{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 2772,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.021645021645021644,
      "grad_norm": 21.6130829466809,
      "learning_rate": 3.237410071942446e-07,
      "loss": 1.1818,
      "step": 10
    },
    {
      "epoch": 0.04329004329004329,
      "grad_norm": 12.462538803947291,
      "learning_rate": 6.83453237410072e-07,
      "loss": 1.0504,
      "step": 20
    },
    {
      "epoch": 0.06493506493506493,
      "grad_norm": 6.295933298100434,
      "learning_rate": 1.0431654676258993e-06,
      "loss": 0.7889,
      "step": 30
    },
    {
      "epoch": 0.08658008658008658,
      "grad_norm": 1.3494409752238592,
      "learning_rate": 1.4028776978417265e-06,
      "loss": 0.5112,
      "step": 40
    },
    {
      "epoch": 0.10822510822510822,
      "grad_norm": 1.765837671402791,
      "learning_rate": 1.762589928057554e-06,
      "loss": 0.4519,
      "step": 50
    },
    {
      "epoch": 0.12987012987012986,
      "grad_norm": 1.0536899557865906,
      "learning_rate": 2.1223021582733816e-06,
      "loss": 0.388,
      "step": 60
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 0.7021481288321632,
      "learning_rate": 2.4820143884892088e-06,
      "loss": 0.3343,
      "step": 70
    },
    {
      "epoch": 0.17316017316017315,
      "grad_norm": 0.9558076935775874,
      "learning_rate": 2.8417266187050364e-06,
      "loss": 0.3002,
      "step": 80
    },
    {
      "epoch": 0.19480519480519481,
      "grad_norm": 0.835577453148521,
      "learning_rate": 3.2014388489208636e-06,
      "loss": 0.307,
      "step": 90
    },
    {
      "epoch": 0.21645021645021645,
      "grad_norm": 1.3188742148893606,
      "learning_rate": 3.561151079136691e-06,
      "loss": 0.2805,
      "step": 100
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 1.0977505332246897,
      "learning_rate": 3.920863309352518e-06,
      "loss": 0.2826,
      "step": 110
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 0.7000307895570216,
      "learning_rate": 4.280575539568346e-06,
      "loss": 0.2763,
      "step": 120
    },
    {
      "epoch": 0.2813852813852814,
      "grad_norm": 1.0414405497390593,
      "learning_rate": 4.640287769784173e-06,
      "loss": 0.2505,
      "step": 130
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 0.7915502943989008,
      "learning_rate": 5e-06,
      "loss": 0.2761,
      "step": 140
    },
    {
      "epoch": 0.3246753246753247,
      "grad_norm": 0.9431870433125923,
      "learning_rate": 5.359712230215828e-06,
      "loss": 0.2536,
      "step": 150
    },
    {
      "epoch": 0.3463203463203463,
      "grad_norm": 0.7481791833089895,
      "learning_rate": 5.719424460431655e-06,
      "loss": 0.2432,
      "step": 160
    },
    {
      "epoch": 0.36796536796536794,
      "grad_norm": 0.7511586521137743,
      "learning_rate": 6.079136690647483e-06,
      "loss": 0.2536,
      "step": 170
    },
    {
      "epoch": 0.38961038961038963,
      "grad_norm": 1.0239907790583915,
      "learning_rate": 6.43884892086331e-06,
      "loss": 0.2422,
      "step": 180
    },
    {
      "epoch": 0.41125541125541126,
      "grad_norm": 0.8566874220063369,
      "learning_rate": 6.798561151079137e-06,
      "loss": 0.2405,
      "step": 190
    },
    {
      "epoch": 0.4329004329004329,
      "grad_norm": 1.0792660019423714,
      "learning_rate": 7.1582733812949644e-06,
      "loss": 0.2332,
      "step": 200
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.7353477255015513,
      "learning_rate": 7.517985611510792e-06,
      "loss": 0.2405,
      "step": 210
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 0.6940022504660209,
      "learning_rate": 7.877697841726619e-06,
      "loss": 0.2226,
      "step": 220
    },
    {
      "epoch": 0.49783549783549785,
      "grad_norm": 0.8431689629414103,
      "learning_rate": 8.237410071942446e-06,
      "loss": 0.2371,
      "step": 230
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 0.7166499608676279,
      "learning_rate": 8.597122302158273e-06,
      "loss": 0.2354,
      "step": 240
    },
    {
      "epoch": 0.5411255411255411,
      "grad_norm": 1.0659232363389686,
      "learning_rate": 8.956834532374102e-06,
      "loss": 0.2357,
      "step": 250
    },
    {
      "epoch": 0.5627705627705628,
      "grad_norm": 0.7857337520941757,
      "learning_rate": 9.31654676258993e-06,
      "loss": 0.2292,
      "step": 260
    },
    {
      "epoch": 0.5844155844155844,
      "grad_norm": 0.6081658029198361,
      "learning_rate": 9.676258992805757e-06,
      "loss": 0.2187,
      "step": 270
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.8845808525217291,
      "learning_rate": 9.999996033140686e-06,
      "loss": 0.2422,
      "step": 280
    },
    {
      "epoch": 0.6277056277056277,
      "grad_norm": 1.4927043302351601,
      "learning_rate": 9.999520017639167e-06,
      "loss": 0.2354,
      "step": 290
    },
    {
      "epoch": 0.6493506493506493,
      "grad_norm": 1.034918808249545,
      "learning_rate": 9.99825071682044e-06,
      "loss": 0.2403,
      "step": 300
    },
    {
      "epoch": 0.670995670995671,
      "grad_norm": 0.635476520424799,
      "learning_rate": 9.996188332087375e-06,
      "loss": 0.2269,
      "step": 310
    },
    {
      "epoch": 0.6926406926406926,
      "grad_norm": 0.5788245381561646,
      "learning_rate": 9.993333190683291e-06,
      "loss": 0.2202,
      "step": 320
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.8510563302651948,
      "learning_rate": 9.989685745640035e-06,
      "loss": 0.2434,
      "step": 330
    },
    {
      "epoch": 0.7359307359307359,
      "grad_norm": 0.8155683277044621,
      "learning_rate": 9.985246575706076e-06,
      "loss": 0.233,
      "step": 340
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.9250498569711657,
      "learning_rate": 9.980016385254703e-06,
      "loss": 0.2266,
      "step": 350
    },
    {
