{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9868043602983362,
  "eval_steps": 500,
  "global_step": 4300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00022948938611589215,
      "grad_norm": 0.2832958400249481,
      "learning_rate": 4.587155963302753e-07,
      "loss": 1.1374,
      "step": 1
    },
    {
      "epoch": 0.0011474469305794606,
      "grad_norm": 0.3270454704761505,
      "learning_rate": 2.2935779816513764e-06,
      "loss": 1.1705,
      "step": 5
    },
    {
      "epoch": 0.002294893861158921,
      "grad_norm": 0.27594515681266785,
      "learning_rate": 4.587155963302753e-06,
      "loss": 1.1193,
      "step": 10
    },
    {
      "epoch": 0.0034423407917383822,
      "grad_norm": 0.2871370315551758,
      "learning_rate": 6.880733944954129e-06,
      "loss": 1.1392,
      "step": 15
    },
    {
      "epoch": 0.004589787722317842,
      "grad_norm": 0.2703236937522888,
      "learning_rate": 9.174311926605506e-06,
      "loss": 1.1746,
      "step": 20
    },
    {
      "epoch": 0.005737234652897304,
      "grad_norm": 0.2527748644351959,
      "learning_rate": 1.1467889908256882e-05,
      "loss": 1.129,
      "step": 25
    },
    {
      "epoch": 0.0068846815834767644,
      "grad_norm": 0.1984025090932846,
      "learning_rate": 1.3761467889908258e-05,
      "loss": 1.0819,
      "step": 30
    },
    {
      "epoch": 0.008032128514056224,
      "grad_norm": 0.21797247231006622,
      "learning_rate": 1.6055045871559634e-05,
      "loss": 1.0604,
      "step": 35
    },
    {
      "epoch": 0.009179575444635685,
      "grad_norm": 0.1990118771791458,
      "learning_rate": 1.834862385321101e-05,
      "loss": 1.0285,
      "step": 40
    },
    {
      "epoch": 0.010327022375215147,
      "grad_norm": 0.19321753084659576,
      "learning_rate": 2.0642201834862388e-05,
      "loss": 1.0389,
      "step": 45
    },
    {
      "epoch": 0.011474469305794608,
      "grad_norm": 0.18540368974208832,
      "learning_rate": 2.2935779816513765e-05,
      "loss": 1.0454,
      "step": 50
    },
    {
      "epoch": 0.012621916236374068,
      "grad_norm": 0.18497976660728455,
      "learning_rate": 2.5229357798165138e-05,
      "loss": 1.0578,
      "step": 55
    },
    {
      "epoch": 0.013769363166953529,
      "grad_norm": 0.14606232941150665,
      "learning_rate": 2.7522935779816515e-05,
      "loss": 1.064,
      "step": 60
    },
    {
      "epoch": 0.01491681009753299,
      "grad_norm": 0.1588371992111206,
      "learning_rate": 2.9816513761467892e-05,
      "loss": 1.0231,
      "step": 65
    },
    {
      "epoch": 0.01606425702811245,
      "grad_norm": 0.1489713490009308,
      "learning_rate": 3.211009174311927e-05,
      "loss": 0.9939,
      "step": 70
    },
    {
      "epoch": 0.01721170395869191,
      "grad_norm": 0.1601668745279312,
      "learning_rate": 3.4403669724770645e-05,
      "loss": 1.04,
      "step": 75
    },
    {
      "epoch": 0.01835915088927137,
      "grad_norm": 0.17821019887924194,
      "learning_rate": 3.669724770642202e-05,
      "loss": 1.0147,
      "step": 80
    },
    {
      "epoch": 0.019506597819850834,
      "grad_norm": 0.15738487243652344,
      "learning_rate": 3.89908256880734e-05,
      "loss": 1.0404,
      "step": 85
    },
    {
      "epoch": 0.020654044750430294,
      "grad_norm": 0.16155536472797394,
      "learning_rate": 4.1284403669724776e-05,
      "loss": 1.0388,
      "step": 90
    },
    {
      "epoch": 0.021801491681009755,
      "grad_norm": 0.1788385957479477,
      "learning_rate": 4.3577981651376146e-05,
      "loss": 0.9844,
      "step": 95
    },
    {
      "epoch": 0.022948938611589215,
      "grad_norm": 0.2024046927690506,
      "learning_rate": 4.587155963302753e-05,
      "loss": 0.9918,
      "step": 100
    },
    {
      "epoch": 0.024096385542168676,
      "grad_norm": 0.16764964163303375,
      "learning_rate": 4.81651376146789e-05,
      "loss": 0.9971,
      "step": 105
    },
    {
      "epoch": 0.025243832472748137,
      "grad_norm": 0.17267030477523804,
      "learning_rate": 5.0458715596330276e-05,
      "loss": 1.0152,
      "step": 110
    },
    {
      "epoch": 0.026391279403327597,
      "grad_norm": 0.1706738919019699,
      "learning_rate": 5.2752293577981646e-05,
      "loss": 1.0135,
      "step": 115
    },
    {
      "epoch": 0.027538726333907058,
      "grad_norm": 0.1741032898426056,
      "learning_rate": 5.504587155963303e-05,
      "loss": 0.9995,
      "step": 120
    },
    {
      "epoch": 0.02868617326448652,
      "grad_norm": 0.17674313485622406,
      "learning_rate": 5.733944954128441e-05,
      "loss": 0.9853,
      "step": 125
    },
    {
      "epoch": 0.02983362019506598,
      "grad_norm": 0.16848498582839966,
      "learning_rate": 5.9633027522935784e-05,
      "loss": 1.0349,
      "step": 130
    },
    {
      "epoch": 0.03098106712564544,
      "grad_norm": 0.17360493540763855,
      "learning_rate": 6.192660550458716e-05,
      "loss": 0.9893,
      "step": 135
    },
    {
      "epoch": 0.0321285140562249,
      "grad_norm": 0.191062331199646,
      "learning_rate": 6.422018348623854e-05,
      "loss": 0.9922,
      "step": 140
    },
    {
      "epoch": 0.03327596098680436,
      "grad_norm": 0.1799137443304062,
      "learning_rate": 6.651376146788991e-05,
      "loss": 0.9998,
      "step": 145
    },
    {
      "epoch": 0.03442340791738382,
      "grad_norm": 0.19225718080997467,
      "learning_rate": 6.880733944954129e-05,
      "loss": 1.0426,
      "step": 150
    },
    {
      "epoch": 0.03557085484796328,
      "grad_norm": 0.20365940034389496,
      "learning_rate": 7.110091743119265e-05,
      "loss": 0.9869,
      "step": 155
    },
    {
      "epoch": 0.03671830177854274,
      "grad_norm": 0.18115347623825073,
      "learning_rate": 7.339449541284404e-05,
      "loss": 0.9939,
      "step": 160
    },
    {
      "epoch": 0.0378657487091222,
      "grad_norm": 0.16979213058948517,
      "learning_rate": 7.568807339449542e-05,
      "loss": 0.9788,
      "step": 165
    },
    {
      "epoch": 0.03901319563970167,
      "grad_norm": 0.16331952810287476,
      "learning_rate": 7.79816513761468e-05,
      "loss": 1.0014,
      "step": 170
    },
    {
      "epoch": 0.040160642570281124,
      "grad_norm": 0.20406360924243927,
      "learning_rate": 8.027522935779816e-05,
      "loss": 0.988,
      "step": 175
    },
    {
      "epoch": 0.04130808950086059,
      "grad_norm": 0.17396636307239532,
      "learning_rate": 8.256880733944955e-05,
      "loss": 0.9756,
      "step": 180
    },
    {
      "epoch": 0.042455536431440045,
      "grad_norm": 0.19402877986431122,
      "learning_rate": 8.486238532110093e-05,
      "loss": 1.0017,
      "step": 185
    },
    {
      "epoch": 0.04360298336201951,
      "grad_norm": 0.16275173425674438,
      "learning_rate": 8.715596330275229e-05,
      "loss": 0.9881,
      "step": 190
    },
    {
      "epoch": 0.04475043029259897,
      "grad_norm": 0.16075727343559265,
      "learning_rate": 8.944954128440367e-05,
      "loss": 0.9818,
      "step": 195
    },
    {
      "epoch": 0.04589787722317843,
      "grad_norm": 0.17031803727149963,
      "learning_rate": 9.174311926605506e-05,
      "loss": 0.9813,
      "step": 200
    },
    {
      "epoch": 0.04704532415375789,
      "grad_norm": 0.16039007902145386,
      "learning_rate": 9.403669724770642e-05,
      "loss": 0.9655,
      "step": 205
    },
    {
      "epoch": 0.04819277108433735,
      "grad_norm": 0.17182576656341553,
      "learning_rate": 9.63302752293578e-05,
      "loss": 1.0034,
      "step": 210
    },
    {
      "epoch": 0.04934021801491681,
      "grad_norm": 0.16478177905082703,
      "learning_rate": 9.862385321100918e-05,
      "loss": 0.9993,
      "step": 215
    },
    {
      "epoch": 0.05048766494549627,
      "grad_norm": 0.16542178392410278,
      "learning_rate": 0.00010091743119266055,
      "loss": 1.005,
      "step": 220
    },
    {
      "epoch": 0.05163511187607573,
      "grad_norm": 0.1568165123462677,
      "learning_rate": 0.00010321100917431193,
      "loss": 1.0148,
      "step": 225
    },
    {
      "epoch": 0.052782558806655194,
      "grad_norm": 0.15992042422294617,
      "learning_rate": 0.00010550458715596329,
      "loss": 1.0084,
      "step": 230
    },
    {
      "epoch": 0.05393000573723465,
      "grad_norm": 0.16262054443359375,
      "learning_rate": 0.0001077981651376147,
      "loss": 0.9871,
      "step": 235
    },
    {
      "epoch": 0.055077452667814115,
      "grad_norm": 0.15893304347991943,
      "learning_rate": 0.00011009174311926606,
      "loss": 0.9609,
      "step": 240
    },
    {
      "epoch": 0.05622489959839357,
      "grad_norm": 0.1627625972032547,
      "learning_rate": 0.00011238532110091744,
      "loss": 0.9571,
      "step": 245
    },
    {
      "epoch": 0.05737234652897304,
      "grad_norm": 0.15538789331912994,
      "learning_rate": 0.00011467889908256881,
      "loss": 0.9632,
      "step": 250
    },
    {
      "epoch": 0.058519793459552494,
      "grad_norm": 0.14804205298423767,
      "learning_rate": 0.00011697247706422019,
      "loss": 1.0001,
      "step": 255
    },
    {
      "epoch": 0.05966724039013196,
      "grad_norm": 0.15267077088356018,
      "learning_rate": 0.00011926605504587157,
      "loss": 0.9697,
      "step": 260
    },
    {
      "epoch": 0.060814687320711415,
      "grad_norm": 0.14952810108661652,
      "learning_rate": 0.00012155963302752293,
      "loss": 1.0034,
      "step": 265
    },
    {
      "epoch": 0.06196213425129088,
      "grad_norm": 0.15380804240703583,
      "learning_rate": 0.00012385321100917432,
      "loss": 0.9614,
      "step": 270
    },
    {
      "epoch": 0.06310958118187034,
      "grad_norm": 0.144377201795578,
      "learning_rate": 0.0001261467889908257,
      "loss": 0.965,
      "step": 275
    },
    {
      "epoch": 0.0642570281124498,
      "grad_norm": 0.13880601525306702,
      "learning_rate": 0.00012844036697247707,
      "loss": 0.9586,
      "step": 280
    },
    {
      "epoch": 0.06540447504302926,
      "grad_norm": 0.1473158597946167,
      "learning_rate": 0.00013073394495412844,
      "loss": 0.994,
      "step": 285
    },
    {
      "epoch": 0.06655192197360872,
      "grad_norm": 0.14506657421588898,
      "learning_rate": 0.00013302752293577983,
      "loss": 1.0082,
      "step": 290
    },
    {
      "epoch": 0.06769936890418818,
      "grad_norm": 0.1545286476612091,
      "learning_rate": 0.0001353211009174312,
      "loss": 0.9639,
      "step": 295
    },
    {
      "epoch": 0.06884681583476764,
      "grad_norm": 0.14149315655231476,
      "learning_rate": 0.00013761467889908258,
      "loss": 0.9682,
      "step": 300
    },
    {
      "epoch": 0.0699942627653471,
      "grad_norm": 0.1382177472114563,
      "learning_rate": 0.00013990825688073395,
      "loss": 0.9534,
      "step": 305
    },
    {
      "epoch": 0.07114170969592656,
      "grad_norm": 0.1412287950515747,
      "learning_rate": 0.0001422018348623853,
      "loss": 0.9463,
      "step": 310
    },
    {
      "epoch": 0.07228915662650602,
      "grad_norm": 0.13654330372810364,
      "learning_rate": 0.00014449541284403673,
      "loss": 0.9851,
      "step": 315
    },
    {
      "epoch": 0.07343660355708548,
      "grad_norm": 0.13656587898731232,
      "learning_rate": 0.0001467889908256881,
      "loss": 0.9938,
      "step": 320
    },
    {
      "epoch": 0.07458405048766495,
      "grad_norm": 0.13453873991966248,
      "learning_rate": 0.00014908256880733945,
      "loss": 0.9847,
      "step": 325
    },
    {
      "epoch": 0.0757314974182444,
      "grad_norm": 0.13527260720729828,
      "learning_rate": 0.00015137614678899084,
      "loss": 0.9436,
      "step": 330
    },
    {
      "epoch": 0.07687894434882386,
      "grad_norm": 0.13411866128444672,
      "learning_rate": 0.0001536697247706422,
      "loss": 0.9737,
      "step": 335
    },
    {
      "epoch": 0.07802639127940333,
      "grad_norm": 0.13352230191230774,
      "learning_rate": 0.0001559633027522936,
      "loss": 0.977,
      "step": 340
    },
    {
      "epoch": 0.07917383820998279,
      "grad_norm": 0.14445261657238007,
      "learning_rate": 0.00015825688073394496,
      "loss": 0.9802,
      "step": 345
    },
    {
      "epoch": 0.08032128514056225,
      "grad_norm": 0.13070940971374512,
      "learning_rate": 0.00016055045871559632,
      "loss": 0.9509,
      "step": 350
    },
    {
      "epoch": 0.0814687320711417,
      "grad_norm": 0.13417434692382812,
      "learning_rate": 0.0001628440366972477,
      "loss": 1.0253,
      "step": 355
    },
    {
      "epoch": 0.08261617900172118,
      "grad_norm": 0.15762361884117126,
      "learning_rate": 0.0001651376146788991,
      "loss": 0.9936,
      "step": 360
    },
    {
      "epoch": 0.08376362593230063,
      "grad_norm": 0.13492801785469055,
      "learning_rate": 0.00016743119266055047,
      "loss": 0.9601,
      "step": 365
    },
    {
      "epoch": 0.08491107286288009,
      "grad_norm": 0.1378079354763031,
      "learning_rate": 0.00016972477064220186,
      "loss": 1.0292,
      "step": 370
    },
    {
      "epoch": 0.08605851979345955,
      "grad_norm": 0.1303025484085083,
      "learning_rate": 0.00017201834862385322,
      "loss": 0.9491,
      "step": 375
    },
    {
      "epoch": 0.08720596672403902,
      "grad_norm": 0.13195742666721344,
      "learning_rate": 0.00017431192660550458,
      "loss": 1.0134,
      "step": 380
    },
    {
      "epoch": 0.08835341365461848,
      "grad_norm": 0.13493070006370544,
      "learning_rate": 0.00017660550458715597,
      "loss": 0.9563,
      "step": 385
    },
    {
      "epoch": 0.08950086058519793,
      "grad_norm": 0.12674984335899353,
      "learning_rate": 0.00017889908256880734,
      "loss": 0.9772,
      "step": 390
    },
    {
      "epoch": 0.09064830751577739,
      "grad_norm": 0.12925055623054504,
      "learning_rate": 0.00018119266055045873,
      "loss": 0.9445,
      "step": 395
    },
    {
      "epoch": 0.09179575444635686,
      "grad_norm": 0.13387279212474823,
      "learning_rate": 0.00018348623853211012,
      "loss": 0.9724,
      "step": 400
    },
    {
      "epoch": 0.09294320137693632,
      "grad_norm": 0.13336487114429474,
      "learning_rate": 0.00018577981651376148,
      "loss": 0.9938,
      "step": 405
    },
    {
      "epoch": 0.09409064830751578,
      "grad_norm": 0.13736997544765472,
      "learning_rate": 0.00018807339449541284,
      "loss": 0.9719,
      "step": 410
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 0.13675136864185333,
      "learning_rate": 0.00019036697247706424,
      "loss": 0.9648,
      "step": 415
    },
    {
      "epoch": 0.0963855421686747,
      "grad_norm": 0.13273541629314423,
      "learning_rate": 0.0001926605504587156,
      "loss": 0.9717,
      "step": 420
    },
    {
      "epoch": 0.09753298909925416,
      "grad_norm": 0.13400444388389587,
      "learning_rate": 0.000194954128440367,
      "loss": 0.9819,
      "step": 425
    },
    {
      "epoch": 0.09868043602983362,
      "grad_norm": 0.1314323991537094,
      "learning_rate": 0.00019724770642201835,
      "loss": 0.984,
      "step": 430
    },
    {
      "epoch": 0.09982788296041308,
      "grad_norm": 0.13610613346099854,
      "learning_rate": 0.00019954128440366972,
      "loss": 0.9681,
      "step": 435
    },
    {
      "epoch": 0.10097532989099255,
      "grad_norm": 0.13332229852676392,
      "learning_rate": 0.00019999948643469536,
      "loss": 0.9801,
      "step": 440
    },
    {
      "epoch": 0.102122776821572,
      "grad_norm": 0.12949495017528534,
      "learning_rate": 0.00019999740008468594,
      "loss": 0.9846,
      "step": 445
    },
    {
      "epoch": 0.10327022375215146,
      "grad_norm": 0.13086913526058197,
      "learning_rate": 0.00019999370888559804,
