{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 615,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008130081300813009,
      "grad_norm": 6.402216722115984,
      "learning_rate": 1.2903225806451614e-06,
      "loss": 1.1862,
      "step": 1
    },
    {
      "epoch": 0.016260162601626018,
      "grad_norm": 6.432510596947495,
      "learning_rate": 2.580645161290323e-06,
      "loss": 1.1976,
      "step": 2
    },
    {
      "epoch": 0.024390243902439025,
      "grad_norm": 6.335452114935431,
      "learning_rate": 3.870967741935484e-06,
      "loss": 1.1916,
      "step": 3
    },
    {
      "epoch": 0.032520325203252036,
      "grad_norm": 4.880163551949025,
      "learning_rate": 5.161290322580646e-06,
      "loss": 1.144,
      "step": 4
    },
    {
      "epoch": 0.04065040650406504,
      "grad_norm": 2.5680754481932984,
      "learning_rate": 6.451612903225806e-06,
      "loss": 1.0762,
      "step": 5
    },
    {
      "epoch": 0.04878048780487805,
      "grad_norm": 2.079489007841418,
      "learning_rate": 7.741935483870968e-06,
      "loss": 1.0423,
      "step": 6
    },
    {
      "epoch": 0.056910569105691054,
      "grad_norm": 4.440780043047891,
      "learning_rate": 9.03225806451613e-06,
      "loss": 1.0447,
      "step": 7
    },
    {
      "epoch": 0.06504065040650407,
      "grad_norm": 4.70770956056196,
      "learning_rate": 1.0322580645161291e-05,
      "loss": 1.0355,
      "step": 8
    },
    {
      "epoch": 0.07317073170731707,
      "grad_norm": 4.209097118108501,
      "learning_rate": 1.1612903225806453e-05,
      "loss": 0.9958,
      "step": 9
    },
    {
      "epoch": 0.08130081300813008,
      "grad_norm": 4.3551218329872725,
      "learning_rate": 1.2903225806451613e-05,
      "loss": 1.0079,
      "step": 10
    },
    {
      "epoch": 0.08943089430894309,
      "grad_norm": 2.875813896184828,
      "learning_rate": 1.4193548387096776e-05,
      "loss": 0.9634,
      "step": 11
    },
    {
      "epoch": 0.0975609756097561,
      "grad_norm": 2.548512871412817,
      "learning_rate": 1.5483870967741936e-05,
      "loss": 0.9365,
      "step": 12
    },
    {
      "epoch": 0.10569105691056911,
      "grad_norm": 2.6849379031600322,
      "learning_rate": 1.6774193548387098e-05,
      "loss": 0.9152,
      "step": 13
    },
    {
      "epoch": 0.11382113821138211,
      "grad_norm": 1.925902162044746,
      "learning_rate": 1.806451612903226e-05,
      "loss": 0.9115,
      "step": 14
    },
    {
      "epoch": 0.12195121951219512,
      "grad_norm": 1.495815577719039,
      "learning_rate": 1.935483870967742e-05,
      "loss": 0.887,
      "step": 15
    },
    {
      "epoch": 0.13008130081300814,
      "grad_norm": 1.5784022678221434,
      "learning_rate": 2.0645161290322582e-05,
      "loss": 0.8764,
      "step": 16
    },
    {
      "epoch": 0.13821138211382114,
      "grad_norm": 1.3427756691833383,
      "learning_rate": 2.193548387096774e-05,
      "loss": 0.8731,
      "step": 17
    },
    {
      "epoch": 0.14634146341463414,
      "grad_norm": 1.2052691893102891,
      "learning_rate": 2.3225806451612906e-05,
      "loss": 0.8631,
      "step": 18
    },
    {
      "epoch": 0.15447154471544716,
      "grad_norm": 1.1924982503304744,
      "learning_rate": 2.4516129032258067e-05,
      "loss": 0.8555,
      "step": 19
    },
    {
      "epoch": 0.16260162601626016,
      "grad_norm": 1.019331240895464,
      "learning_rate": 2.5806451612903226e-05,
      "loss": 0.8385,
      "step": 20
    },
    {
      "epoch": 0.17073170731707318,
      "grad_norm": 0.8065229407863459,
      "learning_rate": 2.7096774193548387e-05,
      "loss": 0.8349,
      "step": 21
    },
    {
      "epoch": 0.17886178861788618,
      "grad_norm": 0.8855397868826815,
      "learning_rate": 2.8387096774193552e-05,
      "loss": 0.8219,
      "step": 22
    },
    {
      "epoch": 0.18699186991869918,
      "grad_norm": 0.9816411741025896,
      "learning_rate": 2.9677419354838714e-05,
      "loss": 0.819,
      "step": 23
    },
    {
      "epoch": 0.1951219512195122,
      "grad_norm": 0.7940009846237075,
      "learning_rate": 3.096774193548387e-05,
      "loss": 0.8102,
      "step": 24
    },
    {
      "epoch": 0.2032520325203252,
      "grad_norm": 1.0151129025596008,
      "learning_rate": 3.2258064516129034e-05,
      "loss": 0.8087,
      "step": 25
    },
    {
      "epoch": 0.21138211382113822,
      "grad_norm": 1.6594609242846412,
      "learning_rate": 3.3548387096774195e-05,
      "loss": 0.8082,
      "step": 26
    },
    {
      "epoch": 0.21951219512195122,
      "grad_norm": 1.0111709011142294,
      "learning_rate": 3.483870967741936e-05,
      "loss": 0.8011,
      "step": 27
    },
    {
      "epoch": 0.22764227642276422,
      "grad_norm": 1.6860488068700434,
      "learning_rate": 3.612903225806452e-05,
      "loss": 0.7955,
      "step": 28
    },
    {
      "epoch": 0.23577235772357724,
      "grad_norm": 1.6644800225150644,
      "learning_rate": 3.741935483870968e-05,
      "loss": 0.7905,
      "step": 29
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 1.164001921392695,
      "learning_rate": 3.870967741935484e-05,
      "loss": 0.7967,
      "step": 30
    },
    {
      "epoch": 0.25203252032520324,
      "grad_norm": 3.0141988139246143,
      "learning_rate": 4e-05,
      "loss": 0.7966,
      "step": 31
    },
    {
      "epoch": 0.2601626016260163,
      "grad_norm": 2.013503884430171,
      "learning_rate": 4.1290322580645165e-05,
      "loss": 0.7897,
      "step": 32
    },
    {
      "epoch": 0.2682926829268293,
      "grad_norm": 2.517434579137323,
      "learning_rate": 4.2580645161290327e-05,
      "loss": 0.7973,
      "step": 33
    },
    {
      "epoch": 0.2764227642276423,
      "grad_norm": 2.280298943389002,
      "learning_rate": 4.387096774193548e-05,
      "loss": 0.7864,
      "step": 34
    },
    {
      "epoch": 0.2845528455284553,
      "grad_norm": 1.9409391084913032,
      "learning_rate": 4.5161290322580657e-05,
      "loss": 0.7829,
      "step": 35
    },
    {
      "epoch": 0.2926829268292683,
      "grad_norm": 1.7435273412874077,
      "learning_rate": 4.645161290322581e-05,
      "loss": 0.7747,
      "step": 36
    },
    {
      "epoch": 0.3008130081300813,
      "grad_norm": 2.304936749421656,
      "learning_rate": 4.774193548387097e-05,
      "loss": 0.7872,
      "step": 37
    },
    {
      "epoch": 0.3089430894308943,
      "grad_norm": 1.8552436642521173,
      "learning_rate": 4.9032258064516135e-05,
      "loss": 0.7712,
      "step": 38
    },
    {
      "epoch": 0.3170731707317073,
      "grad_norm": 2.3233734721418977,
      "learning_rate": 5.0322580645161296e-05,
      "loss": 0.7683,
      "step": 39
    },
    {
      "epoch": 0.3252032520325203,
      "grad_norm": 2.1826657096322246,
      "learning_rate": 5.161290322580645e-05,
      "loss": 0.7676,
      "step": 40
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 3.095174274286662,
      "learning_rate": 5.290322580645161e-05,
      "loss": 0.7502,
      "step": 41
    },
    {
      "epoch": 0.34146341463414637,
      "grad_norm": 2.8070895753469585,
      "learning_rate": 5.4193548387096774e-05,
      "loss": 0.7714,
      "step": 42
    },
    {
      "epoch": 0.34959349593495936,
      "grad_norm": 1.0021414815386867,
      "learning_rate": 5.548387096774194e-05,
      "loss": 0.7494,
      "step": 43
    },
    {
      "epoch": 0.35772357723577236,
      "grad_norm": 323.78462892927485,
      "learning_rate": 5.6774193548387104e-05,
      "loss": 0.9245,
      "step": 44
    },
    {
      "epoch": 0.36585365853658536,
      "grad_norm": 38.611418944019455,
      "learning_rate": 5.8064516129032266e-05,
      "loss": 0.9615,
      "step": 45
    },
    {
      "epoch": 0.37398373983739835,
      "grad_norm": 9.484201470967708,
      "learning_rate": 5.935483870967743e-05,
      "loss": 1.0603,
      "step": 46
    },
    {
      "epoch": 0.3821138211382114,
      "grad_norm": 13.953853718983199,
      "learning_rate": 6.064516129032258e-05,
      "loss": 0.9098,
      "step": 47
    },
    {
      "epoch": 0.3902439024390244,
      "grad_norm": 28.972100526178775,
      "learning_rate": 6.193548387096774e-05,
      "loss": 0.8934,
      "step": 48
    },
    {
      "epoch": 0.3983739837398374,
      "grad_norm": 3.062835300182985,
      "learning_rate": 6.32258064516129e-05,
      "loss": 0.8308,
      "step": 49
    },
    {
      "epoch": 0.4065040650406504,
      "grad_norm": 1.8359307093784598,
      "learning_rate": 6.451612903225807e-05,
      "loss": 0.7809,
      "step": 50
    },
    {
      "epoch": 0.4146341463414634,
      "grad_norm": 1.9856153953431313,
      "learning_rate": 6.580645161290323e-05,
      "loss": 0.7764,
      "step": 51
    },
    {
      "epoch": 0.42276422764227645,
      "grad_norm": 1.8616941754001088,
      "learning_rate": 6.709677419354839e-05,
      "loss": 0.7869,
      "step": 52
    },
    {
      "epoch": 0.43089430894308944,
      "grad_norm": 1.3601792290338877,
      "learning_rate": 6.838709677419355e-05,
      "loss": 0.7535,
      "step": 53
    },
    {
      "epoch": 0.43902439024390244,
      "grad_norm": 1.5501876361900562,
      "learning_rate": 6.967741935483871e-05,
      "loss": 0.7693,
      "step": 54
    },
    {
      "epoch": 0.44715447154471544,
      "grad_norm": 1.8972787541309308,
      "learning_rate": 7.096774193548388e-05,
      "loss": 0.7741,
      "step": 55
    },
    {
      "epoch": 0.45528455284552843,
      "grad_norm": 0.9172361519689148,
      "learning_rate": 7.225806451612904e-05,
      "loss": 0.7643,
      "step": 56
    },
    {
      "epoch": 0.4634146341463415,
      "grad_norm": 1.5421950244625509,
      "learning_rate": 7.35483870967742e-05,
      "loss": 0.7621,
      "step": 57
    },
    {
      "epoch": 0.4715447154471545,
      "grad_norm": 1.8169585969167579,
      "learning_rate": 7.483870967741936e-05,
      "loss": 0.7716,
      "step": 58
    },
    {
      "epoch": 0.4796747967479675,
      "grad_norm": 1.143466822875425,
      "learning_rate": 7.612903225806452e-05,
      "loss": 0.7515,
      "step": 59
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 2.1570876012440934,
      "learning_rate": 7.741935483870968e-05,
      "loss": 0.7569,
      "step": 60
    },
    {
      "epoch": 0.4959349593495935,
      "grad_norm": 1.2848579533107485,
      "learning_rate": 7.870967741935484e-05,
      "loss": 0.7476,
      "step": 61
    },
    {
      "epoch": 0.5040650406504065,
      "grad_norm": 1.5112865885263438,
      "learning_rate": 8e-05,
      "loss": 0.7519,
      "step": 62
    },
    {
      "epoch": 0.5121951219512195,
      "grad_norm": 1.4230311824961481,
      "learning_rate": 7.999935452665835e-05,
      "loss": 0.7451,
      "step": 63
    },
    {
      "epoch": 0.5203252032520326,
      "grad_norm": 1.9248878120715245,
      "learning_rate": 7.999741812746515e-05,
      "loss": 0.7431,
      "step": 64
    },
    {
      "epoch": 0.5284552845528455,
      "grad_norm": 1.0707837991969031,
      "learning_rate": 7.999419086491513e-05,
      "loss": 0.7382,
      "step": 65
    },
    {
      "epoch": 0.5365853658536586,
      "grad_norm": 2.512202332944163,
      "learning_rate": 7.998967284316387e-05,
      "loss": 0.7465,
      "step": 66
    },
    {
      "epoch": 0.5447154471544715,
      "grad_norm": 1.48775725259063,
      "learning_rate": 7.998386420802452e-05,
      "loss": 0.7432,
      "step": 67
    },
    {
      "epoch": 0.5528455284552846,
      "grad_norm": 2.121636852058249,
      "learning_rate": 7.997676514696301e-05,
      "loss": 0.7409,
      "step": 68
    },
    {
      "epoch": 0.5609756097560976,
      "grad_norm": 1.6783577853464693,
      "learning_rate": 7.996837588909207e-05,
      "loss": 0.7432,
      "step": 69
    },
    {
      "epoch": 0.5691056910569106,
      "grad_norm": 1.285435691666245,
      "learning_rate": 7.995869670516387e-05,
      "loss": 0.7384,
      "step": 70
    },
    {
      "epoch": 0.5772357723577236,
      "grad_norm": 1.9461106213021069,
      "learning_rate": 7.994772790756112e-05,
      "loss": 0.7374,
      "step": 71
    },
    {
      "epoch": 0.5853658536585366,
      "grad_norm": 1.3500678654823275,
      "learning_rate": 7.993546985028714e-05,
      "loss": 0.745,
      "step": 72
    },
    {
      "epoch": 0.5934959349593496,
      "grad_norm": 1.7778999763178327,
      "learning_rate": 7.992192292895441e-05,
      "loss": 0.7268,
      "step": 73
    },
    {
      "epoch": 0.6016260162601627,
      "grad_norm": 1.1760779956597847,
      "learning_rate": 7.990708758077177e-05,
      "loss": 0.727,
      "step": 74
    },
    {
      "epoch": 0.6097560975609756,
      "grad_norm": 1.5431774749680818,
      "learning_rate": 7.989096428453027e-05,
      "loss": 0.721,
      "step": 75
    },
    {
      "epoch": 0.6178861788617886,
      "grad_norm": 1.089238490144563,
      "learning_rate": 7.987355356058784e-05,
      "loss": 0.7109,
      "step": 76
    },
    {
      "epoch": 0.6260162601626016,
      "grad_norm": 0.9660825661045613,
      "learning_rate": 7.985485597085238e-05,
      "loss": 0.719,
      "step": 77
    },
    {
      "epoch": 0.6341463414634146,
