{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 454.58181818181816,
  "eval_steps": 10,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09696969696969697,
      "grad_norm": 0.1921696960926056,
      "learning_rate": 4.999e-05,
      "loss": 2.8765,
      "step": 1
    },
    {
      "epoch": 0.19393939393939394,
      "grad_norm": 0.19308380782604218,
      "learning_rate": 4.9980000000000006e-05,
      "loss": 2.8673,
      "step": 2
    },
    {
      "epoch": 0.2909090909090909,
      "grad_norm": 0.19083072245121002,
      "learning_rate": 4.997e-05,
      "loss": 2.8733,
      "step": 3
    },
    {
      "epoch": 0.3878787878787879,
      "grad_norm": 0.19004836678504944,
      "learning_rate": 4.996e-05,
      "loss": 2.892,
      "step": 4
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 0.18908675014972687,
      "learning_rate": 4.995e-05,
      "loss": 2.88,
      "step": 5
    },
    {
      "epoch": 0.5818181818181818,
      "grad_norm": 0.19075514376163483,
      "learning_rate": 4.9940000000000006e-05,
      "loss": 2.8799,
      "step": 6
    },
    {
      "epoch": 0.6787878787878788,
      "grad_norm": 0.19584377110004425,
      "learning_rate": 4.9930000000000005e-05,
      "loss": 2.8752,
      "step": 7
    },
    {
      "epoch": 0.7757575757575758,
      "grad_norm": 0.19357410073280334,
      "learning_rate": 4.992e-05,
      "loss": 2.8753,
      "step": 8
    },
    {
      "epoch": 0.8727272727272727,
      "grad_norm": 0.193776935338974,
      "learning_rate": 4.991e-05,
      "loss": 2.8731,
      "step": 9
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.19457189738750458,
      "learning_rate": 4.99e-05,
      "loss": 2.8538,
      "step": 10
    },
    {
      "epoch": 0.9696969696969697,
      "eval_loss": 2.8589916229248047,
      "eval_runtime": 2.0031,
      "eval_samples_per_second": 27.458,
      "eval_steps_per_second": 3.495,
      "step": 10
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.191086545586586,
      "learning_rate": 4.9890000000000005e-05,
      "loss": 2.8476,
      "step": 11
    },
    {
      "epoch": 1.096969696969697,
      "grad_norm": 0.19667789340019226,
      "learning_rate": 4.9880000000000004e-05,
      "loss": 2.8696,
      "step": 12
    },
    {
      "epoch": 1.1939393939393939,
      "grad_norm": 0.2008863240480423,
      "learning_rate": 4.987e-05,
      "loss": 2.8536,
      "step": 13
    },
    {
      "epoch": 1.290909090909091,
      "grad_norm": 0.19403468072414398,
      "learning_rate": 4.986e-05,
      "loss": 2.8426,
      "step": 14
    },
    {
      "epoch": 1.387878787878788,
      "grad_norm": 0.2005588710308075,
      "learning_rate": 4.9850000000000006e-05,
      "loss": 2.8555,
      "step": 15
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 0.20173998177051544,
      "learning_rate": 4.9840000000000004e-05,
      "loss": 2.852,
      "step": 16
    },
    {
      "epoch": 1.5818181818181818,
      "grad_norm": 0.19781255722045898,
      "learning_rate": 4.983e-05,
      "loss": 2.8481,
      "step": 17
    },
    {
      "epoch": 1.6787878787878787,
      "grad_norm": 0.19945280253887177,
      "learning_rate": 4.982e-05,
      "loss": 2.8432,
      "step": 18
    },
    {
      "epoch": 1.7757575757575759,
      "grad_norm": 0.2086201161146164,
      "learning_rate": 4.981e-05,
      "loss": 2.8603,
      "step": 19
    },
    {
      "epoch": 1.8727272727272726,
      "grad_norm": 0.20707911252975464,
      "learning_rate": 4.9800000000000004e-05,
      "loss": 2.8205,
      "step": 20
    },
    {
      "epoch": 1.8727272727272726,
      "eval_loss": 2.8336637020111084,
      "eval_runtime": 2.097,
      "eval_samples_per_second": 26.228,
      "eval_steps_per_second": 3.338,
      "step": 20
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.20440326631069183,
      "learning_rate": 4.979e-05,
      "loss": 2.8464,
      "step": 21
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.21748438477516174,
      "learning_rate": 4.978e-05,
      "loss": 2.8079,
      "step": 22
    },
    {
      "epoch": 2.096969696969697,
      "grad_norm": 0.21118205785751343,
      "learning_rate": 4.977e-05,
      "loss": 2.8453,
      "step": 23
    },
    {
      "epoch": 2.193939393939394,
      "grad_norm": 0.2079344540834427,
      "learning_rate": 4.976e-05,
      "loss": 2.8295,
      "step": 24
    },
    {
      "epoch": 2.290909090909091,
      "grad_norm": 0.21478907763957977,
      "learning_rate": 4.975e-05,
      "loss": 2.8252,
      "step": 25
    },
    {
      "epoch": 2.3878787878787877,
      "grad_norm": 0.21348024904727936,
      "learning_rate": 4.974e-05,
      "loss": 2.8124,
      "step": 26
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 0.22032339870929718,
      "learning_rate": 4.973000000000001e-05,
      "loss": 2.8064,
      "step": 27
    },
    {
      "epoch": 2.581818181818182,
      "grad_norm": 0.21923890709877014,
      "learning_rate": 4.972e-05,
      "loss": 2.8035,
      "step": 28
    },
    {
      "epoch": 2.6787878787878787,
      "grad_norm": 0.22203513979911804,
      "learning_rate": 4.9710000000000003e-05,
      "loss": 2.8098,
      "step": 29
    },
    {
      "epoch": 2.775757575757576,
      "grad_norm": 0.22282050549983978,
      "learning_rate": 4.97e-05,
      "loss": 2.8159,
      "step": 30
    },
    {
      "epoch": 2.775757575757576,
      "eval_loss": 2.804253339767456,
      "eval_runtime": 2.1435,
      "eval_samples_per_second": 25.659,
      "eval_steps_per_second": 3.266,
      "step": 30
    },
    {
      "epoch": 2.8727272727272726,
      "grad_norm": 0.22722773253917694,
      "learning_rate": 4.969e-05,
      "loss": 2.8072,
      "step": 31
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 0.2269333004951477,
      "learning_rate": 4.9680000000000005e-05,
      "loss": 2.8199,
      "step": 32
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.23603735864162445,
      "learning_rate": 4.967e-05,
      "loss": 2.8051,
      "step": 33
    },
    {
      "epoch": 3.096969696969697,
      "grad_norm": 0.226282000541687,
      "learning_rate": 4.966e-05,
      "loss": 2.8064,
      "step": 34
    },
    {
      "epoch": 3.193939393939394,
      "grad_norm": 0.23394860327243805,
      "learning_rate": 4.965e-05,
      "loss": 2.8158,
      "step": 35
    },
    {
      "epoch": 3.290909090909091,
      "grad_norm": 0.22885440289974213,
      "learning_rate": 4.9640000000000006e-05,
      "loss": 2.7904,
      "step": 36
    },
    {
      "epoch": 3.3878787878787877,
      "grad_norm": 0.23805929720401764,
      "learning_rate": 4.9630000000000004e-05,
      "loss": 2.7877,
      "step": 37
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.23020139336585999,
      "learning_rate": 4.962e-05,
      "loss": 2.787,
      "step": 38
    },
    {
      "epoch": 3.581818181818182,
      "grad_norm": 0.2334854155778885,
      "learning_rate": 4.961e-05,
      "loss": 2.7586,
      "step": 39
    },
    {
      "epoch": 3.6787878787878787,
      "grad_norm": 0.23337151110172272,
      "learning_rate": 4.96e-05,
      "loss": 2.7849,
      "step": 40
    },
    {
      "epoch": 3.6787878787878787,
      "eval_loss": 2.7703146934509277,
      "eval_runtime": 2.1916,
      "eval_samples_per_second": 25.096,
      "eval_steps_per_second": 3.194,
      "step": 40
    },
    {
      "epoch": 3.775757575757576,
      "grad_norm": 0.23387908935546875,
      "learning_rate": 4.9590000000000005e-05,
      "loss": 2.7605,
      "step": 41
    },
    {
      "epoch": 3.8727272727272726,
      "grad_norm": 0.23480385541915894,
      "learning_rate": 4.958e-05,
      "loss": 2.7698,
      "step": 42
    },
    {
      "epoch": 3.9696969696969697,
      "grad_norm": 0.23578350245952606,
      "learning_rate": 4.957e-05,
      "loss": 2.7539,
      "step": 43
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.24619624018669128,
      "learning_rate": 4.956e-05,
      "loss": 2.769,
      "step": 44
    },
    {
      "epoch": 4.096969696969697,
      "grad_norm": 0.23596659302711487,
      "learning_rate": 4.9550000000000005e-05,
      "loss": 2.7477,
      "step": 45
    },
    {
      "epoch": 4.193939393939394,
      "grad_norm": 0.23750188946723938,
      "learning_rate": 4.9540000000000003e-05,
      "loss": 2.7477,
      "step": 46
    },
    {
      "epoch": 4.290909090909091,
      "grad_norm": 0.24240151047706604,
      "learning_rate": 4.953e-05,
      "loss": 2.7516,
      "step": 47
    },
    {
      "epoch": 4.387878787878788,
      "grad_norm": 0.23533636331558228,
      "learning_rate": 4.952e-05,
      "loss": 2.7578,
      "step": 48
    },
    {
      "epoch": 4.484848484848484,
      "grad_norm": 0.2335629016160965,
      "learning_rate": 4.951e-05,
      "loss": 2.7476,
      "step": 49
    },
    {
      "epoch": 4.581818181818182,
      "grad_norm": 0.2371414303779602,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 2.7223,
      "step": 50
    },
    {
      "epoch": 4.581818181818182,
      "eval_loss": 2.7336995601654053,
      "eval_runtime": 2.1993,
      "eval_samples_per_second": 25.008,
      "eval_steps_per_second": 3.183,
      "step": 50
    },
    {
      "epoch": 4.678787878787879,
      "grad_norm": 0.2299228310585022,
      "learning_rate": 4.949e-05,
      "loss": 2.756,
      "step": 51
    },
    {
      "epoch": 4.775757575757575,
      "grad_norm": 0.23399955034255981,
      "learning_rate": 4.948000000000001e-05,
      "loss": 2.7664,
      "step": 52
    },
    {
      "epoch": 4.872727272727273,
      "grad_norm": 0.23219211399555206,
      "learning_rate": 4.947e-05,
      "loss": 2.7277,
      "step": 53
    },
    {
      "epoch": 4.96969696969697,
      "grad_norm": 0.23143132030963898,
      "learning_rate": 4.946e-05,
      "loss": 2.6945,
      "step": 54
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.24724680185317993,
      "learning_rate": 4.945e-05,
      "loss": 2.7077,
      "step": 55
    },
    {
      "epoch": 5.096969696969697,
      "grad_norm": 0.23197874426841736,
      "learning_rate": 4.944e-05,
      "loss": 2.7047,
      "step": 56
    },
    {
      "epoch": 5.193939393939394,
      "grad_norm": 0.23464535176753998,
      "learning_rate": 4.9430000000000006e-05,
      "loss": 2.7158,
      "step": 57
    },
    {
      "epoch": 5.290909090909091,
      "grad_norm": 0.23159249126911163,
      "learning_rate": 4.942e-05,
      "loss": 2.7076,
      "step": 58
    },
    {
      "epoch": 5.387878787878788,
      "grad_norm": 0.23467811942100525,
      "learning_rate": 4.941e-05,
      "loss": 2.7272,
      "step": 59
    },
    {
      "epoch": 5.484848484848484,
      "grad_norm": 0.23297619819641113,
      "learning_rate": 4.94e-05,
      "loss": 2.6964,
      "step": 60
    },
    {
      "epoch": 5.484848484848484,
      "eval_loss": 2.694591760635376,
      "eval_runtime": 2.2142,
      "eval_samples_per_second": 24.84,
      "eval_steps_per_second": 3.161,
      "step": 60
    },
    {
      "epoch": 5.581818181818182,
      "grad_norm": 0.23715738952159882,
      "learning_rate": 4.939e-05,
      "loss": 2.6981,
      "step": 61
    },
    {
      "epoch": 5.678787878787879,
      "grad_norm": 0.23313502967357635,
      "learning_rate": 4.9380000000000005e-05,
      "loss": 2.6859,
      "step": 62
    },
    {
      "epoch": 5.775757575757575,
      "grad_norm": 0.23508723080158234,
      "learning_rate": 4.937e-05,
      "loss": 2.6977,
      "step": 63
    },
    {
      "epoch": 5.872727272727273,
      "grad_norm": 0.2312891036272049,
      "learning_rate": 4.936e-05,
      "loss": 2.6921,
      "step": 64
    },
    {
      "epoch": 5.96969696969697,
      "grad_norm": 0.232807457447052,
      "learning_rate": 4.935e-05,
      "loss": 2.6618,
      "step": 65
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.23697008192539215,
      "learning_rate": 4.9340000000000005e-05,
      "loss": 2.6795,
      "step": 66
    },
    {
      "epoch": 6.096969696969697,
      "grad_norm": 0.23080474138259888,
      "learning_rate": 4.9330000000000004e-05,
      "loss": 2.6681,
      "step": 67
    },
    {
      "epoch": 6.193939393939394,
      "grad_norm": 0.2305583357810974,
      "learning_rate": 4.932e-05,
      "loss": 2.6906,
      "step": 68
    },
    {
      "epoch": 6.290909090909091,
      "grad_norm": 0.2397168129682541,
      "learning_rate": 4.931e-05,
      "loss": 2.6724,
      "step": 69
    },
    {
      "epoch": 6.387878787878788,
      "grad_norm": 0.23490968346595764,
      "learning_rate": 4.93e-05,
      "loss": 2.6539,
      "step": 70
    },
    {
      "epoch": 6.387878787878788,
      "eval_loss": 2.654001474380493,
      "eval_runtime": 2.1567,
      "eval_samples_per_second": 25.502,
      "eval_steps_per_second": 3.246,
      "step": 70
    },
    {
      "epoch": 6.484848484848484,
      "grad_norm": 0.23402239382266998,
      "learning_rate": 4.9290000000000004e-05,
      "loss": 2.6689,
      "step": 71
    },
    {
      "epoch": 6.581818181818182,
      "grad_norm": 0.23379398882389069,
      "learning_rate": 4.928e-05,
      "loss": 2.6511,
      "step": 72
    },
    {
      "epoch": 6.678787878787879,
      "grad_norm": 0.23643533885478973,
      "learning_rate": 4.927000000000001e-05,
      "loss": 2.6132,
      "step": 73
    },
    {
      "epoch": 6.775757575757575,
      "grad_norm": 0.23277723789215088,
      "learning_rate": 4.926e-05,
      "loss": 2.6498,
      "step": 74
    },
    {
      "epoch": 6.872727272727273,
      "grad_norm": 0.23537500202655792,
      "learning_rate": 4.9250000000000004e-05,
      "loss": 2.631,
      "step": 75
    },
    {
      "epoch": 6.96969696969697,
      "grad_norm": 0.22895337641239166,
      "learning_rate": 4.924e-05,
      "loss": 2.6461,
      "step": 76
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.24864304065704346,
      "learning_rate": 4.923e-05,
      "loss": 2.6042,
      "step": 77
    },
    {
      "epoch": 7.096969696969697,
      "grad_norm": 0.23936185240745544,
      "learning_rate": 4.9220000000000006e-05,
      "loss": 2.6338,
      "step": 78
    },
    {
      "epoch": 7.193939393939394,
      "grad_norm": 0.2343270629644394,
      "learning_rate": 4.921e-05,
      "loss": 2.5853,
      "step": 79
    },
    {
      "epoch": 7.290909090909091,
      "grad_norm": 0.2339862585067749,
      "learning_rate": 4.92e-05,
      "loss": 2.6097,
      "step": 80
    },
    {
      "epoch": 7.290909090909091,
      "eval_loss": 2.6114232540130615,
      "eval_runtime": 2.154,
      "eval_samples_per_second": 25.534,
      "eval_steps_per_second": 3.25,
      "step": 80
    },
    {
      "epoch": 7.387878787878788,
      "grad_norm": 0.2315565049648285,
      "learning_rate": 4.919e-05,
      "loss": 2.6263,
      "step": 81
    },
    {
      "epoch": 7.484848484848484,
      "grad_norm": 0.2334234118461609,
      "learning_rate": 4.918000000000001e-05,
      "loss": 2.6175,
      "step": 82
    },
    {
      "epoch": 7.581818181818182,
      "grad_norm": 0.23795263469219208,
      "learning_rate": 4.9170000000000005e-05,
      "loss": 2.6318,
      "step": 83
    },
    {
      "epoch": 7.678787878787879,
      "grad_norm": 0.24352413415908813,
      "learning_rate": 4.9160000000000004e-05,
      "loss": 2.6086,
      "step": 84
    },
    {
      "epoch": 7.775757575757575,
      "grad_norm": 0.23900386691093445,
      "learning_rate": 4.915e-05,
      "loss": 2.5974,
      "step": 85
    },
    {
      "epoch": 7.872727272727273,
      "grad_norm": 0.24548974633216858,
      "learning_rate": 4.914e-05,
      "loss": 2.5798,
      "step": 86
    },
    {
      "epoch": 7.96969696969697,
      "grad_norm": 0.2442977875471115,
      "learning_rate": 4.9130000000000006e-05,
      "loss": 2.5816,
      "step": 87
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.24905452132225037,
      "learning_rate": 4.9120000000000004e-05,
      "loss": 2.55,
      "step": 88
    },
    {
      "epoch": 8.096969696969698,
      "grad_norm": 0.23655293881893158,
      "learning_rate": 4.911e-05,
      "loss": 2.563,
      "step": 89
    },
    {
      "epoch": 8.193939393939393,
      "grad_norm": 0.24191634356975555,
      "learning_rate": 4.91e-05,
      "loss": 2.585,
      "step": 90
    },
    {
      "epoch": 8.193939393939393,
      "eval_loss": 2.5664913654327393,
      "eval_runtime": 2.1414,
      "eval_samples_per_second": 25.684,
      "eval_steps_per_second": 3.269,
      "step": 90
    },
    {
      "epoch": 8.290909090909091,
      "grad_norm": 0.2470662146806717,
      "learning_rate": 4.9090000000000006e-05,
      "loss": 2.5824,
      "step": 91
    },
    {
      "epoch": 8.387878787878789,
      "grad_norm": 0.24474897980690002,
      "learning_rate": 4.9080000000000004e-05,
      "loss": 2.5554,
      "step": 92
    },
    {
      "epoch": 8.484848484848484,
      "grad_norm": 0.2424492985010147,
      "learning_rate": 4.907e-05,
      "loss": 2.552,
      "step": 93
    },
    {
      "epoch": 8.581818181818182,
      "grad_norm": 0.24509666860103607,
      "learning_rate": 4.906e-05,
      "loss": 2.5625,
      "step": 94
    },
    {
      "epoch": 8.67878787878788,
      "grad_norm": 0.2482563704252243,
      "learning_rate": 4.905e-05,
      "loss": 2.5262,
      "step": 95
    },
    {
      "epoch": 8.775757575757575,
      "grad_norm": 0.2569093704223633,
      "learning_rate": 4.9040000000000005e-05,
      "loss": 2.5401,
      "step": 96
    },
    {
      "epoch": 8.872727272727273,
      "grad_norm": 0.2476951777935028,
      "learning_rate": 4.903e-05,
      "loss": 2.5603,
      "step": 97
    },
    {
      "epoch": 8.969696969696969,
      "grad_norm": 0.250521719455719,
      "learning_rate": 4.902e-05,
      "loss": 2.5349,
      "step": 98
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.2586294114589691,
      "learning_rate": 4.901e-05,
      "loss": 2.5142,
      "step": 99
    },
    {
      "epoch": 9.096969696969698,
      "grad_norm": 0.2525769770145416,
      "learning_rate": 4.9e-05,
      "loss": 2.5204,
      "step": 100
    },
    {
      "epoch": 9.096969696969698,
      "eval_loss": 2.5186212062835693,
      "eval_runtime": 2.1549,
      "eval_samples_per_second": 25.523,
      "eval_steps_per_second": 3.248,
      "step": 100
    },
    {
      "epoch": 9.193939393939393,
      "grad_norm": 0.26334884762763977,
      "learning_rate": 4.8990000000000004e-05,
      "loss": 2.5295,
      "step": 101
    },
    {
      "epoch": 9.290909090909091,
      "grad_norm": 0.25765350461006165,
      "learning_rate": 4.898e-05,
      "loss": 2.5146,
      "step": 102
    },
    {
      "epoch": 9.387878787878789,
      "grad_norm": 0.2612263560295105,
      "learning_rate": 4.897000000000001e-05,
      "loss": 2.512,
      "step": 103
    },
    {
      "epoch": 9.484848484848484,
      "grad_norm": 0.2547607123851776,
      "learning_rate": 4.896e-05,
      "loss": 2.5108,
      "step": 104
    },
    {
      "epoch": 9.581818181818182,
      "grad_norm": 0.2574792504310608,
      "learning_rate": 4.8950000000000004e-05,
      "loss": 2.4964,
      "step": 105
    },
    {
      "epoch": 9.67878787878788,
      "grad_norm": 0.25030583143234253,
      "learning_rate": 4.894e-05,
      "loss": 2.488,
      "step": 106
    },
    {
      "epoch": 9.775757575757575,
      "grad_norm": 0.2604466676712036,
      "learning_rate": 4.893e-05,
      "loss": 2.4789,
      "step": 107
    },
    {
      "epoch": 9.872727272727273,
      "grad_norm": 0.25026217103004456,
      "learning_rate": 4.8920000000000006e-05,
      "loss": 2.5173,
      "step": 108
    },
    {
      "epoch": 9.969696969696969,
      "grad_norm": 0.26027733087539673,
      "learning_rate": 4.891e-05,
      "loss": 2.4682,
      "step": 109
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.2752644717693329,
      "learning_rate": 4.89e-05,
      "loss": 2.4277,
      "step": 110
    },
    {
      "epoch": 10.0,
      "eval_loss": 2.4674863815307617,
      "eval_runtime": 2.1562,
      "eval_samples_per_second": 25.508,
      "eval_steps_per_second": 3.247,
      "step": 110
    },
    {
      "epoch": 10.096969696969698,
      "grad_norm": 0.2690877318382263,
      "learning_rate": 4.889e-05,
      "loss": 2.5009,
      "step": 111
    },
    {
      "epoch": 10.193939393939393,
      "grad_norm": 0.26510998606681824,
      "learning_rate": 4.8880000000000006e-05,
      "loss": 2.4636,
      "step": 112
    },
    {
      "epoch": 10.290909090909091,
      "grad_norm": 0.26340264081954956,
      "learning_rate": 4.8870000000000005e-05,
      "loss": 2.4617,
      "step": 113
    },
    {
      "epoch": 10.387878787878789,
      "grad_norm": 0.26563766598701477,
      "learning_rate": 4.886e-05,
      "loss": 2.4468,
      "step": 114
    },
    {
      "epoch": 10.484848484848484,
      "grad_norm": 0.2594900131225586,
      "learning_rate": 4.885e-05,
      "loss": 2.4572,
      "step": 115
    },
    {
      "epoch": 10.581818181818182,
      "grad_norm": 0.26996803283691406,
      "learning_rate": 4.884e-05,
      "loss": 2.4254,
      "step": 116
    },
    {
      "epoch": 10.67878787878788,
      "grad_norm": 0.26760613918304443,
      "learning_rate": 4.8830000000000005e-05,
      "loss": 2.4523,
      "step": 117
    },
    {
      "epoch": 10.775757575757575,
      "grad_norm": 0.27425920963287354,
      "learning_rate": 4.8820000000000004e-05,
      "loss": 2.4343,
      "step": 118
    },
    {
      "epoch": 10.872727272727273,
      "grad_norm": 0.27467113733291626,
      "learning_rate": 4.881e-05,
      "loss": 2.4095,
      "step": 119
    },
    {
      "epoch": 10.969696969696969,
      "grad_norm": 0.2764286398887634,
      "learning_rate": 4.88e-05,
      "loss": 2.3839,
      "step": 120
    },
    {
      "epoch": 10.969696969696969,
      "eval_loss": 2.412783145904541,
      "eval_runtime": 2.1539,
      "eval_samples_per_second": 25.535,
      "eval_steps_per_second": 3.25,
      "step": 120
    },
    {
      "epoch": 11.0,
      "grad_norm": 0.27834805846214294,
      "learning_rate": 4.8790000000000006e-05,
      "loss": 2.4266,
      "step": 121
    },
    {
      "epoch": 11.096969696969698,
      "grad_norm": 0.2820325791835785,
      "learning_rate": 4.8780000000000004e-05,
      "loss": 2.4043,
      "step": 122
    },
    {
      "epoch": 11.193939393939393,
      "grad_norm": 0.28312867879867554,
      "learning_rate": 4.877e-05,
      "loss": 2.4158,
      "step": 123
    },
    {
      "epoch": 11.290909090909091,
      "grad_norm": 0.28191468119621277,
      "learning_rate": 4.876e-05,
      "loss": 2.3882,
      "step": 124
    },
    {
      "epoch": 11.387878787878789,
      "grad_norm": 0.2714318037033081,
      "learning_rate": 4.875e-05,
      "loss": 2.3846,
      "step": 125
    },
    {
      "epoch": 11.484848484848484,
      "grad_norm": 0.27267444133758545,
      "learning_rate": 4.8740000000000004e-05,
      "loss": 2.4015,
      "step": 126
    },
    {
      "epoch": 11.581818181818182,
      "grad_norm": 0.28706297278404236,
      "learning_rate": 4.873e-05,
      "loss": 2.3745,
      "step": 127
    },
    {
      "epoch": 11.67878787878788,
      "grad_norm": 0.2960398495197296,
      "learning_rate": 4.872000000000001e-05,
      "loss": 2.3496,
      "step": 128
    },
    {
      "epoch": 11.775757575757575,
      "grad_norm": 0.2834423780441284,
      "learning_rate": 4.871e-05,
      "loss": 2.3626,
      "step": 129
    },
    {
      "epoch": 11.872727272727273,
      "grad_norm": 0.28782230615615845,
      "learning_rate": 4.87e-05,
      "loss": 2.3588,
      "step": 130
    },
    {
      "epoch": 11.872727272727273,
      "eval_loss": 2.3537094593048096,
      "eval_runtime": 2.154,
      "eval_samples_per_second": 25.534,
      "eval_steps_per_second": 3.25,
      "step": 130
    },
    {
      "epoch": 11.969696969696969,
      "grad_norm": 0.2884594202041626,
      "learning_rate": 4.869e-05,
      "loss": 2.3706,
      "step": 131
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.302971750497818,
      "learning_rate": 4.868e-05,
      "loss": 2.3574,
      "step": 132
    },
    {
      "epoch": 12.096969696969698,
      "grad_norm": 0.2976221740245819,
      "learning_rate": 4.867000000000001e-05,
      "loss": 2.342,
      "step": 133
    },
    {
      "epoch": 12.193939393939393,
      "grad_norm": 0.2907634973526001,
      "learning_rate": 4.866e-05,
      "loss": 2.3474,
      "step": 134
    },
    {
      "epoch": 12.290909090909091,
      "grad_norm": 0.3033353388309479,
      "learning_rate": 4.8650000000000003e-05,
      "loss": 2.3137,
      "step": 135
    },
    {
      "epoch": 12.387878787878789,
      "grad_norm": 0.3008555769920349,
      "learning_rate": 4.864e-05,
      "loss": 2.3045,
      "step": 136
    },
    {
      "epoch": 12.484848484848484,
      "grad_norm": 0.3030095398426056,
      "learning_rate": 4.863e-05,
      "loss": 2.325,
      "step": 137
    },
    {
      "epoch": 12.581818181818182,
      "grad_norm": 0.302406907081604,
      "learning_rate": 4.8620000000000005e-05,
      "loss": 2.3364,
      "step": 138
    },
    {
      "epoch": 12.67878787878788,
      "grad_norm": 0.3113637864589691,
      "learning_rate": 4.861e-05,
      "loss": 2.2955,
      "step": 139
    },
    {
      "epoch": 12.775757575757575,
      "grad_norm": 0.32233142852783203,
      "learning_rate": 4.86e-05,
      "loss": 2.2928,
      "step": 140
    },
    {
      "epoch": 12.775757575757575,
      "eval_loss": 2.288254499435425,
      "eval_runtime": 2.1646,
      "eval_samples_per_second": 25.409,
      "eval_steps_per_second": 3.234,
      "step": 140
    },
    {
      "epoch": 12.872727272727273,
      "grad_norm": 0.3165872395038605,
      "learning_rate": 4.859e-05,
      "loss": 2.2981,
      "step": 141
    },
    {
      "epoch": 12.969696969696969,
      "grad_norm": 0.31966012716293335,
      "learning_rate": 4.8580000000000006e-05,
      "loss": 2.2695,
      "step": 142
    },
    {
      "epoch": 13.0,
      "grad_norm": 0.3387593924999237,
      "learning_rate": 4.8570000000000004e-05,
      "loss": 2.2426,
      "step": 143
    },
    {
      "epoch": 13.096969696969698,
      "grad_norm": 0.3338097929954529,
      "learning_rate": 4.856e-05,
      "loss": 2.2611,
      "step": 144
    },
    {
      "epoch": 13.193939393939393,
      "grad_norm": 0.3384833037853241,
      "learning_rate": 4.855e-05,
      "loss": 2.2649,
      "step": 145
    },
    {
      "epoch": 13.290909090909091,
      "grad_norm": 0.32794836163520813,
      "learning_rate": 4.854e-05,
      "loss": 2.2717,
      "step": 146
    },
    {
      "epoch": 13.387878787878789,
      "grad_norm": 0.33190327882766724,
      "learning_rate": 4.8530000000000005e-05,
      "loss": 2.2387,
      "step": 147
    },
    {
      "epoch": 13.484848484848484,
      "grad_norm": 0.3333049714565277,
      "learning_rate": 4.852e-05,
      "loss": 2.2707,
      "step": 148
    },
    {
      "epoch": 13.581818181818182,
      "grad_norm": 0.3635169267654419,
      "learning_rate": 4.851e-05,
      "loss": 2.1971,
      "step": 149
    },
    {
      "epoch": 13.67878787878788,
      "grad_norm": 0.3630567193031311,
      "learning_rate": 4.85e-05,
      "loss": 2.2174,
      "step": 150
    },
    {
      "epoch": 13.67878787878788,
      "eval_loss": 2.2112526893615723,
      "eval_runtime": 2.1483,
      "eval_samples_per_second": 25.602,
      "eval_steps_per_second": 3.258,
      "step": 150
    },
    {
      "epoch": 13.775757575757575,
      "grad_norm": 0.37875112891197205,
      "learning_rate": 4.8490000000000005e-05,
      "loss": 2.1765,
      "step": 151
    },
    {
      "epoch": 13.872727272727273,
      "grad_norm": 0.3608453869819641,
      "learning_rate": 4.8480000000000003e-05,
      "loss": 2.2112,
      "step": 152
    },
    {
      "epoch": 13.969696969696969,
      "grad_norm": 0.3816240131855011,
      "learning_rate": 4.847e-05,
      "loss": 2.1969,
      "step": 153
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.39412960410118103,
      "learning_rate": 4.846e-05,
      "loss": 2.1902,
      "step": 154
    },
    {
      "epoch": 14.096969696969698,
      "grad_norm": 0.3680770993232727,
      "learning_rate": 4.845e-05,
      "loss": 2.1605,
      "step": 155
    },
    {
      "epoch": 14.193939393939393,
      "grad_norm": 0.348919540643692,
      "learning_rate": 4.8440000000000004e-05,
      "loss": 2.1788,
      "step": 156
    },
    {
      "epoch": 14.290909090909091,
      "grad_norm": 0.3487190306186676,
      "learning_rate": 4.843e-05,
      "loss": 2.1753,
      "step": 157
    },
    {
      "epoch": 14.387878787878789,
      "grad_norm": 0.3936024606227875,
      "learning_rate": 4.842000000000001e-05,
      "loss": 2.1322,
      "step": 158
    },
    {
      "epoch": 14.484848484848484,
      "grad_norm": 0.45259958505630493,
      "learning_rate": 4.841e-05,
      "loss": 2.1273,
      "step": 159
    },
    {
      "epoch": 14.581818181818182,
      "grad_norm": 0.4806135296821594,
      "learning_rate": 4.8400000000000004e-05,
      "loss": 2.1488,
      "step": 160
    },
    {
      "epoch": 14.581818181818182,
      "eval_loss": 2.125792980194092,
      "eval_runtime": 2.1611,
      "eval_samples_per_second": 25.45,
      "eval_steps_per_second": 3.239,
      "step": 160
    },
    {
      "epoch": 14.67878787878788,
      "grad_norm": 0.4466473162174225,
      "learning_rate": 4.839e-05,
      "loss": 2.0891,
      "step": 161
    },
    {
      "epoch": 14.775757575757575,
      "grad_norm": 0.38130316138267517,
      "learning_rate": 4.838e-05,
      "loss": 2.1258,
      "step": 162
    },
    {
      "epoch": 14.872727272727273,
      "grad_norm": 0.36415424942970276,
      "learning_rate": 4.8370000000000006e-05,
      "loss": 2.1096,
      "step": 163
    },
    {
      "epoch": 14.969696969696969,
      "grad_norm": 0.3677339255809784,
      "learning_rate": 4.836e-05,
      "loss": 2.1179,
      "step": 164
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.4070127308368683,
      "learning_rate": 4.835e-05,
      "loss": 2.0842,
      "step": 165
    },
    {
      "epoch": 15.096969696969698,
      "grad_norm": 0.41008260846138,
      "learning_rate": 4.834e-05,
      "loss": 2.1046,
      "step": 166
    },
    {
      "epoch": 15.193939393939393,
      "grad_norm": 0.4382018744945526,
      "learning_rate": 4.833e-05,
      "loss": 2.0424,
      "step": 167
    },
    {
      "epoch": 15.290909090909091,
      "grad_norm": 0.4082714319229126,
      "learning_rate": 4.8320000000000005e-05,
      "loss": 2.0994,
      "step": 168
    },
    {
      "epoch": 15.387878787878789,
      "grad_norm": 0.3896704912185669,
      "learning_rate": 4.8309999999999997e-05,
      "loss": 2.0825,
      "step": 169
    },
    {
      "epoch": 15.484848484848484,
      "grad_norm": 0.38054773211479187,
      "learning_rate": 4.83e-05,
      "loss": 2.0364,
      "step": 170
    },
    {
      "epoch": 15.484848484848484,
      "eval_loss": 2.0379815101623535,
      "eval_runtime": 2.1425,
      "eval_samples_per_second": 25.67,
      "eval_steps_per_second": 3.267,
      "step": 170
    },
    {
      "epoch": 15.581818181818182,
      "grad_norm": 0.3640633225440979,
      "learning_rate": 4.829e-05,
      "loss": 2.0312,
      "step": 171
    },
    {
      "epoch": 15.67878787878788,
      "grad_norm": 0.35890400409698486,
      "learning_rate": 4.8280000000000005e-05,
      "loss": 2.0358,
      "step": 172
    },
    {
      "epoch": 15.775757575757575,
      "grad_norm": 0.41220933198928833,
      "learning_rate": 4.8270000000000004e-05,
      "loss": 2.0001,
      "step": 173
    },
    {
      "epoch": 15.872727272727273,
      "grad_norm": 0.4162037968635559,
      "learning_rate": 4.826e-05,
      "loss": 2.0126,
      "step": 174
    },
    {
      "epoch": 15.969696969696969,
      "grad_norm": 0.391644150018692,
      "learning_rate": 4.825e-05,
      "loss": 1.9568,
      "step": 175
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.3766094744205475,
      "learning_rate": 4.824e-05,
      "loss": 2.0131,
      "step": 176
    },
    {
      "epoch": 16.096969696969698,
      "grad_norm": 0.3483814597129822,
      "learning_rate": 4.8230000000000004e-05,
      "loss": 1.9951,
      "step": 177
    },
    {
      "epoch": 16.193939393939395,
      "grad_norm": 0.39426183700561523,
      "learning_rate": 4.822e-05,
      "loss": 1.9671,
      "step": 178
    },
    {
      "epoch": 16.29090909090909,
      "grad_norm": 0.4302196800708771,
      "learning_rate": 4.821e-05,
      "loss": 1.933,
      "step": 179
    },
    {
      "epoch": 16.387878787878787,
      "grad_norm": 0.35731634497642517,
      "learning_rate": 4.82e-05,
      "loss": 2.016,
      "step": 180
    },
    {
      "epoch": 16.387878787878787,
      "eval_loss": 1.9507142305374146,
      "eval_runtime": 2.1627,
      "eval_samples_per_second": 25.432,
      "eval_steps_per_second": 3.237,
      "step": 180
    },
    {
      "epoch": 16.484848484848484,
      "grad_norm": 0.40772905945777893,
      "learning_rate": 4.8190000000000004e-05,
      "loss": 1.9514,
      "step": 181
    },
    {
      "epoch": 16.581818181818182,
      "grad_norm": 0.37485066056251526,
      "learning_rate": 4.818e-05,
      "loss": 1.9189,
      "step": 182
    },
    {
      "epoch": 16.67878787878788,
      "grad_norm": 0.36008599400520325,
      "learning_rate": 4.817e-05,
      "loss": 1.933,
      "step": 183
    },
    {
      "epoch": 16.775757575757577,
      "grad_norm": 0.376492977142334,
      "learning_rate": 4.816e-05,
      "loss": 1.9066,
      "step": 184
    },
    {
      "epoch": 16.87272727272727,
      "grad_norm": 0.41807034611701965,
      "learning_rate": 4.815e-05,
      "loss": 1.9273,
      "step": 185
    },
    {
      "epoch": 16.96969696969697,
      "grad_norm": 0.3658643066883087,
      "learning_rate": 4.814e-05,
      "loss": 1.8925,
      "step": 186
    },
    {
      "epoch": 17.0,
      "grad_norm": 0.36926499009132385,
      "learning_rate": 4.813e-05,
      "loss": 1.8963,
      "step": 187
    },
    {
      "epoch": 17.096969696969698,
      "grad_norm": 0.34902480244636536,
      "learning_rate": 4.812000000000001e-05,
      "loss": 1.9145,
      "step": 188
    },
    {
      "epoch": 17.193939393939395,
      "grad_norm": 0.41031065583229065,
      "learning_rate": 4.8110000000000005e-05,
      "loss": 1.8789,
      "step": 189
    },
    {
      "epoch": 17.29090909090909,
      "grad_norm": 0.3900549113750458,
      "learning_rate": 4.8100000000000004e-05,
      "loss": 1.8519,
      "step": 190
    },
    {
      "epoch": 17.29090909090909,
      "eval_loss": 1.863952398300171,
      "eval_runtime": 2.1371,
      "eval_samples_per_second": 25.736,
      "eval_steps_per_second": 3.275,
      "step": 190
    },
    {
      "epoch": 17.387878787878787,
      "grad_norm": 0.38358181715011597,
      "learning_rate": 4.809e-05,
      "loss": 1.8534,
      "step": 191
    },
    {
      "epoch": 17.484848484848484,
      "grad_norm": 0.3578030467033386,
      "learning_rate": 4.808e-05,
      "loss": 1.8474,
      "step": 192
    },
    {
      "epoch": 17.581818181818182,
      "grad_norm": 0.39505457878112793,
      "learning_rate": 4.8070000000000006e-05,
      "loss": 1.8345,
      "step": 193
    },
    {
      "epoch": 17.67878787878788,
      "grad_norm": 0.41133230924606323,
      "learning_rate": 4.8060000000000004e-05,
      "loss": 1.8128,
      "step": 194
    },
    {
      "epoch": 17.775757575757577,
      "grad_norm": 0.3547358214855194,
      "learning_rate": 4.805e-05,
      "loss": 1.8453,
      "step": 195
    },
    {
      "epoch": 17.87272727272727,
      "grad_norm": 0.3567870855331421,
      "learning_rate": 4.804e-05,
      "loss": 1.8447,
      "step": 196
    },
    {
      "epoch": 17.96969696969697,
      "grad_norm": 0.4133602976799011,
      "learning_rate": 4.8030000000000006e-05,
      "loss": 1.7853,
      "step": 197
    },
    {
      "epoch": 18.0,
      "grad_norm": 0.35908421874046326,
      "learning_rate": 4.8020000000000004e-05,
      "loss": 1.8628,
      "step": 198
    },
    {
      "epoch": 18.096969696969698,
      "grad_norm": 0.35534003376960754,
      "learning_rate": 4.801e-05,
      "loss": 1.8196,
      "step": 199
    },
    {
      "epoch": 18.193939393939395,
      "grad_norm": 0.34279873967170715,
      "learning_rate": 4.8e-05,
      "loss": 1.7631,
      "step": 200
    },
    {
      "epoch": 18.193939393939395,
      "eval_loss": 1.7806957960128784,
      "eval_runtime": 2.1552,
      "eval_samples_per_second": 25.52,
      "eval_steps_per_second": 3.248,
      "step": 200
    },
    {
      "epoch": 18.29090909090909,
      "grad_norm": 0.32960304617881775,
      "learning_rate": 4.799e-05,
      "loss": 1.8037,
      "step": 201
    },
    {
      "epoch": 18.387878787878787,
      "grad_norm": 0.3589719831943512,
      "learning_rate": 4.7980000000000005e-05,
      "loss": 1.7687,
      "step": 202
    },
    {
      "epoch": 18.484848484848484,
      "grad_norm": 0.34671100974082947,
      "learning_rate": 4.797e-05,
      "loss": 1.759,
      "step": 203
    },
    {
      "epoch": 18.581818181818182,
      "grad_norm": 0.3515738546848297,
      "learning_rate": 4.796e-05,
      "loss": 1.6964,
      "step": 204
    },
    {
      "epoch": 18.67878787878788,
      "grad_norm": 0.3296724259853363,
      "learning_rate": 4.795e-05,
      "loss": 1.7632,
      "step": 205
    },
    {
      "epoch": 18.775757575757577,
      "grad_norm": 0.3740043640136719,
      "learning_rate": 4.794e-05,
      "loss": 1.7256,
      "step": 206
    },
    {
      "epoch": 18.87272727272727,
      "grad_norm": 0.3721851110458374,
      "learning_rate": 4.7930000000000004e-05,
      "loss": 1.76,
      "step": 207
    },
    {
      "epoch": 18.96969696969697,
      "grad_norm": 0.3484649360179901,
      "learning_rate": 4.792e-05,
      "loss": 1.7084,
      "step": 208
    },
    {
      "epoch": 19.0,
      "grad_norm": 0.3665394186973572,
      "learning_rate": 4.791000000000001e-05,
      "loss": 1.7428,
      "step": 209
    },
    {
      "epoch": 19.096969696969698,
      "grad_norm": 0.344052791595459,
      "learning_rate": 4.79e-05,
      "loss": 1.7106,
      "step": 210
    },
    {
      "epoch": 19.096969696969698,
      "eval_loss": 1.7033495903015137,
      "eval_runtime": 2.1657,
      "eval_samples_per_second": 25.396,
      "eval_steps_per_second": 3.232,
      "step": 210
    },
    {
      "epoch": 19.193939393939395,
      "grad_norm": 0.31811317801475525,
      "learning_rate": 4.7890000000000004e-05,
      "loss": 1.7242,
      "step": 211
    },
    {
      "epoch": 19.29090909090909,
      "grad_norm": 0.3290516436100006,
      "learning_rate": 4.788e-05,
      "loss": 1.6554,
      "step": 212
    },
    {
      "epoch": 19.387878787878787,
      "grad_norm": 0.3231726288795471,
      "learning_rate": 4.787e-05,
      "loss": 1.6793,
      "step": 213
    },
    {
      "epoch": 19.484848484848484,
      "grad_norm": 0.3180878460407257,
      "learning_rate": 4.7860000000000006e-05,
      "loss": 1.6603,
      "step": 214
    },
    {
      "epoch": 19.581818181818182,
      "grad_norm": 0.3123716711997986,
      "learning_rate": 4.785e-05,
      "loss": 1.693,
      "step": 215
    },
    {
      "epoch": 19.67878787878788,
      "grad_norm": 0.3156301975250244,
      "learning_rate": 4.784e-05,
      "loss": 1.6663,
      "step": 216
    },
    {
      "epoch": 19.775757575757577,
      "grad_norm": 0.3017955720424652,
      "learning_rate": 4.783e-05,
      "loss": 1.6787,
      "step": 217
    },
    {
      "epoch": 19.87272727272727,
      "grad_norm": 0.3095964193344116,
      "learning_rate": 4.7820000000000006e-05,
      "loss": 1.6693,
      "step": 218
    },
    {
      "epoch": 19.96969696969697,
      "grad_norm": 0.29991716146469116,
      "learning_rate": 4.7810000000000005e-05,
      "loss": 1.629,
      "step": 219
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.32141411304473877,
      "learning_rate": 4.78e-05,
      "loss": 1.6009,
      "step": 220
    },
    {
      "epoch": 20.0,
      "eval_loss": 1.635511040687561,
      "eval_runtime": 2.1483,
      "eval_samples_per_second": 25.602,
      "eval_steps_per_second": 3.258,
      "step": 220
    },
    {
      "epoch": 20.096969696969698,
      "grad_norm": 0.30841124057769775,
      "learning_rate": 4.779e-05,
      "loss": 1.6056,
      "step": 221
    },
    {
      "epoch": 20.193939393939395,
      "grad_norm": 0.2942273020744324,
      "learning_rate": 4.778e-05,
      "loss": 1.6703,
      "step": 222
    },
    {
      "epoch": 20.29090909090909,
      "grad_norm": 0.2849220931529999,
      "learning_rate": 4.7770000000000005e-05,
      "loss": 1.6272,
      "step": 223
    },
    {
      "epoch": 20.387878787878787,
      "grad_norm": 0.2896654009819031,
      "learning_rate": 4.7760000000000004e-05,
      "loss": 1.5896,
      "step": 224
    },
    {
      "epoch": 20.484848484848484,
      "grad_norm": 0.29047784209251404,
      "learning_rate": 4.775e-05,
      "loss": 1.6155,
      "step": 225
    },
    {
      "epoch": 20.581818181818182,
      "grad_norm": 0.2764190137386322,
      "learning_rate": 4.774e-05,
      "loss": 1.6085,
      "step": 226
    },
    {
      "epoch": 20.67878787878788,
      "grad_norm": 0.28020909428596497,
      "learning_rate": 4.7730000000000005e-05,
      "loss": 1.5519,
      "step": 227
    },
    {
      "epoch": 20.775757575757577,
      "grad_norm": 0.27882450819015503,
      "learning_rate": 4.7720000000000004e-05,
      "loss": 1.615,
      "step": 228
    },
    {
      "epoch": 20.87272727272727,
      "grad_norm": 0.2711680233478546,
      "learning_rate": 4.771e-05,
      "loss": 1.6046,
      "step": 229
    },
    {
      "epoch": 20.96969696969697,
      "grad_norm": 0.28352415561676025,
      "learning_rate": 4.77e-05,
      "loss": 1.5786,
      "step": 230
    },
    {
      "epoch": 20.96969696969697,
      "eval_loss": 1.5802624225616455,
      "eval_runtime": 2.1482,
      "eval_samples_per_second": 25.603,
      "eval_steps_per_second": 3.259,
      "step": 230
    },
    {
      "epoch": 21.0,
      "grad_norm": 0.3100228011608124,
      "learning_rate": 4.769e-05,
      "loss": 1.5409,
      "step": 231
    },
    {
      "epoch": 21.096969696969698,
      "grad_norm": 0.2626677453517914,
      "learning_rate": 4.7680000000000004e-05,
      "loss": 1.5711,
      "step": 232
    },
    {
      "epoch": 21.193939393939395,
      "grad_norm": 0.257974237203598,
      "learning_rate": 4.767e-05,
      "loss": 1.5816,
      "step": 233
    },
    {
      "epoch": 21.29090909090909,
      "grad_norm": 0.27121230959892273,
      "learning_rate": 4.766000000000001e-05,
      "loss": 1.5381,
      "step": 234
    },
    {
      "epoch": 21.387878787878787,
      "grad_norm": 0.2541174292564392,
      "learning_rate": 4.765e-05,
      "loss": 1.5695,
      "step": 235
    },
    {
      "epoch": 21.484848484848484,
      "grad_norm": 0.25187304615974426,
      "learning_rate": 4.7640000000000005e-05,
      "loss": 1.5354,
      "step": 236
    },
    {
      "epoch": 21.581818181818182,
      "grad_norm": 0.2590962052345276,
      "learning_rate": 4.763e-05,
      "loss": 1.5459,
      "step": 237
    },
    {
      "epoch": 21.67878787878788,
      "grad_norm": 0.2600695788860321,
      "learning_rate": 4.762e-05,
      "loss": 1.5432,
      "step": 238
    },
    {
      "epoch": 21.775757575757577,
      "grad_norm": 0.2626539468765259,
      "learning_rate": 4.761000000000001e-05,
      "loss": 1.5048,
      "step": 239
    },
    {
      "epoch": 21.87272727272727,
      "grad_norm": 0.2486988604068756,
      "learning_rate": 4.76e-05,
      "loss": 1.5393,
      "step": 240
    },
    {
      "epoch": 21.87272727272727,
      "eval_loss": 1.534799337387085,
      "eval_runtime": 2.1443,
      "eval_samples_per_second": 25.649,
      "eval_steps_per_second": 3.264,
      "step": 240
    },
    {
      "epoch": 21.96969696969697,
      "grad_norm": 0.2648991644382477,
      "learning_rate": 4.7590000000000003e-05,
      "loss": 1.5515,
      "step": 241
    },
    {
      "epoch": 22.0,
      "grad_norm": 0.28589197993278503,
      "learning_rate": 4.758e-05,
      "loss": 1.5614,
      "step": 242
    },
    {
      "epoch": 22.096969696969698,
      "grad_norm": 0.2596450746059418,
      "learning_rate": 4.757e-05,
      "loss": 1.5398,
      "step": 243
    },
    {
      "epoch": 22.193939393939395,
      "grad_norm": 0.24288393557071686,
      "learning_rate": 4.7560000000000005e-05,
      "loss": 1.4971,
      "step": 244
    },
    {
      "epoch": 22.29090909090909,
      "grad_norm": 0.2506029009819031,
      "learning_rate": 4.755e-05,
      "loss": 1.4791,
      "step": 245
    },
    {
      "epoch": 22.387878787878787,
      "grad_norm": 0.23702731728553772,
      "learning_rate": 4.754e-05,
      "loss": 1.558,
      "step": 246
    },
    {
      "epoch": 22.484848484848484,
      "grad_norm": 0.2387530505657196,
      "learning_rate": 4.753e-05,
      "loss": 1.4622,
      "step": 247
    },
    {
      "epoch": 22.581818181818182,
      "grad_norm": 0.23143506050109863,
      "learning_rate": 4.7520000000000006e-05,
      "loss": 1.5016,
      "step": 248
    },
    {
      "epoch": 22.67878787878788,
      "grad_norm": 0.24316754937171936,
      "learning_rate": 4.7510000000000004e-05,
      "loss": 1.5126,
      "step": 249
    },
    {
      "epoch": 22.775757575757577,
      "grad_norm": 0.22900234162807465,
      "learning_rate": 4.75e-05,
      "loss": 1.477,
      "step": 250
    },
    {
      "epoch": 22.775757575757577,
      "eval_loss": 1.4960139989852905,
      "eval_runtime": 2.1603,
      "eval_samples_per_second": 25.459,
      "eval_steps_per_second": 3.24,
      "step": 250
    },
    {
      "epoch": 22.87272727272727,
      "grad_norm": 0.22426266968250275,
      "learning_rate": 4.749e-05,
      "loss": 1.5176,
      "step": 251
    },
    {
      "epoch": 22.96969696969697,
      "grad_norm": 0.21992135047912598,
      "learning_rate": 4.748e-05,
      "loss": 1.4845,
      "step": 252
    },
    {
      "epoch": 23.0,
      "grad_norm": 0.27798858284950256,
      "learning_rate": 4.7470000000000005e-05,
      "loss": 1.456,
      "step": 253
    },
    {
      "epoch": 23.096969696969698,
      "grad_norm": 0.23832809925079346,
      "learning_rate": 4.746e-05,
      "loss": 1.4871,
      "step": 254
    },
    {
      "epoch": 23.193939393939395,
      "grad_norm": 0.2412337064743042,
      "learning_rate": 4.745e-05,
      "loss": 1.4993,
      "step": 255
    },
    {
      "epoch": 23.29090909090909,
      "grad_norm": 0.22906090319156647,
      "learning_rate": 4.744e-05,
      "loss": 1.4443,
      "step": 256
    },
    {
      "epoch": 23.387878787878787,
      "grad_norm": 0.22417926788330078,
      "learning_rate": 4.7430000000000005e-05,
      "loss": 1.5002,
      "step": 257
    },
    {
      "epoch": 23.484848484848484,
      "grad_norm": 0.22132626175880432,
      "learning_rate": 4.742e-05,
      "loss": 1.4494,
      "step": 258
    },
    {
      "epoch": 23.581818181818182,
      "grad_norm": 0.2173018604516983,
      "learning_rate": 4.741e-05,
      "loss": 1.4401,
      "step": 259
    },
    {
      "epoch": 23.67878787878788,
      "grad_norm": 0.23397614061832428,
      "learning_rate": 4.74e-05,
      "loss": 1.4618,
      "step": 260
    },
    {
      "epoch": 23.67878787878788,
      "eval_loss": 1.462380051612854,
      "eval_runtime": 2.1544,
      "eval_samples_per_second": 25.529,
      "eval_steps_per_second": 3.249,
      "step": 260
    },
    {
      "epoch": 23.775757575757577,
      "grad_norm": 0.219797283411026,
      "learning_rate": 4.739e-05,
      "loss": 1.4642,
      "step": 261
    },
    {
      "epoch": 23.87272727272727,
      "grad_norm": 0.23293839395046234,
      "learning_rate": 4.7380000000000004e-05,
      "loss": 1.4228,
      "step": 262
    },
    {
      "epoch": 23.96969696969697,
      "grad_norm": 0.22269245982170105,
      "learning_rate": 4.737e-05,
      "loss": 1.4503,
      "step": 263
    },
    {
      "epoch": 24.0,
      "grad_norm": 0.2525419592857361,
      "learning_rate": 4.736000000000001e-05,
      "loss": 1.4629,
      "step": 264
    },
    {
      "epoch": 24.096969696969698,
      "grad_norm": 0.2224574089050293,
      "learning_rate": 4.735e-05,
      "loss": 1.4174,
      "step": 265
    },
    {
      "epoch": 24.193939393939395,
      "grad_norm": 0.21884828805923462,
      "learning_rate": 4.7340000000000004e-05,
      "loss": 1.436,
      "step": 266
    },
    {
      "epoch": 24.29090909090909,
      "grad_norm": 0.21463195979595184,
      "learning_rate": 4.733e-05,
      "loss": 1.431,
      "step": 267
    },
    {
      "epoch": 24.387878787878787,
      "grad_norm": 0.20648930966854095,
      "learning_rate": 4.732e-05,
      "loss": 1.4178,
      "step": 268
    },
    {
      "epoch": 24.484848484848484,
      "grad_norm": 0.2181093543767929,
      "learning_rate": 4.7310000000000006e-05,
      "loss": 1.4447,
      "step": 269
    },
    {
      "epoch": 24.581818181818182,
      "grad_norm": 0.205455020070076,
      "learning_rate": 4.73e-05,
      "loss": 1.4116,
      "step": 270
    },
    {
      "epoch": 24.581818181818182,
      "eval_loss": 1.432446002960205,
      "eval_runtime": 2.1882,
      "eval_samples_per_second": 25.135,
      "eval_steps_per_second": 3.199,
      "step": 270
    },
    {
      "epoch": 24.67878787878788,
      "grad_norm": 0.21049101650714874,
      "learning_rate": 4.729e-05,
      "loss": 1.4514,
      "step": 271
    },
    {
      "epoch": 24.775757575757577,
      "grad_norm": 0.2380298376083374,
      "learning_rate": 4.728e-05,
      "loss": 1.4535,
      "step": 272
    },
    {
      "epoch": 24.87272727272727,
      "grad_norm": 0.20930226147174835,
      "learning_rate": 4.7270000000000007e-05,
      "loss": 1.4293,
      "step": 273
    },
    {
      "epoch": 24.96969696969697,
      "grad_norm": 0.24193055927753448,
      "learning_rate": 4.7260000000000005e-05,
      "loss": 1.3748,
      "step": 274
    },
    {
      "epoch": 25.0,
      "grad_norm": 0.27937281131744385,
      "learning_rate": 4.7249999999999997e-05,
      "loss": 1.4503,
      "step": 275
    },
    {
      "epoch": 25.096969696969698,
      "grad_norm": 0.20097865164279938,
      "learning_rate": 4.724e-05,
      "loss": 1.3747,
      "step": 276
    },
    {
      "epoch": 25.193939393939395,
      "grad_norm": 0.20190516114234924,
      "learning_rate": 4.723e-05,
      "loss": 1.406,
      "step": 277
    },
    {
      "epoch": 25.29090909090909,
      "grad_norm": 0.2020610272884369,
      "learning_rate": 4.7220000000000005e-05,
      "loss": 1.4089,
      "step": 278
    },
    {
      "epoch": 25.387878787878787,
      "grad_norm": 0.21105410158634186,
      "learning_rate": 4.7210000000000004e-05,
      "loss": 1.3805,
      "step": 279
    },
    {
      "epoch": 25.484848484848484,
      "grad_norm": 0.22814784944057465,
      "learning_rate": 4.72e-05,
      "loss": 1.4185,
      "step": 280
    },
    {
      "epoch": 25.484848484848484,
      "eval_loss": 1.405413269996643,
      "eval_runtime": 2.1372,
      "eval_samples_per_second": 25.735,
      "eval_steps_per_second": 3.275,
      "step": 280
    },
    {
      "epoch": 25.581818181818182,
      "grad_norm": 0.21340583264827728,
      "learning_rate": 4.719e-05,
      "loss": 1.4186,
      "step": 281
    },
    {
      "epoch": 25.67878787878788,
      "grad_norm": 0.19968931376934052,
      "learning_rate": 4.718e-05,
      "loss": 1.4559,
      "step": 282
    },
    {
      "epoch": 25.775757575757577,
      "grad_norm": 0.23844146728515625,
      "learning_rate": 4.7170000000000004e-05,
      "loss": 1.3384,
      "step": 283
    },
    {
      "epoch": 25.87272727272727,
      "grad_norm": 0.21829138696193695,
      "learning_rate": 4.716e-05,
      "loss": 1.4191,
      "step": 284
    },
    {
      "epoch": 25.96969696969697,
      "grad_norm": 0.2160981446504593,
      "learning_rate": 4.715e-05,
      "loss": 1.378,
      "step": 285
    },
    {
      "epoch": 26.0,
      "grad_norm": 0.26061224937438965,
      "learning_rate": 4.714e-05,
      "loss": 1.2759,
      "step": 286
    },
    {
      "epoch": 26.096969696969698,
      "grad_norm": 0.20122674107551575,
      "learning_rate": 4.7130000000000004e-05,
      "loss": 1.3796,
      "step": 287
    },
    {
      "epoch": 26.193939393939395,
      "grad_norm": 0.19445091485977173,
      "learning_rate": 4.712e-05,
      "loss": 1.3533,
      "step": 288
    },
    {
      "epoch": 26.29090909090909,
      "grad_norm": 0.224958136677742,
      "learning_rate": 4.711e-05,
      "loss": 1.4088,
      "step": 289
    },
    {
      "epoch": 26.387878787878787,
      "grad_norm": 0.20992790162563324,
      "learning_rate": 4.71e-05,
      "loss": 1.3567,
      "step": 290
    },
    {
      "epoch": 26.387878787878787,
      "eval_loss": 1.379786491394043,
      "eval_runtime": 2.159,
      "eval_samples_per_second": 25.475,
      "eval_steps_per_second": 3.242,
      "step": 290
    },
    {
      "epoch": 26.484848484848484,
      "grad_norm": 0.20614048838615417,
      "learning_rate": 4.709e-05,
      "loss": 1.3977,
      "step": 291
    },
    {
      "epoch": 26.581818181818182,
      "grad_norm": 0.19888976216316223,
      "learning_rate": 4.708e-05,
      "loss": 1.3461,
      "step": 292
    },
    {
      "epoch": 26.67878787878788,
      "grad_norm": 0.19490261375904083,
      "learning_rate": 4.707e-05,
      "loss": 1.3935,
      "step": 293
    },
    {
      "epoch": 26.775757575757577,
      "grad_norm": 0.19813090562820435,
      "learning_rate": 4.706000000000001e-05,
      "loss": 1.3782,
      "step": 294
    },
    {
      "epoch": 26.87272727272727,
      "grad_norm": 0.23433801531791687,
      "learning_rate": 4.705e-05,
      "loss": 1.3338,
      "step": 295
    },
    {
      "epoch": 26.96969696969697,
      "grad_norm": 0.23675957322120667,
      "learning_rate": 4.7040000000000004e-05,
      "loss": 1.3322,
      "step": 296
    },
    {
      "epoch": 27.0,
      "grad_norm": 0.27610862255096436,
      "learning_rate": 4.703e-05,
      "loss": 1.352,
      "step": 297
    },
    {
      "epoch": 27.096969696969698,
      "grad_norm": 0.2032589614391327,
      "learning_rate": 4.702e-05,
      "loss": 1.3404,
      "step": 298
    },
    {
      "epoch": 27.193939393939395,
      "grad_norm": 0.19773827493190765,
      "learning_rate": 4.7010000000000006e-05,
      "loss": 1.3318,
      "step": 299
    },
    {
      "epoch": 27.29090909090909,
      "grad_norm": 0.26485851407051086,
      "learning_rate": 4.7e-05,
      "loss": 1.3848,
      "step": 300
    },
    {
      "epoch": 27.29090909090909,
      "eval_loss": 1.3541288375854492,
      "eval_runtime": 2.1296,
      "eval_samples_per_second": 25.827,
      "eval_steps_per_second": 3.287,
      "step": 300
    },
    {
      "epoch": 27.387878787878787,
      "grad_norm": 0.21880951523780823,
      "learning_rate": 4.699e-05,
      "loss": 1.3282,
      "step": 301
    },
    {
      "epoch": 27.484848484848484,
      "grad_norm": 0.20713409781455994,
      "learning_rate": 4.698e-05,
      "loss": 1.3377,
      "step": 302
    },
    {
      "epoch": 27.581818181818182,
      "grad_norm": 0.20707346498966217,
      "learning_rate": 4.6970000000000006e-05,
      "loss": 1.3616,
      "step": 303
    },
    {
      "epoch": 27.67878787878788,
      "grad_norm": 0.20824404060840607,
      "learning_rate": 4.6960000000000004e-05,
      "loss": 1.347,
      "step": 304
    },
    {
      "epoch": 27.775757575757577,
      "grad_norm": 0.2190723717212677,
      "learning_rate": 4.695e-05,
      "loss": 1.334,
      "step": 305
    },
    {
      "epoch": 27.87272727272727,
      "grad_norm": 0.23655390739440918,
      "learning_rate": 4.694e-05,
      "loss": 1.3241,
      "step": 306
    },
    {
      "epoch": 27.96969696969697,
      "grad_norm": 0.2129470258951187,
      "learning_rate": 4.693e-05,
      "loss": 1.2995,
      "step": 307
    },
    {
      "epoch": 28.0,
      "grad_norm": 0.2688191533088684,
      "learning_rate": 4.6920000000000005e-05,
      "loss": 1.3388,
      "step": 308
    },
    {
      "epoch": 28.096969696969698,
      "grad_norm": 0.2122860997915268,
      "learning_rate": 4.691e-05,
      "loss": 1.3415,
      "step": 309
    },
    {
      "epoch": 28.193939393939395,
      "grad_norm": 0.21898894011974335,
      "learning_rate": 4.69e-05,
      "loss": 1.3114,
      "step": 310
    },
    {
      "epoch": 28.193939393939395,
      "eval_loss": 1.328099250793457,
      "eval_runtime": 2.136,
      "eval_samples_per_second": 25.749,
      "eval_steps_per_second": 3.277,
      "step": 310
    },
    {
      "epoch": 28.29090909090909,
      "grad_norm": 0.21611572802066803,
      "learning_rate": 4.689e-05,
      "loss": 1.3003,
      "step": 311
    },
    {
      "epoch": 28.387878787878787,
      "grad_norm": 0.2057645171880722,
      "learning_rate": 4.688e-05,
      "loss": 1.2691,
      "step": 312
    },
    {
      "epoch": 28.484848484848484,
      "grad_norm": 0.20616444945335388,
      "learning_rate": 4.6870000000000004e-05,
      "loss": 1.2799,
      "step": 313
    },
    {
      "epoch": 28.581818181818182,
      "grad_norm": 0.21262364089488983,
      "learning_rate": 4.686e-05,
      "loss": 1.3319,
      "step": 314
    },
    {
      "epoch": 28.67878787878788,
      "grad_norm": 0.2020464539527893,
      "learning_rate": 4.685000000000001e-05,
      "loss": 1.3029,
      "step": 315
    },
    {
      "epoch": 28.775757575757577,
      "grad_norm": 0.21590180695056915,
      "learning_rate": 4.684e-05,
      "loss": 1.3115,
      "step": 316
    },
    {
      "epoch": 28.87272727272727,
      "grad_norm": 0.2041538804769516,
      "learning_rate": 4.6830000000000004e-05,
      "loss": 1.3516,
      "step": 317
    },
    {
      "epoch": 28.96969696969697,
      "grad_norm": 0.2074449360370636,
      "learning_rate": 4.682e-05,
      "loss": 1.2955,
      "step": 318
    },
    {
      "epoch": 29.0,
      "grad_norm": 0.3378038704395294,
      "learning_rate": 4.681e-05,
      "loss": 1.3059,
      "step": 319
    },
    {
      "epoch": 29.096969696969698,
      "grad_norm": 0.21807034313678741,
      "learning_rate": 4.6800000000000006e-05,
      "loss": 1.2858,
      "step": 320
    },
    {
      "epoch": 29.096969696969698,
      "eval_loss": 1.3011435270309448,
      "eval_runtime": 2.161,
      "eval_samples_per_second": 25.451,
      "eval_steps_per_second": 3.239,
      "step": 320
    },
    {
      "epoch": 29.193939393939395,
      "grad_norm": 0.2208559513092041,
      "learning_rate": 4.679e-05,
      "loss": 1.2667,
      "step": 321
    },
    {
      "epoch": 29.29090909090909,
      "grad_norm": 0.20598140358924866,
      "learning_rate": 4.678e-05,
      "loss": 1.3218,
      "step": 322
    },
    {
      "epoch": 29.387878787878787,
      "grad_norm": 0.23158182203769684,
      "learning_rate": 4.677e-05,
      "loss": 1.3501,
      "step": 323
    },
    {
      "epoch": 29.484848484848484,
      "grad_norm": 0.21017974615097046,
      "learning_rate": 4.6760000000000006e-05,
      "loss": 1.2647,
      "step": 324
    },
    {
      "epoch": 29.581818181818182,
      "grad_norm": 0.22472354769706726,
      "learning_rate": 4.6750000000000005e-05,
      "loss": 1.2315,
      "step": 325
    },
    {
      "epoch": 29.67878787878788,
      "grad_norm": 0.22486887872219086,
      "learning_rate": 4.674e-05,
      "loss": 1.2892,
      "step": 326
    },
    {
      "epoch": 29.775757575757577,
      "grad_norm": 0.22963334619998932,
      "learning_rate": 4.673e-05,
      "loss": 1.2826,
      "step": 327
    },
    {
      "epoch": 29.87272727272727,
      "grad_norm": 0.24132588505744934,
      "learning_rate": 4.672e-05,
      "loss": 1.2428,
      "step": 328
    },
    {
      "epoch": 29.96969696969697,
      "grad_norm": 0.2201637327671051,
      "learning_rate": 4.6710000000000005e-05,
      "loss": 1.2668,
      "step": 329
    },
    {
      "epoch": 30.0,
      "grad_norm": 0.29462799429893494,
      "learning_rate": 4.6700000000000003e-05,
      "loss": 1.2142,
      "step": 330
    },
    {
      "epoch": 30.0,
      "eval_loss": 1.2728815078735352,
      "eval_runtime": 2.1763,
      "eval_samples_per_second": 25.272,
      "eval_steps_per_second": 3.216,
      "step": 330
    },
    {
      "epoch": 30.096969696969698,
      "grad_norm": 0.2325703203678131,
      "learning_rate": 4.669e-05,
      "loss": 1.2378,
      "step": 331
    },
    {
      "epoch": 30.193939393939395,
      "grad_norm": 0.2217150777578354,
      "learning_rate": 4.668e-05,
      "loss": 1.274,
      "step": 332
    },
    {
      "epoch": 30.29090909090909,
      "grad_norm": 0.22642824053764343,
      "learning_rate": 4.6670000000000005e-05,
      "loss": 1.2656,
      "step": 333
    },
    {
      "epoch": 30.387878787878787,
      "grad_norm": 0.22929802536964417,
      "learning_rate": 4.6660000000000004e-05,
      "loss": 1.2234,
      "step": 334
    },
    {
      "epoch": 30.484848484848484,
      "grad_norm": 0.21472522616386414,
      "learning_rate": 4.665e-05,
      "loss": 1.2359,
      "step": 335
    },
    {
      "epoch": 30.581818181818182,
      "grad_norm": 0.21877072751522064,
      "learning_rate": 4.664e-05,
      "loss": 1.2275,
      "step": 336
    },
    {
      "epoch": 30.67878787878788,
      "grad_norm": 0.24395674467086792,
      "learning_rate": 4.663e-05,
      "loss": 1.2313,
      "step": 337
    },
    {
      "epoch": 30.775757575757577,
      "grad_norm": 0.2251306176185608,
      "learning_rate": 4.6620000000000004e-05,
      "loss": 1.2631,
      "step": 338
    },
    {
      "epoch": 30.87272727272727,
      "grad_norm": 0.2529394030570984,
      "learning_rate": 4.661e-05,
      "loss": 1.2571,
      "step": 339
    },
    {
      "epoch": 30.96969696969697,
      "grad_norm": 0.24227924644947052,
      "learning_rate": 4.660000000000001e-05,
      "loss": 1.2417,
      "step": 340
    },
    {
      "epoch": 30.96969696969697,
      "eval_loss": 1.2429355382919312,
      "eval_runtime": 2.175,
      "eval_samples_per_second": 25.287,
      "eval_steps_per_second": 3.218,
      "step": 340
    },
    {
      "epoch": 31.0,
      "grad_norm": 0.2832590639591217,
      "learning_rate": 4.659e-05,
      "loss": 1.2431,
      "step": 341
    },
    {
      "epoch": 31.096969696969698,
      "grad_norm": 0.23116494715213776,
      "learning_rate": 4.6580000000000005e-05,
      "loss": 1.2152,
      "step": 342
    },
    {
      "epoch": 31.193939393939395,
      "grad_norm": 0.22689251601696014,
      "learning_rate": 4.657e-05,
      "loss": 1.2302,
      "step": 343
    },
    {
      "epoch": 31.29090909090909,
      "grad_norm": 0.2684100270271301,
      "learning_rate": 4.656e-05,
      "loss": 1.1755,
      "step": 344
    },
    {
      "epoch": 31.387878787878787,
      "grad_norm": 0.2274557501077652,
      "learning_rate": 4.655000000000001e-05,
      "loss": 1.2127,
      "step": 345
    },
    {
      "epoch": 31.484848484848484,
      "grad_norm": 0.2353113293647766,
      "learning_rate": 4.654e-05,
      "loss": 1.2531,
      "step": 346
    },
    {
      "epoch": 31.581818181818182,
      "grad_norm": 0.2506870627403259,
      "learning_rate": 4.6530000000000003e-05,
      "loss": 1.1684,
      "step": 347
    },
    {
      "epoch": 31.67878787878788,
      "grad_norm": 0.25880929827690125,
      "learning_rate": 4.652e-05,
      "loss": 1.2229,
      "step": 348
    },
    {
      "epoch": 31.775757575757577,
      "grad_norm": 0.23619729280471802,
      "learning_rate": 4.651e-05,
      "loss": 1.2227,
      "step": 349
    },
    {
      "epoch": 31.87272727272727,
      "grad_norm": 0.23269791901111603,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 1.2458,
      "step": 350
    },
    {
      "epoch": 31.87272727272727,
      "eval_loss": 1.211027979850769,
      "eval_runtime": 2.153,
      "eval_samples_per_second": 25.545,
      "eval_steps_per_second": 3.251,
      "step": 350
    },
    {
      "epoch": 31.96969696969697,
      "grad_norm": 0.22882823646068573,
      "learning_rate": 4.649e-05,
      "loss": 1.1747,
      "step": 351
    },
    {
      "epoch": 32.0,
      "grad_norm": 0.37140440940856934,
      "learning_rate": 4.648e-05,
      "loss": 1.153,
      "step": 352
    },
    {
      "epoch": 32.096969696969694,
      "grad_norm": 0.24536997079849243,
      "learning_rate": 4.647e-05,
      "loss": 1.1431,
      "step": 353
    },
    {
      "epoch": 32.193939393939395,
      "grad_norm": 0.24964717030525208,
      "learning_rate": 4.6460000000000006e-05,
      "loss": 1.188,
      "step": 354
    },
    {
      "epoch": 32.29090909090909,
      "grad_norm": 0.26268887519836426,
      "learning_rate": 4.6450000000000004e-05,
      "loss": 1.1864,
      "step": 355
    },
    {
      "epoch": 32.38787878787879,
      "grad_norm": 0.24667908251285553,
      "learning_rate": 4.644e-05,
      "loss": 1.1578,
      "step": 356
    },
    {
      "epoch": 32.484848484848484,
      "grad_norm": 0.2375129610300064,
      "learning_rate": 4.643e-05,
      "loss": 1.2097,
      "step": 357
    },
    {
      "epoch": 32.58181818181818,
      "grad_norm": 0.27233636379241943,
      "learning_rate": 4.642e-05,
      "loss": 1.134,
      "step": 358
    },
    {
      "epoch": 32.67878787878788,
      "grad_norm": 0.24921810626983643,
      "learning_rate": 4.6410000000000005e-05,
      "loss": 1.1822,
      "step": 359
    },
    {
      "epoch": 32.775757575757574,
      "grad_norm": 0.24470624327659607,
      "learning_rate": 4.64e-05,
      "loss": 1.1689,
      "step": 360
    },
    {
      "epoch": 32.775757575757574,
      "eval_loss": 1.1777660846710205,
      "eval_runtime": 2.1615,
      "eval_samples_per_second": 25.445,
      "eval_steps_per_second": 3.239,
      "step": 360
    },
    {
      "epoch": 32.872727272727275,
      "grad_norm": 0.23806406557559967,
      "learning_rate": 4.639e-05,
      "loss": 1.1702,
      "step": 361
    },
    {
      "epoch": 32.96969696969697,
      "grad_norm": 0.24547907710075378,
      "learning_rate": 4.638e-05,
      "loss": 1.1973,
      "step": 362
    },
    {
      "epoch": 33.0,
      "grad_norm": 0.36119315028190613,
      "learning_rate": 4.6370000000000005e-05,
      "loss": 1.1615,
      "step": 363
    },
    {
      "epoch": 33.096969696969694,
      "grad_norm": 0.2657982110977173,
      "learning_rate": 4.636e-05,
      "loss": 1.1637,
      "step": 364
    },
    {
      "epoch": 33.193939393939395,
      "grad_norm": 0.2563134729862213,
      "learning_rate": 4.635e-05,
      "loss": 1.1516,
      "step": 365
    },
    {
      "epoch": 33.29090909090909,
      "grad_norm": 0.29967230558395386,
      "learning_rate": 4.634e-05,
      "loss": 1.1549,
      "step": 366
    },
    {
      "epoch": 33.38787878787879,
      "grad_norm": 0.2566681504249573,
      "learning_rate": 4.633e-05,
      "loss": 1.1483,
      "step": 367
    },
    {
      "epoch": 33.484848484848484,
      "grad_norm": 0.24460069835186005,
      "learning_rate": 4.6320000000000004e-05,
      "loss": 1.1149,
      "step": 368
    },
    {
      "epoch": 33.58181818181818,
      "grad_norm": 0.263212114572525,
      "learning_rate": 4.631e-05,
      "loss": 1.1419,
      "step": 369
    },
    {
      "epoch": 33.67878787878788,
      "grad_norm": 0.24054227769374847,
      "learning_rate": 4.630000000000001e-05,
      "loss": 1.1073,
      "step": 370
    },
    {
      "epoch": 33.67878787878788,
      "eval_loss": 1.1430851221084595,
      "eval_runtime": 2.1402,
      "eval_samples_per_second": 25.699,
      "eval_steps_per_second": 3.271,
      "step": 370
    },
    {
      "epoch": 33.775757575757574,
      "grad_norm": 0.26968780159950256,
      "learning_rate": 4.629e-05,
      "loss": 1.1465,
      "step": 371
    },
    {
      "epoch": 33.872727272727275,
      "grad_norm": 0.26295608282089233,
      "learning_rate": 4.6280000000000004e-05,
      "loss": 1.1351,
      "step": 372
    },
    {
      "epoch": 33.96969696969697,
      "grad_norm": 0.2707311511039734,
      "learning_rate": 4.627e-05,
      "loss": 1.0995,
      "step": 373
    },
    {
      "epoch": 34.0,
      "grad_norm": 0.37232154607772827,
      "learning_rate": 4.626e-05,
      "loss": 1.0998,
      "step": 374
    },
    {
      "epoch": 34.096969696969694,
      "grad_norm": 0.26393061876296997,
      "learning_rate": 4.6250000000000006e-05,
      "loss": 1.1008,
      "step": 375
    },
    {
      "epoch": 34.193939393939395,
      "grad_norm": 0.25553223490715027,
      "learning_rate": 4.624e-05,
      "loss": 1.1265,
      "step": 376
    },
    {
      "epoch": 34.29090909090909,
      "grad_norm": 0.2520650625228882,
      "learning_rate": 4.623e-05,
      "loss": 1.0521,
      "step": 377
    },
    {
      "epoch": 34.38787878787879,
      "grad_norm": 0.2646835744380951,
      "learning_rate": 4.622e-05,
      "loss": 1.1236,
      "step": 378
    },
    {
      "epoch": 34.484848484848484,
      "grad_norm": 0.31807929277420044,
      "learning_rate": 4.6210000000000006e-05,
      "loss": 1.1201,
      "step": 379
    },
    {
      "epoch": 34.58181818181818,
      "grad_norm": 0.2697524428367615,
      "learning_rate": 4.6200000000000005e-05,
      "loss": 1.1187,
      "step": 380
    },
    {
      "epoch": 34.58181818181818,
      "eval_loss": 1.1067144870758057,
      "eval_runtime": 2.1541,
      "eval_samples_per_second": 25.533,
      "eval_steps_per_second": 3.25,
      "step": 380
    },
    {
      "epoch": 34.67878787878788,
      "grad_norm": 0.28524914383888245,
      "learning_rate": 4.619e-05,
      "loss": 1.0797,
      "step": 381
    },
    {
      "epoch": 34.775757575757574,
      "grad_norm": 0.26566869020462036,
      "learning_rate": 4.618e-05,
      "loss": 1.0689,
      "step": 382
    },
    {
      "epoch": 34.872727272727275,
      "grad_norm": 0.2833225727081299,
      "learning_rate": 4.617e-05,
      "loss": 1.0731,
      "step": 383
    },
    {
      "epoch": 34.96969696969697,
      "grad_norm": 0.2912672758102417,
      "learning_rate": 4.6160000000000005e-05,
      "loss": 1.0997,
      "step": 384
    },
    {
      "epoch": 35.0,
      "grad_norm": 0.3306087255477905,
      "learning_rate": 4.6150000000000004e-05,
      "loss": 1.0326,
      "step": 385
    },
    {
      "epoch": 35.096969696969694,
      "grad_norm": 0.2548561990261078,
      "learning_rate": 4.614e-05,
      "loss": 1.1003,
      "step": 386
    },
    {
      "epoch": 35.193939393939395,
      "grad_norm": 0.2752056121826172,
      "learning_rate": 4.613e-05,
      "loss": 1.0245,
      "step": 387
    },
    {
      "epoch": 35.29090909090909,
      "grad_norm": 0.2673061490058899,
      "learning_rate": 4.612e-05,
      "loss": 1.1164,
      "step": 388
    },
    {
      "epoch": 35.38787878787879,
      "grad_norm": 0.2884056568145752,
      "learning_rate": 4.6110000000000004e-05,
      "loss": 1.0179,
      "step": 389
    },
    {
      "epoch": 35.484848484848484,
      "grad_norm": 0.27470582723617554,
      "learning_rate": 4.61e-05,
      "loss": 1.0875,
      "step": 390
    },
    {
      "epoch": 35.484848484848484,
      "eval_loss": 1.068583369255066,
      "eval_runtime": 2.153,
      "eval_samples_per_second": 25.545,
      "eval_steps_per_second": 3.251,
      "step": 390
    },
    {
      "epoch": 35.58181818181818,
      "grad_norm": 0.3073423206806183,
      "learning_rate": 4.609e-05,
      "loss": 1.047,
      "step": 391
    },
    {
      "epoch": 35.67878787878788,
      "grad_norm": 0.25945353507995605,
      "learning_rate": 4.608e-05,
      "loss": 1.0403,
      "step": 392
    },
    {
      "epoch": 35.775757575757574,
      "grad_norm": 0.303960382938385,
      "learning_rate": 4.6070000000000004e-05,
      "loss": 1.1043,
      "step": 393
    },
    {
      "epoch": 35.872727272727275,
      "grad_norm": 0.26745638251304626,
      "learning_rate": 4.606e-05,
      "loss": 0.9779,
      "step": 394
    },
    {
      "epoch": 35.96969696969697,
      "grad_norm": 0.309352844953537,
      "learning_rate": 4.605e-05,
      "loss": 1.0151,
      "step": 395
    },
    {
      "epoch": 36.0,
      "grad_norm": 0.3577602505683899,
      "learning_rate": 4.604e-05,
      "loss": 1.002,
      "step": 396
    },
    {
      "epoch": 36.096969696969694,
      "grad_norm": 0.262930691242218,
      "learning_rate": 4.603e-05,
      "loss": 1.052,
      "step": 397
    },
    {
      "epoch": 36.193939393939395,
      "grad_norm": 0.2592763602733612,
      "learning_rate": 4.602e-05,
      "loss": 1.0056,
      "step": 398
    },
    {
      "epoch": 36.29090909090909,
      "grad_norm": 0.3038873076438904,
      "learning_rate": 4.601e-05,
      "loss": 1.0478,
      "step": 399
    },
    {
      "epoch": 36.38787878787879,
      "grad_norm": 0.27877938747406006,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.0019,
      "step": 400
    },
    {
      "epoch": 36.38787878787879,
      "eval_loss": 1.029101848602295,
      "eval_runtime": 2.1629,
      "eval_samples_per_second": 25.428,
      "eval_steps_per_second": 3.236,
      "step": 400
    },
    {
      "epoch": 36.484848484848484,
      "grad_norm": 0.29765585064888,
      "learning_rate": 4.599e-05,
      "loss": 0.968,
      "step": 401
    },
    {
      "epoch": 36.58181818181818,
      "grad_norm": 0.31292471289634705,
      "learning_rate": 4.5980000000000004e-05,
      "loss": 0.9544,
      "step": 402
    },
    {
      "epoch": 36.67878787878788,
      "grad_norm": 0.29842785000801086,
      "learning_rate": 4.597e-05,
      "loss": 1.0749,
      "step": 403
    },
    {
      "epoch": 36.775757575757574,
      "grad_norm": 0.32552585005760193,
      "learning_rate": 4.596e-05,
      "loss": 0.9925,
      "step": 404
    },
    {
      "epoch": 36.872727272727275,
      "grad_norm": 0.2806715965270996,
      "learning_rate": 4.5950000000000006e-05,
      "loss": 0.9852,
      "step": 405
    },
    {
      "epoch": 36.96969696969697,
      "grad_norm": 0.2873181402683258,
      "learning_rate": 4.594e-05,
      "loss": 0.9868,
      "step": 406
    },
    {
      "epoch": 37.0,
      "grad_norm": 0.3760148286819458,
      "learning_rate": 4.593e-05,
      "loss": 1.0563,
      "step": 407
    },
    {
      "epoch": 37.096969696969694,
      "grad_norm": 0.29149961471557617,
      "learning_rate": 4.592e-05,
      "loss": 0.9826,
      "step": 408
    },
    {
      "epoch": 37.193939393939395,
      "grad_norm": 0.3023413121700287,
      "learning_rate": 4.5910000000000006e-05,
      "loss": 0.9696,
      "step": 409
    },
    {
      "epoch": 37.29090909090909,
      "grad_norm": 0.31005623936653137,
      "learning_rate": 4.5900000000000004e-05,
      "loss": 0.9835,
      "step": 410
    },
    {
      "epoch": 37.29090909090909,
      "eval_loss": 0.9891466498374939,
      "eval_runtime": 2.1609,
      "eval_samples_per_second": 25.452,
      "eval_steps_per_second": 3.239,
      "step": 410
    },
    {
      "epoch": 37.38787878787879,
      "grad_norm": 0.30315306782722473,
      "learning_rate": 4.589e-05,
      "loss": 0.9656,
      "step": 411
    },
    {
      "epoch": 37.484848484848484,
      "grad_norm": 0.30635392665863037,
      "learning_rate": 4.588e-05,
      "loss": 0.9448,
      "step": 412
    },
    {
      "epoch": 37.58181818181818,
      "grad_norm": 0.2933935225009918,
      "learning_rate": 4.587e-05,
      "loss": 0.9165,
      "step": 413
    },
    {
      "epoch": 37.67878787878788,
      "grad_norm": 0.28974059224128723,
      "learning_rate": 4.5860000000000005e-05,
      "loss": 1.006,
      "step": 414
    },
    {
      "epoch": 37.775757575757574,
      "grad_norm": 0.27250662446022034,
      "learning_rate": 4.585e-05,
      "loss": 0.9674,
      "step": 415
    },
    {
      "epoch": 37.872727272727275,
      "grad_norm": 0.29652273654937744,
      "learning_rate": 4.584e-05,
      "loss": 0.9554,
      "step": 416
    },
    {
      "epoch": 37.96969696969697,
      "grad_norm": 0.3232543170452118,
      "learning_rate": 4.583e-05,
      "loss": 0.9759,
      "step": 417
    },
    {
      "epoch": 38.0,
      "grad_norm": 0.3857976794242859,
      "learning_rate": 4.5820000000000005e-05,
      "loss": 0.88,
      "step": 418
    },
    {
      "epoch": 38.096969696969694,
      "grad_norm": 0.2974799573421478,
      "learning_rate": 4.5810000000000004e-05,
      "loss": 0.9273,
      "step": 419
    },
    {
      "epoch": 38.193939393939395,
      "grad_norm": 0.30977925658226013,
      "learning_rate": 4.58e-05,
      "loss": 0.9634,
      "step": 420
    },
    {
      "epoch": 38.193939393939395,
      "eval_loss": 0.9490349292755127,
      "eval_runtime": 2.1484,
      "eval_samples_per_second": 25.6,
      "eval_steps_per_second": 3.258,
      "step": 420
    },
    {
      "epoch": 38.29090909090909,
      "grad_norm": 0.3215106725692749,
      "learning_rate": 4.579e-05,
      "loss": 0.9166,
      "step": 421
    },
    {
      "epoch": 38.38787878787879,
      "grad_norm": 0.3041662871837616,
      "learning_rate": 4.578e-05,
      "loss": 0.9446,
      "step": 422
    },
    {
      "epoch": 38.484848484848484,
      "grad_norm": 0.3011530637741089,
      "learning_rate": 4.5770000000000004e-05,
      "loss": 0.9306,
      "step": 423
    },
    {
      "epoch": 38.58181818181818,
      "grad_norm": 0.310140997171402,
      "learning_rate": 4.576e-05,
      "loss": 0.8723,
      "step": 424
    },
    {
      "epoch": 38.67878787878788,
      "grad_norm": 0.2920651137828827,
      "learning_rate": 4.575e-05,
      "loss": 0.9286,
      "step": 425
    },
    {
      "epoch": 38.775757575757574,
      "grad_norm": 0.2789269685745239,
      "learning_rate": 4.574e-05,
      "loss": 0.9178,
      "step": 426
    },
    {
      "epoch": 38.872727272727275,
      "grad_norm": 0.30473312735557556,
      "learning_rate": 4.573e-05,
      "loss": 0.9003,
      "step": 427
    },
    {
      "epoch": 38.96969696969697,
      "grad_norm": 0.2925727963447571,
      "learning_rate": 4.572e-05,
      "loss": 0.9238,
      "step": 428
    },
    {
      "epoch": 39.0,
      "grad_norm": 0.4265953600406647,
      "learning_rate": 4.571e-05,
      "loss": 0.8359,
      "step": 429
    },
    {
      "epoch": 39.096969696969694,
      "grad_norm": 0.3189105689525604,
      "learning_rate": 4.5700000000000006e-05,
      "loss": 0.8581,
      "step": 430
    },
    {
      "epoch": 39.096969696969694,
      "eval_loss": 0.9099050760269165,
      "eval_runtime": 2.1501,
      "eval_samples_per_second": 25.58,
      "eval_steps_per_second": 3.256,
      "step": 430
    },
    {
      "epoch": 39.193939393939395,
      "grad_norm": 0.29996800422668457,
      "learning_rate": 4.569e-05,
      "loss": 0.9016,
      "step": 431
    },
    {
      "epoch": 39.29090909090909,
      "grad_norm": 0.3168741464614868,
      "learning_rate": 4.568e-05,
      "loss": 0.9345,
      "step": 432
    },
    {
      "epoch": 39.38787878787879,
      "grad_norm": 0.2974991500377655,
      "learning_rate": 4.567e-05,
      "loss": 0.9088,
      "step": 433
    },
    {
      "epoch": 39.484848484848484,
      "grad_norm": 0.3686419725418091,
      "learning_rate": 4.566e-05,
      "loss": 0.8999,
      "step": 434
    },
    {
      "epoch": 39.58181818181818,
      "grad_norm": 0.3079046308994293,
      "learning_rate": 4.5650000000000005e-05,
      "loss": 0.852,
      "step": 435
    },
    {
      "epoch": 39.67878787878788,
      "grad_norm": 0.3118555247783661,
      "learning_rate": 4.564e-05,
      "loss": 0.8909,
      "step": 436
    },
    {
      "epoch": 39.775757575757574,
      "grad_norm": 0.31345677375793457,
      "learning_rate": 4.563e-05,
      "loss": 0.8366,
      "step": 437
    },
    {
      "epoch": 39.872727272727275,
      "grad_norm": 0.3082335293292999,
      "learning_rate": 4.562e-05,
      "loss": 0.8281,
      "step": 438
    },
    {
      "epoch": 39.96969696969697,
      "grad_norm": 0.3386876881122589,
      "learning_rate": 4.5610000000000005e-05,
      "loss": 0.8827,
      "step": 439
    },
    {
      "epoch": 40.0,
      "grad_norm": 0.46113550662994385,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 0.8334,
      "step": 440
    },
    {
      "epoch": 40.0,
      "eval_loss": 0.8736940622329712,
      "eval_runtime": 2.1491,
      "eval_samples_per_second": 25.592,
      "eval_steps_per_second": 3.257,
      "step": 440
    },
    {
      "epoch": 40.096969696969694,
      "grad_norm": 0.29570069909095764,
      "learning_rate": 4.559e-05,
      "loss": 0.8461,
      "step": 441
    },
    {
      "epoch": 40.193939393939395,
      "grad_norm": 0.33492523431777954,
      "learning_rate": 4.558e-05,
      "loss": 0.8443,
      "step": 442
    },
    {
      "epoch": 40.29090909090909,
      "grad_norm": 0.29383033514022827,
      "learning_rate": 4.557e-05,
      "loss": 0.805,
      "step": 443
    },
    {
      "epoch": 40.38787878787879,
      "grad_norm": 0.3032298684120178,
      "learning_rate": 4.5560000000000004e-05,
      "loss": 0.862,
      "step": 444
    },
    {
      "epoch": 40.484848484848484,
      "grad_norm": 0.3141234815120697,
      "learning_rate": 4.555e-05,
      "loss": 0.8511,
      "step": 445
    },
    {
      "epoch": 40.58181818181818,
      "grad_norm": 0.29081517457962036,
      "learning_rate": 4.554000000000001e-05,
      "loss": 0.8437,
      "step": 446
    },
    {
      "epoch": 40.67878787878788,
      "grad_norm": 0.2950437664985657,
      "learning_rate": 4.553e-05,
      "loss": 0.8204,
      "step": 447
    },
    {
      "epoch": 40.775757575757574,
      "grad_norm": 0.3393695056438446,
      "learning_rate": 4.5520000000000005e-05,
      "loss": 0.8213,
      "step": 448
    },
    {
      "epoch": 40.872727272727275,
      "grad_norm": 0.331230491399765,
      "learning_rate": 4.551e-05,
      "loss": 0.8446,
      "step": 449
    },
    {
      "epoch": 40.96969696969697,
      "grad_norm": 0.33039379119873047,
      "learning_rate": 4.55e-05,
      "loss": 0.8527,
      "step": 450
    },
    {
      "epoch": 40.96969696969697,
      "eval_loss": 0.8403151035308838,
      "eval_runtime": 2.1512,
      "eval_samples_per_second": 25.567,
      "eval_steps_per_second": 3.254,
      "step": 450
    },
    {
      "epoch": 41.0,
      "grad_norm": 0.47446638345718384,
      "learning_rate": 4.549000000000001e-05,
      "loss": 0.8452,
      "step": 451
    },
    {
      "epoch": 41.096969696969694,
      "grad_norm": 0.2963954210281372,
      "learning_rate": 4.548e-05,
      "loss": 0.8243,
      "step": 452
    },
    {
      "epoch": 41.193939393939395,
      "grad_norm": 0.32614511251449585,
      "learning_rate": 4.5470000000000003e-05,
      "loss": 0.7441,
      "step": 453
    },
    {
      "epoch": 41.29090909090909,
      "grad_norm": 0.2976125478744507,
      "learning_rate": 4.546e-05,
      "loss": 0.8023,
      "step": 454
    },
    {
      "epoch": 41.38787878787879,
      "grad_norm": 0.3358374834060669,
      "learning_rate": 4.545000000000001e-05,
      "loss": 0.805,
      "step": 455
    },
    {
      "epoch": 41.484848484848484,
      "grad_norm": 0.28136196732521057,
      "learning_rate": 4.5440000000000005e-05,
      "loss": 0.7902,
      "step": 456
    },
    {
      "epoch": 41.58181818181818,
      "grad_norm": 0.28420117497444153,
      "learning_rate": 4.543e-05,
      "loss": 0.8673,
      "step": 457
    },
    {
      "epoch": 41.67878787878788,
      "grad_norm": 0.32107681035995483,
      "learning_rate": 4.542e-05,
      "loss": 0.7885,
      "step": 458
    },
    {
      "epoch": 41.775757575757574,
      "grad_norm": 0.30052316188812256,
      "learning_rate": 4.541e-05,
      "loss": 0.857,
      "step": 459
    },
    {
      "epoch": 41.872727272727275,
      "grad_norm": 0.3333911895751953,
      "learning_rate": 4.5400000000000006e-05,
      "loss": 0.7672,
      "step": 460
    },
    {
      "epoch": 41.872727272727275,
      "eval_loss": 0.8103982210159302,
      "eval_runtime": 2.1739,
      "eval_samples_per_second": 25.301,
      "eval_steps_per_second": 3.22,
      "step": 460
    },
    {
      "epoch": 41.96969696969697,
      "grad_norm": 0.32705608010292053,
      "learning_rate": 4.5390000000000004e-05,
      "loss": 0.8136,
      "step": 461
    },
    {
      "epoch": 42.0,
      "grad_norm": 0.3649839460849762,
      "learning_rate": 4.538e-05,
      "loss": 0.773,
      "step": 462
    },
    {
      "epoch": 42.096969696969694,
      "grad_norm": 0.313376784324646,
      "learning_rate": 4.537e-05,
      "loss": 0.7613,
      "step": 463
    },
    {
      "epoch": 42.193939393939395,
      "grad_norm": 0.3517065644264221,
      "learning_rate": 4.536e-05,
      "loss": 0.7411,
      "step": 464
    },
    {
      "epoch": 42.29090909090909,
      "grad_norm": 0.3445081114768982,
      "learning_rate": 4.5350000000000005e-05,
      "loss": 0.8118,
      "step": 465
    },
    {
      "epoch": 42.38787878787879,
      "grad_norm": 0.2820802330970764,
      "learning_rate": 4.534e-05,
      "loss": 0.7625,
      "step": 466
    },
    {
      "epoch": 42.484848484848484,
      "grad_norm": 0.31504544615745544,
      "learning_rate": 4.533e-05,
      "loss": 0.7453,
      "step": 467
    },
    {
      "epoch": 42.58181818181818,
      "grad_norm": 0.36062583327293396,
      "learning_rate": 4.532e-05,
      "loss": 0.7759,
      "step": 468
    },
    {
      "epoch": 42.67878787878788,
      "grad_norm": 0.2900105118751526,
      "learning_rate": 4.5310000000000005e-05,
      "loss": 0.7919,
      "step": 469
    },
    {
      "epoch": 42.775757575757574,
      "grad_norm": 0.2746892273426056,
      "learning_rate": 4.53e-05,
      "loss": 0.7571,
      "step": 470
    },
    {
      "epoch": 42.775757575757574,
      "eval_loss": 0.7850269079208374,
      "eval_runtime": 2.1528,
      "eval_samples_per_second": 25.549,
      "eval_steps_per_second": 3.252,
      "step": 470
    },
    {
      "epoch": 42.872727272727275,
      "grad_norm": 0.33745649456977844,
      "learning_rate": 4.529e-05,
      "loss": 0.7894,
      "step": 471
    },
    {
      "epoch": 42.96969696969697,
      "grad_norm": 0.36385348439216614,
      "learning_rate": 4.528e-05,
      "loss": 0.8226,
      "step": 472
    },
    {
      "epoch": 43.0,
      "grad_norm": 0.42536821961402893,
      "learning_rate": 4.527e-05,
      "loss": 0.7393,
      "step": 473
    },
    {
      "epoch": 43.096969696969694,
      "grad_norm": 0.2656398117542267,
      "learning_rate": 4.5260000000000004e-05,
      "loss": 0.7442,
      "step": 474
    },
    {
      "epoch": 43.193939393939395,
      "grad_norm": 0.28830406069755554,
      "learning_rate": 4.525e-05,
      "loss": 0.7747,
      "step": 475
    },
    {
      "epoch": 43.29090909090909,
      "grad_norm": 0.36671680212020874,
      "learning_rate": 4.524000000000001e-05,
      "loss": 0.8123,
      "step": 476
    },
    {
      "epoch": 43.38787878787879,
      "grad_norm": 0.3925282657146454,
      "learning_rate": 4.523e-05,
      "loss": 0.6966,
      "step": 477
    },
    {
      "epoch": 43.484848484848484,
      "grad_norm": 0.3425087332725525,
      "learning_rate": 4.5220000000000004e-05,
      "loss": 0.7403,
      "step": 478
    },
    {
      "epoch": 43.58181818181818,
      "grad_norm": 0.33888742327690125,
      "learning_rate": 4.521e-05,
      "loss": 0.7721,
      "step": 479
    },
    {
      "epoch": 43.67878787878788,
      "grad_norm": 0.3055487871170044,
      "learning_rate": 4.52e-05,
      "loss": 0.7711,
      "step": 480
    },
    {
      "epoch": 43.67878787878788,
      "eval_loss": 0.7622922658920288,
      "eval_runtime": 2.1454,
      "eval_samples_per_second": 25.637,
      "eval_steps_per_second": 3.263,
      "step": 480
    },
    {
      "epoch": 43.775757575757574,
      "grad_norm": 0.3457646667957306,
      "learning_rate": 4.5190000000000006e-05,
      "loss": 0.6743,
      "step": 481
    },
    {
      "epoch": 43.872727272727275,
      "grad_norm": 0.31191420555114746,
      "learning_rate": 4.518e-05,
      "loss": 0.7242,
      "step": 482
    },
    {
      "epoch": 43.96969696969697,
      "grad_norm": 0.2905394434928894,
      "learning_rate": 4.517e-05,
      "loss": 0.7729,
      "step": 483
    },
    {
      "epoch": 44.0,
      "grad_norm": 0.5394179821014404,
      "learning_rate": 4.516e-05,
      "loss": 0.758,
      "step": 484
    },
    {
      "epoch": 44.096969696969694,
      "grad_norm": 0.31047454476356506,
      "learning_rate": 4.5150000000000006e-05,
      "loss": 0.7289,
      "step": 485
    },
    {
      "epoch": 44.193939393939395,
      "grad_norm": 0.4167446792125702,
      "learning_rate": 4.5140000000000005e-05,
      "loss": 0.6943,
      "step": 486
    },
    {
      "epoch": 44.29090909090909,
      "grad_norm": 0.41475990414619446,
      "learning_rate": 4.513e-05,
      "loss": 0.6968,
      "step": 487
    },
    {
      "epoch": 44.38787878787879,
      "grad_norm": 0.3056599795818329,
      "learning_rate": 4.512e-05,
      "loss": 0.6853,
      "step": 488
    },
    {
      "epoch": 44.484848484848484,
      "grad_norm": 0.3095908761024475,
      "learning_rate": 4.511e-05,
      "loss": 0.6931,
      "step": 489
    },
    {
      "epoch": 44.58181818181818,
      "grad_norm": 0.44538062810897827,
      "learning_rate": 4.5100000000000005e-05,
      "loss": 0.8197,
      "step": 490
    },
    {
      "epoch": 44.58181818181818,
      "eval_loss": 0.7430852651596069,
      "eval_runtime": 2.1626,
      "eval_samples_per_second": 25.433,
      "eval_steps_per_second": 3.237,
      "step": 490
    },
    {
      "epoch": 44.67878787878788,
      "grad_norm": 0.37403160333633423,
      "learning_rate": 4.5090000000000004e-05,
      "loss": 0.75,
      "step": 491
    },
    {
      "epoch": 44.775757575757574,
      "grad_norm": 0.33197081089019775,
      "learning_rate": 4.508e-05,
      "loss": 0.7516,
      "step": 492
    },
    {
      "epoch": 44.872727272727275,
      "grad_norm": 0.4307508170604706,
      "learning_rate": 4.507e-05,
      "loss": 0.7692,
      "step": 493
    },
    {
      "epoch": 44.96969696969697,
      "grad_norm": 0.3568163812160492,
      "learning_rate": 4.506e-05,
      "loss": 0.7184,
      "step": 494
    },
    {
      "epoch": 45.0,
      "grad_norm": 0.5191144347190857,
      "learning_rate": 4.5050000000000004e-05,
      "loss": 0.5799,
      "step": 495
    },
    {
      "epoch": 45.096969696969694,
      "grad_norm": 0.29018500447273254,
      "learning_rate": 4.504e-05,
      "loss": 0.6881,
      "step": 496
    },
    {
      "epoch": 45.193939393939395,
      "grad_norm": 0.3260664641857147,
      "learning_rate": 4.503e-05,
      "loss": 0.6887,
      "step": 497
    },
    {
      "epoch": 45.29090909090909,
      "grad_norm": 0.33074823021888733,
      "learning_rate": 4.502e-05,
      "loss": 0.7306,
      "step": 498
    },
    {
      "epoch": 45.38787878787879,
      "grad_norm": 0.3127662241458893,
      "learning_rate": 4.5010000000000004e-05,
      "loss": 0.6712,
      "step": 499
    },
    {
      "epoch": 45.484848484848484,
      "grad_norm": 0.4097149074077606,
      "learning_rate": 4.5e-05,
      "loss": 0.7328,
      "step": 500
    },
    {
      "epoch": 45.484848484848484,
      "eval_loss": 0.725439190864563,
      "eval_runtime": 2.171,
      "eval_samples_per_second": 25.333,
      "eval_steps_per_second": 3.224,
      "step": 500
    },
    {
      "epoch": 45.58181818181818,
      "grad_norm": 0.4050314426422119,
      "learning_rate": 4.499e-05,
      "loss": 0.7028,
      "step": 501
    },
    {
      "epoch": 45.67878787878788,
      "grad_norm": 0.2748536467552185,
      "learning_rate": 4.498e-05,
      "loss": 0.7184,
      "step": 502
    },
    {
      "epoch": 45.775757575757574,
      "grad_norm": 0.3762427568435669,
      "learning_rate": 4.497e-05,
      "loss": 0.7085,
      "step": 503
    },
    {
      "epoch": 45.872727272727275,
      "grad_norm": 0.3729848265647888,
      "learning_rate": 4.496e-05,
      "loss": 0.7336,
      "step": 504
    },
    {
      "epoch": 45.96969696969697,
      "grad_norm": 0.38429006934165955,
      "learning_rate": 4.495e-05,
      "loss": 0.7051,
      "step": 505
    },
    {
      "epoch": 46.0,
      "grad_norm": 0.4606812298297882,
      "learning_rate": 4.494000000000001e-05,
      "loss": 0.6661,
      "step": 506
    },
    {
      "epoch": 46.096969696969694,
      "grad_norm": 0.3665672242641449,
      "learning_rate": 4.493e-05,
      "loss": 0.6646,
      "step": 507
    },
    {
      "epoch": 46.193939393939395,
      "grad_norm": 0.3234887421131134,
      "learning_rate": 4.4920000000000004e-05,
      "loss": 0.7053,
      "step": 508
    },
    {
      "epoch": 46.29090909090909,
      "grad_norm": 0.34196731448173523,
      "learning_rate": 4.491e-05,
      "loss": 0.6993,
      "step": 509
    },
    {
      "epoch": 46.38787878787879,
      "grad_norm": 0.2899550497531891,
      "learning_rate": 4.49e-05,
      "loss": 0.6612,
      "step": 510
    },
    {
      "epoch": 46.38787878787879,
      "eval_loss": 0.7102727293968201,
      "eval_runtime": 2.1663,
      "eval_samples_per_second": 25.389,
      "eval_steps_per_second": 3.231,
      "step": 510
    },
    {
      "epoch": 46.484848484848484,
      "grad_norm": 0.42077502608299255,
      "learning_rate": 4.4890000000000006e-05,
      "loss": 0.7378,
      "step": 511
    },
    {
      "epoch": 46.58181818181818,
      "grad_norm": 0.34420883655548096,
      "learning_rate": 4.488e-05,
      "loss": 0.7084,
      "step": 512
    },
    {
      "epoch": 46.67878787878788,
      "grad_norm": 0.31922397017478943,
      "learning_rate": 4.487e-05,
      "loss": 0.7148,
      "step": 513
    },
    {
      "epoch": 46.775757575757574,
      "grad_norm": 0.3793002665042877,
      "learning_rate": 4.486e-05,
      "loss": 0.6306,
      "step": 514
    },
    {
      "epoch": 46.872727272727275,
      "grad_norm": 0.33293092250823975,
      "learning_rate": 4.4850000000000006e-05,
      "loss": 0.717,
      "step": 515
    },
    {
      "epoch": 46.96969696969697,
      "grad_norm": 0.31802603602409363,
      "learning_rate": 4.4840000000000004e-05,
      "loss": 0.6785,
      "step": 516
    },
    {
      "epoch": 47.0,
      "grad_norm": 0.38046082854270935,
      "learning_rate": 4.483e-05,
      "loss": 0.6203,
      "step": 517
    },
    {
      "epoch": 47.096969696969694,
      "grad_norm": 0.3020094037055969,
      "learning_rate": 4.482e-05,
      "loss": 0.6657,
      "step": 518
    },
    {
      "epoch": 47.193939393939395,
      "grad_norm": 0.30682918429374695,
      "learning_rate": 4.481e-05,
      "loss": 0.6845,
      "step": 519
    },
    {
      "epoch": 47.29090909090909,
      "grad_norm": 0.3144380450248718,
      "learning_rate": 4.4800000000000005e-05,
      "loss": 0.6945,
      "step": 520
    },
    {
      "epoch": 47.29090909090909,
      "eval_loss": 0.6961836218833923,
      "eval_runtime": 2.1618,
      "eval_samples_per_second": 25.442,
      "eval_steps_per_second": 3.238,
      "step": 520
    },
    {
      "epoch": 47.38787878787879,
      "grad_norm": 0.33046039938926697,
      "learning_rate": 4.479e-05,
      "loss": 0.7259,
      "step": 521
    },
    {
      "epoch": 47.484848484848484,
      "grad_norm": 0.2844446897506714,
      "learning_rate": 4.478e-05,
      "loss": 0.6932,
      "step": 522
    },
    {
      "epoch": 47.58181818181818,
      "grad_norm": 0.3066650927066803,
      "learning_rate": 4.477e-05,
      "loss": 0.6395,
      "step": 523
    },
    {
      "epoch": 47.67878787878788,
      "grad_norm": 0.2580311596393585,
      "learning_rate": 4.4760000000000005e-05,
      "loss": 0.6321,
      "step": 524
    },
    {
      "epoch": 47.775757575757574,
      "grad_norm": 0.28855329751968384,
      "learning_rate": 4.4750000000000004e-05,
      "loss": 0.6919,
      "step": 525
    },
    {
      "epoch": 47.872727272727275,
      "grad_norm": 0.28060483932495117,
      "learning_rate": 4.474e-05,
      "loss": 0.6473,
      "step": 526
    },
    {
      "epoch": 47.96969696969697,
      "grad_norm": 0.3266681730747223,
      "learning_rate": 4.473e-05,
      "loss": 0.6605,
      "step": 527
    },
    {
      "epoch": 48.0,
      "grad_norm": 0.5188031792640686,
      "learning_rate": 4.472e-05,
      "loss": 0.722,
      "step": 528
    },
    {
      "epoch": 48.096969696969694,
      "grad_norm": 0.3377719223499298,
      "learning_rate": 4.4710000000000004e-05,
      "loss": 0.6797,
      "step": 529
    },
    {
      "epoch": 48.193939393939395,
      "grad_norm": 0.35471999645233154,
      "learning_rate": 4.47e-05,
      "loss": 0.6555,
      "step": 530
    },
    {
      "epoch": 48.193939393939395,
      "eval_loss": 0.6851030588150024,
      "eval_runtime": 2.1559,
      "eval_samples_per_second": 25.512,
      "eval_steps_per_second": 3.247,
      "step": 530
    },
    {
      "epoch": 48.29090909090909,
      "grad_norm": 0.480877548456192,
      "learning_rate": 4.469e-05,
      "loss": 0.6919,
      "step": 531
    },
    {
      "epoch": 48.38787878787879,
      "grad_norm": 0.2577911913394928,
      "learning_rate": 4.468e-05,
      "loss": 0.6644,
      "step": 532
    },
    {
      "epoch": 48.484848484848484,
      "grad_norm": 0.3358922302722931,
      "learning_rate": 4.467e-05,
      "loss": 0.6243,
      "step": 533
    },
    {
      "epoch": 48.58181818181818,
      "grad_norm": 0.42016032338142395,
      "learning_rate": 4.466e-05,
      "loss": 0.7167,
      "step": 534
    },
    {
      "epoch": 48.67878787878788,
      "grad_norm": 0.34250354766845703,
      "learning_rate": 4.465e-05,
      "loss": 0.6782,
      "step": 535
    },
    {
      "epoch": 48.775757575757574,
      "grad_norm": 0.3065449893474579,
      "learning_rate": 4.4640000000000006e-05,
      "loss": 0.6649,
      "step": 536
    },
    {
      "epoch": 48.872727272727275,
      "grad_norm": 0.23833245038986206,
      "learning_rate": 4.463e-05,
      "loss": 0.6054,
      "step": 537
    },
    {
      "epoch": 48.96969696969697,
      "grad_norm": 0.29932355880737305,
      "learning_rate": 4.462e-05,
      "loss": 0.6268,
      "step": 538
    },
    {
      "epoch": 49.0,
      "grad_norm": 0.4887162148952484,
      "learning_rate": 4.461e-05,
      "loss": 0.7053,
      "step": 539
    },
    {
      "epoch": 49.096969696969694,
      "grad_norm": 0.29336071014404297,
      "learning_rate": 4.46e-05,
      "loss": 0.7038,
      "step": 540
    },
    {
      "epoch": 49.096969696969694,
      "eval_loss": 0.6746473908424377,
      "eval_runtime": 2.1499,
      "eval_samples_per_second": 25.583,
      "eval_steps_per_second": 3.256,
      "step": 540
    },
    {
      "epoch": 49.193939393939395,
      "grad_norm": 0.2892014682292938,
      "learning_rate": 4.4590000000000005e-05,
      "loss": 0.6504,
      "step": 541
    },
    {
      "epoch": 49.29090909090909,
      "grad_norm": 0.30927982926368713,
      "learning_rate": 4.458e-05,
      "loss": 0.646,
      "step": 542
    },
    {
      "epoch": 49.38787878787879,
      "grad_norm": 0.2865374684333801,
      "learning_rate": 4.457e-05,
      "loss": 0.6393,
      "step": 543
    },
    {
      "epoch": 49.484848484848484,
      "grad_norm": 0.3668179512023926,
      "learning_rate": 4.456e-05,
      "loss": 0.6251,
      "step": 544
    },
    {
      "epoch": 49.58181818181818,
      "grad_norm": 0.3081008195877075,
      "learning_rate": 4.4550000000000005e-05,
      "loss": 0.6287,
      "step": 545
    },
    {
      "epoch": 49.67878787878788,
      "grad_norm": 0.30829358100891113,
      "learning_rate": 4.4540000000000004e-05,
      "loss": 0.6336,
      "step": 546
    },
    {
      "epoch": 49.775757575757574,
      "grad_norm": 0.34429988265037537,
      "learning_rate": 4.453e-05,
      "loss": 0.6214,
      "step": 547
    },
    {
      "epoch": 49.872727272727275,
      "grad_norm": 0.31597399711608887,
      "learning_rate": 4.452e-05,
      "loss": 0.629,
      "step": 548
    },
    {
      "epoch": 49.96969696969697,
      "grad_norm": 0.3678438067436218,
      "learning_rate": 4.451e-05,
      "loss": 0.7262,
      "step": 549
    },
    {
      "epoch": 50.0,
      "grad_norm": 0.4396074414253235,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 0.6663,
      "step": 550
    },
    {
      "epoch": 50.0,
      "eval_loss": 0.6663048267364502,
      "eval_runtime": 2.1749,
      "eval_samples_per_second": 25.289,
      "eval_steps_per_second": 3.219,
      "step": 550
    },
    {
      "epoch": 50.096969696969694,
      "grad_norm": 0.32783371210098267,
      "learning_rate": 4.449e-05,
      "loss": 0.6205,
      "step": 551
    },
    {
      "epoch": 50.193939393939395,
      "grad_norm": 0.3056533634662628,
      "learning_rate": 4.448e-05,
      "loss": 0.6235,
      "step": 552
    },
    {
      "epoch": 50.29090909090909,
      "grad_norm": 0.2603963017463684,
      "learning_rate": 4.447e-05,
      "loss": 0.6332,
      "step": 553
    },
    {
      "epoch": 50.38787878787879,
      "grad_norm": 0.365771621465683,
      "learning_rate": 4.4460000000000005e-05,
      "loss": 0.667,
      "step": 554
    },
    {
      "epoch": 50.484848484848484,
      "grad_norm": 0.33529090881347656,
      "learning_rate": 4.445e-05,
      "loss": 0.6358,
      "step": 555
    },
    {
      "epoch": 50.58181818181818,
      "grad_norm": 0.3174441456794739,
      "learning_rate": 4.444e-05,
      "loss": 0.6445,
      "step": 556
    },
    {
      "epoch": 50.67878787878788,
      "grad_norm": 0.30779361724853516,
      "learning_rate": 4.443e-05,
      "loss": 0.6842,
      "step": 557
    },
    {
      "epoch": 50.775757575757574,
      "grad_norm": 0.2936500906944275,
      "learning_rate": 4.442e-05,
      "loss": 0.6634,
      "step": 558
    },
    {
      "epoch": 50.872727272727275,
      "grad_norm": 0.295730859041214,
      "learning_rate": 4.4410000000000003e-05,
      "loss": 0.6013,
      "step": 559
    },
    {
      "epoch": 50.96969696969697,
      "grad_norm": 0.2970040738582611,
      "learning_rate": 4.44e-05,
      "loss": 0.6143,
      "step": 560
    },
    {
      "epoch": 50.96969696969697,
      "eval_loss": 0.658263623714447,
      "eval_runtime": 2.1643,
      "eval_samples_per_second": 25.412,
      "eval_steps_per_second": 3.234,
      "step": 560
    },
    {
      "epoch": 51.0,
      "grad_norm": 0.4583488404750824,
      "learning_rate": 4.439000000000001e-05,
      "loss": 0.7217,
      "step": 561
    },
    {
      "epoch": 51.096969696969694,
      "grad_norm": 0.3247191309928894,
      "learning_rate": 4.438e-05,
      "loss": 0.6468,
      "step": 562
    },
    {
      "epoch": 51.193939393939395,
      "grad_norm": 0.2687683701515198,
      "learning_rate": 4.4370000000000004e-05,
      "loss": 0.6075,
      "step": 563
    },
    {
      "epoch": 51.29090909090909,
      "grad_norm": 0.28057968616485596,
      "learning_rate": 4.436e-05,
      "loss": 0.6772,
      "step": 564
    },
    {
      "epoch": 51.38787878787879,
      "grad_norm": 0.333745539188385,
      "learning_rate": 4.435e-05,
      "loss": 0.629,
      "step": 565
    },
    {
      "epoch": 51.484848484848484,
      "grad_norm": 0.2607054114341736,
      "learning_rate": 4.4340000000000006e-05,
      "loss": 0.6077,
      "step": 566
    },
    {
      "epoch": 51.58181818181818,
      "grad_norm": 0.3412008583545685,
      "learning_rate": 4.4330000000000004e-05,
      "loss": 0.6605,
      "step": 567
    },
    {
      "epoch": 51.67878787878788,
      "grad_norm": 0.2872512936592102,
      "learning_rate": 4.432e-05,
      "loss": 0.661,
      "step": 568
    },
    {
      "epoch": 51.775757575757574,
      "grad_norm": 0.23332110047340393,
      "learning_rate": 4.431e-05,
      "loss": 0.5995,
      "step": 569
    },
    {
      "epoch": 51.872727272727275,
      "grad_norm": 0.3014121651649475,
      "learning_rate": 4.43e-05,
      "loss": 0.5836,
      "step": 570
    },
    {
      "epoch": 51.872727272727275,
      "eval_loss": 0.6509251594543457,
      "eval_runtime": 2.1649,
      "eval_samples_per_second": 25.406,
      "eval_steps_per_second": 3.233,
      "step": 570
    },
    {
      "epoch": 51.96969696969697,
      "grad_norm": 0.27421829104423523,
      "learning_rate": 4.4290000000000005e-05,
      "loss": 0.6791,
      "step": 571
    },
    {
      "epoch": 52.0,
      "grad_norm": 0.4212527871131897,
      "learning_rate": 4.428e-05,
      "loss": 0.5194,
      "step": 572
    },
    {
      "epoch": 52.096969696969694,
      "grad_norm": 0.2564944326877594,
      "learning_rate": 4.427e-05,
      "loss": 0.5887,
      "step": 573
    },
    {
      "epoch": 52.193939393939395,
      "grad_norm": 0.271538645029068,
      "learning_rate": 4.426e-05,
      "loss": 0.6401,
      "step": 574
    },
    {
      "epoch": 52.29090909090909,
      "grad_norm": 0.30780258774757385,
      "learning_rate": 4.4250000000000005e-05,
      "loss": 0.676,
      "step": 575
    },
    {
      "epoch": 52.38787878787879,
      "grad_norm": 0.25906750559806824,
      "learning_rate": 4.424e-05,
      "loss": 0.6563,
      "step": 576
    },
    {
      "epoch": 52.484848484848484,
      "grad_norm": 0.3498859107494354,
      "learning_rate": 4.423e-05,
      "loss": 0.6287,
      "step": 577
    },
    {
      "epoch": 52.58181818181818,
      "grad_norm": 0.31031695008277893,
      "learning_rate": 4.422e-05,
      "loss": 0.6062,
      "step": 578
    },
    {
      "epoch": 52.67878787878788,
      "grad_norm": 0.2525545060634613,
      "learning_rate": 4.421e-05,
      "loss": 0.5684,
      "step": 579
    },
    {
      "epoch": 52.775757575757574,
      "grad_norm": 0.25607815384864807,
      "learning_rate": 4.4200000000000004e-05,
      "loss": 0.6153,
      "step": 580
    },
    {
      "epoch": 52.775757575757574,
      "eval_loss": 0.6439118385314941,
      "eval_runtime": 2.1484,
      "eval_samples_per_second": 25.6,
      "eval_steps_per_second": 3.258,
      "step": 580
    },
    {
      "epoch": 52.872727272727275,
      "grad_norm": 0.30004850029945374,
      "learning_rate": 4.419e-05,
      "loss": 0.6104,
      "step": 581
    },
    {
      "epoch": 52.96969696969697,
      "grad_norm": 0.27988260984420776,
      "learning_rate": 4.418000000000001e-05,
      "loss": 0.6384,
      "step": 582
    },
    {
      "epoch": 53.0,
      "grad_norm": 0.46435651183128357,
      "learning_rate": 4.417e-05,
      "loss": 0.655,
      "step": 583
    },
    {
      "epoch": 53.096969696969694,
      "grad_norm": 0.2364508956670761,
      "learning_rate": 4.4160000000000004e-05,
      "loss": 0.6002,
      "step": 584
    },
    {
      "epoch": 53.193939393939395,
      "grad_norm": 0.2596590220928192,
      "learning_rate": 4.415e-05,
      "loss": 0.6007,
      "step": 585
    },
    {
      "epoch": 53.29090909090909,
      "grad_norm": 0.32126492261886597,
      "learning_rate": 4.414e-05,
      "loss": 0.5991,
      "step": 586
    },
    {
      "epoch": 53.38787878787879,
      "grad_norm": 0.2967338263988495,
      "learning_rate": 4.4130000000000006e-05,
      "loss": 0.6089,
      "step": 587
    },
    {
      "epoch": 53.484848484848484,
      "grad_norm": 0.22388562560081482,
      "learning_rate": 4.412e-05,
      "loss": 0.6334,
      "step": 588
    },
    {
      "epoch": 53.58181818181818,
      "grad_norm": 0.34703031182289124,
      "learning_rate": 4.411e-05,
      "loss": 0.6915,
      "step": 589
    },
    {
      "epoch": 53.67878787878788,
      "grad_norm": 0.3233970105648041,
      "learning_rate": 4.41e-05,
      "loss": 0.6571,
      "step": 590
    },
    {
      "epoch": 53.67878787878788,
      "eval_loss": 0.638057291507721,
      "eval_runtime": 2.1659,
      "eval_samples_per_second": 25.394,
      "eval_steps_per_second": 3.232,
      "step": 590
    },
    {
      "epoch": 53.775757575757574,
      "grad_norm": 0.28742972016334534,
      "learning_rate": 4.4090000000000006e-05,
      "loss": 0.5957,
      "step": 591
    },
    {
      "epoch": 53.872727272727275,
      "grad_norm": 0.22206705808639526,
      "learning_rate": 4.4080000000000005e-05,
      "loss": 0.5216,
      "step": 592
    },
    {
      "epoch": 53.96969696969697,
      "grad_norm": 0.26034870743751526,
      "learning_rate": 4.407e-05,
      "loss": 0.6628,
      "step": 593
    },
    {
      "epoch": 54.0,
      "grad_norm": 0.5107381343841553,
      "learning_rate": 4.406e-05,
      "loss": 0.5909,
      "step": 594
    },
    {
      "epoch": 54.096969696969694,
      "grad_norm": 0.32183170318603516,
      "learning_rate": 4.405e-05,
      "loss": 0.6119,
      "step": 595
    },
    {
      "epoch": 54.193939393939395,
      "grad_norm": 0.31215551495552063,
      "learning_rate": 4.4040000000000005e-05,
      "loss": 0.6399,
      "step": 596
    },
    {
      "epoch": 54.29090909090909,
      "grad_norm": 0.2621861696243286,
      "learning_rate": 4.4030000000000004e-05,
      "loss": 0.5847,
      "step": 597
    },
    {
      "epoch": 54.38787878787879,
      "grad_norm": 0.25500035285949707,
      "learning_rate": 4.402e-05,
      "loss": 0.6267,
      "step": 598
    },
    {
      "epoch": 54.484848484848484,
      "grad_norm": 0.30199798941612244,
      "learning_rate": 4.401e-05,
      "loss": 0.589,
      "step": 599
    },
    {
      "epoch": 54.58181818181818,
      "grad_norm": 0.2709701359272003,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6297,
      "step": 600
    },
    {
      "epoch": 54.58181818181818,
      "eval_loss": 0.6322523355484009,
      "eval_runtime": 2.1489,
      "eval_samples_per_second": 25.594,
      "eval_steps_per_second": 3.257,
      "step": 600
    },
    {
      "epoch": 54.67878787878788,
      "grad_norm": 0.3060154914855957,
      "learning_rate": 4.3990000000000004e-05,
      "loss": 0.5995,
      "step": 601
    },
    {
      "epoch": 54.775757575757574,
      "grad_norm": 0.30862957239151,
      "learning_rate": 4.398e-05,
      "loss": 0.6008,
      "step": 602
    },
    {
      "epoch": 54.872727272727275,
      "grad_norm": 0.32024699449539185,
      "learning_rate": 4.397e-05,
      "loss": 0.5772,
      "step": 603
    },
    {
      "epoch": 54.96969696969697,
      "grad_norm": 0.32759636640548706,
      "learning_rate": 4.396e-05,
      "loss": 0.6523,
      "step": 604
    },
    {
      "epoch": 55.0,
      "grad_norm": 0.39472898840904236,
      "learning_rate": 4.3950000000000004e-05,
      "loss": 0.5785,
      "step": 605
    },
    {
      "epoch": 55.096969696969694,
      "grad_norm": 0.24479424953460693,
      "learning_rate": 4.394e-05,
      "loss": 0.5679,
      "step": 606
    },
    {
      "epoch": 55.193939393939395,
      "grad_norm": 0.2995283007621765,
      "learning_rate": 4.393e-05,
      "loss": 0.6243,
      "step": 607
    },
    {
      "epoch": 55.29090909090909,
      "grad_norm": 0.29041364789009094,
      "learning_rate": 4.392e-05,
      "loss": 0.6334,
      "step": 608
    },
    {
      "epoch": 55.38787878787879,
      "grad_norm": 0.31214168667793274,
      "learning_rate": 4.391e-05,
      "loss": 0.6299,
      "step": 609
    },
    {
      "epoch": 55.484848484848484,
      "grad_norm": 0.2819170653820038,
      "learning_rate": 4.39e-05,
      "loss": 0.5911,
      "step": 610
    },
    {
      "epoch": 55.484848484848484,
      "eval_loss": 0.6271098852157593,
      "eval_runtime": 2.1464,
      "eval_samples_per_second": 25.625,
      "eval_steps_per_second": 3.261,
      "step": 610
    },
    {
      "epoch": 55.58181818181818,
      "grad_norm": 0.27817094326019287,
      "learning_rate": 4.389e-05,
      "loss": 0.5705,
      "step": 611
    },
    {
      "epoch": 55.67878787878788,
      "grad_norm": 0.25310567021369934,
      "learning_rate": 4.388000000000001e-05,
      "loss": 0.6376,
      "step": 612
    },
    {
      "epoch": 55.775757575757574,
      "grad_norm": 0.253987580537796,
      "learning_rate": 4.387e-05,
      "loss": 0.581,
      "step": 613
    },
    {
      "epoch": 55.872727272727275,
      "grad_norm": 0.2909943759441376,
      "learning_rate": 4.3860000000000004e-05,
      "loss": 0.581,
      "step": 614
    },
    {
      "epoch": 55.96969696969697,
      "grad_norm": 0.3695921301841736,
      "learning_rate": 4.385e-05,
      "loss": 0.6361,
      "step": 615
    },
    {
      "epoch": 56.0,
      "grad_norm": 0.4382748305797577,
      "learning_rate": 4.384e-05,
      "loss": 0.5501,
      "step": 616
    },
    {
      "epoch": 56.096969696969694,
      "grad_norm": 0.29598528146743774,
      "learning_rate": 4.3830000000000006e-05,
      "loss": 0.6328,
      "step": 617
    },
    {
      "epoch": 56.193939393939395,
      "grad_norm": 0.3592718541622162,
      "learning_rate": 4.382e-05,
      "loss": 0.5913,
      "step": 618
    },
    {
      "epoch": 56.29090909090909,
      "grad_norm": 0.30886024236679077,
      "learning_rate": 4.381e-05,
      "loss": 0.6639,
      "step": 619
    },
    {
      "epoch": 56.38787878787879,
      "grad_norm": 0.32863330841064453,
      "learning_rate": 4.38e-05,
      "loss": 0.5666,
      "step": 620
    },
    {
      "epoch": 56.38787878787879,
      "eval_loss": 0.6228953003883362,
      "eval_runtime": 2.1556,
      "eval_samples_per_second": 25.515,
      "eval_steps_per_second": 3.247,
      "step": 620
    },
    {
      "epoch": 56.484848484848484,
      "grad_norm": 0.3019230365753174,
      "learning_rate": 4.3790000000000006e-05,
      "loss": 0.5589,
      "step": 621
    },
    {
      "epoch": 56.58181818181818,
      "grad_norm": 0.30603864789009094,
      "learning_rate": 4.3780000000000004e-05,
      "loss": 0.5846,
      "step": 622
    },
    {
      "epoch": 56.67878787878788,
      "grad_norm": 0.3238994777202606,
      "learning_rate": 4.377e-05,
      "loss": 0.5891,
      "step": 623
    },
    {
      "epoch": 56.775757575757574,
      "grad_norm": 0.2910254895687103,
      "learning_rate": 4.376e-05,
      "loss": 0.561,
      "step": 624
    },
    {
      "epoch": 56.872727272727275,
      "grad_norm": 0.28653401136398315,
      "learning_rate": 4.375e-05,
      "loss": 0.6044,
      "step": 625
    },
    {
      "epoch": 56.96969696969697,
      "grad_norm": 0.28131261467933655,
      "learning_rate": 4.3740000000000005e-05,
      "loss": 0.6116,
      "step": 626
    },
    {
      "epoch": 57.0,
      "grad_norm": 0.4516832232475281,
      "learning_rate": 4.373e-05,
      "loss": 0.6444,
      "step": 627
    },
    {
      "epoch": 57.096969696969694,
      "grad_norm": 0.24059724807739258,
      "learning_rate": 4.372e-05,
      "loss": 0.608,
      "step": 628
    },
    {
      "epoch": 57.193939393939395,
      "grad_norm": 0.30921855568885803,
      "learning_rate": 4.371e-05,