      "epoch": 0.7792207792207793,
      "grad_norm": 0.7328479876744679,
      "learning_rate": 9.973996004172239e-06,
      "loss": 0.2269,
      "step": 360
    },
    {
      "epoch": 0.8008658008658008,
      "grad_norm": 0.5442130516646619,
      "learning_rate": 9.967186387726369e-06,
      "loss": 0.2113,
      "step": 370
    },
    {
      "epoch": 0.8225108225108225,
      "grad_norm": 0.7359546086968223,
      "learning_rate": 9.959588616414566e-06,
      "loss": 0.2269,
      "step": 380
    },
    {
      "epoch": 0.8441558441558441,
      "grad_norm": 0.6608458698340268,
      "learning_rate": 9.951203895792643e-06,
      "loss": 0.2225,
      "step": 390
    },
    {
      "epoch": 0.8658008658008658,
      "grad_norm": 0.9714074804725528,
      "learning_rate": 9.94203355628347e-06,
      "loss": 0.2405,
      "step": 400
    },
    {
      "epoch": 0.8874458874458875,
      "grad_norm": 0.6677259786185178,
      "learning_rate": 9.932079052965863e-06,
      "loss": 0.2175,
      "step": 410
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.5841154893855143,
      "learning_rate": 9.921341965343717e-06,
      "loss": 0.2193,
      "step": 420
    },
    {
      "epoch": 0.9307359307359307,
      "grad_norm": 0.7396266989205127,
      "learning_rate": 9.909823997095372e-06,
      "loss": 0.2087,
      "step": 430
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 0.7490459569561692,
      "learning_rate": 9.897526975803283e-06,
      "loss": 0.2203,
      "step": 440
    },
    {
      "epoch": 0.974025974025974,
      "grad_norm": 0.7615162606056478,
      "learning_rate": 9.884452852664048e-06,
      "loss": 0.2141,
      "step": 450
    },
    {
      "epoch": 0.9956709956709957,
      "grad_norm": 0.8128792517019994,
      "learning_rate": 9.870603702178799e-06,
      "loss": 0.2236,
      "step": 460
    },
    {
      "epoch": 1.0173160173160174,
      "grad_norm": 0.6469548170151056,
      "learning_rate": 9.855981721824029e-06,
      "loss": 0.1826,
      "step": 470
    },
    {
      "epoch": 1.0389610389610389,
      "grad_norm": 0.9656679492103972,
      "learning_rate": 9.840589231702925e-06,
      "loss": 0.1566,
      "step": 480
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.7712490729184256,
      "learning_rate": 9.82442867417723e-06,
      "loss": 0.1754,
      "step": 490
    },
    {
      "epoch": 1.0822510822510822,
      "grad_norm": 0.570334629971979,
      "learning_rate": 9.8075026134797e-06,
      "loss": 0.1659,
      "step": 500
    },
    {
      "epoch": 1.103896103896104,
      "grad_norm": 0.7802114570523205,
      "learning_rate": 9.789813735307236e-06,
      "loss": 0.1674,
      "step": 510
    },
    {
      "epoch": 1.1255411255411256,
      "grad_norm": 0.7645514501672953,
      "learning_rate": 9.771364846394746e-06,
      "loss": 0.1667,
      "step": 520
    },
    {
      "epoch": 1.1471861471861473,
      "grad_norm": 0.7071258289607386,
      "learning_rate": 9.752158874069778e-06,
      "loss": 0.1688,
      "step": 530
    },
    {
      "epoch": 1.1688311688311688,
      "grad_norm": 0.6106661557735052,
      "learning_rate": 9.732198865788047e-06,
      "loss": 0.1734,
      "step": 540
    },
    {
      "epoch": 1.1904761904761905,
      "grad_norm": 0.6854627830117066,
      "learning_rate": 9.711487988649886e-06,
      "loss": 0.173,
      "step": 550
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.7138191196723566,
      "learning_rate": 9.69002952889771e-06,
      "loss": 0.185,
      "step": 560
    },
    {
      "epoch": 1.2337662337662338,
      "grad_norm": 0.6738929556977785,
      "learning_rate": 9.66782689139458e-06,
      "loss": 0.171,
      "step": 570
    },
    {
      "epoch": 1.2554112554112553,
      "grad_norm": 0.8476956697456646,
      "learning_rate": 9.644883599083959e-06,
      "loss": 0.1679,
      "step": 580
    },
    {
      "epoch": 1.277056277056277,
      "grad_norm": 0.7260647180052996,
      "learning_rate": 9.621203292430698e-06,
      "loss": 0.1601,
      "step": 590
    },
    {
      "epoch": 1.2987012987012987,
      "grad_norm": 0.8582998389685774,
      "learning_rate": 9.596789728843412e-06,
      "loss": 0.1619,
      "step": 600
    },
    {
      "epoch": 1.3203463203463204,
      "grad_norm": 0.7547970534486574,
      "learning_rate": 9.571646782078277e-06,
      "loss": 0.1568,
      "step": 610
    },
    {
      "epoch": 1.341991341991342,
      "grad_norm": 0.7216308401113449,
      "learning_rate": 9.545778441624365e-06,
      "loss": 0.1766,
      "step": 620
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.6934281952083672,
      "learning_rate": 9.519188812070638e-06,
      "loss": 0.1835,
      "step": 630
    },
    {
      "epoch": 1.3852813852813852,
      "grad_norm": 0.6199405211237684,
      "learning_rate": 9.491882112454651e-06,
      "loss": 0.1897,
      "step": 640
    },
    {
      "epoch": 1.406926406926407,
      "grad_norm": 0.6726407355147255,
      "learning_rate": 9.463862675593111e-06,
      "loss": 0.1736,
      "step": 650
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.6078220589609316,
      "learning_rate": 9.435134947394385e-06,
      "loss": 0.1698,
      "step": 660
    },
    {
      "epoch": 1.4502164502164503,
      "grad_norm": 0.6433686232441512,
      "learning_rate": 9.405703486153046e-06,
      "loss": 0.1772,
      "step": 670
    },
    {
      "epoch": 1.4718614718614718,
      "grad_norm": 0.7692727564366963,
      "learning_rate": 9.375572961826607e-06,
      "loss": 0.1666,
      "step": 680
    },
    {
      "epoch": 1.4935064935064934,
      "grad_norm": 0.6755185900891589,
      "learning_rate": 9.344748155294524e-06,
      "loss": 0.1687,
      "step": 690
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.6374029645418328,
      "learning_rate": 9.313233957599608e-06,
      "loss": 0.1634,
      "step": 700
    },
    {
      "epoch": 1.5367965367965368,
      "grad_norm": 0.6886054373713666,