      "grad_norm": 1.5632316157906598,
      "learning_rate": 7.983487211876365e-05,
      "loss": 0.7108,
      "step": 78
    },
    {
      "epoch": 0.6422764227642277,
      "grad_norm": 1.4325964755345229,
      "learning_rate": 7.981360264927388e-05,
      "loss": 0.7199,
      "step": 79
    },
    {
      "epoch": 0.6504065040650406,
      "grad_norm": 1.294940416051851,
      "learning_rate": 7.979104824882682e-05,
      "loss": 0.7113,
      "step": 80
    },
    {
      "epoch": 0.6585365853658537,
      "grad_norm": 1.0520625697449317,
      "learning_rate": 7.976720964533569e-05,
      "loss": 0.7118,
      "step": 81
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.8594821446498264,
      "learning_rate": 7.974208760815964e-05,
      "loss": 0.7202,
      "step": 82
    },
    {
      "epoch": 0.6747967479674797,
      "grad_norm": 1.1558963148579111,
      "learning_rate": 7.971568294807894e-05,
      "loss": 0.7147,
      "step": 83
    },
    {
      "epoch": 0.6829268292682927,
      "grad_norm": 1.6549183632889652,
      "learning_rate": 7.968799651726879e-05,
      "loss": 0.7148,
      "step": 84
    },
    {
      "epoch": 0.6910569105691057,
      "grad_norm": 0.9880687041801252,
      "learning_rate": 7.965902920927186e-05,
      "loss": 0.7012,
      "step": 85
    },
    {
      "epoch": 0.6991869918699187,
      "grad_norm": 1.7751627165774728,
      "learning_rate": 7.962878195896938e-05,
      "loss": 0.7124,
      "step": 86
    },
    {
      "epoch": 0.7073170731707317,
      "grad_norm": 1.3238133684103306,
      "learning_rate": 7.959725574255106e-05,
      "loss": 0.7029,
      "step": 87
    },
    {
      "epoch": 0.7154471544715447,
      "grad_norm": 1.1346678088235465,
      "learning_rate": 7.956445157748347e-05,
      "loss": 0.7022,
      "step": 88
    },
    {
      "epoch": 0.7235772357723578,
      "grad_norm": 1.3810825094344603,
      "learning_rate": 7.953037052247737e-05,
      "loss": 0.7102,
      "step": 89
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 1.340331365140534,
      "learning_rate": 7.949501367745335e-05,
      "loss": 0.6909,
      "step": 90
    },
    {
      "epoch": 0.7398373983739838,
      "grad_norm": 1.0105470891883848,
      "learning_rate": 7.945838218350645e-05,
      "loss": 0.7012,
      "step": 91
    },
    {
      "epoch": 0.7479674796747967,
      "grad_norm": 1.2567105454974923,
      "learning_rate": 7.942047722286933e-05,
      "loss": 0.6854,
      "step": 92
    },
    {
      "epoch": 0.7560975609756098,
      "grad_norm": 1.017559556241571,
      "learning_rate": 7.938130001887406e-05,
      "loss": 0.7018,
      "step": 93
    },
    {
      "epoch": 0.7642276422764228,
      "grad_norm": 1.02671040396442,
      "learning_rate": 7.934085183591268e-05,
      "loss": 0.687,
      "step": 94
    },
    {
      "epoch": 0.7723577235772358,
      "grad_norm": 1.5054957743058477,
      "learning_rate": 7.929913397939639e-05,
      "loss": 0.7024,
      "step": 95
    },
    {
      "epoch": 0.7804878048780488,
      "grad_norm": 2.0257707931527804,
      "learning_rate": 7.925614779571337e-05,
      "loss": 0.6975,
      "step": 96
    },
    {
      "epoch": 0.7886178861788617,
      "grad_norm": 1.1576757181160013,
      "learning_rate": 7.921189467218546e-05,
      "loss": 0.6977,
      "step": 97
    },
    {
      "epoch": 0.7967479674796748,
      "grad_norm": 3.000271466612812,
      "learning_rate": 7.916637603702318e-05,
      "loss": 0.7114,
      "step": 98
    },
    {
      "epoch": 0.8048780487804879,
      "grad_norm": 2.5600149472436415,
      "learning_rate": 7.911959335927983e-05,
      "loss": 0.7142,
      "step": 99
    },
    {
      "epoch": 0.8130081300813008,
      "grad_norm": 1.87246290040283,
      "learning_rate": 7.907154814880398e-05,
      "loss": 0.7033,
      "step": 100
    },
    {
      "epoch": 0.8211382113821138,
      "grad_norm": 1.8997229736806345,
      "learning_rate": 7.902224195619076e-05,
      "loss": 0.701,
      "step": 101
    },
    {
      "epoch": 0.8292682926829268,
      "grad_norm": 1.5049890620524067,
      "learning_rate": 7.89716763727318e-05,
      "loss": 0.6906,
      "step": 102
    },
    {
      "epoch": 0.8373983739837398,
      "grad_norm": 1.4143332235378092,
      "learning_rate": 7.891985303036391e-05,
      "loss": 0.6994,
      "step": 103
    },
    {
      "epoch": 0.8455284552845529,
      "grad_norm": 1.208065924350305,
      "learning_rate": 7.88667736016164e-05,
      "loss": 0.6834,
      "step": 104
    },
    {
      "epoch": 0.8536585365853658,
      "grad_norm": 1.3350236491232776,
      "learning_rate": 7.881243979955709e-05,
      "loss": 0.6876,
      "step": 105
    },
    {
      "epoch": 0.8617886178861789,
      "grad_norm": 0.8731535998085196,
      "learning_rate": 7.8756853377737e-05,
      "loss": 0.6727,
      "step": 106
    },
    {
      "epoch": 0.8699186991869918,
      "grad_norm": 1.0574561140041197,
      "learning_rate": 7.87000161301338e-05,
      "loss": 0.6832,
      "step": 107
    },
    {
      "epoch": 0.8780487804878049,
      "grad_norm": 1.146455048677622,
      "learning_rate": 7.86419298910939e-05,
      "loss": 0.669,
      "step": 108
    },
    {
      "epoch": 0.8861788617886179,
      "grad_norm": 1.3579230002533929,
      "learning_rate": 7.858259653527326e-05,
      "loss": 0.6774,
      "step": 109
    },
    {
      "epoch": 0.8943089430894309,
      "grad_norm": 0.6266775665026565,
      "learning_rate": 7.852201797757683e-05,
      "loss": 0.6707,
      "step": 110
    },
    {
      "epoch": 0.9024390243902439,
      "grad_norm": 1.0619614573527578,
      "learning_rate": 7.846019617309683e-05,
      "loss": 0.668,
      "step": 111
    },
    {
      "epoch": 0.9105691056910569,
      "grad_norm": 1.4967099125879249,
      "learning_rate": 7.839713311704957e-05,
      "loss": 0.671,
      "step": 112
    },
    {
      "epoch": 0.9186991869918699,
      "grad_norm": 0.9503481549704464,
      "learning_rate": 7.833283084471117e-05,
      "loss": 0.6744,
      "step": 113
    },
    {
      "epoch": 0.926829268292683,
      "grad_norm": 1.4950455746004616,
      "learning_rate": 7.826729143135172e-05,
      "loss": 0.6698,
      "step": 114
    },
    {
      "epoch": 0.9349593495934959,
      "grad_norm": 0.8348799449902218,
      "learning_rate": 7.820051699216844e-05,
      "loss": 0.6734,
      "step": 115
    },
    {
      "epoch": 0.943089430894309,
      "grad_norm": 1.4171902801073721,
      "learning_rate": 7.813250968221736e-05,
      "loss": 0.6675,
      "step": 116
    },
    {
      "epoch": 0.9512195121951219,
      "grad_norm": 0.9763758996292249,
      "learning_rate": 7.806327169634374e-05,
      "loss": 0.6681,
      "step": 117
    },
    {
      "epoch": 0.959349593495935,
      "grad_norm": 1.3478372065980402,
      "learning_rate": 7.799280526911132e-05,
      "loss": 0.6716,
      "step": 118
    },
    {
      "epoch": 0.967479674796748,
      "grad_norm": 1.0949409276953657,
      "learning_rate": 7.792111267473008e-05,
      "loss": 0.6695,
      "step": 119
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.8573172924086485,
      "learning_rate": 7.784819622698294e-05,
      "loss": 0.6667,
      "step": 120
    },
    {
      "epoch": 0.983739837398374,
      "grad_norm": 1.0611187540935,
      "learning_rate": 7.77740582791511e-05,
      "loss": 0.6698,
      "step": 121
    },
    {
      "epoch": 0.991869918699187,
      "grad_norm": 1.5133992736831428,
      "learning_rate": 7.769870122393797e-05,
      "loss": 0.6774,
      "step": 122
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.019040704327121,
      "learning_rate": 7.762212749339206e-05,
      "loss": 0.6651,
      "step": 123
    },
    {
      "epoch": 1.008130081300813,
      "grad_norm": 0.9212106703322404,
      "learning_rate": 7.754433955882848e-05,
      "loss": 0.6326,
      "step": 124
    },
    {
      "epoch": 1.016260162601626,
      "grad_norm": 1.1962859058487172,
      "learning_rate": 7.746533993074912e-05,
      "loss": 0.6349,
      "step": 125
    },
    {
      "epoch": 1.024390243902439,
      "grad_norm": 1.2340639746620863,
      "learning_rate": 7.738513115876169e-05,
      "loss": 0.6325,
      "step": 126
    },
    {
      "epoch": 1.032520325203252,
      "grad_norm": 0.9617021735862622,
      "learning_rate": 7.730371583149736e-05,
      "loss": 0.6314,
      "step": 127
    },
    {
      "epoch": 1.040650406504065,
      "grad_norm": 1.4790672646400322,
      "learning_rate": 7.722109657652735e-05,
      "loss": 0.6278,
      "step": 128
    },
    {
      "epoch": 1.048780487804878,
      "grad_norm": 0.6367366305993373,
      "learning_rate": 7.713727606027794e-05,
      "loss": 0.6199,
      "step": 129
    },
    {
      "epoch": 1.056910569105691,
      "grad_norm": 1.5769862399748125,
      "learning_rate": 7.70522569879446e-05,
      "loss": 0.6312,
      "step": 130
    },
    {
      "epoch": 1.065040650406504,
      "grad_norm": 0.8327281340132524,
      "learning_rate": 7.696604210340455e-05,
      "loss": 0.6352,
      "step": 131
    },
    {
      "epoch": 1.0731707317073171,
      "grad_norm": 1.2396634934167161,
      "learning_rate": 7.687863418912827e-05,
      "loss": 0.6241,
      "step": 132
    },
    {
      "epoch": 1.08130081300813,
      "grad_norm": 0.8469385712980174,
      "learning_rate": 7.679003606608969e-05,
      "loss": 0.6299,
      "step": 133
    },
    {
      "epoch": 1.089430894308943,
      "grad_norm": 1.1070793879736758,
      "learning_rate": 7.670025059367515e-05,
      "loss": 0.6217,
      "step": 134
    },
    {
      "epoch": 1.0975609756097562,
      "grad_norm": 1.1332796092627617,
      "learning_rate": 7.660928066959108e-05,
      "loss": 0.6309,
      "step": 135
    },
    {
      "epoch": 1.1056910569105691,
      "grad_norm": 0.7552437419135463,
      "learning_rate": 7.651712922977052e-05,
      "loss": 0.6264,
      "step": 136
    },
    {
      "epoch": 1.113821138211382,
      "grad_norm": 0.8823781612742758,
      "learning_rate": 7.642379924827837e-05,
      "loss": 0.6259,
      "step": 137
    },
    {
      "epoch": 1.1219512195121952,
      "grad_norm": 1.1666484171615628,
      "learning_rate": 7.632929373721537e-05,
      "loss": 0.622,
      "step": 138
    },
    {
      "epoch": 1.1300813008130082,
      "grad_norm": 1.077735976988468,
      "learning_rate": 7.623361574662094e-05,
      "loss": 0.6226,
      "step": 139
    },
    {
      "epoch": 1.1382113821138211,
      "grad_norm": 1.0878258999152588,
      "learning_rate": 7.613676836437468e-05,
      "loss": 0.6248,
      "step": 140
    },
    {
      "epoch": 1.146341463414634,
      "grad_norm": 0.9671818648460052,
      "learning_rate": 7.603875471609677e-05,
      "loss": 0.6202,
      "step": 141
    },
    {
      "epoch": 1.1544715447154472,
      "grad_norm": 1.065900625442339,
      "learning_rate": 7.593957796504706e-05,
      "loss": 0.6175,
      "step": 142
    },
    {
      "epoch": 1.1626016260162602,
      "grad_norm": 1.4044297826770507,
      "learning_rate": 7.5839241312023e-05,
      "loss": 0.6206,
      "step": 143
    },
    {
      "epoch": 1.170731707317073,
      "grad_norm": 0.5931542892986451,
      "learning_rate": 7.573774799525635e-05,
      "loss": 0.6292,
      "step": 144
    },
    {
      "epoch": 1.1788617886178863,
      "grad_norm": 1.0234454430626854,
      "learning_rate": 7.563510129030857e-05,
      "loss": 0.6093,
      "step": 145
    },
    {
      "epoch": 1.1869918699186992,
      "grad_norm": 1.2734517830728405,
      "learning_rate": 7.553130450996528e-05,
      "loss": 0.6236,
      "step": 146
    },
    {
      "epoch": 1.1951219512195121,
      "grad_norm": 1.1632612507017766,
      "learning_rate": 7.542636100412921e-05,
      "loss": 0.616,
      "step": 147
    },
    {
      "epoch": 1.203252032520325,
      "grad_norm": 0.7000887219441574,
      "learning_rate": 7.532027415971213e-05,
      "loss": 0.6149,
      "step": 148
    },
    {
      "epoch": 1.2113821138211383,
      "grad_norm": 0.6278237668714192,
      "learning_rate": 7.521304740052553e-05,
      "loss": 0.6204,
      "step": 149
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 1.0708488655432584,
      "learning_rate": 7.510468418717015e-05,
      "loss": 0.6112,
      "step": 150
    },
    {
      "epoch": 1.2276422764227641,
      "grad_norm": 0.9769446406238423,
      "learning_rate": 7.499518801692424e-05,
      "loss": 0.6175,
      "step": 151
    },
    {
      "epoch": 1.2357723577235773,
      "grad_norm": 0.9159892303761836,
      "learning_rate": 7.488456242363078e-05,
      "loss": 0.6126,
      "step": 152
    },
    {
      "epoch": 1.2439024390243902,
      "grad_norm": 1.1254859008274773,
      "learning_rate": 7.477281097758331e-05,
      "loss": 0.6134,
      "step": 153
    },
    {
      "epoch": 1.2520325203252032,
      "grad_norm": 1.0072854995448057,
      "learning_rate": 7.46599372854108e-05,
      "loss": 0.6078,
      "step": 154
    },
    {
      "epoch": 1.2601626016260163,
      "grad_norm": 0.9275505839822971,
      "learning_rate": 7.454594498996124e-05,
      "loss": 0.6209,
      "step": 155