      "loss": 0.559,
      "step": 629
    },
    {
      "epoch": 57.29090909090909,
      "grad_norm": 0.3092626929283142,
      "learning_rate": 4.3700000000000005e-05,
      "loss": 0.6223,
      "step": 630
    },
    {
      "epoch": 57.29090909090909,
      "eval_loss": 0.6180606484413147,
      "eval_runtime": 2.1784,
      "eval_samples_per_second": 25.248,
      "eval_steps_per_second": 3.213,
      "step": 630
    },
    {
      "epoch": 57.38787878787879,
      "grad_norm": 0.2981205880641937,
      "learning_rate": 4.3690000000000004e-05,
      "loss": 0.5294,
      "step": 631
    },
    {
      "epoch": 57.484848484848484,
      "grad_norm": 0.2812468707561493,
      "learning_rate": 4.368e-05,
      "loss": 0.6457,
      "step": 632
    },
    {
      "epoch": 57.58181818181818,
      "grad_norm": 0.24225398898124695,
      "learning_rate": 4.367e-05,
      "loss": 0.6397,
      "step": 633
    },
    {
      "epoch": 57.67878787878788,
      "grad_norm": 0.2751910388469696,
      "learning_rate": 4.366e-05,
      "loss": 0.5841,
      "step": 634
    },
    {
      "epoch": 57.775757575757574,
      "grad_norm": 0.2612401843070984,
      "learning_rate": 4.3650000000000004e-05,
      "loss": 0.5602,
      "step": 635
    },
    {
      "epoch": 57.872727272727275,
      "grad_norm": 0.27459433674812317,
      "learning_rate": 4.364e-05,
      "loss": 0.6028,
      "step": 636
    },
    {
      "epoch": 57.96969696969697,
      "grad_norm": 0.27292782068252563,
      "learning_rate": 4.363000000000001e-05,
      "loss": 0.5603,
      "step": 637
    },
    {
      "epoch": 58.0,
      "grad_norm": 0.3346533179283142,
      "learning_rate": 4.362e-05,
      "loss": 0.6219,
      "step": 638
    },
    {
      "epoch": 58.096969696969694,
      "grad_norm": 0.3203403949737549,
      "learning_rate": 4.361e-05,
      "loss": 0.6183,
      "step": 639
    },
    {
      "epoch": 58.193939393939395,
      "grad_norm": 0.27702099084854126,
      "learning_rate": 4.36e-05,
      "loss": 0.5564,
      "step": 640
    },
    {
      "epoch": 58.193939393939395,
      "eval_loss": 0.6141034960746765,
      "eval_runtime": 2.1576,
      "eval_samples_per_second": 25.491,
      "eval_steps_per_second": 3.244,
      "step": 640
    },
    {
      "epoch": 58.29090909090909,
      "grad_norm": 0.3335529863834381,
      "learning_rate": 4.359e-05,
      "loss": 0.6158,
      "step": 641
    },
    {
      "epoch": 58.38787878787879,
      "grad_norm": 0.2782766819000244,
      "learning_rate": 4.3580000000000006e-05,
      "loss": 0.5614,
      "step": 642
    },
    {
      "epoch": 58.484848484848484,
      "grad_norm": 0.2564246356487274,
      "learning_rate": 4.357e-05,
      "loss": 0.5689,
      "step": 643
    },
    {
      "epoch": 58.58181818181818,
      "grad_norm": 0.28409498929977417,
      "learning_rate": 4.356e-05,
      "loss": 0.5503,
      "step": 644
    },
    {
      "epoch": 58.67878787878788,
      "grad_norm": 0.28188493847846985,
      "learning_rate": 4.355e-05,
      "loss": 0.5662,
      "step": 645
    },
    {
      "epoch": 58.775757575757574,
      "grad_norm": 0.36505699157714844,
      "learning_rate": 4.354e-05,
      "loss": 0.6206,
      "step": 646
    },
    {
      "epoch": 58.872727272727275,
      "grad_norm": 0.2862570881843567,
      "learning_rate": 4.3530000000000005e-05,
      "loss": 0.6148,
      "step": 647
    },
    {
      "epoch": 58.96969696969697,
      "grad_norm": 0.26443371176719666,
      "learning_rate": 4.352e-05,
      "loss": 0.5892,
      "step": 648
    },
    {
      "epoch": 59.0,
      "grad_norm": 0.4998515844345093,
      "learning_rate": 4.351e-05,
      "loss": 0.6026,
      "step": 649
    },
    {
      "epoch": 59.096969696969694,
      "grad_norm": 0.28331199288368225,
      "learning_rate": 4.35e-05,
      "loss": 0.5927,
      "step": 650
    },
    {
      "epoch": 59.096969696969694,
      "eval_loss": 0.6094768643379211,
      "eval_runtime": 2.1683,
      "eval_samples_per_second": 25.366,
      "eval_steps_per_second": 3.228,
      "step": 650
    },
    {
      "epoch": 59.193939393939395,
      "grad_norm": 0.34278160333633423,
      "learning_rate": 4.3490000000000005e-05,
      "loss": 0.5871,
      "step": 651
    },
    {
      "epoch": 59.29090909090909,
      "grad_norm": 0.22800803184509277,
      "learning_rate": 4.3480000000000004e-05,
      "loss": 0.5934,
      "step": 652
    },
    {
      "epoch": 59.38787878787879,
      "grad_norm": 0.2810174822807312,
      "learning_rate": 4.347e-05,
      "loss": 0.5211,
      "step": 653
    },
    {
      "epoch": 59.484848484848484,
      "grad_norm": 0.31853628158569336,
      "learning_rate": 4.346e-05,
      "loss": 0.6027,
      "step": 654
    },
    {
      "epoch": 59.58181818181818,
      "grad_norm": 0.31586185097694397,
      "learning_rate": 4.345e-05,
      "loss": 0.6238,
      "step": 655
    },
    {
      "epoch": 59.67878787878788,
      "grad_norm": 0.29778969287872314,
      "learning_rate": 4.3440000000000004e-05,
      "loss": 0.5836,
      "step": 656
    },
    {
      "epoch": 59.775757575757574,
      "grad_norm": 0.25551241636276245,
      "learning_rate": 4.343e-05,
      "loss": 0.5566,
      "step": 657
    },
    {
      "epoch": 59.872727272727275,
      "grad_norm": 0.2709314227104187,
      "learning_rate": 4.342e-05,
      "loss": 0.5779,
      "step": 658
    },
    {
      "epoch": 59.96969696969697,
      "grad_norm": 0.24351809918880463,
      "learning_rate": 4.341e-05,
      "loss": 0.5947,
      "step": 659
    },
    {
      "epoch": 60.0,
      "grad_norm": 0.4438367486000061,
      "learning_rate": 4.3400000000000005e-05,
      "loss": 0.534,
      "step": 660
    },
    {
      "epoch": 60.0,
      "eval_loss": 0.605474591255188,
      "eval_runtime": 2.1507,
      "eval_samples_per_second": 25.574,
      "eval_steps_per_second": 3.255,
      "step": 660
    },
    {
      "epoch": 60.096969696969694,
      "grad_norm": 0.23856693506240845,
      "learning_rate": 4.339e-05,
      "loss": 0.5635,
      "step": 661
    },
    {
      "epoch": 60.193939393939395,
      "grad_norm": 0.32172146439552307,
      "learning_rate": 4.338e-05,
      "loss": 0.6082,
      "step": 662
    },
    {
      "epoch": 60.29090909090909,
      "grad_norm": 0.26954877376556396,
      "learning_rate": 4.337e-05,
      "loss": 0.5741,
      "step": 663
    },
    {
      "epoch": 60.38787878787879,
      "grad_norm": 0.37685203552246094,
      "learning_rate": 4.336e-05,
      "loss": 0.5493,
      "step": 664
    },
    {
      "epoch": 60.484848484848484,
      "grad_norm": 0.2673247456550598,
      "learning_rate": 4.335e-05,
      "loss": 0.6096,
      "step": 665
    },
    {
      "epoch": 60.58181818181818,
      "grad_norm": 0.2932179868221283,
      "learning_rate": 4.334e-05,
      "loss": 0.5901,
      "step": 666
    },
    {
      "epoch": 60.67878787878788,
      "grad_norm": 0.2641594409942627,
      "learning_rate": 4.333000000000001e-05,
      "loss": 0.5539,
      "step": 667
    },
    {
      "epoch": 60.775757575757574,
      "grad_norm": 0.26737964153289795,
      "learning_rate": 4.332e-05,
      "loss": 0.5452,
      "step": 668
    },
    {
      "epoch": 60.872727272727275,
      "grad_norm": 0.3418692946434021,
      "learning_rate": 4.3310000000000004e-05,
      "loss": 0.632,
      "step": 669
    },
    {
      "epoch": 60.96969696969697,
      "grad_norm": 0.37048324942588806,
      "learning_rate": 4.33e-05,
      "loss": 0.5266,
      "step": 670
    },
    {
      "epoch": 60.96969696969697,
      "eval_loss": 0.6022237539291382,
      "eval_runtime": 2.1773,
      "eval_samples_per_second": 25.261,
      "eval_steps_per_second": 3.215,
      "step": 670
    },
    {
      "epoch": 61.0,
      "grad_norm": 0.4496045410633087,
      "learning_rate": 4.329e-05,
      "loss": 0.6288,
      "step": 671
    },
    {
      "epoch": 61.096969696969694,
      "grad_norm": 0.2302197962999344,
      "learning_rate": 4.3280000000000006e-05,
      "loss": 0.5253,
      "step": 672
    },
    {
      "epoch": 61.193939393939395,
      "grad_norm": 0.2643318474292755,
      "learning_rate": 4.327e-05,
      "loss": 0.5668,
      "step": 673
    },
    {
      "epoch": 61.29090909090909,
      "grad_norm": 0.23105840384960175,
      "learning_rate": 4.326e-05,
      "loss": 0.6022,
      "step": 674
    },
    {
      "epoch": 61.38787878787879,
      "grad_norm": 0.2893432080745697,
      "learning_rate": 4.325e-05,
      "loss": 0.5638,
      "step": 675
    },
    {
      "epoch": 61.484848484848484,
      "grad_norm": 0.37326863408088684,
      "learning_rate": 4.324e-05,
      "loss": 0.5725,
      "step": 676
    },
    {
      "epoch": 61.58181818181818,
      "grad_norm": 0.26676344871520996,
      "learning_rate": 4.3230000000000005e-05,
      "loss": 0.6257,
      "step": 677
    },
    {
      "epoch": 61.67878787878788,
      "grad_norm": 0.25675010681152344,
      "learning_rate": 4.3219999999999996e-05,
      "loss": 0.5766,
      "step": 678
    },
    {
      "epoch": 61.775757575757574,
      "grad_norm": 0.228947252035141,
      "learning_rate": 4.321e-05,
      "loss": 0.5876,
      "step": 679
    },
    {
      "epoch": 61.872727272727275,
      "grad_norm": 0.22656583786010742,
      "learning_rate": 4.32e-05,
      "loss": 0.5509,
      "step": 680
    },
    {
      "epoch": 61.872727272727275,
      "eval_loss": 0.5982839465141296,
      "eval_runtime": 2.1524,
      "eval_samples_per_second": 25.552,
      "eval_steps_per_second": 3.252,
      "step": 680
    },
    {
      "epoch": 61.96969696969697,
      "grad_norm": 0.2663673460483551,
      "learning_rate": 4.3190000000000005e-05,
      "loss": 0.5563,
      "step": 681
    },
    {
      "epoch": 62.0,
      "grad_norm": 0.35887378454208374,
      "learning_rate": 4.318e-05,
      "loss": 0.5491,
      "step": 682
    },
    {
      "epoch": 62.096969696969694,
      "grad_norm": 0.2571672797203064,
      "learning_rate": 4.317e-05,
      "loss": 0.5039,
      "step": 683
    },
    {
      "epoch": 62.193939393939395,
      "grad_norm": 0.2171117067337036,
      "learning_rate": 4.316e-05,
      "loss": 0.5455,
      "step": 684
    },
    {
      "epoch": 62.29090909090909,
      "grad_norm": 0.2166251391172409,
      "learning_rate": 4.315e-05,
      "loss": 0.5581,
      "step": 685
    },
    {
      "epoch": 62.38787878787879,
      "grad_norm": 0.2635202407836914,
      "learning_rate": 4.3140000000000004e-05,
      "loss": 0.5444,
      "step": 686
    },
    {
      "epoch": 62.484848484848484,
      "grad_norm": 0.28910040855407715,
      "learning_rate": 4.313e-05,
      "loss": 0.5821,
      "step": 687
    },
    {
      "epoch": 62.58181818181818,
      "grad_norm": 0.30902335047721863,
      "learning_rate": 4.312000000000001e-05,
      "loss": 0.6414,
      "step": 688
    },
    {
      "epoch": 62.67878787878788,
      "grad_norm": 0.29418280720710754,
      "learning_rate": 4.311e-05,
      "loss": 0.6108,
      "step": 689
    },
    {
      "epoch": 62.775757575757574,
      "grad_norm": 0.3029976785182953,
      "learning_rate": 4.3100000000000004e-05,
      "loss": 0.5546,
      "step": 690
    },
    {
      "epoch": 62.775757575757574,
      "eval_loss": 0.5950291752815247,
      "eval_runtime": 2.1501,
      "eval_samples_per_second": 25.58,
      "eval_steps_per_second": 3.256,
      "step": 690
    },
    {
      "epoch": 62.872727272727275,
      "grad_norm": 0.31214746832847595,
      "learning_rate": 4.309e-05,
      "loss": 0.5521,
      "step": 691
    },
    {
      "epoch": 62.96969696969697,
      "grad_norm": 0.2938581705093384,
      "learning_rate": 4.308e-05,
      "loss": 0.5875,
      "step": 692
    },
    {
      "epoch": 63.0,
      "grad_norm": 0.34210625290870667,
      "learning_rate": 4.3070000000000006e-05,
      "loss": 0.5625,
      "step": 693
    },
    {
      "epoch": 63.096969696969694,
      "grad_norm": 0.26499512791633606,
      "learning_rate": 4.306e-05,
      "loss": 0.5297,
      "step": 694
    },
    {
      "epoch": 63.193939393939395,
      "grad_norm": 0.24259193241596222,
      "learning_rate": 4.305e-05,
      "loss": 0.5309,
      "step": 695
    },
    {
      "epoch": 63.29090909090909,
      "grad_norm": 0.24082423746585846,
      "learning_rate": 4.304e-05,
      "loss": 0.5689,
      "step": 696
    },
    {
      "epoch": 63.38787878787879,
      "grad_norm": 0.2728228271007538,
      "learning_rate": 4.3030000000000006e-05,
      "loss": 0.5717,
      "step": 697
    },
    {
      "epoch": 63.484848484848484,
      "grad_norm": 0.2863039970397949,
      "learning_rate": 4.3020000000000005e-05,
      "loss": 0.5788,
      "step": 698
    },
    {
      "epoch": 63.58181818181818,
      "grad_norm": 0.24509799480438232,
      "learning_rate": 4.301e-05,
      "loss": 0.6085,
      "step": 699
    },
    {
      "epoch": 63.67878787878788,
      "grad_norm": 0.2451586127281189,
      "learning_rate": 4.3e-05,
      "loss": 0.551,
      "step": 700
    },
    {
      "epoch": 63.67878787878788,
      "eval_loss": 0.5922297835350037,
      "eval_runtime": 2.1427,
      "eval_samples_per_second": 25.669,
      "eval_steps_per_second": 3.267,
      "step": 700
    },
    {
      "epoch": 63.775757575757574,
      "grad_norm": 0.2422385960817337,
      "learning_rate": 4.299e-05,
      "loss": 0.5275,
      "step": 701
    },
    {
      "epoch": 63.872727272727275,
      "grad_norm": 0.34540262818336487,
      "learning_rate": 4.2980000000000005e-05,
      "loss": 0.5706,
      "step": 702
    },
    {
      "epoch": 63.96969696969697,
      "grad_norm": 0.19919060170650482,
      "learning_rate": 4.2970000000000004e-05,
      "loss": 0.5926,
      "step": 703
    },
    {
      "epoch": 64.0,
      "grad_norm": 0.3330257833003998,
      "learning_rate": 4.296e-05,
      "loss": 0.601,
      "step": 704
    },
    {
      "epoch": 64.0969696969697,
      "grad_norm": 0.25846561789512634,
      "learning_rate": 4.295e-05,
      "loss": 0.5277,
      "step": 705
    },
    {
      "epoch": 64.19393939393939,
      "grad_norm": 0.31724169850349426,
      "learning_rate": 4.2940000000000006e-05,
      "loss": 0.5923,
      "step": 706
    },
    {
      "epoch": 64.2909090909091,
      "grad_norm": 0.23587939143180847,
      "learning_rate": 4.2930000000000004e-05,
      "loss": 0.5633,
      "step": 707
    },
    {
      "epoch": 64.38787878787879,
      "grad_norm": 0.2762688398361206,
      "learning_rate": 4.292e-05,
      "loss": 0.5784,
      "step": 708
    },
    {
      "epoch": 64.48484848484848,
      "grad_norm": 0.23367071151733398,
      "learning_rate": 4.291e-05,
      "loss": 0.6072,
      "step": 709
    },
    {
      "epoch": 64.58181818181818,
      "grad_norm": 0.2732716500759125,
      "learning_rate": 4.29e-05,
      "loss": 0.533,
      "step": 710
    },
    {
      "epoch": 64.58181818181818,
      "eval_loss": 0.5896597504615784,
      "eval_runtime": 2.1619,
      "eval_samples_per_second": 25.441,
      "eval_steps_per_second": 3.238,
      "step": 710
    },
    {
      "epoch": 64.67878787878787,
      "grad_norm": 0.22921384871006012,
      "learning_rate": 4.2890000000000004e-05,
      "loss": 0.5534,
      "step": 711
    },
    {
      "epoch": 64.77575757575758,
      "grad_norm": 0.27668488025665283,
      "learning_rate": 4.288e-05,
      "loss": 0.5378,
      "step": 712
    },
    {
      "epoch": 64.87272727272727,
      "grad_norm": 0.2813219428062439,
      "learning_rate": 4.287000000000001e-05,
      "loss": 0.5711,
      "step": 713
    },
    {
      "epoch": 64.96969696969697,
      "grad_norm": 0.25303152203559875,
      "learning_rate": 4.286e-05,
      "loss": 0.5697,
      "step": 714
    },
    {
      "epoch": 65.0,
      "grad_norm": 0.47778254747390747,
      "learning_rate": 4.285e-05,
      "loss": 0.4676,
      "step": 715
    },
    {
      "epoch": 65.0969696969697,
      "grad_norm": 0.24541884660720825,
      "learning_rate": 4.284e-05,
      "loss": 0.6171,
      "step": 716
    },
    {
      "epoch": 65.19393939393939,
      "grad_norm": 0.2891412079334259,
      "learning_rate": 4.283e-05,
      "loss": 0.5985,
      "step": 717
    },
    {
      "epoch": 65.2909090909091,
      "grad_norm": 0.2899450957775116,
      "learning_rate": 4.282000000000001e-05,
      "loss": 0.5589,
      "step": 718
    },
    {
      "epoch": 65.38787878787879,
      "grad_norm": 0.2739652395248413,
      "learning_rate": 4.281e-05,
      "loss": 0.5468,
      "step": 719
    },
    {
      "epoch": 65.48484848484848,
      "grad_norm": 0.24000035226345062,
      "learning_rate": 4.2800000000000004e-05,
      "loss": 0.5106,
      "step": 720
    },
    {
      "epoch": 65.48484848484848,
      "eval_loss": 0.5866174697875977,
      "eval_runtime": 2.1673,
      "eval_samples_per_second": 25.377,
      "eval_steps_per_second": 3.23,
      "step": 720
    },
    {
      "epoch": 65.58181818181818,
      "grad_norm": 0.2736685872077942,
      "learning_rate": 4.279e-05,
      "loss": 0.5149,
      "step": 721
    },
    {
      "epoch": 65.67878787878787,
      "grad_norm": 0.3003578782081604,
      "learning_rate": 4.278e-05,
      "loss": 0.567,
      "step": 722
    },
    {
      "epoch": 65.77575757575758,
      "grad_norm": 0.2917529046535492,
      "learning_rate": 4.2770000000000006e-05,
      "loss": 0.571,
      "step": 723
    },
    {
      "epoch": 65.87272727272727,
      "grad_norm": 0.20763424038887024,
      "learning_rate": 4.276e-05,
      "loss": 0.5544,
      "step": 724
    },
    {
      "epoch": 65.96969696969697,
      "grad_norm": 0.214829683303833,
      "learning_rate": 4.275e-05,
      "loss": 0.5496,
      "step": 725
    },
    {
      "epoch": 66.0,
      "grad_norm": 0.39429301023483276,
      "learning_rate": 4.274e-05,
      "loss": 0.496,
      "step": 726
    },
    {
      "epoch": 66.0969696969697,
      "grad_norm": 0.24492646753787994,
      "learning_rate": 4.2730000000000006e-05,
      "loss": 0.5959,
      "step": 727
    },
    {
      "epoch": 66.19393939393939,
      "grad_norm": 0.27110999822616577,
      "learning_rate": 4.2720000000000004e-05,
      "loss": 0.5544,
      "step": 728
    },
    {
      "epoch": 66.2909090909091,
      "grad_norm": 0.21391001343727112,
      "learning_rate": 4.271e-05,
      "loss": 0.4939,
      "step": 729
    },
    {
      "epoch": 66.38787878787879,
      "grad_norm": 0.2673359215259552,
      "learning_rate": 4.27e-05,
      "loss": 0.619,
      "step": 730
    },
    {
      "epoch": 66.38787878787879,
      "eval_loss": 0.5843937993049622,
      "eval_runtime": 2.1541,
      "eval_samples_per_second": 25.533,
      "eval_steps_per_second": 3.25,
      "step": 730
    },
    {
      "epoch": 66.48484848484848,
      "grad_norm": 0.34252795577049255,
      "learning_rate": 4.269e-05,
      "loss": 0.5253,
      "step": 731
    },
    {
      "epoch": 66.58181818181818,
      "grad_norm": 0.21049198508262634,
      "learning_rate": 4.2680000000000005e-05,
      "loss": 0.5423,
      "step": 732
    },
    {
      "epoch": 66.67878787878787,
      "grad_norm": 0.213708758354187,
      "learning_rate": 4.267e-05,
      "loss": 0.5479,
      "step": 733
    },
    {
      "epoch": 66.77575757575758,
      "grad_norm": 0.2658591568470001,
      "learning_rate": 4.266e-05,
      "loss": 0.5763,
      "step": 734
    },
    {
      "epoch": 66.87272727272727,
      "grad_norm": 0.2451457381248474,
      "learning_rate": 4.265e-05,
      "loss": 0.5478,
      "step": 735
    },
    {
      "epoch": 66.96969696969697,
      "grad_norm": 0.28571853041648865,
      "learning_rate": 4.2640000000000005e-05,
      "loss": 0.5563,
      "step": 736
    },
    {
      "epoch": 67.0,
      "grad_norm": 0.4764930009841919,
      "learning_rate": 4.2630000000000004e-05,
      "loss": 0.4799,
      "step": 737
    },
    {
      "epoch": 67.0969696969697,
      "grad_norm": 0.21275725960731506,
      "learning_rate": 4.262e-05,
      "loss": 0.563,
      "step": 738
    },
    {
      "epoch": 67.19393939393939,
      "grad_norm": 0.24314363300800323,
      "learning_rate": 4.261e-05,
      "loss": 0.5669,
      "step": 739
    },
    {
      "epoch": 67.2909090909091,
      "grad_norm": 0.23035164177417755,
      "learning_rate": 4.26e-05,
      "loss": 0.5501,
      "step": 740
    },
    {
      "epoch": 67.2909090909091,
      "eval_loss": 0.5817498564720154,
      "eval_runtime": 2.1583,
      "eval_samples_per_second": 25.483,
      "eval_steps_per_second": 3.243,
      "step": 740
    },
    {
      "epoch": 67.38787878787879,
      "grad_norm": 0.2700410485267639,
      "learning_rate": 4.2590000000000004e-05,
      "loss": 0.5613,
      "step": 741
    },
    {
      "epoch": 67.48484848484848,
      "grad_norm": 0.35602664947509766,
      "learning_rate": 4.258e-05,
      "loss": 0.5576,
      "step": 742
    },
    {
      "epoch": 67.58181818181818,
      "grad_norm": 0.26432085037231445,
      "learning_rate": 4.257000000000001e-05,
      "loss": 0.5495,
      "step": 743
    },
    {
      "epoch": 67.67878787878787,
      "grad_norm": 0.2376573383808136,
      "learning_rate": 4.256e-05,
      "loss": 0.5318,
      "step": 744
    },
    {
      "epoch": 67.77575757575758,
      "grad_norm": 0.3016679883003235,
      "learning_rate": 4.2550000000000004e-05,
      "loss": 0.5655,
      "step": 745
    },
    {
      "epoch": 67.87272727272727,
      "grad_norm": 0.2729600965976715,
      "learning_rate": 4.254e-05,
      "loss": 0.4898,
      "step": 746
    },
    {
      "epoch": 67.96969696969697,
      "grad_norm": 0.3013279139995575,
      "learning_rate": 4.253e-05,
      "loss": 0.5625,
      "step": 747
    },
    {
      "epoch": 68.0,
      "grad_norm": 0.4911724030971527,
      "learning_rate": 4.2520000000000006e-05,
      "loss": 0.5731,
      "step": 748
    },
    {
      "epoch": 68.0969696969697,
      "grad_norm": 0.23142361640930176,
      "learning_rate": 4.251e-05,
      "loss": 0.5799,
      "step": 749
    },
    {
      "epoch": 68.19393939393939,
      "grad_norm": 0.22808052599430084,
      "learning_rate": 4.25e-05,
      "loss": 0.5393,
      "step": 750
    },
    {
      "epoch": 68.19393939393939,
      "eval_loss": 0.5790581703186035,
      "eval_runtime": 2.17,
      "eval_samples_per_second": 25.346,
      "eval_steps_per_second": 3.226,
      "step": 750
    },
    {
      "epoch": 68.2909090909091,
      "grad_norm": 0.23808452486991882,
      "learning_rate": 4.249e-05,
      "loss": 0.5646,
      "step": 751
    },
    {
      "epoch": 68.38787878787879,
      "grad_norm": 0.2526884973049164,
      "learning_rate": 4.248e-05,
      "loss": 0.5756,
      "step": 752
    },
    {
      "epoch": 68.48484848484848,
      "grad_norm": 0.2327769547700882,
      "learning_rate": 4.2470000000000005e-05,
      "loss": 0.5109,
      "step": 753
    },
    {
      "epoch": 68.58181818181818,
      "grad_norm": 0.2696152627468109,
      "learning_rate": 4.246e-05,
      "loss": 0.5549,
      "step": 754
    },
    {
      "epoch": 68.67878787878787,
      "grad_norm": 0.33442422747612,
      "learning_rate": 4.245e-05,
      "loss": 0.5413,
      "step": 755
    },
    {
      "epoch": 68.77575757575758,
      "grad_norm": 0.23822557926177979,
      "learning_rate": 4.244e-05,
      "loss": 0.5018,
      "step": 756
    },
    {
      "epoch": 68.87272727272727,
      "grad_norm": 0.26610106229782104,
      "learning_rate": 4.2430000000000005e-05,
      "loss": 0.5691,
      "step": 757
    },
    {
      "epoch": 68.96969696969697,
      "grad_norm": 0.3304058909416199,
      "learning_rate": 4.2420000000000004e-05,
      "loss": 0.5273,
      "step": 758
    },
    {
      "epoch": 69.0,
      "grad_norm": 0.485503613948822,
      "learning_rate": 4.241e-05,
      "loss": 0.568,
      "step": 759
    },
    {
      "epoch": 69.0969696969697,
      "grad_norm": 0.20574645698070526,
      "learning_rate": 4.24e-05,
      "loss": 0.5653,
      "step": 760
    },
    {
      "epoch": 69.0969696969697,
      "eval_loss": 0.5764835476875305,
      "eval_runtime": 2.1761,
      "eval_samples_per_second": 25.275,
      "eval_steps_per_second": 3.217,
      "step": 760
    },
    {
      "epoch": 69.19393939393939,
      "grad_norm": 0.2699494957923889,
      "learning_rate": 4.239e-05,
      "loss": 0.516,
      "step": 761
    },
    {
      "epoch": 69.2909090909091,
      "grad_norm": 0.24947266280651093,
      "learning_rate": 4.2380000000000004e-05,
      "loss": 0.5476,
      "step": 762
    },
    {
      "epoch": 69.38787878787879,
      "grad_norm": 0.2272285521030426,
      "learning_rate": 4.237e-05,
      "loss": 0.5626,
      "step": 763
    },
    {
      "epoch": 69.48484848484848,
      "grad_norm": 0.24354349076747894,
      "learning_rate": 4.236e-05,
      "loss": 0.5629,
      "step": 764
    },
    {
      "epoch": 69.58181818181818,
      "grad_norm": 0.27001243829727173,
      "learning_rate": 4.235e-05,
      "loss": 0.5477,
      "step": 765
    },
    {
      "epoch": 69.67878787878787,
      "grad_norm": 0.23391012847423553,
      "learning_rate": 4.2340000000000005e-05,
      "loss": 0.5328,
      "step": 766
    },
    {
      "epoch": 69.77575757575758,
      "grad_norm": 0.24374918639659882,
      "learning_rate": 4.233e-05,
      "loss": 0.5185,
      "step": 767
    },
    {
      "epoch": 69.87272727272727,
      "grad_norm": 0.2527277171611786,
      "learning_rate": 4.232e-05,
      "loss": 0.5759,
      "step": 768
    },
    {
      "epoch": 69.96969696969697,
      "grad_norm": 0.26951363682746887,
      "learning_rate": 4.231e-05,
      "loss": 0.5184,
      "step": 769
    },
    {
      "epoch": 70.0,
      "grad_norm": 0.281485378742218,
      "learning_rate": 4.23e-05,
      "loss": 0.5212,
      "step": 770
    },
    {
      "epoch": 70.0,
      "eval_loss": 0.5742422342300415,
      "eval_runtime": 2.1603,
      "eval_samples_per_second": 25.46,
      "eval_steps_per_second": 3.24,
      "step": 770
    },
    {
      "epoch": 70.0969696969697,
      "grad_norm": 0.2688864469528198,
      "learning_rate": 4.229e-05,
      "loss": 0.5516,
      "step": 771
    },
    {
      "epoch": 70.19393939393939,
      "grad_norm": 0.2755700349807739,
      "learning_rate": 4.228e-05,
      "loss": 0.512,
      "step": 772
    },
    {
      "epoch": 70.2909090909091,
      "grad_norm": 0.2601207196712494,
      "learning_rate": 4.227000000000001e-05,
      "loss": 0.5696,
      "step": 773
    },
    {
      "epoch": 70.38787878787879,
      "grad_norm": 0.25220006704330444,
      "learning_rate": 4.226e-05,
      "loss": 0.5386,
      "step": 774
    },
    {
      "epoch": 70.48484848484848,
      "grad_norm": 0.31067171692848206,
      "learning_rate": 4.2250000000000004e-05,
      "loss": 0.5292,
      "step": 775
    },
    {
      "epoch": 70.58181818181818,
      "grad_norm": 0.30006277561187744,
      "learning_rate": 4.224e-05,
      "loss": 0.5568,
      "step": 776
    },
    {
      "epoch": 70.67878787878787,
      "grad_norm": 0.2238987237215042,
      "learning_rate": 4.223e-05,
      "loss": 0.5705,
      "step": 777
    },
    {
      "epoch": 70.77575757575758,
      "grad_norm": 0.25496816635131836,
      "learning_rate": 4.2220000000000006e-05,
      "loss": 0.5572,
      "step": 778
    },
    {
      "epoch": 70.87272727272727,
      "grad_norm": 0.21116836369037628,
      "learning_rate": 4.221e-05,
      "loss": 0.5138,
      "step": 779
    },
    {
      "epoch": 70.96969696969697,
      "grad_norm": 0.25126564502716064,
      "learning_rate": 4.22e-05,
      "loss": 0.5241,
      "step": 780
    },
    {
      "epoch": 70.96969696969697,
      "eval_loss": 0.5723573565483093,
      "eval_runtime": 2.1754,
      "eval_samples_per_second": 25.283,
      "eval_steps_per_second": 3.218,
      "step": 780
    },
    {
      "epoch": 71.0,
      "grad_norm": 0.36982303857803345,
      "learning_rate": 4.219e-05,
      "loss": 0.5135,
      "step": 781
    },
    {
      "epoch": 71.0969696969697,
      "grad_norm": 0.2597285211086273,
      "learning_rate": 4.2180000000000006e-05,
      "loss": 0.5315,
      "step": 782
    },
    {
      "epoch": 71.19393939393939,
      "grad_norm": 0.2919541895389557,
      "learning_rate": 4.2170000000000005e-05,
      "loss": 0.5667,
      "step": 783
    },
    {
      "epoch": 71.2909090909091,
      "grad_norm": 0.26001623272895813,
      "learning_rate": 4.2159999999999996e-05,
      "loss": 0.5545,
      "step": 784
    },
    {
      "epoch": 71.38787878787879,
      "grad_norm": 0.25698697566986084,
      "learning_rate": 4.215e-05,
      "loss": 0.5441,
      "step": 785
    },
    {
      "epoch": 71.48484848484848,
      "grad_norm": 0.2391366958618164,
      "learning_rate": 4.214e-05,
      "loss": 0.4832,
      "step": 786
    },
    {
      "epoch": 71.58181818181818,
      "grad_norm": 0.191857248544693,
      "learning_rate": 4.2130000000000005e-05,
      "loss": 0.546,
      "step": 787
    },
    {
      "epoch": 71.67878787878787,
      "grad_norm": 0.27090173959732056,
      "learning_rate": 4.212e-05,
      "loss": 0.5516,
      "step": 788
    },
    {
      "epoch": 71.77575757575758,
      "grad_norm": 0.2738029956817627,
      "learning_rate": 4.211e-05,
      "loss": 0.5072,
      "step": 789
    },
    {
      "epoch": 71.87272727272727,
      "grad_norm": 0.320273220539093,
      "learning_rate": 4.21e-05,
      "loss": 0.5563,
      "step": 790
    },
    {
      "epoch": 71.87272727272727,
      "eval_loss": 0.5696360468864441,
      "eval_runtime": 2.2051,
      "eval_samples_per_second": 24.942,
      "eval_steps_per_second": 3.174,
      "step": 790
    },
    {
      "epoch": 71.96969696969697,
      "grad_norm": 0.26015132665634155,
      "learning_rate": 4.209e-05,
      "loss": 0.5632,
      "step": 791
    },
    {
      "epoch": 72.0,
      "grad_norm": 0.41450807452201843,
      "learning_rate": 4.2080000000000004e-05,
      "loss": 0.4652,
      "step": 792
    },
    {
      "epoch": 72.0969696969697,
      "grad_norm": 0.24714349210262299,
      "learning_rate": 4.207e-05,
      "loss": 0.5923,
      "step": 793
    },
    {
      "epoch": 72.19393939393939,
      "grad_norm": 0.2587253451347351,
      "learning_rate": 4.206e-05,
      "loss": 0.5267,
      "step": 794
    },
    {
      "epoch": 72.2909090909091,
      "grad_norm": 0.23800162971019745,
      "learning_rate": 4.205e-05,
      "loss": 0.5081,
      "step": 795
    },
    {
      "epoch": 72.38787878787879,
      "grad_norm": 0.2461039423942566,
      "learning_rate": 4.2040000000000004e-05,
      "loss": 0.5792,
      "step": 796
    },
    {
      "epoch": 72.48484848484848,
      "grad_norm": 0.2287796139717102,
      "learning_rate": 4.203e-05,
      "loss": 0.554,
      "step": 797
    },
    {
      "epoch": 72.58181818181818,
      "grad_norm": 0.24021528661251068,
      "learning_rate": 4.202e-05,
      "loss": 0.4999,
      "step": 798
    },
    {
      "epoch": 72.67878787878787,
      "grad_norm": 0.2695189416408539,
      "learning_rate": 4.201e-05,
      "loss": 0.5849,
      "step": 799
    },
    {
      "epoch": 72.77575757575758,
      "grad_norm": 0.21474483609199524,
      "learning_rate": 4.2e-05,
      "loss": 0.5188,
      "step": 800
    },
    {
      "epoch": 72.77575757575758,
      "eval_loss": 0.567756175994873,
      "eval_runtime": 2.1668,
      "eval_samples_per_second": 25.383,
      "eval_steps_per_second": 3.231,
      "step": 800
    },
    {
      "epoch": 72.87272727272727,
      "grad_norm": 0.3127596378326416,
      "learning_rate": 4.199e-05,
      "loss": 0.5497,
      "step": 801
    },
    {
      "epoch": 72.96969696969697,
      "grad_norm": 0.2873506546020508,
      "learning_rate": 4.198e-05,
      "loss": 0.481,
      "step": 802
    },
    {
      "epoch": 73.0,
      "grad_norm": 0.3801302909851074,
      "learning_rate": 4.1970000000000006e-05,
      "loss": 0.4043,
      "step": 803
    },
    {
      "epoch": 73.0969696969697,
      "grad_norm": 0.3030167818069458,
      "learning_rate": 4.196e-05,
      "loss": 0.4762,
      "step": 804
    },
    {
      "epoch": 73.19393939393939,
      "grad_norm": 0.2462480664253235,
      "learning_rate": 4.195e-05,
      "loss": 0.5248,
      "step": 805
    },
    {
      "epoch": 73.2909090909091,
      "grad_norm": 0.23730331659317017,
      "learning_rate": 4.194e-05,
      "loss": 0.5396,
      "step": 806
    },
    {
      "epoch": 73.38787878787879,
      "grad_norm": 0.2629089057445526,
      "learning_rate": 4.193e-05,
      "loss": 0.5263,
      "step": 807
    },
    {
      "epoch": 73.48484848484848,
      "grad_norm": 0.27913814783096313,
      "learning_rate": 4.1920000000000005e-05,
      "loss": 0.5425,
      "step": 808
    },
    {
      "epoch": 73.58181818181818,
      "grad_norm": 0.289802610874176,
      "learning_rate": 4.191e-05,
      "loss": 0.5229,
      "step": 809
    },
    {
      "epoch": 73.67878787878787,
      "grad_norm": 0.31583303213119507,
      "learning_rate": 4.19e-05,
      "loss": 0.5544,
      "step": 810
    },
    {
      "epoch": 73.67878787878787,
      "eval_loss": 0.5661746859550476,
      "eval_runtime": 2.1844,
      "eval_samples_per_second": 25.178,
      "eval_steps_per_second": 3.204,
      "step": 810
    },
    {
      "epoch": 73.77575757575758,
      "grad_norm": 0.256304532289505,
      "learning_rate": 4.189e-05,
      "loss": 0.5881,
      "step": 811
    },
    {
      "epoch": 73.87272727272727,
      "grad_norm": 0.2496311515569687,
      "learning_rate": 4.1880000000000006e-05,
      "loss": 0.5106,
      "step": 812
    },
    {
      "epoch": 73.96969696969697,
      "grad_norm": 0.24245764315128326,
      "learning_rate": 4.1870000000000004e-05,
      "loss": 0.5596,
      "step": 813
    },
    {
      "epoch": 74.0,
      "grad_norm": 0.4063602089881897,
      "learning_rate": 4.186e-05,
      "loss": 0.4897,
      "step": 814
    },
    {
      "epoch": 74.0969696969697,
      "grad_norm": 0.2712928056716919,
      "learning_rate": 4.185e-05,
      "loss": 0.581,
      "step": 815
    },
    {
      "epoch": 74.19393939393939,
      "grad_norm": 0.25334519147872925,
      "learning_rate": 4.184e-05,
      "loss": 0.5512,
      "step": 816
    },
    {
      "epoch": 74.2909090909091,
      "grad_norm": 0.23247885704040527,
      "learning_rate": 4.1830000000000004e-05,
      "loss": 0.5292,
      "step": 817
    },
    {
      "epoch": 74.38787878787879,
      "grad_norm": 0.2378424108028412,
      "learning_rate": 4.182e-05,
      "loss": 0.511,
      "step": 818
    },
    {
      "epoch": 74.48484848484848,
      "grad_norm": 0.27492162585258484,
      "learning_rate": 4.181000000000001e-05,
      "loss": 0.5589,
      "step": 819
    },
    {
      "epoch": 74.58181818181818,
      "grad_norm": 0.24575957655906677,
      "learning_rate": 4.18e-05,
      "loss": 0.5062,
      "step": 820
    },
    {
      "epoch": 74.58181818181818,
      "eval_loss": 0.5639890432357788,
      "eval_runtime": 2.1559,
      "eval_samples_per_second": 25.512,
      "eval_steps_per_second": 3.247,
      "step": 820
    },
    {
      "epoch": 74.67878787878787,
      "grad_norm": 0.19967587292194366,
      "learning_rate": 4.179e-05,
      "loss": 0.5208,
      "step": 821
    },
    {
      "epoch": 74.77575757575758,
      "grad_norm": 0.22234216332435608,
      "learning_rate": 4.178e-05,
      "loss": 0.5334,
      "step": 822
    },
    {
      "epoch": 74.87272727272727,
      "grad_norm": 0.25841695070266724,
      "learning_rate": 4.177e-05,
      "loss": 0.5316,
      "step": 823
    },
    {
      "epoch": 74.96969696969697,
      "grad_norm": 0.21328817307949066,
      "learning_rate": 4.176000000000001e-05,
      "loss": 0.5106,
      "step": 824
    },
    {
      "epoch": 75.0,
      "grad_norm": 0.4093509316444397,
      "learning_rate": 4.175e-05,
      "loss": 0.4296,
      "step": 825
    },
    {
      "epoch": 75.0969696969697,
      "grad_norm": 0.26131540536880493,
      "learning_rate": 4.1740000000000004e-05,
      "loss": 0.5628,
      "step": 826
    },
    {
      "epoch": 75.19393939393939,
      "grad_norm": 0.28853708505630493,
      "learning_rate": 4.173e-05,
      "loss": 0.5622,
      "step": 827
    },
    {
      "epoch": 75.2909090909091,
      "grad_norm": 0.2535465657711029,
      "learning_rate": 4.172e-05,
      "loss": 0.5123,
      "step": 828
    },
    {
      "epoch": 75.38787878787879,
      "grad_norm": 0.2491854429244995,
      "learning_rate": 4.1710000000000006e-05,
      "loss": 0.5434,
      "step": 829
    },
    {
      "epoch": 75.48484848484848,
      "grad_norm": 0.24380576610565186,
      "learning_rate": 4.17e-05,
      "loss": 0.5345,
      "step": 830
    },
    {
      "epoch": 75.48484848484848,
      "eval_loss": 0.5620233416557312,
      "eval_runtime": 2.1695,
      "eval_samples_per_second": 25.352,
      "eval_steps_per_second": 3.227,
      "step": 830
    },
    {
      "epoch": 75.58181818181818,
      "grad_norm": 0.22149287164211273,
      "learning_rate": 4.169e-05,
      "loss": 0.5565,
      "step": 831
    },
    {
      "epoch": 75.67878787878787,
      "grad_norm": 0.2667548954486847,
      "learning_rate": 4.168e-05,
      "loss": 0.4894,
      "step": 832
    },
    {
      "epoch": 75.77575757575758,
      "grad_norm": 0.28772029280662537,
      "learning_rate": 4.1670000000000006e-05,
      "loss": 0.4504,
      "step": 833
    },
    {
      "epoch": 75.87272727272727,
      "grad_norm": 0.1839408576488495,
      "learning_rate": 4.1660000000000004e-05,
      "loss": 0.4976,
      "step": 834
    },
    {
      "epoch": 75.96969696969697,
      "grad_norm": 0.2373735010623932,
      "learning_rate": 4.165e-05,
      "loss": 0.5722,
      "step": 835
    },
    {
      "epoch": 76.0,
      "grad_norm": 0.37163248658180237,
      "learning_rate": 4.164e-05,
      "loss": 0.5103,
      "step": 836
    },
    {
      "epoch": 76.0969696969697,
      "grad_norm": 0.25069040060043335,
      "learning_rate": 4.163e-05,
      "loss": 0.5518,
      "step": 837
    },
    {
      "epoch": 76.19393939393939,
      "grad_norm": 0.2810257077217102,
      "learning_rate": 4.1620000000000005e-05,
      "loss": 0.5277,
      "step": 838
    },
    {
      "epoch": 76.2909090909091,
      "grad_norm": 0.24961601197719574,
      "learning_rate": 4.161e-05,
      "loss": 0.5557,
      "step": 839
    },
    {
      "epoch": 76.38787878787879,
      "grad_norm": 0.18921363353729248,
      "learning_rate": 4.16e-05,
      "loss": 0.5246,
      "step": 840
    },
    {
      "epoch": 76.38787878787879,
      "eval_loss": 0.5598552823066711,
      "eval_runtime": 2.1634,
      "eval_samples_per_second": 25.422,
      "eval_steps_per_second": 3.236,
      "step": 840
    },
    {
      "epoch": 76.48484848484848,
      "grad_norm": 0.20288659632205963,
      "learning_rate": 4.159e-05,
      "loss": 0.5274,
      "step": 841
    },
    {
      "epoch": 76.58181818181818,
      "grad_norm": 0.20671656727790833,
      "learning_rate": 4.1580000000000005e-05,
      "loss": 0.5385,
      "step": 842
    },
    {
      "epoch": 76.67878787878787,
      "grad_norm": 0.3161730468273163,
      "learning_rate": 4.1570000000000003e-05,
      "loss": 0.4516,
      "step": 843
    },
    {
      "epoch": 76.77575757575758,
      "grad_norm": 0.18657682836055756,
      "learning_rate": 4.156e-05,
      "loss": 0.5144,
      "step": 844
    },
    {
      "epoch": 76.87272727272727,
      "grad_norm": 0.22428153455257416,
      "learning_rate": 4.155e-05,
      "loss": 0.5353,
      "step": 845
    },
    {
      "epoch": 76.96969696969697,
      "grad_norm": 0.269551157951355,
      "learning_rate": 4.154e-05,
      "loss": 0.5185,
      "step": 846
    },
    {
      "epoch": 77.0,
      "grad_norm": 0.4164291322231293,
      "learning_rate": 4.1530000000000004e-05,
      "loss": 0.5249,
      "step": 847
    },
    {
      "epoch": 77.0969696969697,
      "grad_norm": 0.24406081438064575,
      "learning_rate": 4.152e-05,
      "loss": 0.5158,
      "step": 848
    },
    {
      "epoch": 77.19393939393939,
      "grad_norm": 0.25519075989723206,
      "learning_rate": 4.151000000000001e-05,
      "loss": 0.5357,
      "step": 849
    },
    {
      "epoch": 77.2909090909091,
      "grad_norm": 0.20665471255779266,
      "learning_rate": 4.15e-05,
      "loss": 0.5698,
      "step": 850
    },
    {
      "epoch": 77.2909090909091,
      "eval_loss": 0.5581657886505127,
      "eval_runtime": 2.1723,
      "eval_samples_per_second": 25.318,
      "eval_steps_per_second": 3.222,
      "step": 850
    },
    {
      "epoch": 77.38787878787879,
      "grad_norm": 0.2132198065519333,
      "learning_rate": 4.1490000000000004e-05,
      "loss": 0.5195,
      "step": 851
    },
    {
      "epoch": 77.48484848484848,
      "grad_norm": 0.32972055673599243,
      "learning_rate": 4.148e-05,
      "loss": 0.459,
      "step": 852
    },
    {
      "epoch": 77.58181818181818,
      "grad_norm": 0.23783457279205322,
      "learning_rate": 4.147e-05,
      "loss": 0.5747,
      "step": 853
    },
    {
      "epoch": 77.67878787878787,
      "grad_norm": 0.2641775906085968,
      "learning_rate": 4.1460000000000006e-05,
      "loss": 0.5261,
      "step": 854
    },
    {
      "epoch": 77.77575757575758,
      "grad_norm": 0.22427310049533844,
      "learning_rate": 4.145e-05,
      "loss": 0.5283,
      "step": 855
    },
    {
      "epoch": 77.87272727272727,
      "grad_norm": 0.2677212059497833,
      "learning_rate": 4.144e-05,
      "loss": 0.5173,
      "step": 856
    },
    {
      "epoch": 77.96969696969697,
      "grad_norm": 0.23267583549022675,
      "learning_rate": 4.143e-05,
      "loss": 0.5134,
      "step": 857
    },
    {
      "epoch": 78.0,
      "grad_norm": 0.36065569519996643,
      "learning_rate": 4.142000000000001e-05,
      "loss": 0.4203,
      "step": 858
    },
    {
      "epoch": 78.0969696969697,
      "grad_norm": 0.22041353583335876,
      "learning_rate": 4.1410000000000005e-05,
      "loss": 0.5188,
      "step": 859
    },
    {
      "epoch": 78.19393939393939,
      "grad_norm": 0.2389429658651352,
      "learning_rate": 4.14e-05,
      "loss": 0.5348,
      "step": 860
    },
    {
      "epoch": 78.19393939393939,
      "eval_loss": 0.5565226078033447,
      "eval_runtime": 2.1747,
      "eval_samples_per_second": 25.291,
      "eval_steps_per_second": 3.219,
      "step": 860
    },
    {
      "epoch": 78.2909090909091,
      "grad_norm": 0.22565747797489166,
      "learning_rate": 4.139e-05,
      "loss": 0.5359,
      "step": 861
    },
    {
      "epoch": 78.38787878787879,
      "grad_norm": 0.20705768465995789,
      "learning_rate": 4.138e-05,
      "loss": 0.5281,
      "step": 862
    },
    {
      "epoch": 78.48484848484848,
      "grad_norm": 0.2351296991109848,
      "learning_rate": 4.1370000000000005e-05,
      "loss": 0.4949,
      "step": 863
    },
    {
      "epoch": 78.58181818181818,
      "grad_norm": 0.2614632248878479,
      "learning_rate": 4.1360000000000004e-05,
      "loss": 0.5063,
      "step": 864
    },
    {
      "epoch": 78.67878787878787,
      "grad_norm": 0.21661315858364105,
      "learning_rate": 4.135e-05,
      "loss": 0.5488,
      "step": 865
    },
    {
      "epoch": 78.77575757575758,
      "grad_norm": 0.21293285489082336,
      "learning_rate": 4.134e-05,
      "loss": 0.4951,
      "step": 866
    },
    {
      "epoch": 78.87272727272727,
      "grad_norm": 0.20552313327789307,
      "learning_rate": 4.133e-05,
      "loss": 0.5106,
      "step": 867
    },
    {
      "epoch": 78.96969696969697,
      "grad_norm": 0.23161394894123077,
      "learning_rate": 4.1320000000000004e-05,
      "loss": 0.5271,
      "step": 868
    },
    {
      "epoch": 79.0,
      "grad_norm": 0.36716896295547485,
      "learning_rate": 4.131e-05,
      "loss": 0.5289,
      "step": 869
    },
    {
      "epoch": 79.0969696969697,
      "grad_norm": 0.2597758173942566,
      "learning_rate": 4.13e-05,
      "loss": 0.4919,
      "step": 870
    },
    {
      "epoch": 79.0969696969697,
      "eval_loss": 0.5550717711448669,
      "eval_runtime": 2.1786,
      "eval_samples_per_second": 25.246,
      "eval_steps_per_second": 3.213,
      "step": 870
    },
    {
      "epoch": 79.19393939393939,
      "grad_norm": 0.27578315138816833,
      "learning_rate": 4.129e-05,
      "loss": 0.4752,
      "step": 871
    },
    {
      "epoch": 79.2909090909091,
      "grad_norm": 0.27824288606643677,
      "learning_rate": 4.1280000000000005e-05,
      "loss": 0.5307,
      "step": 872
    },
    {
      "epoch": 79.38787878787879,
      "grad_norm": 0.2966783046722412,
      "learning_rate": 4.127e-05,
      "loss": 0.4939,
      "step": 873
    },
    {
      "epoch": 79.48484848484848,
      "grad_norm": 0.33290615677833557,
      "learning_rate": 4.126e-05,
      "loss": 0.5487,
      "step": 874
    },
    {
      "epoch": 79.58181818181818,
      "grad_norm": 0.23399081826210022,
      "learning_rate": 4.125e-05,
      "loss": 0.5429,
      "step": 875
    },
    {
      "epoch": 79.67878787878787,
      "grad_norm": 0.2639947831630707,
      "learning_rate": 4.124e-05,
      "loss": 0.5252,
      "step": 876
    },
    {
      "epoch": 79.77575757575758,
      "grad_norm": 0.2671773433685303,
      "learning_rate": 4.123e-05,
      "loss": 0.5471,
      "step": 877
    },
    {
      "epoch": 79.87272727272727,
      "grad_norm": 0.259935200214386,
      "learning_rate": 4.122e-05,
      "loss": 0.5292,
      "step": 878
    },
    {
      "epoch": 79.96969696969697,
      "grad_norm": 0.26900020241737366,
      "learning_rate": 4.121000000000001e-05,
      "loss": 0.5124,
      "step": 879
    },
    {
      "epoch": 80.0,
      "grad_norm": 0.33917108178138733,
      "learning_rate": 4.12e-05,
      "loss": 0.4579,
      "step": 880
    },
    {
      "epoch": 80.0,
      "eval_loss": 0.5527308583259583,
      "eval_runtime": 2.1784,
      "eval_samples_per_second": 25.248,
      "eval_steps_per_second": 3.213,
      "step": 880
    },
    {
      "epoch": 80.0969696969697,
      "grad_norm": 0.27525821328163147,
      "learning_rate": 4.1190000000000004e-05,
      "loss": 0.5072,
      "step": 881
    },
    {
      "epoch": 80.19393939393939,
      "grad_norm": 0.3157818615436554,
      "learning_rate": 4.118e-05,
      "loss": 0.5834,
      "step": 882
    },
    {
      "epoch": 80.2909090909091,
      "grad_norm": 0.2659405767917633,
      "learning_rate": 4.117e-05,
      "loss": 0.5384,
      "step": 883
    },
    {
      "epoch": 80.38787878787879,
      "grad_norm": 0.22492793202400208,
      "learning_rate": 4.1160000000000006e-05,
      "loss": 0.5353,
      "step": 884
    },
    {
      "epoch": 80.48484848484848,
      "grad_norm": 0.21968522667884827,
      "learning_rate": 4.115e-05,
      "loss": 0.5242,
      "step": 885
    },
    {
      "epoch": 80.58181818181818,
      "grad_norm": 0.257949560880661,
      "learning_rate": 4.114e-05,
      "loss": 0.4986,
      "step": 886
    },
    {
      "epoch": 80.67878787878787,
      "grad_norm": 0.22139108180999756,
      "learning_rate": 4.113e-05,
      "loss": 0.4976,
      "step": 887
    },
    {
      "epoch": 80.77575757575758,
      "grad_norm": 0.2856335937976837,
      "learning_rate": 4.1120000000000006e-05,
      "loss": 0.4869,
      "step": 888
    },
    {
      "epoch": 80.87272727272727,
      "grad_norm": 0.3159530460834503,
      "learning_rate": 4.1110000000000005e-05,
      "loss": 0.5073,
      "step": 889
    },
    {
      "epoch": 80.96969696969697,
      "grad_norm": 0.26196572184562683,
      "learning_rate": 4.11e-05,
      "loss": 0.4938,
      "step": 890
    },
    {
      "epoch": 80.96969696969697,
      "eval_loss": 0.5511890649795532,
      "eval_runtime": 2.1854,
      "eval_samples_per_second": 25.167,
      "eval_steps_per_second": 3.203,
      "step": 890
    },
    {
      "epoch": 81.0,
      "grad_norm": 0.3370470702648163,
      "learning_rate": 4.109e-05,
      "loss": 0.4691,
      "step": 891
    },
    {
      "epoch": 81.0969696969697,
      "grad_norm": 0.26037460565567017,
      "learning_rate": 4.108e-05,
      "loss": 0.5557,
      "step": 892
    },
    {
      "epoch": 81.19393939393939,
      "grad_norm": 0.2754685878753662,
      "learning_rate": 4.1070000000000005e-05,
      "loss": 0.5391,
      "step": 893
    },
    {
      "epoch": 81.2909090909091,
      "grad_norm": 0.2869008481502533,
      "learning_rate": 4.106e-05,
      "loss": 0.4685,
      "step": 894
    },
    {
      "epoch": 81.38787878787879,
      "grad_norm": 0.31535306572914124,
      "learning_rate": 4.105e-05,
      "loss": 0.5187,
      "step": 895
    },
    {
      "epoch": 81.48484848484848,
      "grad_norm": 0.3106493353843689,
      "learning_rate": 4.104e-05,
      "loss": 0.5059,
      "step": 896
    },
    {
      "epoch": 81.58181818181818,
      "grad_norm": 0.2399621158838272,
      "learning_rate": 4.103e-05,
      "loss": 0.5069,
      "step": 897
    },
    {
      "epoch": 81.67878787878787,
      "grad_norm": 0.22072452306747437,
      "learning_rate": 4.1020000000000004e-05,
      "loss": 0.51,
      "step": 898
    },
    {
      "epoch": 81.77575757575758,
      "grad_norm": 0.23221911489963531,
      "learning_rate": 4.101e-05,
      "loss": 0.5053,
      "step": 899
    },
    {
      "epoch": 81.87272727272727,
      "grad_norm": 0.2527901530265808,
      "learning_rate": 4.1e-05,
      "loss": 0.5118,
      "step": 900
    },
    {
      "epoch": 81.87272727272727,
      "eval_loss": 0.5499869585037231,
      "eval_runtime": 2.1811,
      "eval_samples_per_second": 25.217,
      "eval_steps_per_second": 3.209,
      "step": 900
    },
    {
      "epoch": 81.96969696969697,
      "grad_norm": 0.29236477613449097,
      "learning_rate": 4.099e-05,
      "loss": 0.5386,
      "step": 901
    },
    {
      "epoch": 82.0,
      "grad_norm": 0.36997607350349426,
      "learning_rate": 4.0980000000000004e-05,
      "loss": 0.4296,
      "step": 902
    },
    {
      "epoch": 82.0969696969697,
      "grad_norm": 0.24829061329364777,
      "learning_rate": 4.097e-05,
      "loss": 0.5552,
      "step": 903
    },
    {
      "epoch": 82.19393939393939,
      "grad_norm": 0.25277000665664673,
      "learning_rate": 4.096e-05,
      "loss": 0.5306,
      "step": 904
    },
    {
      "epoch": 82.2909090909091,
      "grad_norm": 0.24040213227272034,
      "learning_rate": 4.095e-05,
      "loss": 0.5043,
      "step": 905
    },
    {
      "epoch": 82.38787878787879,
      "grad_norm": 0.26400551199913025,
      "learning_rate": 4.094e-05,
      "loss": 0.5272,
      "step": 906
    },
    {
      "epoch": 82.48484848484848,
      "grad_norm": 0.2064426690340042,
      "learning_rate": 4.093e-05,
      "loss": 0.4846,
      "step": 907
    },
    {
      "epoch": 82.58181818181818,
      "grad_norm": 0.22036471962928772,
      "learning_rate": 4.092e-05,
      "loss": 0.494,
      "step": 908
    },
    {
      "epoch": 82.67878787878787,
      "grad_norm": 0.2294999212026596,
      "learning_rate": 4.0910000000000006e-05,
      "loss": 0.5316,
      "step": 909
    },
    {
      "epoch": 82.77575757575758,
      "grad_norm": 0.29374563694000244,
      "learning_rate": 4.09e-05,
      "loss": 0.4586,
      "step": 910
    },
    {
      "epoch": 82.77575757575758,
      "eval_loss": 0.5479086637496948,
      "eval_runtime": 2.1669,
      "eval_samples_per_second": 25.382,
      "eval_steps_per_second": 3.23,
      "step": 910
    },
    {
      "epoch": 82.87272727272727,
      "grad_norm": 0.20776768028736115,
      "learning_rate": 4.089e-05,
      "loss": 0.5163,
      "step": 911
    },
    {
      "epoch": 82.96969696969697,
      "grad_norm": 0.23485852777957916,
      "learning_rate": 4.088e-05,
      "loss": 0.4974,
      "step": 912
    },
    {
      "epoch": 83.0,
      "grad_norm": 0.4788278639316559,
      "learning_rate": 4.087e-05,
      "loss": 0.5461,
      "step": 913
    },
    {
      "epoch": 83.0969696969697,
      "grad_norm": 0.20169512927532196,
      "learning_rate": 4.0860000000000005e-05,
      "loss": 0.4889,
      "step": 914
    },
    {
      "epoch": 83.19393939393939,
      "grad_norm": 0.20205354690551758,
      "learning_rate": 4.085e-05,
      "loss": 0.5038,
      "step": 915
    },
    {
      "epoch": 83.2909090909091,
      "grad_norm": 0.2369161695241928,
      "learning_rate": 4.084e-05,
      "loss": 0.538,
      "step": 916
    },
    {
      "epoch": 83.38787878787879,
      "grad_norm": 0.2776532471179962,
      "learning_rate": 4.083e-05,
      "loss": 0.5042,
      "step": 917
    },
    {
      "epoch": 83.48484848484848,
      "grad_norm": 0.27464279532432556,
      "learning_rate": 4.0820000000000006e-05,
      "loss": 0.5034,
      "step": 918
    },
    {
      "epoch": 83.58181818181818,
      "grad_norm": 0.2643680274486542,
      "learning_rate": 4.0810000000000004e-05,
      "loss": 0.5154,
      "step": 919
    },
    {
      "epoch": 83.67878787878787,
      "grad_norm": 0.2726304531097412,
      "learning_rate": 4.08e-05,
      "loss": 0.4715,
      "step": 920
    },
    {
      "epoch": 83.67878787878787,
      "eval_loss": 0.5463842749595642,
      "eval_runtime": 2.1765,
      "eval_samples_per_second": 25.269,
      "eval_steps_per_second": 3.216,
      "step": 920
    },
    {
      "epoch": 83.77575757575758,
      "grad_norm": 0.3199925422668457,
      "learning_rate": 4.079e-05,
      "loss": 0.5715,
      "step": 921
    },
    {
      "epoch": 83.87272727272727,
      "grad_norm": 0.19012175500392914,
      "learning_rate": 4.078e-05,
      "loss": 0.5037,
      "step": 922
    },
    {
      "epoch": 83.96969696969697,
      "grad_norm": 0.2449841946363449,
      "learning_rate": 4.0770000000000004e-05,
      "loss": 0.5269,
      "step": 923
    },
    {
      "epoch": 84.0,
      "grad_norm": 0.36201179027557373,
      "learning_rate": 4.076e-05,
      "loss": 0.3858,
      "step": 924
    },
    {
      "epoch": 84.0969696969697,
      "grad_norm": 0.20386870205402374,
      "learning_rate": 4.075e-05,
      "loss": 0.477,
      "step": 925
    },
    {
      "epoch": 84.19393939393939,
      "grad_norm": 0.25486937165260315,
      "learning_rate": 4.074e-05,
      "loss": 0.4891,
      "step": 926
    },
    {
      "epoch": 84.2909090909091,
      "grad_norm": 0.2490653544664383,
      "learning_rate": 4.0730000000000005e-05,
      "loss": 0.5217,
      "step": 927
    },
    {
      "epoch": 84.38787878787879,
      "grad_norm": 0.33604300022125244,
      "learning_rate": 4.072e-05,
      "loss": 0.5314,
      "step": 928
    },
    {
      "epoch": 84.48484848484848,
      "grad_norm": 0.21965400874614716,
      "learning_rate": 4.071e-05,
      "loss": 0.5255,
      "step": 929
    },
    {
      "epoch": 84.58181818181818,
      "grad_norm": 0.23073619604110718,
      "learning_rate": 4.07e-05,
      "loss": 0.456,
      "step": 930
    },
    {
      "epoch": 84.58181818181818,
      "eval_loss": 0.5450170636177063,
      "eval_runtime": 2.1803,
      "eval_samples_per_second": 25.226,
      "eval_steps_per_second": 3.211,
      "step": 930
    },
    {
      "epoch": 84.67878787878787,
      "grad_norm": 0.27645522356033325,
      "learning_rate": 4.069e-05,
      "loss": 0.4772,
      "step": 931
    },
    {
      "epoch": 84.77575757575758,
      "grad_norm": 0.27022767066955566,
      "learning_rate": 4.0680000000000004e-05,
      "loss": 0.5755,
      "step": 932
    },
    {
      "epoch": 84.87272727272727,
      "grad_norm": 0.24547351896762848,
      "learning_rate": 4.067e-05,
      "loss": 0.4969,
      "step": 933
    },
    {
      "epoch": 84.96969696969697,
      "grad_norm": 0.2812168598175049,
      "learning_rate": 4.066e-05,
      "loss": 0.5265,
      "step": 934
    },
    {
      "epoch": 85.0,
      "grad_norm": 0.36018022894859314,
      "learning_rate": 4.065e-05,
      "loss": 0.4702,
      "step": 935
    },
    {
      "epoch": 85.0969696969697,
      "grad_norm": 0.2520100176334381,
      "learning_rate": 4.064e-05,
      "loss": 0.4896,
      "step": 936
    },
    {
      "epoch": 85.19393939393939,
      "grad_norm": 0.3157733082771301,
      "learning_rate": 4.063e-05,
      "loss": 0.5089,
      "step": 937
    },
    {
      "epoch": 85.2909090909091,
      "grad_norm": 0.2828407883644104,
      "learning_rate": 4.062e-05,
      "loss": 0.4741,
      "step": 938
    },
    {
      "epoch": 85.38787878787879,
      "grad_norm": 0.29748284816741943,
      "learning_rate": 4.0610000000000006e-05,
      "loss": 0.5395,
      "step": 939
    },
    {
      "epoch": 85.48484848484848,
      "grad_norm": 0.2942025363445282,
      "learning_rate": 4.0600000000000004e-05,
      "loss": 0.5289,
      "step": 940
    },
    {
      "epoch": 85.48484848484848,
      "eval_loss": 0.5434708595275879,
      "eval_runtime": 2.1583,
      "eval_samples_per_second": 25.483,
      "eval_steps_per_second": 3.243,
      "step": 940
    },
    {
      "epoch": 85.58181818181818,
      "grad_norm": 0.22913864254951477,
      "learning_rate": 4.059e-05,
      "loss": 0.5366,
      "step": 941
    },
    {
      "epoch": 85.67878787878787,
      "grad_norm": 0.22568920254707336,
      "learning_rate": 4.058e-05,
      "loss": 0.4572,
      "step": 942
    },
    {
      "epoch": 85.77575757575758,
      "grad_norm": 0.31184083223342896,
      "learning_rate": 4.057e-05,
      "loss": 0.5577,
      "step": 943
    },
    {
      "epoch": 85.87272727272727,
      "grad_norm": 0.2624908983707428,
      "learning_rate": 4.0560000000000005e-05,
      "loss": 0.5237,
      "step": 944
    },
    {
      "epoch": 85.96969696969697,
      "grad_norm": 0.22147274017333984,
      "learning_rate": 4.055e-05,
      "loss": 0.4643,
      "step": 945
    },
    {
      "epoch": 86.0,
      "grad_norm": 0.3841789960861206,
      "learning_rate": 4.054e-05,
      "loss": 0.3985,
      "step": 946
    },
    {
      "epoch": 86.0969696969697,
      "grad_norm": 0.24668338894844055,
      "learning_rate": 4.053e-05,
      "loss": 0.5623,
      "step": 947
    },
    {
      "epoch": 86.19393939393939,
      "grad_norm": 0.2290932983160019,
      "learning_rate": 4.0520000000000005e-05,
      "loss": 0.4791,
      "step": 948
    },
    {
      "epoch": 86.2909090909091,
      "grad_norm": 0.25622695684432983,
      "learning_rate": 4.0510000000000003e-05,
      "loss": 0.458,
      "step": 949
    },
    {
      "epoch": 86.38787878787879,
      "grad_norm": 0.27793192863464355,
      "learning_rate": 4.05e-05,
      "loss": 0.5233,
      "step": 950
    },
    {
      "epoch": 86.38787878787879,
      "eval_loss": 0.5420384407043457,
      "eval_runtime": 2.1543,
      "eval_samples_per_second": 25.53,
      "eval_steps_per_second": 3.249,
      "step": 950
    },
    {
      "epoch": 86.48484848484848,
      "grad_norm": 0.290298193693161,
      "learning_rate": 4.049e-05,
      "loss": 0.5017,
      "step": 951
    },
    {
      "epoch": 86.58181818181818,
      "grad_norm": 0.24654896557331085,
      "learning_rate": 4.048e-05,
      "loss": 0.4799,
      "step": 952
    },
    {
      "epoch": 86.67878787878787,
      "grad_norm": 0.34982696175575256,
      "learning_rate": 4.0470000000000004e-05,
      "loss": 0.5292,
      "step": 953
    },
    {
      "epoch": 86.77575757575758,
      "grad_norm": 0.2249320149421692,
      "learning_rate": 4.046e-05,
      "loss": 0.5497,
      "step": 954
    },
    {
      "epoch": 86.87272727272727,
      "grad_norm": 0.31159865856170654,
      "learning_rate": 4.045000000000001e-05,
      "loss": 0.4766,
      "step": 955
    },
    {
      "epoch": 86.96969696969697,
      "grad_norm": 0.3017675578594208,
      "learning_rate": 4.044e-05,
      "loss": 0.4851,
      "step": 956
    },
    {
      "epoch": 87.0,
      "grad_norm": 0.42811262607574463,
      "learning_rate": 4.0430000000000004e-05,
      "loss": 0.4453,
      "step": 957
    },
    {
      "epoch": 87.0969696969697,
      "grad_norm": 0.25056886672973633,
      "learning_rate": 4.042e-05,
      "loss": 0.4505,
      "step": 958
    },
    {
      "epoch": 87.19393939393939,
      "grad_norm": 0.23695793747901917,
      "learning_rate": 4.041e-05,
      "loss": 0.5128,
      "step": 959
    },
    {
      "epoch": 87.2909090909091,
      "grad_norm": 0.23140308260917664,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 0.4859,
      "step": 960
    },
    {
      "epoch": 87.2909090909091,
      "eval_loss": 0.5406550765037537,
      "eval_runtime": 2.1825,
      "eval_samples_per_second": 25.2,
      "eval_steps_per_second": 3.207,
      "step": 960
    },
    {
      "epoch": 87.38787878787879,
      "grad_norm": 0.2655631899833679,
      "learning_rate": 4.039e-05,
      "loss": 0.4867,
      "step": 961
    },
    {
      "epoch": 87.48484848484848,
      "grad_norm": 0.41470491886138916,
      "learning_rate": 4.038e-05,
      "loss": 0.528,
      "step": 962
    },
    {
      "epoch": 87.58181818181818,
      "grad_norm": 0.29415225982666016,
      "learning_rate": 4.037e-05,
      "loss": 0.4967,
      "step": 963
    },
    {
      "epoch": 87.67878787878787,
      "grad_norm": 0.2565857470035553,
      "learning_rate": 4.0360000000000007e-05,
      "loss": 0.4967,
      "step": 964
    },
    {
      "epoch": 87.77575757575758,
      "grad_norm": 0.23314039409160614,
      "learning_rate": 4.0350000000000005e-05,
      "loss": 0.5357,
      "step": 965
    },
    {
      "epoch": 87.87272727272727,
      "grad_norm": 0.2893287241458893,
      "learning_rate": 4.034e-05,
      "loss": 0.4635,
      "step": 966
    },
    {
      "epoch": 87.96969696969697,
      "grad_norm": 0.2738717198371887,
      "learning_rate": 4.033e-05,
      "loss": 0.5388,
      "step": 967
    },
    {
      "epoch": 88.0,
      "grad_norm": 0.3843998610973358,
      "learning_rate": 4.032e-05,
      "loss": 0.5291,
      "step": 968
    },
    {
      "epoch": 88.0969696969697,
      "grad_norm": 0.2288815975189209,
      "learning_rate": 4.0310000000000005e-05,
      "loss": 0.5184,
      "step": 969
    },
    {
      "epoch": 88.19393939393939,
      "grad_norm": 0.2845372259616852,
      "learning_rate": 4.0300000000000004e-05,
      "loss": 0.5061,
      "step": 970
    },
    {
      "epoch": 88.19393939393939,
      "eval_loss": 0.5385352373123169,
      "eval_runtime": 2.175,
      "eval_samples_per_second": 25.287,
      "eval_steps_per_second": 3.218,
      "step": 970
    },
    {
      "epoch": 88.2909090909091,
      "grad_norm": 0.24731485545635223,
      "learning_rate": 4.029e-05,
      "loss": 0.5337,
      "step": 971
    },
    {
      "epoch": 88.38787878787879,
      "grad_norm": 0.29482629895210266,
      "learning_rate": 4.028e-05,
      "loss": 0.4525,
      "step": 972
    },
    {
      "epoch": 88.48484848484848,
      "grad_norm": 0.30363136529922485,
      "learning_rate": 4.027e-05,
      "loss": 0.5321,
      "step": 973
    },
    {
      "epoch": 88.58181818181818,
      "grad_norm": 0.2996557056903839,
      "learning_rate": 4.0260000000000004e-05,
      "loss": 0.4991,
      "step": 974
    },
    {
      "epoch": 88.67878787878787,
      "grad_norm": 0.2432321459054947,
      "learning_rate": 4.025e-05,
      "loss": 0.5032,
      "step": 975
    },
    {
      "epoch": 88.77575757575758,
      "grad_norm": 0.24066884815692902,
      "learning_rate": 4.024e-05,
      "loss": 0.5177,
      "step": 976
    },
    {
      "epoch": 88.87272727272727,
      "grad_norm": 0.2924744188785553,
      "learning_rate": 4.023e-05,
      "loss": 0.4963,
      "step": 977
    },
    {
      "epoch": 88.96969696969697,
      "grad_norm": 0.26147395372390747,
      "learning_rate": 4.0220000000000005e-05,
      "loss": 0.445,
      "step": 978
    },
    {
      "epoch": 89.0,
      "grad_norm": 0.4114329218864441,
      "learning_rate": 4.021e-05,
      "loss": 0.4469,
      "step": 979
    },
    {
      "epoch": 89.0969696969697,
      "grad_norm": 0.2708848714828491,
      "learning_rate": 4.02e-05,
      "loss": 0.5171,
      "step": 980
    },
    {
      "epoch": 89.0969696969697,
      "eval_loss": 0.5373245477676392,
      "eval_runtime": 2.1668,
      "eval_samples_per_second": 25.383,
      "eval_steps_per_second": 3.231,
      "step": 980
    },
    {
      "epoch": 89.19393939393939,
      "grad_norm": 0.24809399247169495,
      "learning_rate": 4.019e-05,
      "loss": 0.4804,
      "step": 981
    },
    {
      "epoch": 89.2909090909091,
      "grad_norm": 0.244582399725914,
      "learning_rate": 4.018e-05,
      "loss": 0.4539,
      "step": 982
    },
    {
      "epoch": 89.38787878787879,
      "grad_norm": 0.238327294588089,
      "learning_rate": 4.017e-05,
      "loss": 0.4755,
      "step": 983
    },
    {
      "epoch": 89.48484848484848,
      "grad_norm": 0.27144134044647217,
      "learning_rate": 4.016e-05,
      "loss": 0.4962,
      "step": 984
    },
    {
      "epoch": 89.58181818181818,
      "grad_norm": 0.27538859844207764,
      "learning_rate": 4.015000000000001e-05,
      "loss": 0.4898,
      "step": 985
    },
    {
      "epoch": 89.67878787878787,
      "grad_norm": 0.23520852625370026,
      "learning_rate": 4.014e-05,
      "loss": 0.5117,
      "step": 986
    },
    {
      "epoch": 89.77575757575758,
      "grad_norm": 0.3188869059085846,
      "learning_rate": 4.0130000000000004e-05,
      "loss": 0.4805,
      "step": 987
    },
    {
      "epoch": 89.87272727272727,
      "grad_norm": 0.25604870915412903,
      "learning_rate": 4.012e-05,
      "loss": 0.5397,
      "step": 988
    },
    {
      "epoch": 89.96969696969697,
      "grad_norm": 0.23663835227489471,
      "learning_rate": 4.011e-05,
      "loss": 0.5087,
      "step": 989
    },
    {
      "epoch": 90.0,
      "grad_norm": 0.42565521597862244,
      "learning_rate": 4.0100000000000006e-05,
      "loss": 0.5336,
      "step": 990
    },
    {
      "epoch": 90.0,
      "eval_loss": 0.5357187390327454,
      "eval_runtime": 2.1759,
      "eval_samples_per_second": 25.277,
      "eval_steps_per_second": 3.217,
      "step": 990
    },
    {
      "epoch": 90.0969696969697,
      "grad_norm": 0.21845504641532898,
      "learning_rate": 4.009e-05,
      "loss": 0.4928,
      "step": 991
    },
    {
      "epoch": 90.19393939393939,
      "grad_norm": 0.22380469739437103,
      "learning_rate": 4.008e-05,
      "loss": 0.4825,
      "step": 992
    },
    {
      "epoch": 90.2909090909091,
      "grad_norm": 0.19843821227550507,
      "learning_rate": 4.007e-05,
      "loss": 0.4954,
      "step": 993
    },
    {
      "epoch": 90.38787878787879,
      "grad_norm": 0.21744690835475922,
      "learning_rate": 4.0060000000000006e-05,
      "loss": 0.4988,
      "step": 994
    },
    {
      "epoch": 90.48484848484848,
      "grad_norm": 0.3063651919364929,
      "learning_rate": 4.0050000000000004e-05,
      "loss": 0.5554,
      "step": 995
    },
    {
      "epoch": 90.58181818181818,
      "grad_norm": 0.20970290899276733,
      "learning_rate": 4.004e-05,
      "loss": 0.4729,
      "step": 996
    },
    {
      "epoch": 90.67878787878787,
      "grad_norm": 0.2944852113723755,
      "learning_rate": 4.003e-05,
      "loss": 0.4766,
      "step": 997
    },
    {
      "epoch": 90.77575757575758,
      "grad_norm": 0.30164456367492676,
      "learning_rate": 4.002e-05,
      "loss": 0.5203,
      "step": 998
    },
    {
      "epoch": 90.87272727272727,
      "grad_norm": 0.2369176745414734,
      "learning_rate": 4.0010000000000005e-05,
      "loss": 0.4602,
      "step": 999
    },
    {
      "epoch": 90.96969696969697,
      "grad_norm": 0.2505776286125183,
      "learning_rate": 4e-05,
      "loss": 0.4625,
      "step": 1000
    },
    {
      "epoch": 90.96969696969697,
      "eval_loss": 0.53392094373703,
      "eval_runtime": 2.1907,
      "eval_samples_per_second": 25.107,
      "eval_steps_per_second": 3.195,
      "step": 1000
    },
    {
      "epoch": 91.0,
      "grad_norm": 0.4616742730140686,
      "learning_rate": 3.999e-05,
      "loss": 0.5795,
      "step": 1001
    },
    {
      "epoch": 91.0969696969697,
      "grad_norm": 0.22923487424850464,
      "learning_rate": 3.998e-05,
      "loss": 0.4707,
      "step": 1002
    },
    {
      "epoch": 91.19393939393939,
      "grad_norm": 0.2569350600242615,
      "learning_rate": 3.9970000000000005e-05,
      "loss": 0.5103,
      "step": 1003
    },
    {
      "epoch": 91.2909090909091,
      "grad_norm": 0.2466253936290741,
      "learning_rate": 3.9960000000000004e-05,
      "loss": 0.53,
      "step": 1004
    },
    {
      "epoch": 91.38787878787879,
      "grad_norm": 0.2187167853116989,
      "learning_rate": 3.995e-05,
      "loss": 0.4855,
      "step": 1005
    },
    {
      "epoch": 91.48484848484848,
      "grad_norm": 0.2226039469242096,
      "learning_rate": 3.994e-05,
      "loss": 0.4987,
      "step": 1006
    },
    {
      "epoch": 91.58181818181818,
      "grad_norm": 0.3207626938819885,
      "learning_rate": 3.993e-05,
      "loss": 0.5082,
      "step": 1007
    },
    {
      "epoch": 91.67878787878787,
      "grad_norm": 0.2705640196800232,
      "learning_rate": 3.9920000000000004e-05,
      "loss": 0.4889,
      "step": 1008
    },
    {
      "epoch": 91.77575757575758,
      "grad_norm": 0.2449875921010971,
      "learning_rate": 3.991e-05,
      "loss": 0.4749,
      "step": 1009
    },
    {
      "epoch": 91.87272727272727,
      "grad_norm": 0.21206451952457428,
      "learning_rate": 3.99e-05,
      "loss": 0.4607,
      "step": 1010
    },
    {
      "epoch": 91.87272727272727,
      "eval_loss": 0.5326685905456543,
      "eval_runtime": 2.188,
      "eval_samples_per_second": 25.137,
      "eval_steps_per_second": 3.199,
      "step": 1010
    },
    {
      "epoch": 91.96969696969697,
      "grad_norm": 0.26551949977874756,
      "learning_rate": 3.989e-05,
      "loss": 0.4945,
      "step": 1011
    },
    {
      "epoch": 92.0,
      "grad_norm": 0.37353646755218506,
      "learning_rate": 3.988e-05,
      "loss": 0.4931,
      "step": 1012
    },
    {
      "epoch": 92.0969696969697,
      "grad_norm": 0.21740274131298065,
      "learning_rate": 3.987e-05,
      "loss": 0.5023,
      "step": 1013
    },
    {
      "epoch": 92.19393939393939,
      "grad_norm": 0.2506203055381775,
      "learning_rate": 3.986e-05,
      "loss": 0.509,
      "step": 1014
    },
    {
      "epoch": 92.2909090909091,
      "grad_norm": 0.22760456800460815,
      "learning_rate": 3.9850000000000006e-05,
      "loss": 0.5169,
      "step": 1015
    },
    {
      "epoch": 92.38787878787879,
      "grad_norm": 0.30753859877586365,
      "learning_rate": 3.984e-05,
      "loss": 0.5013,
      "step": 1016
    },
    {
      "epoch": 92.48484848484848,
      "grad_norm": 0.24482981860637665,
      "learning_rate": 3.983e-05,
      "loss": 0.5132,
      "step": 1017
    },
    {
      "epoch": 92.58181818181818,
      "grad_norm": 0.31413623690605164,
      "learning_rate": 3.982e-05,
      "loss": 0.4936,
      "step": 1018
    },
    {
      "epoch": 92.67878787878787,
      "grad_norm": 0.22564786672592163,
      "learning_rate": 3.981e-05,
      "loss": 0.4823,
      "step": 1019
    },
    {
      "epoch": 92.77575757575758,
      "grad_norm": 0.3035551607608795,
      "learning_rate": 3.9800000000000005e-05,
      "loss": 0.4578,
      "step": 1020
    },
    {
      "epoch": 92.77575757575758,
      "eval_loss": 0.5315625071525574,
      "eval_runtime": 2.1756,
      "eval_samples_per_second": 25.281,
      "eval_steps_per_second": 3.218,
      "step": 1020
    },
    {
      "epoch": 92.87272727272727,
      "grad_norm": 0.2384045273065567,
      "learning_rate": 3.979e-05,
      "loss": 0.4449,
      "step": 1021
    },
    {
      "epoch": 92.96969696969697,
      "grad_norm": 0.25220656394958496,
      "learning_rate": 3.978e-05,
      "loss": 0.4847,
      "step": 1022
    },
    {
      "epoch": 93.0,
      "grad_norm": 0.2742287218570709,
      "learning_rate": 3.977e-05,
      "loss": 0.4884,
      "step": 1023
    },
    {
      "epoch": 93.0969696969697,
      "grad_norm": 0.25580084323883057,
      "learning_rate": 3.9760000000000006e-05,
      "loss": 0.4627,
      "step": 1024
    },
    {
      "epoch": 93.19393939393939,
      "grad_norm": 0.29719841480255127,
      "learning_rate": 3.9750000000000004e-05,
      "loss": 0.4855,
      "step": 1025
    },
    {
      "epoch": 93.2909090909091,
      "grad_norm": 0.3261563777923584,
      "learning_rate": 3.974e-05,
      "loss": 0.4764,
      "step": 1026
    },
    {
      "epoch": 93.38787878787879,
      "grad_norm": 0.2409381866455078,
      "learning_rate": 3.973e-05,
      "loss": 0.4672,
      "step": 1027
    },
    {
      "epoch": 93.48484848484848,
      "grad_norm": 0.26164528727531433,
      "learning_rate": 3.972e-05,
      "loss": 0.5152,
      "step": 1028
    },
    {
      "epoch": 93.58181818181818,
      "grad_norm": 0.2876894474029541,
      "learning_rate": 3.9710000000000004e-05,
      "loss": 0.521,
      "step": 1029
    },
    {
      "epoch": 93.67878787878787,
      "grad_norm": 0.22141848504543304,
      "learning_rate": 3.97e-05,
      "loss": 0.4715,
      "step": 1030
    },
    {
      "epoch": 93.67878787878787,
      "eval_loss": 0.5300446152687073,
      "eval_runtime": 2.1679,
      "eval_samples_per_second": 25.37,
      "eval_steps_per_second": 3.229,
      "step": 1030
    },
    {
      "epoch": 93.77575757575758,
      "grad_norm": 0.29566359519958496,
      "learning_rate": 3.969e-05,
      "loss": 0.5235,
      "step": 1031
    },
    {
      "epoch": 93.87272727272727,
      "grad_norm": 0.34124600887298584,
      "learning_rate": 3.968e-05,
      "loss": 0.4517,
      "step": 1032
    },
    {
      "epoch": 93.96969696969697,
      "grad_norm": 0.27476051449775696,
      "learning_rate": 3.9670000000000005e-05,
      "loss": 0.5173,
      "step": 1033
    },
    {
      "epoch": 94.0,
      "grad_norm": 0.38910970091819763,
      "learning_rate": 3.966e-05,
      "loss": 0.466,
      "step": 1034
    },
    {
      "epoch": 94.0969696969697,
      "grad_norm": 0.25072184205055237,
      "learning_rate": 3.965e-05,
      "loss": 0.5121,
      "step": 1035
    },
    {
      "epoch": 94.19393939393939,
      "grad_norm": 0.34341293573379517,
      "learning_rate": 3.964e-05,
      "loss": 0.435,
      "step": 1036
    },
    {
      "epoch": 94.2909090909091,
      "grad_norm": 0.2264951914548874,
      "learning_rate": 3.963e-05,
      "loss": 0.4919,
      "step": 1037
    },
    {
      "epoch": 94.38787878787879,
      "grad_norm": 0.24091920256614685,
      "learning_rate": 3.9620000000000004e-05,
      "loss": 0.4843,
      "step": 1038
    },
    {
      "epoch": 94.48484848484848,
      "grad_norm": 0.2943407893180847,
      "learning_rate": 3.961e-05,
      "loss": 0.5102,
      "step": 1039
    },
    {
      "epoch": 94.58181818181818,
      "grad_norm": 0.26254013180732727,
      "learning_rate": 3.960000000000001e-05,
      "loss": 0.4832,
      "step": 1040
    },
    {
      "epoch": 94.58181818181818,
      "eval_loss": 0.5287523865699768,
      "eval_runtime": 2.1727,
      "eval_samples_per_second": 25.314,
      "eval_steps_per_second": 3.222,
      "step": 1040
    },
    {
      "epoch": 94.67878787878787,
      "grad_norm": 0.3030388057231903,
      "learning_rate": 3.959e-05,
      "loss": 0.4926,
      "step": 1041
    },
    {
      "epoch": 94.77575757575758,
      "grad_norm": 0.29376158118247986,
      "learning_rate": 3.958e-05,
      "loss": 0.51,
      "step": 1042
    },
    {
      "epoch": 94.87272727272727,
      "grad_norm": 0.26268187165260315,
      "learning_rate": 3.957e-05,
      "loss": 0.4774,
      "step": 1043
    },
    {
      "epoch": 94.96969696969697,
      "grad_norm": 0.2799707055091858,
      "learning_rate": 3.956e-05,
      "loss": 0.464,
      "step": 1044
    },
    {
      "epoch": 95.0,
      "grad_norm": 0.5625638961791992,
      "learning_rate": 3.9550000000000006e-05,
      "loss": 0.5032,
      "step": 1045
    },
    {
      "epoch": 95.0969696969697,
      "grad_norm": 0.2317277193069458,
      "learning_rate": 3.954e-05,
      "loss": 0.4852,
      "step": 1046
    },
    {
      "epoch": 95.19393939393939,
      "grad_norm": 0.3576275706291199,
      "learning_rate": 3.953e-05,
      "loss": 0.4789,
      "step": 1047
    },
    {
      "epoch": 95.2909090909091,
      "grad_norm": 0.26975786685943604,
      "learning_rate": 3.952e-05,
      "loss": 0.5058,
      "step": 1048
    },
    {
      "epoch": 95.38787878787879,
      "grad_norm": 0.23840966820716858,
      "learning_rate": 3.951e-05,
      "loss": 0.541,
      "step": 1049
    },
    {
      "epoch": 95.48484848484848,
      "grad_norm": 0.30056631565093994,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 0.4606,
      "step": 1050
    },
    {
      "epoch": 95.48484848484848,
      "eval_loss": 0.5273352861404419,
      "eval_runtime": 2.2113,
      "eval_samples_per_second": 24.872,
      "eval_steps_per_second": 3.166,
      "step": 1050
    },
    {
      "epoch": 95.58181818181818,
      "grad_norm": 0.22875840961933136,
      "learning_rate": 3.9489999999999996e-05,
      "loss": 0.4746,
      "step": 1051
    },
    {
      "epoch": 95.67878787878787,
      "grad_norm": 0.29144978523254395,
      "learning_rate": 3.948e-05,
      "loss": 0.4952,
      "step": 1052
    },
    {
      "epoch": 95.77575757575758,
      "grad_norm": 0.3198563754558563,
      "learning_rate": 3.947e-05,
      "loss": 0.4659,
      "step": 1053
    },
    {
      "epoch": 95.87272727272727,
      "grad_norm": 0.27438458800315857,
      "learning_rate": 3.9460000000000005e-05,
      "loss": 0.4623,
      "step": 1054
    },
    {
      "epoch": 95.96969696969697,
      "grad_norm": 0.2391417771577835,
      "learning_rate": 3.9450000000000003e-05,
      "loss": 0.5001,
      "step": 1055
    },
    {
      "epoch": 96.0,
      "grad_norm": 0.40089306235313416,
      "learning_rate": 3.944e-05,
      "loss": 0.415,
      "step": 1056
    },
    {
      "epoch": 96.0969696969697,
      "grad_norm": 0.26197531819343567,
      "learning_rate": 3.943e-05,
      "loss": 0.5019,
      "step": 1057
    },
    {
      "epoch": 96.19393939393939,
      "grad_norm": 0.3073684573173523,
      "learning_rate": 3.942e-05,
      "loss": 0.4635,
      "step": 1058
    },
    {
      "epoch": 96.2909090909091,
      "grad_norm": 0.2256932556629181,
      "learning_rate": 3.9410000000000004e-05,
      "loss": 0.535,
      "step": 1059
    },
    {
      "epoch": 96.38787878787879,
      "grad_norm": 0.30590319633483887,
      "learning_rate": 3.94e-05,
      "loss": 0.4781,
      "step": 1060
    },
    {
      "epoch": 96.38787878787879,
      "eval_loss": 0.5255498290061951,
      "eval_runtime": 2.1691,
      "eval_samples_per_second": 25.356,
      "eval_steps_per_second": 3.227,
      "step": 1060
    },
    {
      "epoch": 96.48484848484848,
      "grad_norm": 0.28140392899513245,
      "learning_rate": 3.939e-05,
      "loss": 0.4884,
      "step": 1061
    },
    {
      "epoch": 96.58181818181818,
      "grad_norm": 0.229629784822464,
      "learning_rate": 3.938e-05,
      "loss": 0.4675,
      "step": 1062
    },
    {
      "epoch": 96.67878787878787,
      "grad_norm": 0.24081185460090637,
      "learning_rate": 3.9370000000000004e-05,
      "loss": 0.4826,
      "step": 1063
    },
    {
      "epoch": 96.77575757575758,
      "grad_norm": 0.2174087017774582,
      "learning_rate": 3.936e-05,
      "loss": 0.4502,
      "step": 1064
    },
    {
      "epoch": 96.87272727272727,
      "grad_norm": 0.24293029308319092,
      "learning_rate": 3.935e-05,
      "loss": 0.4529,
      "step": 1065
    },
    {
      "epoch": 96.96969696969697,
      "grad_norm": 0.26654404401779175,
      "learning_rate": 3.9340000000000006e-05,
      "loss": 0.4891,
      "step": 1066
    },
    {
      "epoch": 97.0,
      "grad_norm": 0.4864083230495453,
      "learning_rate": 3.933e-05,
      "loss": 0.5446,
      "step": 1067
    },
    {
      "epoch": 97.0969696969697,
      "grad_norm": 0.24581874907016754,
      "learning_rate": 3.932e-05,
      "loss": 0.5138,
      "step": 1068
    },
    {
      "epoch": 97.19393939393939,
      "grad_norm": 0.24638426303863525,
      "learning_rate": 3.931e-05,
      "loss": 0.5056,
      "step": 1069
    },
    {
      "epoch": 97.2909090909091,
      "grad_norm": 0.24548788368701935,
      "learning_rate": 3.9300000000000007e-05,
      "loss": 0.5096,
      "step": 1070
    },
    {
      "epoch": 97.2909090909091,
      "eval_loss": 0.524466335773468,
      "eval_runtime": 2.1898,
      "eval_samples_per_second": 25.117,
      "eval_steps_per_second": 3.197,
      "step": 1070
    },
    {
      "epoch": 97.38787878787879,
      "grad_norm": 0.3515244424343109,
      "learning_rate": 3.9290000000000005e-05,
      "loss": 0.4682,
      "step": 1071
    },
    {
      "epoch": 97.48484848484848,
      "grad_norm": 0.334836483001709,
      "learning_rate": 3.9280000000000003e-05,
      "loss": 0.4996,
      "step": 1072
    },
    {
      "epoch": 97.58181818181818,
      "grad_norm": 0.3351961076259613,
      "learning_rate": 3.927e-05,
      "loss": 0.435,
      "step": 1073
    },
    {
      "epoch": 97.67878787878787,
      "grad_norm": 0.2502618134021759,
      "learning_rate": 3.926e-05,
      "loss": 0.4569,
      "step": 1074
    },
    {
      "epoch": 97.77575757575758,
      "grad_norm": 0.25060755014419556,
      "learning_rate": 3.9250000000000005e-05,
      "loss": 0.472,
      "step": 1075
    },
    {
      "epoch": 97.87272727272727,
      "grad_norm": 0.29755687713623047,
      "learning_rate": 3.9240000000000004e-05,
      "loss": 0.4862,
      "step": 1076
    },
    {
      "epoch": 97.96969696969697,
      "grad_norm": 0.44427555799484253,
      "learning_rate": 3.923e-05,
      "loss": 0.4797,
      "step": 1077
    },
    {
      "epoch": 98.0,
      "grad_norm": 0.3197348415851593,
      "learning_rate": 3.922e-05,
      "loss": 0.4355,
      "step": 1078
    },
    {
      "epoch": 98.0969696969697,
      "grad_norm": 0.2131693959236145,
      "learning_rate": 3.921e-05,
      "loss": 0.4879,
      "step": 1079
    },
    {
      "epoch": 98.19393939393939,
      "grad_norm": 0.2979012131690979,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 0.4585,
      "step": 1080
    },
    {
      "epoch": 98.19393939393939,
      "eval_loss": 0.5229376554489136,
      "eval_runtime": 2.1836,
      "eval_samples_per_second": 25.187,
      "eval_steps_per_second": 3.206,
      "step": 1080
    },
    {
      "epoch": 98.2909090909091,
      "grad_norm": 0.3021116554737091,
      "learning_rate": 3.919e-05,
      "loss": 0.4985,
      "step": 1081
    },
    {
      "epoch": 98.38787878787879,
      "grad_norm": 0.28391754627227783,
      "learning_rate": 3.918e-05,
      "loss": 0.434,
      "step": 1082
    },
    {
      "epoch": 98.48484848484848,
      "grad_norm": 0.32781583070755005,
      "learning_rate": 3.917e-05,
      "loss": 0.4682,
      "step": 1083
    },
    {
      "epoch": 98.58181818181818,
      "grad_norm": 0.28897055983543396,
      "learning_rate": 3.9160000000000005e-05,
      "loss": 0.4744,
      "step": 1084
    },
    {
      "epoch": 98.67878787878787,
      "grad_norm": 0.33773258328437805,
      "learning_rate": 3.915e-05,
      "loss": 0.5519,
      "step": 1085
    },
    {
      "epoch": 98.77575757575758,
      "grad_norm": 0.23259370028972626,
      "learning_rate": 3.914e-05,
      "loss": 0.4688,
      "step": 1086
    },
    {
      "epoch": 98.87272727272727,
      "grad_norm": 0.27739864587783813,
      "learning_rate": 3.913e-05,
      "loss": 0.4605,
      "step": 1087
    },
    {
      "epoch": 98.96969696969697,
      "grad_norm": 0.2524653375148773,
      "learning_rate": 3.912e-05,
      "loss": 0.4761,
      "step": 1088
    },
    {
      "epoch": 99.0,
      "grad_norm": 0.5492696762084961,
      "learning_rate": 3.911e-05,
      "loss": 0.5231,
      "step": 1089
    },
    {
      "epoch": 99.0969696969697,
      "grad_norm": 0.24189777672290802,
      "learning_rate": 3.91e-05,
      "loss": 0.4844,
      "step": 1090
    },
    {
      "epoch": 99.0969696969697,
      "eval_loss": 0.5221452713012695,
      "eval_runtime": 2.1659,
      "eval_samples_per_second": 25.394,
      "eval_steps_per_second": 3.232,
      "step": 1090
    },
    {
      "epoch": 99.19393939393939,
      "grad_norm": 0.2546369135379791,
      "learning_rate": 3.909000000000001e-05,
      "loss": 0.5451,
      "step": 1091
    },
    {
      "epoch": 99.2909090909091,
      "grad_norm": 0.343545526266098,
      "learning_rate": 3.908e-05,
      "loss": 0.4831,
      "step": 1092
    },
    {
      "epoch": 99.38787878787879,
      "grad_norm": 0.2641466557979584,
      "learning_rate": 3.9070000000000004e-05,
      "loss": 0.4715,
      "step": 1093
    },
    {
      "epoch": 99.48484848484848,
      "grad_norm": 0.2601754665374756,
      "learning_rate": 3.906e-05,
      "loss": 0.4429,
      "step": 1094
    },
    {
      "epoch": 99.58181818181818,
      "grad_norm": 0.2309015393257141,
      "learning_rate": 3.905e-05,
      "loss": 0.4734,
      "step": 1095
    },
    {
      "epoch": 99.67878787878787,
      "grad_norm": 0.24893411993980408,
      "learning_rate": 3.9040000000000006e-05,
      "loss": 0.4558,
      "step": 1096
    },
    {
      "epoch": 99.77575757575758,
      "grad_norm": 0.2772034704685211,
      "learning_rate": 3.903e-05,
      "loss": 0.4914,
      "step": 1097
    },
    {
      "epoch": 99.87272727272727,
      "grad_norm": 0.2813556492328644,
      "learning_rate": 3.902e-05,
      "loss": 0.5145,
      "step": 1098
    },
    {
      "epoch": 99.96969696969697,
      "grad_norm": 0.29801681637763977,
      "learning_rate": 3.901e-05,
      "loss": 0.4345,
      "step": 1099
    },
    {
      "epoch": 100.0,
      "grad_norm": 0.46238136291503906,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.4123,
      "step": 1100
    },
    {
      "epoch": 100.0,
      "eval_loss": 0.5208865404129028,
      "eval_runtime": 2.1746,
      "eval_samples_per_second": 25.292,
      "eval_steps_per_second": 3.219,
      "step": 1100
    },
    {
      "epoch": 100.0969696969697,
      "grad_norm": 0.2547599673271179,
      "learning_rate": 3.8990000000000004e-05,
      "loss": 0.4441,
      "step": 1101
    },
    {
      "epoch": 100.19393939393939,
      "grad_norm": 0.2973734736442566,
      "learning_rate": 3.898e-05,
      "loss": 0.5059,
      "step": 1102
    },
    {
      "epoch": 100.2909090909091,
      "grad_norm": 0.2850284278392792,
      "learning_rate": 3.897e-05,
      "loss": 0.5007,
      "step": 1103
    },
    {
      "epoch": 100.38787878787879,
      "grad_norm": 0.2912881076335907,
      "learning_rate": 3.896e-05,
      "loss": 0.4585,
      "step": 1104
    },
    {
      "epoch": 100.48484848484848,
      "grad_norm": 0.295784592628479,
      "learning_rate": 3.8950000000000005e-05,
      "loss": 0.4647,
      "step": 1105
    },
    {
      "epoch": 100.58181818181818,
      "grad_norm": 0.22379592061042786,
      "learning_rate": 3.894e-05,
      "loss": 0.4693,
      "step": 1106
    },
    {
      "epoch": 100.67878787878787,
      "grad_norm": 0.21056662499904633,
      "learning_rate": 3.893e-05,
      "loss": 0.4755,
      "step": 1107
    },
    {
      "epoch": 100.77575757575758,
      "grad_norm": 0.23073583841323853,
      "learning_rate": 3.892e-05,
      "loss": 0.4556,
      "step": 1108
    },
    {
      "epoch": 100.87272727272727,
      "grad_norm": 0.327334463596344,
      "learning_rate": 3.8910000000000005e-05,
      "loss": 0.4741,
      "step": 1109
    },
    {
      "epoch": 100.96969696969697,
      "grad_norm": 0.30131009221076965,
      "learning_rate": 3.8900000000000004e-05,
      "loss": 0.511,
      "step": 1110
    },
    {
      "epoch": 100.96969696969697,
      "eval_loss": 0.519778847694397,
      "eval_runtime": 2.1764,
      "eval_samples_per_second": 25.272,
      "eval_steps_per_second": 3.216,
      "step": 1110
    },
    {
      "epoch": 101.0,
      "grad_norm": 0.3921085298061371,
      "learning_rate": 3.889e-05,
      "loss": 0.4754,
      "step": 1111
    },
    {
      "epoch": 101.0969696969697,
      "grad_norm": 0.22144660353660583,
      "learning_rate": 3.888e-05,
      "loss": 0.4728,
      "step": 1112
    },
    {
      "epoch": 101.19393939393939,
      "grad_norm": 0.3088531196117401,
      "learning_rate": 3.887e-05,
      "loss": 0.5127,
      "step": 1113
    },
    {
      "epoch": 101.2909090909091,
      "grad_norm": 0.2640801668167114,
      "learning_rate": 3.8860000000000004e-05,
      "loss": 0.4583,
      "step": 1114
    },
    {
      "epoch": 101.38787878787879,
      "grad_norm": 0.2525095045566559,
      "learning_rate": 3.885e-05,
      "loss": 0.5007,
      "step": 1115
    },
    {
      "epoch": 101.48484848484848,
      "grad_norm": 0.26294276118278503,
      "learning_rate": 3.884e-05,
      "loss": 0.4766,
      "step": 1116
    },
    {
      "epoch": 101.58181818181818,
      "grad_norm": 0.21842017769813538,
      "learning_rate": 3.883e-05,
      "loss": 0.4665,
      "step": 1117
    },
    {
      "epoch": 101.67878787878787,
      "grad_norm": 0.30087506771087646,
      "learning_rate": 3.882e-05,
      "loss": 0.5296,
      "step": 1118
    },
    {
      "epoch": 101.77575757575758,
      "grad_norm": 0.2511281967163086,
      "learning_rate": 3.881e-05,
      "loss": 0.4662,
      "step": 1119
    },
    {
      "epoch": 101.87272727272727,
      "grad_norm": 0.2348358929157257,
      "learning_rate": 3.88e-05,
      "loss": 0.4281,
      "step": 1120
    },
    {
      "epoch": 101.87272727272727,
      "eval_loss": 0.5181769132614136,
      "eval_runtime": 2.1723,
      "eval_samples_per_second": 25.319,
      "eval_steps_per_second": 3.222,
      "step": 1120
    },
    {
      "epoch": 101.96969696969697,
      "grad_norm": 0.20796290040016174,
      "learning_rate": 3.8790000000000006e-05,
      "loss": 0.4557,
      "step": 1121
    },
    {
      "epoch": 102.0,
      "grad_norm": 0.3797859251499176,
      "learning_rate": 3.878e-05,
      "loss": 0.3847,
      "step": 1122
    },
    {
      "epoch": 102.0969696969697,
      "grad_norm": 0.25829294323921204,
      "learning_rate": 3.877e-05,
      "loss": 0.4655,
      "step": 1123
    },
    {
      "epoch": 102.19393939393939,
      "grad_norm": 0.2667910158634186,
      "learning_rate": 3.876e-05,
      "loss": 0.4809,
      "step": 1124
    },
    {
      "epoch": 102.2909090909091,
      "grad_norm": 0.30939432978630066,
      "learning_rate": 3.875e-05,
      "loss": 0.4723,
      "step": 1125
    },
    {
      "epoch": 102.38787878787879,
      "grad_norm": 0.2525250315666199,
      "learning_rate": 3.8740000000000005e-05,
      "loss": 0.4678,
      "step": 1126
    },
    {
      "epoch": 102.48484848484848,
      "grad_norm": 0.2383662611246109,
      "learning_rate": 3.873e-05,
      "loss": 0.4835,
      "step": 1127
    },
    {
      "epoch": 102.58181818181818,
      "grad_norm": 0.23342494666576385,
      "learning_rate": 3.872e-05,
      "loss": 0.4867,
      "step": 1128
    },
    {
      "epoch": 102.67878787878787,
      "grad_norm": 0.2512975037097931,
      "learning_rate": 3.871e-05,
      "loss": 0.4751,
      "step": 1129
    },
    {
      "epoch": 102.77575757575758,
      "grad_norm": 0.2931779623031616,
      "learning_rate": 3.8700000000000006e-05,
      "loss": 0.5153,
      "step": 1130
    },
    {
      "epoch": 102.77575757575758,
      "eval_loss": 0.5171474814414978,
      "eval_runtime": 2.1713,
      "eval_samples_per_second": 25.331,
      "eval_steps_per_second": 3.224,
      "step": 1130
    },
    {
      "epoch": 102.87272727272727,
      "grad_norm": 0.3106895089149475,
      "learning_rate": 3.8690000000000004e-05,
      "loss": 0.4588,
      "step": 1131
    },
    {
      "epoch": 102.96969696969697,
      "grad_norm": 0.24597759544849396,
      "learning_rate": 3.868e-05,
      "loss": 0.4184,
      "step": 1132
    },
    {
      "epoch": 103.0,
      "grad_norm": 0.3818826973438263,
      "learning_rate": 3.867e-05,
      "loss": 0.4709,
      "step": 1133
    },
    {
      "epoch": 103.0969696969697,
      "grad_norm": 0.23651278018951416,
      "learning_rate": 3.866e-05,
      "loss": 0.4789,
      "step": 1134
    },
    {
      "epoch": 103.19393939393939,
      "grad_norm": 0.3078266680240631,
      "learning_rate": 3.8650000000000004e-05,
      "loss": 0.5128,
      "step": 1135
    },
    {
      "epoch": 103.2909090909091,
      "grad_norm": 0.329988569021225,
      "learning_rate": 3.864e-05,
      "loss": 0.412,
      "step": 1136
    },
    {
      "epoch": 103.38787878787879,
      "grad_norm": 0.28413721919059753,
      "learning_rate": 3.863e-05,
      "loss": 0.4525,
      "step": 1137
    },
    {
      "epoch": 103.48484848484848,
      "grad_norm": 0.24670186638832092,
      "learning_rate": 3.862e-05,
      "loss": 0.5062,
      "step": 1138
    },
    {
      "epoch": 103.58181818181818,
      "grad_norm": 0.24704158306121826,
      "learning_rate": 3.8610000000000005e-05,
      "loss": 0.4857,
      "step": 1139
    },
    {
      "epoch": 103.67878787878787,
      "grad_norm": 0.20425106585025787,
      "learning_rate": 3.86e-05,
      "loss": 0.4667,
      "step": 1140
    },
    {
      "epoch": 103.67878787878787,
      "eval_loss": 0.5160987973213196,
      "eval_runtime": 2.1933,
      "eval_samples_per_second": 25.076,
      "eval_steps_per_second": 3.191,
      "step": 1140
    },
    {
      "epoch": 103.77575757575758,
      "grad_norm": 0.2687723934650421,
      "learning_rate": 3.859e-05,
      "loss": 0.4881,
      "step": 1141
    },
    {
      "epoch": 103.87272727272727,
      "grad_norm": 0.21118372678756714,
      "learning_rate": 3.858e-05,
      "loss": 0.4545,
      "step": 1142
    },
    {
      "epoch": 103.96969696969697,
      "grad_norm": 0.24843823909759521,
      "learning_rate": 3.857e-05,
      "loss": 0.4524,
      "step": 1143
    },
    {
      "epoch": 104.0,
      "grad_norm": 0.441841185092926,
      "learning_rate": 3.8560000000000004e-05,
      "loss": 0.4624,
      "step": 1144
    },
    {
      "epoch": 104.0969696969697,
      "grad_norm": 0.2640858590602875,
      "learning_rate": 3.855e-05,
      "loss": 0.4837,
      "step": 1145
    },
    {
      "epoch": 104.19393939393939,
      "grad_norm": 0.35844892263412476,
      "learning_rate": 3.854000000000001e-05,
      "loss": 0.5218,
      "step": 1146
    },
    {
      "epoch": 104.2909090909091,
      "grad_norm": 0.26051416993141174,
      "learning_rate": 3.853e-05,
      "loss": 0.4579,
      "step": 1147
    },
    {
      "epoch": 104.38787878787879,
      "grad_norm": 0.20667533576488495,
      "learning_rate": 3.8520000000000004e-05,
      "loss": 0.4353,
      "step": 1148
    },
    {
      "epoch": 104.48484848484848,
      "grad_norm": 0.20429357886314392,
      "learning_rate": 3.851e-05,
      "loss": 0.4358,
      "step": 1149
    },
    {
      "epoch": 104.58181818181818,
      "grad_norm": 0.2379342019557953,
      "learning_rate": 3.85e-05,
      "loss": 0.4862,
      "step": 1150
    },
    {
      "epoch": 104.58181818181818,
      "eval_loss": 0.5155452489852905,
      "eval_runtime": 2.1929,
      "eval_samples_per_second": 25.081,
      "eval_steps_per_second": 3.192,
      "step": 1150
    },
    {
      "epoch": 104.67878787878787,
      "grad_norm": 0.3056175112724304,
      "learning_rate": 3.8490000000000006e-05,
      "loss": 0.4735,
      "step": 1151
    },
    {
      "epoch": 104.77575757575758,
      "grad_norm": 0.31699028611183167,
      "learning_rate": 3.848e-05,
      "loss": 0.4544,
      "step": 1152
    },
    {
      "epoch": 104.87272727272727,
      "grad_norm": 0.24532760679721832,
      "learning_rate": 3.847e-05,
      "loss": 0.4553,
      "step": 1153
    },
    {
      "epoch": 104.96969696969697,
      "grad_norm": 0.22634100914001465,
      "learning_rate": 3.846e-05,
      "loss": 0.4921,
      "step": 1154
    },
    {
      "epoch": 105.0,
      "grad_norm": 0.4279729723930359,
      "learning_rate": 3.845e-05,
      "loss": 0.4548,
      "step": 1155
    },
    {
      "epoch": 105.0969696969697,
      "grad_norm": 0.27833861112594604,
      "learning_rate": 3.8440000000000005e-05,
      "loss": 0.4647,
      "step": 1156
    },
    {
      "epoch": 105.19393939393939,
      "grad_norm": 0.28585150837898254,
      "learning_rate": 3.8429999999999996e-05,
      "loss": 0.4608,
      "step": 1157
    },
    {
      "epoch": 105.2909090909091,
      "grad_norm": 0.23402553796768188,
      "learning_rate": 3.842e-05,
      "loss": 0.4832,
      "step": 1158
    },
    {
      "epoch": 105.38787878787879,
      "grad_norm": 0.23086602985858917,
      "learning_rate": 3.841e-05,
      "loss": 0.3979,
      "step": 1159
    },
    {
      "epoch": 105.48484848484848,
      "grad_norm": 0.2298888862133026,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 0.4567,
      "step": 1160
    },
    {
      "epoch": 105.48484848484848,
      "eval_loss": 0.5147726535797119,
      "eval_runtime": 2.1655,
      "eval_samples_per_second": 25.398,
      "eval_steps_per_second": 3.232,
      "step": 1160
    },
    {
      "epoch": 105.58181818181818,
      "grad_norm": 0.3209298849105835,
      "learning_rate": 3.8390000000000003e-05,
      "loss": 0.4767,
      "step": 1161
    },
    {
      "epoch": 105.67878787878787,
      "grad_norm": 0.3022819459438324,
      "learning_rate": 3.838e-05,
      "loss": 0.4814,
      "step": 1162
    },
    {
      "epoch": 105.77575757575758,
      "grad_norm": 0.25658342242240906,
      "learning_rate": 3.837e-05,
      "loss": 0.4787,
      "step": 1163
    },
    {
      "epoch": 105.87272727272727,
      "grad_norm": 0.21234485507011414,
      "learning_rate": 3.836e-05,
      "loss": 0.4734,
      "step": 1164
    },
    {
      "epoch": 105.96969696969697,
      "grad_norm": 0.23950757086277008,
      "learning_rate": 3.8350000000000004e-05,
      "loss": 0.5033,
      "step": 1165
    },
    {
      "epoch": 106.0,
      "grad_norm": 0.4254544675350189,
      "learning_rate": 3.834e-05,
      "loss": 0.4682,
      "step": 1166
    },
    {
      "epoch": 106.0969696969697,
      "grad_norm": 0.21771714091300964,
      "learning_rate": 3.833e-05,
      "loss": 0.4728,
      "step": 1167
    },
    {
      "epoch": 106.19393939393939,
      "grad_norm": 0.2927077114582062,
      "learning_rate": 3.832e-05,
      "loss": 0.4664,
      "step": 1168
    },
    {
      "epoch": 106.2909090909091,
      "grad_norm": 0.2586457133293152,
      "learning_rate": 3.8310000000000004e-05,
      "loss": 0.4815,
      "step": 1169
    },
    {
      "epoch": 106.38787878787879,
      "grad_norm": 0.24301891028881073,
      "learning_rate": 3.83e-05,
      "loss": 0.4689,
      "step": 1170
    },
    {
      "epoch": 106.38787878787879,
      "eval_loss": 0.513548731803894,
      "eval_runtime": 2.1846,
      "eval_samples_per_second": 25.177,
      "eval_steps_per_second": 3.204,
      "step": 1170
    },
    {
      "epoch": 106.48484848484848,
      "grad_norm": 0.25214365124702454,
      "learning_rate": 3.829e-05,
      "loss": 0.455,
      "step": 1171
    },
    {
      "epoch": 106.58181818181818,
      "grad_norm": 0.28379714488983154,
      "learning_rate": 3.828e-05,
      "loss": 0.458,
      "step": 1172
    },
    {
      "epoch": 106.67878787878787,
      "grad_norm": 0.3102569282054901,
      "learning_rate": 3.827e-05,
      "loss": 0.4689,
      "step": 1173
    },
    {
      "epoch": 106.77575757575758,
      "grad_norm": 0.22973500192165375,
      "learning_rate": 3.826e-05,
      "loss": 0.4404,
      "step": 1174
    },
    {
      "epoch": 106.87272727272727,
      "grad_norm": 0.2559613585472107,
      "learning_rate": 3.825e-05,
      "loss": 0.4613,
      "step": 1175
    },
    {
      "epoch": 106.96969696969697,
      "grad_norm": 0.21462607383728027,
      "learning_rate": 3.8240000000000007e-05,
      "loss": 0.4623,
      "step": 1176
    },
    {
      "epoch": 107.0,
      "grad_norm": 0.4708159565925598,
      "learning_rate": 3.823e-05,
      "loss": 0.5488,
      "step": 1177
    },
    {
      "epoch": 107.0969696969697,
      "grad_norm": 0.24319447576999664,
      "learning_rate": 3.822e-05,
      "loss": 0.4784,
      "step": 1178
    },
    {
      "epoch": 107.19393939393939,
      "grad_norm": 0.32229968905448914,
      "learning_rate": 3.821e-05,
      "loss": 0.4585,
      "step": 1179
    },
    {
      "epoch": 107.2909090909091,
      "grad_norm": 0.2953487038612366,
      "learning_rate": 3.82e-05,
      "loss": 0.4459,
      "step": 1180
    },
    {
      "epoch": 107.2909090909091,
      "eval_loss": 0.5121553540229797,
      "eval_runtime": 2.177,
      "eval_samples_per_second": 25.264,
      "eval_steps_per_second": 3.215,
      "step": 1180
    },
    {
      "epoch": 107.38787878787879,
      "grad_norm": 0.2409444898366928,
      "learning_rate": 3.8190000000000005e-05,
      "loss": 0.4875,
      "step": 1181
    },
    {
      "epoch": 107.48484848484848,
      "grad_norm": 0.2681051194667816,
      "learning_rate": 3.818e-05,
      "loss": 0.4251,
      "step": 1182
    },
    {
      "epoch": 107.58181818181818,
      "grad_norm": 0.3105945289134979,
      "learning_rate": 3.817e-05,
      "loss": 0.4656,
      "step": 1183
    },
    {
      "epoch": 107.67878787878787,
      "grad_norm": 0.27236083149909973,
      "learning_rate": 3.816e-05,
      "loss": 0.4998,
      "step": 1184
    },
    {
      "epoch": 107.77575757575758,
      "grad_norm": 0.27155423164367676,
      "learning_rate": 3.8150000000000006e-05,
      "loss": 0.482,
      "step": 1185
    },
    {
      "epoch": 107.87272727272727,
      "grad_norm": 0.23208624124526978,
      "learning_rate": 3.8140000000000004e-05,
      "loss": 0.4508,
      "step": 1186
    },
    {
      "epoch": 107.96969696969697,
      "grad_norm": 0.2420530915260315,
      "learning_rate": 3.8129999999999996e-05,
      "loss": 0.4653,
      "step": 1187
    },
    {
      "epoch": 108.0,
      "grad_norm": 0.3872806131839752,
      "learning_rate": 3.812e-05,
      "loss": 0.423,
      "step": 1188
    },
    {
      "epoch": 108.0969696969697,
      "grad_norm": 0.29809868335723877,
      "learning_rate": 3.811e-05,
      "loss": 0.4195,
      "step": 1189
    },
    {
      "epoch": 108.19393939393939,
      "grad_norm": 0.2518772482872009,
      "learning_rate": 3.8100000000000005e-05,
      "loss": 0.4722,
      "step": 1190
    },
    {
      "epoch": 108.19393939393939,
      "eval_loss": 0.51094651222229,
      "eval_runtime": 2.1706,
      "eval_samples_per_second": 25.339,
      "eval_steps_per_second": 3.225,
      "step": 1190
    },
    {
      "epoch": 108.2909090909091,
      "grad_norm": 0.33083200454711914,
      "learning_rate": 3.809e-05,
      "loss": 0.434,
      "step": 1191
    },
    {
      "epoch": 108.38787878787879,
      "grad_norm": 0.2682957351207733,
      "learning_rate": 3.808e-05,
      "loss": 0.4792,
      "step": 1192
    },
    {
      "epoch": 108.48484848484848,
      "grad_norm": 0.2922620177268982,
      "learning_rate": 3.807e-05,
      "loss": 0.4703,
      "step": 1193
    },
    {
      "epoch": 108.58181818181818,
      "grad_norm": 0.33325645327568054,
      "learning_rate": 3.806e-05,
      "loss": 0.529,
      "step": 1194
    },
    {
      "epoch": 108.67878787878787,
      "grad_norm": 0.2711305022239685,
      "learning_rate": 3.805e-05,
      "loss": 0.4517,
      "step": 1195
    },
    {
      "epoch": 108.77575757575758,
      "grad_norm": 0.2560056149959564,
      "learning_rate": 3.804e-05,
      "loss": 0.4792,
      "step": 1196
    },
    {
      "epoch": 108.87272727272727,
      "grad_norm": 0.25109291076660156,
      "learning_rate": 3.803000000000001e-05,
      "loss": 0.4449,
      "step": 1197
    },
    {
      "epoch": 108.96969696969697,
      "grad_norm": 0.23504619300365448,
      "learning_rate": 3.802e-05,
      "loss": 0.466,
      "step": 1198
    },
    {
      "epoch": 109.0,
      "grad_norm": 0.45923951268196106,
      "learning_rate": 3.8010000000000004e-05,
      "loss": 0.406,
      "step": 1199
    },
    {
      "epoch": 109.0969696969697,
      "grad_norm": 0.2102024257183075,
      "learning_rate": 3.8e-05,
      "loss": 0.4256,
      "step": 1200
    },
    {
      "epoch": 109.0969696969697,
      "eval_loss": 0.5096712112426758,
      "eval_runtime": 2.183,
      "eval_samples_per_second": 25.195,
      "eval_steps_per_second": 3.207,
      "step": 1200
    },
    {
      "epoch": 109.19393939393939,
      "grad_norm": 0.24149499833583832,
      "learning_rate": 3.799e-05,
      "loss": 0.4755,
      "step": 1201
    },
    {
      "epoch": 109.2909090909091,
      "grad_norm": 0.20073340833187103,
      "learning_rate": 3.7980000000000006e-05,
      "loss": 0.4245,
      "step": 1202
    },
    {
      "epoch": 109.38787878787879,
      "grad_norm": 0.24485673010349274,
      "learning_rate": 3.797e-05,
      "loss": 0.4567,
      "step": 1203
    },
    {
      "epoch": 109.48484848484848,
      "grad_norm": 0.26820191740989685,
      "learning_rate": 3.796e-05,
      "loss": 0.4445,
      "step": 1204
    },
    {
      "epoch": 109.58181818181818,
      "grad_norm": 0.24449220299720764,
      "learning_rate": 3.795e-05,
      "loss": 0.5022,
      "step": 1205
    },
    {
      "epoch": 109.67878787878787,
      "grad_norm": 0.259132444858551,
      "learning_rate": 3.7940000000000006e-05,
      "loss": 0.4861,
      "step": 1206
    },
    {
      "epoch": 109.77575757575758,
      "grad_norm": 0.28169506788253784,
      "learning_rate": 3.7930000000000004e-05,
      "loss": 0.4712,
      "step": 1207
    },
    {
      "epoch": 109.87272727272727,
      "grad_norm": 0.2922845780849457,
      "learning_rate": 3.792e-05,
      "loss": 0.4567,
      "step": 1208
    },
    {
      "epoch": 109.96969696969697,
      "grad_norm": 0.25783708691596985,
      "learning_rate": 3.791e-05,
      "loss": 0.4752,
      "step": 1209
    },
    {
      "epoch": 110.0,
      "grad_norm": 0.38848626613616943,
      "learning_rate": 3.79e-05,
      "loss": 0.4511,
      "step": 1210
    },
    {
      "epoch": 110.0,
      "eval_loss": 0.5089406967163086,
      "eval_runtime": 2.178,
      "eval_samples_per_second": 25.253,
      "eval_steps_per_second": 3.214,
      "step": 1210
    },
    {
      "epoch": 110.0969696969697,
      "grad_norm": 0.24029256403446198,
      "learning_rate": 3.7890000000000005e-05,
      "loss": 0.4672,
      "step": 1211
    },
    {
      "epoch": 110.19393939393939,
      "grad_norm": 0.25455084443092346,
      "learning_rate": 3.788e-05,
      "loss": 0.4671,
      "step": 1212
    },
    {
      "epoch": 110.2909090909091,
      "grad_norm": 0.29211434721946716,
      "learning_rate": 3.787e-05,
      "loss": 0.4523,
      "step": 1213
    },
    {
      "epoch": 110.38787878787879,
      "grad_norm": 0.31663522124290466,
      "learning_rate": 3.786e-05,
      "loss": 0.4581,
      "step": 1214
    },
    {
      "epoch": 110.48484848484848,
      "grad_norm": 0.23833386600017548,
      "learning_rate": 3.7850000000000005e-05,
      "loss": 0.4409,
      "step": 1215
    },
    {
      "epoch": 110.58181818181818,
      "grad_norm": 0.3347303569316864,
      "learning_rate": 3.7840000000000004e-05,
      "loss": 0.512,
      "step": 1216
    },
    {
      "epoch": 110.67878787878787,
      "grad_norm": 0.26010632514953613,
      "learning_rate": 3.783e-05,
      "loss": 0.4411,
      "step": 1217
    },
    {
      "epoch": 110.77575757575758,
      "grad_norm": 0.27888065576553345,
      "learning_rate": 3.782e-05,
      "loss": 0.4628,
      "step": 1218
    },
    {
      "epoch": 110.87272727272727,
      "grad_norm": 0.2746562957763672,
      "learning_rate": 3.781e-05,
      "loss": 0.4374,
      "step": 1219
    },
    {
      "epoch": 110.96969696969697,
      "grad_norm": 0.27418023347854614,
      "learning_rate": 3.7800000000000004e-05,
      "loss": 0.4701,
      "step": 1220
    },
    {
      "epoch": 110.96969696969697,
      "eval_loss": 0.5076990723609924,
      "eval_runtime": 2.1853,
      "eval_samples_per_second": 25.168,
      "eval_steps_per_second": 3.203,
      "step": 1220
    },
    {
      "epoch": 111.0,
      "grad_norm": 0.4324119985103607,
      "learning_rate": 3.779e-05,
      "loss": 0.4461,
      "step": 1221
    },
    {
      "epoch": 111.0969696969697,
      "grad_norm": 0.26746389269828796,
      "learning_rate": 3.778000000000001e-05,
      "loss": 0.4694,
      "step": 1222
    },
    {
      "epoch": 111.19393939393939,
      "grad_norm": 0.2583480477333069,
      "learning_rate": 3.777e-05,
      "loss": 0.459,
      "step": 1223
    },
    {
      "epoch": 111.2909090909091,
      "grad_norm": 0.25809216499328613,
      "learning_rate": 3.776e-05,
      "loss": 0.4464,
      "step": 1224
    },
    {
      "epoch": 111.38787878787879,
      "grad_norm": 0.2889714539051056,
      "learning_rate": 3.775e-05,
      "loss": 0.4636,
      "step": 1225
    },
    {
      "epoch": 111.48484848484848,
      "grad_norm": 0.30283504724502563,
      "learning_rate": 3.774e-05,
      "loss": 0.4838,
      "step": 1226
    },
    {
      "epoch": 111.58181818181818,
      "grad_norm": 0.24876466393470764,
      "learning_rate": 3.7730000000000006e-05,
      "loss": 0.4623,
      "step": 1227
    },
    {
      "epoch": 111.67878787878787,
      "grad_norm": 0.23016254603862762,
      "learning_rate": 3.772e-05,
      "loss": 0.4898,
      "step": 1228
    },
    {
      "epoch": 111.77575757575758,
      "grad_norm": 0.2642112970352173,
      "learning_rate": 3.771e-05,
      "loss": 0.4576,
      "step": 1229
    },
    {
      "epoch": 111.87272727272727,
      "grad_norm": 0.24856284260749817,
      "learning_rate": 3.77e-05,
      "loss": 0.4255,
      "step": 1230
    },
    {
      "epoch": 111.87272727272727,
      "eval_loss": 0.5072154998779297,
      "eval_runtime": 2.1804,
      "eval_samples_per_second": 25.225,
      "eval_steps_per_second": 3.21,
      "step": 1230
    },
    {
      "epoch": 111.96969696969697,
      "grad_norm": 0.2232726514339447,
      "learning_rate": 3.769e-05,
      "loss": 0.4518,
      "step": 1231
    },
    {
      "epoch": 112.0,
      "grad_norm": 0.38369572162628174,
      "learning_rate": 3.7680000000000005e-05,
      "loss": 0.3925,
      "step": 1232
    },
    {
      "epoch": 112.0969696969697,
      "grad_norm": 0.2862841784954071,
      "learning_rate": 3.767e-05,
      "loss": 0.4098,
      "step": 1233
    },
    {
      "epoch": 112.19393939393939,
      "grad_norm": 0.28138405084609985,
      "learning_rate": 3.766e-05,
      "loss": 0.5203,
      "step": 1234
    },
    {
      "epoch": 112.2909090909091,
      "grad_norm": 0.26404833793640137,
      "learning_rate": 3.765e-05,
      "loss": 0.4698,
      "step": 1235
    },
    {
      "epoch": 112.38787878787879,
      "grad_norm": 0.21245059370994568,
      "learning_rate": 3.7640000000000006e-05,
      "loss": 0.4848,
      "step": 1236
    },
    {
      "epoch": 112.48484848484848,
      "grad_norm": 0.2753354609012604,
      "learning_rate": 3.7630000000000004e-05,
      "loss": 0.4127,
      "step": 1237
    },
    {
      "epoch": 112.58181818181818,
      "grad_norm": 0.3698137700557709,
      "learning_rate": 3.762e-05,
      "loss": 0.3973,
      "step": 1238
    },
    {
      "epoch": 112.67878787878787,
      "grad_norm": 0.23089006543159485,
      "learning_rate": 3.761e-05,
      "loss": 0.4356,
      "step": 1239
    },
    {
      "epoch": 112.77575757575758,
      "grad_norm": 0.24908600747585297,
      "learning_rate": 3.76e-05,
      "loss": 0.4809,
      "step": 1240
    },
    {
      "epoch": 112.77575757575758,
      "eval_loss": 0.5066096782684326,
      "eval_runtime": 2.1764,
      "eval_samples_per_second": 25.271,
      "eval_steps_per_second": 3.216,
      "step": 1240
    },
    {
      "epoch": 112.87272727272727,
      "grad_norm": 0.2229243665933609,
      "learning_rate": 3.7590000000000004e-05,
      "loss": 0.4359,
      "step": 1241
    },
    {
      "epoch": 112.96969696969697,
      "grad_norm": 0.3743974268436432,
      "learning_rate": 3.758e-05,
      "loss": 0.4945,
      "step": 1242
    },
    {
      "epoch": 113.0,
      "grad_norm": 0.6298211812973022,
      "learning_rate": 3.757e-05,
      "loss": 0.5496,
      "step": 1243
    },
    {
      "epoch": 113.0969696969697,
      "grad_norm": 0.3047841787338257,
      "learning_rate": 3.756e-05,
      "loss": 0.4694,
      "step": 1244
    },
    {
      "epoch": 113.19393939393939,
      "grad_norm": 0.22005103528499603,
      "learning_rate": 3.7550000000000005e-05,
      "loss": 0.4298,
      "step": 1245
    },
    {
      "epoch": 113.2909090909091,
      "grad_norm": 0.265865296125412,
      "learning_rate": 3.754e-05,
      "loss": 0.4511,
      "step": 1246
    },
    {
      "epoch": 113.38787878787879,
      "grad_norm": 0.2695404887199402,
      "learning_rate": 3.753e-05,
      "loss": 0.4889,
      "step": 1247
    },
    {
      "epoch": 113.48484848484848,
      "grad_norm": 0.26106733083724976,
      "learning_rate": 3.752e-05,
      "loss": 0.4412,
      "step": 1248
    },
    {
      "epoch": 113.58181818181818,
      "grad_norm": 0.27255240082740784,
      "learning_rate": 3.751e-05,
      "loss": 0.4569,
      "step": 1249
    },
    {
      "epoch": 113.67878787878787,
      "grad_norm": 0.23935070633888245,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.45,
      "step": 1250
    },
    {
      "epoch": 113.67878787878787,
      "eval_loss": 0.5050410032272339,
      "eval_runtime": 2.1859,
      "eval_samples_per_second": 25.161,
      "eval_steps_per_second": 3.202,
      "step": 1250
    },
    {
      "epoch": 113.77575757575758,
      "grad_norm": 0.25609126687049866,
      "learning_rate": 3.749e-05,
      "loss": 0.4751,
      "step": 1251
    },
    {
      "epoch": 113.87272727272727,
      "grad_norm": 0.28254762291908264,
      "learning_rate": 3.748000000000001e-05,
      "loss": 0.4178,
      "step": 1252
    },
    {
      "epoch": 113.96969696969697,
      "grad_norm": 0.2685804069042206,
      "learning_rate": 3.747e-05,
      "loss": 0.4401,
      "step": 1253
    },
    {
      "epoch": 114.0,
      "grad_norm": 0.6137757301330566,
      "learning_rate": 3.7460000000000004e-05,
      "loss": 0.5856,
      "step": 1254
    },
    {
      "epoch": 114.0969696969697,
      "grad_norm": 0.25804847478866577,
      "learning_rate": 3.745e-05,
      "loss": 0.4107,
      "step": 1255
    },
    {
      "epoch": 114.19393939393939,
      "grad_norm": 0.24832597374916077,
      "learning_rate": 3.744e-05,
      "loss": 0.5157,
      "step": 1256
    },
    {
      "epoch": 114.2909090909091,
      "grad_norm": 0.30176836252212524,
      "learning_rate": 3.7430000000000006e-05,