      "learning_rate": 9.281035369171943e-06,
      "loss": 0.1799,
      "step": 710
    },
    {
      "epoch": 1.5584415584415585,
      "grad_norm": 0.5888358024006713,
      "learning_rate": 9.248157499035466e-06,
      "loss": 0.1873,
      "step": 720
    },
    {
      "epoch": 1.5800865800865802,
      "grad_norm": 0.9779311458316267,
      "learning_rate": 9.214605563997298e-06,
      "loss": 0.1691,
      "step": 730
    },
    {
      "epoch": 1.601731601731602,
      "grad_norm": 0.6663591886189619,
      "learning_rate": 9.180384887819992e-06,
      "loss": 0.157,
      "step": 740
    },
    {
      "epoch": 1.6233766233766234,
      "grad_norm": 0.5844708530261586,
      "learning_rate": 9.145500900376783e-06,
      "loss": 0.1699,
      "step": 750
    },
    {
      "epoch": 1.645021645021645,
      "grad_norm": 0.6642647622742827,
      "learning_rate": 9.109959136790036e-06,
      "loss": 0.1718,
      "step": 760
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.5769482517355071,
      "learning_rate": 9.073765236552958e-06,
      "loss": 0.166,
      "step": 770
    },
    {
      "epoch": 1.6883116883116882,
      "grad_norm": 0.6982869352657255,
      "learning_rate": 9.036924942634785e-06,
      "loss": 0.1572,
      "step": 780
    },
    {
      "epoch": 1.70995670995671,
      "grad_norm": 0.6773676553730691,
      "learning_rate": 8.999444100569514e-06,
      "loss": 0.1797,
      "step": 790
    },
    {
      "epoch": 1.7316017316017316,
      "grad_norm": 0.7017565176525642,
      "learning_rate": 8.96132865752839e-06,
      "loss": 0.1773,
      "step": 800
    },
    {
      "epoch": 1.7532467532467533,
      "grad_norm": 0.6485222375150361,
      "learning_rate": 8.922584661376252e-06,
      "loss": 0.181,
      "step": 810
    },
    {
      "epoch": 1.774891774891775,
      "grad_norm": 0.630541034397996,
      "learning_rate": 8.883218259711905e-06,
      "loss": 0.1821,
      "step": 820
    },
    {
      "epoch": 1.7965367965367967,
      "grad_norm": 0.7323687940857571,
      "learning_rate": 8.843235698892661e-06,
      "loss": 0.1577,
      "step": 830
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.6828712375429067,
      "learning_rate": 8.802643323043223e-06,
      "loss": 0.1708,
      "step": 840
    },
    {
      "epoch": 1.8398268398268398,
      "grad_norm": 0.7804989692606259,
      "learning_rate": 8.761447573049045e-06,
      "loss": 0.1534,
      "step": 850
    },
    {
      "epoch": 1.8614718614718615,
      "grad_norm": 0.6932069441953983,
      "learning_rate": 8.719654985534337e-06,
      "loss": 0.1593,
      "step": 860
    },
    {
      "epoch": 1.883116883116883,
      "grad_norm": 0.8135263260805993,
      "learning_rate": 8.677272191824894e-06,
      "loss": 0.1631,
      "step": 870
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 0.7550079894916347,
      "learning_rate": 8.634305916895879e-06,
      "loss": 0.1545,
      "step": 880
    },
    {
      "epoch": 1.9264069264069263,
      "grad_norm": 0.6267041974455337,
      "learning_rate": 8.590762978304763e-06,
      "loss": 0.1765,
      "step": 890
    },
    {
      "epoch": 1.948051948051948,
      "grad_norm": 0.7318774760471034,
      "learning_rate": 8.54665028510956e-06,
      "loss": 0.1874,
      "step": 900
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.7737851944600341,
      "learning_rate": 8.501974836772557e-06,
      "loss": 0.1629,
      "step": 910
    },
    {
      "epoch": 1.9913419913419914,
      "grad_norm": 0.6490553623838053,
      "learning_rate": 8.456743722049691e-06,
      "loss": 0.1645,
      "step": 920
    },
    {
      "epoch": 2.012987012987013,
      "grad_norm": 0.7803848982141369,
      "learning_rate": 8.410964117865759e-06,
      "loss": 0.1189,
      "step": 930
    },
    {
      "epoch": 2.034632034632035,
      "grad_norm": 0.7264830805438321,
      "learning_rate": 8.364643288175639e-06,
      "loss": 0.1154,
      "step": 940
    },
    {
      "epoch": 2.0562770562770565,
      "grad_norm": 0.7799453665545895,
      "learning_rate": 8.317788582811706e-06,
      "loss": 0.0991,
      "step": 950
    },
    {
      "epoch": 2.0779220779220777,
      "grad_norm": 0.8442301797019349,
      "learning_rate": 8.270407436317612e-06,
      "loss": 0.0973,
      "step": 960
    },
    {
      "epoch": 2.0995670995670994,
      "grad_norm": 0.637078023754458,
      "learning_rate": 8.22250736676863e-06,
      "loss": 0.1039,
      "step": 970
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.8229285449511264,
      "learning_rate": 8.174095974578745e-06,
      "loss": 0.0963,
      "step": 980
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.6349130429375007,
      "learning_rate": 8.125180941294682e-06,
      "loss": 0.1023,
      "step": 990
    },
    {
      "epoch": 2.1645021645021645,
      "grad_norm": 0.5544498156462541,
      "learning_rate": 8.075770028377052e-06,
      "loss": 0.102,
      "step": 1000
    },
    {
      "epoch": 2.186147186147186,
      "grad_norm": 0.7034272146785523,
      "learning_rate": 8.025871075968828e-06,
      "loss": 0.1011,
      "step": 1010
    },
    {
      "epoch": 2.207792207792208,
      "grad_norm": 0.6846201764673089,
      "learning_rate": 7.975492001651325e-06,
      "loss": 0.099,
      "step": 1020
    },
    {
      "epoch": 2.2294372294372296,
      "grad_norm": 0.9379364522959263,
      "learning_rate": 7.924640799187908e-06,
      "loss": 0.0935,
      "step": 1030
    },
    {
      "epoch": 2.2510822510822512,
      "grad_norm": 0.6905029028460162,
      "learning_rate": 7.873325537255598e-06,
      "loss": 0.1173,
      "step": 1040
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.9718225863996551,
      "learning_rate": 7.821554358164794e-06,
      "loss": 0.1028,
      "step": 1050
    },
    {
      "epoch": 2.2943722943722946,
      "grad_norm": 0.665120172279867,
      "learning_rate": 7.769335476567319e-06,
      "loss": 0.1072,