    },
    {
      "epoch": 1.2682926829268293,
      "grad_norm": 1.0909383506629609,
      "learning_rate": 7.443083777018399e-05,
      "loss": 0.6169,
      "step": 156
    },
    {
      "epoch": 1.2764227642276422,
      "grad_norm": 1.1961241528408824,
      "learning_rate": 7.431461934101114e-05,
      "loss": 0.6172,
      "step": 157
    },
    {
      "epoch": 1.2845528455284554,
      "grad_norm": 0.964346891253909,
      "learning_rate": 7.41972934532376e-05,
      "loss": 0.6121,
      "step": 158
    },
    {
      "epoch": 1.2926829268292683,
      "grad_norm": 1.059614943095973,
      "learning_rate": 7.407886389340002e-05,
      "loss": 0.6106,
      "step": 159
    },
    {
      "epoch": 1.3008130081300813,
      "grad_norm": 1.0589560624317955,
      "learning_rate": 7.395933448365457e-05,
      "loss": 0.6109,
      "step": 160
    },
    {
      "epoch": 1.3089430894308944,
      "grad_norm": 1.0834162965880576,
      "learning_rate": 7.383870908165362e-05,
      "loss": 0.6094,
      "step": 161
    },
    {
      "epoch": 1.3170731707317074,
      "grad_norm": 1.1964363361879669,
      "learning_rate": 7.371699158042126e-05,
      "loss": 0.6136,
      "step": 162
    },
    {
      "epoch": 1.3252032520325203,
      "grad_norm": 0.8041230840504787,
      "learning_rate": 7.359418590822757e-05,
      "loss": 0.613,
      "step": 163
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.6390583765223274,
      "learning_rate": 7.347029602846197e-05,
      "loss": 0.6141,
      "step": 164
    },
    {
      "epoch": 1.3414634146341464,
      "grad_norm": 0.5037164959095444,
      "learning_rate": 7.334532593950516e-05,
      "loss": 0.6131,
      "step": 165
    },
    {
      "epoch": 1.3495934959349594,
      "grad_norm": 0.7367317148829811,
      "learning_rate": 7.32192796746002e-05,
      "loss": 0.6141,
      "step": 166
    },
    {
      "epoch": 1.3577235772357723,
      "grad_norm": 1.0032916384773496,
      "learning_rate": 7.30921613017223e-05,
      "loss": 0.6162,
      "step": 167
    },
    {
      "epoch": 1.3658536585365852,
      "grad_norm": 1.231319125166182,
      "learning_rate": 7.296397492344747e-05,
      "loss": 0.6118,
      "step": 168
    },
    {
      "epoch": 1.3739837398373984,
      "grad_norm": 0.8265860631829808,
      "learning_rate": 7.283472467682023e-05,
      "loss": 0.613,
      "step": 169
    },
    {
      "epoch": 1.3821138211382114,
      "grad_norm": 1.0185111102530426,
      "learning_rate": 7.270441473322e-05,
      "loss": 0.615,
      "step": 170
    },
    {
      "epoch": 1.3902439024390243,
      "grad_norm": 0.8729003187697826,
      "learning_rate": 7.257304929822652e-05,
      "loss": 0.611,
      "step": 171
    },
    {
      "epoch": 1.3983739837398375,
      "grad_norm": 0.677877200935124,
      "learning_rate": 7.244063261148413e-05,
      "loss": 0.6222,
      "step": 172
    },
    {
      "epoch": 1.4065040650406504,
      "grad_norm": 0.8431896720942657,
      "learning_rate": 7.230716894656485e-05,
      "loss": 0.6157,
      "step": 173
    },
    {
      "epoch": 1.4146341463414633,
      "grad_norm": 0.988893030982757,
      "learning_rate": 7.21726626108306e-05,
      "loss": 0.5955,
      "step": 174
    },
    {
      "epoch": 1.4227642276422765,
      "grad_norm": 1.4022544718308547,
      "learning_rate": 7.203711794529406e-05,
      "loss": 0.6155,
      "step": 175
    },
    {
      "epoch": 1.4308943089430894,
      "grad_norm": 0.6768873931220528,
      "learning_rate": 7.190053932447866e-05,
      "loss": 0.6162,
      "step": 176
    },
    {
      "epoch": 1.4390243902439024,
      "grad_norm": 0.9972285933897289,
      "learning_rate": 7.176293115627734e-05,
      "loss": 0.6049,
      "step": 177
    },
    {
      "epoch": 1.4471544715447155,
      "grad_norm": 1.6382694606341253,
      "learning_rate": 7.162429788181029e-05,
      "loss": 0.6191,
      "step": 178
    },
    {
      "epoch": 1.4552845528455285,
      "grad_norm": 0.8195977129318162,
      "learning_rate": 7.148464397528165e-05,
      "loss": 0.6142,
      "step": 179
    },
    {
      "epoch": 1.4634146341463414,
      "grad_norm": 1.7478054797315026,
      "learning_rate": 7.134397394383512e-05,
      "loss": 0.6199,
      "step": 180
    },
    {
      "epoch": 1.4715447154471546,
      "grad_norm": 1.244098806863729,
      "learning_rate": 7.12022923274085e-05,
      "loss": 0.6091,
      "step": 181
    },
    {
      "epoch": 1.4796747967479675,
      "grad_norm": 1.7274368559172673,
      "learning_rate": 7.105960369858705e-05,
      "loss": 0.6236,
      "step": 182
    },
    {
      "epoch": 1.4878048780487805,
      "grad_norm": 1.3313060515171165,
      "learning_rate": 7.09159126624561e-05,
      "loss": 0.6169,
      "step": 183
    },
    {
      "epoch": 1.4959349593495934,
      "grad_norm": 1.3879424412033876,
      "learning_rate": 7.077122385645231e-05,
      "loss": 0.609,
      "step": 184
    },
    {
      "epoch": 1.5040650406504064,
      "grad_norm": 1.3211432085244545,
      "learning_rate": 7.062554195021404e-05,
      "loss": 0.6101,
      "step": 185
    },
    {
      "epoch": 1.5121951219512195,
      "grad_norm": 1.395082952018036,
      "learning_rate": 7.047887164543064e-05,
      "loss": 0.6091,
      "step": 186
    },
    {
      "epoch": 1.5203252032520327,
      "grad_norm": 1.1225542141625535,
      "learning_rate": 7.033121767569067e-05,
      "loss": 0.6133,
      "step": 187
    },
    {
      "epoch": 1.5284552845528454,
      "grad_norm": 1.4739585703307818,
      "learning_rate": 7.018258480632922e-05,
      "loss": 0.6047,
      "step": 188
    },
    {
      "epoch": 1.5365853658536586,
      "grad_norm": 1.0517090529732016,
      "learning_rate": 7.003297783427401e-05,
      "loss": 0.616,
      "step": 189
    },
    {
      "epoch": 1.5447154471544715,
      "grad_norm": 1.5451168035096672,
      "learning_rate": 6.988240158789067e-05,
      "loss": 0.6137,
      "step": 190
    },
    {
      "epoch": 1.5528455284552845,
      "grad_norm": 1.2311554839846426,
      "learning_rate": 6.973086092682684e-05,
      "loss": 0.5978,
      "step": 191
    },
    {
      "epoch": 1.5609756097560976,
      "grad_norm": 1.2134545032700186,
      "learning_rate": 6.957836074185536e-05,
      "loss": 0.6065,
      "step": 192
    },
    {
      "epoch": 1.5691056910569106,
      "grad_norm": 0.9267589307380317,
      "learning_rate": 6.942490595471643e-05,
      "loss": 0.6039,
      "step": 193
    },
    {
      "epoch": 1.5772357723577235,
      "grad_norm": 0.9444157108627714,
      "learning_rate": 6.927050151795877e-05,
      "loss": 0.6064,
      "step": 194
    },
    {
      "epoch": 1.5853658536585367,
      "grad_norm": 0.9721699968973354,
      "learning_rate": 6.911515241477978e-05,
      "loss": 0.6052,
      "step": 195
    },
    {
      "epoch": 1.5934959349593496,
      "grad_norm": 1.2266911426285843,
      "learning_rate": 6.895886365886465e-05,
      "loss": 0.6133,
      "step": 196
    },
    {
      "epoch": 1.6016260162601625,
      "grad_norm": 0.8522539272402457,
      "learning_rate": 6.880164029422471e-05,
      "loss": 0.6035,
      "step": 197
    },
    {
      "epoch": 1.6097560975609757,
      "grad_norm": 0.7816317498625392,
      "learning_rate": 6.864348739503446e-05,
      "loss": 0.5984,
      "step": 198
    },
    {
      "epoch": 1.6178861788617886,
      "grad_norm": 0.8795998716654578,
      "learning_rate": 6.848441006546793e-05,
      "loss": 0.6094,
      "step": 199
    },
    {
      "epoch": 1.6260162601626016,
      "grad_norm": 1.1706920762745012,
      "learning_rate": 6.832441343953389e-05,
      "loss": 0.6021,
      "step": 200
    },
    {
      "epoch": 1.6341463414634148,
      "grad_norm": 0.9315837812702746,
      "learning_rate": 6.816350268091017e-05,
      "loss": 0.6064,
      "step": 201
    },
    {
      "epoch": 1.6422764227642277,
      "grad_norm": 0.9089304196640906,
      "learning_rate": 6.800168298277702e-05,
      "loss": 0.6048,
      "step": 202
    },
    {
      "epoch": 1.6504065040650406,
      "grad_norm": 1.0148216395778233,
      "learning_rate": 6.783895956764954e-05,
      "loss": 0.5984,
      "step": 203
    },
    {
      "epoch": 1.6585365853658538,
      "grad_norm": 0.9681350244016154,
      "learning_rate": 6.767533768720901e-05,
      "loss": 0.6024,
      "step": 204
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.7699177077891188,
      "learning_rate": 6.751082262213357e-05,
      "loss": 0.5991,
      "step": 205
    },
    {
      "epoch": 1.6747967479674797,
      "grad_norm": 0.6759903611884919,
      "learning_rate": 6.734541968192762e-05,
      "loss": 0.6049,
      "step": 206
    },
    {
      "epoch": 1.6829268292682928,
      "grad_norm": 0.6915667814486538,
      "learning_rate": 6.717913420475064e-05,
      "loss": 0.6092,
      "step": 207
    },
    {
      "epoch": 1.6910569105691056,
      "grad_norm": 0.5478125571897071,
      "learning_rate": 6.70119715572447e-05,
      "loss": 0.5984,
      "step": 208
    },
    {
      "epoch": 1.6991869918699187,
      "grad_norm": 0.4996414308447817,
      "learning_rate": 6.684393713436148e-05,
      "loss": 0.5943,
      "step": 209
    },
    {
      "epoch": 1.7073170731707317,
      "grad_norm": 0.6433726707857171,
      "learning_rate": 6.667503635918797e-05,
      "loss": 0.6015,
      "step": 210
    },
    {
      "epoch": 1.7154471544715446,
      "grad_norm": 0.6257402481144321,
      "learning_rate": 6.650527468277158e-05,
      "loss": 0.5997,
      "step": 211
    },
    {
      "epoch": 1.7235772357723578,
      "grad_norm": 0.5424184403651616,
      "learning_rate": 6.633465758394414e-05,
      "loss": 0.5973,
      "step": 212
    },
    {
      "epoch": 1.7317073170731707,
      "grad_norm": 0.5591513194711705,
      "learning_rate": 6.616319056914508e-05,
      "loss": 0.6046,
      "step": 213
    },
    {
      "epoch": 1.7398373983739837,
      "grad_norm": 0.57254458506494,
      "learning_rate": 6.599087917224376e-05,
      "loss": 0.5984,
      "step": 214
    },
    {
      "epoch": 1.7479674796747968,
      "grad_norm": 0.43357042648269545,
      "learning_rate": 6.581772895436085e-05,
      "loss": 0.6021,
      "step": 215
    },
    {
      "epoch": 1.7560975609756098,
      "grad_norm": 0.3575399607048083,
      "learning_rate": 6.56437455036888e-05,
      "loss": 0.596,
      "step": 216
    },
    {
      "epoch": 1.7642276422764227,
      "grad_norm": 0.45183981321384764,
      "learning_rate": 6.546893443531163e-05,
      "loss": 0.5932,
      "step": 217
    },
    {
      "epoch": 1.7723577235772359,
      "grad_norm": 0.4903172407933418,
      "learning_rate": 6.529330139102351e-05,
      "loss": 0.5981,
      "step": 218
    },
    {
      "epoch": 1.7804878048780488,
      "grad_norm": 0.518187172258692,
      "learning_rate": 6.511685203914687e-05,
      "loss": 0.5895,
      "step": 219
    },
    {
      "epoch": 1.7886178861788617,
      "grad_norm": 0.4826885272909795,
      "learning_rate": 6.493959207434934e-05,
      "loss": 0.5965,
      "step": 220
    },
    {
      "epoch": 1.796747967479675,
      "grad_norm": 0.5653326321329663,
      "learning_rate": 6.476152721746004e-05,
      "loss": 0.6067,
      "step": 221
    },
    {
      "epoch": 1.8048780487804879,
      "grad_norm": 0.6547145493149195,
      "learning_rate": 6.458266321528481e-05,
      "loss": 0.6007,
      "step": 222
    },
    {
      "epoch": 1.8130081300813008,
      "grad_norm": 0.7897341657692217,
      "learning_rate": 6.440300584042097e-05,
      "loss": 0.5965,
      "step": 223
    },
    {
      "epoch": 1.821138211382114,
      "grad_norm": 0.9204215163643266,
      "learning_rate": 6.422256089107084e-05,
      "loss": 0.5966,
      "step": 224
    },
    {
      "epoch": 1.8292682926829267,
      "grad_norm": 1.03087914505003,
      "learning_rate": 6.404133419085456e-05,
      "loss": 0.5936,
      "step": 225
    },
    {
      "epoch": 1.8373983739837398,
      "grad_norm": 1.0063011644823814,
      "learning_rate": 6.38593315886224e-05,
      "loss": 0.5944,
      "step": 226
    },
    {
      "epoch": 1.845528455284553,
      "grad_norm": 0.7289508275753798,
      "learning_rate": 6.367655895826573e-05,
      "loss": 0.6001,
      "step": 227
    },
    {
      "epoch": 1.8536585365853657,
      "grad_norm": 0.46673228908691156,
      "learning_rate": 6.349302219852755e-05,
      "loss": 0.5988,
      "step": 228
    },
    {
      "epoch": 1.8617886178861789,
      "grad_norm": 0.6932616598562896,
      "learning_rate": 6.330872723281216e-05,
      "loss": 0.6071,
      "step": 229
    },
    {
      "epoch": 1.8699186991869918,
      "grad_norm": 0.7132023054080954,
      "learning_rate": 6.312368000899394e-05,
      "loss": 0.5947,
      "step": 230
    },
    {
      "epoch": 1.8780487804878048,
      "grad_norm": 0.5002665338730061,
      "learning_rate": 6.293788649922537e-05,
      "loss": 0.6087,
      "step": 231
    },
    {
      "epoch": 1.886178861788618,
      "grad_norm": 0.5269839074780791,
      "learning_rate": 6.275135269974434e-05,
      "loss": 0.599,
      "step": 232
    },
    {
      "epoch": 1.8943089430894309,