      "loss": 0.923,
      "step": 450
    },
    {
      "epoch": 0.10441767068273092,
      "grad_norm": 0.1352614462375641,
      "learning_rate": 0.0001999884128966714,
      "loss": 0.9459,
      "step": 455
    },
    {
      "epoch": 0.10556511761331039,
      "grad_norm": 0.13356110453605652,
      "learning_rate": 0.00019998151220290082,
      "loss": 0.9859,
      "step": 460
    },
    {
      "epoch": 0.10671256454388985,
      "grad_norm": 0.13900211453437805,
      "learning_rate": 0.00019997300691503497,
      "loss": 0.9691,
      "step": 465
    },
    {
      "epoch": 0.1078600114744693,
      "grad_norm": 0.12718099355697632,
      "learning_rate": 0.0001999628971695744,
      "loss": 0.9618,
      "step": 470
    },
    {
      "epoch": 0.10900745840504876,
      "grad_norm": 0.12601250410079956,
      "learning_rate": 0.00019995118312876944,
      "loss": 0.9603,
      "step": 475
    },
    {
      "epoch": 0.11015490533562823,
      "grad_norm": 0.13364988565444946,
      "learning_rate": 0.0001999378649806177,
      "loss": 0.9366,
      "step": 480
    },
    {
      "epoch": 0.11130235226620769,
      "grad_norm": 0.13241152465343475,
      "learning_rate": 0.00019992294293886095,
      "loss": 0.9565,
      "step": 485
    },
    {
      "epoch": 0.11244979919678715,
      "grad_norm": 0.1273089498281479,
      "learning_rate": 0.00019990641724298156,
      "loss": 0.96,
      "step": 490
    },
    {
      "epoch": 0.11359724612736662,
      "grad_norm": 0.13082469999790192,
      "learning_rate": 0.000199888288158199,
      "loss": 0.9767,
      "step": 495
    },
    {
      "epoch": 0.11474469305794607,
      "grad_norm": 0.14526137709617615,
      "learning_rate": 0.00019986855597546526,
      "loss": 0.9641,
      "step": 500
    },
    {
      "epoch": 0.11589213998852553,
      "grad_norm": 0.13247860968112946,
      "learning_rate": 0.00019984722101146029,
      "loss": 0.9676,
      "step": 505
    },
    {
      "epoch": 0.11703958691910499,
      "grad_norm": 0.14177852869033813,
      "learning_rate": 0.000199824283608587,
      "loss": 0.9406,
      "step": 510
    },
    {
      "epoch": 0.11818703384968446,
      "grad_norm": 0.1286071240901947,
      "learning_rate": 0.00019979974413496566,
      "loss": 0.9689,
      "step": 515
    },
    {
      "epoch": 0.11933448078026392,
      "grad_norm": 0.1299976259469986,
      "learning_rate": 0.00019977360298442803,
      "loss": 0.9623,
      "step": 520
    },
    {
      "epoch": 0.12048192771084337,
      "grad_norm": 0.12755008041858673,
      "learning_rate": 0.00019974586057651102,
      "loss": 0.9139,
      "step": 525
    },
    {
      "epoch": 0.12162937464142283,
      "grad_norm": 0.13445819914340973,
      "learning_rate": 0.00019971651735644995,
      "loss": 0.9513,
      "step": 530
    },
    {
      "epoch": 0.1227768215720023,
      "grad_norm": 0.13108594715595245,
      "learning_rate": 0.00019968557379517152,
      "loss": 0.9492,
      "step": 535
    },
    {
      "epoch": 0.12392426850258176,
      "grad_norm": 0.124118372797966,
      "learning_rate": 0.00019965303038928608,
      "loss": 0.9411,
      "step": 540
    },
    {
      "epoch": 0.12507171543316123,
      "grad_norm": 0.12611638009548187,
      "learning_rate": 0.00019961888766107972,
      "loss": 0.976,
      "step": 545
    },
    {
      "epoch": 0.12621916236374067,
      "grad_norm": 0.13510450720787048,
      "learning_rate": 0.00019958314615850598,
      "loss": 0.9632,
      "step": 550
    },
    {
      "epoch": 0.12736660929432014,
      "grad_norm": 0.12858641147613525,
      "learning_rate": 0.00019954580645517697,
      "loss": 0.9491,
      "step": 555
    },
    {
      "epoch": 0.1285140562248996,
      "grad_norm": 0.13497839868068695,
      "learning_rate": 0.0001995068691503541,
      "loss": 0.9604,
      "step": 560
    },
    {
      "epoch": 0.12966150315547906,
      "grad_norm": 0.12632782757282257,
      "learning_rate": 0.00019946633486893865,
      "loss": 0.9363,
      "step": 565
    },
    {
      "epoch": 0.13080895008605853,
      "grad_norm": 0.12541595101356506,
      "learning_rate": 0.00019942420426146153,
      "loss": 0.9897,
      "step": 570
    },
    {
      "epoch": 0.13195639701663797,
      "grad_norm": 0.1276572197675705,
      "learning_rate": 0.00019938047800407302,
      "loss": 0.9535,
      "step": 575
    },
    {
      "epoch": 0.13310384394721744,
      "grad_norm": 0.12992243468761444,
      "learning_rate": 0.00019933515679853182,
      "loss": 0.9879,
      "step": 580
    },
    {
      "epoch": 0.1342512908777969,
      "grad_norm": 0.1361103504896164,
      "learning_rate": 0.0001992882413721937,
      "loss": 0.9684,
      "step": 585
    },
    {
      "epoch": 0.13539873780837636,
      "grad_norm": 0.13442201912403107,
      "learning_rate": 0.0001992397324780001,
      "loss": 0.9427,
      "step": 590
    },
    {
      "epoch": 0.13654618473895583,
      "grad_norm": 0.1316874623298645,
      "learning_rate": 0.00019918963089446577,
      "loss": 0.9795,
      "step": 595
    },
    {
      "epoch": 0.13769363166953527,
      "grad_norm": 0.12683913111686707,
      "learning_rate": 0.00019913793742566647,
      "loss": 0.9517,
      "step": 600
    },
    {
      "epoch": 0.13884107860011474,
      "grad_norm": 0.12805037200450897,
      "learning_rate": 0.00019908465290122585,
      "loss": 0.9728,
      "step": 605
    },
    {
      "epoch": 0.1399885255306942,
      "grad_norm": 0.13088442385196686,
      "learning_rate": 0.00019902977817630243,
      "loss": 0.9695,
      "step": 610
    },
    {
      "epoch": 0.14113597246127366,
      "grad_norm": 0.1324809193611145,
      "learning_rate": 0.00019897331413157548,
      "loss": 0.9996,
      "step": 615
    },
    {
      "epoch": 0.14228341939185313,
      "grad_norm": 0.12211659550666809,
      "learning_rate": 0.00019891526167323145,
      "loss": 0.989,
      "step": 620
    },
    {
      "epoch": 0.1434308663224326,
      "grad_norm": 0.13171370327472687,
      "learning_rate": 0.0001988556217329488,
      "loss": 0.9714,
      "step": 625
    },
    {
      "epoch": 0.14457831325301204,
      "grad_norm": 0.12398378551006317,
      "learning_rate": 0.00019879439526788341,
      "loss": 0.9717,
      "step": 630
    },
    {
      "epoch": 0.1457257601835915,
      "grad_norm": 0.12855181097984314,
      "learning_rate": 0.00019873158326065327,
      "loss": 0.9635,
      "step": 635
    },
    {
      "epoch": 0.14687320711417096,
      "grad_norm": 0.13017693161964417,
      "learning_rate": 0.00019866718671932249,
      "loss": 0.9804,
      "step": 640
    },
    {
      "epoch": 0.14802065404475043,
      "grad_norm": 0.12396004796028137,
      "learning_rate": 0.00019860120667738516,
      "loss": 0.9985,
      "step": 645
    },
    {
      "epoch": 0.1491681009753299,
      "grad_norm": 0.12556292116641998,
      "learning_rate": 0.00019853364419374902,
      "loss": 0.9498,
      "step": 650
    },
    {
      "epoch": 0.15031554790590934,
      "grad_norm": 0.12642225623130798,
      "learning_rate": 0.00019846450035271808,
      "loss": 1.0224,
      "step": 655
    },
    {
      "epoch": 0.1514629948364888,
      "grad_norm": 0.12983497977256775,
      "learning_rate": 0.00019839377626397554,
      "loss": 0.9604,
      "step": 660
    },
    {
      "epoch": 0.15261044176706828,
      "grad_norm": 0.13211478292942047,
      "learning_rate": 0.00019832147306256576,
      "loss": 0.9864,
      "step": 665
    },
    {
      "epoch": 0.15375788869764773,
      "grad_norm": 0.1218995749950409,
      "learning_rate": 0.00019824759190887622,
      "loss": 0.937,
      "step": 670
    },
    {
      "epoch": 0.1549053356282272,
      "grad_norm": 0.13377314805984497,
      "learning_rate": 0.00019817213398861866,
      "loss": 0.9704,
      "step": 675
    },
    {
      "epoch": 0.15605278255880667,
      "grad_norm": 0.12591040134429932,
      "learning_rate": 0.0001980951005128104,
      "loss": 0.9874,
      "step": 680
    },
    {
      "epoch": 0.1572002294893861,
      "grad_norm": 0.12856173515319824,
      "learning_rate": 0.00019801649271775459,
      "loss": 0.9469,
      "step": 685
    },
    {
      "epoch": 0.15834767641996558,
      "grad_norm": 0.1214781105518341,
      "learning_rate": 0.00019793631186502047,
      "loss": 0.9739,
      "step": 690
    },
    {
      "epoch": 0.15949512335054503,
      "grad_norm": 0.12588472664356232,
      "learning_rate": 0.00019785455924142318,
      "loss": 0.9629,
      "step": 695
    },
    {
      "epoch": 0.1606425702811245,
      "grad_norm": 0.12644240260124207,
      "learning_rate": 0.0001977712361590031,
      "loss": 1.0156,
      "step": 700
    },
    {
      "epoch": 0.16179001721170397,
      "grad_norm": 0.12597164511680603,
      "learning_rate": 0.00019768634395500465,
      "loss": 0.9728,
      "step": 705
    },
    {
      "epoch": 0.1629374641422834,
      "grad_norm": 0.12837018072605133,
      "learning_rate": 0.00019759988399185505,
      "loss": 0.9771,
      "step": 710
    },
    {
      "epoch": 0.16408491107286288,
      "grad_norm": 0.13782522082328796,
      "learning_rate": 0.00019751185765714234,
      "loss": 0.9618,
      "step": 715
    },
    {
      "epoch": 0.16523235800344235,
      "grad_norm": 0.12514767050743103,
      "learning_rate": 0.00019742226636359296,
      "loss": 0.9594,
      "step": 720
    },
    {
      "epoch": 0.1663798049340218,
      "grad_norm": 0.1280808001756668,
      "learning_rate": 0.00019733111154904943,
      "loss": 0.9856,
      "step": 725
    },
    {
      "epoch": 0.16752725186460127,
      "grad_norm": 0.12929755449295044,
      "learning_rate": 0.00019723839467644699,
      "loss": 0.9382,
      "step": 730
    },
    {
      "epoch": 0.1686746987951807,
      "grad_norm": 0.5807048678398132,
      "learning_rate": 0.00019714411723379015,
      "loss": 0.9602,
      "step": 735
    },
    {
      "epoch": 0.16982214572576018,
      "grad_norm": 0.13152436912059784,
      "learning_rate": 0.0001970482807341289,
      "loss": 1.003,
      "step": 740
    },
    {
      "epoch": 0.17096959265633965,
      "grad_norm": 0.12607312202453613,
      "learning_rate": 0.0001969508867155345,
      "loss": 0.953,
      "step": 745
    },
    {
      "epoch": 0.1721170395869191,
      "grad_norm": 0.13689285516738892,
      "learning_rate": 0.00019685193674107452,
      "loss": 0.974,
      "step": 750
    },
    {
      "epoch": 0.17326448651749857,
      "grad_norm": 0.1324600875377655,
      "learning_rate": 0.00019675143239878805,
      "loss": 0.9953,
      "step": 755
    },
    {
      "epoch": 0.17441193344807804,
      "grad_norm": 0.13404534757137299,
      "learning_rate": 0.00019664937530166002,
      "loss": 0.9812,
      "step": 760
    },
    {
      "epoch": 0.17555938037865748,
      "grad_norm": 0.13384674489498138,
      "learning_rate": 0.00019654576708759538,
      "loss": 0.9632,
      "step": 765
    },
    {
      "epoch": 0.17670682730923695,
      "grad_norm": 0.1277248114347458,
      "learning_rate": 0.00019644060941939286,
      "loss": 0.9361,
      "step": 770
    },
    {
      "epoch": 0.1778542742398164,
      "grad_norm": 0.13291795551776886,
      "learning_rate": 0.00019633390398471817,
      "loss": 0.9762,
      "step": 775
    },
    {
      "epoch": 0.17900172117039587,
      "grad_norm": 0.13213355839252472,
      "learning_rate": 0.00019622565249607704,
      "loss": 0.9474,
      "step": 780
    },
    {
      "epoch": 0.18014916810097534,
      "grad_norm": 0.12764747440814972,
      "learning_rate": 0.0001961158566907877,
      "loss": 0.9329,
      "step": 785
    },
    {
      "epoch": 0.18129661503155478,
      "grad_norm": 0.13760751485824585,
      "learning_rate": 0.00019600451833095287,
      "loss": 0.9512,
      "step": 790
    },
    {
      "epoch": 0.18244406196213425,
      "grad_norm": 0.13257338106632233,
      "learning_rate": 0.00019589163920343163,
      "loss": 0.944,
      "step": 795
    },
    {
      "epoch": 0.18359150889271372,
      "grad_norm": 0.13907872140407562,
      "learning_rate": 0.00019577722111981078,
      "loss": 0.9839,
      "step": 800
    },
    {
      "epoch": 0.18473895582329317,
      "grad_norm": 0.1304904818534851,
      "learning_rate": 0.00019566126591637558,
      "loss": 0.9763,
      "step": 805
    },
    {
      "epoch": 0.18588640275387264,
      "grad_norm": 0.13839825987815857,
      "learning_rate": 0.0001955437754540805,
      "loss": 0.96,
      "step": 810
    },
    {
      "epoch": 0.18703384968445208,
      "grad_norm": 0.12638252973556519,
      "learning_rate": 0.00019542475161851906,
      "loss": 0.9421,
      "step": 815
    },
    {
      "epoch": 0.18818129661503155,
      "grad_norm": 0.12756413221359253,
      "learning_rate": 0.00019530419631989392,
      "loss": 0.9593,
      "step": 820
    },
    {
      "epoch": 0.18932874354561102,
      "grad_norm": 0.12539435923099518,
      "learning_rate": 0.00019518211149298595,
      "loss": 0.9643,
      "step": 825
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 0.12664757668972015,
      "learning_rate": 0.00019505849909712332,
      "loss": 0.9575,
      "step": 830
    },
    {
      "epoch": 0.19162363740676994,
      "grad_norm": 0.12583203613758087,
      "learning_rate": 0.00019493336111615003,
      "loss": 0.9277,
      "step": 835
    },
    {
      "epoch": 0.1927710843373494,
      "grad_norm": 0.1292329877614975,
      "learning_rate": 0.00019480669955839402,
      "loss": 0.9797,
      "step": 840
    },
    {
      "epoch": 0.19391853126792885,
      "grad_norm": 0.12693488597869873,
      "learning_rate": 0.00019467851645663494,
      "loss": 0.9749,
      "step": 845
    },
    {
      "epoch": 0.19506597819850832,
      "grad_norm": 0.13166065514087677,
      "learning_rate": 0.00019454881386807163,
      "loss": 0.9645,
      "step": 850
    },
    {
      "epoch": 0.1962134251290878,
      "grad_norm": 0.12855538725852966,
      "learning_rate": 0.00019441759387428903,
      "loss": 0.9463,
      "step": 855
    },
    {
      "epoch": 0.19736087205966724,
      "grad_norm": 0.12849357724189758,
      "learning_rate": 0.00019428485858122472,
      "loss": 0.988,
      "step": 860
    },
    {
      "epoch": 0.1985083189902467,
      "grad_norm": 0.12299315631389618,
      "learning_rate": 0.00019415061011913523,
      "loss": 0.9733,
      "step": 865
    },
    {
      "epoch": 0.19965576592082615,
      "grad_norm": 0.13390155136585236,
      "learning_rate": 0.00019401485064256176,
      "loss": 0.9502,
      "step": 870
    },
    {
      "epoch": 0.20080321285140562,
      "grad_norm": 0.13177447021007538,
      "learning_rate": 0.0001938775823302957,
      "loss": 0.9706,
      "step": 875
    },
    {
      "epoch": 0.2019506597819851,
      "grad_norm": 0.13233035802841187,
      "learning_rate": 0.00019373880738534358,
      "loss": 1.0028,
      "step": 880
    },
    {
      "epoch": 0.20309810671256454,
      "grad_norm": 0.12969984114170074,
      "learning_rate": 0.00019359852803489168,
      "loss": 0.9423,
      "step": 885
    },
    {
      "epoch": 0.204245553643144,
      "grad_norm": 0.13504785299301147,
      "learning_rate": 0.0001934567465302704,
      "loss": 0.9708,
      "step": 890
    },
    {
      "epoch": 0.20539300057372348,
      "grad_norm": 0.12698131799697876,
      "learning_rate": 0.00019331346514691813,
      "loss": 0.9548,
      "step": 895
    },
    {
      "epoch": 0.20654044750430292,
      "grad_norm": 0.1307503879070282,
      "learning_rate": 0.00019316868618434455,
      "loss": 0.9595,
      "step": 900
    },
    {
      "epoch": 0.2076878944348824,
      "grad_norm": 0.12671291828155518,
      "learning_rate": 0.00019302241196609397,
      "loss": 0.9456,
      "step": 905
    },
    {
      "epoch": 0.20883534136546184,