      "step": 1060
    },
    {
      "epoch": 2.316017316017316,
      "grad_norm": 1.502009444702413,
      "learning_rate": 7.716677178152983e-06,
      "loss": 0.101,
      "step": 1070
    },
    {
      "epoch": 2.3376623376623376,
      "grad_norm": 0.6759614647266734,
      "learning_rate": 7.663587818334866e-06,
      "loss": 0.0953,
      "step": 1080
    },
    {
      "epoch": 2.3593073593073592,
      "grad_norm": 0.7558040128418608,
      "learning_rate": 7.610075820923555e-06,
      "loss": 0.1001,
      "step": 1090
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 0.5152733963165321,
      "learning_rate": 7.556149676790509e-06,
      "loss": 0.0888,
      "step": 1100
    },
    {
      "epoch": 2.4025974025974026,
      "grad_norm": 0.6716489766953482,
      "learning_rate": 7.501817942520804e-06,
      "loss": 0.1054,
      "step": 1110
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.7494097887159353,
      "learning_rate": 7.447089239055428e-06,
      "loss": 0.1028,
      "step": 1120
    },
    {
      "epoch": 2.445887445887446,
      "grad_norm": 0.8148892781002129,
      "learning_rate": 7.391972250323388e-06,
      "loss": 0.0981,
      "step": 1130
    },
    {
      "epoch": 2.4675324675324677,
      "grad_norm": 0.6730754521280989,
      "learning_rate": 7.3364757218638e-06,
      "loss": 0.1084,
      "step": 1140
    },
    {
      "epoch": 2.4891774891774894,
      "grad_norm": 0.5689712970878772,
      "learning_rate": 7.2806084594382266e-06,
      "loss": 0.1053,
      "step": 1150
    },
    {
      "epoch": 2.5108225108225106,
      "grad_norm": 0.7800300504629694,
      "learning_rate": 7.224379327633433e-06,
      "loss": 0.0956,
      "step": 1160
    },
    {
      "epoch": 2.5324675324675323,
      "grad_norm": 0.7265602117822915,
      "learning_rate": 7.167797248454832e-06,
      "loss": 0.1024,
      "step": 1170
    },
    {
      "epoch": 2.554112554112554,
      "grad_norm": 0.7247049116468843,
      "learning_rate": 7.110871199910807e-06,
      "loss": 0.0882,
      "step": 1180
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.5733710433141679,
      "learning_rate": 7.0536102145881506e-06,
      "loss": 0.098,
      "step": 1190
    },
    {
      "epoch": 2.5974025974025974,
      "grad_norm": 0.5060187668661191,
      "learning_rate": 6.996023378218842e-06,
      "loss": 0.0912,
      "step": 1200
    },
    {
      "epoch": 2.619047619047619,
      "grad_norm": 0.7825691518054259,
      "learning_rate": 6.938119828238394e-06,
      "loss": 0.116,
      "step": 1210
    },
    {
      "epoch": 2.6406926406926408,
      "grad_norm": 0.6123837959808874,
      "learning_rate": 6.8799087523359955e-06,
      "loss": 0.1062,
      "step": 1220
    },
    {
      "epoch": 2.6623376623376624,
      "grad_norm": 0.7307482540727773,
      "learning_rate": 6.821399386996679e-06,
      "loss": 0.1031,
      "step": 1230
    },
    {
      "epoch": 2.683982683982684,
      "grad_norm": 0.7264578828619572,
      "learning_rate": 6.762601016035746e-06,
      "loss": 0.1038,
      "step": 1240
    },
    {
      "epoch": 2.7056277056277054,
      "grad_norm": 0.5418388645800419,
      "learning_rate": 6.7035229691256845e-06,
      "loss": 0.1027,
      "step": 1250
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.7882770142623279,
      "learning_rate": 6.644174620315804e-06,
      "loss": 0.098,
      "step": 1260
    },
    {
      "epoch": 2.7489177489177488,
      "grad_norm": 0.9287906476335954,
      "learning_rate": 6.584565386544845e-06,
      "loss": 0.1061,
      "step": 1270
    },
    {
      "epoch": 2.7705627705627704,
      "grad_norm": 0.8348936347989188,
      "learning_rate": 6.5247047261467535e-06,
      "loss": 0.0982,
      "step": 1280
    },
    {
      "epoch": 2.792207792207792,
      "grad_norm": 0.660624145726457,
      "learning_rate": 6.464602137349922e-06,
      "loss": 0.1116,
      "step": 1290
    },
    {
      "epoch": 2.813852813852814,
      "grad_norm": 0.6530027214588501,
      "learning_rate": 6.404267156770076e-06,
      "loss": 0.1019,
      "step": 1300
    },
    {
      "epoch": 2.8354978354978355,
      "grad_norm": 0.6994109716069091,
      "learning_rate": 6.343709357897081e-06,
      "loss": 0.1074,
      "step": 1310
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.6896266109853648,
      "learning_rate": 6.282938349575886e-06,
      "loss": 0.094,
      "step": 1320
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.8638777272684736,
      "learning_rate": 6.221963774481879e-06,
      "loss": 0.0948,
      "step": 1330
    },
    {
      "epoch": 2.9004329004329006,
      "grad_norm": 0.5707988613176481,
      "learning_rate": 6.16079530759085e-06,
      "loss": 0.103,
      "step": 1340
    },
    {
      "epoch": 2.9220779220779223,
      "grad_norm": 0.6209377352786277,
      "learning_rate": 6.0994426546438504e-06,
      "loss": 0.0965,
      "step": 1350
    },
    {
      "epoch": 2.9437229437229435,
      "grad_norm": 0.6718983119243808,
      "learning_rate": 6.037915550607153e-06,
      "loss": 0.1195,
      "step": 1360
    },
    {
      "epoch": 2.965367965367965,
      "grad_norm": 0.822673753539792,
      "learning_rate": 5.976223758127585e-06,
      "loss": 0.1164,
      "step": 1370
    },
    {
      "epoch": 2.987012987012987,
      "grad_norm": 0.8766056966999247,
      "learning_rate": 5.9143770659834686e-06,
      "loss": 0.0964,
      "step": 1380
    },
    {
      "epoch": 3.0086580086580086,
      "grad_norm": 0.5477610793298274,
      "learning_rate": 5.852385287531411e-06,
      "loss": 0.083,
      "step": 1390
    },
    {
      "epoch": 3.0303030303030303,
      "grad_norm": 0.8309714696169959,
      "learning_rate": 5.790258259149195e-06,
      "loss": 0.0537,
      "step": 1400
    },
    {
      "epoch": 3.051948051948052,
      "grad_norm": 0.5824827947812137,
      "learning_rate": 5.728005838675026e-06,
      "loss": 0.0529,
      "step": 1410
