learning_rate": 0.0002,
      "loss": 1.5474,
      "step": 877
    },
    {
      "epoch": 0.32043795620437954,
      "grad_norm": 0.13409192860126495,
      "learning_rate": 0.0002,
      "loss": 1.6163,
      "step": 878
    },
    {
      "epoch": 0.3208029197080292,
      "grad_norm": 0.1281881481409073,
      "learning_rate": 0.0002,
      "loss": 1.5928,
      "step": 879
    },
    {
      "epoch": 0.32116788321167883,
      "grad_norm": 0.1375834196805954,
      "learning_rate": 0.0002,
      "loss": 1.52,
      "step": 880
    },
    {
      "epoch": 0.32153284671532845,
      "grad_norm": 0.13428941369056702,
      "learning_rate": 0.0002,
      "loss": 1.4875,
      "step": 881
    },
    {
      "epoch": 0.3218978102189781,
      "grad_norm": 0.1299605667591095,
      "learning_rate": 0.0002,
      "loss": 1.5637,
      "step": 882
    },
    {
      "epoch": 0.32226277372262774,
      "grad_norm": 0.13886813819408417,
      "learning_rate": 0.0002,
      "loss": 1.6228,
      "step": 883
    },
    {
      "epoch": 0.32262773722627736,
      "grad_norm": 0.1320122480392456,
      "learning_rate": 0.0002,
      "loss": 1.6124,
      "step": 884
    },
    {
      "epoch": 0.32299270072992703,
      "grad_norm": 0.13286706805229187,
      "learning_rate": 0.0002,
      "loss": 1.5626,
      "step": 885
    },
    {
      "epoch": 0.32335766423357665,
      "grad_norm": 0.12869195640087128,
      "learning_rate": 0.0002,
      "loss": 1.5601,
      "step": 886
    },
    {
      "epoch": 0.32372262773722627,
      "grad_norm": 0.13522611558437347,
      "learning_rate": 0.0002,
      "loss": 1.6511,
      "step": 887
    },
    {
      "epoch": 0.32408759124087594,
      "grad_norm": 0.13006940484046936,
      "learning_rate": 0.0002,
      "loss": 1.5086,
      "step": 888
    },
    {
      "epoch": 0.32445255474452556,
      "grad_norm": 0.14511261880397797,
      "learning_rate": 0.0002,
      "loss": 1.6215,
      "step": 889
    },
    {
      "epoch": 0.3248175182481752,
      "grad_norm": 0.13274237513542175,
      "learning_rate": 0.0002,
      "loss": 1.5996,
      "step": 890
    },
    {
      "epoch": 0.3251824817518248,
      "grad_norm": 0.1374097466468811,
      "learning_rate": 0.0002,
      "loss": 1.5609,
      "step": 891
    },
    {
      "epoch": 0.32554744525547447,
      "grad_norm": 0.135096475481987,
      "learning_rate": 0.0002,
      "loss": 1.5464,
      "step": 892
    },
    {
      "epoch": 0.3259124087591241,
      "grad_norm": 0.1359672248363495,
      "learning_rate": 0.0002,
      "loss": 1.6007,
      "step": 893
    },
    {
      "epoch": 0.3262773722627737,
      "grad_norm": 0.13862727582454681,
      "learning_rate": 0.0002,
      "loss": 1.5852,
      "step": 894
    },
    {
      "epoch": 0.3266423357664234,
      "grad_norm": 0.13472682237625122,
      "learning_rate": 0.0002,
      "loss": 1.5357,
      "step": 895
    },
    {
      "epoch": 0.327007299270073,
      "grad_norm": 0.14301525056362152,
      "learning_rate": 0.0002,
      "loss": 1.6282,
      "step": 896
    },
    {
      "epoch": 0.3273722627737226,
      "grad_norm": 0.13611853122711182,
      "learning_rate": 0.0002,
      "loss": 1.5184,
      "step": 897
    },
    {
      "epoch": 0.3277372262773723,
      "grad_norm": 0.14719192683696747,
      "learning_rate": 0.0002,
      "loss": 1.6022,
      "step": 898
    },
    {
      "epoch": 0.3281021897810219,
      "grad_norm": 0.12869606912136078,
      "learning_rate": 0.0002,
      "loss": 1.4819,
      "step": 899
    },
    {
      "epoch": 0.3284671532846715,
      "grad_norm": 0.14684858918190002,
      "learning_rate": 0.0002,
      "loss": 1.5521,
      "step": 900
    },
    {
      "epoch": 0.3288321167883212,
      "grad_norm": 0.13548345863819122,
      "learning_rate": 0.0002,
      "loss": 1.6435,
      "step": 901
    },
    {
      "epoch": 0.3291970802919708,
      "grad_norm": 0.14584629237651825,
      "learning_rate": 0.0002,
      "loss": 1.5597,
      "step": 902
    },
    {
      "epoch": 0.32956204379562043,
      "grad_norm": 0.12583377957344055,
      "learning_rate": 0.0002,
      "loss": 1.5357,
      "step": 903
    },
    {
      "epoch": 0.32992700729927005,
      "grad_norm": 0.13808879256248474,
      "learning_rate": 0.0002,
      "loss": 1.5584,
      "step": 904
    },
    {
      "epoch": 0.3302919708029197,
      "grad_norm": 0.1320180743932724,
      "learning_rate": 0.0002,
      "loss": 1.5972,
      "step": 905
    },
    {
      "epoch": 0.33065693430656934,
      "grad_norm": 0.1340479701757431,
      "learning_rate": 0.0002,
      "loss": 1.5459,
      "step": 906
    },
    {
      "epoch": 0.33102189781021896,
      "grad_norm": 0.12893833220005035,
      "learning_rate": 0.0002,
      "loss": 1.5647,
      "step": 907
    },
    {
      "epoch": 0.33138686131386863,
      "grad_norm": 0.14115484058856964,
      "learning_rate": 0.0002,
      "loss": 1.6559,
      "step": 908
    },
    {
      "epoch": 0.33175182481751825,
      "grad_norm": 0.147894024848938,
      "learning_rate": 0.0002,
      "loss": 1.6128,
      "step": 909
    },
    {
      "epoch": 0.33211678832116787,
      "grad_norm": 0.14696002006530762,
      "learning_rate": 0.0002,
      "loss": 1.5776,
      "step": 910
    },
    {
      "epoch": 0.33248175182481754,
      "grad_norm": 0.13661448657512665,
      "learning_rate": 0.0002,
      "loss": 1.5762,
      "step": 911
    },
    {
      "epoch": 0.33284671532846716,
      "grad_norm": 0.13727769255638123,
      "learning_rate": 0.0002,
      "loss": 1.4746,
      "step": 912
    },
    {
      "epoch": 0.3332116788321168,
      "grad_norm": 0.13155414164066315,
      "learning_rate": 0.0002,
      "loss": 1.5771,
      "step": 913
    },
    {
      "epoch": 0.33357664233576645,
      "grad_norm": 0.14095626771450043,
      "learning_rate": 0.0002,
      "loss": 1.5812,
      "step": 914
    },
    {
      "epoch": 0.33394160583941607,
      "grad_norm": 0.13072773814201355,
      "learning_rate": 0.0002,
      "loss": 1.4811,
      "step": 915
    },
    {
      "epoch": 0.3343065693430657,
      "grad_norm": 0.14289386570453644,
      "learning_rate": 0.0002,
      "loss": 1.6005,
      "step": 916
    },
    {
      "epoch": 0.3346715328467153,
      "grad_norm": 0.13968724012374878,
      "learning_rate": 0.0002,
      "loss": 1.5748,
      "step": 917
    },
    {
      "epoch": 0.335036496350365,
      "grad_norm": 0.14404451847076416,
      "learning_rate": 0.0002,
      "loss": 1.5475,
      "step": 918
    },
    {
      "epoch": 0.3354014598540146,
      "grad_norm": 0.12679162621498108,
      "learning_rate": 0.0002,
      "loss": 1.4903,
      "step": 919
    },
    {
      "epoch": 0.3357664233576642,
      "grad_norm": 0.13299092650413513,
      "learning_rate": 0.0002,
      "loss": 1.4898,
      "step": 920
    },
    {
      "epoch": 0.3361313868613139,
      "grad_norm": 0.1343211978673935,
      "learning_rate": 0.0002,
      "loss": 1.5938,
      "step": 921
    },
    {
      "epoch": 0.3364963503649635,
      "grad_norm": 0.13752199709415436,
      "learning_rate": 0.0002,
      "loss": 1.5435,
      "step": 922
    },
    {
      "epoch": 0.3368613138686131,
      "grad_norm": 0.13225357234477997,
      "learning_rate": 0.0002,
      "loss": 1.565,
      "step": 923
    },
    {
      "epoch": 0.3372262773722628,
      "grad_norm": 0.1349058747291565,
      "learning_rate": 0.0002,
      "loss": 1.5151,
      "step": 924
    },
    {
      "epoch": 0.3375912408759124,
      "grad_norm": 0.13787759840488434,
      "learning_rate": 0.0002,
      "loss": 1.4992,
      "step": 925
    },
    {
      "epoch": 0.33795620437956203,
      "grad_norm": 0.14459609985351562,
      "learning_rate": 0.0002,
      "loss": 1.5783,
      "step": 926
    },
    {
      "epoch": 0.3383211678832117,
      "grad_norm": 0.12590748071670532,
      "learning_rate": 0.0002,
      "loss": 1.5157,
      "step": 927
    },
    {
      "epoch": 0.3386861313868613,
      "grad_norm": 0.13707707822322845,
      "learning_rate": 0.0002,
      "loss": 1.5511,
      "step": 928
    },
    {
      "epoch": 0.33905109489051094,
      "grad_norm": 0.13401585817337036,
      "learning_rate": 0.0002,
      "loss": 1.6275,
      "step": 929
    },
    {
      "epoch": 0.33941605839416056,
      "grad_norm": 0.13518193364143372,
      "learning_rate": 0.0002,
      "loss": 1.5691,
      "step": 930
    },
    {
      "epoch": 0.33978102189781023,
      "grad_norm": 0.13464201986789703,
      "learning_rate": 0.0002,
      "loss": 1.6213,
      "step": 931
    },
    {
      "epoch": 0.34014598540145985,
      "grad_norm": 0.13671407103538513,
      "learning_rate": 0.0002,
      "loss": 1.5593,
      "step": 932
    },
    {
      "epoch": 0.34051094890510947,
      "grad_norm": 0.13407012820243835,
      "learning_rate": 0.0002,
      "loss": 1.5506,
      "step": 933
    },
    {
      "epoch": 0.34087591240875914,
      "grad_norm": 0.13287101686000824,
      "learning_rate": 0.0002,
      "loss": 1.5503,
      "step": 934
    },
    {
      "epoch": 0.34124087591240876,
      "grad_norm": 0.13110074400901794,
      "learning_rate": 0.0002,
      "loss": 1.5029,
      "step": 935
    },
    {
      "epoch": 0.3416058394160584,
      "grad_norm": 0.13166947662830353,
      "learning_rate": 0.0002,
      "loss": 1.5492,
      "step": 936
    },
    {
      "epoch": 0.34197080291970805,
      "grad_norm": 0.13468553125858307,
      "learning_rate": 0.0002,
      "loss": 1.5987,
      "step": 937
    },
    {
      "epoch": 0.34233576642335767,
      "grad_norm": 0.13085296750068665,
      "learning_rate": 0.0002,
      "loss": 1.5727,
      "step": 938
    },
    {
      "epoch": 0.3427007299270073,
      "grad_norm": 0.13228477537631989,
      "learning_rate": 0.0002,
      "loss": 1.5754,
      "step": 939
    },
    {
      "epoch": 0.34306569343065696,
      "grad_norm": 0.13589254021644592,
      "learning_rate": 0.0002,
      "loss": 1.5645,
      "step": 940
    },
    {
      "epoch": 0.3434306569343066,
      "grad_norm": 0.14200060069561005,
      "learning_rate": 0.0002,
      "loss": 1.6408,
      "step": 941
    },
    {
      "epoch": 0.3437956204379562,
      "grad_norm": 0.13422203063964844,
      "learning_rate": 0.0002,
      "loss": 1.5764,
      "step": 942
    },
    {
      "epoch": 0.3441605839416058,
      "grad_norm": 0.13272887468338013,
      "learning_rate": 0.0002,
      "loss": 1.5747,
      "step": 943
    },
    {
      "epoch": 0.3445255474452555,
      "grad_norm": 0.14351576566696167,
      "learning_rate": 0.0002,
      "loss": 1.5845,
      "step": 944
    },
    {
      "epoch": 0.3448905109489051,
      "grad_norm": 0.12829773128032684,
      "learning_rate": 0.0002,
      "loss": 1.4224,
      "step": 945
    },
    {
      "epoch": 0.3452554744525547,
      "grad_norm": 0.13019244372844696,
      "learning_rate": 0.0002,
      "loss": 1.5793,
      "step": 946
    },
    {
      "epoch": 0.3456204379562044,
      "grad_norm": 0.1384381651878357,
      "learning_rate": 0.0002,
      "loss": 1.6389,
      "step": 947
    },
    {
      "epoch": 0.345985401459854,
      "grad_norm": 0.13548138737678528,
      "learning_rate": 0.0002,
      "loss": 1.5775,
      "step": 948
    },
    {
      "epoch": 0.34635036496350363,
      "grad_norm": 0.13311879336833954,
      "learning_rate": 0.0002,
      "loss": 1.5864,
      "step": 949
    },
    {
      "epoch": 0.3467153284671533,
      "grad_norm": 0.12944196164608002,
      "learning_rate": 0.0002,
      "loss": 1.5375,
      "step": 950
    },
    {
      "epoch": 0.3470802919708029,
      "grad_norm": 0.1353893280029297,
      "learning_rate": 0.0002,
      "loss": 1.6082,
      "step": 951
    },
    {
      "epoch": 0.34744525547445254,
      "grad_norm": 0.1385529339313507,
      "learning_rate": 0.0002,
      "loss": 1.5264,
      "step": 952
    },
    {
      "epoch": 0.3478102189781022,
      "grad_norm": 0.13875916600227356,
      "learning_rate": 0.0002,
      "loss": 1.5472,
      "step": 953
    },
    {
      "epoch": 0.34817518248175183,
      "grad_norm": 0.13275901973247528,
      "learning_rate": 0.0002,
      "loss": 1.5655,
      "step": 954
    },
    {
      "epoch": 0.34854014598540145,
      "grad_norm": 0.13849909603595734,
      "learning_rate": 0.0002,
      "loss": 1.5487,
      "step": 955
    },
    {
      "epoch": 0.34890510948905107,
      "grad_norm": 0.1340492069721222,
      "learning_rate": 0.0002,
      "loss": 1.5792,
      "step": 956
    },
    {
      "epoch": 0.34927007299270074,
      "grad_norm": 0.13207924365997314,
      "learning_rate": 0.0002,
      "loss": 1.5556,
      "step": 957
    },
    {
      "epoch": 0.34963503649635036,
      "grad_norm": 0.12910471856594086,
      "learning_rate": 0.0002,
      "loss": 1.5341,
      "step": 958
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.1342688798904419,
      "learning_rate": 0.0002,
      "loss": 1.6048,
      "step": 959
    },
    {
      "epoch": 0.35036496350364965,
      "grad_norm": 0.13399052619934082,
      "learning_rate": 0.0002,
      "loss": 1.6512,
      "step": 960
    },
    {
      "epoch": 0.35072992700729927,
      "grad_norm": 0.13228823244571686,
      "learning_rate": 0.0002,
      "loss": 1.5401,
      "step": 961
    },
    {
      "epoch": 0.3510948905109489,
      "grad_norm": 0.13953416049480438,
      "learning_rate": 0.0002,
      "loss": 1.6175,
      "step": 962
    },
    {
      "epoch": 0.35145985401459856,
      "grad_norm": 0.13289642333984375,
      "learning_rate": 0.0002,
      "loss": 1.4731,
      "step": 963
    },
    {
      "epoch": 0.3518248175182482,
      "grad_norm": 0.14261065423488617,
      "learning_rate": 0.0002,
      "loss": 1.5812,
      "step": 964
    },
    {
      "epoch": 0.3521897810218978,
      "grad_norm": 0.13481825590133667,
      "learning_rate": 0.0002,
      "loss": 1.5748,
      "step": 965
    },
    {
      "epoch": 0.35255474452554747,
      "grad_norm": 0.14046631753444672,
      "learning_rate": 0.0002,
      "loss": 1.6248,
      "step": 966
    },
    {
      "epoch": 0.3529197080291971,
      "grad_norm": 0.13219691812992096,
      "learning_rate": 0.0002,
      "loss": 1.5979,
      "step": 967
    },
    {
      "epoch": 0.3532846715328467,
      "grad_norm": 0.139474019408226,
      "learning_rate": 0.0002,
      "loss": 1.5551,
      "step": 968
    },
    {
      "epoch": 0.3536496350364963,
      "grad_norm": 0.13077028095722198,
      "learning_rate": 0.0002,
      "loss": 1.5047,
      "step": 969
    },
    {
      "epoch": 0.354014598540146,
      "grad_norm": 0.13471755385398865,
      "learning_rate": 0.0002,
      "loss": 1.6016,
      "step": 970
    },
    {
      "epoch": 0.3543795620437956,
      "grad_norm": 0.1340324729681015,
      "learning_rate": 0.0002,
      "loss": 1.6247,
      "step": 971
    },
    {
      "epoch": 0.35474452554744523,
      "grad_norm": 0.12820014357566833,
      "learning_rate": 0.0002,
      "loss": 1.4884,
      "step": 972
    },
    {
      "epoch": 0.3551094890510949,
      "grad_norm": 0.12773315608501434,
      "learning_rate": 0.0002,
      "loss": 1.5617,
      "step": 973
    },
    {
      "epoch": 0.3554744525547445,
      "grad_norm": 0.13053463399410248,
      "learning_rate": 0.0002,
      "loss": 1.5345,
      "step": 974
    },
    {
      "epoch": 0.35583941605839414,
      "grad_norm": 0.13560952246189117,
      "learning_rate": 0.0002,
      "loss": 1.6682,
      "step": 975
    },
    {
      "epoch": 0.3562043795620438,
      "grad_norm": 0.1307869702577591,
      "learning_rate": 0.0002,
      "loss": 1.5461,
      "step": 976
    },
    {
      "epoch": 0.35656934306569343,
      "grad_norm": 0.13059425354003906,
      "learning_rate": 0.0002,
      "loss": 1.5918,
      "step": 977
    },
    {
      "epoch": 0.35693430656934305,
      "grad_norm": 0.13018213212490082,
      "learning_rate": 0.0002,
      "loss": 1.4367,
      "step": 978
    },
    {
      "epoch": 0.3572992700729927,
      "grad_norm": 0.13409702479839325,
      "learning_rate": 0.0002,
      "loss": 1.5875,
      "step": 979
    },
    {
      "epoch": 0.35766423357664234,
      "grad_norm": 0.13513357937335968,
      "learning_rate": 0.0002,
      "loss": 1.4842,
      "step": 980
    },
    {
      "epoch": 0.35802919708029196,
      "grad_norm": 0.14140664041042328,
      "learning_rate": 0.0002,
      "loss": 1.6396,
      "step": 981
    },
    {
      "epoch": 0.3583941605839416,
      "grad_norm": 0.14239054918289185,
      "learning_rate": 0.0002,
      "loss": 1.5062,
      "step": 982
    },
    {
      "epoch": 0.35875912408759125,
      "grad_norm": 0.13062842190265656,
      "learning_rate": 0.0002,
      "loss": 1.5236,
      "step": 983
    },
    {
      "epoch": 0.35912408759124087,
      "grad_norm": 0.13725702464580536,
      "learning_rate": 0.0002,
      "loss": 1.5365,
      "step": 984
    },
    {
      "epoch": 0.3594890510948905,
      "grad_norm": 0.13594689965248108,
      "learning_rate": 0.0002,
      "loss": 1.5931,
      "step": 985
    },
    {
      "epoch": 0.35985401459854016,
      "grad_norm": 0.13411976397037506,
      "learning_rate": 0.0002,
      "loss": 1.5559,
      "step": 986
    },
    {
      "epoch": 0.3602189781021898,
      "grad_norm": 0.1319122314453125,
      "learning_rate": 0.0002,
      "loss": 1.535,
      "step": 987
    },
    {
      "epoch": 0.3605839416058394,
      "grad_norm": 0.132018581032753,
      "learning_rate": 0.0002,
      "loss": 1.4862,
      "step": 988
    },
    {
      "epoch": 0.36094890510948907,
      "grad_norm": 0.13655546307563782,
      "learning_rate": 0.0002,
      "loss": 1.6169,
      "step": 989
    },
    {
      "epoch": 0.3613138686131387,
      "grad_norm": 0.13056759536266327,
      "learning_rate": 0.0002,
      "loss": 1.5027,
      "step": 990
    },
    {
      "epoch": 0.3616788321167883,
      "grad_norm": 0.143804132938385,
      "learning_rate": 0.0002,
      "loss": 1.6497,
      "step": 991
    },
    {
      "epoch": 0.362043795620438,
      "grad_norm": 0.1434449851512909,
      "learning_rate": 0.0002,
      "loss": 1.6178,
      "step": 992
    },
    {
      "epoch": 0.3624087591240876,
      "grad_norm": 0.13931259512901306,
      "learning_rate": 0.0002,
      "loss": 1.58,
      "step": 993
    },
    {
      "epoch": 0.3627737226277372,
      "grad_norm": 0.13367675244808197,
      "learning_rate": 0.0002,
      "loss": 1.5644,
      "step": 994
    },
    {
      "epoch": 0.36313868613138683,
      "grad_norm": 0.12481454759836197,
      "learning_rate": 0.0002,
      "loss": 1.4349,
      "step": 995
    },
    {
      "epoch": 0.3635036496350365,
      "grad_norm": 0.1399364322423935,
      "learning_rate": 0.0002,
      "loss": 1.5495,
      "step": 996
    },
    {
      "epoch": 0.3638686131386861,
      "grad_norm": 0.1417096108198166,
      "learning_rate": 0.0002,
      "loss": 1.5408,
      "step": 997
    },
    {
      "epoch": 0.36423357664233574,
      "grad_norm": 0.13380427658557892,
      "learning_rate": 0.0002,
      "loss": 1.5287,
      "step": 998
    },
    {
      "epoch": 0.3645985401459854,
      "grad_norm": 0.14945568144321442,
      "learning_rate": 0.0002,
      "loss": 1.568,
      "step": 999
    },
    {
      "epoch": 0.36496350364963503,
      "grad_norm": 0.13691669702529907,
      "learning_rate": 0.0002,
      "loss": 1.6233,
      "step": 1000
    },
    {
      "epoch": 0.36532846715328465,
      "grad_norm": 0.13686369359493256,
      "learning_rate": 0.0002,
      "loss": 1.4809,
      "step": 1001
    },
    {
      "epoch": 0.3656934306569343,
      "grad_norm": 0.14579208195209503,
      "learning_rate": 0.0002,
      "loss": 1.5057,
      "step": 1002
    },
    {
      "epoch": 0.36605839416058394,
      "grad_norm": 0.13766419887542725,
      "learning_rate": 0.0002,
      "loss": 1.5683,
      "step": 1003
    },
    {
      "epoch": 0.36642335766423356,
      "grad_norm": 0.13893380761146545,
      "learning_rate": 0.0002,
      "loss": 1.4822,
      "step": 1004
    },
    {
      "epoch": 0.36678832116788324,
      "grad_norm": 0.14090770483016968,
      "learning_rate": 0.0002,
      "loss": 1.6369,
      "step": 1005
    },
    {
      "epoch": 0.36715328467153285,
      "grad_norm": 0.14497576653957367,
      "learning_rate": 0.0002,
      "loss": 1.583,
      "step": 1006
    },
    {
      "epoch": 0.36751824817518247,
      "grad_norm": 0.1413169801235199,
      "learning_rate": 0.0002,
      "loss": 1.6061,
      "step": 1007
    },
    {
      "epoch": 0.3678832116788321,
      "grad_norm": 0.14227408170700073,
      "learning_rate": 0.0002,
      "loss": 1.6376,
      "step": 1008
    },
    {
      "epoch": 0.36824817518248176,
      "grad_norm": 0.12946757674217224,
      "learning_rate": 0.0002,
      "loss": 1.5183,
      "step": 1009
    },
    {
      "epoch": 0.3686131386861314,
      "grad_norm": 0.1349058747291565,
      "learning_rate": 0.0002,
      "loss": 1.5608,
      "step": 1010
    },
    {
      "epoch": 0.368978102189781,
      "grad_norm": 0.13192005455493927,
      "learning_rate": 0.0002,
      "loss": 1.5196,
      "step": 1011
    },
    {
      "epoch": 0.36934306569343067,
      "grad_norm": 0.13507862389087677,
      "learning_rate": 0.0002,
      "loss": 1.5684,
      "step": 1012
    },
    {
      "epoch": 0.3697080291970803,
      "grad_norm": 0.1369432657957077,
      "learning_rate": 0.0002,
      "loss": 1.5563,
      "step": 1013
    },
    {
      "epoch": 0.3700729927007299,
      "grad_norm": 0.13944852352142334,
      "learning_rate": 0.0002,
      "loss": 1.5891,
      "step": 1014
    },
    {
      "epoch": 0.3704379562043796,
      "grad_norm": 0.1363823562860489,
      "learning_rate": 0.0002,
      "loss": 1.5322,
      "step": 1015
    },
    {
      "epoch": 0.3708029197080292,
      "grad_norm": 0.136458158493042,
      "learning_rate": 0.0002,
      "loss": 1.5703,
      "step": 1016
    },
    {
      "epoch": 0.3711678832116788,
      "grad_norm": 0.13440953195095062,
      "learning_rate": 0.0002,
      "loss": 1.5507,
      "step": 1017
    },
    {
      "epoch": 0.3715328467153285,
      "grad_norm": 0.13421760499477386,
      "learning_rate": 0.0002,
      "loss": 1.5452,
      "step": 1018
    },
    {
      "epoch": 0.3718978102189781,
      "grad_norm": 0.13289391994476318,
      "learning_rate": 0.0002,
      "loss": 1.5639,
      "step": 1019
    },
    {
      "epoch": 0.3722627737226277,
      "grad_norm": 0.13232095539569855,
      "learning_rate": 0.0002,
      "loss": 1.4828,
      "step": 1020
    },
    {
      "epoch": 0.37262773722627734,
      "grad_norm": 0.1331855207681656,
      "learning_rate": 0.0002,
      "loss": 1.5522,
      "step": 1021
    },
    {
      "epoch": 0.372992700729927,
      "grad_norm": 0.13882647454738617,
      "learning_rate": 0.0002,
      "loss": 1.5758,
      "step": 1022
    },
    {
      "epoch": 0.37335766423357664,
      "grad_norm": 0.13548830151557922,
      "learning_rate": 0.0002,
      "loss": 1.5727,
      "step": 1023
    },
    {
      "epoch": 0.37372262773722625,
      "grad_norm": 0.1376398652791977,
      "learning_rate": 0.0002,
      "loss": 1.6185,
      "step": 1024
    },
    {
      "epoch": 0.3740875912408759,
      "grad_norm": 0.13582783937454224,
      "learning_rate": 0.0002,
      "loss": 1.5684,
      "step": 1025
    },
    {
      "epoch": 0.37445255474452555,
      "grad_norm": 0.1420818269252777,
      "learning_rate": 0.0002,
      "loss": 1.533,
      "step": 1026
    },
    {
      "epoch": 0.37481751824817516,
      "grad_norm": 0.14288930594921112,
      "learning_rate": 0.0002,
      "loss": 1.5262,
      "step": 1027
    },
    {
      "epoch": 0.37518248175182484,
      "grad_norm": 0.14483119547367096,
      "learning_rate": 0.0002,
      "loss": 1.6119,
      "step": 1028
    },
    {
      "epoch": 0.37554744525547445,
      "grad_norm": 0.14891919493675232,
      "learning_rate": 0.0002,
      "loss": 1.5426,
      "step": 1029
    },
    {
      "epoch": 0.3759124087591241,
      "grad_norm": 0.13630136847496033,
      "learning_rate": 0.0002,
      "loss": 1.5778,
      "step": 1030
    },
    {
      "epoch": 0.37627737226277375,
      "grad_norm": 0.14587748050689697,
      "learning_rate": 0.0002,
      "loss": 1.5906,
      "step": 1031
    },
    {
      "epoch": 0.37664233576642336,
      "grad_norm": 0.12906284630298615,
      "learning_rate": 0.0002,
      "loss": 1.5378,
      "step": 1032
    },
    {
      "epoch": 0.377007299270073,
      "grad_norm": 0.14113305509090424,
      "learning_rate": 0.0002,
      "loss": 1.5399,
      "step": 1033
    },
    {
      "epoch": 0.37737226277372266,
      "grad_norm": 0.141463041305542,
      "learning_rate": 0.0002,
      "loss": 1.5807,
      "step": 1034
    },
    {
      "epoch": 0.3777372262773723,
      "grad_norm": 0.13525699079036713,
      "learning_rate": 0.0002,
      "loss": 1.5181,
      "step": 1035
    },
    {
      "epoch": 0.3781021897810219,
      "grad_norm": 0.1431453377008438,
      "learning_rate": 0.0002,
      "loss": 1.6307,
      "step": 1036
    },
    {
      "epoch": 0.3784671532846715,
      "grad_norm": 0.13314220309257507,
      "learning_rate": 0.0002,
      "loss": 1.524,
      "step": 1037
    },
    {
      "epoch": 0.3788321167883212,
      "grad_norm": 0.14402706921100616,
      "learning_rate": 0.0002,
      "loss": 1.5647,
      "step": 1038
    },
    {
      "epoch": 0.3791970802919708,
      "grad_norm": 0.13923339545726776,
      "learning_rate": 0.0002,
      "loss": 1.5803,
      "step": 1039
    },
    {
      "epoch": 0.3795620437956204,
      "grad_norm": 0.13599994778633118,
      "learning_rate": 0.0002,
      "loss": 1.5248,
      "step": 1040
    },
    {
      "epoch": 0.3799270072992701,
      "grad_norm": 0.13988648355007172,
      "learning_rate": 0.0002,
      "loss": 1.5758,
      "step": 1041
    },
    {
      "epoch": 0.3802919708029197,
      "grad_norm": 0.1425476372241974,
      "learning_rate": 0.0002,
      "loss": 1.6171,
      "step": 1042
    },
    {
      "epoch": 0.38065693430656933,
      "grad_norm": 0.14013944566249847,
      "learning_rate": 0.0002,
      "loss": 1.5546,
      "step": 1043
    },
    {
      "epoch": 0.381021897810219,
      "grad_norm": 0.13543270528316498,
      "learning_rate": 0.0002,
      "loss": 1.5801,
      "step": 1044
    },
    {
      "epoch": 0.3813868613138686,
      "grad_norm": 0.13897669315338135,
      "learning_rate": 0.0002,
      "loss": 1.6159,
      "step": 1045
    },
    {
      "epoch": 0.38175182481751824,
      "grad_norm": 0.14069108664989471,
      "learning_rate": 0.0002,
      "loss": 1.499,
      "step": 1046
    },
    {
      "epoch": 0.3821167883211679,
      "grad_norm": 0.13422967493534088,
      "learning_rate": 0.0002,
      "loss": 1.4808,
      "step": 1047
    },
    {
      "epoch": 0.38248175182481753,
      "grad_norm": 0.1446090042591095,
      "learning_rate": 0.0002,
      "loss": 1.5566,
      "step": 1048
    },
    {
      "epoch": 0.38284671532846715,
      "grad_norm": 0.1389555037021637,
      "learning_rate": 0.0002,
      "loss": 1.6424,
      "step": 1049
    },
    {
      "epoch": 0.38321167883211676,
      "grad_norm": 0.1430179327726364,
      "learning_rate": 0.0002,
      "loss": 1.5331,
      "step": 1050
    },
    {
      "epoch": 0.38357664233576644,
      "grad_norm": 0.13515399396419525,
      "learning_rate": 0.0002,
      "loss": 1.5886,
      "step": 1051
    },
    {
      "epoch": 0.38394160583941606,
      "grad_norm": 0.1500077098608017,
      "learning_rate": 0.0002,
      "loss": 1.5055,
      "step": 1052
    },
    {
      "epoch": 0.3843065693430657,
      "grad_norm": 0.14432315528392792,
      "learning_rate": 0.0002,
      "loss": 1.5353,
      "step": 1053
    },
    {
      "epoch": 0.38467153284671535,
      "grad_norm": 0.13615576922893524,
      "learning_rate": 0.0002,
      "loss": 1.5256,
      "step": 1054
    },
    {
      "epoch": 0.38503649635036497,
      "grad_norm": 0.13361120223999023,
      "learning_rate": 0.0002,
      "loss": 1.5302,
      "step": 1055
    },
    {
      "epoch": 0.3854014598540146,
      "grad_norm": 0.14258600771427155,
      "learning_rate": 0.0002,
      "loss": 1.5977,
      "step": 1056
    },
    {
      "epoch": 0.38576642335766426,
      "grad_norm": 0.1377772092819214,
      "learning_rate": 0.0002,
      "loss": 1.4823,
      "step": 1057
    },
    {
      "epoch": 0.3861313868613139,
      "grad_norm": 0.14837057888507843,
      "learning_rate": 0.0002,
      "loss": 1.5273,
      "step": 1058
    },
    {
      "epoch": 0.3864963503649635,
      "grad_norm": 0.14119952917099,
      "learning_rate": 0.0002,
      "loss": 1.5527,
      "step": 1059
    },
    {
      "epoch": 0.38686131386861317,
      "grad_norm": 0.1325782835483551,
      "learning_rate": 0.0002,
      "loss": 1.5149,
      "step": 1060
    },
    {
      "epoch": 0.3872262773722628,
      "grad_norm": 0.1384919434785843,
      "learning_rate": 0.0002,
      "loss": 1.5617,
      "step": 1061
    },
    {
      "epoch": 0.3875912408759124,
      "grad_norm": 0.13949382305145264,
      "learning_rate": 0.0002,
      "loss": 1.597,
      "step": 1062
    },
    {
      "epoch": 0.387956204379562,
      "grad_norm": 0.1370716691017151,
      "learning_rate": 0.0002,
      "loss": 1.5716,
      "step": 1063
    },
    {
      "epoch": 0.3883211678832117,
      "grad_norm": 0.13323594629764557,
      "learning_rate": 0.0002,
      "loss": 1.502,
      "step": 1064
    },
    {
      "epoch": 0.3886861313868613,
      "grad_norm": 0.14397089183330536,
      "learning_rate": 0.0002,
      "loss": 1.551,
      "step": 1065
    },
    {
      "epoch": 0.38905109489051093,
      "grad_norm": 0.13238875567913055,
      "learning_rate": 0.0002,
      "loss": 1.5161,
      "step": 1066
    },
    {
      "epoch": 0.3894160583941606,
      "grad_norm": 0.134956955909729,
      "learning_rate": 0.0002,
      "loss": 1.5403,
      "step": 1067
    },
    {
      "epoch": 0.3897810218978102,
      "grad_norm": 0.13498876988887787,
      "learning_rate": 0.0002,
      "loss": 1.5578,
      "step": 1068
    },
    {
      "epoch": 0.39014598540145984,
      "grad_norm": 0.13566215336322784,
      "learning_rate": 0.0002,
      "loss": 1.4977,
      "step": 1069
    },
    {
      "epoch": 0.3905109489051095,
      "grad_norm": 0.1470859795808792,
      "learning_rate": 0.0002,
      "loss": 1.5366,
      "step": 1070
    },
    {
      "epoch": 0.39087591240875913,
      "grad_norm": 0.14386233687400818,
      "learning_rate": 0.0002,
      "loss": 1.5292,
      "step": 1071
    },
    {
      "epoch": 0.39124087591240875,
      "grad_norm": 0.1357048898935318,
      "learning_rate": 0.0002,
      "loss": 1.5305,
      "step": 1072
    },
    {
      "epoch": 0.3916058394160584,
      "grad_norm": 0.13078467547893524,
      "learning_rate": 0.0002,
      "loss": 1.4587,
      "step": 1073
    },
    {
      "epoch": 0.39197080291970804,
      "grad_norm": 0.1353101283311844,
      "learning_rate": 0.0002,
      "loss": 1.5682,
      "step": 1074
    },
    {
      "epoch": 0.39233576642335766,
      "grad_norm": 0.1443767249584198,
      "learning_rate": 0.0002,
      "loss": 1.6107,
      "step": 1075
    },
    {
      "epoch": 0.3927007299270073,
      "grad_norm": 0.13262800872325897,
      "learning_rate": 0.0002,
      "loss": 1.5458,
      "step": 1076
    },
    {
      "epoch": 0.39306569343065695,
      "grad_norm": 0.13317617774009705,
      "learning_rate": 0.0002,
      "loss": 1.5163,
      "step": 1077
    },
    {
      "epoch": 0.39343065693430657,
      "grad_norm": 0.13710258901119232,
      "learning_rate": 0.0002,
      "loss": 1.5336,
      "step": 1078
    },
    {
      "epoch": 0.3937956204379562,
      "grad_norm": 0.1334623545408249,
      "learning_rate": 0.0002,
      "loss": 1.5234,
      "step": 1079
    },
    {
      "epoch": 0.39416058394160586,
      "grad_norm": 0.13603362441062927,
      "learning_rate": 0.0002,
      "loss": 1.5446,
      "step": 1080
    },
    {
      "epoch": 0.3945255474452555,
      "grad_norm": 0.13301493227481842,
      "learning_rate": 0.0002,
      "loss": 1.5199,
      "step": 1081
    },
    {
      "epoch": 0.3948905109489051,
      "grad_norm": 0.1388171911239624,
      "learning_rate": 0.0002,
      "loss": 1.6136,
      "step": 1082
    },
    {
      "epoch": 0.39525547445255477,
      "grad_norm": 0.1399710774421692,
      "learning_rate": 0.0002,
      "loss": 1.4982,
      "step": 1083
    },
    {
      "epoch": 0.3956204379562044,
      "grad_norm": 0.13580727577209473,
      "learning_rate": 0.0002,
      "loss": 1.5099,
      "step": 1084
    },
    {
      "epoch": 0.395985401459854,
      "grad_norm": 0.14123041927814484,
      "learning_rate": 0.0002,
      "loss": 1.6337,
      "step": 1085
    },
    {
      "epoch": 0.3963503649635037,
      "grad_norm": 0.1365749090909958,
      "learning_rate": 0.0002,
      "loss": 1.5152,
      "step": 1086
    },
    {
      "epoch": 0.3967153284671533,
      "grad_norm": 0.13587340712547302,
      "learning_rate": 0.0002,
      "loss": 1.5634,
      "step": 1087
    },
    {
      "epoch": 0.3970802919708029,
      "grad_norm": 0.14085541665554047,
      "learning_rate": 0.0002,
      "loss": 1.5702,
      "step": 1088
    },
    {
      "epoch": 0.39744525547445253,
      "grad_norm": 0.14121417701244354,
      "learning_rate": 0.0002,
      "loss": 1.4942,
      "step": 1089
    },
    {
      "epoch": 0.3978102189781022,
      "grad_norm": 0.1375901997089386,
      "learning_rate": 0.0002,
      "loss": 1.5157,
      "step": 1090
    },
    {
      "epoch": 0.3981751824817518,
      "grad_norm": 0.14686931669712067,
      "learning_rate": 0.0002,
      "loss": 1.5759,
      "step": 1091
    },
    {
      "epoch": 0.39854014598540144,
      "grad_norm": 0.13673318922519684,
      "learning_rate": 0.0002,
      "loss": 1.6021,
      "step": 1092
    },
    {
      "epoch": 0.3989051094890511,
      "grad_norm": 0.15267592668533325,
      "learning_rate": 0.0002,
      "loss": 1.5715,
      "step": 1093
    },
    {
      "epoch": 0.39927007299270073,
      "grad_norm": 0.14170889556407928,
      "learning_rate": 0.0002,
      "loss": 1.5136,
      "step": 1094
    },
    {
      "epoch": 0.39963503649635035,
      "grad_norm": 0.14355076849460602,
      "learning_rate": 0.0002,
      "loss": 1.6174,
      "step": 1095
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.14519143104553223,
      "learning_rate": 0.0002,
      "loss": 1.531,
      "step": 1096
    },
    {
      "epoch": 0.40036496350364964,
      "grad_norm": 0.1410042941570282,
      "learning_rate": 0.0002,
      "loss": 1.5619,
      "step": 1097
    },
    {
      "epoch": 0.40072992700729926,
      "grad_norm": 0.14007176458835602,
      "learning_rate": 0.0002,
      "loss": 1.5713,
      "step": 1098
    },
    {
      "epoch": 0.40109489051094893,
      "grad_norm": 0.13386785984039307,
      "learning_rate": 0.0002,
      "loss": 1.4655,
      "step": 1099
    },
    {
      "epoch": 0.40145985401459855,
      "grad_norm": 0.13994398713111877,
      "learning_rate": 0.0002,
      "loss": 1.5282,
      "step": 1100
    },
    {
      "epoch": 0.40182481751824817,
      "grad_norm": 0.13837845623493195,
      "learning_rate": 0.0002,
      "loss": 1.5893,
      "step": 1101
    },
    {
      "epoch": 0.4021897810218978,
      "grad_norm": 0.13389818370342255,
      "learning_rate": 0.0002,
      "loss": 1.5713,
      "step": 1102
    },
    {
      "epoch": 0.40255474452554746,
      "grad_norm": 0.13901787996292114,
      "learning_rate": 0.0002,
      "loss": 1.5324,
      "step": 1103
    },
    {
      "epoch": 0.4029197080291971,
      "grad_norm": 0.13574528694152832,
      "learning_rate": 0.0002,
      "loss": 1.5176,
      "step": 1104
    },
    {
      "epoch": 0.4032846715328467,
      "grad_norm": 0.1355007141828537,
      "learning_rate": 0.0002,
      "loss": 1.5297,
      "step": 1105
    },
    {
      "epoch": 0.40364963503649637,
      "grad_norm": 0.1292802095413208,
      "learning_rate": 0.0002,
      "loss": 1.4666,
      "step": 1106
    },
    {
      "epoch": 0.404014598540146,
      "grad_norm": 0.13992749154567719,
      "learning_rate": 0.0002,
      "loss": 1.5689,
      "step": 1107
    },
    {
      "epoch": 0.4043795620437956,
      "grad_norm": 0.1357426643371582,
      "learning_rate": 0.0002,
      "loss": 1.5896,
      "step": 1108
    },
    {
      "epoch": 0.4047445255474453,
      "grad_norm": 0.14008718729019165,
      "learning_rate": 0.0002,
      "loss": 1.5201,
      "step": 1109
    },
    {
      "epoch": 0.4051094890510949,
      "grad_norm": 0.1355096399784088,
      "learning_rate": 0.0002,
      "loss": 1.5446,
      "step": 1110
    },
    {
      "epoch": 0.4054744525547445,
      "grad_norm": 0.13563284277915955,
      "learning_rate": 0.0002,
      "loss": 1.4853,
      "step": 1111
    },
    {
      "epoch": 0.4058394160583942,
      "grad_norm": 0.1444336175918579,
      "learning_rate": 0.0002,
      "loss": 1.5898,
      "step": 1112
    },
    {
      "epoch": 0.4062043795620438,
      "grad_norm": 0.1369090974330902,
      "learning_rate": 0.0002,
      "loss": 1.4933,
      "step": 1113
    },
    {
      "epoch": 0.4065693430656934,
      "grad_norm": 0.1418190598487854,
      "learning_rate": 0.0002,
      "loss": 1.5636,
      "step": 1114
    },
    {
      "epoch": 0.40693430656934304,
      "grad_norm": 0.13634535670280457,
      "learning_rate": 0.0002,
      "loss": 1.5217,
      "step": 1115
    },
    {
      "epoch": 0.4072992700729927,
      "grad_norm": 0.1382862627506256,
      "learning_rate": 0.0002,
      "loss": 1.6065,
      "step": 1116
    },
    {
      "epoch": 0.40766423357664233,
      "grad_norm": 0.13778282701969147,
      "learning_rate": 0.0002,
      "loss": 1.6298,
      "step": 1117
    },
    {
      "epoch": 0.40802919708029195,
      "grad_norm": 0.14385798573493958,
      "learning_rate": 0.0002,
      "loss": 1.6256,
      "step": 1118
    },
    {
      "epoch": 0.4083941605839416,
      "grad_norm": 0.1335172951221466,
      "learning_rate": 0.0002,
      "loss": 1.5589,
      "step": 1119
    },
    {
      "epoch": 0.40875912408759124,
      "grad_norm": 0.13923251628875732,
      "learning_rate": 0.0002,
      "loss": 1.5654,
      "step": 1120
    },
    {
      "epoch": 0.40912408759124086,
      "grad_norm": 0.13355012238025665,
      "learning_rate": 0.0002,
      "loss": 1.5638,
      "step": 1121
    },
    {
      "epoch": 0.40948905109489053,
      "grad_norm": 0.14127318561077118,
      "learning_rate": 0.0002,
      "loss": 1.633,
      "step": 1122
    },
    {
      "epoch": 0.40985401459854015,
      "grad_norm": 0.14448651671409607,
      "learning_rate": 0.0002,
      "loss": 1.5468,
      "step": 1123
    },
    {
      "epoch": 0.41021897810218977,
      "grad_norm": 0.13894985616207123,
      "learning_rate": 0.0002,
      "loss": 1.4935,
      "step": 1124
    },
    {
      "epoch": 0.41058394160583944,
      "grad_norm": 0.13718965649604797,
      "learning_rate": 0.0002,
      "loss": 1.5637,
      "step": 1125
    },
    {
      "epoch": 0.41094890510948906,
      "grad_norm": 0.14198271930217743,
      "learning_rate": 0.0002,
      "loss": 1.5138,
      "step": 1126
    },
    {
      "epoch": 0.4113138686131387,
      "grad_norm": 0.1334524005651474,
      "learning_rate": 0.0002,
      "loss": 1.5604,
      "step": 1127
    },
    {
      "epoch": 0.4116788321167883,
      "grad_norm": 0.14538952708244324,
      "learning_rate": 0.0002,
      "loss": 1.5824,
      "step": 1128
    },
    {
      "epoch": 0.41204379562043797,
      "grad_norm": 0.13651980459690094,
      "learning_rate": 0.0002,
      "loss": 1.5036,
      "step": 1129
    },
    {
      "epoch": 0.4124087591240876,
      "grad_norm": 0.13425470888614655,
      "learning_rate": 0.0002,
      "loss": 1.5071,
      "step": 1130
    },
    {
      "epoch": 0.4127737226277372,
      "grad_norm": 0.13964763283729553,
      "learning_rate": 0.0002,
      "loss": 1.5145,
      "step": 1131
    },
    {
      "epoch": 0.4131386861313869,
      "grad_norm": 0.14207245409488678,
      "learning_rate": 0.0002,
      "loss": 1.5438,
      "step": 1132
    },
    {
      "epoch": 0.4135036496350365,
      "grad_norm": 0.1380464732646942,
      "learning_rate": 0.0002,
      "loss": 1.5215,
      "step": 1133
    },
    {
      "epoch": 0.4138686131386861,
      "grad_norm": 0.14121858775615692,
      "learning_rate": 0.0002,
      "loss": 1.5938,
      "step": 1134
    },
    {
      "epoch": 0.4142335766423358,
      "grad_norm": 0.1410834640264511,
      "learning_rate": 0.0002,
      "loss": 1.5943,
      "step": 1135
    },
    {
      "epoch": 0.4145985401459854,
      "grad_norm": 0.1388968974351883,
      "learning_rate": 0.0002,
      "loss": 1.5211,
      "step": 1136
    },
    {
      "epoch": 0.414963503649635,
      "grad_norm": 0.1368446797132492,
      "learning_rate": 0.0002,
      "loss": 1.5678,
      "step": 1137
    },
    {
      "epoch": 0.4153284671532847,
      "grad_norm": 0.1324273645877838,
      "learning_rate": 0.0002,
      "loss": 1.4638,
      "step": 1138
    },
    {
      "epoch": 0.4156934306569343,
      "grad_norm": 0.13303260505199432,
      "learning_rate": 0.0002,
      "loss": 1.5298,
      "step": 1139
    },
    {
      "epoch": 0.41605839416058393,
      "grad_norm": 0.13546206057071686,
      "learning_rate": 0.0002,
      "loss": 1.5523,
      "step": 1140
    },
    {
      "epoch": 0.41642335766423355,
      "grad_norm": 0.13300347328186035,
      "learning_rate": 0.0002,
      "loss": 1.4596,
      "step": 1141
    },
    {
      "epoch": 0.4167883211678832,
      "grad_norm": 0.14920662343502045,
      "learning_rate": 0.0002,
      "loss": 1.6664,
      "step": 1142
    },