      "grad_norm": 0.5757875212642624,
      "learning_rate": 6.256408463068056e-05,
      "loss": 0.5944,
      "step": 233
    },
    {
      "epoch": 1.9024390243902438,
      "grad_norm": 0.5450448896040407,
      "learning_rate": 6.237608833586139e-05,
      "loss": 0.5909,
      "step": 234
    },
    {
      "epoch": 1.910569105691057,
      "grad_norm": 0.4830754142972218,
      "learning_rate": 6.218736988261665e-05,
      "loss": 0.5944,
      "step": 235
    },
    {
      "epoch": 1.91869918699187,
      "grad_norm": 0.5131232843314816,
      "learning_rate": 6.199793536158288e-05,
      "loss": 0.602,
      "step": 236
    },
    {
      "epoch": 1.9268292682926829,
      "grad_norm": 0.4899990167501269,
      "learning_rate": 6.18077908865067e-05,
      "loss": 0.5978,
      "step": 237
    },
    {
      "epoch": 1.934959349593496,
      "grad_norm": 0.44036673915181546,
      "learning_rate": 6.161694259404767e-05,
      "loss": 0.5913,
      "step": 238
    },
    {
      "epoch": 1.943089430894309,
      "grad_norm": 0.39351397715880726,
      "learning_rate": 6.142539664357998e-05,
      "loss": 0.5932,
      "step": 239
    },
    {
      "epoch": 1.951219512195122,
      "grad_norm": 0.5193857150099743,
      "learning_rate": 6.123315921699389e-05,
      "loss": 0.5829,
      "step": 240
    },
    {
      "epoch": 1.959349593495935,
      "grad_norm": 0.707069854547125,
      "learning_rate": 6.104023651849613e-05,
      "loss": 0.592,
      "step": 241
    },
    {
      "epoch": 1.967479674796748,
      "grad_norm": 0.747904453430461,
      "learning_rate": 6.084663477440959e-05,
      "loss": 0.595,
      "step": 242
    },
    {
      "epoch": 1.975609756097561,
      "grad_norm": 0.9137073234795543,
      "learning_rate": 6.065236023297257e-05,
      "loss": 0.5942,
      "step": 243
    },
    {
      "epoch": 1.9837398373983741,
      "grad_norm": 0.9781689328844325,
      "learning_rate": 6.04574191641369e-05,
      "loss": 0.5913,
      "step": 244
    },
    {
      "epoch": 1.9918699186991868,
      "grad_norm": 0.8852310484823096,
      "learning_rate": 6.026181785936573e-05,
      "loss": 0.5922,
      "step": 245
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9016500552690595,
      "learning_rate": 6.006556263143049e-05,
      "loss": 0.5732,
      "step": 246
    },
    {
      "epoch": 2.008130081300813,
      "grad_norm": 0.9723703769285958,
      "learning_rate": 5.986865981420704e-05,
      "loss": 0.536,
      "step": 247
    },
    {
      "epoch": 2.016260162601626,
      "grad_norm": 1.1541920606472482,
      "learning_rate": 5.967111576247136e-05,
      "loss": 0.5402,
      "step": 248
    },
    {
      "epoch": 2.024390243902439,
      "grad_norm": 1.0421619086814218,
      "learning_rate": 5.94729368516944e-05,
      "loss": 0.5456,
      "step": 249
    },
    {
      "epoch": 2.032520325203252,
      "grad_norm": 1.1160479251888653,
      "learning_rate": 5.927412947783637e-05,
      "loss": 0.5354,
      "step": 250
    },
    {
      "epoch": 2.040650406504065,
      "grad_norm": 0.8373870264986623,
      "learning_rate": 5.907470005714025e-05,
      "loss": 0.533,
      "step": 251
    },
    {
      "epoch": 2.048780487804878,
      "grad_norm": 0.7071784566304017,
      "learning_rate": 5.887465502592477e-05,
      "loss": 0.5303,
      "step": 252
    },
    {
      "epoch": 2.0569105691056913,
      "grad_norm": 0.7354187469417341,
      "learning_rate": 5.867400084037668e-05,
      "loss": 0.5381,
      "step": 253
    },
    {
      "epoch": 2.065040650406504,
      "grad_norm": 0.8790103813070679,
      "learning_rate": 5.847274397634236e-05,
      "loss": 0.5376,
      "step": 254
    },
    {
      "epoch": 2.073170731707317,
      "grad_norm": 0.8805570141141628,
      "learning_rate": 5.827089092911884e-05,
      "loss": 0.5399,
      "step": 255
    },
    {
      "epoch": 2.08130081300813,
      "grad_norm": 0.6475906131727288,
      "learning_rate": 5.806844821324415e-05,
      "loss": 0.5335,
      "step": 256
    },
    {
      "epoch": 2.089430894308943,
      "grad_norm": 0.7245502270140599,
      "learning_rate": 5.786542236228712e-05,
      "loss": 0.5317,
      "step": 257
    },
    {
      "epoch": 2.097560975609756,
      "grad_norm": 0.9595256719293759,
      "learning_rate": 5.7661819928636484e-05,
      "loss": 0.5363,
      "step": 258
    },
    {
      "epoch": 2.105691056910569,
      "grad_norm": 0.6898661145738713,
      "learning_rate": 5.7457647483289385e-05,
      "loss": 0.536,
      "step": 259
    },
    {
      "epoch": 2.113821138211382,
      "grad_norm": 0.6975949936441868,
      "learning_rate": 5.725291161563935e-05,
      "loss": 0.5332,
      "step": 260
    },
    {
      "epoch": 2.1219512195121952,
      "grad_norm": 1.099825340268433,
      "learning_rate": 5.704761893326361e-05,
      "loss": 0.5359,
      "step": 261
    },
    {
      "epoch": 2.130081300813008,
      "grad_norm": 0.40647182721562153,
      "learning_rate": 5.684177606170987e-05,
      "loss": 0.5356,
      "step": 262
    },
    {
      "epoch": 2.138211382113821,
      "grad_norm": 0.8777090101435912,
      "learning_rate": 5.663538964428242e-05,
      "loss": 0.5283,
      "step": 263
    },
    {
      "epoch": 2.1463414634146343,
      "grad_norm": 0.8577158143321598,
      "learning_rate": 5.6428466341827793e-05,
      "loss": 0.5358,
      "step": 264
    },
    {
      "epoch": 2.154471544715447,
      "grad_norm": 0.48701405439531464,
      "learning_rate": 5.622101283251976e-05,
      "loss": 0.5291,
      "step": 265
    },
    {
      "epoch": 2.16260162601626,
      "grad_norm": 0.6626166399792842,
      "learning_rate": 5.6013035811643826e-05,
      "loss": 0.5309,
      "step": 266
    },
    {
      "epoch": 2.1707317073170733,
      "grad_norm": 0.5227912113875113,
      "learning_rate": 5.5804541991381113e-05,
      "loss": 0.5327,
      "step": 267
    },
    {
      "epoch": 2.178861788617886,
      "grad_norm": 0.4158136906713258,
      "learning_rate": 5.5595538100591766e-05,
      "loss": 0.534,
      "step": 268
    },
    {
      "epoch": 2.186991869918699,
      "grad_norm": 0.47664685732069456,
      "learning_rate": 5.538603088459779e-05,
      "loss": 0.5361,
      "step": 269
    },
    {
      "epoch": 2.1951219512195124,
      "grad_norm": 0.3734192550943469,
      "learning_rate": 5.5176027104965314e-05,
      "loss": 0.5391,
      "step": 270
    },
    {
      "epoch": 2.203252032520325,
      "grad_norm": 0.4778794183452955,
      "learning_rate": 5.49655335392864e-05,
      "loss": 0.5251,
      "step": 271
    },
    {
      "epoch": 2.2113821138211383,
      "grad_norm": 0.39564615482444043,
      "learning_rate": 5.4754556980960314e-05,
      "loss": 0.5268,
      "step": 272
    },
    {
      "epoch": 2.2195121951219514,
      "grad_norm": 0.39120283159648844,
      "learning_rate": 5.454310423897427e-05,
      "loss": 0.5366,
      "step": 273
    },
    {
      "epoch": 2.227642276422764,
      "grad_norm": 0.41322606498743786,
      "learning_rate": 5.433118213768367e-05,
      "loss": 0.5298,
      "step": 274
    },
    {
      "epoch": 2.2357723577235773,
      "grad_norm": 0.27724552260370505,
      "learning_rate": 5.411879751659184e-05,
      "loss": 0.536,
      "step": 275
    },
    {
      "epoch": 2.2439024390243905,
      "grad_norm": 0.29429845941495025,
      "learning_rate": 5.390595723012936e-05,
      "loss": 0.53,
      "step": 276
    },
    {
      "epoch": 2.252032520325203,
      "grad_norm": 0.27865793012748635,
      "learning_rate": 5.3692668147432746e-05,
      "loss": 0.5307,
      "step": 277
    },
    {
      "epoch": 2.2601626016260163,
      "grad_norm": 0.2540380014859299,
      "learning_rate": 5.347893715212287e-05,
      "loss": 0.5304,
      "step": 278
    },
    {
      "epoch": 2.2682926829268295,
      "grad_norm": 0.282931990712346,
      "learning_rate": 5.326477114208272e-05,
      "loss": 0.5316,
      "step": 279
    },
    {
      "epoch": 2.2764227642276422,
      "grad_norm": 0.26353913769078685,
      "learning_rate": 5.3050177029234786e-05,
      "loss": 0.5284,
      "step": 280
    },
    {
      "epoch": 2.2845528455284554,
      "grad_norm": 0.3629912739071986,
      "learning_rate": 5.2835161739318036e-05,
      "loss": 0.5332,
      "step": 281
    },
    {
      "epoch": 2.292682926829268,
      "grad_norm": 0.35734340348374305,
      "learning_rate": 5.261973221166436e-05,
      "loss": 0.5351,
      "step": 282
    },
    {
      "epoch": 2.3008130081300813,
      "grad_norm": 0.27171156829481286,
      "learning_rate": 5.240389539897461e-05,
      "loss": 0.5327,
      "step": 283
    },
    {
      "epoch": 2.3089430894308944,
      "grad_norm": 0.298189595247889,
      "learning_rate": 5.21876582670942e-05,
      "loss": 0.5365,
      "step": 284
    },
    {
      "epoch": 2.317073170731707,
      "grad_norm": 0.24567752772372758,
      "learning_rate": 5.197102779478837e-05,
      "loss": 0.5287,
      "step": 285
    },
    {
      "epoch": 2.3252032520325203,
      "grad_norm": 0.2892030523640998,
      "learning_rate": 5.175401097351685e-05,
      "loss": 0.536,
      "step": 286
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.28659354790171204,
      "learning_rate": 5.1536614807208266e-05,
      "loss": 0.5283,
      "step": 287
    },
    {
      "epoch": 2.341463414634146,
      "grad_norm": 0.33667599271093507,
      "learning_rate": 5.131884631203413e-05,
      "loss": 0.5312,
      "step": 288
    },
    {
      "epoch": 2.3495934959349594,
      "grad_norm": 0.3158259688118238,
      "learning_rate": 5.110071251618236e-05,
      "loss": 0.5286,
      "step": 289
    },
    {
      "epoch": 2.3577235772357725,
      "grad_norm": 0.23778752717896398,
      "learning_rate": 5.088222045963047e-05,
      "loss": 0.5317,
      "step": 290
    },
    {
      "epoch": 2.3658536585365852,
      "grad_norm": 0.23420322468409785,
      "learning_rate": 5.0663377193918324e-05,
      "loss": 0.5283,
      "step": 291
    },
    {
      "epoch": 2.3739837398373984,
      "grad_norm": 0.28535070489549674,
      "learning_rate": 5.044418978192064e-05,
      "loss": 0.5366,
      "step": 292
    },
    {
      "epoch": 2.3821138211382116,
      "grad_norm": 0.31107022928444744,
      "learning_rate": 5.022466529761899e-05,
      "loss": 0.5337,
      "step": 293
    },
    {
      "epoch": 2.3902439024390243,
      "grad_norm": 0.3004204484587811,
      "learning_rate": 5.0004810825873495e-05,
      "loss": 0.5261,
      "step": 294
    },
    {
      "epoch": 2.3983739837398375,
      "grad_norm": 0.26155629778163336,
      "learning_rate": 4.978463346219417e-05,
      "loss": 0.5278,
      "step": 295
    },
    {
      "epoch": 2.40650406504065,
      "grad_norm": 0.2389875445240219,
      "learning_rate": 4.956414031251194e-05,
      "loss": 0.5272,
      "step": 296
    },
    {
      "epoch": 2.4146341463414633,
      "grad_norm": 0.23785046936227025,
      "learning_rate": 4.934333849294934e-05,
      "loss": 0.5319,
      "step": 297
    },
    {
      "epoch": 2.4227642276422765,
      "grad_norm": 0.22741010006900195,
      "learning_rate": 4.9122235129590774e-05,
      "loss": 0.5355,
      "step": 298
    },
    {
      "epoch": 2.430894308943089,
      "grad_norm": 0.23525720231043418,
      "learning_rate": 4.890083735825258e-05,
      "loss": 0.5245,
      "step": 299
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 0.2216881604997441,
      "learning_rate": 4.8679152324252725e-05,
      "loss": 0.5297,
      "step": 300
    },
    {
      "epoch": 2.4471544715447155,
      "grad_norm": 0.24234098875815116,
      "learning_rate": 4.84571871821802e-05,
      "loss": 0.5281,
      "step": 301
    },
    {
      "epoch": 2.4552845528455283,
      "grad_norm": 0.29179921859610297,
      "learning_rate": 4.823494909566409e-05,
      "loss": 0.532,
      "step": 302
    },
    {
      "epoch": 2.4634146341463414,
      "grad_norm": 0.23911830549425322,
      "learning_rate": 4.801244523714243e-05,
      "loss": 0.5277,
      "step": 303
    },
    {
      "epoch": 2.4715447154471546,
      "grad_norm": 0.2679960722238244,
      "learning_rate": 4.778968278763065e-05,
      "loss": 0.5317,
      "step": 304
    },
    {
      "epoch": 2.4796747967479673,
      "grad_norm": 0.33164413438561197,
      "learning_rate": 4.7566668936489905e-05,
      "loss": 0.5273,
      "step": 305
    },
    {
      "epoch": 2.4878048780487805,
      "grad_norm": 0.388473350189468,
      "learning_rate": 4.7343410881194984e-05,
      "loss": 0.5371,
      "step": 306
    },
    {
      "epoch": 2.4959349593495936,
      "grad_norm": 0.3300349865680693,
      "learning_rate": 4.711991582710203e-05,
      "loss": 0.5303,
      "step": 307
    },
    {
      "epoch": 2.5040650406504064,
      "grad_norm": 0.20547630575911513,
      "learning_rate": 4.689619098721599e-05,
      "loss": 0.5276,
      "step": 308
    },
    {
      "epoch": 2.5121951219512195,
      "grad_norm": 0.2634684897680853,
      "learning_rate": 4.667224358195791e-05,
      "loss": 0.5272,
      "step": 309
    },
    {
      "epoch": 2.5203252032520327,
      "grad_norm": 0.38486711188151485,