    },
    {
      "epoch": 3.0735930735930737,
      "grad_norm": 0.7977639386423778,
      "learning_rate": 5.6656379038433504e-06,
      "loss": 0.0572,
      "step": 1420
    },
    {
      "epoch": 3.0952380952380953,
      "grad_norm": 0.741197338844836,
      "learning_rate": 5.603164350717553e-06,
      "loss": 0.0526,
      "step": 1430
    },
    {
      "epoch": 3.116883116883117,
      "grad_norm": 0.7070850303349854,
      "learning_rate": 5.540595092119709e-06,
      "loss": 0.0592,
      "step": 1440
    },
    {
      "epoch": 3.1385281385281387,
      "grad_norm": 0.686978086711389,
      "learning_rate": 5.477940056057706e-06,
      "loss": 0.0565,
      "step": 1450
    },
    {
      "epoch": 3.16017316017316,
      "grad_norm": 0.6175802132679667,
      "learning_rate": 5.415209184149932e-06,
      "loss": 0.0549,
      "step": 1460
    },
    {
      "epoch": 3.1818181818181817,
      "grad_norm": 0.7977986087387449,
      "learning_rate": 5.35241243004783e-06,
      "loss": 0.0599,
      "step": 1470
    },
    {
      "epoch": 3.2034632034632033,
      "grad_norm": 0.7125393131726135,
      "learning_rate": 5.289559757856515e-06,
      "loss": 0.0539,
      "step": 1480
    },
    {
      "epoch": 3.225108225108225,
      "grad_norm": 0.640867908203246,
      "learning_rate": 5.226661140553759e-06,
      "loss": 0.0578,
      "step": 1490
    },
    {
      "epoch": 3.2467532467532467,
      "grad_norm": 0.5353048545017236,
      "learning_rate": 5.163726558407549e-06,
      "loss": 0.0557,
      "step": 1500
    },
    {
      "epoch": 3.2683982683982684,
      "grad_norm": 0.6000038629811708,
      "learning_rate": 5.100765997392494e-06,
      "loss": 0.0551,
      "step": 1510
    },
    {
      "epoch": 3.29004329004329,
      "grad_norm": 0.652363315743491,
      "learning_rate": 5.037789447605334e-06,
      "loss": 0.0526,
      "step": 1520
    },
    {
      "epoch": 3.311688311688312,
      "grad_norm": 0.8326482069739805,
      "learning_rate": 4.974806901679778e-06,
      "loss": 0.0539,
      "step": 1530
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.6296614766757893,
      "learning_rate": 4.911828353200961e-06,
      "loss": 0.0539,
      "step": 1540
    },
    {
      "epoch": 3.354978354978355,
      "grad_norm": 0.7515024041218878,
      "learning_rate": 4.8488637951197335e-06,
      "loss": 0.0551,
      "step": 1550
    },
    {
      "epoch": 3.3766233766233764,
      "grad_norm": 0.8152999039420703,
      "learning_rate": 4.785923218167057e-06,
      "loss": 0.0793,
      "step": 1560
    },
    {
      "epoch": 3.398268398268398,
      "grad_norm": 0.7413829422122867,
      "learning_rate": 4.723016609268749e-06,
      "loss": 0.0621,
      "step": 1570
    },
    {
      "epoch": 3.41991341991342,
      "grad_norm": 0.8299685702069385,
      "learning_rate": 4.660153949960844e-06,
      "loss": 0.0518,
      "step": 1580
    },
    {
      "epoch": 3.4415584415584415,
      "grad_norm": 0.8282394179919524,
      "learning_rate": 4.597345214805788e-06,
      "loss": 0.0529,
      "step": 1590
    },
    {
      "epoch": 3.463203463203463,
      "grad_norm": 0.9955615715020707,
      "learning_rate": 4.534600369809761e-06,
      "loss": 0.067,
      "step": 1600
    },
    {
      "epoch": 3.484848484848485,
      "grad_norm": 0.5485339860266549,
      "learning_rate": 4.471929370841346e-06,
      "loss": 0.0583,
      "step": 1610
    },
    {
      "epoch": 3.5064935064935066,
      "grad_norm": 0.5809551200356223,
      "learning_rate": 4.409342162051804e-06,
      "loss": 0.0545,
      "step": 1620
    },
    {
      "epoch": 3.5281385281385282,
      "grad_norm": 1.0592816793394937,
      "learning_rate": 4.3468486742972186e-06,
      "loss": 0.0617,
      "step": 1630
    },
    {
      "epoch": 3.54978354978355,
      "grad_norm": 0.6501265272385851,
      "learning_rate": 4.284458823562744e-06,
      "loss": 0.0568,
      "step": 1640
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 0.7223069822816002,
      "learning_rate": 4.222182509389201e-06,
      "loss": 0.0493,
      "step": 1650
    },
    {
      "epoch": 3.5930735930735933,
      "grad_norm": 0.8530295472389439,
      "learning_rate": 4.160029613302315e-06,
      "loss": 0.0568,
      "step": 1660
    },
    {
      "epoch": 3.6147186147186146,
      "grad_norm": 0.6456059934125514,
      "learning_rate": 4.098009997244772e-06,
      "loss": 0.0572,
      "step": 1670
    },
    {
      "epoch": 3.6363636363636362,
      "grad_norm": 0.7751150443747835,
      "learning_rate": 4.036133502011415e-06,
      "loss": 0.052,
      "step": 1680
    },
    {
      "epoch": 3.658008658008658,
      "grad_norm": 0.5965938690746012,
      "learning_rate": 3.974409945687775e-06,
      "loss": 0.0616,
      "step": 1690
    },
    {
      "epoch": 3.6796536796536796,
      "grad_norm": 0.6908278998742275,
      "learning_rate": 3.912849122092216e-06,
      "loss": 0.0546,
      "step": 1700
    },
    {
      "epoch": 3.7012987012987013,
      "grad_norm": 0.6397138152963864,
      "learning_rate": 3.851460799221924e-06,
      "loss": 0.0512,
      "step": 1710
    },
    {
      "epoch": 3.722943722943723,
      "grad_norm": 0.6414741252074939,
      "learning_rate": 3.7902547177030046e-06,
      "loss": 0.0515,
      "step": 1720
    },
    {
      "epoch": 3.7445887445887447,
      "grad_norm": 1.1161499636662198,
      "learning_rate": 3.7292405892449026e-06,
      "loss": 0.0556,
      "step": 1730
    },
    {
      "epoch": 3.7662337662337664,
      "grad_norm": 0.6413652260876448,
      "learning_rate": 3.668428095099437e-06,
      "loss": 0.0614,
      "step": 1740
    },
    {
      "epoch": 3.787878787878788,
      "grad_norm": 0.564086939258874,
      "learning_rate": 3.60782688452464e-06,
      "loss": 0.0538,
      "step": 1750
    },
    {
      "epoch": 3.8095238095238093,
      "grad_norm": 0.7864095464740084,
      "learning_rate": 3.5474465732537016e-06,