    {
      "epoch": 0.41715328467153284,
      "grad_norm": 0.13858824968338013,
      "learning_rate": 0.0002,
      "loss": 1.5432,
      "step": 1143
    },
    {
      "epoch": 0.41751824817518246,
      "grad_norm": 0.14495238661766052,
      "learning_rate": 0.0002,
      "loss": 1.5997,
      "step": 1144
    },
    {
      "epoch": 0.41788321167883213,
      "grad_norm": 0.1335591822862625,
      "learning_rate": 0.0002,
      "loss": 1.4958,
      "step": 1145
    },
    {
      "epoch": 0.41824817518248175,
      "grad_norm": 0.144364595413208,
      "learning_rate": 0.0002,
      "loss": 1.6203,
      "step": 1146
    },
    {
      "epoch": 0.41861313868613137,
      "grad_norm": 0.13621263206005096,
      "learning_rate": 0.0002,
      "loss": 1.5521,
      "step": 1147
    },
    {
      "epoch": 0.41897810218978104,
      "grad_norm": 0.14502960443496704,
      "learning_rate": 0.0002,
      "loss": 1.5902,
      "step": 1148
    },
    {
      "epoch": 0.41934306569343066,
      "grad_norm": 0.13748349249362946,
      "learning_rate": 0.0002,
      "loss": 1.5196,
      "step": 1149
    },
    {
      "epoch": 0.4197080291970803,
      "grad_norm": 0.14028848707675934,
      "learning_rate": 0.0002,
      "loss": 1.5479,
      "step": 1150
    },
    {
      "epoch": 0.42007299270072995,
      "grad_norm": 0.13015945255756378,
      "learning_rate": 0.0002,
      "loss": 1.4775,
      "step": 1151
    },
    {
      "epoch": 0.42043795620437957,
      "grad_norm": 0.13673484325408936,
      "learning_rate": 0.0002,
      "loss": 1.5426,
      "step": 1152
    },
    {
      "epoch": 0.4208029197080292,
      "grad_norm": 0.14188914000988007,
      "learning_rate": 0.0002,
      "loss": 1.6039,
      "step": 1153
    },
    {
      "epoch": 0.4211678832116788,
      "grad_norm": 0.1349174678325653,
      "learning_rate": 0.0002,
      "loss": 1.498,
      "step": 1154
    },
    {
      "epoch": 0.4215328467153285,
      "grad_norm": 0.1406264305114746,
      "learning_rate": 0.0002,
      "loss": 1.584,
      "step": 1155
    },
    {
      "epoch": 0.4218978102189781,
      "grad_norm": 0.1388646811246872,
      "learning_rate": 0.0002,
      "loss": 1.5303,
      "step": 1156
    },
    {
      "epoch": 0.4222627737226277,
      "grad_norm": 0.1416744589805603,
      "learning_rate": 0.0002,
      "loss": 1.5938,
      "step": 1157
    },
    {
      "epoch": 0.4226277372262774,
      "grad_norm": 0.1373978555202484,
      "learning_rate": 0.0002,
      "loss": 1.5215,
      "step": 1158
    },
    {
      "epoch": 0.422992700729927,
      "grad_norm": 0.1414439082145691,
      "learning_rate": 0.0002,
      "loss": 1.585,
      "step": 1159
    },
    {
      "epoch": 0.4233576642335766,
      "grad_norm": 0.1384410709142685,
      "learning_rate": 0.0002,
      "loss": 1.5487,
      "step": 1160
    },
    {
      "epoch": 0.4237226277372263,
      "grad_norm": 0.1379985809326172,
      "learning_rate": 0.0002,
      "loss": 1.5658,
      "step": 1161
    },
    {
      "epoch": 0.4240875912408759,
      "grad_norm": 0.14754989743232727,
      "learning_rate": 0.0002,
      "loss": 1.6112,
      "step": 1162
    },
    {
      "epoch": 0.42445255474452553,
      "grad_norm": 0.14544297754764557,
      "learning_rate": 0.0002,
      "loss": 1.5458,
      "step": 1163
    },
    {
      "epoch": 0.4248175182481752,
      "grad_norm": 0.1381036639213562,
      "learning_rate": 0.0002,
      "loss": 1.5198,
      "step": 1164
    },
    {
      "epoch": 0.4251824817518248,
      "grad_norm": 0.1456703245639801,
      "learning_rate": 0.0002,
      "loss": 1.5646,
      "step": 1165
    },
    {
      "epoch": 0.42554744525547444,
      "grad_norm": 0.147505521774292,
      "learning_rate": 0.0002,
      "loss": 1.6428,
      "step": 1166
    },
    {
      "epoch": 0.42591240875912406,
      "grad_norm": 0.14098583161830902,
      "learning_rate": 0.0002,
      "loss": 1.5447,
      "step": 1167
    },
    {
      "epoch": 0.42627737226277373,
      "grad_norm": 0.13693299889564514,
      "learning_rate": 0.0002,
      "loss": 1.5029,
      "step": 1168
    },
    {
      "epoch": 0.42664233576642335,
      "grad_norm": 0.13770417869091034,
      "learning_rate": 0.0002,
      "loss": 1.5021,
      "step": 1169
    },
    {
      "epoch": 0.42700729927007297,
      "grad_norm": 0.1386016607284546,
      "learning_rate": 0.0002,
      "loss": 1.4721,
      "step": 1170
    },
    {
      "epoch": 0.42737226277372264,
      "grad_norm": 0.14536960422992706,
      "learning_rate": 0.0002,
      "loss": 1.5193,
      "step": 1171
    },
    {
      "epoch": 0.42773722627737226,
      "grad_norm": 0.1453949213027954,
      "learning_rate": 0.0002,
      "loss": 1.5541,
      "step": 1172
    },
    {
      "epoch": 0.4281021897810219,
      "grad_norm": 0.14861036837100983,
      "learning_rate": 0.0002,
      "loss": 1.6185,
      "step": 1173
    },
    {
      "epoch": 0.42846715328467155,
      "grad_norm": 0.1559533327817917,
      "learning_rate": 0.0002,
      "loss": 1.6618,
      "step": 1174
    },
    {
      "epoch": 0.42883211678832117,
      "grad_norm": 0.13631470501422882,
      "learning_rate": 0.0002,
      "loss": 1.5627,
      "step": 1175
    },
    {
      "epoch": 0.4291970802919708,
      "grad_norm": 0.1408136487007141,
      "learning_rate": 0.0002,
      "loss": 1.4745,
      "step": 1176
    },
    {
      "epoch": 0.42956204379562046,
      "grad_norm": 0.13827139139175415,
      "learning_rate": 0.0002,
      "loss": 1.5647,
      "step": 1177
    },
    {
      "epoch": 0.4299270072992701,
      "grad_norm": 0.14211687445640564,
      "learning_rate": 0.0002,
      "loss": 1.5972,
      "step": 1178
    },
    {
      "epoch": 0.4302919708029197,
      "grad_norm": 0.14037221670150757,
      "learning_rate": 0.0002,
      "loss": 1.5634,
      "step": 1179
    },
    {
      "epoch": 0.4306569343065693,
      "grad_norm": 0.13586118817329407,
      "learning_rate": 0.0002,
      "loss": 1.5225,
      "step": 1180
    },
    {
      "epoch": 0.431021897810219,
      "grad_norm": 0.14081813395023346,
      "learning_rate": 0.0002,
      "loss": 1.5802,
      "step": 1181
    },
    {
      "epoch": 0.4313868613138686,
      "grad_norm": 0.13559667766094208,
      "learning_rate": 0.0002,
      "loss": 1.5042,
      "step": 1182
    },
    {
      "epoch": 0.4317518248175182,
      "grad_norm": 0.1443997025489807,
      "learning_rate": 0.0002,
      "loss": 1.6144,
      "step": 1183
    },
    {
      "epoch": 0.4321167883211679,
      "grad_norm": 0.14187034964561462,
      "learning_rate": 0.0002,
      "loss": 1.6273,
      "step": 1184
    },
    {
      "epoch": 0.4324817518248175,
      "grad_norm": 0.13649185001850128,
      "learning_rate": 0.0002,
      "loss": 1.5424,
      "step": 1185
    },
    {
      "epoch": 0.43284671532846714,
      "grad_norm": 0.1370546668767929,
      "learning_rate": 0.0002,
      "loss": 1.541,
      "step": 1186
    },
    {
      "epoch": 0.4332116788321168,
      "grad_norm": 0.14395225048065186,
      "learning_rate": 0.0002,
      "loss": 1.5433,
      "step": 1187
    },
    {
      "epoch": 0.4335766423357664,
      "grad_norm": 0.14330485463142395,
      "learning_rate": 0.0002,
      "loss": 1.5689,
      "step": 1188
    },
    {
      "epoch": 0.43394160583941604,
      "grad_norm": 0.1381494253873825,
      "learning_rate": 0.0002,
      "loss": 1.5895,
      "step": 1189
    },
    {
      "epoch": 0.4343065693430657,
      "grad_norm": 0.1341620683670044,
      "learning_rate": 0.0002,
      "loss": 1.516,
      "step": 1190
    },
    {
      "epoch": 0.43467153284671534,
      "grad_norm": 0.14151190221309662,
      "learning_rate": 0.0002,
      "loss": 1.5816,
      "step": 1191
    },
    {
      "epoch": 0.43503649635036495,
      "grad_norm": 0.13608698546886444,
      "learning_rate": 0.0002,
      "loss": 1.5924,
      "step": 1192
    },
    {
      "epoch": 0.43540145985401457,
      "grad_norm": 0.1350543200969696,
      "learning_rate": 0.0002,
      "loss": 1.5647,
      "step": 1193
    },
    {
      "epoch": 0.43576642335766425,
      "grad_norm": 0.14096246659755707,
      "learning_rate": 0.0002,
      "loss": 1.5384,
      "step": 1194
    },
    {
      "epoch": 0.43613138686131386,
      "grad_norm": 0.13685357570648193,
      "learning_rate": 0.0002,
      "loss": 1.5095,
      "step": 1195
    },
    {
      "epoch": 0.4364963503649635,
      "grad_norm": 0.14377324283123016,
      "learning_rate": 0.0002,
      "loss": 1.5661,
      "step": 1196
    },
    {
      "epoch": 0.43686131386861315,
      "grad_norm": 0.14045393466949463,
      "learning_rate": 0.0002,
      "loss": 1.5733,
      "step": 1197
    },
    {
      "epoch": 0.4372262773722628,
      "grad_norm": 0.14112839102745056,
      "learning_rate": 0.0002,
      "loss": 1.5694,
      "step": 1198
    },
    {
      "epoch": 0.4375912408759124,
      "grad_norm": 0.1403147131204605,
      "learning_rate": 0.0002,
      "loss": 1.6187,
      "step": 1199
    },
    {
      "epoch": 0.43795620437956206,
      "grad_norm": 0.149155855178833,
      "learning_rate": 0.0002,
      "loss": 1.5813,
      "step": 1200
    },
    {
      "epoch": 0.4383211678832117,
      "grad_norm": 0.1331096887588501,
      "learning_rate": 0.0002,
      "loss": 1.5135,
      "step": 1201
    },
    {
      "epoch": 0.4386861313868613,
      "grad_norm": 0.14569178223609924,
      "learning_rate": 0.0002,
      "loss": 1.5949,
      "step": 1202
    },
    {
      "epoch": 0.439051094890511,
      "grad_norm": 0.13567396998405457,
      "learning_rate": 0.0002,
      "loss": 1.5099,
      "step": 1203
    },
    {
      "epoch": 0.4394160583941606,
      "grad_norm": 0.14446289837360382,
      "learning_rate": 0.0002,
      "loss": 1.5047,
      "step": 1204
    },
    {
      "epoch": 0.4397810218978102,
      "grad_norm": 0.13527429103851318,
      "learning_rate": 0.0002,
      "loss": 1.5378,
      "step": 1205
    },
    {
      "epoch": 0.4401459854014598,
      "grad_norm": 0.1380535364151001,
      "learning_rate": 0.0002,
      "loss": 1.5218,
      "step": 1206
    },
    {
      "epoch": 0.4405109489051095,
      "grad_norm": 0.13606858253479004,
      "learning_rate": 0.0002,
      "loss": 1.5409,
      "step": 1207
    },
    {
      "epoch": 0.4408759124087591,
      "grad_norm": 0.13873758912086487,
      "learning_rate": 0.0002,
      "loss": 1.5138,
      "step": 1208
    },
    {
      "epoch": 0.44124087591240874,
      "grad_norm": 0.13520547747612,
      "learning_rate": 0.0002,
      "loss": 1.4934,
      "step": 1209
    },
    {
      "epoch": 0.4416058394160584,
      "grad_norm": 0.13886326551437378,
      "learning_rate": 0.0002,
      "loss": 1.5299,
      "step": 1210
    },
    {
      "epoch": 0.44197080291970803,
      "grad_norm": 0.14521658420562744,
      "learning_rate": 0.0002,
      "loss": 1.6335,
      "step": 1211
    },
    {
      "epoch": 0.44233576642335765,
      "grad_norm": 0.14181935787200928,
      "learning_rate": 0.0002,
      "loss": 1.5897,
      "step": 1212
    },
    {
      "epoch": 0.4427007299270073,
      "grad_norm": 0.14346447587013245,
      "learning_rate": 0.0002,
      "loss": 1.6002,
      "step": 1213
    },
    {
      "epoch": 0.44306569343065694,
      "grad_norm": 0.14468444883823395,
      "learning_rate": 0.0002,
      "loss": 1.6342,
      "step": 1214
    },
    {
      "epoch": 0.44343065693430656,
      "grad_norm": 0.14132562279701233,
      "learning_rate": 0.0002,
      "loss": 1.5177,
      "step": 1215
    },
    {
      "epoch": 0.44379562043795623,
      "grad_norm": 0.13532762229442596,
      "learning_rate": 0.0002,
      "loss": 1.5451,
      "step": 1216
    },
    {
      "epoch": 0.44416058394160585,
      "grad_norm": 0.13688690960407257,
      "learning_rate": 0.0002,
      "loss": 1.5064,
      "step": 1217
    },
    {
      "epoch": 0.44452554744525546,
      "grad_norm": 0.13426612317562103,
      "learning_rate": 0.0002,
      "loss": 1.5316,
      "step": 1218
    },
    {
      "epoch": 0.4448905109489051,
      "grad_norm": 0.14322076737880707,
      "learning_rate": 0.0002,
      "loss": 1.5905,
      "step": 1219
    },
    {
      "epoch": 0.44525547445255476,
      "grad_norm": 0.14106355607509613,
      "learning_rate": 0.0002,
      "loss": 1.5633,
      "step": 1220
    },
    {
      "epoch": 0.4456204379562044,
      "grad_norm": 0.1447700560092926,
      "learning_rate": 0.0002,
      "loss": 1.5667,
      "step": 1221
    },
    {
      "epoch": 0.445985401459854,
      "grad_norm": 0.13977955281734467,
      "learning_rate": 0.0002,
      "loss": 1.5143,
      "step": 1222
    },
    {
      "epoch": 0.44635036496350367,
      "grad_norm": 0.1485152691602707,
      "learning_rate": 0.0002,
      "loss": 1.5395,
      "step": 1223
    },
    {
      "epoch": 0.4467153284671533,
      "grad_norm": 0.13944393396377563,
      "learning_rate": 0.0002,
      "loss": 1.5283,
      "step": 1224
    },
    {
      "epoch": 0.4470802919708029,
      "grad_norm": 0.14249053597450256,
      "learning_rate": 0.0002,
      "loss": 1.5467,
      "step": 1225
    },
    {
      "epoch": 0.4474452554744526,
      "grad_norm": 0.13913418352603912,
      "learning_rate": 0.0002,
      "loss": 1.5416,
      "step": 1226
    },
    {
      "epoch": 0.4478102189781022,
      "grad_norm": 0.14426615834236145,
      "learning_rate": 0.0002,
      "loss": 1.4809,
      "step": 1227
    },
    {
      "epoch": 0.4481751824817518,
      "grad_norm": 0.140080988407135,
      "learning_rate": 0.0002,
      "loss": 1.5656,
      "step": 1228
    },
    {
      "epoch": 0.4485401459854015,
      "grad_norm": 0.1470908671617508,
      "learning_rate": 0.0002,
      "loss": 1.594,
      "step": 1229
    },
    {
      "epoch": 0.4489051094890511,
      "grad_norm": 0.14607945084571838,
      "learning_rate": 0.0002,
      "loss": 1.5174,
      "step": 1230
    },
    {
      "epoch": 0.4492700729927007,
      "grad_norm": 0.14084866642951965,
      "learning_rate": 0.0002,
      "loss": 1.5508,
      "step": 1231
    },
    {
      "epoch": 0.44963503649635034,
      "grad_norm": 0.14586275815963745,
      "learning_rate": 0.0002,
      "loss": 1.5417,
      "step": 1232
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.1417161077260971,
      "learning_rate": 0.0002,
      "loss": 1.5599,
      "step": 1233
    },
    {
      "epoch": 0.45036496350364963,
      "grad_norm": 0.14179855585098267,
      "learning_rate": 0.0002,
      "loss": 1.5237,
      "step": 1234
    },
    {
      "epoch": 0.45072992700729925,
      "grad_norm": 0.14542655646800995,
      "learning_rate": 0.0002,
      "loss": 1.5364,
      "step": 1235
    },
    {
      "epoch": 0.4510948905109489,
      "grad_norm": 0.14230026304721832,
      "learning_rate": 0.0002,
      "loss": 1.462,
      "step": 1236
    },
    {
      "epoch": 0.45145985401459854,
      "grad_norm": 0.14053215086460114,
      "learning_rate": 0.0002,
      "loss": 1.5771,
      "step": 1237
    },
    {
      "epoch": 0.45182481751824816,
      "grad_norm": 0.14945510029792786,
      "learning_rate": 0.0002,
      "loss": 1.4723,
      "step": 1238
    },
    {
      "epoch": 0.45218978102189783,
      "grad_norm": 0.14607320725917816,
      "learning_rate": 0.0002,
      "loss": 1.5538,
      "step": 1239
    },
    {
      "epoch": 0.45255474452554745,
      "grad_norm": 0.13515004515647888,
      "learning_rate": 0.0002,
      "loss": 1.5021,
      "step": 1240
    },
    {
      "epoch": 0.45291970802919707,
      "grad_norm": 0.1434289813041687,
      "learning_rate": 0.0002,
      "loss": 1.5331,
      "step": 1241
    },
    {
      "epoch": 0.45328467153284674,
      "grad_norm": 0.15340344607830048,
      "learning_rate": 0.0002,
      "loss": 1.5829,
      "step": 1242
    },
    {
      "epoch": 0.45364963503649636,
      "grad_norm": 0.13555632531642914,
      "learning_rate": 0.0002,
      "loss": 1.5475,
      "step": 1243
    },
    {
      "epoch": 0.454014598540146,
      "grad_norm": 0.1385466456413269,
      "learning_rate": 0.0002,
      "loss": 1.6074,
      "step": 1244
    },
    {
      "epoch": 0.4543795620437956,
      "grad_norm": 0.13884568214416504,
      "learning_rate": 0.0002,
      "loss": 1.5606,
      "step": 1245
    },
    {
      "epoch": 0.45474452554744527,
      "grad_norm": 0.13896356523036957,
      "learning_rate": 0.0002,
      "loss": 1.5109,
      "step": 1246
    },
    {
      "epoch": 0.4551094890510949,
      "grad_norm": 0.14803259074687958,
      "learning_rate": 0.0002,
      "loss": 1.5719,
      "step": 1247
    },
    {
      "epoch": 0.4554744525547445,
      "grad_norm": 0.143818199634552,
      "learning_rate": 0.0002,
      "loss": 1.5671,
      "step": 1248
    },
    {
      "epoch": 0.4558394160583942,
      "grad_norm": 0.1388106793165207,
      "learning_rate": 0.0002,
      "loss": 1.5462,
      "step": 1249
    },
    {
      "epoch": 0.4562043795620438,
      "grad_norm": 0.14293359220027924,
      "learning_rate": 0.0002,
      "loss": 1.5719,
      "step": 1250
    },
    {
      "epoch": 0.4565693430656934,
      "grad_norm": 0.1525656282901764,
      "learning_rate": 0.0002,
      "loss": 1.6219,
      "step": 1251
    },
    {
      "epoch": 0.4569343065693431,
      "grad_norm": 0.1448962241411209,
      "learning_rate": 0.0002,
      "loss": 1.5744,
      "step": 1252
    },
    {
      "epoch": 0.4572992700729927,
      "grad_norm": 0.1427375078201294,
      "learning_rate": 0.0002,
      "loss": 1.5052,
      "step": 1253
    },
    {
      "epoch": 0.4576642335766423,
      "grad_norm": 0.1440911442041397,
      "learning_rate": 0.0002,
      "loss": 1.5297,
      "step": 1254
    },
    {
      "epoch": 0.458029197080292,
      "grad_norm": 0.142386794090271,
      "learning_rate": 0.0002,
      "loss": 1.5701,
      "step": 1255
    },
    {
      "epoch": 0.4583941605839416,
      "grad_norm": 0.1558191031217575,
      "learning_rate": 0.0002,
      "loss": 1.5588,
      "step": 1256
    },
    {
      "epoch": 0.45875912408759123,
      "grad_norm": 0.1407562494277954,
      "learning_rate": 0.0002,
      "loss": 1.4956,
      "step": 1257
    },
    {
      "epoch": 0.4591240875912409,
      "grad_norm": 0.1365852653980255,
      "learning_rate": 0.0002,
      "loss": 1.536,
      "step": 1258
    },
    {
      "epoch": 0.4594890510948905,
      "grad_norm": 0.13735376298427582,
      "learning_rate": 0.0002,
      "loss": 1.5207,
      "step": 1259
    },
    {
      "epoch": 0.45985401459854014,
      "grad_norm": 0.14341118931770325,
      "learning_rate": 0.0002,
      "loss": 1.5451,
      "step": 1260
    },
    {
      "epoch": 0.46021897810218976,
      "grad_norm": 0.13702426850795746,
      "learning_rate": 0.0002,
      "loss": 1.5203,
      "step": 1261
    },
    {
      "epoch": 0.46058394160583943,
      "grad_norm": 0.13522765040397644,
      "learning_rate": 0.0002,
      "loss": 1.4739,
      "step": 1262
    },
    {
      "epoch": 0.46094890510948905,
      "grad_norm": 0.13788940012454987,
      "learning_rate": 0.0002,
      "loss": 1.4084,
      "step": 1263
    },
    {
      "epoch": 0.46131386861313867,
      "grad_norm": 0.1516128033399582,
      "learning_rate": 0.0002,
      "loss": 1.5329,
      "step": 1264
    },
    {
      "epoch": 0.46167883211678834,
      "grad_norm": 0.1387404352426529,
      "learning_rate": 0.0002,
      "loss": 1.5695,
      "step": 1265
    },
    {
      "epoch": 0.46204379562043796,
      "grad_norm": 0.14633028209209442,
      "learning_rate": 0.0002,
      "loss": 1.5748,
      "step": 1266
    },
    {
      "epoch": 0.4624087591240876,
      "grad_norm": 0.13612833619117737,
      "learning_rate": 0.0002,
      "loss": 1.5516,
      "step": 1267
    },
    {
      "epoch": 0.46277372262773725,
      "grad_norm": 0.13944987952709198,
      "learning_rate": 0.0002,
      "loss": 1.5356,
      "step": 1268
    },
    {
      "epoch": 0.46313868613138687,
      "grad_norm": 0.13883937895298004,
      "learning_rate": 0.0002,
      "loss": 1.5261,
      "step": 1269
    },
    {
      "epoch": 0.4635036496350365,
      "grad_norm": 0.15365077555179596,
      "learning_rate": 0.0002,
      "loss": 1.5348,
      "step": 1270
    },
    {
      "epoch": 0.46386861313868616,
      "grad_norm": 0.13766808807849884,
      "learning_rate": 0.0002,
      "loss": 1.4708,
      "step": 1271
    },
    {
      "epoch": 0.4642335766423358,
      "grad_norm": 0.14412592351436615,
      "learning_rate": 0.0002,
      "loss": 1.5555,
      "step": 1272
    },
    {
      "epoch": 0.4645985401459854,
      "grad_norm": 0.1350487768650055,
      "learning_rate": 0.0002,
      "loss": 1.4872,
      "step": 1273
    },
    {
      "epoch": 0.464963503649635,
      "grad_norm": 0.14102935791015625,
      "learning_rate": 0.0002,
      "loss": 1.6023,
      "step": 1274
    },
    {
      "epoch": 0.4653284671532847,
      "grad_norm": 0.13394318521022797,
      "learning_rate": 0.0002,
      "loss": 1.4643,
      "step": 1275
    },
    {
      "epoch": 0.4656934306569343,
      "grad_norm": 0.13883759081363678,
      "learning_rate": 0.0002,
      "loss": 1.4752,
      "step": 1276
    },
    {
      "epoch": 0.4660583941605839,
      "grad_norm": 0.1420651376247406,
      "learning_rate": 0.0002,
      "loss": 1.5624,
      "step": 1277
    },
    {
      "epoch": 0.4664233576642336,
      "grad_norm": 0.13440601527690887,
      "learning_rate": 0.0002,
      "loss": 1.5644,
      "step": 1278
    },
    {
      "epoch": 0.4667883211678832,
      "grad_norm": 0.14965100586414337,
      "learning_rate": 0.0002,
      "loss": 1.556,
      "step": 1279
    },
    {
      "epoch": 0.46715328467153283,
      "grad_norm": 0.13507254421710968,
      "learning_rate": 0.0002,
      "loss": 1.5268,
      "step": 1280
    },
    {
      "epoch": 0.4675182481751825,
      "grad_norm": 0.14349453151226044,
      "learning_rate": 0.0002,
      "loss": 1.5033,
      "step": 1281
    },
    {
      "epoch": 0.4678832116788321,
      "grad_norm": 0.14425434172153473,
      "learning_rate": 0.0002,
      "loss": 1.5625,
      "step": 1282
    },
    {
      "epoch": 0.46824817518248174,
      "grad_norm": 0.1459636092185974,
      "learning_rate": 0.0002,
      "loss": 1.5493,
      "step": 1283
    },
    {
      "epoch": 0.4686131386861314,
      "grad_norm": 0.14194689691066742,
      "learning_rate": 0.0002,
      "loss": 1.534,
      "step": 1284
    },
    {
      "epoch": 0.46897810218978103,
      "grad_norm": 0.14540015161037445,
      "learning_rate": 0.0002,
      "loss": 1.5131,
      "step": 1285
    },
    {
      "epoch": 0.46934306569343065,
      "grad_norm": 0.14391468465328217,
      "learning_rate": 0.0002,
      "loss": 1.5264,
      "step": 1286
    },
    {
      "epoch": 0.46970802919708027,
      "grad_norm": 0.1434810757637024,
      "learning_rate": 0.0002,
      "loss": 1.5459,
      "step": 1287
    },
    {
      "epoch": 0.47007299270072994,
      "grad_norm": 0.13951091468334198,
      "learning_rate": 0.0002,
      "loss": 1.5541,
      "step": 1288
    },
    {
      "epoch": 0.47043795620437956,
      "grad_norm": 0.13550563156604767,
      "learning_rate": 0.0002,
      "loss": 1.4701,
      "step": 1289
    },
    {
      "epoch": 0.4708029197080292,
      "grad_norm": 0.1541942059993744,
      "learning_rate": 0.0002,
      "loss": 1.6092,
      "step": 1290
    },
    {
      "epoch": 0.47116788321167885,
      "grad_norm": 0.14570224285125732,
      "learning_rate": 0.0002,
      "loss": 1.6338,
      "step": 1291
    },
    {
      "epoch": 0.47153284671532847,
      "grad_norm": 0.15525241196155548,
      "learning_rate": 0.0002,
      "loss": 1.5705,
      "step": 1292
    },
    {
      "epoch": 0.4718978102189781,
      "grad_norm": 0.13773047924041748,
      "learning_rate": 0.0002,
      "loss": 1.5503,
      "step": 1293
    },
    {
      "epoch": 0.47226277372262776,
      "grad_norm": 0.14515553414821625,
      "learning_rate": 0.0002,
      "loss": 1.5401,
      "step": 1294
    },
    {
      "epoch": 0.4726277372262774,
      "grad_norm": 0.13599850237369537,
      "learning_rate": 0.0002,
      "loss": 1.5293,
      "step": 1295
    },
    {
      "epoch": 0.472992700729927,
      "grad_norm": 0.1547880321741104,
      "learning_rate": 0.0002,
      "loss": 1.5077,
      "step": 1296
    },
    {
      "epoch": 0.47335766423357667,
      "grad_norm": 0.1425085812807083,
      "learning_rate": 0.0002,
      "loss": 1.5926,
      "step": 1297
    },
    {
      "epoch": 0.4737226277372263,
      "grad_norm": 0.1425679624080658,
      "learning_rate": 0.0002,
      "loss": 1.6066,
      "step": 1298
    },
    {
      "epoch": 0.4740875912408759,
      "grad_norm": 0.14013376832008362,
      "learning_rate": 0.0002,
      "loss": 1.4994,
      "step": 1299
    },
    {
      "epoch": 0.4744525547445255,
      "grad_norm": 0.13755522668361664,
      "learning_rate": 0.0002,
      "loss": 1.4942,
      "step": 1300
    },
    {
      "epoch": 0.4748175182481752,
      "grad_norm": 0.1457519382238388,
      "learning_rate": 0.0002,
      "loss": 1.5561,
      "step": 1301
    },
    {
      "epoch": 0.4751824817518248,
      "grad_norm": 0.1449151337146759,
      "learning_rate": 0.0002,
      "loss": 1.5395,
      "step": 1302
    },
    {
      "epoch": 0.47554744525547443,
      "grad_norm": 0.14026185870170593,
      "learning_rate": 0.0002,
      "loss": 1.525,
      "step": 1303
    },
    {
      "epoch": 0.4759124087591241,
      "grad_norm": 0.14454710483551025,
      "learning_rate": 0.0002,
      "loss": 1.5468,
      "step": 1304
    },
    {
      "epoch": 0.4762773722627737,
      "grad_norm": 0.1408848762512207,
      "learning_rate": 0.0002,
      "loss": 1.5737,
      "step": 1305
    },
    {
      "epoch": 0.47664233576642334,
      "grad_norm": 0.14418385922908783,
      "learning_rate": 0.0002,
      "loss": 1.5488,
      "step": 1306
    },
    {
      "epoch": 0.477007299270073,
      "grad_norm": 0.1389208734035492,
      "learning_rate": 0.0002,
      "loss": 1.6053,
      "step": 1307
    },
    {
      "epoch": 0.47737226277372263,
      "grad_norm": 0.1470184326171875,
      "learning_rate": 0.0002,
      "loss": 1.5841,
      "step": 1308
    },
    {
      "epoch": 0.47773722627737225,
      "grad_norm": 0.14073319733142853,
      "learning_rate": 0.0002,
      "loss": 1.499,
      "step": 1309
    },
    {
      "epoch": 0.4781021897810219,
      "grad_norm": 0.14321047067642212,
      "learning_rate": 0.0002,
      "loss": 1.6263,
      "step": 1310
    },
    {
      "epoch": 0.47846715328467154,
      "grad_norm": 0.14019696414470673,
      "learning_rate": 0.0002,
      "loss": 1.548,
      "step": 1311
    },
    {
      "epoch": 0.47883211678832116,
      "grad_norm": 0.14383907616138458,
      "learning_rate": 0.0002,
      "loss": 1.6388,
      "step": 1312
    },
    {
      "epoch": 0.4791970802919708,
      "grad_norm": 0.13995461165905,
      "learning_rate": 0.0002,
      "loss": 1.5373,
      "step": 1313
    },
    {
      "epoch": 0.47956204379562045,
      "grad_norm": 0.1428891122341156,
      "learning_rate": 0.0002,
      "loss": 1.6036,
      "step": 1314
    },
    {
      "epoch": 0.47992700729927007,
      "grad_norm": 0.14514166116714478,
      "learning_rate": 0.0002,
      "loss": 1.5439,
      "step": 1315
    },
    {
      "epoch": 0.4802919708029197,
      "grad_norm": 0.14264371991157532,
      "learning_rate": 0.0002,
      "loss": 1.5115,
      "step": 1316
    },
    {
      "epoch": 0.48065693430656936,
      "grad_norm": 0.1450612097978592,
      "learning_rate": 0.0002,
      "loss": 1.6245,
      "step": 1317
    },
    {
      "epoch": 0.481021897810219,
      "grad_norm": 0.13961923122406006,
      "learning_rate": 0.0002,
      "loss": 1.5367,
      "step": 1318
    },
    {
      "epoch": 0.4813868613138686,
      "grad_norm": 0.1448686569929123,
      "learning_rate": 0.0002,
      "loss": 1.507,
      "step": 1319
    },
    {
      "epoch": 0.48175182481751827,
      "grad_norm": 0.1462506353855133,
      "learning_rate": 0.0002,
      "loss": 1.5197,
      "step": 1320
    },
    {
      "epoch": 0.4821167883211679,
      "grad_norm": 0.14809611439704895,
      "learning_rate": 0.0002,
      "loss": 1.5445,
      "step": 1321
    },
    {
      "epoch": 0.4824817518248175,
      "grad_norm": 0.14799737930297852,
      "learning_rate": 0.0002,
      "loss": 1.578,
      "step": 1322
    },
    {
      "epoch": 0.4828467153284672,
      "grad_norm": 0.15556024014949799,
      "learning_rate": 0.0002,
      "loss": 1.6054,
      "step": 1323
    },
    {
      "epoch": 0.4832116788321168,
      "grad_norm": 0.14923416078090668,
      "learning_rate": 0.0002,
      "loss": 1.6042,
      "step": 1324
    },
    {
      "epoch": 0.4835766423357664,
      "grad_norm": 0.14249694347381592,
      "learning_rate": 0.0002,
      "loss": 1.528,
      "step": 1325
    },
    {
      "epoch": 0.48394160583941603,
      "grad_norm": 0.14372067153453827,
      "learning_rate": 0.0002,
      "loss": 1.5394,
      "step": 1326
    },
    {
      "epoch": 0.4843065693430657,
      "grad_norm": 0.15322162210941315,
      "learning_rate": 0.0002,
      "loss": 1.6328,
      "step": 1327
    },
    {
      "epoch": 0.4846715328467153,
      "grad_norm": 0.1400681436061859,
      "learning_rate": 0.0002,
      "loss": 1.5504,
      "step": 1328
    },
    {
      "epoch": 0.48503649635036494,
      "grad_norm": 0.14974567294120789,
      "learning_rate": 0.0002,
      "loss": 1.578,
      "step": 1329
    },
    {
      "epoch": 0.4854014598540146,
      "grad_norm": 0.14469189941883087,
      "learning_rate": 0.0002,
      "loss": 1.5294,
      "step": 1330
    },
    {
      "epoch": 0.48576642335766423,
      "grad_norm": 0.1519039273262024,
      "learning_rate": 0.0002,
      "loss": 1.5372,
      "step": 1331
    },
    {
      "epoch": 0.48613138686131385,
      "grad_norm": 0.141452357172966,
      "learning_rate": 0.0002,
      "loss": 1.5784,
      "step": 1332
    },
    {
      "epoch": 0.4864963503649635,
      "grad_norm": 0.16022619605064392,
      "learning_rate": 0.0002,
      "loss": 1.577,
      "step": 1333
    },
    {
      "epoch": 0.48686131386861314,
      "grad_norm": 0.14563214778900146,
      "learning_rate": 0.0002,
      "loss": 1.5742,
      "step": 1334
    },
    {
      "epoch": 0.48722627737226276,
      "grad_norm": 0.15028531849384308,
      "learning_rate": 0.0002,
      "loss": 1.556,
      "step": 1335
    },
    {
      "epoch": 0.48759124087591244,
      "grad_norm": 0.14298510551452637,
      "learning_rate": 0.0002,
      "loss": 1.5164,
      "step": 1336
    },
    {
      "epoch": 0.48795620437956205,
      "grad_norm": 0.14748181402683258,
      "learning_rate": 0.0002,
      "loss": 1.4475,
      "step": 1337
    },
    {
      "epoch": 0.48832116788321167,
      "grad_norm": 0.1428547203540802,
      "learning_rate": 0.0002,
      "loss": 1.5902,
      "step": 1338
    },
    {
      "epoch": 0.4886861313868613,
      "grad_norm": 0.14427773654460907,
      "learning_rate": 0.0002,
      "loss": 1.5665,
      "step": 1339
    },
    {
      "epoch": 0.48905109489051096,
      "grad_norm": 0.14466895163059235,
      "learning_rate": 0.0002,
      "loss": 1.5287,
      "step": 1340
    },
    {
      "epoch": 0.4894160583941606,
      "grad_norm": 0.14506585896015167,
      "learning_rate": 0.0002,
      "loss": 1.5617,
      "step": 1341
    },
    {
      "epoch": 0.4897810218978102,
      "grad_norm": 0.14108645915985107,
      "learning_rate": 0.0002,
      "loss": 1.5696,
      "step": 1342
    },
    {
      "epoch": 0.49014598540145987,
      "grad_norm": 0.14050480723381042,
      "learning_rate": 0.0002,
      "loss": 1.5051,
      "step": 1343
    },
    {
      "epoch": 0.4905109489051095,
      "grad_norm": 0.1499432921409607,
      "learning_rate": 0.0002,
      "loss": 1.554,
      "step": 1344
    },
    {
      "epoch": 0.4908759124087591,
      "grad_norm": 0.1439749151468277,
      "learning_rate": 0.0002,
      "loss": 1.4758,
      "step": 1345
    },
    {
      "epoch": 0.4912408759124088,
      "grad_norm": 0.142302468419075,
      "learning_rate": 0.0002,
      "loss": 1.4782,
      "step": 1346
    },
    {
      "epoch": 0.4916058394160584,
      "grad_norm": 0.14722028374671936,
      "learning_rate": 0.0002,
      "loss": 1.5883,
      "step": 1347
    },
    {
      "epoch": 0.491970802919708,
      "grad_norm": 0.16787558794021606,
      "learning_rate": 0.0002,
      "loss": 1.6216,
      "step": 1348
    },
    {
      "epoch": 0.4923357664233577,
      "grad_norm": 0.13706082105636597,
      "learning_rate": 0.0002,
      "loss": 1.5333,
      "step": 1349
    },
    {
      "epoch": 0.4927007299270073,
      "grad_norm": 0.157490074634552,
      "learning_rate": 0.0002,
      "loss": 1.5951,
      "step": 1350
    },
    {
      "epoch": 0.4930656934306569,
      "grad_norm": 0.14200635254383087,
      "learning_rate": 0.0002,
      "loss": 1.5874,
      "step": 1351
    },
    {
      "epoch": 0.49343065693430654,
      "grad_norm": 0.14692270755767822,
      "learning_rate": 0.0002,
      "loss": 1.5259,
      "step": 1352
    },
    {
      "epoch": 0.4937956204379562,
      "grad_norm": 0.14393900334835052,
      "learning_rate": 0.0002,
      "loss": 1.5531,
      "step": 1353
    },
    {
      "epoch": 0.49416058394160584,
      "grad_norm": 0.15199580788612366,
      "learning_rate": 0.0002,
      "loss": 1.5025,
      "step": 1354
    },
    {
      "epoch": 0.49452554744525545,
      "grad_norm": 0.1430264562368393,
      "learning_rate": 0.0002,
      "loss": 1.5708,
      "step": 1355
    },
    {
      "epoch": 0.4948905109489051,
      "grad_norm": 0.15903516113758087,
      "learning_rate": 0.0002,
      "loss": 1.5768,
      "step": 1356
    },
    {
      "epoch": 0.49525547445255474,
      "grad_norm": 0.1371280997991562,
      "learning_rate": 0.0002,
      "loss": 1.5123,
      "step": 1357
    },
    {
      "epoch": 0.49562043795620436,
      "grad_norm": 0.155189648270607,
      "learning_rate": 0.0002,
      "loss": 1.5283,
      "step": 1358
    },
    {
      "epoch": 0.49598540145985404,
      "grad_norm": 0.14270073175430298,
      "learning_rate": 0.0002,
      "loss": 1.4728,
      "step": 1359
    },
    {
      "epoch": 0.49635036496350365,
      "grad_norm": 0.14564581215381622,
      "learning_rate": 0.0002,
      "loss": 1.5111,
      "step": 1360
    },
    {
      "epoch": 0.49671532846715327,
      "grad_norm": 0.14370794594287872,
      "learning_rate": 0.0002,
      "loss": 1.57,
      "step": 1361
    },
    {
      "epoch": 0.49708029197080295,
      "grad_norm": 0.14522604644298553,
      "learning_rate": 0.0002,
      "loss": 1.5526,
      "step": 1362
    },
    {
      "epoch": 0.49744525547445256,
      "grad_norm": 0.14272421598434448,
      "learning_rate": 0.0002,
      "loss": 1.5223,
      "step": 1363
    },
    {
      "epoch": 0.4978102189781022,
      "grad_norm": 0.14297039806842804,
      "learning_rate": 0.0002,
      "loss": 1.5425,
      "step": 1364
    },
    {
      "epoch": 0.4981751824817518,
      "grad_norm": 0.14181937277317047,
      "learning_rate": 0.0002,
      "loss": 1.544,
      "step": 1365
    },
    {
      "epoch": 0.4985401459854015,
      "grad_norm": 0.14614765346050262,
      "learning_rate": 0.0002,
      "loss": 1.6157,
      "step": 1366
    },
    {
      "epoch": 0.4989051094890511,
      "grad_norm": 0.14182087779045105,
      "learning_rate": 0.0002,
      "loss": 1.5882,
      "step": 1367
    },
    {
      "epoch": 0.4992700729927007,
      "grad_norm": 0.14247983694076538,
      "learning_rate": 0.0002,
      "loss": 1.5521,
      "step": 1368
    },
    {
      "epoch": 0.4996350364963504,
      "grad_norm": 0.14417298138141632,
      "learning_rate": 0.0002,
      "loss": 1.5735,
      "step": 1369
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1436476707458496,
      "learning_rate": 0.0002,
      "loss": 1.5932,
      "step": 1370
    },
    {
      "epoch": 0.5003649635036497,
      "grad_norm": 0.14210639894008636,
      "learning_rate": 0.0002,
      "loss": 1.5349,
      "step": 1371
    },
    {
      "epoch": 0.5007299270072992,
      "grad_norm": 0.1424606591463089,
      "learning_rate": 0.0002,
      "loss": 1.5371,
      "step": 1372
    },
    {
      "epoch": 0.5010948905109489,
      "grad_norm": 0.14976432919502258,
      "learning_rate": 0.0002,
      "loss": 1.5438,
      "step": 1373
    },
    {
      "epoch": 0.5014598540145986,
      "grad_norm": 0.14661212265491486,
      "learning_rate": 0.0002,
      "loss": 1.5074,
      "step": 1374
    },
    {
      "epoch": 0.5018248175182481,
      "grad_norm": 0.1430431455373764,
      "learning_rate": 0.0002,
      "loss": 1.5293,
      "step": 1375
    },
    {
      "epoch": 0.5021897810218978,
      "grad_norm": 0.1461697220802307,
      "learning_rate": 0.0002,
      "loss": 1.5249,
      "step": 1376
    },
    {
      "epoch": 0.5025547445255475,
      "grad_norm": 0.14215901494026184,
      "learning_rate": 0.0002,
      "loss": 1.5945,
      "step": 1377
    },
    {
      "epoch": 0.502919708029197,
      "grad_norm": 0.1483655571937561,
      "learning_rate": 0.0002,
      "loss": 1.5554,
      "step": 1378
    },
    {
      "epoch": 0.5032846715328467,
      "grad_norm": 0.1459675282239914,
      "learning_rate": 0.0002,
      "loss": 1.5734,
      "step": 1379
    },
    {
      "epoch": 0.5036496350364964,
      "grad_norm": 0.14046071469783783,
      "learning_rate": 0.0002,
      "loss": 1.5233,
      "step": 1380
    },
    {
      "epoch": 0.504014598540146,
      "grad_norm": 0.151759535074234,
      "learning_rate": 0.0002,
      "loss": 1.5343,
      "step": 1381
    },
    {
      "epoch": 0.5043795620437956,
      "grad_norm": 0.14420148730278015,
      "learning_rate": 0.0002,
      "loss": 1.5219,
      "step": 1382
    },
    {
      "epoch": 0.5047445255474453,
      "grad_norm": 0.1466650515794754,
      "learning_rate": 0.0002,
      "loss": 1.6138,
      "step": 1383
    },
    {
      "epoch": 0.5051094890510949,
      "grad_norm": 0.1441303938627243,
      "learning_rate": 0.0002,
      "loss": 1.5121,
      "step": 1384
    },
    {
      "epoch": 0.5054744525547445,
      "grad_norm": 0.14182178676128387,
      "learning_rate": 0.0002,
      "loss": 1.5143,
      "step": 1385
    },
    {
      "epoch": 0.5058394160583941,
      "grad_norm": 0.14212572574615479,
      "learning_rate": 0.0002,
      "loss": 1.5819,
      "step": 1386
    },
    {
      "epoch": 0.5062043795620438,
      "grad_norm": 0.141070157289505,
      "learning_rate": 0.0002,
      "loss": 1.58,
      "step": 1387
    },
    {
      "epoch": 0.5065693430656935,
      "grad_norm": 0.14122633635997772,
      "learning_rate": 0.0002,
      "loss": 1.5519,
      "step": 1388
    },
    {
      "epoch": 0.506934306569343,
      "grad_norm": 0.14164310693740845,
      "learning_rate": 0.0002,
      "loss": 1.4511,
      "step": 1389
    },
    {
      "epoch": 0.5072992700729927,
      "grad_norm": 0.1417575478553772,
      "learning_rate": 0.0002,
      "loss": 1.554,
      "step": 1390
    },
    {
      "epoch": 0.5076642335766424,
      "grad_norm": 0.14424169063568115,
      "learning_rate": 0.0002,
      "loss": 1.5386,
      "step": 1391
    },
    {
      "epoch": 0.5080291970802919,
      "grad_norm": 0.1428794264793396,
      "learning_rate": 0.0002,
      "loss": 1.4933,
      "step": 1392
    },
    {
      "epoch": 0.5083941605839416,
      "grad_norm": 0.14521543681621552,
      "learning_rate": 0.0002,
      "loss": 1.5775,
      "step": 1393
    },
    {
      "epoch": 0.5087591240875913,
      "grad_norm": 0.14567610621452332,
      "learning_rate": 0.0002,
      "loss": 1.5683,
      "step": 1394
    },
    {
      "epoch": 0.5091240875912408,
      "grad_norm": 0.14322175085544586,
      "learning_rate": 0.0002,
      "loss": 1.513,
      "step": 1395
    },
    {
      "epoch": 0.5094890510948905,
      "grad_norm": 0.1477554440498352,
      "learning_rate": 0.0002,
      "loss": 1.5822,
      "step": 1396
    },
    {
      "epoch": 0.5098540145985402,
      "grad_norm": 0.14574214816093445,
      "learning_rate": 0.0002,
      "loss": 1.4716,
      "step": 1397
    },
    {
      "epoch": 0.5102189781021897,
      "grad_norm": 0.1411321610212326,
      "learning_rate": 0.0002,
      "loss": 1.4872,
      "step": 1398
    },
    {
      "epoch": 0.5105839416058394,
      "grad_norm": 0.14788571000099182,
      "learning_rate": 0.0002,
      "loss": 1.568,
      "step": 1399
    },
    {
      "epoch": 0.5109489051094891,
      "grad_norm": 0.1436443030834198,
      "learning_rate": 0.0002,
      "loss": 1.5713,
      "step": 1400
    },
    {
      "epoch": 0.5113138686131387,
      "grad_norm": 0.14331839978694916,
      "learning_rate": 0.0002,
      "loss": 1.5362,
      "step": 1401
    },
    {
      "epoch": 0.5116788321167883,
      "grad_norm": 0.14069515466690063,
      "learning_rate": 0.0002,
      "loss": 1.4805,
      "step": 1402
    },
    {
      "epoch": 0.512043795620438,
      "grad_norm": 0.14688001573085785,
      "learning_rate": 0.0002,
      "loss": 1.5384,
      "step": 1403
    },
    {
      "epoch": 0.5124087591240876,
      "grad_norm": 0.14640820026397705,
      "learning_rate": 0.0002,
      "loss": 1.6104,
      "step": 1404
    },
    {
      "epoch": 0.5127737226277372,
      "grad_norm": 0.15085360407829285,
      "learning_rate": 0.0002,
      "loss": 1.571,
      "step": 1405
    },
    {
      "epoch": 0.5131386861313869,
      "grad_norm": 0.13795189559459686,
      "learning_rate": 0.0002,
      "loss": 1.5036,
      "step": 1406
    },
    {
      "epoch": 0.5135036496350365,
      "grad_norm": 0.14130617678165436,
      "learning_rate": 0.0002,
      "loss": 1.5847,
      "step": 1407
    },
    {