      "learning_rate": 4.6448080838931747e-05,
      "loss": 0.5339,
      "step": 310
    },
    {
      "epoch": 2.5284552845528454,
      "grad_norm": 0.28226181384239807,
      "learning_rate": 4.622370999269128e-05,
      "loss": 0.5268,
      "step": 311
    },
    {
      "epoch": 2.5365853658536586,
      "grad_norm": 0.2031111975110838,
      "learning_rate": 4.599913828450648e-05,
      "loss": 0.5378,
      "step": 312
    },
    {
      "epoch": 2.5447154471544717,
      "grad_norm": 0.17978982748236894,
      "learning_rate": 4.57743729621299e-05,
      "loss": 0.5289,
      "step": 313
    },
    {
      "epoch": 2.5528455284552845,
      "grad_norm": 0.24813035414167875,
      "learning_rate": 4.554942127956274e-05,
      "loss": 0.5243,
      "step": 314
    },
    {
      "epoch": 2.5609756097560976,
      "grad_norm": 0.29968563816817967,
      "learning_rate": 4.5324290496820685e-05,
      "loss": 0.5294,
      "step": 315
    },
    {
      "epoch": 2.569105691056911,
      "grad_norm": 0.23702583096780408,
      "learning_rate": 4.509898787969968e-05,
      "loss": 0.5297,
      "step": 316
    },
    {
      "epoch": 2.5772357723577235,
      "grad_norm": 0.21620550138477387,
      "learning_rate": 4.48735206995414e-05,
      "loss": 0.5291,
      "step": 317
    },
    {
      "epoch": 2.5853658536585367,
      "grad_norm": 0.3221679789621951,
      "learning_rate": 4.4647896232998536e-05,
      "loss": 0.5362,
      "step": 318
    },
    {
      "epoch": 2.59349593495935,
      "grad_norm": 0.33970012722501114,
      "learning_rate": 4.442212176179999e-05,
      "loss": 0.5364,
      "step": 319
    },
    {
      "epoch": 2.6016260162601625,
      "grad_norm": 0.2956828338734781,
      "learning_rate": 4.4196204572515925e-05,
      "loss": 0.5409,
      "step": 320
    },
    {
      "epoch": 2.6097560975609757,
      "grad_norm": 0.20369855697868933,
      "learning_rate": 4.397015195632247e-05,
      "loss": 0.5302,
      "step": 321
    },
    {
      "epoch": 2.617886178861789,
      "grad_norm": 0.2175999461612837,
      "learning_rate": 4.3743971208766504e-05,
      "loss": 0.5333,
      "step": 322
    },
    {
      "epoch": 2.6260162601626016,
      "grad_norm": 0.2853175766731205,
      "learning_rate": 4.3517669629530174e-05,
      "loss": 0.5303,
      "step": 323
    },
    {
      "epoch": 2.6341463414634148,
      "grad_norm": 0.24780824372167998,
      "learning_rate": 4.329125452219532e-05,
      "loss": 0.5277,
      "step": 324
    },
    {
      "epoch": 2.642276422764228,
      "grad_norm": 0.222106856265205,
      "learning_rate": 4.306473319400773e-05,
      "loss": 0.5322,
      "step": 325
    },
    {
      "epoch": 2.6504065040650406,
      "grad_norm": 0.2380311860697225,
      "learning_rate": 4.283811295564133e-05,
      "loss": 0.5304,
      "step": 326
    },
    {
      "epoch": 2.658536585365854,
      "grad_norm": 0.26177899498639123,
      "learning_rate": 4.261140112096226e-05,
      "loss": 0.5315,
      "step": 327
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.2341850600490066,
      "learning_rate": 4.23846050067928e-05,
      "loss": 0.5268,
      "step": 328
    },
    {
      "epoch": 2.6747967479674797,
      "grad_norm": 0.22160594107308393,
      "learning_rate": 4.215773193267522e-05,
      "loss": 0.526,
      "step": 329
    },
    {
      "epoch": 2.682926829268293,
      "grad_norm": 0.2566244722839592,
      "learning_rate": 4.19307892206356e-05,
      "loss": 0.5235,
      "step": 330
    },
    {
      "epoch": 2.6910569105691056,
      "grad_norm": 0.25517888682360007,
      "learning_rate": 4.170378419494745e-05,
      "loss": 0.5334,
      "step": 331
    },
    {
      "epoch": 2.6991869918699187,
      "grad_norm": 0.1796798853410341,
      "learning_rate": 4.1476724181895414e-05,
      "loss": 0.5275,
      "step": 332
    },
    {
      "epoch": 2.7073170731707314,
      "grad_norm": 0.25043942237954464,
      "learning_rate": 4.124961650953875e-05,
      "loss": 0.5367,
      "step": 333
    },
    {
      "epoch": 2.7154471544715446,
      "grad_norm": 0.23694092605685546,
      "learning_rate": 4.102246850747489e-05,
      "loss": 0.5328,
      "step": 334
    },
    {
      "epoch": 2.7235772357723578,
      "grad_norm": 0.18253661953677602,
      "learning_rate": 4.0795287506602806e-05,
      "loss": 0.5277,
      "step": 335
    },
    {
      "epoch": 2.7317073170731705,
      "grad_norm": 0.27410762176779596,
      "learning_rate": 4.056808083888651e-05,
      "loss": 0.5309,
      "step": 336
    },
    {
      "epoch": 2.7398373983739837,
      "grad_norm": 0.31611763782023194,
      "learning_rate": 4.034085583711834e-05,
      "loss": 0.5262,
      "step": 337
    },
    {
      "epoch": 2.747967479674797,
      "grad_norm": 0.23183757249601128,
      "learning_rate": 4.011361983468236e-05,
      "loss": 0.5263,
      "step": 338
    },
    {
      "epoch": 2.7560975609756095,
      "grad_norm": 0.17193227922357013,
      "learning_rate": 3.988638016531765e-05,
      "loss": 0.526,
      "step": 339
    },
    {
      "epoch": 2.7642276422764227,
      "grad_norm": 0.27288245357739827,
      "learning_rate": 3.965914416288168e-05,
      "loss": 0.5308,
      "step": 340
    },
    {
      "epoch": 2.772357723577236,
      "grad_norm": 0.2811884633390003,
      "learning_rate": 3.94319191611135e-05,
      "loss": 0.532,
      "step": 341
    },
    {
      "epoch": 2.7804878048780486,
      "grad_norm": 0.17752906471060148,
      "learning_rate": 3.9204712493397194e-05,
      "loss": 0.5269,
      "step": 342
    },
    {
      "epoch": 2.7886178861788617,
      "grad_norm": 0.2845250664435178,
      "learning_rate": 3.8977531492525124e-05,
      "loss": 0.5277,
      "step": 343
    },
    {
      "epoch": 2.796747967479675,
      "grad_norm": 0.26958285232807416,
      "learning_rate": 3.875038349046126e-05,
      "loss": 0.5312,
      "step": 344
    },
    {
      "epoch": 2.8048780487804876,
      "grad_norm": 0.1694246640194251,
      "learning_rate": 3.85232758181046e-05,
      "loss": 0.5313,
      "step": 345
    },
    {
      "epoch": 2.813008130081301,
      "grad_norm": 0.25631133199827844,
      "learning_rate": 3.829621580505257e-05,
      "loss": 0.539,
      "step": 346
    },
    {
      "epoch": 2.821138211382114,
      "grad_norm": 0.22885516076169457,
      "learning_rate": 3.806921077936442e-05,
      "loss": 0.5282,
      "step": 347
    },
    {
      "epoch": 2.8292682926829267,
      "grad_norm": 0.19392775868370438,
      "learning_rate": 3.7842268067324795e-05,
      "loss": 0.5304,
      "step": 348
    },
    {
      "epoch": 2.83739837398374,
      "grad_norm": 0.22176824539495676,
      "learning_rate": 3.761539499320721e-05,
      "loss": 0.5306,
      "step": 349
    },
    {
      "epoch": 2.845528455284553,
      "grad_norm": 0.2328255603372444,
      "learning_rate": 3.738859887903774e-05,
      "loss": 0.5303,
      "step": 350
    },
    {
      "epoch": 2.8536585365853657,
      "grad_norm": 0.17418593862162282,
      "learning_rate": 3.716188704435868e-05,
      "loss": 0.532,
      "step": 351
    },
    {
      "epoch": 2.861788617886179,
      "grad_norm": 0.17720731252605912,
      "learning_rate": 3.693526680599228e-05,
      "loss": 0.5306,
      "step": 352
    },
    {
      "epoch": 2.869918699186992,
      "grad_norm": 0.21283471852007732,
      "learning_rate": 3.6708745477804685e-05,
      "loss": 0.5288,
      "step": 353
    },
    {
      "epoch": 2.8780487804878048,
      "grad_norm": 0.22341240295628834,
      "learning_rate": 3.648233037046984e-05,
      "loss": 0.5288,
      "step": 354
    },
    {
      "epoch": 2.886178861788618,
      "grad_norm": 0.1899880100703667,
      "learning_rate": 3.62560287912335e-05,
      "loss": 0.5297,
      "step": 355
    },
    {
      "epoch": 2.894308943089431,
      "grad_norm": 0.18494888521312322,
      "learning_rate": 3.602984804367755e-05,
      "loss": 0.5308,
      "step": 356
    },
    {
      "epoch": 2.902439024390244,
      "grad_norm": 0.19479364419658737,
      "learning_rate": 3.580379542748408e-05,
      "loss": 0.532,
      "step": 357
    },
    {
      "epoch": 2.910569105691057,
      "grad_norm": 0.1834923395085625,
      "learning_rate": 3.557787823820001e-05,
      "loss": 0.5218,
      "step": 358
    },
    {
      "epoch": 2.91869918699187,
      "grad_norm": 0.22192106700410788,
      "learning_rate": 3.535210376700149e-05,
      "loss": 0.5309,
      "step": 359
    },
    {
      "epoch": 2.926829268292683,
      "grad_norm": 0.1915909338191775,
      "learning_rate": 3.512647930045861e-05,
      "loss": 0.5237,
      "step": 360
    },
    {
      "epoch": 2.934959349593496,
      "grad_norm": 0.18636867296795742,
      "learning_rate": 3.4901012120300325e-05,
      "loss": 0.5315,
      "step": 361
    },
    {
      "epoch": 2.943089430894309,
      "grad_norm": 0.21589078585324512,
      "learning_rate": 3.4675709503179335e-05,
      "loss": 0.5309,
      "step": 362
    },
    {
      "epoch": 2.951219512195122,
      "grad_norm": 0.1709811819449636,
      "learning_rate": 3.445057872043728e-05,
      "loss": 0.5226,
      "step": 363
    },
    {
      "epoch": 2.959349593495935,
      "grad_norm": 0.2082868229632365,
      "learning_rate": 3.42256270378701e-05,
      "loss": 0.5282,
      "step": 364
    },
    {
      "epoch": 2.9674796747967482,
      "grad_norm": 0.19362176685936341,
      "learning_rate": 3.4000861715493537e-05,
      "loss": 0.5271,
      "step": 365
    },
    {
      "epoch": 2.975609756097561,
      "grad_norm": 0.1981983051979009,
      "learning_rate": 3.377629000730873e-05,
      "loss": 0.5294,
      "step": 366
    },
    {
      "epoch": 2.983739837398374,
      "grad_norm": 0.15869095471418823,
      "learning_rate": 3.355191916106826e-05,
      "loss": 0.5223,
      "step": 367
    },
    {
      "epoch": 2.991869918699187,
      "grad_norm": 0.20900409434261094,
      "learning_rate": 3.332775641804211e-05,
      "loss": 0.5306,
      "step": 368
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.21096310689948075,
      "learning_rate": 3.3103809012784015e-05,
      "loss": 0.5071,
      "step": 369
    },
    {
      "epoch": 3.008130081300813,
      "grad_norm": 0.265901908762846,
      "learning_rate": 3.2880084172898e-05,
      "loss": 0.4704,
      "step": 370
    },
    {
      "epoch": 3.016260162601626,
      "grad_norm": 0.3006900752910841,
      "learning_rate": 3.265658911880503e-05,
      "loss": 0.473,
      "step": 371
    },
    {
      "epoch": 3.024390243902439,
      "grad_norm": 0.30633936102055226,
      "learning_rate": 3.24333310635101e-05,
      "loss": 0.4793,
      "step": 372
    },
    {
      "epoch": 3.032520325203252,
      "grad_norm": 0.3006461258794499,
      "learning_rate": 3.221031721236936e-05,
      "loss": 0.4688,
      "step": 373
    },
    {
      "epoch": 3.040650406504065,
      "grad_norm": 0.3079683975369383,
      "learning_rate": 3.1987554762857584e-05,
      "loss": 0.4733,
      "step": 374
    },
    {
      "epoch": 3.048780487804878,
      "grad_norm": 0.22088913517836137,
      "learning_rate": 3.176505090433593e-05,
      "loss": 0.4691,
      "step": 375
    },
    {
      "epoch": 3.0569105691056913,
      "grad_norm": 0.3232828929471117,
      "learning_rate": 3.1542812817819816e-05,
      "loss": 0.4699,
      "step": 376
    },
    {
      "epoch": 3.065040650406504,
      "grad_norm": 0.28902113111471356,
      "learning_rate": 3.132084767574728e-05,
      "loss": 0.4711,
      "step": 377
    },
    {
      "epoch": 3.073170731707317,
      "grad_norm": 0.22724829822476866,
      "learning_rate": 3.109916264174743e-05,
      "loss": 0.468,
      "step": 378
    },
    {
      "epoch": 3.08130081300813,
      "grad_norm": 0.3114223293758183,
      "learning_rate": 3.087776487040924e-05,
      "loss": 0.4671,
      "step": 379
    },
    {
      "epoch": 3.089430894308943,
      "grad_norm": 0.23075906909201893,
      "learning_rate": 3.065666150705067e-05,
      "loss": 0.4729,
      "step": 380
    },
    {
      "epoch": 3.097560975609756,
      "grad_norm": 0.20677747951901723,
      "learning_rate": 3.043585968748807e-05,
      "loss": 0.4667,
      "step": 381
    },
    {
      "epoch": 3.105691056910569,
      "grad_norm": 0.2559830030921859,
      "learning_rate": 3.021536653780585e-05,
      "loss": 0.4659,
      "step": 382
    },
    {
      "epoch": 3.113821138211382,
      "grad_norm": 0.23278567962741312,
      "learning_rate": 2.999518917412653e-05,
      "loss": 0.4746,
      "step": 383
    },
    {
      "epoch": 3.1219512195121952,
      "grad_norm": 0.2090312196411212,
      "learning_rate": 2.9775334702381017e-05,
      "loss": 0.4753,
      "step": 384
    },
    {
      "epoch": 3.130081300813008,
      "grad_norm": 0.2526705855972774,
      "learning_rate": 2.9555810218079356e-05,
      "loss": 0.4701,
      "step": 385
    },
    {
      "epoch": 3.138211382113821,
      "grad_norm": 0.1902189983709351,
      "learning_rate": 2.933662280608169e-05,
      "loss": 0.4705,
      "step": 386
    },
    {
      "epoch": 3.1463414634146343,
      "grad_norm": 0.2270380458930752,