      "grad_norm": 0.1356162577867508,
      "learning_rate": 0.0001928746448397078,
      "loss": 0.9805,
      "step": 910
    },
    {
      "epoch": 0.2099827882960413,
      "grad_norm": 0.1294623613357544,
      "learning_rate": 0.00019272538717668715,
      "loss": 0.9831,
      "step": 915
    },
    {
      "epoch": 0.21113023522662078,
      "grad_norm": 0.13363958895206451,
      "learning_rate": 0.00019257464137245446,
      "loss": 0.9971,
      "step": 920
    },
    {
      "epoch": 0.21227768215720022,
      "grad_norm": 0.1361180692911148,
      "learning_rate": 0.0001924224098463153,
      "loss": 0.9393,
      "step": 925
    },
    {
      "epoch": 0.2134251290877797,
      "grad_norm": 0.1273190975189209,
      "learning_rate": 0.00019226869504141943,
      "loss": 0.9613,
      "step": 930
    },
    {
      "epoch": 0.21457257601835916,
      "grad_norm": 0.1308612823486328,
      "learning_rate": 0.00019211349942472165,
      "loss": 0.9703,
      "step": 935
    },
    {
      "epoch": 0.2157200229489386,
      "grad_norm": 0.13174669444561005,
      "learning_rate": 0.00019195682548694208,
      "loss": 0.9693,
      "step": 940
    },
    {
      "epoch": 0.21686746987951808,
      "grad_norm": 0.1349845975637436,
      "learning_rate": 0.00019179867574252638,
      "loss": 0.986,
      "step": 945
    },
    {
      "epoch": 0.21801491681009752,
      "grad_norm": 0.13183628022670746,
      "learning_rate": 0.00019163905272960528,
      "loss": 0.9804,
      "step": 950
    },
    {
      "epoch": 0.219162363740677,
      "grad_norm": 0.1328561156988144,
      "learning_rate": 0.0001914779590099538,
      "loss": 0.9582,
      "step": 955
    },
    {
      "epoch": 0.22030981067125646,
      "grad_norm": 0.1274995058774948,
      "learning_rate": 0.00019131539716895024,
      "loss": 0.9245,
      "step": 960
    },
    {
      "epoch": 0.2214572576018359,
      "grad_norm": 0.13776043057441711,
      "learning_rate": 0.00019115136981553464,
      "loss": 0.9683,
      "step": 965
    },
    {
      "epoch": 0.22260470453241538,
      "grad_norm": 0.1258002072572708,
      "learning_rate": 0.00019098587958216688,
      "loss": 0.968,
      "step": 970
    },
    {
      "epoch": 0.22375215146299485,
      "grad_norm": 0.13454443216323853,
      "learning_rate": 0.00019081892912478456,
      "loss": 0.9407,
      "step": 975
    },
    {
      "epoch": 0.2248995983935743,
      "grad_norm": 0.13280132412910461,
      "learning_rate": 0.00019065052112276018,
      "loss": 0.9714,
      "step": 980
    },
    {
      "epoch": 0.22604704532415376,
      "grad_norm": 0.1270975023508072,
      "learning_rate": 0.00019048065827885827,
      "loss": 0.942,
      "step": 985
    },
    {
      "epoch": 0.22719449225473323,
      "grad_norm": 0.13367241621017456,
      "learning_rate": 0.000190309343319192,
      "loss": 0.981,
      "step": 990
    },
    {
      "epoch": 0.22834193918531268,
      "grad_norm": 0.13318173587322235,
      "learning_rate": 0.00019013657899317942,
      "loss": 0.9668,
      "step": 995
    },
    {
      "epoch": 0.22948938611589215,
      "grad_norm": 0.1365250200033188,
      "learning_rate": 0.0001899623680734993,
      "loss": 0.939,
      "step": 1000
    },
    {
      "epoch": 0.2306368330464716,
      "grad_norm": 0.13613171875476837,
      "learning_rate": 0.00018978671335604665,
      "loss": 0.9375,
      "step": 1005
    },
    {
      "epoch": 0.23178427997705106,
      "grad_norm": 0.12359146773815155,
      "learning_rate": 0.00018960961765988792,
      "loss": 0.9686,
      "step": 1010
    },
    {
      "epoch": 0.23293172690763053,
      "grad_norm": 0.12440812587738037,
      "learning_rate": 0.00018943108382721562,
      "loss": 0.9284,
      "step": 1015
    },
    {
      "epoch": 0.23407917383820998,
      "grad_norm": 0.13303610682487488,
      "learning_rate": 0.00018925111472330283,
      "loss": 0.9631,
      "step": 1020
    },
    {
      "epoch": 0.23522662076878945,
      "grad_norm": 0.13873472809791565,
      "learning_rate": 0.00018906971323645713,
      "loss": 0.9854,
      "step": 1025
    },
    {
      "epoch": 0.23637406769936892,
      "grad_norm": 0.13052476942539215,
      "learning_rate": 0.00018888688227797432,
      "loss": 0.9422,
      "step": 1030
    },
    {
      "epoch": 0.23752151462994836,
      "grad_norm": 0.1390550136566162,
      "learning_rate": 0.00018870262478209163,
      "loss": 0.9847,
      "step": 1035
    },
    {
      "epoch": 0.23866896156052783,
      "grad_norm": 0.13032066822052002,
      "learning_rate": 0.00018851694370594069,
      "loss": 0.9539,
      "step": 1040
    },
    {
      "epoch": 0.23981640849110727,
      "grad_norm": 0.13494379818439484,
      "learning_rate": 0.00018832984202949996,
      "loss": 0.9427,
      "step": 1045
    },
    {
      "epoch": 0.24096385542168675,
      "grad_norm": 0.13439466059207916,
      "learning_rate": 0.00018814132275554713,
      "loss": 0.992,
      "step": 1050
    },
    {
      "epoch": 0.24211130235226622,
      "grad_norm": 0.1290155053138733,
      "learning_rate": 0.0001879513889096106,
      "loss": 0.9436,
      "step": 1055
    },
    {
      "epoch": 0.24325874928284566,
      "grad_norm": 0.13683564960956573,
      "learning_rate": 0.00018776004353992124,
      "loss": 0.9595,
      "step": 1060
    },
    {
      "epoch": 0.24440619621342513,
      "grad_norm": 0.12635931372642517,
      "learning_rate": 0.00018756728971736327,
      "loss": 0.9596,
      "step": 1065
    },
    {
      "epoch": 0.2455536431440046,
      "grad_norm": 0.12940926849842072,
      "learning_rate": 0.00018737313053542512,
      "loss": 0.976,
      "step": 1070
    },
    {
      "epoch": 0.24670109007458405,
      "grad_norm": 0.13413465023040771,
      "learning_rate": 0.0001871775691101496,
      "loss": 0.9374,
      "step": 1075
    },
    {
      "epoch": 0.24784853700516352,
      "grad_norm": 0.14798474311828613,
      "learning_rate": 0.00018698060858008403,
      "loss": 0.9331,
      "step": 1080
    },
    {
      "epoch": 0.24899598393574296,
      "grad_norm": 0.13452093303203583,
      "learning_rate": 0.00018678225210622986,
      "loss": 0.9893,
      "step": 1085
    },
    {
      "epoch": 0.25014343086632246,
      "grad_norm": 0.13061241805553436,
      "learning_rate": 0.00018658250287199196,
      "loss": 0.9793,
      "step": 1090
    },
    {
      "epoch": 0.2512908777969019,
      "grad_norm": 0.13494077324867249,
      "learning_rate": 0.00018638136408312728,
      "loss": 0.9793,
      "step": 1095
    },
    {
      "epoch": 0.25243832472748134,
      "grad_norm": 0.12755267322063446,
      "learning_rate": 0.0001861788389676939,
      "loss": 0.957,
      "step": 1100
    },
    {
      "epoch": 0.2535857716580608,
      "grad_norm": 0.13325680792331696,
      "learning_rate": 0.00018597493077599867,
      "loss": 0.9778,
      "step": 1105
    },
    {
      "epoch": 0.2547332185886403,
      "grad_norm": 0.13293135166168213,
      "learning_rate": 0.00018576964278054544,
      "loss": 0.9538,
      "step": 1110
    },
    {
      "epoch": 0.25588066551921973,
      "grad_norm": 0.13516201078891754,
      "learning_rate": 0.00018556297827598242,
      "loss": 0.9815,
      "step": 1115
    },
    {
      "epoch": 0.2570281124497992,
      "grad_norm": 0.12776044011116028,
      "learning_rate": 0.00018535494057904915,
      "loss": 0.9513,
      "step": 1120
    },
    {
      "epoch": 0.25817555938037867,
      "grad_norm": 0.12933199107646942,
      "learning_rate": 0.00018514553302852356,
      "loss": 0.9749,
      "step": 1125
    },
    {
      "epoch": 0.2593230063109581,
      "grad_norm": 0.13189588487148285,
      "learning_rate": 0.00018493475898516813,
      "loss": 0.9604,
      "step": 1130
    },
    {
      "epoch": 0.26047045324153756,
      "grad_norm": 0.12965457141399384,
      "learning_rate": 0.00018472262183167614,
      "loss": 0.9882,
      "step": 1135
    },
    {
      "epoch": 0.26161790017211706,
      "grad_norm": 0.13817626237869263,
      "learning_rate": 0.00018450912497261723,
      "loss": 0.9508,
      "step": 1140
    },
    {
      "epoch": 0.2627653471026965,
      "grad_norm": 0.12874102592468262,
      "learning_rate": 0.00018429427183438288,
      "loss": 0.9871,
      "step": 1145
    },
    {
      "epoch": 0.26391279403327594,
      "grad_norm": 0.12571579217910767,
      "learning_rate": 0.00018407806586513134,
      "loss": 0.9739,
      "step": 1150
    },
    {
      "epoch": 0.26506024096385544,
      "grad_norm": 0.13456568121910095,
      "learning_rate": 0.00018386051053473232,
      "loss": 0.9752,
      "step": 1155
    },
    {
      "epoch": 0.2662076878944349,
      "grad_norm": 0.1293318271636963,
      "learning_rate": 0.00018364160933471134,
      "loss": 0.9483,
      "step": 1160
    },
    {
      "epoch": 0.26735513482501433,
      "grad_norm": 0.13392923772335052,
      "learning_rate": 0.0001834213657781936,
      "loss": 0.9443,
      "step": 1165
    },
    {
      "epoch": 0.2685025817555938,
      "grad_norm": 0.1313287764787674,
      "learning_rate": 0.00018319978339984767,
      "loss": 0.9485,
      "step": 1170
    },
    {
      "epoch": 0.26965002868617327,
      "grad_norm": 0.13383394479751587,
      "learning_rate": 0.0001829768657558288,
      "loss": 0.9647,
      "step": 1175
    },
    {
      "epoch": 0.2707974756167527,
      "grad_norm": 0.12766633927822113,
      "learning_rate": 0.00018275261642372175,
      "loss": 0.9627,
      "step": 1180
    },
    {
      "epoch": 0.2719449225473322,
      "grad_norm": 0.12958547472953796,
      "learning_rate": 0.0001825270390024834,
      "loss": 0.9628,
      "step": 1185
    },
    {
      "epoch": 0.27309236947791166,
      "grad_norm": 0.1323728859424591,
      "learning_rate": 0.00018230013711238513,
      "loss": 0.9905,
      "step": 1190
    },
    {
      "epoch": 0.2742398164084911,
      "grad_norm": 0.13171570003032684,
      "learning_rate": 0.00018207191439495438,
      "loss": 0.9514,
      "step": 1195
    },
    {
      "epoch": 0.27538726333907054,
      "grad_norm": 0.13061773777008057,
      "learning_rate": 0.00018184237451291665,
      "loss": 0.9644,
      "step": 1200
    },
    {
      "epoch": 0.27653471026965004,
      "grad_norm": 0.1268092542886734,
      "learning_rate": 0.00018161152115013637,
      "loss": 0.9399,
      "step": 1205
    },
    {
      "epoch": 0.2776821572002295,
      "grad_norm": 0.12714898586273193,
      "learning_rate": 0.00018137935801155794,
      "loss": 1.0103,
      "step": 1210
    },
    {
      "epoch": 0.27882960413080893,
      "grad_norm": 0.16152754426002502,
      "learning_rate": 0.0001811458888231462,
      "loss": 0.9428,
      "step": 1215
    },
    {
      "epoch": 0.2799770510613884,
      "grad_norm": 0.13298989832401276,
      "learning_rate": 0.0001809111173318267,
      "loss": 0.9532,
      "step": 1220
    },
    {
      "epoch": 0.28112449799196787,
      "grad_norm": 0.1261643022298813,
      "learning_rate": 0.00018067504730542551,
      "loss": 1.0076,
      "step": 1225
    },
    {
      "epoch": 0.2822719449225473,
      "grad_norm": 0.13440310955047607,
      "learning_rate": 0.0001804376825326088,
      "loss": 0.962,
      "step": 1230
    },
    {
      "epoch": 0.2834193918531268,
      "grad_norm": 0.13871833682060242,
      "learning_rate": 0.00018019902682282193,
      "loss": 0.9581,
      "step": 1235
    },
    {
      "epoch": 0.28456683878370626,
      "grad_norm": 0.12701858580112457,
      "learning_rate": 0.0001799590840062285,
      "loss": 0.9703,
      "step": 1240
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.13460960984230042,
      "learning_rate": 0.00017971785793364866,
      "loss": 0.9483,
      "step": 1245
    },
    {
      "epoch": 0.2868617326448652,
      "grad_norm": 0.12756766378879547,
      "learning_rate": 0.0001794753524764975,
      "loss": 0.9693,
      "step": 1250
    },
    {
      "epoch": 0.28800917957544464,
      "grad_norm": 0.12904788553714752,
      "learning_rate": 0.00017923157152672278,
      "loss": 0.9874,
      "step": 1255
    },
    {
      "epoch": 0.2891566265060241,
      "grad_norm": 0.1294935941696167,
      "learning_rate": 0.00017898651899674254,
      "loss": 0.9746,
      "step": 1260
    },
    {
      "epoch": 0.2903040734366036,
      "grad_norm": 0.13519194722175598,
      "learning_rate": 0.00017874019881938233,
      "loss": 0.9489,
      "step": 1265
    },
    {
      "epoch": 0.291451520367183,
      "grad_norm": 0.1270766407251358,
      "learning_rate": 0.000178492614947812,
      "loss": 0.9573,
      "step": 1270
    },
    {
      "epoch": 0.29259896729776247,
      "grad_norm": 0.1287229061126709,
      "learning_rate": 0.00017824377135548236,
      "loss": 0.9544,
      "step": 1275
    },
    {
      "epoch": 0.2937464142283419,
      "grad_norm": 0.1388266533613205,
      "learning_rate": 0.00017799367203606128,
      "loss": 0.9781,
      "step": 1280
    },
    {
      "epoch": 0.2948938611589214,
      "grad_norm": 0.1323292851448059,
      "learning_rate": 0.00017774232100336982,
      "loss": 0.9783,
      "step": 1285
    },
    {
      "epoch": 0.29604130808950085,
      "grad_norm": 0.12682774662971497,
      "learning_rate": 0.00017748972229131757,
      "loss": 0.9363,
      "step": 1290
    },
    {
      "epoch": 0.2971887550200803,
      "grad_norm": 0.12563958764076233,
      "learning_rate": 0.000177235879953838,
      "loss": 0.972,
      "step": 1295
    },
    {
      "epoch": 0.2983362019506598,
      "grad_norm": 0.13384099304676056,
      "learning_rate": 0.00017698079806482343,
      "loss": 0.9991,
      "step": 1300
    },
    {
      "epoch": 0.29948364888123924,
      "grad_norm": 0.12890778481960297,
      "learning_rate": 0.0001767244807180597,
      "loss": 0.9437,
      "step": 1305
    },
    {
      "epoch": 0.3006310958118187,
      "grad_norm": 0.13081791996955872,
      "learning_rate": 0.00017646693202716033,
      "loss": 0.9583,
      "step": 1310
    },
    {
      "epoch": 0.3017785427423982,
      "grad_norm": 0.13217350840568542,
      "learning_rate": 0.0001762081561255005,
      "loss": 0.9599,
      "step": 1315
    },
    {
      "epoch": 0.3029259896729776,
      "grad_norm": 0.1300850510597229,
      "learning_rate": 0.00017594815716615093,
      "loss": 0.96,
      "step": 1320
    },
    {
      "epoch": 0.30407343660355707,
      "grad_norm": 0.13618028163909912,
      "learning_rate": 0.000175686939321811,
      "loss": 0.9466,
      "step": 1325
    },
    {
      "epoch": 0.30522088353413657,
      "grad_norm": 0.13294175267219543,
      "learning_rate": 0.00017542450678474184,
      "loss": 0.9794,
      "step": 1330
    },
    {
      "epoch": 0.306368330464716,
      "grad_norm": 0.12827391922473907,
      "learning_rate": 0.00017516086376669917,
      "loss": 0.9388,
      "step": 1335
    },
    {
      "epoch": 0.30751577739529545,
      "grad_norm": 0.1285424679517746,
      "learning_rate": 0.00017489601449886547,
      "loss": 0.9533,
      "step": 1340
    },
    {
      "epoch": 0.30866322432587495,
      "grad_norm": 0.13624240458011627,
      "learning_rate": 0.00017462996323178235,
      "loss": 0.93,
      "step": 1345
    },
    {
      "epoch": 0.3098106712564544,
      "grad_norm": 0.13076668977737427,
      "learning_rate": 0.00017436271423528206,
      "loss": 0.9547,
      "step": 1350
    },
    {
      "epoch": 0.31095811818703384,
      "grad_norm": 0.12973956763744354,
      "learning_rate": 0.0001740942717984192,
      "loss": 1.0114,
      "step": 1355
    },
    {
      "epoch": 0.31210556511761334,
      "grad_norm": 0.13649657368659973,
      "learning_rate": 0.00017382464022940182,
      "loss": 0.9735,
      "step": 1360
    },
    {
      "epoch": 0.3132530120481928,
      "grad_norm": 0.13023148477077484,
      "learning_rate": 0.00017355382385552206,