      "epoch": 0.5138686131386861,
      "grad_norm": 0.14934630692005157,
      "learning_rate": 0.0002,
      "loss": 1.5346,
      "step": 1408
    },
    {
      "epoch": 0.5142335766423358,
      "grad_norm": 0.1543421596288681,
      "learning_rate": 0.0002,
      "loss": 1.4965,
      "step": 1409
    },
    {
      "epoch": 0.5145985401459854,
      "grad_norm": 0.14345374703407288,
      "learning_rate": 0.0002,
      "loss": 1.4808,
      "step": 1410
    },
    {
      "epoch": 0.5149635036496351,
      "grad_norm": 0.14470990002155304,
      "learning_rate": 0.0002,
      "loss": 1.4742,
      "step": 1411
    },
    {
      "epoch": 0.5153284671532846,
      "grad_norm": 0.15001381933689117,
      "learning_rate": 0.0002,
      "loss": 1.5631,
      "step": 1412
    },
    {
      "epoch": 0.5156934306569343,
      "grad_norm": 0.1413411647081375,
      "learning_rate": 0.0002,
      "loss": 1.5135,
      "step": 1413
    },
    {
      "epoch": 0.516058394160584,
      "grad_norm": 0.15023396909236908,
      "learning_rate": 0.0002,
      "loss": 1.4963,
      "step": 1414
    },
    {
      "epoch": 0.5164233576642335,
      "grad_norm": 0.14387841522693634,
      "learning_rate": 0.0002,
      "loss": 1.5714,
      "step": 1415
    },
    {
      "epoch": 0.5167883211678832,
      "grad_norm": 0.14165225625038147,
      "learning_rate": 0.0002,
      "loss": 1.5834,
      "step": 1416
    },
    {
      "epoch": 0.5171532846715329,
      "grad_norm": 0.14866867661476135,
      "learning_rate": 0.0002,
      "loss": 1.5635,
      "step": 1417
    },
    {
      "epoch": 0.5175182481751824,
      "grad_norm": 0.1406414657831192,
      "learning_rate": 0.0002,
      "loss": 1.4692,
      "step": 1418
    },
    {
      "epoch": 0.5178832116788321,
      "grad_norm": 0.14628347754478455,
      "learning_rate": 0.0002,
      "loss": 1.4413,
      "step": 1419
    },
    {
      "epoch": 0.5182481751824818,
      "grad_norm": 0.1481773853302002,
      "learning_rate": 0.0002,
      "loss": 1.5441,
      "step": 1420
    },
    {
      "epoch": 0.5186131386861313,
      "grad_norm": 0.14823172986507416,
      "learning_rate": 0.0002,
      "loss": 1.5782,
      "step": 1421
    },
    {
      "epoch": 0.518978102189781,
      "grad_norm": 0.15497855842113495,
      "learning_rate": 0.0002,
      "loss": 1.5844,
      "step": 1422
    },
    {
      "epoch": 0.5193430656934307,
      "grad_norm": 0.13245084881782532,
      "learning_rate": 0.0002,
      "loss": 1.4142,
      "step": 1423
    },
    {
      "epoch": 0.5197080291970803,
      "grad_norm": 0.15496160089969635,
      "learning_rate": 0.0002,
      "loss": 1.5402,
      "step": 1424
    },
    {
      "epoch": 0.5200729927007299,
      "grad_norm": 0.15435543656349182,
      "learning_rate": 0.0002,
      "loss": 1.5944,
      "step": 1425
    },
    {
      "epoch": 0.5204379562043796,
      "grad_norm": 0.14302101731300354,
      "learning_rate": 0.0002,
      "loss": 1.5374,
      "step": 1426
    },
    {
      "epoch": 0.5208029197080292,
      "grad_norm": 0.15192078053951263,
      "learning_rate": 0.0002,
      "loss": 1.5054,
      "step": 1427
    },
    {
      "epoch": 0.5211678832116788,
      "grad_norm": 0.14225758612155914,
      "learning_rate": 0.0002,
      "loss": 1.5175,
      "step": 1428
    },
    {
      "epoch": 0.5215328467153285,
      "grad_norm": 0.15240885317325592,
      "learning_rate": 0.0002,
      "loss": 1.6224,
      "step": 1429
    },
    {
      "epoch": 0.5218978102189781,
      "grad_norm": 0.14327700436115265,
      "learning_rate": 0.0002,
      "loss": 1.5603,
      "step": 1430
    },
    {
      "epoch": 0.5222627737226277,
      "grad_norm": 0.15717336535453796,
      "learning_rate": 0.0002,
      "loss": 1.5374,
      "step": 1431
    },
    {
      "epoch": 0.5226277372262774,
      "grad_norm": 0.14663656055927277,
      "learning_rate": 0.0002,
      "loss": 1.6338,
      "step": 1432
    },
    {
      "epoch": 0.522992700729927,
      "grad_norm": 0.14126400649547577,
      "learning_rate": 0.0002,
      "loss": 1.5466,
      "step": 1433
    },
    {
      "epoch": 0.5233576642335767,
      "grad_norm": 0.14555731415748596,
      "learning_rate": 0.0002,
      "loss": 1.4896,
      "step": 1434
    },
    {
      "epoch": 0.5237226277372263,
      "grad_norm": 0.14315859973430634,
      "learning_rate": 0.0002,
      "loss": 1.5335,
      "step": 1435
    },
    {
      "epoch": 0.5240875912408759,
      "grad_norm": 0.14671078324317932,
      "learning_rate": 0.0002,
      "loss": 1.61,
      "step": 1436
    },
    {
      "epoch": 0.5244525547445256,
      "grad_norm": 0.14550861716270447,
      "learning_rate": 0.0002,
      "loss": 1.5263,
      "step": 1437
    },
    {
      "epoch": 0.5248175182481751,
      "grad_norm": 0.15268807113170624,
      "learning_rate": 0.0002,
      "loss": 1.5755,
      "step": 1438
    },
    {
      "epoch": 0.5251824817518248,
      "grad_norm": 0.13968442380428314,
      "learning_rate": 0.0002,
      "loss": 1.4323,
      "step": 1439
    },
    {
      "epoch": 0.5255474452554745,
      "grad_norm": 0.1430366337299347,
      "learning_rate": 0.0002,
      "loss": 1.554,
      "step": 1440
    },
    {
      "epoch": 0.525912408759124,
      "grad_norm": 0.14740513265132904,
      "learning_rate": 0.0002,
      "loss": 1.6276,
      "step": 1441
    },
    {
      "epoch": 0.5262773722627737,
      "grad_norm": 0.1434871107339859,
      "learning_rate": 0.0002,
      "loss": 1.5689,
      "step": 1442
    },
    {
      "epoch": 0.5266423357664234,
      "grad_norm": 0.14541637897491455,
      "learning_rate": 0.0002,
      "loss": 1.5429,
      "step": 1443
    },
    {
      "epoch": 0.527007299270073,
      "grad_norm": 0.14462392032146454,
      "learning_rate": 0.0002,
      "loss": 1.5545,
      "step": 1444
    },
    {
      "epoch": 0.5273722627737226,
      "grad_norm": 0.1523827165365219,
      "learning_rate": 0.0002,
      "loss": 1.5924,
      "step": 1445
    },
    {
      "epoch": 0.5277372262773723,
      "grad_norm": 0.1449674665927887,
      "learning_rate": 0.0002,
      "loss": 1.5082,
      "step": 1446
    },
    {
      "epoch": 0.5281021897810219,
      "grad_norm": 0.14989814162254333,
      "learning_rate": 0.0002,
      "loss": 1.5334,
      "step": 1447
    },
    {
      "epoch": 0.5284671532846715,
      "grad_norm": 0.1402965486049652,
      "learning_rate": 0.0002,
      "loss": 1.5038,
      "step": 1448
    },
    {
      "epoch": 0.5288321167883212,
      "grad_norm": 0.14613866806030273,
      "learning_rate": 0.0002,
      "loss": 1.4967,
      "step": 1449
    },
    {
      "epoch": 0.5291970802919708,
      "grad_norm": 0.1483517289161682,
      "learning_rate": 0.0002,
      "loss": 1.5282,
      "step": 1450
    },
    {
      "epoch": 0.5295620437956204,
      "grad_norm": 0.14213435351848602,
      "learning_rate": 0.0002,
      "loss": 1.5001,
      "step": 1451
    },
    {
      "epoch": 0.5299270072992701,
      "grad_norm": 0.14853952825069427,
      "learning_rate": 0.0002,
      "loss": 1.6056,
      "step": 1452
    },
    {
      "epoch": 0.5302919708029197,
      "grad_norm": 0.15149171650409698,
      "learning_rate": 0.0002,
      "loss": 1.6423,
      "step": 1453
    },
    {
      "epoch": 0.5306569343065694,
      "grad_norm": 0.1368187963962555,
      "learning_rate": 0.0002,
      "loss": 1.4876,
      "step": 1454
    },
    {
      "epoch": 0.531021897810219,
      "grad_norm": 0.14291512966156006,
      "learning_rate": 0.0002,
      "loss": 1.5066,
      "step": 1455
    },
    {
      "epoch": 0.5313868613138686,
      "grad_norm": 0.15009012818336487,
      "learning_rate": 0.0002,
      "loss": 1.5793,
      "step": 1456
    },
    {
      "epoch": 0.5317518248175183,
      "grad_norm": 0.16029541194438934,
      "learning_rate": 0.0002,
      "loss": 1.5948,
      "step": 1457
    },
    {
      "epoch": 0.5321167883211679,
      "grad_norm": 0.14566195011138916,
      "learning_rate": 0.0002,
      "loss": 1.5203,
      "step": 1458
    },
    {
      "epoch": 0.5324817518248175,
      "grad_norm": 0.14309531450271606,
      "learning_rate": 0.0002,
      "loss": 1.6351,
      "step": 1459
    },
    {
      "epoch": 0.5328467153284672,
      "grad_norm": 0.160929873585701,
      "learning_rate": 0.0002,
      "loss": 1.5922,
      "step": 1460
    },
    {
      "epoch": 0.5332116788321168,
      "grad_norm": 0.1424490213394165,
      "learning_rate": 0.0002,
      "loss": 1.5083,
      "step": 1461
    },
    {
      "epoch": 0.5335766423357664,
      "grad_norm": 0.14685146510601044,
      "learning_rate": 0.0002,
      "loss": 1.5153,
      "step": 1462
    },
    {
      "epoch": 0.5339416058394161,
      "grad_norm": 0.14617018401622772,
      "learning_rate": 0.0002,
      "loss": 1.4905,
      "step": 1463
    },
    {
      "epoch": 0.5343065693430656,
      "grad_norm": 0.14849825203418732,
      "learning_rate": 0.0002,
      "loss": 1.5153,
      "step": 1464
    },
    {
      "epoch": 0.5346715328467153,
      "grad_norm": 0.1640341579914093,
      "learning_rate": 0.0002,
      "loss": 1.608,
      "step": 1465
    },
    {
      "epoch": 0.535036496350365,
      "grad_norm": 0.15562793612480164,
      "learning_rate": 0.0002,
      "loss": 1.5788,
      "step": 1466
    },
    {
      "epoch": 0.5354014598540145,
      "grad_norm": 0.15221478044986725,
      "learning_rate": 0.0002,
      "loss": 1.5815,
      "step": 1467
    },
    {
      "epoch": 0.5357664233576642,
      "grad_norm": 0.13986995816230774,
      "learning_rate": 0.0002,
      "loss": 1.5707,
      "step": 1468
    },
    {
      "epoch": 0.5361313868613139,
      "grad_norm": 0.14589530229568481,
      "learning_rate": 0.0002,
      "loss": 1.5299,
      "step": 1469
    },
    {
      "epoch": 0.5364963503649635,
      "grad_norm": 0.14563673734664917,
      "learning_rate": 0.0002,
      "loss": 1.5558,
      "step": 1470
    },
    {
      "epoch": 0.5368613138686131,
      "grad_norm": 0.15010710060596466,
      "learning_rate": 0.0002,
      "loss": 1.5022,
      "step": 1471
    },
    {
      "epoch": 0.5372262773722628,
      "grad_norm": 0.15333884954452515,
      "learning_rate": 0.0002,
      "loss": 1.5521,
      "step": 1472
    },
    {
      "epoch": 0.5375912408759124,
      "grad_norm": 0.14684328436851501,
      "learning_rate": 0.0002,
      "loss": 1.4947,
      "step": 1473
    },
    {
      "epoch": 0.537956204379562,
      "grad_norm": 0.1490885615348816,
      "learning_rate": 0.0002,
      "loss": 1.5827,
      "step": 1474
    },
    {
      "epoch": 0.5383211678832117,
      "grad_norm": 0.15195229649543762,
      "learning_rate": 0.0002,
      "loss": 1.6335,
      "step": 1475
    },
    {
      "epoch": 0.5386861313868613,
      "grad_norm": 0.15622113645076752,
      "learning_rate": 0.0002,
      "loss": 1.4857,
      "step": 1476
    },
    {
      "epoch": 0.539051094890511,
      "grad_norm": 0.14525312185287476,
      "learning_rate": 0.0002,
      "loss": 1.5225,
      "step": 1477
    },
    {
      "epoch": 0.5394160583941606,
      "grad_norm": 0.16221848130226135,
      "learning_rate": 0.0002,
      "loss": 1.5171,
      "step": 1478
    },
    {
      "epoch": 0.5397810218978102,
      "grad_norm": 0.14893865585327148,
      "learning_rate": 0.0002,
      "loss": 1.6258,
      "step": 1479
    },
    {
      "epoch": 0.5401459854014599,
      "grad_norm": 0.17982861399650574,
      "learning_rate": 0.0002,
      "loss": 1.5557,
      "step": 1480
    },
    {
      "epoch": 0.5405109489051095,
      "grad_norm": 0.14447461068630219,
      "learning_rate": 0.0002,
      "loss": 1.5196,
      "step": 1481
    },
    {
      "epoch": 0.5408759124087591,
      "grad_norm": 0.1700318455696106,
      "learning_rate": 0.0002,
      "loss": 1.4763,
      "step": 1482
    },
    {
      "epoch": 0.5412408759124088,
      "grad_norm": 0.15758249163627625,
      "learning_rate": 0.0002,
      "loss": 1.5429,
      "step": 1483
    },
    {
      "epoch": 0.5416058394160584,
      "grad_norm": 0.17124861478805542,
      "learning_rate": 0.0002,
      "loss": 1.5621,
      "step": 1484
    },
    {
      "epoch": 0.541970802919708,
      "grad_norm": 0.17093044519424438,
      "learning_rate": 0.0002,
      "loss": 1.5909,
      "step": 1485
    },
    {
      "epoch": 0.5423357664233577,
      "grad_norm": 0.14535580575466156,
      "learning_rate": 0.0002,
      "loss": 1.5271,
      "step": 1486
    },
    {
      "epoch": 0.5427007299270074,
      "grad_norm": 0.18277332186698914,
      "learning_rate": 0.0002,
      "loss": 1.5723,
      "step": 1487
    },
    {
      "epoch": 0.5430656934306569,
      "grad_norm": 0.14439459145069122,
      "learning_rate": 0.0002,
      "loss": 1.5214,
      "step": 1488
    },
    {
      "epoch": 0.5434306569343066,
      "grad_norm": 0.17486318945884705,
      "learning_rate": 0.0002,
      "loss": 1.5474,
      "step": 1489
    },
    {
      "epoch": 0.5437956204379562,
      "grad_norm": 0.16188158094882965,
      "learning_rate": 0.0002,
      "loss": 1.5336,
      "step": 1490
    },
    {
      "epoch": 0.5441605839416058,
      "grad_norm": 0.1676185429096222,
      "learning_rate": 0.0002,
      "loss": 1.5508,
      "step": 1491
    },
    {
      "epoch": 0.5445255474452555,
      "grad_norm": 0.16415934264659882,
      "learning_rate": 0.0002,
      "loss": 1.5215,
      "step": 1492
    },
    {
      "epoch": 0.5448905109489051,
      "grad_norm": 0.1473916620016098,
      "learning_rate": 0.0002,
      "loss": 1.5467,
      "step": 1493
    },
    {
      "epoch": 0.5452554744525547,
      "grad_norm": 0.16913889348506927,
      "learning_rate": 0.0002,
      "loss": 1.5767,
      "step": 1494
    },
    {
      "epoch": 0.5456204379562044,
      "grad_norm": 0.14560498297214508,
      "learning_rate": 0.0002,
      "loss": 1.5666,
      "step": 1495
    },
    {
      "epoch": 0.545985401459854,
      "grad_norm": 0.16293300688266754,
      "learning_rate": 0.0002,
      "loss": 1.4362,
      "step": 1496
    },
    {
      "epoch": 0.5463503649635036,
      "grad_norm": 0.13691197335720062,
      "learning_rate": 0.0002,
      "loss": 1.5197,
      "step": 1497
    },
    {
      "epoch": 0.5467153284671533,
      "grad_norm": 0.1552390307188034,
      "learning_rate": 0.0002,
      "loss": 1.5144,
      "step": 1498
    },
    {
      "epoch": 0.5470802919708029,
      "grad_norm": 0.1576056331396103,
      "learning_rate": 0.0002,
      "loss": 1.5837,
      "step": 1499
    },
    {
      "epoch": 0.5474452554744526,
      "grad_norm": 0.14970625936985016,
      "learning_rate": 0.0002,
      "loss": 1.5271,
      "step": 1500
    },
    {
      "epoch": 0.5478102189781022,
      "grad_norm": 0.1490573137998581,
      "learning_rate": 0.0002,
      "loss": 1.4696,
      "step": 1501
    },
    {
      "epoch": 0.5481751824817518,
      "grad_norm": 0.146397203207016,
      "learning_rate": 0.0002,
      "loss": 1.5027,
      "step": 1502
    },
    {
      "epoch": 0.5485401459854015,
      "grad_norm": 0.16165632009506226,
      "learning_rate": 0.0002,
      "loss": 1.564,
      "step": 1503
    },
    {
      "epoch": 0.5489051094890511,
      "grad_norm": 0.13937240839004517,
      "learning_rate": 0.0002,
      "loss": 1.5173,
      "step": 1504
    },
    {
      "epoch": 0.5492700729927007,
      "grad_norm": 0.15102897584438324,
      "learning_rate": 0.0002,
      "loss": 1.4791,
      "step": 1505
    },
    {
      "epoch": 0.5496350364963504,
      "grad_norm": 0.14758263528347015,
      "learning_rate": 0.0002,
      "loss": 1.527,
      "step": 1506
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.14474795758724213,
      "learning_rate": 0.0002,
      "loss": 1.5697,
      "step": 1507
    },
    {
      "epoch": 0.5503649635036496,
      "grad_norm": 0.15219804644584656,
      "learning_rate": 0.0002,
      "loss": 1.5464,
      "step": 1508
    },
    {
      "epoch": 0.5507299270072993,
      "grad_norm": 0.1394841969013214,
      "learning_rate": 0.0002,
      "loss": 1.5452,
      "step": 1509
    },
    {
      "epoch": 0.551094890510949,
      "grad_norm": 0.15191805362701416,
      "learning_rate": 0.0002,
      "loss": 1.5745,
      "step": 1510
    },
    {
      "epoch": 0.5514598540145985,
      "grad_norm": 0.16653740406036377,
      "learning_rate": 0.0002,
      "loss": 1.6298,
      "step": 1511
    },
    {
      "epoch": 0.5518248175182482,
      "grad_norm": 0.1485438495874405,
      "learning_rate": 0.0002,
      "loss": 1.521,
      "step": 1512
    },
    {
      "epoch": 0.5521897810218979,
      "grad_norm": 0.15017135441303253,
      "learning_rate": 0.0002,
      "loss": 1.5373,
      "step": 1513
    },
    {
      "epoch": 0.5525547445255474,
      "grad_norm": 0.14504367113113403,
      "learning_rate": 0.0002,
      "loss": 1.553,
      "step": 1514
    },
    {
      "epoch": 0.5529197080291971,
      "grad_norm": 0.1617913693189621,
      "learning_rate": 0.0002,
      "loss": 1.4668,
      "step": 1515
    },
    {
      "epoch": 0.5532846715328467,
      "grad_norm": 0.1418709009885788,
      "learning_rate": 0.0002,
      "loss": 1.4797,
      "step": 1516
    },
    {
      "epoch": 0.5536496350364963,
      "grad_norm": 0.1563291996717453,
      "learning_rate": 0.0002,
      "loss": 1.6021,
      "step": 1517
    },
    {
      "epoch": 0.554014598540146,
      "grad_norm": 0.14263997972011566,
      "learning_rate": 0.0002,
      "loss": 1.5561,
      "step": 1518
    },
    {
      "epoch": 0.5543795620437956,
      "grad_norm": 0.15715628862380981,
      "learning_rate": 0.0002,
      "loss": 1.6089,
      "step": 1519
    },
    {
      "epoch": 0.5547445255474452,
      "grad_norm": 0.14187920093536377,
      "learning_rate": 0.0002,
      "loss": 1.438,
      "step": 1520
    },
    {
      "epoch": 0.5551094890510949,
      "grad_norm": 0.14415620267391205,
      "learning_rate": 0.0002,
      "loss": 1.5483,
      "step": 1521
    },
    {
      "epoch": 0.5554744525547445,
      "grad_norm": 0.1551968902349472,
      "learning_rate": 0.0002,
      "loss": 1.4987,
      "step": 1522
    },
    {
      "epoch": 0.5558394160583942,
      "grad_norm": 0.1510506570339203,
      "learning_rate": 0.0002,
      "loss": 1.5476,
      "step": 1523
    },
    {
      "epoch": 0.5562043795620438,
      "grad_norm": 0.14491407573223114,
      "learning_rate": 0.0002,
      "loss": 1.5376,
      "step": 1524
    },
    {
      "epoch": 0.5565693430656934,
      "grad_norm": 0.1543980836868286,
      "learning_rate": 0.0002,
      "loss": 1.5833,
      "step": 1525
    },
    {
      "epoch": 0.5569343065693431,
      "grad_norm": 0.14973662793636322,
      "learning_rate": 0.0002,
      "loss": 1.5399,
      "step": 1526
    },
    {
      "epoch": 0.5572992700729927,
      "grad_norm": 0.14524134993553162,
      "learning_rate": 0.0002,
      "loss": 1.5248,
      "step": 1527
    },
    {
      "epoch": 0.5576642335766423,
      "grad_norm": 0.145666241645813,
      "learning_rate": 0.0002,
      "loss": 1.5165,
      "step": 1528
    },
    {
      "epoch": 0.558029197080292,
      "grad_norm": 0.14676910638809204,
      "learning_rate": 0.0002,
      "loss": 1.5339,
      "step": 1529
    },
    {
      "epoch": 0.5583941605839416,
      "grad_norm": 0.1458742767572403,
      "learning_rate": 0.0002,
      "loss": 1.5494,
      "step": 1530
    },
    {
      "epoch": 0.5587591240875912,
      "grad_norm": 0.1531999111175537,
      "learning_rate": 0.0002,
      "loss": 1.5785,
      "step": 1531
    },
    {
      "epoch": 0.5591240875912409,
      "grad_norm": 0.14575308561325073,
      "learning_rate": 0.0002,
      "loss": 1.554,
      "step": 1532
    },
    {
      "epoch": 0.5594890510948906,
      "grad_norm": 0.14387720823287964,
      "learning_rate": 0.0002,
      "loss": 1.5284,
      "step": 1533
    },
    {
      "epoch": 0.5598540145985401,
      "grad_norm": 0.1459895670413971,
      "learning_rate": 0.0002,
      "loss": 1.4932,
      "step": 1534
    },
    {
      "epoch": 0.5602189781021898,
      "grad_norm": 0.14624802768230438,
      "learning_rate": 0.0002,
      "loss": 1.5154,
      "step": 1535
    },
    {
      "epoch": 0.5605839416058395,
      "grad_norm": 0.14317132532596588,
      "learning_rate": 0.0002,
      "loss": 1.5516,
      "step": 1536
    },
    {
      "epoch": 0.560948905109489,
      "grad_norm": 0.1417359858751297,
      "learning_rate": 0.0002,
      "loss": 1.5075,
      "step": 1537
    },
    {
      "epoch": 0.5613138686131387,
      "grad_norm": 0.14538337290287018,
      "learning_rate": 0.0002,
      "loss": 1.5528,
      "step": 1538
    },
    {
      "epoch": 0.5616788321167884,
      "grad_norm": 0.15181632339954376,
      "learning_rate": 0.0002,
      "loss": 1.6214,
      "step": 1539
    },
    {
      "epoch": 0.5620437956204379,
      "grad_norm": 0.13802820444107056,
      "learning_rate": 0.0002,
      "loss": 1.5351,
      "step": 1540
    },
    {
      "epoch": 0.5624087591240876,
      "grad_norm": 0.14750130474567413,
      "learning_rate": 0.0002,
      "loss": 1.575,
      "step": 1541
    },
    {
      "epoch": 0.5627737226277372,
      "grad_norm": 0.1430201232433319,
      "learning_rate": 0.0002,
      "loss": 1.5362,
      "step": 1542
    },
    {
      "epoch": 0.5631386861313868,
      "grad_norm": 0.14187805354595184,
      "learning_rate": 0.0002,
      "loss": 1.5781,
      "step": 1543
    },
    {
      "epoch": 0.5635036496350365,
      "grad_norm": 0.1518574059009552,
      "learning_rate": 0.0002,
      "loss": 1.5917,
      "step": 1544
    },
    {
      "epoch": 0.5638686131386861,
      "grad_norm": 0.14134760200977325,
      "learning_rate": 0.0002,
      "loss": 1.5772,
      "step": 1545
    },
    {
      "epoch": 0.5642335766423358,
      "grad_norm": 0.14532986283302307,
      "learning_rate": 0.0002,
      "loss": 1.5283,
      "step": 1546
    },
    {
      "epoch": 0.5645985401459854,
      "grad_norm": 0.13770711421966553,
      "learning_rate": 0.0002,
      "loss": 1.5164,
      "step": 1547
    },
    {
      "epoch": 0.564963503649635,
      "grad_norm": 0.1435277760028839,
      "learning_rate": 0.0002,
      "loss": 1.5234,
      "step": 1548
    },
    {
      "epoch": 0.5653284671532847,
      "grad_norm": 0.1536143720149994,
      "learning_rate": 0.0002,
      "loss": 1.6014,
      "step": 1549
    },
    {
      "epoch": 0.5656934306569343,
      "grad_norm": 0.15163132548332214,
      "learning_rate": 0.0002,
      "loss": 1.5656,
      "step": 1550
    },
    {
      "epoch": 0.5660583941605839,
      "grad_norm": 0.1525932252407074,
      "learning_rate": 0.0002,
      "loss": 1.5131,
      "step": 1551
    },
    {
      "epoch": 0.5664233576642336,
      "grad_norm": 0.13972796499729156,
      "learning_rate": 0.0002,
      "loss": 1.4164,
      "step": 1552
    },
    {
      "epoch": 0.5667883211678832,
      "grad_norm": 0.15311194956302643,
      "learning_rate": 0.0002,
      "loss": 1.5094,
      "step": 1553
    },
    {
      "epoch": 0.5671532846715328,
      "grad_norm": 0.15165290236473083,
      "learning_rate": 0.0002,
      "loss": 1.5442,
      "step": 1554
    },
    {
      "epoch": 0.5675182481751825,
      "grad_norm": 0.1594899743795395,
      "learning_rate": 0.0002,
      "loss": 1.5519,
      "step": 1555
    },
    {
      "epoch": 0.5678832116788322,
      "grad_norm": 0.14014659821987152,
      "learning_rate": 0.0002,
      "loss": 1.4512,
      "step": 1556
    },
    {
      "epoch": 0.5682481751824817,
      "grad_norm": 0.16296252608299255,
      "learning_rate": 0.0002,
      "loss": 1.5838,
      "step": 1557
    },
    {
      "epoch": 0.5686131386861314,
      "grad_norm": 0.1426456719636917,
      "learning_rate": 0.0002,
      "loss": 1.5723,
      "step": 1558
    },
    {
      "epoch": 0.5689781021897811,
      "grad_norm": 0.15538828074932098,
      "learning_rate": 0.0002,
      "loss": 1.5686,
      "step": 1559
    },
    {
      "epoch": 0.5693430656934306,
      "grad_norm": 0.1543959677219391,
      "learning_rate": 0.0002,
      "loss": 1.5468,
      "step": 1560
    },
    {
      "epoch": 0.5697080291970803,
      "grad_norm": 0.15348027646541595,
      "learning_rate": 0.0002,
      "loss": 1.5189,
      "step": 1561
    },
    {
      "epoch": 0.57007299270073,
      "grad_norm": 0.1558728814125061,
      "learning_rate": 0.0002,
      "loss": 1.5993,
      "step": 1562
    },
    {
      "epoch": 0.5704379562043795,
      "grad_norm": 0.15484938025474548,
      "learning_rate": 0.0002,
      "loss": 1.5077,
      "step": 1563
    },
    {
      "epoch": 0.5708029197080292,
      "grad_norm": 0.1438615471124649,
      "learning_rate": 0.0002,
      "loss": 1.5396,
      "step": 1564
    },
    {
      "epoch": 0.5711678832116789,
      "grad_norm": 0.1507718712091446,
      "learning_rate": 0.0002,
      "loss": 1.511,
      "step": 1565
    },
    {
      "epoch": 0.5715328467153284,
      "grad_norm": 0.15944881737232208,
      "learning_rate": 0.0002,
      "loss": 1.5358,
      "step": 1566
    },
    {
      "epoch": 0.5718978102189781,
      "grad_norm": 0.1424209624528885,
      "learning_rate": 0.0002,
      "loss": 1.4835,
      "step": 1567
    },
    {
      "epoch": 0.5722627737226277,
      "grad_norm": 0.15021011233329773,
      "learning_rate": 0.0002,
      "loss": 1.4469,
      "step": 1568
    },
    {
      "epoch": 0.5726277372262774,
      "grad_norm": 0.14689122140407562,
      "learning_rate": 0.0002,
      "loss": 1.4791,
      "step": 1569
    },
    {
      "epoch": 0.572992700729927,
      "grad_norm": 0.15941111743450165,
      "learning_rate": 0.0002,
      "loss": 1.5537,
      "step": 1570
    },
    {
      "epoch": 0.5733576642335766,
      "grad_norm": 0.14347757399082184,
      "learning_rate": 0.0002,
      "loss": 1.5759,
      "step": 1571
    },
    {
      "epoch": 0.5737226277372263,
      "grad_norm": 0.14420434832572937,
      "learning_rate": 0.0002,
      "loss": 1.4728,
      "step": 1572
    },
    {
      "epoch": 0.5740875912408759,
      "grad_norm": 0.14002388715744019,
      "learning_rate": 0.0002,
      "loss": 1.4753,
      "step": 1573
    },
    {
      "epoch": 0.5744525547445255,
      "grad_norm": 0.16281281411647797,
      "learning_rate": 0.0002,
      "loss": 1.5524,
      "step": 1574
    },
    {
      "epoch": 0.5748175182481752,
      "grad_norm": 0.14894481003284454,
      "learning_rate": 0.0002,
      "loss": 1.5024,
      "step": 1575
    },
    {
      "epoch": 0.5751824817518248,
      "grad_norm": 0.14908082783222198,
      "learning_rate": 0.0002,
      "loss": 1.5577,
      "step": 1576
    },
    {
      "epoch": 0.5755474452554744,
      "grad_norm": 0.13620726764202118,
      "learning_rate": 0.0002,
      "loss": 1.4162,
      "step": 1577
    },
    {
      "epoch": 0.5759124087591241,
      "grad_norm": 0.14078481495380402,
      "learning_rate": 0.0002,
      "loss": 1.5612,
      "step": 1578
    },
    {
      "epoch": 0.5762773722627738,
      "grad_norm": 0.15204910933971405,
      "learning_rate": 0.0002,
      "loss": 1.5028,
      "step": 1579
    },
    {
      "epoch": 0.5766423357664233,
      "grad_norm": 0.1652355045080185,
      "learning_rate": 0.0002,
      "loss": 1.514,
      "step": 1580
    },
    {
      "epoch": 0.577007299270073,
      "grad_norm": 0.14811913669109344,
      "learning_rate": 0.0002,
      "loss": 1.5894,
      "step": 1581
    },
    {
      "epoch": 0.5773722627737227,
      "grad_norm": 0.14432281255722046,
      "learning_rate": 0.0002,
      "loss": 1.4739,
      "step": 1582
    },
    {
      "epoch": 0.5777372262773722,
      "grad_norm": 0.15464159846305847,
      "learning_rate": 0.0002,
      "loss": 1.5923,
      "step": 1583
    },
    {
      "epoch": 0.5781021897810219,
      "grad_norm": 0.14613766968250275,
      "learning_rate": 0.0002,
      "loss": 1.5223,
      "step": 1584
    },
    {
      "epoch": 0.5784671532846716,
      "grad_norm": 0.1461004763841629,
      "learning_rate": 0.0002,
      "loss": 1.574,
      "step": 1585
    },
    {
      "epoch": 0.5788321167883211,
      "grad_norm": 0.14416202902793884,
      "learning_rate": 0.0002,
      "loss": 1.519,
      "step": 1586
    },
    {
      "epoch": 0.5791970802919708,
      "grad_norm": 0.1524561047554016,
      "learning_rate": 0.0002,
      "loss": 1.5653,
      "step": 1587
    },
    {
      "epoch": 0.5795620437956205,
      "grad_norm": 0.13999834656715393,
      "learning_rate": 0.0002,
      "loss": 1.4537,
      "step": 1588
    },
    {
      "epoch": 0.57992700729927,
      "grad_norm": 0.14300422370433807,
      "learning_rate": 0.0002,
      "loss": 1.4452,
      "step": 1589
    },
    {
      "epoch": 0.5802919708029197,
      "grad_norm": 0.14635343849658966,
      "learning_rate": 0.0002,
      "loss": 1.5645,
      "step": 1590
    },
    {
      "epoch": 0.5806569343065694,
      "grad_norm": 0.1450841873884201,
      "learning_rate": 0.0002,
      "loss": 1.447,
      "step": 1591
    },
    {
      "epoch": 0.581021897810219,
      "grad_norm": 0.1474352926015854,
      "learning_rate": 0.0002,
      "loss": 1.5669,
      "step": 1592
    },
    {
      "epoch": 0.5813868613138686,
      "grad_norm": 0.15958189964294434,
      "learning_rate": 0.0002,
      "loss": 1.6234,
      "step": 1593
    },
    {
      "epoch": 0.5817518248175182,
      "grad_norm": 0.16490836441516876,
      "learning_rate": 0.0002,
      "loss": 1.5843,
      "step": 1594
    },
    {
      "epoch": 0.5821167883211679,
      "grad_norm": 0.14691637456417084,
      "learning_rate": 0.0002,
      "loss": 1.5072,
      "step": 1595
    },
    {
      "epoch": 0.5824817518248175,
      "grad_norm": 0.14931736886501312,
      "learning_rate": 0.0002,
      "loss": 1.5217,
      "step": 1596
    },
    {
      "epoch": 0.5828467153284671,
      "grad_norm": 0.14678077399730682,
      "learning_rate": 0.0002,
      "loss": 1.5086,
      "step": 1597
    },
    {
      "epoch": 0.5832116788321168,
      "grad_norm": 0.1558760404586792,
      "learning_rate": 0.0002,
      "loss": 1.5453,
      "step": 1598
    },
    {
      "epoch": 0.5835766423357664,
      "grad_norm": 0.14891064167022705,
      "learning_rate": 0.0002,
      "loss": 1.5441,
      "step": 1599
    },
    {
      "epoch": 0.583941605839416,
      "grad_norm": 0.1427258849143982,
      "learning_rate": 0.0002,
      "loss": 1.5038,
      "step": 1600
    },
    {
      "epoch": 0.5843065693430657,
      "grad_norm": 0.14083349704742432,
      "learning_rate": 0.0002,
      "loss": 1.4931,
      "step": 1601
    },
    {
      "epoch": 0.5846715328467154,
      "grad_norm": 0.1403832882642746,
      "learning_rate": 0.0002,
      "loss": 1.4993,
      "step": 1602
    },
    {
      "epoch": 0.5850364963503649,
      "grad_norm": 0.1431397795677185,
      "learning_rate": 0.0002,
      "loss": 1.5412,
      "step": 1603
    },
    {
      "epoch": 0.5854014598540146,
      "grad_norm": 0.14802637696266174,
      "learning_rate": 0.0002,
      "loss": 1.539,
      "step": 1604
    },
    {
      "epoch": 0.5857664233576643,
      "grad_norm": 0.15701542794704437,
      "learning_rate": 0.0002,
      "loss": 1.5668,
      "step": 1605
    },
    {
      "epoch": 0.5861313868613138,
      "grad_norm": 0.1481952667236328,
      "learning_rate": 0.0002,
      "loss": 1.5166,
      "step": 1606
    },
    {
      "epoch": 0.5864963503649635,
      "grad_norm": 0.15412180125713348,
      "learning_rate": 0.0002,
      "loss": 1.5616,
      "step": 1607
    },
    {
      "epoch": 0.5868613138686132,
      "grad_norm": 0.15030193328857422,
      "learning_rate": 0.0002,
      "loss": 1.5128,
      "step": 1608
    },
    {
      "epoch": 0.5872262773722627,
      "grad_norm": 0.1430676281452179,
      "learning_rate": 0.0002,
      "loss": 1.5096,
      "step": 1609
    },
    {
      "epoch": 0.5875912408759124,
      "grad_norm": 0.1514965295791626,
      "learning_rate": 0.0002,
      "loss": 1.5415,
      "step": 1610
    },
    {
      "epoch": 0.5879562043795621,
      "grad_norm": 0.15042665600776672,
      "learning_rate": 0.0002,
      "loss": 1.4955,
      "step": 1611
    },
    {
      "epoch": 0.5883211678832116,
      "grad_norm": 0.14858083426952362,
      "learning_rate": 0.0002,
      "loss": 1.5961,
      "step": 1612
    },
    {
      "epoch": 0.5886861313868613,
      "grad_norm": 0.15292055904865265,
      "learning_rate": 0.0002,
      "loss": 1.5166,
      "step": 1613
    },
    {
      "epoch": 0.589051094890511,
      "grad_norm": 0.14459823071956635,
      "learning_rate": 0.0002,
      "loss": 1.504,
      "step": 1614
    },
    {
      "epoch": 0.5894160583941606,
      "grad_norm": 0.16127653419971466,
      "learning_rate": 0.0002,
      "loss": 1.581,
      "step": 1615
    },
    {
      "epoch": 0.5897810218978102,
      "grad_norm": 0.14814192056655884,
      "learning_rate": 0.0002,
      "loss": 1.54,
      "step": 1616
    },
    {
      "epoch": 0.5901459854014599,
      "grad_norm": 0.1507745087146759,
      "learning_rate": 0.0002,
      "loss": 1.4891,
      "step": 1617
    },
    {
      "epoch": 0.5905109489051095,
      "grad_norm": 0.15411219000816345,
      "learning_rate": 0.0002,
      "loss": 1.5368,
      "step": 1618
    },
    {
      "epoch": 0.5908759124087591,
      "grad_norm": 0.14444802701473236,
      "learning_rate": 0.0002,
      "loss": 1.5355,
      "step": 1619
    },
    {
      "epoch": 0.5912408759124088,
      "grad_norm": 0.15506042540073395,
      "learning_rate": 0.0002,
      "loss": 1.511,
      "step": 1620
    },
    {
      "epoch": 0.5916058394160584,
      "grad_norm": 0.1432209461927414,
      "learning_rate": 0.0002,
      "loss": 1.4409,
      "step": 1621
    },
    {
      "epoch": 0.591970802919708,
      "grad_norm": 0.14219534397125244,
      "learning_rate": 0.0002,
      "loss": 1.5578,
      "step": 1622
    },
    {
      "epoch": 0.5923357664233576,
      "grad_norm": 0.1586819291114807,
      "learning_rate": 0.0002,
      "loss": 1.5002,
      "step": 1623
    },
    {
      "epoch": 0.5927007299270073,
      "grad_norm": 0.1479002833366394,
      "learning_rate": 0.0002,
      "loss": 1.5232,
      "step": 1624
    },
    {
      "epoch": 0.593065693430657,
      "grad_norm": 0.14417928457260132,
      "learning_rate": 0.0002,
      "loss": 1.5338,
      "step": 1625
    },
    {
      "epoch": 0.5934306569343065,
      "grad_norm": 0.1486423909664154,
      "learning_rate": 0.0002,
      "loss": 1.492,
      "step": 1626
    },
    {
      "epoch": 0.5937956204379562,
      "grad_norm": 0.14356744289398193,
      "learning_rate": 0.0002,
      "loss": 1.5763,
      "step": 1627
    },
    {
      "epoch": 0.5941605839416059,
      "grad_norm": 0.15280532836914062,
      "learning_rate": 0.0002,
      "loss": 1.6307,
      "step": 1628
    },
    {
      "epoch": 0.5945255474452554,
      "grad_norm": 0.15000854432582855,
      "learning_rate": 0.0002,
      "loss": 1.4925,
      "step": 1629
    },
    {
      "epoch": 0.5948905109489051,
      "grad_norm": 0.1528661549091339,
      "learning_rate": 0.0002,
      "loss": 1.5323,
      "step": 1630
    },
    {
      "epoch": 0.5952554744525548,
      "grad_norm": 0.1417846530675888,
      "learning_rate": 0.0002,
      "loss": 1.5446,
      "step": 1631
    },
    {
      "epoch": 0.5956204379562043,
      "grad_norm": 0.1442185342311859,
      "learning_rate": 0.0002,
      "loss": 1.4828,
      "step": 1632
    },
    {
      "epoch": 0.595985401459854,
      "grad_norm": 0.1452329009771347,
      "learning_rate": 0.0002,
      "loss": 1.4621,
      "step": 1633
    },
    {
      "epoch": 0.5963503649635037,
      "grad_norm": 0.14758780598640442,
      "learning_rate": 0.0002,
      "loss": 1.5905,
      "step": 1634
    },
    {
      "epoch": 0.5967153284671532,
      "grad_norm": 0.15098930895328522,
      "learning_rate": 0.0002,
      "loss": 1.5493,
      "step": 1635
    },
    {
      "epoch": 0.5970802919708029,
      "grad_norm": 0.14864489436149597,
      "learning_rate": 0.0002,
      "loss": 1.5783,
      "step": 1636
    },
    {
      "epoch": 0.5974452554744526,
      "grad_norm": 0.14712050557136536,
      "learning_rate": 0.0002,
      "loss": 1.4869,
      "step": 1637
    },
    {
      "epoch": 0.5978102189781022,
      "grad_norm": 0.14295360445976257,
      "learning_rate": 0.0002,
      "loss": 1.4926,
      "step": 1638
    },
    {
      "epoch": 0.5981751824817518,
      "grad_norm": 0.14787009358406067,
      "learning_rate": 0.0002,
      "loss": 1.5903,
      "step": 1639
    },
    {
      "epoch": 0.5985401459854015,
      "grad_norm": 0.14039547741413116,
      "learning_rate": 0.0002,
      "loss": 1.4745,
      "step": 1640
    },
    {
      "epoch": 0.5989051094890511,
      "grad_norm": 0.15318597853183746,
      "learning_rate": 0.0002,
      "loss": 1.4809,
      "step": 1641
    },
    {
      "epoch": 0.5992700729927007,
      "grad_norm": 0.1496783196926117,
      "learning_rate": 0.0002,
      "loss": 1.5235,
      "step": 1642
    },
    {
      "epoch": 0.5996350364963504,
      "grad_norm": 0.15337395668029785,
      "learning_rate": 0.0002,
      "loss": 1.5274,
      "step": 1643
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.14359980821609497,
      "learning_rate": 0.0002,
      "loss": 1.5811,
      "step": 1644
    },
    {
      "epoch": 0.6003649635036497,
      "grad_norm": 0.15223738551139832,
      "learning_rate": 0.0002,
      "loss": 1.5048,
      "step": 1645
    },
    {
      "epoch": 0.6007299270072993,
      "grad_norm": 0.14601877331733704,
      "learning_rate": 0.0002,
      "loss": 1.5583,
      "step": 1646
    },
    {
      "epoch": 0.6010948905109489,
      "grad_norm": 0.14701956510543823,
      "learning_rate": 0.0002,
      "loss": 1.4801,
      "step": 1647
    },
    {
      "epoch": 0.6014598540145986,
      "grad_norm": 0.15017391741275787,
      "learning_rate": 0.0002,
      "loss": 1.5225,
      "step": 1648
    },
    {
      "epoch": 0.6018248175182481,
      "grad_norm": 0.149431973695755,
      "learning_rate": 0.0002,
      "loss": 1.5436,
      "step": 1649
    },
    {
      "epoch": 0.6021897810218978,
      "grad_norm": 0.15032467246055603,
      "learning_rate": 0.0002,
      "loss": 1.4852,
      "step": 1650
    },
    {
      "epoch": 0.6025547445255475,
      "grad_norm": 0.15596425533294678,
      "learning_rate": 0.0002,
      "loss": 1.6173,
      "step": 1651
    },
    {
      "epoch": 0.602919708029197,
      "grad_norm": 0.1619851291179657,
      "learning_rate": 0.0002,
      "loss": 1.5624,
      "step": 1652
    },
    {
      "epoch": 0.6032846715328467,
      "grad_norm": 0.14284424483776093,
      "learning_rate": 0.0002,
      "loss": 1.4981,
      "step": 1653
    },
    {
      "epoch": 0.6036496350364964,
      "grad_norm": 0.15262828767299652,
      "learning_rate": 0.0002,
      "loss": 1.5746,
      "step": 1654
    },
    {
      "epoch": 0.6040145985401459,
      "grad_norm": 0.15323911607265472,
      "learning_rate": 0.0002,
      "loss": 1.5367,
      "step": 1655
    },
    {
      "epoch": 0.6043795620437956,
      "grad_norm": 0.14814059436321259,
      "learning_rate": 0.0002,
      "loss": 1.5028,
      "step": 1656
    },
    {
      "epoch": 0.6047445255474453,
      "grad_norm": 0.14697305858135223,
      "learning_rate": 0.0002,
      "loss": 1.5219,
      "step": 1657
    },
    {
      "epoch": 0.6051094890510949,
      "grad_norm": 0.1441735327243805,
      "learning_rate": 0.0002,
      "loss": 1.4746,
      "step": 1658
    },
    {
      "epoch": 0.6054744525547445,
      "grad_norm": 0.14877024292945862,
      "learning_rate": 0.0002,
      "loss": 1.5068,
      "step": 1659
    },
    {
      "epoch": 0.6058394160583942,
      "grad_norm": 0.14579226076602936,
      "learning_rate": 0.0002,
      "loss": 1.4554,
      "step": 1660
    },
    {
      "epoch": 0.6062043795620438,
      "grad_norm": 0.1516817808151245,
      "learning_rate": 0.0002,
      "loss": 1.5462,
      "step": 1661
    },
    {
      "epoch": 0.6065693430656934,
      "grad_norm": 0.15068697929382324,
      "learning_rate": 0.0002,
      "loss": 1.5483,
      "step": 1662
    },
    {
      "epoch": 0.6069343065693431,
      "grad_norm": 0.15561826527118683,
      "learning_rate": 0.0002,
      "loss": 1.5641,
      "step": 1663
    },
    {
      "epoch": 0.6072992700729927,
      "grad_norm": 0.1643374264240265,
      "learning_rate": 0.0002,
      "loss": 1.4921,
      "step": 1664
    },
    {
      "epoch": 0.6076642335766423,
      "grad_norm": 0.1455034464597702,
      "learning_rate": 0.0002,
      "loss": 1.4887,
      "step": 1665
    },
    {
      "epoch": 0.608029197080292,
      "grad_norm": 0.14925020933151245,
      "learning_rate": 0.0002,
      "loss": 1.5045,
      "step": 1666
    },
    {
      "epoch": 0.6083941605839416,
      "grad_norm": 0.16973891854286194,
      "learning_rate": 0.0002,
      "loss": 1.586,
      "step": 1667
    },
    {
      "epoch": 0.6087591240875913,
      "grad_norm": 0.15231965482234955,
      "learning_rate": 0.0002,
      "loss": 1.5842,
      "step": 1668
    },
    {
      "epoch": 0.6091240875912409,
      "grad_norm": 0.16132211685180664,
      "learning_rate": 0.0002,
      "loss": 1.5323,
      "step": 1669
    },
    {
      "epoch": 0.6094890510948905,
      "grad_norm": 0.14981502294540405,
      "learning_rate": 0.0002,
      "loss": 1.5148,
      "step": 1670
    },
    {
      "epoch": 0.6098540145985402,
      "grad_norm": 0.16185474395751953,
      "learning_rate": 0.0002,
      "loss": 1.5501,
      "step": 1671
    },
    {
      "epoch": 0.6102189781021898,
      "grad_norm": 0.14550945162773132,
      "learning_rate": 0.0002,
      "loss": 1.538,
      "step": 1672
    },
    {
      "epoch": 0.6105839416058394,
      "grad_norm": 0.14802230894565582,
      "learning_rate": 0.0002,