      "learning_rate": 2.9117779540369544e-05,
      "loss": 0.4658,
      "step": 387
    },
    {
      "epoch": 3.154471544715447,
      "grad_norm": 0.20950532575639283,
      "learning_rate": 2.8899287483817638e-05,
      "loss": 0.4657,
      "step": 388
    },
    {
      "epoch": 3.16260162601626,
      "grad_norm": 0.21759937053339654,
      "learning_rate": 2.8681153687965883e-05,
      "loss": 0.4707,
      "step": 389
    },
    {
      "epoch": 3.1707317073170733,
      "grad_norm": 0.22081730272344247,
      "learning_rate": 2.8463385192791748e-05,
      "loss": 0.4781,
      "step": 390
    },
    {
      "epoch": 3.178861788617886,
      "grad_norm": 0.16461623065529493,
      "learning_rate": 2.8245989026483173e-05,
      "loss": 0.4717,
      "step": 391
    },
    {
      "epoch": 3.186991869918699,
      "grad_norm": 0.17363189599367604,
      "learning_rate": 2.802897220521164e-05,
      "loss": 0.466,
      "step": 392
    },
    {
      "epoch": 3.1951219512195124,
      "grad_norm": 0.1696365139425646,
      "learning_rate": 2.78123417329058e-05,
      "loss": 0.4709,
      "step": 393
    },
    {
      "epoch": 3.203252032520325,
      "grad_norm": 0.1546039383110101,
      "learning_rate": 2.7596104601025414e-05,
      "loss": 0.4702,
      "step": 394
    },
    {
      "epoch": 3.2113821138211383,
      "grad_norm": 0.1473333328250618,
      "learning_rate": 2.7380267788335654e-05,
      "loss": 0.4706,
      "step": 395
    },
    {
      "epoch": 3.2195121951219514,
      "grad_norm": 0.15687150923782062,
      "learning_rate": 2.7164838260681964e-05,
      "loss": 0.4664,
      "step": 396
    },
    {
      "epoch": 3.227642276422764,
      "grad_norm": 0.15371561503224995,
      "learning_rate": 2.694982297076523e-05,
      "loss": 0.4649,
      "step": 397
    },
    {
      "epoch": 3.2357723577235773,
      "grad_norm": 0.16004131967753452,
      "learning_rate": 2.673522885791729e-05,
      "loss": 0.474,
      "step": 398
    },
    {
      "epoch": 3.2439024390243905,
      "grad_norm": 0.16458541623514825,
      "learning_rate": 2.6521062847877143e-05,
      "loss": 0.4706,
      "step": 399
    },
    {
      "epoch": 3.252032520325203,
      "grad_norm": 0.1846493973614995,
      "learning_rate": 2.6307331852567263e-05,
      "loss": 0.4684,
      "step": 400
    },
    {
      "epoch": 3.2601626016260163,
      "grad_norm": 0.15147442641270048,
      "learning_rate": 2.609404276987065e-05,
      "loss": 0.4667,
      "step": 401
    },
    {
      "epoch": 3.2682926829268295,
      "grad_norm": 0.179576690049958,
      "learning_rate": 2.5881202483408165e-05,
      "loss": 0.4727,
      "step": 402
    },
    {
      "epoch": 3.2764227642276422,
      "grad_norm": 0.14471848092132056,
      "learning_rate": 2.5668817862316332e-05,
      "loss": 0.4655,
      "step": 403
    },
    {
      "epoch": 3.2845528455284554,
      "grad_norm": 0.16304136988755918,
      "learning_rate": 2.5456895761025723e-05,
      "loss": 0.4757,
      "step": 404
    },
    {
      "epoch": 3.292682926829268,
      "grad_norm": 0.14395926252809935,
      "learning_rate": 2.5245443019039693e-05,
      "loss": 0.4727,
      "step": 405
    },
    {
      "epoch": 3.3008130081300813,
      "grad_norm": 0.14931563464208047,
      "learning_rate": 2.503446646071361e-05,
      "loss": 0.4743,
      "step": 406
    },
    {
      "epoch": 3.3089430894308944,
      "grad_norm": 0.15028010975234346,
      "learning_rate": 2.482397289503471e-05,
      "loss": 0.4648,
      "step": 407
    },
    {
      "epoch": 3.317073170731707,
      "grad_norm": 0.16320220341314107,
      "learning_rate": 2.461396911540222e-05,
      "loss": 0.4677,
      "step": 408
    },
    {
      "epoch": 3.3252032520325203,
      "grad_norm": 0.1455790902797688,
      "learning_rate": 2.4404461899408233e-05,
      "loss": 0.4717,
      "step": 409
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.1490317501026984,
      "learning_rate": 2.4195458008618903e-05,
      "loss": 0.4709,
      "step": 410
    },
    {
      "epoch": 3.341463414634146,
      "grad_norm": 0.15826842157186718,
      "learning_rate": 2.398696418835619e-05,
      "loss": 0.4679,
      "step": 411
    },
    {
      "epoch": 3.3495934959349594,
      "grad_norm": 0.16786430961492443,
      "learning_rate": 2.377898716748025e-05,
      "loss": 0.4663,
      "step": 412
    },
    {
      "epoch": 3.3577235772357725,
      "grad_norm": 0.16742815781293757,
      "learning_rate": 2.3571533658172223e-05,
      "loss": 0.4641,
      "step": 413
    },
    {
      "epoch": 3.3658536585365852,
      "grad_norm": 0.1761744139689136,
      "learning_rate": 2.33646103557176e-05,
      "loss": 0.4725,
      "step": 414
    },
    {
      "epoch": 3.3739837398373984,
      "grad_norm": 0.13795109133132294,
      "learning_rate": 2.3158223938290147e-05,
      "loss": 0.4688,
      "step": 415
    },
    {
      "epoch": 3.3821138211382116,
      "grad_norm": 0.16055313592474532,
      "learning_rate": 2.295238106673641e-05,
      "loss": 0.4716,
      "step": 416
    },
    {
      "epoch": 3.3902439024390243,
      "grad_norm": 0.15611992956704845,
      "learning_rate": 2.2747088384360673e-05,
      "loss": 0.4688,
      "step": 417
    },
    {
      "epoch": 3.3983739837398375,
      "grad_norm": 0.14069846656439636,
      "learning_rate": 2.254235251671063e-05,
      "loss": 0.4747,
      "step": 418
    },
    {
      "epoch": 3.40650406504065,
      "grad_norm": 0.17087093694215974,
      "learning_rate": 2.2338180071363543e-05,
      "loss": 0.4758,
      "step": 419
    },
    {
      "epoch": 3.4146341463414633,
      "grad_norm": 0.12781453366172704,
      "learning_rate": 2.213457763771289e-05,
      "loss": 0.4655,
      "step": 420
    },
    {
      "epoch": 3.4227642276422765,
      "grad_norm": 0.1540153350457091,
      "learning_rate": 2.1931551786755855e-05,
      "loss": 0.4685,
      "step": 421
    },
    {
      "epoch": 3.430894308943089,
      "grad_norm": 0.13722376812630757,
      "learning_rate": 2.1729109070881167e-05,
      "loss": 0.4738,
      "step": 422
    },
    {
      "epoch": 3.4390243902439024,
      "grad_norm": 0.1292960618276643,
      "learning_rate": 2.152725602365764e-05,
      "loss": 0.4692,
      "step": 423
    },
    {
      "epoch": 3.4471544715447155,
      "grad_norm": 0.13322872769981725,
      "learning_rate": 2.1325999159623328e-05,
      "loss": 0.4684,
      "step": 424
    },
    {
      "epoch": 3.4552845528455283,
      "grad_norm": 0.1372014593818,
      "learning_rate": 2.112534497407523e-05,
      "loss": 0.4685,
      "step": 425
    },
    {
      "epoch": 3.4634146341463414,
      "grad_norm": 0.13846694949870228,
      "learning_rate": 2.0925299942859768e-05,
      "loss": 0.4761,
      "step": 426
    },
    {
      "epoch": 3.4715447154471546,
      "grad_norm": 0.12719349794368842,
      "learning_rate": 2.0725870522163644e-05,
      "loss": 0.4683,
      "step": 427
    },
    {
      "epoch": 3.4796747967479673,
      "grad_norm": 0.12657263895255097,
      "learning_rate": 2.05270631483056e-05,
      "loss": 0.4651,
      "step": 428
    },
    {
      "epoch": 3.4878048780487805,
      "grad_norm": 0.13810077282440186,
      "learning_rate": 2.0328884237528656e-05,
      "loss": 0.4764,
      "step": 429
    },
    {
      "epoch": 3.4959349593495936,
      "grad_norm": 0.1320750927081112,
      "learning_rate": 2.0131340185792975e-05,
      "loss": 0.4696,
      "step": 430
    },
    {
      "epoch": 3.5040650406504064,
      "grad_norm": 0.1397450418596229,
      "learning_rate": 1.9934437368569515e-05,
      "loss": 0.4727,
      "step": 431
    },
    {
      "epoch": 3.5121951219512195,
      "grad_norm": 0.14936926379247653,
      "learning_rate": 1.9738182140634277e-05,
      "loss": 0.4757,
      "step": 432
    },
    {
      "epoch": 3.5203252032520327,
      "grad_norm": 0.12498940156789684,
      "learning_rate": 1.9542580835863117e-05,
      "loss": 0.4656,
      "step": 433
    },
    {
      "epoch": 3.5284552845528454,
      "grad_norm": 0.14983212889464834,
      "learning_rate": 1.9347639767027454e-05,
      "loss": 0.4662,
      "step": 434
    },
    {
      "epoch": 3.5365853658536586,
      "grad_norm": 0.1376883853956337,
      "learning_rate": 1.9153365225590413e-05,
      "loss": 0.4713,
      "step": 435
    },
    {
      "epoch": 3.5447154471544717,
      "grad_norm": 0.13419456309222105,
      "learning_rate": 1.8959763481503888e-05,
      "loss": 0.468,
      "step": 436
    },
    {
      "epoch": 3.5528455284552845,
      "grad_norm": 0.14807955417003565,
      "learning_rate": 1.8766840783006124e-05,
      "loss": 0.4691,
      "step": 437
    },
    {
      "epoch": 3.5609756097560976,
      "grad_norm": 0.1329631024981258,
      "learning_rate": 1.8574603356420034e-05,
      "loss": 0.4727,
      "step": 438
    },
    {
      "epoch": 3.569105691056911,
      "grad_norm": 0.1431243701218343,
      "learning_rate": 1.8383057405952347e-05,
      "loss": 0.4713,
      "step": 439
    },
    {
      "epoch": 3.5772357723577235,
      "grad_norm": 0.13389352804169835,
      "learning_rate": 1.8192209113493307e-05,
      "loss": 0.4665,
      "step": 440
    },
    {
      "epoch": 3.5853658536585367,
      "grad_norm": 0.1482020586528074,
      "learning_rate": 1.800206463841714e-05,
      "loss": 0.4705,
      "step": 441
    },
    {
      "epoch": 3.59349593495935,
      "grad_norm": 0.13038607420889628,
      "learning_rate": 1.7812630117383355e-05,
      "loss": 0.4664,
      "step": 442
    },
    {
      "epoch": 3.6016260162601625,
      "grad_norm": 0.1665186484592332,
      "learning_rate": 1.762391166413861e-05,
      "loss": 0.4689,
      "step": 443
    },
    {
      "epoch": 3.6097560975609757,
      "grad_norm": 0.13601314972353779,
      "learning_rate": 1.7435915369319434e-05,
      "loss": 0.4715,
      "step": 444
    },
    {
      "epoch": 3.617886178861789,
      "grad_norm": 0.1484890153404192,
      "learning_rate": 1.724864730025568e-05,
      "loss": 0.4716,
      "step": 445
    },
    {
      "epoch": 3.6260162601626016,
      "grad_norm": 0.13586982378564544,
      "learning_rate": 1.7062113500774636e-05,
      "loss": 0.4696,
      "step": 446
    },
    {
      "epoch": 3.6341463414634148,
      "grad_norm": 0.13301578455247703,
      "learning_rate": 1.687631999100606e-05,
      "loss": 0.4696,
      "step": 447
    },
    {
      "epoch": 3.642276422764228,
      "grad_norm": 0.12603570021540972,
      "learning_rate": 1.669127276718784e-05,
      "loss": 0.4733,
      "step": 448
    },
    {
      "epoch": 3.6504065040650406,
      "grad_norm": 0.1263329246098734,
      "learning_rate": 1.650697780147245e-05,
      "loss": 0.4661,
      "step": 449
    },
    {
      "epoch": 3.658536585365854,
      "grad_norm": 0.14191486316301302,
      "learning_rate": 1.6323441041734288e-05,
      "loss": 0.4798,
      "step": 450
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.12444771302681085,
      "learning_rate": 1.6140668411377602e-05,
      "loss": 0.4768,
      "step": 451
    },
    {
      "epoch": 3.6747967479674797,
      "grad_norm": 0.13038608427131376,
      "learning_rate": 1.5958665809145432e-05,
      "loss": 0.4757,
      "step": 452
    },
    {
      "epoch": 3.682926829268293,
      "grad_norm": 0.1387892663233294,
      "learning_rate": 1.5777439108929188e-05,
      "loss": 0.4701,
      "step": 453
    },
    {
      "epoch": 3.6910569105691056,
      "grad_norm": 0.1307681340035132,
      "learning_rate": 1.5596994159579025e-05,
      "loss": 0.4745,
      "step": 454
    },
    {
      "epoch": 3.6991869918699187,
      "grad_norm": 0.15066531142752398,
      "learning_rate": 1.541733678471519e-05,
      "loss": 0.4744,
      "step": 455
    },
    {
      "epoch": 3.7073170731707314,
      "grad_norm": 0.13862066367370007,
      "learning_rate": 1.523847278253999e-05,
      "loss": 0.4749,
      "step": 456
    },
    {
      "epoch": 3.7154471544715446,
      "grad_norm": 0.16153595851631308,
      "learning_rate": 1.5060407925650662e-05,
      "loss": 0.4766,
      "step": 457
    },
    {
      "epoch": 3.7235772357723578,
      "grad_norm": 0.15403271466806268,
      "learning_rate": 1.4883147960853145e-05,
      "loss": 0.4668,
      "step": 458
    },
    {
      "epoch": 3.7317073170731705,
      "grad_norm": 0.1433640158402133,
      "learning_rate": 1.4706698608976501e-05,
      "loss": 0.4742,
      "step": 459
    },
    {
      "epoch": 3.7398373983739837,
      "grad_norm": 0.1657739517712798,
      "learning_rate": 1.4531065564688387e-05,
      "loss": 0.4712,
      "step": 460
    },
    {
      "epoch": 3.747967479674797,
      "grad_norm": 0.1313220912730872,
      "learning_rate": 1.4356254496311208e-05,
      "loss": 0.4649,
      "step": 461
    },
    {
      "epoch": 3.7560975609756095,
      "grad_norm": 0.14040664039810835,
      "learning_rate": 1.4182271045639167e-05,
      "loss": 0.4744,
      "step": 462
    },
    {
      "epoch": 3.7642276422764227,
      "grad_norm": 0.13503999954911183,
      "learning_rate": 1.4009120827756242e-05,
      "loss": 0.4745,
      "step": 463
    },
    {
      "epoch": 3.772357723577236,