      "loss": 0.9856,
      "step": 1365
    },
    {
      "epoch": 0.3144004589787722,
      "grad_norm": 0.1284298449754715,
      "learning_rate": 0.0001732818270230871,
      "loss": 0.966,
      "step": 1370
    },
    {
      "epoch": 0.31554790590935167,
      "grad_norm": 0.1332443803548813,
      "learning_rate": 0.000173008654097349,
      "loss": 0.9429,
      "step": 1375
    },
    {
      "epoch": 0.31669535283993117,
      "grad_norm": 0.13395999372005463,
      "learning_rate": 0.000172734309462435,
      "loss": 0.9748,
      "step": 1380
    },
    {
      "epoch": 0.3178427997705106,
      "grad_norm": 0.12902694940567017,
      "learning_rate": 0.00017245879752127692,
      "loss": 0.9386,
      "step": 1385
    },
    {
      "epoch": 0.31899024670109005,
      "grad_norm": 0.12902621924877167,
      "learning_rate": 0.0001721821226955405,
      "loss": 0.9216,
      "step": 1390
    },
    {
      "epoch": 0.32013769363166955,
      "grad_norm": 0.1328728049993515,
      "learning_rate": 0.00017190428942555463,
      "loss": 0.943,
      "step": 1395
    },
    {
      "epoch": 0.321285140562249,
      "grad_norm": 0.12963657081127167,
      "learning_rate": 0.0001716253021702399,
      "loss": 0.9254,
      "step": 1400
    },
    {
      "epoch": 0.32243258749282844,
      "grad_norm": 0.12427925318479538,
      "learning_rate": 0.0001713451654070371,
      "loss": 0.9665,
      "step": 1405
    },
    {
      "epoch": 0.32358003442340794,
      "grad_norm": 0.12566518783569336,
      "learning_rate": 0.0001710638836318354,
      "loss": 0.9778,
      "step": 1410
    },
    {
      "epoch": 0.3247274813539874,
      "grad_norm": 0.1280212700366974,
      "learning_rate": 0.00017078146135890014,
      "loss": 0.9761,
      "step": 1415
    },
    {
      "epoch": 0.3258749282845668,
      "grad_norm": 0.14105786383152008,
      "learning_rate": 0.0001704979031208004,
      "loss": 0.9658,
      "step": 1420
    },
    {
      "epoch": 0.3270223752151463,
      "grad_norm": 0.1319873183965683,
      "learning_rate": 0.0001702132134683363,
      "loss": 0.96,
      "step": 1425
    },
    {
      "epoch": 0.32816982214572576,
      "grad_norm": 0.12675578892230988,
      "learning_rate": 0.00016992739697046586,
      "loss": 0.9478,
      "step": 1430
    },
    {
      "epoch": 0.3293172690763052,
      "grad_norm": 0.1259857565164566,
      "learning_rate": 0.00016964045821423178,
      "loss": 0.9258,
      "step": 1435
    },
    {
      "epoch": 0.3304647160068847,
      "grad_norm": 0.1366124451160431,
      "learning_rate": 0.00016935240180468775,
      "loss": 0.9556,
      "step": 1440
    },
    {
      "epoch": 0.33161216293746415,
      "grad_norm": 0.13446134328842163,
      "learning_rate": 0.00016906323236482465,
      "loss": 0.9675,
      "step": 1445
    },
    {
      "epoch": 0.3327596098680436,
      "grad_norm": 0.13119171559810638,
      "learning_rate": 0.00016877295453549614,
      "loss": 0.9208,
      "step": 1450
    },
    {
      "epoch": 0.33390705679862304,
      "grad_norm": 0.13426977396011353,
      "learning_rate": 0.00016848157297534453,
      "loss": 0.9755,
      "step": 1455
    },
    {
      "epoch": 0.33505450372920254,
      "grad_norm": 0.1286066770553589,
      "learning_rate": 0.0001681890923607256,
      "loss": 0.9624,
      "step": 1460
    },
    {
      "epoch": 0.336201950659782,
      "grad_norm": 0.1409013718366623,
      "learning_rate": 0.00016789551738563384,
      "loss": 0.9616,
      "step": 1465
    },
    {
      "epoch": 0.3373493975903614,
      "grad_norm": 0.1301148682832718,
      "learning_rate": 0.00016760085276162708,
      "loss": 0.9543,
      "step": 1470
    },
    {
      "epoch": 0.3384968445209409,
      "grad_norm": 0.12759418785572052,
      "learning_rate": 0.00016730510321775075,
      "loss": 0.9395,
      "step": 1475
    },
    {
      "epoch": 0.33964429145152036,
      "grad_norm": 0.12628066539764404,
      "learning_rate": 0.00016700827350046206,
      "loss": 0.951,
      "step": 1480
    },
    {
      "epoch": 0.3407917383820998,
      "grad_norm": 0.13049326837062836,
      "learning_rate": 0.00016671036837355386,
      "loss": 0.9897,
      "step": 1485
    },
    {
      "epoch": 0.3419391853126793,
      "grad_norm": 0.13148042559623718,
      "learning_rate": 0.00016641139261807818,
      "loss": 0.9477,
      "step": 1490
    },
    {
      "epoch": 0.34308663224325875,
      "grad_norm": 0.1383514553308487,
      "learning_rate": 0.00016611135103226937,
      "loss": 0.9836,
      "step": 1495
    },
    {
      "epoch": 0.3442340791738382,
      "grad_norm": 0.13166822493076324,
      "learning_rate": 0.00016581024843146725,
      "loss": 0.9857,
      "step": 1500
    },
    {
      "epoch": 0.3453815261044177,
      "grad_norm": 0.1284855753183365,
      "learning_rate": 0.00016550808964803978,
      "loss": 0.9568,
      "step": 1505
    },
    {
      "epoch": 0.34652897303499713,
      "grad_norm": 0.13784931600093842,
      "learning_rate": 0.00016520487953130552,
      "loss": 0.9829,
      "step": 1510
    },
    {
      "epoch": 0.3476764199655766,
      "grad_norm": 0.1401350200176239,
      "learning_rate": 0.00016490062294745571,
      "loss": 0.9468,
      "step": 1515
    },
    {
      "epoch": 0.3488238668961561,
      "grad_norm": 0.13035733997821808,
      "learning_rate": 0.00016459532477947634,
      "loss": 0.9681,
      "step": 1520
    },
    {
      "epoch": 0.3499713138267355,
      "grad_norm": 0.12921980023384094,
      "learning_rate": 0.00016428898992706955,
      "loss": 0.9839,
      "step": 1525
    },
    {
      "epoch": 0.35111876075731496,
      "grad_norm": 0.12521222233772278,
      "learning_rate": 0.00016398162330657533,
      "loss": 0.97,
      "step": 1530
    },
    {
      "epoch": 0.35226620768789446,
      "grad_norm": 0.13276994228363037,
      "learning_rate": 0.0001636732298508922,
      "loss": 0.9453,
      "step": 1535
    },
    {
      "epoch": 0.3534136546184739,
      "grad_norm": 0.1314193159341812,
      "learning_rate": 0.0001633638145093984,
      "loss": 0.9544,
      "step": 1540
    },
    {
      "epoch": 0.35456110154905335,
      "grad_norm": 0.13540154695510864,
      "learning_rate": 0.00016305338224787235,
      "loss": 0.9958,
      "step": 1545
    },
    {
      "epoch": 0.3557085484796328,
      "grad_norm": 0.13043038547039032,
      "learning_rate": 0.0001627419380484128,
      "loss": 0.9166,
      "step": 1550
    },
    {
      "epoch": 0.3568559954102123,
      "grad_norm": 0.13303428888320923,
      "learning_rate": 0.00016242948690935912,
      "loss": 0.9615,
      "step": 1555
    },
    {
      "epoch": 0.35800344234079173,
      "grad_norm": 0.13199162483215332,
      "learning_rate": 0.00016211603384521083,
      "loss": 0.9632,
      "step": 1560
    },
    {
      "epoch": 0.3591508892713712,
      "grad_norm": 0.1341785341501236,
      "learning_rate": 0.00016180158388654742,
      "loss": 0.9744,
      "step": 1565
    },
    {
      "epoch": 0.3602983362019507,
      "grad_norm": 0.13043268024921417,
      "learning_rate": 0.00016148614207994735,
      "loss": 0.9494,
      "step": 1570
    },
    {
      "epoch": 0.3614457831325301,
      "grad_norm": 0.13018083572387695,
      "learning_rate": 0.00016116971348790712,
      "loss": 0.9418,
      "step": 1575
    },
    {
      "epoch": 0.36259323006310956,
      "grad_norm": 0.1259637027978897,
      "learning_rate": 0.0001608523031887601,
      "loss": 0.9399,
      "step": 1580
    },
    {
      "epoch": 0.36374067699368906,
      "grad_norm": 0.13221842050552368,
      "learning_rate": 0.00016053391627659505,
      "loss": 0.9549,
      "step": 1585
    },
    {
      "epoch": 0.3648881239242685,
      "grad_norm": 0.14202679693698883,
      "learning_rate": 0.0001602145578611742,
      "loss": 0.9279,
      "step": 1590
    },
    {
      "epoch": 0.36603557085484795,
      "grad_norm": 0.13095086812973022,
      "learning_rate": 0.00015989423306785142,
      "loss": 0.9788,
      "step": 1595
    },
    {
      "epoch": 0.36718301778542745,
      "grad_norm": 0.12878695130348206,
      "learning_rate": 0.00015957294703748982,
      "loss": 0.9564,
      "step": 1600
    },
    {
      "epoch": 0.3683304647160069,
      "grad_norm": 0.12988603115081787,
      "learning_rate": 0.00015925070492637944,
      "loss": 0.9672,
      "step": 1605
    },
    {
      "epoch": 0.36947791164658633,
      "grad_norm": 0.13579297065734863,
      "learning_rate": 0.0001589275119061542,
      "loss": 0.9381,
      "step": 1610
    },
    {
      "epoch": 0.37062535857716583,
      "grad_norm": 0.12680459022521973,
      "learning_rate": 0.00015860337316370916,
      "loss": 0.9512,
      "step": 1615
    },
    {
      "epoch": 0.3717728055077453,
      "grad_norm": 0.12983225286006927,
      "learning_rate": 0.0001582782939011173,
      "loss": 0.9539,
      "step": 1620
    },
    {
      "epoch": 0.3729202524383247,
      "grad_norm": 0.13091208040714264,
      "learning_rate": 0.00015795227933554568,
      "loss": 0.9675,
      "step": 1625
    },
    {
      "epoch": 0.37406769936890416,
      "grad_norm": 0.13072653114795685,
      "learning_rate": 0.00015762533469917216,
      "loss": 0.9327,
      "step": 1630
    },
    {
      "epoch": 0.37521514629948366,
      "grad_norm": 0.13025379180908203,
      "learning_rate": 0.00015729746523910113,
      "loss": 0.9623,
      "step": 1635
    },
    {
      "epoch": 0.3763625932300631,
      "grad_norm": 0.14888666570186615,
      "learning_rate": 0.00015696867621727942,
      "loss": 0.9436,
      "step": 1640
    },
    {
      "epoch": 0.37751004016064255,
      "grad_norm": 0.1301380842924118,
      "learning_rate": 0.00015663897291041175,
      "loss": 0.9613,
      "step": 1645
    },
    {
      "epoch": 0.37865748709122204,
      "grad_norm": 0.12864987552165985,
      "learning_rate": 0.00015630836060987624,
      "loss": 0.9701,
      "step": 1650
    },
    {
      "epoch": 0.3798049340218015,
      "grad_norm": 0.13158966600894928,
      "learning_rate": 0.00015597684462163923,
      "loss": 1.0041,
      "step": 1655
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 0.12787610292434692,
      "learning_rate": 0.0001556444302661704,
      "loss": 0.942,
      "step": 1660
    },
    {
      "epoch": 0.38209982788296043,
      "grad_norm": 0.1329614371061325,
      "learning_rate": 0.00015531112287835717,
      "loss": 0.9607,
      "step": 1665
    },
    {
      "epoch": 0.3832472748135399,
      "grad_norm": 0.13437345623970032,
      "learning_rate": 0.00015497692780741908,
      "loss": 0.9657,
      "step": 1670
    },
    {
      "epoch": 0.3843947217441193,
      "grad_norm": 0.13434943556785583,
      "learning_rate": 0.0001546418504168222,
      "loss": 0.9635,
      "step": 1675
    },
    {
      "epoch": 0.3855421686746988,
      "grad_norm": 0.129192054271698,
      "learning_rate": 0.00015430589608419264,
      "loss": 0.9527,
      "step": 1680
    },
    {
      "epoch": 0.38668961560527826,
      "grad_norm": 0.1313447207212448,
      "learning_rate": 0.00015396907020123068,
      "loss": 0.942,
      "step": 1685
    },
    {
      "epoch": 0.3878370625358577,
      "grad_norm": 0.13218559324741364,
      "learning_rate": 0.00015363137817362392,
      "loss": 0.9602,
      "step": 1690
    },
    {
      "epoch": 0.3889845094664372,
      "grad_norm": 0.14416509866714478,
      "learning_rate": 0.00015329282542096064,
      "loss": 0.9419,
      "step": 1695
    },
    {
      "epoch": 0.39013195639701664,
      "grad_norm": 0.14515794813632965,
      "learning_rate": 0.00015295341737664285,
      "loss": 0.9381,
      "step": 1700
    },
    {
      "epoch": 0.3912794033275961,
      "grad_norm": 0.13562296330928802,
      "learning_rate": 0.000152613159487799,
      "loss": 0.9594,
      "step": 1705
    },
    {
      "epoch": 0.3924268502581756,
      "grad_norm": 0.13049104809761047,
      "learning_rate": 0.00015227205721519675,
      "loss": 0.9472,
      "step": 1710
    },
    {
      "epoch": 0.39357429718875503,
      "grad_norm": 0.13842852413654327,
      "learning_rate": 0.00015193011603315503,
      "loss": 0.995,
      "step": 1715
    },
    {
      "epoch": 0.3947217441193345,
      "grad_norm": 0.14135673642158508,
      "learning_rate": 0.00015158734142945644,
      "loss": 0.9992,
      "step": 1720
    },
    {
      "epoch": 0.3958691910499139,
      "grad_norm": 0.13362735509872437,
      "learning_rate": 0.0001512437389052591,
      "loss": 0.9279,
      "step": 1725
    },
    {
      "epoch": 0.3970166379804934,
      "grad_norm": 0.1268196702003479,
      "learning_rate": 0.0001508993139750083,
      "loss": 0.9369,
      "step": 1730
    },
    {
      "epoch": 0.39816408491107286,
      "grad_norm": 0.12576265633106232,
      "learning_rate": 0.0001505540721663481,
      "loss": 0.9777,
      "step": 1735
    },
    {
      "epoch": 0.3993115318416523,
      "grad_norm": 0.1377476304769516,
      "learning_rate": 0.0001502080190200325,
      "loss": 0.96,
      "step": 1740
    },
    {
      "epoch": 0.4004589787722318,
      "grad_norm": 0.12954863905906677,
      "learning_rate": 0.00014986116008983664,
      "loss": 0.9612,
      "step": 1745
    },
    {
      "epoch": 0.40160642570281124,
      "grad_norm": 0.13743706047534943,
      "learning_rate": 0.00014951350094246762,
      "loss": 0.977,
      "step": 1750
    },
    {
      "epoch": 0.4027538726333907,
      "grad_norm": 0.1272636353969574,
      "learning_rate": 0.0001491650471574751,
      "loss": 0.9603,
      "step": 1755
    },
    {
      "epoch": 0.4039013195639702,
      "grad_norm": 0.13247370719909668,
      "learning_rate": 0.00014881580432716182,
      "loss": 0.9251,
      "step": 1760
    },
    {
      "epoch": 0.40504876649454963,
      "grad_norm": 0.13983896374702454,
      "learning_rate": 0.00014846577805649388,
      "loss": 0.9556,
      "step": 1765
    },
    {
      "epoch": 0.40619621342512907,
      "grad_norm": 0.1289101541042328,
      "learning_rate": 0.00014811497396301072,
      "loss": 0.9391,
      "step": 1770
    },
    {
      "epoch": 0.40734366035570857,
      "grad_norm": 0.12569357454776764,
      "learning_rate": 0.00014776339767673491,
      "loss": 0.9301,
      "step": 1775
    },
    {
      "epoch": 0.408491107286288,
      "grad_norm": 0.1362708956003189,
      "learning_rate": 0.000147411054840082,
      "loss": 0.9802,
      "step": 1780
    },
    {
      "epoch": 0.40963855421686746,
      "grad_norm": 0.13561737537384033,
      "learning_rate": 0.00014705795110776974,
      "loss": 0.9624,
      "step": 1785
    },
    {
      "epoch": 0.41078600114744696,
      "grad_norm": 0.13607259094715118,
      "learning_rate": 0.0001467040921467275,
      "loss": 0.9543,
      "step": 1790
    },
    {
      "epoch": 0.4119334480780264,
      "grad_norm": 0.13344360888004303,
      "learning_rate": 0.00014634948363600518,
      "loss": 0.959,
      "step": 1795
    },
    {
      "epoch": 0.41308089500860584,
      "grad_norm": 0.14161522686481476,
      "learning_rate": 0.00014599413126668213,
      "loss": 0.9432,
      "step": 1800
    },
    {
      "epoch": 0.4142283419391853,
      "grad_norm": 0.13088949024677277,
      "learning_rate": 0.00014563804074177588,
      "loss": 0.9647,
      "step": 1805
    },
    {
      "epoch": 0.4153757888697648,
      "grad_norm": 0.13591496646404266,
      "learning_rate": 0.00014528121777615058,
      "loss": 0.9531,
      "step": 1810
    },
    {
      "epoch": 0.4165232358003442,
      "grad_norm": 0.1339530646800995,
      "learning_rate": 0.0001449236680964251,
      "loss": 0.9549,
      "step": 1815
    },
    {
      "epoch": 0.41767068273092367,
      "grad_norm": 0.12666349112987518,
      "learning_rate": 0.00014456539744088138,
      "loss": 0.9424,
      "step": 1820