      "loss": 0.0546,
      "step": 1760
    },
    {
      "epoch": 3.8311688311688314,
      "grad_norm": 0.6780702802746946,
      "learning_rate": 3.4872967419692115e-06,
      "loss": 0.0627,
      "step": 1770
    },
    {
      "epoch": 3.8528138528138527,
      "grad_norm": 1.013816343667498,
      "learning_rate": 3.427386934782965e-06,
      "loss": 0.0515,
      "step": 1780
    },
    {
      "epoch": 3.8744588744588744,
      "grad_norm": 0.6457581674761906,
      "learning_rate": 3.3677266577215874e-06,
      "loss": 0.0608,
      "step": 1790
    },
    {
      "epoch": 3.896103896103896,
      "grad_norm": 0.664257894449446,
      "learning_rate": 3.308325377218188e-06,
      "loss": 0.058,
      "step": 1800
    },
    {
      "epoch": 3.9177489177489178,
      "grad_norm": 0.7878599732314846,
      "learning_rate": 3.2491925186102935e-06,
      "loss": 0.0539,
      "step": 1810
    },
    {
      "epoch": 3.9393939393939394,
      "grad_norm": 0.6855292465105594,
      "learning_rate": 3.190337464644315e-06,
      "loss": 0.0578,
      "step": 1820
    },
    {
      "epoch": 3.961038961038961,
      "grad_norm": 0.6580843656540549,
      "learning_rate": 3.1317695539867606e-06,
      "loss": 0.0548,
      "step": 1830
    },
    {
      "epoch": 3.982683982683983,
      "grad_norm": 0.6284573974473094,
      "learning_rate": 3.0734980797424534e-06,
      "loss": 0.0518,
      "step": 1840
    },
    {
      "epoch": 4.004329004329004,
      "grad_norm": 0.45758562409415,
      "learning_rate": 3.0155322879799727e-06,
      "loss": 0.0486,
      "step": 1850
    },
    {
      "epoch": 4.025974025974026,
      "grad_norm": 0.5354514289102192,
      "learning_rate": 2.9578813762645473e-06,
      "loss": 0.0278,
      "step": 1860
    },
    {
      "epoch": 4.0476190476190474,
      "grad_norm": 0.7126835055872589,
      "learning_rate": 2.9005544921986774e-06,
      "loss": 0.0307,
      "step": 1870
    },
    {
      "epoch": 4.06926406926407,
      "grad_norm": 1.0262145261283995,
      "learning_rate": 2.843560731970645e-06,
      "loss": 0.0298,
      "step": 1880
    },
    {
      "epoch": 4.090909090909091,
      "grad_norm": 0.49319696078847103,
      "learning_rate": 2.786909138911211e-06,
      "loss": 0.0258,
      "step": 1890
    },
    {
      "epoch": 4.112554112554113,
      "grad_norm": 0.5956134322124683,
      "learning_rate": 2.7306087020586847e-06,
      "loss": 0.0318,
      "step": 1900
    },
    {
      "epoch": 4.134199134199134,
      "grad_norm": 0.7168459202076095,
      "learning_rate": 2.674668354732616e-06,
      "loss": 0.0236,
      "step": 1910
    },
    {
      "epoch": 4.1558441558441555,
      "grad_norm": 0.4814478877486977,
      "learning_rate": 2.619096973116318e-06,
      "loss": 0.0313,
      "step": 1920
    },
    {
      "epoch": 4.177489177489178,
      "grad_norm": 0.6072849390729453,
      "learning_rate": 2.5639033748484787e-06,
      "loss": 0.0258,
      "step": 1930
    },
    {
      "epoch": 4.199134199134199,
      "grad_norm": 0.7094493306757976,
      "learning_rate": 2.5090963176240205e-06,
      "loss": 0.0294,
      "step": 1940
    },
    {
      "epoch": 4.220779220779221,
      "grad_norm": 0.6286730295957046,
      "learning_rate": 2.4546844978045237e-06,
      "loss": 0.0326,
      "step": 1950
    },
    {
      "epoch": 4.242424242424242,
      "grad_norm": 0.631819735791692,
      "learning_rate": 2.4006765490383355e-06,
      "loss": 0.0279,
      "step": 1960
    },
    {
      "epoch": 4.264069264069264,
      "grad_norm": 0.6059937950462203,
      "learning_rate": 2.3470810408906553e-06,
      "loss": 0.0362,
      "step": 1970
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.45307102037361907,
      "learning_rate": 2.2939064774837916e-06,
      "loss": 0.0233,
      "step": 1980
    },
    {
      "epoch": 4.307359307359308,
      "grad_norm": 0.6181158133076236,
      "learning_rate": 2.2411612961477704e-06,
      "loss": 0.0256,
      "step": 1990
    },
    {
      "epoch": 4.329004329004329,
      "grad_norm": 0.7086393893216901,
      "learning_rate": 2.188853866081592e-06,
      "loss": 0.0284,
      "step": 2000
    },
    {
      "epoch": 4.35064935064935,
      "grad_norm": 0.5562746518641271,
      "learning_rate": 2.136992487025252e-06,
      "loss": 0.0274,
      "step": 2010
    },
    {
      "epoch": 4.372294372294372,
      "grad_norm": 0.6128690650004339,
      "learning_rate": 2.085585387942808e-06,
      "loss": 0.0304,
      "step": 2020
    },
    {
      "epoch": 4.393939393939394,
      "grad_norm": 0.6275091600798839,
      "learning_rate": 2.034640725716671e-06,
      "loss": 0.0249,
      "step": 2030
    },
    {
      "epoch": 4.415584415584416,
      "grad_norm": 0.4906068889023691,
      "learning_rate": 1.9841665838533412e-06,
      "loss": 0.0307,
      "step": 2040
    },
    {
      "epoch": 4.437229437229437,
      "grad_norm": 0.5899586093556849,
      "learning_rate": 1.9341709712007607e-06,
      "loss": 0.0296,
      "step": 2050
    },
    {
      "epoch": 4.458874458874459,
      "grad_norm": 0.5036318191324448,
      "learning_rate": 1.8846618206775547e-06,
      "loss": 0.0235,
      "step": 2060
    },
    {
      "epoch": 4.48051948051948,
      "grad_norm": 0.9455976004118218,
      "learning_rate": 1.8356469880142802e-06,
      "loss": 0.0312,
      "step": 2070
    },
    {
      "epoch": 4.5021645021645025,
      "grad_norm": 0.6295727625857045,
      "learning_rate": 1.7871342505069439e-06,
      "loss": 0.0274,
      "step": 2080
    },
    {
      "epoch": 4.523809523809524,
      "grad_norm": 0.5853942942786021,
      "learning_rate": 1.7391313057829678e-06,
      "loss": 0.0267,
      "step": 2090
    },
    {
      "epoch": 4.545454545454545,
      "grad_norm": 0.45075157504721464,
      "learning_rate": 1.6916457705797723e-06,
      "loss": 0.0257,
      "step": 2100
    },
    {
      "epoch": 4.567099567099567,