      "grad_norm": 0.14514854546052497,
      "learning_rate": 1.383680943085492e-05,
      "loss": 0.4671,
      "step": 464
    },
    {
      "epoch": 3.7804878048780486,
      "grad_norm": 0.12599219530305003,
      "learning_rate": 1.3665342416055873e-05,
      "loss": 0.4719,
      "step": 465
    },
    {
      "epoch": 3.7886178861788617,
      "grad_norm": 0.13775622977450616,
      "learning_rate": 1.3494725317228423e-05,
      "loss": 0.4711,
      "step": 466
    },
    {
      "epoch": 3.796747967479675,
      "grad_norm": 0.133763800356663,
      "learning_rate": 1.332496364081203e-05,
      "loss": 0.4719,
      "step": 467
    },
    {
      "epoch": 3.8048780487804876,
      "grad_norm": 0.14586869612736364,
      "learning_rate": 1.3156062865638526e-05,
      "loss": 0.4746,
      "step": 468
    },
    {
      "epoch": 3.813008130081301,
      "grad_norm": 0.1377873741303538,
      "learning_rate": 1.2988028442755307e-05,
      "loss": 0.4675,
      "step": 469
    },
    {
      "epoch": 3.821138211382114,
      "grad_norm": 0.12651895742756655,
      "learning_rate": 1.2820865795249372e-05,
      "loss": 0.4687,
      "step": 470
    },
    {
      "epoch": 3.8292682926829267,
      "grad_norm": 0.1188126211389867,
      "learning_rate": 1.2654580318072372e-05,
      "loss": 0.4767,
      "step": 471
    },
    {
      "epoch": 3.83739837398374,
      "grad_norm": 0.12222647609841775,
      "learning_rate": 1.2489177377866444e-05,
      "loss": 0.4708,
      "step": 472
    },
    {
      "epoch": 3.845528455284553,
      "grad_norm": 0.11861659063225727,
      "learning_rate": 1.2324662312790992e-05,
      "loss": 0.4693,
      "step": 473
    },
    {
      "epoch": 3.8536585365853657,
      "grad_norm": 0.11770245587575613,
      "learning_rate": 1.2161040432350472e-05,
      "loss": 0.4727,
      "step": 474
    },
    {
      "epoch": 3.861788617886179,
      "grad_norm": 0.12597805240086682,
      "learning_rate": 1.1998317017222992e-05,
      "loss": 0.4684,
      "step": 475
    },
    {
      "epoch": 3.869918699186992,
      "grad_norm": 0.11715886956884065,
      "learning_rate": 1.183649731908985e-05,
      "loss": 0.4666,
      "step": 476
    },
    {
      "epoch": 3.8780487804878048,
      "grad_norm": 0.11383133406509714,
      "learning_rate": 1.1675586560466137e-05,
      "loss": 0.4696,
      "step": 477
    },
    {
      "epoch": 3.886178861788618,
      "grad_norm": 0.12625827696349007,
      "learning_rate": 1.1515589934532088e-05,
      "loss": 0.4665,
      "step": 478
    },
    {
      "epoch": 3.894308943089431,
      "grad_norm": 0.11244226765728364,
      "learning_rate": 1.1356512604965544e-05,
      "loss": 0.472,
      "step": 479
    },
    {
      "epoch": 3.902439024390244,
      "grad_norm": 0.11376001709369589,
      "learning_rate": 1.1198359705775306e-05,
      "loss": 0.4714,
      "step": 480
    },
    {
      "epoch": 3.910569105691057,
      "grad_norm": 0.11689897982807441,
      "learning_rate": 1.1041136341135354e-05,
      "loss": 0.4644,
      "step": 481
    },
    {
      "epoch": 3.91869918699187,
      "grad_norm": 0.10796060424980164,
      "learning_rate": 1.0884847585220237e-05,
      "loss": 0.4694,
      "step": 482
    },
    {
      "epoch": 3.926829268292683,
      "grad_norm": 0.11238024777805515,
      "learning_rate": 1.0729498482041238e-05,
      "loss": 0.4707,
      "step": 483
    },
    {
      "epoch": 3.934959349593496,
      "grad_norm": 0.12259019342589682,
      "learning_rate": 1.0575094045283576e-05,
      "loss": 0.4687,
      "step": 484
    },
    {
      "epoch": 3.943089430894309,
      "grad_norm": 0.11186041699103712,
      "learning_rate": 1.042163925814466e-05,
      "loss": 0.4724,
      "step": 485
    },
    {
      "epoch": 3.951219512195122,
      "grad_norm": 0.1101113360054757,
      "learning_rate": 1.0269139073173177e-05,
      "loss": 0.4687,
      "step": 486
    },
    {
      "epoch": 3.959349593495935,
      "grad_norm": 0.1188727154785549,
      "learning_rate": 1.011759841210934e-05,
      "loss": 0.4618,
      "step": 487
    },
    {
      "epoch": 3.9674796747967482,
      "grad_norm": 0.10777217289944659,
      "learning_rate": 9.967022165725999e-06,
      "loss": 0.4707,
      "step": 488
    },
    {
      "epoch": 3.975609756097561,
      "grad_norm": 0.10517450382525091,
      "learning_rate": 9.817415193670791e-06,
      "loss": 0.4669,
      "step": 489
    },
    {
      "epoch": 3.983739837398374,
      "grad_norm": 0.1030414776848964,
      "learning_rate": 9.668782324309344e-06,
      "loss": 0.4705,
      "step": 490
    },
    {
      "epoch": 3.991869918699187,
      "grad_norm": 0.10946750285682907,
      "learning_rate": 9.521128354569375e-06,
      "loss": 0.4662,
      "step": 491
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.13514607298308584,
      "learning_rate": 9.374458049785962e-06,
      "loss": 0.4456,
      "step": 492
    },
    {
      "epoch": 4.008130081300813,
      "grad_norm": 0.19234663637384036,
      "learning_rate": 9.2287761435477e-06,
      "loss": 0.4337,
      "step": 493
    },
    {
      "epoch": 4.016260162601626,
      "grad_norm": 0.13229554188334003,
      "learning_rate": 9.084087337543911e-06,
      "loss": 0.4341,
      "step": 494
    },
    {
      "epoch": 4.024390243902439,
      "grad_norm": 0.15854748112255954,
      "learning_rate": 8.94039630141296e-06,
      "loss": 0.437,
      "step": 495
    },
    {
      "epoch": 4.032520325203252,
      "grad_norm": 0.2059109204451756,
      "learning_rate": 8.79770767259152e-06,
      "loss": 0.4341,
      "step": 496
    },
    {
      "epoch": 4.040650406504065,
      "grad_norm": 0.17409785007280926,
      "learning_rate": 8.656026056164872e-06,
      "loss": 0.4343,
      "step": 497
    },
    {
      "epoch": 4.048780487804878,
      "grad_norm": 0.14390829701534977,
      "learning_rate": 8.515356024718354e-06,
      "loss": 0.4346,
      "step": 498
    },
    {
      "epoch": 4.056910569105691,
      "grad_norm": 0.1438214101794298,
      "learning_rate": 8.375702118189735e-06,
      "loss": 0.4344,
      "step": 499
    },
    {
      "epoch": 4.065040650406504,
      "grad_norm": 0.1612407559469925,
      "learning_rate": 8.237068843722675e-06,
      "loss": 0.4308,
      "step": 500
    },
    {
      "epoch": 4.073170731707317,
      "grad_norm": 0.15882191025051573,
      "learning_rate": 8.09946067552135e-06,
      "loss": 0.4332,
      "step": 501
    },
    {
      "epoch": 4.08130081300813,
      "grad_norm": 0.13056607206312038,
      "learning_rate": 7.962882054705945e-06,
      "loss": 0.4355,
      "step": 502
    },
    {
      "epoch": 4.0894308943089435,
      "grad_norm": 0.1312593312804541,
      "learning_rate": 7.82733738916941e-06,
      "loss": 0.4395,
      "step": 503
    },
    {
      "epoch": 4.097560975609756,
      "grad_norm": 0.14229686113576429,
      "learning_rate": 7.692831053435163e-06,
      "loss": 0.4334,
      "step": 504
    },
    {
      "epoch": 4.105691056910569,
      "grad_norm": 0.14467590717195544,
      "learning_rate": 7.559367388515886e-06,
      "loss": 0.4356,
      "step": 505
    },
    {
      "epoch": 4.1138211382113825,
      "grad_norm": 0.13156872819712617,
      "learning_rate": 7.426950701773474e-06,
      "loss": 0.4408,
      "step": 506
    },
    {
      "epoch": 4.121951219512195,
      "grad_norm": 0.13378484036964042,
      "learning_rate": 7.295585266780008e-06,
      "loss": 0.4286,
      "step": 507
    },
    {
      "epoch": 4.130081300813008,
      "grad_norm": 0.13642799897305052,
      "learning_rate": 7.165275323179779e-06,
      "loss": 0.4428,
      "step": 508
    },
    {
      "epoch": 4.138211382113822,
      "grad_norm": 0.13772045002771224,
      "learning_rate": 7.036025076552535e-06,
      "loss": 0.4375,
      "step": 509
    },
    {
      "epoch": 4.146341463414634,
      "grad_norm": 0.1307223728872294,
      "learning_rate": 6.907838698277718e-06,
      "loss": 0.4315,
      "step": 510
    },
    {
      "epoch": 4.154471544715447,
      "grad_norm": 0.12693291651294142,
      "learning_rate": 6.780720325399803e-06,
      "loss": 0.4322,
      "step": 511
    },
    {
      "epoch": 4.16260162601626,
      "grad_norm": 0.12200923341460292,
      "learning_rate": 6.654674060494848e-06,
      "loss": 0.4333,
      "step": 512
    },
    {
      "epoch": 4.170731707317073,
      "grad_norm": 0.1355940352541764,
      "learning_rate": 6.529703971538039e-06,
      "loss": 0.4308,
      "step": 513
    },
    {
      "epoch": 4.178861788617886,
      "grad_norm": 0.12360188165547299,
      "learning_rate": 6.405814091772429e-06,
      "loss": 0.4335,
      "step": 514
    },
    {
      "epoch": 4.186991869918699,
      "grad_norm": 0.11241951311129478,
      "learning_rate": 6.283008419578758e-06,
      "loss": 0.4366,
      "step": 515
    },
    {
      "epoch": 4.195121951219512,
      "grad_norm": 0.11187105275422908,
      "learning_rate": 6.161290918346386e-06,
      "loss": 0.4318,
      "step": 516
    },
    {
      "epoch": 4.203252032520325,
      "grad_norm": 0.12210713771734147,
      "learning_rate": 6.040665516345439e-06,
      "loss": 0.4337,
      "step": 517
    },
    {
      "epoch": 4.211382113821138,
      "grad_norm": 0.10691274770111754,
      "learning_rate": 5.921136106599989e-06,
      "loss": 0.4383,
      "step": 518
    },
    {
      "epoch": 4.219512195121951,
      "grad_norm": 0.10869638865671052,
      "learning_rate": 5.8027065467624e-06,
      "loss": 0.4317,
      "step": 519
    },
    {
      "epoch": 4.227642276422764,
      "grad_norm": 0.11270202189820085,
      "learning_rate": 5.685380658988871e-06,
      "loss": 0.4376,
      "step": 520
    },
    {
      "epoch": 4.235772357723577,
      "grad_norm": 0.10225727344345675,
      "learning_rate": 5.569162229816027e-06,
      "loss": 0.4286,
      "step": 521
    },
    {
      "epoch": 4.2439024390243905,
      "grad_norm": 0.10428657960711893,
      "learning_rate": 5.454055010038776e-06,
      "loss": 0.4347,
      "step": 522
    },
    {
      "epoch": 4.252032520325203,
      "grad_norm": 0.11195468226426815,
      "learning_rate": 5.3400627145892046e-06,
      "loss": 0.4362,
      "step": 523
    },
    {
      "epoch": 4.260162601626016,
      "grad_norm": 0.11015348796500203,
      "learning_rate": 5.227189022416701e-06,
      "loss": 0.4339,
      "step": 524
    },
    {
      "epoch": 4.2682926829268295,
      "grad_norm": 0.10344126610786018,
      "learning_rate": 5.115437576369231e-06,
      "loss": 0.4288,
      "step": 525
    },
    {
      "epoch": 4.276422764227642,
      "grad_norm": 0.10460704526448507,
      "learning_rate": 5.004811983075768e-06,
      "loss": 0.4346,
      "step": 526
    },
    {
      "epoch": 4.284552845528455,
      "grad_norm": 0.11269380643731797,
      "learning_rate": 4.895315812829866e-06,
      "loss": 0.4327,
      "step": 527
    },
    {
      "epoch": 4.2926829268292686,
      "grad_norm": 0.10633291234198551,
      "learning_rate": 4.786952599474486e-06,
      "loss": 0.4316,
      "step": 528
    },
    {
      "epoch": 4.300813008130081,
      "grad_norm": 0.11157311343328219,
      "learning_rate": 4.679725840287885e-06,
      "loss": 0.4341,
      "step": 529
    },
    {
      "epoch": 4.308943089430894,
      "grad_norm": 0.1031060823781168,
      "learning_rate": 4.573638995870795e-06,
      "loss": 0.4347,
      "step": 530
    },
    {
      "epoch": 4.317073170731708,
      "grad_norm": 0.10602957854741088,
      "learning_rate": 4.468695490034733e-06,
      "loss": 0.4341,
      "step": 531
    },
    {
      "epoch": 4.32520325203252,
      "grad_norm": 0.10600472094993332,
      "learning_rate": 4.364898709691443e-06,
      "loss": 0.4318,
      "step": 532
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.1111788195089162,
      "learning_rate": 4.26225200474367e-06,
      "loss": 0.4293,
      "step": 533
    },
    {
      "epoch": 4.341463414634147,
      "grad_norm": 0.10444995212924384,
      "learning_rate": 4.160758687976994e-06,
      "loss": 0.433,
      "step": 534
    },
    {
      "epoch": 4.349593495934959,
      "grad_norm": 0.10608364581240926,
      "learning_rate": 4.060422034952942e-06,
      "loss": 0.4361,
      "step": 535
    },
    {
      "epoch": 4.357723577235772,
      "grad_norm": 0.10766877201382892,
      "learning_rate": 3.961245283903239e-06,
      "loss": 0.4352,
      "step": 536
    },
    {
      "epoch": 4.365853658536586,
      "grad_norm": 0.11354105620535715,
      "learning_rate": 3.863231635625328e-06,
      "loss": 0.4322,
      "step": 537
    },
    {
      "epoch": 4.373983739837398,
      "grad_norm": 0.10631416844578713,
      "learning_rate": 3.7663842533790697e-06,
      "loss": 0.4357,
      "step": 538
    },
    {
      "epoch": 4.382113821138211,
      "grad_norm": 0.10328798173757889,
      "learning_rate": 3.6707062627846425e-06,
      "loss": 0.4343,
      "step": 539
    },
    {
      "epoch": 4.390243902439025,
      "grad_norm": 0.10392661605869524,
      "learning_rate": 3.576200751721648e-06,
      "loss": 0.4347,
      "step": 540
    },
    {
      "epoch": 4.3983739837398375,