      "loss": 1.5236,
      "step": 1673
    },
    {
      "epoch": 0.6109489051094891,
      "grad_norm": 0.15031422674655914,
      "learning_rate": 0.0002,
      "loss": 1.5736,
      "step": 1674
    },
    {
      "epoch": 0.6113138686131386,
      "grad_norm": 0.14818523824214935,
      "learning_rate": 0.0002,
      "loss": 1.563,
      "step": 1675
    },
    {
      "epoch": 0.6116788321167883,
      "grad_norm": 0.15101830661296844,
      "learning_rate": 0.0002,
      "loss": 1.5053,
      "step": 1676
    },
    {
      "epoch": 0.612043795620438,
      "grad_norm": 0.1591547429561615,
      "learning_rate": 0.0002,
      "loss": 1.5835,
      "step": 1677
    },
    {
      "epoch": 0.6124087591240875,
      "grad_norm": 0.1575150489807129,
      "learning_rate": 0.0002,
      "loss": 1.5978,
      "step": 1678
    },
    {
      "epoch": 0.6127737226277372,
      "grad_norm": 0.1473587304353714,
      "learning_rate": 0.0002,
      "loss": 1.4801,
      "step": 1679
    },
    {
      "epoch": 0.6131386861313869,
      "grad_norm": 0.1465386301279068,
      "learning_rate": 0.0002,
      "loss": 1.5379,
      "step": 1680
    },
    {
      "epoch": 0.6135036496350365,
      "grad_norm": 0.14544913172721863,
      "learning_rate": 0.0002,
      "loss": 1.4509,
      "step": 1681
    },
    {
      "epoch": 0.6138686131386861,
      "grad_norm": 0.14975157380104065,
      "learning_rate": 0.0002,
      "loss": 1.5351,
      "step": 1682
    },
    {
      "epoch": 0.6142335766423358,
      "grad_norm": 0.1463102400302887,
      "learning_rate": 0.0002,
      "loss": 1.5023,
      "step": 1683
    },
    {
      "epoch": 0.6145985401459854,
      "grad_norm": 0.15139378607273102,
      "learning_rate": 0.0002,
      "loss": 1.5651,
      "step": 1684
    },
    {
      "epoch": 0.614963503649635,
      "grad_norm": 0.14316824078559875,
      "learning_rate": 0.0002,
      "loss": 1.4758,
      "step": 1685
    },
    {
      "epoch": 0.6153284671532847,
      "grad_norm": 0.14461883902549744,
      "learning_rate": 0.0002,
      "loss": 1.4302,
      "step": 1686
    },
    {
      "epoch": 0.6156934306569343,
      "grad_norm": 0.14871154725551605,
      "learning_rate": 0.0002,
      "loss": 1.444,
      "step": 1687
    },
    {
      "epoch": 0.6160583941605839,
      "grad_norm": 0.1483636498451233,
      "learning_rate": 0.0002,
      "loss": 1.5395,
      "step": 1688
    },
    {
      "epoch": 0.6164233576642336,
      "grad_norm": 0.14860273897647858,
      "learning_rate": 0.0002,
      "loss": 1.5316,
      "step": 1689
    },
    {
      "epoch": 0.6167883211678832,
      "grad_norm": 0.14946942031383514,
      "learning_rate": 0.0002,
      "loss": 1.6206,
      "step": 1690
    },
    {
      "epoch": 0.6171532846715329,
      "grad_norm": 0.15888768434524536,
      "learning_rate": 0.0002,
      "loss": 1.5604,
      "step": 1691
    },
    {
      "epoch": 0.6175182481751825,
      "grad_norm": 0.16070418059825897,
      "learning_rate": 0.0002,
      "loss": 1.6327,
      "step": 1692
    },
    {
      "epoch": 0.6178832116788321,
      "grad_norm": 0.14780502021312714,
      "learning_rate": 0.0002,
      "loss": 1.5728,
      "step": 1693
    },
    {
      "epoch": 0.6182481751824818,
      "grad_norm": 0.1536567062139511,
      "learning_rate": 0.0002,
      "loss": 1.4821,
      "step": 1694
    },
    {
      "epoch": 0.6186131386861314,
      "grad_norm": 0.15179897844791412,
      "learning_rate": 0.0002,
      "loss": 1.5268,
      "step": 1695
    },
    {
      "epoch": 0.618978102189781,
      "grad_norm": 0.16133224964141846,
      "learning_rate": 0.0002,
      "loss": 1.54,
      "step": 1696
    },
    {
      "epoch": 0.6193430656934307,
      "grad_norm": 0.14570356905460358,
      "learning_rate": 0.0002,
      "loss": 1.4869,
      "step": 1697
    },
    {
      "epoch": 0.6197080291970803,
      "grad_norm": 0.14815129339694977,
      "learning_rate": 0.0002,
      "loss": 1.5231,
      "step": 1698
    },
    {
      "epoch": 0.6200729927007299,
      "grad_norm": 0.14624793827533722,
      "learning_rate": 0.0002,
      "loss": 1.5592,
      "step": 1699
    },
    {
      "epoch": 0.6204379562043796,
      "grad_norm": 0.14839771389961243,
      "learning_rate": 0.0002,
      "loss": 1.5541,
      "step": 1700
    },
    {
      "epoch": 0.6208029197080291,
      "grad_norm": 0.15723435580730438,
      "learning_rate": 0.0002,
      "loss": 1.5216,
      "step": 1701
    },
    {
      "epoch": 0.6211678832116788,
      "grad_norm": 0.14696279168128967,
      "learning_rate": 0.0002,
      "loss": 1.449,
      "step": 1702
    },
    {
      "epoch": 0.6215328467153285,
      "grad_norm": 0.14744721353054047,
      "learning_rate": 0.0002,
      "loss": 1.5311,
      "step": 1703
    },
    {
      "epoch": 0.621897810218978,
      "grad_norm": 0.15904685854911804,
      "learning_rate": 0.0002,
      "loss": 1.5025,
      "step": 1704
    },
    {
      "epoch": 0.6222627737226277,
      "grad_norm": 0.1461707204580307,
      "learning_rate": 0.0002,
      "loss": 1.4936,
      "step": 1705
    },
    {
      "epoch": 0.6226277372262774,
      "grad_norm": 0.14879336953163147,
      "learning_rate": 0.0002,
      "loss": 1.5083,
      "step": 1706
    },
    {
      "epoch": 0.622992700729927,
      "grad_norm": 0.1499076932668686,
      "learning_rate": 0.0002,
      "loss": 1.5061,
      "step": 1707
    },
    {
      "epoch": 0.6233576642335766,
      "grad_norm": 0.15982608497142792,
      "learning_rate": 0.0002,
      "loss": 1.5919,
      "step": 1708
    },
    {
      "epoch": 0.6237226277372263,
      "grad_norm": 0.14896990358829498,
      "learning_rate": 0.0002,
      "loss": 1.4999,
      "step": 1709
    },
    {
      "epoch": 0.6240875912408759,
      "grad_norm": 0.1585961878299713,
      "learning_rate": 0.0002,
      "loss": 1.5862,
      "step": 1710
    },
    {
      "epoch": 0.6244525547445255,
      "grad_norm": 0.15127073228359222,
      "learning_rate": 0.0002,
      "loss": 1.5106,
      "step": 1711
    },
    {
      "epoch": 0.6248175182481752,
      "grad_norm": 0.14916592836380005,
      "learning_rate": 0.0002,
      "loss": 1.5335,
      "step": 1712
    },
    {
      "epoch": 0.6251824817518248,
      "grad_norm": 0.14769122004508972,
      "learning_rate": 0.0002,
      "loss": 1.5205,
      "step": 1713
    },
    {
      "epoch": 0.6255474452554745,
      "grad_norm": 0.14720481634140015,
      "learning_rate": 0.0002,
      "loss": 1.5292,
      "step": 1714
    },
    {
      "epoch": 0.6259124087591241,
      "grad_norm": 0.14446903765201569,
      "learning_rate": 0.0002,
      "loss": 1.4229,
      "step": 1715
    },
    {
      "epoch": 0.6262773722627737,
      "grad_norm": 0.15459735691547394,
      "learning_rate": 0.0002,
      "loss": 1.5612,
      "step": 1716
    },
    {
      "epoch": 0.6266423357664234,
      "grad_norm": 0.14367833733558655,
      "learning_rate": 0.0002,
      "loss": 1.5086,
      "step": 1717
    },
    {
      "epoch": 0.627007299270073,
      "grad_norm": 0.15038259327411652,
      "learning_rate": 0.0002,
      "loss": 1.5212,
      "step": 1718
    },
    {
      "epoch": 0.6273722627737226,
      "grad_norm": 0.14840519428253174,
      "learning_rate": 0.0002,
      "loss": 1.4761,
      "step": 1719
    },
    {
      "epoch": 0.6277372262773723,
      "grad_norm": 0.15820124745368958,
      "learning_rate": 0.0002,
      "loss": 1.5448,
      "step": 1720
    },
    {
      "epoch": 0.628102189781022,
      "grad_norm": 0.14968444406986237,
      "learning_rate": 0.0002,
      "loss": 1.5355,
      "step": 1721
    },
    {
      "epoch": 0.6284671532846715,
      "grad_norm": 0.15642733871936798,
      "learning_rate": 0.0002,
      "loss": 1.5427,
      "step": 1722
    },
    {
      "epoch": 0.6288321167883212,
      "grad_norm": 0.15007038414478302,
      "learning_rate": 0.0002,
      "loss": 1.4618,
      "step": 1723
    },
    {
      "epoch": 0.6291970802919709,
      "grad_norm": 0.15174324810504913,
      "learning_rate": 0.0002,
      "loss": 1.5058,
      "step": 1724
    },
    {
      "epoch": 0.6295620437956204,
      "grad_norm": 0.16173364222049713,
      "learning_rate": 0.0002,
      "loss": 1.5443,
      "step": 1725
    },
    {
      "epoch": 0.6299270072992701,
      "grad_norm": 0.15059955418109894,
      "learning_rate": 0.0002,
      "loss": 1.5103,
      "step": 1726
    },
    {
      "epoch": 0.6302919708029197,
      "grad_norm": 0.15490096807479858,
      "learning_rate": 0.0002,
      "loss": 1.5605,
      "step": 1727
    },
    {
      "epoch": 0.6306569343065693,
      "grad_norm": 0.15008287131786346,
      "learning_rate": 0.0002,
      "loss": 1.4891,
      "step": 1728
    },
    {
      "epoch": 0.631021897810219,
      "grad_norm": 0.14691118896007538,
      "learning_rate": 0.0002,
      "loss": 1.4356,
      "step": 1729
    },
    {
      "epoch": 0.6313868613138686,
      "grad_norm": 0.14695879817008972,
      "learning_rate": 0.0002,
      "loss": 1.4709,
      "step": 1730
    },
    {
      "epoch": 0.6317518248175182,
      "grad_norm": 0.1526099294424057,
      "learning_rate": 0.0002,
      "loss": 1.562,
      "step": 1731
    },
    {
      "epoch": 0.6321167883211679,
      "grad_norm": 0.14319702982902527,
      "learning_rate": 0.0002,
      "loss": 1.5228,
      "step": 1732
    },
    {
      "epoch": 0.6324817518248175,
      "grad_norm": 0.14752674102783203,
      "learning_rate": 0.0002,
      "loss": 1.4786,
      "step": 1733
    },
    {
      "epoch": 0.6328467153284671,
      "grad_norm": 0.1520233154296875,
      "learning_rate": 0.0002,
      "loss": 1.4907,
      "step": 1734
    },
    {
      "epoch": 0.6332116788321168,
      "grad_norm": 0.15107610821723938,
      "learning_rate": 0.0002,
      "loss": 1.5017,
      "step": 1735
    },
    {
      "epoch": 0.6335766423357664,
      "grad_norm": 0.14665192365646362,
      "learning_rate": 0.0002,
      "loss": 1.4913,
      "step": 1736
    },
    {
      "epoch": 0.6339416058394161,
      "grad_norm": 0.14739574491977692,
      "learning_rate": 0.0002,
      "loss": 1.5454,
      "step": 1737
    },
    {
      "epoch": 0.6343065693430657,
      "grad_norm": 0.1503191739320755,
      "learning_rate": 0.0002,
      "loss": 1.5563,
      "step": 1738
    },
    {
      "epoch": 0.6346715328467153,
      "grad_norm": 0.15174901485443115,
      "learning_rate": 0.0002,
      "loss": 1.5631,
      "step": 1739
    },
    {
      "epoch": 0.635036496350365,
      "grad_norm": 0.15039384365081787,
      "learning_rate": 0.0002,
      "loss": 1.4969,
      "step": 1740
    },
    {
      "epoch": 0.6354014598540146,
      "grad_norm": 0.14737218618392944,
      "learning_rate": 0.0002,
      "loss": 1.5467,
      "step": 1741
    },
    {
      "epoch": 0.6357664233576642,
      "grad_norm": 0.14912404119968414,
      "learning_rate": 0.0002,
      "loss": 1.5321,
      "step": 1742
    },
    {
      "epoch": 0.6361313868613139,
      "grad_norm": 0.15059010684490204,
      "learning_rate": 0.0002,
      "loss": 1.5308,
      "step": 1743
    },
    {
      "epoch": 0.6364963503649635,
      "grad_norm": 0.15068691968917847,
      "learning_rate": 0.0002,
      "loss": 1.5709,
      "step": 1744
    },
    {
      "epoch": 0.6368613138686131,
      "grad_norm": 0.15062014758586884,
      "learning_rate": 0.0002,
      "loss": 1.4576,
      "step": 1745
    },
    {
      "epoch": 0.6372262773722628,
      "grad_norm": 0.1526610255241394,
      "learning_rate": 0.0002,
      "loss": 1.5605,
      "step": 1746
    },
    {
      "epoch": 0.6375912408759125,
      "grad_norm": 0.15358155965805054,
      "learning_rate": 0.0002,
      "loss": 1.57,
      "step": 1747
    },
    {
      "epoch": 0.637956204379562,
      "grad_norm": 0.15403412282466888,
      "learning_rate": 0.0002,
      "loss": 1.5461,
      "step": 1748
    },
    {
      "epoch": 0.6383211678832117,
      "grad_norm": 0.14670927822589874,
      "learning_rate": 0.0002,
      "loss": 1.5347,
      "step": 1749
    },
    {
      "epoch": 0.6386861313868614,
      "grad_norm": 0.14684735238552094,
      "learning_rate": 0.0002,
      "loss": 1.5731,
      "step": 1750
    },
    {
      "epoch": 0.6390510948905109,
      "grad_norm": 0.15669676661491394,
      "learning_rate": 0.0002,
      "loss": 1.5422,
      "step": 1751
    },
    {
      "epoch": 0.6394160583941606,
      "grad_norm": 0.1490134745836258,
      "learning_rate": 0.0002,
      "loss": 1.4903,
      "step": 1752
    },
    {
      "epoch": 0.6397810218978102,
      "grad_norm": 0.15787550806999207,
      "learning_rate": 0.0002,
      "loss": 1.5866,
      "step": 1753
    },
    {
      "epoch": 0.6401459854014598,
      "grad_norm": 0.15512911975383759,
      "learning_rate": 0.0002,
      "loss": 1.5772,
      "step": 1754
    },
    {
      "epoch": 0.6405109489051095,
      "grad_norm": 0.16660617291927338,
      "learning_rate": 0.0002,
      "loss": 1.5579,
      "step": 1755
    },
    {
      "epoch": 0.6408759124087591,
      "grad_norm": 0.1570538431406021,
      "learning_rate": 0.0002,
      "loss": 1.5702,
      "step": 1756
    },
    {
      "epoch": 0.6412408759124087,
      "grad_norm": 0.14692534506320953,
      "learning_rate": 0.0002,
      "loss": 1.4844,
      "step": 1757
    },
    {
      "epoch": 0.6416058394160584,
      "grad_norm": 0.14471369981765747,
      "learning_rate": 0.0002,
      "loss": 1.5457,
      "step": 1758
    },
    {
      "epoch": 0.641970802919708,
      "grad_norm": 0.14582712948322296,
      "learning_rate": 0.0002,
      "loss": 1.4898,
      "step": 1759
    },
    {
      "epoch": 0.6423357664233577,
      "grad_norm": 0.1591958999633789,
      "learning_rate": 0.0002,
      "loss": 1.5077,
      "step": 1760
    },
    {
      "epoch": 0.6427007299270073,
      "grad_norm": 0.14830170571804047,
      "learning_rate": 0.0002,
      "loss": 1.5198,
      "step": 1761
    },
    {
      "epoch": 0.6430656934306569,
      "grad_norm": 0.1492864489555359,
      "learning_rate": 0.0002,
      "loss": 1.4941,
      "step": 1762
    },
    {
      "epoch": 0.6434306569343066,
      "grad_norm": 0.14228138327598572,
      "learning_rate": 0.0002,
      "loss": 1.4823,
      "step": 1763
    },
    {
      "epoch": 0.6437956204379562,
      "grad_norm": 0.14651428163051605,
      "learning_rate": 0.0002,
      "loss": 1.5123,
      "step": 1764
    },
    {
      "epoch": 0.6441605839416058,
      "grad_norm": 0.14776110649108887,
      "learning_rate": 0.0002,
      "loss": 1.5207,
      "step": 1765
    },
    {
      "epoch": 0.6445255474452555,
      "grad_norm": 0.15335030853748322,
      "learning_rate": 0.0002,
      "loss": 1.4693,
      "step": 1766
    },
    {
      "epoch": 0.6448905109489051,
      "grad_norm": 0.1512148082256317,
      "learning_rate": 0.0002,
      "loss": 1.479,
      "step": 1767
    },
    {
      "epoch": 0.6452554744525547,
      "grad_norm": 0.1437731236219406,
      "learning_rate": 0.0002,
      "loss": 1.46,
      "step": 1768
    },
    {
      "epoch": 0.6456204379562044,
      "grad_norm": 0.14817383885383606,
      "learning_rate": 0.0002,
      "loss": 1.5629,
      "step": 1769
    },
    {
      "epoch": 0.6459854014598541,
      "grad_norm": 0.14361713826656342,
      "learning_rate": 0.0002,
      "loss": 1.4456,
      "step": 1770
    },
    {
      "epoch": 0.6463503649635036,
      "grad_norm": 0.15642237663269043,
      "learning_rate": 0.0002,
      "loss": 1.5319,
      "step": 1771
    },
    {
      "epoch": 0.6467153284671533,
      "grad_norm": 0.16091004014015198,
      "learning_rate": 0.0002,
      "loss": 1.5228,
      "step": 1772
    },
    {
      "epoch": 0.647080291970803,
      "grad_norm": 0.1552465409040451,
      "learning_rate": 0.0002,
      "loss": 1.468,
      "step": 1773
    },
    {
      "epoch": 0.6474452554744525,
      "grad_norm": 0.14960750937461853,
      "learning_rate": 0.0002,
      "loss": 1.4935,
      "step": 1774
    },
    {
      "epoch": 0.6478102189781022,
      "grad_norm": 0.14738021790981293,
      "learning_rate": 0.0002,
      "loss": 1.5323,
      "step": 1775
    },
    {
      "epoch": 0.6481751824817519,
      "grad_norm": 0.14530931413173676,
      "learning_rate": 0.0002,
      "loss": 1.5348,
      "step": 1776
    },
    {
      "epoch": 0.6485401459854014,
      "grad_norm": 0.1554272621870041,
      "learning_rate": 0.0002,
      "loss": 1.5647,
      "step": 1777
    },
    {
      "epoch": 0.6489051094890511,
      "grad_norm": 0.1468043327331543,
      "learning_rate": 0.0002,
      "loss": 1.5365,
      "step": 1778
    },
    {
      "epoch": 0.6492700729927007,
      "grad_norm": 0.15215426683425903,
      "learning_rate": 0.0002,
      "loss": 1.5195,
      "step": 1779
    },
    {
      "epoch": 0.6496350364963503,
      "grad_norm": 0.1512681096792221,
      "learning_rate": 0.0002,
      "loss": 1.5942,
      "step": 1780
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.15007442235946655,
      "learning_rate": 0.0002,
      "loss": 1.5198,
      "step": 1781
    },
    {
      "epoch": 0.6503649635036496,
      "grad_norm": 0.15618249773979187,
      "learning_rate": 0.0002,
      "loss": 1.5306,
      "step": 1782
    },
    {
      "epoch": 0.6507299270072993,
      "grad_norm": 0.1492588222026825,
      "learning_rate": 0.0002,
      "loss": 1.5392,
      "step": 1783
    },
    {
      "epoch": 0.6510948905109489,
      "grad_norm": 0.14574159681797028,
      "learning_rate": 0.0002,
      "loss": 1.4449,
      "step": 1784
    },
    {
      "epoch": 0.6514598540145985,
      "grad_norm": 0.14489562809467316,
      "learning_rate": 0.0002,
      "loss": 1.5235,
      "step": 1785
    },
    {
      "epoch": 0.6518248175182482,
      "grad_norm": 0.15218301117420197,
      "learning_rate": 0.0002,
      "loss": 1.5755,
      "step": 1786
    },
    {
      "epoch": 0.6521897810218978,
      "grad_norm": 0.14621035754680634,
      "learning_rate": 0.0002,
      "loss": 1.4739,
      "step": 1787
    },
    {
      "epoch": 0.6525547445255474,
      "grad_norm": 0.15417440235614777,
      "learning_rate": 0.0002,
      "loss": 1.5132,
      "step": 1788
    },
    {
      "epoch": 0.6529197080291971,
      "grad_norm": 0.14422665536403656,
      "learning_rate": 0.0002,
      "loss": 1.5516,
      "step": 1789
    },
    {
      "epoch": 0.6532846715328468,
      "grad_norm": 0.15233705937862396,
      "learning_rate": 0.0002,
      "loss": 1.4786,
      "step": 1790
    },
    {
      "epoch": 0.6536496350364963,
      "grad_norm": 0.1493826061487198,
      "learning_rate": 0.0002,
      "loss": 1.5261,
      "step": 1791
    },
    {
      "epoch": 0.654014598540146,
      "grad_norm": 0.15102985501289368,
      "learning_rate": 0.0002,
      "loss": 1.4868,
      "step": 1792
    },
    {
      "epoch": 0.6543795620437957,
      "grad_norm": 0.15218611061573029,
      "learning_rate": 0.0002,
      "loss": 1.5655,
      "step": 1793
    },
    {
      "epoch": 0.6547445255474452,
      "grad_norm": 0.14489345252513885,
      "learning_rate": 0.0002,
      "loss": 1.4883,
      "step": 1794
    },
    {
      "epoch": 0.6551094890510949,
      "grad_norm": 0.14699922502040863,
      "learning_rate": 0.0002,
      "loss": 1.4981,
      "step": 1795
    },
    {
      "epoch": 0.6554744525547446,
      "grad_norm": 0.15241490304470062,
      "learning_rate": 0.0002,
      "loss": 1.5282,
      "step": 1796
    },
    {
      "epoch": 0.6558394160583941,
      "grad_norm": 0.15205129981040955,
      "learning_rate": 0.0002,
      "loss": 1.4409,
      "step": 1797
    },
    {
      "epoch": 0.6562043795620438,
      "grad_norm": 0.15404455363750458,
      "learning_rate": 0.0002,
      "loss": 1.5526,
      "step": 1798
    },
    {
      "epoch": 0.6565693430656935,
      "grad_norm": 0.15189717710018158,
      "learning_rate": 0.0002,
      "loss": 1.579,
      "step": 1799
    },
    {
      "epoch": 0.656934306569343,
      "grad_norm": 0.1537383794784546,
      "learning_rate": 0.0002,
      "loss": 1.5144,
      "step": 1800
    },
    {
      "epoch": 0.6572992700729927,
      "grad_norm": 0.15773405134677887,
      "learning_rate": 0.0002,
      "loss": 1.5524,
      "step": 1801
    },
    {
      "epoch": 0.6576642335766424,
      "grad_norm": 0.1555570662021637,
      "learning_rate": 0.0002,
      "loss": 1.5207,
      "step": 1802
    },
    {
      "epoch": 0.658029197080292,
      "grad_norm": 0.15459375083446503,
      "learning_rate": 0.0002,
      "loss": 1.4788,
      "step": 1803
    },
    {
      "epoch": 0.6583941605839416,
      "grad_norm": 0.17016124725341797,
      "learning_rate": 0.0002,
      "loss": 1.5799,
      "step": 1804
    },
    {
      "epoch": 0.6587591240875912,
      "grad_norm": 0.15209294855594635,
      "learning_rate": 0.0002,
      "loss": 1.5447,
      "step": 1805
    },
    {
      "epoch": 0.6591240875912409,
      "grad_norm": 0.14991869032382965,
      "learning_rate": 0.0002,
      "loss": 1.5131,
      "step": 1806
    },
    {
      "epoch": 0.6594890510948905,
      "grad_norm": 0.14774098992347717,
      "learning_rate": 0.0002,
      "loss": 1.5053,
      "step": 1807
    },
    {
      "epoch": 0.6598540145985401,
      "grad_norm": 0.16745632886886597,
      "learning_rate": 0.0002,
      "loss": 1.5714,
      "step": 1808
    },
    {
      "epoch": 0.6602189781021898,
      "grad_norm": 0.14453540742397308,
      "learning_rate": 0.0002,
      "loss": 1.5339,
      "step": 1809
    },
    {
      "epoch": 0.6605839416058394,
      "grad_norm": 0.1567346453666687,
      "learning_rate": 0.0002,
      "loss": 1.4449,
      "step": 1810
    },
    {
      "epoch": 0.660948905109489,
      "grad_norm": 0.14827527105808258,
      "learning_rate": 0.0002,
      "loss": 1.5322,
      "step": 1811
    },
    {
      "epoch": 0.6613138686131387,
      "grad_norm": 0.1618945598602295,
      "learning_rate": 0.0002,
      "loss": 1.5339,
      "step": 1812
    },
    {
      "epoch": 0.6616788321167884,
      "grad_norm": 0.15963982045650482,
      "learning_rate": 0.0002,
      "loss": 1.5861,
      "step": 1813
    },
    {
      "epoch": 0.6620437956204379,
      "grad_norm": 0.1676718294620514,
      "learning_rate": 0.0002,
      "loss": 1.5236,
      "step": 1814
    },
    {
      "epoch": 0.6624087591240876,
      "grad_norm": 0.15308934450149536,
      "learning_rate": 0.0002,
      "loss": 1.5312,
      "step": 1815
    },
    {
      "epoch": 0.6627737226277373,
      "grad_norm": 0.17498576641082764,
      "learning_rate": 0.0002,
      "loss": 1.5531,
      "step": 1816
    },
    {
      "epoch": 0.6631386861313868,
      "grad_norm": 0.14718714356422424,
      "learning_rate": 0.0002,
      "loss": 1.5022,
      "step": 1817
    },
    {
      "epoch": 0.6635036496350365,
      "grad_norm": 0.15819130837917328,
      "learning_rate": 0.0002,
      "loss": 1.5372,
      "step": 1818
    },
    {
      "epoch": 0.6638686131386862,
      "grad_norm": 0.15554796159267426,
      "learning_rate": 0.0002,
      "loss": 1.5611,
      "step": 1819
    },
    {
      "epoch": 0.6642335766423357,
      "grad_norm": 0.15229429304599762,
      "learning_rate": 0.0002,
      "loss": 1.569,
      "step": 1820
    },
    {
      "epoch": 0.6645985401459854,
      "grad_norm": 0.1652994006872177,
      "learning_rate": 0.0002,
      "loss": 1.564,
      "step": 1821
    },
    {
      "epoch": 0.6649635036496351,
      "grad_norm": 0.14414121210575104,
      "learning_rate": 0.0002,
      "loss": 1.491,
      "step": 1822
    },
    {
      "epoch": 0.6653284671532846,
      "grad_norm": 0.16653893887996674,
      "learning_rate": 0.0002,
      "loss": 1.5759,
      "step": 1823
    },
    {
      "epoch": 0.6656934306569343,
      "grad_norm": 0.15265034139156342,
      "learning_rate": 0.0002,
      "loss": 1.4982,
      "step": 1824
    },
    {
      "epoch": 0.666058394160584,
      "grad_norm": 0.15224497020244598,
      "learning_rate": 0.0002,
      "loss": 1.4509,
      "step": 1825
    },
    {
      "epoch": 0.6664233576642336,
      "grad_norm": 0.1668117642402649,
      "learning_rate": 0.0002,
      "loss": 1.5896,
      "step": 1826
    },
    {
      "epoch": 0.6667883211678832,
      "grad_norm": 0.1520099639892578,
      "learning_rate": 0.0002,
      "loss": 1.5112,
      "step": 1827
    },
    {
      "epoch": 0.6671532846715329,
      "grad_norm": 0.163130521774292,
      "learning_rate": 0.0002,
      "loss": 1.5699,
      "step": 1828
    },
    {
      "epoch": 0.6675182481751825,
      "grad_norm": 0.14272268116474152,
      "learning_rate": 0.0002,
      "loss": 1.4562,
      "step": 1829
    },
    {
      "epoch": 0.6678832116788321,
      "grad_norm": 0.16231217980384827,
      "learning_rate": 0.0002,
      "loss": 1.5308,
      "step": 1830
    },
    {
      "epoch": 0.6682481751824818,
      "grad_norm": 0.14881698787212372,
      "learning_rate": 0.0002,
      "loss": 1.4639,
      "step": 1831
    },
    {
      "epoch": 0.6686131386861314,
      "grad_norm": 0.14944985508918762,
      "learning_rate": 0.0002,
      "loss": 1.466,
      "step": 1832
    },
    {
      "epoch": 0.668978102189781,
      "grad_norm": 0.15484784543514252,
      "learning_rate": 0.0002,
      "loss": 1.4927,
      "step": 1833
    },
    {
      "epoch": 0.6693430656934306,
      "grad_norm": 0.14923568069934845,
      "learning_rate": 0.0002,
      "loss": 1.4925,
      "step": 1834
    },
    {
      "epoch": 0.6697080291970803,
      "grad_norm": 0.15071994066238403,
      "learning_rate": 0.0002,
      "loss": 1.4117,
      "step": 1835
    },
    {
      "epoch": 0.67007299270073,
      "grad_norm": 0.15442553162574768,
      "learning_rate": 0.0002,
      "loss": 1.5324,
      "step": 1836
    },
    {
      "epoch": 0.6704379562043795,
      "grad_norm": 0.15190666913986206,
      "learning_rate": 0.0002,
      "loss": 1.526,
      "step": 1837
    },
    {
      "epoch": 0.6708029197080292,
      "grad_norm": 0.15609730780124664,
      "learning_rate": 0.0002,
      "loss": 1.4901,
      "step": 1838
    },
    {
      "epoch": 0.6711678832116789,
      "grad_norm": 0.1590009331703186,
      "learning_rate": 0.0002,
      "loss": 1.565,
      "step": 1839
    },
    {
      "epoch": 0.6715328467153284,
      "grad_norm": 0.1463695466518402,
      "learning_rate": 0.0002,
      "loss": 1.493,
      "step": 1840
    },
    {
      "epoch": 0.6718978102189781,
      "grad_norm": 0.14139153063297272,
      "learning_rate": 0.0002,
      "loss": 1.4395,
      "step": 1841
    },
    {
      "epoch": 0.6722627737226278,
      "grad_norm": 0.1511109620332718,
      "learning_rate": 0.0002,
      "loss": 1.4995,
      "step": 1842
    },
    {
      "epoch": 0.6726277372262773,
      "grad_norm": 0.1524236500263214,
      "learning_rate": 0.0002,
      "loss": 1.5453,
      "step": 1843
    },
    {
      "epoch": 0.672992700729927,
      "grad_norm": 0.15084457397460938,
      "learning_rate": 0.0002,
      "loss": 1.3966,
      "step": 1844
    },
    {
      "epoch": 0.6733576642335767,
      "grad_norm": 0.14927984774112701,
      "learning_rate": 0.0002,
      "loss": 1.5465,
      "step": 1845
    },
    {
      "epoch": 0.6737226277372262,
      "grad_norm": 0.16083863377571106,
      "learning_rate": 0.0002,
      "loss": 1.5426,
      "step": 1846
    },
    {
      "epoch": 0.6740875912408759,
      "grad_norm": 0.14679522812366486,
      "learning_rate": 0.0002,
      "loss": 1.4611,
      "step": 1847
    },
    {
      "epoch": 0.6744525547445256,
      "grad_norm": 0.14517006278038025,
      "learning_rate": 0.0002,
      "loss": 1.512,
      "step": 1848
    },
    {
      "epoch": 0.6748175182481752,
      "grad_norm": 0.14649902284145355,
      "learning_rate": 0.0002,
      "loss": 1.5091,
      "step": 1849
    },
    {
      "epoch": 0.6751824817518248,
      "grad_norm": 0.14790134131908417,
      "learning_rate": 0.0002,
      "loss": 1.5882,
      "step": 1850
    },
    {
      "epoch": 0.6755474452554745,
      "grad_norm": 0.14228348433971405,
      "learning_rate": 0.0002,
      "loss": 1.4,
      "step": 1851
    },
    {
      "epoch": 0.6759124087591241,
      "grad_norm": 0.14529410004615784,
      "learning_rate": 0.0002,
      "loss": 1.4504,
      "step": 1852
    },
    {
      "epoch": 0.6762773722627737,
      "grad_norm": 0.1551772654056549,
      "learning_rate": 0.0002,
      "loss": 1.5185,
      "step": 1853
    },
    {
      "epoch": 0.6766423357664234,
      "grad_norm": 0.16397880017757416,
      "learning_rate": 0.0002,
      "loss": 1.5634,
      "step": 1854
    },
    {
      "epoch": 0.677007299270073,
      "grad_norm": 0.14806635677814484,
      "learning_rate": 0.0002,
      "loss": 1.5011,
      "step": 1855
    },
    {
      "epoch": 0.6773722627737226,
      "grad_norm": 0.15624654293060303,
      "learning_rate": 0.0002,
      "loss": 1.6108,
      "step": 1856
    },
    {
      "epoch": 0.6777372262773723,
      "grad_norm": 0.14879940450191498,
      "learning_rate": 0.0002,
      "loss": 1.5987,
      "step": 1857
    },
    {
      "epoch": 0.6781021897810219,
      "grad_norm": 0.15370270609855652,
      "learning_rate": 0.0002,
      "loss": 1.4818,
      "step": 1858
    },
    {
      "epoch": 0.6784671532846716,
      "grad_norm": 0.14946438372135162,
      "learning_rate": 0.0002,
      "loss": 1.4979,
      "step": 1859
    },
    {
      "epoch": 0.6788321167883211,
      "grad_norm": 0.15276986360549927,
      "learning_rate": 0.0002,
      "loss": 1.5209,
      "step": 1860
    },
    {
      "epoch": 0.6791970802919708,
      "grad_norm": 0.1530311405658722,
      "learning_rate": 0.0002,
      "loss": 1.5158,
      "step": 1861
    },
    {
      "epoch": 0.6795620437956205,
      "grad_norm": 0.1619669497013092,
      "learning_rate": 0.0002,
      "loss": 1.5566,
      "step": 1862
    },
    {
      "epoch": 0.67992700729927,
      "grad_norm": 0.14818932116031647,
      "learning_rate": 0.0002,
      "loss": 1.4313,
      "step": 1863
    },
    {
      "epoch": 0.6802919708029197,
      "grad_norm": 0.14641021192073822,
      "learning_rate": 0.0002,
      "loss": 1.5544,
      "step": 1864
    },
    {
      "epoch": 0.6806569343065694,
      "grad_norm": 0.15359161794185638,
      "learning_rate": 0.0002,
      "loss": 1.5699,
      "step": 1865
    },
    {
      "epoch": 0.6810218978102189,
      "grad_norm": 0.15337029099464417,
      "learning_rate": 0.0002,
      "loss": 1.5994,
      "step": 1866
    },
    {
      "epoch": 0.6813868613138686,
      "grad_norm": 0.15055625140666962,
      "learning_rate": 0.0002,
      "loss": 1.4525,
      "step": 1867
    },
    {
      "epoch": 0.6817518248175183,
      "grad_norm": 0.15047842264175415,
      "learning_rate": 0.0002,
      "loss": 1.5224,
      "step": 1868
    },
    {
      "epoch": 0.6821167883211678,
      "grad_norm": 0.1510978490114212,
      "learning_rate": 0.0002,
      "loss": 1.5115,
      "step": 1869
    },
    {
      "epoch": 0.6824817518248175,
      "grad_norm": 0.1455436646938324,
      "learning_rate": 0.0002,
      "loss": 1.4141,
      "step": 1870
    },
    {
      "epoch": 0.6828467153284672,
      "grad_norm": 0.15870904922485352,
      "learning_rate": 0.0002,
      "loss": 1.5271,
      "step": 1871
    },
    {
      "epoch": 0.6832116788321168,
      "grad_norm": 0.14733049273490906,
      "learning_rate": 0.0002,
      "loss": 1.5306,
      "step": 1872
    },
    {
      "epoch": 0.6835766423357664,
      "grad_norm": 0.1565597653388977,
      "learning_rate": 0.0002,
      "loss": 1.5382,
      "step": 1873
    },
    {
      "epoch": 0.6839416058394161,
      "grad_norm": 0.15600650012493134,
      "learning_rate": 0.0002,
      "loss": 1.6217,
      "step": 1874
    },
    {
      "epoch": 0.6843065693430657,
      "grad_norm": 0.14813275635242462,
      "learning_rate": 0.0002,
      "loss": 1.4364,
      "step": 1875
    },
    {
      "epoch": 0.6846715328467153,
      "grad_norm": 0.1566518247127533,
      "learning_rate": 0.0002,
      "loss": 1.4793,
      "step": 1876
    },
    {
      "epoch": 0.685036496350365,
      "grad_norm": 0.1495555192232132,
      "learning_rate": 0.0002,
      "loss": 1.5326,
      "step": 1877
    },
    {
      "epoch": 0.6854014598540146,
      "grad_norm": 0.1485525518655777,
      "learning_rate": 0.0002,
      "loss": 1.4397,
      "step": 1878
    },
    {
      "epoch": 0.6857664233576642,
      "grad_norm": 0.15515998005867004,
      "learning_rate": 0.0002,
      "loss": 1.503,
      "step": 1879
    },
    {
      "epoch": 0.6861313868613139,
      "grad_norm": 0.1579638421535492,
      "learning_rate": 0.0002,
      "loss": 1.5643,
      "step": 1880
    },
    {
      "epoch": 0.6864963503649635,
      "grad_norm": 0.15135124325752258,
      "learning_rate": 0.0002,
      "loss": 1.5035,
      "step": 1881
    },
    {
      "epoch": 0.6868613138686132,
      "grad_norm": 0.15166279673576355,
      "learning_rate": 0.0002,
      "loss": 1.4941,
      "step": 1882
    },
    {
      "epoch": 0.6872262773722628,
      "grad_norm": 0.14846749603748322,
      "learning_rate": 0.0002,
      "loss": 1.5385,
      "step": 1883
    },
    {
      "epoch": 0.6875912408759124,
      "grad_norm": 0.14371734857559204,
      "learning_rate": 0.0002,
      "loss": 1.4454,
      "step": 1884
    },
    {
      "epoch": 0.6879562043795621,
      "grad_norm": 0.15257611870765686,
      "learning_rate": 0.0002,
      "loss": 1.5451,
      "step": 1885
    },
    {
      "epoch": 0.6883211678832116,
      "grad_norm": 0.1596313714981079,
      "learning_rate": 0.0002,
      "loss": 1.5319,
      "step": 1886
    },
    {
      "epoch": 0.6886861313868613,
      "grad_norm": 0.152935191988945,
      "learning_rate": 0.0002,
      "loss": 1.5433,
      "step": 1887
    },
    {
      "epoch": 0.689051094890511,
      "grad_norm": 0.14259544014930725,
      "learning_rate": 0.0002,
      "loss": 1.4227,
      "step": 1888
    },
    {
      "epoch": 0.6894160583941605,
      "grad_norm": 0.15518195927143097,
      "learning_rate": 0.0002,
      "loss": 1.5462,
      "step": 1889
    },
    {
      "epoch": 0.6897810218978102,
      "grad_norm": 0.1480821818113327,
      "learning_rate": 0.0002,
      "loss": 1.5729,
      "step": 1890
    },
    {
      "epoch": 0.6901459854014599,
      "grad_norm": 0.1428493708372116,
      "learning_rate": 0.0002,
      "loss": 1.541,
      "step": 1891
    },
    {
      "epoch": 0.6905109489051094,
      "grad_norm": 0.14981120824813843,
      "learning_rate": 0.0002,
      "loss": 1.5431,
      "step": 1892
    },
    {
      "epoch": 0.6908759124087591,
      "grad_norm": 0.14766275882720947,
      "learning_rate": 0.0002,
      "loss": 1.5148,
      "step": 1893
    },
    {
      "epoch": 0.6912408759124088,
      "grad_norm": 0.16830995678901672,
      "learning_rate": 0.0002,
      "loss": 1.5341,
      "step": 1894
    },
    {
      "epoch": 0.6916058394160584,
      "grad_norm": 0.1515159010887146,
      "learning_rate": 0.0002,
      "loss": 1.4584,
      "step": 1895
    },
    {
      "epoch": 0.691970802919708,
      "grad_norm": 0.15362057089805603,
      "learning_rate": 0.0002,
      "loss": 1.5296,
      "step": 1896
    },
    {
      "epoch": 0.6923357664233577,
      "grad_norm": 0.1526845246553421,
      "learning_rate": 0.0002,
      "loss": 1.5855,
      "step": 1897
    },
    {
      "epoch": 0.6927007299270073,
      "grad_norm": 0.1584462821483612,
      "learning_rate": 0.0002,
      "loss": 1.5646,
      "step": 1898
    },
    {
      "epoch": 0.6930656934306569,
      "grad_norm": 0.15011870861053467,
      "learning_rate": 0.0002,
      "loss": 1.4926,
      "step": 1899
    },
    {
      "epoch": 0.6934306569343066,
      "grad_norm": 0.14715242385864258,
      "learning_rate": 0.0002,
      "loss": 1.446,
      "step": 1900
    },
    {
      "epoch": 0.6937956204379562,
      "grad_norm": 0.14632190763950348,
      "learning_rate": 0.0002,
      "loss": 1.5045,
      "step": 1901
    },
    {
      "epoch": 0.6941605839416058,
      "grad_norm": 0.1650785207748413,
      "learning_rate": 0.0002,
      "loss": 1.5813,
      "step": 1902
    },
    {
      "epoch": 0.6945255474452555,
      "grad_norm": 0.14811666309833527,
      "learning_rate": 0.0002,
      "loss": 1.4747,
      "step": 1903
    },
    {
      "epoch": 0.6948905109489051,
      "grad_norm": 0.15769453346729279,
      "learning_rate": 0.0002,
      "loss": 1.5342,
      "step": 1904
    },
    {
      "epoch": 0.6952554744525548,
      "grad_norm": 0.1487896889448166,
      "learning_rate": 0.0002,
      "loss": 1.5233,
      "step": 1905
    },
    {
      "epoch": 0.6956204379562044,
      "grad_norm": 0.15262185037136078,
      "learning_rate": 0.0002,
      "loss": 1.4867,
      "step": 1906
    },
    {
      "epoch": 0.695985401459854,
      "grad_norm": 0.15086489915847778,
      "learning_rate": 0.0002,
      "loss": 1.5122,
      "step": 1907
    },
    {
      "epoch": 0.6963503649635037,
      "grad_norm": 0.15295080840587616,
      "learning_rate": 0.0002,
      "loss": 1.4471,
      "step": 1908
    },
    {
      "epoch": 0.6967153284671533,
      "grad_norm": 0.15165024995803833,
      "learning_rate": 0.0002,
      "loss": 1.5015,
      "step": 1909
    },
    {
      "epoch": 0.6970802919708029,
      "grad_norm": 0.15193124115467072,
      "learning_rate": 0.0002,
      "loss": 1.5116,
      "step": 1910
    },
    {
      "epoch": 0.6974452554744526,
      "grad_norm": 0.1551562398672104,
      "learning_rate": 0.0002,
      "loss": 1.5144,
      "step": 1911
    },
    {
      "epoch": 0.6978102189781021,
      "grad_norm": 0.14902807772159576,
      "learning_rate": 0.0002,
      "loss": 1.5174,
      "step": 1912
    },
    {
      "epoch": 0.6981751824817518,
      "grad_norm": 0.15042804181575775,
      "learning_rate": 0.0002,
      "loss": 1.488,
      "step": 1913
    },
    {
      "epoch": 0.6985401459854015,
      "grad_norm": 0.14792928099632263,
      "learning_rate": 0.0002,
      "loss": 1.4619,
      "step": 1914
    },
    {
      "epoch": 0.698905109489051,
      "grad_norm": 0.15113969147205353,
      "learning_rate": 0.0002,
      "loss": 1.4751,
      "step": 1915
    },
    {
      "epoch": 0.6992700729927007,
      "grad_norm": 0.1505683809518814,
      "learning_rate": 0.0002,
      "loss": 1.5101,
      "step": 1916
    },
    {
      "epoch": 0.6996350364963504,
      "grad_norm": 0.14781491458415985,
      "learning_rate": 0.0002,
      "loss": 1.4804,
      "step": 1917
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.1675102710723877,
      "learning_rate": 0.0002,
      "loss": 1.5354,
      "step": 1918
    },
    {
      "epoch": 0.7003649635036496,
      "grad_norm": 0.16165867447853088,
      "learning_rate": 0.0002,
      "loss": 1.5868,
      "step": 1919
    },
    {
      "epoch": 0.7007299270072993,
      "grad_norm": 0.15138152241706848,
      "learning_rate": 0.0002,
      "loss": 1.4787,
      "step": 1920
    },
    {
      "epoch": 0.7010948905109489,
      "grad_norm": 0.1521741896867752,
      "learning_rate": 0.0002,
      "loss": 1.4867,
      "step": 1921
    },
    {
      "epoch": 0.7014598540145985,
      "grad_norm": 0.1544054001569748,
      "learning_rate": 0.0002,
      "loss": 1.5269,
      "step": 1922
    },
    {
      "epoch": 0.7018248175182482,
      "grad_norm": 0.16468971967697144,
      "learning_rate": 0.0002,
      "loss": 1.4717,
      "step": 1923
    },
    {
      "epoch": 0.7021897810218978,
      "grad_norm": 0.1589461714029312,
      "learning_rate": 0.0002,
      "loss": 1.5103,
      "step": 1924
    },
    {
      "epoch": 0.7025547445255474,
      "grad_norm": 0.16189616918563843,
      "learning_rate": 0.0002,
      "loss": 1.5556,
      "step": 1925
    },
    {
      "epoch": 0.7029197080291971,
      "grad_norm": 0.14665786921977997,
      "learning_rate": 0.0002,
      "loss": 1.4941,
      "step": 1926
    },
    {
      "epoch": 0.7032846715328467,
      "grad_norm": 0.17471347749233246,
      "learning_rate": 0.0002,
      "loss": 1.569,
      "step": 1927
    },
    {
      "epoch": 0.7036496350364964,
      "grad_norm": 0.14758935570716858,
      "learning_rate": 0.0002,
      "loss": 1.5647,
      "step": 1928
    },
    {
      "epoch": 0.704014598540146,
      "grad_norm": 0.1592218428850174,
      "learning_rate": 0.0002,
      "loss": 1.4824,
      "step": 1929
    },
    {
      "epoch": 0.7043795620437956,
      "grad_norm": 0.16390585899353027,
      "learning_rate": 0.0002,
      "loss": 1.5937,
      "step": 1930
    },
    {
      "epoch": 0.7047445255474453,
      "grad_norm": 0.15607646107673645,
      "learning_rate": 0.0002,
      "loss": 1.5687,
      "step": 1931
    },
    {
      "epoch": 0.7051094890510949,
      "grad_norm": 0.16218575835227966,
      "learning_rate": 0.0002,
      "loss": 1.5518,
      "step": 1932
    },
    {
      "epoch": 0.7054744525547445,
      "grad_norm": 0.1678551584482193,
      "learning_rate": 0.0002,
      "loss": 1.6538,
      "step": 1933
    },
    {
      "epoch": 0.7058394160583942,
      "grad_norm": 0.145615354180336,
      "learning_rate": 0.0002,
      "loss": 1.4595,
      "step": 1934
    },
    {
      "epoch": 0.7062043795620438,
      "grad_norm": 0.14870882034301758,
      "learning_rate": 0.0002,
      "loss": 1.5682,
      "step": 1935
    },
    {
      "epoch": 0.7065693430656934,
      "grad_norm": 0.1540498286485672,
      "learning_rate": 0.0002,
      "loss": 1.4819,
      "step": 1936
    },
    {
      "epoch": 0.7069343065693431,
      "grad_norm": 0.15346108376979828,
      "learning_rate": 0.0002,
      "loss": 1.4491,
      "step": 1937
    },
    {
      "epoch": 0.7072992700729926,
      "grad_norm": 0.15253335237503052,
      "learning_rate": 0.0002,
      "loss": 1.4996,
      "step": 1938
    },
    {
      "epoch": 0.7076642335766423,