    },
    {
      "epoch": 0.41881812966150317,
      "grad_norm": 0.12611252069473267,
      "learning_rate": 0.00014420641155937224,
      "loss": 0.9359,
      "step": 1825
    },
    {
      "epoch": 0.4199655765920826,
      "grad_norm": 0.13051684200763702,
      "learning_rate": 0.00014384671621322915,
      "loss": 0.9488,
      "step": 1830
    },
    {
      "epoch": 0.42111302352266206,
      "grad_norm": 0.13898026943206787,
      "learning_rate": 0.00014348631717516953,
      "loss": 0.9748,
      "step": 1835
    },
    {
      "epoch": 0.42226047045324155,
      "grad_norm": 0.13012805581092834,
      "learning_rate": 0.00014312522022920444,
      "loss": 0.9631,
      "step": 1840
    },
    {
      "epoch": 0.423407917383821,
      "grad_norm": 0.12887822091579437,
      "learning_rate": 0.00014276343117054563,
      "loss": 0.9532,
      "step": 1845
    },
    {
      "epoch": 0.42455536431440044,
      "grad_norm": 0.14254815876483917,
      "learning_rate": 0.00014240095580551234,
      "loss": 1.0004,
      "step": 1850
    },
    {
      "epoch": 0.42570281124497994,
      "grad_norm": 0.13232548534870148,
      "learning_rate": 0.0001420377999514384,
      "loss": 0.9644,
      "step": 1855
    },
    {
      "epoch": 0.4268502581755594,
      "grad_norm": 0.13256525993347168,
      "learning_rate": 0.00014167396943657874,
      "loss": 0.9671,
      "step": 1860
    },
    {
      "epoch": 0.4279977051061388,
      "grad_norm": 0.13868069648742676,
      "learning_rate": 0.0001413094701000158,
      "loss": 0.924,
      "step": 1865
    },
    {
      "epoch": 0.4291451520367183,
      "grad_norm": 0.13542011380195618,
      "learning_rate": 0.00014094430779156597,
      "loss": 0.9751,
      "step": 1870
    },
    {
      "epoch": 0.43029259896729777,
      "grad_norm": 0.15172503888607025,
      "learning_rate": 0.00014057848837168547,
      "loss": 0.9811,
      "step": 1875
    },
    {
      "epoch": 0.4314400458978772,
      "grad_norm": 0.13170887529850006,
      "learning_rate": 0.00014021201771137663,
      "loss": 0.9598,
      "step": 1880
    },
    {
      "epoch": 0.4325874928284567,
      "grad_norm": 0.13413281738758087,
      "learning_rate": 0.00013984490169209333,
      "loss": 0.9251,
      "step": 1885
    },
    {
      "epoch": 0.43373493975903615,
      "grad_norm": 0.13315904140472412,
      "learning_rate": 0.00013947714620564678,
      "loss": 0.9983,
      "step": 1890
    },
    {
      "epoch": 0.4348823866896156,
      "grad_norm": 0.131989523768425,
      "learning_rate": 0.00013910875715411098,
      "loss": 0.9411,
      "step": 1895
    },
    {
      "epoch": 0.43602983362019504,
      "grad_norm": 0.1380765587091446,
      "learning_rate": 0.000138739740449728,
      "loss": 0.9292,
      "step": 1900
    },
    {
      "epoch": 0.43717728055077454,
      "grad_norm": 0.13661834597587585,
      "learning_rate": 0.0001383701020148129,
      "loss": 0.9477,
      "step": 1905
    },
    {
      "epoch": 0.438324727481354,
      "grad_norm": 0.13486482203006744,
      "learning_rate": 0.000137999847781659,
      "loss": 1.0169,
      "step": 1910
    },
    {
      "epoch": 0.4394721744119334,
      "grad_norm": 0.13214510679244995,
      "learning_rate": 0.00013762898369244238,
      "loss": 0.9418,
      "step": 1915
    },
    {
      "epoch": 0.4406196213425129,
      "grad_norm": 0.12652263045310974,
      "learning_rate": 0.00013725751569912682,
      "loss": 0.9679,
      "step": 1920
    },
    {
      "epoch": 0.44176706827309237,
      "grad_norm": 0.13453719019889832,
      "learning_rate": 0.00013688544976336783,
      "loss": 0.9318,
      "step": 1925
    },
    {
      "epoch": 0.4429145152036718,
      "grad_norm": 0.14246058464050293,
      "learning_rate": 0.00013651279185641752,
      "loss": 0.9544,
      "step": 1930
    },
    {
      "epoch": 0.4440619621342513,
      "grad_norm": 0.13308945298194885,
      "learning_rate": 0.0001361395479590283,
      "loss": 0.9817,
      "step": 1935
    },
    {
      "epoch": 0.44520940906483075,
      "grad_norm": 0.12958864867687225,
      "learning_rate": 0.00013576572406135722,
      "loss": 0.9555,
      "step": 1940
    },
    {
      "epoch": 0.4463568559954102,
      "grad_norm": 0.1342882663011551,
      "learning_rate": 0.00013539132616286956,
      "loss": 0.9723,
      "step": 1945
    },
    {
      "epoch": 0.4475043029259897,
      "grad_norm": 0.12933650612831116,
      "learning_rate": 0.00013501636027224282,
      "loss": 0.9516,
      "step": 1950
    },
    {
      "epoch": 0.44865174985656914,
      "grad_norm": 0.1319791078567505,
      "learning_rate": 0.0001346408324072701,
      "loss": 0.9619,
      "step": 1955
    },
    {
      "epoch": 0.4497991967871486,
      "grad_norm": 0.13945066928863525,
      "learning_rate": 0.0001342647485947635,
      "loss": 0.9109,
      "step": 1960
    },
    {
      "epoch": 0.4509466437177281,
      "grad_norm": 0.13085277378559113,
      "learning_rate": 0.00013388811487045766,
      "loss": 0.968,
      "step": 1965
    },
    {
      "epoch": 0.4520940906483075,
      "grad_norm": 0.12879082560539246,
      "learning_rate": 0.0001335109372789125,
      "loss": 0.9263,
      "step": 1970
    },
    {
      "epoch": 0.45324153757888697,
      "grad_norm": 0.13467159867286682,
      "learning_rate": 0.00013313322187341652,
      "loss": 0.9529,
      "step": 1975
    },
    {
      "epoch": 0.45438898450946646,
      "grad_norm": 0.12812353670597076,
      "learning_rate": 0.00013275497471588953,
      "loss": 0.9455,
      "step": 1980
    },
    {
      "epoch": 0.4555364314400459,
      "grad_norm": 0.1305771768093109,
      "learning_rate": 0.0001323762018767854,
      "loss": 0.9627,
      "step": 1985
    },
    {
      "epoch": 0.45668387837062535,
      "grad_norm": 0.14161469042301178,
      "learning_rate": 0.00013199690943499457,
      "loss": 0.9981,
      "step": 1990
    },
    {
      "epoch": 0.4578313253012048,
      "grad_norm": 0.13396026194095612,
      "learning_rate": 0.0001316171034777466,
      "loss": 0.9786,
      "step": 1995
    },
    {
      "epoch": 0.4589787722317843,
      "grad_norm": 0.18823152780532837,
      "learning_rate": 0.00013123679010051232,
      "loss": 0.9294,
      "step": 2000
    },
    {
      "epoch": 0.46012621916236374,
      "grad_norm": 0.137470543384552,
      "learning_rate": 0.00013085597540690618,
      "loss": 0.9393,
      "step": 2005
    },
    {
      "epoch": 0.4612736660929432,
      "grad_norm": 0.1302119791507721,
      "learning_rate": 0.00013047466550858812,
      "loss": 0.9666,
      "step": 2010
    },
    {
      "epoch": 0.4624211130235227,
      "grad_norm": 0.13366107642650604,
      "learning_rate": 0.00013009286652516575,
      "loss": 0.9243,
      "step": 2015
    },
    {
      "epoch": 0.4635685599541021,
      "grad_norm": 0.14005546271800995,
      "learning_rate": 0.00012971058458409576,
      "loss": 0.9415,
      "step": 2020
    },
    {
      "epoch": 0.46471600688468157,
      "grad_norm": 0.1359815001487732,
      "learning_rate": 0.00012932782582058584,
      "loss": 0.9681,
      "step": 2025
    },
    {
      "epoch": 0.46586345381526106,
      "grad_norm": 0.12984395027160645,
      "learning_rate": 0.00012894459637749627,
      "loss": 0.9475,
      "step": 2030
    },
    {
      "epoch": 0.4670109007458405,
      "grad_norm": 0.12904886901378632,
      "learning_rate": 0.0001285609024052411,
      "loss": 0.9584,
      "step": 2035
    },
    {
      "epoch": 0.46815834767641995,
      "grad_norm": 0.13644804060459137,
      "learning_rate": 0.00012817675006168963,
      "loss": 0.9757,
      "step": 2040
    },
    {
      "epoch": 0.46930579460699945,
      "grad_norm": 0.1333729475736618,
      "learning_rate": 0.00012779214551206746,
      "loss": 0.9733,
      "step": 2045
    },
    {
      "epoch": 0.4704532415375789,
      "grad_norm": 0.128279909491539,
      "learning_rate": 0.0001274070949288577,
      "loss": 0.9772,
      "step": 2050
    },
    {
      "epoch": 0.47160068846815834,
      "grad_norm": 0.15798795223236084,
      "learning_rate": 0.00012702160449170165,
      "loss": 0.9499,
      "step": 2055
    },
    {
      "epoch": 0.47274813539873783,
      "grad_norm": 0.12967512011528015,
      "learning_rate": 0.0001266356803873,
      "loss": 0.9532,
      "step": 2060
    },
    {
      "epoch": 0.4738955823293173,
      "grad_norm": 0.1309663951396942,
      "learning_rate": 0.0001262493288093131,
      "loss": 0.9517,
      "step": 2065
    },
    {
      "epoch": 0.4750430292598967,
      "grad_norm": 0.12817169725894928,
      "learning_rate": 0.000125862555958262,
      "loss": 0.98,
      "step": 2070
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 0.1262216567993164,
      "learning_rate": 0.00012547536804142858,
      "loss": 0.9459,
      "step": 2075
    },
    {
      "epoch": 0.47733792312105566,
      "grad_norm": 0.13227002322673798,
      "learning_rate": 0.0001250877712727561,
      "loss": 0.9733,
      "step": 2080
    },
    {
      "epoch": 0.4784853700516351,
      "grad_norm": 0.13167516887187958,
      "learning_rate": 0.00012469977187274945,
      "loss": 0.9611,
      "step": 2085
    },
    {
      "epoch": 0.47963281698221455,
      "grad_norm": 0.12410388141870499,
      "learning_rate": 0.00012431137606837538,
      "loss": 0.9346,
      "step": 2090
    },
    {
      "epoch": 0.48078026391279405,
      "grad_norm": 0.13433882594108582,
      "learning_rate": 0.0001239225900929624,
      "loss": 0.9528,
      "step": 2095
    },
    {
      "epoch": 0.4819277108433735,
      "grad_norm": 0.13042834401130676,
      "learning_rate": 0.00012353342018610084,
      "loss": 0.958,
      "step": 2100
    },
    {
      "epoch": 0.48307515777395293,
      "grad_norm": 0.1367100179195404,
      "learning_rate": 0.00012314387259354282,
      "loss": 0.9665,
      "step": 2105
    },
    {
      "epoch": 0.48422260470453243,
      "grad_norm": 0.13082900643348694,
      "learning_rate": 0.00012275395356710177,
      "loss": 0.9706,
      "step": 2110
    },
    {
      "epoch": 0.4853700516351119,
      "grad_norm": 0.13011516630649567,
      "learning_rate": 0.0001223636693645523,
      "loss": 1.0014,
      "step": 2115
    },
    {
      "epoch": 0.4865174985656913,
      "grad_norm": 0.12624427676200867,
      "learning_rate": 0.00012197302624952971,
      "loss": 0.9607,
      "step": 2120
    },
    {
      "epoch": 0.4876649454962708,
      "grad_norm": 0.13805273175239563,
      "learning_rate": 0.00012158203049142947,
      "loss": 0.9245,
      "step": 2125
    },
    {
      "epoch": 0.48881239242685026,
      "grad_norm": 0.125767320394516,
      "learning_rate": 0.00012119068836530644,
      "loss": 0.9598,
      "step": 2130
    },
    {
      "epoch": 0.4899598393574297,
      "grad_norm": 0.12957511842250824,
      "learning_rate": 0.00012079900615177449,
      "loss": 0.9658,
      "step": 2135
    },
    {
      "epoch": 0.4911072862880092,
      "grad_norm": 0.13400690257549286,
      "learning_rate": 0.00012040699013690543,
      "loss": 0.9033,
      "step": 2140
    },
    {
      "epoch": 0.49225473321858865,
      "grad_norm": 0.1384160965681076,
      "learning_rate": 0.00012001464661212827,
      "loss": 0.9111,
      "step": 2145
    },
    {
      "epoch": 0.4934021801491681,
      "grad_norm": 0.13281968235969543,
      "learning_rate": 0.0001196219818741281,
      "loss": 0.9509,
      "step": 2150
    },
    {
      "epoch": 0.4945496270797476,
      "grad_norm": 0.12926004827022552,
      "learning_rate": 0.00011922900222474523,
      "loss": 0.9744,
      "step": 2155
    },
    {
      "epoch": 0.49569707401032703,
      "grad_norm": 0.13690362870693207,
      "learning_rate": 0.00011883571397087387,
      "loss": 0.9519,
      "step": 2160
    },
    {
      "epoch": 0.4968445209409065,
      "grad_norm": 0.13583113253116608,
      "learning_rate": 0.00011844212342436112,
      "loss": 0.9651,
      "step": 2165
    },
    {
      "epoch": 0.4979919678714859,
      "grad_norm": 0.13219110667705536,
      "learning_rate": 0.00011804823690190538,
      "loss": 0.9724,
      "step": 2170
    },
    {
      "epoch": 0.4991394148020654,
      "grad_norm": 0.13287144899368286,
      "learning_rate": 0.00011765406072495528,
      "loss": 0.95,
      "step": 2175
    },
    {
      "epoch": 0.5002868617326449,
      "grad_norm": 0.12971337139606476,
      "learning_rate": 0.00011725960121960806,
      "loss": 0.933,
      "step": 2180
    },
    {
      "epoch": 0.5014343086632244,
      "grad_norm": 0.13551461696624756,
      "learning_rate": 0.00011686486471650798,
      "loss": 0.9417,
      "step": 2185
    },
    {
      "epoch": 0.5025817555938038,
      "grad_norm": 0.12743504345417023,
      "learning_rate": 0.0001164698575507449,
      "loss": 0.9396,
      "step": 2190
    },
    {
      "epoch": 0.5037292025243832,
      "grad_norm": 0.13198213279247284,
      "learning_rate": 0.00011607458606175253,
      "loss": 0.9139,
      "step": 2195
    },
    {
      "epoch": 0.5048766494549627,
      "grad_norm": 0.13129118084907532,
      "learning_rate": 0.00011567905659320663,
      "loss": 0.94,
      "step": 2200
    },
    {
      "epoch": 0.5060240963855421,
      "grad_norm": 0.13837940990924835,
      "learning_rate": 0.00011528327549292326,
      "loss": 0.9357,
      "step": 2205
    },
    {
      "epoch": 0.5071715433161216,
      "grad_norm": 0.13046391308307648,
      "learning_rate": 0.00011488724911275694,
      "loss": 0.9488,
      "step": 2210
    },
    {
      "epoch": 0.5083189902467011,
      "grad_norm": 0.12679821252822876,
      "learning_rate": 0.00011449098380849858,
      "loss": 0.9627,
      "step": 2215
    },
    {
      "epoch": 0.5094664371772806,
      "grad_norm": 0.12639085948467255,
      "learning_rate": 0.00011409448593977363,
      "loss": 0.9246,
      "step": 2220
    },
    {
      "epoch": 0.51061388410786,
      "grad_norm": 0.13501042127609253,
      "learning_rate": 0.00011369776186993994,
      "loss": 0.9439,
      "step": 2225
    },
    {
      "epoch": 0.5117613310384395,
      "grad_norm": 0.12938356399536133,
      "learning_rate": 0.0001133008179659856,
      "loss": 0.9551,
      "step": 2230
    },
    {
      "epoch": 0.5129087779690189,
      "grad_norm": 0.13302691280841827,
      "learning_rate": 0.00011290366059842683,
      "loss": 0.9475,
      "step": 2235
    },
    {
      "epoch": 0.5140562248995983,
      "grad_norm": 0.13080324232578278,
      "learning_rate": 0.00011250629614120571,
      "loss": 0.9586,
      "step": 2240
    },
    {
      "epoch": 0.5152036718301779,
      "grad_norm": 0.12532858550548553,
      "learning_rate": 0.00011210873097158786,
      "loss": 0.9296,
      "step": 2245
    },
    {
      "epoch": 0.5163511187607573,
      "grad_norm": 0.13357555866241455,
      "learning_rate": 0.00011171097147006013,
      "loss": 0.9692,
      "step": 2250
    },
    {
      "epoch": 0.5174985656913368,
      "grad_norm": 0.13192251324653625,
      "learning_rate": 0.00011131302402022821,
      "loss": 0.9678,
      "step": 2255
    },
    {
      "epoch": 0.5186460126219162,
      "grad_norm": 0.12825413048267365,
      "learning_rate": 0.00011091489500871408,
      "loss": 0.9521,
      "step": 2260
    },
    {
      "epoch": 0.5197934595524957,
      "grad_norm": 0.1338176429271698,
      "learning_rate": 0.00011051659082505366,
      "loss": 0.9664,
      "step": 2265
    },
    {
      "epoch": 0.5209409064830751,
      "grad_norm": 0.13753551244735718,
      "learning_rate": 0.00011011811786159416,
      "loss": 0.9608,
      "step": 2270
    },
    {
      "epoch": 0.5220883534136547,
      "grad_norm": 0.1290527880191803,
      "learning_rate": 0.00010971948251339157,
      "loss": 0.9334,
      "step": 2275
    },
    {
      "epoch": 0.5232358003442341,