      "loss": 0.4617,
      "step": 1257
    },
    {
      "epoch": 114.38787878787879,
      "grad_norm": 0.2873697280883789,
      "learning_rate": 3.742e-05,
      "loss": 0.4844,
      "step": 1258
    },
    {
      "epoch": 114.48484848484848,
      "grad_norm": 0.3384517431259155,
      "learning_rate": 3.741e-05,
      "loss": 0.4438,
      "step": 1259
    },
    {
      "epoch": 114.58181818181818,
      "grad_norm": 0.26792243123054504,
      "learning_rate": 3.74e-05,
      "loss": 0.4231,
      "step": 1260
    },
    {
      "epoch": 114.58181818181818,
      "eval_loss": 0.5039417147636414,
      "eval_runtime": 2.1877,
      "eval_samples_per_second": 25.14,
      "eval_steps_per_second": 3.2,
      "step": 1260
    },
    {
      "epoch": 114.67878787878787,
      "grad_norm": 0.22581951320171356,
      "learning_rate": 3.739e-05,
      "loss": 0.4391,
      "step": 1261
    },
    {
      "epoch": 114.77575757575758,
      "grad_norm": 0.27811703085899353,
      "learning_rate": 3.7380000000000005e-05,
      "loss": 0.4762,
      "step": 1262
    },
    {
      "epoch": 114.87272727272727,
      "grad_norm": 0.28211668133735657,
      "learning_rate": 3.7369999999999996e-05,
      "loss": 0.4486,
      "step": 1263
    },
    {
      "epoch": 114.96969696969697,
      "grad_norm": 0.35375770926475525,
      "learning_rate": 3.736e-05,
      "loss": 0.4255,
      "step": 1264
    },
    {
      "epoch": 115.0,
      "grad_norm": 0.40355023741722107,
      "learning_rate": 3.735e-05,
      "loss": 0.5068,
      "step": 1265
    },
    {
      "epoch": 115.0969696969697,
      "grad_norm": 0.22719883918762207,
      "learning_rate": 3.7340000000000005e-05,
      "loss": 0.4527,
      "step": 1266
    },
    {
      "epoch": 115.19393939393939,
      "grad_norm": 0.2507949471473694,
      "learning_rate": 3.7330000000000003e-05,
      "loss": 0.4177,
      "step": 1267
    },
    {
      "epoch": 115.2909090909091,
      "grad_norm": 0.3059385120868683,
      "learning_rate": 3.732e-05,
      "loss": 0.4506,
      "step": 1268
    },
    {
      "epoch": 115.38787878787879,
      "grad_norm": 0.24009016156196594,
      "learning_rate": 3.731e-05,
      "loss": 0.4618,
      "step": 1269
    },
    {
      "epoch": 115.48484848484848,
      "grad_norm": 0.2662716507911682,
      "learning_rate": 3.73e-05,
      "loss": 0.4522,
      "step": 1270
    },
    {
      "epoch": 115.48484848484848,
      "eval_loss": 0.5031892657279968,
      "eval_runtime": 2.1723,
      "eval_samples_per_second": 25.318,
      "eval_steps_per_second": 3.222,
      "step": 1270
    },
    {
      "epoch": 115.58181818181818,
      "grad_norm": 0.2552890181541443,
      "learning_rate": 3.7290000000000004e-05,
      "loss": 0.4387,
      "step": 1271
    },
    {
      "epoch": 115.67878787878787,
      "grad_norm": 0.304675430059433,
      "learning_rate": 3.728e-05,
      "loss": 0.5211,
      "step": 1272
    },
    {
      "epoch": 115.77575757575758,
      "grad_norm": 0.3080827593803406,
      "learning_rate": 3.727e-05,
      "loss": 0.4216,
      "step": 1273
    },
    {
      "epoch": 115.87272727272727,
      "grad_norm": 0.2266256958246231,
      "learning_rate": 3.726e-05,
      "loss": 0.4419,
      "step": 1274
    },
    {
      "epoch": 115.96969696969697,
      "grad_norm": 0.3163573145866394,
      "learning_rate": 3.7250000000000004e-05,
      "loss": 0.475,
      "step": 1275
    },
    {
      "epoch": 116.0,
      "grad_norm": 0.36579272150993347,
      "learning_rate": 3.724e-05,
      "loss": 0.4423,
      "step": 1276
    },
    {
      "epoch": 116.0969696969697,
      "grad_norm": 0.25317344069480896,
      "learning_rate": 3.723e-05,
      "loss": 0.4364,
      "step": 1277
    },
    {
      "epoch": 116.19393939393939,
      "grad_norm": 0.23825958371162415,
      "learning_rate": 3.722e-05,
      "loss": 0.4653,
      "step": 1278
    },
    {
      "epoch": 116.2909090909091,
      "grad_norm": 0.27158665657043457,
      "learning_rate": 3.721e-05,
      "loss": 0.4417,
      "step": 1279
    },
    {
      "epoch": 116.38787878787879,
      "grad_norm": 0.32222259044647217,
      "learning_rate": 3.72e-05,
      "loss": 0.4636,
      "step": 1280
    },
    {
      "epoch": 116.38787878787879,
      "eval_loss": 0.5025956034660339,
      "eval_runtime": 2.1635,
      "eval_samples_per_second": 25.422,
      "eval_steps_per_second": 3.236,
      "step": 1280
    },
    {
      "epoch": 116.48484848484848,
      "grad_norm": 0.2667633593082428,
      "learning_rate": 3.719e-05,
      "loss": 0.4709,
      "step": 1281
    },
    {
      "epoch": 116.58181818181818,
      "grad_norm": 0.2588888108730316,
      "learning_rate": 3.7180000000000007e-05,
      "loss": 0.4164,
      "step": 1282
    },
    {
      "epoch": 116.67878787878787,
      "grad_norm": 0.2974129617214203,
      "learning_rate": 3.717e-05,
      "loss": 0.432,
      "step": 1283
    },
    {
      "epoch": 116.77575757575758,
      "grad_norm": 0.24473229050636292,
      "learning_rate": 3.716e-05,
      "loss": 0.4942,
      "step": 1284
    },
    {
      "epoch": 116.87272727272727,
      "grad_norm": 0.2632085084915161,
      "learning_rate": 3.715e-05,
      "loss": 0.4465,
      "step": 1285
    },
    {
      "epoch": 116.96969696969697,
      "grad_norm": 0.29160937666893005,
      "learning_rate": 3.714e-05,
      "loss": 0.442,
      "step": 1286
    },
    {
      "epoch": 117.0,
      "grad_norm": 0.40403980016708374,
      "learning_rate": 3.7130000000000005e-05,
      "loss": 0.4771,
      "step": 1287
    },
    {
      "epoch": 117.0969696969697,
      "grad_norm": 0.3258376717567444,
      "learning_rate": 3.712e-05,
      "loss": 0.4886,
      "step": 1288
    },
    {
      "epoch": 117.19393939393939,
      "grad_norm": 0.24444925785064697,
      "learning_rate": 3.711e-05,
      "loss": 0.4539,
      "step": 1289
    },
    {
      "epoch": 117.2909090909091,
      "grad_norm": 0.24981863796710968,
      "learning_rate": 3.71e-05,
      "loss": 0.4257,
      "step": 1290
    },
    {
      "epoch": 117.2909090909091,
      "eval_loss": 0.5017372369766235,
      "eval_runtime": 2.1651,
      "eval_samples_per_second": 25.403,
      "eval_steps_per_second": 3.233,
      "step": 1290
    },
    {
      "epoch": 117.38787878787879,
      "grad_norm": 0.22072580456733704,
      "learning_rate": 3.7090000000000006e-05,
      "loss": 0.4551,
      "step": 1291
    },
    {
      "epoch": 117.48484848484848,
      "grad_norm": 0.24209479987621307,
      "learning_rate": 3.7080000000000004e-05,
      "loss": 0.4591,
      "step": 1292
    },
    {
      "epoch": 117.58181818181818,
      "grad_norm": 0.2738741934299469,
      "learning_rate": 3.707e-05,
      "loss": 0.4762,
      "step": 1293
    },
    {
      "epoch": 117.67878787878787,
      "grad_norm": 0.22845780849456787,
      "learning_rate": 3.706e-05,
      "loss": 0.4312,
      "step": 1294
    },
    {
      "epoch": 117.77575757575758,
      "grad_norm": 0.2801177203655243,
      "learning_rate": 3.705e-05,
      "loss": 0.452,
      "step": 1295
    },
    {
      "epoch": 117.87272727272727,
      "grad_norm": 0.25144457817077637,
      "learning_rate": 3.7040000000000005e-05,
      "loss": 0.4597,
      "step": 1296
    },
    {
      "epoch": 117.96969696969697,
      "grad_norm": 0.22239412367343903,
      "learning_rate": 3.703e-05,
      "loss": 0.4181,
      "step": 1297
    },
    {
      "epoch": 118.0,
      "grad_norm": 0.4170636832714081,
      "learning_rate": 3.702e-05,
      "loss": 0.4162,
      "step": 1298
    },
    {
      "epoch": 118.0969696969697,
      "grad_norm": 0.2747899889945984,
      "learning_rate": 3.701e-05,
      "loss": 0.4957,
      "step": 1299
    },
    {
      "epoch": 118.19393939393939,
      "grad_norm": 0.31099769473075867,
      "learning_rate": 3.7e-05,
      "loss": 0.4575,
      "step": 1300
    },
    {
      "epoch": 118.19393939393939,
      "eval_loss": 0.500615119934082,
      "eval_runtime": 2.1657,
      "eval_samples_per_second": 25.396,
      "eval_steps_per_second": 3.232,
      "step": 1300
    },
    {
      "epoch": 118.2909090909091,
      "grad_norm": 0.3002060055732727,
      "learning_rate": 3.699e-05,
      "loss": 0.4245,
      "step": 1301
    },
    {
      "epoch": 118.38787878787879,
      "grad_norm": 0.21696718037128448,
      "learning_rate": 3.698e-05,
      "loss": 0.5023,
      "step": 1302
    },
    {
      "epoch": 118.48484848484848,
      "grad_norm": 0.2654174566268921,
      "learning_rate": 3.697e-05,
      "loss": 0.4316,
      "step": 1303
    },
    {
      "epoch": 118.58181818181818,
      "grad_norm": 0.28204402327537537,
      "learning_rate": 3.696e-05,
      "loss": 0.4399,
      "step": 1304
    },
    {
      "epoch": 118.67878787878787,
      "grad_norm": 0.25560328364372253,
      "learning_rate": 3.6950000000000004e-05,
      "loss": 0.3992,
      "step": 1305
    },
    {
      "epoch": 118.77575757575758,
      "grad_norm": 0.2698306441307068,
      "learning_rate": 3.694e-05,
      "loss": 0.4552,
      "step": 1306
    },
    {
      "epoch": 118.87272727272727,
      "grad_norm": 0.26093652844429016,
      "learning_rate": 3.693e-05,
      "loss": 0.4587,
      "step": 1307
    },
    {
      "epoch": 118.96969696969697,
      "grad_norm": 0.30971360206604004,
      "learning_rate": 3.692e-05,
      "loss": 0.4194,
      "step": 1308
    },
    {
      "epoch": 119.0,
      "grad_norm": 0.4734983444213867,
      "learning_rate": 3.691e-05,
      "loss": 0.4719,
      "step": 1309
    },
    {
      "epoch": 119.0969696969697,
      "grad_norm": 0.32670915126800537,
      "learning_rate": 3.69e-05,
      "loss": 0.4586,
      "step": 1310
    },
    {
      "epoch": 119.0969696969697,
      "eval_loss": 0.5000902414321899,
      "eval_runtime": 2.2185,
      "eval_samples_per_second": 24.792,
      "eval_steps_per_second": 3.155,
      "step": 1310
    },
    {
      "epoch": 119.19393939393939,
      "grad_norm": 0.22537638247013092,
      "learning_rate": 3.689e-05,
      "loss": 0.477,
      "step": 1311
    },
    {
      "epoch": 119.2909090909091,
      "grad_norm": 0.258887380361557,
      "learning_rate": 3.6880000000000006e-05,
      "loss": 0.4402,
      "step": 1312
    },
    {
      "epoch": 119.38787878787879,
      "grad_norm": 0.26598671078681946,
      "learning_rate": 3.6870000000000004e-05,
      "loss": 0.4254,
      "step": 1313
    },
    {
      "epoch": 119.48484848484848,
      "grad_norm": 0.23355336487293243,
      "learning_rate": 3.686e-05,
      "loss": 0.4387,
      "step": 1314
    },
    {
      "epoch": 119.58181818181818,
      "grad_norm": 0.2759830355644226,
      "learning_rate": 3.685e-05,
      "loss": 0.4834,
      "step": 1315
    },
    {
      "epoch": 119.67878787878787,
      "grad_norm": 0.24471092224121094,
      "learning_rate": 3.684e-05,
      "loss": 0.4632,
      "step": 1316
    },
    {
      "epoch": 119.77575757575758,
      "grad_norm": 0.23647384345531464,
      "learning_rate": 3.6830000000000005e-05,
      "loss": 0.3989,
      "step": 1317
    },
    {
      "epoch": 119.87272727272727,
      "grad_norm": 0.2802410125732422,
      "learning_rate": 3.682e-05,
      "loss": 0.4359,
      "step": 1318
    },
    {
      "epoch": 119.96969696969697,
      "grad_norm": 0.21583354473114014,
      "learning_rate": 3.681e-05,
      "loss": 0.4621,
      "step": 1319
    },
    {
      "epoch": 120.0,
      "grad_norm": 0.40480583906173706,
      "learning_rate": 3.68e-05,
      "loss": 0.4294,
      "step": 1320
    },
    {
      "epoch": 120.0,
      "eval_loss": 0.4989987015724182,
      "eval_runtime": 2.138,
      "eval_samples_per_second": 25.725,
      "eval_steps_per_second": 3.274,
      "step": 1320
    },
    {
      "epoch": 120.0969696969697,
      "grad_norm": 0.3164379894733429,
      "learning_rate": 3.6790000000000005e-05,
      "loss": 0.4802,
      "step": 1321
    },
    {
      "epoch": 120.19393939393939,
      "grad_norm": 0.31919732689857483,
      "learning_rate": 3.6780000000000004e-05,
      "loss": 0.4447,
      "step": 1322
    },
    {
      "epoch": 120.2909090909091,
      "grad_norm": 0.29702770709991455,
      "learning_rate": 3.677e-05,
      "loss": 0.4632,
      "step": 1323
    },
    {
      "epoch": 120.38787878787879,
      "grad_norm": 0.20388232171535492,
      "learning_rate": 3.676e-05,
      "loss": 0.4442,
      "step": 1324
    },
    {
      "epoch": 120.48484848484848,
      "grad_norm": 0.22363580763339996,
      "learning_rate": 3.675e-05,
      "loss": 0.4455,
      "step": 1325
    },
    {
      "epoch": 120.58181818181818,
      "grad_norm": 0.24787390232086182,
      "learning_rate": 3.6740000000000004e-05,
      "loss": 0.4437,
      "step": 1326
    },
    {
      "epoch": 120.67878787878787,
      "grad_norm": 0.23430295288562775,
      "learning_rate": 3.673e-05,
      "loss": 0.3896,
      "step": 1327
    },
    {
      "epoch": 120.77575757575758,
      "grad_norm": 0.24690541625022888,
      "learning_rate": 3.672000000000001e-05,
      "loss": 0.4307,
      "step": 1328
    },
    {
      "epoch": 120.87272727272727,
      "grad_norm": 0.3323303759098053,
      "learning_rate": 3.671e-05,
      "loss": 0.4549,
      "step": 1329
    },
    {
      "epoch": 120.96969696969697,
      "grad_norm": 0.29622218012809753,
      "learning_rate": 3.6700000000000004e-05,
      "loss": 0.4643,
      "step": 1330
    },
    {
      "epoch": 120.96969696969697,
      "eval_loss": 0.4983088970184326,
      "eval_runtime": 2.1516,
      "eval_samples_per_second": 25.563,
      "eval_steps_per_second": 3.253,
      "step": 1330
    },
    {
      "epoch": 121.0,
      "grad_norm": 0.5410318970680237,
      "learning_rate": 3.669e-05,
      "loss": 0.468,
      "step": 1331
    },
    {
      "epoch": 121.0969696969697,
      "grad_norm": 0.20910969376564026,
      "learning_rate": 3.668e-05,
      "loss": 0.4611,
      "step": 1332
    },
    {
      "epoch": 121.19393939393939,
      "grad_norm": 0.39248064160346985,
      "learning_rate": 3.6670000000000006e-05,
      "loss": 0.3951,
      "step": 1333
    },
    {
      "epoch": 121.2909090909091,
      "grad_norm": 0.33087027072906494,
      "learning_rate": 3.666e-05,
      "loss": 0.4323,
      "step": 1334
    },
    {
      "epoch": 121.38787878787879,
      "grad_norm": 0.4120776951313019,
      "learning_rate": 3.665e-05,
      "loss": 0.4498,
      "step": 1335
    },
    {
      "epoch": 121.48484848484848,
      "grad_norm": 0.2515096962451935,
      "learning_rate": 3.664e-05,
      "loss": 0.4624,
      "step": 1336
    },
    {
      "epoch": 121.58181818181818,
      "grad_norm": 0.28713738918304443,
      "learning_rate": 3.663e-05,
      "loss": 0.4362,
      "step": 1337
    },
    {
      "epoch": 121.67878787878787,
      "grad_norm": 0.3005863130092621,
      "learning_rate": 3.6620000000000005e-05,
      "loss": 0.4994,
      "step": 1338
    },
    {
      "epoch": 121.77575757575758,
      "grad_norm": 0.3588126301765442,
      "learning_rate": 3.661e-05,
      "loss": 0.4499,
      "step": 1339
    },
    {
      "epoch": 121.87272727272727,
      "grad_norm": 0.34536463022232056,
      "learning_rate": 3.66e-05,
      "loss": 0.4284,
      "step": 1340
    },
    {
      "epoch": 121.87272727272727,
      "eval_loss": 0.49804064631462097,
      "eval_runtime": 2.1636,
      "eval_samples_per_second": 25.421,
      "eval_steps_per_second": 3.235,
      "step": 1340
    },
    {
      "epoch": 121.96969696969697,
      "grad_norm": 0.28389716148376465,
      "learning_rate": 3.659e-05,
      "loss": 0.4487,
      "step": 1341
    },
    {
      "epoch": 122.0,
      "grad_norm": 0.4554291367530823,
      "learning_rate": 3.6580000000000006e-05,
      "loss": 0.4294,
      "step": 1342
    },
    {
      "epoch": 122.0969696969697,
      "grad_norm": 0.24930481612682343,
      "learning_rate": 3.6570000000000004e-05,
      "loss": 0.4316,
      "step": 1343
    },
    {
      "epoch": 122.19393939393939,
      "grad_norm": 0.25624746084213257,
      "learning_rate": 3.656e-05,
      "loss": 0.4593,
      "step": 1344
    },
    {
      "epoch": 122.2909090909091,
      "grad_norm": 0.28764474391937256,
      "learning_rate": 3.655e-05,
      "loss": 0.4388,
      "step": 1345
    },
    {
      "epoch": 122.38787878787879,
      "grad_norm": 0.23803141713142395,
      "learning_rate": 3.654e-05,
      "loss": 0.4672,
      "step": 1346
    },
    {
      "epoch": 122.48484848484848,
      "grad_norm": 0.4183109700679779,
      "learning_rate": 3.6530000000000004e-05,
      "loss": 0.453,
      "step": 1347
    },
    {
      "epoch": 122.58181818181818,
      "grad_norm": 0.3076031506061554,
      "learning_rate": 3.652e-05,
      "loss": 0.4322,
      "step": 1348
    },
    {
      "epoch": 122.67878787878787,
      "grad_norm": 0.25950199365615845,
      "learning_rate": 3.651e-05,
      "loss": 0.4452,
      "step": 1349
    },
    {
      "epoch": 122.77575757575758,
      "grad_norm": 0.27765706181526184,
      "learning_rate": 3.65e-05,
      "loss": 0.4321,
      "step": 1350
    },
    {
      "epoch": 122.77575757575758,
      "eval_loss": 0.49684277176856995,
      "eval_runtime": 2.17,
      "eval_samples_per_second": 25.345,
      "eval_steps_per_second": 3.226,
      "step": 1350
    },
    {
      "epoch": 122.87272727272727,
      "grad_norm": 0.3624914288520813,
      "learning_rate": 3.6490000000000005e-05,
      "loss": 0.4507,
      "step": 1351
    },
    {
      "epoch": 122.96969696969697,
      "grad_norm": 0.29541799426078796,
      "learning_rate": 3.648e-05,
      "loss": 0.4287,
      "step": 1352
    },
    {
      "epoch": 123.0,
      "grad_norm": 0.4932088255882263,
      "learning_rate": 3.647e-05,
      "loss": 0.4679,
      "step": 1353
    },
    {
      "epoch": 123.0969696969697,
      "grad_norm": 0.4175611138343811,
      "learning_rate": 3.646e-05,
      "loss": 0.4432,
      "step": 1354
    },
    {
      "epoch": 123.19393939393939,
      "grad_norm": 0.3209320902824402,
      "learning_rate": 3.645e-05,
      "loss": 0.4445,
      "step": 1355
    },
    {
      "epoch": 123.2909090909091,
      "grad_norm": 0.378221720457077,
      "learning_rate": 3.6440000000000003e-05,
      "loss": 0.4699,
      "step": 1356
    },
    {
      "epoch": 123.38787878787879,
      "grad_norm": 0.3122067451477051,
      "learning_rate": 3.643e-05,
      "loss": 0.4392,
      "step": 1357
    },
    {
      "epoch": 123.48484848484848,
      "grad_norm": 0.25616055727005005,
      "learning_rate": 3.642000000000001e-05,
      "loss": 0.4524,
      "step": 1358
    },
    {
      "epoch": 123.58181818181818,
      "grad_norm": 0.27498358488082886,
      "learning_rate": 3.641e-05,
      "loss": 0.4341,
      "step": 1359
    },
    {
      "epoch": 123.67878787878787,
      "grad_norm": 0.25121861696243286,
      "learning_rate": 3.6400000000000004e-05,
      "loss": 0.4653,
      "step": 1360
    },
    {
      "epoch": 123.67878787878787,
      "eval_loss": 0.4961063265800476,
      "eval_runtime": 2.1514,
      "eval_samples_per_second": 25.565,
      "eval_steps_per_second": 3.254,
      "step": 1360
    },
    {
      "epoch": 123.77575757575758,
      "grad_norm": 0.3191700875759125,
      "learning_rate": 3.639e-05,
      "loss": 0.4405,
      "step": 1361
    },
    {
      "epoch": 123.87272727272727,
      "grad_norm": 0.30171266198158264,
      "learning_rate": 3.638e-05,
      "loss": 0.4325,
      "step": 1362
    },
    {
      "epoch": 123.96969696969697,
      "grad_norm": 0.3361068069934845,
      "learning_rate": 3.6370000000000006e-05,
      "loss": 0.4203,
      "step": 1363
    },
    {
      "epoch": 124.0,
      "grad_norm": 0.4331413209438324,
      "learning_rate": 3.636e-05,
      "loss": 0.4128,
      "step": 1364
    },
    {
      "epoch": 124.0969696969697,
      "grad_norm": 0.2644399106502533,
      "learning_rate": 3.635e-05,
      "loss": 0.4806,
      "step": 1365
    },
    {
      "epoch": 124.19393939393939,
      "grad_norm": 0.24663595855236053,
      "learning_rate": 3.634e-05,
      "loss": 0.4547,
      "step": 1366
    },
    {
      "epoch": 124.2909090909091,
      "grad_norm": 0.2665797770023346,
      "learning_rate": 3.6330000000000006e-05,
      "loss": 0.4051,
      "step": 1367
    },
    {
      "epoch": 124.38787878787879,
      "grad_norm": 0.3389240801334381,
      "learning_rate": 3.6320000000000005e-05,
      "loss": 0.4469,
      "step": 1368
    },
    {
      "epoch": 124.48484848484848,
      "grad_norm": 0.2690924108028412,
      "learning_rate": 3.6309999999999996e-05,
      "loss": 0.4341,
      "step": 1369
    },
    {
      "epoch": 124.58181818181818,
      "grad_norm": 0.24461796879768372,
      "learning_rate": 3.63e-05,
      "loss": 0.4346,
      "step": 1370
    },
    {
      "epoch": 124.58181818181818,
      "eval_loss": 0.4951164722442627,
      "eval_runtime": 2.1649,
      "eval_samples_per_second": 25.406,
      "eval_steps_per_second": 3.233,
      "step": 1370
    },
    {
      "epoch": 124.67878787878787,
      "grad_norm": 0.3070378601551056,
      "learning_rate": 3.629e-05,
      "loss": 0.4971,
      "step": 1371
    },
    {
      "epoch": 124.77575757575758,
      "grad_norm": 0.2280680537223816,
      "learning_rate": 3.6280000000000005e-05,
      "loss": 0.4394,
      "step": 1372
    },
    {
      "epoch": 124.87272727272727,
      "grad_norm": 0.2670609652996063,
      "learning_rate": 3.6270000000000003e-05,
      "loss": 0.4052,
      "step": 1373
    },
    {
      "epoch": 124.96969696969697,
      "grad_norm": 0.2442789077758789,
      "learning_rate": 3.626e-05,
      "loss": 0.4269,
      "step": 1374
    },
    {
      "epoch": 125.0,
      "grad_norm": 0.38368964195251465,
      "learning_rate": 3.625e-05,
      "loss": 0.4233,
      "step": 1375
    },
    {
      "epoch": 125.0969696969697,
      "grad_norm": 0.23206835985183716,
      "learning_rate": 3.624e-05,
      "loss": 0.4322,
      "step": 1376
    },
    {
      "epoch": 125.19393939393939,
      "grad_norm": 0.27216339111328125,
      "learning_rate": 3.6230000000000004e-05,
      "loss": 0.4493,
      "step": 1377
    },
    {
      "epoch": 125.2909090909091,
      "grad_norm": 0.26469945907592773,
      "learning_rate": 3.622e-05,
      "loss": 0.4055,
      "step": 1378
    },
    {
      "epoch": 125.38787878787879,
      "grad_norm": 0.2685600221157074,
      "learning_rate": 3.621e-05,
      "loss": 0.4496,
      "step": 1379
    },
    {
      "epoch": 125.48484848484848,
      "grad_norm": 0.2712048888206482,
      "learning_rate": 3.62e-05,
      "loss": 0.4677,
      "step": 1380
    },
    {
      "epoch": 125.48484848484848,
      "eval_loss": 0.4940645396709442,
      "eval_runtime": 2.1505,
      "eval_samples_per_second": 25.575,
      "eval_steps_per_second": 3.255,
      "step": 1380
    },
    {
      "epoch": 125.58181818181818,
      "grad_norm": 0.3111526668071747,
      "learning_rate": 3.6190000000000004e-05,
      "loss": 0.4701,
      "step": 1381
    },
    {
      "epoch": 125.67878787878787,
      "grad_norm": 0.2506205141544342,
      "learning_rate": 3.618e-05,
      "loss": 0.4498,
      "step": 1382
    },
    {
      "epoch": 125.77575757575758,
      "grad_norm": 0.347450315952301,
      "learning_rate": 3.617e-05,
      "loss": 0.4418,
      "step": 1383
    },
    {
      "epoch": 125.87272727272727,
      "grad_norm": 0.3053918778896332,
      "learning_rate": 3.616e-05,
      "loss": 0.4602,
      "step": 1384
    },
    {
      "epoch": 125.96969696969697,
      "grad_norm": 0.25132766366004944,
      "learning_rate": 3.615e-05,
      "loss": 0.4036,
      "step": 1385
    },
    {
      "epoch": 126.0,
      "grad_norm": 0.41760462522506714,
      "learning_rate": 3.614e-05,
      "loss": 0.3651,
      "step": 1386
    },
    {
      "epoch": 126.0969696969697,
      "grad_norm": 0.3355599045753479,
      "learning_rate": 3.613e-05,
      "loss": 0.4767,
      "step": 1387
    },
    {
      "epoch": 126.19393939393939,
      "grad_norm": 0.29463788866996765,
      "learning_rate": 3.6120000000000007e-05,
      "loss": 0.4493,
      "step": 1388
    },
    {
      "epoch": 126.2909090909091,
      "grad_norm": 0.3120037317276001,
      "learning_rate": 3.611e-05,
      "loss": 0.4589,
      "step": 1389
    },
    {
      "epoch": 126.38787878787879,
      "grad_norm": 0.22270894050598145,
      "learning_rate": 3.61e-05,
      "loss": 0.4225,
      "step": 1390
    },
    {
      "epoch": 126.38787878787879,
      "eval_loss": 0.4936691224575043,
      "eval_runtime": 2.1584,
      "eval_samples_per_second": 25.481,
      "eval_steps_per_second": 3.243,
      "step": 1390
    },
    {
      "epoch": 126.48484848484848,
      "grad_norm": 0.2702427804470062,
      "learning_rate": 3.609e-05,
      "loss": 0.4303,
      "step": 1391
    },
    {
      "epoch": 126.58181818181818,
      "grad_norm": 0.2349555343389511,
      "learning_rate": 3.608e-05,
      "loss": 0.4264,
      "step": 1392
    },
    {
      "epoch": 126.67878787878787,
      "grad_norm": 0.23798425495624542,
      "learning_rate": 3.6070000000000005e-05,
      "loss": 0.4566,
      "step": 1393
    },
    {
      "epoch": 126.77575757575758,
      "grad_norm": 0.2422553449869156,
      "learning_rate": 3.606e-05,
      "loss": 0.4078,
      "step": 1394
    },
    {
      "epoch": 126.87272727272727,
      "grad_norm": 0.29934629797935486,
      "learning_rate": 3.605e-05,
      "loss": 0.3834,
      "step": 1395
    },
    {
      "epoch": 126.96969696969697,
      "grad_norm": 0.337470680475235,
      "learning_rate": 3.604e-05,
      "loss": 0.4824,
      "step": 1396
    },
    {
      "epoch": 127.0,
      "grad_norm": 0.44956621527671814,
      "learning_rate": 3.6030000000000006e-05,
      "loss": 0.428,
      "step": 1397
    },
    {
      "epoch": 127.0969696969697,
      "grad_norm": 0.2476656585931778,
      "learning_rate": 3.6020000000000004e-05,
      "loss": 0.3799,
      "step": 1398
    },
    {
      "epoch": 127.19393939393939,
      "grad_norm": 0.3032231628894806,
      "learning_rate": 3.601e-05,
      "loss": 0.4688,
      "step": 1399
    },
    {
      "epoch": 127.2909090909091,
      "grad_norm": 0.21109527349472046,
      "learning_rate": 3.6e-05,
      "loss": 0.3841,
      "step": 1400
    },
    {
      "epoch": 127.2909090909091,
      "eval_loss": 0.4930807948112488,
      "eval_runtime": 2.161,
      "eval_samples_per_second": 25.451,
      "eval_steps_per_second": 3.239,
      "step": 1400
    },
    {
      "epoch": 127.38787878787879,
      "grad_norm": 0.3381328284740448,
      "learning_rate": 3.599e-05,
      "loss": 0.4337,
      "step": 1401
    },
    {
      "epoch": 127.48484848484848,
      "grad_norm": 0.2521904408931732,
      "learning_rate": 3.5980000000000004e-05,
      "loss": 0.4433,
      "step": 1402
    },
    {
      "epoch": 127.58181818181818,
      "grad_norm": 0.3311968743801117,
      "learning_rate": 3.597e-05,
      "loss": 0.4591,
      "step": 1403
    },
    {
      "epoch": 127.67878787878787,
      "grad_norm": 0.2518157660961151,
      "learning_rate": 3.596e-05,
      "loss": 0.4364,
      "step": 1404
    },
    {
      "epoch": 127.77575757575758,
      "grad_norm": 0.27200594544410706,
      "learning_rate": 3.595e-05,
      "loss": 0.4533,
      "step": 1405
    },
    {
      "epoch": 127.87272727272727,
      "grad_norm": 0.27561336755752563,
      "learning_rate": 3.594e-05,
      "loss": 0.4142,
      "step": 1406
    },
    {
      "epoch": 127.96969696969697,
      "grad_norm": 0.3234183192253113,
      "learning_rate": 3.593e-05,
      "loss": 0.4772,
      "step": 1407
    },
    {
      "epoch": 128.0,
      "grad_norm": 0.5238689184188843,
      "learning_rate": 3.592e-05,
      "loss": 0.5281,
      "step": 1408
    },
    {
      "epoch": 128.0969696969697,
      "grad_norm": 0.26986992359161377,
      "learning_rate": 3.591e-05,
      "loss": 0.4498,
      "step": 1409
    },
    {
      "epoch": 128.1939393939394,
      "grad_norm": 0.23676660656929016,
      "learning_rate": 3.59e-05,
      "loss": 0.46,
      "step": 1410
    },
    {
      "epoch": 128.1939393939394,
      "eval_loss": 0.4915180802345276,
      "eval_runtime": 2.1483,
      "eval_samples_per_second": 25.601,
      "eval_steps_per_second": 3.258,
      "step": 1410
    },
    {
      "epoch": 128.29090909090908,
      "grad_norm": 0.2594098150730133,
      "learning_rate": 3.5890000000000004e-05,
      "loss": 0.4418,
      "step": 1411
    },
    {
      "epoch": 128.38787878787878,
      "grad_norm": 0.3071761131286621,
      "learning_rate": 3.588e-05,
      "loss": 0.4297,
      "step": 1412
    },
    {
      "epoch": 128.4848484848485,
      "grad_norm": 0.2817282974720001,
      "learning_rate": 3.587e-05,
      "loss": 0.4191,
      "step": 1413
    },
    {
      "epoch": 128.5818181818182,
      "grad_norm": 0.3153073787689209,
      "learning_rate": 3.586e-05,
      "loss": 0.4725,
      "step": 1414
    },
    {
      "epoch": 128.6787878787879,
      "grad_norm": 0.27557572722435,
      "learning_rate": 3.585e-05,
      "loss": 0.4442,
      "step": 1415
    },
    {
      "epoch": 128.77575757575758,
      "grad_norm": 0.3079877495765686,
      "learning_rate": 3.584e-05,
      "loss": 0.3941,
      "step": 1416
    },
    {
      "epoch": 128.87272727272727,
      "grad_norm": 0.2863088548183441,
      "learning_rate": 3.583e-05,
      "loss": 0.4048,
      "step": 1417
    },
    {
      "epoch": 128.96969696969697,
      "grad_norm": 0.27652451395988464,
      "learning_rate": 3.5820000000000006e-05,
      "loss": 0.4715,
      "step": 1418
    },
    {
      "epoch": 129.0,
      "grad_norm": 0.45896708965301514,
      "learning_rate": 3.581e-05,
      "loss": 0.3892,
      "step": 1419
    },
    {
      "epoch": 129.0969696969697,
      "grad_norm": 0.3001515567302704,
      "learning_rate": 3.58e-05,
      "loss": 0.4649,
      "step": 1420
    },
    {
      "epoch": 129.0969696969697,
      "eval_loss": 0.4916605055332184,
      "eval_runtime": 2.1561,
      "eval_samples_per_second": 25.51,
      "eval_steps_per_second": 3.247,
      "step": 1420
    },
    {
      "epoch": 129.1939393939394,
      "grad_norm": 0.28979921340942383,
      "learning_rate": 3.579e-05,
      "loss": 0.4756,
      "step": 1421
    },
    {
      "epoch": 129.29090909090908,
      "grad_norm": 0.22518810629844666,
      "learning_rate": 3.578e-05,
      "loss": 0.3911,
      "step": 1422
    },
    {
      "epoch": 129.38787878787878,
      "grad_norm": 0.33359140157699585,
      "learning_rate": 3.5770000000000005e-05,
      "loss": 0.4618,
      "step": 1423
    },
    {
      "epoch": 129.4848484848485,
      "grad_norm": 0.26979026198387146,
      "learning_rate": 3.5759999999999996e-05,
      "loss": 0.4277,
      "step": 1424
    },
    {
      "epoch": 129.5818181818182,
      "grad_norm": 0.3089786767959595,
      "learning_rate": 3.575e-05,
      "loss": 0.4067,
      "step": 1425
    },
    {
      "epoch": 129.6787878787879,
      "grad_norm": 0.2348090261220932,
      "learning_rate": 3.574e-05,
      "loss": 0.4446,
      "step": 1426
    },
    {
      "epoch": 129.77575757575758,
      "grad_norm": 0.24655956029891968,
      "learning_rate": 3.5730000000000005e-05,
      "loss": 0.425,
      "step": 1427
    },
    {
      "epoch": 129.87272727272727,
      "grad_norm": 0.23908300697803497,
      "learning_rate": 3.5720000000000004e-05,
      "loss": 0.4311,
      "step": 1428
    },
    {
      "epoch": 129.96969696969697,
      "grad_norm": 0.33590126037597656,
      "learning_rate": 3.571e-05,
      "loss": 0.4437,
      "step": 1429
    },
    {
      "epoch": 130.0,
      "grad_norm": 0.5856208801269531,
      "learning_rate": 3.57e-05,
      "loss": 0.3959,
      "step": 1430
    },
    {
      "epoch": 130.0,
      "eval_loss": 0.49039074778556824,
      "eval_runtime": 2.1368,
      "eval_samples_per_second": 25.74,
      "eval_steps_per_second": 3.276,
      "step": 1430
    },
    {
      "epoch": 130.0969696969697,
      "grad_norm": 0.2495311051607132,
      "learning_rate": 3.569e-05,
      "loss": 0.4326,
      "step": 1431
    },
    {
      "epoch": 130.1939393939394,
      "grad_norm": 0.23713161051273346,
      "learning_rate": 3.5680000000000004e-05,
      "loss": 0.4203,
      "step": 1432
    },
    {
      "epoch": 130.29090909090908,
      "grad_norm": 0.2517852187156677,
      "learning_rate": 3.567e-05,
      "loss": 0.4397,
      "step": 1433
    },
    {
      "epoch": 130.38787878787878,
      "grad_norm": 0.3342297673225403,
      "learning_rate": 3.566e-05,
      "loss": 0.4514,
      "step": 1434
    },
    {
      "epoch": 130.4848484848485,
      "grad_norm": 0.27150875329971313,
      "learning_rate": 3.565e-05,
      "loss": 0.4163,
      "step": 1435
    },
    {
      "epoch": 130.5818181818182,
      "grad_norm": 0.3103017807006836,
      "learning_rate": 3.5640000000000004e-05,
      "loss": 0.4216,
      "step": 1436
    },
    {
      "epoch": 130.6787878787879,
      "grad_norm": 0.3338560461997986,
      "learning_rate": 3.563e-05,
      "loss": 0.4755,
      "step": 1437
    },
    {
      "epoch": 130.77575757575758,
      "grad_norm": 0.3014073967933655,
      "learning_rate": 3.562e-05,
      "loss": 0.4329,
      "step": 1438
    },
    {
      "epoch": 130.87272727272727,
      "grad_norm": 0.2600906491279602,
      "learning_rate": 3.5610000000000006e-05,
      "loss": 0.4158,
      "step": 1439
    },
    {
      "epoch": 130.96969696969697,
      "grad_norm": 0.2621995210647583,
      "learning_rate": 3.56e-05,
      "loss": 0.4183,
      "step": 1440
    },
    {
      "epoch": 130.96969696969697,
      "eval_loss": 0.48934677243232727,
      "eval_runtime": 2.1635,
      "eval_samples_per_second": 25.421,
      "eval_steps_per_second": 3.235,
      "step": 1440
    },
    {
      "epoch": 131.0,
      "grad_norm": 0.4798804521560669,
      "learning_rate": 3.559e-05,
      "loss": 0.5173,
      "step": 1441
    },
    {
      "epoch": 131.0969696969697,
      "grad_norm": 0.3138405382633209,
      "learning_rate": 3.558e-05,
      "loss": 0.4675,
      "step": 1442
    },
    {
      "epoch": 131.1939393939394,
      "grad_norm": 0.27567216753959656,
      "learning_rate": 3.557e-05,
      "loss": 0.465,
      "step": 1443
    },
    {
      "epoch": 131.29090909090908,
      "grad_norm": 0.308773934841156,
      "learning_rate": 3.5560000000000005e-05,
      "loss": 0.4491,
      "step": 1444
    },
    {
      "epoch": 131.38787878787878,
      "grad_norm": 0.3976498544216156,
      "learning_rate": 3.555e-05,
      "loss": 0.4124,
      "step": 1445
    },
    {
      "epoch": 131.4848484848485,
      "grad_norm": 0.22643493115901947,
      "learning_rate": 3.554e-05,
      "loss": 0.4143,
      "step": 1446
    },
    {
      "epoch": 131.5818181818182,
      "grad_norm": 0.2682403326034546,
      "learning_rate": 3.553e-05,
      "loss": 0.444,
      "step": 1447
    },
    {
      "epoch": 131.6787878787879,
      "grad_norm": 0.249359592795372,
      "learning_rate": 3.5520000000000006e-05,
      "loss": 0.4481,
      "step": 1448
    },
    {
      "epoch": 131.77575757575758,
      "grad_norm": 0.31316664814949036,
      "learning_rate": 3.5510000000000004e-05,
      "loss": 0.4161,
      "step": 1449
    },
    {
      "epoch": 131.87272727272727,
      "grad_norm": 0.3224125802516937,
      "learning_rate": 3.55e-05,
      "loss": 0.4131,
      "step": 1450
    },
    {
      "epoch": 131.87272727272727,
      "eval_loss": 0.48860692977905273,
      "eval_runtime": 2.166,
      "eval_samples_per_second": 25.392,
      "eval_steps_per_second": 3.232,
      "step": 1450
    },
    {
      "epoch": 131.96969696969697,
      "grad_norm": 0.2814408540725708,
      "learning_rate": 3.549e-05,
      "loss": 0.4149,
      "step": 1451
    },
    {
      "epoch": 132.0,
      "grad_norm": 0.46793943643569946,
      "learning_rate": 3.548e-05,
      "loss": 0.4296,
      "step": 1452
    },
    {
      "epoch": 132.0969696969697,
      "grad_norm": 0.24512231349945068,
      "learning_rate": 3.5470000000000004e-05,
      "loss": 0.4516,
      "step": 1453
    },
    {
      "epoch": 132.1939393939394,
      "grad_norm": 0.2817038893699646,
      "learning_rate": 3.546e-05,
      "loss": 0.4837,
      "step": 1454
    },
    {
      "epoch": 132.29090909090908,
      "grad_norm": 0.2967666983604431,
      "learning_rate": 3.545e-05,
      "loss": 0.3845,
      "step": 1455
    },
    {
      "epoch": 132.38787878787878,
      "grad_norm": 0.28917327523231506,
      "learning_rate": 3.544e-05,
      "loss": 0.467,
      "step": 1456
    },
    {
      "epoch": 132.4848484848485,
      "grad_norm": 0.2414083033800125,
      "learning_rate": 3.5430000000000005e-05,
      "loss": 0.4025,
      "step": 1457
    },
    {
      "epoch": 132.5818181818182,
      "grad_norm": 0.3054693639278412,
      "learning_rate": 3.542e-05,
      "loss": 0.4138,
      "step": 1458
    },
    {
      "epoch": 132.6787878787879,
      "grad_norm": 0.2707059681415558,
      "learning_rate": 3.541e-05,
      "loss": 0.443,
      "step": 1459
    },
    {
      "epoch": 132.77575757575758,
      "grad_norm": 0.3007029592990875,
      "learning_rate": 3.54e-05,
      "loss": 0.4541,
      "step": 1460
    },
    {
      "epoch": 132.77575757575758,
      "eval_loss": 0.4879237115383148,
      "eval_runtime": 2.1713,
      "eval_samples_per_second": 25.331,
      "eval_steps_per_second": 3.224,
      "step": 1460
    },
    {
      "epoch": 132.87272727272727,
      "grad_norm": 0.27469393610954285,
      "learning_rate": 3.539e-05,
      "loss": 0.4028,
      "step": 1461
    },
    {
      "epoch": 132.96969696969697,
      "grad_norm": 0.2680533528327942,
      "learning_rate": 3.5380000000000003e-05,
      "loss": 0.4265,
      "step": 1462
    },
    {
      "epoch": 133.0,
      "grad_norm": 0.46730121970176697,
      "learning_rate": 3.537e-05,
      "loss": 0.4283,
      "step": 1463
    },
    {
      "epoch": 133.0969696969697,
      "grad_norm": 0.29606109857559204,
      "learning_rate": 3.536000000000001e-05,
      "loss": 0.471,
      "step": 1464
    },
    {
      "epoch": 133.1939393939394,
      "grad_norm": 0.2748776078224182,
      "learning_rate": 3.535e-05,
      "loss": 0.4334,
      "step": 1465
    },
    {
      "epoch": 133.29090909090908,
      "grad_norm": 0.2659863233566284,
      "learning_rate": 3.5340000000000004e-05,
      "loss": 0.4235,
      "step": 1466
    },
    {
      "epoch": 133.38787878787878,
      "grad_norm": 0.23002707958221436,
      "learning_rate": 3.533e-05,
      "loss": 0.4527,
      "step": 1467
    },
    {
      "epoch": 133.4848484848485,
      "grad_norm": 0.252360075712204,
      "learning_rate": 3.532e-05,
      "loss": 0.4203,
      "step": 1468
    },
    {
      "epoch": 133.5818181818182,
      "grad_norm": 0.2285648137331009,
      "learning_rate": 3.5310000000000006e-05,
      "loss": 0.4021,
      "step": 1469
    },
    {
      "epoch": 133.6787878787879,
      "grad_norm": 0.308756947517395,
      "learning_rate": 3.53e-05,
      "loss": 0.3883,
      "step": 1470
    },
    {
      "epoch": 133.6787878787879,
      "eval_loss": 0.4869600832462311,
      "eval_runtime": 2.1755,
      "eval_samples_per_second": 25.281,
      "eval_steps_per_second": 3.218,
      "step": 1470
    },
    {
      "epoch": 133.77575757575758,
      "grad_norm": 0.36151477694511414,
      "learning_rate": 3.529e-05,
      "loss": 0.4449,
      "step": 1471
    },
    {
      "epoch": 133.87272727272727,
      "grad_norm": 0.27611294388771057,
      "learning_rate": 3.528e-05,
      "loss": 0.4614,
      "step": 1472
    },
    {
      "epoch": 133.96969696969697,
      "grad_norm": 0.32699742913246155,
      "learning_rate": 3.5270000000000006e-05,
      "loss": 0.42,
      "step": 1473
    },
    {
      "epoch": 134.0,
      "grad_norm": 0.6019296050071716,
      "learning_rate": 3.5260000000000005e-05,
      "loss": 0.4283,
      "step": 1474
    },
    {
      "epoch": 134.0969696969697,
      "grad_norm": 0.279818058013916,
      "learning_rate": 3.525e-05,
      "loss": 0.3954,
      "step": 1475
    },
    {
      "epoch": 134.1939393939394,
      "grad_norm": 0.269252747297287,
      "learning_rate": 3.524e-05,
      "loss": 0.3926,
      "step": 1476
    },
    {
      "epoch": 134.29090909090908,
      "grad_norm": 0.2596481144428253,
      "learning_rate": 3.523e-05,
      "loss": 0.4227,
      "step": 1477
    },
    {
      "epoch": 134.38787878787878,
      "grad_norm": 0.30212676525115967,
      "learning_rate": 3.5220000000000005e-05,
      "loss": 0.4182,
      "step": 1478
    },
    {
      "epoch": 134.4848484848485,
      "grad_norm": 0.3808242678642273,
      "learning_rate": 3.5210000000000003e-05,
      "loss": 0.4378,
      "step": 1479
    },
    {
      "epoch": 134.5818181818182,
      "grad_norm": 0.28346386551856995,
      "learning_rate": 3.52e-05,
      "loss": 0.4492,
      "step": 1480
    },
    {
      "epoch": 134.5818181818182,
      "eval_loss": 0.48657190799713135,
      "eval_runtime": 2.1723,
      "eval_samples_per_second": 25.319,
      "eval_steps_per_second": 3.222,
      "step": 1480
    },
    {
      "epoch": 134.6787878787879,
      "grad_norm": 0.3859547972679138,
      "learning_rate": 3.519e-05,
      "loss": 0.4525,
      "step": 1481
    },
    {
      "epoch": 134.77575757575758,
      "grad_norm": 0.2632039487361908,
      "learning_rate": 3.518e-05,
      "loss": 0.4463,
      "step": 1482
    },
    {
      "epoch": 134.87272727272727,
      "grad_norm": 0.256745308637619,
      "learning_rate": 3.5170000000000004e-05,
      "loss": 0.4196,
      "step": 1483
    },
    {
      "epoch": 134.96969696969697,
      "grad_norm": 0.3404027223587036,
      "learning_rate": 3.516e-05,
      "loss": 0.4425,
      "step": 1484
    },
    {
      "epoch": 135.0,
      "grad_norm": 0.45514726638793945,
      "learning_rate": 3.515e-05,
      "loss": 0.529,
      "step": 1485
    },
    {
      "epoch": 135.0969696969697,
      "grad_norm": 0.2929936945438385,
      "learning_rate": 3.514e-05,
      "loss": 0.4281,
      "step": 1486
    },
    {
      "epoch": 135.1939393939394,
      "grad_norm": 0.3744279444217682,
      "learning_rate": 3.5130000000000004e-05,
      "loss": 0.3906,
      "step": 1487
    },
    {
      "epoch": 135.29090909090908,
      "grad_norm": 0.30043625831604004,
      "learning_rate": 3.512e-05,
      "loss": 0.4551,
      "step": 1488
    },
    {
      "epoch": 135.38787878787878,
      "grad_norm": 0.23369665443897247,
      "learning_rate": 3.511e-05,
      "loss": 0.4127,
      "step": 1489
    },
    {
      "epoch": 135.4848484848485,
      "grad_norm": 0.32619598507881165,
      "learning_rate": 3.51e-05,
      "loss": 0.426,
      "step": 1490
    },
    {
      "epoch": 135.4848484848485,
      "eval_loss": 0.4854676425457001,
      "eval_runtime": 2.136,
      "eval_samples_per_second": 25.749,
      "eval_steps_per_second": 3.277,
      "step": 1490
    },
    {
      "epoch": 135.5818181818182,
      "grad_norm": 0.24827663600444794,
      "learning_rate": 3.509e-05,
      "loss": 0.4259,
      "step": 1491
    },
    {
      "epoch": 135.6787878787879,
      "grad_norm": 0.239982470870018,
      "learning_rate": 3.508e-05,
      "loss": 0.4167,
      "step": 1492
    },
    {
      "epoch": 135.77575757575758,
      "grad_norm": 0.28403112292289734,
      "learning_rate": 3.507e-05,
      "loss": 0.4565,
      "step": 1493
    },
    {
      "epoch": 135.87272727272727,
      "grad_norm": 0.31020352244377136,
      "learning_rate": 3.5060000000000007e-05,
      "loss": 0.4335,
      "step": 1494
    },
    {
      "epoch": 135.96969696969697,
      "grad_norm": 0.2941436767578125,
      "learning_rate": 3.505e-05,
      "loss": 0.4491,
      "step": 1495
    },
    {
      "epoch": 136.0,
      "grad_norm": 0.42165759205818176,
      "learning_rate": 3.504e-05,
      "loss": 0.4343,
      "step": 1496
    },
    {
      "epoch": 136.0969696969697,
      "grad_norm": 0.2817646563053131,
      "learning_rate": 3.503e-05,
      "loss": 0.3876,
      "step": 1497
    },
    {
      "epoch": 136.1939393939394,
      "grad_norm": 0.2687471807003021,
      "learning_rate": 3.502e-05,
      "loss": 0.4572,
      "step": 1498
    },
    {
      "epoch": 136.29090909090908,
      "grad_norm": 0.2752440571784973,
      "learning_rate": 3.5010000000000005e-05,
      "loss": 0.4754,
      "step": 1499
    },
    {
      "epoch": 136.38787878787878,
      "grad_norm": 0.26806142926216125,
      "learning_rate": 3.5e-05,
      "loss": 0.4099,
      "step": 1500
    },
    {
      "epoch": 136.38787878787878,
      "eval_loss": 0.48457300662994385,
      "eval_runtime": 2.1595,
      "eval_samples_per_second": 25.469,
      "eval_steps_per_second": 3.242,
      "step": 1500
    },
    {
      "epoch": 136.4848484848485,
      "grad_norm": 0.26220443844795227,
      "learning_rate": 3.499e-05,
      "loss": 0.4728,
      "step": 1501
    },
    {
      "epoch": 136.5818181818182,
      "grad_norm": 0.29137399792671204,
      "learning_rate": 3.498e-05,
      "loss": 0.4426,
      "step": 1502
    },
    {
      "epoch": 136.6787878787879,
      "grad_norm": 0.28184735774993896,
      "learning_rate": 3.4970000000000006e-05,
      "loss": 0.3705,
      "step": 1503
    },
    {
      "epoch": 136.77575757575758,
      "grad_norm": 0.26435405015945435,
      "learning_rate": 3.4960000000000004e-05,
      "loss": 0.4361,
      "step": 1504
    },
    {
      "epoch": 136.87272727272727,
      "grad_norm": 0.29367130994796753,
      "learning_rate": 3.495e-05,
      "loss": 0.4451,
      "step": 1505
    },
    {
      "epoch": 136.96969696969697,
      "grad_norm": 0.35805878043174744,
      "learning_rate": 3.494e-05,
      "loss": 0.3964,
      "step": 1506
    },
    {
      "epoch": 137.0,
      "grad_norm": 0.45524048805236816,
      "learning_rate": 3.493e-05,
      "loss": 0.4054,
      "step": 1507
    },
    {
      "epoch": 137.0969696969697,
      "grad_norm": 0.26967760920524597,
      "learning_rate": 3.4920000000000004e-05,
      "loss": 0.4508,
      "step": 1508
    },
    {
      "epoch": 137.1939393939394,
      "grad_norm": 0.25061696767807007,
      "learning_rate": 3.491e-05,
      "loss": 0.4245,
      "step": 1509
    },
    {
      "epoch": 137.29090909090908,
      "grad_norm": 0.24153633415699005,
      "learning_rate": 3.49e-05,
      "loss": 0.4106,
      "step": 1510
    },
    {
      "epoch": 137.29090909090908,
      "eval_loss": 0.4839669167995453,
      "eval_runtime": 2.1707,
      "eval_samples_per_second": 25.338,
      "eval_steps_per_second": 3.225,
      "step": 1510
    },
    {
      "epoch": 137.38787878787878,
      "grad_norm": 0.2820562422275543,
      "learning_rate": 3.489e-05,
      "loss": 0.4411,
      "step": 1511
    },
    {
      "epoch": 137.4848484848485,
      "grad_norm": 0.26837873458862305,
      "learning_rate": 3.4880000000000005e-05,
      "loss": 0.4497,
      "step": 1512
    },
    {
      "epoch": 137.5818181818182,
      "grad_norm": 0.2681138813495636,
      "learning_rate": 3.487e-05,
      "loss": 0.4102,
      "step": 1513
    },
    {
      "epoch": 137.6787878787879,
      "grad_norm": 0.27645957469940186,
      "learning_rate": 3.486e-05,
      "loss": 0.4344,
      "step": 1514
    },
    {
      "epoch": 137.77575757575758,
      "grad_norm": 0.25688016414642334,
      "learning_rate": 3.485e-05,
      "loss": 0.4236,
      "step": 1515
    },
    {
      "epoch": 137.87272727272727,
      "grad_norm": 0.3318941593170166,
      "learning_rate": 3.484e-05,
      "loss": 0.4538,
      "step": 1516
    },
    {
      "epoch": 137.96969696969697,
      "grad_norm": 0.22465866804122925,
      "learning_rate": 3.4830000000000004e-05,
      "loss": 0.3973,
      "step": 1517
    },
    {
      "epoch": 138.0,
      "grad_norm": 0.36190637946128845,
      "learning_rate": 3.482e-05,
      "loss": 0.3488,
      "step": 1518
    },
    {
      "epoch": 138.0969696969697,
      "grad_norm": 0.2324100136756897,
      "learning_rate": 3.481e-05,
      "loss": 0.4149,
      "step": 1519
    },
    {
      "epoch": 138.1939393939394,
      "grad_norm": 0.2485891580581665,
      "learning_rate": 3.48e-05,
      "loss": 0.4321,
      "step": 1520
    },
    {
      "epoch": 138.1939393939394,
      "eval_loss": 0.4832867980003357,
      "eval_runtime": 2.1612,
      "eval_samples_per_second": 25.449,
      "eval_steps_per_second": 3.239,
      "step": 1520
    },
    {
      "epoch": 138.29090909090908,
      "grad_norm": 0.2955532371997833,
      "learning_rate": 3.479e-05,
      "loss": 0.4417,
      "step": 1521
    },
    {
      "epoch": 138.38787878787878,
      "grad_norm": 0.2552751302719116,
      "learning_rate": 3.478e-05,
      "loss": 0.4241,
      "step": 1522
    },
    {
      "epoch": 138.4848484848485,
      "grad_norm": 0.27413198351860046,
      "learning_rate": 3.477e-05,
      "loss": 0.4504,
      "step": 1523
    },
    {
      "epoch": 138.5818181818182,
      "grad_norm": 0.23463669419288635,
      "learning_rate": 3.4760000000000006e-05,
      "loss": 0.4139,
      "step": 1524
    },
    {
      "epoch": 138.6787878787879,
      "grad_norm": 0.2654094696044922,
      "learning_rate": 3.475e-05,
      "loss": 0.4482,
      "step": 1525
    },
    {
      "epoch": 138.77575757575758,
      "grad_norm": 0.3383115828037262,
      "learning_rate": 3.474e-05,
      "loss": 0.4053,
      "step": 1526
    },
    {
      "epoch": 138.87272727272727,
      "grad_norm": 0.3001095652580261,
      "learning_rate": 3.473e-05,
      "loss": 0.4237,
      "step": 1527
    },
    {
      "epoch": 138.96969696969697,
      "grad_norm": 0.31264686584472656,
      "learning_rate": 3.472e-05,
      "loss": 0.4034,
      "step": 1528
    },
    {
      "epoch": 139.0,
      "grad_norm": 0.500529944896698,
      "learning_rate": 3.4710000000000005e-05,
      "loss": 0.4396,
      "step": 1529
    },
    {
      "epoch": 139.0969696969697,
      "grad_norm": 0.26789966225624084,
      "learning_rate": 3.4699999999999996e-05,
      "loss": 0.4432,
      "step": 1530
    },
    {
      "epoch": 139.0969696969697,
      "eval_loss": 0.48245444893836975,
      "eval_runtime": 2.1444,
      "eval_samples_per_second": 25.648,
      "eval_steps_per_second": 3.264,
      "step": 1530
    },
    {
      "epoch": 139.1939393939394,
      "grad_norm": 0.29543471336364746,
      "learning_rate": 3.469e-05,
      "loss": 0.4696,
      "step": 1531
    },
    {
      "epoch": 139.29090909090908,
      "grad_norm": 0.2340238094329834,
      "learning_rate": 3.468e-05,
      "loss": 0.3868,
      "step": 1532
    },
    {
      "epoch": 139.38787878787878,
      "grad_norm": 0.2654852271080017,
      "learning_rate": 3.4670000000000005e-05,
      "loss": 0.4385,
      "step": 1533
    },
    {
      "epoch": 139.4848484848485,
      "grad_norm": 0.2928999364376068,
      "learning_rate": 3.4660000000000004e-05,
      "loss": 0.4506,
      "step": 1534
    },
    {
      "epoch": 139.5818181818182,
      "grad_norm": 0.3032337725162506,
      "learning_rate": 3.465e-05,
      "loss": 0.3848,
      "step": 1535
    },
    {
      "epoch": 139.6787878787879,
      "grad_norm": 0.340878963470459,
      "learning_rate": 3.464e-05,
      "loss": 0.3951,
      "step": 1536
    },
    {
      "epoch": 139.77575757575758,
      "grad_norm": 0.25129806995391846,
      "learning_rate": 3.463e-05,
      "loss": 0.4616,
      "step": 1537
    },
    {
      "epoch": 139.87272727272727,
      "grad_norm": 0.30390605330467224,
      "learning_rate": 3.4620000000000004e-05,
      "loss": 0.4478,
      "step": 1538
    },
    {
      "epoch": 139.96969696969697,
      "grad_norm": 0.24877336621284485,
      "learning_rate": 3.461e-05,
      "loss": 0.3916,
      "step": 1539
    },
    {
      "epoch": 140.0,
      "grad_norm": 0.4447900354862213,
      "learning_rate": 3.46e-05,
      "loss": 0.3716,
      "step": 1540
    },
    {
      "epoch": 140.0,
      "eval_loss": 0.4822714030742645,
      "eval_runtime": 2.1583,
      "eval_samples_per_second": 25.483,
      "eval_steps_per_second": 3.243,
      "step": 1540
    },
    {
      "epoch": 140.0969696969697,
      "grad_norm": 0.31359797716140747,
      "learning_rate": 3.459e-05,
      "loss": 0.441,
      "step": 1541
    },
    {
      "epoch": 140.1939393939394,
      "grad_norm": 0.27479588985443115,
      "learning_rate": 3.4580000000000004e-05,
      "loss": 0.4205,
      "step": 1542
    },
    {
      "epoch": 140.29090909090908,
      "grad_norm": 0.2537696361541748,
      "learning_rate": 3.457e-05,
      "loss": 0.4271,
      "step": 1543
    },
    {
      "epoch": 140.38787878787878,
      "grad_norm": 0.30846652388572693,
      "learning_rate": 3.456e-05,
      "loss": 0.4139,
      "step": 1544
    },
    {
      "epoch": 140.4848484848485,
      "grad_norm": 0.3260568082332611,
      "learning_rate": 3.455e-05,
      "loss": 0.4388,
      "step": 1545
    },
    {
      "epoch": 140.5818181818182,
      "grad_norm": 0.27833911776542664,
      "learning_rate": 3.454e-05,
      "loss": 0.422,
      "step": 1546
    },
    {
      "epoch": 140.6787878787879,
      "grad_norm": 0.3193536698818207,
      "learning_rate": 3.453e-05,
      "loss": 0.4182,
      "step": 1547
    },
    {
      "epoch": 140.77575757575758,
      "grad_norm": 0.2909131646156311,
      "learning_rate": 3.452e-05,
      "loss": 0.418,
      "step": 1548
    },
    {
      "epoch": 140.87272727272727,
      "grad_norm": 0.2805783748626709,
      "learning_rate": 3.451000000000001e-05,
      "loss": 0.4181,
      "step": 1549
    },
    {
      "epoch": 140.96969696969697,
      "grad_norm": 0.2544325590133667,
      "learning_rate": 3.45e-05,
      "loss": 0.4336,
      "step": 1550
    },
    {
      "epoch": 140.96969696969697,
      "eval_loss": 0.4808160364627838,
      "eval_runtime": 2.1672,
      "eval_samples_per_second": 25.378,
      "eval_steps_per_second": 3.23,
      "step": 1550
    },
    {
      "epoch": 141.0,
      "grad_norm": 0.5855840444564819,
      "learning_rate": 3.449e-05,
      "loss": 0.4099,
      "step": 1551
    },
    {
      "epoch": 141.0969696969697,
      "grad_norm": 0.2724772095680237,
      "learning_rate": 3.448e-05,
      "loss": 0.4489,
      "step": 1552
    },
    {
      "epoch": 141.1939393939394,
      "grad_norm": 0.2569520175457001,
      "learning_rate": 3.447e-05,
      "loss": 0.4341,
      "step": 1553
    },
    {
      "epoch": 141.29090909090908,
      "grad_norm": 0.24317876994609833,
      "learning_rate": 3.4460000000000005e-05,
      "loss": 0.4252,
      "step": 1554
    },
    {
      "epoch": 141.38787878787878,
      "grad_norm": 0.24204610288143158,
      "learning_rate": 3.445e-05,
      "loss": 0.3995,
      "step": 1555
    },
    {
      "epoch": 141.4848484848485,
      "grad_norm": 0.2994146943092346,
      "learning_rate": 3.444e-05,
      "loss": 0.4007,
      "step": 1556
    },
    {
      "epoch": 141.5818181818182,
      "grad_norm": 0.2774331569671631,
      "learning_rate": 3.443e-05,
      "loss": 0.4026,
      "step": 1557
    },
    {
      "epoch": 141.6787878787879,
      "grad_norm": 0.267837792634964,
      "learning_rate": 3.442e-05,
      "loss": 0.3982,
      "step": 1558
    },
    {
      "epoch": 141.77575757575758,
      "grad_norm": 0.3514941334724426,
      "learning_rate": 3.4410000000000004e-05,
      "loss": 0.4719,
      "step": 1559
    },
    {
      "epoch": 141.87272727272727,
      "grad_norm": 0.3003464937210083,
      "learning_rate": 3.4399999999999996e-05,
      "loss": 0.4297,
      "step": 1560
    },
    {
      "epoch": 141.87272727272727,
      "eval_loss": 0.4802820384502411,
      "eval_runtime": 2.1484,
      "eval_samples_per_second": 25.6,
      "eval_steps_per_second": 3.258,
      "step": 1560
    },
    {
      "epoch": 141.96969696969697,
      "grad_norm": 0.25958773493766785,
      "learning_rate": 3.439e-05,
      "loss": 0.4322,
      "step": 1561
    },
    {
      "epoch": 142.0,
      "grad_norm": 0.45551276206970215,
      "learning_rate": 3.438e-05,
      "loss": 0.388,
      "step": 1562
    },
    {
      "epoch": 142.0969696969697,
      "grad_norm": 0.22718825936317444,
      "learning_rate": 3.4370000000000005e-05,
      "loss": 0.3969,
      "step": 1563
    },
    {
      "epoch": 142.1939393939394,
      "grad_norm": 0.2647274434566498,
      "learning_rate": 3.436e-05,
      "loss": 0.4392,
      "step": 1564
    },
    {
      "epoch": 142.29090909090908,
      "grad_norm": 0.2525234520435333,
      "learning_rate": 3.435e-05,
      "loss": 0.4372,
      "step": 1565
    },
    {
      "epoch": 142.38787878787878,
      "grad_norm": 0.27025994658470154,
      "learning_rate": 3.434e-05,
      "loss": 0.4349,
      "step": 1566
    },
    {
      "epoch": 142.4848484848485,
      "grad_norm": 0.33979251980781555,
      "learning_rate": 3.433e-05,
      "loss": 0.4484,
      "step": 1567
    },
    {
      "epoch": 142.5818181818182,
      "grad_norm": 0.28829941153526306,
      "learning_rate": 3.4320000000000003e-05,
      "loss": 0.4131,
      "step": 1568
    },
    {
      "epoch": 142.6787878787879,
      "grad_norm": 0.29526832699775696,
      "learning_rate": 3.431e-05,
      "loss": 0.4117,
      "step": 1569
    },
    {
      "epoch": 142.77575757575758,
      "grad_norm": 0.3364526331424713,
      "learning_rate": 3.430000000000001e-05,
      "loss": 0.4403,
      "step": 1570
    },
    {
      "epoch": 142.77575757575758,
      "eval_loss": 0.47912371158599854,
      "eval_runtime": 2.1728,
      "eval_samples_per_second": 25.313,
      "eval_steps_per_second": 3.222,
      "step": 1570
    },
    {
      "epoch": 142.87272727272727,
      "grad_norm": 0.27434784173965454,
      "learning_rate": 3.429e-05,
      "loss": 0.3931,
      "step": 1571
    },
    {
      "epoch": 142.96969696969697,
      "grad_norm": 0.2859874963760376,
      "learning_rate": 3.4280000000000004e-05,
      "loss": 0.4199,
      "step": 1572
    },
    {
      "epoch": 143.0,
      "grad_norm": 0.49399182200431824,
      "learning_rate": 3.427e-05,
      "loss": 0.3853,
      "step": 1573
    },
    {
      "epoch": 143.0969696969697,
      "grad_norm": 0.2726399898529053,
      "learning_rate": 3.426e-05,
      "loss": 0.4038,
      "step": 1574
    },
    {
      "epoch": 143.1939393939394,
      "grad_norm": 0.32016149163246155,
      "learning_rate": 3.4250000000000006e-05,
      "loss": 0.4222,
      "step": 1575
    },
    {
      "epoch": 143.29090909090908,
      "grad_norm": 0.32801294326782227,
      "learning_rate": 3.424e-05,
      "loss": 0.4029,
      "step": 1576
    },
    {
      "epoch": 143.38787878787878,
      "grad_norm": 0.3783630132675171,
      "learning_rate": 3.423e-05,
      "loss": 0.4632,
      "step": 1577
    },
    {
      "epoch": 143.4848484848485,
      "grad_norm": 0.503115177154541,
      "learning_rate": 3.422e-05,
      "loss": 0.391,
      "step": 1578
    },
    {
      "epoch": 143.5818181818182,
      "grad_norm": 0.3633951246738434,
      "learning_rate": 3.4210000000000006e-05,
      "loss": 0.4029,
      "step": 1579
    },
    {
      "epoch": 143.6787878787879,
      "grad_norm": 0.21935221552848816,
      "learning_rate": 3.4200000000000005e-05,
      "loss": 0.3888,
      "step": 1580
    },
    {
      "epoch": 143.6787878787879,
      "eval_loss": 0.4786372780799866,
      "eval_runtime": 2.1746,
      "eval_samples_per_second": 25.292,
      "eval_steps_per_second": 3.219,
      "step": 1580
    },
    {
      "epoch": 143.77575757575758,
      "grad_norm": 0.30795830488204956,
      "learning_rate": 3.419e-05,
      "loss": 0.4336,
      "step": 1581
    },
    {
      "epoch": 143.87272727272727,
      "grad_norm": 0.27631527185440063,
      "learning_rate": 3.418e-05,
      "loss": 0.4377,
      "step": 1582
    },
    {
      "epoch": 143.96969696969697,
      "grad_norm": 0.32671844959259033,
      "learning_rate": 3.417e-05,
      "loss": 0.4654,
      "step": 1583
    },
    {
      "epoch": 144.0,
      "grad_norm": 0.5490805506706238,
      "learning_rate": 3.4160000000000005e-05,
      "loss": 0.4344,
      "step": 1584
    },
    {
      "epoch": 144.0969696969697,
      "grad_norm": 0.31319963932037354,
      "learning_rate": 3.415e-05,
      "loss": 0.4453,
      "step": 1585
    },
    {
      "epoch": 144.1939393939394,
      "grad_norm": 0.2572191059589386,
      "learning_rate": 3.414e-05,
      "loss": 0.4076,
      "step": 1586
    },
    {
      "epoch": 144.29090909090908,
      "grad_norm": 0.2999926507472992,
      "learning_rate": 3.413e-05,
      "loss": 0.438,
      "step": 1587
    },
    {
      "epoch": 144.38787878787878,
      "grad_norm": 0.29962679743766785,
      "learning_rate": 3.412e-05,
      "loss": 0.4267,
      "step": 1588
    },
    {
      "epoch": 144.4848484848485,
      "grad_norm": 0.28824809193611145,
      "learning_rate": 3.4110000000000004e-05,
      "loss": 0.4185,
      "step": 1589
    },
    {
      "epoch": 144.5818181818182,
      "grad_norm": 0.3261265456676483,
      "learning_rate": 3.41e-05,
      "loss": 0.4141,
      "step": 1590
    },
    {
      "epoch": 144.5818181818182,
      "eval_loss": 0.47886165976524353,
      "eval_runtime": 2.1489,
      "eval_samples_per_second": 25.595,
      "eval_steps_per_second": 3.258,
      "step": 1590
    },
    {
      "epoch": 144.6787878787879,
      "grad_norm": 0.3977707028388977,
      "learning_rate": 3.409e-05,
      "loss": 0.4051,
      "step": 1591
    },
    {
      "epoch": 144.77575757575758,
      "grad_norm": 0.3052314221858978,
      "learning_rate": 3.408e-05,
      "loss": 0.3797,
      "step": 1592
    },
    {
      "epoch": 144.87272727272727,
      "grad_norm": 0.2947228252887726,
      "learning_rate": 3.4070000000000004e-05,
      "loss": 0.4332,
      "step": 1593
    },
    {
      "epoch": 144.96969696969697,
      "grad_norm": 0.2849595844745636,
      "learning_rate": 3.406e-05,
      "loss": 0.4327,
      "step": 1594
    },
    {
      "epoch": 145.0,
      "grad_norm": 0.45358598232269287,
      "learning_rate": 3.405e-05,
      "loss": 0.4358,
      "step": 1595
    },
    {
      "epoch": 145.0969696969697,
      "grad_norm": 0.30770865082740784,
      "learning_rate": 3.404e-05,
      "loss": 0.3994,
      "step": 1596
    },
    {
      "epoch": 145.1939393939394,
      "grad_norm": 0.4116463363170624,
      "learning_rate": 3.403e-05,
      "loss": 0.408,
      "step": 1597
    },
    {
      "epoch": 145.29090909090908,
      "grad_norm": 0.3829961121082306,
      "learning_rate": 3.402e-05,
      "loss": 0.4197,
      "step": 1598
    },
    {
      "epoch": 145.38787878787878,
      "grad_norm": 0.3223172128200531,
      "learning_rate": 3.401e-05,
      "loss": 0.4271,
      "step": 1599
    },
    {
      "epoch": 145.4848484848485,
      "grad_norm": 0.29979923367500305,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.4279,
      "step": 1600
    },
    {
      "epoch": 145.4848484848485,
      "eval_loss": 0.47824418544769287,
      "eval_runtime": 2.1476,
      "eval_samples_per_second": 25.61,
      "eval_steps_per_second": 3.259,
      "step": 1600
    },
    {
      "epoch": 145.5818181818182,
      "grad_norm": 0.29526492953300476,
      "learning_rate": 3.399e-05,
      "loss": 0.4245,
      "step": 1601
    },
    {
      "epoch": 145.6787878787879,
      "grad_norm": 0.39535799622535706,
      "learning_rate": 3.398e-05,
      "loss": 0.4542,
      "step": 1602
    },
    {
      "epoch": 145.77575757575758,
      "grad_norm": 0.3696836531162262,
      "learning_rate": 3.397e-05,
      "loss": 0.4573,
      "step": 1603
    },
    {
      "epoch": 145.87272727272727,
      "grad_norm": 0.2670351564884186,
      "learning_rate": 3.396e-05,
      "loss": 0.3654,
      "step": 1604
    },
    {
      "epoch": 145.96969696969697,
      "grad_norm": 0.2809720039367676,
      "learning_rate": 3.3950000000000005e-05,
      "loss": 0.4188,
      "step": 1605
    },
    {
      "epoch": 146.0,
      "grad_norm": 0.491420179605484,
      "learning_rate": 3.394e-05,
      "loss": 0.4057,
      "step": 1606
    },
    {
      "epoch": 146.0969696969697,
      "grad_norm": 0.22479571402072906,
      "learning_rate": 3.393e-05,
      "loss": 0.4385,
      "step": 1607
    },
    {
      "epoch": 146.1939393939394,
      "grad_norm": 0.3089130222797394,
      "learning_rate": 3.392e-05,
      "loss": 0.42,
      "step": 1608
    },
    {
      "epoch": 146.29090909090908,
      "grad_norm": 0.3012380301952362,
      "learning_rate": 3.3910000000000006e-05,
      "loss": 0.4482,
      "step": 1609
    },
    {
      "epoch": 146.38787878787878,
      "grad_norm": 0.2956259846687317,
      "learning_rate": 3.3900000000000004e-05,
      "loss": 0.4254,
      "step": 1610
    },
    {
      "epoch": 146.38787878787878,
      "eval_loss": 0.47690457105636597,
      "eval_runtime": 2.1714,
      "eval_samples_per_second": 25.329,
      "eval_steps_per_second": 3.224,
      "step": 1610
    },
    {
      "epoch": 146.4848484848485,
      "grad_norm": 0.2577868103981018,
      "learning_rate": 3.389e-05,
      "loss": 0.378,
      "step": 1611
    },
    {
      "epoch": 146.5818181818182,
      "grad_norm": 0.28192633390426636,
      "learning_rate": 3.388e-05,
      "loss": 0.3782,
      "step": 1612
    },
    {
      "epoch": 146.6787878787879,
      "grad_norm": 0.2914888262748718,
      "learning_rate": 3.387e-05,
      "loss": 0.4122,
      "step": 1613
    },
    {
      "epoch": 146.77575757575758,
      "grad_norm": 0.3398047983646393,
      "learning_rate": 3.3860000000000004e-05,
      "loss": 0.4108,
      "step": 1614
    },
    {
      "epoch": 146.87272727272727,
      "grad_norm": 0.32673320174217224,
      "learning_rate": 3.385e-05,
      "loss": 0.46,
      "step": 1615
    },
    {
      "epoch": 146.96969696969697,
      "grad_norm": 0.29334592819213867,
      "learning_rate": 3.384e-05,
      "loss": 0.4148,
      "step": 1616
    },
    {
      "epoch": 147.0,
      "grad_norm": 0.43029725551605225,
      "learning_rate": 3.383e-05,
      "loss": 0.4231,
      "step": 1617
    },
    {
      "epoch": 147.0969696969697,
      "grad_norm": 0.27203699946403503,
      "learning_rate": 3.3820000000000005e-05,
      "loss": 0.4093,
      "step": 1618
    },
    {
      "epoch": 147.1939393939394,
      "grad_norm": 0.3553725779056549,
      "learning_rate": 3.381e-05,
      "loss": 0.4163,
      "step": 1619
    },
    {
      "epoch": 147.29090909090908,
      "grad_norm": 0.30790579319000244,
      "learning_rate": 3.38e-05,
      "loss": 0.4504,
      "step": 1620
    },
    {
      "epoch": 147.29090909090908,
      "eval_loss": 0.4763954281806946,
      "eval_runtime": 2.1544,
      "eval_samples_per_second": 25.529,
      "eval_steps_per_second": 3.249,
      "step": 1620
    },
    {
      "epoch": 147.38787878787878,
      "grad_norm": 0.29896020889282227,
      "learning_rate": 3.379e-05,
      "loss": 0.4409,
      "step": 1621
    },
    {
      "epoch": 147.4848484848485,
      "grad_norm": 0.300333172082901,
      "learning_rate": 3.378e-05,
      "loss": 0.4207,
      "step": 1622
    },
    {
      "epoch": 147.5818181818182,
      "grad_norm": 0.2776268422603607,
      "learning_rate": 3.3770000000000004e-05,
      "loss": 0.4463,
      "step": 1623
    },
    {
      "epoch": 147.6787878787879,
      "grad_norm": 0.3045647442340851,
      "learning_rate": 3.376e-05,
      "loss": 0.3936,
      "step": 1624
    },
    {
      "epoch": 147.77575757575758,
      "grad_norm": 0.2653445601463318,
      "learning_rate": 3.375000000000001e-05,
      "loss": 0.3902,
      "step": 1625
    },
    {
      "epoch": 147.87272727272727,
      "grad_norm": 0.26968562602996826,
      "learning_rate": 3.374e-05,
      "loss": 0.4058,
      "step": 1626
    },
    {
      "epoch": 147.96969696969697,
      "grad_norm": 0.2962372601032257,
      "learning_rate": 3.373e-05,
      "loss": 0.4248,
      "step": 1627
    },
    {
      "epoch": 148.0,
      "grad_norm": 0.42462968826293945,
      "learning_rate": 3.372e-05,
      "loss": 0.3493,
      "step": 1628
    },
    {
      "epoch": 148.0969696969697,
      "grad_norm": 0.2580646574497223,
      "learning_rate": 3.371e-05,
      "loss": 0.3995,
      "step": 1629
    },
    {
      "epoch": 148.1939393939394,
      "grad_norm": 0.2709900438785553,
      "learning_rate": 3.3700000000000006e-05,
      "loss": 0.4414,
      "step": 1630
    },
    {
      "epoch": 148.1939393939394,
      "eval_loss": 0.476203054189682,
      "eval_runtime": 2.1468,
      "eval_samples_per_second": 25.62,
      "eval_steps_per_second": 3.261,
      "step": 1630
    },
    {
      "epoch": 148.29090909090908,
      "grad_norm": 0.24366122484207153,
      "learning_rate": 3.369e-05,
      "loss": 0.4154,
      "step": 1631
    },
    {
      "epoch": 148.38787878787878,
      "grad_norm": 0.29152488708496094,
      "learning_rate": 3.368e-05,
      "loss": 0.375,
      "step": 1632
    },
    {
      "epoch": 148.4848484848485,
      "grad_norm": 0.285420298576355,
      "learning_rate": 3.367e-05,
      "loss": 0.43,
      "step": 1633
    },
    {
      "epoch": 148.5818181818182,
      "grad_norm": 0.29472947120666504,
      "learning_rate": 3.366e-05,
      "loss": 0.4215,
      "step": 1634
    },
    {
      "epoch": 148.6787878787879,
      "grad_norm": 0.45701467990875244,
      "learning_rate": 3.3650000000000005e-05,
      "loss": 0.3847,
      "step": 1635
    },
    {
      "epoch": 148.77575757575758,
      "grad_norm": 0.2992357313632965,
      "learning_rate": 3.3639999999999996e-05,
      "loss": 0.4468,
      "step": 1636
    },
    {
      "epoch": 148.87272727272727,
      "grad_norm": 0.21454013884067535,
      "learning_rate": 3.363e-05,
      "loss": 0.3987,
      "step": 1637
    },
    {
      "epoch": 148.96969696969697,
      "grad_norm": 0.2769618034362793,
      "learning_rate": 3.362e-05,
      "loss": 0.4434,
      "step": 1638
    },
    {
      "epoch": 149.0,
      "grad_norm": 0.5300103425979614,
      "learning_rate": 3.3610000000000005e-05,
      "loss": 0.4426,
      "step": 1639
    },
    {
      "epoch": 149.0969696969697,
      "grad_norm": 0.2771533131599426,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 0.4,
      "step": 1640
    },
    {
      "epoch": 149.0969696969697,
      "eval_loss": 0.4750164747238159,
      "eval_runtime": 2.159,
      "eval_samples_per_second": 25.475,
      "eval_steps_per_second": 3.242,
      "step": 1640
    },
    {
      "epoch": 149.1939393939394,
      "grad_norm": 0.3044075667858124,
      "learning_rate": 3.359e-05,
      "loss": 0.3797,
      "step": 1641
    },
    {
      "epoch": 149.29090909090908,
      "grad_norm": 0.2780078053474426,
      "learning_rate": 3.358e-05,
      "loss": 0.4281,
      "step": 1642
    },
    {
      "epoch": 149.38787878787878,
      "grad_norm": 0.27099543809890747,
      "learning_rate": 3.357e-05,
      "loss": 0.4213,
      "step": 1643
    },
    {
      "epoch": 149.4848484848485,
      "grad_norm": 0.30960965156555176,
      "learning_rate": 3.3560000000000004e-05,
      "loss": 0.4088,
      "step": 1644
    },
    {
      "epoch": 149.5818181818182,
      "grad_norm": 0.29440176486968994,
      "learning_rate": 3.355e-05,
      "loss": 0.4539,
      "step": 1645
    },
    {
      "epoch": 149.6787878787879,
      "grad_norm": 0.31968364119529724,
      "learning_rate": 3.354e-05,
      "loss": 0.3991,
      "step": 1646
    },
    {
      "epoch": 149.77575757575758,
      "grad_norm": 0.2963279187679291,
      "learning_rate": 3.353e-05,
      "loss": 0.4245,
      "step": 1647
    },
    {
      "epoch": 149.87272727272727,
      "grad_norm": 0.26230543851852417,
      "learning_rate": 3.3520000000000004e-05,
      "loss": 0.4083,
      "step": 1648
    },
    {
      "epoch": 149.96969696969697,
      "grad_norm": 0.30306246876716614,
      "learning_rate": 3.351e-05,
      "loss": 0.4338,
      "step": 1649
    },
    {
      "epoch": 150.0,
      "grad_norm": 0.4665185511112213,
      "learning_rate": 3.35e-05,
      "loss": 0.4254,
      "step": 1650
    },
    {
      "epoch": 150.0,
      "eval_loss": 0.475295752286911,
      "eval_runtime": 2.1709,
      "eval_samples_per_second": 25.335,
      "eval_steps_per_second": 3.225,
      "step": 1650
    },
    {
      "epoch": 150.0969696969697,
      "grad_norm": 0.3292195796966553,
      "learning_rate": 3.349e-05,
      "loss": 0.3975,
      "step": 1651
    },
    {
      "epoch": 150.1939393939394,
      "grad_norm": 0.25012484192848206,
      "learning_rate": 3.348e-05,
      "loss": 0.378,
      "step": 1652
    },
    {
      "epoch": 150.29090909090908,
      "grad_norm": 0.31056174635887146,
      "learning_rate": 3.347e-05,
      "loss": 0.445,
      "step": 1653
    },
    {
      "epoch": 150.38787878787878,
      "grad_norm": 0.3301188051700592,
      "learning_rate": 3.346e-05,
      "loss": 0.4353,
      "step": 1654
    },
    {
      "epoch": 150.4848484848485,
      "grad_norm": 0.3136032819747925,
      "learning_rate": 3.345000000000001e-05,
      "loss": 0.4452,
      "step": 1655
    },
    {
      "epoch": 150.5818181818182,
      "grad_norm": 0.31803566217422485,
      "learning_rate": 3.344e-05,
      "loss": 0.3753,
      "step": 1656
    },
    {
      "epoch": 150.6787878787879,
      "grad_norm": 0.28205081820487976,
      "learning_rate": 3.3430000000000003e-05,
      "loss": 0.4514,
      "step": 1657
    },
    {
      "epoch": 150.77575757575758,
      "grad_norm": 0.27556318044662476,
      "learning_rate": 3.342e-05,
      "loss": 0.4138,
      "step": 1658
    },
    {
      "epoch": 150.87272727272727,
      "grad_norm": 0.25817060470581055,
      "learning_rate": 3.341e-05,
      "loss": 0.3995,
      "step": 1659
    },
    {
      "epoch": 150.96969696969697,
      "grad_norm": 0.35275375843048096,
      "learning_rate": 3.3400000000000005e-05,
      "loss": 0.4112,
      "step": 1660
    },
    {
      "epoch": 150.96969696969697,
      "eval_loss": 0.4743495285511017,
      "eval_runtime": 2.1628,
      "eval_samples_per_second": 25.43,
      "eval_steps_per_second": 3.237,
      "step": 1660
    },
    {
      "epoch": 151.0,
      "grad_norm": 0.4107286334037781,
      "learning_rate": 3.339e-05,
      "loss": 0.3963,
      "step": 1661
    },
    {
      "epoch": 151.0969696969697,
      "grad_norm": 0.3202429711818695,
      "learning_rate": 3.338e-05,
      "loss": 0.456,
      "step": 1662
    },
    {
      "epoch": 151.1939393939394,
      "grad_norm": 0.2752574384212494,
      "learning_rate": 3.337e-05,
      "loss": 0.3696,
      "step": 1663
    },
    {
      "epoch": 151.29090909090908,
      "grad_norm": 0.24195246398448944,
      "learning_rate": 3.336e-05,
      "loss": 0.3814,
      "step": 1664
    },
    {
      "epoch": 151.38787878787878,
      "grad_norm": 0.2573438584804535,
      "learning_rate": 3.3350000000000004e-05,
      "loss": 0.4085,
      "step": 1665
    },
    {
      "epoch": 151.4848484848485,
      "grad_norm": 0.29553788900375366,
      "learning_rate": 3.3339999999999996e-05,
      "loss": 0.4757,
      "step": 1666
    },
    {
      "epoch": 151.5818181818182,
      "grad_norm": 0.29257678985595703,
      "learning_rate": 3.333e-05,
      "loss": 0.4003,
      "step": 1667
    },
    {
      "epoch": 151.6787878787879,
      "grad_norm": 0.28193211555480957,
      "learning_rate": 3.332e-05,
      "loss": 0.3916,
      "step": 1668
    },
    {
      "epoch": 151.77575757575758,
      "grad_norm": 0.2794460654258728,
      "learning_rate": 3.3310000000000005e-05,
      "loss": 0.3924,
      "step": 1669
    },
    {
      "epoch": 151.87272727272727,
      "grad_norm": 0.3853585124015808,
      "learning_rate": 3.33e-05,
      "loss": 0.4365,
      "step": 1670
    },
    {
      "epoch": 151.87272727272727,
      "eval_loss": 0.47319746017456055,
      "eval_runtime": 2.1647,
      "eval_samples_per_second": 25.407,
      "eval_steps_per_second": 3.234,
      "step": 1670
    },
    {
      "epoch": 151.96969696969697,
      "grad_norm": 0.27314481139183044,
      "learning_rate": 3.329e-05,
      "loss": 0.4107,
      "step": 1671
    },
    {
      "epoch": 152.0,
      "grad_norm": 0.4267328083515167,
      "learning_rate": 3.328e-05,
      "loss": 0.4639,
      "step": 1672
    },
    {
      "epoch": 152.0969696969697,
      "grad_norm": 0.38574451208114624,
      "learning_rate": 3.327e-05,
      "loss": 0.4079,
      "step": 1673
    },
    {
      "epoch": 152.1939393939394,
      "grad_norm": 0.29176294803619385,
      "learning_rate": 3.3260000000000003e-05,
      "loss": 0.4114,
      "step": 1674
    },
    {
      "epoch": 152.29090909090908,
      "grad_norm": 0.2615834176540375,
      "learning_rate": 3.325e-05,
      "loss": 0.422,
      "step": 1675
    },
    {
      "epoch": 152.38787878787878,
      "grad_norm": 0.2873329818248749,
      "learning_rate": 3.324e-05,
      "loss": 0.4061,
      "step": 1676
    },
    {
      "epoch": 152.4848484848485,
      "grad_norm": 0.2655068635940552,
      "learning_rate": 3.323e-05,
      "loss": 0.354,
      "step": 1677
    },
    {
      "epoch": 152.5818181818182,
      "grad_norm": 0.2940961420536041,
      "learning_rate": 3.3220000000000004e-05,
      "loss": 0.4342,
      "step": 1678
    },
    {
      "epoch": 152.6787878787879,
      "grad_norm": 0.2879832684993744,
      "learning_rate": 3.321e-05,
      "loss": 0.4436,
      "step": 1679
    },
    {
      "epoch": 152.77575757575758,
      "grad_norm": 0.2613934874534607,
      "learning_rate": 3.32e-05,
      "loss": 0.3811,
      "step": 1680
    },
    {
      "epoch": 152.77575757575758,
      "eval_loss": 0.47243139147758484,
      "eval_runtime": 2.1715,
      "eval_samples_per_second": 25.328,
      "eval_steps_per_second": 3.224,
      "step": 1680
    },
    {
      "epoch": 152.87272727272727,
      "grad_norm": 0.2908010482788086,
      "learning_rate": 3.319e-05,
      "loss": 0.416,
      "step": 1681
    },
    {
      "epoch": 152.96969696969697,
      "grad_norm": 0.28120896220207214,
      "learning_rate": 3.318e-05,
      "loss": 0.4437,
      "step": 1682
    },
    {
      "epoch": 153.0,
      "grad_norm": 0.46632063388824463,
      "learning_rate": 3.317e-05,
      "loss": 0.4474,
      "step": 1683
    },
    {
      "epoch": 153.0969696969697,
      "grad_norm": 0.25776177644729614,
      "learning_rate": 3.316e-05,
      "loss": 0.3701,
      "step": 1684
    },
    {
      "epoch": 153.1939393939394,
      "grad_norm": 0.3227159380912781,
      "learning_rate": 3.3150000000000006e-05,
      "loss": 0.4368,
      "step": 1685
    },
    {
      "epoch": 153.29090909090908,
      "grad_norm": 0.27070197463035583,
      "learning_rate": 3.314e-05,
      "loss": 0.3981,
      "step": 1686
    },
    {
      "epoch": 153.38787878787878,
      "grad_norm": 0.29919660091400146,
      "learning_rate": 3.313e-05,
      "loss": 0.4038,
      "step": 1687
    },
    {
      "epoch": 153.4848484848485,
      "grad_norm": 0.25041720271110535,
      "learning_rate": 3.312e-05,
      "loss": 0.3935,
      "step": 1688
    },
    {
      "epoch": 153.5818181818182,
      "grad_norm": 0.29702797532081604,
      "learning_rate": 3.311e-05,
      "loss": 0.436,
      "step": 1689
    },
    {
      "epoch": 153.6787878787879,
      "grad_norm": 0.2959851026535034,
      "learning_rate": 3.3100000000000005e-05,
      "loss": 0.4883,
      "step": 1690
    },
    {
      "epoch": 153.6787878787879,
      "eval_loss": 0.47213518619537354,
      "eval_runtime": 2.165,
      "eval_samples_per_second": 25.404,
      "eval_steps_per_second": 3.233,
      "step": 1690
    },
    {
      "epoch": 153.77575757575758,
      "grad_norm": 0.2581026256084442,
      "learning_rate": 3.309e-05,
      "loss": 0.4061,
      "step": 1691
    },
    {
      "epoch": 153.87272727272727,
      "grad_norm": 0.2597564458847046,
      "learning_rate": 3.308e-05,
      "loss": 0.4079,
      "step": 1692
    },
    {
      "epoch": 153.96969696969697,
      "grad_norm": 0.2366248518228531,
      "learning_rate": 3.307e-05,
      "loss": 0.3716,
      "step": 1693
    },
    {
      "epoch": 154.0,
      "grad_norm": 0.5981835722923279,
      "learning_rate": 3.3060000000000005e-05,
      "loss": 0.4349,
      "step": 1694
    },
    {
      "epoch": 154.0969696969697,
      "grad_norm": 0.3014335036277771,
      "learning_rate": 3.3050000000000004e-05,
      "loss": 0.4638,
      "step": 1695
    },
    {
      "epoch": 154.1939393939394,
      "grad_norm": 0.24050500988960266,
      "learning_rate": 3.304e-05,
      "loss": 0.386,
      "step": 1696
    },
    {
      "epoch": 154.29090909090908,
      "grad_norm": 0.2561618685722351,
      "learning_rate": 3.303e-05,
      "loss": 0.3917,
      "step": 1697
    },
    {
      "epoch": 154.38787878787878,
      "grad_norm": 0.2533642649650574,
      "learning_rate": 3.302e-05,
      "loss": 0.3893,
      "step": 1698
    },
    {
      "epoch": 154.4848484848485,
      "grad_norm": 0.314961314201355,
      "learning_rate": 3.3010000000000004e-05,
      "loss": 0.4156,
      "step": 1699
    },
    {
      "epoch": 154.5818181818182,
      "grad_norm": 0.28019827604293823,
      "learning_rate": 3.3e-05,
      "loss": 0.4582,
      "step": 1700
    },
    {
      "epoch": 154.5818181818182,
      "eval_loss": 0.47177475690841675,
      "eval_runtime": 2.1504,
      "eval_samples_per_second": 25.577,
      "eval_steps_per_second": 3.255,
      "step": 1700
    },
    {
      "epoch": 154.6787878787879,
      "grad_norm": 0.27833741903305054,
      "learning_rate": 3.299e-05,
      "loss": 0.4155,
      "step": 1701
    },
    {
      "epoch": 154.77575757575758,
      "grad_norm": 0.2910160720348358,
      "learning_rate": 3.298e-05,
      "loss": 0.4179,
      "step": 1702
    },
    {
      "epoch": 154.87272727272727,
      "grad_norm": 0.26338979601860046,
      "learning_rate": 3.297e-05,
      "loss": 0.4019,
      "step": 1703
    },
    {
      "epoch": 154.96969696969697,
      "grad_norm": 0.23985280096530914,
      "learning_rate": 3.296e-05,
      "loss": 0.3711,
      "step": 1704
    },
    {
      "epoch": 155.0,
      "grad_norm": 0.4723038971424103,
      "learning_rate": 3.295e-05,
      "loss": 0.4152,
      "step": 1705
    },
    {
      "epoch": 155.0969696969697,
      "grad_norm": 0.27281108498573303,
      "learning_rate": 3.2940000000000006e-05,
      "loss": 0.4006,
      "step": 1706
    },
    {
      "epoch": 155.1939393939394,
      "grad_norm": 0.22958678007125854,
      "learning_rate": 3.293e-05,
      "loss": 0.3824,
      "step": 1707
    },
    {
      "epoch": 155.29090909090908,
      "grad_norm": 0.23069177567958832,
      "learning_rate": 3.292e-05,
      "loss": 0.4577,
      "step": 1708
    },
    {
      "epoch": 155.38787878787878,
      "grad_norm": 0.3019179403781891,
      "learning_rate": 3.291e-05,
      "loss": 0.4202,
      "step": 1709
    },
    {
      "epoch": 155.4848484848485,
      "grad_norm": 0.3441668450832367,
      "learning_rate": 3.29e-05,
      "loss": 0.4257,
      "step": 1710
    },
    {
      "epoch": 155.4848484848485,
      "eval_loss": 0.4712180495262146,
      "eval_runtime": 2.1716,
      "eval_samples_per_second": 25.327,
      "eval_steps_per_second": 3.223,
      "step": 1710
    },
    {
      "epoch": 155.5818181818182,
      "grad_norm": 0.2304125428199768,
      "learning_rate": 3.2890000000000005e-05,
      "loss": 0.4019,
      "step": 1711
    },
    {
      "epoch": 155.6787878787879,
      "grad_norm": 0.3568699359893799,
      "learning_rate": 3.288e-05,
      "loss": 0.4266,
      "step": 1712
    },
    {
      "epoch": 155.77575757575758,
      "grad_norm": 0.35448890924453735,
      "learning_rate": 3.287e-05,
      "loss": 0.3887,
      "step": 1713
    },
    {
      "epoch": 155.87272727272727,
      "grad_norm": 0.2856895327568054,
      "learning_rate": 3.286e-05,
      "loss": 0.3936,
      "step": 1714
    },
    {
      "epoch": 155.96969696969697,
      "grad_norm": 0.26315397024154663,
      "learning_rate": 3.2850000000000006e-05,
      "loss": 0.4024,
      "step": 1715
    },
    {
      "epoch": 156.0,
      "grad_norm": 0.49158674478530884,
      "learning_rate": 3.2840000000000004e-05,
      "loss": 0.4161,
      "step": 1716
    },
    {
      "epoch": 156.0969696969697,
      "grad_norm": 0.2861306071281433,
      "learning_rate": 3.283e-05,
      "loss": 0.4164,
      "step": 1717
    },
    {
      "epoch": 156.1939393939394,
      "grad_norm": 0.2937961220741272,
      "learning_rate": 3.282e-05,
      "loss": 0.4228,
      "step": 1718
    },
    {
      "epoch": 156.29090909090908,
      "grad_norm": 0.26782360672950745,
      "learning_rate": 3.281e-05,
      "loss": 0.4491,
      "step": 1719
    },
    {
      "epoch": 156.38787878787878,
      "grad_norm": 0.30058830976486206,
      "learning_rate": 3.2800000000000004e-05,
      "loss": 0.3851,
      "step": 1720
    },
    {
      "epoch": 156.38787878787878,
      "eval_loss": 0.47035104036331177,
      "eval_runtime": 2.1673,
      "eval_samples_per_second": 25.377,
      "eval_steps_per_second": 3.23,
      "step": 1720
    },
    {
      "epoch": 156.4848484848485,
      "grad_norm": 0.22291842103004456,
      "learning_rate": 3.279e-05,
      "loss": 0.3677,
      "step": 1721
    },
    {
      "epoch": 156.5818181818182,
      "grad_norm": 0.26373815536499023,
      "learning_rate": 3.278e-05,
      "loss": 0.4028,
      "step": 1722
    },
    {
      "epoch": 156.6787878787879,
      "grad_norm": 0.2856782078742981,
      "learning_rate": 3.277e-05,
      "loss": 0.4263,
      "step": 1723
    },
    {
      "epoch": 156.77575757575758,
      "grad_norm": 0.24891617894172668,
      "learning_rate": 3.2760000000000005e-05,
      "loss": 0.428,
      "step": 1724
    },
    {
      "epoch": 156.87272727272727,
      "grad_norm": 0.3333253860473633,
      "learning_rate": 3.275e-05,
      "loss": 0.4135,
      "step": 1725
    },
    {
      "epoch": 156.96969696969697,
      "grad_norm": 0.2727908194065094,
      "learning_rate": 3.274e-05,
      "loss": 0.3742,
      "step": 1726
    },
    {
      "epoch": 157.0,
      "grad_norm": 0.5115933418273926,
      "learning_rate": 3.273e-05,
      "loss": 0.4297,
      "step": 1727
    },
    {
      "epoch": 157.0969696969697,
      "grad_norm": 0.2451162487268448,
      "learning_rate": 3.272e-05,
      "loss": 0.4076,
      "step": 1728
    },
    {
      "epoch": 157.1939393939394,
      "grad_norm": 0.32220616936683655,
      "learning_rate": 3.2710000000000004e-05,
      "loss": 0.4132,
      "step": 1729
    },
    {
      "epoch": 157.29090909090908,
      "grad_norm": 0.26839929819107056,
      "learning_rate": 3.27e-05,
      "loss": 0.4217,
      "step": 1730
    },
    {
      "epoch": 157.29090909090908,
      "eval_loss": 0.4702255129814148,
      "eval_runtime": 2.1582,
      "eval_samples_per_second": 25.484,
      "eval_steps_per_second": 3.243,
      "step": 1730
    },
    {
      "epoch": 157.38787878787878,
      "grad_norm": 0.26511865854263306,
      "learning_rate": 3.269000000000001e-05,
      "loss": 0.3892,
      "step": 1731
    },
    {
      "epoch": 157.4848484848485,
      "grad_norm": 0.2777853310108185,
      "learning_rate": 3.268e-05,
      "loss": 0.4259,
      "step": 1732
    },
    {
      "epoch": 157.5818181818182,
      "grad_norm": 0.32341113686561584,
      "learning_rate": 3.267e-05,
      "loss": 0.3851,
      "step": 1733
    },
    {
      "epoch": 157.6787878787879,
      "grad_norm": 0.29954448342323303,
      "learning_rate": 3.266e-05,
      "loss": 0.4389,
      "step": 1734
    },
    {
      "epoch": 157.77575757575758,
      "grad_norm": 0.24362747371196747,
      "learning_rate": 3.265e-05,
      "loss": 0.404,
      "step": 1735
    },
    {
      "epoch": 157.87272727272727,
      "grad_norm": 0.27364447712898254,
      "learning_rate": 3.2640000000000006e-05,
      "loss": 0.445,
      "step": 1736
    },
    {
      "epoch": 157.96969696969697,
      "grad_norm": 0.26995348930358887,
      "learning_rate": 3.263e-05,
      "loss": 0.3549,
      "step": 1737
    },
    {
      "epoch": 158.0,
      "grad_norm": 0.5283206105232239,
      "learning_rate": 3.262e-05,
      "loss": 0.4122,
      "step": 1738
    },
    {
      "epoch": 158.0969696969697,
      "grad_norm": 0.30771592259407043,
      "learning_rate": 3.261e-05,
      "loss": 0.4484,
      "step": 1739
    },
    {
      "epoch": 158.1939393939394,
      "grad_norm": 0.2356712818145752,
      "learning_rate": 3.26e-05,
      "loss": 0.4009,
      "step": 1740
    },
    {
      "epoch": 158.1939393939394,
      "eval_loss": 0.46936678886413574,
      "eval_runtime": 2.1604,
      "eval_samples_per_second": 25.458,
      "eval_steps_per_second": 3.24,
      "step": 1740
    },
    {
      "epoch": 158.29090909090908,
      "grad_norm": 0.2755451500415802,
      "learning_rate": 3.2590000000000005e-05,
      "loss": 0.3879,
      "step": 1741
    },
    {
      "epoch": 158.38787878787878,
      "grad_norm": 0.3021290898323059,
      "learning_rate": 3.2579999999999996e-05,
      "loss": 0.3837,
      "step": 1742
    },
    {
      "epoch": 158.4848484848485,
      "grad_norm": 0.33612263202667236,
      "learning_rate": 3.257e-05,
      "loss": 0.4023,
      "step": 1743
    },
    {
      "epoch": 158.5818181818182,
      "grad_norm": 0.29134008288383484,
      "learning_rate": 3.256e-05,
      "loss": 0.3893,
      "step": 1744
    },
    {
      "epoch": 158.6787878787879,
      "grad_norm": 0.28086304664611816,
      "learning_rate": 3.2550000000000005e-05,
      "loss": 0.4277,
      "step": 1745
    },
    {
      "epoch": 158.77575757575758,
      "grad_norm": 0.298905611038208,
      "learning_rate": 3.2540000000000004e-05,
      "loss": 0.4219,
      "step": 1746
    },
    {
      "epoch": 158.87272727272727,
      "grad_norm": 0.23620347678661346,
      "learning_rate": 3.253e-05,
      "loss": 0.3694,
      "step": 1747
    },
    {
      "epoch": 158.96969696969697,
      "grad_norm": 0.34921467304229736,
      "learning_rate": 3.252e-05,
      "loss": 0.4472,
      "step": 1748
    },
    {
      "epoch": 159.0,
      "grad_norm": 0.4663753807544708,
      "learning_rate": 3.251e-05,
      "loss": 0.411,
      "step": 1749
    },
    {
      "epoch": 159.0969696969697,
      "grad_norm": 0.35847172141075134,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.4179,
      "step": 1750
    },
    {
      "epoch": 159.0969696969697,
      "eval_loss": 0.4685215353965759,
      "eval_runtime": 2.1614,
      "eval_samples_per_second": 25.446,
      "eval_steps_per_second": 3.239,
      "step": 1750
    },
    {
      "epoch": 159.1939393939394,
      "grad_norm": 0.282756507396698,
      "learning_rate": 3.249e-05,
      "loss": 0.409,
      "step": 1751
    },
    {
      "epoch": 159.29090909090908,
      "grad_norm": 0.2560710310935974,
      "learning_rate": 3.248e-05,
      "loss": 0.4262,
      "step": 1752
    },
    {
      "epoch": 159.38787878787878,
      "grad_norm": 0.26294830441474915,
      "learning_rate": 3.247e-05,
      "loss": 0.343,
      "step": 1753
    },
    {
      "epoch": 159.4848484848485,
      "grad_norm": 0.3224695324897766,
      "learning_rate": 3.2460000000000004e-05,
      "loss": 0.4206,
      "step": 1754
    },
    {
      "epoch": 159.5818181818182,
      "grad_norm": 0.28522032499313354,
      "learning_rate": 3.245e-05,
      "loss": 0.3707,
      "step": 1755
    },
    {
      "epoch": 159.6787878787879,
      "grad_norm": 0.34134167432785034,
      "learning_rate": 3.244e-05,
      "loss": 0.4514,
      "step": 1756
    },
    {
      "epoch": 159.77575757575758,
      "grad_norm": 0.24785830080509186,
      "learning_rate": 3.243e-05,
      "loss": 0.3801,
      "step": 1757
    },
    {
      "epoch": 159.87272727272727,
      "grad_norm": 0.35771289467811584,
      "learning_rate": 3.242e-05,
      "loss": 0.3888,
      "step": 1758
    },
    {
      "epoch": 159.96969696969697,
      "grad_norm": 0.40290340781211853,
      "learning_rate": 3.241e-05,
      "loss": 0.4668,
      "step": 1759
    },
    {
      "epoch": 160.0,
      "grad_norm": 0.5135341286659241,
      "learning_rate": 3.24e-05,
      "loss": 0.3879,
      "step": 1760
    },
    {
      "epoch": 160.0,
      "eval_loss": 0.4681518077850342,
      "eval_runtime": 2.1812,
      "eval_samples_per_second": 25.215,
      "eval_steps_per_second": 3.209,
      "step": 1760
    },
    {
      "epoch": 160.0969696969697,
      "grad_norm": 0.2755342423915863,
      "learning_rate": 3.239000000000001e-05,
      "loss": 0.3986,
      "step": 1761
    },
    {
      "epoch": 160.1939393939394,
      "grad_norm": 0.34573477506637573,
      "learning_rate": 3.238e-05,
      "loss": 0.4542,
      "step": 1762
    },
    {
      "epoch": 160.29090909090908,
      "grad_norm": 0.31546109914779663,
      "learning_rate": 3.2370000000000003e-05,
      "loss": 0.4341,
      "step": 1763
    },
    {
      "epoch": 160.38787878787878,
      "grad_norm": 0.3111988306045532,
      "learning_rate": 3.236e-05,
      "loss": 0.4386,
      "step": 1764
    },
    {
      "epoch": 160.4848484848485,
      "grad_norm": 0.33897435665130615,
      "learning_rate": 3.235e-05,
      "loss": 0.3434,
      "step": 1765
    },
    {
      "epoch": 160.5818181818182,
      "grad_norm": 0.27282023429870605,
      "learning_rate": 3.2340000000000005e-05,
      "loss": 0.4253,
      "step": 1766
    },
    {
      "epoch": 160.6787878787879,
      "grad_norm": 0.25652244687080383,
      "learning_rate": 3.233e-05,
      "loss": 0.4284,
      "step": 1767
    },
    {
      "epoch": 160.77575757575758,
      "grad_norm": 0.3286043107509613,
      "learning_rate": 3.232e-05,
      "loss": 0.3984,
      "step": 1768
    },
    {
      "epoch": 160.87272727272727,
      "grad_norm": 0.34483659267425537,
      "learning_rate": 3.231e-05,
      "loss": 0.3858,
      "step": 1769
    },
    {
      "epoch": 160.96969696969697,
      "grad_norm": 0.28678274154663086,
      "learning_rate": 3.2300000000000006e-05,
      "loss": 0.347,
      "step": 1770
    },
    {
      "epoch": 160.96969696969697,
      "eval_loss": 0.46771591901779175,
      "eval_runtime": 2.1478,
      "eval_samples_per_second": 25.607,
      "eval_steps_per_second": 3.259,
      "step": 1770
    },
    {
      "epoch": 161.0,
      "grad_norm": 0.4694156348705292,
      "learning_rate": 3.2290000000000004e-05,
      "loss": 0.4208,
      "step": 1771
    },
    {
      "epoch": 161.0969696969697,
      "grad_norm": 0.2683364152908325,
      "learning_rate": 3.2279999999999996e-05,
      "loss": 0.4647,
      "step": 1772
    },
    {
      "epoch": 161.1939393939394,
      "grad_norm": 0.27918240427970886,
      "learning_rate": 3.227e-05,
      "loss": 0.4131,
      "step": 1773
    },
    {
      "epoch": 161.29090909090908,
      "grad_norm": 0.25537046790122986,
      "learning_rate": 3.226e-05,
      "loss": 0.3655,
      "step": 1774
    },
    {
      "epoch": 161.38787878787878,
      "grad_norm": 0.2649289071559906,
      "learning_rate": 3.2250000000000005e-05,
      "loss": 0.4214,
      "step": 1775
    },
    {
      "epoch": 161.4848484848485,
      "grad_norm": 0.3552446663379669,
      "learning_rate": 3.224e-05,
      "loss": 0.416,
      "step": 1776
    },
    {
      "epoch": 161.5818181818182,
      "grad_norm": 0.2456580251455307,
      "learning_rate": 3.223e-05,
      "loss": 0.3932,
      "step": 1777
    },
    {
      "epoch": 161.6787878787879,
      "grad_norm": 0.32637685537338257,
      "learning_rate": 3.222e-05,
      "loss": 0.4071,
      "step": 1778
    },
    {
      "epoch": 161.77575757575758,
      "grad_norm": 0.3025254011154175,
      "learning_rate": 3.221e-05,
      "loss": 0.3814,
      "step": 1779
    },
    {
      "epoch": 161.87272727272727,
      "grad_norm": 0.2737649977207184,
      "learning_rate": 3.2200000000000003e-05,
      "loss": 0.3886,
      "step": 1780
    },
    {
      "epoch": 161.87272727272727,
      "eval_loss": 0.4672313332557678,
      "eval_runtime": 2.1526,
      "eval_samples_per_second": 25.55,
      "eval_steps_per_second": 3.252,
      "step": 1780
    },
    {
      "epoch": 161.96969696969697,
      "grad_norm": 0.25991830229759216,
      "learning_rate": 3.219e-05,
      "loss": 0.4104,
      "step": 1781
    },
    {
      "epoch": 162.0,
      "grad_norm": 0.4407697021961212,
      "learning_rate": 3.218e-05,
      "loss": 0.3704,
      "step": 1782
    },
    {
      "epoch": 162.0969696969697,
      "grad_norm": 0.25895529985427856,
      "learning_rate": 3.217e-05,
      "loss": 0.4313,
      "step": 1783
    },
    {
      "epoch": 162.1939393939394,
      "grad_norm": 0.26714879274368286,
      "learning_rate": 3.2160000000000004e-05,
      "loss": 0.4393,
      "step": 1784
    },
    {
      "epoch": 162.29090909090908,
      "grad_norm": 0.41708916425704956,
      "learning_rate": 3.215e-05,
      "loss": 0.39,
      "step": 1785
    },
    {
      "epoch": 162.38787878787878,
      "grad_norm": 0.28017428517341614,
      "learning_rate": 3.214e-05,
      "loss": 0.4031,
      "step": 1786
    },
    {
      "epoch": 162.4848484848485,
      "grad_norm": 0.3049089312553406,
      "learning_rate": 3.213e-05,
      "loss": 0.3572,
      "step": 1787
    },
    {
      "epoch": 162.5818181818182,
      "grad_norm": 0.2849743664264679,
      "learning_rate": 3.212e-05,
      "loss": 0.3748,
      "step": 1788
    },
    {
      "epoch": 162.6787878787879,
      "grad_norm": 0.2804197371006012,
      "learning_rate": 3.211e-05,
      "loss": 0.4213,
      "step": 1789
    },
    {
      "epoch": 162.77575757575758,
      "grad_norm": 0.24261786043643951,
      "learning_rate": 3.21e-05,
      "loss": 0.4041,
      "step": 1790
    },
    {
      "epoch": 162.77575757575758,
      "eval_loss": 0.46590346097946167,
      "eval_runtime": 2.1701,
      "eval_samples_per_second": 25.345,
      "eval_steps_per_second": 3.226,
      "step": 1790
    },
    {
      "epoch": 162.87272727272727,
      "grad_norm": 0.3327369689941406,
      "learning_rate": 3.2090000000000006e-05,
      "loss": 0.4168,
      "step": 1791
    },
    {
      "epoch": 162.96969696969697,
      "grad_norm": 0.3327149748802185,
      "learning_rate": 3.208e-05,
      "loss": 0.3969,
      "step": 1792
    },
    {
      "epoch": 163.0,
      "grad_norm": 0.7095963358879089,
      "learning_rate": 3.207e-05,
      "loss": 0.4276,
      "step": 1793
    },
    {
      "epoch": 163.0969696969697,
      "grad_norm": 0.2949429452419281,
      "learning_rate": 3.206e-05,
      "loss": 0.4401,
      "step": 1794
    },
    {
      "epoch": 163.1939393939394,
      "grad_norm": 0.2918664813041687,
      "learning_rate": 3.205e-05,
      "loss": 0.4117,
      "step": 1795
    },
    {
      "epoch": 163.29090909090908,
      "grad_norm": 0.3940178453922272,
      "learning_rate": 3.2040000000000005e-05,
      "loss": 0.4197,
      "step": 1796
    },
    {
      "epoch": 163.38787878787878,
      "grad_norm": 0.27344033122062683,
      "learning_rate": 3.2029999999999997e-05,
      "loss": 0.3974,
      "step": 1797
    },
    {
      "epoch": 163.4848484848485,
      "grad_norm": 0.33115363121032715,
      "learning_rate": 3.202e-05,
      "loss": 0.4082,
      "step": 1798
    },
    {
      "epoch": 163.5818181818182,
      "grad_norm": 0.29205378890037537,
      "learning_rate": 3.201e-05,
      "loss": 0.4047,
      "step": 1799
    },
    {
      "epoch": 163.6787878787879,
      "grad_norm": 0.29947832226753235,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.3817,
      "step": 1800
    },
    {
      "epoch": 163.6787878787879,
      "eval_loss": 0.4654845893383026,
      "eval_runtime": 2.1572,
      "eval_samples_per_second": 25.496,
      "eval_steps_per_second": 3.245,
      "step": 1800
    },
    {
      "epoch": 163.77575757575758,
      "grad_norm": 0.32452553510665894,
      "learning_rate": 3.1990000000000004e-05,
      "loss": 0.3824,
      "step": 1801
    },
    {
      "epoch": 163.87272727272727,
      "grad_norm": 0.24157431721687317,
      "learning_rate": 3.198e-05,
      "loss": 0.4015,
      "step": 1802
    },
    {
      "epoch": 163.96969696969697,
      "grad_norm": 0.30334722995758057,
      "learning_rate": 3.197e-05,
      "loss": 0.4231,
      "step": 1803
    },
    {
      "epoch": 164.0,
      "grad_norm": 0.47599658370018005,
      "learning_rate": 3.196e-05,
      "loss": 0.296,
      "step": 1804
    },
    {
      "epoch": 164.0969696969697,
      "grad_norm": 0.23312796652317047,
      "learning_rate": 3.1950000000000004e-05,
      "loss": 0.4002,
      "step": 1805
    },
    {
      "epoch": 164.1939393939394,
      "grad_norm": 0.302068293094635,
      "learning_rate": 3.194e-05,
      "loss": 0.3674,
      "step": 1806
    },
    {
      "epoch": 164.29090909090908,
      "grad_norm": 0.3966183662414551,
      "learning_rate": 3.193e-05,
      "loss": 0.4546,
      "step": 1807
    },
    {
      "epoch": 164.38787878787878,
      "grad_norm": 0.25288668274879456,
      "learning_rate": 3.192e-05,
      "loss": 0.4078,
      "step": 1808
    },
    {
      "epoch": 164.4848484848485,
      "grad_norm": 0.28041747212409973,
      "learning_rate": 3.191e-05,
      "loss": 0.3804,
      "step": 1809
    },
    {
      "epoch": 164.5818181818182,
      "grad_norm": 0.2514358460903168,
      "learning_rate": 3.19e-05,
      "loss": 0.3812,
      "step": 1810
    },
    {
      "epoch": 164.5818181818182,
      "eval_loss": 0.4653727412223816,
      "eval_runtime": 2.1628,
      "eval_samples_per_second": 25.43,
      "eval_steps_per_second": 3.237,
      "step": 1810
    },
    {
      "epoch": 164.6787878787879,
      "grad_norm": 0.3115096092224121,
      "learning_rate": 3.189e-05,
      "loss": 0.4226,
      "step": 1811
    },
    {
      "epoch": 164.77575757575758,
      "grad_norm": 0.28320327401161194,
      "learning_rate": 3.188e-05,
      "loss": 0.3643,
      "step": 1812
    },
    {
      "epoch": 164.87272727272727,
      "grad_norm": 0.3034592866897583,
      "learning_rate": 3.187e-05,
      "loss": 0.4262,
      "step": 1813
    },
    {
      "epoch": 164.96969696969697,
      "grad_norm": 0.33773717284202576,
      "learning_rate": 3.186e-05,
      "loss": 0.4256,
      "step": 1814
    },
    {
      "epoch": 165.0,
      "grad_norm": 0.5290171504020691,
      "learning_rate": 3.185e-05,
      "loss": 0.3871,
      "step": 1815
    },
    {
      "epoch": 165.0969696969697,
      "grad_norm": 0.25624457001686096,
      "learning_rate": 3.184e-05,
      "loss": 0.4394,
      "step": 1816
    },
    {
      "epoch": 165.1939393939394,
      "grad_norm": 0.25744593143463135,
      "learning_rate": 3.1830000000000005e-05,
      "loss": 0.4091,
      "step": 1817
    },
    {
      "epoch": 165.29090909090908,
      "grad_norm": 0.339099258184433,
      "learning_rate": 3.182e-05,
      "loss": 0.3906,
      "step": 1818
    },
    {
      "epoch": 165.38787878787878,
      "grad_norm": 0.30119866132736206,
      "learning_rate": 3.181e-05,
      "loss": 0.4274,
      "step": 1819
    },
    {
      "epoch": 165.4848484848485,
      "grad_norm": 0.32787880301475525,
      "learning_rate": 3.18e-05,
      "loss": 0.441,
      "step": 1820
    },
    {
      "epoch": 165.4848484848485,
      "eval_loss": 0.4649874269962311,
      "eval_runtime": 2.1606,
      "eval_samples_per_second": 25.456,
      "eval_steps_per_second": 3.24,
      "step": 1820
    },
    {
      "epoch": 165.5818181818182,
      "grad_norm": 0.316469669342041,
      "learning_rate": 3.1790000000000006e-05,
      "loss": 0.3948,
      "step": 1821
    },
    {
      "epoch": 165.6787878787879,
      "grad_norm": 0.28067582845687866,
      "learning_rate": 3.1780000000000004e-05,
      "loss": 0.3742,
      "step": 1822
    },
    {
      "epoch": 165.77575757575758,
      "grad_norm": 0.2999702990055084,
      "learning_rate": 3.177e-05,
      "loss": 0.3831,
      "step": 1823
    },
    {
      "epoch": 165.87272727272727,
      "grad_norm": 0.22824245691299438,
      "learning_rate": 3.176e-05,
      "loss": 0.3727,
      "step": 1824
    },
    {
      "epoch": 165.96969696969697,
      "grad_norm": 0.3537152409553528,
      "learning_rate": 3.175e-05,
      "loss": 0.4028,
      "step": 1825
    },
    {
      "epoch": 166.0,
      "grad_norm": 0.49454474449157715,
      "learning_rate": 3.1740000000000004e-05,
      "loss": 0.3595,
      "step": 1826
    },
    {
      "epoch": 166.0969696969697,
      "grad_norm": 0.3131580054759979,
      "learning_rate": 3.173e-05,
      "loss": 0.4085,
      "step": 1827
    },
    {
      "epoch": 166.1939393939394,
      "grad_norm": 0.3912461996078491,
      "learning_rate": 3.172e-05,
      "loss": 0.4287,
      "step": 1828
    },
    {
      "epoch": 166.29090909090908,
      "grad_norm": 0.30655741691589355,
      "learning_rate": 3.171e-05,
      "loss": 0.4043,
      "step": 1829
    },
    {
      "epoch": 166.38787878787878,
      "grad_norm": 0.3822678029537201,
      "learning_rate": 3.1700000000000005e-05,
      "loss": 0.3814,
      "step": 1830
    },
    {
      "epoch": 166.38787878787878,
      "eval_loss": 0.4645639955997467,
      "eval_runtime": 2.1598,
      "eval_samples_per_second": 25.465,
      "eval_steps_per_second": 3.241,
      "step": 1830
    },
    {
      "epoch": 166.4848484848485,
      "grad_norm": 0.346089243888855,
      "learning_rate": 3.169e-05,
      "loss": 0.3884,
      "step": 1831
    },
    {
      "epoch": 166.5818181818182,
      "grad_norm": 0.2965271472930908,
      "learning_rate": 3.168e-05,
      "loss": 0.4098,
      "step": 1832
    },
    {
      "epoch": 166.6787878787879,
      "grad_norm": 0.29764848947525024,
      "learning_rate": 3.167e-05,
      "loss": 0.4053,
      "step": 1833
    },
    {
      "epoch": 166.77575757575758,
      "grad_norm": 0.24490293860435486,
      "learning_rate": 3.166e-05,
      "loss": 0.3439,
      "step": 1834
    },
    {
      "epoch": 166.87272727272727,
      "grad_norm": 0.28927063941955566,
      "learning_rate": 3.1650000000000004e-05,
      "loss": 0.4155,
      "step": 1835
    },
    {
      "epoch": 166.96969696969697,
      "grad_norm": 0.27231332659721375,
      "learning_rate": 3.164e-05,
      "loss": 0.4291,
      "step": 1836
    },
    {
      "epoch": 167.0,
      "grad_norm": 0.5904393792152405,
      "learning_rate": 3.163000000000001e-05,
      "loss": 0.4002,
      "step": 1837
    },
    {
      "epoch": 167.0969696969697,
      "grad_norm": 0.30065590143203735,
      "learning_rate": 3.162e-05,
      "loss": 0.4008,
      "step": 1838
    },
    {
      "epoch": 167.1939393939394,
      "grad_norm": 0.2866343855857849,
      "learning_rate": 3.1610000000000004e-05,
      "loss": 0.3725,
      "step": 1839
    },
    {
      "epoch": 167.29090909090908,
      "grad_norm": 0.2981482744216919,
      "learning_rate": 3.16e-05,
      "loss": 0.4071,
      "step": 1840
    },
    {
      "epoch": 167.29090909090908,
      "eval_loss": 0.46402543783187866,
      "eval_runtime": 2.1685,
      "eval_samples_per_second": 25.363,
      "eval_steps_per_second": 3.228,
      "step": 1840
    },
    {
      "epoch": 167.38787878787878,
      "grad_norm": 0.2863614857196808,
      "learning_rate": 3.159e-05,
      "loss": 0.3969,
      "step": 1841
    },
    {
      "epoch": 167.4848484848485,
      "grad_norm": 0.276218056678772,
      "learning_rate": 3.1580000000000006e-05,
      "loss": 0.4347,
      "step": 1842
    },
    {
      "epoch": 167.5818181818182,
      "grad_norm": 0.29230207204818726,
      "learning_rate": 3.157e-05,
      "loss": 0.4602,
      "step": 1843
    },
    {
      "epoch": 167.6787878787879,
      "grad_norm": 0.36772671341896057,
      "learning_rate": 3.156e-05,
      "loss": 0.399,
      "step": 1844
    },
    {
      "epoch": 167.77575757575758,
      "grad_norm": 0.38044774532318115,
      "learning_rate": 3.155e-05,
      "loss": 0.3632,
      "step": 1845
    },
    {
      "epoch": 167.87272727272727,
      "grad_norm": 0.2674912214279175,
      "learning_rate": 3.154e-05,
      "loss": 0.397,
      "step": 1846
    },
    {
      "epoch": 167.96969696969697,
      "grad_norm": 0.3134581446647644,
      "learning_rate": 3.1530000000000005e-05,
      "loss": 0.3744,
      "step": 1847
    },
    {
      "epoch": 168.0,
      "grad_norm": 0.4266735017299652,
      "learning_rate": 3.1519999999999996e-05,
      "loss": 0.3918,
      "step": 1848
    },
    {
      "epoch": 168.0969696969697,
      "grad_norm": 0.3174222409725189,
      "learning_rate": 3.151e-05,
      "loss": 0.3921,
      "step": 1849
    },
    {
      "epoch": 168.1939393939394,
      "grad_norm": 0.2756790518760681,
      "learning_rate": 3.15e-05,
      "loss": 0.3856,
      "step": 1850
    },
    {
      "epoch": 168.1939393939394,
      "eval_loss": 0.4628044366836548,
      "eval_runtime": 2.209,
      "eval_samples_per_second": 24.898,
      "eval_steps_per_second": 3.169,
      "step": 1850
    },
    {
      "epoch": 168.29090909090908,
      "grad_norm": 0.290212482213974,
      "learning_rate": 3.1490000000000005e-05,
      "loss": 0.4029,
      "step": 1851
    },
    {
      "epoch": 168.38787878787878,
      "grad_norm": 0.3511327803134918,
      "learning_rate": 3.1480000000000004e-05,
      "loss": 0.3939,
      "step": 1852
    },
    {
      "epoch": 168.4848484848485,
      "grad_norm": 0.2698800563812256,
      "learning_rate": 3.147e-05,
      "loss": 0.4087,
      "step": 1853
    },
    {
      "epoch": 168.5818181818182,
      "grad_norm": 0.30846282839775085,
      "learning_rate": 3.146e-05,
      "loss": 0.3729,
      "step": 1854
    },
    {
      "epoch": 168.6787878787879,
      "grad_norm": 0.32614874839782715,
      "learning_rate": 3.145e-05,
      "loss": 0.4484,
      "step": 1855
    },
    {
      "epoch": 168.77575757575758,
      "grad_norm": 0.27173444628715515,
      "learning_rate": 3.1440000000000004e-05,
      "loss": 0.3884,
      "step": 1856
    },
    {
      "epoch": 168.87272727272727,
      "grad_norm": 0.2709703743457794,
      "learning_rate": 3.143e-05,
      "loss": 0.3991,
      "step": 1857
    },
    {
      "epoch": 168.96969696969697,
      "grad_norm": 0.24700695276260376,
      "learning_rate": 3.142e-05,
      "loss": 0.3884,
      "step": 1858
    },
    {
      "epoch": 169.0,
      "grad_norm": 0.48396357893943787,
      "learning_rate": 3.141e-05,
      "loss": 0.4496,
      "step": 1859
    },
    {
      "epoch": 169.0969696969697,
      "grad_norm": 0.285308837890625,
      "learning_rate": 3.1400000000000004e-05,
      "loss": 0.4341,
      "step": 1860
    },
    {
      "epoch": 169.0969696969697,
      "eval_loss": 0.4620818495750427,
      "eval_runtime": 2.157,
      "eval_samples_per_second": 25.499,
      "eval_steps_per_second": 3.245,
      "step": 1860
    },
    {
      "epoch": 169.1939393939394,
      "grad_norm": 0.26520836353302,
      "learning_rate": 3.139e-05,
      "loss": 0.3765,
      "step": 1861
    },
    {
      "epoch": 169.29090909090908,
      "grad_norm": 0.2857014536857605,
      "learning_rate": 3.138e-05,
      "loss": 0.402,
      "step": 1862
    },
    {
      "epoch": 169.38787878787878,
      "grad_norm": 0.3332759737968445,
      "learning_rate": 3.137e-05,
      "loss": 0.3689,
      "step": 1863
    },
    {
      "epoch": 169.4848484848485,
      "grad_norm": 0.29484209418296814,
      "learning_rate": 3.136e-05,
      "loss": 0.4029,
      "step": 1864
    },
    {
      "epoch": 169.5818181818182,
      "grad_norm": 0.3884888291358948,
      "learning_rate": 3.135e-05,
      "loss": 0.4124,
      "step": 1865
    },
    {
      "epoch": 169.6787878787879,
      "grad_norm": 0.31348225474357605,
      "learning_rate": 3.134e-05,
      "loss": 0.3802,
      "step": 1866
    },
    {
      "epoch": 169.77575757575758,
      "grad_norm": 0.2756951153278351,
      "learning_rate": 3.133000000000001e-05,
      "loss": 0.432,
      "step": 1867
    },
    {
      "epoch": 169.87272727272727,
      "grad_norm": 0.2762760519981384,
      "learning_rate": 3.132e-05,
      "loss": 0.3936,
      "step": 1868
    },
    {
      "epoch": 169.96969696969697,
      "grad_norm": 0.27015718817710876,
      "learning_rate": 3.1310000000000003e-05,
      "loss": 0.4193,
      "step": 1869
    },
    {
      "epoch": 170.0,
      "grad_norm": 0.46624138951301575,
      "learning_rate": 3.13e-05,
      "loss": 0.2868,
      "step": 1870
    },
    {
      "epoch": 170.0,
      "eval_loss": 0.4615083634853363,
      "eval_runtime": 2.1721,
      "eval_samples_per_second": 25.321,
      "eval_steps_per_second": 3.223,
      "step": 1870
    },
    {
      "epoch": 170.0969696969697,
      "grad_norm": 0.3341410756111145,
      "learning_rate": 3.129e-05,
      "loss": 0.403,
      "step": 1871
    },
    {
      "epoch": 170.1939393939394,
      "grad_norm": 0.30921101570129395,
      "learning_rate": 3.1280000000000005e-05,
      "loss": 0.3938,
      "step": 1872
    },
    {
      "epoch": 170.29090909090908,
      "grad_norm": 0.3262888789176941,
      "learning_rate": 3.127e-05,
      "loss": 0.4082,
      "step": 1873
    },
    {
      "epoch": 170.38787878787878,
      "grad_norm": 0.29888027906417847,
      "learning_rate": 3.126e-05,
      "loss": 0.3694,
      "step": 1874
    },
    {
      "epoch": 170.4848484848485,
      "grad_norm": 0.30654269456863403,
      "learning_rate": 3.125e-05,
      "loss": 0.3891,
      "step": 1875
    },
    {
      "epoch": 170.5818181818182,
      "grad_norm": 0.2998983561992645,
      "learning_rate": 3.1240000000000006e-05,
      "loss": 0.3869,
      "step": 1876
    },
    {
      "epoch": 170.6787878787879,
      "grad_norm": 0.29455775022506714,
      "learning_rate": 3.1230000000000004e-05,
      "loss": 0.3903,
      "step": 1877
    },
    {
      "epoch": 170.77575757575758,
      "grad_norm": 0.3592742681503296,
      "learning_rate": 3.122e-05,
      "loss": 0.4284,
      "step": 1878
    },
    {
      "epoch": 170.87272727272727,
      "grad_norm": 0.3325028121471405,
      "learning_rate": 3.121e-05,
      "loss": 0.4097,
      "step": 1879
    },
    {
      "epoch": 170.96969696969697,
      "grad_norm": 0.29504528641700745,
      "learning_rate": 3.12e-05,
      "loss": 0.3819,
      "step": 1880
    },
    {
      "epoch": 170.96969696969697,
      "eval_loss": 0.4609839618206024,
      "eval_runtime": 2.1984,
      "eval_samples_per_second": 25.018,
      "eval_steps_per_second": 3.184,
      "step": 1880
    },
    {
      "epoch": 171.0,
      "grad_norm": 0.42094412446022034,
      "learning_rate": 3.1190000000000005e-05,
      "loss": 0.4709,
      "step": 1881
    },
    {
      "epoch": 171.0969696969697,
      "grad_norm": 0.26900404691696167,