      "grad_norm": 0.15672935545444489,
      "learning_rate": 0.0002,
      "loss": 1.4893,
      "step": 1939
    },
    {
      "epoch": 0.708029197080292,
      "grad_norm": 0.15119130909442902,
      "learning_rate": 0.0002,
      "loss": 1.4889,
      "step": 1940
    },
    {
      "epoch": 0.7083941605839416,
      "grad_norm": 0.15770505368709564,
      "learning_rate": 0.0002,
      "loss": 1.4547,
      "step": 1941
    },
    {
      "epoch": 0.7087591240875912,
      "grad_norm": 0.148014098405838,
      "learning_rate": 0.0002,
      "loss": 1.428,
      "step": 1942
    },
    {
      "epoch": 0.7091240875912409,
      "grad_norm": 0.1513221114873886,
      "learning_rate": 0.0002,
      "loss": 1.4327,
      "step": 1943
    },
    {
      "epoch": 0.7094890510948905,
      "grad_norm": 0.15678037703037262,
      "learning_rate": 0.0002,
      "loss": 1.4999,
      "step": 1944
    },
    {
      "epoch": 0.7098540145985401,
      "grad_norm": 0.152816042304039,
      "learning_rate": 0.0002,
      "loss": 1.4725,
      "step": 1945
    },
    {
      "epoch": 0.7102189781021898,
      "grad_norm": 0.15888437628746033,
      "learning_rate": 0.0002,
      "loss": 1.5276,
      "step": 1946
    },
    {
      "epoch": 0.7105839416058394,
      "grad_norm": 0.15351785719394684,
      "learning_rate": 0.0002,
      "loss": 1.4542,
      "step": 1947
    },
    {
      "epoch": 0.710948905109489,
      "grad_norm": 0.16152866184711456,
      "learning_rate": 0.0002,
      "loss": 1.5276,
      "step": 1948
    },
    {
      "epoch": 0.7113138686131387,
      "grad_norm": 0.16255448758602142,
      "learning_rate": 0.0002,
      "loss": 1.5787,
      "step": 1949
    },
    {
      "epoch": 0.7116788321167883,
      "grad_norm": 0.15363222360610962,
      "learning_rate": 0.0002,
      "loss": 1.5192,
      "step": 1950
    },
    {
      "epoch": 0.712043795620438,
      "grad_norm": 0.16223151981830597,
      "learning_rate": 0.0002,
      "loss": 1.5044,
      "step": 1951
    },
    {
      "epoch": 0.7124087591240876,
      "grad_norm": 0.16266924142837524,
      "learning_rate": 0.0002,
      "loss": 1.4878,
      "step": 1952
    },
    {
      "epoch": 0.7127737226277372,
      "grad_norm": 0.15666839480400085,
      "learning_rate": 0.0002,
      "loss": 1.4959,
      "step": 1953
    },
    {
      "epoch": 0.7131386861313869,
      "grad_norm": 0.15953722596168518,
      "learning_rate": 0.0002,
      "loss": 1.4749,
      "step": 1954
    },
    {
      "epoch": 0.7135036496350365,
      "grad_norm": 0.15467338263988495,
      "learning_rate": 0.0002,
      "loss": 1.468,
      "step": 1955
    },
    {
      "epoch": 0.7138686131386861,
      "grad_norm": 0.1527474820613861,
      "learning_rate": 0.0002,
      "loss": 1.5114,
      "step": 1956
    },
    {
      "epoch": 0.7142335766423358,
      "grad_norm": 0.1580076962709427,
      "learning_rate": 0.0002,
      "loss": 1.4654,
      "step": 1957
    },
    {
      "epoch": 0.7145985401459855,
      "grad_norm": 0.15431088209152222,
      "learning_rate": 0.0002,
      "loss": 1.5505,
      "step": 1958
    },
    {
      "epoch": 0.714963503649635,
      "grad_norm": 0.1590520441532135,
      "learning_rate": 0.0002,
      "loss": 1.5257,
      "step": 1959
    },
    {
      "epoch": 0.7153284671532847,
      "grad_norm": 0.16174906492233276,
      "learning_rate": 0.0002,
      "loss": 1.5615,
      "step": 1960
    },
    {
      "epoch": 0.7156934306569344,
      "grad_norm": 0.1591993272304535,
      "learning_rate": 0.0002,
      "loss": 1.509,
      "step": 1961
    },
    {
      "epoch": 0.7160583941605839,
      "grad_norm": 0.18314582109451294,
      "learning_rate": 0.0002,
      "loss": 1.5262,
      "step": 1962
    },
    {
      "epoch": 0.7164233576642336,
      "grad_norm": 0.15491558611392975,
      "learning_rate": 0.0002,
      "loss": 1.5216,
      "step": 1963
    },
    {
      "epoch": 0.7167883211678832,
      "grad_norm": 0.15493892133235931,
      "learning_rate": 0.0002,
      "loss": 1.4308,
      "step": 1964
    },
    {
      "epoch": 0.7171532846715328,
      "grad_norm": 0.15838085114955902,
      "learning_rate": 0.0002,
      "loss": 1.5728,
      "step": 1965
    },
    {
      "epoch": 0.7175182481751825,
      "grad_norm": 0.14755339920520782,
      "learning_rate": 0.0002,
      "loss": 1.4792,
      "step": 1966
    },
    {
      "epoch": 0.7178832116788321,
      "grad_norm": 0.16276602447032928,
      "learning_rate": 0.0002,
      "loss": 1.5434,
      "step": 1967
    },
    {
      "epoch": 0.7182481751824817,
      "grad_norm": 0.1531320959329605,
      "learning_rate": 0.0002,
      "loss": 1.517,
      "step": 1968
    },
    {
      "epoch": 0.7186131386861314,
      "grad_norm": 0.15953712165355682,
      "learning_rate": 0.0002,
      "loss": 1.5592,
      "step": 1969
    },
    {
      "epoch": 0.718978102189781,
      "grad_norm": 0.15087148547172546,
      "learning_rate": 0.0002,
      "loss": 1.5197,
      "step": 1970
    },
    {
      "epoch": 0.7193430656934306,
      "grad_norm": 0.1537655144929886,
      "learning_rate": 0.0002,
      "loss": 1.4785,
      "step": 1971
    },
    {
      "epoch": 0.7197080291970803,
      "grad_norm": 0.14925988018512726,
      "learning_rate": 0.0002,
      "loss": 1.4761,
      "step": 1972
    },
    {
      "epoch": 0.7200729927007299,
      "grad_norm": 0.14662739634513855,
      "learning_rate": 0.0002,
      "loss": 1.516,
      "step": 1973
    },
    {
      "epoch": 0.7204379562043796,
      "grad_norm": 0.1537463515996933,
      "learning_rate": 0.0002,
      "loss": 1.4628,
      "step": 1974
    },
    {
      "epoch": 0.7208029197080292,
      "grad_norm": 0.15679863095283508,
      "learning_rate": 0.0002,
      "loss": 1.5196,
      "step": 1975
    },
    {
      "epoch": 0.7211678832116788,
      "grad_norm": 0.1570269763469696,
      "learning_rate": 0.0002,
      "loss": 1.5319,
      "step": 1976
    },
    {
      "epoch": 0.7215328467153285,
      "grad_norm": 0.15065446496009827,
      "learning_rate": 0.0002,
      "loss": 1.5693,
      "step": 1977
    },
    {
      "epoch": 0.7218978102189781,
      "grad_norm": 0.15408562123775482,
      "learning_rate": 0.0002,
      "loss": 1.5165,
      "step": 1978
    },
    {
      "epoch": 0.7222627737226277,
      "grad_norm": 0.1499413251876831,
      "learning_rate": 0.0002,
      "loss": 1.5362,
      "step": 1979
    },
    {
      "epoch": 0.7226277372262774,
      "grad_norm": 0.1497785747051239,
      "learning_rate": 0.0002,
      "loss": 1.5047,
      "step": 1980
    },
    {
      "epoch": 0.722992700729927,
      "grad_norm": 0.15123912692070007,
      "learning_rate": 0.0002,
      "loss": 1.5506,
      "step": 1981
    },
    {
      "epoch": 0.7233576642335766,
      "grad_norm": 0.14549827575683594,
      "learning_rate": 0.0002,
      "loss": 1.5389,
      "step": 1982
    },
    {
      "epoch": 0.7237226277372263,
      "grad_norm": 0.15464343130588531,
      "learning_rate": 0.0002,
      "loss": 1.546,
      "step": 1983
    },
    {
      "epoch": 0.724087591240876,
      "grad_norm": 0.15633466839790344,
      "learning_rate": 0.0002,
      "loss": 1.5888,
      "step": 1984
    },
    {
      "epoch": 0.7244525547445255,
      "grad_norm": 0.15250952541828156,
      "learning_rate": 0.0002,
      "loss": 1.493,
      "step": 1985
    },
    {
      "epoch": 0.7248175182481752,
      "grad_norm": 0.1480371057987213,
      "learning_rate": 0.0002,
      "loss": 1.483,
      "step": 1986
    },
    {
      "epoch": 0.7251824817518249,
      "grad_norm": 0.15016423165798187,
      "learning_rate": 0.0002,
      "loss": 1.4422,
      "step": 1987
    },
    {
      "epoch": 0.7255474452554744,
      "grad_norm": 0.17016582190990448,
      "learning_rate": 0.0002,
      "loss": 1.5573,
      "step": 1988
    },
    {
      "epoch": 0.7259124087591241,
      "grad_norm": 0.1682599037885666,
      "learning_rate": 0.0002,
      "loss": 1.5592,
      "step": 1989
    },
    {
      "epoch": 0.7262773722627737,
      "grad_norm": 0.1628475934267044,
      "learning_rate": 0.0002,
      "loss": 1.5174,
      "step": 1990
    },
    {
      "epoch": 0.7266423357664233,
      "grad_norm": 0.15951725840568542,
      "learning_rate": 0.0002,
      "loss": 1.5458,
      "step": 1991
    },
    {
      "epoch": 0.727007299270073,
      "grad_norm": 0.1524718850851059,
      "learning_rate": 0.0002,
      "loss": 1.5077,
      "step": 1992
    },
    {
      "epoch": 0.7273722627737226,
      "grad_norm": 0.16679786145687103,
      "learning_rate": 0.0002,
      "loss": 1.4241,
      "step": 1993
    },
    {
      "epoch": 0.7277372262773723,
      "grad_norm": 0.15216270089149475,
      "learning_rate": 0.0002,
      "loss": 1.4556,
      "step": 1994
    },
    {
      "epoch": 0.7281021897810219,
      "grad_norm": 0.17707909643650055,
      "learning_rate": 0.0002,
      "loss": 1.5197,
      "step": 1995
    },
    {
      "epoch": 0.7284671532846715,
      "grad_norm": 0.161293163895607,
      "learning_rate": 0.0002,
      "loss": 1.4955,
      "step": 1996
    },
    {
      "epoch": 0.7288321167883212,
      "grad_norm": 0.16129866242408752,
      "learning_rate": 0.0002,
      "loss": 1.45,
      "step": 1997
    },
    {
      "epoch": 0.7291970802919708,
      "grad_norm": 0.17047792673110962,
      "learning_rate": 0.0002,
      "loss": 1.5567,
      "step": 1998
    },
    {
      "epoch": 0.7295620437956204,
      "grad_norm": 0.16707663238048553,
      "learning_rate": 0.0002,
      "loss": 1.5639,
      "step": 1999
    },
    {
      "epoch": 0.7299270072992701,
      "grad_norm": 0.15770679712295532,
      "learning_rate": 0.0002,
      "loss": 1.4203,
      "step": 2000
    },
    {
      "epoch": 0.7302919708029197,
      "grad_norm": 0.14924786984920502,
      "learning_rate": 0.0002,
      "loss": 1.5385,
      "step": 2001
    },
    {
      "epoch": 0.7306569343065693,
      "grad_norm": 0.16455644369125366,
      "learning_rate": 0.0002,
      "loss": 1.5274,
      "step": 2002
    },
    {
      "epoch": 0.731021897810219,
      "grad_norm": 0.15551413595676422,
      "learning_rate": 0.0002,
      "loss": 1.5414,
      "step": 2003
    },
    {
      "epoch": 0.7313868613138687,
      "grad_norm": 0.16498327255249023,
      "learning_rate": 0.0002,
      "loss": 1.4279,
      "step": 2004
    },
    {
      "epoch": 0.7317518248175182,
      "grad_norm": 0.1514066457748413,
      "learning_rate": 0.0002,
      "loss": 1.4715,
      "step": 2005
    },
    {
      "epoch": 0.7321167883211679,
      "grad_norm": 0.15139488875865936,
      "learning_rate": 0.0002,
      "loss": 1.5113,
      "step": 2006
    },
    {
      "epoch": 0.7324817518248176,
      "grad_norm": 0.15868563950061798,
      "learning_rate": 0.0002,
      "loss": 1.5579,
      "step": 2007
    },
    {
      "epoch": 0.7328467153284671,
      "grad_norm": 0.1579631119966507,
      "learning_rate": 0.0002,
      "loss": 1.6354,
      "step": 2008
    },
    {
      "epoch": 0.7332116788321168,
      "grad_norm": 0.1547684222459793,
      "learning_rate": 0.0002,
      "loss": 1.5634,
      "step": 2009
    },
    {
      "epoch": 0.7335766423357665,
      "grad_norm": 0.15444660186767578,
      "learning_rate": 0.0002,
      "loss": 1.5068,
      "step": 2010
    },
    {
      "epoch": 0.733941605839416,
      "grad_norm": 0.15940849483013153,
      "learning_rate": 0.0002,
      "loss": 1.5862,
      "step": 2011
    },
    {
      "epoch": 0.7343065693430657,
      "grad_norm": 0.15142439305782318,
      "learning_rate": 0.0002,
      "loss": 1.4658,
      "step": 2012
    },
    {
      "epoch": 0.7346715328467154,
      "grad_norm": 0.148756206035614,
      "learning_rate": 0.0002,
      "loss": 1.5046,
      "step": 2013
    },
    {
      "epoch": 0.7350364963503649,
      "grad_norm": 0.15176013112068176,
      "learning_rate": 0.0002,
      "loss": 1.5357,
      "step": 2014
    },
    {
      "epoch": 0.7354014598540146,
      "grad_norm": 0.15212948620319366,
      "learning_rate": 0.0002,
      "loss": 1.5446,
      "step": 2015
    },
    {
      "epoch": 0.7357664233576642,
      "grad_norm": 0.1598239243030548,
      "learning_rate": 0.0002,
      "loss": 1.5532,
      "step": 2016
    },
    {
      "epoch": 0.7361313868613139,
      "grad_norm": 0.15277917683124542,
      "learning_rate": 0.0002,
      "loss": 1.4696,
      "step": 2017
    },
    {
      "epoch": 0.7364963503649635,
      "grad_norm": 0.15416835248470306,
      "learning_rate": 0.0002,
      "loss": 1.6069,
      "step": 2018
    },
    {
      "epoch": 0.7368613138686131,
      "grad_norm": 0.14756669104099274,
      "learning_rate": 0.0002,
      "loss": 1.4376,
      "step": 2019
    },
    {
      "epoch": 0.7372262773722628,
      "grad_norm": 0.15101653337478638,
      "learning_rate": 0.0002,
      "loss": 1.5507,
      "step": 2020
    },
    {
      "epoch": 0.7375912408759124,
      "grad_norm": 0.1560095101594925,
      "learning_rate": 0.0002,
      "loss": 1.5604,
      "step": 2021
    },
    {
      "epoch": 0.737956204379562,
      "grad_norm": 0.15721265971660614,
      "learning_rate": 0.0002,
      "loss": 1.5264,
      "step": 2022
    },
    {
      "epoch": 0.7383211678832117,
      "grad_norm": 0.15381625294685364,
      "learning_rate": 0.0002,
      "loss": 1.4686,
      "step": 2023
    },
    {
      "epoch": 0.7386861313868613,
      "grad_norm": 0.1554683893918991,
      "learning_rate": 0.0002,
      "loss": 1.5333,
      "step": 2024
    },
    {
      "epoch": 0.7390510948905109,
      "grad_norm": 0.1516694575548172,
      "learning_rate": 0.0002,
      "loss": 1.4543,
      "step": 2025
    },
    {
      "epoch": 0.7394160583941606,
      "grad_norm": 0.1599535197019577,
      "learning_rate": 0.0002,
      "loss": 1.4765,
      "step": 2026
    },
    {
      "epoch": 0.7397810218978103,
      "grad_norm": 0.15289005637168884,
      "learning_rate": 0.0002,
      "loss": 1.4554,
      "step": 2027
    },
    {
      "epoch": 0.7401459854014598,
      "grad_norm": 0.1491389125585556,
      "learning_rate": 0.0002,
      "loss": 1.4449,
      "step": 2028
    },
    {
      "epoch": 0.7405109489051095,
      "grad_norm": 0.1500927060842514,
      "learning_rate": 0.0002,
      "loss": 1.4992,
      "step": 2029
    },
    {
      "epoch": 0.7408759124087592,
      "grad_norm": 0.1597473919391632,
      "learning_rate": 0.0002,
      "loss": 1.5136,
      "step": 2030
    },
    {
      "epoch": 0.7412408759124087,
      "grad_norm": 0.1558518409729004,
      "learning_rate": 0.0002,
      "loss": 1.4743,
      "step": 2031
    },
    {
      "epoch": 0.7416058394160584,
      "grad_norm": 0.15535713732242584,
      "learning_rate": 0.0002,
      "loss": 1.4414,
      "step": 2032
    },
    {
      "epoch": 0.7419708029197081,
      "grad_norm": 0.18021608889102936,
      "learning_rate": 0.0002,
      "loss": 1.5626,
      "step": 2033
    },
    {
      "epoch": 0.7423357664233576,
      "grad_norm": 0.16699282824993134,
      "learning_rate": 0.0002,
      "loss": 1.5014,
      "step": 2034
    },
    {
      "epoch": 0.7427007299270073,
      "grad_norm": 0.17014804482460022,
      "learning_rate": 0.0002,
      "loss": 1.5508,
      "step": 2035
    },
    {
      "epoch": 0.743065693430657,
      "grad_norm": 0.1647229790687561,
      "learning_rate": 0.0002,
      "loss": 1.5548,
      "step": 2036
    },
    {
      "epoch": 0.7434306569343065,
      "grad_norm": 0.18091802299022675,
      "learning_rate": 0.0002,
      "loss": 1.5231,
      "step": 2037
    },
    {
      "epoch": 0.7437956204379562,
      "grad_norm": 0.15827687084674835,
      "learning_rate": 0.0002,
      "loss": 1.5628,
      "step": 2038
    },
    {
      "epoch": 0.7441605839416059,
      "grad_norm": 0.17712585628032684,
      "learning_rate": 0.0002,
      "loss": 1.4977,
      "step": 2039
    },
    {
      "epoch": 0.7445255474452555,
      "grad_norm": 0.1567835807800293,
      "learning_rate": 0.0002,
      "loss": 1.4758,
      "step": 2040
    },
    {
      "epoch": 0.7448905109489051,
      "grad_norm": 0.1817864328622818,
      "learning_rate": 0.0002,
      "loss": 1.5258,
      "step": 2041
    },
    {
      "epoch": 0.7452554744525547,
      "grad_norm": 0.16020436584949493,
      "learning_rate": 0.0002,
      "loss": 1.5245,
      "step": 2042
    },
    {
      "epoch": 0.7456204379562044,
      "grad_norm": 0.1563435196876526,
      "learning_rate": 0.0002,
      "loss": 1.5112,
      "step": 2043
    },
    {
      "epoch": 0.745985401459854,
      "grad_norm": 0.15963655710220337,
      "learning_rate": 0.0002,
      "loss": 1.4541,
      "step": 2044
    },
    {
      "epoch": 0.7463503649635036,
      "grad_norm": 0.15102618932724,
      "learning_rate": 0.0002,
      "loss": 1.462,
      "step": 2045
    },
    {
      "epoch": 0.7467153284671533,
      "grad_norm": 0.1702490746974945,
      "learning_rate": 0.0002,
      "loss": 1.4537,
      "step": 2046
    },
    {
      "epoch": 0.747080291970803,
      "grad_norm": 0.15937253832817078,
      "learning_rate": 0.0002,
      "loss": 1.4816,
      "step": 2047
    },
    {
      "epoch": 0.7474452554744525,
      "grad_norm": 0.1584261804819107,
      "learning_rate": 0.0002,
      "loss": 1.5428,
      "step": 2048
    },
    {
      "epoch": 0.7478102189781022,
      "grad_norm": 0.16192352771759033,
      "learning_rate": 0.0002,
      "loss": 1.4215,
      "step": 2049
    },
    {
      "epoch": 0.7481751824817519,
      "grad_norm": 0.15228067338466644,
      "learning_rate": 0.0002,
      "loss": 1.5046,
      "step": 2050
    },
    {
      "epoch": 0.7485401459854014,
      "grad_norm": 0.15824617445468903,
      "learning_rate": 0.0002,
      "loss": 1.4623,
      "step": 2051
    },
    {
      "epoch": 0.7489051094890511,
      "grad_norm": 0.1668606847524643,
      "learning_rate": 0.0002,
      "loss": 1.555,
      "step": 2052
    },
    {
      "epoch": 0.7492700729927008,
      "grad_norm": 0.15289060771465302,
      "learning_rate": 0.0002,
      "loss": 1.4643,
      "step": 2053
    },
    {
      "epoch": 0.7496350364963503,
      "grad_norm": 0.16122867166996002,
      "learning_rate": 0.0002,
      "loss": 1.4606,
      "step": 2054
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1544070839881897,
      "learning_rate": 0.0002,
      "loss": 1.5528,
      "step": 2055
    },
    {
      "epoch": 0.7503649635036497,
      "grad_norm": 0.1587553769350052,
      "learning_rate": 0.0002,
      "loss": 1.5313,
      "step": 2056
    },
    {
      "epoch": 0.7507299270072992,
      "grad_norm": 0.14852963387966156,
      "learning_rate": 0.0002,
      "loss": 1.4676,
      "step": 2057
    },
    {
      "epoch": 0.7510948905109489,
      "grad_norm": 0.14854425191879272,
      "learning_rate": 0.0002,
      "loss": 1.5481,
      "step": 2058
    },
    {
      "epoch": 0.7514598540145986,
      "grad_norm": 0.14560186862945557,
      "learning_rate": 0.0002,
      "loss": 1.4955,
      "step": 2059
    },
    {
      "epoch": 0.7518248175182481,
      "grad_norm": 0.1515301764011383,
      "learning_rate": 0.0002,
      "loss": 1.5195,
      "step": 2060
    },
    {
      "epoch": 0.7521897810218978,
      "grad_norm": 0.15525642037391663,
      "learning_rate": 0.0002,
      "loss": 1.4882,
      "step": 2061
    },
    {
      "epoch": 0.7525547445255475,
      "grad_norm": 0.15467096865177155,
      "learning_rate": 0.0002,
      "loss": 1.566,
      "step": 2062
    },
    {
      "epoch": 0.752919708029197,
      "grad_norm": 0.15532870590686798,
      "learning_rate": 0.0002,
      "loss": 1.5161,
      "step": 2063
    },
    {
      "epoch": 0.7532846715328467,
      "grad_norm": 0.15213552117347717,
      "learning_rate": 0.0002,
      "loss": 1.5074,
      "step": 2064
    },
    {
      "epoch": 0.7536496350364964,
      "grad_norm": 0.15144555270671844,
      "learning_rate": 0.0002,
      "loss": 1.5123,
      "step": 2065
    },
    {
      "epoch": 0.754014598540146,
      "grad_norm": 0.1552143096923828,
      "learning_rate": 0.0002,
      "loss": 1.4925,
      "step": 2066
    },
    {
      "epoch": 0.7543795620437956,
      "grad_norm": 0.1577780395746231,
      "learning_rate": 0.0002,
      "loss": 1.5786,
      "step": 2067
    },
    {
      "epoch": 0.7547445255474453,
      "grad_norm": 0.15597188472747803,
      "learning_rate": 0.0002,
      "loss": 1.5561,
      "step": 2068
    },
    {
      "epoch": 0.7551094890510949,
      "grad_norm": 0.15799134969711304,
      "learning_rate": 0.0002,
      "loss": 1.6,
      "step": 2069
    },
    {
      "epoch": 0.7554744525547445,
      "grad_norm": 0.1498728096485138,
      "learning_rate": 0.0002,
      "loss": 1.4622,
      "step": 2070
    },
    {
      "epoch": 0.7558394160583941,
      "grad_norm": 0.15287485718727112,
      "learning_rate": 0.0002,
      "loss": 1.5071,
      "step": 2071
    },
    {
      "epoch": 0.7562043795620438,
      "grad_norm": 0.15806007385253906,
      "learning_rate": 0.0002,
      "loss": 1.5209,
      "step": 2072
    },
    {
      "epoch": 0.7565693430656935,
      "grad_norm": 0.159692645072937,
      "learning_rate": 0.0002,
      "loss": 1.4779,
      "step": 2073
    },
    {
      "epoch": 0.756934306569343,
      "grad_norm": 0.16158275306224823,
      "learning_rate": 0.0002,
      "loss": 1.574,
      "step": 2074
    },
    {
      "epoch": 0.7572992700729927,
      "grad_norm": 0.1574568748474121,
      "learning_rate": 0.0002,
      "loss": 1.5396,
      "step": 2075
    },
    {
      "epoch": 0.7576642335766424,
      "grad_norm": 0.15079174935817719,
      "learning_rate": 0.0002,
      "loss": 1.5079,
      "step": 2076
    },
    {
      "epoch": 0.7580291970802919,
      "grad_norm": 0.1566835194826126,
      "learning_rate": 0.0002,
      "loss": 1.5753,
      "step": 2077
    },
    {
      "epoch": 0.7583941605839416,
      "grad_norm": 0.15195849537849426,
      "learning_rate": 0.0002,
      "loss": 1.4417,
      "step": 2078
    },
    {
      "epoch": 0.7587591240875913,
      "grad_norm": 0.1486731767654419,
      "learning_rate": 0.0002,
      "loss": 1.4866,
      "step": 2079
    },
    {
      "epoch": 0.7591240875912408,
      "grad_norm": 0.14894594252109528,
      "learning_rate": 0.0002,
      "loss": 1.4347,
      "step": 2080
    },
    {
      "epoch": 0.7594890510948905,
      "grad_norm": 0.15265095233917236,
      "learning_rate": 0.0002,
      "loss": 1.4994,
      "step": 2081
    },
    {
      "epoch": 0.7598540145985402,
      "grad_norm": 0.1555320918560028,
      "learning_rate": 0.0002,
      "loss": 1.5746,
      "step": 2082
    },
    {
      "epoch": 0.7602189781021897,
      "grad_norm": 0.1571255475282669,
      "learning_rate": 0.0002,
      "loss": 1.5216,
      "step": 2083
    },
    {
      "epoch": 0.7605839416058394,
      "grad_norm": 0.15190628170967102,
      "learning_rate": 0.0002,
      "loss": 1.4769,
      "step": 2084
    },
    {
      "epoch": 0.7609489051094891,
      "grad_norm": 0.15482057631015778,
      "learning_rate": 0.0002,
      "loss": 1.5537,
      "step": 2085
    },
    {
      "epoch": 0.7613138686131387,
      "grad_norm": 0.15167760848999023,
      "learning_rate": 0.0002,
      "loss": 1.4674,
      "step": 2086
    },
    {
      "epoch": 0.7616788321167883,
      "grad_norm": 0.15746118128299713,
      "learning_rate": 0.0002,
      "loss": 1.6003,
      "step": 2087
    },
    {
      "epoch": 0.762043795620438,
      "grad_norm": 0.15438441932201385,
      "learning_rate": 0.0002,
      "loss": 1.5036,
      "step": 2088
    },
    {
      "epoch": 0.7624087591240876,
      "grad_norm": 0.15328550338745117,
      "learning_rate": 0.0002,
      "loss": 1.4912,
      "step": 2089
    },
    {
      "epoch": 0.7627737226277372,
      "grad_norm": 0.15592452883720398,
      "learning_rate": 0.0002,
      "loss": 1.5881,
      "step": 2090
    },
    {
      "epoch": 0.7631386861313869,
      "grad_norm": 0.15219539403915405,
      "learning_rate": 0.0002,
      "loss": 1.4584,
      "step": 2091
    },
    {
      "epoch": 0.7635036496350365,
      "grad_norm": 0.1546536684036255,
      "learning_rate": 0.0002,
      "loss": 1.4709,
      "step": 2092
    },
    {
      "epoch": 0.7638686131386861,
      "grad_norm": 0.15590763092041016,
      "learning_rate": 0.0002,
      "loss": 1.4235,
      "step": 2093
    },
    {
      "epoch": 0.7642335766423358,
      "grad_norm": 0.1502070128917694,
      "learning_rate": 0.0002,
      "loss": 1.4579,
      "step": 2094
    },
    {
      "epoch": 0.7645985401459854,
      "grad_norm": 0.1555340588092804,
      "learning_rate": 0.0002,
      "loss": 1.4791,
      "step": 2095
    },
    {
      "epoch": 0.7649635036496351,
      "grad_norm": 0.15622155368328094,
      "learning_rate": 0.0002,
      "loss": 1.5227,
      "step": 2096
    },
    {
      "epoch": 0.7653284671532846,
      "grad_norm": 0.14831191301345825,
      "learning_rate": 0.0002,
      "loss": 1.492,
      "step": 2097
    },
    {
      "epoch": 0.7656934306569343,
      "grad_norm": 0.1546277552843094,
      "learning_rate": 0.0002,
      "loss": 1.5081,
      "step": 2098
    },
    {
      "epoch": 0.766058394160584,
      "grad_norm": 0.1615852564573288,
      "learning_rate": 0.0002,
      "loss": 1.4238,
      "step": 2099
    },
    {
      "epoch": 0.7664233576642335,
      "grad_norm": 0.15894511342048645,
      "learning_rate": 0.0002,
      "loss": 1.5315,
      "step": 2100
    },
    {
      "epoch": 0.7667883211678832,
      "grad_norm": 0.16496749222278595,
      "learning_rate": 0.0002,
      "loss": 1.5249,
      "step": 2101
    },
    {
      "epoch": 0.7671532846715329,
      "grad_norm": 0.16032083332538605,
      "learning_rate": 0.0002,
      "loss": 1.573,
      "step": 2102
    },
    {
      "epoch": 0.7675182481751824,
      "grad_norm": 0.156418114900589,
      "learning_rate": 0.0002,
      "loss": 1.4685,
      "step": 2103
    },
    {
      "epoch": 0.7678832116788321,
      "grad_norm": 0.17000409960746765,
      "learning_rate": 0.0002,
      "loss": 1.6201,
      "step": 2104
    },
    {
      "epoch": 0.7682481751824818,
      "grad_norm": 0.16321685910224915,
      "learning_rate": 0.0002,
      "loss": 1.528,
      "step": 2105
    },
    {
      "epoch": 0.7686131386861313,
      "grad_norm": 0.16541822254657745,
      "learning_rate": 0.0002,
      "loss": 1.5725,
      "step": 2106
    },
    {
      "epoch": 0.768978102189781,
      "grad_norm": 0.15652059018611908,
      "learning_rate": 0.0002,
      "loss": 1.5636,
      "step": 2107
    },
    {
      "epoch": 0.7693430656934307,
      "grad_norm": 0.16083137691020966,
      "learning_rate": 0.0002,
      "loss": 1.4804,
      "step": 2108
    },
    {
      "epoch": 0.7697080291970803,
      "grad_norm": 0.16257058084011078,
      "learning_rate": 0.0002,
      "loss": 1.5681,
      "step": 2109
    },
    {
      "epoch": 0.7700729927007299,
      "grad_norm": 0.1612514853477478,
      "learning_rate": 0.0002,
      "loss": 1.6071,
      "step": 2110
    },
    {
      "epoch": 0.7704379562043796,
      "grad_norm": 0.15261049568653107,
      "learning_rate": 0.0002,
      "loss": 1.4695,
      "step": 2111
    },
    {
      "epoch": 0.7708029197080292,
      "grad_norm": 0.1553247570991516,
      "learning_rate": 0.0002,
      "loss": 1.5402,
      "step": 2112
    },
    {
      "epoch": 0.7711678832116788,
      "grad_norm": 0.15815414488315582,
      "learning_rate": 0.0002,
      "loss": 1.447,
      "step": 2113
    },
    {
      "epoch": 0.7715328467153285,
      "grad_norm": 0.16065990924835205,
      "learning_rate": 0.0002,
      "loss": 1.5056,
      "step": 2114
    },
    {
      "epoch": 0.7718978102189781,
      "grad_norm": 0.15313079953193665,
      "learning_rate": 0.0002,
      "loss": 1.4706,
      "step": 2115
    },
    {
      "epoch": 0.7722627737226277,
      "grad_norm": 0.16512411832809448,
      "learning_rate": 0.0002,
      "loss": 1.4963,
      "step": 2116
    },
    {
      "epoch": 0.7726277372262774,
      "grad_norm": 0.15584588050842285,
      "learning_rate": 0.0002,
      "loss": 1.4895,
      "step": 2117
    },
    {
      "epoch": 0.772992700729927,
      "grad_norm": 0.16135640442371368,
      "learning_rate": 0.0002,
      "loss": 1.6202,
      "step": 2118
    },
    {
      "epoch": 0.7733576642335767,
      "grad_norm": 0.16206656396389008,
      "learning_rate": 0.0002,
      "loss": 1.5739,
      "step": 2119
    },
    {
      "epoch": 0.7737226277372263,
      "grad_norm": 0.15239644050598145,
      "learning_rate": 0.0002,
      "loss": 1.482,
      "step": 2120
    },
    {
      "epoch": 0.7740875912408759,
      "grad_norm": 0.16171157360076904,
      "learning_rate": 0.0002,
      "loss": 1.5584,
      "step": 2121
    },
    {
      "epoch": 0.7744525547445256,
      "grad_norm": 0.15634821355342865,
      "learning_rate": 0.0002,
      "loss": 1.5214,
      "step": 2122
    },
    {
      "epoch": 0.7748175182481751,
      "grad_norm": 0.15074703097343445,
      "learning_rate": 0.0002,
      "loss": 1.4551,
      "step": 2123
    },
    {
      "epoch": 0.7751824817518248,
      "grad_norm": 0.15724031627178192,
      "learning_rate": 0.0002,
      "loss": 1.5318,
      "step": 2124
    },
    {
      "epoch": 0.7755474452554745,
      "grad_norm": 0.14934736490249634,
      "learning_rate": 0.0002,
      "loss": 1.3626,
      "step": 2125
    },
    {
      "epoch": 0.775912408759124,
      "grad_norm": 0.15306633710861206,
      "learning_rate": 0.0002,
      "loss": 1.4307,
      "step": 2126
    },
    {
      "epoch": 0.7762773722627737,
      "grad_norm": 0.15188626945018768,
      "learning_rate": 0.0002,
      "loss": 1.5294,
      "step": 2127
    },
    {
      "epoch": 0.7766423357664234,
      "grad_norm": 0.151889368891716,
      "learning_rate": 0.0002,
      "loss": 1.531,
      "step": 2128
    },
    {
      "epoch": 0.777007299270073,
      "grad_norm": 0.15436860918998718,
      "learning_rate": 0.0002,
      "loss": 1.4647,
      "step": 2129
    },
    {
      "epoch": 0.7773722627737226,
      "grad_norm": 0.15661679208278656,
      "learning_rate": 0.0002,
      "loss": 1.4624,
      "step": 2130
    },
    {
      "epoch": 0.7777372262773723,
      "grad_norm": 0.14946074783802032,
      "learning_rate": 0.0002,
      "loss": 1.526,
      "step": 2131
    },
    {
      "epoch": 0.7781021897810219,
      "grad_norm": 0.15488705039024353,
      "learning_rate": 0.0002,
      "loss": 1.5476,
      "step": 2132
    },
    {
      "epoch": 0.7784671532846715,
      "grad_norm": 0.16147853434085846,
      "learning_rate": 0.0002,
      "loss": 1.5264,
      "step": 2133
    },
    {
      "epoch": 0.7788321167883212,
      "grad_norm": 0.15462516248226166,
      "learning_rate": 0.0002,
      "loss": 1.4399,
      "step": 2134
    },
    {
      "epoch": 0.7791970802919708,
      "grad_norm": 0.15532289445400238,
      "learning_rate": 0.0002,
      "loss": 1.4735,
      "step": 2135
    },
    {
      "epoch": 0.7795620437956204,
      "grad_norm": 0.1656450480222702,
      "learning_rate": 0.0002,
      "loss": 1.5105,
      "step": 2136
    },
    {
      "epoch": 0.7799270072992701,
      "grad_norm": 0.16104644536972046,
      "learning_rate": 0.0002,
      "loss": 1.4493,
      "step": 2137
    },
    {
      "epoch": 0.7802919708029197,
      "grad_norm": 0.16525892913341522,
      "learning_rate": 0.0002,
      "loss": 1.4695,
      "step": 2138
    },
    {
      "epoch": 0.7806569343065694,
      "grad_norm": 0.15531045198440552,
      "learning_rate": 0.0002,
      "loss": 1.4723,
      "step": 2139
    },
    {
      "epoch": 0.781021897810219,
      "grad_norm": 0.15464255213737488,
      "learning_rate": 0.0002,
      "loss": 1.5241,
      "step": 2140
    },
    {
      "epoch": 0.7813868613138686,
      "grad_norm": 0.1542951613664627,
      "learning_rate": 0.0002,
      "loss": 1.5123,
      "step": 2141
    },
    {
      "epoch": 0.7817518248175183,
      "grad_norm": 0.16115862131118774,
      "learning_rate": 0.0002,
      "loss": 1.515,
      "step": 2142
    },
    {
      "epoch": 0.7821167883211679,
      "grad_norm": 0.1541251689195633,
      "learning_rate": 0.0002,
      "loss": 1.4861,
      "step": 2143
    },
    {
      "epoch": 0.7824817518248175,
      "grad_norm": 0.15437598526477814,
      "learning_rate": 0.0002,
      "loss": 1.4154,
      "step": 2144
    },
    {
      "epoch": 0.7828467153284672,
      "grad_norm": 0.1592313051223755,
      "learning_rate": 0.0002,
      "loss": 1.5301,
      "step": 2145
    },
    {
      "epoch": 0.7832116788321168,
      "grad_norm": 0.1587037593126297,
      "learning_rate": 0.0002,
      "loss": 1.5037,
      "step": 2146
    },
    {
      "epoch": 0.7835766423357664,
      "grad_norm": 0.16776834428310394,
      "learning_rate": 0.0002,
      "loss": 1.57,
      "step": 2147
    },
    {
      "epoch": 0.7839416058394161,
      "grad_norm": 0.1605588048696518,
      "learning_rate": 0.0002,
      "loss": 1.5911,
      "step": 2148
    },
    {
      "epoch": 0.7843065693430656,
      "grad_norm": 0.1521119624376297,
      "learning_rate": 0.0002,
      "loss": 1.4671,
      "step": 2149
    },
    {
      "epoch": 0.7846715328467153,
      "grad_norm": 0.1582719385623932,
      "learning_rate": 0.0002,
      "loss": 1.5751,
      "step": 2150
    },
    {
      "epoch": 0.785036496350365,
      "grad_norm": 0.162398561835289,
      "learning_rate": 0.0002,
      "loss": 1.4855,
      "step": 2151
    },
    {
      "epoch": 0.7854014598540145,
      "grad_norm": 0.15684527158737183,
      "learning_rate": 0.0002,
      "loss": 1.5577,
      "step": 2152
    },
    {
      "epoch": 0.7857664233576642,
      "grad_norm": 0.1562509685754776,
      "learning_rate": 0.0002,
      "loss": 1.5019,
      "step": 2153
    },
    {
      "epoch": 0.7861313868613139,
      "grad_norm": 0.16381525993347168,
      "learning_rate": 0.0002,
      "loss": 1.5728,
      "step": 2154
    },
    {
      "epoch": 0.7864963503649635,
      "grad_norm": 0.15197311341762543,
      "learning_rate": 0.0002,
      "loss": 1.499,
      "step": 2155
    },
    {
      "epoch": 0.7868613138686131,
      "grad_norm": 0.15926070511341095,
      "learning_rate": 0.0002,
      "loss": 1.4896,
      "step": 2156
    },
    {
      "epoch": 0.7872262773722628,
      "grad_norm": 0.165689617395401,
      "learning_rate": 0.0002,
      "loss": 1.5943,
      "step": 2157
    },
    {
      "epoch": 0.7875912408759124,
      "grad_norm": 0.15521055459976196,
      "learning_rate": 0.0002,
      "loss": 1.5068,
      "step": 2158
    },
    {
      "epoch": 0.787956204379562,
      "grad_norm": 0.14867708086967468,
      "learning_rate": 0.0002,
      "loss": 1.4244,
      "step": 2159
    },
    {
      "epoch": 0.7883211678832117,
      "grad_norm": 0.15359121561050415,
      "learning_rate": 0.0002,
      "loss": 1.4943,
      "step": 2160
    },
    {
      "epoch": 0.7886861313868613,
      "grad_norm": 0.16918304562568665,
      "learning_rate": 0.0002,
      "loss": 1.4935,
      "step": 2161
    },
    {
      "epoch": 0.789051094890511,
      "grad_norm": 0.1530209630727768,
      "learning_rate": 0.0002,
      "loss": 1.5501,
      "step": 2162
    },
    {
      "epoch": 0.7894160583941606,
      "grad_norm": 0.17515549063682556,
      "learning_rate": 0.0002,
      "loss": 1.4644,
      "step": 2163
    },
    {
      "epoch": 0.7897810218978102,
      "grad_norm": 0.1513419896364212,
      "learning_rate": 0.0002,
      "loss": 1.4823,
      "step": 2164
    },
    {
      "epoch": 0.7901459854014599,
      "grad_norm": 0.16807599365711212,
      "learning_rate": 0.0002,
      "loss": 1.504,
      "step": 2165
    },
    {
      "epoch": 0.7905109489051095,
      "grad_norm": 0.15638227760791779,
      "learning_rate": 0.0002,
      "loss": 1.4712,
      "step": 2166
    },
    {
      "epoch": 0.7908759124087591,
      "grad_norm": 0.15589390695095062,
      "learning_rate": 0.0002,
      "loss": 1.4617,
      "step": 2167
    },
    {
      "epoch": 0.7912408759124088,
      "grad_norm": 0.15778905153274536,
      "learning_rate": 0.0002,
      "loss": 1.5338,
      "step": 2168
    },
    {
      "epoch": 0.7916058394160584,
      "grad_norm": 0.15290994942188263,
      "learning_rate": 0.0002,
      "loss": 1.4736,
      "step": 2169
    },
    {
      "epoch": 0.791970802919708,
      "grad_norm": 0.1509230136871338,
      "learning_rate": 0.0002,
      "loss": 1.4987,
      "step": 2170
    },
    {
      "epoch": 0.7923357664233577,
      "grad_norm": 0.16029399633407593,
      "learning_rate": 0.0002,
      "loss": 1.4866,
      "step": 2171
    },
    {
      "epoch": 0.7927007299270074,
      "grad_norm": 0.15749643743038177,
      "learning_rate": 0.0002,
      "loss": 1.4904,
      "step": 2172
    },
    {
      "epoch": 0.7930656934306569,
      "grad_norm": 0.158694326877594,
      "learning_rate": 0.0002,
      "loss": 1.5414,
      "step": 2173
    },
    {
      "epoch": 0.7934306569343066,
      "grad_norm": 0.1521519422531128,
      "learning_rate": 0.0002,
      "loss": 1.4714,
      "step": 2174
    },
    {
      "epoch": 0.7937956204379562,
      "grad_norm": 0.16072914004325867,
      "learning_rate": 0.0002,
      "loss": 1.552,
      "step": 2175
    },
    {
      "epoch": 0.7941605839416058,
      "grad_norm": 0.15187612175941467,
      "learning_rate": 0.0002,
      "loss": 1.4583,
      "step": 2176
    },
    {
      "epoch": 0.7945255474452555,
      "grad_norm": 0.16186575591564178,
      "learning_rate": 0.0002,
      "loss": 1.5011,
      "step": 2177
    },
    {
      "epoch": 0.7948905109489051,
      "grad_norm": 0.17714788019657135,
      "learning_rate": 0.0002,
      "loss": 1.4962,
      "step": 2178
    },
    {
      "epoch": 0.7952554744525547,
      "grad_norm": 0.15734679996967316,
      "learning_rate": 0.0002,
      "loss": 1.4946,
      "step": 2179
    },
    {
      "epoch": 0.7956204379562044,
      "grad_norm": 0.15723565220832825,
      "learning_rate": 0.0002,
      "loss": 1.5635,
      "step": 2180
    },
    {
      "epoch": 0.795985401459854,
      "grad_norm": 0.15400776267051697,
      "learning_rate": 0.0002,
      "loss": 1.5362,
      "step": 2181
    },
    {
      "epoch": 0.7963503649635036,
      "grad_norm": 0.150944784283638,
      "learning_rate": 0.0002,
      "loss": 1.4381,
      "step": 2182
    },
    {
      "epoch": 0.7967153284671533,
      "grad_norm": 0.17286211252212524,
      "learning_rate": 0.0002,
      "loss": 1.576,
      "step": 2183
    },
    {
      "epoch": 0.7970802919708029,
      "grad_norm": 0.1474877893924713,
      "learning_rate": 0.0002,
      "loss": 1.4734,
      "step": 2184
    },
    {
      "epoch": 0.7974452554744526,
      "grad_norm": 0.15296553075313568,
      "learning_rate": 0.0002,
      "loss": 1.4282,
      "step": 2185
    },
    {
      "epoch": 0.7978102189781022,
      "grad_norm": 0.1566886454820633,
      "learning_rate": 0.0002,
      "loss": 1.4854,
      "step": 2186
    },
    {
      "epoch": 0.7981751824817518,
      "grad_norm": 0.15699657797813416,
      "learning_rate": 0.0002,
      "loss": 1.4988,
      "step": 2187
    },
    {
      "epoch": 0.7985401459854015,
      "grad_norm": 0.1598733514547348,
      "learning_rate": 0.0002,
      "loss": 1.5371,
      "step": 2188
    },
    {
      "epoch": 0.7989051094890511,
      "grad_norm": 0.15715117752552032,
      "learning_rate": 0.0002,
      "loss": 1.503,
      "step": 2189
    },
    {
      "epoch": 0.7992700729927007,
      "grad_norm": 0.1571096032857895,
      "learning_rate": 0.0002,
      "loss": 1.5802,
      "step": 2190
    },
    {
      "epoch": 0.7996350364963504,
      "grad_norm": 0.15845200419425964,
      "learning_rate": 0.0002,
      "loss": 1.4449,
      "step": 2191
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.15734829008579254,
      "learning_rate": 0.0002,
      "loss": 1.4965,
      "step": 2192
    },
    {
      "epoch": 0.8003649635036496,
      "grad_norm": 0.16051490604877472,
      "learning_rate": 0.0002,
      "loss": 1.4933,
      "step": 2193
    },
    {
      "epoch": 0.8007299270072993,
      "grad_norm": 0.1546684056520462,
      "learning_rate": 0.0002,
      "loss": 1.5491,
      "step": 2194
    },
    {
      "epoch": 0.801094890510949,
      "grad_norm": 0.15060491859912872,
      "learning_rate": 0.0002,
      "loss": 1.4687,
      "step": 2195
    },
    {
      "epoch": 0.8014598540145985,
      "grad_norm": 0.15072669088840485,
      "learning_rate": 0.0002,
      "loss": 1.4834,
      "step": 2196
    },
    {
      "epoch": 0.8018248175182482,
      "grad_norm": 0.15237891674041748,
      "learning_rate": 0.0002,
      "loss": 1.4661,
      "step": 2197
    },
    {
      "epoch": 0.8021897810218979,
      "grad_norm": 0.15766143798828125,
      "learning_rate": 0.0002,
      "loss": 1.5175,
      "step": 2198
    },
    {
      "epoch": 0.8025547445255474,
      "grad_norm": 0.15796513855457306,
      "learning_rate": 0.0002,
      "loss": 1.5231,
      "step": 2199
    },
    {
      "epoch": 0.8029197080291971,
      "grad_norm": 0.16044728457927704,
      "learning_rate": 0.0002,
      "loss": 1.6144,
      "step": 2200
    },
    {
      "epoch": 0.8032846715328467,
      "grad_norm": 0.15399840474128723,
      "learning_rate": 0.0002,
      "loss": 1.4719,
      "step": 2201
    },
    {
      "epoch": 0.8036496350364963,
      "grad_norm": 0.16090025007724762,
      "learning_rate": 0.0002,
      "loss": 1.5533,
      "step": 2202
    },
    {
      "epoch": 0.804014598540146,
      "grad_norm": 0.1563398391008377,
      "learning_rate": 0.0002,
      "loss": 1.4857,
      "step": 2203
    },
    {
      "epoch": 0.8043795620437956,
      "grad_norm": 0.16053391993045807,
      "learning_rate": 0.0002,
      "loss": 1.5133,
      "step": 2204