      "grad_norm": 0.13250134885311127,
      "learning_rate": 0.00010932069117810787,
      "loss": 0.9512,
      "step": 2280
    },
    {
      "epoch": 0.5243832472748136,
      "grad_norm": 0.13204516470432281,
      "learning_rate": 0.00010892175025590856,
      "loss": 0.9711,
      "step": 2285
    },
    {
      "epoch": 0.525530694205393,
      "grad_norm": 0.12426961213350296,
      "learning_rate": 0.00010852266614935982,
      "loss": 0.9634,
      "step": 2290
    },
    {
      "epoch": 0.5266781411359724,
      "grad_norm": 0.13476422429084778,
      "learning_rate": 0.00010812344526332578,
      "loss": 0.9345,
      "step": 2295
    },
    {
      "epoch": 0.5278255880665519,
      "grad_norm": 0.1305239349603653,
      "learning_rate": 0.00010772409400486571,
      "loss": 0.973,
      "step": 2300
    },
    {
      "epoch": 0.5289730349971313,
      "grad_norm": 0.12918445467948914,
      "learning_rate": 0.00010732461878313125,
      "loss": 0.9602,
      "step": 2305
    },
    {
      "epoch": 0.5301204819277109,
      "grad_norm": 0.13843993842601776,
      "learning_rate": 0.00010692502600926348,
      "loss": 0.9528,
      "step": 2310
    },
    {
      "epoch": 0.5312679288582903,
      "grad_norm": 0.1355164349079132,
      "learning_rate": 0.00010652532209629011,
      "loss": 0.9423,
      "step": 2315
    },
    {
      "epoch": 0.5324153757888698,
      "grad_norm": 0.1288328468799591,
      "learning_rate": 0.00010612551345902244,
      "loss": 0.9223,
      "step": 2320
    },
    {
      "epoch": 0.5335628227194492,
      "grad_norm": 0.13057994842529297,
      "learning_rate": 0.00010572560651395258,
      "loss": 0.9117,
      "step": 2325
    },
    {
      "epoch": 0.5347102696500287,
      "grad_norm": 0.12813064455986023,
      "learning_rate": 0.0001053256076791503,
      "loss": 0.953,
      "step": 2330
    },
    {
      "epoch": 0.5358577165806081,
      "grad_norm": 0.13439275324344635,
      "learning_rate": 0.00010492552337416007,
      "loss": 0.9552,
      "step": 2335
    },
    {
      "epoch": 0.5370051635111877,
      "grad_norm": 0.12726064026355743,
      "learning_rate": 0.00010452536001989815,
      "loss": 0.9556,
      "step": 2340
    },
    {
      "epoch": 0.5381526104417671,
      "grad_norm": 0.1329938769340515,
      "learning_rate": 0.00010412512403854942,
      "loss": 0.9806,
      "step": 2345
    },
    {
      "epoch": 0.5393000573723465,
      "grad_norm": 0.13323380053043365,
      "learning_rate": 0.00010372482185346435,
      "loss": 0.9542,
      "step": 2350
    },
    {
      "epoch": 0.540447504302926,
      "grad_norm": 0.12690098583698273,
      "learning_rate": 0.00010332445988905586,
      "loss": 0.9443,
      "step": 2355
    },
    {
      "epoch": 0.5415949512335054,
      "grad_norm": 0.12987449765205383,
      "learning_rate": 0.00010292404457069631,
      "loss": 0.9617,
      "step": 2360
    },
    {
      "epoch": 0.5427423981640849,
      "grad_norm": 0.12815946340560913,
      "learning_rate": 0.00010252358232461436,
      "loss": 0.9741,
      "step": 2365
    },
    {
      "epoch": 0.5438898450946644,
      "grad_norm": 0.13437722623348236,
      "learning_rate": 0.00010212307957779173,
      "loss": 0.9798,
      "step": 2370
    },
    {
      "epoch": 0.5450372920252439,
      "grad_norm": 0.14197693765163422,
      "learning_rate": 0.00010172254275786017,
      "loss": 0.9112,
      "step": 2375
    },
    {
      "epoch": 0.5461847389558233,
      "grad_norm": 0.12714983522891998,
      "learning_rate": 0.0001013219782929983,
      "loss": 0.9103,
      "step": 2380
    },
    {
      "epoch": 0.5473321858864028,
      "grad_norm": 0.13637425005435944,
      "learning_rate": 0.0001009213926118284,
      "loss": 0.9564,
      "step": 2385
    },
    {
      "epoch": 0.5484796328169822,
      "grad_norm": 0.13670361042022705,
      "learning_rate": 0.00010052079214331318,
      "loss": 0.9295,
      "step": 2390
    },
    {
      "epoch": 0.5496270797475616,
      "grad_norm": 0.12989582121372223,
      "learning_rate": 0.00010012018331665272,
      "loss": 0.9428,
      "step": 2395
    },
    {
      "epoch": 0.5507745266781411,
      "grad_norm": 0.12808671593666077,
      "learning_rate": 9.971957256118129e-05,
      "loss": 0.9216,
      "step": 2400
    },
    {
      "epoch": 0.5519219736087206,
      "grad_norm": 0.1358453780412674,
      "learning_rate": 9.931896630626402e-05,
      "loss": 0.9711,
      "step": 2405
    },
    {
      "epoch": 0.5530694205393001,
      "grad_norm": 0.1230422705411911,
      "learning_rate": 9.891837098119389e-05,
      "loss": 0.9282,
      "step": 2410
    },
    {
      "epoch": 0.5542168674698795,
      "grad_norm": 0.13326068222522736,
      "learning_rate": 9.851779301508842e-05,
      "loss": 0.9906,
      "step": 2415
    },
    {
      "epoch": 0.555364314400459,
      "grad_norm": 0.12692315876483917,
      "learning_rate": 9.811723883678654e-05,
      "loss": 0.9537,
      "step": 2420
    },
    {
      "epoch": 0.5565117613310384,
      "grad_norm": 0.13576091825962067,
      "learning_rate": 9.771671487474546e-05,
      "loss": 0.9393,
      "step": 2425
    },
    {
      "epoch": 0.5576592082616179,
      "grad_norm": 0.12827935814857483,
      "learning_rate": 9.731622755693737e-05,
      "loss": 0.9014,
      "step": 2430
    },
    {
      "epoch": 0.5588066551921974,
      "grad_norm": 0.1317823976278305,
      "learning_rate": 9.691578331074643e-05,
      "loss": 0.9424,
      "step": 2435
    },
    {
      "epoch": 0.5599541021227769,
      "grad_norm": 0.13604985177516937,
      "learning_rate": 9.651538856286551e-05,
      "loss": 0.9395,
      "step": 2440
    },
    {
      "epoch": 0.5611015490533563,
      "grad_norm": 0.1410413533449173,
      "learning_rate": 9.611504973919311e-05,
      "loss": 0.9711,
      "step": 2445
    },
    {
      "epoch": 0.5622489959839357,
      "grad_norm": 0.1319737732410431,
      "learning_rate": 9.571477326473021e-05,
      "loss": 0.944,
      "step": 2450
    },
    {
      "epoch": 0.5633964429145152,
      "grad_norm": 0.1291595995426178,
      "learning_rate": 9.531456556347712e-05,
      "loss": 0.9621,
      "step": 2455
    },
    {
      "epoch": 0.5645438898450946,
      "grad_norm": 0.12568049132823944,
      "learning_rate": 9.491443305833043e-05,
      "loss": 0.923,
      "step": 2460
    },
    {
      "epoch": 0.5656913367756741,
      "grad_norm": 0.13689230382442474,
      "learning_rate": 9.451438217097994e-05,
      "loss": 0.9335,
      "step": 2465
    },
    {
      "epoch": 0.5668387837062536,
      "grad_norm": 0.12956008315086365,
      "learning_rate": 9.411441932180554e-05,
      "loss": 0.9757,
      "step": 2470
    },
    {
      "epoch": 0.5679862306368331,
      "grad_norm": 0.13027647137641907,
      "learning_rate": 9.371455092977423e-05,
      "loss": 0.9225,
      "step": 2475
    },
    {
      "epoch": 0.5691336775674125,
      "grad_norm": 0.12894025444984436,
      "learning_rate": 9.331478341233706e-05,
      "loss": 0.9508,
      "step": 2480
    },
    {
      "epoch": 0.570281124497992,
      "grad_norm": 0.13095425069332123,
      "learning_rate": 9.291512318532614e-05,
      "loss": 0.9388,
      "step": 2485
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.1269254833459854,
      "learning_rate": 9.251557666285174e-05,
      "loss": 0.9682,
      "step": 2490
    },
    {
      "epoch": 0.5725760183591508,
      "grad_norm": 0.1342654824256897,
      "learning_rate": 9.211615025719919e-05,
      "loss": 0.9229,
      "step": 2495
    },
    {
      "epoch": 0.5737234652897304,
      "grad_norm": 0.1338120847940445,
      "learning_rate": 9.17168503787262e-05,
      "loss": 0.9593,
      "step": 2500
    },
    {
      "epoch": 0.5748709122203098,
      "grad_norm": 0.1292128562927246,
      "learning_rate": 9.131768343575979e-05,
      "loss": 0.9353,
      "step": 2505
    },
    {
      "epoch": 0.5760183591508893,
      "grad_norm": 0.12859606742858887,
      "learning_rate": 9.091865583449351e-05,
      "loss": 0.9279,
      "step": 2510
    },
    {
      "epoch": 0.5771658060814687,
      "grad_norm": 0.12240682542324066,
      "learning_rate": 9.051977397888464e-05,
      "loss": 0.9242,
      "step": 2515
    },
    {
      "epoch": 0.5783132530120482,
      "grad_norm": 0.13055679202079773,
      "learning_rate": 9.01210442705514e-05,
      "loss": 0.9374,
      "step": 2520
    },
    {
      "epoch": 0.5794606999426276,
      "grad_norm": 0.1301288604736328,
      "learning_rate": 8.972247310867027e-05,
      "loss": 0.9403,
      "step": 2525
    },
    {
      "epoch": 0.5806081468732072,
      "grad_norm": 0.1294988989830017,
      "learning_rate": 8.932406688987309e-05,
      "loss": 0.9511,
      "step": 2530
    },
    {
      "epoch": 0.5817555938037866,
      "grad_norm": 0.1301121711730957,
      "learning_rate": 8.892583200814466e-05,
      "loss": 0.9553,
      "step": 2535
    },
    {
      "epoch": 0.582903040734366,
      "grad_norm": 0.12820468842983246,
      "learning_rate": 8.852777485471997e-05,
      "loss": 0.938,
      "step": 2540
    },
    {
      "epoch": 0.5840504876649455,
      "grad_norm": 0.12704700231552124,
      "learning_rate": 8.81299018179817e-05,
      "loss": 0.9188,
      "step": 2545
    },
    {
      "epoch": 0.5851979345955249,
      "grad_norm": 0.12919333577156067,
      "learning_rate": 8.773221928335759e-05,
      "loss": 0.9601,
      "step": 2550
    },
    {
      "epoch": 0.5863453815261044,
      "grad_norm": 0.13542434573173523,
      "learning_rate": 8.73347336332181e-05,
      "loss": 0.9476,
      "step": 2555
    },
    {
      "epoch": 0.5874928284566838,
      "grad_norm": 0.12929943203926086,
      "learning_rate": 8.693745124677386e-05,
      "loss": 0.9397,
      "step": 2560
    },
    {
      "epoch": 0.5886402753872634,
      "grad_norm": 0.130234032869339,
      "learning_rate": 8.654037849997342e-05,
      "loss": 0.9313,
      "step": 2565
    },
    {
      "epoch": 0.5897877223178428,
      "grad_norm": 0.13515686988830566,
      "learning_rate": 8.614352176540067e-05,
      "loss": 0.9503,
      "step": 2570
    },
    {
      "epoch": 0.5909351692484223,
      "grad_norm": 0.1329420804977417,
      "learning_rate": 8.57468874121729e-05,
      "loss": 0.9606,
      "step": 2575
    },
    {
      "epoch": 0.5920826161790017,
      "grad_norm": 0.1323956400156021,
      "learning_rate": 8.535048180583838e-05,
      "loss": 0.9512,
      "step": 2580
    },
    {
      "epoch": 0.5932300631095812,
      "grad_norm": 0.13518303632736206,
      "learning_rate": 8.495431130827422e-05,
      "loss": 0.9377,
      "step": 2585
    },
    {
      "epoch": 0.5943775100401606,
      "grad_norm": 0.13215835392475128,
      "learning_rate": 8.455838227758432e-05,
      "loss": 0.9984,
      "step": 2590
    },
    {
      "epoch": 0.5955249569707401,
      "grad_norm": 0.13005711138248444,
      "learning_rate": 8.416270106799726e-05,
      "loss": 0.9556,
      "step": 2595
    },
    {
      "epoch": 0.5966724039013196,
      "grad_norm": 0.1329335868358612,
      "learning_rate": 8.376727402976447e-05,
      "loss": 0.9351,
      "step": 2600
    },
    {
      "epoch": 0.597819850831899,
      "grad_norm": 0.1298844963312149,
      "learning_rate": 8.3372107509058e-05,
      "loss": 0.9492,
      "step": 2605
    },
    {
      "epoch": 0.5989672977624785,
      "grad_norm": 0.13211120665073395,
      "learning_rate": 8.297720784786906e-05,
      "loss": 0.9495,
      "step": 2610
    },
    {
      "epoch": 0.6001147446930579,
      "grad_norm": 0.12725351750850677,
      "learning_rate": 8.2582581383906e-05,
      "loss": 0.953,
      "step": 2615
    },
    {
      "epoch": 0.6012621916236374,
      "grad_norm": 0.12866829335689545,
      "learning_rate": 8.218823445049265e-05,
      "loss": 0.9536,
      "step": 2620
    },
    {
      "epoch": 0.6024096385542169,
      "grad_norm": 0.1358039677143097,
      "learning_rate": 8.179417337646669e-05,
      "loss": 0.9731,
      "step": 2625
    },
    {
      "epoch": 0.6035570854847964,
      "grad_norm": 0.13068200647830963,
      "learning_rate": 8.140040448607804e-05,
      "loss": 0.9592,
      "step": 2630
    },
    {
      "epoch": 0.6047045324153758,
      "grad_norm": 0.1265789270401001,
      "learning_rate": 8.100693409888748e-05,
      "loss": 0.97,
      "step": 2635
    },
    {
      "epoch": 0.6058519793459552,
      "grad_norm": 0.13472406566143036,
      "learning_rate": 8.061376852966495e-05,
      "loss": 0.9648,
      "step": 2640
    },
    {
      "epoch": 0.6069994262765347,
      "grad_norm": 0.13039621710777283,
      "learning_rate": 8.02209140882886e-05,
      "loss": 0.926,
      "step": 2645
    },
    {
      "epoch": 0.6081468732071141,
      "grad_norm": 0.1332240104675293,
      "learning_rate": 7.982837707964321e-05,
      "loss": 0.9637,
      "step": 2650
    },
    {
      "epoch": 0.6092943201376936,
      "grad_norm": 0.1377446949481964,
      "learning_rate": 7.943616380351913e-05,
      "loss": 0.9556,
      "step": 2655
    },
    {
      "epoch": 0.6104417670682731,
      "grad_norm": 0.12704016268253326,
      "learning_rate": 7.904428055451118e-05,
      "loss": 0.9323,
      "step": 2660
    },
    {
      "epoch": 0.6115892139988526,
      "grad_norm": 0.12996916472911835,
      "learning_rate": 7.865273362191759e-05,
      "loss": 0.9376,
      "step": 2665
    },
    {
      "epoch": 0.612736660929432,
      "grad_norm": 0.13720089197158813,
      "learning_rate": 7.826152928963904e-05,
      "loss": 0.9469,
      "step": 2670
    },
    {
      "epoch": 0.6138841078600115,
      "grad_norm": 0.13306473195552826,
      "learning_rate": 7.787067383607796e-05,
      "loss": 0.9643,
      "step": 2675
    },
    {
      "epoch": 0.6150315547905909,
      "grad_norm": 0.1308506578207016,
      "learning_rate": 7.748017353403748e-05,
      "loss": 0.9476,
      "step": 2680
    },
    {
      "epoch": 0.6161790017211703,
      "grad_norm": 0.12719397246837616,
      "learning_rate": 7.70900346506211e-05,
      "loss": 0.949,
      "step": 2685
    },
    {
      "epoch": 0.6173264486517499,
      "grad_norm": 0.12891539931297302,
      "learning_rate": 7.670026344713189e-05,
      "loss": 0.9486,
      "step": 2690
    },
    {
      "epoch": 0.6184738955823293,
      "grad_norm": 0.12883388996124268,
      "learning_rate": 7.631086617897203e-05,
      "loss": 0.9422,
      "step": 2695
    },
    {
      "epoch": 0.6196213425129088,
      "grad_norm": 0.128887340426445,
      "learning_rate": 7.592184909554245e-05,
      "loss": 0.9712,
      "step": 2700
    },
    {
      "epoch": 0.6207687894434882,
      "grad_norm": 0.12608595192432404,
      "learning_rate": 7.553321844014258e-05,
      "loss": 0.9674,
      "step": 2705
    },
    {
      "epoch": 0.6219162363740677,
      "grad_norm": 0.12851373851299286,
      "learning_rate": 7.514498044987009e-05,
      "loss": 0.9229,
      "step": 2710
    },
    {
      "epoch": 0.6230636833046471,
      "grad_norm": 0.1275022029876709,
      "learning_rate": 7.475714135552074e-05,
      "loss": 0.9108,
      "step": 2715
    },
    {
      "epoch": 0.6242111302352267,
      "grad_norm": 0.13182856142520905,
      "learning_rate": 7.43697073814885e-05,
      "loss": 0.9536,
      "step": 2720
    },
    {
      "epoch": 0.6253585771658061,
      "grad_norm": 0.12938585877418518,
      "learning_rate": 7.39826847456656e-05,
      "loss": 0.9425,
      "step": 2725
    },
    {
      "epoch": 0.6265060240963856,
      "grad_norm": 0.13789427280426025,
      "learning_rate": 7.359607965934274e-05,
      "loss": 0.9692,
      "step": 2730
    },
    {
      "epoch": 0.627653471026965,
      "grad_norm": 0.13062715530395508,
      "learning_rate": 7.32098983271094e-05,
      "loss": 0.9383,
      "step": 2735
    },
    {
      "epoch": 0.6288009179575444,