      "learning_rate": 3.118e-05,
      "loss": 0.3885,
      "step": 1882
    },
    {
      "epoch": 171.1939393939394,
      "grad_norm": 0.36168283224105835,
      "learning_rate": 3.117e-05,
      "loss": 0.3713,
      "step": 1883
    },
    {
      "epoch": 171.29090909090908,
      "grad_norm": 0.28963232040405273,
      "learning_rate": 3.116e-05,
      "loss": 0.3786,
      "step": 1884
    },
    {
      "epoch": 171.38787878787878,
      "grad_norm": 0.30029475688934326,
      "learning_rate": 3.115e-05,
      "loss": 0.3851,
      "step": 1885
    },
    {
      "epoch": 171.4848484848485,
      "grad_norm": 0.2340232878923416,
      "learning_rate": 3.1140000000000003e-05,
      "loss": 0.3724,
      "step": 1886
    },
    {
      "epoch": 171.5818181818182,
      "grad_norm": 0.3094693720340729,
      "learning_rate": 3.113e-05,
      "loss": 0.4172,
      "step": 1887
    },
    {
      "epoch": 171.6787878787879,
      "grad_norm": 0.32682815194129944,
      "learning_rate": 3.112e-05,
      "loss": 0.4379,
      "step": 1888
    },
    {
      "epoch": 171.77575757575758,
      "grad_norm": 0.39302778244018555,
      "learning_rate": 3.111e-05,
      "loss": 0.4312,
      "step": 1889
    },
    {
      "epoch": 171.87272727272727,
      "grad_norm": 0.3695316016674042,
      "learning_rate": 3.1100000000000004e-05,
      "loss": 0.4128,
      "step": 1890
    },
    {
      "epoch": 171.87272727272727,
      "eval_loss": 0.46061545610427856,
      "eval_runtime": 2.1757,
      "eval_samples_per_second": 25.279,
      "eval_steps_per_second": 3.217,
      "step": 1890
    },
    {
      "epoch": 171.96969696969697,
      "grad_norm": 0.3330514430999756,
      "learning_rate": 3.109e-05,
      "loss": 0.3874,
      "step": 1891
    },
    {
      "epoch": 172.0,
      "grad_norm": 0.37093260884284973,
      "learning_rate": 3.108e-05,
      "loss": 0.3705,
      "step": 1892
    },
    {
      "epoch": 172.0969696969697,
      "grad_norm": 0.23208312690258026,
      "learning_rate": 3.107e-05,
      "loss": 0.3781,
      "step": 1893
    },
    {
      "epoch": 172.1939393939394,
      "grad_norm": 0.2550165355205536,
      "learning_rate": 3.106e-05,
      "loss": 0.3863,
      "step": 1894
    },
    {
      "epoch": 172.29090909090908,
      "grad_norm": 0.35588306188583374,
      "learning_rate": 3.105e-05,
      "loss": 0.416,
      "step": 1895
    },
    {
      "epoch": 172.38787878787878,
      "grad_norm": 0.26319488883018494,
      "learning_rate": 3.104e-05,
      "loss": 0.3926,
      "step": 1896
    },
    {
      "epoch": 172.4848484848485,
      "grad_norm": 0.23338590562343597,
      "learning_rate": 3.1030000000000006e-05,
      "loss": 0.3879,
      "step": 1897
    },
    {
      "epoch": 172.5818181818182,
      "grad_norm": 0.3668774366378784,
      "learning_rate": 3.102e-05,
      "loss": 0.3602,
      "step": 1898
    },
    {
      "epoch": 172.6787878787879,
      "grad_norm": 0.29479148983955383,
      "learning_rate": 3.101e-05,
      "loss": 0.3748,
      "step": 1899
    },
    {
      "epoch": 172.77575757575758,
      "grad_norm": 0.291279673576355,
      "learning_rate": 3.1e-05,
      "loss": 0.4041,
      "step": 1900
    },
    {
      "epoch": 172.77575757575758,
      "eval_loss": 0.46023276448249817,
      "eval_runtime": 2.1674,
      "eval_samples_per_second": 25.375,
      "eval_steps_per_second": 3.23,
      "step": 1900
    },
    {
      "epoch": 172.87272727272727,
      "grad_norm": 0.2789687514305115,
      "learning_rate": 3.099e-05,
      "loss": 0.4038,
      "step": 1901
    },
    {
      "epoch": 172.96969696969697,
      "grad_norm": 0.283159077167511,
      "learning_rate": 3.0980000000000005e-05,
      "loss": 0.4683,
      "step": 1902
    },
    {
      "epoch": 173.0,
      "grad_norm": 0.5169331431388855,
      "learning_rate": 3.0969999999999997e-05,
      "loss": 0.3729,
      "step": 1903
    },
    {
      "epoch": 173.0969696969697,
      "grad_norm": 0.30403706431388855,
      "learning_rate": 3.096e-05,
      "loss": 0.4029,
      "step": 1904
    },
    {
      "epoch": 173.1939393939394,
      "grad_norm": 0.2558678686618805,
      "learning_rate": 3.095e-05,
      "loss": 0.4093,
      "step": 1905
    },
    {
      "epoch": 173.29090909090908,
      "grad_norm": 0.34730300307273865,
      "learning_rate": 3.0940000000000005e-05,
      "loss": 0.3644,
      "step": 1906
    },
    {
      "epoch": 173.38787878787878,
      "grad_norm": 0.267134428024292,
      "learning_rate": 3.0930000000000004e-05,
      "loss": 0.4018,
      "step": 1907
    },
    {
      "epoch": 173.4848484848485,
      "grad_norm": 0.22669728100299835,
      "learning_rate": 3.092e-05,
      "loss": 0.3635,
      "step": 1908
    },
    {
      "epoch": 173.5818181818182,
      "grad_norm": 0.3618592619895935,
      "learning_rate": 3.091e-05,
      "loss": 0.3999,
      "step": 1909
    },
    {
      "epoch": 173.6787878787879,
      "grad_norm": 0.3263871669769287,
      "learning_rate": 3.09e-05,
      "loss": 0.4313,
      "step": 1910
    },
    {
      "epoch": 173.6787878787879,
      "eval_loss": 0.4603014588356018,
      "eval_runtime": 2.1594,
      "eval_samples_per_second": 25.471,
      "eval_steps_per_second": 3.242,
      "step": 1910
    },
    {
      "epoch": 173.77575757575758,
      "grad_norm": 0.3253178894519806,
      "learning_rate": 3.0890000000000004e-05,
      "loss": 0.4348,
      "step": 1911
    },
    {
      "epoch": 173.87272727272727,
      "grad_norm": 0.32916221022605896,
      "learning_rate": 3.088e-05,
      "loss": 0.3839,
      "step": 1912
    },
    {
      "epoch": 173.96969696969697,
      "grad_norm": 0.2587529122829437,
      "learning_rate": 3.087e-05,
      "loss": 0.3832,
      "step": 1913
    },
    {
      "epoch": 174.0,
      "grad_norm": 0.45776283740997314,
      "learning_rate": 3.086e-05,
      "loss": 0.3451,
      "step": 1914
    },
    {
      "epoch": 174.0969696969697,
      "grad_norm": 0.2600422203540802,
      "learning_rate": 3.0850000000000004e-05,
      "loss": 0.384,
      "step": 1915
    },
    {
      "epoch": 174.1939393939394,
      "grad_norm": 0.2785820960998535,
      "learning_rate": 3.084e-05,
      "loss": 0.3848,
      "step": 1916
    },
    {
      "epoch": 174.29090909090908,
      "grad_norm": 0.30167168378829956,
      "learning_rate": 3.083e-05,
      "loss": 0.4166,
      "step": 1917
    },
    {
      "epoch": 174.38787878787878,
      "grad_norm": 0.38409626483917236,
      "learning_rate": 3.082e-05,
      "loss": 0.3884,
      "step": 1918
    },
    {
      "epoch": 174.4848484848485,
      "grad_norm": 0.40932610630989075,
      "learning_rate": 3.081e-05,
      "loss": 0.3618,
      "step": 1919
    },
    {
      "epoch": 174.5818181818182,
      "grad_norm": 0.26346346735954285,
      "learning_rate": 3.08e-05,
      "loss": 0.3602,
      "step": 1920
    },
    {
      "epoch": 174.5818181818182,
      "eval_loss": 0.4591878056526184,
      "eval_runtime": 2.152,
      "eval_samples_per_second": 25.558,
      "eval_steps_per_second": 3.253,
      "step": 1920
    },
    {
      "epoch": 174.6787878787879,
      "grad_norm": 0.260904997587204,
      "learning_rate": 3.079e-05,
      "loss": 0.43,
      "step": 1921
    },
    {
      "epoch": 174.77575757575758,
      "grad_norm": 0.3299517333507538,
      "learning_rate": 3.078e-05,
      "loss": 0.4078,
      "step": 1922
    },
    {
      "epoch": 174.87272727272727,
      "grad_norm": 0.38013601303100586,
      "learning_rate": 3.077e-05,
      "loss": 0.422,
      "step": 1923
    },
    {
      "epoch": 174.96969696969697,
      "grad_norm": 0.27372559905052185,
      "learning_rate": 3.076e-05,
      "loss": 0.3849,
      "step": 1924
    },
    {
      "epoch": 175.0,
      "grad_norm": 0.590491533279419,
      "learning_rate": 3.075e-05,
      "loss": 0.4239,
      "step": 1925
    },
    {
      "epoch": 175.0969696969697,
      "grad_norm": 0.25433579087257385,
      "learning_rate": 3.074e-05,
      "loss": 0.404,
      "step": 1926
    },
    {
      "epoch": 175.1939393939394,
      "grad_norm": 0.2625212073326111,
      "learning_rate": 3.0730000000000006e-05,
      "loss": 0.4086,
      "step": 1927
    },
    {
      "epoch": 175.29090909090908,
      "grad_norm": 0.24784494936466217,
      "learning_rate": 3.072e-05,
      "loss": 0.3628,
      "step": 1928
    },
    {
      "epoch": 175.38787878787878,
      "grad_norm": 0.2836151719093323,
      "learning_rate": 3.071e-05,
      "loss": 0.4296,
      "step": 1929
    },
    {
      "epoch": 175.4848484848485,
      "grad_norm": 0.376642644405365,
      "learning_rate": 3.07e-05,
      "loss": 0.4384,
      "step": 1930
    },
    {
      "epoch": 175.4848484848485,
      "eval_loss": 0.4586239159107208,
      "eval_runtime": 2.1663,
      "eval_samples_per_second": 25.389,
      "eval_steps_per_second": 3.231,
      "step": 1930
    },
    {
      "epoch": 175.5818181818182,
      "grad_norm": 0.2923555374145508,
      "learning_rate": 3.069e-05,
      "loss": 0.3755,
      "step": 1931
    },
    {
      "epoch": 175.6787878787879,
      "grad_norm": 0.33017534017562866,
      "learning_rate": 3.0680000000000004e-05,
      "loss": 0.3765,
      "step": 1932
    },
    {
      "epoch": 175.77575757575758,
      "grad_norm": 0.2759827673435211,
      "learning_rate": 3.0669999999999996e-05,
      "loss": 0.3779,
      "step": 1933
    },
    {
      "epoch": 175.87272727272727,
      "grad_norm": 0.25442424416542053,
      "learning_rate": 3.066e-05,
      "loss": 0.4056,
      "step": 1934
    },
    {
      "epoch": 175.96969696969697,
      "grad_norm": 0.3058272898197174,
      "learning_rate": 3.065e-05,
      "loss": 0.382,
      "step": 1935
    },
    {
      "epoch": 176.0,
      "grad_norm": 0.5333647727966309,
      "learning_rate": 3.0640000000000005e-05,
      "loss": 0.3437,
      "step": 1936
    },
    {
      "epoch": 176.0969696969697,
      "grad_norm": 0.2898612320423126,
      "learning_rate": 3.063e-05,
      "loss": 0.3814,
      "step": 1937
    },
    {
      "epoch": 176.1939393939394,
      "grad_norm": 0.2499009072780609,
      "learning_rate": 3.062e-05,
      "loss": 0.4063,
      "step": 1938
    },
    {
      "epoch": 176.29090909090908,
      "grad_norm": 0.30573129653930664,
      "learning_rate": 3.061e-05,
      "loss": 0.4219,
      "step": 1939
    },
    {
      "epoch": 176.38787878787878,
      "grad_norm": 0.31919580698013306,
      "learning_rate": 3.06e-05,
      "loss": 0.4216,
      "step": 1940
    },
    {
      "epoch": 176.38787878787878,
      "eval_loss": 0.45822402834892273,
      "eval_runtime": 2.1441,
      "eval_samples_per_second": 25.651,
      "eval_steps_per_second": 3.265,
      "step": 1940
    },
    {
      "epoch": 176.4848484848485,
      "grad_norm": 0.2452893853187561,
      "learning_rate": 3.0590000000000004e-05,
      "loss": 0.3846,
      "step": 1941
    },
    {
      "epoch": 176.5818181818182,
      "grad_norm": 0.23256346583366394,
      "learning_rate": 3.058e-05,
      "loss": 0.3503,
      "step": 1942
    },
    {
      "epoch": 176.6787878787879,
      "grad_norm": 0.298978716135025,
      "learning_rate": 3.057000000000001e-05,
      "loss": 0.3615,
      "step": 1943
    },
    {
      "epoch": 176.77575757575758,
      "grad_norm": 0.29125311970710754,
      "learning_rate": 3.056e-05,
      "loss": 0.391,
      "step": 1944
    },
    {
      "epoch": 176.87272727272727,
      "grad_norm": 0.3193817436695099,
      "learning_rate": 3.0550000000000004e-05,
      "loss": 0.4168,
      "step": 1945
    },
    {
      "epoch": 176.96969696969697,
      "grad_norm": 0.2996191382408142,
      "learning_rate": 3.054e-05,
      "loss": 0.3894,
      "step": 1946
    },
    {
      "epoch": 177.0,
      "grad_norm": 0.48652327060699463,
      "learning_rate": 3.053e-05,
      "loss": 0.4369,
      "step": 1947
    },
    {
      "epoch": 177.0969696969697,
      "grad_norm": 0.2905808985233307,
      "learning_rate": 3.0520000000000006e-05,
      "loss": 0.3961,
      "step": 1948
    },
    {
      "epoch": 177.1939393939394,
      "grad_norm": 0.3679002821445465,
      "learning_rate": 3.051e-05,
      "loss": 0.4492,
      "step": 1949
    },
    {
      "epoch": 177.29090909090908,
      "grad_norm": 0.2981612980365753,
      "learning_rate": 3.05e-05,
      "loss": 0.393,
      "step": 1950
    },
    {
      "epoch": 177.29090909090908,
      "eval_loss": 0.4579584300518036,
      "eval_runtime": 2.1224,
      "eval_samples_per_second": 25.914,
      "eval_steps_per_second": 3.298,
      "step": 1950
    },
    {
      "epoch": 177.38787878787878,
      "grad_norm": 0.3025275766849518,
      "learning_rate": 3.049e-05,
      "loss": 0.4055,
      "step": 1951
    },
    {
      "epoch": 177.4848484848485,
      "grad_norm": 0.26585206389427185,
      "learning_rate": 3.0480000000000003e-05,
      "loss": 0.3805,
      "step": 1952
    },
    {
      "epoch": 177.5818181818182,
      "grad_norm": 0.2868068516254425,
      "learning_rate": 3.0470000000000005e-05,
      "loss": 0.3709,
      "step": 1953
    },
    {
      "epoch": 177.6787878787879,
      "grad_norm": 0.28243038058280945,
      "learning_rate": 3.046e-05,
      "loss": 0.3904,
      "step": 1954
    },
    {
      "epoch": 177.77575757575758,
      "grad_norm": 0.32643869519233704,
      "learning_rate": 3.045e-05,
      "loss": 0.3602,
      "step": 1955
    },
    {
      "epoch": 177.87272727272727,
      "grad_norm": 0.36092421412467957,
      "learning_rate": 3.0440000000000003e-05,
      "loss": 0.3731,
      "step": 1956
    },
    {
      "epoch": 177.96969696969697,
      "grad_norm": 0.25088557600975037,
      "learning_rate": 3.0430000000000002e-05,
      "loss": 0.4032,
      "step": 1957
    },
    {
      "epoch": 178.0,
      "grad_norm": 0.4627428948879242,
      "learning_rate": 3.0420000000000004e-05,
      "loss": 0.4199,
      "step": 1958
    },
    {
      "epoch": 178.0969696969697,
      "grad_norm": 0.24005559086799622,
      "learning_rate": 3.041e-05,
      "loss": 0.3832,
      "step": 1959
    },
    {
      "epoch": 178.1939393939394,
      "grad_norm": 0.3331547677516937,
      "learning_rate": 3.04e-05,
      "loss": 0.4401,
      "step": 1960
    },
    {
      "epoch": 178.1939393939394,
      "eval_loss": 0.45817330479621887,
      "eval_runtime": 2.1406,
      "eval_samples_per_second": 25.693,
      "eval_steps_per_second": 3.27,
      "step": 1960
    },
    {
      "epoch": 178.29090909090908,
      "grad_norm": 0.3940174877643585,
      "learning_rate": 3.0390000000000002e-05,
      "loss": 0.3819,
      "step": 1961
    },
    {
      "epoch": 178.38787878787878,
      "grad_norm": 0.29317134618759155,
      "learning_rate": 3.0380000000000004e-05,
      "loss": 0.3818,
      "step": 1962
    },
    {
      "epoch": 178.4848484848485,
      "grad_norm": 0.2608102560043335,
      "learning_rate": 3.0370000000000006e-05,
      "loss": 0.4172,
      "step": 1963
    },
    {
      "epoch": 178.5818181818182,
      "grad_norm": 0.2783978283405304,
      "learning_rate": 3.036e-05,
      "loss": 0.3945,
      "step": 1964
    },
    {
      "epoch": 178.6787878787879,
      "grad_norm": 0.2786385715007782,
      "learning_rate": 3.035e-05,
      "loss": 0.4058,
      "step": 1965
    },
    {
      "epoch": 178.77575757575758,
      "grad_norm": 0.3624214828014374,
      "learning_rate": 3.034e-05,
      "loss": 0.38,
      "step": 1966
    },
    {
      "epoch": 178.87272727272727,
      "grad_norm": 0.23733411729335785,
      "learning_rate": 3.0330000000000003e-05,
      "loss": 0.3582,
      "step": 1967
    },
    {
      "epoch": 178.96969696969697,
      "grad_norm": 0.3076571822166443,
      "learning_rate": 3.0320000000000004e-05,
      "loss": 0.3797,
      "step": 1968
    },
    {
      "epoch": 179.0,
      "grad_norm": 0.47409960627555847,
      "learning_rate": 3.031e-05,
      "loss": 0.3919,
      "step": 1969
    },
    {
      "epoch": 179.0969696969697,
      "grad_norm": 0.2679012715816498,
      "learning_rate": 3.03e-05,
      "loss": 0.3471,
      "step": 1970
    },
    {
      "epoch": 179.0969696969697,
      "eval_loss": 0.4579102694988251,
      "eval_runtime": 2.151,
      "eval_samples_per_second": 25.57,
      "eval_steps_per_second": 3.254,
      "step": 1970
    },
    {
      "epoch": 179.1939393939394,
      "grad_norm": 0.37176313996315,
      "learning_rate": 3.0290000000000003e-05,
      "loss": 0.3834,
      "step": 1971
    },
    {
      "epoch": 179.29090909090908,
      "grad_norm": 0.36216312646865845,
      "learning_rate": 3.028e-05,
      "loss": 0.4412,
      "step": 1972
    },
    {
      "epoch": 179.38787878787878,
      "grad_norm": 0.36082252860069275,
      "learning_rate": 3.0270000000000003e-05,
      "loss": 0.381,
      "step": 1973
    },
    {
      "epoch": 179.4848484848485,
      "grad_norm": 0.3448978662490845,
      "learning_rate": 3.0259999999999998e-05,
      "loss": 0.3909,
      "step": 1974
    },
    {
      "epoch": 179.5818181818182,
      "grad_norm": 0.2553895115852356,
      "learning_rate": 3.025e-05,
      "loss": 0.3737,
      "step": 1975
    },
    {
      "epoch": 179.6787878787879,
      "grad_norm": 0.2908379137516022,
      "learning_rate": 3.0240000000000002e-05,
      "loss": 0.4023,
      "step": 1976
    },
    {
      "epoch": 179.77575757575758,
      "grad_norm": 0.2847512662410736,
      "learning_rate": 3.0230000000000004e-05,
      "loss": 0.3828,
      "step": 1977
    },
    {
      "epoch": 179.87272727272727,
      "grad_norm": 0.31389573216438293,
      "learning_rate": 3.0220000000000005e-05,
      "loss": 0.4101,
      "step": 1978
    },
    {
      "epoch": 179.96969696969697,
      "grad_norm": 0.27856436371803284,
      "learning_rate": 3.021e-05,
      "loss": 0.3984,
      "step": 1979
    },
    {
      "epoch": 180.0,
      "grad_norm": 0.6234988570213318,
      "learning_rate": 3.02e-05,
      "loss": 0.411,
      "step": 1980
    },
    {
      "epoch": 180.0,
      "eval_loss": 0.45648694038391113,
      "eval_runtime": 2.1436,
      "eval_samples_per_second": 25.657,
      "eval_steps_per_second": 3.265,
      "step": 1980
    },
    {
      "epoch": 180.0969696969697,
      "grad_norm": 0.2653573751449585,
      "learning_rate": 3.019e-05,
      "loss": 0.3949,
      "step": 1981
    },
    {
      "epoch": 180.1939393939394,
      "grad_norm": 0.38016894459724426,
      "learning_rate": 3.0180000000000002e-05,
      "loss": 0.3905,
      "step": 1982
    },
    {
      "epoch": 180.29090909090908,
      "grad_norm": 0.35457536578178406,
      "learning_rate": 3.0170000000000004e-05,
      "loss": 0.4022,
      "step": 1983
    },
    {
      "epoch": 180.38787878787878,
      "grad_norm": 0.2664441764354706,
      "learning_rate": 3.016e-05,
      "loss": 0.3634,
      "step": 1984
    },
    {
      "epoch": 180.4848484848485,
      "grad_norm": 0.37940555810928345,
      "learning_rate": 3.015e-05,
      "loss": 0.3936,
      "step": 1985
    },
    {
      "epoch": 180.5818181818182,
      "grad_norm": 0.27955812215805054,
      "learning_rate": 3.0140000000000003e-05,
      "loss": 0.3521,
      "step": 1986
    },
    {
      "epoch": 180.6787878787879,
      "grad_norm": 0.3075166344642639,
      "learning_rate": 3.013e-05,
      "loss": 0.3979,
      "step": 1987
    },
    {
      "epoch": 180.77575757575758,
      "grad_norm": 0.4995088279247284,
      "learning_rate": 3.0120000000000003e-05,
      "loss": 0.4299,
      "step": 1988
    },
    {
      "epoch": 180.87272727272727,
      "grad_norm": 0.30396339297294617,
      "learning_rate": 3.0109999999999998e-05,
      "loss": 0.4124,
      "step": 1989
    },
    {
      "epoch": 180.96969696969697,
      "grad_norm": 0.26562339067459106,
      "learning_rate": 3.01e-05,
      "loss": 0.3467,
      "step": 1990
    },
    {
      "epoch": 180.96969696969697,
      "eval_loss": 0.4557584822177887,
      "eval_runtime": 2.1466,
      "eval_samples_per_second": 25.622,
      "eval_steps_per_second": 3.261,
      "step": 1990
    },
    {
      "epoch": 181.0,
      "grad_norm": 0.4487966001033783,
      "learning_rate": 3.009e-05,
      "loss": 0.4811,
      "step": 1991
    },
    {
      "epoch": 181.0969696969697,
      "grad_norm": 0.3044120967388153,
      "learning_rate": 3.0080000000000003e-05,
      "loss": 0.386,
      "step": 1992
    },
    {
      "epoch": 181.1939393939394,
      "grad_norm": 0.3281792104244232,
      "learning_rate": 3.0070000000000005e-05,
      "loss": 0.409,
      "step": 1993
    },
    {
      "epoch": 181.29090909090908,
      "grad_norm": 0.33631032705307007,
      "learning_rate": 3.006e-05,
      "loss": 0.3704,
      "step": 1994
    },
    {
      "epoch": 181.38787878787878,
      "grad_norm": 0.2445061206817627,
      "learning_rate": 3.0050000000000002e-05,
      "loss": 0.3865,
      "step": 1995
    },
    {
      "epoch": 181.4848484848485,
      "grad_norm": 0.35595229268074036,
      "learning_rate": 3.004e-05,
      "loss": 0.3441,
      "step": 1996
    },
    {
      "epoch": 181.5818181818182,
      "grad_norm": 0.26783713698387146,
      "learning_rate": 3.0030000000000002e-05,
      "loss": 0.3794,
      "step": 1997
    },
    {
      "epoch": 181.6787878787879,
      "grad_norm": 0.2763480544090271,
      "learning_rate": 3.0020000000000004e-05,
      "loss": 0.3984,
      "step": 1998
    },
    {
      "epoch": 181.77575757575758,
      "grad_norm": 0.28086158633232117,
      "learning_rate": 3.001e-05,
      "loss": 0.406,
      "step": 1999
    },
    {
      "epoch": 181.87272727272727,
      "grad_norm": 0.24243846535682678,
      "learning_rate": 3e-05,
      "loss": 0.3838,
      "step": 2000
    },
    {
      "epoch": 181.87272727272727,
      "eval_loss": 0.45567676424980164,
      "eval_runtime": 2.1323,
      "eval_samples_per_second": 25.794,
      "eval_steps_per_second": 3.283,
      "step": 2000
    },
    {
      "epoch": 181.96969696969697,
      "grad_norm": 0.37673303484916687,
      "learning_rate": 2.9990000000000003e-05,
      "loss": 0.4429,
      "step": 2001
    },
    {
      "epoch": 182.0,
      "grad_norm": 0.39634788036346436,
      "learning_rate": 2.998e-05,
      "loss": 0.3811,
      "step": 2002
    },
    {
      "epoch": 182.0969696969697,
      "grad_norm": 0.2639450430870056,
      "learning_rate": 2.9970000000000003e-05,
      "loss": 0.3787,
      "step": 2003
    },
    {
      "epoch": 182.1939393939394,
      "grad_norm": 0.2618735730648041,
      "learning_rate": 2.9959999999999998e-05,
      "loss": 0.4243,
      "step": 2004
    },
    {
      "epoch": 182.29090909090908,
      "grad_norm": 0.2818409502506256,
      "learning_rate": 2.995e-05,
      "loss": 0.3799,
      "step": 2005
    },
    {
      "epoch": 182.38787878787878,
      "grad_norm": 0.2500140070915222,
      "learning_rate": 2.994e-05,
      "loss": 0.3458,
      "step": 2006
    },
    {
      "epoch": 182.4848484848485,
      "grad_norm": 0.28353533148765564,
      "learning_rate": 2.9930000000000003e-05,
      "loss": 0.4154,
      "step": 2007
    },
    {
      "epoch": 182.5818181818182,
      "grad_norm": 0.26767697930336,
      "learning_rate": 2.9920000000000005e-05,
      "loss": 0.4024,
      "step": 2008
    },
    {
      "epoch": 182.6787878787879,
      "grad_norm": 0.26572415232658386,
      "learning_rate": 2.991e-05,
      "loss": 0.3788,
      "step": 2009
    },
    {
      "epoch": 182.77575757575758,
      "grad_norm": 0.3166835308074951,
      "learning_rate": 2.9900000000000002e-05,
      "loss": 0.4086,
      "step": 2010
    },
    {
      "epoch": 182.77575757575758,
      "eval_loss": 0.45494434237480164,
      "eval_runtime": 2.1299,
      "eval_samples_per_second": 25.823,
      "eval_steps_per_second": 3.287,
      "step": 2010
    },
    {
      "epoch": 182.87272727272727,
      "grad_norm": 0.3935854434967041,
      "learning_rate": 2.989e-05,
      "loss": 0.3866,
      "step": 2011
    },
    {
      "epoch": 182.96969696969697,
      "grad_norm": 0.2760485112667084,
      "learning_rate": 2.9880000000000002e-05,
      "loss": 0.3652,
      "step": 2012
    },
    {
      "epoch": 183.0,
      "grad_norm": 0.4534105360507965,
      "learning_rate": 2.9870000000000004e-05,
      "loss": 0.4278,
      "step": 2013
    },
    {
      "epoch": 183.0969696969697,
      "grad_norm": 0.2943841814994812,
      "learning_rate": 2.986e-05,
      "loss": 0.375,
      "step": 2014
    },
    {
      "epoch": 183.1939393939394,
      "grad_norm": 0.3969006836414337,
      "learning_rate": 2.985e-05,
      "loss": 0.3654,
      "step": 2015
    },
    {
      "epoch": 183.29090909090908,
      "grad_norm": 0.3221188187599182,
      "learning_rate": 2.9840000000000002e-05,
      "loss": 0.3803,
      "step": 2016
    },
    {
      "epoch": 183.38787878787878,
      "grad_norm": 0.2897550165653229,
      "learning_rate": 2.9830000000000004e-05,
      "loss": 0.4101,
      "step": 2017
    },
    {
      "epoch": 183.4848484848485,
      "grad_norm": 0.256380170583725,
      "learning_rate": 2.9820000000000002e-05,
      "loss": 0.4126,
      "step": 2018
    },
    {
      "epoch": 183.5818181818182,
      "grad_norm": 0.28354132175445557,
      "learning_rate": 2.9809999999999997e-05,
      "loss": 0.3683,
      "step": 2019
    },
    {
      "epoch": 183.6787878787879,
      "grad_norm": 0.2842443287372589,
      "learning_rate": 2.98e-05,
      "loss": 0.405,
      "step": 2020
    },
    {
      "epoch": 183.6787878787879,
      "eval_loss": 0.454367995262146,
      "eval_runtime": 2.1417,
      "eval_samples_per_second": 25.68,
      "eval_steps_per_second": 3.268,
      "step": 2020
    },
    {
      "epoch": 183.77575757575758,
      "grad_norm": 0.29224082827568054,
      "learning_rate": 2.979e-05,
      "loss": 0.4091,
      "step": 2021
    },
    {
      "epoch": 183.87272727272727,
      "grad_norm": 0.2452160269021988,
      "learning_rate": 2.9780000000000003e-05,
      "loss": 0.3958,
      "step": 2022
    },
    {
      "epoch": 183.96969696969697,
      "grad_norm": 0.2455267459154129,
      "learning_rate": 2.9770000000000005e-05,
      "loss": 0.3598,
      "step": 2023
    },
    {
      "epoch": 184.0,
      "grad_norm": 0.5353454351425171,
      "learning_rate": 2.976e-05,
      "loss": 0.4148,
      "step": 2024
    },
    {
      "epoch": 184.0969696969697,
      "grad_norm": 0.23789070546627045,
      "learning_rate": 2.975e-05,
      "loss": 0.3749,
      "step": 2025
    },
    {
      "epoch": 184.1939393939394,
      "grad_norm": 0.27197128534317017,
      "learning_rate": 2.974e-05,
      "loss": 0.3735,
      "step": 2026
    },
    {
      "epoch": 184.29090909090908,
      "grad_norm": 0.32617834210395813,
      "learning_rate": 2.973e-05,
      "loss": 0.4064,
      "step": 2027
    },
    {
      "epoch": 184.38787878787878,
      "grad_norm": 0.33649569749832153,
      "learning_rate": 2.9720000000000003e-05,
      "loss": 0.3779,
      "step": 2028
    },
    {
      "epoch": 184.4848484848485,
      "grad_norm": 0.3602868914604187,
      "learning_rate": 2.971e-05,
      "loss": 0.3829,
      "step": 2029
    },
    {
      "epoch": 184.5818181818182,
      "grad_norm": 0.2936662435531616,
      "learning_rate": 2.97e-05,
      "loss": 0.3844,
      "step": 2030
    },
    {
      "epoch": 184.5818181818182,
      "eval_loss": 0.45421287417411804,
      "eval_runtime": 2.1497,
      "eval_samples_per_second": 25.584,
      "eval_steps_per_second": 3.256,
      "step": 2030
    },
    {
      "epoch": 184.6787878787879,
      "grad_norm": 0.2704029977321625,
      "learning_rate": 2.9690000000000002e-05,
      "loss": 0.3553,
      "step": 2031
    },
    {
      "epoch": 184.77575757575758,
      "grad_norm": 0.2969801723957062,
      "learning_rate": 2.9680000000000004e-05,
      "loss": 0.4152,
      "step": 2032
    },
    {
      "epoch": 184.87272727272727,
      "grad_norm": 0.36191806197166443,
      "learning_rate": 2.9670000000000002e-05,
      "loss": 0.4175,
      "step": 2033
    },
    {
      "epoch": 184.96969696969697,
      "grad_norm": 0.32152867317199707,
      "learning_rate": 2.9659999999999997e-05,
      "loss": 0.387,
      "step": 2034
    },
    {
      "epoch": 185.0,
      "grad_norm": 0.42188048362731934,
      "learning_rate": 2.965e-05,
      "loss": 0.4182,
      "step": 2035
    },
    {
      "epoch": 185.0969696969697,
      "grad_norm": 0.28894734382629395,
      "learning_rate": 2.964e-05,
      "loss": 0.4227,
      "step": 2036
    },
    {
      "epoch": 185.1939393939394,
      "grad_norm": 0.3921942412853241,
      "learning_rate": 2.9630000000000003e-05,
      "loss": 0.3479,
      "step": 2037
    },
    {
      "epoch": 185.29090909090908,
      "grad_norm": 0.31626754999160767,
      "learning_rate": 2.9620000000000004e-05,
      "loss": 0.3833,
      "step": 2038
    },
    {
      "epoch": 185.38787878787878,
      "grad_norm": 0.2705192565917969,
      "learning_rate": 2.961e-05,
      "loss": 0.3865,
      "step": 2039
    },
    {
      "epoch": 185.4848484848485,
      "grad_norm": 0.2821842133998871,
      "learning_rate": 2.96e-05,
      "loss": 0.3941,
      "step": 2040
    },
    {
      "epoch": 185.4848484848485,
      "eval_loss": 0.45374125242233276,
      "eval_runtime": 2.1472,
      "eval_samples_per_second": 25.615,
      "eval_steps_per_second": 3.26,
      "step": 2040
    },
    {
      "epoch": 185.5818181818182,
      "grad_norm": 0.26259809732437134,
      "learning_rate": 2.959e-05,
      "loss": 0.4028,
      "step": 2041
    },
    {
      "epoch": 185.6787878787879,
      "grad_norm": 0.3034418225288391,
      "learning_rate": 2.958e-05,
      "loss": 0.3733,
      "step": 2042
    },
    {
      "epoch": 185.77575757575758,
      "grad_norm": 0.2680684030056,
      "learning_rate": 2.9570000000000003e-05,
      "loss": 0.382,
      "step": 2043
    },
    {
      "epoch": 185.87272727272727,
      "grad_norm": 0.224136620759964,
      "learning_rate": 2.9559999999999998e-05,
      "loss": 0.348,
      "step": 2044
    },
    {
      "epoch": 185.96969696969697,
      "grad_norm": 0.28951144218444824,
      "learning_rate": 2.955e-05,
      "loss": 0.4173,
      "step": 2045
    },
    {
      "epoch": 186.0,
      "grad_norm": 0.6578705906867981,
      "learning_rate": 2.9540000000000002e-05,
      "loss": 0.4462,
      "step": 2046
    },
    {
      "epoch": 186.0969696969697,
      "grad_norm": 0.37858498096466064,
      "learning_rate": 2.9530000000000004e-05,
      "loss": 0.3548,
      "step": 2047
    },
    {
      "epoch": 186.1939393939394,
      "grad_norm": 0.25204986333847046,
      "learning_rate": 2.9520000000000002e-05,
      "loss": 0.4039,
      "step": 2048
    },
    {
      "epoch": 186.29090909090908,
      "grad_norm": 0.36383822560310364,
      "learning_rate": 2.951e-05,
      "loss": 0.3992,
      "step": 2049
    },
    {
      "epoch": 186.38787878787878,
      "grad_norm": 0.2452254593372345,
      "learning_rate": 2.95e-05,
      "loss": 0.409,
      "step": 2050
    },
    {
      "epoch": 186.38787878787878,
      "eval_loss": 0.45269790291786194,
      "eval_runtime": 2.1403,
      "eval_samples_per_second": 25.698,
      "eval_steps_per_second": 3.271,
      "step": 2050
    },
    {
      "epoch": 186.4848484848485,
      "grad_norm": 0.24856916069984436,
      "learning_rate": 2.949e-05,
      "loss": 0.3842,
      "step": 2051
    },
    {
      "epoch": 186.5818181818182,
      "grad_norm": 0.27388259768486023,
      "learning_rate": 2.9480000000000002e-05,
      "loss": 0.3932,
      "step": 2052
    },
    {
      "epoch": 186.6787878787879,
      "grad_norm": 0.24813078343868256,
      "learning_rate": 2.9470000000000004e-05,
      "loss": 0.3902,
      "step": 2053
    },
    {
      "epoch": 186.77575757575758,
      "grad_norm": 0.2655707895755768,
      "learning_rate": 2.946e-05,
      "loss": 0.3691,
      "step": 2054
    },
    {
      "epoch": 186.87272727272727,
      "grad_norm": 0.3013978600502014,
      "learning_rate": 2.945e-05,
      "loss": 0.3864,
      "step": 2055
    },
    {
      "epoch": 186.96969696969697,
      "grad_norm": 0.32454273104667664,
      "learning_rate": 2.944e-05,
      "loss": 0.3617,
      "step": 2056
    },
    {
      "epoch": 187.0,
      "grad_norm": 0.5444182753562927,
      "learning_rate": 2.943e-05,
      "loss": 0.4437,
      "step": 2057
    },
    {
      "epoch": 187.0969696969697,
      "grad_norm": 0.2726549804210663,
      "learning_rate": 2.9420000000000003e-05,
      "loss": 0.4053,
      "step": 2058
    },
    {
      "epoch": 187.1939393939394,
      "grad_norm": 0.2990370988845825,
      "learning_rate": 2.9409999999999998e-05,
      "loss": 0.3343,
      "step": 2059
    },
    {
      "epoch": 187.29090909090908,
      "grad_norm": 0.33290451765060425,
      "learning_rate": 2.94e-05,
      "loss": 0.4081,
      "step": 2060
    },
    {
      "epoch": 187.29090909090908,
      "eval_loss": 0.4533584713935852,
      "eval_runtime": 2.1628,
      "eval_samples_per_second": 25.43,
      "eval_steps_per_second": 3.237,
      "step": 2060
    },
    {
      "epoch": 187.38787878787878,
      "grad_norm": 0.313956081867218,
      "learning_rate": 2.939e-05,
      "loss": 0.3903,
      "step": 2061
    },
    {
      "epoch": 187.4848484848485,
      "grad_norm": 0.282910019159317,
      "learning_rate": 2.9380000000000003e-05,
      "loss": 0.4115,
      "step": 2062
    },
    {
      "epoch": 187.5818181818182,
      "grad_norm": 0.301270067691803,
      "learning_rate": 2.9370000000000002e-05,
      "loss": 0.3726,
      "step": 2063
    },
    {
      "epoch": 187.6787878787879,
      "grad_norm": 0.30559325218200684,
      "learning_rate": 2.9360000000000003e-05,
      "loss": 0.388,
      "step": 2064
    },
    {
      "epoch": 187.77575757575758,
      "grad_norm": 0.31332719326019287,
      "learning_rate": 2.935e-05,
      "loss": 0.3771,
      "step": 2065
    },
    {
      "epoch": 187.87272727272727,
      "grad_norm": 0.31585612893104553,
      "learning_rate": 2.934e-05,
      "loss": 0.3978,
      "step": 2066
    },
    {
      "epoch": 187.96969696969697,
      "grad_norm": 0.30112504959106445,
      "learning_rate": 2.9330000000000002e-05,
      "loss": 0.3806,
      "step": 2067
    },
    {
      "epoch": 188.0,
      "grad_norm": 0.43080729246139526,
      "learning_rate": 2.9320000000000004e-05,
      "loss": 0.3825,
      "step": 2068
    },
    {
      "epoch": 188.0969696969697,
      "grad_norm": 0.29388001561164856,
      "learning_rate": 2.9310000000000006e-05,
      "loss": 0.3544,
      "step": 2069
    },
    {
      "epoch": 188.1939393939394,
      "grad_norm": 0.3585991859436035,
      "learning_rate": 2.93e-05,
      "loss": 0.4224,
      "step": 2070
    },
    {
      "epoch": 188.1939393939394,
      "eval_loss": 0.45311594009399414,
      "eval_runtime": 2.1539,
      "eval_samples_per_second": 25.536,
      "eval_steps_per_second": 3.25,
      "step": 2070
    },
    {
      "epoch": 188.29090909090908,
      "grad_norm": 0.27885720133781433,
      "learning_rate": 2.929e-05,
      "loss": 0.3926,
      "step": 2071
    },
    {
      "epoch": 188.38787878787878,
      "grad_norm": 0.34820637106895447,
      "learning_rate": 2.928e-05,
      "loss": 0.401,
      "step": 2072
    },
    {
      "epoch": 188.4848484848485,
      "grad_norm": 0.3126968741416931,
      "learning_rate": 2.9270000000000003e-05,
      "loss": 0.3889,
      "step": 2073
    },
    {
      "epoch": 188.5818181818182,
      "grad_norm": 0.2805291414260864,
      "learning_rate": 2.9260000000000004e-05,
      "loss": 0.3369,
      "step": 2074
    },
    {
      "epoch": 188.6787878787879,
      "grad_norm": 0.26318931579589844,
      "learning_rate": 2.925e-05,
      "loss": 0.3957,
      "step": 2075
    },
    {
      "epoch": 188.77575757575758,
      "grad_norm": 0.2963278293609619,
      "learning_rate": 2.924e-05,
      "loss": 0.3928,
      "step": 2076
    },
    {
      "epoch": 188.87272727272727,
      "grad_norm": 0.2633761167526245,
      "learning_rate": 2.9230000000000003e-05,
      "loss": 0.3823,
      "step": 2077
    },
    {
      "epoch": 188.96969696969697,
      "grad_norm": 0.37855419516563416,
      "learning_rate": 2.922e-05,
      "loss": 0.3874,
      "step": 2078
    },
    {
      "epoch": 189.0,
      "grad_norm": 0.5898931622505188,
      "learning_rate": 2.9210000000000003e-05,
      "loss": 0.396,
      "step": 2079
    },
    {
      "epoch": 189.0969696969697,
      "grad_norm": 0.3189804255962372,
      "learning_rate": 2.9199999999999998e-05,
      "loss": 0.3979,
      "step": 2080
    },
    {
      "epoch": 189.0969696969697,
      "eval_loss": 0.4517754912376404,
      "eval_runtime": 2.1255,
      "eval_samples_per_second": 25.876,
      "eval_steps_per_second": 3.293,
      "step": 2080
    },
    {
      "epoch": 189.1939393939394,
      "grad_norm": 0.3477194309234619,
      "learning_rate": 2.919e-05,
      "loss": 0.4225,
      "step": 2081
    },
    {
      "epoch": 189.29090909090908,
      "grad_norm": 0.3043428957462311,
      "learning_rate": 2.9180000000000002e-05,
      "loss": 0.3858,
      "step": 2082
    },
    {
      "epoch": 189.38787878787878,
      "grad_norm": 0.30492085218429565,
      "learning_rate": 2.9170000000000004e-05,
      "loss": 0.4159,
      "step": 2083
    },
    {
      "epoch": 189.4848484848485,
      "grad_norm": 0.30474886298179626,
      "learning_rate": 2.9160000000000005e-05,
      "loss": 0.3593,
      "step": 2084
    },
    {
      "epoch": 189.5818181818182,
      "grad_norm": 0.28546661138534546,
      "learning_rate": 2.915e-05,
      "loss": 0.3774,
      "step": 2085
    },
    {
      "epoch": 189.6787878787879,
      "grad_norm": 0.2670576870441437,
      "learning_rate": 2.9140000000000002e-05,
      "loss": 0.361,
      "step": 2086
    },
    {
      "epoch": 189.77575757575758,
      "grad_norm": 0.25876742601394653,
      "learning_rate": 2.913e-05,
      "loss": 0.346,
      "step": 2087
    },
    {
      "epoch": 189.87272727272727,
      "grad_norm": 0.2909092605113983,
      "learning_rate": 2.9120000000000002e-05,
      "loss": 0.4011,
      "step": 2088
    },
    {
      "epoch": 189.96969696969697,
      "grad_norm": 0.3466051518917084,
      "learning_rate": 2.9110000000000004e-05,
      "loss": 0.3702,
      "step": 2089
    },
    {
      "epoch": 190.0,
      "grad_norm": 0.37901443243026733,
      "learning_rate": 2.91e-05,
      "loss": 0.4301,
      "step": 2090
    },
    {
      "epoch": 190.0,
      "eval_loss": 0.4515250325202942,
      "eval_runtime": 2.1422,
      "eval_samples_per_second": 25.674,
      "eval_steps_per_second": 3.268,
      "step": 2090
    },
    {
      "epoch": 190.0969696969697,
      "grad_norm": 0.30174320936203003,
      "learning_rate": 2.909e-05,
      "loss": 0.4002,
      "step": 2091
    },
    {
      "epoch": 190.1939393939394,
      "grad_norm": 0.299001544713974,
      "learning_rate": 2.9080000000000003e-05,
      "loss": 0.4014,
      "step": 2092
    },
    {
      "epoch": 190.29090909090908,
      "grad_norm": 0.2551984488964081,
      "learning_rate": 2.907e-05,
      "loss": 0.3411,
      "step": 2093
    },
    {
      "epoch": 190.38787878787878,
      "grad_norm": 0.28803738951683044,
      "learning_rate": 2.9060000000000003e-05,
      "loss": 0.3405,
      "step": 2094
    },
    {
      "epoch": 190.4848484848485,
      "grad_norm": 0.240384042263031,
      "learning_rate": 2.9049999999999998e-05,
      "loss": 0.3615,
      "step": 2095
    },
    {
      "epoch": 190.5818181818182,
      "grad_norm": 0.27672815322875977,
      "learning_rate": 2.904e-05,
      "loss": 0.3987,
      "step": 2096
    },
    {
      "epoch": 190.6787878787879,
      "grad_norm": 0.29123345017433167,
      "learning_rate": 2.903e-05,
      "loss": 0.3931,
      "step": 2097
    },
    {
      "epoch": 190.77575757575758,
      "grad_norm": 0.30591174960136414,
      "learning_rate": 2.9020000000000003e-05,
      "loss": 0.4157,
      "step": 2098
    },
    {
      "epoch": 190.87272727272727,
      "grad_norm": 0.2984998822212219,
      "learning_rate": 2.9010000000000005e-05,
      "loss": 0.3936,
      "step": 2099
    },
    {
      "epoch": 190.96969696969697,
      "grad_norm": 0.2587094008922577,
      "learning_rate": 2.9e-05,
      "loss": 0.3863,
      "step": 2100
    },
    {
      "epoch": 190.96969696969697,
      "eval_loss": 0.4510568380355835,
      "eval_runtime": 2.1397,
      "eval_samples_per_second": 25.704,
      "eval_steps_per_second": 3.271,
      "step": 2100
    },
    {
      "epoch": 191.0,
      "grad_norm": 0.41683897376060486,
      "learning_rate": 2.8990000000000002e-05,
      "loss": 0.4114,
      "step": 2101
    },
    {
      "epoch": 191.0969696969697,
      "grad_norm": 0.27058327198028564,
      "learning_rate": 2.898e-05,
      "loss": 0.38,
      "step": 2102
    },
    {
      "epoch": 191.1939393939394,
      "grad_norm": 0.24893786013126373,
      "learning_rate": 2.8970000000000002e-05,
      "loss": 0.4132,
      "step": 2103
    },
    {
      "epoch": 191.29090909090908,
      "grad_norm": 0.3115560710430145,
      "learning_rate": 2.8960000000000004e-05,
      "loss": 0.3814,
      "step": 2104
    },
    {
      "epoch": 191.38787878787878,
      "grad_norm": 0.2888801395893097,
      "learning_rate": 2.895e-05,
      "loss": 0.4135,
      "step": 2105
    },
    {
      "epoch": 191.4848484848485,
      "grad_norm": 0.306196391582489,
      "learning_rate": 2.894e-05,
      "loss": 0.3731,
      "step": 2106
    },
    {
      "epoch": 191.5818181818182,
      "grad_norm": 0.2939477860927582,
      "learning_rate": 2.8930000000000003e-05,
      "loss": 0.3631,
      "step": 2107
    },
    {
      "epoch": 191.6787878787879,
      "grad_norm": 0.26490044593811035,
      "learning_rate": 2.8920000000000004e-05,
      "loss": 0.3973,
      "step": 2108
    },
    {
      "epoch": 191.77575757575758,
      "grad_norm": 0.24801471829414368,
      "learning_rate": 2.8910000000000003e-05,
      "loss": 0.3534,
      "step": 2109
    },
    {
      "epoch": 191.87272727272727,
      "grad_norm": 0.29833361506462097,
      "learning_rate": 2.8899999999999998e-05,
      "loss": 0.3721,
      "step": 2110
    },
    {
      "epoch": 191.87272727272727,
      "eval_loss": 0.4510841369628906,
      "eval_runtime": 2.1402,
      "eval_samples_per_second": 25.699,
      "eval_steps_per_second": 3.271,
      "step": 2110
    },
    {
      "epoch": 191.96969696969697,
      "grad_norm": 0.3207988142967224,
      "learning_rate": 2.889e-05,
      "loss": 0.3783,
      "step": 2111
    },
    {
      "epoch": 192.0,
      "grad_norm": 0.46868717670440674,
      "learning_rate": 2.888e-05,
      "loss": 0.4238,
      "step": 2112
    },
    {
      "epoch": 192.0969696969697,
      "grad_norm": 0.24790030717849731,
      "learning_rate": 2.8870000000000003e-05,
      "loss": 0.3819,
      "step": 2113
    },
    {
      "epoch": 192.1939393939394,
      "grad_norm": 0.36050891876220703,
      "learning_rate": 2.8860000000000005e-05,
      "loss": 0.3866,
      "step": 2114
    },
    {
      "epoch": 192.29090909090908,
      "grad_norm": 0.29458919167518616,
      "learning_rate": 2.885e-05,
      "loss": 0.3591,
      "step": 2115
    },
    {
      "epoch": 192.38787878787878,
      "grad_norm": 0.2597523331642151,
      "learning_rate": 2.8840000000000002e-05,
      "loss": 0.3779,
      "step": 2116
    },
    {
      "epoch": 192.4848484848485,
      "grad_norm": 0.26730599999427795,
      "learning_rate": 2.883e-05,
      "loss": 0.4044,
      "step": 2117
    },
    {
      "epoch": 192.5818181818182,
      "grad_norm": 0.3013802766799927,
      "learning_rate": 2.8820000000000002e-05,
      "loss": 0.3506,
      "step": 2118
    },
    {
      "epoch": 192.6787878787879,
      "grad_norm": 0.30980682373046875,
      "learning_rate": 2.8810000000000004e-05,
      "loss": 0.3812,
      "step": 2119
    },
    {
      "epoch": 192.77575757575758,
      "grad_norm": 0.284757137298584,
      "learning_rate": 2.88e-05,
      "loss": 0.3879,
      "step": 2120
    },
    {
      "epoch": 192.77575757575758,
      "eval_loss": 0.45064395666122437,
      "eval_runtime": 2.1303,
      "eval_samples_per_second": 25.817,
      "eval_steps_per_second": 3.286,
      "step": 2120
    },
    {
      "epoch": 192.87272727272727,
      "grad_norm": 0.31305962800979614,
      "learning_rate": 2.879e-05,
      "loss": 0.41,
      "step": 2121
    },
    {
      "epoch": 192.96969696969697,
      "grad_norm": 0.3607388734817505,
      "learning_rate": 2.8780000000000002e-05,
      "loss": 0.3922,
      "step": 2122
    },
    {
      "epoch": 193.0,
      "grad_norm": 0.5440636277198792,
      "learning_rate": 2.8770000000000004e-05,
      "loss": 0.3857,
      "step": 2123
    },
    {
      "epoch": 193.0969696969697,
      "grad_norm": 0.25659671425819397,
      "learning_rate": 2.8760000000000002e-05,
      "loss": 0.3733,
      "step": 2124
    },
    {
      "epoch": 193.1939393939394,
      "grad_norm": 0.2667579650878906,
      "learning_rate": 2.8749999999999997e-05,
      "loss": 0.3897,
      "step": 2125
    },
    {
      "epoch": 193.29090909090908,
      "grad_norm": 0.3180757761001587,
      "learning_rate": 2.874e-05,
      "loss": 0.4107,
      "step": 2126
    },
    {
      "epoch": 193.38787878787878,
      "grad_norm": 0.2646811902523041,
      "learning_rate": 2.873e-05,
      "loss": 0.353,
      "step": 2127
    },
    {
      "epoch": 193.4848484848485,
      "grad_norm": 0.27166447043418884,
      "learning_rate": 2.8720000000000003e-05,
      "loss": 0.3432,
      "step": 2128
    },
    {
      "epoch": 193.5818181818182,
      "grad_norm": 0.3005528151988983,
      "learning_rate": 2.8710000000000005e-05,
      "loss": 0.4227,
      "step": 2129
    },
    {
      "epoch": 193.6787878787879,
      "grad_norm": 0.29820331931114197,
      "learning_rate": 2.87e-05,
      "loss": 0.3951,
      "step": 2130
    },
    {
      "epoch": 193.6787878787879,
      "eval_loss": 0.45010998845100403,
      "eval_runtime": 2.1348,
      "eval_samples_per_second": 25.764,
      "eval_steps_per_second": 3.279,
      "step": 2130
    },
    {
      "epoch": 193.77575757575758,
      "grad_norm": 0.28893670439720154,
      "learning_rate": 2.869e-05,
      "loss": 0.3715,
      "step": 2131
    },
    {
      "epoch": 193.87272727272727,
      "grad_norm": 0.30619752407073975,
      "learning_rate": 2.868e-05,
      "loss": 0.4044,
      "step": 2132
    },
    {
      "epoch": 193.96969696969697,
      "grad_norm": 0.3117333650588989,
      "learning_rate": 2.867e-05,
      "loss": 0.3557,
      "step": 2133
    },
    {
      "epoch": 194.0,
      "grad_norm": 0.5674205422401428,
      "learning_rate": 2.8660000000000003e-05,
      "loss": 0.4006,
      "step": 2134
    },
    {
      "epoch": 194.0969696969697,
      "grad_norm": 0.24678772687911987,
      "learning_rate": 2.865e-05,
      "loss": 0.4015,
      "step": 2135
    },
    {
      "epoch": 194.1939393939394,
      "grad_norm": 0.30944305658340454,
      "learning_rate": 2.864e-05,
      "loss": 0.3721,
      "step": 2136
    },
    {
      "epoch": 194.29090909090908,
      "grad_norm": 0.301979124546051,
      "learning_rate": 2.8630000000000002e-05,
      "loss": 0.3505,
      "step": 2137
    },
    {
      "epoch": 194.38787878787878,
      "grad_norm": 0.23500828444957733,
      "learning_rate": 2.8620000000000004e-05,
      "loss": 0.36,
      "step": 2138
    },
    {
      "epoch": 194.4848484848485,
      "grad_norm": 0.27564796805381775,
      "learning_rate": 2.8610000000000002e-05,
      "loss": 0.3744,
      "step": 2139
    },
    {
      "epoch": 194.5818181818182,
      "grad_norm": 0.28560516238212585,
      "learning_rate": 2.86e-05,
      "loss": 0.3813,
      "step": 2140
    },
    {
      "epoch": 194.5818181818182,
      "eval_loss": 0.44997575879096985,
      "eval_runtime": 2.1293,
      "eval_samples_per_second": 25.83,
      "eval_steps_per_second": 3.287,
      "step": 2140
    },
    {
      "epoch": 194.6787878787879,
      "grad_norm": 0.3202413022518158,
      "learning_rate": 2.859e-05,
      "loss": 0.4007,
      "step": 2141
    },
    {
      "epoch": 194.77575757575758,
      "grad_norm": 0.2470165342092514,
      "learning_rate": 2.858e-05,
      "loss": 0.3691,
      "step": 2142
    },
    {
      "epoch": 194.87272727272727,
      "grad_norm": 0.3099730312824249,
      "learning_rate": 2.8570000000000003e-05,
      "loss": 0.3977,
      "step": 2143
    },
    {
      "epoch": 194.96969696969697,
      "grad_norm": 0.33016228675842285,
      "learning_rate": 2.8560000000000004e-05,
      "loss": 0.409,
      "step": 2144
    },
    {
      "epoch": 195.0,
      "grad_norm": 0.5346285104751587,
      "learning_rate": 2.855e-05,
      "loss": 0.383,
      "step": 2145
    },
    {
      "epoch": 195.0969696969697,
      "grad_norm": 0.3117358088493347,
      "learning_rate": 2.854e-05,
      "loss": 0.4108,
      "step": 2146
    },
    {
      "epoch": 195.1939393939394,
      "grad_norm": 0.3187137246131897,
      "learning_rate": 2.853e-05,
      "loss": 0.3952,
      "step": 2147
    },
    {
      "epoch": 195.29090909090908,
      "grad_norm": 0.25608211755752563,
      "learning_rate": 2.852e-05,
      "loss": 0.3825,
      "step": 2148
    },
    {
      "epoch": 195.38787878787878,
      "grad_norm": 0.2572425901889801,
      "learning_rate": 2.8510000000000003e-05,
      "loss": 0.3897,
      "step": 2149
    },
    {
      "epoch": 195.4848484848485,
      "grad_norm": 0.24927036464214325,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 0.3956,
      "step": 2150
    },
    {
      "epoch": 195.4848484848485,
      "eval_loss": 0.4496622681617737,
      "eval_runtime": 2.156,
      "eval_samples_per_second": 25.51,
      "eval_steps_per_second": 3.247,
      "step": 2150
    },
    {
      "epoch": 195.5818181818182,
      "grad_norm": 0.29352957010269165,
      "learning_rate": 2.849e-05,
      "loss": 0.3445,
      "step": 2151
    },
    {
      "epoch": 195.6787878787879,
      "grad_norm": 0.2819279730319977,
      "learning_rate": 2.8480000000000002e-05,
      "loss": 0.3872,
      "step": 2152
    },
    {
      "epoch": 195.77575757575758,
      "grad_norm": 0.26512467861175537,
      "learning_rate": 2.8470000000000004e-05,
      "loss": 0.3793,
      "step": 2153
    },
    {
      "epoch": 195.87272727272727,
      "grad_norm": 0.29471302032470703,
      "learning_rate": 2.8460000000000002e-05,
      "loss": 0.3611,
      "step": 2154
    },
    {
      "epoch": 195.96969696969697,
      "grad_norm": 0.3353519141674042,
      "learning_rate": 2.845e-05,
      "loss": 0.3636,
      "step": 2155
    },
    {
      "epoch": 196.0,
      "grad_norm": 0.5895360708236694,
      "learning_rate": 2.844e-05,
      "loss": 0.3931,
      "step": 2156
    },
    {
      "epoch": 196.0969696969697,
      "grad_norm": 0.296806275844574,
      "learning_rate": 2.843e-05,
      "loss": 0.3933,
      "step": 2157
    },
    {
      "epoch": 196.1939393939394,
      "grad_norm": 0.27521470189094543,
      "learning_rate": 2.8420000000000002e-05,
      "loss": 0.3828,
      "step": 2158
    },
    {
      "epoch": 196.29090909090908,
      "grad_norm": 0.3265436887741089,
      "learning_rate": 2.8410000000000004e-05,
      "loss": 0.3615,
      "step": 2159
    },
    {
      "epoch": 196.38787878787878,
      "grad_norm": 0.32005420327186584,
      "learning_rate": 2.84e-05,
      "loss": 0.4026,
      "step": 2160
    },
    {
      "epoch": 196.38787878787878,
      "eval_loss": 0.4497809112071991,
      "eval_runtime": 2.1349,
      "eval_samples_per_second": 25.762,
      "eval_steps_per_second": 3.279,
      "step": 2160
    },
    {
      "epoch": 196.4848484848485,
      "grad_norm": 0.31602486968040466,
      "learning_rate": 2.839e-05,
      "loss": 0.3344,
      "step": 2161
    },
    {
      "epoch": 196.5818181818182,
      "grad_norm": 0.29071107506752014,
      "learning_rate": 2.8380000000000003e-05,
      "loss": 0.428,
      "step": 2162
    },
    {
      "epoch": 196.6787878787879,
      "grad_norm": 0.3036842942237854,
      "learning_rate": 2.837e-05,
      "loss": 0.4099,
      "step": 2163
    },
    {
      "epoch": 196.77575757575758,
      "grad_norm": 0.2522382140159607,
      "learning_rate": 2.8360000000000003e-05,
      "loss": 0.4009,
      "step": 2164
    },
    {
      "epoch": 196.87272727272727,
      "grad_norm": 0.2813572883605957,
      "learning_rate": 2.8349999999999998e-05,
      "loss": 0.3761,
      "step": 2165
    },
    {
      "epoch": 196.96969696969697,
      "grad_norm": 0.2621469795703888,
      "learning_rate": 2.834e-05,
      "loss": 0.3328,
      "step": 2166
    },
    {
      "epoch": 197.0,
      "grad_norm": 0.5489789843559265,
      "learning_rate": 2.833e-05,
      "loss": 0.3352,
      "step": 2167
    },
    {
      "epoch": 197.0969696969697,
      "grad_norm": 0.24636216461658478,
      "learning_rate": 2.8320000000000003e-05,
      "loss": 0.3706,
      "step": 2168
    },
    {
      "epoch": 197.1939393939394,
      "grad_norm": 0.2669205665588379,
      "learning_rate": 2.8310000000000002e-05,
      "loss": 0.3886,
      "step": 2169
    },
    {
      "epoch": 197.29090909090908,
      "grad_norm": 0.36443209648132324,
      "learning_rate": 2.83e-05,
      "loss": 0.3641,
      "step": 2170
    },
    {
      "epoch": 197.29090909090908,
      "eval_loss": 0.44957584142684937,
      "eval_runtime": 2.1169,
      "eval_samples_per_second": 25.981,
      "eval_steps_per_second": 3.307,
      "step": 2170
    },
    {
      "epoch": 197.38787878787878,
      "grad_norm": 0.3616332411766052,
      "learning_rate": 2.829e-05,
      "loss": 0.3687,
      "step": 2171
    },
    {
      "epoch": 197.4848484848485,
      "grad_norm": 0.29667091369628906,
      "learning_rate": 2.828e-05,
      "loss": 0.3602,
      "step": 2172
    },
    {
      "epoch": 197.5818181818182,
      "grad_norm": 0.34419506788253784,
      "learning_rate": 2.8270000000000002e-05,
      "loss": 0.3893,
      "step": 2173
    },
    {
      "epoch": 197.6787878787879,
      "grad_norm": 0.2823171019554138,
      "learning_rate": 2.8260000000000004e-05,
      "loss": 0.3839,
      "step": 2174
    },
    {
      "epoch": 197.77575757575758,
      "grad_norm": 0.33764100074768066,
      "learning_rate": 2.825e-05,
      "loss": 0.4157,
      "step": 2175
    },
    {
      "epoch": 197.87272727272727,
      "grad_norm": 0.30144527554512024,
      "learning_rate": 2.824e-05,
      "loss": 0.4137,
      "step": 2176
    },
    {
      "epoch": 197.96969696969697,
      "grad_norm": 0.30979812145233154,
      "learning_rate": 2.8230000000000002e-05,
      "loss": 0.3674,
      "step": 2177
    },
    {
      "epoch": 198.0,
      "grad_norm": 0.43826353549957275,
      "learning_rate": 2.822e-05,
      "loss": 0.3122,
      "step": 2178
    },
    {
      "epoch": 198.0969696969697,
      "grad_norm": 0.2648268938064575,
      "learning_rate": 2.8210000000000003e-05,
      "loss": 0.3007,
      "step": 2179
    },
    {
      "epoch": 198.1939393939394,
      "grad_norm": 0.2786037027835846,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 0.3952,
      "step": 2180
    },
    {
      "epoch": 198.1939393939394,
      "eval_loss": 0.448219358921051,
      "eval_runtime": 2.1609,
      "eval_samples_per_second": 25.452,
      "eval_steps_per_second": 3.239,
      "step": 2180
    },
    {
      "epoch": 198.29090909090908,
      "grad_norm": 0.31549665331840515,
      "learning_rate": 2.819e-05,
      "loss": 0.3769,
      "step": 2181
    },
    {
      "epoch": 198.38787878787878,
      "grad_norm": 0.3384656310081482,
      "learning_rate": 2.818e-05,
      "loss": 0.4068,
      "step": 2182
    },
    {
      "epoch": 198.4848484848485,
      "grad_norm": 0.33379441499710083,
      "learning_rate": 2.8170000000000003e-05,
      "loss": 0.4166,
      "step": 2183
    },
    {
      "epoch": 198.5818181818182,
      "grad_norm": 0.4237292408943176,
      "learning_rate": 2.816e-05,
      "loss": 0.3673,
      "step": 2184
    },
    {
      "epoch": 198.6787878787879,
      "grad_norm": 0.306974858045578,
      "learning_rate": 2.815e-05,
      "loss": 0.3976,
      "step": 2185
    },
    {
      "epoch": 198.77575757575758,
      "grad_norm": 0.28793197870254517,
      "learning_rate": 2.8139999999999998e-05,
      "loss": 0.3793,
      "step": 2186
    },
    {
      "epoch": 198.87272727272727,
      "grad_norm": 0.28818053007125854,
      "learning_rate": 2.813e-05,
      "loss": 0.3847,
      "step": 2187
    },
    {
      "epoch": 198.96969696969697,
      "grad_norm": 0.34341973066329956,
      "learning_rate": 2.8120000000000002e-05,
      "loss": 0.3795,
      "step": 2188
    },
    {
      "epoch": 199.0,
      "grad_norm": 0.46411457657814026,
      "learning_rate": 2.8110000000000004e-05,
      "loss": 0.3738,
      "step": 2189
    },
    {
      "epoch": 199.0969696969697,
      "grad_norm": 0.32759159803390503,
      "learning_rate": 2.8100000000000005e-05,
      "loss": 0.39,
      "step": 2190
    },
    {
      "epoch": 199.0969696969697,
      "eval_loss": 0.4479880928993225,
      "eval_runtime": 2.1443,
      "eval_samples_per_second": 25.649,
      "eval_steps_per_second": 3.264,
      "step": 2190
    },
    {
      "epoch": 199.1939393939394,
      "grad_norm": 0.2490648478269577,
      "learning_rate": 2.809e-05,
      "loss": 0.3804,
      "step": 2191
    },
    {
      "epoch": 199.29090909090908,
      "grad_norm": 0.3017059564590454,
      "learning_rate": 2.8080000000000002e-05,
      "loss": 0.3681,
      "step": 2192
    },
    {
      "epoch": 199.38787878787878,
      "grad_norm": 0.3729502856731415,
      "learning_rate": 2.807e-05,
      "loss": 0.3944,
      "step": 2193
    },
    {
      "epoch": 199.4848484848485,
      "grad_norm": 0.32608261704444885,
      "learning_rate": 2.8060000000000002e-05,
      "loss": 0.3895,
      "step": 2194
    },
    {
      "epoch": 199.5818181818182,
      "grad_norm": 0.23186321556568146,
      "learning_rate": 2.8050000000000004e-05,
      "loss": 0.3592,
      "step": 2195
    },
    {
      "epoch": 199.6787878787879,
      "grad_norm": 0.3033948540687561,
      "learning_rate": 2.804e-05,
      "loss": 0.396,
      "step": 2196
    },
    {
      "epoch": 199.77575757575758,
      "grad_norm": 0.30141010880470276,
      "learning_rate": 2.803e-05,
      "loss": 0.3821,
      "step": 2197
    },
    {
      "epoch": 199.87272727272727,
      "grad_norm": 0.3135485351085663,
      "learning_rate": 2.8020000000000003e-05,
      "loss": 0.3644,
      "step": 2198
    },
    {
      "epoch": 199.96969696969697,
      "grad_norm": 0.27928251028060913,
      "learning_rate": 2.8010000000000005e-05,
      "loss": 0.3543,
      "step": 2199
    },
    {
      "epoch": 200.0,
      "grad_norm": 0.4405915439128876,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.419,
      "step": 2200
    },
    {
      "epoch": 200.0,
      "eval_loss": 0.4478053152561188,
      "eval_runtime": 2.1288,
      "eval_samples_per_second": 25.836,
      "eval_steps_per_second": 3.288,
      "step": 2200
    },
    {
      "epoch": 200.0969696969697,
      "grad_norm": 0.25227150321006775,
      "learning_rate": 2.7989999999999998e-05,
      "loss": 0.3807,
      "step": 2201
    },
    {
      "epoch": 200.1939393939394,
      "grad_norm": 0.27379876375198364,
      "learning_rate": 2.798e-05,
      "loss": 0.3747,
      "step": 2202
    },
    {
      "epoch": 200.29090909090908,
      "grad_norm": 0.3297831118106842,
      "learning_rate": 2.797e-05,
      "loss": 0.3817,
      "step": 2203
    },
    {
      "epoch": 200.38787878787878,
      "grad_norm": 0.3016502261161804,
      "learning_rate": 2.7960000000000003e-05,
      "loss": 0.3829,
      "step": 2204
    },
    {
      "epoch": 200.4848484848485,
      "grad_norm": 0.3145178258419037,
      "learning_rate": 2.7950000000000005e-05,
      "loss": 0.3586,
      "step": 2205
    },
    {
      "epoch": 200.5818181818182,
      "grad_norm": 0.3010222017765045,
      "learning_rate": 2.794e-05,
      "loss": 0.3441,
      "step": 2206
    },
    {
      "epoch": 200.6787878787879,
      "grad_norm": 0.27790072560310364,
      "learning_rate": 2.7930000000000002e-05,
      "loss": 0.3872,
      "step": 2207
    },
    {
      "epoch": 200.77575757575758,
      "grad_norm": 0.306438148021698,
      "learning_rate": 2.792e-05,
      "loss": 0.402,
      "step": 2208
    },
    {
      "epoch": 200.87272727272727,
      "grad_norm": 0.3268650472164154,
      "learning_rate": 2.7910000000000002e-05,
      "loss": 0.4085,
      "step": 2209
    },
    {
      "epoch": 200.96969696969697,
      "grad_norm": 0.28407397866249084,
      "learning_rate": 2.7900000000000004e-05,
      "loss": 0.3667,
      "step": 2210
    },
    {
      "epoch": 200.96969696969697,
      "eval_loss": 0.44716876745224,
      "eval_runtime": 2.1402,
      "eval_samples_per_second": 25.698,
      "eval_steps_per_second": 3.271,
      "step": 2210
    },
    {
      "epoch": 201.0,
      "grad_norm": 0.537050724029541,
      "learning_rate": 2.789e-05,
      "loss": 0.3747,
      "step": 2211
    },
    {
      "epoch": 201.0969696969697,
      "grad_norm": 0.34605276584625244,
      "learning_rate": 2.788e-05,
      "loss": 0.3457,
      "step": 2212
    },
    {
      "epoch": 201.1939393939394,
      "grad_norm": 0.3041094243526459,
      "learning_rate": 2.7870000000000003e-05,
      "loss": 0.3857,
      "step": 2213
    },
    {
      "epoch": 201.29090909090908,
      "grad_norm": 0.2671644389629364,
      "learning_rate": 2.7860000000000004e-05,
      "loss": 0.4177,
      "step": 2214
    },
    {
      "epoch": 201.38787878787878,
      "grad_norm": 0.29513826966285706,
      "learning_rate": 2.7850000000000003e-05,
      "loss": 0.4293,
      "step": 2215
    },
    {
      "epoch": 201.4848484848485,
      "grad_norm": 0.34477704763412476,
      "learning_rate": 2.7839999999999998e-05,
      "loss": 0.3838,
      "step": 2216
    },
    {
      "epoch": 201.5818181818182,
      "grad_norm": 0.2400050014257431,
      "learning_rate": 2.783e-05,
      "loss": 0.3408,
      "step": 2217
    },
    {
      "epoch": 201.6787878787879,
      "grad_norm": 0.25592660903930664,
      "learning_rate": 2.782e-05,
      "loss": 0.3827,
      "step": 2218
    },
    {
      "epoch": 201.77575757575758,
      "grad_norm": 0.3099137842655182,
      "learning_rate": 2.7810000000000003e-05,
      "loss": 0.3614,
      "step": 2219
    },
    {
      "epoch": 201.87272727272727,
      "grad_norm": 0.3049536645412445,
      "learning_rate": 2.7800000000000005e-05,
      "loss": 0.3757,
      "step": 2220
    },
    {
      "epoch": 201.87272727272727,
      "eval_loss": 0.4471142888069153,
      "eval_runtime": 2.1643,
      "eval_samples_per_second": 25.413,
      "eval_steps_per_second": 3.234,
      "step": 2220
    },
    {
      "epoch": 201.96969696969697,
      "grad_norm": 0.26360929012298584,
      "learning_rate": 2.779e-05,
      "loss": 0.3576,
      "step": 2221
    },
    {
      "epoch": 202.0,
      "grad_norm": 0.5084606409072876,
      "learning_rate": 2.778e-05,
      "loss": 0.3695,
      "step": 2222
    },
    {
      "epoch": 202.0969696969697,
      "grad_norm": 0.306815505027771,
      "learning_rate": 2.777e-05,
      "loss": 0.4109,
      "step": 2223
    },
    {
      "epoch": 202.1939393939394,
      "grad_norm": 0.28837597370147705,
      "learning_rate": 2.7760000000000002e-05,
      "loss": 0.3717,
      "step": 2224
    },
    {
      "epoch": 202.29090909090908,
      "grad_norm": 0.2985439896583557,
      "learning_rate": 2.7750000000000004e-05,
      "loss": 0.3379,
      "step": 2225
    },
    {
      "epoch": 202.38787878787878,
      "grad_norm": 0.34035661816596985,
      "learning_rate": 2.774e-05,
      "loss": 0.3933,
      "step": 2226
    },
    {
      "epoch": 202.4848484848485,
      "grad_norm": 0.31377413868904114,
      "learning_rate": 2.773e-05,
      "loss": 0.359,
      "step": 2227
    },
    {
      "epoch": 202.5818181818182,
      "grad_norm": 0.3064294457435608,
      "learning_rate": 2.7720000000000002e-05,
      "loss": 0.387,
      "step": 2228
    },
    {
      "epoch": 202.6787878787879,
      "grad_norm": 0.27735260128974915,
      "learning_rate": 2.7710000000000004e-05,
      "loss": 0.3856,
      "step": 2229
    },
    {
      "epoch": 202.77575757575758,
      "grad_norm": 0.2875250279903412,
      "learning_rate": 2.7700000000000002e-05,
      "loss": 0.3819,
      "step": 2230
    },
    {
      "epoch": 202.77575757575758,
      "eval_loss": 0.4466860294342041,
      "eval_runtime": 2.1516,
      "eval_samples_per_second": 25.563,
      "eval_steps_per_second": 3.253,
      "step": 2230
    },
    {
      "epoch": 202.87272727272727,
      "grad_norm": 0.26599356532096863,
      "learning_rate": 2.769e-05,
      "loss": 0.3431,
      "step": 2231
    },
    {
      "epoch": 202.96969696969697,
      "grad_norm": 0.27645865082740784,
      "learning_rate": 2.768e-05,
      "loss": 0.3793,
      "step": 2232
    },
    {
      "epoch": 203.0,
      "grad_norm": 0.6768845915794373,
      "learning_rate": 2.767e-05,
      "loss": 0.4634,
      "step": 2233
    },
    {
      "epoch": 203.0969696969697,
      "grad_norm": 0.2488522231578827,
      "learning_rate": 2.7660000000000003e-05,
      "loss": 0.3871,
      "step": 2234
    },
    {
      "epoch": 203.1939393939394,
      "grad_norm": 0.25670602917671204,
      "learning_rate": 2.7650000000000005e-05,
      "loss": 0.3764,
      "step": 2235
    },
    {
      "epoch": 203.29090909090908,
      "grad_norm": 0.2679460346698761,
      "learning_rate": 2.764e-05,
      "loss": 0.3713,
      "step": 2236
    },
    {
      "epoch": 203.38787878787878,
      "grad_norm": 0.2824065685272217,
      "learning_rate": 2.763e-05,
      "loss": 0.4196,
      "step": 2237
    },
    {
      "epoch": 203.4848484848485,
      "grad_norm": 0.3017754852771759,
      "learning_rate": 2.762e-05,
      "loss": 0.4152,
      "step": 2238
    },
    {
      "epoch": 203.5818181818182,
      "grad_norm": 0.31364500522613525,
      "learning_rate": 2.761e-05,
      "loss": 0.3406,
      "step": 2239
    },
    {
      "epoch": 203.6787878787879,
      "grad_norm": 0.3167931139469147,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 0.365,
      "step": 2240
    },
    {
      "epoch": 203.6787878787879,
      "eval_loss": 0.44658520817756653,
      "eval_runtime": 2.1274,
      "eval_samples_per_second": 25.853,
      "eval_steps_per_second": 3.29,
      "step": 2240
    },
    {
      "epoch": 203.77575757575758,
      "grad_norm": 0.2750272750854492,
      "learning_rate": 2.759e-05,
      "loss": 0.3776,
      "step": 2241
    },
    {
      "epoch": 203.87272727272727,
      "grad_norm": 0.3103097081184387,
      "learning_rate": 2.758e-05,
      "loss": 0.3668,
      "step": 2242
    },
    {
      "epoch": 203.96969696969697,
      "grad_norm": 0.2923380732536316,
      "learning_rate": 2.7570000000000002e-05,
      "loss": 0.3329,
      "step": 2243
    },
    {
      "epoch": 204.0,
      "grad_norm": 0.4510895609855652,
      "learning_rate": 2.7560000000000004e-05,
      "loss": 0.427,
      "step": 2244
    },
    {
      "epoch": 204.0969696969697,
      "grad_norm": 0.3134826123714447,
      "learning_rate": 2.7550000000000002e-05,
      "loss": 0.3567,
      "step": 2245
    },
    {
      "epoch": 204.1939393939394,
      "grad_norm": 0.35793009400367737,
      "learning_rate": 2.754e-05,
      "loss": 0.4098,
      "step": 2246
    },
    {
      "epoch": 204.29090909090908,
      "grad_norm": 0.2968956530094147,
      "learning_rate": 2.753e-05,
      "loss": 0.363,
      "step": 2247
    },
    {
      "epoch": 204.38787878787878,
      "grad_norm": 0.28569701313972473,
      "learning_rate": 2.752e-05,
      "loss": 0.3713,
      "step": 2248
    },
    {
      "epoch": 204.4848484848485,
      "grad_norm": 0.3336655795574188,
      "learning_rate": 2.7510000000000003e-05,
      "loss": 0.3785,
      "step": 2249
    },
    {
      "epoch": 204.5818181818182,
      "grad_norm": 0.3033703565597534,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.3699,
      "step": 2250
    },
    {
      "epoch": 204.5818181818182,
      "eval_loss": 0.44568753242492676,
      "eval_runtime": 2.1588,
      "eval_samples_per_second": 25.478,
      "eval_steps_per_second": 3.243,
      "step": 2250
    },
    {
      "epoch": 204.6787878787879,
      "grad_norm": 0.31997036933898926,
      "learning_rate": 2.749e-05,
      "loss": 0.3259,
      "step": 2251
    },
    {
      "epoch": 204.77575757575758,
      "grad_norm": 0.281178742647171,
      "learning_rate": 2.748e-05,
      "loss": 0.4069,
      "step": 2252
    },
    {
      "epoch": 204.87272727272727,
      "grad_norm": 0.3126077950000763,
      "learning_rate": 2.7470000000000003e-05,
      "loss": 0.3975,
      "step": 2253
    },
    {
      "epoch": 204.96969696969697,
      "grad_norm": 0.31442731618881226,
      "learning_rate": 2.746e-05,
      "loss": 0.3947,
      "step": 2254
    },
    {
      "epoch": 205.0,
      "grad_norm": 0.6187658309936523,
      "learning_rate": 2.7450000000000003e-05,
      "loss": 0.3549,
      "step": 2255
    },
    {
      "epoch": 205.0969696969697,
      "grad_norm": 0.32099851965904236,
      "learning_rate": 2.7439999999999998e-05,
      "loss": 0.3769,
      "step": 2256
    },
    {
      "epoch": 205.1939393939394,
      "grad_norm": 0.4058302342891693,
      "learning_rate": 2.743e-05,
      "loss": 0.3651,
      "step": 2257
    },
    {
      "epoch": 205.29090909090908,
      "grad_norm": 0.3068258464336395,
      "learning_rate": 2.7420000000000002e-05,
      "loss": 0.4019,
      "step": 2258
    },
    {
      "epoch": 205.38787878787878,
      "grad_norm": 0.5158912539482117,
      "learning_rate": 2.7410000000000004e-05,
      "loss": 0.4054,
      "step": 2259
    },
    {
      "epoch": 205.4848484848485,
      "grad_norm": 0.28603148460388184,
      "learning_rate": 2.7400000000000002e-05,
      "loss": 0.3759,
      "step": 2260
    },
    {
      "epoch": 205.4848484848485,
      "eval_loss": 0.44500473141670227,
      "eval_runtime": 2.155,
      "eval_samples_per_second": 25.522,
      "eval_steps_per_second": 3.248,
      "step": 2260
    },
    {
      "epoch": 205.5818181818182,
      "grad_norm": 0.27636051177978516,
      "learning_rate": 2.739e-05,
      "loss": 0.3705,
      "step": 2261
    },
    {
      "epoch": 205.6787878787879,
      "grad_norm": 0.31432798504829407,
      "learning_rate": 2.738e-05,
      "loss": 0.3597,
      "step": 2262
    },
    {
      "epoch": 205.77575757575758,
      "grad_norm": 0.3602944016456604,
      "learning_rate": 2.737e-05,
      "loss": 0.3889,
      "step": 2263
    },
    {
      "epoch": 205.87272727272727,
      "grad_norm": 0.38743069767951965,
      "learning_rate": 2.7360000000000002e-05,
      "loss": 0.3703,
      "step": 2264
    },
    {
      "epoch": 205.96969696969697,
      "grad_norm": 0.30791616439819336,
      "learning_rate": 2.7350000000000004e-05,
      "loss": 0.3417,
      "step": 2265
    },
    {
      "epoch": 206.0,
      "grad_norm": 0.46710509061813354,
      "learning_rate": 2.734e-05,
      "loss": 0.3919,
      "step": 2266
    },
    {
      "epoch": 206.0969696969697,
      "grad_norm": 0.23429720103740692,
      "learning_rate": 2.733e-05,
      "loss": 0.3503,
      "step": 2267
    },
    {
      "epoch": 206.1939393939394,
      "grad_norm": 0.2622058391571045,
      "learning_rate": 2.7320000000000003e-05,
      "loss": 0.3682,
      "step": 2268
    },
    {
      "epoch": 206.29090909090908,
      "grad_norm": 0.341849684715271,
      "learning_rate": 2.731e-05,
      "loss": 0.3715,
      "step": 2269
    },
    {
      "epoch": 206.38787878787878,
      "grad_norm": 0.31176722049713135,
      "learning_rate": 2.7300000000000003e-05,
      "loss": 0.4036,
      "step": 2270
    },
    {
      "epoch": 206.38787878787878,
      "eval_loss": 0.4462737441062927,
      "eval_runtime": 2.141,
      "eval_samples_per_second": 25.689,
      "eval_steps_per_second": 3.27,
      "step": 2270
    },
    {
      "epoch": 206.4848484848485,
      "grad_norm": 0.30896806716918945,
      "learning_rate": 2.7289999999999998e-05,
      "loss": 0.3979,
      "step": 2271
    },
    {
      "epoch": 206.5818181818182,
      "grad_norm": 0.35733941197395325,
      "learning_rate": 2.728e-05,
      "loss": 0.3668,
      "step": 2272
    },
    {
      "epoch": 206.6787878787879,
      "grad_norm": 0.3435685932636261,
      "learning_rate": 2.727e-05,
      "loss": 0.3879,
      "step": 2273
    },
    {
      "epoch": 206.77575757575758,
      "grad_norm": 0.2502520978450775,
      "learning_rate": 2.7260000000000003e-05,
      "loss": 0.3613,
      "step": 2274
    },
    {
      "epoch": 206.87272727272727,
      "grad_norm": 0.3206669092178345,
      "learning_rate": 2.725e-05,
      "loss": 0.4247,
      "step": 2275
    },
    {
      "epoch": 206.96969696969697,
      "grad_norm": 0.2648768126964569,
      "learning_rate": 2.724e-05,
      "loss": 0.3348,
      "step": 2276
    },
    {
      "epoch": 207.0,
      "grad_norm": 0.6231780052185059,
      "learning_rate": 2.723e-05,
      "loss": 0.3228,
      "step": 2277
    },
    {
      "epoch": 207.0969696969697,
      "grad_norm": 0.3182898163795471,
      "learning_rate": 2.722e-05,
      "loss": 0.352,
      "step": 2278
    },
    {
      "epoch": 207.1939393939394,
      "grad_norm": 0.2684725224971771,
      "learning_rate": 2.7210000000000002e-05,
      "loss": 0.3666,
      "step": 2279
    },
    {
      "epoch": 207.29090909090908,
      "grad_norm": 0.2992083728313446,
      "learning_rate": 2.7200000000000004e-05,
      "loss": 0.4296,
      "step": 2280
    },
    {
      "epoch": 207.29090909090908,
      "eval_loss": 0.44510698318481445,
      "eval_runtime": 2.1509,
      "eval_samples_per_second": 25.571,
      "eval_steps_per_second": 3.254,
      "step": 2280
    },
    {
      "epoch": 207.38787878787878,
      "grad_norm": 0.26230400800704956,
      "learning_rate": 2.719e-05,
      "loss": 0.3608,
      "step": 2281
    },
    {
      "epoch": 207.4848484848485,
      "grad_norm": 0.3085038363933563,
      "learning_rate": 2.718e-05,
      "loss": 0.3991,
      "step": 2282
    },
    {
      "epoch": 207.5818181818182,
      "grad_norm": 0.33652815222740173,
      "learning_rate": 2.7170000000000002e-05,
      "loss": 0.3472,
      "step": 2283
    },
    {
      "epoch": 207.6787878787879,
      "grad_norm": 0.3844667375087738,
      "learning_rate": 2.716e-05,
      "loss": 0.3846,
      "step": 2284
    },
    {
      "epoch": 207.77575757575758,
      "grad_norm": 0.29387590289115906,
      "learning_rate": 2.7150000000000003e-05,
      "loss": 0.3576,
      "step": 2285
    },
    {
      "epoch": 207.87272727272727,
      "grad_norm": 0.30108508467674255,
      "learning_rate": 2.7139999999999998e-05,
      "loss": 0.3944,
      "step": 2286
    },
    {
      "epoch": 207.96969696969697,
      "grad_norm": 0.30221885442733765,
      "learning_rate": 2.713e-05,
      "loss": 0.3693,
      "step": 2287
    },
    {
      "epoch": 208.0,
      "grad_norm": 0.44697195291519165,
      "learning_rate": 2.712e-05,
      "loss": 0.3378,
      "step": 2288
    },
    {
      "epoch": 208.0969696969697,
      "grad_norm": 0.25601184368133545,
      "learning_rate": 2.7110000000000003e-05,
      "loss": 0.3761,
      "step": 2289
    },
    {
      "epoch": 208.1939393939394,
      "grad_norm": 0.283724844455719,
      "learning_rate": 2.7100000000000005e-05,
      "loss": 0.3876,
      "step": 2290
    },
    {
      "epoch": 208.1939393939394,
      "eval_loss": 0.4447232186794281,
      "eval_runtime": 2.1305,
      "eval_samples_per_second": 25.815,
      "eval_steps_per_second": 3.286,
      "step": 2290
    },
    {
      "epoch": 208.29090909090908,
      "grad_norm": 0.3504049479961395,
      "learning_rate": 2.709e-05,
      "loss": 0.3737,
      "step": 2291
    },
    {
      "epoch": 208.38787878787878,
      "grad_norm": 0.2954561114311218,
      "learning_rate": 2.7079999999999998e-05,
      "loss": 0.3726,
      "step": 2292
    },
    {
      "epoch": 208.4848484848485,
      "grad_norm": 0.27079299092292786,
      "learning_rate": 2.707e-05,
      "loss": 0.4073,
      "step": 2293
    },
    {
      "epoch": 208.5818181818182,
      "grad_norm": 0.2923456132411957,
      "learning_rate": 2.7060000000000002e-05,
      "loss": 0.3635,
      "step": 2294
    },
    {
      "epoch": 208.6787878787879,
      "grad_norm": 0.28545117378234863,
      "learning_rate": 2.7050000000000004e-05,
      "loss": 0.3716,
      "step": 2295
    },
    {
      "epoch": 208.77575757575758,
      "grad_norm": 0.28535234928131104,
      "learning_rate": 2.704e-05,
      "loss": 0.3699,
      "step": 2296
    },
    {
      "epoch": 208.87272727272727,
      "grad_norm": 0.3026406764984131,
      "learning_rate": 2.703e-05,
      "loss": 0.3638,
      "step": 2297
    },
    {
      "epoch": 208.96969696969697,
      "grad_norm": 0.3343816101551056,
      "learning_rate": 2.7020000000000002e-05,
      "loss": 0.369,
      "step": 2298
    },
    {
      "epoch": 209.0,
      "grad_norm": 0.46312960982322693,
      "learning_rate": 2.701e-05,
      "loss": 0.3242,
      "step": 2299
    },
    {
      "epoch": 209.0969696969697,
      "grad_norm": 0.2601545453071594,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 0.3709,
      "step": 2300
    },
    {
      "epoch": 209.0969696969697,
      "eval_loss": 0.4444558620452881,
      "eval_runtime": 2.1319,
      "eval_samples_per_second": 25.799,
      "eval_steps_per_second": 3.284,
      "step": 2300
    },
    {
      "epoch": 209.1939393939394,
      "grad_norm": 0.2880859076976776,
      "learning_rate": 2.6989999999999997e-05,
      "loss": 0.375,
      "step": 2301
    },
    {
      "epoch": 209.29090909090908,
      "grad_norm": 0.4282890558242798,
      "learning_rate": 2.698e-05,
      "loss": 0.3903,
      "step": 2302
    },
    {
      "epoch": 209.38787878787878,
      "grad_norm": 0.28130725026130676,
      "learning_rate": 2.697e-05,
      "loss": 0.3495,
      "step": 2303
    },
    {
      "epoch": 209.4848484848485,
      "grad_norm": 0.33825841546058655,
      "learning_rate": 2.6960000000000003e-05,
      "loss": 0.392,
      "step": 2304
    },
    {
      "epoch": 209.5818181818182,
      "grad_norm": 0.34332937002182007,
      "learning_rate": 2.6950000000000005e-05,
      "loss": 0.3506,
      "step": 2305
    },
    {
      "epoch": 209.6787878787879,
      "grad_norm": 0.2920333445072174,
      "learning_rate": 2.694e-05,
      "loss": 0.3493,
      "step": 2306
    },
    {
      "epoch": 209.77575757575758,
      "grad_norm": 0.29194843769073486,
      "learning_rate": 2.693e-05,
      "loss": 0.4088,
      "step": 2307
    },
    {
      "epoch": 209.87272727272727,
      "grad_norm": 0.2841552495956421,
      "learning_rate": 2.692e-05,
      "loss": 0.4067,
      "step": 2308
    },
    {
      "epoch": 209.96969696969697,
      "grad_norm": 0.3399902582168579,
      "learning_rate": 2.691e-05,
      "loss": 0.3592,
      "step": 2309
    },
    {
      "epoch": 210.0,
      "grad_norm": 0.4557587206363678,
      "learning_rate": 2.6900000000000003e-05,
      "loss": 0.3301,
      "step": 2310
    },
    {
      "epoch": 210.0,
      "eval_loss": 0.44406941533088684,
      "eval_runtime": 2.1725,
      "eval_samples_per_second": 25.317,
      "eval_steps_per_second": 3.222,
      "step": 2310
    },
    {
      "epoch": 210.0969696969697,
      "grad_norm": 0.29916703701019287,
      "learning_rate": 2.689e-05,
      "loss": 0.3669,
      "step": 2311
    },
    {
      "epoch": 210.1939393939394,
      "grad_norm": 0.3633650541305542,
      "learning_rate": 2.688e-05,
      "loss": 0.3665,
      "step": 2312
    },
    {
      "epoch": 210.29090909090908,
      "grad_norm": 0.29734817147254944,
      "learning_rate": 2.6870000000000002e-05,
      "loss": 0.3624,
      "step": 2313
    },
    {
      "epoch": 210.38787878787878,
      "grad_norm": 0.2979733347892761,
      "learning_rate": 2.686e-05,
      "loss": 0.3705,
      "step": 2314
    },
    {
      "epoch": 210.4848484848485,
      "grad_norm": 0.3146786689758301,
      "learning_rate": 2.6850000000000002e-05,
      "loss": 0.3792,
      "step": 2315
    },
    {
      "epoch": 210.5818181818182,
      "grad_norm": 0.2864646017551422,
      "learning_rate": 2.6840000000000004e-05,
      "loss": 0.4047,
      "step": 2316
    },
    {
      "epoch": 210.6787878787879,
      "grad_norm": 0.24151170253753662,
      "learning_rate": 2.683e-05,
      "loss": 0.3796,
      "step": 2317
    },
    {
      "epoch": 210.77575757575758,
      "grad_norm": 0.2813226580619812,
      "learning_rate": 2.682e-05,
      "loss": 0.3795,
      "step": 2318
    },
    {
      "epoch": 210.87272727272727,
      "grad_norm": 0.28772884607315063,
      "learning_rate": 2.6810000000000003e-05,
      "loss": 0.3447,
      "step": 2319
    },
    {
      "epoch": 210.96969696969697,
      "grad_norm": 0.33018478751182556,
      "learning_rate": 2.6800000000000004e-05,
      "loss": 0.357,
      "step": 2320
    },
    {
      "epoch": 210.96969696969697,
      "eval_loss": 0.44407859444618225,
      "eval_runtime": 2.16,
      "eval_samples_per_second": 25.463,
      "eval_steps_per_second": 3.241,
      "step": 2320
    },
    {
      "epoch": 211.0,
      "grad_norm": 0.6959975361824036,
      "learning_rate": 2.6790000000000003e-05,
      "loss": 0.4455,
      "step": 2321
    },
    {
      "epoch": 211.0969696969697,
      "grad_norm": 0.26903417706489563,
      "learning_rate": 2.678e-05,
      "loss": 0.3935,
      "step": 2322
    },
    {
      "epoch": 211.1939393939394,
      "grad_norm": 0.2566149830818176,
      "learning_rate": 2.677e-05,
      "loss": 0.4047,
      "step": 2323
    },
    {
      "epoch": 211.29090909090908,
      "grad_norm": 0.3153493106365204,
      "learning_rate": 2.676e-05,
      "loss": 0.3853,
      "step": 2324
    },
    {
      "epoch": 211.38787878787878,
      "grad_norm": 0.33052006363868713,
      "learning_rate": 2.6750000000000003e-05,
      "loss": 0.3401,
      "step": 2325
    },
    {
      "epoch": 211.4848484848485,
      "grad_norm": 0.28321343660354614,
      "learning_rate": 2.6740000000000005e-05,
      "loss": 0.3642,
      "step": 2326
    },
    {
      "epoch": 211.5818181818182,
      "grad_norm": 0.2920601963996887,
      "learning_rate": 2.673e-05,
      "loss": 0.3674,
      "step": 2327
    },
    {
      "epoch": 211.6787878787879,
      "grad_norm": 0.30511558055877686,
      "learning_rate": 2.672e-05,
      "loss": 0.3397,
      "step": 2328
    },
    {
      "epoch": 211.77575757575758,
      "grad_norm": 0.2932232916355133,
      "learning_rate": 2.671e-05,
      "loss": 0.372,
      "step": 2329
    },
    {
      "epoch": 211.87272727272727,
      "grad_norm": 0.26344335079193115,
      "learning_rate": 2.6700000000000002e-05,
      "loss": 0.3836,
      "step": 2330
    },
    {
      "epoch": 211.87272727272727,
      "eval_loss": 0.44368764758110046,
      "eval_runtime": 2.1428,
      "eval_samples_per_second": 25.667,
      "eval_steps_per_second": 3.267,
      "step": 2330
    },
    {
      "epoch": 211.96969696969697,
      "grad_norm": 0.302018940448761,
      "learning_rate": 2.6690000000000004e-05,
      "loss": 0.3491,
      "step": 2331
    },
    {
      "epoch": 212.0,
      "grad_norm": 0.5806761980056763,
      "learning_rate": 2.668e-05,
      "loss": 0.4508,
      "step": 2332
    },
    {
      "epoch": 212.0969696969697,
      "grad_norm": 0.2934873402118683,
      "learning_rate": 2.667e-05,
      "loss": 0.3778,
      "step": 2333
    },
    {
      "epoch": 212.1939393939394,
      "grad_norm": 0.30157631635665894,
      "learning_rate": 2.6660000000000002e-05,
      "loss": 0.3947,
      "step": 2334
    },
    {
      "epoch": 212.29090909090908,
      "grad_norm": 0.34195688366889954,
      "learning_rate": 2.6650000000000004e-05,
      "loss": 0.3918,
      "step": 2335
    },
    {
      "epoch": 212.38787878787878,
      "grad_norm": 0.2895645797252655,
      "learning_rate": 2.6640000000000002e-05,
      "loss": 0.3873,
      "step": 2336
    },
    {
      "epoch": 212.4848484848485,
      "grad_norm": 0.3008175492286682,
      "learning_rate": 2.663e-05,
      "loss": 0.3488,
      "step": 2337
    },
    {
      "epoch": 212.5818181818182,
      "grad_norm": 0.3135222792625427,
      "learning_rate": 2.662e-05,
      "loss": 0.3687,
      "step": 2338
    },
    {
      "epoch": 212.6787878787879,
      "grad_norm": 0.3346012234687805,
      "learning_rate": 2.661e-05,
      "loss": 0.3364,
      "step": 2339
    },
    {
      "epoch": 212.77575757575758,
      "grad_norm": 0.2940306067466736,
      "learning_rate": 2.6600000000000003e-05,
      "loss": 0.3738,
      "step": 2340
    },
    {
      "epoch": 212.77575757575758,
      "eval_loss": 0.44305384159088135,
      "eval_runtime": 2.1483,
      "eval_samples_per_second": 25.602,
      "eval_steps_per_second": 3.258,
      "step": 2340
    },
    {
      "epoch": 212.87272727272727,
      "grad_norm": 0.3285348117351532,
      "learning_rate": 2.6590000000000005e-05,
      "loss": 0.362,
      "step": 2341
    },
    {
      "epoch": 212.96969696969697,
      "grad_norm": 0.32656341791152954,
      "learning_rate": 2.658e-05,
      "loss": 0.3652,
      "step": 2342
    },
    {
      "epoch": 213.0,
      "grad_norm": 0.48844948410987854,
      "learning_rate": 2.657e-05,
      "loss": 0.4201,
      "step": 2343
    },
    {
      "epoch": 213.0969696969697,
      "grad_norm": 0.29274335503578186,
      "learning_rate": 2.6560000000000003e-05,
      "loss": 0.3844,
      "step": 2344
    },
    {
      "epoch": 213.1939393939394,
      "grad_norm": 0.31116437911987305,
      "learning_rate": 2.655e-05,
      "loss": 0.3413,
      "step": 2345
    },
    {
      "epoch": 213.29090909090908,
      "grad_norm": 0.30364781618118286,
      "learning_rate": 2.6540000000000003e-05,
      "loss": 0.3795,
      "step": 2346
    },
    {
      "epoch": 213.38787878787878,
      "grad_norm": 0.2625868320465088,
      "learning_rate": 2.653e-05,
      "loss": 0.38,
      "step": 2347
    },
    {
      "epoch": 213.4848484848485,
      "grad_norm": 0.2870539128780365,
      "learning_rate": 2.652e-05,
      "loss": 0.3586,
      "step": 2348
    },
    {
      "epoch": 213.5818181818182,
      "grad_norm": 0.3897850215435028,
      "learning_rate": 2.6510000000000002e-05,
      "loss": 0.3713,
      "step": 2349
    },
    {
      "epoch": 213.6787878787879,
      "grad_norm": 0.3048601746559143,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 0.4012,
      "step": 2350
    },
    {
      "epoch": 213.6787878787879,
      "eval_loss": 0.4433639645576477,
      "eval_runtime": 2.1461,
      "eval_samples_per_second": 25.627,
      "eval_steps_per_second": 3.262,
      "step": 2350
    },
    {
      "epoch": 213.77575757575758,
      "grad_norm": 0.26275619864463806,
      "learning_rate": 2.6490000000000002e-05,
      "loss": 0.3418,
      "step": 2351
    },
    {
      "epoch": 213.87272727272727,
      "grad_norm": 0.36444348096847534,
      "learning_rate": 2.648e-05,
      "loss": 0.4155,
      "step": 2352
    },
    {
      "epoch": 213.96969696969697,
      "grad_norm": 0.2950473427772522,
      "learning_rate": 2.647e-05,
      "loss": 0.3565,
      "step": 2353
    },
    {
      "epoch": 214.0,
      "grad_norm": 0.6876822113990784,
      "learning_rate": 2.646e-05,
      "loss": 0.3346,
      "step": 2354
    },
    {
      "epoch": 214.0969696969697,
      "grad_norm": 0.2745935916900635,
      "learning_rate": 2.6450000000000003e-05,
      "loss": 0.3661,
      "step": 2355
    },
    {
      "epoch": 214.1939393939394,
      "grad_norm": 0.29313480854034424,
      "learning_rate": 2.6440000000000004e-05,
      "loss": 0.3596,
      "step": 2356
    },
    {
      "epoch": 214.29090909090908,
      "grad_norm": 0.27624714374542236,
      "learning_rate": 2.643e-05,
      "loss": 0.3754,
      "step": 2357
    },
    {
      "epoch": 214.38787878787878,
      "grad_norm": 0.3175128996372223,
      "learning_rate": 2.642e-05,
      "loss": 0.3611,
      "step": 2358
    },
    {
      "epoch": 214.4848484848485,
      "grad_norm": 0.2875557541847229,
      "learning_rate": 2.6410000000000003e-05,
      "loss": 0.3296,
      "step": 2359
    },
    {
      "epoch": 214.5818181818182,
      "grad_norm": 0.32324209809303284,
      "learning_rate": 2.64e-05,
      "loss": 0.3988,
      "step": 2360
    },
    {
      "epoch": 214.5818181818182,
      "eval_loss": 0.442745566368103,
      "eval_runtime": 2.1394,
      "eval_samples_per_second": 25.708,
      "eval_steps_per_second": 3.272,
      "step": 2360
    },
    {
      "epoch": 214.6787878787879,
      "grad_norm": 0.28028708696365356,
      "learning_rate": 2.6390000000000003e-05,
      "loss": 0.3799,
      "step": 2361
    },
    {
      "epoch": 214.77575757575758,
      "grad_norm": 0.2773652672767639,
      "learning_rate": 2.6379999999999998e-05,
      "loss": 0.3755,
      "step": 2362
    },
    {
      "epoch": 214.87272727272727,
      "grad_norm": 0.3084127902984619,
      "learning_rate": 2.637e-05,
      "loss": 0.3815,
      "step": 2363
    },
    {
      "epoch": 214.96969696969697,
      "grad_norm": 0.30936163663864136,
      "learning_rate": 2.6360000000000002e-05,
      "loss": 0.3916,
      "step": 2364
    },
    {
      "epoch": 215.0,
      "grad_norm": 0.5225114822387695,
      "learning_rate": 2.6350000000000004e-05,
      "loss": 0.3516,
      "step": 2365
    },
    {
      "epoch": 215.0969696969697,
      "grad_norm": 0.3306688964366913,
      "learning_rate": 2.6340000000000002e-05,
      "loss": 0.3982,
      "step": 2366
    },
    {
      "epoch": 215.1939393939394,
      "grad_norm": 0.3688218891620636,
      "learning_rate": 2.633e-05,
      "loss": 0.3643,
      "step": 2367
    },
    {
      "epoch": 215.29090909090908,
      "grad_norm": 0.2552226483821869,
      "learning_rate": 2.632e-05,
      "loss": 0.3626,
      "step": 2368
    },
    {
      "epoch": 215.38787878787878,
      "grad_norm": 0.28460806608200073,
      "learning_rate": 2.631e-05,
      "loss": 0.3908,
      "step": 2369
    },
    {
      "epoch": 215.4848484848485,
      "grad_norm": 0.3210669755935669,
      "learning_rate": 2.6300000000000002e-05,
      "loss": 0.3495,
      "step": 2370
    },
    {
      "epoch": 215.4848484848485,
      "eval_loss": 0.4424516260623932,
      "eval_runtime": 2.1474,
      "eval_samples_per_second": 25.612,
      "eval_steps_per_second": 3.26,
      "step": 2370
    },
    {
      "epoch": 215.5818181818182,
      "grad_norm": 0.2910328209400177,
      "learning_rate": 2.6290000000000004e-05,
      "loss": 0.3904,
      "step": 2371
    },
    {
      "epoch": 215.6787878787879,
      "grad_norm": 0.2920510768890381,
      "learning_rate": 2.628e-05,
      "loss": 0.3432,
      "step": 2372
    },
    {
      "epoch": 215.77575757575758,
      "grad_norm": 0.2948629856109619,
      "learning_rate": 2.627e-05,
      "loss": 0.3663,
      "step": 2373
    },
    {
      "epoch": 215.87272727272727,
      "grad_norm": 0.2976299226284027,
      "learning_rate": 2.6260000000000003e-05,
      "loss": 0.3917,
      "step": 2374
    },
    {
      "epoch": 215.96969696969697,
      "grad_norm": 0.2646479308605194,
      "learning_rate": 2.625e-05,
      "loss": 0.3427,
      "step": 2375
    },
    {
      "epoch": 216.0,
      "grad_norm": 0.5189900994300842,
      "learning_rate": 2.6240000000000003e-05,
      "loss": 0.386,
      "step": 2376
    },
    {
      "epoch": 216.0969696969697,
      "grad_norm": 0.28399553894996643,
      "learning_rate": 2.6229999999999998e-05,
      "loss": 0.3693,
      "step": 2377
    },
    {
      "epoch": 216.1939393939394,
      "grad_norm": 0.23609177768230438,
      "learning_rate": 2.622e-05,
      "loss": 0.3611,
      "step": 2378
    },
    {
      "epoch": 216.29090909090908,
      "grad_norm": 0.3684936463832855,
      "learning_rate": 2.621e-05,
      "loss": 0.3526,
      "step": 2379
    },
    {
      "epoch": 216.38787878787878,
      "grad_norm": 0.28553107380867004,
      "learning_rate": 2.6200000000000003e-05,
      "loss": 0.3951,
      "step": 2380
    },
    {
      "epoch": 216.38787878787878,
      "eval_loss": 0.44313111901283264,
      "eval_runtime": 2.1413,
      "eval_samples_per_second": 25.686,
      "eval_steps_per_second": 3.269,
      "step": 2380
    },
    {
      "epoch": 216.4848484848485,
      "grad_norm": 0.3319821357727051,
      "learning_rate": 2.6190000000000005e-05,
      "loss": 0.3954,
      "step": 2381
    },
    {
      "epoch": 216.5818181818182,
      "grad_norm": 0.3223443925380707,
      "learning_rate": 2.618e-05,
      "loss": 0.3477,
      "step": 2382
    },
    {
      "epoch": 216.6787878787879,
      "grad_norm": 0.28395891189575195,
      "learning_rate": 2.617e-05,
      "loss": 0.349,
      "step": 2383
    },
    {
      "epoch": 216.77575757575758,
      "grad_norm": 0.2875801920890808,
      "learning_rate": 2.616e-05,
      "loss": 0.383,
      "step": 2384
    },
    {
      "epoch": 216.87272727272727,
      "grad_norm": 0.2594863772392273,
      "learning_rate": 2.6150000000000002e-05,
      "loss": 0.3727,
      "step": 2385
    },
    {
      "epoch": 216.96969696969697,
      "grad_norm": 0.2604277431964874,
      "learning_rate": 2.6140000000000004e-05,
      "loss": 0.3848,
      "step": 2386
    },
    {
      "epoch": 217.0,
      "grad_norm": 0.5404815673828125,
      "learning_rate": 2.613e-05,
      "loss": 0.3486,
      "step": 2387
    },
    {
      "epoch": 217.0969696969697,
      "grad_norm": 0.3194997012615204,
      "learning_rate": 2.612e-05,
      "loss": 0.4172,
      "step": 2388
    },
    {
      "epoch": 217.1939393939394,
      "grad_norm": 0.29180049896240234,
      "learning_rate": 2.6110000000000002e-05,
      "loss": 0.3639,
      "step": 2389
    },
    {
      "epoch": 217.29090909090908,
      "grad_norm": 0.24527181684970856,
      "learning_rate": 2.61e-05,
      "loss": 0.3385,
      "step": 2390
    },
    {
      "epoch": 217.29090909090908,
      "eval_loss": 0.44199734926223755,
      "eval_runtime": 2.1449,
      "eval_samples_per_second": 25.642,
      "eval_steps_per_second": 3.263,
      "step": 2390
    },
    {
      "epoch": 217.38787878787878,
      "grad_norm": 0.3221138119697571,
      "learning_rate": 2.6090000000000003e-05,
      "loss": 0.3591,
      "step": 2391
    },
    {
      "epoch": 217.4848484848485,
      "grad_norm": 0.326957643032074,
      "learning_rate": 2.6079999999999998e-05,
      "loss": 0.3872,
      "step": 2392
    },
    {
      "epoch": 217.5818181818182,
      "grad_norm": 0.30668917298316956,
      "learning_rate": 2.607e-05,
      "loss": 0.3616,
      "step": 2393
    },
    {
      "epoch": 217.6787878787879,
      "grad_norm": 0.29979145526885986,
      "learning_rate": 2.606e-05,
      "loss": 0.3654,
      "step": 2394
    },
    {
      "epoch": 217.77575757575758,
      "grad_norm": 0.24426676332950592,
      "learning_rate": 2.6050000000000003e-05,
      "loss": 0.3708,
      "step": 2395
    },
    {
      "epoch": 217.87272727272727,
      "grad_norm": 0.32726359367370605,
      "learning_rate": 2.6040000000000005e-05,
      "loss": 0.3694,
      "step": 2396
    },
    {
      "epoch": 217.96969696969697,
      "grad_norm": 0.27767452597618103,
      "learning_rate": 2.603e-05,
      "loss": 0.3707,
      "step": 2397
    },
    {
      "epoch": 218.0,
      "grad_norm": 0.4858188331127167,
      "learning_rate": 2.602e-05,
      "loss": 0.3486,
      "step": 2398
    },
    {
      "epoch": 218.0969696969697,
      "grad_norm": 0.2870570123195648,
      "learning_rate": 2.601e-05,
      "loss": 0.3497,
      "step": 2399
    },
    {
      "epoch": 218.1939393939394,
      "grad_norm": 0.3260629177093506,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.3686,
      "step": 2400
    },
    {
      "epoch": 218.1939393939394,
      "eval_loss": 0.44243302941322327,
      "eval_runtime": 2.1191,
      "eval_samples_per_second": 25.954,
      "eval_steps_per_second": 3.303,
      "step": 2400
    },
    {
      "epoch": 218.29090909090908,
      "grad_norm": 0.2818470597267151,
      "learning_rate": 2.5990000000000004e-05,
      "loss": 0.3307,
      "step": 2401
    },
    {
      "epoch": 218.38787878787878,
      "grad_norm": 0.2645418047904968,
      "learning_rate": 2.598e-05,
      "loss": 0.3749,
      "step": 2402
    },
    {
      "epoch": 218.4848484848485,
      "grad_norm": 0.28982293605804443,
      "learning_rate": 2.597e-05,
      "loss": 0.3416,
      "step": 2403
    },
    {
      "epoch": 218.5818181818182,
      "grad_norm": 0.32691192626953125,
      "learning_rate": 2.5960000000000002e-05,
      "loss": 0.3879,
      "step": 2404
    },
    {
      "epoch": 218.6787878787879,
      "grad_norm": 0.39779746532440186,
      "learning_rate": 2.595e-05,
      "loss": 0.339,
      "step": 2405
    },
    {
      "epoch": 218.77575757575758,
      "grad_norm": 0.3154836893081665,
      "learning_rate": 2.5940000000000002e-05,
      "loss": 0.3836,
      "step": 2406
    },
    {
      "epoch": 218.87272727272727,
      "grad_norm": 0.26961442828178406,
      "learning_rate": 2.5929999999999997e-05,
      "loss": 0.4092,
      "step": 2407
    },
    {
      "epoch": 218.96969696969697,
      "grad_norm": 0.2830916941165924,
      "learning_rate": 2.592e-05,
      "loss": 0.3925,
      "step": 2408
    },
    {
      "epoch": 219.0,
      "grad_norm": 0.4212952256202698,
      "learning_rate": 2.591e-05,
      "loss": 0.4103,
      "step": 2409
    },
    {
      "epoch": 219.0969696969697,
      "grad_norm": 0.2978769540786743,
      "learning_rate": 2.5900000000000003e-05,
      "loss": 0.3848,
      "step": 2410
    },
    {
      "epoch": 219.0969696969697,
      "eval_loss": 0.4409344792366028,
      "eval_runtime": 2.1366,
      "eval_samples_per_second": 25.741,
      "eval_steps_per_second": 3.276,
      "step": 2410
    },
    {
      "epoch": 219.1939393939394,
      "grad_norm": 0.30883023142814636,
      "learning_rate": 2.5890000000000005e-05,
      "loss": 0.3975,
      "step": 2411
    },
    {
      "epoch": 219.29090909090908,
      "grad_norm": 0.28611019253730774,
      "learning_rate": 2.588e-05,
      "loss": 0.3675,
      "step": 2412
    },
    {
      "epoch": 219.38787878787878,
      "grad_norm": 0.31392866373062134,
      "learning_rate": 2.587e-05,
      "loss": 0.387,
      "step": 2413
    },
    {
      "epoch": 219.4848484848485,
      "grad_norm": 0.3225494623184204,
      "learning_rate": 2.586e-05,
      "loss": 0.3736,
      "step": 2414
    },
    {
      "epoch": 219.5818181818182,
      "grad_norm": 0.35923585295677185,
      "learning_rate": 2.585e-05,
      "loss": 0.386,
      "step": 2415
    },
    {
      "epoch": 219.6787878787879,
      "grad_norm": 0.3180157244205475,
      "learning_rate": 2.5840000000000003e-05,
      "loss": 0.3752,
      "step": 2416
    },
    {
      "epoch": 219.77575757575758,
      "grad_norm": 0.275451123714447,
      "learning_rate": 2.583e-05,
      "loss": 0.3636,
      "step": 2417
    },
    {
      "epoch": 219.87272727272727,
      "grad_norm": 0.25374045968055725,
      "learning_rate": 2.582e-05,
      "loss": 0.3155,
      "step": 2418
    },
    {
      "epoch": 219.96969696969697,
      "grad_norm": 0.2972322404384613,
      "learning_rate": 2.5810000000000002e-05,
      "loss": 0.3356,
      "step": 2419
    },
    {
      "epoch": 220.0,
      "grad_norm": 0.4347028434276581,
      "learning_rate": 2.58e-05,
      "loss": 0.3743,
      "step": 2420
    },
    {
      "epoch": 220.0,
      "eval_loss": 0.441646546125412,
      "eval_runtime": 2.1241,
      "eval_samples_per_second": 25.893,
      "eval_steps_per_second": 3.295,
      "step": 2420
    },
    {
      "epoch": 220.0969696969697,
      "grad_norm": 0.32299545407295227,
      "learning_rate": 2.5790000000000002e-05,
      "loss": 0.345,
      "step": 2421
    },
    {
      "epoch": 220.1939393939394,
      "grad_norm": 0.3762829601764679,
      "learning_rate": 2.5779999999999997e-05,
      "loss": 0.3814,
      "step": 2422
    },
    {
      "epoch": 220.29090909090908,
      "grad_norm": 0.2898702025413513,
      "learning_rate": 2.577e-05,
      "loss": 0.3791,
      "step": 2423
    },
    {
      "epoch": 220.38787878787878,
      "grad_norm": 0.267051637172699,
      "learning_rate": 2.576e-05,
      "loss": 0.3161,
      "step": 2424
    },
    {
      "epoch": 220.4848484848485,
      "grad_norm": 0.24987362325191498,
      "learning_rate": 2.5750000000000002e-05,
      "loss": 0.3901,
      "step": 2425
    },
    {
      "epoch": 220.5818181818182,
      "grad_norm": 0.26668789982795715,
      "learning_rate": 2.5740000000000004e-05,
      "loss": 0.3786,
      "step": 2426
    },
    {
      "epoch": 220.6787878787879,
      "grad_norm": 0.2954038679599762,
      "learning_rate": 2.573e-05,
      "loss": 0.3828,
      "step": 2427
    },
    {
      "epoch": 220.77575757575758,
      "grad_norm": 0.2518426775932312,
      "learning_rate": 2.572e-05,
      "loss": 0.3689,
      "step": 2428
    },
    {
      "epoch": 220.87272727272727,
      "grad_norm": 0.34907805919647217,
      "learning_rate": 2.571e-05,
      "loss": 0.3817,
      "step": 2429
    },
    {
      "epoch": 220.96969696969697,
      "grad_norm": 0.27415215969085693,
      "learning_rate": 2.57e-05,
      "loss": 0.3515,
      "step": 2430
    },
    {
      "epoch": 220.96969696969697,
      "eval_loss": 0.44070032238960266,
      "eval_runtime": 2.1477,
      "eval_samples_per_second": 25.609,
      "eval_steps_per_second": 3.259,
      "step": 2430
    },
    {
      "epoch": 221.0,
      "grad_norm": 0.4006612300872803,
      "learning_rate": 2.5690000000000003e-05,
      "loss": 0.397,
      "step": 2431
    },
    {
      "epoch": 221.0969696969697,
      "grad_norm": 0.2683582901954651,
      "learning_rate": 2.5679999999999998e-05,
      "loss": 0.3842,
      "step": 2432
    },
    {
      "epoch": 221.1939393939394,
      "grad_norm": 0.30595675110816956,
      "learning_rate": 2.567e-05,
      "loss": 0.3886,
      "step": 2433
    },
    {
      "epoch": 221.29090909090908,
      "grad_norm": 0.2578451931476593,
      "learning_rate": 2.566e-05,
      "loss": 0.3531,
      "step": 2434
    },
    {
      "epoch": 221.38787878787878,
      "grad_norm": 0.2802889049053192,
      "learning_rate": 2.5650000000000003e-05,
      "loss": 0.3652,
      "step": 2435
    },
    {
      "epoch": 221.4848484848485,
      "grad_norm": 0.3214050829410553,
      "learning_rate": 2.5640000000000002e-05,
      "loss": 0.3754,
      "step": 2436
    },
    {
      "epoch": 221.5818181818182,
      "grad_norm": 0.2835405766963959,
      "learning_rate": 2.5629999999999997e-05,
      "loss": 0.3837,
      "step": 2437
    },
    {
      "epoch": 221.6787878787879,
      "grad_norm": 0.24525125324726105,
      "learning_rate": 2.562e-05,
      "loss": 0.3381,
      "step": 2438
    },
    {
      "epoch": 221.77575757575758,
      "grad_norm": 0.31164395809173584,
      "learning_rate": 2.561e-05,
      "loss": 0.3975,
      "step": 2439
    },
    {
      "epoch": 221.87272727272727,
      "grad_norm": 0.2807594835758209,
      "learning_rate": 2.5600000000000002e-05,
      "loss": 0.351,
      "step": 2440
    },
    {
      "epoch": 221.87272727272727,
      "eval_loss": 0.4406813085079193,
      "eval_runtime": 2.1591,
      "eval_samples_per_second": 25.474,
      "eval_steps_per_second": 3.242,
      "step": 2440
    },
    {
      "epoch": 221.96969696969697,
      "grad_norm": 0.2830051779747009,
      "learning_rate": 2.5590000000000004e-05,
      "loss": 0.33,
      "step": 2441
    },
    {
      "epoch": 222.0,
      "grad_norm": 0.43788567185401917,
      "learning_rate": 2.5580000000000002e-05,
      "loss": 0.4056,
      "step": 2442
    },
    {
      "epoch": 222.0969696969697,
      "grad_norm": 0.26734405755996704,
      "learning_rate": 2.557e-05,
      "loss": 0.3465,
      "step": 2443
    },
    {
      "epoch": 222.1939393939394,
      "grad_norm": 0.28306853771209717,
      "learning_rate": 2.556e-05,
      "loss": 0.3666,
      "step": 2444
    },
    {
      "epoch": 222.29090909090908,
      "grad_norm": 0.2725483179092407,
      "learning_rate": 2.555e-05,
      "loss": 0.378,
      "step": 2445
    },
    {
      "epoch": 222.38787878787878,
      "grad_norm": 0.31050193309783936,
      "learning_rate": 2.5540000000000003e-05,
      "loss": 0.3966,
      "step": 2446
    },
    {
      "epoch": 222.4848484848485,
      "grad_norm": 0.2954963743686676,
      "learning_rate": 2.5530000000000005e-05,
      "loss": 0.3789,
      "step": 2447
    },
    {
      "epoch": 222.5818181818182,
      "grad_norm": 0.27009761333465576,
      "learning_rate": 2.552e-05,
      "loss": 0.3528,
      "step": 2448
    },
    {
      "epoch": 222.6787878787879,
      "grad_norm": 0.2876873314380646,
      "learning_rate": 2.551e-05,
      "loss": 0.3723,
      "step": 2449
    },
    {
      "epoch": 222.77575757575758,
      "grad_norm": 0.2781578302383423,
      "learning_rate": 2.5500000000000003e-05,
      "loss": 0.3448,
      "step": 2450
    },
    {
      "epoch": 222.77575757575758,
      "eval_loss": 0.4403097629547119,
      "eval_runtime": 2.1503,
      "eval_samples_per_second": 25.578,
      "eval_steps_per_second": 3.255,
      "step": 2450
    },
    {
      "epoch": 222.87272727272727,
      "grad_norm": 0.32250741124153137,
      "learning_rate": 2.549e-05,
      "loss": 0.3737,
      "step": 2451
    },
    {
      "epoch": 222.96969696969697,
      "grad_norm": 0.25998982787132263,
      "learning_rate": 2.5480000000000003e-05,
      "loss": 0.3628,
      "step": 2452
    },
    {
      "epoch": 223.0,
      "grad_norm": 0.6483612656593323,
      "learning_rate": 2.547e-05,
      "loss": 0.3658,
      "step": 2453
    },
    {
      "epoch": 223.0969696969697,
      "grad_norm": 0.27064916491508484,
      "learning_rate": 2.546e-05,
      "loss": 0.324,
      "step": 2454
    },
    {
      "epoch": 223.1939393939394,
      "grad_norm": 0.3874616324901581,
      "learning_rate": 2.5450000000000002e-05,
      "loss": 0.3746,
      "step": 2455
    },
    {
      "epoch": 223.29090909090908,
      "grad_norm": 0.2780890464782715,
      "learning_rate": 2.5440000000000004e-05,
      "loss": 0.3736,
      "step": 2456
    },
    {
      "epoch": 223.38787878787878,
      "grad_norm": 0.2915639877319336,
      "learning_rate": 2.5430000000000002e-05,
      "loss": 0.3724,
      "step": 2457
    },
    {
      "epoch": 223.4848484848485,
      "grad_norm": 0.2976807951927185,
      "learning_rate": 2.542e-05,
      "loss": 0.3596,
      "step": 2458
    },
    {
      "epoch": 223.5818181818182,
      "grad_norm": 0.27504071593284607,
      "learning_rate": 2.541e-05,
      "loss": 0.326,
      "step": 2459
    },
    {
      "epoch": 223.6787878787879,
      "grad_norm": 0.2684337794780731,
      "learning_rate": 2.54e-05,
      "loss": 0.3707,
      "step": 2460
    },
    {
      "epoch": 223.6787878787879,
      "eval_loss": 0.44031795859336853,
      "eval_runtime": 2.1228,
      "eval_samples_per_second": 25.91,
      "eval_steps_per_second": 3.298,
      "step": 2460
    },
    {
      "epoch": 223.77575757575758,
      "grad_norm": 0.2877907156944275,
      "learning_rate": 2.5390000000000003e-05,
      "loss": 0.3672,
      "step": 2461
    },
    {
      "epoch": 223.87272727272727,
      "grad_norm": 0.32059311866760254,
      "learning_rate": 2.5380000000000004e-05,
      "loss": 0.4111,
      "step": 2462
    },
    {
      "epoch": 223.96969696969697,
      "grad_norm": 0.27829068899154663,
      "learning_rate": 2.537e-05,
      "loss": 0.3862,
      "step": 2463
    },
    {
      "epoch": 224.0,
      "grad_norm": 0.5866357684135437,
      "learning_rate": 2.536e-05,
      "loss": 0.3742,
      "step": 2464
    },
    {
      "epoch": 224.0969696969697,
      "grad_norm": 0.2550725042819977,
      "learning_rate": 2.5350000000000003e-05,
      "loss": 0.3502,
      "step": 2465
    },
    {
      "epoch": 224.1939393939394,
      "grad_norm": 0.2818206250667572,
      "learning_rate": 2.534e-05,
      "loss": 0.3404,
      "step": 2466
    },
    {
      "epoch": 224.29090909090908,
      "grad_norm": 0.3186753988265991,
      "learning_rate": 2.5330000000000003e-05,
      "loss": 0.3526,
      "step": 2467
    },
    {
      "epoch": 224.38787878787878,
      "grad_norm": 0.2827731668949127,
      "learning_rate": 2.5319999999999998e-05,
      "loss": 0.3615,
      "step": 2468
    },
    {
      "epoch": 224.4848484848485,
      "grad_norm": 0.2541167438030243,
      "learning_rate": 2.531e-05,
      "loss": 0.3522,
      "step": 2469
    },
    {
      "epoch": 224.5818181818182,
      "grad_norm": 0.2639825642108917,
      "learning_rate": 2.5300000000000002e-05,
      "loss": 0.3718,
      "step": 2470
    },
    {
      "epoch": 224.5818181818182,
      "eval_loss": 0.43977347016334534,
      "eval_runtime": 2.1394,
      "eval_samples_per_second": 25.708,
      "eval_steps_per_second": 3.272,
      "step": 2470
    },
    {
      "epoch": 224.6787878787879,
      "grad_norm": 0.2934424579143524,
      "learning_rate": 2.5290000000000004e-05,
      "loss": 0.414,
      "step": 2471
    },
    {
      "epoch": 224.77575757575758,
      "grad_norm": 0.3221575915813446,
      "learning_rate": 2.5280000000000005e-05,
      "loss": 0.3758,
      "step": 2472
    },
    {
      "epoch": 224.87272727272727,
      "grad_norm": 0.2983260452747345,
      "learning_rate": 2.527e-05,
      "loss": 0.3965,
      "step": 2473
    },
    {
      "epoch": 224.96969696969697,
      "grad_norm": 0.28553321957588196,
      "learning_rate": 2.526e-05,
      "loss": 0.3533,
      "step": 2474
    },
    {
      "epoch": 225.0,
      "grad_norm": 0.5285137891769409,
      "learning_rate": 2.525e-05,
      "loss": 0.3625,
      "step": 2475
    },
    {
      "epoch": 225.0969696969697,
      "grad_norm": 0.2885406017303467,
      "learning_rate": 2.5240000000000002e-05,
      "loss": 0.3286,
      "step": 2476
    },
    {
      "epoch": 225.1939393939394,
      "grad_norm": 0.28469815850257874,
      "learning_rate": 2.5230000000000004e-05,
      "loss": 0.356,
      "step": 2477
    },
    {
      "epoch": 225.29090909090908,
      "grad_norm": 0.29516857862472534,
      "learning_rate": 2.522e-05,
      "loss": 0.328,
      "step": 2478
    },
    {
      "epoch": 225.38787878787878,
      "grad_norm": 0.2980527877807617,
      "learning_rate": 2.521e-05,
      "loss": 0.3609,
      "step": 2479
    },
    {
      "epoch": 225.4848484848485,
      "grad_norm": 0.32728075981140137,
      "learning_rate": 2.5200000000000003e-05,
      "loss": 0.3466,
      "step": 2480
    },
    {
      "epoch": 225.4848484848485,
      "eval_loss": 0.44033753871917725,
      "eval_runtime": 2.1249,
      "eval_samples_per_second": 25.884,
      "eval_steps_per_second": 3.294,
      "step": 2480
    },
    {
      "epoch": 225.5818181818182,
      "grad_norm": 0.36214345693588257,
      "learning_rate": 2.519e-05,
      "loss": 0.4097,
      "step": 2481
    },
    {
      "epoch": 225.6787878787879,
      "grad_norm": 0.38986948132514954,
      "learning_rate": 2.5180000000000003e-05,
      "loss": 0.3869,
      "step": 2482
    },
    {
      "epoch": 225.77575757575758,
      "grad_norm": 0.27240610122680664,
      "learning_rate": 2.5169999999999998e-05,
      "loss": 0.3773,
      "step": 2483
    },
    {
      "epoch": 225.87272727272727,
      "grad_norm": 0.2969878017902374,
      "learning_rate": 2.516e-05,
      "loss": 0.4009,
      "step": 2484
    },
    {
      "epoch": 225.96969696969697,
      "grad_norm": 0.3425680994987488,
      "learning_rate": 2.515e-05,
      "loss": 0.3723,
      "step": 2485
    },
    {
      "epoch": 226.0,
      "grad_norm": 0.47936302423477173,
      "learning_rate": 2.5140000000000003e-05,
      "loss": 0.3564,
      "step": 2486
    },
    {
      "epoch": 226.0969696969697,
      "grad_norm": 0.3240700364112854,
      "learning_rate": 2.5130000000000005e-05,
      "loss": 0.3674,
      "step": 2487
    },
    {
      "epoch": 226.1939393939394,
      "grad_norm": 0.2965087890625,
      "learning_rate": 2.512e-05,
      "loss": 0.3758,
      "step": 2488
    },
    {
      "epoch": 226.29090909090908,
      "grad_norm": 0.2678585350513458,
      "learning_rate": 2.5110000000000002e-05,
      "loss": 0.3782,
      "step": 2489
    },
    {
      "epoch": 226.38787878787878,
      "grad_norm": 0.3514326512813568,
      "learning_rate": 2.51e-05,
      "loss": 0.2942,
      "step": 2490
    },
    {
      "epoch": 226.38787878787878,
      "eval_loss": 0.4388936758041382,
      "eval_runtime": 2.1285,
      "eval_samples_per_second": 25.84,
      "eval_steps_per_second": 3.289,
      "step": 2490
    },
    {
      "epoch": 226.4848484848485,
      "grad_norm": 0.343416303396225,
      "learning_rate": 2.5090000000000002e-05,
      "loss": 0.3645,
      "step": 2491
    },
    {
      "epoch": 226.5818181818182,
      "grad_norm": 0.30410099029541016,
      "learning_rate": 2.5080000000000004e-05,
      "loss": 0.3294,
      "step": 2492
    },
    {
      "epoch": 226.6787878787879,
      "grad_norm": 0.26536622643470764,
      "learning_rate": 2.507e-05,
      "loss": 0.3534,
      "step": 2493
    },
    {
      "epoch": 226.77575757575758,
      "grad_norm": 0.27707982063293457,
      "learning_rate": 2.506e-05,
      "loss": 0.401,
      "step": 2494
    },
    {
      "epoch": 226.87272727272727,
      "grad_norm": 0.294041246175766,
      "learning_rate": 2.5050000000000002e-05,
      "loss": 0.4049,
      "step": 2495
    },
    {
      "epoch": 226.96969696969697,
      "grad_norm": 0.31618446111679077,
      "learning_rate": 2.504e-05,
      "loss": 0.4065,
      "step": 2496
    },
    {
      "epoch": 227.0,
      "grad_norm": 0.6172340512275696,
      "learning_rate": 2.5030000000000003e-05,
      "loss": 0.3122,
      "step": 2497
    },
    {
      "epoch": 227.0969696969697,
      "grad_norm": 0.24138444662094116,
      "learning_rate": 2.5019999999999998e-05,
      "loss": 0.3704,
      "step": 2498
    },
    {
      "epoch": 227.1939393939394,
      "grad_norm": 0.33306029438972473,
      "learning_rate": 2.501e-05,
      "loss": 0.4177,
      "step": 2499
    },
    {
      "epoch": 227.29090909090908,
      "grad_norm": 0.40666037797927856,
      "learning_rate": 2.5e-05,
      "loss": 0.3513,