      "grad_norm": 0.6326635145123081,
      "learning_rate": 1.6446851795362322e-06,
      "loss": 0.0234,
      "step": 2110
    },
    {
      "epoch": 4.588744588744589,
      "grad_norm": 0.5001726765299896,
      "learning_rate": 1.5982569839971212e-06,
      "loss": 0.0328,
      "step": 2120
    },
    {
      "epoch": 4.6103896103896105,
      "grad_norm": 0.5951949515519321,
      "learning_rate": 1.5523685508307972e-06,
      "loss": 0.0246,
      "step": 2130
    },
    {
      "epoch": 4.632034632034632,
      "grad_norm": 0.6069999960014675,
      "learning_rate": 1.5070271612602848e-06,
      "loss": 0.0262,
      "step": 2140
    },
    {
      "epoch": 4.653679653679654,
      "grad_norm": 0.5064167687288715,
      "learning_rate": 1.4622400097079448e-06,
      "loss": 0.0271,
      "step": 2150
    },
    {
      "epoch": 4.675324675324675,
      "grad_norm": 0.5554020565918719,
      "learning_rate": 1.4180142026539173e-06,
      "loss": 0.0244,
      "step": 2160
    },
    {
      "epoch": 4.696969696969697,
      "grad_norm": 0.6009149042653651,
      "learning_rate": 1.3743567575085282e-06,
      "loss": 0.0262,
      "step": 2170
    },
    {
      "epoch": 4.7186147186147185,
      "grad_norm": 0.6482435687628327,
      "learning_rate": 1.3312746014988092e-06,
      "loss": 0.0275,
      "step": 2180
    },
    {
      "epoch": 4.740259740259741,
      "grad_norm": 0.9926857132905802,
      "learning_rate": 1.288774570569345e-06,
      "loss": 0.028,
      "step": 2190
    },
    {
      "epoch": 4.761904761904762,
      "grad_norm": 0.6026293047805794,
      "learning_rate": 1.2468634082976022e-06,
      "loss": 0.0316,
      "step": 2200
    },
    {
      "epoch": 4.783549783549784,
      "grad_norm": 0.6177570798867973,
      "learning_rate": 1.2055477648238977e-06,
      "loss": 0.0256,
      "step": 2210
    },
    {
      "epoch": 4.805194805194805,
      "grad_norm": 0.5222307371108272,
      "learning_rate": 1.1648341957962227e-06,
      "loss": 0.0437,
      "step": 2220
    },
    {
      "epoch": 4.8268398268398265,
      "grad_norm": 0.6513459538355414,
      "learning_rate": 1.1247291613300254e-06,
      "loss": 0.027,
      "step": 2230
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.5203541849723394,
      "learning_rate": 1.0852390249831858e-06,
      "loss": 0.0327,
      "step": 2240
    },
    {
      "epoch": 4.87012987012987,
      "grad_norm": 0.6689998908887559,
      "learning_rate": 1.046370052746284e-06,
      "loss": 0.0263,
      "step": 2250
    },
    {
      "epoch": 4.891774891774892,
      "grad_norm": 0.8022386692319375,
      "learning_rate": 1.0081284120483642e-06,
      "loss": 0.0235,
      "step": 2260
    },
    {
      "epoch": 4.913419913419913,
      "grad_norm": 0.4831778612797812,
      "learning_rate": 9.705201707783368e-07,
      "loss": 0.0302,
      "step": 2270
    },
    {
      "epoch": 4.935064935064935,
      "grad_norm": 0.5477549170800028,
      "learning_rate": 9.335512963221732e-07,
      "loss": 0.0243,
      "step": 2280
    },
    {
      "epoch": 4.956709956709957,
      "grad_norm": 0.697634375938823,
      "learning_rate": 8.972276546160408e-07,
      "loss": 0.0278,
      "step": 2290
    },
    {
      "epoch": 4.978354978354979,
      "grad_norm": 0.5458987126265021,
      "learning_rate": 8.615550092155478e-07,
      "loss": 0.0275,
      "step": 2300
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5442210565078133,
      "learning_rate": 8.265390203812229e-07,
      "loss": 0.0283,
      "step": 2310
    },
    {
      "epoch": 5.021645021645021,
      "grad_norm": 0.6988479841532632,
      "learning_rate": 7.921852441803907e-07,
      "loss": 0.0169,
      "step": 2320
    },
    {
      "epoch": 5.043290043290043,
      "grad_norm": 0.3740945023049904,
      "learning_rate": 7.584991316055801e-07,
      "loss": 0.0154,
      "step": 2330
    },
    {
      "epoch": 5.064935064935065,
      "grad_norm": 0.5738683386264282,
      "learning_rate": 7.254860277095932e-07,
      "loss": 0.0158,
      "step": 2340
    },
    {
      "epoch": 5.086580086580087,
      "grad_norm": 0.5683588054694332,
      "learning_rate": 6.931511707574073e-07,
      "loss": 0.0148,
      "step": 2350
    },
    {
      "epoch": 5.108225108225108,
      "grad_norm": 0.4551682863092917,
      "learning_rate": 6.614996913949961e-07,
      "loss": 0.0128,
      "step": 2360
    },
    {
      "epoch": 5.12987012987013,
      "grad_norm": 0.4227712388354827,
      "learning_rate": 6.305366118352435e-07,
      "loss": 0.0167,
      "step": 2370
    },
    {
      "epoch": 5.151515151515151,
      "grad_norm": 0.5897280075104119,
      "learning_rate": 6.002668450610566e-07,
      "loss": 0.0147,
      "step": 2380
    },
    {
      "epoch": 5.1731601731601735,
      "grad_norm": 0.2302221690058311,
      "learning_rate": 5.706951940458111e-07,
      "loss": 0.0132,
      "step": 2390
    },
    {
      "epoch": 5.194805194805195,
      "grad_norm": 0.5086444609250476,
      "learning_rate": 5.418263509912536e-07,
      "loss": 0.0191,
      "step": 2400
    },
    {
      "epoch": 5.216450216450216,
      "grad_norm": 0.4784802998644711,
      "learning_rate": 5.136648965829782e-07,
      "loss": 0.0125,
      "step": 2410
    },
    {
      "epoch": 5.238095238095238,
      "grad_norm": 0.35595572655337715,
      "learning_rate": 4.862152992635976e-07,
      "loss": 0.015,
      "step": 2420
    },
    {
      "epoch": 5.259740259740259,
      "grad_norm": 0.4297636612354201,
      "learning_rate": 4.594819145237256e-07,
      "loss": 0.0161,
      "step": 2430
    },
    {
      "epoch": 5.2813852813852815,
      "grad_norm": 0.2751220931796682,
      "learning_rate": 4.334689842108869e-07,
      "loss": 0.0133,
      "step": 2440
    },
    {
      "epoch": 5.303030303030303,
      "grad_norm": 0.4197137279285618,
      "learning_rate": 4.0818063585644054e-07,
      "loss": 0.014,
      "step": 2450
    },
    {