    },
    {
      "epoch": 0.8047445255474452,
      "grad_norm": 0.15220069885253906,
      "learning_rate": 0.0002,
      "loss": 1.5204,
      "step": 2205
    },
    {
      "epoch": 0.8051094890510949,
      "grad_norm": 0.15326455235481262,
      "learning_rate": 0.0002,
      "loss": 1.5267,
      "step": 2206
    },
    {
      "epoch": 0.8054744525547445,
      "grad_norm": 0.1585809737443924,
      "learning_rate": 0.0002,
      "loss": 1.5078,
      "step": 2207
    },
    {
      "epoch": 0.8058394160583942,
      "grad_norm": 0.176558718085289,
      "learning_rate": 0.0002,
      "loss": 1.5246,
      "step": 2208
    },
    {
      "epoch": 0.8062043795620438,
      "grad_norm": 0.16207051277160645,
      "learning_rate": 0.0002,
      "loss": 1.4968,
      "step": 2209
    },
    {
      "epoch": 0.8065693430656934,
      "grad_norm": 0.16076402366161346,
      "learning_rate": 0.0002,
      "loss": 1.4695,
      "step": 2210
    },
    {
      "epoch": 0.8069343065693431,
      "grad_norm": 0.156129851937294,
      "learning_rate": 0.0002,
      "loss": 1.4467,
      "step": 2211
    },
    {
      "epoch": 0.8072992700729927,
      "grad_norm": 0.16084560751914978,
      "learning_rate": 0.0002,
      "loss": 1.5382,
      "step": 2212
    },
    {
      "epoch": 0.8076642335766423,
      "grad_norm": 0.15734794735908508,
      "learning_rate": 0.0002,
      "loss": 1.4916,
      "step": 2213
    },
    {
      "epoch": 0.808029197080292,
      "grad_norm": 0.16180066764354706,
      "learning_rate": 0.0002,
      "loss": 1.5602,
      "step": 2214
    },
    {
      "epoch": 0.8083941605839416,
      "grad_norm": 0.15757577121257782,
      "learning_rate": 0.0002,
      "loss": 1.5114,
      "step": 2215
    },
    {
      "epoch": 0.8087591240875912,
      "grad_norm": 0.15597617626190186,
      "learning_rate": 0.0002,
      "loss": 1.4773,
      "step": 2216
    },
    {
      "epoch": 0.8091240875912409,
      "grad_norm": 0.15361930429935455,
      "learning_rate": 0.0002,
      "loss": 1.4909,
      "step": 2217
    },
    {
      "epoch": 0.8094890510948906,
      "grad_norm": 0.156340554356575,
      "learning_rate": 0.0002,
      "loss": 1.5231,
      "step": 2218
    },
    {
      "epoch": 0.8098540145985401,
      "grad_norm": 0.15457271039485931,
      "learning_rate": 0.0002,
      "loss": 1.5003,
      "step": 2219
    },
    {
      "epoch": 0.8102189781021898,
      "grad_norm": 0.15572312474250793,
      "learning_rate": 0.0002,
      "loss": 1.5108,
      "step": 2220
    },
    {
      "epoch": 0.8105839416058395,
      "grad_norm": 0.15656065940856934,
      "learning_rate": 0.0002,
      "loss": 1.5536,
      "step": 2221
    },
    {
      "epoch": 0.810948905109489,
      "grad_norm": 0.1575278788805008,
      "learning_rate": 0.0002,
      "loss": 1.6052,
      "step": 2222
    },
    {
      "epoch": 0.8113138686131387,
      "grad_norm": 0.1580163687467575,
      "learning_rate": 0.0002,
      "loss": 1.5842,
      "step": 2223
    },
    {
      "epoch": 0.8116788321167884,
      "grad_norm": 0.16402053833007812,
      "learning_rate": 0.0002,
      "loss": 1.5288,
      "step": 2224
    },
    {
      "epoch": 0.8120437956204379,
      "grad_norm": 0.15927176177501678,
      "learning_rate": 0.0002,
      "loss": 1.5454,
      "step": 2225
    },
    {
      "epoch": 0.8124087591240876,
      "grad_norm": 0.16466377675533295,
      "learning_rate": 0.0002,
      "loss": 1.5729,
      "step": 2226
    },
    {
      "epoch": 0.8127737226277372,
      "grad_norm": 0.15202534198760986,
      "learning_rate": 0.0002,
      "loss": 1.494,
      "step": 2227
    },
    {
      "epoch": 0.8131386861313868,
      "grad_norm": 0.1546994298696518,
      "learning_rate": 0.0002,
      "loss": 1.4478,
      "step": 2228
    },
    {
      "epoch": 0.8135036496350365,
      "grad_norm": 0.15925002098083496,
      "learning_rate": 0.0002,
      "loss": 1.4587,
      "step": 2229
    },
    {
      "epoch": 0.8138686131386861,
      "grad_norm": 0.14904777705669403,
      "learning_rate": 0.0002,
      "loss": 1.4245,
      "step": 2230
    },
    {
      "epoch": 0.8142335766423358,
      "grad_norm": 0.1533837616443634,
      "learning_rate": 0.0002,
      "loss": 1.5371,
      "step": 2231
    },
    {
      "epoch": 0.8145985401459854,
      "grad_norm": 0.15266254544258118,
      "learning_rate": 0.0002,
      "loss": 1.459,
      "step": 2232
    },
    {
      "epoch": 0.814963503649635,
      "grad_norm": 0.1594044715166092,
      "learning_rate": 0.0002,
      "loss": 1.4865,
      "step": 2233
    },
    {
      "epoch": 0.8153284671532847,
      "grad_norm": 0.15857721865177155,
      "learning_rate": 0.0002,
      "loss": 1.4536,
      "step": 2234
    },
    {
      "epoch": 0.8156934306569343,
      "grad_norm": 0.15945102274417877,
      "learning_rate": 0.0002,
      "loss": 1.553,
      "step": 2235
    },
    {
      "epoch": 0.8160583941605839,
      "grad_norm": 0.15498164296150208,
      "learning_rate": 0.0002,
      "loss": 1.4872,
      "step": 2236
    },
    {
      "epoch": 0.8164233576642336,
      "grad_norm": 0.16246281564235687,
      "learning_rate": 0.0002,
      "loss": 1.5385,
      "step": 2237
    },
    {
      "epoch": 0.8167883211678832,
      "grad_norm": 0.15449301898479462,
      "learning_rate": 0.0002,
      "loss": 1.4632,
      "step": 2238
    },
    {
      "epoch": 0.8171532846715328,
      "grad_norm": 0.16152016818523407,
      "learning_rate": 0.0002,
      "loss": 1.4769,
      "step": 2239
    },
    {
      "epoch": 0.8175182481751825,
      "grad_norm": 0.15214724838733673,
      "learning_rate": 0.0002,
      "loss": 1.4221,
      "step": 2240
    },
    {
      "epoch": 0.8178832116788322,
      "grad_norm": 0.15313735604286194,
      "learning_rate": 0.0002,
      "loss": 1.5,
      "step": 2241
    },
    {
      "epoch": 0.8182481751824817,
      "grad_norm": 0.1534382402896881,
      "learning_rate": 0.0002,
      "loss": 1.4533,
      "step": 2242
    },
    {
      "epoch": 0.8186131386861314,
      "grad_norm": 0.16452684998512268,
      "learning_rate": 0.0002,
      "loss": 1.5231,
      "step": 2243
    },
    {
      "epoch": 0.8189781021897811,
      "grad_norm": 0.16418226063251495,
      "learning_rate": 0.0002,
      "loss": 1.5833,
      "step": 2244
    },
    {
      "epoch": 0.8193430656934306,
      "grad_norm": 0.16079388558864594,
      "learning_rate": 0.0002,
      "loss": 1.5256,
      "step": 2245
    },
    {
      "epoch": 0.8197080291970803,
      "grad_norm": 0.1510297656059265,
      "learning_rate": 0.0002,
      "loss": 1.437,
      "step": 2246
    },
    {
      "epoch": 0.82007299270073,
      "grad_norm": 0.15781307220458984,
      "learning_rate": 0.0002,
      "loss": 1.4886,
      "step": 2247
    },
    {
      "epoch": 0.8204379562043795,
      "grad_norm": 0.16229836642742157,
      "learning_rate": 0.0002,
      "loss": 1.557,
      "step": 2248
    },
    {
      "epoch": 0.8208029197080292,
      "grad_norm": 0.15896330773830414,
      "learning_rate": 0.0002,
      "loss": 1.4648,
      "step": 2249
    },
    {
      "epoch": 0.8211678832116789,
      "grad_norm": 0.17146839201450348,
      "learning_rate": 0.0002,
      "loss": 1.6071,
      "step": 2250
    },
    {
      "epoch": 0.8215328467153284,
      "grad_norm": 0.15192079544067383,
      "learning_rate": 0.0002,
      "loss": 1.4971,
      "step": 2251
    },
    {
      "epoch": 0.8218978102189781,
      "grad_norm": 0.1811387538909912,
      "learning_rate": 0.0002,
      "loss": 1.5211,
      "step": 2252
    },
    {
      "epoch": 0.8222627737226277,
      "grad_norm": 0.1489838808774948,
      "learning_rate": 0.0002,
      "loss": 1.4726,
      "step": 2253
    },
    {
      "epoch": 0.8226277372262774,
      "grad_norm": 0.1619056761264801,
      "learning_rate": 0.0002,
      "loss": 1.5187,
      "step": 2254
    },
    {
      "epoch": 0.822992700729927,
      "grad_norm": 0.15924441814422607,
      "learning_rate": 0.0002,
      "loss": 1.4734,
      "step": 2255
    },
    {
      "epoch": 0.8233576642335766,
      "grad_norm": 0.18053877353668213,
      "learning_rate": 0.0002,
      "loss": 1.4916,
      "step": 2256
    },
    {
      "epoch": 0.8237226277372263,
      "grad_norm": 0.15883943438529968,
      "learning_rate": 0.0002,
      "loss": 1.4871,
      "step": 2257
    },
    {
      "epoch": 0.8240875912408759,
      "grad_norm": 0.17340844869613647,
      "learning_rate": 0.0002,
      "loss": 1.5752,
      "step": 2258
    },
    {
      "epoch": 0.8244525547445255,
      "grad_norm": 0.16305063664913177,
      "learning_rate": 0.0002,
      "loss": 1.423,
      "step": 2259
    },
    {
      "epoch": 0.8248175182481752,
      "grad_norm": 0.1598449945449829,
      "learning_rate": 0.0002,
      "loss": 1.5386,
      "step": 2260
    },
    {
      "epoch": 0.8251824817518248,
      "grad_norm": 0.1709093302488327,
      "learning_rate": 0.0002,
      "loss": 1.5526,
      "step": 2261
    },
    {
      "epoch": 0.8255474452554744,
      "grad_norm": 0.155319944024086,
      "learning_rate": 0.0002,
      "loss": 1.4555,
      "step": 2262
    },
    {
      "epoch": 0.8259124087591241,
      "grad_norm": 0.1527390033006668,
      "learning_rate": 0.0002,
      "loss": 1.4809,
      "step": 2263
    },
    {
      "epoch": 0.8262773722627738,
      "grad_norm": 0.159617617726326,
      "learning_rate": 0.0002,
      "loss": 1.524,
      "step": 2264
    },
    {
      "epoch": 0.8266423357664233,
      "grad_norm": 0.16119500994682312,
      "learning_rate": 0.0002,
      "loss": 1.5338,
      "step": 2265
    },
    {
      "epoch": 0.827007299270073,
      "grad_norm": 0.16080860793590546,
      "learning_rate": 0.0002,
      "loss": 1.5382,
      "step": 2266
    },
    {
      "epoch": 0.8273722627737227,
      "grad_norm": 0.15743327140808105,
      "learning_rate": 0.0002,
      "loss": 1.4784,
      "step": 2267
    },
    {
      "epoch": 0.8277372262773722,
      "grad_norm": 0.15680499374866486,
      "learning_rate": 0.0002,
      "loss": 1.4973,
      "step": 2268
    },
    {
      "epoch": 0.8281021897810219,
      "grad_norm": 0.1563030332326889,
      "learning_rate": 0.0002,
      "loss": 1.5249,
      "step": 2269
    },
    {
      "epoch": 0.8284671532846716,
      "grad_norm": 0.16259720921516418,
      "learning_rate": 0.0002,
      "loss": 1.4916,
      "step": 2270
    },
    {
      "epoch": 0.8288321167883211,
      "grad_norm": 0.1595313847064972,
      "learning_rate": 0.0002,
      "loss": 1.5298,
      "step": 2271
    },
    {
      "epoch": 0.8291970802919708,
      "grad_norm": 0.16097380220890045,
      "learning_rate": 0.0002,
      "loss": 1.4588,
      "step": 2272
    },
    {
      "epoch": 0.8295620437956205,
      "grad_norm": 0.159635990858078,
      "learning_rate": 0.0002,
      "loss": 1.5441,
      "step": 2273
    },
    {
      "epoch": 0.82992700729927,
      "grad_norm": 0.16618938744068146,
      "learning_rate": 0.0002,
      "loss": 1.61,
      "step": 2274
    },
    {
      "epoch": 0.8302919708029197,
      "grad_norm": 0.15680310130119324,
      "learning_rate": 0.0002,
      "loss": 1.5187,
      "step": 2275
    },
    {
      "epoch": 0.8306569343065694,
      "grad_norm": 0.1538371741771698,
      "learning_rate": 0.0002,
      "loss": 1.449,
      "step": 2276
    },
    {
      "epoch": 0.831021897810219,
      "grad_norm": 0.1644904613494873,
      "learning_rate": 0.0002,
      "loss": 1.5204,
      "step": 2277
    },
    {
      "epoch": 0.8313868613138686,
      "grad_norm": 0.1620727777481079,
      "learning_rate": 0.0002,
      "loss": 1.4893,
      "step": 2278
    },
    {
      "epoch": 0.8317518248175182,
      "grad_norm": 0.17021270096302032,
      "learning_rate": 0.0002,
      "loss": 1.5077,
      "step": 2279
    },
    {
      "epoch": 0.8321167883211679,
      "grad_norm": 0.15933747589588165,
      "learning_rate": 0.0002,
      "loss": 1.5087,
      "step": 2280
    },
    {
      "epoch": 0.8324817518248175,
      "grad_norm": 0.15967759490013123,
      "learning_rate": 0.0002,
      "loss": 1.4326,
      "step": 2281
    },
    {
      "epoch": 0.8328467153284671,
      "grad_norm": 0.15833355486392975,
      "learning_rate": 0.0002,
      "loss": 1.5451,
      "step": 2282
    },
    {
      "epoch": 0.8332116788321168,
      "grad_norm": 0.15707454085350037,
      "learning_rate": 0.0002,
      "loss": 1.427,
      "step": 2283
    },
    {
      "epoch": 0.8335766423357664,
      "grad_norm": 0.15656037628650665,
      "learning_rate": 0.0002,
      "loss": 1.4873,
      "step": 2284
    },
    {
      "epoch": 0.833941605839416,
      "grad_norm": 0.16039277613162994,
      "learning_rate": 0.0002,
      "loss": 1.4486,
      "step": 2285
    },
    {
      "epoch": 0.8343065693430657,
      "grad_norm": 0.1561787873506546,
      "learning_rate": 0.0002,
      "loss": 1.5036,
      "step": 2286
    },
    {
      "epoch": 0.8346715328467154,
      "grad_norm": 0.16879738867282867,
      "learning_rate": 0.0002,
      "loss": 1.5073,
      "step": 2287
    },
    {
      "epoch": 0.8350364963503649,
      "grad_norm": 0.15154311060905457,
      "learning_rate": 0.0002,
      "loss": 1.4591,
      "step": 2288
    },
    {
      "epoch": 0.8354014598540146,
      "grad_norm": 0.17462129890918732,
      "learning_rate": 0.0002,
      "loss": 1.5597,
      "step": 2289
    },
    {
      "epoch": 0.8357664233576643,
      "grad_norm": 0.16567908227443695,
      "learning_rate": 0.0002,
      "loss": 1.4429,
      "step": 2290
    },
    {
      "epoch": 0.8361313868613138,
      "grad_norm": 0.15452058613300323,
      "learning_rate": 0.0002,
      "loss": 1.4545,
      "step": 2291
    },
    {
      "epoch": 0.8364963503649635,
      "grad_norm": 0.16518020629882812,
      "learning_rate": 0.0002,
      "loss": 1.4623,
      "step": 2292
    },
    {
      "epoch": 0.8368613138686132,
      "grad_norm": 0.15965904295444489,
      "learning_rate": 0.0002,
      "loss": 1.564,
      "step": 2293
    },
    {
      "epoch": 0.8372262773722627,
      "grad_norm": 0.16197268664836884,
      "learning_rate": 0.0002,
      "loss": 1.4973,
      "step": 2294
    },
    {
      "epoch": 0.8375912408759124,
      "grad_norm": 0.1560663878917694,
      "learning_rate": 0.0002,
      "loss": 1.4834,
      "step": 2295
    },
    {
      "epoch": 0.8379562043795621,
      "grad_norm": 0.16666880249977112,
      "learning_rate": 0.0002,
      "loss": 1.5144,
      "step": 2296
    },
    {
      "epoch": 0.8383211678832116,
      "grad_norm": 0.16623884439468384,
      "learning_rate": 0.0002,
      "loss": 1.5627,
      "step": 2297
    },
    {
      "epoch": 0.8386861313868613,
      "grad_norm": 0.16150391101837158,
      "learning_rate": 0.0002,
      "loss": 1.546,
      "step": 2298
    },
    {
      "epoch": 0.839051094890511,
      "grad_norm": 0.16383549571037292,
      "learning_rate": 0.0002,
      "loss": 1.4712,
      "step": 2299
    },
    {
      "epoch": 0.8394160583941606,
      "grad_norm": 0.16150294244289398,
      "learning_rate": 0.0002,
      "loss": 1.4816,
      "step": 2300
    },
    {
      "epoch": 0.8397810218978102,
      "grad_norm": 0.15559813380241394,
      "learning_rate": 0.0002,
      "loss": 1.4783,
      "step": 2301
    },
    {
      "epoch": 0.8401459854014599,
      "grad_norm": 0.18212434649467468,
      "learning_rate": 0.0002,
      "loss": 1.5345,
      "step": 2302
    },
    {
      "epoch": 0.8405109489051095,
      "grad_norm": 0.15881288051605225,
      "learning_rate": 0.0002,
      "loss": 1.4864,
      "step": 2303
    },
    {
      "epoch": 0.8408759124087591,
      "grad_norm": 0.1602386087179184,
      "learning_rate": 0.0002,
      "loss": 1.5039,
      "step": 2304
    },
    {
      "epoch": 0.8412408759124088,
      "grad_norm": 0.15861576795578003,
      "learning_rate": 0.0002,
      "loss": 1.5078,
      "step": 2305
    },
    {
      "epoch": 0.8416058394160584,
      "grad_norm": 0.15379668772220612,
      "learning_rate": 0.0002,
      "loss": 1.4794,
      "step": 2306
    },
    {
      "epoch": 0.841970802919708,
      "grad_norm": 0.1641872227191925,
      "learning_rate": 0.0002,
      "loss": 1.552,
      "step": 2307
    },
    {
      "epoch": 0.8423357664233576,
      "grad_norm": 0.15960709750652313,
      "learning_rate": 0.0002,
      "loss": 1.5059,
      "step": 2308
    },
    {
      "epoch": 0.8427007299270073,
      "grad_norm": 0.16700872778892517,
      "learning_rate": 0.0002,
      "loss": 1.5949,
      "step": 2309
    },
    {
      "epoch": 0.843065693430657,
      "grad_norm": 0.16384482383728027,
      "learning_rate": 0.0002,
      "loss": 1.5121,
      "step": 2310
    },
    {
      "epoch": 0.8434306569343065,
      "grad_norm": 0.1719949096441269,
      "learning_rate": 0.0002,
      "loss": 1.487,
      "step": 2311
    },
    {
      "epoch": 0.8437956204379562,
      "grad_norm": 0.16556832194328308,
      "learning_rate": 0.0002,
      "loss": 1.4591,
      "step": 2312
    },
    {
      "epoch": 0.8441605839416059,
      "grad_norm": 0.17521554231643677,
      "learning_rate": 0.0002,
      "loss": 1.5092,
      "step": 2313
    },
    {
      "epoch": 0.8445255474452554,
      "grad_norm": 0.159409761428833,
      "learning_rate": 0.0002,
      "loss": 1.5175,
      "step": 2314
    },
    {
      "epoch": 0.8448905109489051,
      "grad_norm": 0.17051677405834198,
      "learning_rate": 0.0002,
      "loss": 1.4863,
      "step": 2315
    },
    {
      "epoch": 0.8452554744525548,
      "grad_norm": 0.15669485926628113,
      "learning_rate": 0.0002,
      "loss": 1.4869,
      "step": 2316
    },
    {
      "epoch": 0.8456204379562043,
      "grad_norm": 0.16043908894062042,
      "learning_rate": 0.0002,
      "loss": 1.4878,
      "step": 2317
    },
    {
      "epoch": 0.845985401459854,
      "grad_norm": 0.16012299060821533,
      "learning_rate": 0.0002,
      "loss": 1.519,
      "step": 2318
    },
    {
      "epoch": 0.8463503649635037,
      "grad_norm": 0.15601202845573425,
      "learning_rate": 0.0002,
      "loss": 1.4684,
      "step": 2319
    },
    {
      "epoch": 0.8467153284671532,
      "grad_norm": 0.16038863360881805,
      "learning_rate": 0.0002,
      "loss": 1.4797,
      "step": 2320
    },
    {
      "epoch": 0.8470802919708029,
      "grad_norm": 0.14965301752090454,
      "learning_rate": 0.0002,
      "loss": 1.5122,
      "step": 2321
    },
    {
      "epoch": 0.8474452554744526,
      "grad_norm": 0.15605370700359344,
      "learning_rate": 0.0002,
      "loss": 1.5029,
      "step": 2322
    },
    {
      "epoch": 0.8478102189781022,
      "grad_norm": 0.1543460488319397,
      "learning_rate": 0.0002,
      "loss": 1.5271,
      "step": 2323
    },
    {
      "epoch": 0.8481751824817518,
      "grad_norm": 0.158338263630867,
      "learning_rate": 0.0002,
      "loss": 1.4763,
      "step": 2324
    },
    {
      "epoch": 0.8485401459854015,
      "grad_norm": 0.15912003815174103,
      "learning_rate": 0.0002,
      "loss": 1.5328,
      "step": 2325
    },
    {
      "epoch": 0.8489051094890511,
      "grad_norm": 0.17063118517398834,
      "learning_rate": 0.0002,
      "loss": 1.5334,
      "step": 2326
    },
    {
      "epoch": 0.8492700729927007,
      "grad_norm": 0.16123458743095398,
      "learning_rate": 0.0002,
      "loss": 1.5518,
      "step": 2327
    },
    {
      "epoch": 0.8496350364963504,
      "grad_norm": 0.15864123404026031,
      "learning_rate": 0.0002,
      "loss": 1.521,
      "step": 2328
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.18683871626853943,
      "learning_rate": 0.0002,
      "loss": 1.5382,
      "step": 2329
    },
    {
      "epoch": 0.8503649635036497,
      "grad_norm": 0.17189975082874298,
      "learning_rate": 0.0002,
      "loss": 1.5273,
      "step": 2330
    },
    {
      "epoch": 0.8507299270072993,
      "grad_norm": 0.16707713901996613,
      "learning_rate": 0.0002,
      "loss": 1.5589,
      "step": 2331
    },
    {
      "epoch": 0.8510948905109489,
      "grad_norm": 0.15353228151798248,
      "learning_rate": 0.0002,
      "loss": 1.491,
      "step": 2332
    },
    {
      "epoch": 0.8514598540145986,
      "grad_norm": 0.16234935820102692,
      "learning_rate": 0.0002,
      "loss": 1.4843,
      "step": 2333
    },
    {
      "epoch": 0.8518248175182481,
      "grad_norm": 0.1549072265625,
      "learning_rate": 0.0002,
      "loss": 1.4756,
      "step": 2334
    },
    {
      "epoch": 0.8521897810218978,
      "grad_norm": 0.18108859658241272,
      "learning_rate": 0.0002,
      "loss": 1.5495,
      "step": 2335
    },
    {
      "epoch": 0.8525547445255475,
      "grad_norm": 0.1683167964220047,
      "learning_rate": 0.0002,
      "loss": 1.4908,
      "step": 2336
    },
    {
      "epoch": 0.852919708029197,
      "grad_norm": 0.17790882289409637,
      "learning_rate": 0.0002,
      "loss": 1.5581,
      "step": 2337
    },
    {
      "epoch": 0.8532846715328467,
      "grad_norm": 0.15481743216514587,
      "learning_rate": 0.0002,
      "loss": 1.472,
      "step": 2338
    },
    {
      "epoch": 0.8536496350364964,
      "grad_norm": 0.17493991553783417,
      "learning_rate": 0.0002,
      "loss": 1.505,
      "step": 2339
    },
    {
      "epoch": 0.8540145985401459,
      "grad_norm": 0.17816923558712006,
      "learning_rate": 0.0002,
      "loss": 1.4817,
      "step": 2340
    },
    {
      "epoch": 0.8543795620437956,
      "grad_norm": 0.15632985532283783,
      "learning_rate": 0.0002,
      "loss": 1.4454,
      "step": 2341
    },
    {
      "epoch": 0.8547445255474453,
      "grad_norm": 0.16928304731845856,
      "learning_rate": 0.0002,
      "loss": 1.4714,
      "step": 2342
    },
    {
      "epoch": 0.8551094890510949,
      "grad_norm": 0.16253970563411713,
      "learning_rate": 0.0002,
      "loss": 1.5992,
      "step": 2343
    },
    {
      "epoch": 0.8554744525547445,
      "grad_norm": 0.16390997171401978,
      "learning_rate": 0.0002,
      "loss": 1.5054,
      "step": 2344
    },
    {
      "epoch": 0.8558394160583942,
      "grad_norm": 0.1627093106508255,
      "learning_rate": 0.0002,
      "loss": 1.542,
      "step": 2345
    },
    {
      "epoch": 0.8562043795620438,
      "grad_norm": 0.16612571477890015,
      "learning_rate": 0.0002,
      "loss": 1.5672,
      "step": 2346
    },
    {
      "epoch": 0.8565693430656934,
      "grad_norm": 0.1653878539800644,
      "learning_rate": 0.0002,
      "loss": 1.6359,
      "step": 2347
    },
    {
      "epoch": 0.8569343065693431,
      "grad_norm": 0.1702175885438919,
      "learning_rate": 0.0002,
      "loss": 1.5283,
      "step": 2348
    },
    {
      "epoch": 0.8572992700729927,
      "grad_norm": 0.15709930658340454,
      "learning_rate": 0.0002,
      "loss": 1.4215,
      "step": 2349
    },
    {
      "epoch": 0.8576642335766423,
      "grad_norm": 0.16643859446048737,
      "learning_rate": 0.0002,
      "loss": 1.4777,
      "step": 2350
    },
    {
      "epoch": 0.858029197080292,
      "grad_norm": 0.15706242620944977,
      "learning_rate": 0.0002,
      "loss": 1.4446,
      "step": 2351
    },
    {
      "epoch": 0.8583941605839416,
      "grad_norm": 0.15836533904075623,
      "learning_rate": 0.0002,
      "loss": 1.4585,
      "step": 2352
    },
    {
      "epoch": 0.8587591240875913,
      "grad_norm": 0.15292388200759888,
      "learning_rate": 0.0002,
      "loss": 1.408,
      "step": 2353
    },
    {
      "epoch": 0.8591240875912409,
      "grad_norm": 0.16259470582008362,
      "learning_rate": 0.0002,
      "loss": 1.4696,
      "step": 2354
    },
    {
      "epoch": 0.8594890510948905,
      "grad_norm": 0.158892422914505,
      "learning_rate": 0.0002,
      "loss": 1.4808,
      "step": 2355
    },
    {
      "epoch": 0.8598540145985402,
      "grad_norm": 0.15600568056106567,
      "learning_rate": 0.0002,
      "loss": 1.4608,
      "step": 2356
    },
    {
      "epoch": 0.8602189781021898,
      "grad_norm": 0.15368296205997467,
      "learning_rate": 0.0002,
      "loss": 1.5871,
      "step": 2357
    },
    {
      "epoch": 0.8605839416058394,
      "grad_norm": 0.15102575719356537,
      "learning_rate": 0.0002,
      "loss": 1.3856,
      "step": 2358
    },
    {
      "epoch": 0.8609489051094891,
      "grad_norm": 0.16131892800331116,
      "learning_rate": 0.0002,
      "loss": 1.4717,
      "step": 2359
    },
    {
      "epoch": 0.8613138686131386,
      "grad_norm": 0.16567210853099823,
      "learning_rate": 0.0002,
      "loss": 1.5379,
      "step": 2360
    },
    {
      "epoch": 0.8616788321167883,
      "grad_norm": 0.15733885765075684,
      "learning_rate": 0.0002,
      "loss": 1.4697,
      "step": 2361
    },
    {
      "epoch": 0.862043795620438,
      "grad_norm": 0.16568511724472046,
      "learning_rate": 0.0002,
      "loss": 1.5368,
      "step": 2362
    },
    {
      "epoch": 0.8624087591240875,
      "grad_norm": 0.17110677063465118,
      "learning_rate": 0.0002,
      "loss": 1.5885,
      "step": 2363
    },
    {
      "epoch": 0.8627737226277372,
      "grad_norm": 0.1571027785539627,
      "learning_rate": 0.0002,
      "loss": 1.4906,
      "step": 2364
    },
    {
      "epoch": 0.8631386861313869,
      "grad_norm": 0.16544929146766663,
      "learning_rate": 0.0002,
      "loss": 1.5602,
      "step": 2365
    },
    {
      "epoch": 0.8635036496350365,
      "grad_norm": 0.15627048909664154,
      "learning_rate": 0.0002,
      "loss": 1.4717,
      "step": 2366
    },
    {
      "epoch": 0.8638686131386861,
      "grad_norm": 0.16342099010944366,
      "learning_rate": 0.0002,
      "loss": 1.4659,
      "step": 2367
    },
    {
      "epoch": 0.8642335766423358,
      "grad_norm": 0.16413341462612152,
      "learning_rate": 0.0002,
      "loss": 1.4417,
      "step": 2368
    },
    {
      "epoch": 0.8645985401459854,
      "grad_norm": 0.16964882612228394,
      "learning_rate": 0.0002,
      "loss": 1.497,
      "step": 2369
    },
    {
      "epoch": 0.864963503649635,
      "grad_norm": 0.1596495509147644,
      "learning_rate": 0.0002,
      "loss": 1.4762,
      "step": 2370
    },
    {
      "epoch": 0.8653284671532847,
      "grad_norm": 0.15869098901748657,
      "learning_rate": 0.0002,
      "loss": 1.4684,
      "step": 2371
    },
    {
      "epoch": 0.8656934306569343,
      "grad_norm": 0.1614452749490738,
      "learning_rate": 0.0002,
      "loss": 1.4598,
      "step": 2372
    },
    {
      "epoch": 0.8660583941605839,
      "grad_norm": 0.16864795982837677,
      "learning_rate": 0.0002,
      "loss": 1.5255,
      "step": 2373
    },
    {
      "epoch": 0.8664233576642336,
      "grad_norm": 0.1597328633069992,
      "learning_rate": 0.0002,
      "loss": 1.5061,
      "step": 2374
    },
    {
      "epoch": 0.8667883211678832,
      "grad_norm": 0.16680952906608582,
      "learning_rate": 0.0002,
      "loss": 1.5037,
      "step": 2375
    },
    {
      "epoch": 0.8671532846715329,
      "grad_norm": 0.16517363488674164,
      "learning_rate": 0.0002,
      "loss": 1.555,
      "step": 2376
    },
    {
      "epoch": 0.8675182481751825,
      "grad_norm": 0.15717652440071106,
      "learning_rate": 0.0002,
      "loss": 1.4982,
      "step": 2377
    },
    {
      "epoch": 0.8678832116788321,
      "grad_norm": 0.15927942097187042,
      "learning_rate": 0.0002,
      "loss": 1.5347,
      "step": 2378
    },
    {
      "epoch": 0.8682481751824818,
      "grad_norm": 0.15520413219928741,
      "learning_rate": 0.0002,
      "loss": 1.5053,
      "step": 2379
    },
    {
      "epoch": 0.8686131386861314,
      "grad_norm": 0.15824003517627716,
      "learning_rate": 0.0002,
      "loss": 1.4341,
      "step": 2380
    },
    {
      "epoch": 0.868978102189781,
      "grad_norm": 0.16770531237125397,
      "learning_rate": 0.0002,
      "loss": 1.5262,
      "step": 2381
    },
    {
      "epoch": 0.8693430656934307,
      "grad_norm": 0.17245754599571228,
      "learning_rate": 0.0002,
      "loss": 1.4353,
      "step": 2382
    },
    {
      "epoch": 0.8697080291970803,
      "grad_norm": 0.1593446135520935,
      "learning_rate": 0.0002,
      "loss": 1.5106,
      "step": 2383
    },
    {
      "epoch": 0.8700729927007299,
      "grad_norm": 0.16178858280181885,
      "learning_rate": 0.0002,
      "loss": 1.4804,
      "step": 2384
    },
    {
      "epoch": 0.8704379562043796,
      "grad_norm": 0.1548338681459427,
      "learning_rate": 0.0002,
      "loss": 1.4708,
      "step": 2385
    },
    {
      "epoch": 0.8708029197080291,
      "grad_norm": 0.17035585641860962,
      "learning_rate": 0.0002,
      "loss": 1.4879,
      "step": 2386
    },
    {
      "epoch": 0.8711678832116788,
      "grad_norm": 0.16489295661449432,
      "learning_rate": 0.0002,
      "loss": 1.4442,
      "step": 2387
    },
    {
      "epoch": 0.8715328467153285,
      "grad_norm": 0.16197368502616882,
      "learning_rate": 0.0002,
      "loss": 1.4598,
      "step": 2388
    },
    {
      "epoch": 0.871897810218978,
      "grad_norm": 0.15466010570526123,
      "learning_rate": 0.0002,
      "loss": 1.45,
      "step": 2389
    },
    {
      "epoch": 0.8722627737226277,
      "grad_norm": 0.15518301725387573,
      "learning_rate": 0.0002,
      "loss": 1.4166,
      "step": 2390
    },
    {
      "epoch": 0.8726277372262774,
      "grad_norm": 0.15693876147270203,
      "learning_rate": 0.0002,
      "loss": 1.459,
      "step": 2391
    },
    {
      "epoch": 0.872992700729927,
      "grad_norm": 0.16240152716636658,
      "learning_rate": 0.0002,
      "loss": 1.5437,
      "step": 2392
    },
    {
      "epoch": 0.8733576642335766,
      "grad_norm": 0.16576439142227173,
      "learning_rate": 0.0002,
      "loss": 1.5118,
      "step": 2393
    },
    {
      "epoch": 0.8737226277372263,
      "grad_norm": 0.16430111229419708,
      "learning_rate": 0.0002,
      "loss": 1.4836,
      "step": 2394
    },
    {
      "epoch": 0.8740875912408759,
      "grad_norm": 0.16535474359989166,
      "learning_rate": 0.0002,
      "loss": 1.5536,
      "step": 2395
    },
    {
      "epoch": 0.8744525547445255,
      "grad_norm": 0.1597190499305725,
      "learning_rate": 0.0002,
      "loss": 1.4799,
      "step": 2396
    },
    {
      "epoch": 0.8748175182481752,
      "grad_norm": 0.15948303043842316,
      "learning_rate": 0.0002,
      "loss": 1.4914,
      "step": 2397
    },
    {
      "epoch": 0.8751824817518248,
      "grad_norm": 0.16773371398448944,
      "learning_rate": 0.0002,
      "loss": 1.5468,
      "step": 2398
    },
    {
      "epoch": 0.8755474452554745,
      "grad_norm": 0.16035176813602448,
      "learning_rate": 0.0002,
      "loss": 1.5933,
      "step": 2399
    },
    {
      "epoch": 0.8759124087591241,
      "grad_norm": 0.15977783501148224,
      "learning_rate": 0.0002,
      "loss": 1.5085,
      "step": 2400
    },
    {
      "epoch": 0.8762773722627737,
      "grad_norm": 0.1566234529018402,
      "learning_rate": 0.0002,
      "loss": 1.4731,
      "step": 2401
    },
    {
      "epoch": 0.8766423357664234,
      "grad_norm": 0.15607596933841705,
      "learning_rate": 0.0002,
      "loss": 1.5039,
      "step": 2402
    },
    {
      "epoch": 0.877007299270073,
      "grad_norm": 0.16120702028274536,
      "learning_rate": 0.0002,
      "loss": 1.464,
      "step": 2403
    },
    {
      "epoch": 0.8773722627737226,
      "grad_norm": 0.1574815958738327,
      "learning_rate": 0.0002,
      "loss": 1.4356,
      "step": 2404
    },
    {
      "epoch": 0.8777372262773723,
      "grad_norm": 0.1630537062883377,
      "learning_rate": 0.0002,
      "loss": 1.4417,
      "step": 2405
    },
    {
      "epoch": 0.878102189781022,
      "grad_norm": 0.1571120321750641,
      "learning_rate": 0.0002,
      "loss": 1.4714,
      "step": 2406
    },
    {
      "epoch": 0.8784671532846715,
      "grad_norm": 0.16036511957645416,
      "learning_rate": 0.0002,
      "loss": 1.4976,
      "step": 2407
    },
    {
      "epoch": 0.8788321167883212,
      "grad_norm": 0.15866465866565704,
      "learning_rate": 0.0002,
      "loss": 1.5346,
      "step": 2408
    },
    {
      "epoch": 0.8791970802919709,
      "grad_norm": 0.15605930984020233,
      "learning_rate": 0.0002,
      "loss": 1.4112,
      "step": 2409
    },
    {
      "epoch": 0.8795620437956204,
      "grad_norm": 0.16651196777820587,
      "learning_rate": 0.0002,
      "loss": 1.5193,
      "step": 2410
    },
    {
      "epoch": 0.8799270072992701,
      "grad_norm": 0.1699402928352356,
      "learning_rate": 0.0002,
      "loss": 1.5219,
      "step": 2411
    },
    {
      "epoch": 0.8802919708029197,
      "grad_norm": 0.1696481555700302,
      "learning_rate": 0.0002,
      "loss": 1.4865,
      "step": 2412
    },
    {
      "epoch": 0.8806569343065693,
      "grad_norm": 0.16886992752552032,
      "learning_rate": 0.0002,
      "loss": 1.5162,
      "step": 2413
    },
    {
      "epoch": 0.881021897810219,
      "grad_norm": 0.15679000318050385,
      "learning_rate": 0.0002,
      "loss": 1.4569,
      "step": 2414
    },
    {
      "epoch": 0.8813868613138686,
      "grad_norm": 0.15927092730998993,
      "learning_rate": 0.0002,
      "loss": 1.5124,
      "step": 2415
    },
    {
      "epoch": 0.8817518248175182,
      "grad_norm": 0.16112136840820312,
      "learning_rate": 0.0002,
      "loss": 1.5123,
      "step": 2416
    },
    {
      "epoch": 0.8821167883211679,
      "grad_norm": 0.16158726811408997,
      "learning_rate": 0.0002,
      "loss": 1.4942,
      "step": 2417
    },
    {
      "epoch": 0.8824817518248175,
      "grad_norm": 0.1650429219007492,
      "learning_rate": 0.0002,
      "loss": 1.58,
      "step": 2418
    },
    {
      "epoch": 0.8828467153284671,
      "grad_norm": 0.1628686934709549,
      "learning_rate": 0.0002,
      "loss": 1.4706,
      "step": 2419
    },
    {
      "epoch": 0.8832116788321168,
      "grad_norm": 0.16590768098831177,
      "learning_rate": 0.0002,
      "loss": 1.4466,
      "step": 2420
    },
    {
      "epoch": 0.8835766423357664,
      "grad_norm": 0.16133752465248108,
      "learning_rate": 0.0002,
      "loss": 1.5241,
      "step": 2421
    },
    {
      "epoch": 0.8839416058394161,
      "grad_norm": 0.16249948740005493,
      "learning_rate": 0.0002,
      "loss": 1.5102,
      "step": 2422
    },
    {
      "epoch": 0.8843065693430657,
      "grad_norm": 0.15699654817581177,
      "learning_rate": 0.0002,
      "loss": 1.4701,
      "step": 2423
    },
    {
      "epoch": 0.8846715328467153,
      "grad_norm": 0.1508307158946991,
      "learning_rate": 0.0002,
      "loss": 1.4631,
      "step": 2424
    },
    {
      "epoch": 0.885036496350365,
      "grad_norm": 0.15701407194137573,
      "learning_rate": 0.0002,
      "loss": 1.4308,
      "step": 2425
    },
    {
      "epoch": 0.8854014598540146,
      "grad_norm": 0.16453422605991364,
      "learning_rate": 0.0002,
      "loss": 1.5548,
      "step": 2426
    },
    {
      "epoch": 0.8857664233576642,
      "grad_norm": 0.16434268653392792,
      "learning_rate": 0.0002,
      "loss": 1.5103,
      "step": 2427
    },
    {
      "epoch": 0.8861313868613139,
      "grad_norm": 0.16782088577747345,
      "learning_rate": 0.0002,
      "loss": 1.5634,
      "step": 2428
    },
    {
      "epoch": 0.8864963503649635,
      "grad_norm": 0.15514881908893585,
      "learning_rate": 0.0002,
      "loss": 1.4986,
      "step": 2429
    },
    {
      "epoch": 0.8868613138686131,
      "grad_norm": 0.1578177958726883,
      "learning_rate": 0.0002,
      "loss": 1.5001,
      "step": 2430
    },
    {
      "epoch": 0.8872262773722628,
      "grad_norm": 0.16008447110652924,
      "learning_rate": 0.0002,
      "loss": 1.4955,
      "step": 2431
    },
    {
      "epoch": 0.8875912408759125,
      "grad_norm": 0.16505426168441772,
      "learning_rate": 0.0002,
      "loss": 1.5067,
      "step": 2432
    },
    {
      "epoch": 0.887956204379562,
      "grad_norm": 0.16591113805770874,
      "learning_rate": 0.0002,
      "loss": 1.4943,
      "step": 2433
    },
    {
      "epoch": 0.8883211678832117,
      "grad_norm": 0.16543015837669373,
      "learning_rate": 0.0002,
      "loss": 1.6112,
      "step": 2434
    },
    {
      "epoch": 0.8886861313868614,
      "grad_norm": 0.16147854924201965,
      "learning_rate": 0.0002,
      "loss": 1.5477,
      "step": 2435
    },
    {
      "epoch": 0.8890510948905109,
      "grad_norm": 0.15685655176639557,
      "learning_rate": 0.0002,
      "loss": 1.4437,
      "step": 2436
    },
    {
      "epoch": 0.8894160583941606,
      "grad_norm": 0.1576976776123047,
      "learning_rate": 0.0002,
      "loss": 1.4702,
      "step": 2437
    },
    {
      "epoch": 0.8897810218978102,
      "grad_norm": 0.16120463609695435,
      "learning_rate": 0.0002,
      "loss": 1.4903,
      "step": 2438
    },
    {
      "epoch": 0.8901459854014598,
      "grad_norm": 0.16302640736103058,
      "learning_rate": 0.0002,
      "loss": 1.4765,
      "step": 2439
    },
    {
      "epoch": 0.8905109489051095,
      "grad_norm": 0.15952904522418976,
      "learning_rate": 0.0002,
      "loss": 1.4767,
      "step": 2440
    },
    {
      "epoch": 0.8908759124087591,
      "grad_norm": 0.18589071929454803,
      "learning_rate": 0.0002,
      "loss": 1.5292,
      "step": 2441
    },
    {
      "epoch": 0.8912408759124087,
      "grad_norm": 0.16066139936447144,
      "learning_rate": 0.0002,
      "loss": 1.3947,
      "step": 2442
    },
    {
      "epoch": 0.8916058394160584,
      "grad_norm": 0.1631733477115631,
      "learning_rate": 0.0002,
      "loss": 1.5127,
      "step": 2443
    },
    {
      "epoch": 0.891970802919708,
      "grad_norm": 0.17466677725315094,
      "learning_rate": 0.0002,
      "loss": 1.4972,
      "step": 2444
    },
    {
      "epoch": 0.8923357664233577,
      "grad_norm": 0.16400554776191711,
      "learning_rate": 0.0002,
      "loss": 1.574,
      "step": 2445
    },
    {
      "epoch": 0.8927007299270073,
      "grad_norm": 0.1617157906293869,
      "learning_rate": 0.0002,
      "loss": 1.5536,
      "step": 2446
    },
    {
      "epoch": 0.8930656934306569,
      "grad_norm": 0.16294118762016296,
      "learning_rate": 0.0002,
      "loss": 1.493,
      "step": 2447
    },
    {
      "epoch": 0.8934306569343066,
      "grad_norm": 0.15469758212566376,
      "learning_rate": 0.0002,
      "loss": 1.5212,
      "step": 2448
    },
    {
      "epoch": 0.8937956204379562,
      "grad_norm": 0.15943771600723267,
      "learning_rate": 0.0002,
      "loss": 1.4621,
      "step": 2449
    },
    {
      "epoch": 0.8941605839416058,
      "grad_norm": 0.15956605970859528,
      "learning_rate": 0.0002,
      "loss": 1.5044,
      "step": 2450
    },
    {
      "epoch": 0.8945255474452555,
      "grad_norm": 0.16085542738437653,
      "learning_rate": 0.0002,
      "loss": 1.473,
      "step": 2451
    },
    {
      "epoch": 0.8948905109489051,
      "grad_norm": 0.16324806213378906,
      "learning_rate": 0.0002,
      "loss": 1.4782,
      "step": 2452
    },
    {
      "epoch": 0.8952554744525547,
      "grad_norm": 0.18025197088718414,
      "learning_rate": 0.0002,
      "loss": 1.5304,
      "step": 2453
    },
    {
      "epoch": 0.8956204379562044,
      "grad_norm": 0.16391408443450928,
      "learning_rate": 0.0002,
      "loss": 1.5221,
      "step": 2454
    },
    {
      "epoch": 0.8959854014598541,
      "grad_norm": 0.1678081452846527,
      "learning_rate": 0.0002,
      "loss": 1.4851,
      "step": 2455
    },
    {
      "epoch": 0.8963503649635036,
      "grad_norm": 0.15725000202655792,
      "learning_rate": 0.0002,
      "loss": 1.5074,
      "step": 2456
    },
    {
      "epoch": 0.8967153284671533,
      "grad_norm": 0.170259490609169,
      "learning_rate": 0.0002,
      "loss": 1.4305,
      "step": 2457
    },
    {
      "epoch": 0.897080291970803,
      "grad_norm": 0.16391447186470032,
      "learning_rate": 0.0002,
      "loss": 1.5227,
      "step": 2458
    },
    {
      "epoch": 0.8974452554744525,
      "grad_norm": 0.15701158344745636,
      "learning_rate": 0.0002,
      "loss": 1.3953,
      "step": 2459
    },
    {
      "epoch": 0.8978102189781022,
      "grad_norm": 0.16566221415996552,
      "learning_rate": 0.0002,
      "loss": 1.4976,
      "step": 2460
    },
    {
      "epoch": 0.8981751824817519,
      "grad_norm": 0.16520942747592926,
      "learning_rate": 0.0002,
      "loss": 1.5183,
      "step": 2461
    },
    {
      "epoch": 0.8985401459854014,
      "grad_norm": 0.16356204450130463,
      "learning_rate": 0.0002,
      "loss": 1.4917,
      "step": 2462
    },
    {
      "epoch": 0.8989051094890511,
      "grad_norm": 0.16227205097675323,
      "learning_rate": 0.0002,
      "loss": 1.5257,
      "step": 2463
    },
    {
      "epoch": 0.8992700729927007,
      "grad_norm": 0.16640114784240723,
      "learning_rate": 0.0002,
      "loss": 1.5532,
      "step": 2464
    },
    {
      "epoch": 0.8996350364963503,
      "grad_norm": 0.1651475429534912,
      "learning_rate": 0.0002,
      "loss": 1.5108,
      "step": 2465
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1697779893875122,
      "learning_rate": 0.0002,
      "loss": 1.5535,
      "step": 2466
    },
    {
      "epoch": 0.9003649635036496,
      "grad_norm": 0.1668228656053543,
      "learning_rate": 0.0002,
      "loss": 1.4697,
      "step": 2467
    },
    {
      "epoch": 0.9007299270072993,
      "grad_norm": 0.16157253086566925,
      "learning_rate": 0.0002,
      "loss": 1.5431,
      "step": 2468
    },
    {
      "epoch": 0.9010948905109489,
      "grad_norm": 0.1637316197156906,
      "learning_rate": 0.0002,
      "loss": 1.5438,
      "step": 2469
    },
    {
      "epoch": 0.9014598540145985,
      "grad_norm": 0.15963096916675568,
      "learning_rate": 0.0002,