      "step": 2500
    },
    {
      "epoch": 227.29090909090908,
      "eval_loss": 0.4396165609359741,
      "eval_runtime": 2.1224,
      "eval_samples_per_second": 25.914,
      "eval_steps_per_second": 3.298,
      "step": 2500
    },
    {
      "epoch": 227.38787878787878,
      "grad_norm": 0.29784542322158813,
      "learning_rate": 2.4990000000000003e-05,
      "loss": 0.3711,
      "step": 2501
    },
    {
      "epoch": 227.4848484848485,
      "grad_norm": 0.46334195137023926,
      "learning_rate": 2.498e-05,
      "loss": 0.3346,
      "step": 2502
    },
    {
      "epoch": 227.5818181818182,
      "grad_norm": 0.3165861964225769,
      "learning_rate": 2.4970000000000003e-05,
      "loss": 0.333,
      "step": 2503
    },
    {
      "epoch": 227.6787878787879,
      "grad_norm": 0.30741187930107117,
      "learning_rate": 2.496e-05,
      "loss": 0.372,
      "step": 2504
    },
    {
      "epoch": 227.77575757575758,
      "grad_norm": 0.2982431948184967,
      "learning_rate": 2.495e-05,
      "loss": 0.3405,
      "step": 2505
    },
    {
      "epoch": 227.87272727272727,
      "grad_norm": 0.32411104440689087,
      "learning_rate": 2.4940000000000002e-05,
      "loss": 0.4046,
      "step": 2506
    },
    {
      "epoch": 227.96969696969697,
      "grad_norm": 0.3008289635181427,
      "learning_rate": 2.493e-05,
      "loss": 0.3571,
      "step": 2507
    },
    {
      "epoch": 228.0,
      "grad_norm": 0.4273081123828888,
      "learning_rate": 2.4920000000000002e-05,
      "loss": 0.3727,
      "step": 2508
    },
    {
      "epoch": 228.0969696969697,
      "grad_norm": 0.29499098658561707,
      "learning_rate": 2.491e-05,
      "loss": 0.386,
      "step": 2509
    },
    {
      "epoch": 228.1939393939394,
      "grad_norm": 0.263407438993454,
      "learning_rate": 2.4900000000000002e-05,
      "loss": 0.3541,
      "step": 2510
    },
    {
      "epoch": 228.1939393939394,
      "eval_loss": 0.43920692801475525,
      "eval_runtime": 2.114,
      "eval_samples_per_second": 26.018,
      "eval_steps_per_second": 3.311,
      "step": 2510
    },
    {
      "epoch": 228.29090909090908,
      "grad_norm": 0.2716670036315918,
      "learning_rate": 2.489e-05,
      "loss": 0.3323,
      "step": 2511
    },
    {
      "epoch": 228.38787878787878,
      "grad_norm": 0.3222666084766388,
      "learning_rate": 2.488e-05,
      "loss": 0.3869,
      "step": 2512
    },
    {
      "epoch": 228.4848484848485,
      "grad_norm": 0.2719743251800537,
      "learning_rate": 2.487e-05,
      "loss": 0.3589,
      "step": 2513
    },
    {
      "epoch": 228.5818181818182,
      "grad_norm": 0.35152000188827515,
      "learning_rate": 2.486e-05,
      "loss": 0.3539,
      "step": 2514
    },
    {
      "epoch": 228.6787878787879,
      "grad_norm": 0.2843145728111267,
      "learning_rate": 2.485e-05,
      "loss": 0.3452,
      "step": 2515
    },
    {
      "epoch": 228.77575757575758,
      "grad_norm": 0.293863981962204,
      "learning_rate": 2.4840000000000003e-05,
      "loss": 0.3967,
      "step": 2516
    },
    {
      "epoch": 228.87272727272727,
      "grad_norm": 0.26122814416885376,
      "learning_rate": 2.483e-05,
      "loss": 0.3519,
      "step": 2517
    },
    {
      "epoch": 228.96969696969697,
      "grad_norm": 0.28403446078300476,
      "learning_rate": 2.4820000000000003e-05,
      "loss": 0.3486,
      "step": 2518
    },
    {
      "epoch": 229.0,
      "grad_norm": 0.5891252756118774,
      "learning_rate": 2.481e-05,
      "loss": 0.4587,
      "step": 2519
    },
    {
      "epoch": 229.0969696969697,
      "grad_norm": 0.26723888516426086,
      "learning_rate": 2.48e-05,
      "loss": 0.3911,
      "step": 2520
    },
    {
      "epoch": 229.0969696969697,
      "eval_loss": 0.43881770968437195,
      "eval_runtime": 2.132,
      "eval_samples_per_second": 25.797,
      "eval_steps_per_second": 3.283,
      "step": 2520
    },
    {
      "epoch": 229.1939393939394,
      "grad_norm": 0.30375170707702637,
      "learning_rate": 2.479e-05,
      "loss": 0.3853,
      "step": 2521
    },
    {
      "epoch": 229.29090909090908,
      "grad_norm": 0.2493750900030136,
      "learning_rate": 2.478e-05,
      "loss": 0.364,
      "step": 2522
    },
    {
      "epoch": 229.38787878787878,
      "grad_norm": 0.2770415246486664,
      "learning_rate": 2.4770000000000002e-05,
      "loss": 0.3524,
      "step": 2523
    },
    {
      "epoch": 229.4848484848485,
      "grad_norm": 0.3257941007614136,
      "learning_rate": 2.476e-05,
      "loss": 0.3724,
      "step": 2524
    },
    {
      "epoch": 229.5818181818182,
      "grad_norm": 0.29938164353370667,
      "learning_rate": 2.4750000000000002e-05,
      "loss": 0.3349,
      "step": 2525
    },
    {
      "epoch": 229.6787878787879,
      "grad_norm": 0.28248274326324463,
      "learning_rate": 2.4740000000000004e-05,
      "loss": 0.3504,
      "step": 2526
    },
    {
      "epoch": 229.77575757575758,
      "grad_norm": 0.27092963457107544,
      "learning_rate": 2.473e-05,
      "loss": 0.3698,
      "step": 2527
    },
    {
      "epoch": 229.87272727272727,
      "grad_norm": 0.28257593512535095,
      "learning_rate": 2.472e-05,
      "loss": 0.3456,
      "step": 2528
    },
    {
      "epoch": 229.96969696969697,
      "grad_norm": 0.2728407680988312,
      "learning_rate": 2.471e-05,
      "loss": 0.3744,
      "step": 2529
    },
    {
      "epoch": 230.0,
      "grad_norm": 0.6111608147621155,
      "learning_rate": 2.47e-05,
      "loss": 0.378,
      "step": 2530
    },
    {
      "epoch": 230.0,
      "eval_loss": 0.438978910446167,
      "eval_runtime": 2.1304,
      "eval_samples_per_second": 25.817,
      "eval_steps_per_second": 3.286,
      "step": 2530
    },
    {
      "epoch": 230.0969696969697,
      "grad_norm": 0.2594529688358307,
      "learning_rate": 2.4690000000000002e-05,
      "loss": 0.369,
      "step": 2531
    },
    {
      "epoch": 230.1939393939394,
      "grad_norm": 0.25597700476646423,
      "learning_rate": 2.468e-05,
      "loss": 0.344,
      "step": 2532
    },
    {
      "epoch": 230.29090909090908,
      "grad_norm": 0.307706356048584,
      "learning_rate": 2.4670000000000003e-05,
      "loss": 0.3455,
      "step": 2533
    },
    {
      "epoch": 230.38787878787878,
      "grad_norm": 0.2459164261817932,
      "learning_rate": 2.466e-05,
      "loss": 0.3559,
      "step": 2534
    },
    {
      "epoch": 230.4848484848485,
      "grad_norm": 0.3313330113887787,
      "learning_rate": 2.465e-05,
      "loss": 0.3781,
      "step": 2535
    },
    {
      "epoch": 230.5818181818182,
      "grad_norm": 0.3011106550693512,
      "learning_rate": 2.464e-05,
      "loss": 0.3497,
      "step": 2536
    },
    {
      "epoch": 230.6787878787879,
      "grad_norm": 0.3560205399990082,
      "learning_rate": 2.463e-05,
      "loss": 0.3793,
      "step": 2537
    },
    {
      "epoch": 230.77575757575758,
      "grad_norm": 0.2969333529472351,
      "learning_rate": 2.462e-05,
      "loss": 0.3979,
      "step": 2538
    },
    {
      "epoch": 230.87272727272727,
      "grad_norm": 0.32418274879455566,
      "learning_rate": 2.4610000000000003e-05,
      "loss": 0.3774,
      "step": 2539
    },
    {
      "epoch": 230.96969696969697,
      "grad_norm": 0.27839162945747375,
      "learning_rate": 2.46e-05,
      "loss": 0.3585,
      "step": 2540
    },
    {
      "epoch": 230.96969696969697,
      "eval_loss": 0.4386548399925232,
      "eval_runtime": 2.1313,
      "eval_samples_per_second": 25.806,
      "eval_steps_per_second": 3.284,
      "step": 2540
    },
    {
      "epoch": 231.0,
      "grad_norm": 0.6563290953636169,
      "learning_rate": 2.4590000000000003e-05,
      "loss": 0.3148,
      "step": 2541
    },
    {
      "epoch": 231.0969696969697,
      "grad_norm": 0.28322136402130127,
      "learning_rate": 2.4580000000000002e-05,
      "loss": 0.3882,
      "step": 2542
    },
    {
      "epoch": 231.1939393939394,
      "grad_norm": 0.4256789982318878,
      "learning_rate": 2.457e-05,
      "loss": 0.3609,
      "step": 2543
    },
    {
      "epoch": 231.29090909090908,
      "grad_norm": 0.30935171246528625,
      "learning_rate": 2.4560000000000002e-05,
      "loss": 0.3702,
      "step": 2544
    },
    {
      "epoch": 231.38787878787878,
      "grad_norm": 0.3351457417011261,
      "learning_rate": 2.455e-05,
      "loss": 0.3499,
      "step": 2545
    },
    {
      "epoch": 231.4848484848485,
      "grad_norm": 0.38497939705848694,
      "learning_rate": 2.4540000000000002e-05,
      "loss": 0.3414,
      "step": 2546
    },
    {
      "epoch": 231.5818181818182,
      "grad_norm": 0.33501312136650085,
      "learning_rate": 2.453e-05,
      "loss": 0.3999,
      "step": 2547
    },
    {
      "epoch": 231.6787878787879,
      "grad_norm": 0.321706086397171,
      "learning_rate": 2.4520000000000002e-05,
      "loss": 0.3548,
      "step": 2548
    },
    {
      "epoch": 231.77575757575758,
      "grad_norm": 0.2958731949329376,
      "learning_rate": 2.451e-05,
      "loss": 0.3296,
      "step": 2549
    },
    {
      "epoch": 231.87272727272727,
      "grad_norm": 0.2821996510028839,
      "learning_rate": 2.45e-05,
      "loss": 0.3653,
      "step": 2550
    },
    {
      "epoch": 231.87272727272727,
      "eval_loss": 0.43812915682792664,
      "eval_runtime": 2.134,
      "eval_samples_per_second": 25.773,
      "eval_steps_per_second": 3.28,
      "step": 2550
    },
    {
      "epoch": 231.96969696969697,
      "grad_norm": 0.26832205057144165,
      "learning_rate": 2.449e-05,
      "loss": 0.3869,
      "step": 2551
    },
    {
      "epoch": 232.0,
      "grad_norm": 0.442310094833374,
      "learning_rate": 2.448e-05,
      "loss": 0.3305,
      "step": 2552
    },
    {
      "epoch": 232.0969696969697,
      "grad_norm": 0.2917024791240692,
      "learning_rate": 2.447e-05,
      "loss": 0.3526,
      "step": 2553
    },
    {
      "epoch": 232.1939393939394,
      "grad_norm": 0.2630578279495239,
      "learning_rate": 2.4460000000000003e-05,
      "loss": 0.3713,
      "step": 2554
    },
    {
      "epoch": 232.29090909090908,
      "grad_norm": 0.2642268240451813,
      "learning_rate": 2.445e-05,
      "loss": 0.3782,
      "step": 2555
    },
    {
      "epoch": 232.38787878787878,
      "grad_norm": 0.33103933930397034,
      "learning_rate": 2.4440000000000003e-05,
      "loss": 0.4088,
      "step": 2556
    },
    {
      "epoch": 232.4848484848485,
      "grad_norm": 0.33370935916900635,
      "learning_rate": 2.443e-05,
      "loss": 0.352,
      "step": 2557
    },
    {
      "epoch": 232.5818181818182,
      "grad_norm": 0.31459617614746094,
      "learning_rate": 2.442e-05,
      "loss": 0.3471,
      "step": 2558
    },
    {
      "epoch": 232.6787878787879,
      "grad_norm": 0.315646231174469,
      "learning_rate": 2.4410000000000002e-05,
      "loss": 0.3721,
      "step": 2559
    },
    {
      "epoch": 232.77575757575758,
      "grad_norm": 0.27737879753112793,
      "learning_rate": 2.44e-05,
      "loss": 0.3508,
      "step": 2560
    },
    {
      "epoch": 232.77575757575758,
      "eval_loss": 0.43808242678642273,
      "eval_runtime": 2.1296,
      "eval_samples_per_second": 25.827,
      "eval_steps_per_second": 3.287,
      "step": 2560
    },
    {
      "epoch": 232.87272727272727,
      "grad_norm": 0.32893404364585876,
      "learning_rate": 2.4390000000000002e-05,
      "loss": 0.3614,
      "step": 2561
    },
    {
      "epoch": 232.96969696969697,
      "grad_norm": 0.28458207845687866,
      "learning_rate": 2.438e-05,
      "loss": 0.3461,
      "step": 2562
    },
    {
      "epoch": 233.0,
      "grad_norm": 0.5089003443717957,
      "learning_rate": 2.4370000000000002e-05,
      "loss": 0.3414,
      "step": 2563
    },
    {
      "epoch": 233.0969696969697,
      "grad_norm": 0.32916566729545593,
      "learning_rate": 2.4360000000000004e-05,
      "loss": 0.3614,
      "step": 2564
    },
    {
      "epoch": 233.1939393939394,
      "grad_norm": 0.30260559916496277,
      "learning_rate": 2.435e-05,
      "loss": 0.363,
      "step": 2565
    },
    {
      "epoch": 233.29090909090908,
      "grad_norm": 0.2912161350250244,
      "learning_rate": 2.434e-05,
      "loss": 0.365,
      "step": 2566
    },
    {
      "epoch": 233.38787878787878,
      "grad_norm": 0.2847774922847748,
      "learning_rate": 2.433e-05,
      "loss": 0.3574,
      "step": 2567
    },
    {
      "epoch": 233.4848484848485,
      "grad_norm": 0.3008267879486084,
      "learning_rate": 2.432e-05,
      "loss": 0.3781,
      "step": 2568
    },
    {
      "epoch": 233.5818181818182,
      "grad_norm": 0.29215264320373535,
      "learning_rate": 2.4310000000000003e-05,
      "loss": 0.3462,
      "step": 2569
    },
    {
      "epoch": 233.6787878787879,
      "grad_norm": 0.39694666862487793,
      "learning_rate": 2.43e-05,
      "loss": 0.3789,
      "step": 2570
    },
    {
      "epoch": 233.6787878787879,
      "eval_loss": 0.4377782642841339,
      "eval_runtime": 2.1486,
      "eval_samples_per_second": 25.599,
      "eval_steps_per_second": 3.258,
      "step": 2570
    },
    {
      "epoch": 233.77575757575758,
      "grad_norm": 0.2967980206012726,
      "learning_rate": 2.4290000000000003e-05,
      "loss": 0.3664,
      "step": 2571
    },
    {
      "epoch": 233.87272727272727,
      "grad_norm": 0.2728116512298584,
      "learning_rate": 2.428e-05,
      "loss": 0.3658,
      "step": 2572
    },
    {
      "epoch": 233.96969696969697,
      "grad_norm": 0.3128872811794281,
      "learning_rate": 2.427e-05,
      "loss": 0.3539,
      "step": 2573
    },
    {
      "epoch": 234.0,
      "grad_norm": 0.36063623428344727,
      "learning_rate": 2.426e-05,
      "loss": 0.3449,
      "step": 2574
    },
    {
      "epoch": 234.0969696969697,
      "grad_norm": 0.24417349696159363,
      "learning_rate": 2.425e-05,
      "loss": 0.3253,
      "step": 2575
    },
    {
      "epoch": 234.1939393939394,
      "grad_norm": 0.40090200304985046,
      "learning_rate": 2.4240000000000002e-05,
      "loss": 0.3953,
      "step": 2576
    },
    {
      "epoch": 234.29090909090908,
      "grad_norm": 0.2958609461784363,
      "learning_rate": 2.423e-05,
      "loss": 0.3593,
      "step": 2577
    },
    {
      "epoch": 234.38787878787878,
      "grad_norm": 0.2797829210758209,
      "learning_rate": 2.4220000000000002e-05,
      "loss": 0.3864,
      "step": 2578
    },
    {
      "epoch": 234.4848484848485,
      "grad_norm": 0.2914266288280487,
      "learning_rate": 2.4210000000000004e-05,
      "loss": 0.3434,
      "step": 2579
    },
    {
      "epoch": 234.5818181818182,
      "grad_norm": 0.27718478441238403,
      "learning_rate": 2.4200000000000002e-05,
      "loss": 0.3629,
      "step": 2580
    },
    {
      "epoch": 234.5818181818182,
      "eval_loss": 0.43738406896591187,
      "eval_runtime": 2.1359,
      "eval_samples_per_second": 25.751,
      "eval_steps_per_second": 3.277,
      "step": 2580
    },
    {
      "epoch": 234.6787878787879,
      "grad_norm": 0.2699889540672302,
      "learning_rate": 2.419e-05,
      "loss": 0.3434,
      "step": 2581
    },
    {
      "epoch": 234.77575757575758,
      "grad_norm": 0.2644093930721283,
      "learning_rate": 2.418e-05,
      "loss": 0.3681,
      "step": 2582
    },
    {
      "epoch": 234.87272727272727,
      "grad_norm": 0.2939375638961792,
      "learning_rate": 2.417e-05,
      "loss": 0.3763,
      "step": 2583
    },
    {
      "epoch": 234.96969696969697,
      "grad_norm": 0.2855975925922394,
      "learning_rate": 2.4160000000000002e-05,
      "loss": 0.359,
      "step": 2584
    },
    {
      "epoch": 235.0,
      "grad_norm": 0.4643716514110565,
      "learning_rate": 2.415e-05,
      "loss": 0.3744,
      "step": 2585
    },
    {
      "epoch": 235.0969696969697,
      "grad_norm": 0.34113290905952454,
      "learning_rate": 2.4140000000000003e-05,
      "loss": 0.4155,
      "step": 2586
    },
    {
      "epoch": 235.1939393939394,
      "grad_norm": 0.273038774728775,
      "learning_rate": 2.413e-05,
      "loss": 0.3692,
      "step": 2587
    },
    {
      "epoch": 235.29090909090908,
      "grad_norm": 0.3686050772666931,
      "learning_rate": 2.412e-05,
      "loss": 0.3418,
      "step": 2588
    },
    {
      "epoch": 235.38787878787878,
      "grad_norm": 0.27432775497436523,
      "learning_rate": 2.411e-05,
      "loss": 0.3567,
      "step": 2589
    },
    {
      "epoch": 235.4848484848485,
      "grad_norm": 0.27832579612731934,
      "learning_rate": 2.41e-05,
      "loss": 0.3441,
      "step": 2590
    },
    {
      "epoch": 235.4848484848485,
      "eval_loss": 0.4376099109649658,
      "eval_runtime": 2.1414,
      "eval_samples_per_second": 25.685,
      "eval_steps_per_second": 3.269,
      "step": 2590
    },
    {
      "epoch": 235.5818181818182,
      "grad_norm": 0.29833900928497314,
      "learning_rate": 2.409e-05,
      "loss": 0.3467,
      "step": 2591
    },
    {
      "epoch": 235.6787878787879,
      "grad_norm": 0.29584160447120667,
      "learning_rate": 2.408e-05,
      "loss": 0.377,
      "step": 2592
    },
    {
      "epoch": 235.77575757575758,
      "grad_norm": 0.3445207178592682,
      "learning_rate": 2.407e-05,
      "loss": 0.3735,
      "step": 2593
    },
    {
      "epoch": 235.87272727272727,
      "grad_norm": 0.304282009601593,
      "learning_rate": 2.4060000000000003e-05,
      "loss": 0.3469,
      "step": 2594
    },
    {
      "epoch": 235.96969696969697,
      "grad_norm": 0.24318461120128632,
      "learning_rate": 2.4050000000000002e-05,
      "loss": 0.3524,
      "step": 2595
    },
    {
      "epoch": 236.0,
      "grad_norm": 0.5731469392776489,
      "learning_rate": 2.404e-05,
      "loss": 0.3424,
      "step": 2596
    },
    {
      "epoch": 236.0969696969697,
      "grad_norm": 0.2819751501083374,
      "learning_rate": 2.4030000000000002e-05,
      "loss": 0.3142,
      "step": 2597
    },
    {
      "epoch": 236.1939393939394,
      "grad_norm": 0.2591220438480377,
      "learning_rate": 2.402e-05,
      "loss": 0.3156,
      "step": 2598
    },
    {
      "epoch": 236.29090909090908,
      "grad_norm": 0.29000306129455566,
      "learning_rate": 2.4010000000000002e-05,
      "loss": 0.3941,
      "step": 2599
    },
    {
      "epoch": 236.38787878787878,
      "grad_norm": 0.2846730947494507,
      "learning_rate": 2.4e-05,
      "loss": 0.3397,
      "step": 2600
    },
    {
      "epoch": 236.38787878787878,
      "eval_loss": 0.437418133020401,
      "eval_runtime": 2.123,
      "eval_samples_per_second": 25.907,
      "eval_steps_per_second": 3.297,
      "step": 2600
    },
    {
      "epoch": 236.4848484848485,
      "grad_norm": 0.32599303126335144,
      "learning_rate": 2.3990000000000002e-05,
      "loss": 0.3638,
      "step": 2601
    },
    {
      "epoch": 236.5818181818182,
      "grad_norm": 0.2827683687210083,
      "learning_rate": 2.398e-05,
      "loss": 0.3915,
      "step": 2602
    },
    {
      "epoch": 236.6787878787879,
      "grad_norm": 0.26976168155670166,
      "learning_rate": 2.397e-05,
      "loss": 0.366,
      "step": 2603
    },
    {
      "epoch": 236.77575757575758,
      "grad_norm": 0.33612337708473206,
      "learning_rate": 2.396e-05,
      "loss": 0.3472,
      "step": 2604
    },
    {
      "epoch": 236.87272727272727,
      "grad_norm": 0.24638354778289795,
      "learning_rate": 2.395e-05,
      "loss": 0.3924,
      "step": 2605
    },
    {
      "epoch": 236.96969696969697,
      "grad_norm": 0.3176722228527069,
      "learning_rate": 2.394e-05,
      "loss": 0.3878,
      "step": 2606
    },
    {
      "epoch": 237.0,
      "grad_norm": 0.49755457043647766,
      "learning_rate": 2.3930000000000003e-05,
      "loss": 0.3706,
      "step": 2607
    },
    {
      "epoch": 237.0969696969697,
      "grad_norm": 0.26535603404045105,
      "learning_rate": 2.392e-05,
      "loss": 0.3736,
      "step": 2608
    },
    {
      "epoch": 237.1939393939394,
      "grad_norm": 0.2597862184047699,
      "learning_rate": 2.3910000000000003e-05,
      "loss": 0.3434,
      "step": 2609
    },
    {
      "epoch": 237.29090909090908,
      "grad_norm": 0.30801284313201904,
      "learning_rate": 2.39e-05,
      "loss": 0.3423,
      "step": 2610
    },
    {
      "epoch": 237.29090909090908,
      "eval_loss": 0.4371388256549835,
      "eval_runtime": 2.1469,
      "eval_samples_per_second": 25.618,
      "eval_steps_per_second": 3.26,
      "step": 2610
    },
    {
      "epoch": 237.38787878787878,
      "grad_norm": 0.29568618535995483,
      "learning_rate": 2.389e-05,
      "loss": 0.353,
      "step": 2611
    },
    {
      "epoch": 237.4848484848485,
      "grad_norm": 0.2596912384033203,
      "learning_rate": 2.3880000000000002e-05,
      "loss": 0.394,
      "step": 2612
    },
    {
      "epoch": 237.5818181818182,
      "grad_norm": 0.2487613558769226,
      "learning_rate": 2.387e-05,
      "loss": 0.3593,
      "step": 2613
    },
    {
      "epoch": 237.6787878787879,
      "grad_norm": 0.30508145689964294,
      "learning_rate": 2.3860000000000002e-05,
      "loss": 0.3823,
      "step": 2614
    },
    {
      "epoch": 237.77575757575758,
      "grad_norm": 0.3193148970603943,
      "learning_rate": 2.385e-05,
      "loss": 0.3531,
      "step": 2615
    },
    {
      "epoch": 237.87272727272727,
      "grad_norm": 0.3358408808708191,
      "learning_rate": 2.3840000000000002e-05,
      "loss": 0.3581,
      "step": 2616
    },
    {
      "epoch": 237.96969696969697,
      "grad_norm": 0.3238699436187744,
      "learning_rate": 2.3830000000000004e-05,
      "loss": 0.3644,
      "step": 2617
    },
    {
      "epoch": 238.0,
      "grad_norm": 0.5226176381111145,
      "learning_rate": 2.3820000000000002e-05,
      "loss": 0.3302,
      "step": 2618
    },
    {
      "epoch": 238.0969696969697,
      "grad_norm": 0.2687991261482239,
      "learning_rate": 2.381e-05,
      "loss": 0.3478,
      "step": 2619
    },
    {
      "epoch": 238.1939393939394,
      "grad_norm": 0.2733170688152313,
      "learning_rate": 2.38e-05,
      "loss": 0.3572,
      "step": 2620
    },
    {
      "epoch": 238.1939393939394,
      "eval_loss": 0.4368661344051361,
      "eval_runtime": 2.1339,
      "eval_samples_per_second": 25.774,
      "eval_steps_per_second": 3.28,
      "step": 2620
    },
    {
      "epoch": 238.29090909090908,
      "grad_norm": 0.2491229921579361,
      "learning_rate": 2.379e-05,
      "loss": 0.3534,
      "step": 2621
    },
    {
      "epoch": 238.38787878787878,
      "grad_norm": 0.3663552701473236,
      "learning_rate": 2.3780000000000003e-05,
      "loss": 0.3496,
      "step": 2622
    },
    {
      "epoch": 238.4848484848485,
      "grad_norm": 0.28013721108436584,
      "learning_rate": 2.377e-05,
      "loss": 0.3667,
      "step": 2623
    },
    {
      "epoch": 238.5818181818182,
      "grad_norm": 0.3751683533191681,
      "learning_rate": 2.3760000000000003e-05,
      "loss": 0.3921,
      "step": 2624
    },
    {
      "epoch": 238.6787878787879,
      "grad_norm": 0.2832871675491333,
      "learning_rate": 2.375e-05,
      "loss": 0.3444,
      "step": 2625
    },
    {
      "epoch": 238.77575757575758,
      "grad_norm": 0.3930876851081848,
      "learning_rate": 2.374e-05,
      "loss": 0.3599,
      "step": 2626
    },
    {
      "epoch": 238.87272727272727,
      "grad_norm": 0.3104434013366699,
      "learning_rate": 2.373e-05,
      "loss": 0.351,
      "step": 2627
    },
    {
      "epoch": 238.96969696969697,
      "grad_norm": 0.36164671182632446,
      "learning_rate": 2.372e-05,
      "loss": 0.388,
      "step": 2628
    },
    {
      "epoch": 239.0,
      "grad_norm": 0.4879319965839386,
      "learning_rate": 2.371e-05,
      "loss": 0.362,
      "step": 2629
    },
    {
      "epoch": 239.0969696969697,
      "grad_norm": 0.32071059942245483,
      "learning_rate": 2.37e-05,
      "loss": 0.3739,
      "step": 2630
    },
    {
      "epoch": 239.0969696969697,
      "eval_loss": 0.43735435605049133,
      "eval_runtime": 2.1416,
      "eval_samples_per_second": 25.682,
      "eval_steps_per_second": 3.269,
      "step": 2630
    },
    {
      "epoch": 239.1939393939394,
      "grad_norm": 0.27180489897727966,
      "learning_rate": 2.3690000000000002e-05,
      "loss": 0.3268,
      "step": 2631
    },
    {
      "epoch": 239.29090909090908,
      "grad_norm": 0.44245511293411255,
      "learning_rate": 2.3680000000000004e-05,
      "loss": 0.3708,
      "step": 2632
    },
    {
      "epoch": 239.38787878787878,
      "grad_norm": 0.350221186876297,
      "learning_rate": 2.3670000000000002e-05,
      "loss": 0.3692,
      "step": 2633
    },
    {
      "epoch": 239.4848484848485,
      "grad_norm": 0.34308651089668274,
      "learning_rate": 2.366e-05,
      "loss": 0.3924,
      "step": 2634
    },
    {
      "epoch": 239.5818181818182,
      "grad_norm": 0.29386603832244873,
      "learning_rate": 2.365e-05,
      "loss": 0.3458,
      "step": 2635
    },
    {
      "epoch": 239.6787878787879,
      "grad_norm": 0.29850178956985474,
      "learning_rate": 2.364e-05,
      "loss": 0.3525,
      "step": 2636
    },
    {
      "epoch": 239.77575757575758,
      "grad_norm": 0.302652508020401,
      "learning_rate": 2.3630000000000002e-05,
      "loss": 0.3375,
      "step": 2637
    },
    {
      "epoch": 239.87272727272727,
      "grad_norm": 0.2891862392425537,
      "learning_rate": 2.362e-05,
      "loss": 0.3687,
      "step": 2638
    },
    {
      "epoch": 239.96969696969697,
      "grad_norm": 0.3072037100791931,
      "learning_rate": 2.3610000000000003e-05,
      "loss": 0.3664,
      "step": 2639
    },
    {
      "epoch": 240.0,
      "grad_norm": 0.5286477208137512,
      "learning_rate": 2.36e-05,
      "loss": 0.368,
      "step": 2640
    },
    {
      "epoch": 240.0,
      "eval_loss": 0.4361574351787567,
      "eval_runtime": 2.1758,
      "eval_samples_per_second": 25.278,
      "eval_steps_per_second": 3.217,
      "step": 2640
    },
    {
      "epoch": 240.0969696969697,
      "grad_norm": 0.2846533954143524,
      "learning_rate": 2.359e-05,
      "loss": 0.3299,
      "step": 2641
    },
    {
      "epoch": 240.1939393939394,
      "grad_norm": 0.282564252614975,
      "learning_rate": 2.358e-05,
      "loss": 0.3832,
      "step": 2642
    },
    {
      "epoch": 240.29090909090908,
      "grad_norm": 0.27177733182907104,
      "learning_rate": 2.357e-05,
      "loss": 0.3328,
      "step": 2643
    },
    {
      "epoch": 240.38787878787878,
      "grad_norm": 0.2859629988670349,
      "learning_rate": 2.356e-05,
      "loss": 0.3656,
      "step": 2644
    },
    {
      "epoch": 240.4848484848485,
      "grad_norm": 0.2732296884059906,
      "learning_rate": 2.355e-05,
      "loss": 0.3545,
      "step": 2645
    },
    {
      "epoch": 240.5818181818182,
      "grad_norm": 0.2838834524154663,
      "learning_rate": 2.354e-05,
      "loss": 0.331,
      "step": 2646
    },
    {
      "epoch": 240.6787878787879,
      "grad_norm": 0.3037766218185425,
      "learning_rate": 2.3530000000000003e-05,
      "loss": 0.3786,
      "step": 2647
    },
    {
      "epoch": 240.77575757575758,
      "grad_norm": 0.3013867139816284,
      "learning_rate": 2.3520000000000002e-05,
      "loss": 0.3472,
      "step": 2648
    },
    {
      "epoch": 240.87272727272727,
      "grad_norm": 0.2989252209663391,
      "learning_rate": 2.351e-05,
      "loss": 0.37,
      "step": 2649
    },
    {
      "epoch": 240.96969696969697,
      "grad_norm": 0.2740340828895569,
      "learning_rate": 2.35e-05,
      "loss": 0.3863,
      "step": 2650
    },
    {
      "epoch": 240.96969696969697,
      "eval_loss": 0.435860276222229,
      "eval_runtime": 2.1361,
      "eval_samples_per_second": 25.748,
      "eval_steps_per_second": 3.277,
      "step": 2650
    },
    {
      "epoch": 241.0,
      "grad_norm": 0.48094940185546875,
      "learning_rate": 2.349e-05,
      "loss": 0.4181,
      "step": 2651
    },
    {
      "epoch": 241.0969696969697,
      "grad_norm": 0.25698959827423096,
      "learning_rate": 2.3480000000000002e-05,
      "loss": 0.3279,
      "step": 2652
    },
    {
      "epoch": 241.1939393939394,
      "grad_norm": 0.27691131830215454,
      "learning_rate": 2.347e-05,
      "loss": 0.3487,
      "step": 2653
    },
    {
      "epoch": 241.29090909090908,
      "grad_norm": 0.24899674952030182,
      "learning_rate": 2.3460000000000002e-05,
      "loss": 0.3505,
      "step": 2654
    },
    {
      "epoch": 241.38787878787878,
      "grad_norm": 0.2850962281227112,
      "learning_rate": 2.345e-05,
      "loss": 0.3669,
      "step": 2655
    },
    {
      "epoch": 241.4848484848485,
      "grad_norm": 0.3749915659427643,
      "learning_rate": 2.344e-05,
      "loss": 0.3641,
      "step": 2656
    },
    {
      "epoch": 241.5818181818182,
      "grad_norm": 0.35326889157295227,
      "learning_rate": 2.343e-05,
      "loss": 0.3402,
      "step": 2657
    },
    {
      "epoch": 241.6787878787879,
      "grad_norm": 0.3130856156349182,
      "learning_rate": 2.342e-05,
      "loss": 0.3617,
      "step": 2658
    },
    {
      "epoch": 241.77575757575758,
      "grad_norm": 0.32955002784729004,
      "learning_rate": 2.341e-05,
      "loss": 0.3988,
      "step": 2659
    },
    {
      "epoch": 241.87272727272727,
      "grad_norm": 0.3192631006240845,
      "learning_rate": 2.3400000000000003e-05,
      "loss": 0.3683,
      "step": 2660
    },
    {
      "epoch": 241.87272727272727,
      "eval_loss": 0.43651309609413147,
      "eval_runtime": 2.1336,
      "eval_samples_per_second": 25.778,
      "eval_steps_per_second": 3.281,
      "step": 2660
    },
    {
      "epoch": 241.96969696969697,
      "grad_norm": 0.2528250515460968,
      "learning_rate": 2.339e-05,
      "loss": 0.3671,
      "step": 2661
    },
    {
      "epoch": 242.0,
      "grad_norm": 0.4293440282344818,
      "learning_rate": 2.3380000000000003e-05,
      "loss": 0.3697,
      "step": 2662
    },
    {
      "epoch": 242.0969696969697,
      "grad_norm": 0.3195413649082184,
      "learning_rate": 2.337e-05,
      "loss": 0.3752,
      "step": 2663
    },
    {
      "epoch": 242.1939393939394,
      "grad_norm": 0.25986403226852417,
      "learning_rate": 2.336e-05,
      "loss": 0.3317,
      "step": 2664
    },
    {
      "epoch": 242.29090909090908,
      "grad_norm": 0.33819544315338135,
      "learning_rate": 2.3350000000000002e-05,
      "loss": 0.3416,
      "step": 2665
    },
    {
      "epoch": 242.38787878787878,
      "grad_norm": 0.3044179379940033,
      "learning_rate": 2.334e-05,
      "loss": 0.3542,
      "step": 2666
    },
    {
      "epoch": 242.4848484848485,
      "grad_norm": 0.39956021308898926,
      "learning_rate": 2.3330000000000002e-05,
      "loss": 0.3439,
      "step": 2667
    },
    {
      "epoch": 242.5818181818182,
      "grad_norm": 0.3181312084197998,
      "learning_rate": 2.332e-05,
      "loss": 0.3304,
      "step": 2668
    },
    {
      "epoch": 242.6787878787879,
      "grad_norm": 0.2646390497684479,
      "learning_rate": 2.3310000000000002e-05,
      "loss": 0.3445,
      "step": 2669
    },
    {
      "epoch": 242.77575757575758,
      "grad_norm": 0.33343034982681274,
      "learning_rate": 2.3300000000000004e-05,
      "loss": 0.3675,
      "step": 2670
    },
    {
      "epoch": 242.77575757575758,
      "eval_loss": 0.43568772077560425,
      "eval_runtime": 2.176,
      "eval_samples_per_second": 25.275,
      "eval_steps_per_second": 3.217,
      "step": 2670
    },
    {
      "epoch": 242.87272727272727,
      "grad_norm": 0.300836980342865,
      "learning_rate": 2.3290000000000002e-05,
      "loss": 0.3774,
      "step": 2671
    },
    {
      "epoch": 242.96969696969697,
      "grad_norm": 0.27857717871665955,
      "learning_rate": 2.328e-05,
      "loss": 0.3975,
      "step": 2672
    },
    {
      "epoch": 243.0,
      "grad_norm": 0.605607271194458,
      "learning_rate": 2.327e-05,
      "loss": 0.4599,
      "step": 2673
    },
    {
      "epoch": 243.0969696969697,
      "grad_norm": 0.31283333897590637,
      "learning_rate": 2.326e-05,
      "loss": 0.3753,
      "step": 2674
    },
    {
      "epoch": 243.1939393939394,
      "grad_norm": 0.2685428261756897,
      "learning_rate": 2.3250000000000003e-05,
      "loss": 0.3631,
      "step": 2675
    },
    {
      "epoch": 243.29090909090908,
      "grad_norm": 0.2813699245452881,
      "learning_rate": 2.324e-05,
      "loss": 0.3413,
      "step": 2676
    },
    {
      "epoch": 243.38787878787878,
      "grad_norm": 0.26216891407966614,
      "learning_rate": 2.3230000000000003e-05,
      "loss": 0.3513,
      "step": 2677
    },
    {
      "epoch": 243.4848484848485,
      "grad_norm": 0.2780851125717163,
      "learning_rate": 2.322e-05,
      "loss": 0.3912,
      "step": 2678
    },
    {
      "epoch": 243.5818181818182,
      "grad_norm": 0.27967149019241333,
      "learning_rate": 2.321e-05,
      "loss": 0.3524,
      "step": 2679
    },
    {
      "epoch": 243.6787878787879,
      "grad_norm": 0.3546093702316284,
      "learning_rate": 2.32e-05,
      "loss": 0.3769,
      "step": 2680
    },
    {
      "epoch": 243.6787878787879,
      "eval_loss": 0.4360385835170746,
      "eval_runtime": 2.1281,
      "eval_samples_per_second": 25.844,
      "eval_steps_per_second": 3.289,
      "step": 2680
    },
    {
      "epoch": 243.77575757575758,
      "grad_norm": 0.3281584084033966,
      "learning_rate": 2.319e-05,
      "loss": 0.3355,
      "step": 2681
    },
    {
      "epoch": 243.87272727272727,
      "grad_norm": 0.2678728997707367,
      "learning_rate": 2.318e-05,
      "loss": 0.3333,
      "step": 2682
    },
    {
      "epoch": 243.96969696969697,
      "grad_norm": 0.33740246295928955,
      "learning_rate": 2.317e-05,
      "loss": 0.3729,
      "step": 2683
    },
    {
      "epoch": 244.0,
      "grad_norm": 0.49675431847572327,
      "learning_rate": 2.3160000000000002e-05,
      "loss": 0.3465,
      "step": 2684
    },
    {
      "epoch": 244.0969696969697,
      "grad_norm": 0.29415813088417053,
      "learning_rate": 2.3150000000000004e-05,
      "loss": 0.3735,
      "step": 2685
    },
    {
      "epoch": 244.1939393939394,
      "grad_norm": 0.2935345768928528,
      "learning_rate": 2.3140000000000002e-05,
      "loss": 0.3458,
      "step": 2686
    },
    {
      "epoch": 244.29090909090908,
      "grad_norm": 0.29703962802886963,
      "learning_rate": 2.313e-05,
      "loss": 0.3842,
      "step": 2687
    },
    {
      "epoch": 244.38787878787878,
      "grad_norm": 0.3192014694213867,
      "learning_rate": 2.312e-05,
      "loss": 0.3941,
      "step": 2688
    },
    {
      "epoch": 244.4848484848485,
      "grad_norm": 0.29394468665122986,
      "learning_rate": 2.311e-05,
      "loss": 0.3421,
      "step": 2689
    },
    {
      "epoch": 244.5818181818182,
      "grad_norm": 0.24139085412025452,
      "learning_rate": 2.3100000000000002e-05,
      "loss": 0.3456,
      "step": 2690
    },
    {
      "epoch": 244.5818181818182,
      "eval_loss": 0.4357967674732208,
      "eval_runtime": 2.1326,
      "eval_samples_per_second": 25.79,
      "eval_steps_per_second": 3.282,
      "step": 2690
    },
    {
      "epoch": 244.6787878787879,
      "grad_norm": 0.2768908143043518,
      "learning_rate": 2.309e-05,
      "loss": 0.3778,
      "step": 2691
    },
    {
      "epoch": 244.77575757575758,
      "grad_norm": 0.2763884663581848,
      "learning_rate": 2.3080000000000003e-05,
      "loss": 0.3425,
      "step": 2692
    },
    {
      "epoch": 244.87272727272727,
      "grad_norm": 0.310211181640625,
      "learning_rate": 2.307e-05,
      "loss": 0.3269,
      "step": 2693
    },
    {
      "epoch": 244.96969696969697,
      "grad_norm": 0.27987420558929443,
      "learning_rate": 2.306e-05,
      "loss": 0.3573,
      "step": 2694
    },
    {
      "epoch": 245.0,
      "grad_norm": 0.49357059597969055,
      "learning_rate": 2.305e-05,
      "loss": 0.3383,
      "step": 2695
    },
    {
      "epoch": 245.0969696969697,
      "grad_norm": 0.29769405722618103,
      "learning_rate": 2.304e-05,
      "loss": 0.3586,
      "step": 2696
    },
    {
      "epoch": 245.1939393939394,
      "grad_norm": 0.29475346207618713,
      "learning_rate": 2.303e-05,
      "loss": 0.3454,
      "step": 2697
    },
    {
      "epoch": 245.29090909090908,
      "grad_norm": 0.2872675359249115,
      "learning_rate": 2.302e-05,
      "loss": 0.3577,
      "step": 2698
    },
    {
      "epoch": 245.38787878787878,
      "grad_norm": 0.27988383173942566,
      "learning_rate": 2.301e-05,
      "loss": 0.3627,
      "step": 2699
    },
    {
      "epoch": 245.4848484848485,
      "grad_norm": 0.3422905206680298,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 0.38,
      "step": 2700
    },
    {
      "epoch": 245.4848484848485,
      "eval_loss": 0.43530699610710144,
      "eval_runtime": 2.1081,
      "eval_samples_per_second": 26.09,
      "eval_steps_per_second": 3.32,
      "step": 2700
    },
    {
      "epoch": 245.5818181818182,
      "grad_norm": 0.3205651044845581,
      "learning_rate": 2.2990000000000002e-05,
      "loss": 0.3462,
      "step": 2701
    },
    {
      "epoch": 245.6787878787879,
      "grad_norm": 0.2794988751411438,
      "learning_rate": 2.298e-05,
      "loss": 0.378,
      "step": 2702
    },
    {
      "epoch": 245.77575757575758,
      "grad_norm": 0.2612413465976715,
      "learning_rate": 2.297e-05,
      "loss": 0.3647,
      "step": 2703
    },
    {
      "epoch": 245.87272727272727,
      "grad_norm": 0.2854641377925873,
      "learning_rate": 2.296e-05,
      "loss": 0.3289,
      "step": 2704
    },
    {
      "epoch": 245.96969696969697,
      "grad_norm": 0.29106345772743225,
      "learning_rate": 2.2950000000000002e-05,
      "loss": 0.349,
      "step": 2705
    },
    {
      "epoch": 246.0,
      "grad_norm": 0.5645086169242859,
      "learning_rate": 2.294e-05,
      "loss": 0.3966,
      "step": 2706
    },
    {
      "epoch": 246.0969696969697,
      "grad_norm": 0.33425596356391907,
      "learning_rate": 2.2930000000000002e-05,
      "loss": 0.351,
      "step": 2707
    },
    {
      "epoch": 246.1939393939394,
      "grad_norm": 0.2670089900493622,
      "learning_rate": 2.292e-05,
      "loss": 0.3298,
      "step": 2708
    },
    {
      "epoch": 246.29090909090908,
      "grad_norm": 0.2702706456184387,
      "learning_rate": 2.2910000000000003e-05,
      "loss": 0.3676,
      "step": 2709
    },
    {
      "epoch": 246.38787878787878,
      "grad_norm": 0.30324020981788635,
      "learning_rate": 2.29e-05,
      "loss": 0.3569,
      "step": 2710
    },
    {
      "epoch": 246.38787878787878,
      "eval_loss": 0.43549418449401855,
      "eval_runtime": 2.1254,
      "eval_samples_per_second": 25.877,
      "eval_steps_per_second": 3.293,
      "step": 2710
    },
    {
      "epoch": 246.4848484848485,
      "grad_norm": 0.3084642291069031,
      "learning_rate": 2.289e-05,
      "loss": 0.3725,
      "step": 2711
    },
    {
      "epoch": 246.5818181818182,
      "grad_norm": 0.28131335973739624,
      "learning_rate": 2.288e-05,
      "loss": 0.3642,
      "step": 2712
    },
    {
      "epoch": 246.6787878787879,
      "grad_norm": 0.28382378816604614,
      "learning_rate": 2.287e-05,
      "loss": 0.3469,
      "step": 2713
    },
    {
      "epoch": 246.77575757575758,
      "grad_norm": 0.3298587203025818,
      "learning_rate": 2.286e-05,
      "loss": 0.3785,
      "step": 2714
    },
    {
      "epoch": 246.87272727272727,
      "grad_norm": 0.27626752853393555,
      "learning_rate": 2.2850000000000003e-05,
      "loss": 0.3784,
      "step": 2715
    },
    {
      "epoch": 246.96969696969697,
      "grad_norm": 0.28574326634407043,
      "learning_rate": 2.284e-05,
      "loss": 0.3418,
      "step": 2716
    },
    {
      "epoch": 247.0,
      "grad_norm": 0.4947167634963989,
      "learning_rate": 2.283e-05,
      "loss": 0.3356,
      "step": 2717
    },
    {
      "epoch": 247.0969696969697,
      "grad_norm": 0.3042255640029907,
      "learning_rate": 2.282e-05,
      "loss": 0.3509,
      "step": 2718
    },
    {
      "epoch": 247.1939393939394,
      "grad_norm": 0.2821086347103119,
      "learning_rate": 2.281e-05,
      "loss": 0.3148,
      "step": 2719
    },
    {
      "epoch": 247.29090909090908,
      "grad_norm": 0.3053955137729645,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 0.3743,
      "step": 2720
    },
    {
      "epoch": 247.29090909090908,
      "eval_loss": 0.43521028757095337,
      "eval_runtime": 2.1152,
      "eval_samples_per_second": 26.002,
      "eval_steps_per_second": 3.309,
      "step": 2720
    },
    {
      "epoch": 247.38787878787878,
      "grad_norm": 0.2838248312473297,
      "learning_rate": 2.279e-05,
      "loss": 0.3703,
      "step": 2721
    },
    {
      "epoch": 247.4848484848485,
      "grad_norm": 0.2984090745449066,
      "learning_rate": 2.2780000000000002e-05,
      "loss": 0.3624,
      "step": 2722
    },
    {
      "epoch": 247.5818181818182,
      "grad_norm": 0.25600433349609375,
      "learning_rate": 2.2770000000000004e-05,
      "loss": 0.3801,
      "step": 2723
    },
    {
      "epoch": 247.6787878787879,
      "grad_norm": 0.3806086480617523,
      "learning_rate": 2.2760000000000002e-05,
      "loss": 0.3568,
      "step": 2724
    },
    {
      "epoch": 247.77575757575758,
      "grad_norm": 0.2735169231891632,
      "learning_rate": 2.275e-05,
      "loss": 0.3429,
      "step": 2725
    },
    {
      "epoch": 247.87272727272727,
      "grad_norm": 0.2780759036540985,
      "learning_rate": 2.274e-05,
      "loss": 0.3573,
      "step": 2726
    },
    {
      "epoch": 247.96969696969697,
      "grad_norm": 0.24951213598251343,
      "learning_rate": 2.273e-05,
      "loss": 0.3584,
      "step": 2727
    },
    {
      "epoch": 248.0,
      "grad_norm": 0.4915936589241028,
      "learning_rate": 2.2720000000000003e-05,
      "loss": 0.3818,
      "step": 2728
    },
    {
      "epoch": 248.0969696969697,
      "grad_norm": 0.3129208981990814,
      "learning_rate": 2.271e-05,
      "loss": 0.4144,
      "step": 2729
    },
    {
      "epoch": 248.1939393939394,
      "grad_norm": 0.24280574917793274,
      "learning_rate": 2.2700000000000003e-05,
      "loss": 0.3706,
      "step": 2730
    },
    {
      "epoch": 248.1939393939394,
      "eval_loss": 0.43458279967308044,
      "eval_runtime": 2.1181,
      "eval_samples_per_second": 25.966,
      "eval_steps_per_second": 3.305,
      "step": 2730
    },
    {
      "epoch": 248.29090909090908,
      "grad_norm": 0.2954522967338562,
      "learning_rate": 2.269e-05,
      "loss": 0.3396,
      "step": 2731
    },
    {
      "epoch": 248.38787878787878,
      "grad_norm": 0.2967293858528137,
      "learning_rate": 2.268e-05,
      "loss": 0.3365,
      "step": 2732
    },
    {
      "epoch": 248.4848484848485,
      "grad_norm": 0.3011590540409088,
      "learning_rate": 2.267e-05,
      "loss": 0.3521,
      "step": 2733
    },
    {
      "epoch": 248.5818181818182,
      "grad_norm": 0.2972559928894043,
      "learning_rate": 2.266e-05,
      "loss": 0.3605,
      "step": 2734
    },
    {
      "epoch": 248.6787878787879,
      "grad_norm": 0.30652526021003723,
      "learning_rate": 2.265e-05,
      "loss": 0.334,
      "step": 2735
    },
    {
      "epoch": 248.77575757575758,
      "grad_norm": 0.2733524739742279,
      "learning_rate": 2.264e-05,
      "loss": 0.3715,
      "step": 2736
    },
    {
      "epoch": 248.87272727272727,
      "grad_norm": 0.29794397950172424,
      "learning_rate": 2.2630000000000002e-05,
      "loss": 0.3531,
      "step": 2737
    },
    {
      "epoch": 248.96969696969697,
      "grad_norm": 0.26180553436279297,
      "learning_rate": 2.2620000000000004e-05,
      "loss": 0.3529,
      "step": 2738
    },
    {
      "epoch": 249.0,
      "grad_norm": 0.3981349468231201,
      "learning_rate": 2.2610000000000002e-05,
      "loss": 0.3109,
      "step": 2739
    },
    {
      "epoch": 249.0969696969697,
      "grad_norm": 0.30066075921058655,
      "learning_rate": 2.26e-05,
      "loss": 0.3637,
      "step": 2740
    },
    {
      "epoch": 249.0969696969697,
      "eval_loss": 0.4350603520870209,
      "eval_runtime": 2.1392,
      "eval_samples_per_second": 25.711,
      "eval_steps_per_second": 3.272,
      "step": 2740
    },
    {
      "epoch": 249.1939393939394,
      "grad_norm": 0.2947355806827545,
      "learning_rate": 2.259e-05,
      "loss": 0.4017,
      "step": 2741
    },
    {
      "epoch": 249.29090909090908,
      "grad_norm": 0.26995307207107544,
      "learning_rate": 2.258e-05,
      "loss": 0.3106,
      "step": 2742
    },
    {
      "epoch": 249.38787878787878,
      "grad_norm": 0.24598020315170288,
      "learning_rate": 2.2570000000000002e-05,
      "loss": 0.3411,
      "step": 2743
    },
    {
      "epoch": 249.4848484848485,
      "grad_norm": 0.28865739703178406,
      "learning_rate": 2.256e-05,
      "loss": 0.3884,
      "step": 2744
    },
    {
      "epoch": 249.5818181818182,
      "grad_norm": 0.3034263253211975,
      "learning_rate": 2.2550000000000003e-05,
      "loss": 0.3637,
      "step": 2745
    },
    {
      "epoch": 249.6787878787879,
      "grad_norm": 0.30831438302993774,
      "learning_rate": 2.254e-05,
      "loss": 0.3672,
      "step": 2746
    },
    {
      "epoch": 249.77575757575758,
      "grad_norm": 0.2989354133605957,
      "learning_rate": 2.253e-05,
      "loss": 0.3489,
      "step": 2747
    },
    {
      "epoch": 249.87272727272727,
      "grad_norm": 0.2784823775291443,
      "learning_rate": 2.252e-05,
      "loss": 0.3226,
      "step": 2748
    },
    {
      "epoch": 249.96969696969697,
      "grad_norm": 0.33569571375846863,
      "learning_rate": 2.251e-05,
      "loss": 0.3456,
      "step": 2749
    },
    {
      "epoch": 250.0,
      "grad_norm": 0.5049756765365601,
      "learning_rate": 2.25e-05,
      "loss": 0.3994,
      "step": 2750
    },
    {
      "epoch": 250.0,
      "eval_loss": 0.4343967139720917,
      "eval_runtime": 2.123,
      "eval_samples_per_second": 25.906,
      "eval_steps_per_second": 3.297,
      "step": 2750
    },
    {
      "epoch": 250.0969696969697,
      "grad_norm": 0.3116196393966675,
      "learning_rate": 2.249e-05,
      "loss": 0.3805,
      "step": 2751
    },
    {
      "epoch": 250.1939393939394,
      "grad_norm": 0.31933361291885376,
      "learning_rate": 2.248e-05,
      "loss": 0.3548,
      "step": 2752
    },
    {
      "epoch": 250.29090909090908,
      "grad_norm": 0.3103712201118469,
      "learning_rate": 2.2470000000000003e-05,
      "loss": 0.3475,
      "step": 2753
    },
    {
      "epoch": 250.38787878787878,
      "grad_norm": 0.30291813611984253,
      "learning_rate": 2.2460000000000002e-05,
      "loss": 0.3643,
      "step": 2754
    },
    {
      "epoch": 250.4848484848485,
      "grad_norm": 0.31556010246276855,
      "learning_rate": 2.245e-05,
      "loss": 0.364,
      "step": 2755
    },
    {
      "epoch": 250.5818181818182,
      "grad_norm": 0.25011295080184937,
      "learning_rate": 2.244e-05,
      "loss": 0.3281,
      "step": 2756
    },
    {
      "epoch": 250.6787878787879,
      "grad_norm": 0.30427756905555725,
      "learning_rate": 2.243e-05,
      "loss": 0.3667,
      "step": 2757
    },
    {
      "epoch": 250.77575757575758,
      "grad_norm": 0.28063085675239563,
      "learning_rate": 2.2420000000000002e-05,
      "loss": 0.386,
      "step": 2758
    },
    {
      "epoch": 250.87272727272727,
      "grad_norm": 0.3346632719039917,
      "learning_rate": 2.241e-05,
      "loss": 0.2989,
      "step": 2759
    },
    {
      "epoch": 250.96969696969697,
      "grad_norm": 0.2454753965139389,
      "learning_rate": 2.2400000000000002e-05,
      "loss": 0.3819,
      "step": 2760
    },
    {
      "epoch": 250.96969696969697,
      "eval_loss": 0.43441030383110046,
      "eval_runtime": 2.1286,
      "eval_samples_per_second": 25.839,
      "eval_steps_per_second": 3.289,
      "step": 2760
    },
    {
      "epoch": 251.0,
      "grad_norm": 0.3684418499469757,
      "learning_rate": 2.239e-05,
      "loss": 0.335,
      "step": 2761
    },
    {
      "epoch": 251.0969696969697,
      "grad_norm": 0.2791552245616913,
      "learning_rate": 2.2380000000000003e-05,
      "loss": 0.3631,
      "step": 2762
    },
    {
      "epoch": 251.1939393939394,
      "grad_norm": 0.3262360394001007,
      "learning_rate": 2.237e-05,
      "loss": 0.3232,
      "step": 2763
    },
    {
      "epoch": 251.29090909090908,
      "grad_norm": 0.3078155815601349,
      "learning_rate": 2.236e-05,
      "loss": 0.3687,
      "step": 2764
    },
    {
      "epoch": 251.38787878787878,
      "grad_norm": 0.2574249804019928,
      "learning_rate": 2.235e-05,
      "loss": 0.3558,
      "step": 2765
    },
    {
      "epoch": 251.4848484848485,
      "grad_norm": 0.2514949440956116,
      "learning_rate": 2.234e-05,
      "loss": 0.3739,
      "step": 2766
    },
    {
      "epoch": 251.5818181818182,
      "grad_norm": 0.3091491460800171,
      "learning_rate": 2.233e-05,
      "loss": 0.3153,
      "step": 2767
    },
    {
      "epoch": 251.6787878787879,
      "grad_norm": 0.3235987722873688,
      "learning_rate": 2.2320000000000003e-05,
      "loss": 0.3392,
      "step": 2768
    },
    {
      "epoch": 251.77575757575758,
      "grad_norm": 0.29184114933013916,
      "learning_rate": 2.231e-05,
      "loss": 0.3874,
      "step": 2769
    },
    {
      "epoch": 251.87272727272727,
      "grad_norm": 0.28661590814590454,
      "learning_rate": 2.23e-05,
      "loss": 0.3721,
      "step": 2770
    },
    {
      "epoch": 251.87272727272727,
      "eval_loss": 0.43438997864723206,
      "eval_runtime": 2.1219,
      "eval_samples_per_second": 25.92,
      "eval_steps_per_second": 3.299,
      "step": 2770
    },
    {
      "epoch": 251.96969696969697,
      "grad_norm": 0.29083678126335144,
      "learning_rate": 2.229e-05,
      "loss": 0.348,
      "step": 2771
    },
    {
      "epoch": 252.0,
      "grad_norm": 0.48070859909057617,
      "learning_rate": 2.228e-05,
      "loss": 0.4007,
      "step": 2772
    },
    {
      "epoch": 252.0969696969697,
      "grad_norm": 0.2264174222946167,
      "learning_rate": 2.2270000000000002e-05,
      "loss": 0.3252,
      "step": 2773
    },
    {
      "epoch": 252.1939393939394,
      "grad_norm": 0.2599073648452759,
      "learning_rate": 2.226e-05,
      "loss": 0.3747,
      "step": 2774
    },
    {
      "epoch": 252.29090909090908,
      "grad_norm": 0.3252253830432892,
      "learning_rate": 2.2250000000000002e-05,
      "loss": 0.3797,
      "step": 2775
    },
    {
      "epoch": 252.38787878787878,
      "grad_norm": 0.33481475710868835,
      "learning_rate": 2.224e-05,
      "loss": 0.3327,
      "step": 2776
    },
    {
      "epoch": 252.4848484848485,
      "grad_norm": 0.2517988085746765,
      "learning_rate": 2.2230000000000002e-05,
      "loss": 0.3472,
      "step": 2777
    },
    {
      "epoch": 252.5818181818182,
      "grad_norm": 0.28708556294441223,
      "learning_rate": 2.222e-05,
      "loss": 0.3387,
      "step": 2778
    },
    {
      "epoch": 252.6787878787879,
      "grad_norm": 0.31490185856819153,
      "learning_rate": 2.221e-05,
      "loss": 0.3629,
      "step": 2779
    },
    {
      "epoch": 252.77575757575758,
      "grad_norm": 0.34092140197753906,
      "learning_rate": 2.22e-05,
      "loss": 0.3281,
      "step": 2780
    },
    {
      "epoch": 252.77575757575758,
      "eval_loss": 0.4341490864753723,
      "eval_runtime": 2.132,
      "eval_samples_per_second": 25.797,
      "eval_steps_per_second": 3.283,
      "step": 2780
    },
    {
      "epoch": 252.87272727272727,
      "grad_norm": 0.31832852959632874,
      "learning_rate": 2.219e-05,
      "loss": 0.3817,
      "step": 2781
    },
    {
      "epoch": 252.96969696969697,
      "grad_norm": 0.28311994671821594,
      "learning_rate": 2.218e-05,
      "loss": 0.3705,
      "step": 2782
    },
    {
      "epoch": 253.0,
      "grad_norm": 0.4966646432876587,
      "learning_rate": 2.2170000000000003e-05,
      "loss": 0.4161,
      "step": 2783
    },
    {
      "epoch": 253.0969696969697,
      "grad_norm": 0.2603364586830139,
      "learning_rate": 2.216e-05,
      "loss": 0.3614,
      "step": 2784
    },
    {
      "epoch": 253.1939393939394,
      "grad_norm": 0.2830662429332733,
      "learning_rate": 2.215e-05,
      "loss": 0.3586,
      "step": 2785
    },
    {
      "epoch": 253.29090909090908,
      "grad_norm": 0.32937613129615784,
      "learning_rate": 2.214e-05,
      "loss": 0.3782,
      "step": 2786
    },
    {
      "epoch": 253.38787878787878,
      "grad_norm": 0.33842575550079346,
      "learning_rate": 2.213e-05,
      "loss": 0.368,
      "step": 2787
    },
    {
      "epoch": 253.4848484848485,
      "grad_norm": 0.3619191348552704,
      "learning_rate": 2.212e-05,
      "loss": 0.357,
      "step": 2788
    },
    {
      "epoch": 253.5818181818182,
      "grad_norm": 0.27501344680786133,
      "learning_rate": 2.211e-05,
      "loss": 0.3425,
      "step": 2789
    },
    {
      "epoch": 253.6787878787879,
      "grad_norm": 0.24685457348823547,
      "learning_rate": 2.2100000000000002e-05,
      "loss": 0.3064,
      "step": 2790
    },
    {
      "epoch": 253.6787878787879,
      "eval_loss": 0.4338873028755188,
      "eval_runtime": 2.1219,
      "eval_samples_per_second": 25.921,
      "eval_steps_per_second": 3.299,
      "step": 2790
    },
    {
      "epoch": 253.77575757575758,
      "grad_norm": 0.3157132565975189,
      "learning_rate": 2.2090000000000004e-05,
      "loss": 0.339,
      "step": 2791
    },
    {
      "epoch": 253.87272727272727,
      "grad_norm": 0.3351724445819855,
      "learning_rate": 2.2080000000000002e-05,
      "loss": 0.3762,
      "step": 2792
    },
    {
      "epoch": 253.96969696969697,
      "grad_norm": 0.29916393756866455,
      "learning_rate": 2.207e-05,
      "loss": 0.3566,
      "step": 2793
    },
    {
      "epoch": 254.0,
      "grad_norm": 0.4145132899284363,
      "learning_rate": 2.206e-05,
      "loss": 0.3867,
      "step": 2794
    },
    {
      "epoch": 254.0969696969697,
      "grad_norm": 0.3179309368133545,
      "learning_rate": 2.205e-05,
      "loss": 0.3995,
      "step": 2795
    },
    {
      "epoch": 254.1939393939394,
      "grad_norm": 0.30933842062950134,
      "learning_rate": 2.2040000000000002e-05,
      "loss": 0.3161,
      "step": 2796
    },
    {
      "epoch": 254.29090909090908,
      "grad_norm": 0.24094711244106293,
      "learning_rate": 2.203e-05,
      "loss": 0.3451,
      "step": 2797
    },
    {
      "epoch": 254.38787878787878,
      "grad_norm": 0.34217455983161926,
      "learning_rate": 2.2020000000000003e-05,
      "loss": 0.3428,
      "step": 2798
    },
    {
      "epoch": 254.4848484848485,
      "grad_norm": 0.3295113146305084,
      "learning_rate": 2.201e-05,
      "loss": 0.3441,
      "step": 2799
    },
    {
      "epoch": 254.5818181818182,
      "grad_norm": 0.293735146522522,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.3684,
      "step": 2800
    },
    {
      "epoch": 254.5818181818182,
      "eval_loss": 0.4341706931591034,
      "eval_runtime": 2.1323,
      "eval_samples_per_second": 25.794,
      "eval_steps_per_second": 3.283,
      "step": 2800
    },
    {
      "epoch": 254.6787878787879,
      "grad_norm": 0.3140919804573059,
      "learning_rate": 2.199e-05,
      "loss": 0.3704,
      "step": 2801
    },
    {
      "epoch": 254.77575757575758,
      "grad_norm": 0.3150195777416229,
      "learning_rate": 2.198e-05,
      "loss": 0.3247,
      "step": 2802
    },
    {
      "epoch": 254.87272727272727,
      "grad_norm": 0.36488816142082214,
      "learning_rate": 2.197e-05,
      "loss": 0.347,
      "step": 2803
    },
    {
      "epoch": 254.96969696969697,
      "grad_norm": 0.2660673260688782,
      "learning_rate": 2.196e-05,
      "loss": 0.3861,
      "step": 2804
    },
    {
      "epoch": 255.0,
      "grad_norm": 0.43914130330085754,
      "learning_rate": 2.195e-05,
      "loss": 0.3764,
      "step": 2805
    },
    {
      "epoch": 255.0969696969697,
      "grad_norm": 0.2545498311519623,
      "learning_rate": 2.1940000000000003e-05,
      "loss": 0.3369,
      "step": 2806
    },
    {
      "epoch": 255.1939393939394,
      "grad_norm": 0.30860552191734314,
      "learning_rate": 2.1930000000000002e-05,
      "loss": 0.3233,
      "step": 2807
    },
    {
      "epoch": 255.29090909090908,
      "grad_norm": 0.2711668610572815,
      "learning_rate": 2.192e-05,
      "loss": 0.3579,
      "step": 2808
    },
    {
      "epoch": 255.38787878787878,
      "grad_norm": 0.2572686970233917,
      "learning_rate": 2.191e-05,
      "loss": 0.365,
      "step": 2809
    },
    {
      "epoch": 255.4848484848485,
      "grad_norm": 0.2472272664308548,
      "learning_rate": 2.19e-05,
      "loss": 0.3652,
      "step": 2810
    },
    {
      "epoch": 255.4848484848485,
      "eval_loss": 0.43348821997642517,
      "eval_runtime": 2.127,
      "eval_samples_per_second": 25.858,
      "eval_steps_per_second": 3.291,
      "step": 2810
    },
    {
      "epoch": 255.5818181818182,
      "grad_norm": 0.31041258573532104,
      "learning_rate": 2.1890000000000002e-05,
      "loss": 0.3725,
      "step": 2811
    },
    {
      "epoch": 255.6787878787879,
      "grad_norm": 0.3136427700519562,
      "learning_rate": 2.188e-05,
      "loss": 0.3256,
      "step": 2812
    },
    {
      "epoch": 255.77575757575758,
      "grad_norm": 0.3329264521598816,
      "learning_rate": 2.1870000000000002e-05,
      "loss": 0.3629,
      "step": 2813
    },
    {
      "epoch": 255.87272727272727,
      "grad_norm": 0.43352991342544556,
      "learning_rate": 2.186e-05,
      "loss": 0.3633,
      "step": 2814
    },
    {
      "epoch": 255.96969696969697,
      "grad_norm": 0.29008254408836365,
      "learning_rate": 2.1850000000000003e-05,
      "loss": 0.3756,
      "step": 2815
    },
    {
      "epoch": 256.0,
      "grad_norm": 0.5636740326881409,
      "learning_rate": 2.184e-05,
      "loss": 0.3529,
      "step": 2816
    },
    {
      "epoch": 256.0969696969697,
      "grad_norm": 0.28044673800468445,
      "learning_rate": 2.183e-05,
      "loss": 0.3452,
      "step": 2817
    },
    {
      "epoch": 256.1939393939394,
      "grad_norm": 0.3173394501209259,
      "learning_rate": 2.182e-05,
      "loss": 0.3842,
      "step": 2818
    },
    {
      "epoch": 256.2909090909091,
      "grad_norm": 0.25740739703178406,
      "learning_rate": 2.181e-05,
      "loss": 0.3507,
      "step": 2819
    },
    {
      "epoch": 256.3878787878788,
      "grad_norm": 0.24135111272335052,
      "learning_rate": 2.18e-05,
      "loss": 0.3481,
      "step": 2820
    },
    {
      "epoch": 256.3878787878788,
      "eval_loss": 0.43362826108932495,
      "eval_runtime": 2.1052,
      "eval_samples_per_second": 26.126,
      "eval_steps_per_second": 3.325,
      "step": 2820
    },
    {
      "epoch": 256.4848484848485,
      "grad_norm": 0.36606621742248535,
      "learning_rate": 2.1790000000000003e-05,
      "loss": 0.3631,
      "step": 2821
    },
    {
      "epoch": 256.58181818181816,
      "grad_norm": 0.34213078022003174,
      "learning_rate": 2.178e-05,
      "loss": 0.3221,
      "step": 2822
    },
    {
      "epoch": 256.6787878787879,
      "grad_norm": 0.36660224199295044,
      "learning_rate": 2.177e-05,
      "loss": 0.3657,
      "step": 2823
    },
    {
      "epoch": 256.77575757575755,
      "grad_norm": 0.2867613136768341,
      "learning_rate": 2.176e-05,
      "loss": 0.3337,
      "step": 2824
    },
    {
      "epoch": 256.8727272727273,
      "grad_norm": 0.28834688663482666,
      "learning_rate": 2.175e-05,
      "loss": 0.3719,
      "step": 2825
    },
    {
      "epoch": 256.969696969697,
      "grad_norm": 0.28708332777023315,
      "learning_rate": 2.1740000000000002e-05,
      "loss": 0.371,
      "step": 2826
    },
    {
      "epoch": 257.0,
      "grad_norm": 0.5373139381408691,
      "learning_rate": 2.173e-05,
      "loss": 0.3228,
      "step": 2827
    },
    {
      "epoch": 257.0969696969697,
      "grad_norm": 0.33356258273124695,
      "learning_rate": 2.1720000000000002e-05,
      "loss": 0.3629,
      "step": 2828
    },
    {
      "epoch": 257.1939393939394,
      "grad_norm": 0.2881753444671631,
      "learning_rate": 2.171e-05,
      "loss": 0.3258,
      "step": 2829
    },
    {
      "epoch": 257.2909090909091,
      "grad_norm": 0.27659082412719727,
      "learning_rate": 2.1700000000000002e-05,
      "loss": 0.338,
      "step": 2830
    },
    {
      "epoch": 257.2909090909091,
      "eval_loss": 0.43408533930778503,
      "eval_runtime": 2.1289,
      "eval_samples_per_second": 25.835,
      "eval_steps_per_second": 3.288,
      "step": 2830
    },
    {
      "epoch": 257.3878787878788,
      "grad_norm": 0.39573410153388977,
      "learning_rate": 2.169e-05,
      "loss": 0.3303,
      "step": 2831
    },
    {
      "epoch": 257.4848484848485,
      "grad_norm": 0.3405710458755493,
      "learning_rate": 2.168e-05,
      "loss": 0.3754,
      "step": 2832
    },
    {
      "epoch": 257.58181818181816,
      "grad_norm": 0.24880316853523254,
      "learning_rate": 2.167e-05,
      "loss": 0.3733,
      "step": 2833
    },
    {
      "epoch": 257.6787878787879,
      "grad_norm": 0.2995625436306,
      "learning_rate": 2.166e-05,
      "loss": 0.3231,
      "step": 2834
    },
    {
      "epoch": 257.77575757575755,
      "grad_norm": 0.3127087354660034,
      "learning_rate": 2.165e-05,
      "loss": 0.3844,
      "step": 2835
    },
    {
      "epoch": 257.8727272727273,
      "grad_norm": 0.32488930225372314,
      "learning_rate": 2.1640000000000003e-05,
      "loss": 0.3644,
      "step": 2836
    },
    {
      "epoch": 257.969696969697,
      "grad_norm": 0.28634101152420044,
      "learning_rate": 2.163e-05,
      "loss": 0.3553,
      "step": 2837
    },
    {
      "epoch": 258.0,
      "grad_norm": 0.4247503876686096,
      "learning_rate": 2.162e-05,
      "loss": 0.3823,
      "step": 2838
    },
    {
      "epoch": 258.0969696969697,
      "grad_norm": 0.29246553778648376,
      "learning_rate": 2.1609999999999998e-05,
      "loss": 0.3658,
      "step": 2839
    },
    {
      "epoch": 258.1939393939394,
      "grad_norm": 0.32681018114089966,
      "learning_rate": 2.16e-05,
      "loss": 0.3263,
      "step": 2840
    },
    {
      "epoch": 258.1939393939394,
      "eval_loss": 0.43289804458618164,
      "eval_runtime": 2.1171,
      "eval_samples_per_second": 25.978,
      "eval_steps_per_second": 3.306,
      "step": 2840
    },
    {
      "epoch": 258.2909090909091,
      "grad_norm": 0.4305499196052551,
      "learning_rate": 2.159e-05,
      "loss": 0.3872,
      "step": 2841
    },
    {
      "epoch": 258.3878787878788,
      "grad_norm": 0.2609012722969055,
      "learning_rate": 2.158e-05,
      "loss": 0.333,
      "step": 2842
    },
    {
      "epoch": 258.4848484848485,
      "grad_norm": 0.3244587779045105,
      "learning_rate": 2.1570000000000002e-05,
      "loss": 0.3555,
      "step": 2843
    },
    {
      "epoch": 258.58181818181816,
      "grad_norm": 0.26547229290008545,
      "learning_rate": 2.1560000000000004e-05,
      "loss": 0.3297,
      "step": 2844
    },
    {
      "epoch": 258.6787878787879,
      "grad_norm": 0.2793107330799103,
      "learning_rate": 2.1550000000000002e-05,
      "loss": 0.3694,
      "step": 2845
    },
    {
      "epoch": 258.77575757575755,
      "grad_norm": 0.3379392921924591,
      "learning_rate": 2.154e-05,
      "loss": 0.3293,
      "step": 2846
    },
    {
      "epoch": 258.8727272727273,
      "grad_norm": 0.2868521809577942,
      "learning_rate": 2.153e-05,
      "loss": 0.3364,
      "step": 2847
    },
    {
      "epoch": 258.969696969697,
      "grad_norm": 0.3567075729370117,
      "learning_rate": 2.152e-05,
      "loss": 0.3914,
      "step": 2848
    },
    {
      "epoch": 259.0,
      "grad_norm": 0.47458121180534363,
      "learning_rate": 2.1510000000000002e-05,
      "loss": 0.4021,
      "step": 2849
    },
    {
      "epoch": 259.0969696969697,
      "grad_norm": 0.39531266689300537,
      "learning_rate": 2.15e-05,
      "loss": 0.3842,
      "step": 2850
    },
    {
      "epoch": 259.0969696969697,
      "eval_loss": 0.43324747681617737,
      "eval_runtime": 2.1122,
      "eval_samples_per_second": 26.039,
      "eval_steps_per_second": 3.314,
      "step": 2850
    },
    {
      "epoch": 259.1939393939394,
      "grad_norm": 0.2564777135848999,
      "learning_rate": 2.1490000000000003e-05,
      "loss": 0.3599,
      "step": 2851
    },
    {
      "epoch": 259.2909090909091,
      "grad_norm": 0.27107998728752136,
      "learning_rate": 2.148e-05,
      "loss": 0.3548,
      "step": 2852
    },
    {
      "epoch": 259.3878787878788,
      "grad_norm": 0.3113803267478943,
      "learning_rate": 2.1470000000000003e-05,
      "loss": 0.3208,
      "step": 2853
    },
    {
      "epoch": 259.4848484848485,
      "grad_norm": 0.28532272577285767,
      "learning_rate": 2.146e-05,
      "loss": 0.3793,
      "step": 2854
    },
    {
      "epoch": 259.58181818181816,
      "grad_norm": 0.30134227871894836,
      "learning_rate": 2.145e-05,
      "loss": 0.357,
      "step": 2855
    },
    {
      "epoch": 259.6787878787879,
      "grad_norm": 0.3014954626560211,
      "learning_rate": 2.144e-05,
      "loss": 0.3464,
      "step": 2856
    },
    {
      "epoch": 259.77575757575755,
      "grad_norm": 0.30127498507499695,
      "learning_rate": 2.143e-05,
      "loss": 0.2996,
      "step": 2857
    },
    {
      "epoch": 259.8727272727273,
      "grad_norm": 0.30472543835639954,
      "learning_rate": 2.142e-05,
      "loss": 0.3568,
      "step": 2858
    },
    {
      "epoch": 259.969696969697,
      "grad_norm": 0.2586653530597687,
      "learning_rate": 2.1410000000000003e-05,
      "loss": 0.3592,
      "step": 2859
    },
    {
      "epoch": 260.0,
      "grad_norm": 0.5651054978370667,
      "learning_rate": 2.1400000000000002e-05,
      "loss": 0.4092,
      "step": 2860
    },
    {
      "epoch": 260.0,
      "eval_loss": 0.43332797288894653,
      "eval_runtime": 2.1223,
      "eval_samples_per_second": 25.916,
      "eval_steps_per_second": 3.298,
      "step": 2860
    },
    {
      "epoch": 260.0969696969697,
      "grad_norm": 0.30095285177230835,
      "learning_rate": 2.139e-05,
      "loss": 0.3493,
      "step": 2861
    },
    {
      "epoch": 260.1939393939394,
      "grad_norm": 0.26430273056030273,
      "learning_rate": 2.138e-05,
      "loss": 0.3582,
      "step": 2862
    },
    {
      "epoch": 260.2909090909091,
      "grad_norm": 0.2739563584327698,
      "learning_rate": 2.137e-05,
      "loss": 0.3702,
      "step": 2863
    },
    {
      "epoch": 260.3878787878788,
      "grad_norm": 0.36289191246032715,
      "learning_rate": 2.1360000000000002e-05,
      "loss": 0.359,
      "step": 2864
    },
    {
      "epoch": 260.4848484848485,
      "grad_norm": 0.3343953490257263,
      "learning_rate": 2.135e-05,
      "loss": 0.3568,
      "step": 2865
    },
    {
      "epoch": 260.58181818181816,
      "grad_norm": 0.29125627875328064,
      "learning_rate": 2.1340000000000002e-05,
      "loss": 0.3117,
      "step": 2866
    },
    {
      "epoch": 260.6787878787879,
      "grad_norm": 0.4203968346118927,
      "learning_rate": 2.133e-05,
      "loss": 0.3383,
      "step": 2867
    },
    {
      "epoch": 260.77575757575755,
      "grad_norm": 0.26528501510620117,
      "learning_rate": 2.1320000000000003e-05,
      "loss": 0.3634,
      "step": 2868
    },
    {
      "epoch": 260.8727272727273,
      "grad_norm": 0.2988184988498688,
      "learning_rate": 2.131e-05,
      "loss": 0.3406,
      "step": 2869
    },
    {
      "epoch": 260.969696969697,
      "grad_norm": 0.45513859391212463,
      "learning_rate": 2.13e-05,
      "loss": 0.3663,
      "step": 2870
    },
    {
      "epoch": 260.969696969697,
      "eval_loss": 0.43273502588272095,
      "eval_runtime": 2.1264,
      "eval_samples_per_second": 25.866,
      "eval_steps_per_second": 3.292,
      "step": 2870
    },
    {
      "epoch": 261.0,
      "grad_norm": 0.5738509893417358,
      "learning_rate": 2.129e-05,
      "loss": 0.4098,
      "step": 2871
    },
    {
      "epoch": 261.0969696969697,
      "grad_norm": 0.32520735263824463,
      "learning_rate": 2.128e-05,
      "loss": 0.3516,
      "step": 2872
    },
    {
      "epoch": 261.1939393939394,
      "grad_norm": 0.2425633817911148,
      "learning_rate": 2.127e-05,
      "loss": 0.3491,
      "step": 2873
    },
    {
      "epoch": 261.2909090909091,
      "grad_norm": 0.27022457122802734,
      "learning_rate": 2.1260000000000003e-05,
      "loss": 0.4148,
      "step": 2874
    },
    {
      "epoch": 261.3878787878788,
      "grad_norm": 0.31663596630096436,
      "learning_rate": 2.125e-05,
      "loss": 0.3379,
      "step": 2875
    },
    {
      "epoch": 261.4848484848485,
      "grad_norm": 0.3284948468208313,
      "learning_rate": 2.124e-05,
      "loss": 0.3272,
      "step": 2876
    },
    {
      "epoch": 261.58181818181816,
      "grad_norm": 0.32757529616355896,
      "learning_rate": 2.123e-05,
      "loss": 0.3585,
      "step": 2877
    },
    {
      "epoch": 261.6787878787879,
      "grad_norm": 0.2892166078090668,
      "learning_rate": 2.122e-05,
      "loss": 0.3985,
      "step": 2878
    },
    {
      "epoch": 261.77575757575755,
      "grad_norm": 0.29307177662849426,
      "learning_rate": 2.1210000000000002e-05,
      "loss": 0.3348,
      "step": 2879
    },
    {
      "epoch": 261.8727272727273,
      "grad_norm": 0.34251949191093445,
      "learning_rate": 2.12e-05,
      "loss": 0.3241,
      "step": 2880
    },
    {
      "epoch": 261.8727272727273,
      "eval_loss": 0.4326820373535156,
      "eval_runtime": 2.1437,
      "eval_samples_per_second": 25.657,
      "eval_steps_per_second": 3.265,
      "step": 2880
    },
    {
      "epoch": 261.969696969697,
      "grad_norm": 0.28630900382995605,
      "learning_rate": 2.1190000000000002e-05,
      "loss": 0.3342,
      "step": 2881
    },
    {
      "epoch": 262.0,
      "grad_norm": 0.47029659152030945,
      "learning_rate": 2.118e-05,
      "loss": 0.3435,
      "step": 2882
    },
    {
      "epoch": 262.0969696969697,
      "grad_norm": 0.2874227464199066,
      "learning_rate": 2.1170000000000002e-05,
      "loss": 0.365,
      "step": 2883
    },
    {
      "epoch": 262.1939393939394,
      "grad_norm": 0.2977810502052307,
      "learning_rate": 2.116e-05,
      "loss": 0.3589,
      "step": 2884
    },
    {
      "epoch": 262.2909090909091,
      "grad_norm": 0.2793772518634796,
      "learning_rate": 2.115e-05,
      "loss": 0.38,
      "step": 2885
    },
    {
      "epoch": 262.3878787878788,
      "grad_norm": 0.2945645749568939,
      "learning_rate": 2.114e-05,
      "loss": 0.3388,
      "step": 2886
    },
    {
      "epoch": 262.4848484848485,
      "grad_norm": 0.2820453643798828,
      "learning_rate": 2.113e-05,
      "loss": 0.3481,
      "step": 2887
    },
    {
      "epoch": 262.58181818181816,
      "grad_norm": 0.30547210574150085,
      "learning_rate": 2.112e-05,
      "loss": 0.3654,
      "step": 2888
    },
    {
      "epoch": 262.6787878787879,
      "grad_norm": 0.32220014929771423,
      "learning_rate": 2.1110000000000003e-05,
      "loss": 0.347,
      "step": 2889
    },
    {
      "epoch": 262.77575757575755,
      "grad_norm": 0.28820616006851196,
      "learning_rate": 2.11e-05,
      "loss": 0.368,
      "step": 2890
    },
    {
      "epoch": 262.77575757575755,
      "eval_loss": 0.4324910044670105,
      "eval_runtime": 2.1073,
      "eval_samples_per_second": 26.1,
      "eval_steps_per_second": 3.322,
      "step": 2890
    },
    {
      "epoch": 262.8727272727273,
      "grad_norm": 0.3436179459095001,
      "learning_rate": 2.1090000000000003e-05,
      "loss": 0.3484,
      "step": 2891
    },
    {
      "epoch": 262.969696969697,
      "grad_norm": 0.28017234802246094,
      "learning_rate": 2.1079999999999998e-05,
      "loss": 0.3152,
      "step": 2892
    },
    {
      "epoch": 263.0,
      "grad_norm": 0.4805695116519928,
      "learning_rate": 2.107e-05,
      "loss": 0.3203,
      "step": 2893
    },
    {
      "epoch": 263.0969696969697,
      "grad_norm": 0.2792573869228363,
      "learning_rate": 2.106e-05,
      "loss": 0.3718,
      "step": 2894
    },
    {
      "epoch": 263.1939393939394,
      "grad_norm": 0.27789148688316345,
      "learning_rate": 2.105e-05,
      "loss": 0.3342,
      "step": 2895
    },
    {
      "epoch": 263.2909090909091,
      "grad_norm": 0.29735344648361206,
      "learning_rate": 2.1040000000000002e-05,
      "loss": 0.3603,
      "step": 2896
    },
    {
      "epoch": 263.3878787878788,
      "grad_norm": 0.28099167346954346,
      "learning_rate": 2.103e-05,
      "loss": 0.3775,
      "step": 2897
    },
    {
      "epoch": 263.4848484848485,
      "grad_norm": 0.2803812325000763,
      "learning_rate": 2.1020000000000002e-05,
      "loss": 0.3377,
      "step": 2898
    },
    {
      "epoch": 263.58181818181816,
      "grad_norm": 0.37615877389907837,
      "learning_rate": 2.101e-05,
      "loss": 0.3467,
      "step": 2899
    },
    {
      "epoch": 263.6787878787879,
      "grad_norm": 0.27375122904777527,
      "learning_rate": 2.1e-05,
      "loss": 0.3341,
      "step": 2900
    },
    {
      "epoch": 263.6787878787879,
      "eval_loss": 0.4324527084827423,
      "eval_runtime": 2.1016,
      "eval_samples_per_second": 26.17,
      "eval_steps_per_second": 3.331,
      "step": 2900
    },
    {
      "epoch": 263.77575757575755,
      "grad_norm": 0.30880647897720337,
      "learning_rate": 2.099e-05,
      "loss": 0.3396,
      "step": 2901
    },
    {
      "epoch": 263.8727272727273,
      "grad_norm": 0.32526326179504395,
      "learning_rate": 2.098e-05,
      "loss": 0.3562,
      "step": 2902
    },
    {
      "epoch": 263.969696969697,
      "grad_norm": 0.2949479818344116,
      "learning_rate": 2.097e-05,
      "loss": 0.3509,
      "step": 2903
    },
    {
      "epoch": 264.0,
      "grad_norm": 0.6180709600448608,
      "learning_rate": 2.0960000000000003e-05,
      "loss": 0.3928,
      "step": 2904
    },
    {
      "epoch": 264.0969696969697,
      "grad_norm": 0.2905566990375519,
      "learning_rate": 2.095e-05,
      "loss": 0.3352,
      "step": 2905
    },
    {
      "epoch": 264.1939393939394,
      "grad_norm": 0.28563404083251953,
      "learning_rate": 2.0940000000000003e-05,
      "loss": 0.3304,
      "step": 2906
    },
    {
      "epoch": 264.2909090909091,
      "grad_norm": 0.29980409145355225,
      "learning_rate": 2.093e-05,
      "loss": 0.408,
      "step": 2907
    },
    {
      "epoch": 264.3878787878788,
      "grad_norm": 0.2590050995349884,
      "learning_rate": 2.092e-05,
      "loss": 0.345,
      "step": 2908
    },
    {
      "epoch": 264.4848484848485,
      "grad_norm": 0.321993887424469,
      "learning_rate": 2.091e-05,
      "loss": 0.328,
      "step": 2909
    },
    {
      "epoch": 264.58181818181816,
      "grad_norm": 0.23598870635032654,
      "learning_rate": 2.09e-05,
      "loss": 0.3337,
      "step": 2910
    },
    {
      "epoch": 264.58181818181816,
      "eval_loss": 0.4328293204307556,
      "eval_runtime": 2.0996,
      "eval_samples_per_second": 26.195,
      "eval_steps_per_second": 3.334,
      "step": 2910
    },
    {
      "epoch": 264.6787878787879,
      "grad_norm": 0.304935485124588,
      "learning_rate": 2.089e-05,
      "loss": 0.3495,
      "step": 2911
    },
    {
      "epoch": 264.77575757575755,
      "grad_norm": 0.2994263768196106,
      "learning_rate": 2.0880000000000003e-05,
      "loss": 0.3487,
      "step": 2912
    },
    {
      "epoch": 264.8727272727273,
      "grad_norm": 0.29449424147605896,
      "learning_rate": 2.0870000000000002e-05,
      "loss": 0.3425,
      "step": 2913
    },
    {
      "epoch": 264.969696969697,
      "grad_norm": 0.3142696022987366,
      "learning_rate": 2.086e-05,
      "loss": 0.4075,
      "step": 2914
    },
    {
      "epoch": 265.0,
      "grad_norm": 0.40163326263427734,
      "learning_rate": 2.085e-05,
      "loss": 0.3102,
      "step": 2915
    },
    {
      "epoch": 265.0969696969697,
      "grad_norm": 0.30728578567504883,
      "learning_rate": 2.084e-05,
      "loss": 0.3383,
      "step": 2916
    },
    {
      "epoch": 265.1939393939394,
      "grad_norm": 0.3148532211780548,
      "learning_rate": 2.0830000000000002e-05,
      "loss": 0.3846,
      "step": 2917
    },
    {
      "epoch": 265.2909090909091,
      "grad_norm": 0.27514949440956116,
      "learning_rate": 2.082e-05,
      "loss": 0.3516,
      "step": 2918
    },
    {
      "epoch": 265.3878787878788,
      "grad_norm": 0.2847309112548828,
      "learning_rate": 2.0810000000000002e-05,
      "loss": 0.2964,
      "step": 2919
    },
    {
      "epoch": 265.4848484848485,
      "grad_norm": 0.2686462104320526,
      "learning_rate": 2.08e-05,
      "loss": 0.3298,
      "step": 2920
    },
    {
      "epoch": 265.4848484848485,
      "eval_loss": 0.43240609765052795,
      "eval_runtime": 2.1075,
      "eval_samples_per_second": 26.097,
      "eval_steps_per_second": 3.321,
      "step": 2920
    },
    {
      "epoch": 265.58181818181816,
      "grad_norm": 0.2931896150112152,
      "learning_rate": 2.0790000000000003e-05,
      "loss": 0.3495,
      "step": 2921
    },
    {
      "epoch": 265.6787878787879,
      "grad_norm": 0.34141311049461365,
      "learning_rate": 2.078e-05,
      "loss": 0.3247,
      "step": 2922
    },
    {
      "epoch": 265.77575757575755,
      "grad_norm": 0.3040899336338043,
      "learning_rate": 2.077e-05,
      "loss": 0.38,
      "step": 2923
    },
    {
      "epoch": 265.8727272727273,
      "grad_norm": 0.3186739981174469,
      "learning_rate": 2.076e-05,
      "loss": 0.3767,
      "step": 2924
    },
    {
      "epoch": 265.969696969697,
      "grad_norm": 0.3134538233280182,
      "learning_rate": 2.075e-05,
      "loss": 0.3834,
      "step": 2925
    },
    {
      "epoch": 266.0,
      "grad_norm": 0.44522473216056824,
      "learning_rate": 2.074e-05,
      "loss": 0.3597,
      "step": 2926
    },
    {
      "epoch": 266.0969696969697,
      "grad_norm": 0.3128274381160736,
      "learning_rate": 2.0730000000000003e-05,
      "loss": 0.3552,
      "step": 2927
    },
    {
      "epoch": 266.1939393939394,
      "grad_norm": 0.27766191959381104,
      "learning_rate": 2.072e-05,
      "loss": 0.358,
      "step": 2928
    },
    {
      "epoch": 266.2909090909091,
      "grad_norm": 0.3204481303691864,
      "learning_rate": 2.0710000000000003e-05,
      "loss": 0.3259,
      "step": 2929
    },
    {
      "epoch": 266.3878787878788,
      "grad_norm": 0.2819754183292389,
      "learning_rate": 2.07e-05,
      "loss": 0.3445,
      "step": 2930
    },
    {
      "epoch": 266.3878787878788,
      "eval_loss": 0.43232443928718567,
      "eval_runtime": 2.1158,
      "eval_samples_per_second": 25.995,
      "eval_steps_per_second": 3.308,
      "step": 2930
    },
    {
      "epoch": 266.4848484848485,
      "grad_norm": 0.33887726068496704,
      "learning_rate": 2.069e-05,
      "loss": 0.3622,
      "step": 2931
    },
    {
      "epoch": 266.58181818181816,
      "grad_norm": 0.27619192004203796,
      "learning_rate": 2.0680000000000002e-05,
      "loss": 0.3591,
      "step": 2932
    },
    {
      "epoch": 266.6787878787879,
      "grad_norm": 0.2761163115501404,
      "learning_rate": 2.067e-05,
      "loss": 0.3481,
      "step": 2933
    },
    {
      "epoch": 266.77575757575755,
      "grad_norm": 0.2995469272136688,
      "learning_rate": 2.0660000000000002e-05,
      "loss": 0.3578,
      "step": 2934
    },
    {
      "epoch": 266.8727272727273,
      "grad_norm": 0.26977553963661194,
      "learning_rate": 2.065e-05,
      "loss": 0.3536,
      "step": 2935
    },
    {
      "epoch": 266.969696969697,
      "grad_norm": 0.2937801480293274,
      "learning_rate": 2.0640000000000002e-05,
      "loss": 0.3335,
      "step": 2936
    },
    {
      "epoch": 267.0,
      "grad_norm": 0.5464254021644592,
      "learning_rate": 2.063e-05,
      "loss": 0.4013,
      "step": 2937
    },
    {
      "epoch": 267.0969696969697,
      "grad_norm": 0.2871507704257965,
      "learning_rate": 2.062e-05,
      "loss": 0.3432,
      "step": 2938
    },
    {
      "epoch": 267.1939393939394,
      "grad_norm": 0.3833047151565552,
      "learning_rate": 2.061e-05,
      "loss": 0.3692,
      "step": 2939
    },
    {
      "epoch": 267.2909090909091,
      "grad_norm": 0.2943490445613861,
      "learning_rate": 2.06e-05,
      "loss": 0.3036,
      "step": 2940
    },
    {
      "epoch": 267.2909090909091,
      "eval_loss": 0.43172433972358704,
      "eval_runtime": 2.1026,
      "eval_samples_per_second": 26.158,
      "eval_steps_per_second": 3.329,
      "step": 2940
    },
    {
      "epoch": 267.3878787878788,
      "grad_norm": 0.2477860301733017,
      "learning_rate": 2.059e-05,
      "loss": 0.372,
      "step": 2941
    },
    {
      "epoch": 267.4848484848485,
      "grad_norm": 0.2954455018043518,
      "learning_rate": 2.0580000000000003e-05,
      "loss": 0.3343,
      "step": 2942
    },
    {
      "epoch": 267.58181818181816,
      "grad_norm": 0.2852739989757538,
      "learning_rate": 2.057e-05,
      "loss": 0.329,
      "step": 2943
    },
    {
      "epoch": 267.6787878787879,
      "grad_norm": 0.3311695158481598,
      "learning_rate": 2.0560000000000003e-05,
      "loss": 0.3626,
      "step": 2944
    },
    {
      "epoch": 267.77575757575755,
      "grad_norm": 0.3276229798793793,
      "learning_rate": 2.055e-05,
      "loss": 0.3642,
      "step": 2945
    },
    {
      "epoch": 267.8727272727273,
      "grad_norm": 0.35247862339019775,
      "learning_rate": 2.054e-05,
      "loss": 0.3787,
      "step": 2946
    },
    {
      "epoch": 267.969696969697,
      "grad_norm": 0.24730032682418823,
      "learning_rate": 2.053e-05,
      "loss": 0.3486,
      "step": 2947
    },
    {
      "epoch": 268.0,
      "grad_norm": 0.4214022755622864,
      "learning_rate": 2.052e-05,
      "loss": 0.3604,
      "step": 2948
    },
    {
      "epoch": 268.0969696969697,
      "grad_norm": 0.28409886360168457,
      "learning_rate": 2.0510000000000002e-05,
      "loss": 0.3571,
      "step": 2949
    },
    {
      "epoch": 268.1939393939394,
      "grad_norm": 0.27273720502853394,
      "learning_rate": 2.05e-05,
      "loss": 0.3489,
      "step": 2950
    },
    {
      "epoch": 268.1939393939394,
      "eval_loss": 0.4321860373020172,
      "eval_runtime": 2.1649,
      "eval_samples_per_second": 25.405,
      "eval_steps_per_second": 3.233,
      "step": 2950
    },
    {
      "epoch": 268.2909090909091,
      "grad_norm": 0.3224973976612091,
      "learning_rate": 2.0490000000000002e-05,
      "loss": 0.3593,
      "step": 2951
    },
    {
      "epoch": 268.3878787878788,
      "grad_norm": 0.29232099652290344,
      "learning_rate": 2.048e-05,
      "loss": 0.3588,
      "step": 2952
    },
    {
      "epoch": 268.4848484848485,
      "grad_norm": 0.30897924304008484,
      "learning_rate": 2.047e-05,
      "loss": 0.3953,
      "step": 2953
    },
    {
      "epoch": 268.58181818181816,
      "grad_norm": 0.3023488223552704,
      "learning_rate": 2.046e-05,
      "loss": 0.3967,
      "step": 2954
    },
    {
      "epoch": 268.6787878787879,
      "grad_norm": 0.2875419557094574,
      "learning_rate": 2.045e-05,
      "loss": 0.3137,
      "step": 2955
    },
    {
      "epoch": 268.77575757575755,
      "grad_norm": 0.30910709500312805,
      "learning_rate": 2.044e-05,
      "loss": 0.3439,
      "step": 2956
    },
    {
      "epoch": 268.8727272727273,
      "grad_norm": 0.3412000834941864,
      "learning_rate": 2.0430000000000003e-05,
      "loss": 0.3213,
      "step": 2957
    },
    {
      "epoch": 268.969696969697,
      "grad_norm": 0.27936646342277527,
      "learning_rate": 2.042e-05,
      "loss": 0.3279,
      "step": 2958
    },
    {
      "epoch": 269.0,
      "grad_norm": 0.4301643371582031,
      "learning_rate": 2.0410000000000003e-05,
      "loss": 0.2894,
      "step": 2959
    },
    {
      "epoch": 269.0969696969697,
      "grad_norm": 0.26649555563926697,
      "learning_rate": 2.04e-05,
      "loss": 0.3296,
      "step": 2960
    },
    {
      "epoch": 269.0969696969697,
      "eval_loss": 0.43194034695625305,
      "eval_runtime": 2.1128,
      "eval_samples_per_second": 26.031,
      "eval_steps_per_second": 3.313,
      "step": 2960
    },
    {
      "epoch": 269.1939393939394,
      "grad_norm": 0.28858742117881775,
      "learning_rate": 2.039e-05,
      "loss": 0.3605,
      "step": 2961
    },
    {
      "epoch": 269.2909090909091,
      "grad_norm": 0.3377971053123474,
      "learning_rate": 2.038e-05,
      "loss": 0.3219,
      "step": 2962
    },
    {
      "epoch": 269.3878787878788,
      "grad_norm": 0.3358132839202881,
      "learning_rate": 2.037e-05,
      "loss": 0.343,
      "step": 2963
    },
    {
      "epoch": 269.4848484848485,
      "grad_norm": 0.3058205544948578,
      "learning_rate": 2.036e-05,
      "loss": 0.3609,
      "step": 2964
    },
    {
      "epoch": 269.58181818181816,
      "grad_norm": 0.3367607295513153,
      "learning_rate": 2.035e-05,
      "loss": 0.347,
      "step": 2965
    },
    {
      "epoch": 269.6787878787879,
      "grad_norm": 0.33359044790267944,
      "learning_rate": 2.0340000000000002e-05,
      "loss": 0.3899,
      "step": 2966
    },
    {
      "epoch": 269.77575757575755,
      "grad_norm": 0.28831222653388977,
      "learning_rate": 2.033e-05,
      "loss": 0.3559,
      "step": 2967
    },
    {
      "epoch": 269.8727272727273,
      "grad_norm": 0.24753743410110474,
      "learning_rate": 2.032e-05,
      "loss": 0.3609,
      "step": 2968
    },
    {
      "epoch": 269.969696969697,
      "grad_norm": 0.3068210482597351,
      "learning_rate": 2.031e-05,
      "loss": 0.3495,
      "step": 2969
    },
    {
      "epoch": 270.0,
      "grad_norm": 0.5752652883529663,
      "learning_rate": 2.0300000000000002e-05,
      "loss": 0.3073,
      "step": 2970
    },
    {
      "epoch": 270.0,
      "eval_loss": 0.43105173110961914,
      "eval_runtime": 2.1156,
      "eval_samples_per_second": 25.997,
      "eval_steps_per_second": 3.309,
      "step": 2970
    },
    {
      "epoch": 270.0969696969697,
      "grad_norm": 0.289946973323822,
      "learning_rate": 2.029e-05,
      "loss": 0.3672,
      "step": 2971
    },
    {
      "epoch": 270.1939393939394,
      "grad_norm": 0.2912636399269104,
      "learning_rate": 2.0280000000000002e-05,
      "loss": 0.3491,
      "step": 2972
    },
    {
      "epoch": 270.2909090909091,
      "grad_norm": 0.4258723258972168,
      "learning_rate": 2.027e-05,
      "loss": 0.3439,
      "step": 2973
    },
    {
      "epoch": 270.3878787878788,
      "grad_norm": 0.24976497888565063,
      "learning_rate": 2.0260000000000003e-05,
      "loss": 0.3305,
      "step": 2974
    },
    {
      "epoch": 270.4848484848485,
      "grad_norm": 0.3167341351509094,
      "learning_rate": 2.025e-05,
      "loss": 0.3744,
      "step": 2975
    },
    {
      "epoch": 270.58181818181816,
      "grad_norm": 0.3470234274864197,
      "learning_rate": 2.024e-05,
      "loss": 0.3286,
      "step": 2976
    },
    {
      "epoch": 270.6787878787879,
      "grad_norm": 0.30521225929260254,
      "learning_rate": 2.023e-05,
      "loss": 0.3427,
      "step": 2977
    },
    {
      "epoch": 270.77575757575755,
      "grad_norm": 0.3303682804107666,
      "learning_rate": 2.022e-05,
      "loss": 0.371,
      "step": 2978
    },
    {
      "epoch": 270.8727272727273,
      "grad_norm": 0.31925997138023376,
      "learning_rate": 2.021e-05,
      "loss": 0.3558,
      "step": 2979
    },
    {
      "epoch": 270.969696969697,
      "grad_norm": 0.3036620616912842,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 0.3245,
      "step": 2980
    },
    {
      "epoch": 270.969696969697,
      "eval_loss": 0.4314921498298645,
      "eval_runtime": 2.1128,
      "eval_samples_per_second": 26.032,
      "eval_steps_per_second": 3.313,
      "step": 2980
    },
    {
      "epoch": 271.0,
      "grad_norm": 0.5975680351257324,
      "learning_rate": 2.019e-05,
      "loss": 0.3874,
      "step": 2981
    },
    {
      "epoch": 271.0969696969697,
      "grad_norm": 0.2873639464378357,
      "learning_rate": 2.0180000000000003e-05,
      "loss": 0.3234,
      "step": 2982
    },
    {
      "epoch": 271.1939393939394,
      "grad_norm": 0.2703818082809448,
      "learning_rate": 2.017e-05,
      "loss": 0.345,
      "step": 2983
    },
    {
      "epoch": 271.2909090909091,
      "grad_norm": 0.3004109561443329,
      "learning_rate": 2.016e-05,
      "loss": 0.3726,
      "step": 2984
    },
    {
      "epoch": 271.3878787878788,
      "grad_norm": 0.29286929965019226,
      "learning_rate": 2.0150000000000002e-05,
      "loss": 0.3264,
      "step": 2985
    },
    {
      "epoch": 271.4848484848485,
      "grad_norm": 0.3061460554599762,
      "learning_rate": 2.014e-05,
      "loss": 0.338,
      "step": 2986
    },
    {
      "epoch": 271.58181818181816,
      "grad_norm": 0.3210451304912567,
      "learning_rate": 2.0130000000000002e-05,
      "loss": 0.3491,
      "step": 2987
    },
    {
      "epoch": 271.6787878787879,
      "grad_norm": 0.3391363024711609,
      "learning_rate": 2.012e-05,
      "loss": 0.3642,
      "step": 2988
    },
    {
      "epoch": 271.77575757575755,
      "grad_norm": 0.3371747136116028,
      "learning_rate": 2.0110000000000002e-05,
      "loss": 0.3192,
      "step": 2989
    },
    {
      "epoch": 271.8727272727273,
      "grad_norm": 0.29019173979759216,
      "learning_rate": 2.01e-05,
      "loss": 0.3577,
      "step": 2990
    },
    {
      "epoch": 271.8727272727273,
      "eval_loss": 0.43195751309394836,
      "eval_runtime": 2.1447,
      "eval_samples_per_second": 25.645,
      "eval_steps_per_second": 3.264,
      "step": 2990
    },
    {
      "epoch": 271.969696969697,
      "grad_norm": 0.3430919945240021,
      "learning_rate": 2.009e-05,
      "loss": 0.3814,
      "step": 2991
    },
    {
      "epoch": 272.0,
      "grad_norm": 0.6806743144989014,
      "learning_rate": 2.008e-05,
      "loss": 0.4202,
      "step": 2992
    },
    {
      "epoch": 272.0969696969697,
      "grad_norm": 0.323353111743927,
      "learning_rate": 2.007e-05,
      "loss": 0.3634,
      "step": 2993
    },
    {
      "epoch": 272.1939393939394,
      "grad_norm": 0.3096074163913727,
      "learning_rate": 2.006e-05,
      "loss": 0.3412,
      "step": 2994
    },
    {
      "epoch": 272.2909090909091,
      "grad_norm": 0.2603781819343567,
      "learning_rate": 2.0050000000000003e-05,
      "loss": 0.3313,
      "step": 2995
    },
    {
      "epoch": 272.3878787878788,
      "grad_norm": 0.38184812664985657,
      "learning_rate": 2.004e-05,
      "loss": 0.3773,
      "step": 2996
    },
    {
      "epoch": 272.4848484848485,
      "grad_norm": 0.28190943598747253,
      "learning_rate": 2.0030000000000003e-05,
      "loss": 0.3858,
      "step": 2997
    },
    {
      "epoch": 272.58181818181816,
      "grad_norm": 0.29366445541381836,
      "learning_rate": 2.002e-05,
      "loss": 0.3654,
      "step": 2998
    },
    {
      "epoch": 272.6787878787879,
      "grad_norm": 0.3181076645851135,
      "learning_rate": 2.001e-05,
      "loss": 0.3091,
      "step": 2999
    },
    {
      "epoch": 272.77575757575755,
      "grad_norm": 0.31197789311408997,
      "learning_rate": 2e-05,
      "loss": 0.3291,
      "step": 3000
    },
    {
      "epoch": 272.77575757575755,
      "eval_loss": 0.43125858902931213,
      "eval_runtime": 2.1098,
      "eval_samples_per_second": 26.068,
      "eval_steps_per_second": 3.318,
      "step": 3000
    },
    {
      "epoch": 272.8727272727273,
      "grad_norm": 0.27228426933288574,
      "learning_rate": 1.999e-05,
      "loss": 0.314,
      "step": 3001
    },
    {
      "epoch": 272.969696969697,
      "grad_norm": 0.27844688296318054,
      "learning_rate": 1.9980000000000002e-05,
      "loss": 0.381,
      "step": 3002
    },
    {
      "epoch": 273.0,
      "grad_norm": 0.5656706094741821,
      "learning_rate": 1.997e-05,
      "loss": 0.3438,
      "step": 3003
    },
    {
      "epoch": 273.0969696969697,
      "grad_norm": 0.3047666549682617,
      "learning_rate": 1.9960000000000002e-05,
      "loss": 0.3674,
      "step": 3004
    },
    {
      "epoch": 273.1939393939394,
      "grad_norm": 0.2743180990219116,
      "learning_rate": 1.995e-05,
      "loss": 0.3293,
      "step": 3005
    },
    {
      "epoch": 273.2909090909091,
      "grad_norm": 0.3302759826183319,
      "learning_rate": 1.994e-05,
      "loss": 0.3426,
      "step": 3006
    },
    {
      "epoch": 273.3878787878788,
      "grad_norm": 0.3315967321395874,
      "learning_rate": 1.993e-05,
      "loss": 0.3581,
      "step": 3007
    },
    {
      "epoch": 273.4848484848485,
      "grad_norm": 0.3116733133792877,
      "learning_rate": 1.992e-05,
      "loss": 0.3407,
      "step": 3008
    },
    {
      "epoch": 273.58181818181816,
      "grad_norm": 0.27811938524246216,
      "learning_rate": 1.991e-05,
      "loss": 0.3487,
      "step": 3009
    },
    {
      "epoch": 273.6787878787879,
      "grad_norm": 0.29493939876556396,
      "learning_rate": 1.9900000000000003e-05,
      "loss": 0.3717,
      "step": 3010
    },
    {
      "epoch": 273.6787878787879,
      "eval_loss": 0.4310474693775177,
      "eval_runtime": 2.1139,
      "eval_samples_per_second": 26.018,
      "eval_steps_per_second": 3.311,
      "step": 3010
    },
    {
      "epoch": 273.77575757575755,
      "grad_norm": 0.2869589626789093,
      "learning_rate": 1.989e-05,
      "loss": 0.3577,
      "step": 3011
    },
    {
      "epoch": 273.8727272727273,
      "grad_norm": 0.32067158818244934,
      "learning_rate": 1.9880000000000003e-05,
      "loss": 0.3186,
      "step": 3012
    },
    {
      "epoch": 273.969696969697,
      "grad_norm": 0.3775586485862732,
      "learning_rate": 1.987e-05,
      "loss": 0.3697,
      "step": 3013
    },
    {
      "epoch": 274.0,
      "grad_norm": 0.4237777888774872,
      "learning_rate": 1.986e-05,
      "loss": 0.313,
      "step": 3014
    },
    {
      "epoch": 274.0969696969697,
      "grad_norm": 0.2924163341522217,
      "learning_rate": 1.985e-05,
      "loss": 0.3671,
      "step": 3015
    },
    {
      "epoch": 274.1939393939394,
      "grad_norm": 0.29473182559013367,
      "learning_rate": 1.984e-05,
      "loss": 0.376,
      "step": 3016
    },
    {
      "epoch": 274.2909090909091,
      "grad_norm": 0.35198548436164856,
      "learning_rate": 1.983e-05,
      "loss": 0.3578,
      "step": 3017
    },
    {
      "epoch": 274.3878787878788,
      "grad_norm": 0.2765541672706604,
      "learning_rate": 1.982e-05,
      "loss": 0.3727,
      "step": 3018
    },
    {
      "epoch": 274.4848484848485,
      "grad_norm": 0.3288250267505646,
      "learning_rate": 1.9810000000000002e-05,
      "loss": 0.329,
      "step": 3019
    },
    {
      "epoch": 274.58181818181816,
      "grad_norm": 0.3038225769996643,
      "learning_rate": 1.9800000000000004e-05,
      "loss": 0.3089,
      "step": 3020
    },
    {
      "epoch": 274.58181818181816,
      "eval_loss": 0.43139633536338806,
      "eval_runtime": 2.1041,
      "eval_samples_per_second": 26.139,
      "eval_steps_per_second": 3.327,
      "step": 3020
    },
    {
      "epoch": 274.6787878787879,
      "grad_norm": 0.28620851039886475,
      "learning_rate": 1.979e-05,
      "loss": 0.3792,
      "step": 3021
    },
    {
      "epoch": 274.77575757575755,
      "grad_norm": 0.24123451113700867,
      "learning_rate": 1.978e-05,
      "loss": 0.3362,
      "step": 3022
    },
    {
      "epoch": 274.8727272727273,
      "grad_norm": 0.2994822859764099,
      "learning_rate": 1.977e-05,
      "loss": 0.3286,
      "step": 3023
    },
    {
      "epoch": 274.969696969697,
      "grad_norm": 0.2507471442222595,
      "learning_rate": 1.976e-05,
      "loss": 0.3346,
      "step": 3024
    },
    {
      "epoch": 275.0,
      "grad_norm": 0.49847498536109924,
      "learning_rate": 1.9750000000000002e-05,
      "loss": 0.3411,
      "step": 3025
    },
    {
      "epoch": 275.0969696969697,
      "grad_norm": 0.30484265089035034,
      "learning_rate": 1.974e-05,
      "loss": 0.3269,
      "step": 3026
    },
    {
      "epoch": 275.1939393939394,
      "grad_norm": 0.3385094702243805,
      "learning_rate": 1.9730000000000003e-05,
      "loss": 0.3751,
      "step": 3027
    },
    {
      "epoch": 275.2909090909091,
      "grad_norm": 0.27394574880599976,
      "learning_rate": 1.972e-05,
      "loss": 0.3457,
      "step": 3028
    },
    {
      "epoch": 275.3878787878788,
      "grad_norm": 0.3241487443447113,
      "learning_rate": 1.971e-05,
      "loss": 0.3625,
      "step": 3029
    },
    {
      "epoch": 275.4848484848485,
      "grad_norm": 0.2895892262458801,
      "learning_rate": 1.97e-05,
      "loss": 0.3576,
      "step": 3030
    },
    {
      "epoch": 275.4848484848485,
      "eval_loss": 0.43097203969955444,
      "eval_runtime": 2.1081,
      "eval_samples_per_second": 26.09,
      "eval_steps_per_second": 3.321,
      "step": 3030
    },
    {
      "epoch": 275.58181818181816,
      "grad_norm": 0.3027297556400299,
      "learning_rate": 1.969e-05,
      "loss": 0.3301,
      "step": 3031
    },
    {
      "epoch": 275.6787878787879,
      "grad_norm": 0.34131574630737305,
      "learning_rate": 1.968e-05,
      "loss": 0.3459,
      "step": 3032
    },
    {
      "epoch": 275.77575757575755,
      "grad_norm": 0.2701811194419861,
      "learning_rate": 1.9670000000000003e-05,
      "loss": 0.345,
      "step": 3033
    },
    {
      "epoch": 275.8727272727273,
      "grad_norm": 0.337098628282547,
      "learning_rate": 1.966e-05,
      "loss": 0.3745,
      "step": 3034
    },
    {
      "epoch": 275.969696969697,
      "grad_norm": 0.3441219627857208,
      "learning_rate": 1.9650000000000003e-05,
      "loss": 0.3189,
      "step": 3035
    },
    {
      "epoch": 276.0,
      "grad_norm": 0.717963457107544,
      "learning_rate": 1.9640000000000002e-05,
      "loss": 0.3628,
      "step": 3036
    },
    {
      "epoch": 276.0969696969697,
      "grad_norm": 0.27845311164855957,
      "learning_rate": 1.963e-05,
      "loss": 0.3544,
      "step": 3037
    },
    {
      "epoch": 276.1939393939394,
      "grad_norm": 0.36221420764923096,
      "learning_rate": 1.9620000000000002e-05,
      "loss": 0.3806,
      "step": 3038
    },
    {
      "epoch": 276.2909090909091,
      "grad_norm": 0.34527480602264404,
      "learning_rate": 1.961e-05,
      "loss": 0.3743,
      "step": 3039
    },
    {
      "epoch": 276.3878787878788,
      "grad_norm": 0.3114314675331116,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.3565,
      "step": 3040
    },
    {
      "epoch": 276.3878787878788,
      "eval_loss": 0.43109825253486633,
      "eval_runtime": 2.131,
      "eval_samples_per_second": 25.81,
      "eval_steps_per_second": 3.285,
      "step": 3040
    },
    {
      "epoch": 276.4848484848485,
      "grad_norm": 0.3695688843727112,
      "learning_rate": 1.959e-05,
      "loss": 0.3316,
      "step": 3041
    },
    {
      "epoch": 276.58181818181816,
      "grad_norm": 0.29446908831596375,
      "learning_rate": 1.9580000000000002e-05,
      "loss": 0.3476,
      "step": 3042
    },
    {
      "epoch": 276.6787878787879,
      "grad_norm": 0.30297887325286865,
      "learning_rate": 1.957e-05,
      "loss": 0.3114,
      "step": 3043
    },
    {
      "epoch": 276.77575757575755,
      "grad_norm": 0.30303916335105896,
      "learning_rate": 1.956e-05,
      "loss": 0.3298,
      "step": 3044
    },
    {
      "epoch": 276.8727272727273,
      "grad_norm": 0.28460606932640076,
      "learning_rate": 1.955e-05,
      "loss": 0.3494,
      "step": 3045
    },
    {
      "epoch": 276.969696969697,
      "grad_norm": 0.3434138596057892,
      "learning_rate": 1.954e-05,
      "loss": 0.3583,
      "step": 3046
    },
    {
      "epoch": 277.0,
      "grad_norm": 0.4618801474571228,
      "learning_rate": 1.953e-05,
      "loss": 0.3074,
      "step": 3047
    },
    {
      "epoch": 277.0969696969697,
      "grad_norm": 0.32681795954704285,
      "learning_rate": 1.9520000000000003e-05,
      "loss": 0.318,
      "step": 3048
    },
    {
      "epoch": 277.1939393939394,
      "grad_norm": 0.2953447699546814,
      "learning_rate": 1.951e-05,
      "loss": 0.374,
      "step": 3049
    },
    {
      "epoch": 277.2909090909091,
      "grad_norm": 0.28921085596084595,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 0.3511,
      "step": 3050
    },
    {
      "epoch": 277.2909090909091,
      "eval_loss": 0.4307308495044708,
      "eval_runtime": 2.1197,
      "eval_samples_per_second": 25.947,
      "eval_steps_per_second": 3.302,
      "step": 3050
    },
    {
      "epoch": 277.3878787878788,
      "grad_norm": 0.3594936430454254,
      "learning_rate": 1.949e-05,
      "loss": 0.3635,
      "step": 3051
    },
    {
      "epoch": 277.4848484848485,
      "grad_norm": 0.34557226300239563,
      "learning_rate": 1.948e-05,
      "loss": 0.3457,
      "step": 3052
    },
    {
      "epoch": 277.58181818181816,
      "grad_norm": 0.2568785548210144,
      "learning_rate": 1.947e-05,
      "loss": 0.3378,
      "step": 3053
    },
    {
      "epoch": 277.6787878787879,
      "grad_norm": 0.27184462547302246,
      "learning_rate": 1.946e-05,
      "loss": 0.3379,
      "step": 3054
    },
    {
      "epoch": 277.77575757575755,
      "grad_norm": 0.29772868752479553,
      "learning_rate": 1.9450000000000002e-05,
      "loss": 0.3628,
      "step": 3055
    },
    {
      "epoch": 277.8727272727273,
      "grad_norm": 0.35863181948661804,
      "learning_rate": 1.944e-05,
      "loss": 0.3368,
      "step": 3056
    },
    {
      "epoch": 277.969696969697,
      "grad_norm": 0.28443828225135803,
      "learning_rate": 1.9430000000000002e-05,
      "loss": 0.3488,
      "step": 3057
    },
    {
      "epoch": 278.0,
      "grad_norm": 0.4689065217971802,
      "learning_rate": 1.942e-05,
      "loss": 0.3654,
      "step": 3058
    },
    {
      "epoch": 278.0969696969697,
      "grad_norm": 0.27923905849456787,
      "learning_rate": 1.941e-05,
      "loss": 0.3318,
      "step": 3059
    },
    {
      "epoch": 278.1939393939394,
      "grad_norm": 0.28366586565971375,
      "learning_rate": 1.94e-05,
      "loss": 0.381,
      "step": 3060
    },
    {
      "epoch": 278.1939393939394,
      "eval_loss": 0.43080776929855347,
      "eval_runtime": 2.106,
      "eval_samples_per_second": 26.116,
      "eval_steps_per_second": 3.324,
      "step": 3060
    },
    {
      "epoch": 278.2909090909091,
      "grad_norm": 0.29033416509628296,
      "learning_rate": 1.939e-05,
      "loss": 0.2994,
      "step": 3061
    },
    {
      "epoch": 278.3878787878788,
      "grad_norm": 0.29525431990623474,
      "learning_rate": 1.938e-05,
      "loss": 0.3606,
      "step": 3062
    },
    {
      "epoch": 278.4848484848485,
      "grad_norm": 0.29698970913887024,
      "learning_rate": 1.9370000000000003e-05,
      "loss": 0.3392,
      "step": 3063
    },
    {
      "epoch": 278.58181818181816,
      "grad_norm": 0.27126365900039673,
      "learning_rate": 1.936e-05,
      "loss": 0.3429,
      "step": 3064
    },
    {
      "epoch": 278.6787878787879,
      "grad_norm": 0.3190431594848633,
      "learning_rate": 1.9350000000000003e-05,
      "loss": 0.3631,
      "step": 3065
    },
    {
      "epoch": 278.77575757575755,
      "grad_norm": 0.3183770179748535,
      "learning_rate": 1.934e-05,
      "loss": 0.3582,
      "step": 3066
    },
    {
      "epoch": 278.8727272727273,
      "grad_norm": 0.28862932324409485,
      "learning_rate": 1.933e-05,
      "loss": 0.3427,
      "step": 3067
    },
    {
      "epoch": 278.969696969697,
      "grad_norm": 0.31062668561935425,
      "learning_rate": 1.932e-05,
      "loss": 0.3639,
      "step": 3068
    },
    {
      "epoch": 279.0,
      "grad_norm": 0.3792564570903778,
      "learning_rate": 1.931e-05,
      "loss": 0.3284,
      "step": 3069
    },
    {
      "epoch": 279.0969696969697,
      "grad_norm": 0.26903533935546875,
      "learning_rate": 1.93e-05,
      "loss": 0.3243,
      "step": 3070
    },
    {
      "epoch": 279.0969696969697,
      "eval_loss": 0.4307677447795868,
      "eval_runtime": 2.0965,
      "eval_samples_per_second": 26.234,
      "eval_steps_per_second": 3.339,
      "step": 3070
    },
    {
      "epoch": 279.1939393939394,
      "grad_norm": 0.28425025939941406,
      "learning_rate": 1.929e-05,
      "loss": 0.3522,
      "step": 3071
    },
    {
      "epoch": 279.2909090909091,
      "grad_norm": 0.2826736867427826,
      "learning_rate": 1.9280000000000002e-05,
      "loss": 0.3038,
      "step": 3072
    },
    {
      "epoch": 279.3878787878788,
      "grad_norm": 0.3037801682949066,
      "learning_rate": 1.9270000000000004e-05,
      "loss": 0.3718,
      "step": 3073
    },
    {
      "epoch": 279.4848484848485,
      "grad_norm": 0.319252610206604,
      "learning_rate": 1.9260000000000002e-05,
      "loss": 0.3388,
      "step": 3074
    },
    {
      "epoch": 279.58181818181816,
      "grad_norm": 0.29738590121269226,
      "learning_rate": 1.925e-05,
      "loss": 0.3716,
      "step": 3075
    },
    {
      "epoch": 279.6787878787879,
      "grad_norm": 0.28336021304130554,
      "learning_rate": 1.924e-05,
      "loss": 0.3868,
      "step": 3076
    },
    {
      "epoch": 279.77575757575755,
      "grad_norm": 0.29788967967033386,
      "learning_rate": 1.923e-05,
      "loss": 0.3595,
      "step": 3077
    },
    {
      "epoch": 279.8727272727273,
      "grad_norm": 0.2871606647968292,
      "learning_rate": 1.9220000000000002e-05,
      "loss": 0.3453,
      "step": 3078
    },
    {
      "epoch": 279.969696969697,
      "grad_norm": 0.32829487323760986,
      "learning_rate": 1.921e-05,
      "loss": 0.3364,
      "step": 3079
    },
    {
      "epoch": 280.0,
      "grad_norm": 0.43252795934677124,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.2944,
      "step": 3080
    },
    {
      "epoch": 280.0,
      "eval_loss": 0.4305260479450226,
      "eval_runtime": 2.1118,
      "eval_samples_per_second": 26.044,
      "eval_steps_per_second": 3.315,
      "step": 3080
    },
    {
      "epoch": 280.0969696969697,
      "grad_norm": 0.28369900584220886,
      "learning_rate": 1.919e-05,
      "loss": 0.3107,
      "step": 3081
    },
    {
      "epoch": 280.1939393939394,
      "grad_norm": 0.2864631116390228,
      "learning_rate": 1.918e-05,
      "loss": 0.3431,
      "step": 3082
    },
    {
      "epoch": 280.2909090909091,
      "grad_norm": 0.2793765962123871,
      "learning_rate": 1.917e-05,
      "loss": 0.3503,
      "step": 3083
    },
    {
      "epoch": 280.3878787878788,
      "grad_norm": 0.2945725619792938,
      "learning_rate": 1.916e-05,
      "loss": 0.355,
      "step": 3084
    },
    {
      "epoch": 280.4848484848485,
      "grad_norm": 0.29189518094062805,
      "learning_rate": 1.915e-05,
      "loss": 0.3404,
      "step": 3085
    },
    {
      "epoch": 280.58181818181816,
      "grad_norm": 0.3202930986881256,
      "learning_rate": 1.914e-05,
      "loss": 0.3533,
      "step": 3086
    },
    {
      "epoch": 280.6787878787879,
      "grad_norm": 0.3237863779067993,
      "learning_rate": 1.913e-05,
      "loss": 0.3396,
      "step": 3087
    },
    {
      "epoch": 280.77575757575755,
      "grad_norm": 0.42770758271217346,
      "learning_rate": 1.9120000000000003e-05,
      "loss": 0.3605,
      "step": 3088
    },
    {
      "epoch": 280.8727272727273,
      "grad_norm": 0.25378715991973877,
      "learning_rate": 1.911e-05,
      "loss": 0.3603,
      "step": 3089
    },
    {
      "epoch": 280.969696969697,
      "grad_norm": 0.28680214285850525,
      "learning_rate": 1.91e-05,
      "loss": 0.3551,
      "step": 3090
    },
    {
      "epoch": 280.969696969697,
      "eval_loss": 0.4307253956794739,
      "eval_runtime": 2.108,
      "eval_samples_per_second": 26.092,
      "eval_steps_per_second": 3.321,
      "step": 3090
    },
    {
      "epoch": 281.0,
      "grad_norm": 0.5535325407981873,
      "learning_rate": 1.909e-05,
      "loss": 0.3546,
      "step": 3091
    },
    {
      "epoch": 281.0969696969697,
      "grad_norm": 0.27855050563812256,
      "learning_rate": 1.908e-05,
      "loss": 0.3554,
      "step": 3092
    },
    {
      "epoch": 281.1939393939394,
      "grad_norm": 0.29888078570365906,
      "learning_rate": 1.9070000000000002e-05,
      "loss": 0.3379,
      "step": 3093
    },
    {
      "epoch": 281.2909090909091,
      "grad_norm": 0.28423425555229187,
      "learning_rate": 1.906e-05,
      "loss": 0.372,
      "step": 3094
    },
    {
      "epoch": 281.3878787878788,
      "grad_norm": 0.2922607958316803,
      "learning_rate": 1.9050000000000002e-05,
      "loss": 0.347,
      "step": 3095
    },
    {
      "epoch": 281.4848484848485,
      "grad_norm": 0.2930501699447632,
      "learning_rate": 1.904e-05,
      "loss": 0.3176,
      "step": 3096
    },
    {
      "epoch": 281.58181818181816,
      "grad_norm": 0.32645270228385925,
      "learning_rate": 1.903e-05,
      "loss": 0.3394,
      "step": 3097
    },
    {
      "epoch": 281.6787878787879,
      "grad_norm": 0.28562331199645996,
      "learning_rate": 1.902e-05,
      "loss": 0.3122,
      "step": 3098
    },
    {
      "epoch": 281.77575757575755,
      "grad_norm": 0.3255389630794525,
      "learning_rate": 1.901e-05,
      "loss": 0.3787,
      "step": 3099
    },
    {
      "epoch": 281.8727272727273,
      "grad_norm": 0.29790711402893066,
      "learning_rate": 1.9e-05,
      "loss": 0.3318,
      "step": 3100
    },
    {
      "epoch": 281.8727272727273,
      "eval_loss": 0.43041619658470154,
      "eval_runtime": 2.0976,
      "eval_samples_per_second": 26.22,
      "eval_steps_per_second": 3.337,
      "step": 3100
    },
    {
      "epoch": 281.969696969697,
      "grad_norm": 0.29323795437812805,
      "learning_rate": 1.8990000000000003e-05,
      "loss": 0.372,
      "step": 3101
    },
    {
      "epoch": 282.0,
      "grad_norm": 0.5759214162826538,
      "learning_rate": 1.898e-05,
      "loss": 0.3692,
      "step": 3102
    },
    {
      "epoch": 282.0969696969697,
      "grad_norm": 0.26442188024520874,
      "learning_rate": 1.8970000000000003e-05,
      "loss": 0.3629,
      "step": 3103
    },
    {
      "epoch": 282.1939393939394,
      "grad_norm": 0.34483394026756287,
      "learning_rate": 1.896e-05,
      "loss": 0.3764,
      "step": 3104
    },
    {
      "epoch": 282.2909090909091,
      "grad_norm": 0.3741251230239868,
      "learning_rate": 1.895e-05,
      "loss": 0.3287,
      "step": 3105
    },
    {
      "epoch": 282.3878787878788,
      "grad_norm": 0.28175756335258484,
      "learning_rate": 1.894e-05,
      "loss": 0.3671,
      "step": 3106
    },
    {
      "epoch": 282.4848484848485,
      "grad_norm": 0.3492465019226074,
      "learning_rate": 1.893e-05,
      "loss": 0.3595,
      "step": 3107
    },
    {
      "epoch": 282.58181818181816,
      "grad_norm": 0.2611752450466156,
      "learning_rate": 1.8920000000000002e-05,
      "loss": 0.301,
      "step": 3108
    },
    {
      "epoch": 282.6787878787879,
      "grad_norm": 0.3798886835575104,
      "learning_rate": 1.891e-05,
      "loss": 0.3393,
      "step": 3109
    },
    {
      "epoch": 282.77575757575755,
      "grad_norm": 0.34574228525161743,
      "learning_rate": 1.8900000000000002e-05,
      "loss": 0.3288,
      "step": 3110
    },
    {
      "epoch": 282.77575757575755,
      "eval_loss": 0.43000346422195435,
      "eval_runtime": 2.1018,
      "eval_samples_per_second": 26.168,
      "eval_steps_per_second": 3.33,
      "step": 3110
    },
    {
      "epoch": 282.8727272727273,
      "grad_norm": 0.301952600479126,
      "learning_rate": 1.8890000000000004e-05,
      "loss": 0.3288,
      "step": 3111
    },
    {
      "epoch": 282.969696969697,
      "grad_norm": 0.26290667057037354,
      "learning_rate": 1.888e-05,
      "loss": 0.3848,
      "step": 3112
    },
    {
      "epoch": 283.0,
      "grad_norm": 0.46030494570732117,
      "learning_rate": 1.887e-05,
      "loss": 0.3212,
      "step": 3113
    },
    {
      "epoch": 283.0969696969697,
      "grad_norm": 0.3023940324783325,
      "learning_rate": 1.886e-05,
      "loss": 0.3092,
      "step": 3114
    },
    {
      "epoch": 283.1939393939394,
      "grad_norm": 0.32762202620506287,
      "learning_rate": 1.885e-05,
      "loss": 0.3419,
      "step": 3115
    },
    {
      "epoch": 283.2909090909091,
      "grad_norm": 0.28179922699928284,
      "learning_rate": 1.8840000000000003e-05,
      "loss": 0.339,
      "step": 3116
    },
    {
      "epoch": 283.3878787878788,
      "grad_norm": 0.2876780331134796,
      "learning_rate": 1.883e-05,
      "loss": 0.3731,
      "step": 3117
    },
    {
      "epoch": 283.4848484848485,
      "grad_norm": 0.2742939889431,
      "learning_rate": 1.8820000000000003e-05,
      "loss": 0.2887,
      "step": 3118
    },
    {
      "epoch": 283.58181818181816,
      "grad_norm": 0.3161603808403015,
      "learning_rate": 1.881e-05,
      "loss": 0.3623,
      "step": 3119
    },
    {
      "epoch": 283.6787878787879,
      "grad_norm": 0.2728952467441559,
      "learning_rate": 1.88e-05,
      "loss": 0.3722,
      "step": 3120
    },
    {
      "epoch": 283.6787878787879,
      "eval_loss": 0.4307185113430023,
      "eval_runtime": 2.1149,
      "eval_samples_per_second": 26.006,
      "eval_steps_per_second": 3.31,
      "step": 3120
    },
    {
      "epoch": 283.77575757575755,
      "grad_norm": 0.23937064409255981,
      "learning_rate": 1.879e-05,
      "loss": 0.3412,
      "step": 3121
    },
    {
      "epoch": 283.8727272727273,
      "grad_norm": 0.2939877212047577,
      "learning_rate": 1.878e-05,
      "loss": 0.343,
      "step": 3122
    },
    {
      "epoch": 283.969696969697,
      "grad_norm": 0.3187495768070221,
      "learning_rate": 1.877e-05,
      "loss": 0.3873,
      "step": 3123
    },
    {
      "epoch": 284.0,
      "grad_norm": 0.5333037972450256,
      "learning_rate": 1.876e-05,
      "loss": 0.3612,
      "step": 3124