      "epoch": 5.324675324675325,
      "grad_norm": 0.44050347181839344,
      "learning_rate": 3.836208820206705e-07,
      "loss": 0.0123,
      "step": 2460
    },
    {
      "epoch": 5.346320346320346,
      "grad_norm": 0.3065455679871206,
      "learning_rate": 3.5979361965609247e-07,
      "loss": 0.0154,
      "step": 2470
    },
    {
      "epoch": 5.367965367965368,
      "grad_norm": 0.5550653575279008,
      "learning_rate": 3.367026294891185e-07,
      "loss": 0.0142,
      "step": 2480
    },
    {
      "epoch": 5.3896103896103895,
      "grad_norm": 0.41580544263505936,
      "learning_rate": 3.14351575420162e-07,
      "loss": 0.0135,
      "step": 2490
    },
    {
      "epoch": 5.411255411255412,
      "grad_norm": 0.5248393848680003,
      "learning_rate": 2.9274400394227744e-07,
      "loss": 0.015,
      "step": 2500
    },
    {
      "epoch": 5.432900432900433,
      "grad_norm": 0.5986448734424771,
      "learning_rate": 2.7188334357843014e-07,
      "loss": 0.0134,
      "step": 2510
    },
    {
      "epoch": 5.454545454545454,
      "grad_norm": 0.5309670182797535,
      "learning_rate": 2.517729043374845e-07,
      "loss": 0.0136,
      "step": 2520
    },
    {
      "epoch": 5.476190476190476,
      "grad_norm": 0.35573191068127813,
      "learning_rate": 2.3241587718899561e-07,
      "loss": 0.0154,
      "step": 2530
    },
    {
      "epoch": 5.4978354978354975,
      "grad_norm": 0.38132026018602805,
      "learning_rate": 2.1381533355689433e-07,
      "loss": 0.0192,
      "step": 2540
    },
    {
      "epoch": 5.51948051948052,
      "grad_norm": 0.64254201796522,
      "learning_rate": 1.95974224832135e-07,
      "loss": 0.0153,
      "step": 2550
    },
    {
      "epoch": 5.541125541125541,
      "grad_norm": 0.6213757843049649,
      "learning_rate": 1.7889538190439094e-07,
      "loss": 0.0138,
      "step": 2560
    },
    {
      "epoch": 5.562770562770563,
      "grad_norm": 0.2987300993361835,
      "learning_rate": 1.6258151471287397e-07,
      "loss": 0.0155,
      "step": 2570
    },
    {
      "epoch": 5.584415584415584,
      "grad_norm": 0.6582102856789737,
      "learning_rate": 1.4703521181633763e-07,
      "loss": 0.0129,
      "step": 2580
    },
    {
      "epoch": 5.606060606060606,
      "grad_norm": 0.6500246549678568,
      "learning_rate": 1.3225893998234872e-07,
      "loss": 0.0134,
      "step": 2590
    },
    {
      "epoch": 5.627705627705628,
      "grad_norm": 0.6168417770544529,
      "learning_rate": 1.1825504379587627e-07,
      "loss": 0.0142,
      "step": 2600
    },
    {
      "epoch": 5.64935064935065,
      "grad_norm": 0.6079522411622211,
      "learning_rate": 1.0502574528727217e-07,
      "loss": 0.012,
      "step": 2610
    },
    {
      "epoch": 5.670995670995671,
      "grad_norm": 0.4169031844641975,
      "learning_rate": 9.257314357970004e-08,
      "loss": 0.0185,
      "step": 2620
    },
    {
      "epoch": 5.692640692640692,
      "grad_norm": 0.45113712338804446,
      "learning_rate": 8.089921455605499e-08,
      "loss": 0.0123,
      "step": 2630
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 0.38621085058053184,
      "learning_rate": 7.000581054545208e-08,
      "loss": 0.0145,
      "step": 2640
    },
    {
      "epoch": 5.735930735930736,
      "grad_norm": 0.5628992022349025,
      "learning_rate": 5.989466002931044e-08,
      "loss": 0.0127,
      "step": 2650
    },
    {
      "epoch": 5.757575757575758,
      "grad_norm": 0.5189409732855611,
      "learning_rate": 5.056736736709256e-08,
      "loss": 0.0119,
      "step": 2660
    },
    {
      "epoch": 5.779220779220779,
      "grad_norm": 0.4525527972738947,
      "learning_rate": 4.2025412541734645e-08,
      "loss": 0.0148,
      "step": 2670
    },
    {
      "epoch": 5.800865800865801,
      "grad_norm": 0.353943217698887,
      "learning_rate": 3.427015092481667e-08,
      "loss": 0.0141,
      "step": 2680
    },
    {
      "epoch": 5.822510822510822,
      "grad_norm": 0.4205930002985954,
      "learning_rate": 2.7302813061501067e-08,
      "loss": 0.013,
      "step": 2690
    },
    {
      "epoch": 5.8441558441558445,
      "grad_norm": 0.4448324879851497,
      "learning_rate": 2.1124504475278918e-08,
      "loss": 0.0161,
      "step": 2700
    },
    {
      "epoch": 5.865800865800866,
      "grad_norm": 0.5602095769779442,
      "learning_rate": 1.5736205492555835e-08,
      "loss": 0.0123,
      "step": 2710
    },
    {
      "epoch": 5.887445887445887,
      "grad_norm": 0.44832482766932447,
      "learning_rate": 1.113877108709971e-08,
      "loss": 0.0162,
      "step": 2720
    },
    {
      "epoch": 5.909090909090909,
      "grad_norm": 0.4347478134914692,
      "learning_rate": 7.332930744380906e-09,
      "loss": 0.0116,
      "step": 2730
    },
    {
      "epoch": 5.93073593073593,
      "grad_norm": 0.3159206854507724,
      "learning_rate": 4.31928834582207e-09,
      "loss": 0.012,
      "step": 2740
    },
    {
      "epoch": 5.9523809523809526,
      "grad_norm": 0.44786630476707673,
      "learning_rate": 2.098322072980885e-09,
      "loss": 0.0317,
      "step": 2750
    },
    {
      "epoch": 5.974025974025974,
      "grad_norm": 0.4014542976497695,
      "learning_rate": 6.70384331674101e-10,
      "loss": 0.0128,
      "step": 2760
    },
    {
      "epoch": 5.995670995670996,
      "grad_norm": 0.42608574655523024,
      "learning_rate": 3.5701696060042036e-11,
      "loss": 0.0153,
      "step": 2770
    },
    {
      "epoch": 6.0,
      "step": 2772,
      "total_flos": 665846364340224.0,
      "train_loss": 0.11275661619404306,
      "train_runtime": 28497.6157,
      "train_samples_per_second": 0.777,
      "train_steps_per_second": 0.097
    }
  ],
  "logging_steps": 10,
  "max_steps": 2772,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 50000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 665846364340224.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}