      "loss": 1.4488,
      "step": 2470
    },
    {
      "epoch": 0.9018248175182482,
      "grad_norm": 0.16991889476776123,
      "learning_rate": 0.0002,
      "loss": 1.5556,
      "step": 2471
    },
    {
      "epoch": 0.9021897810218978,
      "grad_norm": 0.1639627069234848,
      "learning_rate": 0.0002,
      "loss": 1.5254,
      "step": 2472
    },
    {
      "epoch": 0.9025547445255474,
      "grad_norm": 0.1590580940246582,
      "learning_rate": 0.0002,
      "loss": 1.4551,
      "step": 2473
    },
    {
      "epoch": 0.9029197080291971,
      "grad_norm": 0.15971213579177856,
      "learning_rate": 0.0002,
      "loss": 1.421,
      "step": 2474
    },
    {
      "epoch": 0.9032846715328468,
      "grad_norm": 0.1554863154888153,
      "learning_rate": 0.0002,
      "loss": 1.4379,
      "step": 2475
    },
    {
      "epoch": 0.9036496350364963,
      "grad_norm": 0.1614917814731598,
      "learning_rate": 0.0002,
      "loss": 1.5094,
      "step": 2476
    },
    {
      "epoch": 0.904014598540146,
      "grad_norm": 0.1596604436635971,
      "learning_rate": 0.0002,
      "loss": 1.4696,
      "step": 2477
    },
    {
      "epoch": 0.9043795620437957,
      "grad_norm": 0.15971754491329193,
      "learning_rate": 0.0002,
      "loss": 1.5007,
      "step": 2478
    },
    {
      "epoch": 0.9047445255474452,
      "grad_norm": 0.15824519097805023,
      "learning_rate": 0.0002,
      "loss": 1.4506,
      "step": 2479
    },
    {
      "epoch": 0.9051094890510949,
      "grad_norm": 0.1628393679857254,
      "learning_rate": 0.0002,
      "loss": 1.4848,
      "step": 2480
    },
    {
      "epoch": 0.9054744525547446,
      "grad_norm": 0.17192164063453674,
      "learning_rate": 0.0002,
      "loss": 1.5108,
      "step": 2481
    },
    {
      "epoch": 0.9058394160583941,
      "grad_norm": 0.1555435061454773,
      "learning_rate": 0.0002,
      "loss": 1.473,
      "step": 2482
    },
    {
      "epoch": 0.9062043795620438,
      "grad_norm": 0.16192342340946198,
      "learning_rate": 0.0002,
      "loss": 1.4454,
      "step": 2483
    },
    {
      "epoch": 0.9065693430656935,
      "grad_norm": 0.16099943220615387,
      "learning_rate": 0.0002,
      "loss": 1.5092,
      "step": 2484
    },
    {
      "epoch": 0.906934306569343,
      "grad_norm": 0.16427081823349,
      "learning_rate": 0.0002,
      "loss": 1.5686,
      "step": 2485
    },
    {
      "epoch": 0.9072992700729927,
      "grad_norm": 0.15842734277248383,
      "learning_rate": 0.0002,
      "loss": 1.4352,
      "step": 2486
    },
    {
      "epoch": 0.9076642335766424,
      "grad_norm": 0.16636206209659576,
      "learning_rate": 0.0002,
      "loss": 1.4777,
      "step": 2487
    },
    {
      "epoch": 0.908029197080292,
      "grad_norm": 0.160578653216362,
      "learning_rate": 0.0002,
      "loss": 1.4749,
      "step": 2488
    },
    {
      "epoch": 0.9083941605839416,
      "grad_norm": 0.16006852686405182,
      "learning_rate": 0.0002,
      "loss": 1.5228,
      "step": 2489
    },
    {
      "epoch": 0.9087591240875912,
      "grad_norm": 0.16462931036949158,
      "learning_rate": 0.0002,
      "loss": 1.501,
      "step": 2490
    },
    {
      "epoch": 0.9091240875912409,
      "grad_norm": 0.16356563568115234,
      "learning_rate": 0.0002,
      "loss": 1.447,
      "step": 2491
    },
    {
      "epoch": 0.9094890510948905,
      "grad_norm": 0.16262589395046234,
      "learning_rate": 0.0002,
      "loss": 1.4455,
      "step": 2492
    },
    {
      "epoch": 0.9098540145985401,
      "grad_norm": 0.17037273943424225,
      "learning_rate": 0.0002,
      "loss": 1.4763,
      "step": 2493
    },
    {
      "epoch": 0.9102189781021898,
      "grad_norm": 0.15950211882591248,
      "learning_rate": 0.0002,
      "loss": 1.4605,
      "step": 2494
    },
    {
      "epoch": 0.9105839416058394,
      "grad_norm": 0.16635762155056,
      "learning_rate": 0.0002,
      "loss": 1.4702,
      "step": 2495
    },
    {
      "epoch": 0.910948905109489,
      "grad_norm": 0.16502083837985992,
      "learning_rate": 0.0002,
      "loss": 1.5365,
      "step": 2496
    },
    {
      "epoch": 0.9113138686131387,
      "grad_norm": 0.16077646613121033,
      "learning_rate": 0.0002,
      "loss": 1.5172,
      "step": 2497
    },
    {
      "epoch": 0.9116788321167884,
      "grad_norm": 0.16205154359340668,
      "learning_rate": 0.0002,
      "loss": 1.5649,
      "step": 2498
    },
    {
      "epoch": 0.9120437956204379,
      "grad_norm": 0.15633723139762878,
      "learning_rate": 0.0002,
      "loss": 1.4023,
      "step": 2499
    },
    {
      "epoch": 0.9124087591240876,
      "grad_norm": 0.17277608811855316,
      "learning_rate": 0.0002,
      "loss": 1.5251,
      "step": 2500
    },
    {
      "epoch": 0.9127737226277373,
      "grad_norm": 0.16403469443321228,
      "learning_rate": 0.0002,
      "loss": 1.4659,
      "step": 2501
    },
    {
      "epoch": 0.9131386861313868,
      "grad_norm": 0.16576442122459412,
      "learning_rate": 0.0002,
      "loss": 1.5217,
      "step": 2502
    },
    {
      "epoch": 0.9135036496350365,
      "grad_norm": 0.16019345819950104,
      "learning_rate": 0.0002,
      "loss": 1.4377,
      "step": 2503
    },
    {
      "epoch": 0.9138686131386862,
      "grad_norm": 0.165249302983284,
      "learning_rate": 0.0002,
      "loss": 1.5209,
      "step": 2504
    },
    {
      "epoch": 0.9142335766423357,
      "grad_norm": 0.17250987887382507,
      "learning_rate": 0.0002,
      "loss": 1.4985,
      "step": 2505
    },
    {
      "epoch": 0.9145985401459854,
      "grad_norm": 0.16643929481506348,
      "learning_rate": 0.0002,
      "loss": 1.5145,
      "step": 2506
    },
    {
      "epoch": 0.9149635036496351,
      "grad_norm": 0.16755516827106476,
      "learning_rate": 0.0002,
      "loss": 1.556,
      "step": 2507
    },
    {
      "epoch": 0.9153284671532846,
      "grad_norm": 0.1536182314157486,
      "learning_rate": 0.0002,
      "loss": 1.4546,
      "step": 2508
    },
    {
      "epoch": 0.9156934306569343,
      "grad_norm": 0.16669775545597076,
      "learning_rate": 0.0002,
      "loss": 1.4642,
      "step": 2509
    },
    {
      "epoch": 0.916058394160584,
      "grad_norm": 0.1576424092054367,
      "learning_rate": 0.0002,
      "loss": 1.456,
      "step": 2510
    },
    {
      "epoch": 0.9164233576642336,
      "grad_norm": 0.16155290603637695,
      "learning_rate": 0.0002,
      "loss": 1.4959,
      "step": 2511
    },
    {
      "epoch": 0.9167883211678832,
      "grad_norm": 0.16688187420368195,
      "learning_rate": 0.0002,
      "loss": 1.5381,
      "step": 2512
    },
    {
      "epoch": 0.9171532846715329,
      "grad_norm": 0.1873917132616043,
      "learning_rate": 0.0002,
      "loss": 1.4249,
      "step": 2513
    },
    {
      "epoch": 0.9175182481751825,
      "grad_norm": 0.15992550551891327,
      "learning_rate": 0.0002,
      "loss": 1.5189,
      "step": 2514
    },
    {
      "epoch": 0.9178832116788321,
      "grad_norm": 0.17123663425445557,
      "learning_rate": 0.0002,
      "loss": 1.5404,
      "step": 2515
    },
    {
      "epoch": 0.9182481751824818,
      "grad_norm": 0.16687163710594177,
      "learning_rate": 0.0002,
      "loss": 1.4898,
      "step": 2516
    },
    {
      "epoch": 0.9186131386861314,
      "grad_norm": 0.1627236008644104,
      "learning_rate": 0.0002,
      "loss": 1.5171,
      "step": 2517
    },
    {
      "epoch": 0.918978102189781,
      "grad_norm": 0.17087994515895844,
      "learning_rate": 0.0002,
      "loss": 1.554,
      "step": 2518
    },
    {
      "epoch": 0.9193430656934306,
      "grad_norm": 0.17231978476047516,
      "learning_rate": 0.0002,
      "loss": 1.4426,
      "step": 2519
    },
    {
      "epoch": 0.9197080291970803,
      "grad_norm": 0.16648684442043304,
      "learning_rate": 0.0002,
      "loss": 1.4857,
      "step": 2520
    },
    {
      "epoch": 0.92007299270073,
      "grad_norm": 0.16370485723018646,
      "learning_rate": 0.0002,
      "loss": 1.409,
      "step": 2521
    },
    {
      "epoch": 0.9204379562043795,
      "grad_norm": 0.17218996584415436,
      "learning_rate": 0.0002,
      "loss": 1.5112,
      "step": 2522
    },
    {
      "epoch": 0.9208029197080292,
      "grad_norm": 0.17276902496814728,
      "learning_rate": 0.0002,
      "loss": 1.4788,
      "step": 2523
    },
    {
      "epoch": 0.9211678832116789,
      "grad_norm": 0.16965170204639435,
      "learning_rate": 0.0002,
      "loss": 1.5196,
      "step": 2524
    },
    {
      "epoch": 0.9215328467153284,
      "grad_norm": 0.16335782408714294,
      "learning_rate": 0.0002,
      "loss": 1.5055,
      "step": 2525
    },
    {
      "epoch": 0.9218978102189781,
      "grad_norm": 0.17167262732982635,
      "learning_rate": 0.0002,
      "loss": 1.4679,
      "step": 2526
    },
    {
      "epoch": 0.9222627737226278,
      "grad_norm": 0.16349290311336517,
      "learning_rate": 0.0002,
      "loss": 1.4153,
      "step": 2527
    },
    {
      "epoch": 0.9226277372262773,
      "grad_norm": 0.17494849860668182,
      "learning_rate": 0.0002,
      "loss": 1.4923,
      "step": 2528
    },
    {
      "epoch": 0.922992700729927,
      "grad_norm": 0.17214685678482056,
      "learning_rate": 0.0002,
      "loss": 1.5199,
      "step": 2529
    },
    {
      "epoch": 0.9233576642335767,
      "grad_norm": 0.15945856273174286,
      "learning_rate": 0.0002,
      "loss": 1.464,
      "step": 2530
    },
    {
      "epoch": 0.9237226277372262,
      "grad_norm": 0.16745683550834656,
      "learning_rate": 0.0002,
      "loss": 1.5107,
      "step": 2531
    },
    {
      "epoch": 0.9240875912408759,
      "grad_norm": 0.16383004188537598,
      "learning_rate": 0.0002,
      "loss": 1.5324,
      "step": 2532
    },
    {
      "epoch": 0.9244525547445256,
      "grad_norm": 0.15960979461669922,
      "learning_rate": 0.0002,
      "loss": 1.4612,
      "step": 2533
    },
    {
      "epoch": 0.9248175182481752,
      "grad_norm": 0.1688227653503418,
      "learning_rate": 0.0002,
      "loss": 1.5234,
      "step": 2534
    },
    {
      "epoch": 0.9251824817518248,
      "grad_norm": 0.16016429662704468,
      "learning_rate": 0.0002,
      "loss": 1.4738,
      "step": 2535
    },
    {
      "epoch": 0.9255474452554745,
      "grad_norm": 0.16196270287036896,
      "learning_rate": 0.0002,
      "loss": 1.5224,
      "step": 2536
    },
    {
      "epoch": 0.9259124087591241,
      "grad_norm": 0.16171318292617798,
      "learning_rate": 0.0002,
      "loss": 1.4957,
      "step": 2537
    },
    {
      "epoch": 0.9262773722627737,
      "grad_norm": 0.16211725771427155,
      "learning_rate": 0.0002,
      "loss": 1.4493,
      "step": 2538
    },
    {
      "epoch": 0.9266423357664234,
      "grad_norm": 0.16216647624969482,
      "learning_rate": 0.0002,
      "loss": 1.4601,
      "step": 2539
    },
    {
      "epoch": 0.927007299270073,
      "grad_norm": 0.161648228764534,
      "learning_rate": 0.0002,
      "loss": 1.4957,
      "step": 2540
    },
    {
      "epoch": 0.9273722627737226,
      "grad_norm": 0.16507236659526825,
      "learning_rate": 0.0002,
      "loss": 1.5298,
      "step": 2541
    },
    {
      "epoch": 0.9277372262773723,
      "grad_norm": 0.16315415501594543,
      "learning_rate": 0.0002,
      "loss": 1.5272,
      "step": 2542
    },
    {
      "epoch": 0.9281021897810219,
      "grad_norm": 0.16288232803344727,
      "learning_rate": 0.0002,
      "loss": 1.5026,
      "step": 2543
    },
    {
      "epoch": 0.9284671532846716,
      "grad_norm": 0.16895364224910736,
      "learning_rate": 0.0002,
      "loss": 1.5397,
      "step": 2544
    },
    {
      "epoch": 0.9288321167883211,
      "grad_norm": 0.16603031754493713,
      "learning_rate": 0.0002,
      "loss": 1.5058,
      "step": 2545
    },
    {
      "epoch": 0.9291970802919708,
      "grad_norm": 0.1692887842655182,
      "learning_rate": 0.0002,
      "loss": 1.523,
      "step": 2546
    },
    {
      "epoch": 0.9295620437956205,
      "grad_norm": 0.1594434678554535,
      "learning_rate": 0.0002,
      "loss": 1.5393,
      "step": 2547
    },
    {
      "epoch": 0.92992700729927,
      "grad_norm": 0.17692303657531738,
      "learning_rate": 0.0002,
      "loss": 1.6183,
      "step": 2548
    },
    {
      "epoch": 0.9302919708029197,
      "grad_norm": 0.15984998643398285,
      "learning_rate": 0.0002,
      "loss": 1.5493,
      "step": 2549
    },
    {
      "epoch": 0.9306569343065694,
      "grad_norm": 0.15916268527507782,
      "learning_rate": 0.0002,
      "loss": 1.4571,
      "step": 2550
    },
    {
      "epoch": 0.9310218978102189,
      "grad_norm": 0.16497759521007538,
      "learning_rate": 0.0002,
      "loss": 1.51,
      "step": 2551
    },
    {
      "epoch": 0.9313868613138686,
      "grad_norm": 0.16636699438095093,
      "learning_rate": 0.0002,
      "loss": 1.4919,
      "step": 2552
    },
    {
      "epoch": 0.9317518248175183,
      "grad_norm": 0.16724666953086853,
      "learning_rate": 0.0002,
      "loss": 1.522,
      "step": 2553
    },
    {
      "epoch": 0.9321167883211678,
      "grad_norm": 0.17091615498065948,
      "learning_rate": 0.0002,
      "loss": 1.4944,
      "step": 2554
    },
    {
      "epoch": 0.9324817518248175,
      "grad_norm": 0.1595361828804016,
      "learning_rate": 0.0002,
      "loss": 1.5544,
      "step": 2555
    },
    {
      "epoch": 0.9328467153284672,
      "grad_norm": 0.15928024053573608,
      "learning_rate": 0.0002,
      "loss": 1.4395,
      "step": 2556
    },
    {
      "epoch": 0.9332116788321168,
      "grad_norm": 0.1656642109155655,
      "learning_rate": 0.0002,
      "loss": 1.4986,
      "step": 2557
    },
    {
      "epoch": 0.9335766423357664,
      "grad_norm": 0.16868048906326294,
      "learning_rate": 0.0002,
      "loss": 1.4799,
      "step": 2558
    },
    {
      "epoch": 0.9339416058394161,
      "grad_norm": 0.16069982945919037,
      "learning_rate": 0.0002,
      "loss": 1.5588,
      "step": 2559
    },
    {
      "epoch": 0.9343065693430657,
      "grad_norm": 0.16057930886745453,
      "learning_rate": 0.0002,
      "loss": 1.4843,
      "step": 2560
    },
    {
      "epoch": 0.9346715328467153,
      "grad_norm": 0.1591300517320633,
      "learning_rate": 0.0002,
      "loss": 1.4184,
      "step": 2561
    },
    {
      "epoch": 0.935036496350365,
      "grad_norm": 0.15567919611930847,
      "learning_rate": 0.0002,
      "loss": 1.5087,
      "step": 2562
    },
    {
      "epoch": 0.9354014598540146,
      "grad_norm": 0.1615423858165741,
      "learning_rate": 0.0002,
      "loss": 1.49,
      "step": 2563
    },
    {
      "epoch": 0.9357664233576642,
      "grad_norm": 0.16434410214424133,
      "learning_rate": 0.0002,
      "loss": 1.4916,
      "step": 2564
    },
    {
      "epoch": 0.9361313868613139,
      "grad_norm": 0.165253683924675,
      "learning_rate": 0.0002,
      "loss": 1.5294,
      "step": 2565
    },
    {
      "epoch": 0.9364963503649635,
      "grad_norm": 0.16211415827274323,
      "learning_rate": 0.0002,
      "loss": 1.5153,
      "step": 2566
    },
    {
      "epoch": 0.9368613138686132,
      "grad_norm": 0.15923842787742615,
      "learning_rate": 0.0002,
      "loss": 1.482,
      "step": 2567
    },
    {
      "epoch": 0.9372262773722628,
      "grad_norm": 0.1640588790178299,
      "learning_rate": 0.0002,
      "loss": 1.4331,
      "step": 2568
    },
    {
      "epoch": 0.9375912408759124,
      "grad_norm": 0.16685856878757477,
      "learning_rate": 0.0002,
      "loss": 1.5202,
      "step": 2569
    },
    {
      "epoch": 0.9379562043795621,
      "grad_norm": 0.16382430493831635,
      "learning_rate": 0.0002,
      "loss": 1.478,
      "step": 2570
    },
    {
      "epoch": 0.9383211678832116,
      "grad_norm": 0.16192719340324402,
      "learning_rate": 0.0002,
      "loss": 1.4988,
      "step": 2571
    },
    {
      "epoch": 0.9386861313868613,
      "grad_norm": 0.16881732642650604,
      "learning_rate": 0.0002,
      "loss": 1.5006,
      "step": 2572
    },
    {
      "epoch": 0.939051094890511,
      "grad_norm": 0.15907448530197144,
      "learning_rate": 0.0002,
      "loss": 1.4647,
      "step": 2573
    },
    {
      "epoch": 0.9394160583941605,
      "grad_norm": 0.17061035335063934,
      "learning_rate": 0.0002,
      "loss": 1.5018,
      "step": 2574
    },
    {
      "epoch": 0.9397810218978102,
      "grad_norm": 0.16687995195388794,
      "learning_rate": 0.0002,
      "loss": 1.4432,
      "step": 2575
    },
    {
      "epoch": 0.9401459854014599,
      "grad_norm": 0.1617455631494522,
      "learning_rate": 0.0002,
      "loss": 1.5253,
      "step": 2576
    },
    {
      "epoch": 0.9405109489051094,
      "grad_norm": 0.18492820858955383,
      "learning_rate": 0.0002,
      "loss": 1.4625,
      "step": 2577
    },
    {
      "epoch": 0.9408759124087591,
      "grad_norm": 0.15722785890102386,
      "learning_rate": 0.0002,
      "loss": 1.438,
      "step": 2578
    },
    {
      "epoch": 0.9412408759124088,
      "grad_norm": 0.1609824001789093,
      "learning_rate": 0.0002,
      "loss": 1.4819,
      "step": 2579
    },
    {
      "epoch": 0.9416058394160584,
      "grad_norm": 0.17463000118732452,
      "learning_rate": 0.0002,
      "loss": 1.4889,
      "step": 2580
    },
    {
      "epoch": 0.941970802919708,
      "grad_norm": 0.16820397973060608,
      "learning_rate": 0.0002,
      "loss": 1.5131,
      "step": 2581
    },
    {
      "epoch": 0.9423357664233577,
      "grad_norm": 0.1700102835893631,
      "learning_rate": 0.0002,
      "loss": 1.4598,
      "step": 2582
    },
    {
      "epoch": 0.9427007299270073,
      "grad_norm": 0.1668480932712555,
      "learning_rate": 0.0002,
      "loss": 1.5164,
      "step": 2583
    },
    {
      "epoch": 0.9430656934306569,
      "grad_norm": 0.17528818547725677,
      "learning_rate": 0.0002,
      "loss": 1.5232,
      "step": 2584
    },
    {
      "epoch": 0.9434306569343066,
      "grad_norm": 0.1635499894618988,
      "learning_rate": 0.0002,
      "loss": 1.5361,
      "step": 2585
    },
    {
      "epoch": 0.9437956204379562,
      "grad_norm": 0.16383320093154907,
      "learning_rate": 0.0002,
      "loss": 1.5263,
      "step": 2586
    },
    {
      "epoch": 0.9441605839416058,
      "grad_norm": 0.16177332401275635,
      "learning_rate": 0.0002,
      "loss": 1.4831,
      "step": 2587
    },
    {
      "epoch": 0.9445255474452555,
      "grad_norm": 0.16188402473926544,
      "learning_rate": 0.0002,
      "loss": 1.5133,
      "step": 2588
    },
    {
      "epoch": 0.9448905109489051,
      "grad_norm": 0.18248501420021057,
      "learning_rate": 0.0002,
      "loss": 1.5307,
      "step": 2589
    },
    {
      "epoch": 0.9452554744525548,
      "grad_norm": 0.166423499584198,
      "learning_rate": 0.0002,
      "loss": 1.5164,
      "step": 2590
    },
    {
      "epoch": 0.9456204379562044,
      "grad_norm": 0.1664702594280243,
      "learning_rate": 0.0002,
      "loss": 1.4529,
      "step": 2591
    },
    {
      "epoch": 0.945985401459854,
      "grad_norm": 0.17995983362197876,
      "learning_rate": 0.0002,
      "loss": 1.5417,
      "step": 2592
    },
    {
      "epoch": 0.9463503649635037,
      "grad_norm": 0.169575035572052,
      "learning_rate": 0.0002,
      "loss": 1.4392,
      "step": 2593
    },
    {
      "epoch": 0.9467153284671533,
      "grad_norm": 0.19120202958583832,
      "learning_rate": 0.0002,
      "loss": 1.515,
      "step": 2594
    },
    {
      "epoch": 0.9470802919708029,
      "grad_norm": 0.16946770250797272,
      "learning_rate": 0.0002,
      "loss": 1.5537,
      "step": 2595
    },
    {
      "epoch": 0.9474452554744526,
      "grad_norm": 0.18357348442077637,
      "learning_rate": 0.0002,
      "loss": 1.4875,
      "step": 2596
    },
    {
      "epoch": 0.9478102189781021,
      "grad_norm": 0.15908583998680115,
      "learning_rate": 0.0002,
      "loss": 1.4533,
      "step": 2597
    },
    {
      "epoch": 0.9481751824817518,
      "grad_norm": 0.17905563116073608,
      "learning_rate": 0.0002,
      "loss": 1.5205,
      "step": 2598
    },
    {
      "epoch": 0.9485401459854015,
      "grad_norm": 0.16896726191043854,
      "learning_rate": 0.0002,
      "loss": 1.5488,
      "step": 2599
    },
    {
      "epoch": 0.948905109489051,
      "grad_norm": 0.17171034216880798,
      "learning_rate": 0.0002,
      "loss": 1.5384,
      "step": 2600
    },
    {
      "epoch": 0.9492700729927007,
      "grad_norm": 0.16640429198741913,
      "learning_rate": 0.0002,
      "loss": 1.4567,
      "step": 2601
    },
    {
      "epoch": 0.9496350364963504,
      "grad_norm": 0.16722437739372253,
      "learning_rate": 0.0002,
      "loss": 1.427,
      "step": 2602
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.15880778431892395,
      "learning_rate": 0.0002,
      "loss": 1.4254,
      "step": 2603
    },
    {
      "epoch": 0.9503649635036496,
      "grad_norm": 0.16618353128433228,
      "learning_rate": 0.0002,
      "loss": 1.6337,
      "step": 2604
    },
    {
      "epoch": 0.9507299270072993,
      "grad_norm": 0.1587599664926529,
      "learning_rate": 0.0002,
      "loss": 1.4516,
      "step": 2605
    },
    {
      "epoch": 0.9510948905109489,
      "grad_norm": 0.15957415103912354,
      "learning_rate": 0.0002,
      "loss": 1.4291,
      "step": 2606
    },
    {
      "epoch": 0.9514598540145985,
      "grad_norm": 0.1599472314119339,
      "learning_rate": 0.0002,
      "loss": 1.4468,
      "step": 2607
    },
    {
      "epoch": 0.9518248175182482,
      "grad_norm": 0.16853396594524384,
      "learning_rate": 0.0002,
      "loss": 1.5323,
      "step": 2608
    },
    {
      "epoch": 0.9521897810218978,
      "grad_norm": 0.17310065031051636,
      "learning_rate": 0.0002,
      "loss": 1.5523,
      "step": 2609
    },
    {
      "epoch": 0.9525547445255474,
      "grad_norm": 0.1671569049358368,
      "learning_rate": 0.0002,
      "loss": 1.4331,
      "step": 2610
    },
    {
      "epoch": 0.9529197080291971,
      "grad_norm": 0.16725647449493408,
      "learning_rate": 0.0002,
      "loss": 1.5397,
      "step": 2611
    },
    {
      "epoch": 0.9532846715328467,
      "grad_norm": 0.15943627059459686,
      "learning_rate": 0.0002,
      "loss": 1.4083,
      "step": 2612
    },
    {
      "epoch": 0.9536496350364964,
      "grad_norm": 0.16376885771751404,
      "learning_rate": 0.0002,
      "loss": 1.4767,
      "step": 2613
    },
    {
      "epoch": 0.954014598540146,
      "grad_norm": 0.17040078341960907,
      "learning_rate": 0.0002,
      "loss": 1.4865,
      "step": 2614
    },
    {
      "epoch": 0.9543795620437956,
      "grad_norm": 0.1637021005153656,
      "learning_rate": 0.0002,
      "loss": 1.4068,
      "step": 2615
    },
    {
      "epoch": 0.9547445255474453,
      "grad_norm": 0.16056060791015625,
      "learning_rate": 0.0002,
      "loss": 1.5061,
      "step": 2616
    },
    {
      "epoch": 0.9551094890510949,
      "grad_norm": 0.16133427619934082,
      "learning_rate": 0.0002,
      "loss": 1.4427,
      "step": 2617
    },
    {
      "epoch": 0.9554744525547445,
      "grad_norm": 0.1605280190706253,
      "learning_rate": 0.0002,
      "loss": 1.4236,
      "step": 2618
    },
    {
      "epoch": 0.9558394160583942,
      "grad_norm": 0.17509835958480835,
      "learning_rate": 0.0002,
      "loss": 1.4411,
      "step": 2619
    },
    {
      "epoch": 0.9562043795620438,
      "grad_norm": 0.18138298392295837,
      "learning_rate": 0.0002,
      "loss": 1.5523,
      "step": 2620
    },
    {
      "epoch": 0.9565693430656934,
      "grad_norm": 0.1543637216091156,
      "learning_rate": 0.0002,
      "loss": 1.3976,
      "step": 2621
    },
    {
      "epoch": 0.9569343065693431,
      "grad_norm": 0.16988366842269897,
      "learning_rate": 0.0002,
      "loss": 1.4587,
      "step": 2622
    },
    {
      "epoch": 0.9572992700729926,
      "grad_norm": 0.15944379568099976,
      "learning_rate": 0.0002,
      "loss": 1.4813,
      "step": 2623
    },
    {
      "epoch": 0.9576642335766423,
      "grad_norm": 0.1653096228837967,
      "learning_rate": 0.0002,
      "loss": 1.4705,
      "step": 2624
    },
    {
      "epoch": 0.958029197080292,
      "grad_norm": 0.16263969242572784,
      "learning_rate": 0.0002,
      "loss": 1.4869,
      "step": 2625
    },
    {
      "epoch": 0.9583941605839416,
      "grad_norm": 0.16910004615783691,
      "learning_rate": 0.0002,
      "loss": 1.5067,
      "step": 2626
    },
    {
      "epoch": 0.9587591240875912,
      "grad_norm": 0.15625591576099396,
      "learning_rate": 0.0002,
      "loss": 1.4941,
      "step": 2627
    },
    {
      "epoch": 0.9591240875912409,
      "grad_norm": 0.15989892184734344,
      "learning_rate": 0.0002,
      "loss": 1.4092,
      "step": 2628
    },
    {
      "epoch": 0.9594890510948905,
      "grad_norm": 0.1691388636827469,
      "learning_rate": 0.0002,
      "loss": 1.4904,
      "step": 2629
    },
    {
      "epoch": 0.9598540145985401,
      "grad_norm": 0.1743697226047516,
      "learning_rate": 0.0002,
      "loss": 1.4594,
      "step": 2630
    },
    {
      "epoch": 0.9602189781021898,
      "grad_norm": 0.16778066754341125,
      "learning_rate": 0.0002,
      "loss": 1.5173,
      "step": 2631
    },
    {
      "epoch": 0.9605839416058394,
      "grad_norm": 0.1733938455581665,
      "learning_rate": 0.0002,
      "loss": 1.5339,
      "step": 2632
    },
    {
      "epoch": 0.960948905109489,
      "grad_norm": 0.1595926731824875,
      "learning_rate": 0.0002,
      "loss": 1.4374,
      "step": 2633
    },
    {
      "epoch": 0.9613138686131387,
      "grad_norm": 0.16470901668071747,
      "learning_rate": 0.0002,
      "loss": 1.5396,
      "step": 2634
    },
    {
      "epoch": 0.9616788321167883,
      "grad_norm": 0.1768956184387207,
      "learning_rate": 0.0002,
      "loss": 1.5701,
      "step": 2635
    },
    {
      "epoch": 0.962043795620438,
      "grad_norm": 0.1660374253988266,
      "learning_rate": 0.0002,
      "loss": 1.5033,
      "step": 2636
    },
    {
      "epoch": 0.9624087591240876,
      "grad_norm": 0.17001494765281677,
      "learning_rate": 0.0002,
      "loss": 1.5665,
      "step": 2637
    },
    {
      "epoch": 0.9627737226277372,
      "grad_norm": 0.1607896089553833,
      "learning_rate": 0.0002,
      "loss": 1.4712,
      "step": 2638
    },
    {
      "epoch": 0.9631386861313869,
      "grad_norm": 0.16273950040340424,
      "learning_rate": 0.0002,
      "loss": 1.4793,
      "step": 2639
    },
    {
      "epoch": 0.9635036496350365,
      "grad_norm": 0.16146887838840485,
      "learning_rate": 0.0002,
      "loss": 1.4662,
      "step": 2640
    },
    {
      "epoch": 0.9638686131386861,
      "grad_norm": 0.16855330765247345,
      "learning_rate": 0.0002,
      "loss": 1.5473,
      "step": 2641
    },
    {
      "epoch": 0.9642335766423358,
      "grad_norm": 0.16091258823871613,
      "learning_rate": 0.0002,
      "loss": 1.5322,
      "step": 2642
    },
    {
      "epoch": 0.9645985401459855,
      "grad_norm": 0.17266736924648285,
      "learning_rate": 0.0002,
      "loss": 1.5465,
      "step": 2643
    },
    {
      "epoch": 0.964963503649635,
      "grad_norm": 0.16533157229423523,
      "learning_rate": 0.0002,
      "loss": 1.4602,
      "step": 2644
    },
    {
      "epoch": 0.9653284671532847,
      "grad_norm": 0.1568583995103836,
      "learning_rate": 0.0002,
      "loss": 1.3778,
      "step": 2645
    },
    {
      "epoch": 0.9656934306569344,
      "grad_norm": 0.16645345091819763,
      "learning_rate": 0.0002,
      "loss": 1.4702,
      "step": 2646
    },
    {
      "epoch": 0.9660583941605839,
      "grad_norm": 0.16603653132915497,
      "learning_rate": 0.0002,
      "loss": 1.5247,
      "step": 2647
    },
    {
      "epoch": 0.9664233576642336,
      "grad_norm": 0.1663249284029007,
      "learning_rate": 0.0002,
      "loss": 1.4781,
      "step": 2648
    },
    {
      "epoch": 0.9667883211678832,
      "grad_norm": 0.1706889569759369,
      "learning_rate": 0.0002,
      "loss": 1.4701,
      "step": 2649
    },
    {
      "epoch": 0.9671532846715328,
      "grad_norm": 0.16227003931999207,
      "learning_rate": 0.0002,
      "loss": 1.5177,
      "step": 2650
    },
    {
      "epoch": 0.9675182481751825,
      "grad_norm": 0.16958026587963104,
      "learning_rate": 0.0002,
      "loss": 1.5189,
      "step": 2651
    },
    {
      "epoch": 0.9678832116788321,
      "grad_norm": 0.17030663788318634,
      "learning_rate": 0.0002,
      "loss": 1.4489,
      "step": 2652
    },
    {
      "epoch": 0.9682481751824817,
      "grad_norm": 0.16460777819156647,
      "learning_rate": 0.0002,
      "loss": 1.4259,
      "step": 2653
    },
    {
      "epoch": 0.9686131386861314,
      "grad_norm": 0.16712139546871185,
      "learning_rate": 0.0002,
      "loss": 1.47,
      "step": 2654
    },
    {
      "epoch": 0.968978102189781,
      "grad_norm": 0.16881325840950012,
      "learning_rate": 0.0002,
      "loss": 1.4292,
      "step": 2655
    },
    {
      "epoch": 0.9693430656934306,
      "grad_norm": 0.16799800097942352,
      "learning_rate": 0.0002,
      "loss": 1.5288,
      "step": 2656
    },
    {
      "epoch": 0.9697080291970803,
      "grad_norm": 0.16528736054897308,
      "learning_rate": 0.0002,
      "loss": 1.5415,
      "step": 2657
    },
    {
      "epoch": 0.9700729927007299,
      "grad_norm": 0.1646011918783188,
      "learning_rate": 0.0002,
      "loss": 1.4808,
      "step": 2658
    },
    {
      "epoch": 0.9704379562043796,
      "grad_norm": 0.16652622818946838,
      "learning_rate": 0.0002,
      "loss": 1.4224,
      "step": 2659
    },
    {
      "epoch": 0.9708029197080292,
      "grad_norm": 0.16427253186702728,
      "learning_rate": 0.0002,
      "loss": 1.5033,
      "step": 2660
    },
    {
      "epoch": 0.9711678832116788,
      "grad_norm": 0.16977131366729736,
      "learning_rate": 0.0002,
      "loss": 1.5314,
      "step": 2661
    },
    {
      "epoch": 0.9715328467153285,
      "grad_norm": 0.1619565188884735,
      "learning_rate": 0.0002,
      "loss": 1.4671,
      "step": 2662
    },
    {
      "epoch": 0.9718978102189781,
      "grad_norm": 0.1577938199043274,
      "learning_rate": 0.0002,
      "loss": 1.4619,
      "step": 2663
    },
    {
      "epoch": 0.9722627737226277,
      "grad_norm": 0.15977534651756287,
      "learning_rate": 0.0002,
      "loss": 1.484,
      "step": 2664
    },
    {
      "epoch": 0.9726277372262774,
      "grad_norm": 0.16409330070018768,
      "learning_rate": 0.0002,
      "loss": 1.5189,
      "step": 2665
    },
    {
      "epoch": 0.972992700729927,
      "grad_norm": 0.15828025341033936,
      "learning_rate": 0.0002,
      "loss": 1.4263,
      "step": 2666
    },
    {
      "epoch": 0.9733576642335766,
      "grad_norm": 0.1650678515434265,
      "learning_rate": 0.0002,
      "loss": 1.5047,
      "step": 2667
    },
    {
      "epoch": 0.9737226277372263,
      "grad_norm": 0.16040046513080597,
      "learning_rate": 0.0002,
      "loss": 1.4798,
      "step": 2668
    },
    {
      "epoch": 0.974087591240876,
      "grad_norm": 0.162154883146286,
      "learning_rate": 0.0002,
      "loss": 1.503,
      "step": 2669
    },
    {
      "epoch": 0.9744525547445255,
      "grad_norm": 0.15932995080947876,
      "learning_rate": 0.0002,
      "loss": 1.4793,
      "step": 2670
    },
    {
      "epoch": 0.9748175182481752,
      "grad_norm": 0.16092489659786224,
      "learning_rate": 0.0002,
      "loss": 1.471,
      "step": 2671
    },
    {
      "epoch": 0.9751824817518249,
      "grad_norm": 0.1646498739719391,
      "learning_rate": 0.0002,
      "loss": 1.4917,
      "step": 2672
    },
    {
      "epoch": 0.9755474452554744,
      "grad_norm": 0.16179852187633514,
      "learning_rate": 0.0002,
      "loss": 1.4324,
      "step": 2673
    },
    {
      "epoch": 0.9759124087591241,
      "grad_norm": 0.1644614189863205,
      "learning_rate": 0.0002,
      "loss": 1.4627,
      "step": 2674
    },
    {
      "epoch": 0.9762773722627737,
      "grad_norm": 0.16447153687477112,
      "learning_rate": 0.0002,
      "loss": 1.4491,
      "step": 2675
    },
    {
      "epoch": 0.9766423357664233,
      "grad_norm": 0.16249555349349976,
      "learning_rate": 0.0002,
      "loss": 1.487,
      "step": 2676
    },
    {
      "epoch": 0.977007299270073,
      "grad_norm": 0.16603465378284454,
      "learning_rate": 0.0002,
      "loss": 1.4913,
      "step": 2677
    },
    {
      "epoch": 0.9773722627737226,
      "grad_norm": 0.16248942911624908,
      "learning_rate": 0.0002,
      "loss": 1.4566,
      "step": 2678
    },
    {
      "epoch": 0.9777372262773723,
      "grad_norm": 0.16778512299060822,
      "learning_rate": 0.0002,
      "loss": 1.5165,
      "step": 2679
    },
    {
      "epoch": 0.9781021897810219,
      "grad_norm": 0.15797843039035797,
      "learning_rate": 0.0002,
      "loss": 1.4001,
      "step": 2680
    },
    {
      "epoch": 0.9784671532846715,
      "grad_norm": 0.17571523785591125,
      "learning_rate": 0.0002,
      "loss": 1.5116,
      "step": 2681
    },
    {
      "epoch": 0.9788321167883212,
      "grad_norm": 0.17001955211162567,
      "learning_rate": 0.0002,
      "loss": 1.4705,
      "step": 2682
    },
    {
      "epoch": 0.9791970802919708,
      "grad_norm": 0.16982561349868774,
      "learning_rate": 0.0002,
      "loss": 1.5089,
      "step": 2683
    },
    {
      "epoch": 0.9795620437956204,
      "grad_norm": 0.1614050567150116,
      "learning_rate": 0.0002,
      "loss": 1.46,
      "step": 2684
    },
    {
      "epoch": 0.9799270072992701,
      "grad_norm": 0.16552594304084778,
      "learning_rate": 0.0002,
      "loss": 1.5648,
      "step": 2685
    },
    {
      "epoch": 0.9802919708029197,
      "grad_norm": 0.1645481437444687,
      "learning_rate": 0.0002,
      "loss": 1.4903,
      "step": 2686
    },
    {
      "epoch": 0.9806569343065693,
      "grad_norm": 0.17247942090034485,
      "learning_rate": 0.0002,
      "loss": 1.5206,
      "step": 2687
    },
    {
      "epoch": 0.981021897810219,
      "grad_norm": 0.16435781121253967,
      "learning_rate": 0.0002,
      "loss": 1.4041,
      "step": 2688
    },
    {
      "epoch": 0.9813868613138687,
      "grad_norm": 0.16200537979602814,
      "learning_rate": 0.0002,
      "loss": 1.5434,
      "step": 2689
    },
    {
      "epoch": 0.9817518248175182,
      "grad_norm": 0.17101001739501953,
      "learning_rate": 0.0002,
      "loss": 1.5058,
      "step": 2690
    },
    {
      "epoch": 0.9821167883211679,
      "grad_norm": 0.16365015506744385,
      "learning_rate": 0.0002,
      "loss": 1.5401,
      "step": 2691
    },
    {
      "epoch": 0.9824817518248176,
      "grad_norm": 0.1653309017419815,
      "learning_rate": 0.0002,
      "loss": 1.472,
      "step": 2692
    },
    {
      "epoch": 0.9828467153284671,
      "grad_norm": 0.16835634410381317,
      "learning_rate": 0.0002,
      "loss": 1.4759,
      "step": 2693
    },
    {
      "epoch": 0.9832116788321168,
      "grad_norm": 0.1644730269908905,
      "learning_rate": 0.0002,
      "loss": 1.4835,
      "step": 2694
    },
    {
      "epoch": 0.9835766423357665,
      "grad_norm": 0.17060862481594086,
      "learning_rate": 0.0002,
      "loss": 1.5302,
      "step": 2695
    },
    {
      "epoch": 0.983941605839416,
      "grad_norm": 0.16324584186077118,
      "learning_rate": 0.0002,
      "loss": 1.4435,
      "step": 2696
    },
    {
      "epoch": 0.9843065693430657,
      "grad_norm": 0.15935496985912323,
      "learning_rate": 0.0002,
      "loss": 1.476,
      "step": 2697
    },
    {
      "epoch": 0.9846715328467154,
      "grad_norm": 0.1649622917175293,
      "learning_rate": 0.0002,
      "loss": 1.5505,
      "step": 2698
    },
    {
      "epoch": 0.9850364963503649,
      "grad_norm": 0.16700305044651031,
      "learning_rate": 0.0002,
      "loss": 1.504,
      "step": 2699
    },
    {
      "epoch": 0.9854014598540146,
      "grad_norm": 0.1668970137834549,
      "learning_rate": 0.0002,
      "loss": 1.4809,
      "step": 2700
    },
    {
      "epoch": 0.9857664233576642,
      "grad_norm": 0.16346511244773865,
      "learning_rate": 0.0002,
      "loss": 1.4519,
      "step": 2701
    },
    {
      "epoch": 0.9861313868613139,
      "grad_norm": 0.16543687880039215,
      "learning_rate": 0.0002,
      "loss": 1.538,
      "step": 2702
    },
    {
      "epoch": 0.9864963503649635,
      "grad_norm": 0.1641186773777008,
      "learning_rate": 0.0002,
      "loss": 1.4943,
      "step": 2703
    },
    {
      "epoch": 0.9868613138686131,
      "grad_norm": 0.16319482028484344,
      "learning_rate": 0.0002,
      "loss": 1.5034,
      "step": 2704
    },
    {
      "epoch": 0.9872262773722628,
      "grad_norm": 0.1639380306005478,
      "learning_rate": 0.0002,
      "loss": 1.5655,
      "step": 2705
    },
    {
      "epoch": 0.9875912408759124,
      "grad_norm": 0.16000205278396606,
      "learning_rate": 0.0002,
      "loss": 1.5073,
      "step": 2706
    },
    {
      "epoch": 0.987956204379562,
      "grad_norm": 0.1698179841041565,
      "learning_rate": 0.0002,
      "loss": 1.5269,
      "step": 2707
    },
    {
      "epoch": 0.9883211678832117,
      "grad_norm": 0.16495373845100403,
      "learning_rate": 0.0002,
      "loss": 1.4838,
      "step": 2708
    },
    {
      "epoch": 0.9886861313868613,
      "grad_norm": 0.1681678295135498,
      "learning_rate": 0.0002,
      "loss": 1.4385,
      "step": 2709
    },
    {
      "epoch": 0.9890510948905109,
      "grad_norm": 0.16897739470005035,
      "learning_rate": 0.0002,
      "loss": 1.5668,
      "step": 2710
    },
    {
      "epoch": 0.9894160583941606,
      "grad_norm": 0.16810882091522217,
      "learning_rate": 0.0002,
      "loss": 1.441,
      "step": 2711
    },
    {
      "epoch": 0.9897810218978103,
      "grad_norm": 0.17517369985580444,
      "learning_rate": 0.0002,
      "loss": 1.506,
      "step": 2712
    },
    {
      "epoch": 0.9901459854014598,
      "grad_norm": 0.16689033806324005,
      "learning_rate": 0.0002,
      "loss": 1.4442,
      "step": 2713
    },
    {
      "epoch": 0.9905109489051095,
      "grad_norm": 0.18461735546588898,
      "learning_rate": 0.0002,
      "loss": 1.4894,
      "step": 2714
    },
    {
      "epoch": 0.9908759124087592,
      "grad_norm": 0.16885003447532654,
      "learning_rate": 0.0002,
      "loss": 1.5482,
      "step": 2715
    },
    {
      "epoch": 0.9912408759124087,
      "grad_norm": 0.17314651608467102,
      "learning_rate": 0.0002,
      "loss": 1.4628,
      "step": 2716
    },
    {
      "epoch": 0.9916058394160584,
      "grad_norm": 0.18896660208702087,
      "learning_rate": 0.0002,
      "loss": 1.4809,
      "step": 2717
    },
    {
      "epoch": 0.9919708029197081,
      "grad_norm": 0.1690961867570877,
      "learning_rate": 0.0002,
      "loss": 1.503,
      "step": 2718
    },
    {
      "epoch": 0.9923357664233576,
      "grad_norm": 0.17377851903438568,
      "learning_rate": 0.0002,
      "loss": 1.4549,
      "step": 2719
    },
    {
      "epoch": 0.9927007299270073,
      "grad_norm": 0.15766246616840363,
      "learning_rate": 0.0002,
      "loss": 1.3856,
      "step": 2720
    },
    {
      "epoch": 0.993065693430657,
      "grad_norm": 0.1867644041776657,
      "learning_rate": 0.0002,
      "loss": 1.5526,
      "step": 2721
    },
    {
      "epoch": 0.9934306569343065,
      "grad_norm": 0.17104551196098328,
      "learning_rate": 0.0002,
      "loss": 1.4918,
      "step": 2722
    },
    {
      "epoch": 0.9937956204379562,
      "grad_norm": 0.16760122776031494,
      "learning_rate": 0.0002,
      "loss": 1.5393,
      "step": 2723
    },
    {
      "epoch": 0.9941605839416059,
      "grad_norm": 0.18323549628257751,
      "learning_rate": 0.0002,
      "loss": 1.4727,
      "step": 2724
    },
    {
      "epoch": 0.9945255474452555,
      "grad_norm": 0.17052264511585236,
      "learning_rate": 0.0002,
      "loss": 1.5567,
      "step": 2725
    },
    {
      "epoch": 0.9948905109489051,
      "grad_norm": 0.18296584486961365,
      "learning_rate": 0.0002,
      "loss": 1.4839,
      "step": 2726
    },
    {
      "epoch": 0.9952554744525547,
      "grad_norm": 0.17054511606693268,
      "learning_rate": 0.0002,
      "loss": 1.5241,
      "step": 2727
    },
    {
      "epoch": 0.9956204379562044,
      "grad_norm": 0.16412486135959625,
      "learning_rate": 0.0002,
      "loss": 1.3988,
      "step": 2728
    },
    {
      "epoch": 0.995985401459854,
      "grad_norm": 0.17159780859947205,
      "learning_rate": 0.0002,
      "loss": 1.4972,
      "step": 2729
    },
    {
      "epoch": 0.9963503649635036,
      "grad_norm": 0.16943159699440002,
      "learning_rate": 0.0002,
      "loss": 1.5266,
      "step": 2730
    },
    {
      "epoch": 0.9967153284671533,
      "grad_norm": 0.1862369328737259,
      "learning_rate": 0.0002,
      "loss": 1.4332,
      "step": 2731
    },
    {
      "epoch": 0.997080291970803,
      "grad_norm": 0.162359818816185,
      "learning_rate": 0.0002,
      "loss": 1.4372,
      "step": 2732
    },
    {
      "epoch": 0.9974452554744525,
      "grad_norm": 0.16765020787715912,
      "learning_rate": 0.0002,
      "loss": 1.5527,
      "step": 2733
    },
    {
      "epoch": 0.9978102189781022,
      "grad_norm": 0.1620924025774002,
      "learning_rate": 0.0002,
      "loss": 1.4792,
      "step": 2734
    },
    {
      "epoch": 0.9981751824817519,
      "grad_norm": 0.16451580822467804,
      "learning_rate": 0.0002,
      "loss": 1.4664,
      "step": 2735
    },
    {
      "epoch": 0.9985401459854014,