    },
    {
      "epoch": 284.0969696969697,
      "grad_norm": 0.26920658349990845,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.3463,
      "step": 3125
    },
    {
      "epoch": 284.1939393939394,
      "grad_norm": 0.2812713086605072,
      "learning_rate": 1.8740000000000004e-05,
      "loss": 0.3433,
      "step": 3126
    },
    {
      "epoch": 284.2909090909091,
      "grad_norm": 0.28586363792419434,
      "learning_rate": 1.8730000000000002e-05,
      "loss": 0.335,
      "step": 3127
    },
    {
      "epoch": 284.3878787878788,
      "grad_norm": 0.2654596269130707,
      "learning_rate": 1.872e-05,
      "loss": 0.36,
      "step": 3128
    },
    {
      "epoch": 284.4848484848485,
      "grad_norm": 0.30611246824264526,
      "learning_rate": 1.871e-05,
      "loss": 0.3661,
      "step": 3129
    },
    {
      "epoch": 284.58181818181816,
      "grad_norm": 0.2727811634540558,
      "learning_rate": 1.87e-05,
      "loss": 0.3221,
      "step": 3130
    },
    {
      "epoch": 284.58181818181816,
      "eval_loss": 0.4300423562526703,
      "eval_runtime": 2.1098,
      "eval_samples_per_second": 26.069,
      "eval_steps_per_second": 3.318,
      "step": 3130
    },
    {
      "epoch": 284.6787878787879,
      "grad_norm": 0.3090428113937378,
      "learning_rate": 1.8690000000000002e-05,
      "loss": 0.3605,
      "step": 3131
    },
    {
      "epoch": 284.77575757575755,
      "grad_norm": 0.276574969291687,
      "learning_rate": 1.868e-05,
      "loss": 0.3638,
      "step": 3132
    },
    {
      "epoch": 284.8727272727273,
      "grad_norm": 0.26750755310058594,
      "learning_rate": 1.8670000000000003e-05,
      "loss": 0.3213,
      "step": 3133
    },
    {
      "epoch": 284.969696969697,
      "grad_norm": 0.3467130959033966,
      "learning_rate": 1.866e-05,
      "loss": 0.353,
      "step": 3134
    },
    {
      "epoch": 285.0,
      "grad_norm": 0.37440377473831177,
      "learning_rate": 1.865e-05,
      "loss": 0.3205,
      "step": 3135
    },
    {
      "epoch": 285.0969696969697,
      "grad_norm": 0.2853250801563263,
      "learning_rate": 1.864e-05,
      "loss": 0.3198,
      "step": 3136
    },
    {
      "epoch": 285.1939393939394,
      "grad_norm": 0.2645166218280792,
      "learning_rate": 1.863e-05,
      "loss": 0.3163,
      "step": 3137
    },
    {
      "epoch": 285.2909090909091,
      "grad_norm": 0.2701696455478668,
      "learning_rate": 1.862e-05,
      "loss": 0.3391,
      "step": 3138
    },
    {
      "epoch": 285.3878787878788,
      "grad_norm": 0.312705397605896,
      "learning_rate": 1.861e-05,
      "loss": 0.353,
      "step": 3139
    },
    {
      "epoch": 285.4848484848485,
      "grad_norm": 0.2816062271595001,
      "learning_rate": 1.86e-05,
      "loss": 0.35,
      "step": 3140
    },
    {
      "epoch": 285.4848484848485,
      "eval_loss": 0.43025192618370056,
      "eval_runtime": 2.108,
      "eval_samples_per_second": 26.091,
      "eval_steps_per_second": 3.321,
      "step": 3140
    },
    {
      "epoch": 285.58181818181816,
      "grad_norm": 0.26278868317604065,
      "learning_rate": 1.8590000000000003e-05,
      "loss": 0.3185,
      "step": 3141
    },
    {
      "epoch": 285.6787878787879,
      "grad_norm": 0.3085387945175171,
      "learning_rate": 1.858e-05,
      "loss": 0.3627,
      "step": 3142
    },
    {
      "epoch": 285.77575757575755,
      "grad_norm": 0.28027740120887756,
      "learning_rate": 1.857e-05,
      "loss": 0.3828,
      "step": 3143
    },
    {
      "epoch": 285.8727272727273,
      "grad_norm": 0.2845478355884552,
      "learning_rate": 1.856e-05,
      "loss": 0.3368,
      "step": 3144
    },
    {
      "epoch": 285.969696969697,
      "grad_norm": 0.3337837755680084,
      "learning_rate": 1.855e-05,
      "loss": 0.3708,
      "step": 3145
    },
    {
      "epoch": 286.0,
      "grad_norm": 0.47991085052490234,
      "learning_rate": 1.8540000000000002e-05,
      "loss": 0.3774,
      "step": 3146
    },
    {
      "epoch": 286.0969696969697,
      "grad_norm": 0.2537936568260193,
      "learning_rate": 1.853e-05,
      "loss": 0.3285,
      "step": 3147
    },
    {
      "epoch": 286.1939393939394,
      "grad_norm": 0.3946389853954315,
      "learning_rate": 1.8520000000000002e-05,
      "loss": 0.3513,
      "step": 3148
    },
    {
      "epoch": 286.2909090909091,
      "grad_norm": 0.2878142297267914,
      "learning_rate": 1.851e-05,
      "loss": 0.3444,
      "step": 3149
    },
    {
      "epoch": 286.3878787878788,
      "grad_norm": 0.28056782484054565,
      "learning_rate": 1.85e-05,
      "loss": 0.3674,
      "step": 3150
    },
    {
      "epoch": 286.3878787878788,
      "eval_loss": 0.42976757884025574,
      "eval_runtime": 2.0736,
      "eval_samples_per_second": 26.524,
      "eval_steps_per_second": 3.376,
      "step": 3150
    },
    {
      "epoch": 286.4848484848485,
      "grad_norm": 0.30548804998397827,
      "learning_rate": 1.849e-05,
      "loss": 0.342,
      "step": 3151
    },
    {
      "epoch": 286.58181818181816,
      "grad_norm": 0.30740660429000854,
      "learning_rate": 1.848e-05,
      "loss": 0.3405,
      "step": 3152
    },
    {
      "epoch": 286.6787878787879,
      "grad_norm": 0.29892516136169434,
      "learning_rate": 1.847e-05,
      "loss": 0.3436,
      "step": 3153
    },
    {
      "epoch": 286.77575757575755,
      "grad_norm": 0.3422331213951111,
      "learning_rate": 1.846e-05,
      "loss": 0.3366,
      "step": 3154
    },
    {
      "epoch": 286.8727272727273,
      "grad_norm": 0.29767316579818726,
      "learning_rate": 1.845e-05,
      "loss": 0.3488,
      "step": 3155
    },
    {
      "epoch": 286.969696969697,
      "grad_norm": 0.26103007793426514,
      "learning_rate": 1.8440000000000003e-05,
      "loss": 0.3691,
      "step": 3156
    },
    {
      "epoch": 287.0,
      "grad_norm": 0.5521721839904785,
      "learning_rate": 1.843e-05,
      "loss": 0.3043,
      "step": 3157
    },
    {
      "epoch": 287.0969696969697,
      "grad_norm": 0.2533706724643707,
      "learning_rate": 1.842e-05,
      "loss": 0.3805,
      "step": 3158
    },
    {
      "epoch": 287.1939393939394,
      "grad_norm": 0.25408628582954407,
      "learning_rate": 1.841e-05,
      "loss": 0.2988,
      "step": 3159
    },
    {
      "epoch": 287.2909090909091,
      "grad_norm": 0.2781979739665985,
      "learning_rate": 1.84e-05,
      "loss": 0.3651,
      "step": 3160
    },
    {
      "epoch": 287.2909090909091,
      "eval_loss": 0.42997777462005615,
      "eval_runtime": 2.1056,
      "eval_samples_per_second": 26.121,
      "eval_steps_per_second": 3.324,
      "step": 3160
    },
    {
      "epoch": 287.3878787878788,
      "grad_norm": 0.2921355664730072,
      "learning_rate": 1.8390000000000002e-05,
      "loss": 0.3701,
      "step": 3161
    },
    {
      "epoch": 287.4848484848485,
      "grad_norm": 0.3304900825023651,
      "learning_rate": 1.838e-05,
      "loss": 0.339,
      "step": 3162
    },
    {
      "epoch": 287.58181818181816,
      "grad_norm": 0.3184772729873657,
      "learning_rate": 1.8370000000000002e-05,
      "loss": 0.3453,
      "step": 3163
    },
    {
      "epoch": 287.6787878787879,
      "grad_norm": 0.2869490683078766,
      "learning_rate": 1.8360000000000004e-05,
      "loss": 0.3134,
      "step": 3164
    },
    {
      "epoch": 287.77575757575755,
      "grad_norm": 0.2816450595855713,
      "learning_rate": 1.8350000000000002e-05,
      "loss": 0.3279,
      "step": 3165
    },
    {
      "epoch": 287.8727272727273,
      "grad_norm": 0.3218279182910919,
      "learning_rate": 1.834e-05,
      "loss": 0.3545,
      "step": 3166
    },
    {
      "epoch": 287.969696969697,
      "grad_norm": 0.29858097434043884,
      "learning_rate": 1.833e-05,
      "loss": 0.3568,
      "step": 3167
    },
    {
      "epoch": 288.0,
      "grad_norm": 0.5968769192695618,
      "learning_rate": 1.832e-05,
      "loss": 0.3491,
      "step": 3168
    },
    {
      "epoch": 288.0969696969697,
      "grad_norm": 0.32061246037483215,
      "learning_rate": 1.8310000000000003e-05,
      "loss": 0.3503,
      "step": 3169
    },
    {
      "epoch": 288.1939393939394,
      "grad_norm": 0.28813618421554565,
      "learning_rate": 1.83e-05,
      "loss": 0.3622,
      "step": 3170
    },
    {
      "epoch": 288.1939393939394,
      "eval_loss": 0.4300357401371002,
      "eval_runtime": 2.105,
      "eval_samples_per_second": 26.129,
      "eval_steps_per_second": 3.325,
      "step": 3170
    },
    {
      "epoch": 288.2909090909091,
      "grad_norm": 0.3164956569671631,
      "learning_rate": 1.8290000000000003e-05,
      "loss": 0.3108,
      "step": 3171
    },
    {
      "epoch": 288.3878787878788,
      "grad_norm": 0.25404638051986694,
      "learning_rate": 1.828e-05,
      "loss": 0.3372,
      "step": 3172
    },
    {
      "epoch": 288.4848484848485,
      "grad_norm": 0.25583022832870483,
      "learning_rate": 1.827e-05,
      "loss": 0.3289,
      "step": 3173
    },
    {
      "epoch": 288.58181818181816,
      "grad_norm": 0.48261046409606934,
      "learning_rate": 1.826e-05,
      "loss": 0.3589,
      "step": 3174
    },
    {
      "epoch": 288.6787878787879,
      "grad_norm": 0.33504432439804077,
      "learning_rate": 1.825e-05,
      "loss": 0.3152,
      "step": 3175
    },
    {
      "epoch": 288.77575757575755,
      "grad_norm": 0.29916203022003174,
      "learning_rate": 1.824e-05,
      "loss": 0.3649,
      "step": 3176
    },
    {
      "epoch": 288.8727272727273,
      "grad_norm": 0.4544450342655182,
      "learning_rate": 1.823e-05,
      "loss": 0.3586,
      "step": 3177
    },
    {
      "epoch": 288.969696969697,
      "grad_norm": 0.3073657155036926,
      "learning_rate": 1.8220000000000002e-05,
      "loss": 0.3652,
      "step": 3178
    },
    {
      "epoch": 289.0,
      "grad_norm": 0.4764782190322876,
      "learning_rate": 1.8210000000000004e-05,
      "loss": 0.3451,
      "step": 3179
    },
    {
      "epoch": 289.0969696969697,
      "grad_norm": 0.2884770929813385,
      "learning_rate": 1.8200000000000002e-05,
      "loss": 0.3436,
      "step": 3180
    },
    {
      "epoch": 289.0969696969697,
      "eval_loss": 0.429638534784317,
      "eval_runtime": 2.102,
      "eval_samples_per_second": 26.165,
      "eval_steps_per_second": 3.33,
      "step": 3180
    },
    {
      "epoch": 289.1939393939394,
      "grad_norm": 0.2942644953727722,
      "learning_rate": 1.819e-05,
      "loss": 0.3399,
      "step": 3181
    },
    {
      "epoch": 289.2909090909091,
      "grad_norm": 0.30701959133148193,
      "learning_rate": 1.818e-05,
      "loss": 0.3338,
      "step": 3182
    },
    {
      "epoch": 289.3878787878788,
      "grad_norm": 0.3285771608352661,
      "learning_rate": 1.817e-05,
      "loss": 0.3458,
      "step": 3183
    },
    {
      "epoch": 289.4848484848485,
      "grad_norm": 0.27240973711013794,
      "learning_rate": 1.8160000000000002e-05,
      "loss": 0.3225,
      "step": 3184
    },
    {
      "epoch": 289.58181818181816,
      "grad_norm": 0.3030914068222046,
      "learning_rate": 1.815e-05,
      "loss": 0.3538,
      "step": 3185
    },
    {
      "epoch": 289.6787878787879,
      "grad_norm": 0.2722974717617035,
      "learning_rate": 1.8140000000000003e-05,
      "loss": 0.3198,
      "step": 3186
    },
    {
      "epoch": 289.77575757575755,
      "grad_norm": 0.3061240017414093,
      "learning_rate": 1.813e-05,
      "loss": 0.366,
      "step": 3187
    },
    {
      "epoch": 289.8727272727273,
      "grad_norm": 0.30942073464393616,
      "learning_rate": 1.812e-05,
      "loss": 0.3788,
      "step": 3188
    },
    {
      "epoch": 289.969696969697,
      "grad_norm": 0.3707157373428345,
      "learning_rate": 1.811e-05,
      "loss": 0.3517,
      "step": 3189
    },
    {
      "epoch": 290.0,
      "grad_norm": 0.4878753423690796,
      "learning_rate": 1.81e-05,
      "loss": 0.3301,
      "step": 3190
    },
    {
      "epoch": 290.0,
      "eval_loss": 0.4299190938472748,
      "eval_runtime": 2.1284,
      "eval_samples_per_second": 25.841,
      "eval_steps_per_second": 3.289,
      "step": 3190
    },
    {
      "epoch": 290.0969696969697,
      "grad_norm": 0.29387012124061584,
      "learning_rate": 1.809e-05,
      "loss": 0.3544,
      "step": 3191
    },
    {
      "epoch": 290.1939393939394,
      "grad_norm": 0.29010263085365295,
      "learning_rate": 1.808e-05,
      "loss": 0.383,
      "step": 3192
    },
    {
      "epoch": 290.2909090909091,
      "grad_norm": 0.27737659215927124,
      "learning_rate": 1.807e-05,
      "loss": 0.3158,
      "step": 3193
    },
    {
      "epoch": 290.3878787878788,
      "grad_norm": 0.27322596311569214,
      "learning_rate": 1.8060000000000003e-05,
      "loss": 0.3413,
      "step": 3194
    },
    {
      "epoch": 290.4848484848485,
      "grad_norm": 0.2959296703338623,
      "learning_rate": 1.805e-05,
      "loss": 0.3251,
      "step": 3195
    },
    {
      "epoch": 290.58181818181816,
      "grad_norm": 0.33172884583473206,
      "learning_rate": 1.804e-05,
      "loss": 0.3314,
      "step": 3196
    },
    {
      "epoch": 290.6787878787879,
      "grad_norm": 0.29233667254447937,
      "learning_rate": 1.803e-05,
      "loss": 0.3403,
      "step": 3197
    },
    {
      "epoch": 290.77575757575755,
      "grad_norm": 0.2946917414665222,
      "learning_rate": 1.802e-05,
      "loss": 0.3321,
      "step": 3198
    },
    {
      "epoch": 290.8727272727273,
      "grad_norm": 0.31235837936401367,
      "learning_rate": 1.8010000000000002e-05,
      "loss": 0.3635,
      "step": 3199
    },
    {
      "epoch": 290.969696969697,
      "grad_norm": 0.29859185218811035,
      "learning_rate": 1.8e-05,
      "loss": 0.3764,
      "step": 3200
    },
    {
      "epoch": 290.969696969697,
      "eval_loss": 0.42965859174728394,
      "eval_runtime": 2.1181,
      "eval_samples_per_second": 25.967,
      "eval_steps_per_second": 3.305,
      "step": 3200
    },
    {
      "epoch": 291.0,
      "grad_norm": 0.7010234594345093,
      "learning_rate": 1.7990000000000002e-05,
      "loss": 0.3011,
      "step": 3201
    },
    {
      "epoch": 291.0969696969697,
      "grad_norm": 0.27527788281440735,
      "learning_rate": 1.798e-05,
      "loss": 0.3411,
      "step": 3202
    },
    {
      "epoch": 291.1939393939394,
      "grad_norm": 0.2753966748714447,
      "learning_rate": 1.797e-05,
      "loss": 0.313,
      "step": 3203
    },
    {
      "epoch": 291.2909090909091,
      "grad_norm": 0.2650189995765686,
      "learning_rate": 1.796e-05,
      "loss": 0.322,
      "step": 3204
    },
    {
      "epoch": 291.3878787878788,
      "grad_norm": 0.2802697718143463,
      "learning_rate": 1.795e-05,
      "loss": 0.3403,
      "step": 3205
    },
    {
      "epoch": 291.4848484848485,
      "grad_norm": 0.33286914229393005,
      "learning_rate": 1.794e-05,
      "loss": 0.3379,
      "step": 3206
    },
    {
      "epoch": 291.58181818181816,
      "grad_norm": 0.30533063411712646,
      "learning_rate": 1.793e-05,
      "loss": 0.3336,
      "step": 3207
    },
    {
      "epoch": 291.6787878787879,
      "grad_norm": 0.3137774169445038,
      "learning_rate": 1.792e-05,
      "loss": 0.3487,
      "step": 3208
    },
    {
      "epoch": 291.77575757575755,
      "grad_norm": 0.37040799856185913,
      "learning_rate": 1.7910000000000003e-05,
      "loss": 0.3926,
      "step": 3209
    },
    {
      "epoch": 291.8727272727273,
      "grad_norm": 0.2896455228328705,
      "learning_rate": 1.79e-05,
      "loss": 0.3406,
      "step": 3210
    },
    {
      "epoch": 291.8727272727273,
      "eval_loss": 0.4292603135108948,
      "eval_runtime": 2.0907,
      "eval_samples_per_second": 26.307,
      "eval_steps_per_second": 3.348,
      "step": 3210
    },
    {
      "epoch": 291.969696969697,
      "grad_norm": 0.2867777347564697,
      "learning_rate": 1.789e-05,
      "loss": 0.3498,
      "step": 3211
    },
    {
      "epoch": 292.0,
      "grad_norm": 0.4879944324493408,
      "learning_rate": 1.7879999999999998e-05,
      "loss": 0.4218,
      "step": 3212
    },
    {
      "epoch": 292.0969696969697,
      "grad_norm": 0.29318252205848694,
      "learning_rate": 1.787e-05,
      "loss": 0.362,
      "step": 3213
    },
    {
      "epoch": 292.1939393939394,
      "grad_norm": 0.282903790473938,
      "learning_rate": 1.7860000000000002e-05,
      "loss": 0.332,
      "step": 3214
    },
    {
      "epoch": 292.2909090909091,
      "grad_norm": 0.31095901131629944,
      "learning_rate": 1.785e-05,
      "loss": 0.3615,
      "step": 3215
    },
    {
      "epoch": 292.3878787878788,
      "grad_norm": 0.30026254057884216,
      "learning_rate": 1.7840000000000002e-05,
      "loss": 0.3512,
      "step": 3216
    },
    {
      "epoch": 292.4848484848485,
      "grad_norm": 0.2917325496673584,
      "learning_rate": 1.783e-05,
      "loss": 0.3206,
      "step": 3217
    },
    {
      "epoch": 292.58181818181816,
      "grad_norm": 0.29843437671661377,
      "learning_rate": 1.7820000000000002e-05,
      "loss": 0.3621,
      "step": 3218
    },
    {
      "epoch": 292.6787878787879,
      "grad_norm": 0.2887900173664093,
      "learning_rate": 1.781e-05,
      "loss": 0.3549,
      "step": 3219
    },
    {
      "epoch": 292.77575757575755,
      "grad_norm": 0.3294726610183716,
      "learning_rate": 1.78e-05,
      "loss": 0.3464,
      "step": 3220
    },
    {
      "epoch": 292.77575757575755,
      "eval_loss": 0.42934826016426086,
      "eval_runtime": 2.1632,
      "eval_samples_per_second": 25.425,
      "eval_steps_per_second": 3.236,
      "step": 3220
    },
    {
      "epoch": 292.8727272727273,
      "grad_norm": 0.3146337866783142,
      "learning_rate": 1.779e-05,
      "loss": 0.3267,
      "step": 3221
    },
    {
      "epoch": 292.969696969697,
      "grad_norm": 0.2778744399547577,
      "learning_rate": 1.7780000000000003e-05,
      "loss": 0.3511,
      "step": 3222
    },
    {
      "epoch": 293.0,
      "grad_norm": 0.4883699417114258,
      "learning_rate": 1.777e-05,
      "loss": 0.2654,
      "step": 3223
    },
    {
      "epoch": 293.0969696969697,
      "grad_norm": 0.30644264817237854,
      "learning_rate": 1.7760000000000003e-05,
      "loss": 0.3088,
      "step": 3224
    },
    {
      "epoch": 293.1939393939394,
      "grad_norm": 0.29900655150413513,
      "learning_rate": 1.775e-05,
      "loss": 0.3547,
      "step": 3225
    },
    {
      "epoch": 293.2909090909091,
      "grad_norm": 0.3363381624221802,
      "learning_rate": 1.774e-05,
      "loss": 0.3218,
      "step": 3226
    },
    {
      "epoch": 293.3878787878788,
      "grad_norm": 0.2873002588748932,
      "learning_rate": 1.773e-05,
      "loss": 0.3535,
      "step": 3227
    },
    {
      "epoch": 293.4848484848485,
      "grad_norm": 0.31082361936569214,
      "learning_rate": 1.772e-05,
      "loss": 0.3491,
      "step": 3228
    },
    {
      "epoch": 293.58181818181816,
      "grad_norm": 0.3087656795978546,
      "learning_rate": 1.771e-05,
      "loss": 0.3448,
      "step": 3229
    },
    {
      "epoch": 293.6787878787879,
      "grad_norm": 0.2696332037448883,
      "learning_rate": 1.77e-05,
      "loss": 0.3501,
      "step": 3230
    },
    {
      "epoch": 293.6787878787879,
      "eval_loss": 0.42914244532585144,
      "eval_runtime": 2.1003,
      "eval_samples_per_second": 26.186,
      "eval_steps_per_second": 3.333,
      "step": 3230
    },
    {
      "epoch": 293.77575757575755,
      "grad_norm": 0.3170471787452698,
      "learning_rate": 1.7690000000000002e-05,
      "loss": 0.3262,
      "step": 3231
    },
    {
      "epoch": 293.8727272727273,
      "grad_norm": 0.2885948121547699,
      "learning_rate": 1.7680000000000004e-05,
      "loss": 0.4026,
      "step": 3232
    },
    {
      "epoch": 293.969696969697,
      "grad_norm": 0.4239996075630188,
      "learning_rate": 1.7670000000000002e-05,
      "loss": 0.3271,
      "step": 3233
    },
    {
      "epoch": 294.0,
      "grad_norm": 0.5607514977455139,
      "learning_rate": 1.766e-05,
      "loss": 0.3531,
      "step": 3234
    },
    {
      "epoch": 294.0969696969697,
      "grad_norm": 0.24723725020885468,
      "learning_rate": 1.765e-05,
      "loss": 0.3625,
      "step": 3235
    },
    {
      "epoch": 294.1939393939394,
      "grad_norm": 0.26862984895706177,
      "learning_rate": 1.764e-05,
      "loss": 0.3426,
      "step": 3236
    },
    {
      "epoch": 294.2909090909091,
      "grad_norm": 0.28407901525497437,
      "learning_rate": 1.7630000000000002e-05,
      "loss": 0.3475,
      "step": 3237
    },
    {
      "epoch": 294.3878787878788,
      "grad_norm": 0.29130661487579346,
      "learning_rate": 1.762e-05,
      "loss": 0.3089,
      "step": 3238
    },
    {
      "epoch": 294.4848484848485,
      "grad_norm": 0.3147640824317932,
      "learning_rate": 1.7610000000000002e-05,
      "loss": 0.3193,
      "step": 3239
    },
    {
      "epoch": 294.58181818181816,
      "grad_norm": 0.3013257384300232,
      "learning_rate": 1.76e-05,
      "loss": 0.3466,
      "step": 3240
    },
    {
      "epoch": 294.58181818181816,
      "eval_loss": 0.4294959604740143,
      "eval_runtime": 2.1214,
      "eval_samples_per_second": 25.927,
      "eval_steps_per_second": 3.3,
      "step": 3240
    },
    {
      "epoch": 294.6787878787879,
      "grad_norm": 0.36485326290130615,
      "learning_rate": 1.759e-05,
      "loss": 0.3797,
      "step": 3241
    },
    {
      "epoch": 294.77575757575755,
      "grad_norm": 0.37424951791763306,
      "learning_rate": 1.758e-05,
      "loss": 0.3067,
      "step": 3242
    },
    {
      "epoch": 294.8727272727273,
      "grad_norm": 0.29747769236564636,
      "learning_rate": 1.757e-05,
      "loss": 0.3655,
      "step": 3243
    },
    {
      "epoch": 294.969696969697,
      "grad_norm": 0.3009466230869293,
      "learning_rate": 1.756e-05,
      "loss": 0.3634,
      "step": 3244
    },
    {
      "epoch": 295.0,
      "grad_norm": 0.6070283651351929,
      "learning_rate": 1.755e-05,
      "loss": 0.3316,
      "step": 3245
    },
    {
      "epoch": 295.0969696969697,
      "grad_norm": 0.27728110551834106,
      "learning_rate": 1.754e-05,
      "loss": 0.3468,
      "step": 3246
    },
    {
      "epoch": 295.1939393939394,
      "grad_norm": 0.3359808027744293,
      "learning_rate": 1.7530000000000003e-05,
      "loss": 0.363,
      "step": 3247
    },
    {
      "epoch": 295.2909090909091,
      "grad_norm": 0.25140830874443054,
      "learning_rate": 1.752e-05,
      "loss": 0.3309,
      "step": 3248
    },
    {
      "epoch": 295.3878787878788,
      "grad_norm": 0.33611467480659485,
      "learning_rate": 1.751e-05,
      "loss": 0.3477,
      "step": 3249
    },
    {
      "epoch": 295.4848484848485,
      "grad_norm": 0.2783098816871643,
      "learning_rate": 1.75e-05,
      "loss": 0.3626,
      "step": 3250
    },
    {
      "epoch": 295.4848484848485,
      "eval_loss": 0.4290139079093933,
      "eval_runtime": 2.1355,
      "eval_samples_per_second": 25.755,
      "eval_steps_per_second": 3.278,
      "step": 3250
    },
    {
      "epoch": 295.58181818181816,
      "grad_norm": 0.295444130897522,
      "learning_rate": 1.749e-05,
      "loss": 0.3462,
      "step": 3251
    },
    {
      "epoch": 295.6787878787879,
      "grad_norm": 0.2505262494087219,
      "learning_rate": 1.7480000000000002e-05,
      "loss": 0.3365,
      "step": 3252
    },
    {
      "epoch": 295.77575757575755,
      "grad_norm": 0.2934752106666565,
      "learning_rate": 1.747e-05,
      "loss": 0.3433,
      "step": 3253
    },
    {
      "epoch": 295.8727272727273,
      "grad_norm": 0.31944575905799866,
      "learning_rate": 1.7460000000000002e-05,
      "loss": 0.3482,
      "step": 3254
    },
    {
      "epoch": 295.969696969697,
      "grad_norm": 0.2714713215827942,
      "learning_rate": 1.745e-05,
      "loss": 0.3014,
      "step": 3255
    },
    {
      "epoch": 296.0,
      "grad_norm": 0.543958842754364,
      "learning_rate": 1.7440000000000002e-05,
      "loss": 0.3689,
      "step": 3256
    },
    {
      "epoch": 296.0969696969697,
      "grad_norm": 0.30188027024269104,
      "learning_rate": 1.743e-05,
      "loss": 0.3403,
      "step": 3257
    },
    {
      "epoch": 296.1939393939394,
      "grad_norm": 0.3010038733482361,
      "learning_rate": 1.742e-05,
      "loss": 0.3497,
      "step": 3258
    },
    {
      "epoch": 296.2909090909091,
      "grad_norm": 0.29702892899513245,
      "learning_rate": 1.741e-05,
      "loss": 0.3126,
      "step": 3259
    },
    {
      "epoch": 296.3878787878788,
      "grad_norm": 0.31002187728881836,
      "learning_rate": 1.74e-05,
      "loss": 0.3697,
      "step": 3260
    },
    {
      "epoch": 296.3878787878788,
      "eval_loss": 0.42891135811805725,
      "eval_runtime": 2.1018,
      "eval_samples_per_second": 26.168,
      "eval_steps_per_second": 3.33,
      "step": 3260
    },
    {
      "epoch": 296.4848484848485,
      "grad_norm": 0.2677341401576996,
      "learning_rate": 1.739e-05,
      "loss": 0.3266,
      "step": 3261
    },
    {
      "epoch": 296.58181818181816,
      "grad_norm": 0.3090519607067108,
      "learning_rate": 1.7380000000000003e-05,
      "loss": 0.3657,
      "step": 3262
    },
    {
      "epoch": 296.6787878787879,
      "grad_norm": 0.2831075191497803,
      "learning_rate": 1.737e-05,
      "loss": 0.3482,
      "step": 3263
    },
    {
      "epoch": 296.77575757575755,
      "grad_norm": 0.3414286673069,
      "learning_rate": 1.736e-05,
      "loss": 0.3999,
      "step": 3264
    },
    {
      "epoch": 296.8727272727273,
      "grad_norm": 0.3059631884098053,
      "learning_rate": 1.7349999999999998e-05,
      "loss": 0.327,
      "step": 3265
    },
    {
      "epoch": 296.969696969697,
      "grad_norm": 0.2777879536151886,
      "learning_rate": 1.734e-05,
      "loss": 0.325,
      "step": 3266
    },
    {
      "epoch": 297.0,
      "grad_norm": 0.5726981163024902,
      "learning_rate": 1.7330000000000002e-05,
      "loss": 0.2346,
      "step": 3267
    },
    {
      "epoch": 297.0969696969697,
      "grad_norm": 0.251736581325531,
      "learning_rate": 1.732e-05,
      "loss": 0.3528,
      "step": 3268
    },
    {
      "epoch": 297.1939393939394,
      "grad_norm": 0.3111477792263031,
      "learning_rate": 1.7310000000000002e-05,
      "loss": 0.3494,
      "step": 3269
    },
    {
      "epoch": 297.2909090909091,
      "grad_norm": 0.26640409231185913,
      "learning_rate": 1.73e-05,
      "loss": 0.3449,
      "step": 3270
    },
    {
      "epoch": 297.2909090909091,
      "eval_loss": 0.4294259548187256,
      "eval_runtime": 2.139,
      "eval_samples_per_second": 25.713,
      "eval_steps_per_second": 3.273,
      "step": 3270
    },
    {
      "epoch": 297.3878787878788,
      "grad_norm": 0.28788143396377563,
      "learning_rate": 1.7290000000000002e-05,
      "loss": 0.3584,
      "step": 3271
    },
    {
      "epoch": 297.4848484848485,
      "grad_norm": 0.3301577866077423,
      "learning_rate": 1.728e-05,
      "loss": 0.3715,
      "step": 3272
    },
    {
      "epoch": 297.58181818181816,
      "grad_norm": 0.3239557147026062,
      "learning_rate": 1.727e-05,
      "loss": 0.326,
      "step": 3273
    },
    {
      "epoch": 297.6787878787879,
      "grad_norm": 0.260246217250824,
      "learning_rate": 1.726e-05,
      "loss": 0.3456,
      "step": 3274
    },
    {
      "epoch": 297.77575757575755,
      "grad_norm": 0.3090617060661316,
      "learning_rate": 1.725e-05,
      "loss": 0.3487,
      "step": 3275
    },
    {
      "epoch": 297.8727272727273,
      "grad_norm": 0.2856177091598511,
      "learning_rate": 1.724e-05,
      "loss": 0.2956,
      "step": 3276
    },
    {
      "epoch": 297.969696969697,
      "grad_norm": 0.31400689482688904,
      "learning_rate": 1.7230000000000003e-05,
      "loss": 0.341,
      "step": 3277
    },
    {
      "epoch": 298.0,
      "grad_norm": 0.5386035442352295,
      "learning_rate": 1.722e-05,
      "loss": 0.3256,
      "step": 3278
    },
    {
      "epoch": 298.0969696969697,
      "grad_norm": 0.31615230441093445,
      "learning_rate": 1.721e-05,
      "loss": 0.3486,
      "step": 3279
    },
    {
      "epoch": 298.1939393939394,
      "grad_norm": 0.3985329568386078,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 0.3398,
      "step": 3280
    },
    {
      "epoch": 298.1939393939394,
      "eval_loss": 0.4285666346549988,
      "eval_runtime": 2.1236,
      "eval_samples_per_second": 25.899,
      "eval_steps_per_second": 3.296,
      "step": 3280
    },
    {
      "epoch": 298.2909090909091,
      "grad_norm": 0.24339041113853455,
      "learning_rate": 1.719e-05,
      "loss": 0.3374,
      "step": 3281
    },
    {
      "epoch": 298.3878787878788,
      "grad_norm": 0.27008259296417236,
      "learning_rate": 1.718e-05,
      "loss": 0.3551,
      "step": 3282
    },
    {
      "epoch": 298.4848484848485,
      "grad_norm": 0.2825574576854706,
      "learning_rate": 1.717e-05,
      "loss": 0.3557,
      "step": 3283
    },
    {
      "epoch": 298.58181818181816,
      "grad_norm": 0.3063613772392273,
      "learning_rate": 1.7160000000000002e-05,
      "loss": 0.3248,
      "step": 3284
    },
    {
      "epoch": 298.6787878787879,
      "grad_norm": 0.3211213946342468,
      "learning_rate": 1.7150000000000004e-05,
      "loss": 0.3411,
      "step": 3285
    },
    {
      "epoch": 298.77575757575755,
      "grad_norm": 0.2771667540073395,
      "learning_rate": 1.7140000000000002e-05,
      "loss": 0.335,
      "step": 3286
    },
    {
      "epoch": 298.8727272727273,
      "grad_norm": 0.3101070821285248,
      "learning_rate": 1.713e-05,
      "loss": 0.3573,
      "step": 3287
    },
    {
      "epoch": 298.969696969697,
      "grad_norm": 0.2912956178188324,
      "learning_rate": 1.712e-05,
      "loss": 0.3304,
      "step": 3288
    },
    {
      "epoch": 299.0,
      "grad_norm": 0.5872663855552673,
      "learning_rate": 1.711e-05,
      "loss": 0.3486,
      "step": 3289
    },
    {
      "epoch": 299.0969696969697,
      "grad_norm": 0.30190402269363403,
      "learning_rate": 1.7100000000000002e-05,
      "loss": 0.3315,
      "step": 3290
    },
    {
      "epoch": 299.0969696969697,
      "eval_loss": 0.42890438437461853,
      "eval_runtime": 2.1137,
      "eval_samples_per_second": 26.02,
      "eval_steps_per_second": 3.312,
      "step": 3290
    },
    {
      "epoch": 299.1939393939394,
      "grad_norm": 0.3438356816768646,
      "learning_rate": 1.709e-05,
      "loss": 0.3138,
      "step": 3291
    },
    {
      "epoch": 299.2909090909091,
      "grad_norm": 0.2754231095314026,
      "learning_rate": 1.7080000000000002e-05,
      "loss": 0.3508,
      "step": 3292
    },
    {
      "epoch": 299.3878787878788,
      "grad_norm": 0.27830561995506287,
      "learning_rate": 1.707e-05,
      "loss": 0.3243,
      "step": 3293
    },
    {
      "epoch": 299.4848484848485,
      "grad_norm": 0.3274482488632202,
      "learning_rate": 1.706e-05,
      "loss": 0.3398,
      "step": 3294
    },
    {
      "epoch": 299.58181818181816,
      "grad_norm": 0.34774988889694214,
      "learning_rate": 1.705e-05,
      "loss": 0.3373,
      "step": 3295
    },
    {
      "epoch": 299.6787878787879,
      "grad_norm": 0.3125561773777008,
      "learning_rate": 1.704e-05,
      "loss": 0.3228,
      "step": 3296
    },
    {
      "epoch": 299.77575757575755,
      "grad_norm": 0.2622632086277008,
      "learning_rate": 1.703e-05,
      "loss": 0.3725,
      "step": 3297
    },
    {
      "epoch": 299.8727272727273,
      "grad_norm": 0.2701694965362549,
      "learning_rate": 1.702e-05,
      "loss": 0.3695,
      "step": 3298
    },
    {
      "epoch": 299.969696969697,
      "grad_norm": 0.3075459897518158,
      "learning_rate": 1.701e-05,
      "loss": 0.354,
      "step": 3299
    },
    {
      "epoch": 300.0,
      "grad_norm": 0.5179641842842102,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 0.3704,
      "step": 3300
    },
    {
      "epoch": 300.0,
      "eval_loss": 0.4286232888698578,
      "eval_runtime": 2.1146,
      "eval_samples_per_second": 26.01,
      "eval_steps_per_second": 3.31,
      "step": 3300
    },
    {
      "epoch": 300.0969696969697,
      "grad_norm": 0.2588394284248352,
      "learning_rate": 1.699e-05,
      "loss": 0.3211,
      "step": 3301
    },
    {
      "epoch": 300.1939393939394,
      "grad_norm": 0.2719418704509735,
      "learning_rate": 1.698e-05,
      "loss": 0.3314,
      "step": 3302
    },
    {
      "epoch": 300.2909090909091,
      "grad_norm": 0.2840977907180786,
      "learning_rate": 1.697e-05,
      "loss": 0.3335,
      "step": 3303
    },
    {
      "epoch": 300.3878787878788,
      "grad_norm": 0.3237716257572174,
      "learning_rate": 1.696e-05,
      "loss": 0.3296,
      "step": 3304
    },
    {
      "epoch": 300.4848484848485,
      "grad_norm": 0.3320935368537903,
      "learning_rate": 1.6950000000000002e-05,
      "loss": 0.3671,
      "step": 3305
    },
    {
      "epoch": 300.58181818181816,
      "grad_norm": 0.5175302028656006,
      "learning_rate": 1.694e-05,
      "loss": 0.3522,
      "step": 3306
    },
    {
      "epoch": 300.6787878787879,
      "grad_norm": 0.3399061858654022,
      "learning_rate": 1.6930000000000002e-05,
      "loss": 0.3604,
      "step": 3307
    },
    {
      "epoch": 300.77575757575755,
      "grad_norm": 0.2818537950515747,
      "learning_rate": 1.692e-05,
      "loss": 0.3422,
      "step": 3308
    },
    {
      "epoch": 300.8727272727273,
      "grad_norm": 0.3524338901042938,
      "learning_rate": 1.6910000000000002e-05,
      "loss": 0.3142,
      "step": 3309
    },
    {
      "epoch": 300.969696969697,
      "grad_norm": 0.284162312746048,
      "learning_rate": 1.69e-05,
      "loss": 0.3543,
      "step": 3310
    },
    {
      "epoch": 300.969696969697,
      "eval_loss": 0.42852863669395447,
      "eval_runtime": 2.1374,
      "eval_samples_per_second": 25.732,
      "eval_steps_per_second": 3.275,
      "step": 3310
    },
    {
      "epoch": 301.0,
      "grad_norm": 0.6364848613739014,
      "learning_rate": 1.689e-05,
      "loss": 0.4003,
      "step": 3311
    },
    {
      "epoch": 301.0969696969697,
      "grad_norm": 0.45382216572761536,
      "learning_rate": 1.688e-05,
      "loss": 0.3158,
      "step": 3312
    },
    {
      "epoch": 301.1939393939394,
      "grad_norm": 0.3232753872871399,
      "learning_rate": 1.687e-05,
      "loss": 0.3746,
      "step": 3313
    },
    {
      "epoch": 301.2909090909091,
      "grad_norm": 0.3339671194553375,
      "learning_rate": 1.686e-05,
      "loss": 0.3444,
      "step": 3314
    },
    {
      "epoch": 301.3878787878788,
      "grad_norm": 0.33689528703689575,
      "learning_rate": 1.6850000000000003e-05,
      "loss": 0.3493,
      "step": 3315
    },
    {
      "epoch": 301.4848484848485,
      "grad_norm": 0.32228925824165344,
      "learning_rate": 1.684e-05,
      "loss": 0.344,
      "step": 3316
    },
    {
      "epoch": 301.58181818181816,
      "grad_norm": 0.2638039290904999,
      "learning_rate": 1.683e-05,
      "loss": 0.3418,
      "step": 3317
    },
    {
      "epoch": 301.6787878787879,
      "grad_norm": 0.3011254072189331,
      "learning_rate": 1.6819999999999998e-05,
      "loss": 0.3614,
      "step": 3318
    },
    {
      "epoch": 301.77575757575755,
      "grad_norm": 0.28206703066825867,
      "learning_rate": 1.681e-05,
      "loss": 0.3241,
      "step": 3319
    },
    {
      "epoch": 301.8727272727273,
      "grad_norm": 0.2938118875026703,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.3496,
      "step": 3320
    },
    {
      "epoch": 301.8727272727273,
      "eval_loss": 0.4284074008464813,
      "eval_runtime": 2.1364,
      "eval_samples_per_second": 25.744,
      "eval_steps_per_second": 3.277,
      "step": 3320
    },
    {
      "epoch": 301.969696969697,
      "grad_norm": 0.28451794385910034,
      "learning_rate": 1.679e-05,
      "loss": 0.3157,
      "step": 3321
    },
    {
      "epoch": 302.0,
      "grad_norm": 0.41703781485557556,
      "learning_rate": 1.6780000000000002e-05,
      "loss": 0.3465,
      "step": 3322
    },
    {
      "epoch": 302.0969696969697,
      "grad_norm": 0.28743523359298706,
      "learning_rate": 1.677e-05,
      "loss": 0.3253,
      "step": 3323
    },
    {
      "epoch": 302.1939393939394,
      "grad_norm": 0.26439985632896423,
      "learning_rate": 1.6760000000000002e-05,
      "loss": 0.3537,
      "step": 3324
    },
    {
      "epoch": 302.2909090909091,
      "grad_norm": 0.28935468196868896,
      "learning_rate": 1.675e-05,
      "loss": 0.3165,
      "step": 3325
    },
    {
      "epoch": 302.3878787878788,
      "grad_norm": 0.2709727883338928,
      "learning_rate": 1.674e-05,
      "loss": 0.3501,
      "step": 3326
    },
    {
      "epoch": 302.4848484848485,
      "grad_norm": 0.3182224631309509,
      "learning_rate": 1.673e-05,
      "loss": 0.3083,
      "step": 3327
    },
    {
      "epoch": 302.58181818181816,
      "grad_norm": 0.3185989558696747,
      "learning_rate": 1.672e-05,
      "loss": 0.3492,
      "step": 3328
    },
    {
      "epoch": 302.6787878787879,
      "grad_norm": 0.2719970643520355,
      "learning_rate": 1.671e-05,
      "loss": 0.353,
      "step": 3329
    },
    {
      "epoch": 302.77575757575755,
      "grad_norm": 0.2939992845058441,
      "learning_rate": 1.6700000000000003e-05,
      "loss": 0.342,
      "step": 3330
    },
    {
      "epoch": 302.77575757575755,
      "eval_loss": 0.4282497465610504,
      "eval_runtime": 2.1569,
      "eval_samples_per_second": 25.5,
      "eval_steps_per_second": 3.245,
      "step": 3330
    },
    {
      "epoch": 302.8727272727273,
      "grad_norm": 0.2789037823677063,
      "learning_rate": 1.669e-05,
      "loss": 0.3246,
      "step": 3331
    },
    {
      "epoch": 302.969696969697,
      "grad_norm": 0.32086724042892456,
      "learning_rate": 1.668e-05,
      "loss": 0.382,
      "step": 3332
    },
    {
      "epoch": 303.0,
      "grad_norm": 0.5400450229644775,
      "learning_rate": 1.6669999999999998e-05,
      "loss": 0.3965,
      "step": 3333
    },
    {
      "epoch": 303.0969696969697,
      "grad_norm": 0.28759241104125977,
      "learning_rate": 1.666e-05,
      "loss": 0.3688,
      "step": 3334
    },
    {
      "epoch": 303.1939393939394,
      "grad_norm": 0.26669740676879883,
      "learning_rate": 1.665e-05,
      "loss": 0.3256,
      "step": 3335
    },
    {
      "epoch": 303.2909090909091,
      "grad_norm": 0.29431137442588806,
      "learning_rate": 1.664e-05,
      "loss": 0.3388,
      "step": 3336
    },
    {
      "epoch": 303.3878787878788,
      "grad_norm": 0.27999913692474365,
      "learning_rate": 1.6630000000000002e-05,
      "loss": 0.352,
      "step": 3337
    },
    {
      "epoch": 303.4848484848485,
      "grad_norm": 0.28299450874328613,
      "learning_rate": 1.662e-05,
      "loss": 0.3548,
      "step": 3338
    },
    {
      "epoch": 303.58181818181816,
      "grad_norm": 0.3314816355705261,
      "learning_rate": 1.6610000000000002e-05,
      "loss": 0.3551,
      "step": 3339
    },
    {
      "epoch": 303.6787878787879,
      "grad_norm": 0.2577263414859772,
      "learning_rate": 1.66e-05,
      "loss": 0.3113,
      "step": 3340
    },
    {
      "epoch": 303.6787878787879,
      "eval_loss": 0.42867740988731384,
      "eval_runtime": 2.1184,
      "eval_samples_per_second": 25.964,
      "eval_steps_per_second": 3.304,
      "step": 3340
    },
    {
      "epoch": 303.77575757575755,
      "grad_norm": 0.27317485213279724,
      "learning_rate": 1.659e-05,
      "loss": 0.3556,
      "step": 3341
    },
    {
      "epoch": 303.8727272727273,
      "grad_norm": 0.3101962208747864,
      "learning_rate": 1.658e-05,
      "loss": 0.3076,
      "step": 3342
    },
    {
      "epoch": 303.969696969697,
      "grad_norm": 0.2733157277107239,
      "learning_rate": 1.657e-05,
      "loss": 0.3224,
      "step": 3343
    },
    {
      "epoch": 304.0,
      "grad_norm": 0.6002190113067627,
      "learning_rate": 1.656e-05,
      "loss": 0.4141,
      "step": 3344
    },
    {
      "epoch": 304.0969696969697,
      "grad_norm": 0.2699577808380127,
      "learning_rate": 1.6550000000000002e-05,
      "loss": 0.3271,
      "step": 3345
    },
    {
      "epoch": 304.1939393939394,
      "grad_norm": 0.26280707120895386,
      "learning_rate": 1.654e-05,
      "loss": 0.3375,
      "step": 3346
    },
    {
      "epoch": 304.2909090909091,
      "grad_norm": 0.28962603211402893,
      "learning_rate": 1.6530000000000003e-05,
      "loss": 0.3411,
      "step": 3347
    },
    {
      "epoch": 304.3878787878788,
      "grad_norm": 0.37396422028541565,
      "learning_rate": 1.652e-05,
      "loss": 0.3239,
      "step": 3348
    },
    {
      "epoch": 304.4848484848485,
      "grad_norm": 0.2670392096042633,
      "learning_rate": 1.651e-05,
      "loss": 0.3541,
      "step": 3349
    },
    {
      "epoch": 304.58181818181816,
      "grad_norm": 0.3426806628704071,
      "learning_rate": 1.65e-05,
      "loss": 0.3417,
      "step": 3350
    },
    {
      "epoch": 304.58181818181816,
      "eval_loss": 0.42846569418907166,
      "eval_runtime": 2.1309,
      "eval_samples_per_second": 25.811,
      "eval_steps_per_second": 3.285,
      "step": 3350
    },
    {
      "epoch": 304.6787878787879,
      "grad_norm": 0.3189926743507385,
      "learning_rate": 1.649e-05,
      "loss": 0.3505,
      "step": 3351
    },
    {
      "epoch": 304.77575757575755,
      "grad_norm": 0.29972100257873535,
      "learning_rate": 1.648e-05,
      "loss": 0.367,
      "step": 3352
    },
    {
      "epoch": 304.8727272727273,
      "grad_norm": 0.3046066164970398,
      "learning_rate": 1.6470000000000003e-05,
      "loss": 0.3114,
      "step": 3353
    },
    {
      "epoch": 304.969696969697,
      "grad_norm": 0.29538267850875854,
      "learning_rate": 1.646e-05,
      "loss": 0.3604,
      "step": 3354
    },
    {
      "epoch": 305.0,
      "grad_norm": 0.48789748549461365,
      "learning_rate": 1.645e-05,
      "loss": 0.3483,
      "step": 3355
    },
    {
      "epoch": 305.0969696969697,
      "grad_norm": 0.293074369430542,
      "learning_rate": 1.644e-05,
      "loss": 0.3444,
      "step": 3356
    },
    {
      "epoch": 305.1939393939394,
      "grad_norm": 0.3257480263710022,
      "learning_rate": 1.643e-05,
      "loss": 0.3487,
      "step": 3357
    },
    {
      "epoch": 305.2909090909091,
      "grad_norm": 0.29099413752555847,
      "learning_rate": 1.6420000000000002e-05,
      "loss": 0.3707,
      "step": 3358
    },
    {
      "epoch": 305.3878787878788,
      "grad_norm": 0.2569948434829712,
      "learning_rate": 1.641e-05,
      "loss": 0.3147,
      "step": 3359
    },
    {
      "epoch": 305.4848484848485,
      "grad_norm": 0.3167359232902527,
      "learning_rate": 1.6400000000000002e-05,
      "loss": 0.339,
      "step": 3360
    },
    {
      "epoch": 305.4848484848485,
      "eval_loss": 0.42803701758384705,
      "eval_runtime": 2.1427,
      "eval_samples_per_second": 25.668,
      "eval_steps_per_second": 3.267,
      "step": 3360
    },
    {
      "epoch": 305.58181818181816,
      "grad_norm": 0.2881450653076172,
      "learning_rate": 1.639e-05,
      "loss": 0.3366,
      "step": 3361
    },
    {
      "epoch": 305.6787878787879,
      "grad_norm": 0.22674117982387543,
      "learning_rate": 1.6380000000000002e-05,
      "loss": 0.3346,
      "step": 3362
    },
    {
      "epoch": 305.77575757575755,
      "grad_norm": 0.34561410546302795,
      "learning_rate": 1.637e-05,
      "loss": 0.3172,
      "step": 3363
    },
    {
      "epoch": 305.8727272727273,
      "grad_norm": 0.3139340281486511,
      "learning_rate": 1.636e-05,
      "loss": 0.3488,
      "step": 3364
    },
    {
      "epoch": 305.969696969697,
      "grad_norm": 0.28778886795043945,
      "learning_rate": 1.635e-05,
      "loss": 0.3472,
      "step": 3365
    },
    {
      "epoch": 306.0,
      "grad_norm": 0.599293053150177,
      "learning_rate": 1.634e-05,
      "loss": 0.3715,
      "step": 3366
    },
    {
      "epoch": 306.0969696969697,
      "grad_norm": 0.33993762731552124,
      "learning_rate": 1.633e-05,
      "loss": 0.3552,
      "step": 3367
    },
    {
      "epoch": 306.1939393939394,
      "grad_norm": 0.29261258244514465,
      "learning_rate": 1.6320000000000003e-05,
      "loss": 0.341,
      "step": 3368
    },
    {
      "epoch": 306.2909090909091,
      "grad_norm": 0.30039000511169434,
      "learning_rate": 1.631e-05,
      "loss": 0.334,
      "step": 3369
    },
    {
      "epoch": 306.3878787878788,
      "grad_norm": 0.3208199739456177,
      "learning_rate": 1.63e-05,
      "loss": 0.3273,
      "step": 3370
    },
    {
      "epoch": 306.3878787878788,
      "eval_loss": 0.4288559854030609,
      "eval_runtime": 2.1164,
      "eval_samples_per_second": 25.987,
      "eval_steps_per_second": 3.307,
      "step": 3370
    },
    {
      "epoch": 306.4848484848485,
      "grad_norm": 0.3807413578033447,
      "learning_rate": 1.6289999999999998e-05,
      "loss": 0.2988,
      "step": 3371
    },
    {
      "epoch": 306.58181818181816,
      "grad_norm": 0.3323957920074463,
      "learning_rate": 1.628e-05,
      "loss": 0.3371,
      "step": 3372
    },
    {
      "epoch": 306.6787878787879,
      "grad_norm": 0.27052778005599976,
      "learning_rate": 1.6270000000000002e-05,
      "loss": 0.3627,
      "step": 3373
    },
    {
      "epoch": 306.77575757575755,
      "grad_norm": 0.3271731734275818,
      "learning_rate": 1.626e-05,
      "loss": 0.3724,
      "step": 3374
    },
    {
      "epoch": 306.8727272727273,
      "grad_norm": 0.26784583926200867,
      "learning_rate": 1.6250000000000002e-05,
      "loss": 0.3475,
      "step": 3375
    },
    {
      "epoch": 306.969696969697,
      "grad_norm": 0.3374277949333191,
      "learning_rate": 1.624e-05,
      "loss": 0.3392,
      "step": 3376
    },
    {
      "epoch": 307.0,
      "grad_norm": 0.4716193675994873,
      "learning_rate": 1.6230000000000002e-05,
      "loss": 0.3315,
      "step": 3377
    },
    {
      "epoch": 307.0969696969697,
      "grad_norm": 0.2882108986377716,
      "learning_rate": 1.622e-05,
      "loss": 0.3224,
      "step": 3378
    },
    {
      "epoch": 307.1939393939394,
      "grad_norm": 0.32774612307548523,
      "learning_rate": 1.621e-05,
      "loss": 0.3578,
      "step": 3379
    },
    {
      "epoch": 307.2909090909091,
      "grad_norm": 0.40872931480407715,
      "learning_rate": 1.62e-05,
      "loss": 0.347,
      "step": 3380
    },
    {
      "epoch": 307.2909090909091,
      "eval_loss": 0.4279690086841583,
      "eval_runtime": 2.1221,
      "eval_samples_per_second": 25.918,
      "eval_steps_per_second": 3.299,
      "step": 3380
    },
    {
      "epoch": 307.3878787878788,
      "grad_norm": 0.317185640335083,
      "learning_rate": 1.619e-05,
      "loss": 0.3236,
      "step": 3381
    },
    {
      "epoch": 307.4848484848485,
      "grad_norm": 0.3067472577095032,
      "learning_rate": 1.618e-05,
      "loss": 0.3414,
      "step": 3382
    },
    {
      "epoch": 307.58181818181816,
      "grad_norm": 0.2706355154514313,
      "learning_rate": 1.6170000000000003e-05,
      "loss": 0.3488,
      "step": 3383
    },
    {
      "epoch": 307.6787878787879,
      "grad_norm": 0.26995623111724854,
      "learning_rate": 1.616e-05,
      "loss": 0.3501,
      "step": 3384
    },
    {
      "epoch": 307.77575757575755,
      "grad_norm": 0.3374771177768707,
      "learning_rate": 1.6150000000000003e-05,
      "loss": 0.331,
      "step": 3385
    },
    {
      "epoch": 307.8727272727273,
      "grad_norm": 0.29605382680892944,
      "learning_rate": 1.6139999999999998e-05,
      "loss": 0.3445,
      "step": 3386
    },
    {
      "epoch": 307.969696969697,
      "grad_norm": 0.3035816550254822,
      "learning_rate": 1.613e-05,
      "loss": 0.3479,
      "step": 3387
    },
    {
      "epoch": 308.0,
      "grad_norm": 0.4600336253643036,
      "learning_rate": 1.612e-05,
      "loss": 0.3264,
      "step": 3388
    },
    {
      "epoch": 308.0969696969697,
      "grad_norm": 0.28409048914909363,
      "learning_rate": 1.611e-05,
      "loss": 0.3152,
      "step": 3389
    },
    {
      "epoch": 308.1939393939394,
      "grad_norm": 0.26470401883125305,
      "learning_rate": 1.6100000000000002e-05,
      "loss": 0.3355,
      "step": 3390
    },
    {
      "epoch": 308.1939393939394,
      "eval_loss": 0.42832431197166443,
      "eval_runtime": 2.1153,
      "eval_samples_per_second": 26.001,
      "eval_steps_per_second": 3.309,
      "step": 3390
    },
    {
      "epoch": 308.2909090909091,
      "grad_norm": 0.33754292130470276,
      "learning_rate": 1.609e-05,
      "loss": 0.3398,
      "step": 3391
    },
    {
      "epoch": 308.3878787878788,
      "grad_norm": 0.32827553153038025,
      "learning_rate": 1.6080000000000002e-05,
      "loss": 0.396,
      "step": 3392
    },
    {
      "epoch": 308.4848484848485,
      "grad_norm": 0.3123801052570343,
      "learning_rate": 1.607e-05,
      "loss": 0.3454,
      "step": 3393
    },
    {
      "epoch": 308.58181818181816,
      "grad_norm": 0.31118044257164,
      "learning_rate": 1.606e-05,
      "loss": 0.3468,
      "step": 3394
    },
    {
      "epoch": 308.6787878787879,
      "grad_norm": 0.32048311829566956,
      "learning_rate": 1.605e-05,
      "loss": 0.3689,
      "step": 3395
    },
    {
      "epoch": 308.77575757575755,
      "grad_norm": 0.3279634416103363,
      "learning_rate": 1.604e-05,
      "loss": 0.3171,
      "step": 3396
    },
    {
      "epoch": 308.8727272727273,
      "grad_norm": 0.2815092206001282,
      "learning_rate": 1.603e-05,
      "loss": 0.3118,
      "step": 3397
    },
    {
      "epoch": 308.969696969697,
      "grad_norm": 0.30864110589027405,
      "learning_rate": 1.6020000000000002e-05,
      "loss": 0.3363,
      "step": 3398
    },
    {
      "epoch": 309.0,
      "grad_norm": 0.55268794298172,
      "learning_rate": 1.601e-05,
      "loss": 0.3085,
      "step": 3399
    },
    {
      "epoch": 309.0969696969697,
      "grad_norm": 0.31190842390060425,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.3552,
      "step": 3400
    },
    {
      "epoch": 309.0969696969697,
      "eval_loss": 0.4276943504810333,
      "eval_runtime": 2.1232,
      "eval_samples_per_second": 25.904,
      "eval_steps_per_second": 3.297,
      "step": 3400
    },
    {
      "epoch": 309.1939393939394,
      "grad_norm": 0.2999674081802368,
      "learning_rate": 1.599e-05,
      "loss": 0.3228,
      "step": 3401
    },
    {
      "epoch": 309.2909090909091,
      "grad_norm": 0.2836794853210449,
      "learning_rate": 1.598e-05,
      "loss": 0.3026,
      "step": 3402
    },
    {
      "epoch": 309.3878787878788,
      "grad_norm": 0.3843519687652588,
      "learning_rate": 1.597e-05,
      "loss": 0.3681,
      "step": 3403
    },
    {
      "epoch": 309.4848484848485,
      "grad_norm": 0.269823282957077,
      "learning_rate": 1.596e-05,
      "loss": 0.3443,
      "step": 3404
    },
    {
      "epoch": 309.58181818181816,
      "grad_norm": 0.291311115026474,
      "learning_rate": 1.595e-05,
      "loss": 0.3341,
      "step": 3405
    },
    {
      "epoch": 309.6787878787879,
      "grad_norm": 0.29707270860671997,
      "learning_rate": 1.594e-05,
      "loss": 0.3522,
      "step": 3406
    },
    {
      "epoch": 309.77575757575755,
      "grad_norm": 0.3027060925960541,
      "learning_rate": 1.593e-05,
      "loss": 0.3636,
      "step": 3407
    },
    {
      "epoch": 309.8727272727273,
      "grad_norm": 0.3122222423553467,
      "learning_rate": 1.592e-05,
      "loss": 0.3476,
      "step": 3408
    },
    {
      "epoch": 309.969696969697,
      "grad_norm": 0.270575612783432,
      "learning_rate": 1.591e-05,
      "loss": 0.3181,
      "step": 3409
    },
    {
      "epoch": 310.0,
      "grad_norm": 0.5845109224319458,
      "learning_rate": 1.59e-05,
      "loss": 0.3362,
      "step": 3410
    },
    {
      "epoch": 310.0,
      "eval_loss": 0.42802172899246216,
      "eval_runtime": 2.127,
      "eval_samples_per_second": 25.858,
      "eval_steps_per_second": 3.291,
      "step": 3410
    },
    {
      "epoch": 310.0969696969697,
      "grad_norm": 0.2749779522418976,
      "learning_rate": 1.5890000000000002e-05,
      "loss": 0.329,
      "step": 3411
    },
    {
      "epoch": 310.1939393939394,
      "grad_norm": 0.3250596821308136,
      "learning_rate": 1.588e-05,
      "loss": 0.3682,
      "step": 3412
    },
    {
      "epoch": 310.2909090909091,
      "grad_norm": 0.3222499489784241,
      "learning_rate": 1.5870000000000002e-05,
      "loss": 0.3753,
      "step": 3413
    },
    {
      "epoch": 310.3878787878788,
      "grad_norm": 0.3307608366012573,
      "learning_rate": 1.586e-05,
      "loss": 0.3522,
      "step": 3414
    },
    {
      "epoch": 310.4848484848485,
      "grad_norm": 0.2511182725429535,
      "learning_rate": 1.5850000000000002e-05,
      "loss": 0.34,
      "step": 3415
    },
    {
      "epoch": 310.58181818181816,
      "grad_norm": 0.2914133071899414,
      "learning_rate": 1.584e-05,
      "loss": 0.297,
      "step": 3416
    },
    {
      "epoch": 310.6787878787879,
      "grad_norm": 0.26581066846847534,
      "learning_rate": 1.583e-05,
      "loss": 0.3568,
      "step": 3417
    },
    {
      "epoch": 310.77575757575755,
      "grad_norm": 0.35770997405052185,
      "learning_rate": 1.582e-05,
      "loss": 0.3209,
      "step": 3418
    },
    {
      "epoch": 310.8727272727273,
      "grad_norm": 0.3471708595752716,
      "learning_rate": 1.581e-05,
      "loss": 0.3288,
      "step": 3419
    },
    {
      "epoch": 310.969696969697,
      "grad_norm": 0.3137297034263611,
      "learning_rate": 1.58e-05,
      "loss": 0.3496,
      "step": 3420
    },
    {
      "epoch": 310.969696969697,
      "eval_loss": 0.4279772639274597,
      "eval_runtime": 2.1187,
      "eval_samples_per_second": 25.959,
      "eval_steps_per_second": 3.304,
      "step": 3420
    },
    {
      "epoch": 311.0,
      "grad_norm": 0.5063175559043884,
      "learning_rate": 1.5790000000000003e-05,
      "loss": 0.2936,
      "step": 3421
    },
    {
      "epoch": 311.0969696969697,
      "grad_norm": 0.299641877412796,
      "learning_rate": 1.578e-05,
      "loss": 0.3563,
      "step": 3422
    },
    {
      "epoch": 311.1939393939394,
      "grad_norm": 0.2759750485420227,
      "learning_rate": 1.577e-05,
      "loss": 0.3226,
      "step": 3423
    },
    {
      "epoch": 311.2909090909091,
      "grad_norm": 0.34390154480934143,
      "learning_rate": 1.5759999999999998e-05,
      "loss": 0.3265,
      "step": 3424
    },
    {
      "epoch": 311.3878787878788,
      "grad_norm": 0.3145259618759155,
      "learning_rate": 1.575e-05,
      "loss": 0.344,
      "step": 3425
    },
    {
      "epoch": 311.4848484848485,
      "grad_norm": 0.3961293697357178,
      "learning_rate": 1.5740000000000002e-05,
      "loss": 0.3178,
      "step": 3426
    },
    {
      "epoch": 311.58181818181816,
      "grad_norm": 0.3132284879684448,
      "learning_rate": 1.573e-05,
      "loss": 0.3608,
      "step": 3427
    },
    {
      "epoch": 311.6787878787879,
      "grad_norm": 0.2846105992794037,
      "learning_rate": 1.5720000000000002e-05,
      "loss": 0.3414,
      "step": 3428
    },
    {
      "epoch": 311.77575757575755,
      "grad_norm": 0.3138156235218048,
      "learning_rate": 1.571e-05,
      "loss": 0.3688,
      "step": 3429
    },
    {
      "epoch": 311.8727272727273,
      "grad_norm": 0.3472890853881836,
      "learning_rate": 1.5700000000000002e-05,
      "loss": 0.3496,
      "step": 3430
    },
    {
      "epoch": 311.8727272727273,
      "eval_loss": 0.4282764494419098,
      "eval_runtime": 2.1266,
      "eval_samples_per_second": 25.863,
      "eval_steps_per_second": 3.292,
      "step": 3430
    },
    {
      "epoch": 311.969696969697,
      "grad_norm": 0.27399343252182007,
      "learning_rate": 1.569e-05,
      "loss": 0.3055,
      "step": 3431
    },
    {
      "epoch": 312.0,
      "grad_norm": 0.5205172896385193,
      "learning_rate": 1.568e-05,
      "loss": 0.3608,
      "step": 3432
    },
    {
      "epoch": 312.0969696969697,
      "grad_norm": 0.30829933285713196,
      "learning_rate": 1.567e-05,
      "loss": 0.306,
      "step": 3433
    },
    {
      "epoch": 312.1939393939394,
      "grad_norm": 0.3162461519241333,
      "learning_rate": 1.566e-05,
      "loss": 0.3345,
      "step": 3434
    },
    {
      "epoch": 312.2909090909091,
      "grad_norm": 0.28259918093681335,
      "learning_rate": 1.565e-05,
      "loss": 0.3672,
      "step": 3435
    },
    {
      "epoch": 312.3878787878788,
      "grad_norm": 0.28280651569366455,
      "learning_rate": 1.5640000000000003e-05,
      "loss": 0.3485,
      "step": 3436
    },
    {
      "epoch": 312.4848484848485,
      "grad_norm": 0.29870906472206116,
      "learning_rate": 1.563e-05,
      "loss": 0.3635,
      "step": 3437
    },
    {
      "epoch": 312.58181818181816,
      "grad_norm": 0.30681851506233215,
      "learning_rate": 1.5620000000000003e-05,
      "loss": 0.3335,
      "step": 3438
    },
    {
      "epoch": 312.6787878787879,
      "grad_norm": 0.29939502477645874,
      "learning_rate": 1.561e-05,
      "loss": 0.3349,
      "step": 3439
    },
    {
      "epoch": 312.77575757575755,
      "grad_norm": 0.3255018889904022,
      "learning_rate": 1.56e-05,
      "loss": 0.3386,
      "step": 3440
    },
    {
      "epoch": 312.77575757575755,
      "eval_loss": 0.4275350272655487,
      "eval_runtime": 2.1392,
      "eval_samples_per_second": 25.711,
      "eval_steps_per_second": 3.272,
      "step": 3440
    },
    {
      "epoch": 312.8727272727273,
      "grad_norm": 0.29633569717407227,
      "learning_rate": 1.559e-05,
      "loss": 0.3466,
      "step": 3441
    },
    {
      "epoch": 312.969696969697,
      "grad_norm": 0.26882505416870117,
      "learning_rate": 1.558e-05,
      "loss": 0.3449,
      "step": 3442
    },
    {
      "epoch": 313.0,
      "grad_norm": 0.43489527702331543,
      "learning_rate": 1.5570000000000002e-05,
      "loss": 0.2805,
      "step": 3443
    },
    {
      "epoch": 313.0969696969697,
      "grad_norm": 0.28184351325035095,
      "learning_rate": 1.556e-05,
      "loss": 0.3146,
      "step": 3444
    },
    {
      "epoch": 313.1939393939394,
      "grad_norm": 0.32311639189720154,
      "learning_rate": 1.5550000000000002e-05,
      "loss": 0.3836,
      "step": 3445
    },
    {
      "epoch": 313.2909090909091,
      "grad_norm": 0.3593761920928955,
      "learning_rate": 1.554e-05,
      "loss": 0.3481,
      "step": 3446
    },
    {
      "epoch": 313.3878787878788,
      "grad_norm": 0.3303178548812866,
      "learning_rate": 1.553e-05,
      "loss": 0.3512,
      "step": 3447
    },
    {
      "epoch": 313.4848484848485,
      "grad_norm": 0.2715561091899872,
      "learning_rate": 1.552e-05,
      "loss": 0.3606,
      "step": 3448
    },
    {
      "epoch": 313.58181818181816,
      "grad_norm": 0.31712865829467773,
      "learning_rate": 1.551e-05,
      "loss": 0.3525,
      "step": 3449
    },
    {
      "epoch": 313.6787878787879,
      "grad_norm": 0.30805760622024536,
      "learning_rate": 1.55e-05,
      "loss": 0.3276,
      "step": 3450
    },
    {
      "epoch": 313.6787878787879,
      "eval_loss": 0.4274803400039673,
      "eval_runtime": 2.1241,
      "eval_samples_per_second": 25.894,
      "eval_steps_per_second": 3.296,
      "step": 3450
    },
    {
      "epoch": 313.77575757575755,
      "grad_norm": 0.26966550946235657,
      "learning_rate": 1.5490000000000002e-05,
      "loss": 0.3137,
      "step": 3451
    },
    {
      "epoch": 313.8727272727273,
      "grad_norm": 0.2930833399295807,
      "learning_rate": 1.548e-05,
      "loss": 0.3352,
      "step": 3452
    },
    {
      "epoch": 313.969696969697,
      "grad_norm": 0.29453831911087036,
      "learning_rate": 1.5470000000000003e-05,
      "loss": 0.3087,
      "step": 3453
    },
    {
      "epoch": 314.0,
      "grad_norm": 0.4948039948940277,
      "learning_rate": 1.546e-05,
      "loss": 0.3371,
      "step": 3454
    },
    {
      "epoch": 314.0969696969697,
      "grad_norm": 0.29488882422447205,
      "learning_rate": 1.545e-05,
      "loss": 0.3489,
      "step": 3455
    },
    {
      "epoch": 314.1939393939394,
      "grad_norm": 0.2348889857530594,
      "learning_rate": 1.544e-05,
      "loss": 0.3255,
      "step": 3456
    },
    {
      "epoch": 314.2909090909091,
      "grad_norm": 0.3093610405921936,
      "learning_rate": 1.543e-05,
      "loss": 0.3239,
      "step": 3457
    },
    {
      "epoch": 314.3878787878788,
      "grad_norm": 0.30039307475090027,
      "learning_rate": 1.542e-05,
      "loss": 0.3257,
      "step": 3458
    },
    {
      "epoch": 314.4848484848485,
      "grad_norm": 0.3127565085887909,
      "learning_rate": 1.541e-05,
      "loss": 0.3782,
      "step": 3459
    },
    {
      "epoch": 314.58181818181816,
      "grad_norm": 0.2688782513141632,
      "learning_rate": 1.54e-05,
      "loss": 0.3159,
      "step": 3460
    },
    {
      "epoch": 314.58181818181816,
      "eval_loss": 0.4277699291706085,
      "eval_runtime": 2.1333,
      "eval_samples_per_second": 25.782,
      "eval_steps_per_second": 3.281,
      "step": 3460
    },
    {
      "epoch": 314.6787878787879,
      "grad_norm": 0.2904873192310333,
      "learning_rate": 1.539e-05,
      "loss": 0.3127,
      "step": 3461
    },
    {
      "epoch": 314.77575757575755,
      "grad_norm": 0.34919148683547974,
      "learning_rate": 1.538e-05,
      "loss": 0.365,
      "step": 3462
    },
    {
      "epoch": 314.8727272727273,
      "grad_norm": 0.3132627010345459,
      "learning_rate": 1.537e-05,
      "loss": 0.363,
      "step": 3463
    },
    {
      "epoch": 314.969696969697,
      "grad_norm": 0.30896368622779846,
      "learning_rate": 1.536e-05,
      "loss": 0.3534,
      "step": 3464
    },
    {
      "epoch": 315.0,
      "grad_norm": 0.4962367117404938,
      "learning_rate": 1.535e-05,
      "loss": 0.272,
      "step": 3465
    },
    {
      "epoch": 315.0969696969697,
      "grad_norm": 0.3301921784877777,
      "learning_rate": 1.5340000000000002e-05,
      "loss": 0.3477,
      "step": 3466
    },
    {
      "epoch": 315.1939393939394,
      "grad_norm": 0.26896604895591736,
      "learning_rate": 1.533e-05,
      "loss": 0.3477,
      "step": 3467
    },
    {
      "epoch": 315.2909090909091,
      "grad_norm": 0.28536197543144226,
      "learning_rate": 1.5320000000000002e-05,
      "loss": 0.3169,
      "step": 3468
    },
    {
      "epoch": 315.3878787878788,
      "grad_norm": 0.49175986647605896,
      "learning_rate": 1.531e-05,
      "loss": 0.3347,
      "step": 3469
    },
    {
      "epoch": 315.4848484848485,
      "grad_norm": 0.28688549995422363,
      "learning_rate": 1.53e-05,
      "loss": 0.3294,
      "step": 3470
    },
    {
      "epoch": 315.4848484848485,
      "eval_loss": 0.4271085262298584,
      "eval_runtime": 2.1397,
      "eval_samples_per_second": 25.705,
      "eval_steps_per_second": 3.272,
      "step": 3470
    },
    {
      "epoch": 315.58181818181816,
      "grad_norm": 0.3667631149291992,
      "learning_rate": 1.529e-05,
      "loss": 0.3546,
      "step": 3471
    },
    {
      "epoch": 315.6787878787879,
      "grad_norm": 0.3338446021080017,
      "learning_rate": 1.528e-05,
      "loss": 0.3329,
      "step": 3472
    },
    {
      "epoch": 315.77575757575755,
      "grad_norm": 0.3099835515022278,
      "learning_rate": 1.527e-05,
      "loss": 0.3548,
      "step": 3473
    },
    {
      "epoch": 315.8727272727273,
      "grad_norm": 0.25447139143943787,
      "learning_rate": 1.5260000000000003e-05,
      "loss": 0.3349,
      "step": 3474
    },
    {
      "epoch": 315.969696969697,
      "grad_norm": 0.3141191899776459,
      "learning_rate": 1.525e-05,
      "loss": 0.3448,
      "step": 3475
    },
    {
      "epoch": 316.0,
      "grad_norm": 0.58322674036026,
      "learning_rate": 1.5240000000000001e-05,
      "loss": 0.3171,
      "step": 3476
    },
    {
      "epoch": 316.0969696969697,
      "grad_norm": 0.26733261346817017,
      "learning_rate": 1.523e-05,
      "loss": 0.3089,
      "step": 3477
    },
    {
      "epoch": 316.1939393939394,
      "grad_norm": 0.28217995166778564,
      "learning_rate": 1.5220000000000002e-05,
      "loss": 0.3472,
      "step": 3478
    },
    {
      "epoch": 316.2909090909091,
      "grad_norm": 0.3269166946411133,
      "learning_rate": 1.5210000000000002e-05,
      "loss": 0.3838,
      "step": 3479
    },
    {
      "epoch": 316.3878787878788,
      "grad_norm": 0.37794098258018494,
      "learning_rate": 1.52e-05,
      "loss": 0.3263,
      "step": 3480
    },
    {
      "epoch": 316.3878787878788,
      "eval_loss": 0.42741304636001587,
      "eval_runtime": 2.1304,
      "eval_samples_per_second": 25.816,
      "eval_steps_per_second": 3.286,
      "step": 3480
    },
    {
      "epoch": 316.4848484848485,
      "grad_norm": 0.3018639087677002,
      "learning_rate": 1.5190000000000002e-05,
      "loss": 0.3679,
      "step": 3481
    },
    {
      "epoch": 316.58181818181816,
      "grad_norm": 0.2624432444572449,
      "learning_rate": 1.518e-05,
      "loss": 0.3288,
      "step": 3482
    },
    {
      "epoch": 316.6787878787879,
      "grad_norm": 0.2726985216140747,
      "learning_rate": 1.517e-05,
      "loss": 0.3474,
      "step": 3483
    },
    {
      "epoch": 316.77575757575755,
      "grad_norm": 0.25825533270835876,
      "learning_rate": 1.5160000000000002e-05,
      "loss": 0.3117,
      "step": 3484
    },
    {
      "epoch": 316.8727272727273,
      "grad_norm": 0.2738543450832367,
      "learning_rate": 1.515e-05,
      "loss": 0.3531,
      "step": 3485
    },
    {
      "epoch": 316.969696969697,
      "grad_norm": 0.30904558300971985,
      "learning_rate": 1.514e-05,
      "loss": 0.3376,
      "step": 3486
    },
    {
      "epoch": 317.0,
      "grad_norm": 0.49280738830566406,
      "learning_rate": 1.5129999999999999e-05,
      "loss": 0.2761,
      "step": 3487
    },
    {
      "epoch": 317.0969696969697,
      "grad_norm": 0.3254148066043854,
      "learning_rate": 1.5120000000000001e-05,
      "loss": 0.3588,
      "step": 3488
    },
    {
      "epoch": 317.1939393939394,
      "grad_norm": 0.2624494135379791,
      "learning_rate": 1.5110000000000003e-05,
      "loss": 0.3318,
      "step": 3489
    },
    {
      "epoch": 317.2909090909091,
      "grad_norm": 0.29126566648483276,
      "learning_rate": 1.51e-05,
      "loss": 0.3851,
      "step": 3490
    },
    {
      "epoch": 317.2909090909091,
      "eval_loss": 0.42778417468070984,
      "eval_runtime": 2.1286,
      "eval_samples_per_second": 25.838,
      "eval_steps_per_second": 3.288,
      "step": 3490
    },
    {
      "epoch": 317.3878787878788,
      "grad_norm": 0.2885343134403229,
      "learning_rate": 1.5090000000000001e-05,
      "loss": 0.3627,
      "step": 3491
    },
    {
      "epoch": 317.4848484848485,
      "grad_norm": 0.2681112289428711,
      "learning_rate": 1.508e-05,
      "loss": 0.3011,
      "step": 3492
    },
    {
      "epoch": 317.58181818181816,
      "grad_norm": 0.325305312871933,
      "learning_rate": 1.5070000000000001e-05,
      "loss": 0.3809,
      "step": 3493
    },
    {
      "epoch": 317.6787878787879,
      "grad_norm": 0.34081873297691345,
      "learning_rate": 1.5060000000000001e-05,
      "loss": 0.3354,
      "step": 3494
    },
    {
      "epoch": 317.77575757575755,
      "grad_norm": 0.2977830469608307,
      "learning_rate": 1.505e-05,
      "loss": 0.3032,
      "step": 3495
    },
    {
      "epoch": 317.8727272727273,
      "grad_norm": 0.2674584686756134,
      "learning_rate": 1.5040000000000002e-05,
      "loss": 0.3162,
      "step": 3496
    },
    {
      "epoch": 317.969696969697,
      "grad_norm": 0.4172142446041107,
      "learning_rate": 1.503e-05,
      "loss": 0.3079,
      "step": 3497
    },
    {
      "epoch": 318.0,
      "grad_norm": 0.47455283999443054,
      "learning_rate": 1.502e-05,
      "loss": 0.3574,
      "step": 3498
    },
    {
      "epoch": 318.0969696969697,
      "grad_norm": 0.29889288544654846,
      "learning_rate": 1.5010000000000002e-05,
      "loss": 0.3538,
      "step": 3499
    },
    {
      "epoch": 318.1939393939394,
      "grad_norm": 0.32577717304229736,
      "learning_rate": 1.5e-05,
      "loss": 0.288,
      "step": 3500
    },
    {
      "epoch": 318.1939393939394,
      "eval_loss": 0.42727982997894287,
      "eval_runtime": 2.1191,
      "eval_samples_per_second": 25.954,
      "eval_steps_per_second": 3.303,
      "step": 3500
    },
    {
      "epoch": 318.2909090909091,
      "grad_norm": 0.2910911440849304,
      "learning_rate": 1.499e-05,
      "loss": 0.3426,
      "step": 3501
    },
    {
      "epoch": 318.3878787878788,
      "grad_norm": 0.3130618929862976,
      "learning_rate": 1.4979999999999999e-05,
      "loss": 0.351,
      "step": 3502
    },
    {
      "epoch": 318.4848484848485,
      "grad_norm": 0.35861071944236755,
      "learning_rate": 1.497e-05,
      "loss": 0.3723,
      "step": 3503
    },
    {
      "epoch": 318.58181818181816,
      "grad_norm": 0.31288236379623413,
      "learning_rate": 1.4960000000000002e-05,
      "loss": 0.3183,
      "step": 3504
    },
    {
      "epoch": 318.6787878787879,
      "grad_norm": 0.2688522934913635,
      "learning_rate": 1.4950000000000001e-05,
      "loss": 0.3124,
      "step": 3505
    },
    {
      "epoch": 318.77575757575755,
      "grad_norm": 0.2826993465423584,
      "learning_rate": 1.4940000000000001e-05,
      "loss": 0.3453,
      "step": 3506
    },
    {
      "epoch": 318.8727272727273,
      "grad_norm": 0.3165665864944458,
      "learning_rate": 1.493e-05,
      "loss": 0.3775,
      "step": 3507
    },
    {
      "epoch": 318.969696969697,
      "grad_norm": 0.2961466610431671,
      "learning_rate": 1.4920000000000001e-05,
      "loss": 0.3331,
      "step": 3508
    },
    {
      "epoch": 319.0,
      "grad_norm": 0.5470519661903381,
      "learning_rate": 1.4910000000000001e-05,
      "loss": 0.3091,
      "step": 3509
    },
    {
      "epoch": 319.0969696969697,
      "grad_norm": 0.28951361775398254,
      "learning_rate": 1.49e-05,
      "loss": 0.3426,
      "step": 3510
    },
    {
      "epoch": 319.0969696969697,
      "eval_loss": 0.4273596704006195,
      "eval_runtime": 2.116,
      "eval_samples_per_second": 25.992,
      "eval_steps_per_second": 3.308,
      "step": 3510
    },
    {
      "epoch": 319.1939393939394,
      "grad_norm": 0.26769697666168213,
      "learning_rate": 1.4890000000000001e-05,
      "loss": 0.3528,
      "step": 3511
    },
    {
      "epoch": 319.2909090909091,
      "grad_norm": 0.30560117959976196,
      "learning_rate": 1.488e-05,
      "loss": 0.3703,
      "step": 3512
    },
    {
      "epoch": 319.3878787878788,
      "grad_norm": 0.26912692189216614,
      "learning_rate": 1.487e-05,
      "loss": 0.336,
      "step": 3513
    },
    {
      "epoch": 319.4848484848485,
      "grad_norm": 0.2910601794719696,
      "learning_rate": 1.4860000000000002e-05,
      "loss": 0.3301,
      "step": 3514
    },
    {
      "epoch": 319.58181818181816,
      "grad_norm": 0.29013335704803467,
      "learning_rate": 1.485e-05,
      "loss": 0.3387,
      "step": 3515
    },
    {
      "epoch": 319.6787878787879,
      "grad_norm": 0.279678612947464,
      "learning_rate": 1.4840000000000002e-05,
      "loss": 0.3256,
      "step": 3516
    },
    {
      "epoch": 319.77575757575755,
      "grad_norm": 0.2691352367401123,
      "learning_rate": 1.4829999999999999e-05,
      "loss": 0.3125,
      "step": 3517
    },
    {
      "epoch": 319.8727272727273,
      "grad_norm": 0.28071603178977966,
      "learning_rate": 1.482e-05,
      "loss": 0.3485,
      "step": 3518
    },
    {
      "epoch": 319.969696969697,
      "grad_norm": 0.32526275515556335,
      "learning_rate": 1.4810000000000002e-05,
      "loss": 0.3243,
      "step": 3519
    },
    {
      "epoch": 320.0,
      "grad_norm": 0.60272616147995,
      "learning_rate": 1.48e-05,
      "loss": 0.3513,
      "step": 3520
    },
    {
      "epoch": 320.0,
      "eval_loss": 0.42688173055648804,
      "eval_runtime": 2.129,
      "eval_samples_per_second": 25.833,
      "eval_steps_per_second": 3.288,
      "step": 3520
    },
    {
      "epoch": 320.0969696969697,
      "grad_norm": 0.32412081956863403,
      "learning_rate": 1.479e-05,
      "loss": 0.3895,
      "step": 3521
    },
    {
      "epoch": 320.1939393939394,
      "grad_norm": 0.2637699246406555,
      "learning_rate": 1.4779999999999999e-05,
      "loss": 0.3519,
      "step": 3522
    },
    {
      "epoch": 320.2909090909091,
      "grad_norm": 0.2972037196159363,
      "learning_rate": 1.4770000000000001e-05,
      "loss": 0.36,
      "step": 3523
    },
    {
      "epoch": 320.3878787878788,
      "grad_norm": 0.3079843819141388,
      "learning_rate": 1.4760000000000001e-05,
      "loss": 0.3174,
      "step": 3524
    },
    {
      "epoch": 320.4848484848485,
      "grad_norm": 0.3062552809715271,
      "learning_rate": 1.475e-05,
      "loss": 0.3088,
      "step": 3525
    },
    {
      "epoch": 320.58181818181816,
      "grad_norm": 0.2870745062828064,
      "learning_rate": 1.4740000000000001e-05,
      "loss": 0.2884,
      "step": 3526
    },
    {
      "epoch": 320.6787878787879,
      "grad_norm": 0.2897455096244812,
      "learning_rate": 1.473e-05,
      "loss": 0.3331,
      "step": 3527
    },
    {
      "epoch": 320.77575757575755,
      "grad_norm": 0.29129114747047424,
      "learning_rate": 1.472e-05,
      "loss": 0.3259,
      "step": 3528
    },
    {
      "epoch": 320.8727272727273,
      "grad_norm": 0.29049405455589294,
      "learning_rate": 1.4710000000000001e-05,
      "loss": 0.3441,
      "step": 3529
    },
    {
      "epoch": 320.969696969697,
      "grad_norm": 0.3123665750026703,
      "learning_rate": 1.47e-05,
      "loss": 0.3592,
      "step": 3530
    },
    {
      "epoch": 320.969696969697,
      "eval_loss": 0.42728039622306824,
      "eval_runtime": 2.1096,
      "eval_samples_per_second": 26.071,
      "eval_steps_per_second": 3.318,
      "step": 3530
    },
    {
      "epoch": 321.0,
      "grad_norm": 0.6494812965393066,
      "learning_rate": 1.4690000000000002e-05,
      "loss": 0.3496,
      "step": 3531
    },
    {
      "epoch": 321.0969696969697,
      "grad_norm": 0.24925000965595245,
      "learning_rate": 1.4680000000000002e-05,
      "loss": 0.3694,
      "step": 3532
    },
    {
      "epoch": 321.1939393939394,
      "grad_norm": 0.283581405878067,
      "learning_rate": 1.467e-05,
      "loss": 0.3035,
      "step": 3533
    },
    {
      "epoch": 321.2909090909091,
      "grad_norm": 0.2899045944213867,
      "learning_rate": 1.4660000000000002e-05,
      "loss": 0.3194,
      "step": 3534
    },
    {
      "epoch": 321.3878787878788,
      "grad_norm": 0.2756378948688507,
      "learning_rate": 1.465e-05,
      "loss": 0.3162,
      "step": 3535
    },
    {
      "epoch": 321.4848484848485,
      "grad_norm": 0.24909161031246185,
      "learning_rate": 1.464e-05,
      "loss": 0.3253,
      "step": 3536
    },
    {
      "epoch": 321.58181818181816,
      "grad_norm": 0.33241987228393555,
      "learning_rate": 1.4630000000000002e-05,
      "loss": 0.3408,
      "step": 3537
    },
    {
      "epoch": 321.6787878787879,
      "grad_norm": 0.33643579483032227,
      "learning_rate": 1.462e-05,
      "loss": 0.3442,
      "step": 3538
    },
    {
      "epoch": 321.77575757575755,
      "grad_norm": 0.24275629222393036,
      "learning_rate": 1.461e-05,
      "loss": 0.3182,
      "step": 3539
    },
    {
      "epoch": 321.8727272727273,
      "grad_norm": 0.3069804310798645,
      "learning_rate": 1.4599999999999999e-05,
      "loss": 0.3636,
      "step": 3540
    },
    {
      "epoch": 321.8727272727273,
      "eval_loss": 0.42699331045150757,
      "eval_runtime": 2.1078,
      "eval_samples_per_second": 26.093,
      "eval_steps_per_second": 3.321,
      "step": 3540
    },
    {
      "epoch": 321.969696969697,
      "grad_norm": 0.3260021209716797,
      "learning_rate": 1.4590000000000001e-05,
      "loss": 0.3606,
      "step": 3541
    },
    {
      "epoch": 322.0,
      "grad_norm": 0.5835338830947876,
      "learning_rate": 1.4580000000000003e-05,
      "loss": 0.3974,
      "step": 3542
    },
    {
      "epoch": 322.0969696969697,
      "grad_norm": 0.2785661220550537,
      "learning_rate": 1.4570000000000001e-05,
      "loss": 0.3258,
      "step": 3543
    },
    {
      "epoch": 322.1939393939394,
      "grad_norm": 0.2921468913555145,
      "learning_rate": 1.4560000000000001e-05,
      "loss": 0.3604,
      "step": 3544
    },
    {
      "epoch": 322.2909090909091,
      "grad_norm": 0.36530375480651855,
      "learning_rate": 1.455e-05,
      "loss": 0.3382,
      "step": 3545
    },
    {
      "epoch": 322.3878787878788,
      "grad_norm": 0.24066120386123657,
      "learning_rate": 1.4540000000000001e-05,
      "loss": 0.2964,
      "step": 3546
    },
    {
      "epoch": 322.4848484848485,
      "grad_norm": 0.30269855260849,
      "learning_rate": 1.4530000000000001e-05,
      "loss": 0.3187,
      "step": 3547
    },
    {
      "epoch": 322.58181818181816,
      "grad_norm": 0.3029041290283203,
      "learning_rate": 1.452e-05,
      "loss": 0.3732,
      "step": 3548
    },
    {
      "epoch": 322.6787878787879,
      "grad_norm": 0.2999378740787506,
      "learning_rate": 1.4510000000000002e-05,
      "loss": 0.3732,
      "step": 3549
    },
    {
      "epoch": 322.77575757575755,
      "grad_norm": 0.288894385099411,
      "learning_rate": 1.45e-05,
      "loss": 0.3201,
      "step": 3550
    },
    {
      "epoch": 322.77575757575755,
      "eval_loss": 0.4270142912864685,
      "eval_runtime": 2.1015,
      "eval_samples_per_second": 26.172,
      "eval_steps_per_second": 3.331,
      "step": 3550
    },
    {
      "epoch": 322.8727272727273,
      "grad_norm": 0.2889610230922699,
      "learning_rate": 1.449e-05,
      "loss": 0.3297,
      "step": 3551
    },
    {
      "epoch": 322.969696969697,
      "grad_norm": 0.3216736614704132,
      "learning_rate": 1.4480000000000002e-05,
      "loss": 0.3422,
      "step": 3552
    },
    {
      "epoch": 323.0,
      "grad_norm": 0.5026072263717651,
      "learning_rate": 1.447e-05,
      "loss": 0.3391,
      "step": 3553
    },
    {
      "epoch": 323.0969696969697,
      "grad_norm": 0.4211641252040863,
      "learning_rate": 1.4460000000000002e-05,
      "loss": 0.3586,
      "step": 3554
    },
    {
      "epoch": 323.1939393939394,
      "grad_norm": 0.30118682980537415,
      "learning_rate": 1.4449999999999999e-05,
      "loss": 0.332,
      "step": 3555
    },
    {
      "epoch": 323.2909090909091,
      "grad_norm": 0.25670328736305237,
      "learning_rate": 1.444e-05,
      "loss": 0.3582,
      "step": 3556
    },
    {
      "epoch": 323.3878787878788,
      "grad_norm": 0.3140673339366913,
      "learning_rate": 1.4430000000000002e-05,
      "loss": 0.3238,
      "step": 3557
    },
    {
      "epoch": 323.4848484848485,
      "grad_norm": 0.32594263553619385,
      "learning_rate": 1.4420000000000001e-05,
      "loss": 0.3079,
      "step": 3558
    },
    {
      "epoch": 323.58181818181816,
      "grad_norm": 0.3219808042049408,
      "learning_rate": 1.4410000000000001e-05,
      "loss": 0.3602,
      "step": 3559
    },
    {
      "epoch": 323.6787878787879,
      "grad_norm": 0.27611875534057617,
      "learning_rate": 1.44e-05,
      "loss": 0.3265,
      "step": 3560
    },
    {
      "epoch": 323.6787878787879,
      "eval_loss": 0.4271848499774933,
      "eval_runtime": 2.1293,
      "eval_samples_per_second": 25.831,
      "eval_steps_per_second": 3.288,
      "step": 3560
    },
    {
      "epoch": 323.77575757575755,
      "grad_norm": 0.29057878255844116,
      "learning_rate": 1.4390000000000001e-05,
      "loss": 0.3325,
      "step": 3561
    },
    {
      "epoch": 323.8727272727273,
      "grad_norm": 0.2801297605037689,
      "learning_rate": 1.4380000000000001e-05,
      "loss": 0.3796,
      "step": 3562
    },
    {
      "epoch": 323.969696969697,
      "grad_norm": 0.3203805983066559,
      "learning_rate": 1.437e-05,
      "loss": 0.3119,
      "step": 3563
    },
    {
      "epoch": 324.0,
      "grad_norm": 0.4999186396598816,
      "learning_rate": 1.4360000000000001e-05,
      "loss": 0.3017,
      "step": 3564
    },
    {
      "epoch": 324.0969696969697,
      "grad_norm": 0.2902086675167084,
      "learning_rate": 1.435e-05,
      "loss": 0.2954,
      "step": 3565
    },
    {
      "epoch": 324.1939393939394,
      "grad_norm": 0.26491281390190125,
      "learning_rate": 1.434e-05,
      "loss": 0.3254,
      "step": 3566
    },
    {
      "epoch": 324.2909090909091,
      "grad_norm": 0.32385510206222534,
      "learning_rate": 1.4330000000000002e-05,
      "loss": 0.3507,
      "step": 3567
    },
    {
      "epoch": 324.3878787878788,
      "grad_norm": 0.2817283570766449,
      "learning_rate": 1.432e-05,
      "loss": 0.3183,
      "step": 3568
    },
    {
      "epoch": 324.4848484848485,
      "grad_norm": 0.32849523425102234,
      "learning_rate": 1.4310000000000002e-05,
      "loss": 0.3303,
      "step": 3569
    },
    {
      "epoch": 324.58181818181816,
      "grad_norm": 0.30101409554481506,
      "learning_rate": 1.43e-05,
      "loss": 0.3005,
      "step": 3570
    },
    {
      "epoch": 324.58181818181816,
      "eval_loss": 0.4270551800727844,
      "eval_runtime": 2.166,
      "eval_samples_per_second": 25.393,
      "eval_steps_per_second": 3.232,
      "step": 3570
    },
    {
      "epoch": 324.6787878787879,
      "grad_norm": 0.2798803448677063,
      "learning_rate": 1.429e-05,
      "loss": 0.3218,
      "step": 3571
    },
    {
      "epoch": 324.77575757575755,
      "grad_norm": 0.30352485179901123,
      "learning_rate": 1.4280000000000002e-05,
      "loss": 0.3478,
      "step": 3572
    },
    {
      "epoch": 324.8727272727273,
      "grad_norm": 0.344475120306015,
      "learning_rate": 1.427e-05,
      "loss": 0.3741,
      "step": 3573
    },
    {
      "epoch": 324.969696969697,
      "grad_norm": 0.3088480234146118,
      "learning_rate": 1.426e-05,
      "loss": 0.3926,
      "step": 3574
    },
    {
      "epoch": 325.0,
      "grad_norm": 0.5807185173034668,
      "learning_rate": 1.4249999999999999e-05,
      "loss": 0.4022,
      "step": 3575
    },
    {
      "epoch": 325.0969696969697,
      "grad_norm": 0.3358258903026581,
      "learning_rate": 1.4240000000000001e-05,
      "loss": 0.2947,
      "step": 3576
    },
    {
      "epoch": 325.1939393939394,
      "grad_norm": 0.32850319147109985,
      "learning_rate": 1.4230000000000001e-05,
      "loss": 0.3275,
      "step": 3577
    },
    {
      "epoch": 325.2909090909091,
      "grad_norm": 0.33393165469169617,
      "learning_rate": 1.422e-05,
      "loss": 0.3291,
      "step": 3578
    },
    {
      "epoch": 325.3878787878788,
      "grad_norm": 0.2724887430667877,
      "learning_rate": 1.4210000000000001e-05,
      "loss": 0.3414,
      "step": 3579
    },
    {
      "epoch": 325.4848484848485,
      "grad_norm": 0.26683491468429565,
      "learning_rate": 1.42e-05,
      "loss": 0.3547,
      "step": 3580
    },
    {
      "epoch": 325.4848484848485,
      "eval_loss": 0.42706215381622314,
      "eval_runtime": 2.1441,
      "eval_samples_per_second": 25.652,
      "eval_steps_per_second": 3.265,
      "step": 3580
    },
    {
      "epoch": 325.58181818181816,
      "grad_norm": 0.2737730145454407,
      "learning_rate": 1.4190000000000001e-05,
      "loss": 0.343,
      "step": 3581
    },
    {
      "epoch": 325.6787878787879,
      "grad_norm": 0.3594973087310791,
      "learning_rate": 1.4180000000000001e-05,
      "loss": 0.3171,
      "step": 3582
    },
    {
      "epoch": 325.77575757575755,
      "grad_norm": 0.2571316659450531,
      "learning_rate": 1.417e-05,
      "loss": 0.3625,
      "step": 3583
    },
    {
      "epoch": 325.8727272727273,
      "grad_norm": 0.30839765071868896,
      "learning_rate": 1.4160000000000002e-05,
      "loss": 0.3514,
      "step": 3584
    },
    {
      "epoch": 325.969696969697,
      "grad_norm": 0.3523865044116974,
      "learning_rate": 1.415e-05,
      "loss": 0.3502,
      "step": 3585
    },
    {
      "epoch": 326.0,
      "grad_norm": 0.449184775352478,
      "learning_rate": 1.414e-05,
      "loss": 0.3377,
      "step": 3586
    },
    {
      "epoch": 326.0969696969697,
      "grad_norm": 0.2980722486972809,
      "learning_rate": 1.4130000000000002e-05,
      "loss": 0.3289,
      "step": 3587
    },
    {
      "epoch": 326.1939393939394,
      "grad_norm": 0.2593381702899933,
      "learning_rate": 1.412e-05,
      "loss": 0.322,
      "step": 3588
    },
    {
      "epoch": 326.2909090909091,
      "grad_norm": 0.32586827874183655,
      "learning_rate": 1.411e-05,
      "loss": 0.344,
      "step": 3589
    },
    {
      "epoch": 326.3878787878788,
      "grad_norm": 0.30395492911338806,
      "learning_rate": 1.4099999999999999e-05,
      "loss": 0.3395,
      "step": 3590
    },
    {
      "epoch": 326.3878787878788,
      "eval_loss": 0.4270731210708618,
      "eval_runtime": 2.1085,
      "eval_samples_per_second": 26.085,
      "eval_steps_per_second": 3.32,
      "step": 3590
    },
    {
      "epoch": 326.4848484848485,
      "grad_norm": 0.25640663504600525,
      "learning_rate": 1.409e-05,
      "loss": 0.3299,
      "step": 3591
    },
    {
      "epoch": 326.58181818181816,
      "grad_norm": 0.28479263186454773,
      "learning_rate": 1.408e-05,
      "loss": 0.3439,
      "step": 3592
    },
    {
      "epoch": 326.6787878787879,
      "grad_norm": 0.3171486556529999,
      "learning_rate": 1.4069999999999999e-05,
      "loss": 0.3525,
      "step": 3593
    },
    {
      "epoch": 326.77575757575755,
      "grad_norm": 0.28105539083480835,
      "learning_rate": 1.4060000000000001e-05,
      "loss": 0.3661,
      "step": 3594
    },
    {
      "epoch": 326.8727272727273,
      "grad_norm": 0.2885766327381134,
      "learning_rate": 1.4050000000000003e-05,
      "loss": 0.3232,
      "step": 3595
    },
    {
      "epoch": 326.969696969697,
      "grad_norm": 0.2773924767971039,
      "learning_rate": 1.4040000000000001e-05,
      "loss": 0.3151,
      "step": 3596
    },
    {
      "epoch": 327.0,
      "grad_norm": 0.5883837938308716,
      "learning_rate": 1.4030000000000001e-05,
      "loss": 0.3596,
      "step": 3597
    },
    {
      "epoch": 327.0969696969697,
      "grad_norm": 0.29262015223503113,
      "learning_rate": 1.402e-05,
      "loss": 0.3514,
      "step": 3598
    },
    {
      "epoch": 327.1939393939394,
      "grad_norm": 0.26522696018218994,
      "learning_rate": 1.4010000000000001e-05,
      "loss": 0.3189,
      "step": 3599
    },
    {
      "epoch": 327.2909090909091,
      "grad_norm": 0.29591551423072815,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.3355,
      "step": 3600
    },
    {
      "epoch": 327.2909090909091,
      "eval_loss": 0.42640984058380127,
      "eval_runtime": 2.1345,
      "eval_samples_per_second": 25.767,
      "eval_steps_per_second": 3.279,
      "step": 3600
    },
    {
      "epoch": 327.3878787878788,
      "grad_norm": 0.2899519205093384,
      "learning_rate": 1.399e-05,
      "loss": 0.3344,
      "step": 3601
    },
    {
      "epoch": 327.4848484848485,
      "grad_norm": 0.31008467078208923,
      "learning_rate": 1.3980000000000002e-05,
      "loss": 0.319,
      "step": 3602
    },
    {
      "epoch": 327.58181818181816,
      "grad_norm": 0.27912476658821106,
      "learning_rate": 1.397e-05,
      "loss": 0.3533,
      "step": 3603
    },
    {
      "epoch": 327.6787878787879,
      "grad_norm": 0.27924835681915283,
      "learning_rate": 1.396e-05,
      "loss": 0.359,
      "step": 3604
    },
    {
      "epoch": 327.77575757575755,
      "grad_norm": 0.28805863857269287,
      "learning_rate": 1.3950000000000002e-05,
      "loss": 0.3701,
      "step": 3605
    },
    {
      "epoch": 327.8727272727273,
      "grad_norm": 0.32875150442123413,
      "learning_rate": 1.394e-05,
      "loss": 0.3139,
      "step": 3606
    },
    {
      "epoch": 327.969696969697,
      "grad_norm": 0.321266233921051,
      "learning_rate": 1.3930000000000002e-05,
      "loss": 0.3262,
      "step": 3607
    },
    {
      "epoch": 328.0,
      "grad_norm": 0.4865063428878784,
      "learning_rate": 1.3919999999999999e-05,
      "loss": 0.2978,
      "step": 3608
    },
    {
      "epoch": 328.0969696969697,
      "grad_norm": 0.3163537085056305,
      "learning_rate": 1.391e-05,
      "loss": 0.3324,
      "step": 3609
    },
    {
      "epoch": 328.1939393939394,
      "grad_norm": 0.28259798884391785,
      "learning_rate": 1.3900000000000002e-05,
      "loss": 0.371,
      "step": 3610
    },
    {
      "epoch": 328.1939393939394,
      "eval_loss": 0.42678430676460266,
      "eval_runtime": 2.1264,
      "eval_samples_per_second": 25.865,
      "eval_steps_per_second": 3.292,
      "step": 3610
    },
    {
      "epoch": 328.2909090909091,
      "grad_norm": 0.3229399025440216,
      "learning_rate": 1.389e-05,
      "loss": 0.3338,
      "step": 3611
    },
    {
      "epoch": 328.3878787878788,
      "grad_norm": 0.24665595591068268,
      "learning_rate": 1.3880000000000001e-05,
      "loss": 0.3561,
      "step": 3612
    },
    {
      "epoch": 328.4848484848485,
      "grad_norm": 0.3246692717075348,
      "learning_rate": 1.387e-05,
      "loss": 0.3467,
      "step": 3613
    },
    {
      "epoch": 328.58181818181816,
      "grad_norm": 0.24859903752803802,
      "learning_rate": 1.3860000000000001e-05,
      "loss": 0.3192,
      "step": 3614
    },
    {
      "epoch": 328.6787878787879,
      "grad_norm": 0.28727802634239197,
      "learning_rate": 1.3850000000000001e-05,
      "loss": 0.3291,
      "step": 3615
    },
    {
      "epoch": 328.77575757575755,
      "grad_norm": 0.31984421610832214,
      "learning_rate": 1.384e-05,
      "loss": 0.3081,
      "step": 3616
    },
    {
      "epoch": 328.8727272727273,
      "grad_norm": 0.24756306409835815,
      "learning_rate": 1.3830000000000001e-05,
      "loss": 0.3564,
      "step": 3617
    },
    {
      "epoch": 328.969696969697,
      "grad_norm": 0.3125709891319275,
      "learning_rate": 1.382e-05,
      "loss": 0.3245,
      "step": 3618
    },
    {
      "epoch": 329.0,
      "grad_norm": 0.5306656956672668,
      "learning_rate": 1.381e-05,
      "loss": 0.304,
      "step": 3619
    },
    {
      "epoch": 329.0969696969697,
      "grad_norm": 0.3231707811355591,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 0.3664,
      "step": 3620
    },
    {
      "epoch": 329.0969696969697,
      "eval_loss": 0.42667579650878906,
      "eval_runtime": 2.1187,
      "eval_samples_per_second": 25.96,
      "eval_steps_per_second": 3.304,
      "step": 3620
    },
    {
      "epoch": 329.1939393939394,
      "grad_norm": 0.27657705545425415,
      "learning_rate": 1.379e-05,
      "loss": 0.3301,
      "step": 3621
    },
    {
      "epoch": 329.2909090909091,
      "grad_norm": 0.2769893705844879,
      "learning_rate": 1.3780000000000002e-05,
      "loss": 0.3481,
      "step": 3622
    },
    {
      "epoch": 329.3878787878788,
      "grad_norm": 0.2747114300727844,
      "learning_rate": 1.377e-05,
      "loss": 0.3256,
      "step": 3623
    },
    {
      "epoch": 329.4848484848485,
      "grad_norm": 0.37332674860954285,
      "learning_rate": 1.376e-05,
      "loss": 0.3255,
      "step": 3624
    },
    {
      "epoch": 329.58181818181816,
      "grad_norm": 0.28072360157966614,
      "learning_rate": 1.3750000000000002e-05,
      "loss": 0.3399,
      "step": 3625
    },
    {
      "epoch": 329.6787878787879,
      "grad_norm": 0.3200606405735016,
      "learning_rate": 1.374e-05,
      "loss": 0.3075,
      "step": 3626
    },
    {
      "epoch": 329.77575757575755,
      "grad_norm": 0.3110363483428955,
      "learning_rate": 1.373e-05,
      "loss": 0.344,
      "step": 3627
    },
    {
      "epoch": 329.8727272727273,
      "grad_norm": 0.3103969693183899,
      "learning_rate": 1.3719999999999999e-05,
      "loss": 0.3484,
      "step": 3628
    },
    {
      "epoch": 329.969696969697,
      "grad_norm": 0.3344944417476654,
      "learning_rate": 1.3710000000000001e-05,
      "loss": 0.334,
      "step": 3629
    },
    {
      "epoch": 330.0,
      "grad_norm": 0.5091283321380615,
      "learning_rate": 1.3700000000000001e-05,
      "loss": 0.329,
      "step": 3630
    },
    {
      "epoch": 330.0,
      "eval_loss": 0.42662423849105835,
      "eval_runtime": 2.1264,
      "eval_samples_per_second": 25.865,
      "eval_steps_per_second": 3.292,
      "step": 3630
    },
    {
      "epoch": 330.0969696969697,
      "grad_norm": 0.2514973282814026,
      "learning_rate": 1.369e-05,
      "loss": 0.2958,
      "step": 3631
    },
    {
      "epoch": 330.1939393939394,
      "grad_norm": 0.295162171125412,
      "learning_rate": 1.3680000000000001e-05,
      "loss": 0.3649,
      "step": 3632
    },
    {
      "epoch": 330.2909090909091,
      "grad_norm": 0.3357464671134949,
      "learning_rate": 1.367e-05,
      "loss": 0.3441,
      "step": 3633
    },
    {
      "epoch": 330.3878787878788,
      "grad_norm": 0.2648813724517822,
      "learning_rate": 1.3660000000000001e-05,
      "loss": 0.3225,
      "step": 3634
    },
    {
      "epoch": 330.4848484848485,
      "grad_norm": 0.3024829924106598,
      "learning_rate": 1.3650000000000001e-05,
      "loss": 0.3277,
      "step": 3635
    },
    {
      "epoch": 330.58181818181816,
      "grad_norm": 0.3073791265487671,
      "learning_rate": 1.364e-05,
      "loss": 0.3478,
      "step": 3636
    },
    {
      "epoch": 330.6787878787879,
      "grad_norm": 0.3012111485004425,
      "learning_rate": 1.3630000000000002e-05,
      "loss": 0.3716,
      "step": 3637
    },
    {
      "epoch": 330.77575757575755,
      "grad_norm": 0.2969381511211395,
      "learning_rate": 1.362e-05,
      "loss": 0.3264,
      "step": 3638
    },
    {
      "epoch": 330.8727272727273,
      "grad_norm": 0.3305814266204834,
      "learning_rate": 1.361e-05,
      "loss": 0.3185,
      "step": 3639
    },
    {
      "epoch": 330.969696969697,
      "grad_norm": 0.29909998178482056,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.3337,
      "step": 3640
    },
    {
      "epoch": 330.969696969697,
      "eval_loss": 0.426724910736084,
      "eval_runtime": 2.1181,
      "eval_samples_per_second": 25.966,
      "eval_steps_per_second": 3.305,
      "step": 3640
    },
    {
      "epoch": 331.0,
      "grad_norm": 0.6468818783760071,
      "learning_rate": 1.359e-05,
      "loss": 0.3671,
      "step": 3641
    },
    {
      "epoch": 331.0969696969697,
      "grad_norm": 0.2924717664718628,
      "learning_rate": 1.358e-05,
      "loss": 0.3459,
      "step": 3642
    },
    {
      "epoch": 331.1939393939394,
      "grad_norm": 0.3029530644416809,
      "learning_rate": 1.3569999999999999e-05,
      "loss": 0.3425,
      "step": 3643
    },
    {
      "epoch": 331.2909090909091,
      "grad_norm": 0.2762484550476074,
      "learning_rate": 1.356e-05,
      "loss": 0.3558,
      "step": 3644
    },
    {
      "epoch": 331.3878787878788,
      "grad_norm": 0.28036603331565857,
      "learning_rate": 1.3550000000000002e-05,
      "loss": 0.3028,
      "step": 3645
    },
    {
      "epoch": 331.4848484848485,
      "grad_norm": 0.2963285446166992,
      "learning_rate": 1.3539999999999999e-05,
      "loss": 0.3174,
      "step": 3646
    },
    {
      "epoch": 331.58181818181816,
      "grad_norm": 0.2616845667362213,
      "learning_rate": 1.3530000000000001e-05,
      "loss": 0.347,
      "step": 3647
    },
    {
      "epoch": 331.6787878787879,
      "grad_norm": 0.30983370542526245,
      "learning_rate": 1.352e-05,
      "loss": 0.3538,
      "step": 3648
    },
    {
      "epoch": 331.77575757575755,
      "grad_norm": 0.28175094723701477,
      "learning_rate": 1.3510000000000001e-05,
      "loss": 0.3375,
      "step": 3649
    },
    {
      "epoch": 331.8727272727273,
      "grad_norm": 0.27941614389419556,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 0.3228,
      "step": 3650
    },
    {
      "epoch": 331.8727272727273,
      "eval_loss": 0.4264315962791443,
      "eval_runtime": 2.1118,
      "eval_samples_per_second": 26.044,
      "eval_steps_per_second": 3.315,
      "step": 3650
    },
    {
      "epoch": 331.969696969697,
      "grad_norm": 0.28335991501808167,
      "learning_rate": 1.349e-05,
      "loss": 0.3275,
      "step": 3651
    },
    {
      "epoch": 332.0,
      "grad_norm": 0.5670108795166016,
      "learning_rate": 1.3480000000000001e-05,
      "loss": 0.3716,
      "step": 3652
    },
    {
      "epoch": 332.0969696969697,
      "grad_norm": 0.29136744141578674,
      "learning_rate": 1.347e-05,
      "loss": 0.3782,
      "step": 3653
    },
    {
      "epoch": 332.1939393939394,
      "grad_norm": 0.2685430645942688,
      "learning_rate": 1.346e-05,
      "loss": 0.3181,
      "step": 3654
    },
    {
      "epoch": 332.2909090909091,
      "grad_norm": 0.2690100073814392,
      "learning_rate": 1.3450000000000002e-05,
      "loss": 0.3499,
      "step": 3655
    },
    {
      "epoch": 332.3878787878788,
      "grad_norm": 0.28610414266586304,
      "learning_rate": 1.344e-05,
      "loss": 0.3198,
      "step": 3656
    },
    {
      "epoch": 332.4848484848485,
      "grad_norm": 0.3067784905433655,
      "learning_rate": 1.343e-05,
      "loss": 0.3223,
      "step": 3657
    },
    {
      "epoch": 332.58181818181816,
      "grad_norm": 0.27512192726135254,
      "learning_rate": 1.3420000000000002e-05,
      "loss": 0.3342,
      "step": 3658
    },
    {
      "epoch": 332.6787878787879,
      "grad_norm": 0.25142213702201843,
      "learning_rate": 1.341e-05,
      "loss": 0.3191,
      "step": 3659
    },
    {
      "epoch": 332.77575757575755,
      "grad_norm": 0.2903703451156616,
      "learning_rate": 1.3400000000000002e-05,
      "loss": 0.3591,
      "step": 3660
    },
    {
      "epoch": 332.77575757575755,
      "eval_loss": 0.426946222782135,
      "eval_runtime": 2.1164,
      "eval_samples_per_second": 25.987,
      "eval_steps_per_second": 3.307,
      "step": 3660
    },
    {
      "epoch": 332.8727272727273,
      "grad_norm": 0.2897012531757355,
      "learning_rate": 1.339e-05,
      "loss": 0.3465,
      "step": 3661
    },
    {
      "epoch": 332.969696969697,
      "grad_norm": 0.3455847203731537,
      "learning_rate": 1.338e-05,
      "loss": 0.3221,
      "step": 3662
    },
    {
      "epoch": 333.0,
      "grad_norm": 0.4463295638561249,
      "learning_rate": 1.3370000000000002e-05,
      "loss": 0.3094,
      "step": 3663
    },
    {
      "epoch": 333.0969696969697,
      "grad_norm": 0.2687438130378723,
      "learning_rate": 1.336e-05,
      "loss": 0.3238,
      "step": 3664
    },
    {
      "epoch": 333.1939393939394,
      "grad_norm": 0.27776917815208435,
      "learning_rate": 1.3350000000000001e-05,
      "loss": 0.344,
      "step": 3665
    },
    {
      "epoch": 333.2909090909091,
      "grad_norm": 0.27730217576026917,
      "learning_rate": 1.334e-05,
      "loss": 0.3635,
      "step": 3666
    },
    {
      "epoch": 333.3878787878788,
      "grad_norm": 0.3098108768463135,
      "learning_rate": 1.3330000000000001e-05,
      "loss": 0.3259,
      "step": 3667
    },
    {
      "epoch": 333.4848484848485,
      "grad_norm": 0.2877723276615143,
      "learning_rate": 1.3320000000000001e-05,
      "loss": 0.3534,
      "step": 3668
    },
    {
      "epoch": 333.58181818181816,
      "grad_norm": 0.30429133772850037,
      "learning_rate": 1.331e-05,
      "loss": 0.3345,
      "step": 3669
    },
    {
      "epoch": 333.6787878787879,
      "grad_norm": 0.4659949839115143,
      "learning_rate": 1.3300000000000001e-05,
      "loss": 0.3406,
      "step": 3670
    },
    {
      "epoch": 333.6787878787879,
      "eval_loss": 0.42636099457740784,
      "eval_runtime": 2.1229,
      "eval_samples_per_second": 25.908,
      "eval_steps_per_second": 3.297,
      "step": 3670
    },
    {
      "epoch": 333.77575757575755,
      "grad_norm": 0.28408676385879517,
      "learning_rate": 1.329e-05,
      "loss": 0.3153,
      "step": 3671
    },
    {
      "epoch": 333.8727272727273,
      "grad_norm": 0.3500078320503235,
      "learning_rate": 1.3280000000000002e-05,
      "loss": 0.3392,
      "step": 3672
    },
    {
      "epoch": 333.969696969697,
      "grad_norm": 0.31569191813468933,
      "learning_rate": 1.3270000000000002e-05,
      "loss": 0.313,
      "step": 3673
    },
    {
      "epoch": 334.0,
      "grad_norm": 0.4757641553878784,
      "learning_rate": 1.326e-05,
      "loss": 0.3553,
      "step": 3674
    },
    {
      "epoch": 334.0969696969697,
      "grad_norm": 0.261425256729126,
      "learning_rate": 1.3250000000000002e-05,
      "loss": 0.3573,
      "step": 3675
    },
    {
      "epoch": 334.1939393939394,
      "grad_norm": 0.3167806565761566,
      "learning_rate": 1.324e-05,
      "loss": 0.3145,
      "step": 3676
    },
    {
      "epoch": 334.2909090909091,
      "grad_norm": 0.2758427858352661,
      "learning_rate": 1.323e-05,
      "loss": 0.3121,
      "step": 3677
    },
    {
      "epoch": 334.3878787878788,
      "grad_norm": 0.32896846532821655,
      "learning_rate": 1.3220000000000002e-05,
      "loss": 0.3737,
      "step": 3678
    },
    {
      "epoch": 334.4848484848485,
      "grad_norm": 0.2690833806991577,
      "learning_rate": 1.321e-05,
      "loss": 0.3399,
      "step": 3679
    },
    {
      "epoch": 334.58181818181816,
      "grad_norm": 0.34167739748954773,
      "learning_rate": 1.32e-05,
      "loss": 0.3438,
      "step": 3680
    },
    {
      "epoch": 334.58181818181816,
      "eval_loss": 0.42669355869293213,
      "eval_runtime": 2.1225,
      "eval_samples_per_second": 25.913,
      "eval_steps_per_second": 3.298,
      "step": 3680
    },
    {
      "epoch": 334.6787878787879,
      "grad_norm": 0.2956273555755615,
      "learning_rate": 1.3189999999999999e-05,
      "loss": 0.3606,
      "step": 3681
    },
    {
      "epoch": 334.77575757575755,
      "grad_norm": 0.2629927694797516,
      "learning_rate": 1.3180000000000001e-05,
      "loss": 0.3291,
      "step": 3682
    },
    {
      "epoch": 334.8727272727273,
      "grad_norm": 0.2564525008201599,
      "learning_rate": 1.3170000000000001e-05,
      "loss": 0.2946,
      "step": 3683
    },
    {
      "epoch": 334.969696969697,
      "grad_norm": 0.30224695801734924,
      "learning_rate": 1.316e-05,
      "loss": 0.3371,
      "step": 3684
    },
    {
      "epoch": 335.0,
      "grad_norm": 0.430014431476593,
      "learning_rate": 1.3150000000000001e-05,
      "loss": 0.3143,
      "step": 3685
    },
    {
      "epoch": 335.0969696969697,
      "grad_norm": 0.2904576063156128,
      "learning_rate": 1.314e-05,
      "loss": 0.3467,
      "step": 3686
    },
    {
      "epoch": 335.1939393939394,
      "grad_norm": 0.286720335483551,
      "learning_rate": 1.3130000000000001e-05,
      "loss": 0.3332,
      "step": 3687
    },
    {
      "epoch": 335.2909090909091,
      "grad_norm": 0.36974719166755676,
      "learning_rate": 1.3120000000000001e-05,
      "loss": 0.3734,
      "step": 3688
    },
    {
      "epoch": 335.3878787878788,
      "grad_norm": 0.2774631381034851,
      "learning_rate": 1.311e-05,
      "loss": 0.3064,
      "step": 3689
    },
    {
      "epoch": 335.4848484848485,
      "grad_norm": 0.30980628728866577,
      "learning_rate": 1.3100000000000002e-05,
      "loss": 0.3694,
      "step": 3690
    },
    {
      "epoch": 335.4848484848485,
      "eval_loss": 0.4261588752269745,
      "eval_runtime": 2.1539,
      "eval_samples_per_second": 25.535,
      "eval_steps_per_second": 3.25,
      "step": 3690
    },
    {
      "epoch": 335.58181818181816,
      "grad_norm": 0.2886945605278015,
      "learning_rate": 1.309e-05,
      "loss": 0.3054,
      "step": 3691
    },
    {
      "epoch": 335.6787878787879,
      "grad_norm": 0.285832017660141,
      "learning_rate": 1.308e-05,
      "loss": 0.3251,
      "step": 3692
    },
    {
      "epoch": 335.77575757575755,
      "grad_norm": 0.2936649024486542,
      "learning_rate": 1.3070000000000002e-05,
      "loss": 0.3398,
      "step": 3693
    },
    {
      "epoch": 335.8727272727273,
      "grad_norm": 0.2469436377286911,
      "learning_rate": 1.306e-05,
      "loss": 0.3497,
      "step": 3694
    },
    {
      "epoch": 335.969696969697,
      "grad_norm": 0.2709849178791046,
      "learning_rate": 1.305e-05,
      "loss": 0.3155,
      "step": 3695
    },
    {
      "epoch": 336.0,
      "grad_norm": 0.5284827947616577,
      "learning_rate": 1.3039999999999999e-05,
      "loss": 0.3105,
      "step": 3696
    },
    {
      "epoch": 336.0969696969697,
      "grad_norm": 0.26851266622543335,
      "learning_rate": 1.303e-05,
      "loss": 0.3332,
      "step": 3697
    },
    {
      "epoch": 336.1939393939394,
      "grad_norm": 0.306275337934494,
      "learning_rate": 1.3020000000000002e-05,
      "loss": 0.3403,
      "step": 3698
    },
    {
      "epoch": 336.2909090909091,
      "grad_norm": 0.29274263978004456,
      "learning_rate": 1.301e-05,
      "loss": 0.3411,
      "step": 3699
    },
    {
      "epoch": 336.3878787878788,
      "grad_norm": 0.2938704788684845,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.3286,
      "step": 3700
    },
    {
      "epoch": 336.3878787878788,
      "eval_loss": 0.4266641438007355,
      "eval_runtime": 2.1234,
      "eval_samples_per_second": 25.902,
      "eval_steps_per_second": 3.297,
      "step": 3700
    },
    {
      "epoch": 336.4848484848485,
      "grad_norm": 0.3236421048641205,
      "learning_rate": 1.299e-05,
      "loss": 0.3311,
      "step": 3701
    },
    {
      "epoch": 336.58181818181816,
      "grad_norm": 0.3449302911758423,
      "learning_rate": 1.2980000000000001e-05,
      "loss": 0.3351,
      "step": 3702
    },
    {
      "epoch": 336.6787878787879,
      "grad_norm": 0.2878248989582062,
      "learning_rate": 1.2970000000000001e-05,
      "loss": 0.3222,
      "step": 3703
    },
    {
      "epoch": 336.77575757575755,
      "grad_norm": 0.27355077862739563,
      "learning_rate": 1.296e-05,
      "loss": 0.3229,
      "step": 3704
    },
    {
      "epoch": 336.8727272727273,
      "grad_norm": 0.285319983959198,
      "learning_rate": 1.2950000000000001e-05,
      "loss": 0.357,
      "step": 3705
    },
    {
      "epoch": 336.969696969697,
      "grad_norm": 0.2839474678039551,
      "learning_rate": 1.294e-05,
      "loss": 0.3399,
      "step": 3706
    },
    {
      "epoch": 337.0,
      "grad_norm": 0.42722317576408386,
      "learning_rate": 1.293e-05,
      "loss": 0.3444,
      "step": 3707
    },
    {
      "epoch": 337.0969696969697,
      "grad_norm": 0.26353755593299866,
      "learning_rate": 1.2920000000000002e-05,
      "loss": 0.3869,
      "step": 3708
    },
    {
      "epoch": 337.1939393939394,
      "grad_norm": 0.26299166679382324,
      "learning_rate": 1.291e-05,
      "loss": 0.3225,
      "step": 3709
    },
    {
      "epoch": 337.2909090909091,
      "grad_norm": 0.3076457679271698,
      "learning_rate": 1.29e-05,
      "loss": 0.3096,
      "step": 3710
    },
    {
      "epoch": 337.2909090909091,
      "eval_loss": 0.4259534180164337,
      "eval_runtime": 2.1593,
      "eval_samples_per_second": 25.471,
      "eval_steps_per_second": 3.242,
      "step": 3710
    },
    {
      "epoch": 337.3878787878788,
      "grad_norm": 0.3430866301059723,
      "learning_rate": 1.2889999999999999e-05,
      "loss": 0.3567,
      "step": 3711
    },
    {
      "epoch": 337.4848484848485,
      "grad_norm": 0.35578009486198425,
      "learning_rate": 1.288e-05,
      "loss": 0.2813,
      "step": 3712
    },
    {
      "epoch": 337.58181818181816,
      "grad_norm": 0.2620140016078949,
      "learning_rate": 1.2870000000000002e-05,
      "loss": 0.3425,
      "step": 3713
    },
    {
      "epoch": 337.6787878787879,
      "grad_norm": 0.32215750217437744,
      "learning_rate": 1.286e-05,
      "loss": 0.326,
      "step": 3714
    },
    {
      "epoch": 337.77575757575755,
      "grad_norm": 0.3033801019191742,
      "learning_rate": 1.285e-05,
      "loss": 0.3361,
      "step": 3715
    },
    {
      "epoch": 337.8727272727273,
      "grad_norm": 0.32087045907974243,
      "learning_rate": 1.2839999999999999e-05,
      "loss": 0.3533,
      "step": 3716
    },
    {
      "epoch": 337.969696969697,
      "grad_norm": 0.3305118680000305,
      "learning_rate": 1.283e-05,
      "loss": 0.3324,
      "step": 3717
    },
    {
      "epoch": 338.0,
      "grad_norm": 0.5343279838562012,
      "learning_rate": 1.2820000000000001e-05,
      "loss": 0.3582,
      "step": 3718
    },
    {
      "epoch": 338.0969696969697,
      "grad_norm": 0.3098539113998413,
      "learning_rate": 1.281e-05,
      "loss": 0.3337,
      "step": 3719
    },
    {
      "epoch": 338.1939393939394,
      "grad_norm": 0.3534613847732544,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.3418,
      "step": 3720
    },
    {
      "epoch": 338.1939393939394,
      "eval_loss": 0.42633816599845886,
      "eval_runtime": 2.1317,
      "eval_samples_per_second": 25.801,
      "eval_steps_per_second": 3.284,
      "step": 3720
    },
    {
      "epoch": 338.2909090909091,
      "grad_norm": 0.27533775568008423,
      "learning_rate": 1.2790000000000001e-05,
      "loss": 0.3369,
      "step": 3721
    },
    {
      "epoch": 338.3878787878788,
      "grad_norm": 0.27337315678596497,
      "learning_rate": 1.278e-05,
      "loss": 0.3259,
      "step": 3722
    },
    {
      "epoch": 338.4848484848485,
      "grad_norm": 0.30268651247024536,
      "learning_rate": 1.2770000000000001e-05,
      "loss": 0.3497,
      "step": 3723
    },
    {
      "epoch": 338.58181818181816,
      "grad_norm": 0.27092444896698,
      "learning_rate": 1.276e-05,
      "loss": 0.3558,
      "step": 3724
    },
    {
      "epoch": 338.6787878787879,
      "grad_norm": 0.2851577401161194,
      "learning_rate": 1.2750000000000002e-05,
      "loss": 0.3132,
      "step": 3725
    },
    {
      "epoch": 338.77575757575755,
      "grad_norm": 0.3387395441532135,
      "learning_rate": 1.2740000000000002e-05,
      "loss": 0.3087,
      "step": 3726
    },
    {
      "epoch": 338.8727272727273,
      "grad_norm": 0.27414950728416443,
      "learning_rate": 1.273e-05,
      "loss": 0.3613,
      "step": 3727
    },
    {
      "epoch": 338.969696969697,
      "grad_norm": 0.2948184609413147,
      "learning_rate": 1.2720000000000002e-05,
      "loss": 0.3251,
      "step": 3728
    },
    {
      "epoch": 339.0,
      "grad_norm": 0.544585645198822,
      "learning_rate": 1.271e-05,
      "loss": 0.3472,
      "step": 3729
    },
    {
      "epoch": 339.0969696969697,
      "grad_norm": 0.2687508463859558,
      "learning_rate": 1.27e-05,
      "loss": 0.3414,
      "step": 3730
    },
    {
      "epoch": 339.0969696969697,
      "eval_loss": 0.4261287748813629,
      "eval_runtime": 2.1147,
      "eval_samples_per_second": 26.008,
      "eval_steps_per_second": 3.31,
      "step": 3730
    },
    {
      "epoch": 339.1939393939394,
      "grad_norm": 0.3146977722644806,
      "learning_rate": 1.2690000000000002e-05,
      "loss": 0.3515,
      "step": 3731
    },
    {
      "epoch": 339.2909090909091,
      "grad_norm": 0.2693532109260559,
      "learning_rate": 1.268e-05,
      "loss": 0.3107,
      "step": 3732
    },
    {
      "epoch": 339.3878787878788,
      "grad_norm": 0.30518874526023865,
      "learning_rate": 1.267e-05,
      "loss": 0.3177,
      "step": 3733
    },
    {
      "epoch": 339.4848484848485,
      "grad_norm": 0.2727731168270111,
      "learning_rate": 1.2659999999999999e-05,
      "loss": 0.3353,
      "step": 3734
    },
    {
      "epoch": 339.58181818181816,
      "grad_norm": 0.31078311800956726,
      "learning_rate": 1.2650000000000001e-05,
      "loss": 0.3035,
      "step": 3735
    },
    {
      "epoch": 339.6787878787879,
      "grad_norm": 0.2725074589252472,
      "learning_rate": 1.2640000000000003e-05,
      "loss": 0.3455,
      "step": 3736
    },
    {
      "epoch": 339.77575757575755,
      "grad_norm": 0.3046586811542511,
      "learning_rate": 1.263e-05,
      "loss": 0.39,
      "step": 3737
    },
    {
      "epoch": 339.8727272727273,
      "grad_norm": 0.3013319671154022,
      "learning_rate": 1.2620000000000001e-05,
      "loss": 0.3466,
      "step": 3738
    },
    {
      "epoch": 339.969696969697,
      "grad_norm": 0.32089748978614807,
      "learning_rate": 1.261e-05,
      "loss": 0.322,
      "step": 3739
    },
    {
      "epoch": 340.0,
      "grad_norm": 0.5079250335693359,
      "learning_rate": 1.2600000000000001e-05,
      "loss": 0.2924,
      "step": 3740
    },
    {
      "epoch": 340.0,
      "eval_loss": 0.42617207765579224,
      "eval_runtime": 2.1379,
      "eval_samples_per_second": 25.726,
      "eval_steps_per_second": 3.274,
      "step": 3740
    },
    {
      "epoch": 340.0969696969697,
      "grad_norm": 0.25629541277885437,
      "learning_rate": 1.2590000000000001e-05,
      "loss": 0.3203,
      "step": 3741
    },
    {
      "epoch": 340.1939393939394,
      "grad_norm": 0.25945302844047546,
      "learning_rate": 1.258e-05,
      "loss": 0.3363,
      "step": 3742
    },
    {
      "epoch": 340.2909090909091,
      "grad_norm": 0.28978466987609863,
      "learning_rate": 1.2570000000000002e-05,
      "loss": 0.3216,
      "step": 3743
    },
    {
      "epoch": 340.3878787878788,
      "grad_norm": 0.2984698712825775,
      "learning_rate": 1.256e-05,
      "loss": 0.3478,
      "step": 3744
    },
    {
      "epoch": 340.4848484848485,
      "grad_norm": 0.3027973473072052,
      "learning_rate": 1.255e-05,
      "loss": 0.3527,
      "step": 3745
    },
    {
      "epoch": 340.58181818181816,
      "grad_norm": 0.2580370306968689,
      "learning_rate": 1.2540000000000002e-05,
      "loss": 0.3283,
      "step": 3746
    },
    {
      "epoch": 340.6787878787879,
      "grad_norm": 0.3044896721839905,
      "learning_rate": 1.253e-05,
      "loss": 0.3256,
      "step": 3747
    },
    {