      "grad_norm": 0.13577692210674286,
      "learning_rate": 7.282414694675426e-05,
      "loss": 0.9281,
      "step": 2740
    },
    {
      "epoch": 0.6299483648881239,
      "grad_norm": 0.13397711515426636,
      "learning_rate": 7.243883170916574e-05,
      "loss": 0.9881,
      "step": 2745
    },
    {
      "epoch": 0.6310958118187033,
      "grad_norm": 0.13154800236225128,
      "learning_rate": 7.205395879823271e-05,
      "loss": 0.9107,
      "step": 2750
    },
    {
      "epoch": 0.6322432587492829,
      "grad_norm": 0.12707136571407318,
      "learning_rate": 7.166953439074504e-05,
      "loss": 0.9728,
      "step": 2755
    },
    {
      "epoch": 0.6333907056798623,
      "grad_norm": 0.13293033838272095,
      "learning_rate": 7.128556465629475e-05,
      "loss": 0.9652,
      "step": 2760
    },
    {
      "epoch": 0.6345381526104418,
      "grad_norm": 0.1299448162317276,
      "learning_rate": 7.090205575717678e-05,
      "loss": 0.9455,
      "step": 2765
    },
    {
      "epoch": 0.6356855995410212,
      "grad_norm": 0.12795931100845337,
      "learning_rate": 7.051901384829021e-05,
      "loss": 0.9289,
      "step": 2770
    },
    {
      "epoch": 0.6368330464716007,
      "grad_norm": 0.13310518860816956,
      "learning_rate": 7.013644507703937e-05,
      "loss": 0.9203,
      "step": 2775
    },
    {
      "epoch": 0.6379804934021801,
      "grad_norm": 0.1404452919960022,
      "learning_rate": 6.975435558323532e-05,
      "loss": 0.9364,
      "step": 2780
    },
    {
      "epoch": 0.6391279403327597,
      "grad_norm": 0.1289965659379959,
      "learning_rate": 6.937275149899725e-05,
      "loss": 0.9313,
      "step": 2785
    },
    {
      "epoch": 0.6402753872633391,
      "grad_norm": 0.12998373806476593,
      "learning_rate": 6.899163894865395e-05,
      "loss": 0.9575,
      "step": 2790
    },
    {
      "epoch": 0.6414228341939185,
      "grad_norm": 0.13112759590148926,
      "learning_rate": 6.86110240486457e-05,
      "loss": 0.9648,
      "step": 2795
    },
    {
      "epoch": 0.642570281124498,
      "grad_norm": 0.12994885444641113,
      "learning_rate": 6.823091290742602e-05,
      "loss": 0.961,
      "step": 2800
    },
    {
      "epoch": 0.6437177280550774,
      "grad_norm": 0.1322287768125534,
      "learning_rate": 6.785131162536374e-05,
      "loss": 0.9878,
      "step": 2805
    },
    {
      "epoch": 0.6448651749856569,
      "grad_norm": 0.12845389544963837,
      "learning_rate": 6.747222629464484e-05,
      "loss": 0.9455,
      "step": 2810
    },
    {
      "epoch": 0.6460126219162364,
      "grad_norm": 0.13705497980117798,
      "learning_rate": 6.709366299917497e-05,
      "loss": 0.9537,
      "step": 2815
    },
    {
      "epoch": 0.6471600688468159,
      "grad_norm": 0.1286800056695938,
      "learning_rate": 6.671562781448166e-05,
      "loss": 0.9477,
      "step": 2820
    },
    {
      "epoch": 0.6483075157773953,
      "grad_norm": 0.13222958147525787,
      "learning_rate": 6.633812680761684e-05,
      "loss": 0.9348,
      "step": 2825
    },
    {
      "epoch": 0.6494549627079748,
      "grad_norm": 0.12600712478160858,
      "learning_rate": 6.59611660370594e-05,
      "loss": 0.9627,
      "step": 2830
    },
    {
      "epoch": 0.6506024096385542,
      "grad_norm": 0.130609929561615,
      "learning_rate": 6.558475155261811e-05,
      "loss": 0.9506,
      "step": 2835
    },
    {
      "epoch": 0.6517498565691336,
      "grad_norm": 0.1283220797777176,
      "learning_rate": 6.52088893953344e-05,
      "loss": 0.9468,
      "step": 2840
    },
    {
      "epoch": 0.6528973034997131,
      "grad_norm": 0.13281899690628052,
      "learning_rate": 6.48335855973855e-05,
      "loss": 0.95,
      "step": 2845
    },
    {
      "epoch": 0.6540447504302926,
      "grad_norm": 0.12522880733013153,
      "learning_rate": 6.445884618198754e-05,
      "loss": 0.9691,
      "step": 2850
    },
    {
      "epoch": 0.6551921973608721,
      "grad_norm": 0.13211198151111603,
      "learning_rate": 6.408467716329894e-05,
      "loss": 0.9636,
      "step": 2855
    },
    {
      "epoch": 0.6563396442914515,
      "grad_norm": 0.13233083486557007,
      "learning_rate": 6.371108454632391e-05,
      "loss": 0.9484,
      "step": 2860
    },
    {
      "epoch": 0.657487091222031,
      "grad_norm": 0.13322043418884277,
      "learning_rate": 6.33380743268159e-05,
      "loss": 0.925,
      "step": 2865
    },
    {
      "epoch": 0.6586345381526104,
      "grad_norm": 0.1293569952249527,
      "learning_rate": 6.29656524911817e-05,
      "loss": 0.9651,
      "step": 2870
    },
    {
      "epoch": 0.6597819850831899,
      "grad_norm": 0.15199431777000427,
      "learning_rate": 6.259382501638509e-05,
      "loss": 0.9831,
      "step": 2875
    },
    {
      "epoch": 0.6609294320137694,
      "grad_norm": 0.1290358006954193,
      "learning_rate": 6.222259786985101e-05,
      "loss": 0.9603,
      "step": 2880
    },
    {
      "epoch": 0.6620768789443489,
      "grad_norm": 0.1347620040178299,
      "learning_rate": 6.185197700936982e-05,
      "loss": 0.9258,
      "step": 2885
    },
    {
      "epoch": 0.6632243258749283,
      "grad_norm": 0.13369546830654144,
      "learning_rate": 6.14819683830016e-05,
      "loss": 0.9471,
      "step": 2890
    },
    {
      "epoch": 0.6643717728055077,
      "grad_norm": 0.1245836541056633,
      "learning_rate": 6.111257792898082e-05,
      "loss": 0.9626,
      "step": 2895
    },
    {
      "epoch": 0.6655192197360872,
      "grad_norm": 0.1284221112728119,
      "learning_rate": 6.0743811575620846e-05,
      "loss": 0.9219,
      "step": 2900
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.1400459110736847,
      "learning_rate": 6.0375675241219e-05,
      "loss": 0.9495,
      "step": 2905
    },
    {
      "epoch": 0.6678141135972461,
      "grad_norm": 0.13489268720149994,
      "learning_rate": 6.000817483396148e-05,
      "loss": 0.9763,
      "step": 2910
    },
    {
      "epoch": 0.6689615605278256,
      "grad_norm": 0.13535645604133606,
      "learning_rate": 5.96413162518285e-05,
      "loss": 0.9263,
      "step": 2915
    },
    {
      "epoch": 0.6701090074584051,
      "grad_norm": 0.1278652399778366,
      "learning_rate": 5.9275105382499694e-05,
      "loss": 0.9649,
      "step": 2920
    },
    {
      "epoch": 0.6712564543889845,
      "grad_norm": 0.13380366563796997,
      "learning_rate": 5.890954810325966e-05,
      "loss": 0.9073,
      "step": 2925
    },
    {
      "epoch": 0.672403901319564,
      "grad_norm": 0.12865275144577026,
      "learning_rate": 5.854465028090355e-05,
      "loss": 0.9642,
      "step": 2930
    },
    {
      "epoch": 0.6735513482501434,
      "grad_norm": 0.13362818956375122,
      "learning_rate": 5.8180417771643006e-05,
      "loss": 0.9398,
      "step": 2935
    },
    {
      "epoch": 0.6746987951807228,
      "grad_norm": 0.1288289725780487,
      "learning_rate": 5.781685642101196e-05,
      "loss": 0.9917,
      "step": 2940
    },
    {
      "epoch": 0.6758462421113024,
      "grad_norm": 0.1349886655807495,
      "learning_rate": 5.7453972063773184e-05,
      "loss": 0.9158,
      "step": 2945
    },
    {
      "epoch": 0.6769936890418818,
      "grad_norm": 0.12944377958774567,
      "learning_rate": 5.7091770523824317e-05,
      "loss": 0.9787,
      "step": 2950
    },
    {
      "epoch": 0.6781411359724613,
      "grad_norm": 0.1282043606042862,
      "learning_rate": 5.673025761410462e-05,
      "loss": 0.9396,
      "step": 2955
    },
    {
      "epoch": 0.6792885829030407,
      "grad_norm": 0.12658511102199554,
      "learning_rate": 5.636943913650147e-05,
      "loss": 0.9507,
      "step": 2960
    },
    {
      "epoch": 0.6804360298336202,
      "grad_norm": 0.13132619857788086,
      "learning_rate": 5.60093208817575e-05,
      "loss": 0.9011,
      "step": 2965
    },
    {
      "epoch": 0.6815834767641996,
      "grad_norm": 0.12832701206207275,
      "learning_rate": 5.564990862937744e-05,
      "loss": 0.9602,
      "step": 2970
    },
    {
      "epoch": 0.6827309236947792,
      "grad_norm": 0.13031287491321564,
      "learning_rate": 5.5291208147535466e-05,
      "loss": 0.934,
      "step": 2975
    },
    {
      "epoch": 0.6838783706253586,
      "grad_norm": 0.1344255656003952,
      "learning_rate": 5.4933225192982586e-05,
      "loss": 0.9709,
      "step": 2980
    },
    {
      "epoch": 0.685025817555938,
      "grad_norm": 0.1263609677553177,
      "learning_rate": 5.457596551095441e-05,
      "loss": 0.9742,
      "step": 2985
    },
    {
      "epoch": 0.6861732644865175,
      "grad_norm": 0.13262014091014862,
      "learning_rate": 5.421943483507863e-05,
      "loss": 0.9751,
      "step": 2990
    },
    {
      "epoch": 0.6873207114170969,
      "grad_norm": 0.12867961823940277,
      "learning_rate": 5.3863638887283364e-05,
      "loss": 0.9876,
      "step": 2995
    },
    {
      "epoch": 0.6884681583476764,
      "grad_norm": 0.12955667078495026,
      "learning_rate": 5.350858337770498e-05,
      "loss": 0.9609,
      "step": 3000
    },
    {
      "epoch": 0.6896156052782558,
      "grad_norm": 0.13022245466709137,
      "learning_rate": 5.315427400459678e-05,
      "loss": 0.9605,
      "step": 3005
    },
    {
      "epoch": 0.6907630522088354,
      "grad_norm": 0.1346377581357956,
      "learning_rate": 5.280071645423726e-05,
      "loss": 0.9811,
      "step": 3010
    },
    {
      "epoch": 0.6919104991394148,
      "grad_norm": 0.1265023946762085,
      "learning_rate": 5.244791640083906e-05,
      "loss": 0.9523,
      "step": 3015
    },
    {
      "epoch": 0.6930579460699943,
      "grad_norm": 0.12777413427829742,
      "learning_rate": 5.2095879506457736e-05,
      "loss": 0.9318,
      "step": 3020
    },
    {
      "epoch": 0.6942053930005737,
      "grad_norm": 0.12877587974071503,
      "learning_rate": 5.174461142090111e-05,
      "loss": 0.911,
      "step": 3025
    },
    {
      "epoch": 0.6953528399311532,
      "grad_norm": 0.1300651729106903,
      "learning_rate": 5.139411778163832e-05,
      "loss": 0.8949,
      "step": 3030
    },
    {
      "epoch": 0.6965002868617326,
      "grad_norm": 0.13519862294197083,
      "learning_rate": 5.104440421370962e-05,
      "loss": 0.9456,
      "step": 3035
    },
    {
      "epoch": 0.6976477337923122,
      "grad_norm": 0.12986309826374054,
      "learning_rate": 5.0695476329635825e-05,
      "loss": 0.9451,
      "step": 3040
    },
    {
      "epoch": 0.6987951807228916,
      "grad_norm": 0.12880775332450867,
      "learning_rate": 5.034733972932855e-05,
      "loss": 0.9474,
      "step": 3045
    },
    {
      "epoch": 0.699942627653471,
      "grad_norm": 0.13083116710186005,
      "learning_rate": 5.000000000000002e-05,
      "loss": 0.9568,
      "step": 3050
    },
    {
      "epoch": 0.7010900745840505,
      "grad_norm": 0.12505587935447693,
      "learning_rate": 4.9653462716073594e-05,
      "loss": 0.9103,
      "step": 3055
    },
    {
      "epoch": 0.7022375215146299,
      "grad_norm": 0.12554022669792175,
      "learning_rate": 4.930773343909434e-05,
      "loss": 0.9495,
      "step": 3060
    },
    {
      "epoch": 0.7033849684452094,
      "grad_norm": 0.12707488238811493,
      "learning_rate": 4.8962817717639555e-05,
      "loss": 0.9487,
      "step": 3065
    },
    {
      "epoch": 0.7045324153757889,
      "grad_norm": 0.127935528755188,
      "learning_rate": 4.8618721087230014e-05,
      "loss": 0.9821,
      "step": 3070
    },
    {
      "epoch": 0.7056798623063684,
      "grad_norm": 0.132330060005188,
      "learning_rate": 4.8275449070240854e-05,
      "loss": 0.9869,
      "step": 3075
    },
    {
      "epoch": 0.7068273092369478,
      "grad_norm": 0.1311495006084442,
      "learning_rate": 4.793300717581308e-05,
      "loss": 0.9842,
      "step": 3080
    },
    {
      "epoch": 0.7079747561675273,
      "grad_norm": 0.13313139975070953,
      "learning_rate": 4.7591400899765234e-05,
      "loss": 0.9559,
      "step": 3085
    },
    {
      "epoch": 0.7091222030981067,
      "grad_norm": 0.13822641968727112,
      "learning_rate": 4.7250635724505e-05,
      "loss": 0.9324,
      "step": 3090
    },
    {
      "epoch": 0.7102696500286861,
      "grad_norm": 0.13737522065639496,
      "learning_rate": 4.6910717118941286e-05,
      "loss": 0.9403,
      "step": 3095
    },
    {
      "epoch": 0.7114170969592656,
      "grad_norm": 0.12806421518325806,
      "learning_rate": 4.6571650538396615e-05,
      "loss": 0.9148,
      "step": 3100
    },
    {
      "epoch": 0.7125645438898451,
      "grad_norm": 0.13433298468589783,
      "learning_rate": 4.6233441424519295e-05,
      "loss": 0.9628,
      "step": 3105
    },
    {
      "epoch": 0.7137119908204246,
      "grad_norm": 0.13459208607673645,
      "learning_rate": 4.5896095205196356e-05,
      "loss": 0.9604,
      "step": 3110
    },
    {
      "epoch": 0.714859437751004,
      "grad_norm": 0.14309802651405334,
      "learning_rate": 4.5559617294466176e-05,
      "loss": 0.943,
      "step": 3115
    },
    {
      "epoch": 0.7160068846815835,
      "grad_norm": 0.1299637407064438,
      "learning_rate": 4.522401309243193e-05,
      "loss": 0.9355,
      "step": 3120
    },
    {
      "epoch": 0.7171543316121629,
      "grad_norm": 0.12573669850826263,
      "learning_rate": 4.488928798517442e-05,
      "loss": 0.9422,
      "step": 3125
    },
    {
      "epoch": 0.7183017785427424,
      "grad_norm": 0.13579024374485016,
      "learning_rate": 4.455544734466624e-05,
      "loss": 0.9349,
      "step": 3130
    },
    {
      "epoch": 0.7194492254733219,
      "grad_norm": 0.12755471467971802,
      "learning_rate": 4.422249652868506e-05,
      "loss": 0.9367,
      "step": 3135
    },
    {
      "epoch": 0.7205966724039014,
      "grad_norm": 0.12943492829799652,
      "learning_rate": 4.389044088072798e-05,
      "loss": 0.9717,
      "step": 3140
    },
    {
      "epoch": 0.7217441193344808,
      "grad_norm": 0.13614961504936218,
      "learning_rate": 4.355928572992547e-05,
      "loss": 0.9643,
      "step": 3145
    },
    {
      "epoch": 0.7228915662650602,
      "grad_norm": 0.12964297831058502,
      "learning_rate": 4.322903639095619e-05,
      "loss": 0.9397,
      "step": 3150
    },
    {
      "epoch": 0.7240390131956397,
      "grad_norm": 0.12502126395702362,
      "learning_rate": 4.289969816396132e-05,
      "loss": 0.9278,
      "step": 3155
    },
    {
      "epoch": 0.7251864601262191,
      "grad_norm": 0.12723030149936676,
      "learning_rate": 4.2571276334459895e-05,
      "loss": 0.9721,
      "step": 3160
    },
    {
      "epoch": 0.7263339070567987,
      "grad_norm": 0.12903666496276855,
      "learning_rate": 4.224377617326353e-05,
      "loss": 0.9325,
      "step": 3165
    },
    {
      "epoch": 0.7274813539873781,
      "grad_norm": 0.12750981748104095,
      "learning_rate": 4.1917202936392265e-05,
      "loss": 0.965,
      "step": 3170
    },
    {
      "epoch": 0.7286288009179576,
      "grad_norm": 0.1284424364566803,
      "learning_rate": 4.15915618649899e-05,
      "loss": 0.9264,
      "step": 3175
    },
    {
      "epoch": 0.729776247848537,
      "grad_norm": 0.12181597203016281,
      "learning_rate": 4.126685818524013e-05,
      "loss": 0.9283,
      "step": 3180
    },
    {
      "epoch": 0.7309236947791165,
      "grad_norm": 0.13070735335350037,
      "learning_rate": 4.094309710828236e-05,
      "loss": 0.9313,
      "step": 3185
    },
    {
      "epoch": 0.7320711417096959,
      "grad_norm": 0.12752236425876617,
      "learning_rate": 4.0620283830128414e-05,
      "loss": 0.9366,
      "step": 3190
    },
    {
      "epoch": 0.7332185886402753,
      "grad_norm": 0.2004683017730713,
      "learning_rate": 4.029842353157888e-05,
      "loss": 0.9166,
      "step": 3195
    },
    {