      "grad_norm": 0.10204660854825877,
      "learning_rate": 3.482870770229494e-06,
      "loss": 0.4341,
      "step": 541
    },
    {
      "epoch": 4.40650406504065,
      "grad_norm": 0.10132350456787592,
      "learning_rate": 3.3907193304089403e-06,
      "loss": 0.4247,
      "step": 542
    },
    {
      "epoch": 4.414634146341464,
      "grad_norm": 0.11656166305731963,
      "learning_rate": 3.2997494063248615e-06,
      "loss": 0.4377,
      "step": 543
    },
    {
      "epoch": 4.4227642276422765,
      "grad_norm": 0.10523482078415457,
      "learning_rate": 3.209963933910314e-06,
      "loss": 0.4313,
      "step": 544
    },
    {
      "epoch": 4.430894308943089,
      "grad_norm": 0.10067041225634954,
      "learning_rate": 3.1213658108717372e-06,
      "loss": 0.4322,
      "step": 545
    },
    {
      "epoch": 4.439024390243903,
      "grad_norm": 0.10555800131037159,
      "learning_rate": 3.0339578965954542e-06,
      "loss": 0.4363,
      "step": 546
    },
    {
      "epoch": 4.4471544715447155,
      "grad_norm": 0.10122569047137354,
      "learning_rate": 2.9477430120554086e-06,
      "loss": 0.4335,
      "step": 547
    },
    {
      "epoch": 4.455284552845528,
      "grad_norm": 0.10624994220616767,
      "learning_rate": 2.862723939722063e-06,
      "loss": 0.4347,
      "step": 548
    },
    {
      "epoch": 4.463414634146342,
      "grad_norm": 0.09908267669871018,
      "learning_rate": 2.778903423472659e-06,
      "loss": 0.4347,
      "step": 549
    },
    {
      "epoch": 4.471544715447155,
      "grad_norm": 0.10201504371802464,
      "learning_rate": 2.696284168502641e-06,
      "loss": 0.431,
      "step": 550
    },
    {
      "epoch": 4.479674796747967,
      "grad_norm": 0.09799805847484688,
      "learning_rate": 2.614868841238325e-06,
      "loss": 0.4365,
      "step": 551
    },
    {
      "epoch": 4.487804878048781,
      "grad_norm": 0.11127919339545263,
      "learning_rate": 2.5346600692508894e-06,
      "loss": 0.4366,
      "step": 552
    },
    {
      "epoch": 4.495934959349594,
      "grad_norm": 0.10037119125840871,
      "learning_rate": 2.4556604411715322e-06,
      "loss": 0.436,
      "step": 553
    },
    {
      "epoch": 4.504065040650406,
      "grad_norm": 0.09734374510859575,
      "learning_rate": 2.377872506607952e-06,
      "loss": 0.4304,
      "step": 554
    },
    {
      "epoch": 4.512195121951219,
      "grad_norm": 0.09811087062331385,
      "learning_rate": 2.301298776062053e-06,
      "loss": 0.4302,
      "step": 555
    },
    {
      "epoch": 4.520325203252033,
      "grad_norm": 0.10038291737534318,
      "learning_rate": 2.225941720848921e-06,
      "loss": 0.4317,
      "step": 556
    },
    {
      "epoch": 4.528455284552845,
      "grad_norm": 0.10199072159335258,
      "learning_rate": 2.151803773017065e-06,
      "loss": 0.4362,
      "step": 557
    },
    {
      "epoch": 4.536585365853659,
      "grad_norm": 0.10396980855728817,
      "learning_rate": 2.078887325269934e-06,
      "loss": 0.4341,
      "step": 558
    },
    {
      "epoch": 4.544715447154472,
      "grad_norm": 0.09873025477330767,
      "learning_rate": 2.007194730888693e-06,
      "loss": 0.4332,
      "step": 559
    },
    {
      "epoch": 4.5528455284552845,
      "grad_norm": 0.09621250084883998,
      "learning_rate": 1.9367283036562632e-06,
      "loss": 0.4377,
      "step": 560
    },
    {
      "epoch": 4.560975609756097,
      "grad_norm": 0.09528596628490693,
      "learning_rate": 1.8674903177826475e-06,
      "loss": 0.4364,
      "step": 561
    },
    {
      "epoch": 4.569105691056911,
      "grad_norm": 0.09765508685386991,
      "learning_rate": 1.799483007831566e-06,
      "loss": 0.4317,
      "step": 562
    },
    {
      "epoch": 4.5772357723577235,
      "grad_norm": 0.0973921120993045,
      "learning_rate": 1.7327085686482937e-06,
      "loss": 0.4303,
      "step": 563
    },
    {
      "epoch": 4.585365853658536,
      "grad_norm": 0.09371077139966895,
      "learning_rate": 1.6671691552888437e-06,
      "loss": 0.4274,
      "step": 564
    },
    {
      "epoch": 4.59349593495935,
      "grad_norm": 0.09386858903830189,
      "learning_rate": 1.602866882950429e-06,
      "loss": 0.427,
      "step": 565
    },
    {
      "epoch": 4.6016260162601625,
      "grad_norm": 0.09551835015171915,
      "learning_rate": 1.5398038269031834e-06,
      "loss": 0.4332,
      "step": 566
    },
    {
      "epoch": 4.609756097560975,
      "grad_norm": 0.09237743337404059,
      "learning_rate": 1.4779820224231788e-06,
      "loss": 0.4367,
      "step": 567
    },
    {
      "epoch": 4.617886178861789,
      "grad_norm": 0.09252488905767112,
      "learning_rate": 1.4174034647267499e-06,
      "loss": 0.4321,
      "step": 568
    },
    {
      "epoch": 4.626016260162602,
      "grad_norm": 0.09512689612613341,
      "learning_rate": 1.3580701089061043e-06,
      "loss": 0.4381,
      "step": 569
    },
    {
      "epoch": 4.634146341463414,
      "grad_norm": 0.0957694046004578,
      "learning_rate": 1.2999838698662105e-06,
      "loss": 0.4355,
      "step": 570
    },
    {
      "epoch": 4.642276422764228,
      "grad_norm": 0.09400332001222222,
      "learning_rate": 1.2431466222630185e-06,
      "loss": 0.432,
      "step": 571
    },
    {
      "epoch": 4.650406504065041,
      "grad_norm": 0.09364356449106392,
      "learning_rate": 1.1875602004429232e-06,
      "loss": 0.4325,
      "step": 572
    },
    {
      "epoch": 4.658536585365853,
      "grad_norm": 0.09829881123868839,
      "learning_rate": 1.1332263983835978e-06,
      "loss": 0.4384,
      "step": 573
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.09355483367575865,
      "learning_rate": 1.0801469696360934e-06,
      "loss": 0.4326,
      "step": 574
    },
    {
      "epoch": 4.67479674796748,
      "grad_norm": 0.0924248378202357,
      "learning_rate": 1.0283236272682128e-06,
      "loss": 0.4266,
      "step": 575
    },
    {
      "epoch": 4.682926829268292,
      "grad_norm": 0.09447921099981366,
      "learning_rate": 9.777580438092537e-07,
      "loss": 0.4368,
      "step": 576
    },
    {
      "epoch": 4.691056910569106,
      "grad_norm": 0.0936529818837489,
      "learning_rate": 9.284518511960283e-07,
      "loss": 0.4348,
      "step": 577
    },
    {
      "epoch": 4.699186991869919,
      "grad_norm": 0.09281565430560301,
      "learning_rate": 8.804066407201772e-07,
      "loss": 0.428,
      "step": 578
    },
    {
      "epoch": 4.7073170731707314,
      "grad_norm": 0.09490187282777024,
      "learning_rate": 8.336239629768283e-07,
      "loss": 0.4412,
      "step": 579
    },
    {
      "epoch": 4.715447154471545,
      "grad_norm": 0.09322107019930924,
      "learning_rate": 7.881053278145567e-07,
      "loss": 0.4324,
      "step": 580
    },
    {
      "epoch": 4.723577235772358,
      "grad_norm": 0.09554874672128658,
      "learning_rate": 7.43852204286628e-07,
      "loss": 0.4363,
      "step": 581
    },
    {
      "epoch": 4.7317073170731705,
      "grad_norm": 0.09634924288844665,
      "learning_rate": 7.008660206036233e-07,
      "loss": 0.4378,
      "step": 582
    },
    {
      "epoch": 4.739837398373984,
      "grad_norm": 0.09301988306633939,
      "learning_rate": 6.591481640873243e-07,
      "loss": 0.4306,
      "step": 583
    },
    {
      "epoch": 4.747967479674797,
      "grad_norm": 0.0955773271965025,
      "learning_rate": 6.186999811259454e-07,
      "loss": 0.4339,
      "step": 584
    },
    {
      "epoch": 4.7560975609756095,
      "grad_norm": 0.09560772426866701,
      "learning_rate": 5.795227771306788e-07,
      "loss": 0.4317,
      "step": 585
    },
    {
      "epoch": 4.764227642276423,
      "grad_norm": 0.09318368133987177,
      "learning_rate": 5.416178164935559e-07,
      "loss": 0.437,
      "step": 586
    },
    {
      "epoch": 4.772357723577236,
      "grad_norm": 0.09214235181380076,
      "learning_rate": 5.049863225466656e-07,
      "loss": 0.4331,
      "step": 587
    },
    {
      "epoch": 4.780487804878049,
      "grad_norm": 0.09728904652532608,
      "learning_rate": 4.6962947752263956e-07,
      "loss": 0.437,
      "step": 588
    },
    {
      "epoch": 4.788617886178862,
      "grad_norm": 0.09517528126859258,
      "learning_rate": 4.3554842251653186e-07,
      "loss": 0.4361,
      "step": 589
    },
    {
      "epoch": 4.796747967479675,
      "grad_norm": 0.0918896668723724,
      "learning_rate": 4.027442574489593e-07,
      "loss": 0.4367,
      "step": 590
    },
    {
      "epoch": 4.804878048780488,
      "grad_norm": 0.09467919967854278,
      "learning_rate": 3.7121804103062765e-07,
      "loss": 0.4308,
      "step": 591
    },
    {
      "epoch": 4.8130081300813,
      "grad_norm": 0.0968626822559012,
      "learning_rate": 3.4097079072814565e-07,
      "loss": 0.4357,
      "step": 592
    },
    {
      "epoch": 4.821138211382114,
      "grad_norm": 0.0936688868676205,
      "learning_rate": 3.1200348273121127e-07,
      "loss": 0.4335,
      "step": 593
    },
    {
      "epoch": 4.829268292682927,
      "grad_norm": 0.09249308909898428,
      "learning_rate": 2.843170519210636e-07,
      "loss": 0.4297,
      "step": 594
    },
    {
      "epoch": 4.83739837398374,
      "grad_norm": 0.09313534867978861,
      "learning_rate": 2.5791239184036474e-07,
      "loss": 0.4296,
      "step": 595
    },
    {
      "epoch": 4.845528455284553,
      "grad_norm": 0.09634815047495345,
      "learning_rate": 2.327903546643162e-07,
      "loss": 0.43,
      "step": 596
    },
    {
      "epoch": 4.853658536585366,
      "grad_norm": 0.09272748034725385,
      "learning_rate": 2.0895175117318756e-07,
      "loss": 0.4327,
      "step": 597
    },
    {
      "epoch": 4.861788617886178,
      "grad_norm": 0.0932498768878636,
      "learning_rate": 1.8639735072612852e-07,
      "loss": 0.4381,
      "step": 598
    },
    {
      "epoch": 4.869918699186992,
      "grad_norm": 0.09371786035357835,
      "learning_rate": 1.6512788123635325e-07,
      "loss": 0.4343,
      "step": 599
    },
    {
      "epoch": 4.878048780487805,
      "grad_norm": 0.0956702717727826,
      "learning_rate": 1.451440291476347e-07,
      "loss": 0.4324,
      "step": 600
    },
    {
      "epoch": 4.886178861788618,
      "grad_norm": 0.09477457919502134,
      "learning_rate": 1.2644643941216673e-07,
      "loss": 0.4328,
      "step": 601
    },
    {
      "epoch": 4.894308943089431,
      "grad_norm": 0.09615053578039959,
      "learning_rate": 1.0903571546973191e-07,
      "loss": 0.4426,
      "step": 602
    },
    {
      "epoch": 4.902439024390244,
      "grad_norm": 0.09402219901230496,
      "learning_rate": 9.291241922824157e-08,
      "loss": 0.4301,
      "step": 603
    },
    {
      "epoch": 4.9105691056910565,
      "grad_norm": 0.09437264344078931,
      "learning_rate": 7.807707104559026e-08,
      "loss": 0.4329,
      "step": 604
    },
    {
      "epoch": 4.91869918699187,
      "grad_norm": 0.09282091883311075,
      "learning_rate": 6.453014971286475e-08,
      "loss": 0.4278,
      "step": 605
    },
    {
      "epoch": 4.926829268292683,
      "grad_norm": 0.09657035596512123,
      "learning_rate": 5.2272092438894194e-08,
      "loss": 0.4327,
      "step": 606
    },
    {
      "epoch": 4.934959349593496,
      "grad_norm": 0.09363835686955926,
      "learning_rate": 4.130329483614137e-08,
      "loss": 0.4334,
      "step": 607
    },
    {
      "epoch": 4.943089430894309,
      "grad_norm": 0.09396144971453399,
      "learning_rate": 3.162411090792627e-08,
      "loss": 0.4355,
      "step": 608
    },
    {
      "epoch": 4.951219512195122,
      "grad_norm": 0.09234809286300218,
      "learning_rate": 2.3234853037008566e-08,
      "loss": 0.435,
      "step": 609
    },
    {
      "epoch": 4.959349593495935,
      "grad_norm": 0.0926664310323078,
      "learning_rate": 1.613579197549786e-08,
      "loss": 0.4319,
      "step": 610
    },
    {
      "epoch": 4.967479674796748,
      "grad_norm": 0.09275825122159159,
      "learning_rate": 1.0327156836140718e-08,
      "loss": 0.4297,
      "step": 611
    },
    {
      "epoch": 4.975609756097561,
      "grad_norm": 0.09351695018990755,
      "learning_rate": 5.809135084882123e-09,
      "loss": 0.4324,
      "step": 612
    },
    {
      "epoch": 4.983739837398374,
      "grad_norm": 0.09720775795174409,
      "learning_rate": 2.5818725348569773e-09,
      "loss": 0.4343,
      "step": 613
    },
    {
      "epoch": 4.991869918699187,
      "grad_norm": 0.09248737373929655,
      "learning_rate": 6.454733416605407e-10,
      "loss": 0.4329,
      "step": 614
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.10012242323902595,
      "learning_rate": 0.0,
      "loss": 0.4256,
      "step": 615
    },
    {
      "epoch": 5.0,
      "step": 615,
      "total_flos": 1.031732748877824e+16,
      "train_loss": 0.5654353130154494,
      "train_runtime": 8268.7983,
      "train_samples_per_second": 37.948,
      "train_steps_per_second": 0.074
    }
  ],
  "logging_steps": 1,
  "max_steps": 615,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.031732748877824e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}