      "epoch": 340.77575757575755,
      "grad_norm": 0.28928688168525696,
      "learning_rate": 1.252e-05,
      "loss": 0.3504,
      "step": 3748
    },
    {
      "epoch": 340.8727272727273,
      "grad_norm": 0.36256060004234314,
      "learning_rate": 1.2509999999999999e-05,
      "loss": 0.3382,
      "step": 3749
    },
    {
      "epoch": 340.969696969697,
      "grad_norm": 0.3101032078266144,
      "learning_rate": 1.25e-05,
      "loss": 0.3296,
      "step": 3750
    },
    {
      "epoch": 340.969696969697,
      "eval_loss": 0.42549657821655273,
      "eval_runtime": 2.1573,
      "eval_samples_per_second": 25.495,
      "eval_steps_per_second": 3.245,
      "step": 3750
    },
    {
      "epoch": 341.0,
      "grad_norm": 0.590671718120575,
      "learning_rate": 1.249e-05,
      "loss": 0.3182,
      "step": 3751
    },
    {
      "epoch": 341.0969696969697,
      "grad_norm": 0.2854500114917755,
      "learning_rate": 1.248e-05,
      "loss": 0.3615,
      "step": 3752
    },
    {
      "epoch": 341.1939393939394,
      "grad_norm": 0.35217300057411194,
      "learning_rate": 1.2470000000000001e-05,
      "loss": 0.3348,
      "step": 3753
    },
    {
      "epoch": 341.2909090909091,
      "grad_norm": 0.2804906666278839,
      "learning_rate": 1.2460000000000001e-05,
      "loss": 0.3354,
      "step": 3754
    },
    {
      "epoch": 341.3878787878788,
      "grad_norm": 0.25322121381759644,
      "learning_rate": 1.2450000000000001e-05,
      "loss": 0.36,
      "step": 3755
    },
    {
      "epoch": 341.4848484848485,
      "grad_norm": 0.31866592168807983,
      "learning_rate": 1.244e-05,
      "loss": 0.3086,
      "step": 3756
    },
    {
      "epoch": 341.58181818181816,
      "grad_norm": 0.31964659690856934,
      "learning_rate": 1.243e-05,
      "loss": 0.3385,
      "step": 3757
    },
    {
      "epoch": 341.6787878787879,
      "grad_norm": 0.2785675823688507,
      "learning_rate": 1.2420000000000001e-05,
      "loss": 0.3326,
      "step": 3758
    },
    {
      "epoch": 341.77575757575755,
      "grad_norm": 0.2687302231788635,
      "learning_rate": 1.2410000000000001e-05,
      "loss": 0.303,
      "step": 3759
    },
    {
      "epoch": 341.8727272727273,
      "grad_norm": 0.2588405907154083,
      "learning_rate": 1.24e-05,
      "loss": 0.3213,
      "step": 3760
    },
    {
      "epoch": 341.8727272727273,
      "eval_loss": 0.42594873905181885,
      "eval_runtime": 2.1086,
      "eval_samples_per_second": 26.084,
      "eval_steps_per_second": 3.32,
      "step": 3760
    },
    {
      "epoch": 341.969696969697,
      "grad_norm": 0.3199463486671448,
      "learning_rate": 1.239e-05,
      "loss": 0.3517,
      "step": 3761
    },
    {
      "epoch": 342.0,
      "grad_norm": 0.4935012757778168,
      "learning_rate": 1.238e-05,
      "loss": 0.3284,
      "step": 3762
    },
    {
      "epoch": 342.0969696969697,
      "grad_norm": 0.36274877190589905,
      "learning_rate": 1.2370000000000002e-05,
      "loss": 0.3371,
      "step": 3763
    },
    {
      "epoch": 342.1939393939394,
      "grad_norm": 0.28404876589775085,
      "learning_rate": 1.236e-05,
      "loss": 0.3289,
      "step": 3764
    },
    {
      "epoch": 342.2909090909091,
      "grad_norm": 0.27230966091156006,
      "learning_rate": 1.235e-05,
      "loss": 0.3121,
      "step": 3765
    },
    {
      "epoch": 342.3878787878788,
      "grad_norm": 0.2236277163028717,
      "learning_rate": 1.234e-05,
      "loss": 0.3464,
      "step": 3766
    },
    {
      "epoch": 342.4848484848485,
      "grad_norm": 0.2575564980506897,
      "learning_rate": 1.233e-05,
      "loss": 0.319,
      "step": 3767
    },
    {
      "epoch": 342.58181818181816,
      "grad_norm": 0.3601349890232086,
      "learning_rate": 1.232e-05,
      "loss": 0.357,
      "step": 3768
    },
    {
      "epoch": 342.6787878787879,
      "grad_norm": 0.3346935510635376,
      "learning_rate": 1.231e-05,
      "loss": 0.3365,
      "step": 3769
    },
    {
      "epoch": 342.77575757575755,
      "grad_norm": 0.3269232213497162,
      "learning_rate": 1.23e-05,
      "loss": 0.3441,
      "step": 3770
    },
    {
      "epoch": 342.77575757575755,
      "eval_loss": 0.42581814527511597,
      "eval_runtime": 2.1304,
      "eval_samples_per_second": 25.816,
      "eval_steps_per_second": 3.286,
      "step": 3770
    },
    {
      "epoch": 342.8727272727273,
      "grad_norm": 0.2760692536830902,
      "learning_rate": 1.2290000000000001e-05,
      "loss": 0.3279,
      "step": 3771
    },
    {
      "epoch": 342.969696969697,
      "grad_norm": 0.3023752272129059,
      "learning_rate": 1.2280000000000001e-05,
      "loss": 0.3521,
      "step": 3772
    },
    {
      "epoch": 343.0,
      "grad_norm": 0.5031427145004272,
      "learning_rate": 1.2270000000000001e-05,
      "loss": 0.2737,
      "step": 3773
    },
    {
      "epoch": 343.0969696969697,
      "grad_norm": 0.2600629925727844,
      "learning_rate": 1.2260000000000001e-05,
      "loss": 0.334,
      "step": 3774
    },
    {
      "epoch": 343.1939393939394,
      "grad_norm": 0.2708774209022522,
      "learning_rate": 1.225e-05,
      "loss": 0.3288,
      "step": 3775
    },
    {
      "epoch": 343.2909090909091,
      "grad_norm": 0.30316439270973206,
      "learning_rate": 1.224e-05,
      "loss": 0.3327,
      "step": 3776
    },
    {
      "epoch": 343.3878787878788,
      "grad_norm": 0.2830888330936432,
      "learning_rate": 1.2230000000000001e-05,
      "loss": 0.3221,
      "step": 3777
    },
    {
      "epoch": 343.4848484848485,
      "grad_norm": 0.33427178859710693,
      "learning_rate": 1.2220000000000002e-05,
      "loss": 0.3008,
      "step": 3778
    },
    {
      "epoch": 343.58181818181816,
      "grad_norm": 0.3453505039215088,
      "learning_rate": 1.221e-05,
      "loss": 0.3728,
      "step": 3779
    },
    {
      "epoch": 343.6787878787879,
      "grad_norm": 0.2651301920413971,
      "learning_rate": 1.22e-05,
      "loss": 0.3563,
      "step": 3780
    },
    {
      "epoch": 343.6787878787879,
      "eval_loss": 0.42604801058769226,
      "eval_runtime": 2.1237,
      "eval_samples_per_second": 25.899,
      "eval_steps_per_second": 3.296,
      "step": 3780
    },
    {
      "epoch": 343.77575757575755,
      "grad_norm": 0.28830796480178833,
      "learning_rate": 1.219e-05,
      "loss": 0.343,
      "step": 3781
    },
    {
      "epoch": 343.8727272727273,
      "grad_norm": 0.3189815878868103,
      "learning_rate": 1.2180000000000002e-05,
      "loss": 0.3287,
      "step": 3782
    },
    {
      "epoch": 343.969696969697,
      "grad_norm": 0.3367035984992981,
      "learning_rate": 1.217e-05,
      "loss": 0.3363,
      "step": 3783
    },
    {
      "epoch": 344.0,
      "grad_norm": 0.5709712505340576,
      "learning_rate": 1.216e-05,
      "loss": 0.2973,
      "step": 3784
    },
    {
      "epoch": 344.0969696969697,
      "grad_norm": 0.30823367834091187,
      "learning_rate": 1.215e-05,
      "loss": 0.305,
      "step": 3785
    },
    {
      "epoch": 344.1939393939394,
      "grad_norm": 0.32099658250808716,
      "learning_rate": 1.214e-05,
      "loss": 0.3073,
      "step": 3786
    },
    {
      "epoch": 344.2909090909091,
      "grad_norm": 0.25584954023361206,
      "learning_rate": 1.213e-05,
      "loss": 0.3347,
      "step": 3787
    },
    {
      "epoch": 344.3878787878788,
      "grad_norm": 0.3256348669528961,
      "learning_rate": 1.2120000000000001e-05,
      "loss": 0.3043,
      "step": 3788
    },
    {
      "epoch": 344.4848484848485,
      "grad_norm": 0.2968614399433136,
      "learning_rate": 1.2110000000000001e-05,
      "loss": 0.3308,
      "step": 3789
    },
    {
      "epoch": 344.58181818181816,
      "grad_norm": 0.28445330262184143,
      "learning_rate": 1.2100000000000001e-05,
      "loss": 0.351,
      "step": 3790
    },
    {
      "epoch": 344.58181818181816,
      "eval_loss": 0.4258812367916107,
      "eval_runtime": 2.1339,
      "eval_samples_per_second": 25.775,
      "eval_steps_per_second": 3.28,
      "step": 3790
    },
    {
      "epoch": 344.6787878787879,
      "grad_norm": 0.3036136329174042,
      "learning_rate": 1.209e-05,
      "loss": 0.3405,
      "step": 3791
    },
    {
      "epoch": 344.77575757575755,
      "grad_norm": 0.28134721517562866,
      "learning_rate": 1.2080000000000001e-05,
      "loss": 0.3625,
      "step": 3792
    },
    {
      "epoch": 344.8727272727273,
      "grad_norm": 0.27972412109375,
      "learning_rate": 1.2070000000000001e-05,
      "loss": 0.3769,
      "step": 3793
    },
    {
      "epoch": 344.969696969697,
      "grad_norm": 0.32446789741516113,
      "learning_rate": 1.206e-05,
      "loss": 0.3437,
      "step": 3794
    },
    {
      "epoch": 345.0,
      "grad_norm": 0.47319990396499634,
      "learning_rate": 1.205e-05,
      "loss": 0.2866,
      "step": 3795
    },
    {
      "epoch": 345.0969696969697,
      "grad_norm": 0.2875915765762329,
      "learning_rate": 1.204e-05,
      "loss": 0.3453,
      "step": 3796
    },
    {
      "epoch": 345.1939393939394,
      "grad_norm": 0.2987116575241089,
      "learning_rate": 1.2030000000000002e-05,
      "loss": 0.3257,
      "step": 3797
    },
    {
      "epoch": 345.2909090909091,
      "grad_norm": 0.31415995955467224,
      "learning_rate": 1.202e-05,
      "loss": 0.3464,
      "step": 3798
    },
    {
      "epoch": 345.3878787878788,
      "grad_norm": 0.3293113708496094,
      "learning_rate": 1.201e-05,
      "loss": 0.3064,
      "step": 3799
    },
    {
      "epoch": 345.4848484848485,
      "grad_norm": 0.2928531765937805,
      "learning_rate": 1.2e-05,
      "loss": 0.3113,
      "step": 3800
    },
    {
      "epoch": 345.4848484848485,
      "eval_loss": 0.4257052540779114,
      "eval_runtime": 2.116,
      "eval_samples_per_second": 25.992,
      "eval_steps_per_second": 3.308,
      "step": 3800
    },
    {
      "epoch": 345.58181818181816,
      "grad_norm": 0.29395541548728943,
      "learning_rate": 1.199e-05,
      "loss": 0.3525,
      "step": 3801
    },
    {
      "epoch": 345.6787878787879,
      "grad_norm": 0.29284462332725525,
      "learning_rate": 1.198e-05,
      "loss": 0.3008,
      "step": 3802
    },
    {
      "epoch": 345.77575757575755,
      "grad_norm": 0.2965705692768097,
      "learning_rate": 1.197e-05,
      "loss": 0.3647,
      "step": 3803
    },
    {
      "epoch": 345.8727272727273,
      "grad_norm": 0.30063632130622864,
      "learning_rate": 1.196e-05,
      "loss": 0.3421,
      "step": 3804
    },
    {
      "epoch": 345.969696969697,
      "grad_norm": 0.2827775478363037,
      "learning_rate": 1.195e-05,
      "loss": 0.3481,
      "step": 3805
    },
    {
      "epoch": 346.0,
      "grad_norm": 0.7182817459106445,
      "learning_rate": 1.1940000000000001e-05,
      "loss": 0.3309,
      "step": 3806
    },
    {
      "epoch": 346.0969696969697,
      "grad_norm": 0.30240872502326965,
      "learning_rate": 1.1930000000000001e-05,
      "loss": 0.3205,
      "step": 3807
    },
    {
      "epoch": 346.1939393939394,
      "grad_norm": 0.2711304724216461,
      "learning_rate": 1.1920000000000001e-05,
      "loss": 0.3334,
      "step": 3808
    },
    {
      "epoch": 346.2909090909091,
      "grad_norm": 0.27359533309936523,
      "learning_rate": 1.1910000000000001e-05,
      "loss": 0.3547,
      "step": 3809
    },
    {
      "epoch": 346.3878787878788,
      "grad_norm": 0.47756561636924744,
      "learning_rate": 1.19e-05,
      "loss": 0.3202,
      "step": 3810
    },
    {
      "epoch": 346.3878787878788,
      "eval_loss": 0.42556026577949524,
      "eval_runtime": 2.1237,
      "eval_samples_per_second": 25.898,
      "eval_steps_per_second": 3.296,
      "step": 3810
    },
    {
      "epoch": 346.4848484848485,
      "grad_norm": 0.3103063106536865,
      "learning_rate": 1.1890000000000001e-05,
      "loss": 0.3288,
      "step": 3811
    },
    {
      "epoch": 346.58181818181816,
      "grad_norm": 0.3046967387199402,
      "learning_rate": 1.1880000000000001e-05,
      "loss": 0.3435,
      "step": 3812
    },
    {
      "epoch": 346.6787878787879,
      "grad_norm": 0.33921241760253906,
      "learning_rate": 1.187e-05,
      "loss": 0.3244,
      "step": 3813
    },
    {
      "epoch": 346.77575757575755,
      "grad_norm": 0.2951901853084564,
      "learning_rate": 1.186e-05,
      "loss": 0.3583,
      "step": 3814
    },
    {
      "epoch": 346.8727272727273,
      "grad_norm": 0.3240387737751007,
      "learning_rate": 1.185e-05,
      "loss": 0.3247,
      "step": 3815
    },
    {
      "epoch": 346.969696969697,
      "grad_norm": 0.2763834595680237,
      "learning_rate": 1.1840000000000002e-05,
      "loss": 0.3465,
      "step": 3816
    },
    {
      "epoch": 347.0,
      "grad_norm": 0.5004387497901917,
      "learning_rate": 1.183e-05,
      "loss": 0.282,
      "step": 3817
    },
    {
      "epoch": 347.0969696969697,
      "grad_norm": 0.3101780116558075,
      "learning_rate": 1.182e-05,
      "loss": 0.3448,
      "step": 3818
    },
    {
      "epoch": 347.1939393939394,
      "grad_norm": 0.2685920000076294,
      "learning_rate": 1.181e-05,
      "loss": 0.3363,
      "step": 3819
    },
    {
      "epoch": 347.2909090909091,
      "grad_norm": 0.27615267038345337,
      "learning_rate": 1.18e-05,
      "loss": 0.328,
      "step": 3820
    },
    {
      "epoch": 347.2909090909091,
      "eval_loss": 0.42572295665740967,
      "eval_runtime": 2.12,
      "eval_samples_per_second": 25.944,
      "eval_steps_per_second": 3.302,
      "step": 3820
    },
    {
      "epoch": 347.3878787878788,
      "grad_norm": 0.2806648015975952,
      "learning_rate": 1.179e-05,
      "loss": 0.3218,
      "step": 3821
    },
    {
      "epoch": 347.4848484848485,
      "grad_norm": 0.2721249461174011,
      "learning_rate": 1.178e-05,
      "loss": 0.3465,
      "step": 3822
    },
    {
      "epoch": 347.58181818181816,
      "grad_norm": 0.2903819978237152,
      "learning_rate": 1.177e-05,
      "loss": 0.3075,
      "step": 3823
    },
    {
      "epoch": 347.6787878787879,
      "grad_norm": 0.2980758845806122,
      "learning_rate": 1.1760000000000001e-05,
      "loss": 0.3462,
      "step": 3824
    },
    {
      "epoch": 347.77575757575755,
      "grad_norm": 0.3511853516101837,
      "learning_rate": 1.175e-05,
      "loss": 0.3159,
      "step": 3825
    },
    {
      "epoch": 347.8727272727273,
      "grad_norm": 0.276050865650177,
      "learning_rate": 1.1740000000000001e-05,
      "loss": 0.3209,
      "step": 3826
    },
    {
      "epoch": 347.969696969697,
      "grad_norm": 0.3628363609313965,
      "learning_rate": 1.1730000000000001e-05,
      "loss": 0.3595,
      "step": 3827
    },
    {
      "epoch": 348.0,
      "grad_norm": 0.6746332049369812,
      "learning_rate": 1.172e-05,
      "loss": 0.3602,
      "step": 3828
    },
    {
      "epoch": 348.0969696969697,
      "grad_norm": 0.2941165268421173,
      "learning_rate": 1.171e-05,
      "loss": 0.3088,
      "step": 3829
    },
    {
      "epoch": 348.1939393939394,
      "grad_norm": 0.28681251406669617,
      "learning_rate": 1.1700000000000001e-05,
      "loss": 0.3529,
      "step": 3830
    },
    {
      "epoch": 348.1939393939394,
      "eval_loss": 0.42591407895088196,
      "eval_runtime": 2.1257,
      "eval_samples_per_second": 25.873,
      "eval_steps_per_second": 3.293,
      "step": 3830
    },
    {
      "epoch": 348.2909090909091,
      "grad_norm": 0.30104222893714905,
      "learning_rate": 1.1690000000000002e-05,
      "loss": 0.3544,
      "step": 3831
    },
    {
      "epoch": 348.3878787878788,
      "grad_norm": 0.27228203415870667,
      "learning_rate": 1.168e-05,
      "loss": 0.3113,
      "step": 3832
    },
    {
      "epoch": 348.4848484848485,
      "grad_norm": 0.2858501374721527,
      "learning_rate": 1.167e-05,
      "loss": 0.3268,
      "step": 3833
    },
    {
      "epoch": 348.58181818181816,
      "grad_norm": 0.2979985475540161,
      "learning_rate": 1.166e-05,
      "loss": 0.3642,
      "step": 3834
    },
    {
      "epoch": 348.6787878787879,
      "grad_norm": 0.29480111598968506,
      "learning_rate": 1.1650000000000002e-05,
      "loss": 0.3566,
      "step": 3835
    },
    {
      "epoch": 348.77575757575755,
      "grad_norm": 0.3030244708061218,
      "learning_rate": 1.164e-05,
      "loss": 0.3377,
      "step": 3836
    },
    {
      "epoch": 348.8727272727273,
      "grad_norm": 0.32541602849960327,
      "learning_rate": 1.163e-05,
      "loss": 0.3218,
      "step": 3837
    },
    {
      "epoch": 348.969696969697,
      "grad_norm": 0.28534772992134094,
      "learning_rate": 1.162e-05,
      "loss": 0.3071,
      "step": 3838
    },
    {
      "epoch": 349.0,
      "grad_norm": 0.5974946022033691,
      "learning_rate": 1.161e-05,
      "loss": 0.3073,
      "step": 3839
    },
    {
      "epoch": 349.0969696969697,
      "grad_norm": 0.31280946731567383,
      "learning_rate": 1.16e-05,
      "loss": 0.3439,
      "step": 3840
    },
    {
      "epoch": 349.0969696969697,
      "eval_loss": 0.42570239305496216,
      "eval_runtime": 2.1285,
      "eval_samples_per_second": 25.839,
      "eval_steps_per_second": 3.289,
      "step": 3840
    },
    {
      "epoch": 349.1939393939394,
      "grad_norm": 0.29928094148635864,
      "learning_rate": 1.159e-05,
      "loss": 0.3078,
      "step": 3841
    },
    {
      "epoch": 349.2909090909091,
      "grad_norm": 0.35683488845825195,
      "learning_rate": 1.1580000000000001e-05,
      "loss": 0.3568,
      "step": 3842
    },
    {
      "epoch": 349.3878787878788,
      "grad_norm": 0.2896963953971863,
      "learning_rate": 1.1570000000000001e-05,
      "loss": 0.356,
      "step": 3843
    },
    {
      "epoch": 349.4848484848485,
      "grad_norm": 0.24926257133483887,
      "learning_rate": 1.156e-05,
      "loss": 0.3207,
      "step": 3844
    },
    {
      "epoch": 349.58181818181816,
      "grad_norm": 0.27918097376823425,
      "learning_rate": 1.1550000000000001e-05,
      "loss": 0.3037,
      "step": 3845
    },
    {
      "epoch": 349.6787878787879,
      "grad_norm": 0.30067959427833557,
      "learning_rate": 1.1540000000000001e-05,
      "loss": 0.3175,
      "step": 3846
    },
    {
      "epoch": 349.77575757575755,
      "grad_norm": 0.29443076252937317,
      "learning_rate": 1.153e-05,
      "loss": 0.305,
      "step": 3847
    },
    {
      "epoch": 349.8727272727273,
      "grad_norm": 0.3009408712387085,
      "learning_rate": 1.152e-05,
      "loss": 0.3323,
      "step": 3848
    },
    {
      "epoch": 349.969696969697,
      "grad_norm": 0.33346986770629883,
      "learning_rate": 1.151e-05,
      "loss": 0.365,
      "step": 3849
    },
    {
      "epoch": 350.0,
      "grad_norm": 0.5449870824813843,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 0.3946,
      "step": 3850
    },
    {
      "epoch": 350.0,
      "eval_loss": 0.4254421591758728,
      "eval_runtime": 2.1379,
      "eval_samples_per_second": 25.727,
      "eval_steps_per_second": 3.274,
      "step": 3850
    },
    {
      "epoch": 350.0969696969697,
      "grad_norm": 0.3212931454181671,
      "learning_rate": 1.149e-05,
      "loss": 0.3496,
      "step": 3851
    },
    {
      "epoch": 350.1939393939394,
      "grad_norm": 0.351112961769104,
      "learning_rate": 1.148e-05,
      "loss": 0.3197,
      "step": 3852
    },
    {
      "epoch": 350.2909090909091,
      "grad_norm": 0.29633641242980957,
      "learning_rate": 1.147e-05,
      "loss": 0.3099,
      "step": 3853
    },
    {
      "epoch": 350.3878787878788,
      "grad_norm": 0.31008249521255493,
      "learning_rate": 1.146e-05,
      "loss": 0.2962,
      "step": 3854
    },
    {
      "epoch": 350.4848484848485,
      "grad_norm": 0.31466251611709595,
      "learning_rate": 1.145e-05,
      "loss": 0.3429,
      "step": 3855
    },
    {
      "epoch": 350.58181818181816,
      "grad_norm": 0.27296024560928345,
      "learning_rate": 1.144e-05,
      "loss": 0.3601,
      "step": 3856
    },
    {
      "epoch": 350.6787878787879,
      "grad_norm": 0.2866796553134918,
      "learning_rate": 1.143e-05,
      "loss": 0.3148,
      "step": 3857
    },
    {
      "epoch": 350.77575757575755,
      "grad_norm": 0.3360688388347626,
      "learning_rate": 1.142e-05,
      "loss": 0.3429,
      "step": 3858
    },
    {
      "epoch": 350.8727272727273,
      "grad_norm": 0.2791575789451599,
      "learning_rate": 1.141e-05,
      "loss": 0.3234,
      "step": 3859
    },
    {
      "epoch": 350.969696969697,
      "grad_norm": 0.310596764087677,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 0.3686,
      "step": 3860
    },
    {
      "epoch": 350.969696969697,
      "eval_loss": 0.42575693130493164,
      "eval_runtime": 2.1423,
      "eval_samples_per_second": 25.673,
      "eval_steps_per_second": 3.267,
      "step": 3860
    },
    {
      "epoch": 351.0,
      "grad_norm": 0.42716941237449646,
      "learning_rate": 1.1390000000000001e-05,
      "loss": 0.3444,
      "step": 3861
    },
    {
      "epoch": 351.0969696969697,
      "grad_norm": 0.28955182433128357,
      "learning_rate": 1.1380000000000001e-05,
      "loss": 0.3347,
      "step": 3862
    },
    {
      "epoch": 351.1939393939394,
      "grad_norm": 0.27090415358543396,
      "learning_rate": 1.137e-05,
      "loss": 0.3043,
      "step": 3863
    },
    {
      "epoch": 351.2909090909091,
      "grad_norm": 0.2701130211353302,
      "learning_rate": 1.1360000000000001e-05,
      "loss": 0.3114,
      "step": 3864
    },
    {
      "epoch": 351.3878787878788,
      "grad_norm": 0.28966429829597473,
      "learning_rate": 1.1350000000000001e-05,
      "loss": 0.305,
      "step": 3865
    },
    {
      "epoch": 351.4848484848485,
      "grad_norm": 0.295930951833725,
      "learning_rate": 1.134e-05,
      "loss": 0.3698,
      "step": 3866
    },
    {
      "epoch": 351.58181818181816,
      "grad_norm": 0.25204113125801086,
      "learning_rate": 1.133e-05,
      "loss": 0.3581,
      "step": 3867
    },
    {
      "epoch": 351.6787878787879,
      "grad_norm": 0.2838861048221588,
      "learning_rate": 1.132e-05,
      "loss": 0.325,
      "step": 3868
    },
    {
      "epoch": 351.77575757575755,
      "grad_norm": 0.3151472210884094,
      "learning_rate": 1.1310000000000002e-05,
      "loss": 0.31,
      "step": 3869
    },
    {
      "epoch": 351.8727272727273,
      "grad_norm": 0.31971457600593567,
      "learning_rate": 1.13e-05,
      "loss": 0.3523,
      "step": 3870
    },
    {
      "epoch": 351.8727272727273,
      "eval_loss": 0.4253227114677429,
      "eval_runtime": 2.1596,
      "eval_samples_per_second": 25.467,
      "eval_steps_per_second": 3.241,
      "step": 3870
    },
    {
      "epoch": 351.969696969697,
      "grad_norm": 0.32695403695106506,
      "learning_rate": 1.129e-05,
      "loss": 0.3688,
      "step": 3871
    },
    {
      "epoch": 352.0,
      "grad_norm": 0.5059196352958679,
      "learning_rate": 1.128e-05,
      "loss": 0.2973,
      "step": 3872
    },
    {
      "epoch": 352.0969696969697,
      "grad_norm": 0.31515827775001526,
      "learning_rate": 1.127e-05,
      "loss": 0.3277,
      "step": 3873
    },
    {
      "epoch": 352.1939393939394,
      "grad_norm": 0.3012741208076477,
      "learning_rate": 1.126e-05,
      "loss": 0.3339,
      "step": 3874
    },
    {
      "epoch": 352.2909090909091,
      "grad_norm": 0.3248525857925415,
      "learning_rate": 1.125e-05,
      "loss": 0.3321,
      "step": 3875
    },
    {
      "epoch": 352.3878787878788,
      "grad_norm": 0.30005964636802673,
      "learning_rate": 1.124e-05,
      "loss": 0.3685,
      "step": 3876
    },
    {
      "epoch": 352.4848484848485,
      "grad_norm": 0.334306001663208,
      "learning_rate": 1.1230000000000001e-05,
      "loss": 0.3317,
      "step": 3877
    },
    {
      "epoch": 352.58181818181816,
      "grad_norm": 0.2881486415863037,
      "learning_rate": 1.122e-05,
      "loss": 0.3174,
      "step": 3878
    },
    {
      "epoch": 352.6787878787879,
      "grad_norm": 0.2741554379463196,
      "learning_rate": 1.1210000000000001e-05,
      "loss": 0.3515,
      "step": 3879
    },
    {
      "epoch": 352.77575757575755,
      "grad_norm": 0.2765088677406311,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.3364,
      "step": 3880
    },
    {
      "epoch": 352.77575757575755,
      "eval_loss": 0.42547720670700073,
      "eval_runtime": 2.1326,
      "eval_samples_per_second": 25.79,
      "eval_steps_per_second": 3.282,
      "step": 3880
    },
    {
      "epoch": 352.8727272727273,
      "grad_norm": 0.21953384578227997,
      "learning_rate": 1.1190000000000001e-05,
      "loss": 0.3198,
      "step": 3881
    },
    {
      "epoch": 352.969696969697,
      "grad_norm": 0.32624542713165283,
      "learning_rate": 1.118e-05,
      "loss": 0.3275,
      "step": 3882
    },
    {
      "epoch": 353.0,
      "grad_norm": 0.4760756492614746,
      "learning_rate": 1.117e-05,
      "loss": 0.2851,
      "step": 3883
    },
    {
      "epoch": 353.0969696969697,
      "grad_norm": 0.26394760608673096,
      "learning_rate": 1.1160000000000002e-05,
      "loss": 0.2947,
      "step": 3884
    },
    {
      "epoch": 353.1939393939394,
      "grad_norm": 0.31982186436653137,
      "learning_rate": 1.115e-05,
      "loss": 0.3338,
      "step": 3885
    },
    {
      "epoch": 353.2909090909091,
      "grad_norm": 0.2912452220916748,
      "learning_rate": 1.114e-05,
      "loss": 0.3035,
      "step": 3886
    },
    {
      "epoch": 353.3878787878788,
      "grad_norm": 0.28226861357688904,
      "learning_rate": 1.113e-05,
      "loss": 0.328,
      "step": 3887
    },
    {
      "epoch": 353.4848484848485,
      "grad_norm": 0.2831850051879883,
      "learning_rate": 1.112e-05,
      "loss": 0.3714,
      "step": 3888
    },
    {
      "epoch": 353.58181818181816,
      "grad_norm": 0.292229562997818,
      "learning_rate": 1.111e-05,
      "loss": 0.3491,
      "step": 3889
    },
    {
      "epoch": 353.6787878787879,
      "grad_norm": 0.2777029573917389,
      "learning_rate": 1.11e-05,
      "loss": 0.3499,
      "step": 3890
    },
    {
      "epoch": 353.6787878787879,
      "eval_loss": 0.42573267221450806,
      "eval_runtime": 2.1309,
      "eval_samples_per_second": 25.811,
      "eval_steps_per_second": 3.285,
      "step": 3890
    },
    {
      "epoch": 353.77575757575755,
      "grad_norm": 0.32899153232574463,
      "learning_rate": 1.109e-05,
      "loss": 0.3612,
      "step": 3891
    },
    {
      "epoch": 353.8727272727273,
      "grad_norm": 0.2697356045246124,
      "learning_rate": 1.108e-05,
      "loss": 0.3014,
      "step": 3892
    },
    {
      "epoch": 353.969696969697,
      "grad_norm": 0.26077619194984436,
      "learning_rate": 1.107e-05,
      "loss": 0.3224,
      "step": 3893
    },
    {
      "epoch": 354.0,
      "grad_norm": 0.5170732736587524,
      "learning_rate": 1.106e-05,
      "loss": 0.3688,
      "step": 3894
    },
    {
      "epoch": 354.0969696969697,
      "grad_norm": 0.2870286703109741,
      "learning_rate": 1.1050000000000001e-05,
      "loss": 0.3059,
      "step": 3895
    },
    {
      "epoch": 354.1939393939394,
      "grad_norm": 0.28191274404525757,
      "learning_rate": 1.1040000000000001e-05,
      "loss": 0.3329,
      "step": 3896
    },
    {
      "epoch": 354.2909090909091,
      "grad_norm": 0.2566986382007599,
      "learning_rate": 1.103e-05,
      "loss": 0.3068,
      "step": 3897
    },
    {
      "epoch": 354.3878787878788,
      "grad_norm": 0.2678109109401703,
      "learning_rate": 1.1020000000000001e-05,
      "loss": 0.317,
      "step": 3898
    },
    {
      "epoch": 354.4848484848485,
      "grad_norm": 0.29949188232421875,
      "learning_rate": 1.1010000000000001e-05,
      "loss": 0.34,
      "step": 3899
    },
    {
      "epoch": 354.58181818181816,
      "grad_norm": 0.330269455909729,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.3302,
      "step": 3900
    },
    {
      "epoch": 354.58181818181816,
      "eval_loss": 0.4255954623222351,
      "eval_runtime": 2.1491,
      "eval_samples_per_second": 25.592,
      "eval_steps_per_second": 3.257,
      "step": 3900
    },
    {
      "epoch": 354.6787878787879,
      "grad_norm": 0.30191609263420105,
      "learning_rate": 1.099e-05,
      "loss": 0.3181,
      "step": 3901
    },
    {
      "epoch": 354.77575757575755,
      "grad_norm": 0.2983716130256653,
      "learning_rate": 1.098e-05,
      "loss": 0.3562,
      "step": 3902
    },
    {
      "epoch": 354.8727272727273,
      "grad_norm": 0.329935222864151,
      "learning_rate": 1.0970000000000002e-05,
      "loss": 0.3566,
      "step": 3903
    },
    {
      "epoch": 354.969696969697,
      "grad_norm": 0.30771416425704956,
      "learning_rate": 1.096e-05,
      "loss": 0.3578,
      "step": 3904
    },
    {
      "epoch": 355.0,
      "grad_norm": 0.46148690581321716,
      "learning_rate": 1.095e-05,
      "loss": 0.3558,
      "step": 3905
    },
    {
      "epoch": 355.0969696969697,
      "grad_norm": 0.30684080719947815,
      "learning_rate": 1.094e-05,
      "loss": 0.3014,
      "step": 3906
    },
    {
      "epoch": 355.1939393939394,
      "grad_norm": 0.26615822315216064,
      "learning_rate": 1.093e-05,
      "loss": 0.3163,
      "step": 3907
    },
    {
      "epoch": 355.2909090909091,
      "grad_norm": 0.2488378882408142,
      "learning_rate": 1.092e-05,
      "loss": 0.3514,
      "step": 3908
    },
    {
      "epoch": 355.3878787878788,
      "grad_norm": 0.29284900426864624,
      "learning_rate": 1.091e-05,
      "loss": 0.3522,
      "step": 3909
    },
    {
      "epoch": 355.4848484848485,
      "grad_norm": 0.2902972102165222,
      "learning_rate": 1.09e-05,
      "loss": 0.3437,
      "step": 3910
    },
    {
      "epoch": 355.4848484848485,
      "eval_loss": 0.42546331882476807,
      "eval_runtime": 2.1237,
      "eval_samples_per_second": 25.898,
      "eval_steps_per_second": 3.296,
      "step": 3910
    },
    {
      "epoch": 355.58181818181816,
      "grad_norm": 0.3538964092731476,
      "learning_rate": 1.089e-05,
      "loss": 0.3273,
      "step": 3911
    },
    {
      "epoch": 355.6787878787879,
      "grad_norm": 0.3419858515262604,
      "learning_rate": 1.088e-05,
      "loss": 0.3456,
      "step": 3912
    },
    {
      "epoch": 355.77575757575755,
      "grad_norm": 0.310893177986145,
      "learning_rate": 1.0870000000000001e-05,
      "loss": 0.3529,
      "step": 3913
    },
    {
      "epoch": 355.8727272727273,
      "grad_norm": 0.2636914551258087,
      "learning_rate": 1.0860000000000001e-05,
      "loss": 0.3095,
      "step": 3914
    },
    {
      "epoch": 355.969696969697,
      "grad_norm": 0.2791852653026581,
      "learning_rate": 1.0850000000000001e-05,
      "loss": 0.3287,
      "step": 3915
    },
    {
      "epoch": 356.0,
      "grad_norm": 0.4313600957393646,
      "learning_rate": 1.084e-05,
      "loss": 0.3258,
      "step": 3916
    },
    {
      "epoch": 356.0969696969697,
      "grad_norm": 0.3125990629196167,
      "learning_rate": 1.083e-05,
      "loss": 0.3627,
      "step": 3917
    },
    {
      "epoch": 356.1939393939394,
      "grad_norm": 0.28080224990844727,
      "learning_rate": 1.0820000000000001e-05,
      "loss": 0.348,
      "step": 3918
    },
    {
      "epoch": 356.2909090909091,
      "grad_norm": 0.27683061361312866,
      "learning_rate": 1.081e-05,
      "loss": 0.3389,
      "step": 3919
    },
    {
      "epoch": 356.3878787878788,
      "grad_norm": 0.509411633014679,
      "learning_rate": 1.08e-05,
      "loss": 0.3431,
      "step": 3920
    },
    {
      "epoch": 356.3878787878788,
      "eval_loss": 0.42507705092430115,
      "eval_runtime": 2.1365,
      "eval_samples_per_second": 25.744,
      "eval_steps_per_second": 3.276,
      "step": 3920
    },
    {
      "epoch": 356.4848484848485,
      "grad_norm": 0.2736632227897644,
      "learning_rate": 1.079e-05,
      "loss": 0.3336,
      "step": 3921
    },
    {
      "epoch": 356.58181818181816,
      "grad_norm": 0.27927494049072266,
      "learning_rate": 1.0780000000000002e-05,
      "loss": 0.3236,
      "step": 3922
    },
    {
      "epoch": 356.6787878787879,
      "grad_norm": 0.23492373526096344,
      "learning_rate": 1.077e-05,
      "loss": 0.3135,
      "step": 3923
    },
    {
      "epoch": 356.77575757575755,
      "grad_norm": 0.33352339267730713,
      "learning_rate": 1.076e-05,
      "loss": 0.2849,
      "step": 3924
    },
    {
      "epoch": 356.8727272727273,
      "grad_norm": 0.2793315649032593,
      "learning_rate": 1.075e-05,
      "loss": 0.3611,
      "step": 3925
    },
    {
      "epoch": 356.969696969697,
      "grad_norm": 0.29622992873191833,
      "learning_rate": 1.074e-05,
      "loss": 0.3215,
      "step": 3926
    },
    {
      "epoch": 357.0,
      "grad_norm": 0.4465533196926117,
      "learning_rate": 1.073e-05,
      "loss": 0.3219,
      "step": 3927
    },
    {
      "epoch": 357.0969696969697,
      "grad_norm": 0.26654234528541565,
      "learning_rate": 1.072e-05,
      "loss": 0.3282,
      "step": 3928
    },
    {
      "epoch": 357.1939393939394,
      "grad_norm": 0.3026706278324127,
      "learning_rate": 1.071e-05,
      "loss": 0.327,
      "step": 3929
    },
    {
      "epoch": 357.2909090909091,
      "grad_norm": 0.2650451362133026,
      "learning_rate": 1.0700000000000001e-05,
      "loss": 0.3291,
      "step": 3930
    },
    {
      "epoch": 357.2909090909091,
      "eval_loss": 0.42528554797172546,
      "eval_runtime": 2.1199,
      "eval_samples_per_second": 25.945,
      "eval_steps_per_second": 3.302,
      "step": 3930
    },
    {
      "epoch": 357.3878787878788,
      "grad_norm": 0.26152563095092773,
      "learning_rate": 1.069e-05,
      "loss": 0.3328,
      "step": 3931
    },
    {
      "epoch": 357.4848484848485,
      "grad_norm": 0.30806395411491394,
      "learning_rate": 1.0680000000000001e-05,
      "loss": 0.3217,
      "step": 3932
    },
    {
      "epoch": 357.58181818181816,
      "grad_norm": 0.2408694475889206,
      "learning_rate": 1.0670000000000001e-05,
      "loss": 0.3296,
      "step": 3933
    },
    {
      "epoch": 357.6787878787879,
      "grad_norm": 0.35017138719558716,
      "learning_rate": 1.0660000000000001e-05,
      "loss": 0.3301,
      "step": 3934
    },
    {
      "epoch": 357.77575757575755,
      "grad_norm": 0.26608020067214966,
      "learning_rate": 1.065e-05,
      "loss": 0.316,
      "step": 3935
    },
    {
      "epoch": 357.8727272727273,
      "grad_norm": 0.3791072964668274,
      "learning_rate": 1.064e-05,
      "loss": 0.3598,
      "step": 3936
    },
    {
      "epoch": 357.969696969697,
      "grad_norm": 0.2872272729873657,
      "learning_rate": 1.0630000000000002e-05,
      "loss": 0.3669,
      "step": 3937
    },
    {
      "epoch": 358.0,
      "grad_norm": 0.5474063754081726,
      "learning_rate": 1.062e-05,
      "loss": 0.2731,
      "step": 3938
    },
    {
      "epoch": 358.0969696969697,
      "grad_norm": 0.30581510066986084,
      "learning_rate": 1.061e-05,
      "loss": 0.3275,
      "step": 3939
    },
    {
      "epoch": 358.1939393939394,
      "grad_norm": 0.3368642032146454,
      "learning_rate": 1.06e-05,
      "loss": 0.367,
      "step": 3940
    },
    {
      "epoch": 358.1939393939394,
      "eval_loss": 0.42523324489593506,
      "eval_runtime": 2.1338,
      "eval_samples_per_second": 25.776,
      "eval_steps_per_second": 3.281,
      "step": 3940
    },
    {
      "epoch": 358.2909090909091,
      "grad_norm": 0.28304001688957214,
      "learning_rate": 1.059e-05,
      "loss": 0.3377,
      "step": 3941
    },
    {
      "epoch": 358.3878787878788,
      "grad_norm": 0.22714465856552124,
      "learning_rate": 1.058e-05,
      "loss": 0.3298,
      "step": 3942
    },
    {
      "epoch": 358.4848484848485,
      "grad_norm": 0.30259737372398376,
      "learning_rate": 1.057e-05,
      "loss": 0.3283,
      "step": 3943
    },
    {
      "epoch": 358.58181818181816,
      "grad_norm": 0.2771393954753876,
      "learning_rate": 1.056e-05,
      "loss": 0.3162,
      "step": 3944
    },
    {
      "epoch": 358.6787878787879,
      "grad_norm": 0.2818312644958496,
      "learning_rate": 1.055e-05,
      "loss": 0.344,
      "step": 3945
    },
    {
      "epoch": 358.77575757575755,
      "grad_norm": 0.26231974363327026,
      "learning_rate": 1.0539999999999999e-05,
      "loss": 0.337,
      "step": 3946
    },
    {
      "epoch": 358.8727272727273,
      "grad_norm": 0.34645920991897583,
      "learning_rate": 1.053e-05,
      "loss": 0.3291,
      "step": 3947
    },
    {
      "epoch": 358.969696969697,
      "grad_norm": 0.29468825459480286,
      "learning_rate": 1.0520000000000001e-05,
      "loss": 0.2825,
      "step": 3948
    },
    {
      "epoch": 359.0,
      "grad_norm": 0.5619099736213684,
      "learning_rate": 1.0510000000000001e-05,
      "loss": 0.3934,
      "step": 3949
    },
    {
      "epoch": 359.0969696969697,
      "grad_norm": 0.258922815322876,
      "learning_rate": 1.05e-05,
      "loss": 0.3313,
      "step": 3950
    },
    {
      "epoch": 359.0969696969697,
      "eval_loss": 0.4250088334083557,
      "eval_runtime": 2.1484,
      "eval_samples_per_second": 25.6,
      "eval_steps_per_second": 3.258,
      "step": 3950
    },
    {
      "epoch": 359.1939393939394,
      "grad_norm": 0.4014686644077301,
      "learning_rate": 1.049e-05,
      "loss": 0.3434,
      "step": 3951
    },
    {
      "epoch": 359.2909090909091,
      "grad_norm": 0.270025372505188,
      "learning_rate": 1.0480000000000001e-05,
      "loss": 0.2918,
      "step": 3952
    },
    {
      "epoch": 359.3878787878788,
      "grad_norm": 0.31015175580978394,
      "learning_rate": 1.0470000000000001e-05,
      "loss": 0.3407,
      "step": 3953
    },
    {
      "epoch": 359.4848484848485,
      "grad_norm": 0.26703396439552307,
      "learning_rate": 1.046e-05,
      "loss": 0.3264,
      "step": 3954
    },
    {
      "epoch": 359.58181818181816,
      "grad_norm": 0.25800076127052307,
      "learning_rate": 1.045e-05,
      "loss": 0.3408,
      "step": 3955
    },
    {
      "epoch": 359.6787878787879,
      "grad_norm": 0.2913641631603241,
      "learning_rate": 1.0440000000000002e-05,
      "loss": 0.3562,
      "step": 3956
    },
    {
      "epoch": 359.77575757575755,
      "grad_norm": 0.350644052028656,
      "learning_rate": 1.043e-05,
      "loss": 0.3556,
      "step": 3957
    },
    {
      "epoch": 359.8727272727273,
      "grad_norm": 0.2973456084728241,
      "learning_rate": 1.042e-05,
      "loss": 0.3085,
      "step": 3958
    },
    {
      "epoch": 359.969696969697,
      "grad_norm": 0.2788713276386261,
      "learning_rate": 1.041e-05,
      "loss": 0.3227,
      "step": 3959
    },
    {
      "epoch": 360.0,
      "grad_norm": 0.42132866382598877,
      "learning_rate": 1.04e-05,
      "loss": 0.3393,
      "step": 3960
    },
    {
      "epoch": 360.0,
      "eval_loss": 0.42503562569618225,
      "eval_runtime": 2.1239,
      "eval_samples_per_second": 25.895,
      "eval_steps_per_second": 3.296,
      "step": 3960
    },
    {
      "epoch": 360.0969696969697,
      "grad_norm": 0.2816360890865326,
      "learning_rate": 1.039e-05,
      "loss": 0.2806,
      "step": 3961
    },
    {
      "epoch": 360.1939393939394,
      "grad_norm": 0.26682475209236145,
      "learning_rate": 1.038e-05,
      "loss": 0.3402,
      "step": 3962
    },
    {
      "epoch": 360.2909090909091,
      "grad_norm": 0.2514760494232178,
      "learning_rate": 1.037e-05,
      "loss": 0.3163,
      "step": 3963
    },
    {
      "epoch": 360.3878787878788,
      "grad_norm": 0.299314022064209,
      "learning_rate": 1.036e-05,
      "loss": 0.3187,
      "step": 3964
    },
    {
      "epoch": 360.4848484848485,
      "grad_norm": 0.2835948169231415,
      "learning_rate": 1.035e-05,
      "loss": 0.304,
      "step": 3965
    },
    {
      "epoch": 360.58181818181816,
      "grad_norm": 0.2949691116809845,
      "learning_rate": 1.0340000000000001e-05,
      "loss": 0.3801,
      "step": 3966
    },
    {
      "epoch": 360.6787878787879,
      "grad_norm": 0.30468055605888367,
      "learning_rate": 1.0330000000000001e-05,
      "loss": 0.3158,
      "step": 3967
    },
    {
      "epoch": 360.77575757575755,
      "grad_norm": 0.3264327347278595,
      "learning_rate": 1.0320000000000001e-05,
      "loss": 0.3169,
      "step": 3968
    },
    {
      "epoch": 360.8727272727273,
      "grad_norm": 0.3753027617931366,
      "learning_rate": 1.031e-05,
      "loss": 0.3805,
      "step": 3969
    },
    {
      "epoch": 360.969696969697,
      "grad_norm": 0.31869491934776306,
      "learning_rate": 1.03e-05,
      "loss": 0.3675,
      "step": 3970
    },
    {
      "epoch": 360.969696969697,
      "eval_loss": 0.4247276782989502,
      "eval_runtime": 2.1286,
      "eval_samples_per_second": 25.838,
      "eval_steps_per_second": 3.288,
      "step": 3970
    },
    {
      "epoch": 361.0,
      "grad_norm": 0.547243595123291,
      "learning_rate": 1.0290000000000001e-05,
      "loss": 0.3337,
      "step": 3971
    },
    {
      "epoch": 361.0969696969697,
      "grad_norm": 0.290824294090271,
      "learning_rate": 1.0280000000000002e-05,
      "loss": 0.3522,
      "step": 3972
    },
    {
      "epoch": 361.1939393939394,
      "grad_norm": 0.24751658737659454,
      "learning_rate": 1.027e-05,
      "loss": 0.3332,
      "step": 3973
    },
    {
      "epoch": 361.2909090909091,
      "grad_norm": 0.2930498421192169,
      "learning_rate": 1.026e-05,
      "loss": 0.3102,
      "step": 3974
    },
    {
      "epoch": 361.3878787878788,
      "grad_norm": 0.28421491384506226,
      "learning_rate": 1.025e-05,
      "loss": 0.3381,
      "step": 3975
    },
    {
      "epoch": 361.4848484848485,
      "grad_norm": 0.31874242424964905,
      "learning_rate": 1.024e-05,
      "loss": 0.3196,
      "step": 3976
    },
    {
      "epoch": 361.58181818181816,
      "grad_norm": 0.3144734501838684,
      "learning_rate": 1.023e-05,
      "loss": 0.3567,
      "step": 3977
    },
    {
      "epoch": 361.6787878787879,
      "grad_norm": 0.27309650182724,
      "learning_rate": 1.022e-05,
      "loss": 0.3337,
      "step": 3978
    },
    {
      "epoch": 361.77575757575755,
      "grad_norm": 0.35673069953918457,
      "learning_rate": 1.021e-05,
      "loss": 0.3244,
      "step": 3979
    },
    {
      "epoch": 361.8727272727273,
      "grad_norm": 0.28269463777542114,
      "learning_rate": 1.02e-05,
      "loss": 0.3315,
      "step": 3980
    },
    {
      "epoch": 361.8727272727273,
      "eval_loss": 0.4251295030117035,
      "eval_runtime": 2.116,
      "eval_samples_per_second": 25.992,
      "eval_steps_per_second": 3.308,
      "step": 3980
    },
    {
      "epoch": 361.969696969697,
      "grad_norm": 0.26234710216522217,
      "learning_rate": 1.019e-05,
      "loss": 0.323,
      "step": 3981
    },
    {
      "epoch": 362.0,
      "grad_norm": 0.4822295010089874,
      "learning_rate": 1.018e-05,
      "loss": 0.3195,
      "step": 3982
    },
    {
      "epoch": 362.0969696969697,
      "grad_norm": 0.2835442125797272,
      "learning_rate": 1.0170000000000001e-05,
      "loss": 0.3206,
      "step": 3983
    },
    {
      "epoch": 362.1939393939394,
      "grad_norm": 0.3007937967777252,
      "learning_rate": 1.016e-05,
      "loss": 0.3322,
      "step": 3984
    },
    {
      "epoch": 362.2909090909091,
      "grad_norm": 0.27044838666915894,
      "learning_rate": 1.0150000000000001e-05,
      "loss": 0.3168,
      "step": 3985
    },
    {
      "epoch": 362.3878787878788,
      "grad_norm": 0.27905598282814026,
      "learning_rate": 1.0140000000000001e-05,
      "loss": 0.3209,
      "step": 3986
    },
    {
      "epoch": 362.4848484848485,
      "grad_norm": 0.2970101535320282,
      "learning_rate": 1.0130000000000001e-05,
      "loss": 0.3481,
      "step": 3987
    },
    {
      "epoch": 362.58181818181816,
      "grad_norm": 0.27853038907051086,
      "learning_rate": 1.012e-05,
      "loss": 0.3228,
      "step": 3988
    },
    {
      "epoch": 362.6787878787879,
      "grad_norm": 0.32778528332710266,
      "learning_rate": 1.011e-05,
      "loss": 0.3385,
      "step": 3989
    },
    {
      "epoch": 362.77575757575755,
      "grad_norm": 0.3114234507083893,
      "learning_rate": 1.0100000000000002e-05,
      "loss": 0.3553,
      "step": 3990
    },
    {
      "epoch": 362.77575757575755,
      "eval_loss": 0.4250686764717102,
      "eval_runtime": 2.1338,
      "eval_samples_per_second": 25.775,
      "eval_steps_per_second": 3.28,
      "step": 3990
    },
    {
      "epoch": 362.8727272727273,
      "grad_norm": 0.2816259264945984,
      "learning_rate": 1.0090000000000002e-05,
      "loss": 0.3652,
      "step": 3991
    },
    {
      "epoch": 362.969696969697,
      "grad_norm": 0.2591400146484375,
      "learning_rate": 1.008e-05,
      "loss": 0.3156,
      "step": 3992
    },
    {
      "epoch": 363.0,
      "grad_norm": 0.47758832573890686,
      "learning_rate": 1.007e-05,
      "loss": 0.2642,
      "step": 3993
    },
    {
      "epoch": 363.0969696969697,
      "grad_norm": 0.33999863266944885,
      "learning_rate": 1.006e-05,
      "loss": 0.2949,
      "step": 3994
    },
    {
      "epoch": 363.1939393939394,
      "grad_norm": 0.2757456302642822,
      "learning_rate": 1.005e-05,
      "loss": 0.3314,
      "step": 3995
    },
    {
      "epoch": 363.2909090909091,
      "grad_norm": 0.2781667113304138,
      "learning_rate": 1.004e-05,
      "loss": 0.3512,
      "step": 3996
    },
    {
      "epoch": 363.3878787878788,
      "grad_norm": 0.30516746640205383,
      "learning_rate": 1.003e-05,
      "loss": 0.3407,
      "step": 3997
    },
    {
      "epoch": 363.4848484848485,
      "grad_norm": 0.3394595682621002,
      "learning_rate": 1.002e-05,
      "loss": 0.3415,
      "step": 3998
    },
    {
      "epoch": 363.58181818181816,
      "grad_norm": 0.3109613060951233,
      "learning_rate": 1.001e-05,
      "loss": 0.324,
      "step": 3999
    },
    {
      "epoch": 363.6787878787879,
      "grad_norm": 0.2772013247013092,
      "learning_rate": 1e-05,
      "loss": 0.3119,
      "step": 4000
    },
    {
      "epoch": 363.6787878787879,
      "eval_loss": 0.4250546395778656,
      "eval_runtime": 2.1465,
      "eval_samples_per_second": 25.623,
      "eval_steps_per_second": 3.261,
      "step": 4000
    },
    {
      "epoch": 363.77575757575755,
      "grad_norm": 0.31099388003349304,
      "learning_rate": 9.990000000000001e-06,
      "loss": 0.3787,
      "step": 4001
    },
    {
      "epoch": 363.8727272727273,
      "grad_norm": 0.29429104924201965,
      "learning_rate": 9.980000000000001e-06,
      "loss": 0.3431,
      "step": 4002
    },
    {
      "epoch": 363.969696969697,
      "grad_norm": 0.27421092987060547,
      "learning_rate": 9.97e-06,
      "loss": 0.2933,
      "step": 4003
    },
    {
      "epoch": 364.0,
      "grad_norm": 0.4427415430545807,
      "learning_rate": 9.96e-06,
      "loss": 0.3564,
      "step": 4004
    },
    {
      "epoch": 364.0969696969697,
      "grad_norm": 0.2868616580963135,
      "learning_rate": 9.950000000000001e-06,
      "loss": 0.3189,
      "step": 4005
    },
    {
      "epoch": 364.1939393939394,
      "grad_norm": 0.30533933639526367,
      "learning_rate": 9.940000000000001e-06,
      "loss": 0.3299,
      "step": 4006
    },
    {
      "epoch": 364.2909090909091,
      "grad_norm": 0.2742302417755127,
      "learning_rate": 9.93e-06,
      "loss": 0.3206,
      "step": 4007
    },
    {
      "epoch": 364.3878787878788,
      "grad_norm": 0.29955291748046875,
      "learning_rate": 9.92e-06,
      "loss": 0.3458,
      "step": 4008
    },
    {
      "epoch": 364.4848484848485,
      "grad_norm": 0.24295401573181152,
      "learning_rate": 9.91e-06,
      "loss": 0.3426,
      "step": 4009
    },
    {
      "epoch": 364.58181818181816,
      "grad_norm": 0.2764863669872284,
      "learning_rate": 9.900000000000002e-06,
      "loss": 0.3547,
      "step": 4010
    },
    {
      "epoch": 364.58181818181816,
      "eval_loss": 0.4247507154941559,
      "eval_runtime": 2.1081,
      "eval_samples_per_second": 26.09,
      "eval_steps_per_second": 3.321,
      "step": 4010
    },
    {
      "epoch": 364.6787878787879,
      "grad_norm": 0.28245973587036133,
      "learning_rate": 9.89e-06,
      "loss": 0.3362,
      "step": 4011
    },
    {
      "epoch": 364.77575757575755,
      "grad_norm": 0.33747348189353943,
      "learning_rate": 9.88e-06,
      "loss": 0.3207,
      "step": 4012
    },
    {
      "epoch": 364.8727272727273,
      "grad_norm": 0.3213714361190796,
      "learning_rate": 9.87e-06,
      "loss": 0.3217,
      "step": 4013
    },
    {
      "epoch": 364.969696969697,
      "grad_norm": 0.2978791892528534,
      "learning_rate": 9.86e-06,
      "loss": 0.3275,
      "step": 4014
    },
    {
      "epoch": 365.0,
      "grad_norm": 0.5955439805984497,
      "learning_rate": 9.85e-06,
      "loss": 0.3338,
      "step": 4015
    },
    {
      "epoch": 365.0969696969697,
      "grad_norm": 0.25707629323005676,
      "learning_rate": 9.84e-06,
      "loss": 0.3284,
      "step": 4016
    },
    {
      "epoch": 365.1939393939394,
      "grad_norm": 0.33836591243743896,
      "learning_rate": 9.83e-06,
      "loss": 0.3375,
      "step": 4017
    },
    {
      "epoch": 365.2909090909091,
      "grad_norm": 0.2548236846923828,
      "learning_rate": 9.820000000000001e-06,
      "loss": 0.329,
      "step": 4018
    },
    {
      "epoch": 365.3878787878788,
      "grad_norm": 0.30328866839408875,
      "learning_rate": 9.810000000000001e-06,
      "loss": 0.3284,
      "step": 4019
    },
    {
      "epoch": 365.4848484848485,
      "grad_norm": 0.29344213008880615,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.3367,
      "step": 4020
    },
    {
      "epoch": 365.4848484848485,
      "eval_loss": 0.4250096082687378,
      "eval_runtime": 2.1186,
      "eval_samples_per_second": 25.961,
      "eval_steps_per_second": 3.304,
      "step": 4020
    },
    {
      "epoch": 365.58181818181816,
      "grad_norm": 0.29405832290649414,
      "learning_rate": 9.790000000000001e-06,
      "loss": 0.3655,
      "step": 4021
    },
    {
      "epoch": 365.6787878787879,
      "grad_norm": 0.31418031454086304,
      "learning_rate": 9.78e-06,
      "loss": 0.3249,
      "step": 4022
    },
    {
      "epoch": 365.77575757575755,
      "grad_norm": 0.3338219225406647,
      "learning_rate": 9.77e-06,
      "loss": 0.2812,
      "step": 4023
    },
    {
      "epoch": 365.8727272727273,
      "grad_norm": 0.26671719551086426,
      "learning_rate": 9.760000000000001e-06,
      "loss": 0.3499,
      "step": 4024
    },
    {
      "epoch": 365.969696969697,
      "grad_norm": 0.32956159114837646,
      "learning_rate": 9.750000000000002e-06,
      "loss": 0.3476,
      "step": 4025
    },
    {
      "epoch": 366.0,
      "grad_norm": 0.4222256541252136,
      "learning_rate": 9.74e-06,
      "loss": 0.2755,
      "step": 4026
    },
    {
      "epoch": 366.0969696969697,
      "grad_norm": 0.27772313356399536,
      "learning_rate": 9.73e-06,
      "loss": 0.3148,
      "step": 4027
    },
    {
      "epoch": 366.1939393939394,
      "grad_norm": 0.32751691341400146,
      "learning_rate": 9.72e-06,
      "loss": 0.3628,
      "step": 4028
    },
    {
      "epoch": 366.2909090909091,
      "grad_norm": 0.3402860462665558,
      "learning_rate": 9.71e-06,
      "loss": 0.3789,
      "step": 4029
    },
    {
      "epoch": 366.3878787878788,
      "grad_norm": 0.2285175919532776,
      "learning_rate": 9.7e-06,
      "loss": 0.3331,
      "step": 4030
    },
    {
      "epoch": 366.3878787878788,
      "eval_loss": 0.42465740442276,
      "eval_runtime": 2.1169,
      "eval_samples_per_second": 25.981,
      "eval_steps_per_second": 3.307,
      "step": 4030
    },
    {
      "epoch": 366.4848484848485,
      "grad_norm": 0.26843827962875366,
      "learning_rate": 9.69e-06,
      "loss": 0.3265,
      "step": 4031
    },
    {
      "epoch": 366.58181818181816,
      "grad_norm": 0.36523428559303284,
      "learning_rate": 9.68e-06,
      "loss": 0.2867,
      "step": 4032
    },
    {
      "epoch": 366.6787878787879,
      "grad_norm": 0.2948589324951172,
      "learning_rate": 9.67e-06,
      "loss": 0.3547,
      "step": 4033
    },
    {
      "epoch": 366.77575757575755,
      "grad_norm": 0.2663814127445221,
      "learning_rate": 9.66e-06,
      "loss": 0.2983,
      "step": 4034
    },
    {
      "epoch": 366.8727272727273,
      "grad_norm": 0.4233630299568176,
      "learning_rate": 9.65e-06,
      "loss": 0.3001,
      "step": 4035
    },
    {
      "epoch": 366.969696969697,
      "grad_norm": 0.2976456582546234,
      "learning_rate": 9.640000000000001e-06,
      "loss": 0.3356,
      "step": 4036
    },
    {
      "epoch": 367.0,
      "grad_norm": 0.5265571475028992,
      "learning_rate": 9.630000000000001e-06,
      "loss": 0.3957,
      "step": 4037
    },
    {
      "epoch": 367.0969696969697,
      "grad_norm": 0.269848108291626,
      "learning_rate": 9.62e-06,
      "loss": 0.349,
      "step": 4038
    },
    {
      "epoch": 367.1939393939394,
      "grad_norm": 0.3301657438278198,
      "learning_rate": 9.610000000000001e-06,
      "loss": 0.3184,
      "step": 4039
    },
    {
      "epoch": 367.2909090909091,
      "grad_norm": 0.3008776307106018,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.3579,
      "step": 4040
    },
    {
      "epoch": 367.2909090909091,
      "eval_loss": 0.4249913692474365,
      "eval_runtime": 2.1317,
      "eval_samples_per_second": 25.801,
      "eval_steps_per_second": 3.284,
      "step": 4040
    },
    {
      "epoch": 367.3878787878788,
      "grad_norm": 0.27042990922927856,
      "learning_rate": 9.59e-06,
      "loss": 0.332,
      "step": 4041
    },
    {
      "epoch": 367.4848484848485,
      "grad_norm": 0.2910686731338501,
      "learning_rate": 9.58e-06,
      "loss": 0.306,
      "step": 4042
    },
    {
      "epoch": 367.58181818181816,
      "grad_norm": 0.32154032588005066,
      "learning_rate": 9.57e-06,
      "loss": 0.3121,
      "step": 4043
    },
    {
      "epoch": 367.6787878787879,
      "grad_norm": 0.303244948387146,
      "learning_rate": 9.560000000000002e-06,
      "loss": 0.3221,
      "step": 4044
    },
    {
      "epoch": 367.77575757575755,
      "grad_norm": 0.292343407869339,
      "learning_rate": 9.55e-06,
      "loss": 0.375,
      "step": 4045
    },
    {
      "epoch": 367.8727272727273,
      "grad_norm": 0.28821298480033875,
      "learning_rate": 9.54e-06,
      "loss": 0.3198,
      "step": 4046
    },
    {
      "epoch": 367.969696969697,
      "grad_norm": 0.3002316653728485,
      "learning_rate": 9.53e-06,
      "loss": 0.3319,
      "step": 4047
    },
    {
      "epoch": 368.0,
      "grad_norm": 0.4897102117538452,
      "learning_rate": 9.52e-06,
      "loss": 0.3124,
      "step": 4048
    },
    {
      "epoch": 368.0969696969697,
      "grad_norm": 0.3233930170536041,
      "learning_rate": 9.51e-06,
      "loss": 0.3134,
      "step": 4049
    },
    {
      "epoch": 368.1939393939394,
      "grad_norm": 0.2641765773296356,
      "learning_rate": 9.5e-06,
      "loss": 0.334,
      "step": 4050
    },
    {
      "epoch": 368.1939393939394,
      "eval_loss": 0.42491304874420166,
      "eval_runtime": 2.126,
      "eval_samples_per_second": 25.87,
      "eval_steps_per_second": 3.292,
      "step": 4050
    },
    {
      "epoch": 368.2909090909091,
      "grad_norm": 0.27830106019973755,
      "learning_rate": 9.49e-06,
      "loss": 0.3238,
      "step": 4051
    },
    {
      "epoch": 368.3878787878788,
      "grad_norm": 0.25568294525146484,
      "learning_rate": 9.48e-06,
      "loss": 0.3427,
      "step": 4052
    },
    {
      "epoch": 368.4848484848485,
      "grad_norm": 0.3005104660987854,
      "learning_rate": 9.47e-06,
      "loss": 0.2992,
      "step": 4053
    },
    {
      "epoch": 368.58181818181816,
      "grad_norm": 0.2774483263492584,
      "learning_rate": 9.460000000000001e-06,
      "loss": 0.3246,
      "step": 4054
    },
    {
      "epoch": 368.6787878787879,
      "grad_norm": 0.32644858956336975,
      "learning_rate": 9.450000000000001e-06,
      "loss": 0.3247,
      "step": 4055
    },
    {
      "epoch": 368.77575757575755,
      "grad_norm": 0.31656840443611145,
      "learning_rate": 9.44e-06,
      "loss": 0.3473,
      "step": 4056
    },
    {
      "epoch": 368.8727272727273,
      "grad_norm": 0.2856542766094208,
      "learning_rate": 9.43e-06,
      "loss": 0.3635,
      "step": 4057
    },
    {
      "epoch": 368.969696969697,
      "grad_norm": 0.25360551476478577,
      "learning_rate": 9.420000000000001e-06,
      "loss": 0.3494,
      "step": 4058
    },
    {
      "epoch": 369.0,
      "grad_norm": 0.5523640513420105,
      "learning_rate": 9.410000000000001e-06,
      "loss": 0.2984,
      "step": 4059
    },
    {
      "epoch": 369.0969696969697,
      "grad_norm": 0.26677101850509644,
      "learning_rate": 9.4e-06,
      "loss": 0.3447,
      "step": 4060
    },
    {
      "epoch": 369.0969696969697,
      "eval_loss": 0.42472049593925476,
      "eval_runtime": 2.1255,
      "eval_samples_per_second": 25.876,
      "eval_steps_per_second": 3.293,
      "step": 4060
    },
    {
      "epoch": 369.1939393939394,
      "grad_norm": 0.279538631439209,
      "learning_rate": 9.39e-06,
      "loss": 0.307,
      "step": 4061
    },
    {
      "epoch": 369.2909090909091,
      "grad_norm": 0.30089130997657776,
      "learning_rate": 9.38e-06,
      "loss": 0.3473,
      "step": 4062
    },
    {
      "epoch": 369.3878787878788,
      "grad_norm": 0.2933594286441803,
      "learning_rate": 9.370000000000002e-06,
      "loss": 0.3001,
      "step": 4063
    },
    {
      "epoch": 369.4848484848485,
      "grad_norm": 0.266652911901474,
      "learning_rate": 9.36e-06,
      "loss": 0.3355,
      "step": 4064
    },
    {
      "epoch": 369.58181818181816,
      "grad_norm": 0.25464680790901184,
      "learning_rate": 9.35e-06,
      "loss": 0.3101,
      "step": 4065
    },
    {
      "epoch": 369.6787878787879,
      "grad_norm": 0.2997622489929199,
      "learning_rate": 9.34e-06,
      "loss": 0.3191,
      "step": 4066
    },
    {
      "epoch": 369.77575757575755,
      "grad_norm": 0.3323077857494354,
      "learning_rate": 9.33e-06,
      "loss": 0.3544,
      "step": 4067
    },
    {
      "epoch": 369.8727272727273,
      "grad_norm": 0.29756712913513184,
      "learning_rate": 9.32e-06,
      "loss": 0.3342,
      "step": 4068
    },
    {
      "epoch": 369.969696969697,
      "grad_norm": 0.2877776026725769,
      "learning_rate": 9.31e-06,
      "loss": 0.3577,
      "step": 4069
    },
    {
      "epoch": 370.0,
      "grad_norm": 0.5464515686035156,
      "learning_rate": 9.3e-06,
      "loss": 0.328,
      "step": 4070
    },
    {
      "epoch": 370.0,
      "eval_loss": 0.4247572720050812,
      "eval_runtime": 2.1217,
      "eval_samples_per_second": 25.922,
      "eval_steps_per_second": 3.299,
      "step": 4070
    },
    {
      "epoch": 370.0969696969697,
      "grad_norm": 0.2983577847480774,
      "learning_rate": 9.29e-06,
      "loss": 0.3391,
      "step": 4071
    },
    {
      "epoch": 370.1939393939394,
      "grad_norm": 0.2986522614955902,
      "learning_rate": 9.28e-06,
      "loss": 0.3325,
      "step": 4072
    },
    {
      "epoch": 370.2909090909091,
      "grad_norm": 0.27389615774154663,
      "learning_rate": 9.270000000000001e-06,
      "loss": 0.3639,
      "step": 4073
    },
    {
      "epoch": 370.3878787878788,
      "grad_norm": 0.2886711061000824,
      "learning_rate": 9.260000000000001e-06,
      "loss": 0.3264,
      "step": 4074
    },
    {
      "epoch": 370.4848484848485,
      "grad_norm": 0.30563750863075256,
      "learning_rate": 9.25e-06,
      "loss": 0.3075,
      "step": 4075
    },
    {
      "epoch": 370.58181818181816,
      "grad_norm": 0.3458840548992157,
      "learning_rate": 9.24e-06,
      "loss": 0.3191,
      "step": 4076
    },
    {
      "epoch": 370.6787878787879,
      "grad_norm": 0.4566315710544586,
      "learning_rate": 9.23e-06,
      "loss": 0.3237,
      "step": 4077
    },
    {
      "epoch": 370.77575757575755,
      "grad_norm": 0.298952579498291,
      "learning_rate": 9.220000000000002e-06,
      "loss": 0.3557,
      "step": 4078
    },
    {
      "epoch": 370.8727272727273,
      "grad_norm": 0.2589184045791626,
      "learning_rate": 9.21e-06,
      "loss": 0.3147,
      "step": 4079
    },
    {
      "epoch": 370.969696969697,
      "grad_norm": 0.3013869822025299,
      "learning_rate": 9.2e-06,
      "loss": 0.3389,
      "step": 4080
    },
    {
      "epoch": 370.969696969697,
      "eval_loss": 0.42480072379112244,
      "eval_runtime": 2.1089,
      "eval_samples_per_second": 26.079,
      "eval_steps_per_second": 3.319,
      "step": 4080
    },
    {
      "epoch": 371.0,
      "grad_norm": 0.5611611604690552,
      "learning_rate": 9.19e-06,
      "loss": 0.3006,
      "step": 4081
    },
    {
      "epoch": 371.0969696969697,
      "grad_norm": 0.28300800919532776,
      "learning_rate": 9.180000000000002e-06,
      "loss": 0.3446,
      "step": 4082
    },
    {
      "epoch": 371.1939393939394,
      "grad_norm": 0.2626231908798218,
      "learning_rate": 9.17e-06,
      "loss": 0.3481,
      "step": 4083
    },
    {
      "epoch": 371.2909090909091,
      "grad_norm": 0.2636497914791107,
      "learning_rate": 9.16e-06,
      "loss": 0.3083,
      "step": 4084
    },
    {
      "epoch": 371.3878787878788,
      "grad_norm": 0.3330760598182678,
      "learning_rate": 9.15e-06,
      "loss": 0.3171,
      "step": 4085
    },
    {
      "epoch": 371.4848484848485,
      "grad_norm": 0.25505438446998596,
      "learning_rate": 9.14e-06,
      "loss": 0.3434,
      "step": 4086
    },
    {
      "epoch": 371.58181818181816,
      "grad_norm": 0.3272264003753662,
      "learning_rate": 9.13e-06,
      "loss": 0.2918,
      "step": 4087
    },
    {
      "epoch": 371.6787878787879,
      "grad_norm": 0.30581533908843994,
      "learning_rate": 9.12e-06,
      "loss": 0.3406,
      "step": 4088
    },
    {
      "epoch": 371.77575757575755,
      "grad_norm": 0.28679990768432617,
      "learning_rate": 9.110000000000001e-06,
      "loss": 0.345,
      "step": 4089
    },
    {
      "epoch": 371.8727272727273,
      "grad_norm": 0.46133214235305786,
      "learning_rate": 9.100000000000001e-06,
      "loss": 0.3456,
      "step": 4090
    },
    {
      "epoch": 371.8727272727273,
      "eval_loss": 0.4248599708080292,
      "eval_runtime": 2.1296,
      "eval_samples_per_second": 25.826,
      "eval_steps_per_second": 3.287,
      "step": 4090
    },
    {
      "epoch": 371.969696969697,
      "grad_norm": 0.2853439748287201,
      "learning_rate": 9.09e-06,
      "loss": 0.3206,
      "step": 4091
    },
    {
      "epoch": 372.0,
      "grad_norm": 0.548027515411377,
      "learning_rate": 9.080000000000001e-06,
      "loss": 0.3337,
      "step": 4092
    },
    {
      "epoch": 372.0969696969697,
      "grad_norm": 0.2653498351573944,
      "learning_rate": 9.070000000000001e-06,
      "loss": 0.3608,
      "step": 4093
    },
    {
      "epoch": 372.1939393939394,
      "grad_norm": 0.27974510192871094,
      "learning_rate": 9.06e-06,
      "loss": 0.3155,
      "step": 4094
    },
    {
      "epoch": 372.2909090909091,
      "grad_norm": 0.26262444257736206,
      "learning_rate": 9.05e-06,
      "loss": 0.3314,
      "step": 4095
    },
    {
      "epoch": 372.3878787878788,
      "grad_norm": 0.2681049704551697,
      "learning_rate": 9.04e-06,
      "loss": 0.2963,
      "step": 4096
    },
    {
      "epoch": 372.4848484848485,
      "grad_norm": 0.3194270730018616,
      "learning_rate": 9.030000000000002e-06,
      "loss": 0.2966,
      "step": 4097
    },
    {
      "epoch": 372.58181818181816,
      "grad_norm": 0.29719874262809753,
      "learning_rate": 9.02e-06,
      "loss": 0.3365,
      "step": 4098
    },
    {
      "epoch": 372.6787878787879,
      "grad_norm": 0.2969561815261841,
      "learning_rate": 9.01e-06,
      "loss": 0.354,
      "step": 4099
    },
    {
      "epoch": 372.77575757575755,
      "grad_norm": 0.31575682759284973,
      "learning_rate": 9e-06,
      "loss": 0.3704,
      "step": 4100
    },
    {
      "epoch": 372.77575757575755,
      "eval_loss": 0.4247245192527771,
      "eval_runtime": 2.1174,
      "eval_samples_per_second": 25.976,
      "eval_steps_per_second": 3.306,
      "step": 4100
    },
    {
      "epoch": 372.8727272727273,
      "grad_norm": 0.3249724507331848,
      "learning_rate": 8.99e-06,
      "loss": 0.311,
      "step": 4101
    },
    {
      "epoch": 372.969696969697,
      "grad_norm": 0.285019189119339,
      "learning_rate": 8.98e-06,
      "loss": 0.3534,
      "step": 4102
    },
    {
      "epoch": 373.0,
      "grad_norm": 0.617175281047821,
      "learning_rate": 8.97e-06,
      "loss": 0.2668,
      "step": 4103
    },
    {
      "epoch": 373.0969696969697,
      "grad_norm": 0.3166021406650543,
      "learning_rate": 8.96e-06,
      "loss": 0.3533,
      "step": 4104
    },
    {
      "epoch": 373.1939393939394,
      "grad_norm": 0.3377653658390045,
      "learning_rate": 8.95e-06,
      "loss": 0.3174,
      "step": 4105
    },
    {
      "epoch": 373.2909090909091,
      "grad_norm": 0.2525980472564697,
      "learning_rate": 8.939999999999999e-06,
      "loss": 0.3036,
      "step": 4106
    },
    {
      "epoch": 373.3878787878788,
      "grad_norm": 0.30190375447273254,
      "learning_rate": 8.930000000000001e-06,
      "loss": 0.3661,
      "step": 4107
    },
    {
      "epoch": 373.4848484848485,
      "grad_norm": 0.32537418603897095,
      "learning_rate": 8.920000000000001e-06,
      "loss": 0.3355,
      "step": 4108
    },
    {
      "epoch": 373.58181818181816,
      "grad_norm": 0.27809938788414,
      "learning_rate": 8.910000000000001e-06,
      "loss": 0.3138,
      "step": 4109
    },
    {
      "epoch": 373.6787878787879,
      "grad_norm": 0.298795610666275,
      "learning_rate": 8.9e-06,
      "loss": 0.3375,
      "step": 4110
    },
    {
      "epoch": 373.6787878787879,
      "eval_loss": 0.4250388443470001,
      "eval_runtime": 2.1164,
      "eval_samples_per_second": 25.988,
      "eval_steps_per_second": 3.308,
      "step": 4110
    },
    {
      "epoch": 373.77575757575755,
      "grad_norm": 0.32072368264198303,
      "learning_rate": 8.890000000000001e-06,
      "loss": 0.3474,
      "step": 4111
    },
    {
      "epoch": 373.8727272727273,
      "grad_norm": 0.3273388743400574,
      "learning_rate": 8.880000000000001e-06,
      "loss": 0.3195,
      "step": 4112
    },
    {
      "epoch": 373.969696969697,
      "grad_norm": 0.3104739785194397,
      "learning_rate": 8.87e-06,
      "loss": 0.2922,
      "step": 4113
    },
    {
      "epoch": 374.0,
      "grad_norm": 0.4112694263458252,
      "learning_rate": 8.86e-06,
      "loss": 0.3919,
      "step": 4114
    },
    {
      "epoch": 374.0969696969697,
      "grad_norm": 0.2891893982887268,
      "learning_rate": 8.85e-06,
      "loss": 0.384,
      "step": 4115
    },
    {
      "epoch": 374.1939393939394,
      "grad_norm": 0.2470197081565857,
      "learning_rate": 8.840000000000002e-06,
      "loss": 0.3046,
      "step": 4116
    },
    {
      "epoch": 374.2909090909091,
      "grad_norm": 0.24681566655635834,
      "learning_rate": 8.83e-06,
      "loss": 0.3291,
      "step": 4117
    },
    {
      "epoch": 374.3878787878788,
      "grad_norm": 0.3531782627105713,
      "learning_rate": 8.82e-06,
      "loss": 0.3499,
      "step": 4118
    },
    {
      "epoch": 374.4848484848485,
      "grad_norm": 0.31936323642730713,
      "learning_rate": 8.81e-06,
      "loss": 0.31,
      "step": 4119
    },
    {
      "epoch": 374.58181818181816,
      "grad_norm": 0.2687115967273712,
      "learning_rate": 8.8e-06,
      "loss": 0.3571,
      "step": 4120
    },
    {
      "epoch": 374.58181818181816,
      "eval_loss": 0.42443034052848816,
      "eval_runtime": 2.1256,
      "eval_samples_per_second": 25.875,
      "eval_steps_per_second": 3.293,
      "step": 4120
    },
    {
      "epoch": 374.6787878787879,
      "grad_norm": 0.345988392829895,
      "learning_rate": 8.79e-06,
      "loss": 0.343,
      "step": 4121
    },
    {
      "epoch": 374.77575757575755,
      "grad_norm": 0.3371879458427429,
      "learning_rate": 8.78e-06,
      "loss": 0.3026,
      "step": 4122
    },
    {
      "epoch": 374.8727272727273,
      "grad_norm": 0.27523255348205566,
      "learning_rate": 8.77e-06,
      "loss": 0.2955,
      "step": 4123
    },
    {
      "epoch": 374.969696969697,
      "grad_norm": 0.3101111054420471,
      "learning_rate": 8.76e-06,
      "loss": 0.3227,
      "step": 4124
    },
    {
      "epoch": 375.0,
      "grad_norm": 0.44706135988235474,
      "learning_rate": 8.75e-06,
      "loss": 0.3418,
      "step": 4125
    },
    {
      "epoch": 375.0969696969697,
      "grad_norm": 0.3133430480957031,
      "learning_rate": 8.740000000000001e-06,
      "loss": 0.3404,
      "step": 4126
    },
    {
      "epoch": 375.1939393939394,
      "grad_norm": 0.2988424599170685,
      "learning_rate": 8.730000000000001e-06,
      "loss": 0.3702,
      "step": 4127
    },
    {
      "epoch": 375.2909090909091,
      "grad_norm": 0.28647005558013916,
      "learning_rate": 8.720000000000001e-06,
      "loss": 0.3317,
      "step": 4128
    },
    {
      "epoch": 375.3878787878788,
      "grad_norm": 0.29622405767440796,
      "learning_rate": 8.71e-06,
      "loss": 0.3167,
      "step": 4129
    },
    {
      "epoch": 375.4848484848485,
      "grad_norm": 0.2939758002758026,
      "learning_rate": 8.7e-06,
      "loss": 0.3244,
      "step": 4130
    },
    {
      "epoch": 375.4848484848485,
      "eval_loss": 0.4248404800891876,
      "eval_runtime": 2.1331,
      "eval_samples_per_second": 25.784,
      "eval_steps_per_second": 3.282,
      "step": 4130
    },
    {
      "epoch": 375.58181818181816,
      "grad_norm": 0.25706255435943604,
      "learning_rate": 8.690000000000002e-06,
      "loss": 0.337,
      "step": 4131
    },
    {
      "epoch": 375.6787878787879,
      "grad_norm": 0.30878663063049316,
      "learning_rate": 8.68e-06,
      "loss": 0.2789,
      "step": 4132
    },
    {
      "epoch": 375.77575757575755,
      "grad_norm": 0.29647648334503174,
      "learning_rate": 8.67e-06,
      "loss": 0.348,
      "step": 4133
    },
    {
      "epoch": 375.8727272727273,
      "grad_norm": 0.31073883175849915,
      "learning_rate": 8.66e-06,
      "loss": 0.2972,
      "step": 4134
    },
    {
      "epoch": 375.969696969697,
      "grad_norm": 0.3301756978034973,
      "learning_rate": 8.65e-06,
      "loss": 0.3553,
      "step": 4135
    },
    {
      "epoch": 376.0,
      "grad_norm": 0.5006003975868225,
      "learning_rate": 8.64e-06,
      "loss": 0.347,
      "step": 4136
    },
    {
      "epoch": 376.0969696969697,
      "grad_norm": 0.2915913462638855,
      "learning_rate": 8.63e-06,
      "loss": 0.3644,
      "step": 4137
    },
    {
      "epoch": 376.1939393939394,
      "grad_norm": 0.26376426219940186,
      "learning_rate": 8.62e-06,
      "loss": 0.3347,
      "step": 4138
    },
    {
      "epoch": 376.2909090909091,
      "grad_norm": 0.31158575415611267,
      "learning_rate": 8.61e-06,
      "loss": 0.3291,
      "step": 4139
    },
    {
      "epoch": 376.3878787878788,
      "grad_norm": 0.2940891981124878,
      "learning_rate": 8.599999999999999e-06,
      "loss": 0.328,
      "step": 4140
    },
    {
      "epoch": 376.3878787878788,
      "eval_loss": 0.424710214138031,
      "eval_runtime": 2.1258,
      "eval_samples_per_second": 25.873,
      "eval_steps_per_second": 3.293,
      "step": 4140
    },
    {
      "epoch": 376.4848484848485,
      "grad_norm": 0.3219124376773834,
      "learning_rate": 8.59e-06,
      "loss": 0.3244,
      "step": 4141
    },
    {
      "epoch": 376.58181818181816,
      "grad_norm": 0.23037515580654144,
      "learning_rate": 8.580000000000001e-06,
      "loss": 0.3396,
      "step": 4142
    },
    {
      "epoch": 376.6787878787879,
      "grad_norm": 0.2835302948951721,
      "learning_rate": 8.570000000000001e-06,
      "loss": 0.3036,
      "step": 4143
    },
    {
      "epoch": 376.77575757575755,
      "grad_norm": 0.30571696162223816,
      "learning_rate": 8.56e-06,
      "loss": 0.3398,
      "step": 4144
    },
    {
      "epoch": 376.8727272727273,
      "grad_norm": 0.3565056622028351,
      "learning_rate": 8.550000000000001e-06,
      "loss": 0.3289,
      "step": 4145
    },
    {
      "epoch": 376.969696969697,
      "grad_norm": 0.30703380703926086,
      "learning_rate": 8.540000000000001e-06,
      "loss": 0.307,
      "step": 4146
    },
    {
      "epoch": 377.0,
      "grad_norm": 0.5746144652366638,
      "learning_rate": 8.53e-06,
      "loss": 0.3417,
      "step": 4147
    },
    {
      "epoch": 377.0969696969697,
      "grad_norm": 0.31260401010513306,
      "learning_rate": 8.52e-06,
      "loss": 0.3397,
      "step": 4148
    },
    {
      "epoch": 377.1939393939394,
      "grad_norm": 0.26729366183280945,
      "learning_rate": 8.51e-06,
      "loss": 0.324,
      "step": 4149
    },
    {
      "epoch": 377.2909090909091,
      "grad_norm": 0.31326571106910706,
      "learning_rate": 8.500000000000002e-06,
      "loss": 0.3467,
      "step": 4150
    },
    {
      "epoch": 377.2909090909091,
      "eval_loss": 0.42445996403694153,
      "eval_runtime": 2.1611,
      "eval_samples_per_second": 25.45,
      "eval_steps_per_second": 3.239,
      "step": 4150
    },
    {
      "epoch": 377.3878787878788,
      "grad_norm": 0.31573301553726196,
      "learning_rate": 8.49e-06,
      "loss": 0.3165,
      "step": 4151
    },
    {
      "epoch": 377.4848484848485,
      "grad_norm": 0.2820153534412384,
      "learning_rate": 8.48e-06,
      "loss": 0.3455,
      "step": 4152
    },
    {
      "epoch": 377.58181818181816,
      "grad_norm": 0.3372909128665924,
      "learning_rate": 8.47e-06,
      "loss": 0.3157,
      "step": 4153
    },
    {
      "epoch": 377.6787878787879,
      "grad_norm": 0.3248935043811798,
      "learning_rate": 8.46e-06,
      "loss": 0.3208,
      "step": 4154
    },
    {
      "epoch": 377.77575757575755,
      "grad_norm": 0.3556617796421051,
      "learning_rate": 8.45e-06,
      "loss": 0.3412,
      "step": 4155
    },
    {
      "epoch": 377.8727272727273,
      "grad_norm": 0.29159489274024963,
      "learning_rate": 8.44e-06,
      "loss": 0.2932,
      "step": 4156
    },
    {
      "epoch": 377.969696969697,
      "grad_norm": 0.27935653924942017,
      "learning_rate": 8.43e-06,
      "loss": 0.3583,
      "step": 4157
    },
    {
      "epoch": 378.0,
      "grad_norm": 0.45205578207969666,
      "learning_rate": 8.42e-06,
      "loss": 0.3296,
      "step": 4158
    },
    {
      "epoch": 378.0969696969697,
      "grad_norm": 0.349948525428772,
      "learning_rate": 8.409999999999999e-06,
      "loss": 0.3178,
      "step": 4159
    },
    {
      "epoch": 378.1939393939394,
      "grad_norm": 0.305345356464386,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.3089,
      "step": 4160
    },
    {
      "epoch": 378.1939393939394,
      "eval_loss": 0.4246861934661865,
      "eval_runtime": 2.1364,
      "eval_samples_per_second": 25.744,
      "eval_steps_per_second": 3.277,
      "step": 4160
    },
    {
      "epoch": 378.2909090909091,
      "grad_norm": 0.3020115792751312,
      "learning_rate": 8.390000000000001e-06,
      "loss": 0.3357,
      "step": 4161
    },
    {
      "epoch": 378.3878787878788,
      "grad_norm": 0.24768730998039246,
      "learning_rate": 8.380000000000001e-06,
      "loss": 0.2948,
      "step": 4162
    },
    {
      "epoch": 378.4848484848485,
      "grad_norm": 0.2680698037147522,
      "learning_rate": 8.37e-06,
      "loss": 0.3254,
      "step": 4163
    },
    {
      "epoch": 378.58181818181816,
      "grad_norm": 0.2712077796459198,
      "learning_rate": 8.36e-06,
      "loss": 0.3501,
      "step": 4164
    },
    {
      "epoch": 378.6787878787879,
      "grad_norm": 0.30424049496650696,
      "learning_rate": 8.350000000000001e-06,
      "loss": 0.2979,
      "step": 4165
    },
    {
      "epoch": 378.77575757575755,
      "grad_norm": 0.2615557312965393,
      "learning_rate": 8.34e-06,
      "loss": 0.3759,
      "step": 4166
    },
    {
      "epoch": 378.8727272727273,
      "grad_norm": 0.2612929940223694,
      "learning_rate": 8.33e-06,
      "loss": 0.3204,
      "step": 4167
    },
    {
      "epoch": 378.969696969697,
      "grad_norm": 0.34027451276779175,
      "learning_rate": 8.32e-06,
      "loss": 0.3751,
      "step": 4168
    },
    {
      "epoch": 379.0,
      "grad_norm": 0.4507916569709778,
      "learning_rate": 8.31e-06,
      "loss": 0.306,
      "step": 4169
    },
    {
      "epoch": 379.0969696969697,
      "grad_norm": 0.3092374801635742,
      "learning_rate": 8.3e-06,
      "loss": 0.333,
      "step": 4170
    },
    {
      "epoch": 379.0969696969697,
      "eval_loss": 0.4247243404388428,
      "eval_runtime": 2.1965,
      "eval_samples_per_second": 25.04,
      "eval_steps_per_second": 3.187,
      "step": 4170
    },
    {
      "epoch": 379.1939393939394,
      "grad_norm": 0.264586478471756,
      "learning_rate": 8.29e-06,
      "loss": 0.3343,
      "step": 4171
    },
    {
      "epoch": 379.2909090909091,
      "grad_norm": 0.3064897060394287,
      "learning_rate": 8.28e-06,
      "loss": 0.3234,
      "step": 4172
    },
    {
      "epoch": 379.3878787878788,
      "grad_norm": 0.33171793818473816,
      "learning_rate": 8.27e-06,
      "loss": 0.3462,
      "step": 4173
    },
    {
      "epoch": 379.4848484848485,
      "grad_norm": 0.28818413615226746,
      "learning_rate": 8.26e-06,
      "loss": 0.3108,
      "step": 4174
    },
    {
      "epoch": 379.58181818181816,
      "grad_norm": 0.3123342990875244,
      "learning_rate": 8.25e-06,
      "loss": 0.3585,
      "step": 4175
    },
    {
      "epoch": 379.6787878787879,
      "grad_norm": 0.3093242347240448,
      "learning_rate": 8.24e-06,
      "loss": 0.3032,
      "step": 4176
    },
    {
      "epoch": 379.77575757575755,
      "grad_norm": 0.28044188022613525,
      "learning_rate": 8.23e-06,
      "loss": 0.3217,
      "step": 4177
    },
    {
      "epoch": 379.8727272727273,
      "grad_norm": 0.3285064399242401,
      "learning_rate": 8.22e-06,
      "loss": 0.3583,
      "step": 4178
    },
    {
      "epoch": 379.969696969697,
      "grad_norm": 0.29143503308296204,
      "learning_rate": 8.210000000000001e-06,
      "loss": 0.3071,
      "step": 4179
    },
    {
      "epoch": 380.0,
      "grad_norm": 0.5048023462295532,
      "learning_rate": 8.200000000000001e-06,
      "loss": 0.3354,
      "step": 4180
    },
    {
      "epoch": 380.0,
      "eval_loss": 0.4245156943798065,
      "eval_runtime": 2.1402,
      "eval_samples_per_second": 25.698,
      "eval_steps_per_second": 3.271,
      "step": 4180
    },
    {
      "epoch": 380.0969696969697,
      "grad_norm": 0.269228458404541,
      "learning_rate": 8.190000000000001e-06,
      "loss": 0.3459,
      "step": 4181
    },
    {
      "epoch": 380.1939393939394,
      "grad_norm": 0.2466476857662201,
      "learning_rate": 8.18e-06,
      "loss": 0.3371,
      "step": 4182
    },
    {
      "epoch": 380.2909090909091,
      "grad_norm": 0.29603132605552673,
      "learning_rate": 8.17e-06,
      "loss": 0.3512,
      "step": 4183
    },
    {
      "epoch": 380.3878787878788,
      "grad_norm": 0.29953810572624207,
      "learning_rate": 8.160000000000001e-06,
      "loss": 0.3179,
      "step": 4184
    },
    {
      "epoch": 380.4848484848485,
      "grad_norm": 0.2953645586967468,
      "learning_rate": 8.15e-06,
      "loss": 0.3459,
      "step": 4185
    },
    {
      "epoch": 380.58181818181816,
      "grad_norm": 0.3172391355037689,
      "learning_rate": 8.14e-06,
      "loss": 0.3277,
      "step": 4186
    },
    {
      "epoch": 380.6787878787879,
      "grad_norm": 0.29355430603027344,
      "learning_rate": 8.13e-06,
      "loss": 0.3377,
      "step": 4187
    },
    {
      "epoch": 380.77575757575755,
      "grad_norm": 0.3100706934928894,
      "learning_rate": 8.12e-06,
      "loss": 0.3176,
      "step": 4188
    },
    {
      "epoch": 380.8727272727273,
      "grad_norm": 0.27939969301223755,
      "learning_rate": 8.11e-06,
      "loss": 0.3162,
      "step": 4189
    },
    {
      "epoch": 380.969696969697,
      "grad_norm": 0.2738233804702759,
      "learning_rate": 8.1e-06,
      "loss": 0.3127,
      "step": 4190
    },
    {
      "epoch": 380.969696969697,
      "eval_loss": 0.4244054853916168,
      "eval_runtime": 2.1388,
      "eval_samples_per_second": 25.716,
      "eval_steps_per_second": 3.273,
      "step": 4190
    },
    {
      "epoch": 381.0,
      "grad_norm": 0.4269275367259979,
      "learning_rate": 8.09e-06,
      "loss": 0.2913,
      "step": 4191
    },
    {
      "epoch": 381.0969696969697,
      "grad_norm": 0.27549001574516296,
      "learning_rate": 8.08e-06,
      "loss": 0.3624,
      "step": 4192
    },
    {
      "epoch": 381.1939393939394,
      "grad_norm": 0.3019556999206543,
      "learning_rate": 8.069999999999999e-06,
      "loss": 0.3107,
      "step": 4193
    },
    {
      "epoch": 381.2909090909091,
      "grad_norm": 0.2658679485321045,
      "learning_rate": 8.06e-06,
      "loss": 0.327,
      "step": 4194
    },
    {
      "epoch": 381.3878787878788,
      "grad_norm": 0.2855111062526703,
      "learning_rate": 8.050000000000001e-06,
      "loss": 0.3253,
      "step": 4195
    },
    {
      "epoch": 381.4848484848485,
      "grad_norm": 0.3003518879413605,
      "learning_rate": 8.040000000000001e-06,
      "loss": 0.3105,
      "step": 4196
    },
    {
      "epoch": 381.58181818181816,
      "grad_norm": 0.31421568989753723,
      "learning_rate": 8.03e-06,
      "loss": 0.3181,
      "step": 4197
    },
    {
      "epoch": 381.6787878787879,
      "grad_norm": 0.257122278213501,
      "learning_rate": 8.02e-06,
      "loss": 0.3346,
      "step": 4198
    },
    {
      "epoch": 381.77575757575755,
      "grad_norm": 0.2668331265449524,
      "learning_rate": 8.010000000000001e-06,
      "loss": 0.3353,
      "step": 4199
    },
    {
      "epoch": 381.8727272727273,
      "grad_norm": 0.2797575294971466,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.3628,
      "step": 4200
    },
    {
      "epoch": 381.8727272727273,
      "eval_loss": 0.42448917031288147,
      "eval_runtime": 2.1343,
      "eval_samples_per_second": 25.769,
      "eval_steps_per_second": 3.28,
      "step": 4200
    },
    {
      "epoch": 381.969696969697,
      "grad_norm": 0.3442399203777313,
      "learning_rate": 7.99e-06,
      "loss": 0.3209,
      "step": 4201
    },
    {
      "epoch": 382.0,
      "grad_norm": 0.5818289518356323,
      "learning_rate": 7.98e-06,
      "loss": 0.297,
      "step": 4202
    },
    {
      "epoch": 382.0969696969697,
      "grad_norm": 0.3574053645133972,
      "learning_rate": 7.97e-06,
      "loss": 0.3459,
      "step": 4203
    },
    {
      "epoch": 382.1939393939394,
      "grad_norm": 0.3037893772125244,
      "learning_rate": 7.96e-06,
      "loss": 0.3341,
      "step": 4204
    },
    {
      "epoch": 382.2909090909091,
      "grad_norm": 0.29630377888679504,
      "learning_rate": 7.95e-06,
      "loss": 0.3139,
      "step": 4205
    },
    {
      "epoch": 382.3878787878788,
      "grad_norm": 0.2979125380516052,
      "learning_rate": 7.94e-06,
      "loss": 0.3363,
      "step": 4206
    },
    {
      "epoch": 382.4848484848485,
      "grad_norm": 0.320667028427124,
      "learning_rate": 7.93e-06,
      "loss": 0.3347,
      "step": 4207
    },
    {
      "epoch": 382.58181818181816,
      "grad_norm": 0.30430322885513306,
      "learning_rate": 7.92e-06,
      "loss": 0.3295,
      "step": 4208
    },
    {
      "epoch": 382.6787878787879,
      "grad_norm": 0.2292478084564209,
      "learning_rate": 7.91e-06,
      "loss": 0.3027,
      "step": 4209
    },
    {
      "epoch": 382.77575757575755,
      "grad_norm": 0.27211201190948486,
      "learning_rate": 7.9e-06,
      "loss": 0.3278,
      "step": 4210
    },
    {
      "epoch": 382.77575757575755,
      "eval_loss": 0.4243857264518738,
      "eval_runtime": 2.1481,
      "eval_samples_per_second": 25.604,
      "eval_steps_per_second": 3.259,
      "step": 4210
    },
    {
      "epoch": 382.8727272727273,
      "grad_norm": 0.3260558247566223,
      "learning_rate": 7.89e-06,
      "loss": 0.3245,
      "step": 4211
    },
    {
      "epoch": 382.969696969697,
      "grad_norm": 0.29133886098861694,
      "learning_rate": 7.879999999999999e-06,
      "loss": 0.3401,
      "step": 4212
    },
    {
      "epoch": 383.0,
      "grad_norm": 0.4845393896102905,
      "learning_rate": 7.870000000000001e-06,
      "loss": 0.3478,
      "step": 4213
    },
    {
      "epoch": 383.0969696969697,
      "grad_norm": 0.3398871421813965,
      "learning_rate": 7.860000000000001e-06,
      "loss": 0.3212,
      "step": 4214
    },
    {
      "epoch": 383.1939393939394,
      "grad_norm": 0.2844429016113281,
      "learning_rate": 7.850000000000001e-06,
      "loss": 0.3588,
      "step": 4215
    },
    {
      "epoch": 383.2909090909091,
      "grad_norm": 0.3390371799468994,
      "learning_rate": 7.84e-06,
      "loss": 0.2931,
      "step": 4216
    },
    {
      "epoch": 383.3878787878788,
      "grad_norm": 0.29903873801231384,
      "learning_rate": 7.83e-06,
      "loss": 0.3449,
      "step": 4217
    },
    {
      "epoch": 383.4848484848485,
      "grad_norm": 0.244550421833992,
      "learning_rate": 7.820000000000001e-06,
      "loss": 0.3379,
      "step": 4218
    },
    {
      "epoch": 383.58181818181816,
      "grad_norm": 0.28961002826690674,
      "learning_rate": 7.810000000000001e-06,
      "loss": 0.3522,
      "step": 4219
    },
    {
      "epoch": 383.6787878787879,
      "grad_norm": 0.2954062521457672,
      "learning_rate": 7.8e-06,
      "loss": 0.269,
      "step": 4220
    },
    {
      "epoch": 383.6787878787879,
      "eval_loss": 0.4247341454029083,
      "eval_runtime": 2.1211,
      "eval_samples_per_second": 25.93,
      "eval_steps_per_second": 3.3,
      "step": 4220
    },
    {
      "epoch": 383.77575757575755,
      "grad_norm": 0.2589881718158722,
      "learning_rate": 7.79e-06,
      "loss": 0.3359,
      "step": 4221
    },
    {
      "epoch": 383.8727272727273,
      "grad_norm": 0.3851045072078705,
      "learning_rate": 7.78e-06,
      "loss": 0.3301,
      "step": 4222
    },
    {
      "epoch": 383.969696969697,
      "grad_norm": 0.26817086338996887,
      "learning_rate": 7.77e-06,
      "loss": 0.3367,
      "step": 4223
    },
    {
      "epoch": 384.0,
      "grad_norm": 0.5171042084693909,
      "learning_rate": 7.76e-06,
      "loss": 0.3863,
      "step": 4224
    },
    {
      "epoch": 384.0969696969697,
      "grad_norm": 0.3027506172657013,
      "learning_rate": 7.75e-06,
      "loss": 0.3258,
      "step": 4225
    },
    {
      "epoch": 384.1939393939394,
      "grad_norm": 0.27451062202453613,
      "learning_rate": 7.74e-06,
      "loss": 0.3417,
      "step": 4226
    },
    {
      "epoch": 384.2909090909091,
      "grad_norm": 0.29753100872039795,
      "learning_rate": 7.73e-06,
      "loss": 0.331,
      "step": 4227
    },
    {
      "epoch": 384.3878787878788,
      "grad_norm": 0.31794247031211853,
      "learning_rate": 7.72e-06,
      "loss": 0.308,
      "step": 4228
    },
    {
      "epoch": 384.4848484848485,
      "grad_norm": 0.30627694725990295,
      "learning_rate": 7.71e-06,
      "loss": 0.3245,
      "step": 4229
    },
    {
      "epoch": 384.58181818181816,
      "grad_norm": 0.3163098692893982,
      "learning_rate": 7.7e-06,
      "loss": 0.3194,
      "step": 4230
    },
    {
      "epoch": 384.58181818181816,
      "eval_loss": 0.424343466758728,
      "eval_runtime": 2.13,
      "eval_samples_per_second": 25.821,
      "eval_steps_per_second": 3.286,
      "step": 4230
    },
    {
      "epoch": 384.6787878787879,
      "grad_norm": 0.2726805806159973,
      "learning_rate": 7.69e-06,
      "loss": 0.341,
      "step": 4231
    },
    {
      "epoch": 384.77575757575755,
      "grad_norm": 0.28615909814834595,
      "learning_rate": 7.68e-06,
      "loss": 0.3394,
      "step": 4232
    },
    {
      "epoch": 384.8727272727273,
      "grad_norm": 0.3741395175457001,
      "learning_rate": 7.670000000000001e-06,
      "loss": 0.337,
      "step": 4233
    },
    {
      "epoch": 384.969696969697,
      "grad_norm": 0.25299376249313354,
      "learning_rate": 7.660000000000001e-06,
      "loss": 0.3416,
      "step": 4234
    },
    {
      "epoch": 385.0,
      "grad_norm": 0.5409793853759766,
      "learning_rate": 7.65e-06,
      "loss": 0.2778,
      "step": 4235
    },
    {
      "epoch": 385.0969696969697,
      "grad_norm": 0.38123613595962524,
      "learning_rate": 7.64e-06,
      "loss": 0.3322,
      "step": 4236
    },
    {
      "epoch": 385.1939393939394,
      "grad_norm": 0.3055660128593445,
      "learning_rate": 7.630000000000001e-06,
      "loss": 0.357,
      "step": 4237
    },
    {
      "epoch": 385.2909090909091,
      "grad_norm": 0.285840779542923,
      "learning_rate": 7.620000000000001e-06,
      "loss": 0.3618,
      "step": 4238
    },
    {
      "epoch": 385.3878787878788,
      "grad_norm": 0.289868026971817,
      "learning_rate": 7.610000000000001e-06,
      "loss": 0.3113,
      "step": 4239
    },
    {
      "epoch": 385.4848484848485,
      "grad_norm": 0.26244574785232544,
      "learning_rate": 7.6e-06,
      "loss": 0.3111,
      "step": 4240
    },
    {
      "epoch": 385.4848484848485,
      "eval_loss": 0.4244261085987091,
      "eval_runtime": 2.1304,
      "eval_samples_per_second": 25.817,
      "eval_steps_per_second": 3.286,
      "step": 4240
    },
    {
      "epoch": 385.58181818181816,
      "grad_norm": 0.3094077408313751,
      "learning_rate": 7.59e-06,
      "loss": 0.3327,
      "step": 4241
    },
    {
      "epoch": 385.6787878787879,
      "grad_norm": 0.27038809657096863,
      "learning_rate": 7.580000000000001e-06,
      "loss": 0.3015,
      "step": 4242
    },
    {
      "epoch": 385.77575757575755,
      "grad_norm": 0.2761613428592682,
      "learning_rate": 7.57e-06,
      "loss": 0.2914,
      "step": 4243
    },
    {
      "epoch": 385.8727272727273,
      "grad_norm": 0.3535533845424652,
      "learning_rate": 7.5600000000000005e-06,
      "loss": 0.3411,
      "step": 4244
    },
    {
      "epoch": 385.969696969697,
      "grad_norm": 0.2971293032169342,
      "learning_rate": 7.55e-06,
      "loss": 0.3395,
      "step": 4245
    },
    {
      "epoch": 386.0,
      "grad_norm": 0.5061172246932983,
      "learning_rate": 7.54e-06,
      "loss": 0.3751,
      "step": 4246
    },
    {
      "epoch": 386.0969696969697,
      "grad_norm": 0.3000218868255615,
      "learning_rate": 7.530000000000001e-06,
      "loss": 0.3364,
      "step": 4247
    },
    {
      "epoch": 386.1939393939394,
      "grad_norm": 0.2683129906654358,
      "learning_rate": 7.520000000000001e-06,
      "loss": 0.3172,
      "step": 4248
    },
    {
      "epoch": 386.2909090909091,
      "grad_norm": 0.26465749740600586,
      "learning_rate": 7.51e-06,
      "loss": 0.2949,
      "step": 4249
    },
    {
      "epoch": 386.3878787878788,
      "grad_norm": 0.2655685544013977,
      "learning_rate": 7.5e-06,
      "loss": 0.3697,
      "step": 4250
    },
    {
      "epoch": 386.3878787878788,
      "eval_loss": 0.424681156873703,
      "eval_runtime": 2.1213,
      "eval_samples_per_second": 25.927,
      "eval_steps_per_second": 3.3,
      "step": 4250
    },
    {
      "epoch": 386.4848484848485,
      "grad_norm": 0.3396849036216736,
      "learning_rate": 7.4899999999999994e-06,
      "loss": 0.3515,
      "step": 4251
    },
    {
      "epoch": 386.58181818181816,
      "grad_norm": 0.31697672605514526,
      "learning_rate": 7.480000000000001e-06,
      "loss": 0.3149,
      "step": 4252
    },
    {
      "epoch": 386.6787878787879,
      "grad_norm": 0.2642710208892822,
      "learning_rate": 7.4700000000000005e-06,
      "loss": 0.2963,
      "step": 4253
    },
    {
      "epoch": 386.77575757575755,
      "grad_norm": 0.3324968218803406,
      "learning_rate": 7.4600000000000006e-06,
      "loss": 0.3494,
      "step": 4254
    },
    {
      "epoch": 386.8727272727273,
      "grad_norm": 0.3071610629558563,
      "learning_rate": 7.45e-06,
      "loss": 0.3504,
      "step": 4255
    },
    {
      "epoch": 386.969696969697,
      "grad_norm": 0.25106191635131836,
      "learning_rate": 7.44e-06,
      "loss": 0.3138,
      "step": 4256
    },
    {
      "epoch": 387.0,
      "grad_norm": 0.4752749800682068,
      "learning_rate": 7.430000000000001e-06,
      "loss": 0.3195,
      "step": 4257
    },
    {
      "epoch": 387.0969696969697,
      "grad_norm": 0.2990110218524933,
      "learning_rate": 7.420000000000001e-06,
      "loss": 0.2864,
      "step": 4258
    },
    {
      "epoch": 387.1939393939394,
      "grad_norm": 0.2590501010417938,
      "learning_rate": 7.41e-06,
      "loss": 0.3396,
      "step": 4259
    },
    {
      "epoch": 387.2909090909091,
      "grad_norm": 0.292054146528244,
      "learning_rate": 7.4e-06,
      "loss": 0.3605,
      "step": 4260
    },
    {
      "epoch": 387.2909090909091,
      "eval_loss": 0.4244232773780823,
      "eval_runtime": 2.13,
      "eval_samples_per_second": 25.822,
      "eval_steps_per_second": 3.286,
      "step": 4260
    },
    {
      "epoch": 387.3878787878788,
      "grad_norm": 0.31396231055259705,
      "learning_rate": 7.3899999999999995e-06,
      "loss": 0.3332,
      "step": 4261
    },
    {
      "epoch": 387.4848484848485,
      "grad_norm": 0.46617692708969116,
      "learning_rate": 7.3800000000000005e-06,
      "loss": 0.3425,
      "step": 4262
    },
    {
      "epoch": 387.58181818181816,
      "grad_norm": 0.3220658004283905,
      "learning_rate": 7.370000000000001e-06,
      "loss": 0.3119,
      "step": 4263
    },
    {
      "epoch": 387.6787878787879,
      "grad_norm": 0.2932121157646179,
      "learning_rate": 7.36e-06,
      "loss": 0.3344,
      "step": 4264
    },
    {
      "epoch": 387.77575757575755,
      "grad_norm": 0.3561127483844757,
      "learning_rate": 7.35e-06,
      "loss": 0.3304,
      "step": 4265
    },
    {
      "epoch": 387.8727272727273,
      "grad_norm": 0.316982626914978,
      "learning_rate": 7.340000000000001e-06,
      "loss": 0.3156,
      "step": 4266
    },
    {
      "epoch": 387.969696969697,
      "grad_norm": 0.34779998660087585,
      "learning_rate": 7.330000000000001e-06,
      "loss": 0.3207,
      "step": 4267
    },
    {
      "epoch": 388.0,
      "grad_norm": 0.618858277797699,
      "learning_rate": 7.32e-06,
      "loss": 0.3917,
      "step": 4268
    },
    {
      "epoch": 388.0969696969697,
      "grad_norm": 0.27004459500312805,
      "learning_rate": 7.31e-06,
      "loss": 0.3537,
      "step": 4269
    },
    {
      "epoch": 388.1939393939394,
      "grad_norm": 0.3083222508430481,
      "learning_rate": 7.2999999999999996e-06,
      "loss": 0.3302,
      "step": 4270
    },
    {
      "epoch": 388.1939393939394,
      "eval_loss": 0.4242541193962097,
      "eval_runtime": 2.1265,
      "eval_samples_per_second": 25.864,
      "eval_steps_per_second": 3.292,
      "step": 4270
    },
    {
      "epoch": 388.2909090909091,
      "grad_norm": 0.2764505445957184,
      "learning_rate": 7.290000000000001e-06,
      "loss": 0.3211,
      "step": 4271
    },
    {
      "epoch": 388.3878787878788,
      "grad_norm": 0.2886534333229065,
      "learning_rate": 7.280000000000001e-06,
      "loss": 0.3287,
      "step": 4272
    },
    {
      "epoch": 388.4848484848485,
      "grad_norm": 0.32928892970085144,
      "learning_rate": 7.270000000000001e-06,
      "loss": 0.3167,
      "step": 4273
    },
    {
      "epoch": 388.58181818181816,
      "grad_norm": 0.2919245660305023,
      "learning_rate": 7.26e-06,
      "loss": 0.337,
      "step": 4274
    },
    {
      "epoch": 388.6787878787879,
      "grad_norm": 0.26534411311149597,
      "learning_rate": 7.25e-06,
      "loss": 0.303,
      "step": 4275
    },
    {
      "epoch": 388.77575757575755,
      "grad_norm": 0.31570714712142944,
      "learning_rate": 7.240000000000001e-06,
      "loss": 0.3339,
      "step": 4276
    },
    {
      "epoch": 388.8727272727273,
      "grad_norm": 0.2628118395805359,
      "learning_rate": 7.230000000000001e-06,
      "loss": 0.3206,
      "step": 4277
    },
    {
      "epoch": 388.969696969697,
      "grad_norm": 0.29525238275527954,
      "learning_rate": 7.22e-06,
      "loss": 0.3514,
      "step": 4278
    },
    {
      "epoch": 389.0,
      "grad_norm": 0.4390280842781067,
      "learning_rate": 7.2100000000000004e-06,
      "loss": 0.3143,
      "step": 4279
    },
    {
      "epoch": 389.0969696969697,
      "grad_norm": 0.27595826983451843,
      "learning_rate": 7.2e-06,
      "loss": 0.3055,
      "step": 4280
    },
    {
      "epoch": 389.0969696969697,
      "eval_loss": 0.4243329167366028,
      "eval_runtime": 2.1301,
      "eval_samples_per_second": 25.82,
      "eval_steps_per_second": 3.286,
      "step": 4280
    },
    {
      "epoch": 389.1939393939394,
      "grad_norm": 0.2787480056285858,
      "learning_rate": 7.190000000000001e-06,
      "loss": 0.3158,
      "step": 4281
    },
    {
      "epoch": 389.2909090909091,
      "grad_norm": 0.3227105736732483,
      "learning_rate": 7.180000000000001e-06,
      "loss": 0.3747,
      "step": 4282
    },
    {
      "epoch": 389.3878787878788,
      "grad_norm": 0.29996225237846375,
      "learning_rate": 7.17e-06,
      "loss": 0.3398,
      "step": 4283
    },
    {
      "epoch": 389.4848484848485,
      "grad_norm": 0.269465833902359,
      "learning_rate": 7.16e-06,
      "loss": 0.2991,
      "step": 4284
    },
    {
      "epoch": 389.58181818181816,
      "grad_norm": 0.3202161490917206,
      "learning_rate": 7.15e-06,
      "loss": 0.335,
      "step": 4285
    },
    {
      "epoch": 389.6787878787879,
      "grad_norm": 0.24681954085826874,
      "learning_rate": 7.140000000000001e-06,
      "loss": 0.3572,
      "step": 4286
    },
    {
      "epoch": 389.77575757575755,
      "grad_norm": 0.2774510681629181,
      "learning_rate": 7.13e-06,
      "loss": 0.3156,
      "step": 4287
    },
    {
      "epoch": 389.8727272727273,
      "grad_norm": 0.3508288562297821,
      "learning_rate": 7.1200000000000004e-06,
      "loss": 0.3452,
      "step": 4288
    },
    {
      "epoch": 389.969696969697,
      "grad_norm": 0.27009424567222595,
      "learning_rate": 7.11e-06,
      "loss": 0.2825,
      "step": 4289
    },
    {
      "epoch": 390.0,
      "grad_norm": 0.6318101286888123,
      "learning_rate": 7.1e-06,
      "loss": 0.4019,
      "step": 4290
    },
    {
      "epoch": 390.0,
      "eval_loss": 0.42439505457878113,
      "eval_runtime": 2.1324,
      "eval_samples_per_second": 25.793,
      "eval_steps_per_second": 3.283,
      "step": 4290
    },
    {
      "epoch": 390.0969696969697,
      "grad_norm": 0.26486244797706604,
      "learning_rate": 7.090000000000001e-06,
      "loss": 0.3427,
      "step": 4291
    },
    {
      "epoch": 390.1939393939394,
      "grad_norm": 0.3479774296283722,
      "learning_rate": 7.080000000000001e-06,
      "loss": 0.3538,
      "step": 4292
    },
    {
      "epoch": 390.2909090909091,
      "grad_norm": 0.2837066352367401,
      "learning_rate": 7.07e-06,
      "loss": 0.3171,
      "step": 4293
    },
    {
      "epoch": 390.3878787878788,
      "grad_norm": 0.2607334852218628,
      "learning_rate": 7.06e-06,
      "loss": 0.2954,
      "step": 4294
    },
    {
      "epoch": 390.4848484848485,
      "grad_norm": 0.3070979118347168,
      "learning_rate": 7.049999999999999e-06,
      "loss": 0.3584,
      "step": 4295
    },
    {
      "epoch": 390.58181818181816,
      "grad_norm": 0.3067142963409424,
      "learning_rate": 7.04e-06,
      "loss": 0.3348,
      "step": 4296
    },
    {
      "epoch": 390.6787878787879,
      "grad_norm": 0.2787237763404846,
      "learning_rate": 7.0300000000000005e-06,
      "loss": 0.3376,
      "step": 4297
    },
    {
      "epoch": 390.77575757575755,
      "grad_norm": 0.2802588641643524,
      "learning_rate": 7.0200000000000006e-06,
      "loss": 0.3186,
      "step": 4298
    },
    {
      "epoch": 390.8727272727273,
      "grad_norm": 0.2827461361885071,
      "learning_rate": 7.01e-06,
      "loss": 0.3325,
      "step": 4299
    },
    {
      "epoch": 390.969696969697,
      "grad_norm": 0.3040320575237274,
      "learning_rate": 7.000000000000001e-06,
      "loss": 0.3013,
      "step": 4300
    },
    {
      "epoch": 390.969696969697,
      "eval_loss": 0.42447808384895325,
      "eval_runtime": 2.1287,
      "eval_samples_per_second": 25.837,
      "eval_steps_per_second": 3.288,
      "step": 4300
    },
    {
      "epoch": 391.0,
      "grad_norm": 0.3805631697177887,
      "learning_rate": 6.990000000000001e-06,
      "loss": 0.3153,
      "step": 4301
    },
    {
      "epoch": 391.0969696969697,
      "grad_norm": 0.30007538199424744,
      "learning_rate": 6.98e-06,
      "loss": 0.3193,
      "step": 4302
    },
    {
      "epoch": 391.1939393939394,
      "grad_norm": 0.25372499227523804,
      "learning_rate": 6.97e-06,
      "loss": 0.3496,
      "step": 4303
    },
    {
      "epoch": 391.2909090909091,
      "grad_norm": 0.5099557042121887,
      "learning_rate": 6.9599999999999994e-06,
      "loss": 0.2949,
      "step": 4304
    },
    {
      "epoch": 391.3878787878788,
      "grad_norm": 0.2995280623435974,
      "learning_rate": 6.950000000000001e-06,
      "loss": 0.3355,
      "step": 4305
    },
    {
      "epoch": 391.4848484848485,
      "grad_norm": 0.26424339413642883,
      "learning_rate": 6.9400000000000005e-06,
      "loss": 0.358,
      "step": 4306
    },
    {
      "epoch": 391.58181818181816,
      "grad_norm": 0.33444005250930786,
      "learning_rate": 6.9300000000000006e-06,
      "loss": 0.3335,
      "step": 4307
    },
    {
      "epoch": 391.6787878787879,
      "grad_norm": 0.29056233167648315,
      "learning_rate": 6.92e-06,
      "loss": 0.3384,
      "step": 4308
    },
    {
      "epoch": 391.77575757575755,
      "grad_norm": 0.3315895199775696,
      "learning_rate": 6.91e-06,
      "loss": 0.3234,
      "step": 4309
    },
    {
      "epoch": 391.8727272727273,
      "grad_norm": 0.2590215504169464,
      "learning_rate": 6.900000000000001e-06,
      "loss": 0.3019,
      "step": 4310
    },
    {
      "epoch": 391.8727272727273,
      "eval_loss": 0.42437347769737244,
      "eval_runtime": 2.1192,
      "eval_samples_per_second": 25.953,
      "eval_steps_per_second": 3.303,
      "step": 4310
    },
    {
      "epoch": 391.969696969697,
      "grad_norm": 0.28518152236938477,
      "learning_rate": 6.890000000000001e-06,
      "loss": 0.3496,
      "step": 4311
    },
    {
      "epoch": 392.0,
      "grad_norm": 0.5114558339118958,
      "learning_rate": 6.88e-06,
      "loss": 0.2863,
      "step": 4312
    },
    {
      "epoch": 392.0969696969697,
      "grad_norm": 0.2683810293674469,
      "learning_rate": 6.87e-06,
      "loss": 0.3313,
      "step": 4313
    },
    {
      "epoch": 392.1939393939394,
      "grad_norm": 0.29235902428627014,
      "learning_rate": 6.8599999999999995e-06,
      "loss": 0.3362,
      "step": 4314
    },
    {
      "epoch": 392.2909090909091,
      "grad_norm": 0.2714664041996002,
      "learning_rate": 6.8500000000000005e-06,
      "loss": 0.3688,
      "step": 4315
    },
    {
      "epoch": 392.3878787878788,
      "grad_norm": 0.2590443193912506,
      "learning_rate": 6.840000000000001e-06,
      "loss": 0.3377,
      "step": 4316
    },
    {
      "epoch": 392.4848484848485,
      "grad_norm": 0.32732295989990234,
      "learning_rate": 6.830000000000001e-06,
      "loss": 0.3176,
      "step": 4317
    },
    {
      "epoch": 392.58181818181816,
      "grad_norm": 0.2644732594490051,
      "learning_rate": 6.82e-06,
      "loss": 0.3194,
      "step": 4318
    },
    {
      "epoch": 392.6787878787879,
      "grad_norm": 0.29884040355682373,
      "learning_rate": 6.81e-06,
      "loss": 0.3386,
      "step": 4319
    },
    {
      "epoch": 392.77575757575755,
      "grad_norm": 0.3094496428966522,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.327,
      "step": 4320
    },
    {
      "epoch": 392.77575757575755,
      "eval_loss": 0.42413949966430664,
      "eval_runtime": 2.1306,
      "eval_samples_per_second": 25.814,
      "eval_steps_per_second": 3.285,
      "step": 4320
    },
    {
      "epoch": 392.8727272727273,
      "grad_norm": 0.2963372766971588,
      "learning_rate": 6.79e-06,
      "loss": 0.2948,
      "step": 4321
    },
    {
      "epoch": 392.969696969697,
      "grad_norm": 0.25973087549209595,
      "learning_rate": 6.78e-06,
      "loss": 0.3123,
      "step": 4322
    },
    {
      "epoch": 393.0,
      "grad_norm": 0.8426999449729919,
      "learning_rate": 6.7699999999999996e-06,
      "loss": 0.3366,
      "step": 4323
    },
    {
      "epoch": 393.0969696969697,
      "grad_norm": 0.26413246989250183,
      "learning_rate": 6.76e-06,
      "loss": 0.362,
      "step": 4324
    },
    {
      "epoch": 393.1939393939394,
      "grad_norm": 0.31809934973716736,
      "learning_rate": 6.750000000000001e-06,
      "loss": 0.3461,
      "step": 4325
    },
    {
      "epoch": 393.2909090909091,
      "grad_norm": 0.28155213594436646,
      "learning_rate": 6.740000000000001e-06,
      "loss": 0.3144,
      "step": 4326
    },
    {
      "epoch": 393.3878787878788,
      "grad_norm": 0.3532550036907196,
      "learning_rate": 6.73e-06,
      "loss": 0.3669,
      "step": 4327
    },
    {
      "epoch": 393.4848484848485,
      "grad_norm": 0.30630025267601013,
      "learning_rate": 6.72e-06,
      "loss": 0.355,
      "step": 4328
    },
    {
      "epoch": 393.58181818181816,
      "grad_norm": 0.2951465845108032,
      "learning_rate": 6.710000000000001e-06,
      "loss": 0.3199,
      "step": 4329
    },
    {
      "epoch": 393.6787878787879,
      "grad_norm": 0.2646746039390564,
      "learning_rate": 6.700000000000001e-06,
      "loss": 0.3071,
      "step": 4330
    },
    {
      "epoch": 393.6787878787879,
      "eval_loss": 0.4244401454925537,
      "eval_runtime": 2.1167,
      "eval_samples_per_second": 25.984,
      "eval_steps_per_second": 3.307,
      "step": 4330
    },
    {
      "epoch": 393.77575757575755,
      "grad_norm": 0.27225181460380554,
      "learning_rate": 6.69e-06,
      "loss": 0.2883,
      "step": 4331
    },
    {
      "epoch": 393.8727272727273,
      "grad_norm": 0.2778588831424713,
      "learning_rate": 6.68e-06,
      "loss": 0.3288,
      "step": 4332
    },
    {
      "epoch": 393.969696969697,
      "grad_norm": 0.3167281746864319,
      "learning_rate": 6.67e-06,
      "loss": 0.2983,
      "step": 4333
    },
    {
      "epoch": 394.0,
      "grad_norm": 0.5620117783546448,
      "learning_rate": 6.660000000000001e-06,
      "loss": 0.3396,
      "step": 4334
    },
    {
      "epoch": 394.0969696969697,
      "grad_norm": 0.33203279972076416,
      "learning_rate": 6.650000000000001e-06,
      "loss": 0.3326,
      "step": 4335
    },
    {
      "epoch": 394.1939393939394,
      "grad_norm": 0.27428722381591797,
      "learning_rate": 6.640000000000001e-06,
      "loss": 0.3455,
      "step": 4336
    },
    {
      "epoch": 394.2909090909091,
      "grad_norm": 0.26442083716392517,
      "learning_rate": 6.63e-06,
      "loss": 0.3129,
      "step": 4337
    },
    {
      "epoch": 394.3878787878788,
      "grad_norm": 0.29959186911582947,
      "learning_rate": 6.62e-06,
      "loss": 0.2967,
      "step": 4338
    },
    {
      "epoch": 394.4848484848485,
      "grad_norm": 0.26450297236442566,
      "learning_rate": 6.610000000000001e-06,
      "loss": 0.3061,
      "step": 4339
    },
    {
      "epoch": 394.58181818181816,
      "grad_norm": 0.3069596290588379,
      "learning_rate": 6.6e-06,
      "loss": 0.3369,
      "step": 4340
    },
    {
      "epoch": 394.58181818181816,
      "eval_loss": 0.42409956455230713,
      "eval_runtime": 2.1323,
      "eval_samples_per_second": 25.794,
      "eval_steps_per_second": 3.283,
      "step": 4340
    },
    {
      "epoch": 394.6787878787879,
      "grad_norm": 0.2847595810890198,
      "learning_rate": 6.5900000000000004e-06,
      "loss": 0.3158,
      "step": 4341
    },
    {
      "epoch": 394.77575757575755,
      "grad_norm": 0.33862245082855225,
      "learning_rate": 6.58e-06,
      "loss": 0.3478,
      "step": 4342
    },
    {
      "epoch": 394.8727272727273,
      "grad_norm": 0.3017573058605194,
      "learning_rate": 6.57e-06,
      "loss": 0.3384,
      "step": 4343
    },
    {
      "epoch": 394.969696969697,
      "grad_norm": 0.2776963710784912,
      "learning_rate": 6.560000000000001e-06,
      "loss": 0.3544,
      "step": 4344
    },
    {
      "epoch": 395.0,
      "grad_norm": 0.5282799005508423,
      "learning_rate": 6.550000000000001e-06,
      "loss": 0.323,
      "step": 4345
    },
    {
      "epoch": 395.0969696969697,
      "grad_norm": 0.31962233781814575,
      "learning_rate": 6.54e-06,
      "loss": 0.3292,
      "step": 4346
    },
    {
      "epoch": 395.1939393939394,
      "grad_norm": 0.28816673159599304,
      "learning_rate": 6.53e-06,
      "loss": 0.357,
      "step": 4347
    },
    {
      "epoch": 395.2909090909091,
      "grad_norm": 0.264009952545166,
      "learning_rate": 6.519999999999999e-06,
      "loss": 0.2984,
      "step": 4348
    },
    {
      "epoch": 395.3878787878788,
      "grad_norm": 0.2885288596153259,
      "learning_rate": 6.510000000000001e-06,
      "loss": 0.3199,
      "step": 4349
    },
    {
      "epoch": 395.4848484848485,
      "grad_norm": 0.32528144121170044,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.3225,
      "step": 4350
    },
    {
      "epoch": 395.4848484848485,
      "eval_loss": 0.4243115186691284,
      "eval_runtime": 2.1296,
      "eval_samples_per_second": 25.826,
      "eval_steps_per_second": 3.287,
      "step": 4350
    },
    {
      "epoch": 395.58181818181816,
      "grad_norm": 0.270365446805954,
      "learning_rate": 6.4900000000000005e-06,
      "loss": 0.3231,
      "step": 4351
    },
    {
      "epoch": 395.6787878787879,
      "grad_norm": 0.2635614275932312,
      "learning_rate": 6.48e-06,
      "loss": 0.3142,
      "step": 4352
    },
    {
      "epoch": 395.77575757575755,
      "grad_norm": 0.2843759059906006,
      "learning_rate": 6.47e-06,
      "loss": 0.3191,
      "step": 4353
    },
    {
      "epoch": 395.8727272727273,
      "grad_norm": 0.441561222076416,
      "learning_rate": 6.460000000000001e-06,
      "loss": 0.3365,
      "step": 4354
    },
    {
      "epoch": 395.969696969697,
      "grad_norm": 0.30812638998031616,
      "learning_rate": 6.45e-06,
      "loss": 0.3594,
      "step": 4355
    },
    {
      "epoch": 396.0,
      "grad_norm": 0.418353796005249,
      "learning_rate": 6.44e-06,
      "loss": 0.3349,
      "step": 4356
    },
    {
      "epoch": 396.0969696969697,
      "grad_norm": 0.27173471450805664,
      "learning_rate": 6.43e-06,
      "loss": 0.309,
      "step": 4357
    },
    {
      "epoch": 396.1939393939394,
      "grad_norm": 0.31804460287094116,
      "learning_rate": 6.4199999999999995e-06,
      "loss": 0.3391,
      "step": 4358
    },
    {
      "epoch": 396.2909090909091,
      "grad_norm": 0.29313576221466064,
      "learning_rate": 6.4100000000000005e-06,
      "loss": 0.3235,
      "step": 4359
    },
    {
      "epoch": 396.3878787878788,
      "grad_norm": 0.2768804430961609,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.3105,
      "step": 4360
    },
    {
      "epoch": 396.3878787878788,
      "eval_loss": 0.42421257495880127,
      "eval_runtime": 2.1208,
      "eval_samples_per_second": 25.934,
      "eval_steps_per_second": 3.301,
      "step": 4360
    },
    {
      "epoch": 396.4848484848485,
      "grad_norm": 0.33066731691360474,
      "learning_rate": 6.39e-06,
      "loss": 0.3546,
      "step": 4361
    },
    {
      "epoch": 396.58181818181816,
      "grad_norm": 0.28355875611305237,
      "learning_rate": 6.38e-06,
      "loss": 0.3148,
      "step": 4362
    },
    {
      "epoch": 396.6787878787879,
      "grad_norm": 0.4002547860145569,
      "learning_rate": 6.370000000000001e-06,
      "loss": 0.3381,
      "step": 4363
    },
    {
      "epoch": 396.77575757575755,
      "grad_norm": 0.25764694809913635,
      "learning_rate": 6.360000000000001e-06,
      "loss": 0.3304,
      "step": 4364
    },
    {
      "epoch": 396.8727272727273,
      "grad_norm": 0.37603357434272766,
      "learning_rate": 6.35e-06,
      "loss": 0.3101,
      "step": 4365
    },
    {
      "epoch": 396.969696969697,
      "grad_norm": 0.31180375814437866,
      "learning_rate": 6.34e-06,
      "loss": 0.3443,
      "step": 4366
    },
    {
      "epoch": 397.0,
      "grad_norm": 0.5192236304283142,
      "learning_rate": 6.3299999999999995e-06,
      "loss": 0.3664,
      "step": 4367
    },
    {
      "epoch": 397.0969696969697,
      "grad_norm": 0.2815007269382477,
      "learning_rate": 6.320000000000001e-06,
      "loss": 0.3054,
      "step": 4368
    },
    {
      "epoch": 397.1939393939394,
      "grad_norm": 0.27915215492248535,
      "learning_rate": 6.3100000000000006e-06,
      "loss": 0.3342,
      "step": 4369
    },
    {
      "epoch": 397.2909090909091,
      "grad_norm": 0.3028716444969177,
      "learning_rate": 6.300000000000001e-06,
      "loss": 0.312,
      "step": 4370
    },
    {
      "epoch": 397.2909090909091,
      "eval_loss": 0.42403414845466614,
      "eval_runtime": 2.1218,
      "eval_samples_per_second": 25.921,
      "eval_steps_per_second": 3.299,
      "step": 4370
    },
    {
      "epoch": 397.3878787878788,
      "grad_norm": 0.26431724429130554,
      "learning_rate": 6.29e-06,
      "loss": 0.3532,
      "step": 4371
    },
    {
      "epoch": 397.4848484848485,
      "grad_norm": 0.26671725511550903,
      "learning_rate": 6.28e-06,
      "loss": 0.3397,
      "step": 4372
    },
    {
      "epoch": 397.58181818181816,
      "grad_norm": 0.26035550236701965,