      "grad_norm": 0.15664073824882507,
      "learning_rate": 0.0002,
      "loss": 1.3755,
      "step": 2736
    },
    {
      "epoch": 0.9989051094890511,
      "grad_norm": 0.1587759405374527,
      "learning_rate": 0.0002,
      "loss": 1.456,
      "step": 2737
    },
    {
      "epoch": 0.9992700729927008,
      "grad_norm": 0.16149666905403137,
      "learning_rate": 0.0002,
      "loss": 1.4712,
      "step": 2738
    },
    {
      "epoch": 0.9996350364963503,
      "grad_norm": 0.1584656685590744,
      "learning_rate": 0.0002,
      "loss": 1.4537,
      "step": 2739
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.17085984349250793,
      "learning_rate": 0.0002,
      "loss": 1.5285,
      "step": 2740
    },
    {
      "epoch": 1.0003649635036496,
      "grad_norm": 0.16723331809043884,
      "learning_rate": 0.0002,
      "loss": 1.3413,
      "step": 2741
    },
    {
      "epoch": 1.0007299270072993,
      "grad_norm": 0.1639166921377182,
      "learning_rate": 0.0002,
      "loss": 1.4296,
      "step": 2742
    },
    {
      "epoch": 1.001094890510949,
      "grad_norm": 0.1750025451183319,
      "learning_rate": 0.0002,
      "loss": 1.4232,
      "step": 2743
    },
    {
      "epoch": 1.0014598540145985,
      "grad_norm": 0.16961973905563354,
      "learning_rate": 0.0002,
      "loss": 1.424,
      "step": 2744
    },
    {
      "epoch": 1.0018248175182483,
      "grad_norm": 0.17182715237140656,
      "learning_rate": 0.0002,
      "loss": 1.3541,
      "step": 2745
    },
    {
      "epoch": 1.0021897810218978,
      "grad_norm": 0.16970954835414886,
      "learning_rate": 0.0002,
      "loss": 1.3353,
      "step": 2746
    },
    {
      "epoch": 1.0025547445255474,
      "grad_norm": 0.182961106300354,
      "learning_rate": 0.0002,
      "loss": 1.4412,
      "step": 2747
    },
    {
      "epoch": 1.0029197080291972,
      "grad_norm": 0.1713067889213562,
      "learning_rate": 0.0002,
      "loss": 1.3461,
      "step": 2748
    },
    {
      "epoch": 1.0032846715328467,
      "grad_norm": 0.17873597145080566,
      "learning_rate": 0.0002,
      "loss": 1.4127,
      "step": 2749
    },
    {
      "epoch": 1.0036496350364963,
      "grad_norm": 0.17537762224674225,
      "learning_rate": 0.0002,
      "loss": 1.3842,
      "step": 2750
    },
    {
      "epoch": 1.004014598540146,
      "grad_norm": 0.1954246461391449,
      "learning_rate": 0.0002,
      "loss": 1.451,
      "step": 2751
    },
    {
      "epoch": 1.0043795620437956,
      "grad_norm": 0.1677187830209732,
      "learning_rate": 0.0002,
      "loss": 1.2639,
      "step": 2752
    },
    {
      "epoch": 1.0047445255474452,
      "grad_norm": 0.18577490746974945,
      "learning_rate": 0.0002,
      "loss": 1.374,
      "step": 2753
    },
    {
      "epoch": 1.005109489051095,
      "grad_norm": 0.170639306306839,
      "learning_rate": 0.0002,
      "loss": 1.342,
      "step": 2754
    },
    {
      "epoch": 1.0054744525547445,
      "grad_norm": 0.16958554089069366,
      "learning_rate": 0.0002,
      "loss": 1.3892,
      "step": 2755
    },
    {
      "epoch": 1.005839416058394,
      "grad_norm": 0.1733635514974594,
      "learning_rate": 0.0002,
      "loss": 1.3659,
      "step": 2756
    },
    {
      "epoch": 1.006204379562044,
      "grad_norm": 0.18436013162136078,
      "learning_rate": 0.0002,
      "loss": 1.3324,
      "step": 2757
    },
    {
      "epoch": 1.0065693430656935,
      "grad_norm": 0.20860178768634796,
      "learning_rate": 0.0002,
      "loss": 1.4819,
      "step": 2758
    },
    {
      "epoch": 1.006934306569343,
      "grad_norm": 0.17786023020744324,
      "learning_rate": 0.0002,
      "loss": 1.3626,
      "step": 2759
    },
    {
      "epoch": 1.0072992700729928,
      "grad_norm": 0.18276774883270264,
      "learning_rate": 0.0002,
      "loss": 1.3321,
      "step": 2760
    },
    {
      "epoch": 1.0076642335766424,
      "grad_norm": 0.17823459208011627,
      "learning_rate": 0.0002,
      "loss": 1.3904,
      "step": 2761
    },
    {
      "epoch": 1.008029197080292,
      "grad_norm": 0.18675021827220917,
      "learning_rate": 0.0002,
      "loss": 1.3171,
      "step": 2762
    },
    {
      "epoch": 1.0083941605839417,
      "grad_norm": 0.1798490285873413,
      "learning_rate": 0.0002,
      "loss": 1.3674,
      "step": 2763
    },
    {
      "epoch": 1.0087591240875913,
      "grad_norm": 0.19856971502304077,
      "learning_rate": 0.0002,
      "loss": 1.3987,
      "step": 2764
    },
    {
      "epoch": 1.0091240875912408,
      "grad_norm": 0.17647667229175568,
      "learning_rate": 0.0002,
      "loss": 1.421,
      "step": 2765
    },
    {
      "epoch": 1.0094890510948904,
      "grad_norm": 0.17845581471920013,
      "learning_rate": 0.0002,
      "loss": 1.3226,
      "step": 2766
    },
    {
      "epoch": 1.0098540145985402,
      "grad_norm": 0.17317746579647064,
      "learning_rate": 0.0002,
      "loss": 1.4012,
      "step": 2767
    },
    {
      "epoch": 1.0102189781021897,
      "grad_norm": 0.17807157337665558,
      "learning_rate": 0.0002,
      "loss": 1.3011,
      "step": 2768
    },
    {
      "epoch": 1.0105839416058393,
      "grad_norm": 0.19425052404403687,
      "learning_rate": 0.0002,
      "loss": 1.3838,
      "step": 2769
    },
    {
      "epoch": 1.010948905109489,
      "grad_norm": 0.18174350261688232,
      "learning_rate": 0.0002,
      "loss": 1.3115,
      "step": 2770
    },
    {
      "epoch": 1.0113138686131387,
      "grad_norm": 0.183743417263031,
      "learning_rate": 0.0002,
      "loss": 1.333,
      "step": 2771
    },
    {
      "epoch": 1.0116788321167882,
      "grad_norm": 0.17381425201892853,
      "learning_rate": 0.0002,
      "loss": 1.3925,
      "step": 2772
    },
    {
      "epoch": 1.012043795620438,
      "grad_norm": 0.17883019149303436,
      "learning_rate": 0.0002,
      "loss": 1.3933,
      "step": 2773
    },
    {
      "epoch": 1.0124087591240876,
      "grad_norm": 0.17295077443122864,
      "learning_rate": 0.0002,
      "loss": 1.3292,
      "step": 2774
    },
    {
      "epoch": 1.0127737226277371,
      "grad_norm": 0.17868399620056152,
      "learning_rate": 0.0002,
      "loss": 1.3349,
      "step": 2775
    },
    {
      "epoch": 1.013138686131387,
      "grad_norm": 0.17760372161865234,
      "learning_rate": 0.0002,
      "loss": 1.3527,
      "step": 2776
    },
    {
      "epoch": 1.0135036496350365,
      "grad_norm": 0.17143170535564423,
      "learning_rate": 0.0002,
      "loss": 1.397,
      "step": 2777
    },
    {
      "epoch": 1.013868613138686,
      "grad_norm": 0.1719256192445755,
      "learning_rate": 0.0002,
      "loss": 1.3705,
      "step": 2778
    },
    {
      "epoch": 1.0142335766423358,
      "grad_norm": 0.1726711243391037,
      "learning_rate": 0.0002,
      "loss": 1.3808,
      "step": 2779
    },
    {
      "epoch": 1.0145985401459854,
      "grad_norm": 0.17461317777633667,
      "learning_rate": 0.0002,
      "loss": 1.4192,
      "step": 2780
    },
    {
      "epoch": 1.014963503649635,
      "grad_norm": 0.16870781779289246,
      "learning_rate": 0.0002,
      "loss": 1.3201,
      "step": 2781
    },
    {
      "epoch": 1.0153284671532847,
      "grad_norm": 0.1705092042684555,
      "learning_rate": 0.0002,
      "loss": 1.2771,
      "step": 2782
    },
    {
      "epoch": 1.0156934306569343,
      "grad_norm": 0.18435268104076385,
      "learning_rate": 0.0002,
      "loss": 1.4337,
      "step": 2783
    },
    {
      "epoch": 1.0160583941605839,
      "grad_norm": 0.1837470382452011,
      "learning_rate": 0.0002,
      "loss": 1.3102,
      "step": 2784
    },
    {
      "epoch": 1.0164233576642336,
      "grad_norm": 0.1864379495382309,
      "learning_rate": 0.0002,
      "loss": 1.3492,
      "step": 2785
    },
    {
      "epoch": 1.0167883211678832,
      "grad_norm": 0.18448898196220398,
      "learning_rate": 0.0002,
      "loss": 1.3522,
      "step": 2786
    },
    {
      "epoch": 1.0171532846715328,
      "grad_norm": 0.17643460631370544,
      "learning_rate": 0.0002,
      "loss": 1.3942,
      "step": 2787
    },
    {
      "epoch": 1.0175182481751825,
      "grad_norm": 0.1865793764591217,
      "learning_rate": 0.0002,
      "loss": 1.3605,
      "step": 2788
    },
    {
      "epoch": 1.0178832116788321,
      "grad_norm": 0.1762734055519104,
      "learning_rate": 0.0002,
      "loss": 1.3823,
      "step": 2789
    },
    {
      "epoch": 1.0182481751824817,
      "grad_norm": 0.1866241991519928,
      "learning_rate": 0.0002,
      "loss": 1.3885,
      "step": 2790
    },
    {
      "epoch": 1.0186131386861315,
      "grad_norm": 0.19213350117206573,
      "learning_rate": 0.0002,
      "loss": 1.3538,
      "step": 2791
    },
    {
      "epoch": 1.018978102189781,
      "grad_norm": 0.1805182248353958,
      "learning_rate": 0.0002,
      "loss": 1.4606,
      "step": 2792
    },
    {
      "epoch": 1.0193430656934306,
      "grad_norm": 0.19520540535449982,
      "learning_rate": 0.0002,
      "loss": 1.4166,
      "step": 2793
    },
    {
      "epoch": 1.0197080291970804,
      "grad_norm": 0.1786096692085266,
      "learning_rate": 0.0002,
      "loss": 1.4027,
      "step": 2794
    },
    {
      "epoch": 1.02007299270073,
      "grad_norm": 0.17543423175811768,
      "learning_rate": 0.0002,
      "loss": 1.3478,
      "step": 2795
    },
    {
      "epoch": 1.0204379562043795,
      "grad_norm": 0.18682558834552765,
      "learning_rate": 0.0002,
      "loss": 1.3669,
      "step": 2796
    },
    {
      "epoch": 1.0208029197080293,
      "grad_norm": 0.1908658891916275,
      "learning_rate": 0.0002,
      "loss": 1.4683,
      "step": 2797
    },
    {
      "epoch": 1.0211678832116788,
      "grad_norm": 0.18187205493450165,
      "learning_rate": 0.0002,
      "loss": 1.4164,
      "step": 2798
    },
    {
      "epoch": 1.0215328467153284,
      "grad_norm": 0.17911934852600098,
      "learning_rate": 0.0002,
      "loss": 1.4372,
      "step": 2799
    },
    {
      "epoch": 1.0218978102189782,
      "grad_norm": 0.18852615356445312,
      "learning_rate": 0.0002,
      "loss": 1.4231,
      "step": 2800
    },
    {
      "epoch": 1.0222627737226277,
      "grad_norm": 0.18145325779914856,
      "learning_rate": 0.0002,
      "loss": 1.3527,
      "step": 2801
    },
    {
      "epoch": 1.0226277372262773,
      "grad_norm": 0.1901959627866745,
      "learning_rate": 0.0002,
      "loss": 1.2997,
      "step": 2802
    },
    {
      "epoch": 1.022992700729927,
      "grad_norm": 0.18002350628376007,
      "learning_rate": 0.0002,
      "loss": 1.3599,
      "step": 2803
    },
    {
      "epoch": 1.0233576642335767,
      "grad_norm": 0.17931945621967316,
      "learning_rate": 0.0002,
      "loss": 1.3003,
      "step": 2804
    },
    {
      "epoch": 1.0237226277372262,
      "grad_norm": 0.18978071212768555,
      "learning_rate": 0.0002,
      "loss": 1.3852,
      "step": 2805
    },
    {
      "epoch": 1.024087591240876,
      "grad_norm": 0.18487615883350372,
      "learning_rate": 0.0002,
      "loss": 1.3511,
      "step": 2806
    },
    {
      "epoch": 1.0244525547445256,
      "grad_norm": 0.18114839494228363,
      "learning_rate": 0.0002,
      "loss": 1.3632,
      "step": 2807
    },
    {
      "epoch": 1.0248175182481751,
      "grad_norm": 0.18243679404258728,
      "learning_rate": 0.0002,
      "loss": 1.409,
      "step": 2808
    },
    {
      "epoch": 1.025182481751825,
      "grad_norm": 0.18227624893188477,
      "learning_rate": 0.0002,
      "loss": 1.3651,
      "step": 2809
    },
    {
      "epoch": 1.0255474452554745,
      "grad_norm": 0.19139815866947174,
      "learning_rate": 0.0002,
      "loss": 1.4261,
      "step": 2810
    },
    {
      "epoch": 1.025912408759124,
      "grad_norm": 0.1803199052810669,
      "learning_rate": 0.0002,
      "loss": 1.3099,
      "step": 2811
    },
    {
      "epoch": 1.0262773722627738,
      "grad_norm": 0.18807640671730042,
      "learning_rate": 0.0002,
      "loss": 1.4074,
      "step": 2812
    },
    {
      "epoch": 1.0266423357664234,
      "grad_norm": 0.18847179412841797,
      "learning_rate": 0.0002,
      "loss": 1.4443,
      "step": 2813
    },
    {
      "epoch": 1.027007299270073,
      "grad_norm": 0.17401430010795593,
      "learning_rate": 0.0002,
      "loss": 1.3454,
      "step": 2814
    },
    {
      "epoch": 1.0273722627737227,
      "grad_norm": 0.1867239773273468,
      "learning_rate": 0.0002,
      "loss": 1.3365,
      "step": 2815
    },
    {
      "epoch": 1.0277372262773723,
      "grad_norm": 0.18103854358196259,
      "learning_rate": 0.0002,
      "loss": 1.4289,
      "step": 2816
    },
    {
      "epoch": 1.0281021897810219,
      "grad_norm": 0.1735541820526123,
      "learning_rate": 0.0002,
      "loss": 1.3266,
      "step": 2817
    },
    {
      "epoch": 1.0284671532846716,
      "grad_norm": 0.17086364328861237,
      "learning_rate": 0.0002,
      "loss": 1.3571,
      "step": 2818
    },
    {
      "epoch": 1.0288321167883212,
      "grad_norm": 0.17741020023822784,
      "learning_rate": 0.0002,
      "loss": 1.3818,
      "step": 2819
    },
    {
      "epoch": 1.0291970802919708,
      "grad_norm": 0.18213814496994019,
      "learning_rate": 0.0002,
      "loss": 1.3538,
      "step": 2820
    },
    {
      "epoch": 1.0295620437956203,
      "grad_norm": 0.17599831521511078,
      "learning_rate": 0.0002,
      "loss": 1.3434,
      "step": 2821
    },
    {
      "epoch": 1.0299270072992701,
      "grad_norm": 0.174272820353508,
      "learning_rate": 0.0002,
      "loss": 1.2651,
      "step": 2822
    },
    {
      "epoch": 1.0302919708029197,
      "grad_norm": 0.1903129667043686,
      "learning_rate": 0.0002,
      "loss": 1.3971,
      "step": 2823
    },
    {
      "epoch": 1.0306569343065692,
      "grad_norm": 0.1811424195766449,
      "learning_rate": 0.0002,
      "loss": 1.3663,
      "step": 2824
    },
    {
      "epoch": 1.031021897810219,
      "grad_norm": 0.1806936115026474,
      "learning_rate": 0.0002,
      "loss": 1.4109,
      "step": 2825
    },
    {
      "epoch": 1.0313868613138686,
      "grad_norm": 0.17854635417461395,
      "learning_rate": 0.0002,
      "loss": 1.3325,
      "step": 2826
    },
    {
      "epoch": 1.0317518248175181,
      "grad_norm": 0.1775699406862259,
      "learning_rate": 0.0002,
      "loss": 1.3641,
      "step": 2827
    },
    {
      "epoch": 1.032116788321168,
      "grad_norm": 0.18899700045585632,
      "learning_rate": 0.0002,
      "loss": 1.4635,
      "step": 2828
    },
    {
      "epoch": 1.0324817518248175,
      "grad_norm": 0.1722145974636078,
      "learning_rate": 0.0002,
      "loss": 1.3657,
      "step": 2829
    },
    {
      "epoch": 1.032846715328467,
      "grad_norm": 0.18217770755290985,
      "learning_rate": 0.0002,
      "loss": 1.3876,
      "step": 2830
    },
    {
      "epoch": 1.0332116788321168,
      "grad_norm": 0.18260079622268677,
      "learning_rate": 0.0002,
      "loss": 1.3484,
      "step": 2831
    },
    {
      "epoch": 1.0335766423357664,
      "grad_norm": 0.17962123453617096,
      "learning_rate": 0.0002,
      "loss": 1.3184,
      "step": 2832
    },
    {
      "epoch": 1.033941605839416,
      "grad_norm": 0.2004845291376114,
      "learning_rate": 0.0002,
      "loss": 1.3948,
      "step": 2833
    },
    {
      "epoch": 1.0343065693430658,
      "grad_norm": 0.1865808516740799,
      "learning_rate": 0.0002,
      "loss": 1.3909,
      "step": 2834
    },
    {
      "epoch": 1.0346715328467153,
      "grad_norm": 0.18416230380535126,
      "learning_rate": 0.0002,
      "loss": 1.3843,
      "step": 2835
    },
    {
      "epoch": 1.0350364963503649,
      "grad_norm": 0.17659170925617218,
      "learning_rate": 0.0002,
      "loss": 1.3508,
      "step": 2836
    },
    {
      "epoch": 1.0354014598540147,
      "grad_norm": 0.18310260772705078,
      "learning_rate": 0.0002,
      "loss": 1.4243,
      "step": 2837
    },
    {
      "epoch": 1.0357664233576642,
      "grad_norm": 0.20768967270851135,
      "learning_rate": 0.0002,
      "loss": 1.4914,
      "step": 2838
    },
    {
      "epoch": 1.0361313868613138,
      "grad_norm": 0.1951708346605301,
      "learning_rate": 0.0002,
      "loss": 1.3502,
      "step": 2839
    },
    {
      "epoch": 1.0364963503649636,
      "grad_norm": 0.1869037002325058,
      "learning_rate": 0.0002,
      "loss": 1.3637,
      "step": 2840
    },
    {
      "epoch": 1.0368613138686131,
      "grad_norm": 0.18238994479179382,
      "learning_rate": 0.0002,
      "loss": 1.3946,
      "step": 2841
    },
    {
      "epoch": 1.0372262773722627,
      "grad_norm": 0.18238504230976105,
      "learning_rate": 0.0002,
      "loss": 1.4874,
      "step": 2842
    },
    {
      "epoch": 1.0375912408759125,
      "grad_norm": 0.2091313600540161,
      "learning_rate": 0.0002,
      "loss": 1.3191,
      "step": 2843
    },
    {
      "epoch": 1.037956204379562,
      "grad_norm": 0.18782348930835724,
      "learning_rate": 0.0002,
      "loss": 1.3511,
      "step": 2844
    },
    {
      "epoch": 1.0383211678832116,
      "grad_norm": 0.18741516768932343,
      "learning_rate": 0.0002,
      "loss": 1.4129,
      "step": 2845
    },
    {
      "epoch": 1.0386861313868614,
      "grad_norm": 0.18542732298374176,
      "learning_rate": 0.0002,
      "loss": 1.363,
      "step": 2846
    },
    {
      "epoch": 1.039051094890511,
      "grad_norm": 0.20187066495418549,
      "learning_rate": 0.0002,
      "loss": 1.4002,
      "step": 2847
    },
    {
      "epoch": 1.0394160583941605,
      "grad_norm": 0.18538090586662292,
      "learning_rate": 0.0002,
      "loss": 1.3577,
      "step": 2848
    },
    {
      "epoch": 1.0397810218978103,
      "grad_norm": 0.19068637490272522,
      "learning_rate": 0.0002,
      "loss": 1.4007,
      "step": 2849
    },
    {
      "epoch": 1.0401459854014599,
      "grad_norm": 0.19518297910690308,
      "learning_rate": 0.0002,
      "loss": 1.4464,
      "step": 2850
    },
    {
      "epoch": 1.0405109489051094,
      "grad_norm": 0.19712775945663452,
      "learning_rate": 0.0002,
      "loss": 1.4044,
      "step": 2851
    },
    {
      "epoch": 1.0408759124087592,
      "grad_norm": 0.1818314492702484,
      "learning_rate": 0.0002,
      "loss": 1.3441,
      "step": 2852
    },
    {
      "epoch": 1.0412408759124088,
      "grad_norm": 0.19436313211917877,
      "learning_rate": 0.0002,
      "loss": 1.3484,
      "step": 2853
    },
    {
      "epoch": 1.0416058394160583,
      "grad_norm": 0.18059328198432922,
      "learning_rate": 0.0002,
      "loss": 1.365,
      "step": 2854
    },
    {
      "epoch": 1.0419708029197081,
      "grad_norm": 0.20097677409648895,
      "learning_rate": 0.0002,
      "loss": 1.3546,
      "step": 2855
    },
    {
      "epoch": 1.0423357664233577,
      "grad_norm": 0.18863606452941895,
      "learning_rate": 0.0002,
      "loss": 1.3938,
      "step": 2856
    },
    {
      "epoch": 1.0427007299270072,
      "grad_norm": 0.19279745221138,
      "learning_rate": 0.0002,
      "loss": 1.3429,
      "step": 2857
    },
    {
      "epoch": 1.043065693430657,
      "grad_norm": 0.18130804598331451,
      "learning_rate": 0.0002,
      "loss": 1.3588,
      "step": 2858
    },
    {
      "epoch": 1.0434306569343066,
      "grad_norm": 0.18833519518375397,
      "learning_rate": 0.0002,
      "loss": 1.3566,
      "step": 2859
    },
    {
      "epoch": 1.0437956204379562,
      "grad_norm": 0.1911589801311493,
      "learning_rate": 0.0002,
      "loss": 1.4182,
      "step": 2860
    },
    {
      "epoch": 1.044160583941606,
      "grad_norm": 0.18112553656101227,
      "learning_rate": 0.0002,
      "loss": 1.2639,
      "step": 2861
    },
    {
      "epoch": 1.0445255474452555,
      "grad_norm": 0.18449513614177704,
      "learning_rate": 0.0002,
      "loss": 1.3395,
      "step": 2862
    },
    {
      "epoch": 1.044890510948905,
      "grad_norm": 0.1831454634666443,
      "learning_rate": 0.0002,
      "loss": 1.35,
      "step": 2863
    },
    {
      "epoch": 1.0452554744525548,
      "grad_norm": 0.18223625421524048,
      "learning_rate": 0.0002,
      "loss": 1.3697,
      "step": 2864
    },
    {
      "epoch": 1.0456204379562044,
      "grad_norm": 0.17459410429000854,
      "learning_rate": 0.0002,
      "loss": 1.3408,
      "step": 2865
    },
    {
      "epoch": 1.045985401459854,
      "grad_norm": 0.1750446856021881,
      "learning_rate": 0.0002,
      "loss": 1.3827,
      "step": 2866
    },
    {
      "epoch": 1.0463503649635038,
      "grad_norm": 0.1795663982629776,
      "learning_rate": 0.0002,
      "loss": 1.3742,
      "step": 2867
    },
    {
      "epoch": 1.0467153284671533,
      "grad_norm": 0.18237198889255524,
      "learning_rate": 0.0002,
      "loss": 1.4069,
      "step": 2868
    },
    {
      "epoch": 1.0470802919708029,
      "grad_norm": 0.18966645002365112,
      "learning_rate": 0.0002,
      "loss": 1.2901,
      "step": 2869
    },
    {
      "epoch": 1.0474452554744524,
      "grad_norm": 0.1793471872806549,
      "learning_rate": 0.0002,
      "loss": 1.3704,
      "step": 2870
    },
    {
      "epoch": 1.0478102189781022,
      "grad_norm": 0.17302463948726654,
      "learning_rate": 0.0002,
      "loss": 1.2792,
      "step": 2871
    },
    {
      "epoch": 1.0481751824817518,
      "grad_norm": 0.1736961156129837,
      "learning_rate": 0.0002,
      "loss": 1.4116,
      "step": 2872
    },
    {
      "epoch": 1.0485401459854016,
      "grad_norm": 0.18082575500011444,
      "learning_rate": 0.0002,
      "loss": 1.4257,
      "step": 2873
    },
    {
      "epoch": 1.0489051094890511,
      "grad_norm": 0.1758885532617569,
      "learning_rate": 0.0002,
      "loss": 1.3887,
      "step": 2874
    },
    {
      "epoch": 1.0492700729927007,
      "grad_norm": 0.18769851326942444,
      "learning_rate": 0.0002,
      "loss": 1.3296,
      "step": 2875
    },
    {
      "epoch": 1.0496350364963503,
      "grad_norm": 0.18509690463542938,
      "learning_rate": 0.0002,
      "loss": 1.2992,
      "step": 2876
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.20282621681690216,
      "learning_rate": 0.0002,
      "loss": 1.4671,
      "step": 2877
    },
    {
      "epoch": 1.0503649635036496,
      "grad_norm": 0.19202186167240143,
      "learning_rate": 0.0002,
      "loss": 1.3495,
      "step": 2878
    },
    {
      "epoch": 1.0507299270072992,
      "grad_norm": 0.20633359253406525,
      "learning_rate": 0.0002,
      "loss": 1.3574,
      "step": 2879
    },
    {
      "epoch": 1.051094890510949,
      "grad_norm": 0.1882936656475067,
      "learning_rate": 0.0002,
      "loss": 1.3912,
      "step": 2880
    },
    {
      "epoch": 1.0514598540145985,
      "grad_norm": 0.17444032430648804,
      "learning_rate": 0.0002,
      "loss": 1.4071,
      "step": 2881
    },
    {
      "epoch": 1.051824817518248,
      "grad_norm": 0.1810348629951477,
      "learning_rate": 0.0002,
      "loss": 1.4088,
      "step": 2882
    },
    {
      "epoch": 1.0521897810218979,
      "grad_norm": 0.20148180425167084,
      "learning_rate": 0.0002,
      "loss": 1.4298,
      "step": 2883
    },
    {
      "epoch": 1.0525547445255474,
      "grad_norm": 0.19182832539081573,
      "learning_rate": 0.0002,
      "loss": 1.4805,
      "step": 2884
    },
    {
      "epoch": 1.052919708029197,
      "grad_norm": 0.2036486715078354,
      "learning_rate": 0.0002,
      "loss": 1.4334,
      "step": 2885
    },
    {
      "epoch": 1.0532846715328468,
      "grad_norm": 0.1718732714653015,
      "learning_rate": 0.0002,
      "loss": 1.3278,
      "step": 2886
    },
    {
      "epoch": 1.0536496350364963,
      "grad_norm": 0.17945221066474915,
      "learning_rate": 0.0002,
      "loss": 1.326,
      "step": 2887
    },
    {
      "epoch": 1.054014598540146,
      "grad_norm": 0.18048256635665894,
      "learning_rate": 0.0002,
      "loss": 1.3248,
      "step": 2888
    },
    {
      "epoch": 1.0543795620437957,
      "grad_norm": 0.17586326599121094,
      "learning_rate": 0.0002,
      "loss": 1.316,
      "step": 2889
    },
    {
      "epoch": 1.0547445255474452,
      "grad_norm": 0.18860934674739838,
      "learning_rate": 0.0002,
      "loss": 1.4298,
      "step": 2890
    },
    {
      "epoch": 1.0551094890510948,
      "grad_norm": 0.20941348373889923,
      "learning_rate": 0.0002,
      "loss": 1.3912,
      "step": 2891
    },
    {
      "epoch": 1.0554744525547446,
      "grad_norm": 0.18297454714775085,
      "learning_rate": 0.0002,
      "loss": 1.3301,
      "step": 2892
    },
    {
      "epoch": 1.0558394160583942,
      "grad_norm": 0.20133012533187866,
      "learning_rate": 0.0002,
      "loss": 1.4069,
      "step": 2893
    },
    {
      "epoch": 1.0562043795620437,
      "grad_norm": 0.18290133774280548,
      "learning_rate": 0.0002,
      "loss": 1.4222,
      "step": 2894
    },
    {
      "epoch": 1.0565693430656935,
      "grad_norm": 0.19148589670658112,
      "learning_rate": 0.0002,
      "loss": 1.3616,
      "step": 2895
    },
    {
      "epoch": 1.056934306569343,
      "grad_norm": 0.1885538250207901,
      "learning_rate": 0.0002,
      "loss": 1.3684,
      "step": 2896
    },
    {
      "epoch": 1.0572992700729926,
      "grad_norm": 0.19833098351955414,
      "learning_rate": 0.0002,
      "loss": 1.3707,
      "step": 2897
    },
    {
      "epoch": 1.0576642335766424,
      "grad_norm": 0.1790136843919754,
      "learning_rate": 0.0002,
      "loss": 1.2765,
      "step": 2898
    },
    {
      "epoch": 1.058029197080292,
      "grad_norm": 0.191654235124588,
      "learning_rate": 0.0002,
      "loss": 1.3916,
      "step": 2899
    },
    {
      "epoch": 1.0583941605839415,
      "grad_norm": 0.18117845058441162,
      "learning_rate": 0.0002,
      "loss": 1.389,
      "step": 2900
    },
    {
      "epoch": 1.0587591240875913,
      "grad_norm": 0.17866508662700653,
      "learning_rate": 0.0002,
      "loss": 1.3289,
      "step": 2901
    },
    {
      "epoch": 1.0591240875912409,
      "grad_norm": 0.18016433715820312,
      "learning_rate": 0.0002,
      "loss": 1.4067,
      "step": 2902
    },
    {
      "epoch": 1.0594890510948904,
      "grad_norm": 0.180988147854805,
      "learning_rate": 0.0002,
      "loss": 1.343,
      "step": 2903
    },
    {
      "epoch": 1.0598540145985402,
      "grad_norm": 0.19753937423229218,
      "learning_rate": 0.0002,
      "loss": 1.3561,
      "step": 2904
    },
    {
      "epoch": 1.0602189781021898,
      "grad_norm": 0.19587931036949158,
      "learning_rate": 0.0002,
      "loss": 1.429,
      "step": 2905
    },
    {
      "epoch": 1.0605839416058394,
      "grad_norm": 0.18821407854557037,
      "learning_rate": 0.0002,
      "loss": 1.3128,
      "step": 2906
    },
    {
      "epoch": 1.0609489051094891,
      "grad_norm": 0.19076694548130035,
      "learning_rate": 0.0002,
      "loss": 1.382,
      "step": 2907
    },
    {
      "epoch": 1.0613138686131387,
      "grad_norm": 0.19640406966209412,
      "learning_rate": 0.0002,
      "loss": 1.4156,
      "step": 2908
    },
    {
      "epoch": 1.0616788321167883,
      "grad_norm": 0.19544292986392975,
      "learning_rate": 0.0002,
      "loss": 1.3736,
      "step": 2909
    },
    {
      "epoch": 1.062043795620438,
      "grad_norm": 0.19213290512561798,
      "learning_rate": 0.0002,
      "loss": 1.3917,
      "step": 2910
    },
    {
      "epoch": 1.0624087591240876,
      "grad_norm": 0.18191197514533997,
      "learning_rate": 0.0002,
      "loss": 1.3621,
      "step": 2911
    },
    {
      "epoch": 1.0627737226277372,
      "grad_norm": 0.19427014887332916,
      "learning_rate": 0.0002,
      "loss": 1.3462,
      "step": 2912
    },
    {
      "epoch": 1.063138686131387,
      "grad_norm": 0.20099769532680511,
      "learning_rate": 0.0002,
      "loss": 1.4008,
      "step": 2913
    },
    {
      "epoch": 1.0635036496350365,
      "grad_norm": 0.1838114708662033,
      "learning_rate": 0.0002,
      "loss": 1.3098,
      "step": 2914
    },
    {
      "epoch": 1.063868613138686,
      "grad_norm": 0.19396701455116272,
      "learning_rate": 0.0002,
      "loss": 1.3845,
      "step": 2915
    },
    {
      "epoch": 1.0642335766423359,
      "grad_norm": 0.19033768773078918,
      "learning_rate": 0.0002,
      "loss": 1.3665,
      "step": 2916
    },
    {
      "epoch": 1.0645985401459854,
      "grad_norm": 0.19423244893550873,
      "learning_rate": 0.0002,
      "loss": 1.4113,
      "step": 2917
    },
    {
      "epoch": 1.064963503649635,
      "grad_norm": 0.18376435339450836,
      "learning_rate": 0.0002,
      "loss": 1.4428,
      "step": 2918
    },
    {
      "epoch": 1.0653284671532848,
      "grad_norm": 0.18093307316303253,
      "learning_rate": 0.0002,
      "loss": 1.3844,
      "step": 2919
    },
    {
      "epoch": 1.0656934306569343,
      "grad_norm": 0.18230699002742767,
      "learning_rate": 0.0002,
      "loss": 1.3756,
      "step": 2920
    },
    {
      "epoch": 1.066058394160584,
      "grad_norm": 0.1793680340051651,
      "learning_rate": 0.0002,
      "loss": 1.356,
      "step": 2921
    },
    {
      "epoch": 1.0664233576642337,
      "grad_norm": 0.18377253413200378,
      "learning_rate": 0.0002,
      "loss": 1.2795,
      "step": 2922
    },
    {
      "epoch": 1.0667883211678832,
      "grad_norm": 0.18726050853729248,
      "learning_rate": 0.0002,
      "loss": 1.3939,
      "step": 2923
    },
    {
      "epoch": 1.0671532846715328,
      "grad_norm": 0.1815202236175537,
      "learning_rate": 0.0002,
      "loss": 1.3561,
      "step": 2924
    },
    {
      "epoch": 1.0675182481751824,
      "grad_norm": 0.18439404666423798,
      "learning_rate": 0.0002,
      "loss": 1.3898,
      "step": 2925
    },
    {
      "epoch": 1.0678832116788322,
      "grad_norm": 0.18233971297740936,
      "learning_rate": 0.0002,
      "loss": 1.3432,
      "step": 2926
    },
    {
      "epoch": 1.0682481751824817,
      "grad_norm": 0.1844305843114853,
      "learning_rate": 0.0002,
      "loss": 1.3999,
      "step": 2927
    },
    {
      "epoch": 1.0686131386861315,
      "grad_norm": 0.18898117542266846,
      "learning_rate": 0.0002,
      "loss": 1.4355,
      "step": 2928
    },
    {
      "epoch": 1.068978102189781,
      "grad_norm": 0.1844037026166916,
      "learning_rate": 0.0002,
      "loss": 1.4037,
      "step": 2929
    },
    {
      "epoch": 1.0693430656934306,
      "grad_norm": 0.1891563981771469,
      "learning_rate": 0.0002,
      "loss": 1.3798,
      "step": 2930
    },
    {
      "epoch": 1.0697080291970802,
      "grad_norm": 0.18778352439403534,
      "learning_rate": 0.0002,
      "loss": 1.4673,
      "step": 2931
    },
    {
      "epoch": 1.07007299270073,
      "grad_norm": 0.18711206316947937,
      "learning_rate": 0.0002,
      "loss": 1.3596,
      "step": 2932
    },
    {
      "epoch": 1.0704379562043795,
      "grad_norm": 0.17663107812404633,
      "learning_rate": 0.0002,
      "loss": 1.3211,
      "step": 2933
    },
    {
      "epoch": 1.070802919708029,
      "grad_norm": 0.18851922452449799,
      "learning_rate": 0.0002,
      "loss": 1.3304,
      "step": 2934
    },
    {
      "epoch": 1.0711678832116789,
      "grad_norm": 0.1916496753692627,
      "learning_rate": 0.0002,
      "loss": 1.3752,
      "step": 2935
    },
    {
      "epoch": 1.0715328467153284,
      "grad_norm": 0.18218298256397247,
      "learning_rate": 0.0002,
      "loss": 1.2698,
      "step": 2936
    },
    {
      "epoch": 1.071897810218978,
      "grad_norm": 0.18727430701255798,
      "learning_rate": 0.0002,
      "loss": 1.3264,
      "step": 2937
    },
    {
      "epoch": 1.0722627737226278,
      "grad_norm": 0.1977178454399109,
      "learning_rate": 0.0002,
      "loss": 1.4256,
      "step": 2938
    },
    {
      "epoch": 1.0726277372262774,
      "grad_norm": 0.19108404219150543,
      "learning_rate": 0.0002,
      "loss": 1.3189,
      "step": 2939
    },
    {
      "epoch": 1.072992700729927,
      "grad_norm": 0.18836070597171783,
      "learning_rate": 0.0002,
      "loss": 1.4142,
      "step": 2940
    },
    {
      "epoch": 1.0733576642335767,
      "grad_norm": 0.17991599440574646,
      "learning_rate": 0.0002,
      "loss": 1.342,
      "step": 2941
    },
    {
      "epoch": 1.0737226277372263,
      "grad_norm": 0.18201829493045807,
      "learning_rate": 0.0002,
      "loss": 1.3118,
      "step": 2942
    },
    {
      "epoch": 1.0740875912408758,
      "grad_norm": 0.18765325844287872,
      "learning_rate": 0.0002,
      "loss": 1.4152,
      "step": 2943
    },
    {
      "epoch": 1.0744525547445256,
      "grad_norm": 0.1885276436805725,
      "learning_rate": 0.0002,
      "loss": 1.3165,
      "step": 2944
    },
    {
      "epoch": 1.0748175182481752,
      "grad_norm": 0.17637817561626434,
      "learning_rate": 0.0002,
      "loss": 1.2821,
      "step": 2945
    },
    {
      "epoch": 1.0751824817518247,
      "grad_norm": 0.19787269830703735,
      "learning_rate": 0.0002,
      "loss": 1.4716,
      "step": 2946
    },
    {
      "epoch": 1.0755474452554745,
      "grad_norm": 0.20585647225379944,
      "learning_rate": 0.0002,
      "loss": 1.3848,
      "step": 2947
    },
    {
      "epoch": 1.075912408759124,
      "grad_norm": 0.19018682837486267,
      "learning_rate": 0.0002,
      "loss": 1.3369,
      "step": 2948
    },
    {
      "epoch": 1.0762773722627736,
      "grad_norm": 0.18739114701747894,
      "learning_rate": 0.0002,
      "loss": 1.3727,
      "step": 2949
    },
    {
      "epoch": 1.0766423357664234,
      "grad_norm": 0.17503373324871063,
      "learning_rate": 0.0002,
      "loss": 1.2772,
      "step": 2950
    },
    {
      "epoch": 1.077007299270073,
      "grad_norm": 0.18489192426204681,
      "learning_rate": 0.0002,
      "loss": 1.314,
      "step": 2951
    },
    {
      "epoch": 1.0773722627737226,
      "grad_norm": 0.18612481653690338,
      "learning_rate": 0.0002,
      "loss": 1.3723,
      "step": 2952
    },
    {
      "epoch": 1.0777372262773723,
      "grad_norm": 0.18309351801872253,
      "learning_rate": 0.0002,
      "loss": 1.3923,
      "step": 2953
    },
    {
      "epoch": 1.078102189781022,
      "grad_norm": 0.18644674122333527,
      "learning_rate": 0.0002,
      "loss": 1.3134,
      "step": 2954
    },
    {
      "epoch": 1.0784671532846715,
      "grad_norm": 0.1848331242799759,
      "learning_rate": 0.0002,
      "loss": 1.3213,
      "step": 2955
    },
    {
      "epoch": 1.0788321167883212,
      "grad_norm": 0.19440576434135437,
      "learning_rate": 0.0002,
      "loss": 1.3546,
      "step": 2956
    },
    {
      "epoch": 1.0791970802919708,
      "grad_norm": 0.18256410956382751,
      "learning_rate": 0.0002,
      "loss": 1.3382,
      "step": 2957
    },
    {
      "epoch": 1.0795620437956204,
      "grad_norm": 0.19742490351200104,
      "learning_rate": 0.0002,
      "loss": 1.3693,
      "step": 2958
    },
    {
      "epoch": 1.0799270072992702,
      "grad_norm": 0.1887078881263733,
      "learning_rate": 0.0002,
      "loss": 1.3837,
      "step": 2959
    },
    {
      "epoch": 1.0802919708029197,
      "grad_norm": 0.18228468298912048,
      "learning_rate": 0.0002,
      "loss": 1.3206,
      "step": 2960
    },
    {
      "epoch": 1.0806569343065693,
      "grad_norm": 0.18590186536312103,
      "learning_rate": 0.0002,
      "loss": 1.421,
      "step": 2961
    },
    {
      "epoch": 1.081021897810219,
      "grad_norm": 0.18091757595539093,
      "learning_rate": 0.0002,
      "loss": 1.3555,
      "step": 2962
    },
    {
      "epoch": 1.0813868613138686,
      "grad_norm": 0.184972882270813,
      "learning_rate": 0.0002,
      "loss": 1.3501,
      "step": 2963
    },
    {
      "epoch": 1.0817518248175182,
      "grad_norm": 0.19001883268356323,
      "learning_rate": 0.0002,
      "loss": 1.4275,
      "step": 2964
    },
    {
      "epoch": 1.082116788321168,
      "grad_norm": 0.18911001086235046,
      "learning_rate": 0.0002,
      "loss": 1.3644,
      "step": 2965
    },
    {
      "epoch": 1.0824817518248175,
      "grad_norm": 0.17874062061309814,
      "learning_rate": 0.0002,
      "loss": 1.2742,
      "step": 2966
    },
    {
      "epoch": 1.082846715328467,
      "grad_norm": 0.1793525665998459,
      "learning_rate": 0.0002,
      "loss": 1.3659,
      "step": 2967
    },
    {
      "epoch": 1.0832116788321169,
      "grad_norm": 0.19276568293571472,
      "learning_rate": 0.0002,
      "loss": 1.4456,
      "step": 2968
    },
    {
      "epoch": 1.0835766423357664,
      "grad_norm": 0.18156202137470245,
      "learning_rate": 0.0002,
      "loss": 1.3754,
      "step": 2969
    },
    {
      "epoch": 1.083941605839416,
      "grad_norm": 0.195973739027977,
      "learning_rate": 0.0002,
      "loss": 1.3677,
      "step": 2970
    },
    {
      "epoch": 1.0843065693430658,
      "grad_norm": 0.18698614835739136,
      "learning_rate": 0.0002,
      "loss": 1.4231,
      "step": 2971
    },
    {
      "epoch": 1.0846715328467154,
      "grad_norm": 0.18434329330921173,
      "learning_rate": 0.0002,
      "loss": 1.3457,
      "step": 2972
    },
    {
      "epoch": 1.085036496350365,
      "grad_norm": 0.1892637461423874,
      "learning_rate": 0.0002,
      "loss": 1.3417,
      "step": 2973
    },
    {
      "epoch": 1.0854014598540145,
      "grad_norm": 0.19298849999904633,
      "learning_rate": 0.0002,
      "loss": 1.3749,
      "step": 2974
    },
    {
      "epoch": 1.0857664233576643,
      "grad_norm": 0.19229689240455627,
      "learning_rate": 0.0002,
      "loss": 1.3821,
      "step": 2975
    },
    {
      "epoch": 1.0861313868613138,
      "grad_norm": 0.1990097165107727,
      "learning_rate": 0.0002,
      "loss": 1.3279,
      "step": 2976
    },
    {
      "epoch": 1.0864963503649636,
      "grad_norm": 0.20265904068946838,
      "learning_rate": 0.0002,
      "loss": 1.3342,
      "step": 2977
    },
    {
      "epoch": 1.0868613138686132,
      "grad_norm": 0.20877724885940552,
      "learning_rate": 0.0002,
      "loss": 1.3549,
      "step": 2978
    },
    {
      "epoch": 1.0872262773722627,
      "grad_norm": 0.1893492192029953,
      "learning_rate": 0.0002,
      "loss": 1.3372,
      "step": 2979
    },
    {
      "epoch": 1.0875912408759123,
      "grad_norm": 0.19266720116138458,
      "learning_rate": 0.0002,
      "loss": 1.3623,
      "step": 2980
    },
    {
      "epoch": 1.087956204379562,
      "grad_norm": 0.1953873634338379,
      "learning_rate": 0.0002,
      "loss": 1.3467,
      "step": 2981
    },
    {
      "epoch": 1.0883211678832116,
      "grad_norm": 0.19460760056972504,
      "learning_rate": 0.0002,
      "loss": 1.3496,
      "step": 2982
    },
    {
      "epoch": 1.0886861313868612,
      "grad_norm": 0.19259989261627197,
      "learning_rate": 0.0002,
      "loss": 1.341,
      "step": 2983
    },
    {
      "epoch": 1.089051094890511,
      "grad_norm": 0.1901838630437851,
      "learning_rate": 0.0002,
      "loss": 1.3315,
      "step": 2984
    },
    {
      "epoch": 1.0894160583941606,
      "grad_norm": 0.19996406137943268,
      "learning_rate": 0.0002,
      "loss": 1.3265,
      "step": 2985
    },
    {
      "epoch": 1.0897810218978101,
      "grad_norm": 0.1888936161994934,
      "learning_rate": 0.0002,
      "loss": 1.3442,
      "step": 2986
    },
    {
      "epoch": 1.09014598540146,
      "grad_norm": 0.19305051863193512,
      "learning_rate": 0.0002,
      "loss": 1.4012,
      "step": 2987
    },
    {
      "epoch": 1.0905109489051095,
      "grad_norm": 0.1818569451570511,
      "learning_rate": 0.0002,
      "loss": 1.3928,
      "step": 2988
    },
    {
      "epoch": 1.090875912408759,
      "grad_norm": 0.19327391684055328,
      "learning_rate": 0.0002,
      "loss": 1.4145,
      "step": 2989
    },
    {
      "epoch": 1.0912408759124088,
      "grad_norm": 0.20359587669372559,
      "learning_rate": 0.0002,
      "loss": 1.4032,
      "step": 2990
    },
    {
      "epoch": 1.0916058394160584,
      "grad_norm": 0.187437042593956,
      "learning_rate": 0.0002,
      "loss": 1.2996,
      "step": 2991
    },
    {
      "epoch": 1.091970802919708,
      "grad_norm": 0.18890738487243652,
      "learning_rate": 0.0002,
      "loss": 1.3795,
      "step": 2992
    },
    {
      "epoch": 1.0923357664233577,
      "grad_norm": 0.19123947620391846,
      "learning_rate": 0.0002,
      "loss": 1.2961,
      "step": 2993
    },
    {
      "epoch": 1.0927007299270073,
      "grad_norm": 0.2009521722793579,
      "learning_rate": 0.0002,
      "loss": 1.4238,
      "step": 2994
    },
    {
      "epoch": 1.0930656934306568,
      "grad_norm": 0.18928052484989166,
      "learning_rate": 0.0002,
      "loss": 1.3021,
      "step": 2995
    },
    {
      "epoch": 1.0934306569343066,
      "grad_norm": 0.18800859153270721,
      "learning_rate": 0.0002,
      "loss": 1.386,
      "step": 2996
    },
    {
      "epoch": 1.0937956204379562,
      "grad_norm": 0.19338972866535187,
      "learning_rate": 0.0002,
      "loss": 1.4031,
      "step": 2997
    },
    {
      "epoch": 1.0941605839416058,
      "grad_norm": 0.18576739728450775,
      "learning_rate": 0.0002,
      "loss": 1.4146,
      "step": 2998
    },
    {
      "epoch": 1.0945255474452555,
      "grad_norm": 0.19094611704349518,
      "learning_rate": 0.0002,
      "loss": 1.4151,
      "step": 2999
    },
    {
      "epoch": 1.094890510948905,
      "grad_norm": 0.18794675171375275,
      "learning_rate": 0.0002,
      "loss": 1.3457,
      "step": 3000
    }
  ],
  "logging_steps": 1,
  "max_steps": 5480,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.7079890935283712e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}