      "learning_rate": 6.270000000000001e-06,
      "loss": 0.2852,
      "step": 4373
    },
    {
      "epoch": 397.6787878787879,
      "grad_norm": 0.29644638299942017,
      "learning_rate": 6.26e-06,
      "loss": 0.307,
      "step": 4374
    },
    {
      "epoch": 397.77575757575755,
      "grad_norm": 0.29415082931518555,
      "learning_rate": 6.25e-06,
      "loss": 0.3362,
      "step": 4375
    },
    {
      "epoch": 397.8727272727273,
      "grad_norm": 0.29705363512039185,
      "learning_rate": 6.24e-06,
      "loss": 0.3451,
      "step": 4376
    },
    {
      "epoch": 397.969696969697,
      "grad_norm": 0.3315047323703766,
      "learning_rate": 6.2300000000000005e-06,
      "loss": 0.3535,
      "step": 4377
    },
    {
      "epoch": 398.0,
      "grad_norm": 0.6073850989341736,
      "learning_rate": 6.22e-06,
      "loss": 0.3566,
      "step": 4378
    },
    {
      "epoch": 398.0969696969697,
      "grad_norm": 0.2666085660457611,
      "learning_rate": 6.210000000000001e-06,
      "loss": 0.3293,
      "step": 4379
    },
    {
      "epoch": 398.1939393939394,
      "grad_norm": 0.2885405421257019,
      "learning_rate": 6.2e-06,
      "loss": 0.346,
      "step": 4380
    },
    {
      "epoch": 398.1939393939394,
      "eval_loss": 0.42399489879608154,
      "eval_runtime": 2.1503,
      "eval_samples_per_second": 25.578,
      "eval_steps_per_second": 3.255,
      "step": 4380
    },
    {
      "epoch": 398.2909090909091,
      "grad_norm": 0.3000956177711487,
      "learning_rate": 6.19e-06,
      "loss": 0.3534,
      "step": 4381
    },
    {
      "epoch": 398.3878787878788,
      "grad_norm": 0.2843852937221527,
      "learning_rate": 6.18e-06,
      "loss": 0.2892,
      "step": 4382
    },
    {
      "epoch": 398.4848484848485,
      "grad_norm": 0.278983473777771,
      "learning_rate": 6.17e-06,
      "loss": 0.3321,
      "step": 4383
    },
    {
      "epoch": 398.58181818181816,
      "grad_norm": 0.24275529384613037,
      "learning_rate": 6.16e-06,
      "loss": 0.3351,
      "step": 4384
    },
    {
      "epoch": 398.6787878787879,
      "grad_norm": 0.2708281874656677,
      "learning_rate": 6.15e-06,
      "loss": 0.3005,
      "step": 4385
    },
    {
      "epoch": 398.77575757575755,
      "grad_norm": 0.32019880414009094,
      "learning_rate": 6.1400000000000005e-06,
      "loss": 0.3409,
      "step": 4386
    },
    {
      "epoch": 398.8727272727273,
      "grad_norm": 0.32802683115005493,
      "learning_rate": 6.130000000000001e-06,
      "loss": 0.3375,
      "step": 4387
    },
    {
      "epoch": 398.969696969697,
      "grad_norm": 0.3256203234195709,
      "learning_rate": 6.12e-06,
      "loss": 0.321,
      "step": 4388
    },
    {
      "epoch": 399.0,
      "grad_norm": 0.49412262439727783,
      "learning_rate": 6.110000000000001e-06,
      "loss": 0.3175,
      "step": 4389
    },
    {
      "epoch": 399.0969696969697,
      "grad_norm": 0.2912564277648926,
      "learning_rate": 6.1e-06,
      "loss": 0.328,
      "step": 4390
    },
    {
      "epoch": 399.0969696969697,
      "eval_loss": 0.4240299165248871,
      "eval_runtime": 2.1219,
      "eval_samples_per_second": 25.92,
      "eval_steps_per_second": 3.299,
      "step": 4390
    },
    {
      "epoch": 399.1939393939394,
      "grad_norm": 0.30961018800735474,
      "learning_rate": 6.090000000000001e-06,
      "loss": 0.2733,
      "step": 4391
    },
    {
      "epoch": 399.2909090909091,
      "grad_norm": 0.284614622592926,
      "learning_rate": 6.08e-06,
      "loss": 0.3479,
      "step": 4392
    },
    {
      "epoch": 399.3878787878788,
      "grad_norm": 0.2767944633960724,
      "learning_rate": 6.07e-06,
      "loss": 0.3286,
      "step": 4393
    },
    {
      "epoch": 399.4848484848485,
      "grad_norm": 0.30359727144241333,
      "learning_rate": 6.0600000000000004e-06,
      "loss": 0.3227,
      "step": 4394
    },
    {
      "epoch": 399.58181818181816,
      "grad_norm": 0.3214990794658661,
      "learning_rate": 6.0500000000000005e-06,
      "loss": 0.3438,
      "step": 4395
    },
    {
      "epoch": 399.6787878787879,
      "grad_norm": 0.28994226455688477,
      "learning_rate": 6.040000000000001e-06,
      "loss": 0.3691,
      "step": 4396
    },
    {
      "epoch": 399.77575757575755,
      "grad_norm": 0.2830333113670349,
      "learning_rate": 6.03e-06,
      "loss": 0.3225,
      "step": 4397
    },
    {
      "epoch": 399.8727272727273,
      "grad_norm": 0.2534017562866211,
      "learning_rate": 6.02e-06,
      "loss": 0.3475,
      "step": 4398
    },
    {
      "epoch": 399.969696969697,
      "grad_norm": 0.2874850630760193,
      "learning_rate": 6.01e-06,
      "loss": 0.3051,
      "step": 4399
    },
    {
      "epoch": 400.0,
      "grad_norm": 0.5695606470108032,
      "learning_rate": 6e-06,
      "loss": 0.3072,
      "step": 4400
    },
    {
      "epoch": 400.0,
      "eval_loss": 0.4242197275161743,
      "eval_runtime": 2.1336,
      "eval_samples_per_second": 25.778,
      "eval_steps_per_second": 3.281,
      "step": 4400
    },
    {
      "epoch": 400.0969696969697,
      "grad_norm": 0.2943519949913025,
      "learning_rate": 5.99e-06,
      "loss": 0.3076,
      "step": 4401
    },
    {
      "epoch": 400.1939393939394,
      "grad_norm": 0.26126110553741455,
      "learning_rate": 5.98e-06,
      "loss": 0.355,
      "step": 4402
    },
    {
      "epoch": 400.2909090909091,
      "grad_norm": 0.2612912654876709,
      "learning_rate": 5.9700000000000004e-06,
      "loss": 0.326,
      "step": 4403
    },
    {
      "epoch": 400.3878787878788,
      "grad_norm": 0.28973740339279175,
      "learning_rate": 5.9600000000000005e-06,
      "loss": 0.34,
      "step": 4404
    },
    {
      "epoch": 400.4848484848485,
      "grad_norm": 0.28152215480804443,
      "learning_rate": 5.95e-06,
      "loss": 0.3691,
      "step": 4405
    },
    {
      "epoch": 400.58181818181816,
      "grad_norm": 0.2883203327655792,
      "learning_rate": 5.940000000000001e-06,
      "loss": 0.3574,
      "step": 4406
    },
    {
      "epoch": 400.6787878787879,
      "grad_norm": 0.3067658841609955,
      "learning_rate": 5.93e-06,
      "loss": 0.2759,
      "step": 4407
    },
    {
      "epoch": 400.77575757575755,
      "grad_norm": 0.265886127948761,
      "learning_rate": 5.920000000000001e-06,
      "loss": 0.3187,
      "step": 4408
    },
    {
      "epoch": 400.8727272727273,
      "grad_norm": 0.34973838925361633,
      "learning_rate": 5.91e-06,
      "loss": 0.3129,
      "step": 4409
    },
    {
      "epoch": 400.969696969697,
      "grad_norm": 0.2948808968067169,
      "learning_rate": 5.9e-06,
      "loss": 0.3399,
      "step": 4410
    },
    {
      "epoch": 400.969696969697,
      "eval_loss": 0.42390140891075134,
      "eval_runtime": 2.1249,
      "eval_samples_per_second": 25.884,
      "eval_steps_per_second": 3.294,
      "step": 4410
    },
    {
      "epoch": 401.0,
      "grad_norm": 0.5547888875007629,
      "learning_rate": 5.89e-06,
      "loss": 0.2722,
      "step": 4411
    },
    {
      "epoch": 401.0969696969697,
      "grad_norm": 0.28168249130249023,
      "learning_rate": 5.8800000000000005e-06,
      "loss": 0.3197,
      "step": 4412
    },
    {
      "epoch": 401.1939393939394,
      "grad_norm": 0.32535097002983093,
      "learning_rate": 5.8700000000000005e-06,
      "loss": 0.3446,
      "step": 4413
    },
    {
      "epoch": 401.2909090909091,
      "grad_norm": 0.3320438265800476,
      "learning_rate": 5.86e-06,
      "loss": 0.3055,
      "step": 4414
    },
    {
      "epoch": 401.3878787878788,
      "grad_norm": 0.33057060837745667,
      "learning_rate": 5.850000000000001e-06,
      "loss": 0.3342,
      "step": 4415
    },
    {
      "epoch": 401.4848484848485,
      "grad_norm": 0.3710538148880005,
      "learning_rate": 5.84e-06,
      "loss": 0.3305,
      "step": 4416
    },
    {
      "epoch": 401.58181818181816,
      "grad_norm": 0.3122580647468567,
      "learning_rate": 5.83e-06,
      "loss": 0.3178,
      "step": 4417
    },
    {
      "epoch": 401.6787878787879,
      "grad_norm": 0.24265459179878235,
      "learning_rate": 5.82e-06,
      "loss": 0.3026,
      "step": 4418
    },
    {
      "epoch": 401.77575757575755,
      "grad_norm": 0.2679807245731354,
      "learning_rate": 5.81e-06,
      "loss": 0.33,
      "step": 4419
    },
    {
      "epoch": 401.8727272727273,
      "grad_norm": 0.28795769810676575,
      "learning_rate": 5.8e-06,
      "loss": 0.342,
      "step": 4420
    },
    {
      "epoch": 401.8727272727273,
      "eval_loss": 0.4240061044692993,
      "eval_runtime": 2.118,
      "eval_samples_per_second": 25.968,
      "eval_steps_per_second": 3.305,
      "step": 4420
    },
    {
      "epoch": 401.969696969697,
      "grad_norm": 0.3102927505970001,
      "learning_rate": 5.7900000000000005e-06,
      "loss": 0.3402,
      "step": 4421
    },
    {
      "epoch": 402.0,
      "grad_norm": 0.4905965328216553,
      "learning_rate": 5.78e-06,
      "loss": 0.3695,
      "step": 4422
    },
    {
      "epoch": 402.0969696969697,
      "grad_norm": 0.2856130003929138,
      "learning_rate": 5.770000000000001e-06,
      "loss": 0.3564,
      "step": 4423
    },
    {
      "epoch": 402.1939393939394,
      "grad_norm": 0.3797096014022827,
      "learning_rate": 5.76e-06,
      "loss": 0.2945,
      "step": 4424
    },
    {
      "epoch": 402.2909090909091,
      "grad_norm": 0.30673947930336,
      "learning_rate": 5.750000000000001e-06,
      "loss": 0.331,
      "step": 4425
    },
    {
      "epoch": 402.3878787878788,
      "grad_norm": 0.32168588042259216,
      "learning_rate": 5.74e-06,
      "loss": 0.3345,
      "step": 4426
    },
    {
      "epoch": 402.4848484848485,
      "grad_norm": 0.3218797743320465,
      "learning_rate": 5.73e-06,
      "loss": 0.3298,
      "step": 4427
    },
    {
      "epoch": 402.58181818181816,
      "grad_norm": 0.2737649977207184,
      "learning_rate": 5.72e-06,
      "loss": 0.3461,
      "step": 4428
    },
    {
      "epoch": 402.6787878787879,
      "grad_norm": 0.28247711062431335,
      "learning_rate": 5.71e-06,
      "loss": 0.2937,
      "step": 4429
    },
    {
      "epoch": 402.77575757575755,
      "grad_norm": 0.33665931224823,
      "learning_rate": 5.7000000000000005e-06,
      "loss": 0.3385,
      "step": 4430
    },
    {
      "epoch": 402.77575757575755,
      "eval_loss": 0.4240177869796753,
      "eval_runtime": 2.1337,
      "eval_samples_per_second": 25.777,
      "eval_steps_per_second": 3.281,
      "step": 4430
    },
    {
      "epoch": 402.8727272727273,
      "grad_norm": 0.2630540430545807,
      "learning_rate": 5.690000000000001e-06,
      "loss": 0.296,
      "step": 4431
    },
    {
      "epoch": 402.969696969697,
      "grad_norm": 0.3100334703922272,
      "learning_rate": 5.680000000000001e-06,
      "loss": 0.3603,
      "step": 4432
    },
    {
      "epoch": 403.0,
      "grad_norm": 0.5338719487190247,
      "learning_rate": 5.67e-06,
      "loss": 0.3289,
      "step": 4433
    },
    {
      "epoch": 403.0969696969697,
      "grad_norm": 0.3095831871032715,
      "learning_rate": 5.66e-06,
      "loss": 0.3444,
      "step": 4434
    },
    {
      "epoch": 403.1939393939394,
      "grad_norm": 0.24900095164775848,
      "learning_rate": 5.65e-06,
      "loss": 0.2999,
      "step": 4435
    },
    {
      "epoch": 403.2909090909091,
      "grad_norm": 0.28118544816970825,
      "learning_rate": 5.64e-06,
      "loss": 0.3081,
      "step": 4436
    },
    {
      "epoch": 403.3878787878788,
      "grad_norm": 0.3060985803604126,
      "learning_rate": 5.63e-06,
      "loss": 0.3327,
      "step": 4437
    },
    {
      "epoch": 403.4848484848485,
      "grad_norm": 0.252957284450531,
      "learning_rate": 5.62e-06,
      "loss": 0.3445,
      "step": 4438
    },
    {
      "epoch": 403.58181818181816,
      "grad_norm": 0.2750088572502136,
      "learning_rate": 5.61e-06,
      "loss": 0.3123,
      "step": 4439
    },
    {
      "epoch": 403.6787878787879,
      "grad_norm": 0.3438875079154968,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.3181,
      "step": 4440
    },
    {
      "epoch": 403.6787878787879,
      "eval_loss": 0.42397940158843994,
      "eval_runtime": 2.1375,
      "eval_samples_per_second": 25.732,
      "eval_steps_per_second": 3.275,
      "step": 4440
    },
    {
      "epoch": 403.77575757575755,
      "grad_norm": 0.2744629681110382,
      "learning_rate": 5.59e-06,
      "loss": 0.3433,
      "step": 4441
    },
    {
      "epoch": 403.8727272727273,
      "grad_norm": 0.29330527782440186,
      "learning_rate": 5.580000000000001e-06,
      "loss": 0.3222,
      "step": 4442
    },
    {
      "epoch": 403.969696969697,
      "grad_norm": 0.2824782431125641,
      "learning_rate": 5.57e-06,
      "loss": 0.3402,
      "step": 4443
    },
    {
      "epoch": 404.0,
      "grad_norm": 0.6952817440032959,
      "learning_rate": 5.56e-06,
      "loss": 0.3728,
      "step": 4444
    },
    {
      "epoch": 404.0969696969697,
      "grad_norm": 0.3297734558582306,
      "learning_rate": 5.55e-06,
      "loss": 0.3097,
      "step": 4445
    },
    {
      "epoch": 404.1939393939394,
      "grad_norm": 0.29496389627456665,
      "learning_rate": 5.54e-06,
      "loss": 0.3369,
      "step": 4446
    },
    {
      "epoch": 404.2909090909091,
      "grad_norm": 0.30634990334510803,
      "learning_rate": 5.53e-06,
      "loss": 0.3165,
      "step": 4447
    },
    {
      "epoch": 404.3878787878788,
      "grad_norm": 0.25977566838264465,
      "learning_rate": 5.5200000000000005e-06,
      "loss": 0.3423,
      "step": 4448
    },
    {
      "epoch": 404.4848484848485,
      "grad_norm": 0.30670416355133057,
      "learning_rate": 5.510000000000001e-06,
      "loss": 0.3214,
      "step": 4449
    },
    {
      "epoch": 404.58181818181816,
      "grad_norm": 0.41724225878715515,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.327,
      "step": 4450
    },
    {
      "epoch": 404.58181818181816,
      "eval_loss": 0.423881471157074,
      "eval_runtime": 2.1176,
      "eval_samples_per_second": 25.973,
      "eval_steps_per_second": 3.306,
      "step": 4450
    },
    {
      "epoch": 404.6787878787879,
      "grad_norm": 0.2812544107437134,
      "learning_rate": 5.49e-06,
      "loss": 0.3541,
      "step": 4451
    },
    {
      "epoch": 404.77575757575755,
      "grad_norm": 0.3451825678348541,
      "learning_rate": 5.48e-06,
      "loss": 0.3401,
      "step": 4452
    },
    {
      "epoch": 404.8727272727273,
      "grad_norm": 0.2523994445800781,
      "learning_rate": 5.47e-06,
      "loss": 0.3159,
      "step": 4453
    },
    {
      "epoch": 404.969696969697,
      "grad_norm": 0.2768281400203705,
      "learning_rate": 5.46e-06,
      "loss": 0.3254,
      "step": 4454
    },
    {
      "epoch": 405.0,
      "grad_norm": 0.5321841239929199,
      "learning_rate": 5.45e-06,
      "loss": 0.3025,
      "step": 4455
    },
    {
      "epoch": 405.0969696969697,
      "grad_norm": 0.30620166659355164,
      "learning_rate": 5.44e-06,
      "loss": 0.3197,
      "step": 4456
    },
    {
      "epoch": 405.1939393939394,
      "grad_norm": 0.3524479866027832,
      "learning_rate": 5.4300000000000005e-06,
      "loss": 0.3438,
      "step": 4457
    },
    {
      "epoch": 405.2909090909091,
      "grad_norm": 0.27902770042419434,
      "learning_rate": 5.42e-06,
      "loss": 0.353,
      "step": 4458
    },
    {
      "epoch": 405.3878787878788,
      "grad_norm": 0.4012645483016968,
      "learning_rate": 5.410000000000001e-06,
      "loss": 0.3244,
      "step": 4459
    },
    {
      "epoch": 405.4848484848485,
      "grad_norm": 0.26941490173339844,
      "learning_rate": 5.4e-06,
      "loss": 0.3164,
      "step": 4460
    },
    {
      "epoch": 405.4848484848485,
      "eval_loss": 0.42388486862182617,
      "eval_runtime": 2.1321,
      "eval_samples_per_second": 25.796,
      "eval_steps_per_second": 3.283,
      "step": 4460
    },
    {
      "epoch": 405.58181818181816,
      "grad_norm": 0.30666786432266235,
      "learning_rate": 5.390000000000001e-06,
      "loss": 0.328,
      "step": 4461
    },
    {
      "epoch": 405.6787878787879,
      "grad_norm": 0.2669554650783539,
      "learning_rate": 5.38e-06,
      "loss": 0.3195,
      "step": 4462
    },
    {
      "epoch": 405.77575757575755,
      "grad_norm": 0.34226134419441223,
      "learning_rate": 5.37e-06,
      "loss": 0.3309,
      "step": 4463
    },
    {
      "epoch": 405.8727272727273,
      "grad_norm": 0.3121098577976227,
      "learning_rate": 5.36e-06,
      "loss": 0.3022,
      "step": 4464
    },
    {
      "epoch": 405.969696969697,
      "grad_norm": 0.2962389588356018,
      "learning_rate": 5.3500000000000004e-06,
      "loss": 0.3432,
      "step": 4465
    },
    {
      "epoch": 406.0,
      "grad_norm": 0.5181310772895813,
      "learning_rate": 5.3400000000000005e-06,
      "loss": 0.3171,
      "step": 4466
    },
    {
      "epoch": 406.0969696969697,
      "grad_norm": 0.2788962721824646,
      "learning_rate": 5.330000000000001e-06,
      "loss": 0.2907,
      "step": 4467
    },
    {
      "epoch": 406.1939393939394,
      "grad_norm": 0.23445305228233337,
      "learning_rate": 5.32e-06,
      "loss": 0.3339,
      "step": 4468
    },
    {
      "epoch": 406.2909090909091,
      "grad_norm": 0.2866305112838745,
      "learning_rate": 5.31e-06,
      "loss": 0.3472,
      "step": 4469
    },
    {
      "epoch": 406.3878787878788,
      "grad_norm": 0.27106624841690063,
      "learning_rate": 5.3e-06,
      "loss": 0.3226,
      "step": 4470
    },
    {
      "epoch": 406.3878787878788,
      "eval_loss": 0.42396581172943115,
      "eval_runtime": 2.1231,
      "eval_samples_per_second": 25.905,
      "eval_steps_per_second": 3.297,
      "step": 4470
    },
    {
      "epoch": 406.4848484848485,
      "grad_norm": 0.2928353548049927,
      "learning_rate": 5.29e-06,
      "loss": 0.2981,
      "step": 4471
    },
    {
      "epoch": 406.58181818181816,
      "grad_norm": 0.27284109592437744,
      "learning_rate": 5.28e-06,
      "loss": 0.3282,
      "step": 4472
    },
    {
      "epoch": 406.6787878787879,
      "grad_norm": 0.33163031935691833,
      "learning_rate": 5.2699999999999995e-06,
      "loss": 0.3396,
      "step": 4473
    },
    {
      "epoch": 406.77575757575755,
      "grad_norm": 0.33143162727355957,
      "learning_rate": 5.2600000000000005e-06,
      "loss": 0.3402,
      "step": 4474
    },
    {
      "epoch": 406.8727272727273,
      "grad_norm": 0.3006884455680847,
      "learning_rate": 5.25e-06,
      "loss": 0.321,
      "step": 4475
    },
    {
      "epoch": 406.969696969697,
      "grad_norm": 0.26408207416534424,
      "learning_rate": 5.240000000000001e-06,
      "loss": 0.3447,
      "step": 4476
    },
    {
      "epoch": 407.0,
      "grad_norm": 0.5091809630393982,
      "learning_rate": 5.23e-06,
      "loss": 0.3562,
      "step": 4477
    },
    {
      "epoch": 407.0969696969697,
      "grad_norm": 0.294497549533844,
      "learning_rate": 5.220000000000001e-06,
      "loss": 0.3708,
      "step": 4478
    },
    {
      "epoch": 407.1939393939394,
      "grad_norm": 0.3147751986980438,
      "learning_rate": 5.21e-06,
      "loss": 0.3433,
      "step": 4479
    },
    {
      "epoch": 407.2909090909091,
      "grad_norm": 0.2634759247303009,
      "learning_rate": 5.2e-06,
      "loss": 0.2847,
      "step": 4480
    },
    {
      "epoch": 407.2909090909091,
      "eval_loss": 0.423898845911026,
      "eval_runtime": 2.118,
      "eval_samples_per_second": 25.967,
      "eval_steps_per_second": 3.305,
      "step": 4480
    },
    {
      "epoch": 407.3878787878788,
      "grad_norm": 0.26809194684028625,
      "learning_rate": 5.19e-06,
      "loss": 0.3095,
      "step": 4481
    },
    {
      "epoch": 407.4848484848485,
      "grad_norm": 0.30752381682395935,
      "learning_rate": 5.18e-06,
      "loss": 0.3521,
      "step": 4482
    },
    {
      "epoch": 407.58181818181816,
      "grad_norm": 0.2728431820869446,
      "learning_rate": 5.1700000000000005e-06,
      "loss": 0.2942,
      "step": 4483
    },
    {
      "epoch": 407.6787878787879,
      "grad_norm": 0.3156808614730835,
      "learning_rate": 5.1600000000000006e-06,
      "loss": 0.3183,
      "step": 4484
    },
    {
      "epoch": 407.77575757575755,
      "grad_norm": 0.28744783997535706,
      "learning_rate": 5.15e-06,
      "loss": 0.3612,
      "step": 4485
    },
    {
      "epoch": 407.8727272727273,
      "grad_norm": 0.3444119691848755,
      "learning_rate": 5.140000000000001e-06,
      "loss": 0.2986,
      "step": 4486
    },
    {
      "epoch": 407.969696969697,
      "grad_norm": 0.33266812562942505,
      "learning_rate": 5.13e-06,
      "loss": 0.3468,
      "step": 4487
    },
    {
      "epoch": 408.0,
      "grad_norm": 0.46602705121040344,
      "learning_rate": 5.12e-06,
      "loss": 0.327,
      "step": 4488
    },
    {
      "epoch": 408.0969696969697,
      "grad_norm": 0.3970167636871338,
      "learning_rate": 5.11e-06,
      "loss": 0.3239,
      "step": 4489
    },
    {
      "epoch": 408.1939393939394,
      "grad_norm": 0.24424688518047333,
      "learning_rate": 5.1e-06,
      "loss": 0.3289,
      "step": 4490
    },
    {
      "epoch": 408.1939393939394,
      "eval_loss": 0.4239845871925354,
      "eval_runtime": 2.1275,
      "eval_samples_per_second": 25.851,
      "eval_steps_per_second": 3.29,
      "step": 4490
    },
    {
      "epoch": 408.2909090909091,
      "grad_norm": 0.26348650455474854,
      "learning_rate": 5.09e-06,
      "loss": 0.3393,
      "step": 4491
    },
    {
      "epoch": 408.3878787878788,
      "grad_norm": 0.2709069550037384,
      "learning_rate": 5.08e-06,
      "loss": 0.3009,
      "step": 4492
    },
    {
      "epoch": 408.4848484848485,
      "grad_norm": 0.2873727083206177,
      "learning_rate": 5.070000000000001e-06,
      "loss": 0.2914,
      "step": 4493
    },
    {
      "epoch": 408.58181818181816,
      "grad_norm": 0.36658114194869995,
      "learning_rate": 5.06e-06,
      "loss": 0.3573,
      "step": 4494
    },
    {
      "epoch": 408.6787878787879,
      "grad_norm": 0.2843858003616333,
      "learning_rate": 5.050000000000001e-06,
      "loss": 0.3047,
      "step": 4495
    },
    {
      "epoch": 408.77575757575755,
      "grad_norm": 0.30994489789009094,
      "learning_rate": 5.04e-06,
      "loss": 0.3566,
      "step": 4496
    },
    {
      "epoch": 408.8727272727273,
      "grad_norm": 0.2767634689807892,
      "learning_rate": 5.03e-06,
      "loss": 0.3307,
      "step": 4497
    },
    {
      "epoch": 408.969696969697,
      "grad_norm": 0.3072282075881958,
      "learning_rate": 5.02e-06,
      "loss": 0.3204,
      "step": 4498
    },
    {
      "epoch": 409.0,
      "grad_norm": 0.6475484371185303,
      "learning_rate": 5.01e-06,
      "loss": 0.3943,
      "step": 4499
    },
    {
      "epoch": 409.0969696969697,
      "grad_norm": 0.3126541078090668,
      "learning_rate": 5e-06,
      "loss": 0.3417,
      "step": 4500
    },
    {
      "epoch": 409.0969696969697,
      "eval_loss": 0.4239867925643921,
      "eval_runtime": 2.1314,
      "eval_samples_per_second": 25.804,
      "eval_steps_per_second": 3.284,
      "step": 4500
    },
    {
      "epoch": 409.1939393939394,
      "grad_norm": 0.31892114877700806,
      "learning_rate": 4.9900000000000005e-06,
      "loss": 0.3238,
      "step": 4501
    },
    {
      "epoch": 409.2909090909091,
      "grad_norm": 0.3017483055591583,
      "learning_rate": 4.98e-06,
      "loss": 0.3444,
      "step": 4502
    },
    {
      "epoch": 409.3878787878788,
      "grad_norm": 0.2635285258293152,
      "learning_rate": 4.970000000000001e-06,
      "loss": 0.311,
      "step": 4503
    },
    {
      "epoch": 409.4848484848485,
      "grad_norm": 0.2573358714580536,
      "learning_rate": 4.96e-06,
      "loss": 0.3069,
      "step": 4504
    },
    {
      "epoch": 409.58181818181816,
      "grad_norm": 0.28138741850852966,
      "learning_rate": 4.950000000000001e-06,
      "loss": 0.3414,
      "step": 4505
    },
    {
      "epoch": 409.6787878787879,
      "grad_norm": 0.24946054816246033,
      "learning_rate": 4.94e-06,
      "loss": 0.3883,
      "step": 4506
    },
    {
      "epoch": 409.77575757575755,
      "grad_norm": 0.33677899837493896,
      "learning_rate": 4.93e-06,
      "loss": 0.3041,
      "step": 4507
    },
    {
      "epoch": 409.8727272727273,
      "grad_norm": 0.30087172985076904,
      "learning_rate": 4.92e-06,
      "loss": 0.3296,
      "step": 4508
    },
    {
      "epoch": 409.969696969697,
      "grad_norm": 0.29604464769363403,
      "learning_rate": 4.9100000000000004e-06,
      "loss": 0.2914,
      "step": 4509
    },
    {
      "epoch": 410.0,
      "grad_norm": 0.5490189790725708,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.3065,
      "step": 4510
    },
    {
      "epoch": 410.0,
      "eval_loss": 0.4238167703151703,
      "eval_runtime": 2.1079,
      "eval_samples_per_second": 26.092,
      "eval_steps_per_second": 3.321,
      "step": 4510
    },
    {
      "epoch": 410.0969696969697,
      "grad_norm": 0.27711138129234314,
      "learning_rate": 4.89e-06,
      "loss": 0.3376,
      "step": 4511
    },
    {
      "epoch": 410.1939393939394,
      "grad_norm": 0.24843066930770874,
      "learning_rate": 4.880000000000001e-06,
      "loss": 0.3168,
      "step": 4512
    },
    {
      "epoch": 410.2909090909091,
      "grad_norm": 0.3280465304851532,
      "learning_rate": 4.87e-06,
      "loss": 0.3036,
      "step": 4513
    },
    {
      "epoch": 410.3878787878788,
      "grad_norm": 0.3124517500400543,
      "learning_rate": 4.86e-06,
      "loss": 0.3217,
      "step": 4514
    },
    {
      "epoch": 410.4848484848485,
      "grad_norm": 0.2721874415874481,
      "learning_rate": 4.85e-06,
      "loss": 0.362,
      "step": 4515
    },
    {
      "epoch": 410.58181818181816,
      "grad_norm": 0.27525606751441956,
      "learning_rate": 4.84e-06,
      "loss": 0.3408,
      "step": 4516
    },
    {
      "epoch": 410.6787878787879,
      "grad_norm": 0.2887386381626129,
      "learning_rate": 4.83e-06,
      "loss": 0.3291,
      "step": 4517
    },
    {
      "epoch": 410.77575757575755,
      "grad_norm": 0.299444317817688,
      "learning_rate": 4.8200000000000004e-06,
      "loss": 0.3377,
      "step": 4518
    },
    {
      "epoch": 410.8727272727273,
      "grad_norm": 0.2675188481807709,
      "learning_rate": 4.81e-06,
      "loss": 0.3157,
      "step": 4519
    },
    {
      "epoch": 410.969696969697,
      "grad_norm": 0.2881135046482086,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.309,
      "step": 4520
    },
    {
      "epoch": 410.969696969697,
      "eval_loss": 0.42392778396606445,
      "eval_runtime": 2.1506,
      "eval_samples_per_second": 25.574,
      "eval_steps_per_second": 3.255,
      "step": 4520
    },
    {
      "epoch": 411.0,
      "grad_norm": 0.521273672580719,
      "learning_rate": 4.79e-06,
      "loss": 0.3303,
      "step": 4521
    },
    {
      "epoch": 411.0969696969697,
      "grad_norm": 0.29321396350860596,
      "learning_rate": 4.780000000000001e-06,
      "loss": 0.3009,
      "step": 4522
    },
    {
      "epoch": 411.1939393939394,
      "grad_norm": 0.26560118794441223,
      "learning_rate": 4.77e-06,
      "loss": 0.3558,
      "step": 4523
    },
    {
      "epoch": 411.2909090909091,
      "grad_norm": 0.2869216501712799,
      "learning_rate": 4.76e-06,
      "loss": 0.3183,
      "step": 4524
    },
    {
      "epoch": 411.3878787878788,
      "grad_norm": 0.27583593130111694,
      "learning_rate": 4.75e-06,
      "loss": 0.3387,
      "step": 4525
    },
    {
      "epoch": 411.4848484848485,
      "grad_norm": 0.30815714597702026,
      "learning_rate": 4.74e-06,
      "loss": 0.3649,
      "step": 4526
    },
    {
      "epoch": 411.58181818181816,
      "grad_norm": 0.30801698565483093,
      "learning_rate": 4.7300000000000005e-06,
      "loss": 0.3698,
      "step": 4527
    },
    {
      "epoch": 411.6787878787879,
      "grad_norm": 0.2815544605255127,
      "learning_rate": 4.72e-06,
      "loss": 0.284,
      "step": 4528
    },
    {
      "epoch": 411.77575757575755,
      "grad_norm": 0.2630556523799896,
      "learning_rate": 4.710000000000001e-06,
      "loss": 0.3308,
      "step": 4529
    },
    {
      "epoch": 411.8727272727273,
      "grad_norm": 0.3057383894920349,
      "learning_rate": 4.7e-06,
      "loss": 0.3192,
      "step": 4530
    },
    {
      "epoch": 411.8727272727273,
      "eval_loss": 0.4240568280220032,
      "eval_runtime": 2.117,
      "eval_samples_per_second": 25.98,
      "eval_steps_per_second": 3.306,
      "step": 4530
    },
    {
      "epoch": 411.969696969697,
      "grad_norm": 0.2713818848133087,
      "learning_rate": 4.69e-06,
      "loss": 0.3087,
      "step": 4531
    },
    {
      "epoch": 412.0,
      "grad_norm": 0.577849268913269,
      "learning_rate": 4.68e-06,
      "loss": 0.2719,
      "step": 4532
    },
    {
      "epoch": 412.0969696969697,
      "grad_norm": 0.2892434597015381,
      "learning_rate": 4.67e-06,
      "loss": 0.3245,
      "step": 4533
    },
    {
      "epoch": 412.1939393939394,
      "grad_norm": 0.25237318873405457,
      "learning_rate": 4.66e-06,
      "loss": 0.3474,
      "step": 4534
    },
    {
      "epoch": 412.2909090909091,
      "grad_norm": 0.3992471992969513,
      "learning_rate": 4.65e-06,
      "loss": 0.2817,
      "step": 4535
    },
    {
      "epoch": 412.3878787878788,
      "grad_norm": 0.3084762692451477,
      "learning_rate": 4.64e-06,
      "loss": 0.3598,
      "step": 4536
    },
    {
      "epoch": 412.4848484848485,
      "grad_norm": 0.31249937415122986,
      "learning_rate": 4.6300000000000006e-06,
      "loss": 0.3552,
      "step": 4537
    },
    {
      "epoch": 412.58181818181816,
      "grad_norm": 0.29190361499786377,
      "learning_rate": 4.62e-06,
      "loss": 0.3039,
      "step": 4538
    },
    {
      "epoch": 412.6787878787879,
      "grad_norm": 0.29947835206985474,
      "learning_rate": 4.610000000000001e-06,
      "loss": 0.2769,
      "step": 4539
    },
    {
      "epoch": 412.77575757575755,
      "grad_norm": 0.30425596237182617,
      "learning_rate": 4.6e-06,
      "loss": 0.342,
      "step": 4540
    },
    {
      "epoch": 412.77575757575755,
      "eval_loss": 0.4239722788333893,
      "eval_runtime": 2.1347,
      "eval_samples_per_second": 25.765,
      "eval_steps_per_second": 3.279,
      "step": 4540
    },
    {
      "epoch": 412.8727272727273,
      "grad_norm": 0.2873094081878662,
      "learning_rate": 4.590000000000001e-06,
      "loss": 0.3116,
      "step": 4541
    },
    {
      "epoch": 412.969696969697,
      "grad_norm": 0.2804364562034607,
      "learning_rate": 4.58e-06,
      "loss": 0.3688,
      "step": 4542
    },
    {
      "epoch": 413.0,
      "grad_norm": 0.5596648454666138,
      "learning_rate": 4.57e-06,
      "loss": 0.3303,
      "step": 4543
    },
    {
      "epoch": 413.0969696969697,
      "grad_norm": 0.27821826934814453,
      "learning_rate": 4.56e-06,
      "loss": 0.3208,
      "step": 4544
    },
    {
      "epoch": 413.1939393939394,
      "grad_norm": 0.2921523451805115,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.3403,
      "step": 4545
    },
    {
      "epoch": 413.2909090909091,
      "grad_norm": 0.2964740991592407,
      "learning_rate": 4.540000000000001e-06,
      "loss": 0.34,
      "step": 4546
    },
    {
      "epoch": 413.3878787878788,
      "grad_norm": 0.2719631493091583,
      "learning_rate": 4.53e-06,
      "loss": 0.3357,
      "step": 4547
    },
    {
      "epoch": 413.4848484848485,
      "grad_norm": 0.2727508246898651,
      "learning_rate": 4.52e-06,
      "loss": 0.3318,
      "step": 4548
    },
    {
      "epoch": 413.58181818181816,
      "grad_norm": 0.27473440766334534,
      "learning_rate": 4.51e-06,
      "loss": 0.3354,
      "step": 4549
    },
    {
      "epoch": 413.6787878787879,
      "grad_norm": 0.3429303467273712,
      "learning_rate": 4.5e-06,
      "loss": 0.3306,
      "step": 4550
    },
    {
      "epoch": 413.6787878787879,
      "eval_loss": 0.4237283170223236,
      "eval_runtime": 2.1253,
      "eval_samples_per_second": 25.879,
      "eval_steps_per_second": 3.294,
      "step": 4550
    },
    {
      "epoch": 413.77575757575755,
      "grad_norm": 0.25924238562583923,
      "learning_rate": 4.49e-06,
      "loss": 0.3218,
      "step": 4551
    },
    {
      "epoch": 413.8727272727273,
      "grad_norm": 0.3040829002857208,
      "learning_rate": 4.48e-06,
      "loss": 0.2817,
      "step": 4552
    },
    {
      "epoch": 413.969696969697,
      "grad_norm": 0.31921207904815674,
      "learning_rate": 4.4699999999999996e-06,
      "loss": 0.3438,
      "step": 4553
    },
    {
      "epoch": 414.0,
      "grad_norm": 0.4776359498500824,
      "learning_rate": 4.4600000000000005e-06,
      "loss": 0.3055,
      "step": 4554
    },
    {
      "epoch": 414.0969696969697,
      "grad_norm": 0.38110432028770447,
      "learning_rate": 4.45e-06,
      "loss": 0.3258,
      "step": 4555
    },
    {
      "epoch": 414.1939393939394,
      "grad_norm": 0.3182128965854645,
      "learning_rate": 4.440000000000001e-06,
      "loss": 0.2973,
      "step": 4556
    },
    {
      "epoch": 414.2909090909091,
      "grad_norm": 0.2799331247806549,
      "learning_rate": 4.43e-06,
      "loss": 0.3438,
      "step": 4557
    },
    {
      "epoch": 414.3878787878788,
      "grad_norm": 0.2912656366825104,
      "learning_rate": 4.420000000000001e-06,
      "loss": 0.3162,
      "step": 4558
    },
    {
      "epoch": 414.4848484848485,
      "grad_norm": 0.2915773391723633,
      "learning_rate": 4.41e-06,
      "loss": 0.3664,
      "step": 4559
    },
    {
      "epoch": 414.58181818181816,
      "grad_norm": 0.2833085060119629,
      "learning_rate": 4.4e-06,
      "loss": 0.3063,
      "step": 4560
    },
    {
      "epoch": 414.58181818181816,
      "eval_loss": 0.4240158498287201,
      "eval_runtime": 2.107,
      "eval_samples_per_second": 26.103,
      "eval_steps_per_second": 3.322,
      "step": 4560
    },
    {
      "epoch": 414.6787878787879,
      "grad_norm": 0.28285571932792664,
      "learning_rate": 4.39e-06,
      "loss": 0.325,
      "step": 4561
    },
    {
      "epoch": 414.77575757575755,
      "grad_norm": 0.2957201898097992,
      "learning_rate": 4.38e-06,
      "loss": 0.3156,
      "step": 4562
    },
    {
      "epoch": 414.8727272727273,
      "grad_norm": 0.28091609477996826,
      "learning_rate": 4.3700000000000005e-06,
      "loss": 0.3411,
      "step": 4563
    },
    {
      "epoch": 414.969696969697,
      "grad_norm": 0.3014490306377411,
      "learning_rate": 4.360000000000001e-06,
      "loss": 0.3379,
      "step": 4564
    },
    {
      "epoch": 415.0,
      "grad_norm": 0.4831256866455078,
      "learning_rate": 4.35e-06,
      "loss": 0.3259,
      "step": 4565
    },
    {
      "epoch": 415.0969696969697,
      "grad_norm": 0.27512794733047485,
      "learning_rate": 4.34e-06,
      "loss": 0.3098,
      "step": 4566
    },
    {
      "epoch": 415.1939393939394,
      "grad_norm": 0.28693804144859314,
      "learning_rate": 4.33e-06,
      "loss": 0.3027,
      "step": 4567
    },
    {
      "epoch": 415.2909090909091,
      "grad_norm": 0.32689154148101807,
      "learning_rate": 4.32e-06,
      "loss": 0.3277,
      "step": 4568
    },
    {
      "epoch": 415.3878787878788,
      "grad_norm": 0.32227641344070435,
      "learning_rate": 4.31e-06,
      "loss": 0.3702,
      "step": 4569
    },
    {
      "epoch": 415.4848484848485,
      "grad_norm": 0.271851122379303,
      "learning_rate": 4.2999999999999995e-06,
      "loss": 0.3373,
      "step": 4570
    },
    {
      "epoch": 415.4848484848485,
      "eval_loss": 0.42395830154418945,
      "eval_runtime": 2.1299,
      "eval_samples_per_second": 25.823,
      "eval_steps_per_second": 3.287,
      "step": 4570
    },
    {
      "epoch": 415.58181818181816,
      "grad_norm": 0.28459352254867554,
      "learning_rate": 4.2900000000000004e-06,
      "loss": 0.3204,
      "step": 4571
    },
    {
      "epoch": 415.6787878787879,
      "grad_norm": 0.2655867636203766,
      "learning_rate": 4.28e-06,
      "loss": 0.2963,
      "step": 4572
    },
    {
      "epoch": 415.77575757575755,
      "grad_norm": 0.29420235753059387,
      "learning_rate": 4.270000000000001e-06,
      "loss": 0.3227,
      "step": 4573
    },
    {
      "epoch": 415.8727272727273,
      "grad_norm": 0.27869269251823425,
      "learning_rate": 4.26e-06,
      "loss": 0.3522,
      "step": 4574
    },
    {
      "epoch": 415.969696969697,
      "grad_norm": 0.27301132678985596,
      "learning_rate": 4.250000000000001e-06,
      "loss": 0.3521,
      "step": 4575
    },
    {
      "epoch": 416.0,
      "grad_norm": 0.517616868019104,
      "learning_rate": 4.24e-06,
      "loss": 0.263,
      "step": 4576
    },
    {
      "epoch": 416.0969696969697,
      "grad_norm": 0.2560671269893646,
      "learning_rate": 4.23e-06,
      "loss": 0.3247,
      "step": 4577
    },
    {
      "epoch": 416.1939393939394,
      "grad_norm": 0.28626900911331177,
      "learning_rate": 4.22e-06,
      "loss": 0.3398,
      "step": 4578
    },
    {
      "epoch": 416.2909090909091,
      "grad_norm": 0.25116318464279175,
      "learning_rate": 4.21e-06,
      "loss": 0.3497,
      "step": 4579
    },
    {
      "epoch": 416.3878787878788,
      "grad_norm": 0.2865499258041382,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.3258,
      "step": 4580
    },
    {
      "epoch": 416.3878787878788,
      "eval_loss": 0.4238404333591461,
      "eval_runtime": 2.1121,
      "eval_samples_per_second": 26.041,
      "eval_steps_per_second": 3.314,
      "step": 4580
    },
    {
      "epoch": 416.4848484848485,
      "grad_norm": 0.2667963206768036,
      "learning_rate": 4.1900000000000005e-06,
      "loss": 0.3043,
      "step": 4581
    },
    {
      "epoch": 416.58181818181816,
      "grad_norm": 0.3104817569255829,
      "learning_rate": 4.18e-06,
      "loss": 0.3103,
      "step": 4582
    },
    {
      "epoch": 416.6787878787879,
      "grad_norm": 0.34828588366508484,
      "learning_rate": 4.17e-06,
      "loss": 0.3354,
      "step": 4583
    },
    {
      "epoch": 416.77575757575755,
      "grad_norm": 0.23135364055633545,
      "learning_rate": 4.16e-06,
      "loss": 0.3084,
      "step": 4584
    },
    {
      "epoch": 416.8727272727273,
      "grad_norm": 0.2692182660102844,
      "learning_rate": 4.15e-06,
      "loss": 0.3481,
      "step": 4585
    },
    {
      "epoch": 416.969696969697,
      "grad_norm": 0.300054132938385,
      "learning_rate": 4.14e-06,
      "loss": 0.3198,
      "step": 4586
    },
    {
      "epoch": 417.0,
      "grad_norm": 0.6591640710830688,
      "learning_rate": 4.13e-06,
      "loss": 0.3431,
      "step": 4587
    },
    {
      "epoch": 417.0969696969697,
      "grad_norm": 0.2736823558807373,
      "learning_rate": 4.12e-06,
      "loss": 0.3235,
      "step": 4588
    },
    {
      "epoch": 417.1939393939394,
      "grad_norm": 0.30306321382522583,
      "learning_rate": 4.11e-06,
      "loss": 0.3591,
      "step": 4589
    },
    {
      "epoch": 417.2909090909091,
      "grad_norm": 0.2945916950702667,
      "learning_rate": 4.1000000000000006e-06,
      "loss": 0.3196,
      "step": 4590
    },
    {
      "epoch": 417.2909090909091,
      "eval_loss": 0.42390501499176025,
      "eval_runtime": 2.1312,
      "eval_samples_per_second": 25.807,
      "eval_steps_per_second": 3.285,
      "step": 4590
    },
    {
      "epoch": 417.3878787878788,
      "grad_norm": 0.3049025535583496,
      "learning_rate": 4.09e-06,
      "loss": 0.2935,
      "step": 4591
    },
    {
      "epoch": 417.4848484848485,
      "grad_norm": 0.2787110507488251,
      "learning_rate": 4.080000000000001e-06,
      "loss": 0.314,
      "step": 4592
    },
    {
      "epoch": 417.58181818181816,
      "grad_norm": 0.2911776006221771,
      "learning_rate": 4.07e-06,
      "loss": 0.3078,
      "step": 4593
    },
    {
      "epoch": 417.6787878787879,
      "grad_norm": 0.26729869842529297,
      "learning_rate": 4.06e-06,
      "loss": 0.3457,
      "step": 4594
    },
    {
      "epoch": 417.77575757575755,
      "grad_norm": 0.2834504544734955,
      "learning_rate": 4.05e-06,
      "loss": 0.3142,
      "step": 4595
    },
    {
      "epoch": 417.8727272727273,
      "grad_norm": 0.28690284490585327,
      "learning_rate": 4.04e-06,
      "loss": 0.3359,
      "step": 4596
    },
    {
      "epoch": 417.969696969697,
      "grad_norm": 0.2627941071987152,
      "learning_rate": 4.03e-06,
      "loss": 0.3468,
      "step": 4597
    },
    {
      "epoch": 418.0,
      "grad_norm": 0.5440768003463745,
      "learning_rate": 4.0200000000000005e-06,
      "loss": 0.3624,
      "step": 4598
    },
    {
      "epoch": 418.0969696969697,
      "grad_norm": 0.27621591091156006,
      "learning_rate": 4.01e-06,
      "loss": 0.3036,
      "step": 4599
    },
    {
      "epoch": 418.1939393939394,
      "grad_norm": 0.29796677827835083,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.336,
      "step": 4600
    },
    {
      "epoch": 418.1939393939394,
      "eval_loss": 0.4240822494029999,
      "eval_runtime": 2.1218,
      "eval_samples_per_second": 25.922,
      "eval_steps_per_second": 3.299,
      "step": 4600
    },
    {
      "epoch": 418.2909090909091,
      "grad_norm": 0.26402226090431213,
      "learning_rate": 3.99e-06,
      "loss": 0.3005,
      "step": 4601
    },
    {
      "epoch": 418.3878787878788,
      "grad_norm": 0.29714128375053406,
      "learning_rate": 3.98e-06,
      "loss": 0.3396,
      "step": 4602
    },
    {
      "epoch": 418.4848484848485,
      "grad_norm": 0.2947816848754883,
      "learning_rate": 3.97e-06,
      "loss": 0.3243,
      "step": 4603
    },
    {
      "epoch": 418.58181818181816,
      "grad_norm": 0.32920148968696594,
      "learning_rate": 3.96e-06,
      "loss": 0.346,
      "step": 4604
    },
    {
      "epoch": 418.6787878787879,
      "grad_norm": 0.3153909146785736,
      "learning_rate": 3.95e-06,
      "loss": 0.3285,
      "step": 4605
    },
    {
      "epoch": 418.77575757575755,
      "grad_norm": 0.28215932846069336,
      "learning_rate": 3.9399999999999995e-06,
      "loss": 0.3619,
      "step": 4606
    },
    {
      "epoch": 418.8727272727273,
      "grad_norm": 0.2899491786956787,
      "learning_rate": 3.9300000000000005e-06,
      "loss": 0.3283,
      "step": 4607
    },
    {
      "epoch": 418.969696969697,
      "grad_norm": 0.2625412940979004,
      "learning_rate": 3.92e-06,
      "loss": 0.2992,
      "step": 4608
    },
    {
      "epoch": 419.0,
      "grad_norm": 0.6752861142158508,
      "learning_rate": 3.910000000000001e-06,
      "loss": 0.3287,
      "step": 4609
    },
    {
      "epoch": 419.0969696969697,
      "grad_norm": 0.2802707254886627,
      "learning_rate": 3.9e-06,
      "loss": 0.339,
      "step": 4610
    },
    {
      "epoch": 419.0969696969697,
      "eval_loss": 0.4236716330051422,
      "eval_runtime": 2.1068,
      "eval_samples_per_second": 26.106,
      "eval_steps_per_second": 3.323,
      "step": 4610
    },
    {
      "epoch": 419.1939393939394,
      "grad_norm": 0.26663121581077576,
      "learning_rate": 3.89e-06,
      "loss": 0.3388,
      "step": 4611
    },
    {
      "epoch": 419.2909090909091,
      "grad_norm": 0.31221041083335876,
      "learning_rate": 3.88e-06,
      "loss": 0.2823,
      "step": 4612
    },
    {
      "epoch": 419.3878787878788,
      "grad_norm": 0.3001995086669922,
      "learning_rate": 3.87e-06,
      "loss": 0.3422,
      "step": 4613
    },
    {
      "epoch": 419.4848484848485,
      "grad_norm": 0.29675009846687317,
      "learning_rate": 3.86e-06,
      "loss": 0.3397,
      "step": 4614
    },
    {
      "epoch": 419.58181818181816,
      "grad_norm": 0.3295038938522339,
      "learning_rate": 3.85e-06,
      "loss": 0.3224,
      "step": 4615
    },
    {
      "epoch": 419.6787878787879,
      "grad_norm": 0.2736305892467499,
      "learning_rate": 3.84e-06,
      "loss": 0.317,
      "step": 4616
    },
    {
      "epoch": 419.77575757575755,
      "grad_norm": 0.32741448283195496,
      "learning_rate": 3.830000000000001e-06,
      "loss": 0.3323,
      "step": 4617
    },
    {
      "epoch": 419.8727272727273,
      "grad_norm": 0.2668072581291199,
      "learning_rate": 3.82e-06,
      "loss": 0.3605,
      "step": 4618
    },
    {
      "epoch": 419.969696969697,
      "grad_norm": 0.29451069235801697,
      "learning_rate": 3.8100000000000004e-06,
      "loss": 0.293,
      "step": 4619
    },
    {
      "epoch": 420.0,
      "grad_norm": 0.40569227933883667,
      "learning_rate": 3.8e-06,
      "loss": 0.3387,
      "step": 4620
    },
    {
      "epoch": 420.0,
      "eval_loss": 0.42370980978012085,
      "eval_runtime": 2.116,
      "eval_samples_per_second": 25.992,
      "eval_steps_per_second": 3.308,
      "step": 4620
    },
    {
      "epoch": 420.0969696969697,
      "grad_norm": 0.27761849761009216,
      "learning_rate": 3.7900000000000006e-06,
      "loss": 0.3157,
      "step": 4621
    },
    {
      "epoch": 420.1939393939394,
      "grad_norm": 0.31935086846351624,
      "learning_rate": 3.7800000000000002e-06,
      "loss": 0.3145,
      "step": 4622
    },
    {
      "epoch": 420.2909090909091,
      "grad_norm": 0.23702049255371094,
      "learning_rate": 3.77e-06,
      "loss": 0.2684,
      "step": 4623
    },
    {
      "epoch": 420.3878787878788,
      "grad_norm": 0.2601977586746216,
      "learning_rate": 3.7600000000000004e-06,
      "loss": 0.3297,
      "step": 4624
    },
    {
      "epoch": 420.4848484848485,
      "grad_norm": 0.26816099882125854,
      "learning_rate": 3.75e-06,
      "loss": 0.3325,
      "step": 4625
    },
    {
      "epoch": 420.58181818181816,
      "grad_norm": 0.2610798180103302,
      "learning_rate": 3.7400000000000006e-06,
      "loss": 0.2982,
      "step": 4626
    },
    {
      "epoch": 420.6787878787879,
      "grad_norm": 0.2762024700641632,
      "learning_rate": 3.7300000000000003e-06,
      "loss": 0.3697,
      "step": 4627
    },
    {
      "epoch": 420.77575757575755,
      "grad_norm": 0.29188522696495056,
      "learning_rate": 3.72e-06,
      "loss": 0.3499,
      "step": 4628
    },
    {
      "epoch": 420.8727272727273,
      "grad_norm": 0.30248361825942993,
      "learning_rate": 3.7100000000000005e-06,
      "loss": 0.3439,
      "step": 4629
    },
    {
      "epoch": 420.969696969697,
      "grad_norm": 0.29372766613960266,
      "learning_rate": 3.7e-06,
      "loss": 0.3216,
      "step": 4630
    },
    {
      "epoch": 420.969696969697,
      "eval_loss": 0.4238877594470978,
      "eval_runtime": 2.1381,
      "eval_samples_per_second": 25.724,
      "eval_steps_per_second": 3.274,
      "step": 4630
    },
    {
      "epoch": 421.0,
      "grad_norm": 0.4670877158641815,
      "learning_rate": 3.6900000000000002e-06,
      "loss": 0.3945,
      "step": 4631
    },
    {
      "epoch": 421.0969696969697,
      "grad_norm": 0.3275664150714874,
      "learning_rate": 3.68e-06,
      "loss": 0.3402,
      "step": 4632
    },
    {
      "epoch": 421.1939393939394,
      "grad_norm": 0.302420437335968,
      "learning_rate": 3.6700000000000004e-06,
      "loss": 0.3289,
      "step": 4633
    },
    {
      "epoch": 421.2909090909091,
      "grad_norm": 0.3075823485851288,
      "learning_rate": 3.66e-06,
      "loss": 0.2901,
      "step": 4634
    },
    {
      "epoch": 421.3878787878788,
      "grad_norm": 0.31915780901908875,
      "learning_rate": 3.6499999999999998e-06,
      "loss": 0.3143,
      "step": 4635
    },
    {
      "epoch": 421.4848484848485,
      "grad_norm": 0.2630278766155243,
      "learning_rate": 3.6400000000000003e-06,
      "loss": 0.3389,
      "step": 4636
    },
    {
      "epoch": 421.58181818181816,
      "grad_norm": 0.26861655712127686,
      "learning_rate": 3.63e-06,
      "loss": 0.3433,
      "step": 4637
    },
    {
      "epoch": 421.6787878787879,
      "grad_norm": 0.27271535992622375,
      "learning_rate": 3.6200000000000005e-06,
      "loss": 0.2828,
      "step": 4638
    },
    {
      "epoch": 421.77575757575755,
      "grad_norm": 0.28074952960014343,
      "learning_rate": 3.61e-06,
      "loss": 0.3431,
      "step": 4639
    },
    {
      "epoch": 421.8727272727273,
      "grad_norm": 0.29417550563812256,
      "learning_rate": 3.6e-06,
      "loss": 0.3302,
      "step": 4640
    },
    {
      "epoch": 421.8727272727273,
      "eval_loss": 0.4238133728504181,
      "eval_runtime": 2.1098,
      "eval_samples_per_second": 26.068,
      "eval_steps_per_second": 3.318,
      "step": 4640
    },
    {
      "epoch": 421.969696969697,
      "grad_norm": 0.29620280861854553,
      "learning_rate": 3.5900000000000004e-06,
      "loss": 0.3514,
      "step": 4641
    },
    {
      "epoch": 422.0,
      "grad_norm": 0.6159494519233704,
      "learning_rate": 3.58e-06,
      "loss": 0.3477,
      "step": 4642
    },
    {
      "epoch": 422.0969696969697,
      "grad_norm": 0.3032165765762329,
      "learning_rate": 3.5700000000000005e-06,
      "loss": 0.343,
      "step": 4643
    },
    {
      "epoch": 422.1939393939394,
      "grad_norm": 0.2550334632396698,
      "learning_rate": 3.5600000000000002e-06,
      "loss": 0.3302,
      "step": 4644
    },
    {
      "epoch": 422.2909090909091,
      "grad_norm": 0.27611789107322693,
      "learning_rate": 3.55e-06,
      "loss": 0.3381,
      "step": 4645
    },
    {
      "epoch": 422.3878787878788,
      "grad_norm": 0.28917229175567627,
      "learning_rate": 3.5400000000000004e-06,
      "loss": 0.326,
      "step": 4646
    },
    {
      "epoch": 422.4848484848485,
      "grad_norm": 0.3703041672706604,
      "learning_rate": 3.53e-06,
      "loss": 0.337,
      "step": 4647
    },
    {
      "epoch": 422.58181818181816,
      "grad_norm": 0.24577577412128448,
      "learning_rate": 3.52e-06,
      "loss": 0.3145,
      "step": 4648
    },
    {
      "epoch": 422.6787878787879,
      "grad_norm": 0.27131393551826477,
      "learning_rate": 3.5100000000000003e-06,
      "loss": 0.3543,
      "step": 4649
    },
    {
      "epoch": 422.77575757575755,
      "grad_norm": 0.24354679882526398,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 0.2833,
      "step": 4650
    },
    {
      "epoch": 422.77575757575755,
      "eval_loss": 0.42374733090400696,
      "eval_runtime": 2.134,
      "eval_samples_per_second": 25.773,
      "eval_steps_per_second": 3.28,
      "step": 4650
    },
    {
      "epoch": 422.8727272727273,
      "grad_norm": 0.2760464549064636,
      "learning_rate": 3.49e-06,
      "loss": 0.309,
      "step": 4651
    },
    {
      "epoch": 422.969696969697,
      "grad_norm": 0.30734682083129883,
      "learning_rate": 3.4799999999999997e-06,
      "loss": 0.3718,
      "step": 4652
    },
    {
      "epoch": 423.0,
      "grad_norm": 0.4714656472206116,
      "learning_rate": 3.4700000000000002e-06,
      "loss": 0.2027,
      "step": 4653
    },
    {
      "epoch": 423.0969696969697,
      "grad_norm": 0.2786318361759186,
      "learning_rate": 3.46e-06,
      "loss": 0.3422,
      "step": 4654
    },
    {
      "epoch": 423.1939393939394,
      "grad_norm": 0.31899377703666687,
      "learning_rate": 3.4500000000000004e-06,
      "loss": 0.3328,
      "step": 4655
    },
    {
      "epoch": 423.2909090909091,
      "grad_norm": 0.2987779378890991,
      "learning_rate": 3.44e-06,
      "loss": 0.3495,
      "step": 4656
    },
    {
      "epoch": 423.3878787878788,
      "grad_norm": 0.3024296462535858,
      "learning_rate": 3.4299999999999998e-06,
      "loss": 0.289,
      "step": 4657
    },
    {
      "epoch": 423.4848484848485,
      "grad_norm": 0.33767056465148926,
      "learning_rate": 3.4200000000000003e-06,
      "loss": 0.3157,
      "step": 4658
    },
    {
      "epoch": 423.58181818181816,
      "grad_norm": 0.2794208228588104,
      "learning_rate": 3.41e-06,
      "loss": 0.3241,
      "step": 4659
    },
    {
      "epoch": 423.6787878787879,
      "grad_norm": 0.2814173698425293,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.332,
      "step": 4660
    },
    {
      "epoch": 423.6787878787879,
      "eval_loss": 0.4236131012439728,
      "eval_runtime": 2.1268,
      "eval_samples_per_second": 25.861,
      "eval_steps_per_second": 3.291,
      "step": 4660
    },
    {
      "epoch": 423.77575757575755,
      "grad_norm": 0.2672629952430725,
      "learning_rate": 3.39e-06,
      "loss": 0.3408,
      "step": 4661
    },
    {
      "epoch": 423.8727272727273,
      "grad_norm": 0.26779016852378845,
      "learning_rate": 3.38e-06,
      "loss": 0.3353,
      "step": 4662
    },
    {
      "epoch": 423.969696969697,
      "grad_norm": 0.27753445506095886,
      "learning_rate": 3.3700000000000003e-06,
      "loss": 0.3237,
      "step": 4663
    },
    {
      "epoch": 424.0,
      "grad_norm": 0.4725082218647003,
      "learning_rate": 3.36e-06,
      "loss": 0.2746,
      "step": 4664
    },
    {
      "epoch": 424.0969696969697,
      "grad_norm": 0.28391534090042114,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.3008,
      "step": 4665
    },
    {
      "epoch": 424.1939393939394,
      "grad_norm": 0.29803451895713806,
      "learning_rate": 3.34e-06,
      "loss": 0.3041,
      "step": 4666
    },
    {
      "epoch": 424.2909090909091,
      "grad_norm": 0.27545270323753357,
      "learning_rate": 3.3300000000000003e-06,
      "loss": 0.308,
      "step": 4667
    },
    {
      "epoch": 424.3878787878788,
      "grad_norm": 0.28638580441474915,
      "learning_rate": 3.3200000000000004e-06,
      "loss": 0.381,
      "step": 4668
    },
    {
      "epoch": 424.4848484848485,
      "grad_norm": 0.24369746446609497,
      "learning_rate": 3.31e-06,
      "loss": 0.298,
      "step": 4669
    },
    {
      "epoch": 424.58181818181816,
      "grad_norm": 0.27538207173347473,
      "learning_rate": 3.3e-06,
      "loss": 0.3337,
      "step": 4670
    },
    {
      "epoch": 424.58181818181816,
      "eval_loss": 0.4239676594734192,
      "eval_runtime": 2.1146,
      "eval_samples_per_second": 26.01,
      "eval_steps_per_second": 3.31,
      "step": 4670
    },
    {
      "epoch": 424.6787878787879,
      "grad_norm": 0.31407785415649414,
      "learning_rate": 3.29e-06,
      "loss": 0.3383,
      "step": 4671
    },
    {
      "epoch": 424.77575757575755,
      "grad_norm": 0.30840155482292175,
      "learning_rate": 3.2800000000000004e-06,
      "loss": 0.306,
      "step": 4672
    },
    {
      "epoch": 424.8727272727273,
      "grad_norm": 0.27344176173210144,
      "learning_rate": 3.27e-06,
      "loss": 0.331,
      "step": 4673
    },
    {
      "epoch": 424.969696969697,
      "grad_norm": 0.3199271857738495,
      "learning_rate": 3.2599999999999997e-06,
      "loss": 0.36,
      "step": 4674
    },
    {
      "epoch": 425.0,
      "grad_norm": 0.4382783770561218,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.35,
      "step": 4675
    },
    {
      "epoch": 425.0969696969697,
      "grad_norm": 0.30849358439445496,
      "learning_rate": 3.24e-06,
      "loss": 0.3223,
      "step": 4676
    },
    {
      "epoch": 425.1939393939394,
      "grad_norm": 0.28263401985168457,
      "learning_rate": 3.2300000000000004e-06,
      "loss": 0.3385,
      "step": 4677
    },
    {
      "epoch": 425.2909090909091,
      "grad_norm": 0.33110207319259644,
      "learning_rate": 3.22e-06,
      "loss": 0.3222,
      "step": 4678
    },
    {
      "epoch": 425.3878787878788,
      "grad_norm": 0.25571611523628235,
      "learning_rate": 3.2099999999999998e-06,
      "loss": 0.3254,
      "step": 4679
    },
    {
      "epoch": 425.4848484848485,
      "grad_norm": 0.3212851285934448,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.3227,
      "step": 4680
    },
    {
      "epoch": 425.4848484848485,
      "eval_loss": 0.4238819181919098,
      "eval_runtime": 2.1317,
      "eval_samples_per_second": 25.801,
      "eval_steps_per_second": 3.284,
      "step": 4680
    },
    {
      "epoch": 425.58181818181816,
      "grad_norm": 0.3157774806022644,
      "learning_rate": 3.19e-06,
      "loss": 0.3397,
      "step": 4681
    },
    {
      "epoch": 425.6787878787879,
      "grad_norm": 0.2688293755054474,
      "learning_rate": 3.1800000000000005e-06,
      "loss": 0.3423,
      "step": 4682
    },
    {
      "epoch": 425.77575757575755,
      "grad_norm": 0.2958885729312897,
      "learning_rate": 3.17e-06,
      "loss": 0.3099,
      "step": 4683
    },
    {
      "epoch": 425.8727272727273,
      "grad_norm": 0.2832617163658142,
      "learning_rate": 3.1600000000000007e-06,
      "loss": 0.3393,
      "step": 4684
    },
    {
      "epoch": 425.969696969697,
      "grad_norm": 0.32423144578933716,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.3021,
      "step": 4685
    },
    {
      "epoch": 426.0,
      "grad_norm": 0.4971548318862915,
      "learning_rate": 3.14e-06,
      "loss": 0.3438,
      "step": 4686
    },
    {
      "epoch": 426.0969696969697,
      "grad_norm": 0.276876837015152,
      "learning_rate": 3.13e-06,
      "loss": 0.3225,
      "step": 4687
    },
    {
      "epoch": 426.1939393939394,
      "grad_norm": 0.2893661856651306,
      "learning_rate": 3.12e-06,
      "loss": 0.3164,
      "step": 4688
    },
    {
      "epoch": 426.2909090909091,
      "grad_norm": 0.3256908357143402,
      "learning_rate": 3.11e-06,
      "loss": 0.3228,
      "step": 4689
    },
    {
      "epoch": 426.3878787878788,
      "grad_norm": 0.30764877796173096,
      "learning_rate": 3.1e-06,
      "loss": 0.3429,
      "step": 4690
    },
    {
      "epoch": 426.3878787878788,
      "eval_loss": 0.4238271415233612,
      "eval_runtime": 2.1183,
      "eval_samples_per_second": 25.965,
      "eval_steps_per_second": 3.305,
      "step": 4690
    },
    {
      "epoch": 426.4848484848485,
      "grad_norm": 0.25312313437461853,
      "learning_rate": 3.09e-06,
      "loss": 0.318,
      "step": 4691
    },
    {
      "epoch": 426.58181818181816,
      "grad_norm": 0.32886362075805664,
      "learning_rate": 3.08e-06,
      "loss": 0.3334,
      "step": 4692
    },
    {
      "epoch": 426.6787878787879,
      "grad_norm": 0.2524215579032898,
      "learning_rate": 3.0700000000000003e-06,
      "loss": 0.3206,
      "step": 4693
    },
    {
      "epoch": 426.77575757575755,
      "grad_norm": 0.31209561228752136,
      "learning_rate": 3.06e-06,
      "loss": 0.324,
      "step": 4694
    },
    {
      "epoch": 426.8727272727273,
      "grad_norm": 0.2784077823162079,
      "learning_rate": 3.05e-06,
      "loss": 0.3411,
      "step": 4695
    },
    {
      "epoch": 426.969696969697,
      "grad_norm": 0.2601439952850342,
      "learning_rate": 3.04e-06,
      "loss": 0.3427,
      "step": 4696
    },
    {
      "epoch": 427.0,
      "grad_norm": 0.5130738019943237,
      "learning_rate": 3.0300000000000002e-06,
      "loss": 0.2719,
      "step": 4697
    },
    {
      "epoch": 427.0969696969697,
      "grad_norm": 0.26737532019615173,
      "learning_rate": 3.0200000000000003e-06,
      "loss": 0.3273,
      "step": 4698
    },
    {
      "epoch": 427.1939393939394,
      "grad_norm": 0.2933454215526581,
      "learning_rate": 3.01e-06,
      "loss": 0.3311,
      "step": 4699
    },
    {
      "epoch": 427.2909090909091,
      "grad_norm": 0.30595117807388306,
      "learning_rate": 3e-06,
      "loss": 0.3515,
      "step": 4700
    },
    {
      "epoch": 427.2909090909091,
      "eval_loss": 0.4236823320388794,
      "eval_runtime": 2.1384,
      "eval_samples_per_second": 25.721,
      "eval_steps_per_second": 3.274,
      "step": 4700
    },
    {
      "epoch": 427.3878787878788,
      "grad_norm": 0.31341302394866943,
      "learning_rate": 2.99e-06,
      "loss": 0.3082,
      "step": 4701
    },
    {
      "epoch": 427.4848484848485,
      "grad_norm": 0.28494933247566223,
      "learning_rate": 2.9800000000000003e-06,
      "loss": 0.3256,
      "step": 4702
    },
    {
      "epoch": 427.58181818181816,
      "grad_norm": 0.2777785062789917,
      "learning_rate": 2.9700000000000004e-06,
      "loss": 0.3099,
      "step": 4703
    },
    {
      "epoch": 427.6787878787879,
      "grad_norm": 0.25094926357269287,
      "learning_rate": 2.9600000000000005e-06,
      "loss": 0.3432,
      "step": 4704
    },
    {
      "epoch": 427.77575757575755,
      "grad_norm": 0.29807037115097046,
      "learning_rate": 2.95e-06,
      "loss": 0.315,
      "step": 4705
    },
    {
      "epoch": 427.8727272727273,
      "grad_norm": 0.3128386437892914,
      "learning_rate": 2.9400000000000002e-06,
      "loss": 0.3334,
      "step": 4706
    },
    {
      "epoch": 427.969696969697,
      "grad_norm": 0.3074195981025696,
      "learning_rate": 2.93e-06,
      "loss": 0.3195,
      "step": 4707
    },
    {
      "epoch": 428.0,
      "grad_norm": 0.4899100959300995,
      "learning_rate": 2.92e-06,
      "loss": 0.3367,
      "step": 4708
    },
    {
      "epoch": 428.0969696969697,
      "grad_norm": 0.2630816400051117,
      "learning_rate": 2.91e-06,
      "loss": 0.3628,
      "step": 4709
    },
    {
      "epoch": 428.1939393939394,
      "grad_norm": 0.3018128573894501,
      "learning_rate": 2.9e-06,
      "loss": 0.3246,
      "step": 4710
    },
    {
      "epoch": 428.1939393939394,
      "eval_loss": 0.42355218529701233,
      "eval_runtime": 2.1113,
      "eval_samples_per_second": 26.05,
      "eval_steps_per_second": 3.315,
      "step": 4710
    },
    {
      "epoch": 428.2909090909091,
      "grad_norm": 0.25907474756240845,
      "learning_rate": 2.89e-06,
      "loss": 0.3417,
      "step": 4711
    },
    {
      "epoch": 428.3878787878788,
      "grad_norm": 0.31002143025398254,
      "learning_rate": 2.88e-06,
      "loss": 0.3136,
      "step": 4712
    },
    {
      "epoch": 428.4848484848485,
      "grad_norm": 0.24571257829666138,
      "learning_rate": 2.87e-06,
      "loss": 0.3607,
      "step": 4713
    },
    {
      "epoch": 428.58181818181816,
      "grad_norm": 0.2842963933944702,
      "learning_rate": 2.86e-06,
      "loss": 0.296,
      "step": 4714
    },
    {
      "epoch": 428.6787878787879,
      "grad_norm": 0.36192312836647034,
      "learning_rate": 2.8500000000000002e-06,
      "loss": 0.3295,
      "step": 4715
    },
    {
      "epoch": 428.77575757575755,
      "grad_norm": 0.2653605043888092,
      "learning_rate": 2.8400000000000003e-06,
      "loss": 0.3109,
      "step": 4716
    },
    {
      "epoch": 428.8727272727273,
      "grad_norm": 0.29872456192970276,
      "learning_rate": 2.83e-06,
      "loss": 0.319,
      "step": 4717
    },
    {
      "epoch": 428.969696969697,
      "grad_norm": 0.2873593866825104,
      "learning_rate": 2.82e-06,
      "loss": 0.3029,
      "step": 4718
    },
    {
      "epoch": 429.0,
      "grad_norm": 0.5015831589698792,
      "learning_rate": 2.81e-06,
      "loss": 0.3301,
      "step": 4719
    },
    {
      "epoch": 429.0969696969697,
      "grad_norm": 0.26784220337867737,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.3132,
      "step": 4720
    },
    {
      "epoch": 429.0969696969697,
      "eval_loss": 0.42372432351112366,
      "eval_runtime": 2.1006,
      "eval_samples_per_second": 26.183,
      "eval_steps_per_second": 3.332,
      "step": 4720
    },
    {
      "epoch": 429.1939393939394,
      "grad_norm": 0.3056507706642151,
      "learning_rate": 2.7900000000000004e-06,
      "loss": 0.3011,
      "step": 4721
    },
    {
      "epoch": 429.2909090909091,
      "grad_norm": 0.30956873297691345,
      "learning_rate": 2.78e-06,
      "loss": 0.3318,
      "step": 4722
    },
    {
      "epoch": 429.3878787878788,
      "grad_norm": 0.28301090002059937,
      "learning_rate": 2.77e-06,
      "loss": 0.3312,
      "step": 4723
    },
    {
      "epoch": 429.4848484848485,
      "grad_norm": 0.30617964267730713,
      "learning_rate": 2.7600000000000003e-06,
      "loss": 0.3547,
      "step": 4724
    },
    {
      "epoch": 429.58181818181816,
      "grad_norm": 0.31053993105888367,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.3213,
      "step": 4725
    },
    {
      "epoch": 429.6787878787879,
      "grad_norm": 0.269156277179718,
      "learning_rate": 2.74e-06,
      "loss": 0.3257,
      "step": 4726
    },
    {
      "epoch": 429.77575757575755,
      "grad_norm": 0.25708848237991333,
      "learning_rate": 2.73e-06,
      "loss": 0.3304,
      "step": 4727
    },
    {
      "epoch": 429.8727272727273,
      "grad_norm": 0.2988244891166687,
      "learning_rate": 2.72e-06,
      "loss": 0.3393,
      "step": 4728
    },
    {
      "epoch": 429.969696969697,
      "grad_norm": 0.2603486180305481,
      "learning_rate": 2.71e-06,
      "loss": 0.3167,
      "step": 4729
    },
    {
      "epoch": 430.0,
      "grad_norm": 0.4041363000869751,
      "learning_rate": 2.7e-06,
      "loss": 0.3297,
      "step": 4730
    },
    {
      "epoch": 430.0,
      "eval_loss": 0.42374521493911743,
      "eval_runtime": 2.1174,
      "eval_samples_per_second": 25.976,
      "eval_steps_per_second": 3.306,
      "step": 4730
    },
    {
      "epoch": 430.0969696969697,
      "grad_norm": 0.27298206090927124,
      "learning_rate": 2.69e-06,
      "loss": 0.3604,
      "step": 4731
    },
    {
      "epoch": 430.1939393939394,
      "grad_norm": 0.2766759693622589,
      "learning_rate": 2.68e-06,
      "loss": 0.3354,
      "step": 4732
    },
    {
      "epoch": 430.2909090909091,
      "grad_norm": 0.26006969809532166,
      "learning_rate": 2.6700000000000003e-06,
      "loss": 0.3173,
      "step": 4733
    },
    {
      "epoch": 430.3878787878788,
      "grad_norm": 0.2893983721733093,
      "learning_rate": 2.66e-06,
      "loss": 0.3294,
      "step": 4734
    },
    {
      "epoch": 430.4848484848485,
      "grad_norm": 0.2810271680355072,
      "learning_rate": 2.65e-06,
      "loss": 0.3395,
      "step": 4735
    },
    {
      "epoch": 430.58181818181816,
      "grad_norm": 0.29783904552459717,
      "learning_rate": 2.64e-06,
      "loss": 0.3079,
      "step": 4736
    },
    {
      "epoch": 430.6787878787879,
      "grad_norm": 0.2597369849681854,
      "learning_rate": 2.6300000000000002e-06,
      "loss": 0.2866,
      "step": 4737
    },
    {
      "epoch": 430.77575757575755,
      "grad_norm": 0.27778446674346924,
      "learning_rate": 2.6200000000000003e-06,
      "loss": 0.3203,
      "step": 4738
    },
    {
      "epoch": 430.8727272727273,
      "grad_norm": 0.34003597497940063,
      "learning_rate": 2.6100000000000004e-06,
      "loss": 0.3483,
      "step": 4739
    },
    {
      "epoch": 430.969696969697,
      "grad_norm": 0.2701445519924164,
      "learning_rate": 2.6e-06,
      "loss": 0.3032,
      "step": 4740
    },
    {
      "epoch": 430.969696969697,
      "eval_loss": 0.4238017797470093,
      "eval_runtime": 2.1068,
      "eval_samples_per_second": 26.106,
      "eval_steps_per_second": 3.323,
      "step": 4740
    },
    {
      "epoch": 431.0,
      "grad_norm": 0.5419566035270691,
      "learning_rate": 2.59e-06,
      "loss": 0.3887,
      "step": 4741
    },
    {
      "epoch": 431.0969696969697,
      "grad_norm": 0.31240081787109375,
      "learning_rate": 2.5800000000000003e-06,
      "loss": 0.3338,
      "step": 4742
    },
    {
      "epoch": 431.1939393939394,
      "grad_norm": 0.28262385725975037,
      "learning_rate": 2.5700000000000004e-06,
      "loss": 0.3193,
      "step": 4743
    },
    {
      "epoch": 431.2909090909091,
      "grad_norm": 0.28926101326942444,
      "learning_rate": 2.56e-06,
      "loss": 0.3167,
      "step": 4744
    },
    {
      "epoch": 431.3878787878788,
      "grad_norm": 0.2820095419883728,
      "learning_rate": 2.55e-06,
      "loss": 0.3244,
      "step": 4745
    },
    {
      "epoch": 431.4848484848485,
      "grad_norm": 0.26937371492385864,
      "learning_rate": 2.54e-06,
      "loss": 0.3157,
      "step": 4746
    },
    {
      "epoch": 431.58181818181816,
      "grad_norm": 0.28882157802581787,
      "learning_rate": 2.53e-06,
      "loss": 0.3016,
      "step": 4747
    },
    {
      "epoch": 431.6787878787879,
      "grad_norm": 0.2726987600326538,
      "learning_rate": 2.52e-06,
      "loss": 0.3514,
      "step": 4748
    },
    {
      "epoch": 431.77575757575755,
      "grad_norm": 0.28297871351242065,
      "learning_rate": 2.51e-06,
      "loss": 0.335,
      "step": 4749
    },
    {
      "epoch": 431.8727272727273,
      "grad_norm": 0.2559443712234497,
      "learning_rate": 2.5e-06,
      "loss": 0.3212,
      "step": 4750
    },
    {
      "epoch": 431.8727272727273,
      "eval_loss": 0.4235769510269165,
      "eval_runtime": 2.108,
      "eval_samples_per_second": 26.091,
      "eval_steps_per_second": 3.321,
      "step": 4750
    },
    {
      "epoch": 431.969696969697,
      "grad_norm": 0.3145511746406555,
      "learning_rate": 2.49e-06,
      "loss": 0.3164,
      "step": 4751
    },
    {
      "epoch": 432.0,
      "grad_norm": 0.5010744333267212,
      "learning_rate": 2.48e-06,
      "loss": 0.4194,
      "step": 4752
    },
    {
      "epoch": 432.0969696969697,
      "grad_norm": 0.30744919180870056,
      "learning_rate": 2.47e-06,
      "loss": 0.3192,
      "step": 4753
    },
    {
      "epoch": 432.1939393939394,
      "grad_norm": 0.2515227794647217,
      "learning_rate": 2.46e-06,
      "loss": 0.3254,
      "step": 4754
    },
    {
      "epoch": 432.2909090909091,
      "grad_norm": 0.3341293931007385,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.3559,
      "step": 4755
    },
    {
      "epoch": 432.3878787878788,
      "grad_norm": 0.2512374222278595,
      "learning_rate": 2.4400000000000004e-06,
      "loss": 0.3434,
      "step": 4756
    },
    {
      "epoch": 432.4848484848485,
      "grad_norm": 0.2560592591762543,
      "learning_rate": 2.43e-06,
      "loss": 0.3145,
      "step": 4757
    },
    {
      "epoch": 432.58181818181816,
      "grad_norm": 0.27512484788894653,
      "learning_rate": 2.42e-06,
      "loss": 0.3072,
      "step": 4758
    },
    {
      "epoch": 432.6787878787879,
      "grad_norm": 0.3230624198913574,
      "learning_rate": 2.4100000000000002e-06,
      "loss": 0.3542,
      "step": 4759
    },
    {
      "epoch": 432.77575757575755,
      "grad_norm": 0.3115209639072418,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.323,
      "step": 4760
    },
    {
      "epoch": 432.77575757575755,
      "eval_loss": 0.42355164885520935,
      "eval_runtime": 2.1163,
      "eval_samples_per_second": 25.989,
      "eval_steps_per_second": 3.308,
      "step": 4760
    },
    {
      "epoch": 432.8727272727273,
      "grad_norm": 0.26935118436813354,
      "learning_rate": 2.3900000000000004e-06,
      "loss": 0.2812,
      "step": 4761
    },
    {
      "epoch": 432.969696969697,
      "grad_norm": 0.2984224557876587,
      "learning_rate": 2.38e-06,
      "loss": 0.3334,
      "step": 4762
    },
    {
      "epoch": 433.0,
      "grad_norm": 0.44937461614608765,
      "learning_rate": 2.37e-06,
      "loss": 0.3498,
      "step": 4763
    },
    {
      "epoch": 433.0969696969697,
      "grad_norm": 0.29262152314186096,
      "learning_rate": 2.36e-06,
      "loss": 0.3392,
      "step": 4764
    },
    {
      "epoch": 433.1939393939394,
      "grad_norm": 0.30966514348983765,
      "learning_rate": 2.35e-06,
      "loss": 0.3217,
      "step": 4765
    },
    {
      "epoch": 433.2909090909091,
      "grad_norm": 0.2939615845680237,
      "learning_rate": 2.34e-06,
      "loss": 0.3118,
      "step": 4766
    },
    {
      "epoch": 433.3878787878788,
      "grad_norm": 0.2691563367843628,
      "learning_rate": 2.33e-06,
      "loss": 0.3595,
      "step": 4767
    },
    {
      "epoch": 433.4848484848485,
      "grad_norm": 0.30199316143989563,
      "learning_rate": 2.32e-06,
      "loss": 0.3523,
      "step": 4768
    },
    {
      "epoch": 433.58181818181816,
      "grad_norm": 0.2727270722389221,
      "learning_rate": 2.31e-06,
      "loss": 0.3091,
      "step": 4769
    },
    {
      "epoch": 433.6787878787879,
      "grad_norm": 0.28975585103034973,
      "learning_rate": 2.3e-06,
      "loss": 0.3097,
      "step": 4770
    },
    {
      "epoch": 433.6787878787879,
      "eval_loss": 0.42369014024734497,
      "eval_runtime": 2.1122,
      "eval_samples_per_second": 26.039,
      "eval_steps_per_second": 3.314,
      "step": 4770
    },
    {
      "epoch": 433.77575757575755,
      "grad_norm": 0.2741400897502899,
      "learning_rate": 2.29e-06,
      "loss": 0.3175,
      "step": 4771
    },
    {
      "epoch": 433.8727272727273,
      "grad_norm": 0.3120625913143158,
      "learning_rate": 2.28e-06,
      "loss": 0.3066,
      "step": 4772
    },
    {
      "epoch": 433.969696969697,
      "grad_norm": 0.3074108362197876,
      "learning_rate": 2.2700000000000003e-06,
      "loss": 0.328,
      "step": 4773
    },
    {
      "epoch": 434.0,
      "grad_norm": 0.5580938458442688,
      "learning_rate": 2.26e-06,
      "loss": 0.3553,
      "step": 4774
    },
    {
      "epoch": 434.0969696969697,
      "grad_norm": 0.2596130073070526,
      "learning_rate": 2.25e-06,
      "loss": 0.3386,
      "step": 4775
    },
    {
      "epoch": 434.1939393939394,
      "grad_norm": 0.31736263632774353,
      "learning_rate": 2.24e-06,
      "loss": 0.3018,
      "step": 4776
    },
    {
      "epoch": 434.2909090909091,
      "grad_norm": 0.2828652560710907,
      "learning_rate": 2.2300000000000002e-06,
      "loss": 0.3408,
      "step": 4777
    },
    {
      "epoch": 434.3878787878788,
      "grad_norm": 0.2554495334625244,
      "learning_rate": 2.2200000000000003e-06,
      "loss": 0.3246,
      "step": 4778
    },
    {
      "epoch": 434.4848484848485,
      "grad_norm": 0.27445319294929504,
      "learning_rate": 2.2100000000000004e-06,
      "loss": 0.3136,
      "step": 4779
    },
    {
      "epoch": 434.58181818181816,
      "grad_norm": 0.3118382692337036,
      "learning_rate": 2.2e-06,
      "loss": 0.3171,
      "step": 4780
    },
    {
      "epoch": 434.58181818181816,
      "eval_loss": 0.42379724979400635,
      "eval_runtime": 2.1206,
      "eval_samples_per_second": 25.937,
      "eval_steps_per_second": 3.301,
      "step": 4780
    },
    {
      "epoch": 434.6787878787879,
      "grad_norm": 0.29958993196487427,
      "learning_rate": 2.19e-06,
      "loss": 0.3275,
      "step": 4781
    },
    {
      "epoch": 434.77575757575755,
      "grad_norm": 0.24797827005386353,
      "learning_rate": 2.1800000000000003e-06,
      "loss": 0.303,
      "step": 4782
    },
    {
      "epoch": 434.8727272727273,
      "grad_norm": 0.2977842688560486,
      "learning_rate": 2.17e-06,
      "loss": 0.3703,
      "step": 4783
    },
    {
      "epoch": 434.969696969697,
      "grad_norm": 0.26860713958740234,
      "learning_rate": 2.16e-06,
      "loss": 0.335,
      "step": 4784
    },
    {
      "epoch": 435.0,
      "grad_norm": 0.47584444284439087,
      "learning_rate": 2.1499999999999997e-06,
      "loss": 0.2958,
      "step": 4785
    },
    {
      "epoch": 435.0969696969697,
      "grad_norm": 0.2772142291069031,
      "learning_rate": 2.14e-06,
      "loss": 0.3721,
      "step": 4786
    },
    {
      "epoch": 435.1939393939394,
      "grad_norm": 0.294428288936615,
      "learning_rate": 2.13e-06,
      "loss": 0.3096,
      "step": 4787
    },
    {
      "epoch": 435.2909090909091,
      "grad_norm": 0.3144077956676483,
      "learning_rate": 2.12e-06,
      "loss": 0.3143,
      "step": 4788
    },
    {
      "epoch": 435.3878787878788,
      "grad_norm": 0.2641150951385498,
      "learning_rate": 2.11e-06,
      "loss": 0.2965,
      "step": 4789
    },
    {
      "epoch": 435.4848484848485,
      "grad_norm": 0.2828153371810913,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.309,
      "step": 4790
    },
    {
      "epoch": 435.4848484848485,
      "eval_loss": 0.4236714243888855,
      "eval_runtime": 2.1076,
      "eval_samples_per_second": 26.096,
      "eval_steps_per_second": 3.321,
      "step": 4790
    },
    {
      "epoch": 435.58181818181816,
      "grad_norm": 0.29111000895500183,
      "learning_rate": 2.09e-06,
      "loss": 0.3343,
      "step": 4791
    },
    {
      "epoch": 435.6787878787879,
      "grad_norm": 0.32535022497177124,
      "learning_rate": 2.08e-06,
      "loss": 0.3533,
      "step": 4792
    },
    {
      "epoch": 435.77575757575755,
      "grad_norm": 0.25760555267333984,
      "learning_rate": 2.07e-06,
      "loss": 0.3249,
      "step": 4793
    },
    {
      "epoch": 435.8727272727273,
      "grad_norm": 0.26453596353530884,
      "learning_rate": 2.06e-06,
      "loss": 0.346,
      "step": 4794
    },
    {
      "epoch": 435.969696969697,
      "grad_norm": 0.29465922713279724,
      "learning_rate": 2.0500000000000003e-06,
      "loss": 0.2883,
      "step": 4795
    },
    {
      "epoch": 436.0,
      "grad_norm": 0.5762606263160706,
      "learning_rate": 2.0400000000000004e-06,
      "loss": 0.3735,
      "step": 4796
    },
    {
      "epoch": 436.0969696969697,
      "grad_norm": 0.2903105914592743,
      "learning_rate": 2.03e-06,
      "loss": 0.3105,
      "step": 4797
    },
    {
      "epoch": 436.1939393939394,
      "grad_norm": 0.2878726124763489,
      "learning_rate": 2.02e-06,
      "loss": 0.3584,
      "step": 4798
    },
    {
      "epoch": 436.2909090909091,
      "grad_norm": 0.2933520972728729,
      "learning_rate": 2.0100000000000002e-06,
      "loss": 0.3282,
      "step": 4799
    },
    {
      "epoch": 436.3878787878788,
      "grad_norm": 0.270224928855896,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.3184,
      "step": 4800
    },
    {
      "epoch": 436.3878787878788,
      "eval_loss": 0.423602432012558,
      "eval_runtime": 2.0973,
      "eval_samples_per_second": 26.225,
      "eval_steps_per_second": 3.338,
      "step": 4800
    },
    {
      "epoch": 436.4848484848485,
      "grad_norm": 0.272318035364151,
      "learning_rate": 1.99e-06,
      "loss": 0.3204,
      "step": 4801
    },
    {
      "epoch": 436.58181818181816,
      "grad_norm": 0.30114665627479553,
      "learning_rate": 1.98e-06,
      "loss": 0.3378,
      "step": 4802
    },
    {
      "epoch": 436.6787878787879,
      "grad_norm": 0.28273192048072815,
      "learning_rate": 1.9699999999999998e-06,
      "loss": 0.3617,
      "step": 4803
    },
    {
      "epoch": 436.77575757575755,
      "grad_norm": 0.31538066267967224,
      "learning_rate": 1.96e-06,
      "loss": 0.3142,
      "step": 4804
    },
    {
      "epoch": 436.8727272727273,
      "grad_norm": 0.37554872035980225,
      "learning_rate": 1.95e-06,
      "loss": 0.3327,
      "step": 4805
    },
    {
      "epoch": 436.969696969697,
      "grad_norm": 0.30792397260665894,
      "learning_rate": 1.94e-06,
      "loss": 0.3026,
      "step": 4806
    },
    {
      "epoch": 437.0,
      "grad_norm": 0.5518594980239868,
      "learning_rate": 1.93e-06,
      "loss": 0.2628,
      "step": 4807
    },
    {
      "epoch": 437.0969696969697,
      "grad_norm": 0.3231973648071289,
      "learning_rate": 1.92e-06,
      "loss": 0.2984,
      "step": 4808
    },
    {
      "epoch": 437.1939393939394,
      "grad_norm": 0.2513819932937622,
      "learning_rate": 1.91e-06,
      "loss": 0.3173,
      "step": 4809
    },
    {
      "epoch": 437.2909090909091,
      "grad_norm": 0.275855153799057,
      "learning_rate": 1.9e-06,
      "loss": 0.3195,
      "step": 4810
    },
    {
      "epoch": 437.2909090909091,
      "eval_loss": 0.42367324233055115,
      "eval_runtime": 2.111,
      "eval_samples_per_second": 26.054,
      "eval_steps_per_second": 3.316,
      "step": 4810
    },
    {
      "epoch": 437.3878787878788,
      "grad_norm": 0.25522032380104065,
      "learning_rate": 1.8900000000000001e-06,
      "loss": 0.3273,
      "step": 4811
    },
    {
      "epoch": 437.4848484848485,
      "grad_norm": 0.29282745718955994,
      "learning_rate": 1.8800000000000002e-06,
      "loss": 0.3043,
      "step": 4812
    },
    {
      "epoch": 437.58181818181816,
      "grad_norm": 0.28065821528434753,
      "learning_rate": 1.8700000000000003e-06,
      "loss": 0.3699,
      "step": 4813
    },
    {
      "epoch": 437.6787878787879,
      "grad_norm": 0.27441632747650146,
      "learning_rate": 1.86e-06,
      "loss": 0.328,
      "step": 4814
    },
    {
      "epoch": 437.77575757575755,
      "grad_norm": 0.3121632933616638,
      "learning_rate": 1.85e-06,
      "loss": 0.3467,
      "step": 4815
    },
    {
      "epoch": 437.8727272727273,
      "grad_norm": 0.3079357147216797,
      "learning_rate": 1.84e-06,
      "loss": 0.3555,
      "step": 4816
    },
    {
      "epoch": 437.969696969697,
      "grad_norm": 0.2577809989452362,
      "learning_rate": 1.83e-06,
      "loss": 0.3061,
      "step": 4817
    },
    {
      "epoch": 438.0,
      "grad_norm": 0.45075249671936035,
      "learning_rate": 1.8200000000000002e-06,
      "loss": 0.2846,
      "step": 4818
    },
    {
      "epoch": 438.0969696969697,
      "grad_norm": 0.37741875648498535,
      "learning_rate": 1.8100000000000002e-06,
      "loss": 0.3472,
      "step": 4819
    },
    {
      "epoch": 438.1939393939394,
      "grad_norm": 0.3882526755332947,
      "learning_rate": 1.8e-06,
      "loss": 0.3513,
      "step": 4820
    },
    {
      "epoch": 438.1939393939394,
      "eval_loss": 0.4238298237323761,
      "eval_runtime": 2.1023,
      "eval_samples_per_second": 26.162,
      "eval_steps_per_second": 3.33,
      "step": 4820
    },
    {
      "epoch": 438.2909090909091,
      "grad_norm": 0.318301260471344,
      "learning_rate": 1.79e-06,
      "loss": 0.3422,
      "step": 4821
    },
    {
      "epoch": 438.3878787878788,
      "grad_norm": 0.2677038013935089,
      "learning_rate": 1.7800000000000001e-06,
      "loss": 0.3271,
      "step": 4822
    },
    {
      "epoch": 438.4848484848485,
      "grad_norm": 0.2726380228996277,
      "learning_rate": 1.7700000000000002e-06,
      "loss": 0.3278,
      "step": 4823
    },
    {
      "epoch": 438.58181818181816,
      "grad_norm": 0.2821105420589447,
      "learning_rate": 1.76e-06,
      "loss": 0.2994,
      "step": 4824
    },
    {
      "epoch": 438.6787878787879,
      "grad_norm": 0.3339035212993622,
      "learning_rate": 1.7500000000000002e-06,
      "loss": 0.2954,
      "step": 4825
    },
    {
      "epoch": 438.77575757575755,
      "grad_norm": 0.2941056489944458,
      "learning_rate": 1.7399999999999999e-06,
      "loss": 0.3239,
      "step": 4826
    },
    {
      "epoch": 438.8727272727273,
      "grad_norm": 0.2761934995651245,
      "learning_rate": 1.73e-06,
      "loss": 0.3361,
      "step": 4827
    },
    {
      "epoch": 438.969696969697,
      "grad_norm": 0.26881179213523865,
      "learning_rate": 1.72e-06,
      "loss": 0.321,
      "step": 4828
    },
    {
      "epoch": 439.0,
      "grad_norm": 0.5755116939544678,
      "learning_rate": 1.7100000000000001e-06,
      "loss": 0.3036,
      "step": 4829
    },
    {
      "epoch": 439.0969696969697,
      "grad_norm": 0.29740414023399353,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.307,
      "step": 4830
    },
    {
      "epoch": 439.0969696969697,
      "eval_loss": 0.4238336682319641,
      "eval_runtime": 2.1193,
      "eval_samples_per_second": 25.952,
      "eval_steps_per_second": 3.303,
      "step": 4830
    },
    {
      "epoch": 439.1939393939394,
      "grad_norm": 0.346254825592041,
      "learning_rate": 1.69e-06,
      "loss": 0.3137,
      "step": 4831
    },
    {
      "epoch": 439.2909090909091,
      "grad_norm": 0.2799615263938904,
      "learning_rate": 1.68e-06,
      "loss": 0.3146,
      "step": 4832
    },
    {
      "epoch": 439.3878787878788,
      "grad_norm": 0.269903302192688,
      "learning_rate": 1.67e-06,
      "loss": 0.3039,
      "step": 4833
    },
    {
      "epoch": 439.4848484848485,
      "grad_norm": 0.2790166735649109,
      "learning_rate": 1.6600000000000002e-06,
      "loss": 0.3179,
      "step": 4834
    },
    {
      "epoch": 439.58181818181816,
      "grad_norm": 0.3139094114303589,
      "learning_rate": 1.65e-06,
      "loss": 0.3358,
      "step": 4835
    },
    {
      "epoch": 439.6787878787879,
      "grad_norm": 0.29021573066711426,
      "learning_rate": 1.6400000000000002e-06,
      "loss": 0.3409,
      "step": 4836
    },
    {
      "epoch": 439.77575757575755,
      "grad_norm": 0.3001808822154999,
      "learning_rate": 1.6299999999999999e-06,
      "loss": 0.3347,
      "step": 4837
    },
    {
      "epoch": 439.8727272727273,
      "grad_norm": 0.2734721601009369,
      "learning_rate": 1.62e-06,
      "loss": 0.3314,
      "step": 4838
    },
    {
      "epoch": 439.969696969697,
      "grad_norm": 0.27748993039131165,
      "learning_rate": 1.61e-06,
      "loss": 0.3715,
      "step": 4839
    },
    {
      "epoch": 440.0,
      "grad_norm": 0.4819665849208832,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.3023,
      "step": 4840
    },
    {
      "epoch": 440.0,
      "eval_loss": 0.42376306653022766,
      "eval_runtime": 2.109,
      "eval_samples_per_second": 26.079,
      "eval_steps_per_second": 3.319,
      "step": 4840
    },
    {
      "epoch": 440.0969696969697,
      "grad_norm": 0.31397852301597595,
      "learning_rate": 1.5900000000000002e-06,
      "loss": 0.3496,
      "step": 4841
    },
    {
      "epoch": 440.1939393939394,
      "grad_norm": 0.3182499408721924,
      "learning_rate": 1.5800000000000003e-06,
      "loss": 0.3132,
      "step": 4842
    },
    {
      "epoch": 440.2909090909091,
      "grad_norm": 0.29344242811203003,
      "learning_rate": 1.57e-06,
      "loss": 0.2864,
      "step": 4843
    },
    {
      "epoch": 440.3878787878788,
      "grad_norm": 0.27054810523986816,
      "learning_rate": 1.56e-06,
      "loss": 0.338,
      "step": 4844
    },
    {
      "epoch": 440.4848484848485,
      "grad_norm": 0.2715929448604584,
      "learning_rate": 1.55e-06,
      "loss": 0.3441,
      "step": 4845
    },
    {
      "epoch": 440.58181818181816,
      "grad_norm": 0.27300938963890076,
      "learning_rate": 1.54e-06,
      "loss": 0.3072,
      "step": 4846
    },
    {
      "epoch": 440.6787878787879,
      "grad_norm": 0.2786642611026764,
      "learning_rate": 1.53e-06,
      "loss": 0.313,
      "step": 4847
    },
    {
      "epoch": 440.77575757575755,
      "grad_norm": 0.269209086894989,
      "learning_rate": 1.52e-06,
      "loss": 0.3338,
      "step": 4848
    },
    {
      "epoch": 440.8727272727273,
      "grad_norm": 0.26223093271255493,
      "learning_rate": 1.5100000000000002e-06,
      "loss": 0.3319,
      "step": 4849
    },
    {
      "epoch": 440.969696969697,
      "grad_norm": 0.28911757469177246,
      "learning_rate": 1.5e-06,
      "loss": 0.3457,
      "step": 4850
    },
    {
      "epoch": 440.969696969697,
      "eval_loss": 0.4236938953399658,
      "eval_runtime": 2.1201,
      "eval_samples_per_second": 25.943,
      "eval_steps_per_second": 3.302,
      "step": 4850
    },
    {
      "epoch": 441.0,
      "grad_norm": 0.5196654796600342,
      "learning_rate": 1.4900000000000001e-06,
      "loss": 0.3324,
      "step": 4851
    },
    {
      "epoch": 441.0969696969697,
      "grad_norm": 0.2900083363056183,
      "learning_rate": 1.4800000000000002e-06,
      "loss": 0.3761,
      "step": 4852
    },
    {
      "epoch": 441.1939393939394,
      "grad_norm": 0.2578149735927582,
      "learning_rate": 1.4700000000000001e-06,
      "loss": 0.3005,
      "step": 4853
    },
    {
      "epoch": 441.2909090909091,
      "grad_norm": 0.27350473403930664,
      "learning_rate": 1.46e-06,
      "loss": 0.3364,
      "step": 4854
    },
    {
      "epoch": 441.3878787878788,
      "grad_norm": 0.28097885847091675,
      "learning_rate": 1.45e-06,
      "loss": 0.3353,
      "step": 4855
    },
    {
      "epoch": 441.4848484848485,
      "grad_norm": 0.3071709871292114,
      "learning_rate": 1.44e-06,
      "loss": 0.3351,
      "step": 4856
    },
    {
      "epoch": 441.58181818181816,
      "grad_norm": 0.382062703371048,
      "learning_rate": 1.43e-06,
      "loss": 0.3114,
      "step": 4857
    },
    {
      "epoch": 441.6787878787879,
      "grad_norm": 0.27680331468582153,
      "learning_rate": 1.4200000000000002e-06,
      "loss": 0.3436,
      "step": 4858
    },
    {
      "epoch": 441.77575757575755,
      "grad_norm": 0.2957818806171417,
      "learning_rate": 1.41e-06,
      "loss": 0.3129,
      "step": 4859
    },
    {
      "epoch": 441.8727272727273,
      "grad_norm": 0.2488706111907959,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.3263,
      "step": 4860
    },
    {
      "epoch": 441.8727272727273,
      "eval_loss": 0.4236072301864624,
      "eval_runtime": 2.1291,
      "eval_samples_per_second": 25.832,
      "eval_steps_per_second": 3.288,
      "step": 4860
    },
    {
      "epoch": 441.969696969697,
      "grad_norm": 0.29431694746017456,
      "learning_rate": 1.39e-06,
      "loss": 0.2953,
      "step": 4861
    },
    {
      "epoch": 442.0,
      "grad_norm": 0.5485882759094238,
      "learning_rate": 1.3800000000000001e-06,
      "loss": 0.2916,
      "step": 4862
    },
    {
      "epoch": 442.0969696969697,
      "grad_norm": 0.2963046729564667,
      "learning_rate": 1.37e-06,
      "loss": 0.301,
      "step": 4863
    },
    {
      "epoch": 442.1939393939394,
      "grad_norm": 0.27449437975883484,
      "learning_rate": 1.36e-06,
      "loss": 0.2883,
      "step": 4864
    },
    {
      "epoch": 442.2909090909091,
      "grad_norm": 0.25899559259414673,
      "learning_rate": 1.35e-06,
      "loss": 0.3357,
      "step": 4865
    },
    {
      "epoch": 442.3878787878788,
      "grad_norm": 0.29228493571281433,
      "learning_rate": 1.34e-06,
      "loss": 0.3331,
      "step": 4866
    },
    {
      "epoch": 442.4848484848485,
      "grad_norm": 0.2560674250125885,
      "learning_rate": 1.33e-06,
      "loss": 0.3252,
      "step": 4867
    },
    {
      "epoch": 442.58181818181816,
      "grad_norm": 0.2929626405239105,
      "learning_rate": 1.32e-06,
      "loss": 0.3488,
      "step": 4868
    },
    {
      "epoch": 442.6787878787879,
      "grad_norm": 0.26246029138565063,
      "learning_rate": 1.3100000000000002e-06,
      "loss": 0.3306,
      "step": 4869
    },
    {
      "epoch": 442.77575757575755,
      "grad_norm": 0.29994064569473267,
      "learning_rate": 1.3e-06,
      "loss": 0.3437,
      "step": 4870
    },
    {
      "epoch": 442.77575757575755,
      "eval_loss": 0.42368239164352417,
      "eval_runtime": 2.1083,
      "eval_samples_per_second": 26.087,
      "eval_steps_per_second": 3.32,
      "step": 4870
    },
    {
      "epoch": 442.8727272727273,
      "grad_norm": 0.336182564496994,
      "learning_rate": 1.2900000000000001e-06,
      "loss": 0.3325,
      "step": 4871
    },
    {
      "epoch": 442.969696969697,
      "grad_norm": 0.2670702636241913,
      "learning_rate": 1.28e-06,
      "loss": 0.3243,
      "step": 4872
    },
    {
      "epoch": 443.0,
      "grad_norm": 0.4987045228481293,
      "learning_rate": 1.27e-06,
      "loss": 0.3315,
      "step": 4873
    },
    {
      "epoch": 443.0969696969697,
      "grad_norm": 0.45853880047798157,
      "learning_rate": 1.26e-06,
      "loss": 0.2985,
      "step": 4874
    },
    {
      "epoch": 443.1939393939394,
      "grad_norm": 0.2865673899650574,
      "learning_rate": 1.25e-06,
      "loss": 0.3484,
      "step": 4875
    },
    {
      "epoch": 443.2909090909091,
      "grad_norm": 0.2569712698459625,
      "learning_rate": 1.24e-06,
      "loss": 0.3291,
      "step": 4876
    },
    {
      "epoch": 443.3878787878788,
      "grad_norm": 0.24614472687244415,
      "learning_rate": 1.23e-06,
      "loss": 0.336,
      "step": 4877
    },
    {
      "epoch": 443.4848484848485,
      "grad_norm": 0.2899981439113617,
      "learning_rate": 1.2200000000000002e-06,
      "loss": 0.3269,
      "step": 4878
    },
    {
      "epoch": 443.58181818181816,
      "grad_norm": 0.28798767924308777,
      "learning_rate": 1.21e-06,
      "loss": 0.351,
      "step": 4879
    },
    {
      "epoch": 443.6787878787879,
      "grad_norm": 0.2734663784503937,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.3152,
      "step": 4880
    },
    {
      "epoch": 443.6787878787879,
      "eval_loss": 0.423577219247818,
      "eval_runtime": 2.0931,
      "eval_samples_per_second": 26.277,
      "eval_steps_per_second": 3.344,
      "step": 4880
    },
    {
      "epoch": 443.77575757575755,
      "grad_norm": 0.26212504506111145,
      "learning_rate": 1.19e-06,
      "loss": 0.2981,
      "step": 4881
    },
    {
      "epoch": 443.8727272727273,
      "grad_norm": 0.2801170349121094,
      "learning_rate": 1.18e-06,
      "loss": 0.3277,
      "step": 4882
    },
    {
      "epoch": 443.969696969697,
      "grad_norm": 0.2888226807117462,
      "learning_rate": 1.17e-06,
      "loss": 0.3287,
      "step": 4883
    },
    {
      "epoch": 444.0,
      "grad_norm": 0.6596548557281494,
      "learning_rate": 1.16e-06,
      "loss": 0.3443,
      "step": 4884
    },
    {
      "epoch": 444.0969696969697,
      "grad_norm": 0.2544076442718506,
      "learning_rate": 1.15e-06,
      "loss": 0.3149,
      "step": 4885
    },
    {
      "epoch": 444.1939393939394,
      "grad_norm": 0.3435005247592926,
      "learning_rate": 1.14e-06,
      "loss": 0.329,
      "step": 4886
    },
    {
      "epoch": 444.2909090909091,
      "grad_norm": 0.31201043725013733,
      "learning_rate": 1.13e-06,
      "loss": 0.3529,
      "step": 4887
    },
    {
      "epoch": 444.3878787878788,
      "grad_norm": 0.27917465567588806,
      "learning_rate": 1.12e-06,
      "loss": 0.3457,
      "step": 4888
    },
    {
      "epoch": 444.4848484848485,
      "grad_norm": 0.3057820498943329,
      "learning_rate": 1.1100000000000002e-06,
      "loss": 0.3285,
      "step": 4889
    },
    {
      "epoch": 444.58181818181816,
      "grad_norm": 0.26686015725135803,
      "learning_rate": 1.1e-06,
      "loss": 0.3251,
      "step": 4890
    },
    {
      "epoch": 444.58181818181816,
      "eval_loss": 0.4236905872821808,
      "eval_runtime": 2.0952,
      "eval_samples_per_second": 26.251,
      "eval_steps_per_second": 3.341,
      "step": 4890
    },
    {
      "epoch": 444.6787878787879,
      "grad_norm": 0.28641071915626526,
      "learning_rate": 1.0900000000000002e-06,
      "loss": 0.3135,
      "step": 4891
    },
    {
      "epoch": 444.77575757575755,
      "grad_norm": 0.2760150730609894,
      "learning_rate": 1.08e-06,
      "loss": 0.3069,
      "step": 4892
    },
    {
      "epoch": 444.8727272727273,
      "grad_norm": 0.266857773065567,
      "learning_rate": 1.07e-06,
      "loss": 0.346,
      "step": 4893
    },
    {
      "epoch": 444.969696969697,
      "grad_norm": 0.32249146699905396,
      "learning_rate": 1.06e-06,
      "loss": 0.317,
      "step": 4894
    },
    {
      "epoch": 445.0,
      "grad_norm": 0.41339409351348877,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 0.2677,
      "step": 4895
    },
    {
      "epoch": 445.0969696969697,
      "grad_norm": 0.27063751220703125,
      "learning_rate": 1.04e-06,
      "loss": 0.3249,
      "step": 4896
    },
    {
      "epoch": 445.1939393939394,
      "grad_norm": 0.2708544135093689,
      "learning_rate": 1.03e-06,
      "loss": 0.3461,
      "step": 4897
    },
    {
      "epoch": 445.2909090909091,
      "grad_norm": 0.3347277343273163,
      "learning_rate": 1.0200000000000002e-06,
      "loss": 0.3332,
      "step": 4898
    },
    {
      "epoch": 445.3878787878788,
      "grad_norm": 0.282702773809433,
      "learning_rate": 1.01e-06,
      "loss": 0.3284,
      "step": 4899
    },
    {
      "epoch": 445.4848484848485,
      "grad_norm": 0.30617809295654297,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.3039,
      "step": 4900
    },
    {
      "epoch": 445.4848484848485,
      "eval_loss": 0.4235691726207733,
      "eval_runtime": 2.1044,
      "eval_samples_per_second": 26.136,
      "eval_steps_per_second": 3.326,
      "step": 4900
    },
    {
      "epoch": 445.58181818181816,
      "grad_norm": 0.3247913122177124,
      "learning_rate": 9.9e-07,
      "loss": 0.3181,
      "step": 4901
    },
    {
      "epoch": 445.6787878787879,
      "grad_norm": 0.30320414900779724,
      "learning_rate": 9.8e-07,
      "loss": 0.3312,
      "step": 4902
    },
    {
      "epoch": 445.77575757575755,
      "grad_norm": 0.27612176537513733,
      "learning_rate": 9.7e-07,
      "loss": 0.3375,
      "step": 4903
    },
    {
      "epoch": 445.8727272727273,
      "grad_norm": 0.31707799434661865,
      "learning_rate": 9.6e-07,
      "loss": 0.3347,
      "step": 4904
    },
    {
      "epoch": 445.969696969697,
      "grad_norm": 0.28800132870674133,
      "learning_rate": 9.5e-07,
      "loss": 0.3175,
      "step": 4905
    },
    {
      "epoch": 446.0,
      "grad_norm": 0.6043686866760254,
      "learning_rate": 9.400000000000001e-07,
      "loss": 0.2893,
      "step": 4906
    },
    {
      "epoch": 446.0969696969697,
      "grad_norm": 0.2963293194770813,
      "learning_rate": 9.3e-07,
      "loss": 0.317,
      "step": 4907
    },
    {
      "epoch": 446.1939393939394,
      "grad_norm": 0.30245649814605713,
      "learning_rate": 9.2e-07,
      "loss": 0.3513,
      "step": 4908
    },
    {
      "epoch": 446.2909090909091,
      "grad_norm": 0.25606197118759155,
      "learning_rate": 9.100000000000001e-07,
      "loss": 0.3136,
      "step": 4909
    },
    {
      "epoch": 446.3878787878788,
      "grad_norm": 0.3276771903038025,
      "learning_rate": 9e-07,
      "loss": 0.3119,
      "step": 4910
    },
    {
      "epoch": 446.3878787878788,
      "eval_loss": 0.42358332872390747,
      "eval_runtime": 2.1075,
      "eval_samples_per_second": 26.098,
      "eval_steps_per_second": 3.322,
      "step": 4910
    },
    {
      "epoch": 446.4848484848485,
      "grad_norm": 0.28884825110435486,
      "learning_rate": 8.900000000000001e-07,
      "loss": 0.3349,
      "step": 4911
    },
    {
      "epoch": 446.58181818181816,
      "grad_norm": 0.29931309819221497,
      "learning_rate": 8.8e-07,
      "loss": 0.3571,
      "step": 4912
    },
    {
      "epoch": 446.6787878787879,
      "grad_norm": 0.30855095386505127,
      "learning_rate": 8.699999999999999e-07,
      "loss": 0.2973,
      "step": 4913
    },
    {
      "epoch": 446.77575757575755,
      "grad_norm": 0.3217792510986328,
      "learning_rate": 8.6e-07,
      "loss": 0.3319,
      "step": 4914
    },
    {
      "epoch": 446.8727272727273,
      "grad_norm": 0.2590404748916626,
      "learning_rate": 8.500000000000001e-07,
      "loss": 0.3124,
      "step": 4915
    },
    {
      "epoch": 446.969696969697,
      "grad_norm": 0.30771511793136597,
      "learning_rate": 8.4e-07,
      "loss": 0.3506,
      "step": 4916
    },
    {
      "epoch": 447.0,
      "grad_norm": 0.5211871862411499,
      "learning_rate": 8.300000000000001e-07,
      "loss": 0.2776,
      "step": 4917
    },
    {
      "epoch": 447.0969696969697,
      "grad_norm": 0.27772092819213867,
      "learning_rate": 8.200000000000001e-07,
      "loss": 0.3135,
      "step": 4918
    },
    {
      "epoch": 447.1939393939394,
      "grad_norm": 0.29392674565315247,
      "learning_rate": 8.1e-07,
      "loss": 0.3452,
      "step": 4919
    },
    {
      "epoch": 447.2909090909091,
      "grad_norm": 0.2605448365211487,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.3271,
      "step": 4920
    },
    {
      "epoch": 447.2909090909091,
      "eval_loss": 0.4235498607158661,
      "eval_runtime": 2.1047,
      "eval_samples_per_second": 26.132,
      "eval_steps_per_second": 3.326,
      "step": 4920
    },
    {
      "epoch": 447.3878787878788,
      "grad_norm": 0.29172196984291077,
      "learning_rate": 7.900000000000002e-07,
      "loss": 0.3317,
      "step": 4921
    },
    {
      "epoch": 447.4848484848485,
      "grad_norm": 0.28964757919311523,
      "learning_rate": 7.8e-07,
      "loss": 0.3499,
      "step": 4922
    },
    {
      "epoch": 447.58181818181816,
      "grad_norm": 0.27365830540657043,
      "learning_rate": 7.7e-07,
      "loss": 0.3366,
      "step": 4923
    },
    {
      "epoch": 447.6787878787879,
      "grad_norm": 0.25957080721855164,
      "learning_rate": 7.6e-07,
      "loss": 0.3317,
      "step": 4924
    },
    {
      "epoch": 447.77575757575755,
      "grad_norm": 0.26374414563179016,
      "learning_rate": 7.5e-07,
      "loss": 0.3006,
      "step": 4925
    },
    {
      "epoch": 447.8727272727273,
      "grad_norm": 0.31097298860549927,
      "learning_rate": 7.400000000000001e-07,
      "loss": 0.3119,
      "step": 4926
    },
    {
      "epoch": 447.969696969697,
      "grad_norm": 0.2407606989145279,
      "learning_rate": 7.3e-07,
      "loss": 0.3228,
      "step": 4927
    },
    {
      "epoch": 448.0,
      "grad_norm": 0.6657671332359314,
      "learning_rate": 7.2e-07,
      "loss": 0.3023,
      "step": 4928
    },
    {
      "epoch": 448.0969696969697,
      "grad_norm": 0.3061230778694153,
      "learning_rate": 7.100000000000001e-07,
      "loss": 0.3364,
      "step": 4929
    },
    {
      "epoch": 448.1939393939394,
      "grad_norm": 0.31951260566711426,
      "learning_rate": 7.000000000000001e-07,
      "loss": 0.3437,
      "step": 4930
    },
    {
      "epoch": 448.1939393939394,
      "eval_loss": 0.4234934151172638,
      "eval_runtime": 2.1175,
      "eval_samples_per_second": 25.975,
      "eval_steps_per_second": 3.306,
      "step": 4930
    },
    {
      "epoch": 448.2909090909091,
      "grad_norm": 0.26566335558891296,
      "learning_rate": 6.900000000000001e-07,
      "loss": 0.3125,
      "step": 4931
    },
    {
      "epoch": 448.3878787878788,
      "grad_norm": 0.26490557193756104,
      "learning_rate": 6.8e-07,
      "loss": 0.3289,
      "step": 4932
    },
    {
      "epoch": 448.4848484848485,
      "grad_norm": 0.3251090943813324,
      "learning_rate": 6.7e-07,
      "loss": 0.3069,
      "step": 4933
    },
    {
      "epoch": 448.58181818181816,
      "grad_norm": 0.28914380073547363,
      "learning_rate": 6.6e-07,
      "loss": 0.3205,
      "step": 4934
    },
    {
      "epoch": 448.6787878787879,
      "grad_norm": 0.30359625816345215,
      "learning_rate": 6.5e-07,
      "loss": 0.3374,
      "step": 4935
    },
    {
      "epoch": 448.77575757575755,
      "grad_norm": 0.2604947090148926,
      "learning_rate": 6.4e-07,
      "loss": 0.3215,
      "step": 4936
    },
    {
      "epoch": 448.8727272727273,
      "grad_norm": 0.2874707281589508,
      "learning_rate": 6.3e-07,
      "loss": 0.3247,
      "step": 4937
    },
    {
      "epoch": 448.969696969697,
      "grad_norm": 0.32386428117752075,
      "learning_rate": 6.2e-07,
      "loss": 0.3289,
      "step": 4938
    },
    {
      "epoch": 449.0,
      "grad_norm": 0.5075844526290894,
      "learning_rate": 6.100000000000001e-07,
      "loss": 0.3356,
      "step": 4939
    },
    {
      "epoch": 449.0969696969697,
      "grad_norm": 0.30247360467910767,
      "learning_rate": 6.000000000000001e-07,
      "loss": 0.3472,
      "step": 4940
    },
    {
      "epoch": 449.0969696969697,
      "eval_loss": 0.42349129915237427,
      "eval_runtime": 2.0915,
      "eval_samples_per_second": 26.297,
      "eval_steps_per_second": 3.347,
      "step": 4940
    },
    {
      "epoch": 449.1939393939394,
      "grad_norm": 0.2983882427215576,
      "learning_rate": 5.9e-07,
      "loss": 0.3281,
      "step": 4941
    },
    {
      "epoch": 449.2909090909091,
      "grad_norm": 0.2906897962093353,
      "learning_rate": 5.8e-07,
      "loss": 0.3541,
      "step": 4942
    },
    {
      "epoch": 449.3878787878788,
      "grad_norm": 0.2708105146884918,
      "learning_rate": 5.7e-07,
      "loss": 0.3005,
      "step": 4943
    },
    {
      "epoch": 449.4848484848485,
      "grad_norm": 0.2866322696208954,
      "learning_rate": 5.6e-07,
      "loss": 0.3287,
      "step": 4944
    },
    {
      "epoch": 449.58181818181816,
      "grad_norm": 0.2833450734615326,
      "learning_rate": 5.5e-07,
      "loss": 0.328,
      "step": 4945
    },
    {
      "epoch": 449.6787878787879,
      "grad_norm": 0.2915904223918915,
      "learning_rate": 5.4e-07,
      "loss": 0.3245,
      "step": 4946
    },
    {
      "epoch": 449.77575757575755,
      "grad_norm": 0.24670271575450897,
      "learning_rate": 5.3e-07,
      "loss": 0.33,
      "step": 4947
    },
    {
      "epoch": 449.8727272727273,
      "grad_norm": 0.31395289301872253,
      "learning_rate": 5.2e-07,
      "loss": 0.3198,
      "step": 4948
    },
    {
      "epoch": 449.969696969697,
      "grad_norm": 0.27524620294570923,
      "learning_rate": 5.100000000000001e-07,
      "loss": 0.3062,
      "step": 4949
    },
    {
      "epoch": 450.0,
      "grad_norm": 0.4141598045825958,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.3101,
      "step": 4950
    },
    {
      "epoch": 450.0,
      "eval_loss": 0.4236139953136444,
      "eval_runtime": 2.1008,
      "eval_samples_per_second": 26.181,
      "eval_steps_per_second": 3.332,
      "step": 4950
    },
    {
      "epoch": 450.0969696969697,
      "grad_norm": 0.26350322365760803,
      "learning_rate": 4.9e-07,
      "loss": 0.3295,
      "step": 4951
    },
    {
      "epoch": 450.1939393939394,
      "grad_norm": 0.243812695145607,
      "learning_rate": 4.8e-07,
      "loss": 0.2898,
      "step": 4952
    },
    {
      "epoch": 450.2909090909091,
      "grad_norm": 0.2724280059337616,
      "learning_rate": 4.7000000000000005e-07,
      "loss": 0.3327,
      "step": 4953
    },
    {
      "epoch": 450.3878787878788,
      "grad_norm": 0.27506184577941895,
      "learning_rate": 4.6e-07,
      "loss": 0.3174,
      "step": 4954
    },
    {
      "epoch": 450.4848484848485,
      "grad_norm": 0.2802964448928833,
      "learning_rate": 4.5e-07,
      "loss": 0.3159,
      "step": 4955
    },
    {
      "epoch": 450.58181818181816,
      "grad_norm": 0.29718518257141113,
      "learning_rate": 4.4e-07,
      "loss": 0.329,
      "step": 4956
    },
    {
      "epoch": 450.6787878787879,
      "grad_norm": 0.28943824768066406,
      "learning_rate": 4.3e-07,
      "loss": 0.3397,
      "step": 4957
    },
    {
      "epoch": 450.77575757575755,
      "grad_norm": 0.2653546631336212,
      "learning_rate": 4.2e-07,
      "loss": 0.3435,
      "step": 4958
    },
    {
      "epoch": 450.8727272727273,
      "grad_norm": 0.25897830724716187,
      "learning_rate": 4.1000000000000004e-07,
      "loss": 0.3195,
      "step": 4959
    },
    {
      "epoch": 450.969696969697,
      "grad_norm": 0.3242824077606201,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.3304,
      "step": 4960
    },
    {
      "epoch": 450.969696969697,
      "eval_loss": 0.42363691329956055,
      "eval_runtime": 2.1169,
      "eval_samples_per_second": 25.981,
      "eval_steps_per_second": 3.307,
      "step": 4960
    },
    {
      "epoch": 451.0,
      "grad_norm": 0.5916183590888977,
      "learning_rate": 3.9e-07,
      "loss": 0.3718,
      "step": 4961
    },
    {
      "epoch": 451.0969696969697,
      "grad_norm": 0.3053007423877716,
      "learning_rate": 3.8e-07,
      "loss": 0.311,
      "step": 4962
    },
    {
      "epoch": 451.1939393939394,
      "grad_norm": 0.26286765933036804,
      "learning_rate": 3.7000000000000006e-07,
      "loss": 0.3738,
      "step": 4963
    },
    {
      "epoch": 451.2909090909091,
      "grad_norm": 0.29149967432022095,
      "learning_rate": 3.6e-07,
      "loss": 0.3599,
      "step": 4964
    },
    {
      "epoch": 451.3878787878788,
      "grad_norm": 0.32165083289146423,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 0.334,
      "step": 4965
    },
    {
      "epoch": 451.4848484848485,
      "grad_norm": 0.26468729972839355,
      "learning_rate": 3.4e-07,
      "loss": 0.3138,
      "step": 4966
    },
    {
      "epoch": 451.58181818181816,
      "grad_norm": 0.2730655372142792,
      "learning_rate": 3.3e-07,
      "loss": 0.2947,
      "step": 4967
    },
    {
      "epoch": 451.6787878787879,
      "grad_norm": 0.286457896232605,
      "learning_rate": 3.2e-07,
      "loss": 0.2962,
      "step": 4968
    },
    {
      "epoch": 451.77575757575755,
      "grad_norm": 0.28551334142684937,
      "learning_rate": 3.1e-07,
      "loss": 0.3417,
      "step": 4969
    },
    {
      "epoch": 451.8727272727273,
      "grad_norm": 0.29090040922164917,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.3109,
      "step": 4970
    },
    {
      "epoch": 451.8727272727273,
      "eval_loss": 0.4235963225364685,
      "eval_runtime": 2.1061,
      "eval_samples_per_second": 26.115,
      "eval_steps_per_second": 3.324,
      "step": 4970
    },
    {
      "epoch": 451.969696969697,
      "grad_norm": 0.3137418031692505,
      "learning_rate": 2.9e-07,
      "loss": 0.3263,
      "step": 4971
    },
    {
      "epoch": 452.0,
      "grad_norm": 0.5446314811706543,
      "learning_rate": 2.8e-07,
      "loss": 0.3331,
      "step": 4972
    },
    {
      "epoch": 452.0969696969697,
      "grad_norm": 0.2853860855102539,
      "learning_rate": 2.7e-07,
      "loss": 0.3425,
      "step": 4973
    },
    {
      "epoch": 452.1939393939394,
      "grad_norm": 0.24901294708251953,
      "learning_rate": 2.6e-07,
      "loss": 0.33,
      "step": 4974
    },
    {
      "epoch": 452.2909090909091,
      "grad_norm": 0.2792257070541382,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.3255,
      "step": 4975
    },
    {
      "epoch": 452.3878787878788,
      "grad_norm": 0.2828456461429596,
      "learning_rate": 2.4e-07,
      "loss": 0.2862,
      "step": 4976
    },
    {
      "epoch": 452.4848484848485,
      "grad_norm": 0.24984979629516602,
      "learning_rate": 2.3e-07,
      "loss": 0.2976,
      "step": 4977
    },
    {
      "epoch": 452.58181818181816,
      "grad_norm": 0.31003209948539734,
      "learning_rate": 2.2e-07,
      "loss": 0.3427,
      "step": 4978
    },
    {
      "epoch": 452.6787878787879,
      "grad_norm": 0.316670298576355,
      "learning_rate": 2.1e-07,
      "loss": 0.3252,
      "step": 4979
    },
    {
      "epoch": 452.77575757575755,
      "grad_norm": 0.3688763976097107,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 0.3448,
      "step": 4980
    },
    {
      "epoch": 452.77575757575755,
      "eval_loss": 0.4235762655735016,
      "eval_runtime": 2.1016,
      "eval_samples_per_second": 26.17,
      "eval_steps_per_second": 3.331,
      "step": 4980
    },
    {
      "epoch": 452.8727272727273,
      "grad_norm": 0.29659217596054077,
      "learning_rate": 1.9e-07,
      "loss": 0.3238,
      "step": 4981
    },
    {
      "epoch": 452.969696969697,
      "grad_norm": 0.2912966310977936,
      "learning_rate": 1.8e-07,
      "loss": 0.3385,
      "step": 4982
    },
    {
      "epoch": 453.0,
      "grad_norm": 0.4333655536174774,
      "learning_rate": 1.7e-07,
      "loss": 0.3468,
      "step": 4983
    },
    {
      "epoch": 453.0969696969697,
      "grad_norm": 0.25823283195495605,
      "learning_rate": 1.6e-07,
      "loss": 0.2911,
      "step": 4984
    },
    {
      "epoch": 453.1939393939394,
      "grad_norm": 0.24722981452941895,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 0.3286,
      "step": 4985
    },
    {
      "epoch": 453.2909090909091,
      "grad_norm": 0.2904161214828491,
      "learning_rate": 1.4e-07,
      "loss": 0.3107,
      "step": 4986
    },
    {
      "epoch": 453.3878787878788,
      "grad_norm": 0.2927333116531372,
      "learning_rate": 1.3e-07,
      "loss": 0.3286,
      "step": 4987
    },
    {
      "epoch": 453.4848484848485,
      "grad_norm": 0.26876407861709595,
      "learning_rate": 1.2e-07,
      "loss": 0.3369,
      "step": 4988
    },
    {
      "epoch": 453.58181818181816,
      "grad_norm": 0.30206063389778137,
      "learning_rate": 1.1e-07,
      "loss": 0.313,
      "step": 4989
    },
    {
      "epoch": 453.6787878787879,
      "grad_norm": 0.2813703119754791,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 0.3211,
      "step": 4990
    },
    {
      "epoch": 453.6787878787879,
      "eval_loss": 0.42354801297187805,
      "eval_runtime": 2.0963,
      "eval_samples_per_second": 26.236,
      "eval_steps_per_second": 3.339,
      "step": 4990
    },
    {
      "epoch": 453.77575757575755,
      "grad_norm": 0.2632094621658325,
      "learning_rate": 9e-08,
      "loss": 0.3469,
      "step": 4991
    },
    {
      "epoch": 453.8727272727273,
      "grad_norm": 0.29159316420555115,
      "learning_rate": 8e-08,
      "loss": 0.3544,
      "step": 4992
    },
    {
      "epoch": 453.969696969697,
      "grad_norm": 0.2889498174190521,
      "learning_rate": 7e-08,
      "loss": 0.3087,
      "step": 4993
    },
    {
      "epoch": 454.0,
      "grad_norm": 0.5088276863098145,
      "learning_rate": 6e-08,
      "loss": 0.3929,
      "step": 4994
    },
    {
      "epoch": 454.0969696969697,
      "grad_norm": 0.29596179723739624,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 0.3085,
      "step": 4995
    },
    {
      "epoch": 454.1939393939394,
      "grad_norm": 0.2667468190193176,
      "learning_rate": 4e-08,
      "loss": 0.3522,
      "step": 4996
    },
    {
      "epoch": 454.2909090909091,
      "grad_norm": 0.25791922211647034,
      "learning_rate": 3e-08,
      "loss": 0.3197,
      "step": 4997
    },
    {
      "epoch": 454.3878787878788,
      "grad_norm": 0.26287326216697693,
      "learning_rate": 2e-08,
      "loss": 0.3218,
      "step": 4998
    },
    {
      "epoch": 454.4848484848485,
      "grad_norm": 0.26132941246032715,
      "learning_rate": 1e-08,
      "loss": 0.2982,
      "step": 4999
    },
    {
      "epoch": 454.58181818181816,
      "grad_norm": 0.2780923545360565,
      "learning_rate": 0.0,
      "loss": 0.3526,
      "step": 5000
    },
    {
      "epoch": 454.58181818181816,
      "eval_loss": 0.4235467314720154,
      "eval_runtime": 2.0906,
      "eval_samples_per_second": 26.308,
      "eval_steps_per_second": 3.348,
      "step": 5000
    }
  ],
  "logging_steps": 1,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 500,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.180595246465024e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}