      "epoch": 0.7343660355708549,
      "grad_norm": 0.12685176730155945,
      "learning_rate": 3.9977521378140084e-05,
      "loss": 0.9164,
      "step": 3200
    },
    {
      "epoch": 0.7355134825014343,
      "grad_norm": 0.1288214921951294,
      "learning_rate": 3.965758251994115e-05,
      "loss": 0.9292,
      "step": 3205
    },
    {
      "epoch": 0.7366609294320138,
      "grad_norm": 0.13338162004947662,
      "learning_rate": 3.933861209165146e-05,
      "loss": 0.9872,
      "step": 3210
    },
    {
      "epoch": 0.7378083763625932,
      "grad_norm": 0.12942497432231903,
      "learning_rate": 3.9020615212398016e-05,
      "loss": 0.9797,
      "step": 3215
    },
    {
      "epoch": 0.7389558232931727,
      "grad_norm": 0.13232511281967163,
      "learning_rate": 3.8703596985683556e-05,
      "loss": 0.9463,
      "step": 3220
    },
    {
      "epoch": 0.7401032702237521,
      "grad_norm": 0.13070105016231537,
      "learning_rate": 3.838756249930439e-05,
      "loss": 0.9215,
      "step": 3225
    },
    {
      "epoch": 0.7412507171543317,
      "grad_norm": 0.12770217657089233,
      "learning_rate": 3.807251682526902e-05,
      "loss": 0.9534,
      "step": 3230
    },
    {
      "epoch": 0.7423981640849111,
      "grad_norm": 0.1268807053565979,
      "learning_rate": 3.775846501971636e-05,
      "loss": 0.9179,
      "step": 3235
    },
    {
      "epoch": 0.7435456110154905,
      "grad_norm": 0.12574085593223572,
      "learning_rate": 3.7445412122835077e-05,
      "loss": 0.9214,
      "step": 3240
    },
    {
      "epoch": 0.74469305794607,
      "grad_norm": 0.13617047667503357,
      "learning_rate": 3.713336315878224e-05,
      "loss": 0.9603,
      "step": 3245
    },
    {
      "epoch": 0.7458405048766494,
      "grad_norm": 0.1332123875617981,
      "learning_rate": 3.6822323135603054e-05,
      "loss": 0.9363,
      "step": 3250
    },
    {
      "epoch": 0.7469879518072289,
      "grad_norm": 0.13324113190174103,
      "learning_rate": 3.651229704515018e-05,
      "loss": 0.9128,
      "step": 3255
    },
    {
      "epoch": 0.7481353987378083,
      "grad_norm": 0.1259755641222,
      "learning_rate": 3.6203289863003905e-05,
      "loss": 0.9431,
      "step": 3260
    },
    {
      "epoch": 0.7492828456683879,
      "grad_norm": 0.12775766849517822,
      "learning_rate": 3.5895306548392005e-05,
      "loss": 0.9707,
      "step": 3265
    },
    {
      "epoch": 0.7504302925989673,
      "grad_norm": 0.130545511841774,
      "learning_rate": 3.558835204411044e-05,
      "loss": 0.946,
      "step": 3270
    },
    {
      "epoch": 0.7515777395295468,
      "grad_norm": 0.13085825741291046,
      "learning_rate": 3.52824312764438e-05,
      "loss": 0.9029,
      "step": 3275
    },
    {
      "epoch": 0.7527251864601262,
      "grad_norm": 0.12603677809238434,
      "learning_rate": 3.497754915508632e-05,
      "loss": 0.9256,
      "step": 3280
    },
    {
      "epoch": 0.7538726333907056,
      "grad_norm": 0.12976112961769104,
      "learning_rate": 3.467371057306318e-05,
      "loss": 0.9324,
      "step": 3285
    },
    {
      "epoch": 0.7550200803212851,
      "grad_norm": 0.13229593634605408,
      "learning_rate": 3.437092040665183e-05,
      "loss": 0.9497,
      "step": 3290
    },
    {
      "epoch": 0.7561675272518646,
      "grad_norm": 0.13209198415279388,
      "learning_rate": 3.406918351530376e-05,
      "loss": 0.9345,
      "step": 3295
    },
    {
      "epoch": 0.7573149741824441,
      "grad_norm": 0.1314823031425476,
      "learning_rate": 3.3768504741566664e-05,
      "loss": 0.9272,
      "step": 3300
    },
    {
      "epoch": 0.7584624211130235,
      "grad_norm": 0.13039475679397583,
      "learning_rate": 3.346888891100649e-05,
      "loss": 0.9358,
      "step": 3305
    },
    {
      "epoch": 0.759609868043603,
      "grad_norm": 0.12750479578971863,
      "learning_rate": 3.3170340832130134e-05,
      "loss": 0.9454,
      "step": 3310
    },
    {
      "epoch": 0.7607573149741824,
      "grad_norm": 0.12613527476787567,
      "learning_rate": 3.287286529630832e-05,
      "loss": 0.9337,
      "step": 3315
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 0.12861678004264832,
      "learning_rate": 3.2576467077698493e-05,
      "loss": 0.9045,
      "step": 3320
    },
    {
      "epoch": 0.7630522088353414,
      "grad_norm": 0.12556509673595428,
      "learning_rate": 3.228115093316848e-05,
      "loss": 0.9095,
      "step": 3325
    },
    {
      "epoch": 0.7641996557659209,
      "grad_norm": 0.13398994505405426,
      "learning_rate": 3.198692160221987e-05,
      "loss": 0.966,
      "step": 3330
    },
    {
      "epoch": 0.7653471026965003,
      "grad_norm": 0.13073302805423737,
      "learning_rate": 3.169378380691218e-05,
      "loss": 0.9506,
      "step": 3335
    },
    {
      "epoch": 0.7664945496270797,
      "grad_norm": 0.13139608502388,
      "learning_rate": 3.140174225178692e-05,
      "loss": 0.9285,
      "step": 3340
    },
    {
      "epoch": 0.7676419965576592,
      "grad_norm": 0.13152596354484558,
      "learning_rate": 3.111080162379215e-05,
      "loss": 0.9519,
      "step": 3345
    },
    {
      "epoch": 0.7687894434882386,
      "grad_norm": 0.12944857776165009,
      "learning_rate": 3.082096659220722e-05,
      "loss": 0.9108,
      "step": 3350
    },
    {
      "epoch": 0.7699368904188181,
      "grad_norm": 0.12854111194610596,
      "learning_rate": 3.0532241808567966e-05,
      "loss": 0.9315,
      "step": 3355
    },
    {
      "epoch": 0.7710843373493976,
      "grad_norm": 0.12585926055908203,
      "learning_rate": 3.0244631906591825e-05,
      "loss": 0.9585,
      "step": 3360
    },
    {
      "epoch": 0.7722317842799771,
      "grad_norm": 0.12900350987911224,
      "learning_rate": 2.9958141502103722e-05,
      "loss": 0.9725,
      "step": 3365
    },
    {
      "epoch": 0.7733792312105565,
      "grad_norm": 0.13046613335609436,
      "learning_rate": 2.9672775192961756e-05,
      "loss": 0.9677,
      "step": 3370
    },
    {
      "epoch": 0.774526678141136,
      "grad_norm": 0.1273292899131775,
      "learning_rate": 2.938853755898364e-05,
      "loss": 0.9494,
      "step": 3375
    },
    {
      "epoch": 0.7756741250717154,
      "grad_norm": 0.13062521815299988,
      "learning_rate": 2.910543316187301e-05,
      "loss": 0.9134,
      "step": 3380
    },
    {
      "epoch": 0.7768215720022948,
      "grad_norm": 0.12806479632854462,
      "learning_rate": 2.882346654514627e-05,
      "loss": 0.9374,
      "step": 3385
    },
    {
      "epoch": 0.7779690189328744,
      "grad_norm": 0.1296064406633377,
      "learning_rate": 2.8542642234059725e-05,
      "loss": 0.9424,
      "step": 3390
    },
    {
      "epoch": 0.7791164658634538,
      "grad_norm": 0.12994979321956635,
      "learning_rate": 2.826296473553697e-05,
      "loss": 0.9855,
      "step": 3395
    },
    {
      "epoch": 0.7802639127940333,
      "grad_norm": 0.13433627784252167,
      "learning_rate": 2.7984438538096392e-05,
      "loss": 0.8971,
      "step": 3400
    },
    {
      "epoch": 0.7814113597246127,
      "grad_norm": 0.13279375433921814,
      "learning_rate": 2.7707068111779377e-05,
      "loss": 0.9418,
      "step": 3405
    },
    {
      "epoch": 0.7825588066551922,
      "grad_norm": 0.14554573595523834,
      "learning_rate": 2.7430857908078345e-05,
      "loss": 0.9549,
      "step": 3410
    },
    {
      "epoch": 0.7837062535857716,
      "grad_norm": 0.12296421080827713,
      "learning_rate": 2.7155812359865517e-05,
      "loss": 0.9126,
      "step": 3415
    },
    {
      "epoch": 0.7848537005163512,
      "grad_norm": 0.12731173634529114,
      "learning_rate": 2.6881935881321563e-05,
      "loss": 0.9698,
      "step": 3420
    },
    {
      "epoch": 0.7860011474469306,
      "grad_norm": 0.13245446979999542,
      "learning_rate": 2.6609232867864896e-05,
      "loss": 0.9658,
      "step": 3425
    },
    {
      "epoch": 0.7871485943775101,
      "grad_norm": 0.1442078948020935,
      "learning_rate": 2.6337707696081094e-05,
      "loss": 0.9532,
      "step": 3430
    },
    {
      "epoch": 0.7882960413080895,
      "grad_norm": 0.12503184378147125,
      "learning_rate": 2.606736472365272e-05,
      "loss": 0.9534,
      "step": 3435
    },
    {
      "epoch": 0.789443488238669,
      "grad_norm": 0.13164812326431274,
      "learning_rate": 2.5798208289289204e-05,
      "loss": 0.9451,
      "step": 3440
    },
    {
      "epoch": 0.7905909351692484,
      "grad_norm": 0.13405010104179382,
      "learning_rate": 2.5530242712657492e-05,
      "loss": 0.9547,
      "step": 3445
    },
    {
      "epoch": 0.7917383820998278,
      "grad_norm": 0.13217557966709137,
      "learning_rate": 2.526347229431242e-05,
      "loss": 0.9647,
      "step": 3450
    },
    {
      "epoch": 0.7928858290304074,
      "grad_norm": 0.13415786623954773,
      "learning_rate": 2.499790131562797e-05,
      "loss": 0.9719,
      "step": 3455
    },
    {
      "epoch": 0.7940332759609868,
      "grad_norm": 0.12753376364707947,
      "learning_rate": 2.4733534038728257e-05,
      "loss": 0.9217,
      "step": 3460
    },
    {
      "epoch": 0.7951807228915663,
      "grad_norm": 0.1291799694299698,
      "learning_rate": 2.4470374706419485e-05,
      "loss": 0.9598,
      "step": 3465
    },
    {
      "epoch": 0.7963281698221457,
      "grad_norm": 0.1334388107061386,
      "learning_rate": 2.4208427542121504e-05,
      "loss": 0.9298,
      "step": 3470
    },
    {
      "epoch": 0.7974756167527252,
      "grad_norm": 0.1315917819738388,
      "learning_rate": 2.394769674980035e-05,
      "loss": 0.9623,
      "step": 3475
    },
    {
      "epoch": 0.7986230636833046,
      "grad_norm": 0.1384865790605545,
      "learning_rate": 2.3688186513900455e-05,
      "loss": 0.9543,
      "step": 3480
    },
    {
      "epoch": 0.7997705106138842,
      "grad_norm": 0.12887543439865112,
      "learning_rate": 2.34299009992778e-05,
      "loss": 0.9742,
      "step": 3485
    },
    {
      "epoch": 0.8009179575444636,
      "grad_norm": 0.12744690477848053,
      "learning_rate": 2.317284435113278e-05,
      "loss": 0.954,
      "step": 3490
    },
    {
      "epoch": 0.802065404475043,
      "grad_norm": 0.12910710275173187,
      "learning_rate": 2.2917020694944023e-05,
      "loss": 0.9107,
      "step": 3495
    },
    {
      "epoch": 0.8032128514056225,
      "grad_norm": 0.1282537430524826,
      "learning_rate": 2.2662434136401722e-05,
      "loss": 0.9795,
      "step": 3500
    },
    {
      "epoch": 0.8043602983362019,
      "grad_norm": 0.13550768792629242,
      "learning_rate": 2.2409088761342235e-05,
      "loss": 0.9415,
      "step": 3505
    },
    {
      "epoch": 0.8055077452667814,
      "grad_norm": 0.12883414328098297,
      "learning_rate": 2.215698863568213e-05,
      "loss": 0.9445,
      "step": 3510
    },
    {
      "epoch": 0.8066551921973609,
      "grad_norm": 0.13625557720661163,
      "learning_rate": 2.1906137805353212e-05,
      "loss": 0.9726,
      "step": 3515
    },
    {
      "epoch": 0.8078026391279404,
      "grad_norm": 0.12835553288459778,
      "learning_rate": 2.1656540296237316e-05,
      "loss": 0.9235,
      "step": 3520
    },
    {
      "epoch": 0.8089500860585198,
      "grad_norm": 0.13299131393432617,
      "learning_rate": 2.1408200114101985e-05,
      "loss": 0.9173,
      "step": 3525
    },
    {
      "epoch": 0.8100975329890993,
      "grad_norm": 0.13075728714466095,
      "learning_rate": 2.116112124453592e-05,
      "loss": 0.9613,
      "step": 3530
    },
    {
      "epoch": 0.8112449799196787,
      "grad_norm": 0.1347227394580841,
      "learning_rate": 2.0915307652885164e-05,
      "loss": 0.9611,
      "step": 3535
    },
    {
      "epoch": 0.8123924268502581,
      "grad_norm": 0.13222533464431763,
      "learning_rate": 2.067076328418949e-05,
      "loss": 0.9492,
      "step": 3540
    },
    {
      "epoch": 0.8135398737808376,
      "grad_norm": 0.12829838693141937,
      "learning_rate": 2.0427492063118935e-05,
      "loss": 0.9843,
      "step": 3545
    },
    {
      "epoch": 0.8146873207114171,
      "grad_norm": 0.13268935680389404,
      "learning_rate": 2.018549789391102e-05,
      "loss": 0.9879,
      "step": 3550
    },
    {
      "epoch": 0.8158347676419966,
      "grad_norm": 0.12710270285606384,
      "learning_rate": 1.994478466030787e-05,
      "loss": 0.8988,
      "step": 3555
    },
    {
      "epoch": 0.816982214572576,
      "grad_norm": 0.13554465770721436,
      "learning_rate": 1.970535622549401e-05,
      "loss": 0.948,
      "step": 3560
    },
    {
      "epoch": 0.8181296615031555,
      "grad_norm": 0.12602022290229797,
      "learning_rate": 1.946721643203443e-05,
      "loss": 0.9437,
      "step": 3565
    },
    {
      "epoch": 0.8192771084337349,
      "grad_norm": 0.1312330663204193,
      "learning_rate": 1.923036910181275e-05,
      "loss": 0.9544,
      "step": 3570
    },
    {
      "epoch": 0.8204245553643144,
      "grad_norm": 0.12853941321372986,
      "learning_rate": 1.8994818035969975e-05,
      "loss": 0.9798,
      "step": 3575
    },
    {
      "epoch": 0.8215720022948939,
      "grad_norm": 0.13181468844413757,
      "learning_rate": 1.8760567014843545e-05,
      "loss": 0.944,
      "step": 3580
    },
    {
      "epoch": 0.8227194492254734,
      "grad_norm": 0.12790626287460327,
      "learning_rate": 1.8527619797906494e-05,
      "loss": 0.901,
      "step": 3585
    },
    {
      "epoch": 0.8238668961560528,
      "grad_norm": 0.12571856379508972,
      "learning_rate": 1.8295980123707357e-05,
      "loss": 0.955,
      "step": 3590
    },
    {
      "epoch": 0.8250143430866322,
      "grad_norm": 0.1319160759449005,
      "learning_rate": 1.8065651709809905e-05,
      "loss": 0.9046,
      "step": 3595
    },
    {
      "epoch": 0.8261617900172117,
      "grad_norm": 0.1291453242301941,
      "learning_rate": 1.783663825273372e-05,
      "loss": 0.9337,
      "step": 3600
    },
    {
      "epoch": 0.8273092369477911,
      "grad_norm": 0.1331816017627716,
      "learning_rate": 1.7608943427894686e-05,
      "loss": 0.9563,
      "step": 3605
    },
    {
      "epoch": 0.8284566838783706,
      "grad_norm": 0.14010034501552582,
      "learning_rate": 1.7382570889546124e-05,
      "loss": 0.9518,
      "step": 3610
    },
    {
      "epoch": 0.8296041308089501,
      "grad_norm": 0.1268058568239212,
      "learning_rate": 1.7157524270720036e-05,
      "loss": 0.912,
      "step": 3615
    },
    {
      "epoch": 0.8307515777395296,
      "grad_norm": 0.1255045235157013,
      "learning_rate": 1.6933807183168994e-05,
      "loss": 0.9146,
      "step": 3620
    },
    {
      "epoch": 0.831899024670109,
      "grad_norm": 0.14171355962753296,
      "learning_rate": 1.6711423217307885e-05,
      "loss": 0.9269,
      "step": 3625
    },
    {
      "epoch": 0.8330464716006885,
      "grad_norm": 0.12760847806930542,
      "learning_rate": 1.64903759421566e-05,
      "loss": 0.9143,
      "step": 3630
    },
    {
      "epoch": 0.8341939185312679,
      "grad_norm": 0.1354866921901703,
      "learning_rate": 1.627066890528247e-05,
      "loss": 0.9634,
      "step": 3635
    },
    {
      "epoch": 0.8353413654618473,
      "grad_norm": 0.1262219399213791,
      "learning_rate": 1.6052305632743592e-05,
      "loss": 0.9785,
      "step": 3640
    },
    {
      "epoch": 0.8364888123924269,
      "grad_norm": 0.1341947317123413,
      "learning_rate": 1.583528962903197e-05,
      "loss": 0.9685,
      "step": 3645
    },
    {
      "epoch": 0.8376362593230063,
      "grad_norm": 0.12864305078983307,
      "learning_rate": 1.5619624377017537e-05,
      "loss": 0.9519,
      "step": 3650
    },
    {
      "epoch": 0.8387837062535858,
      "grad_norm": 0.15601038932800293,
      "learning_rate": 1.540531333789207e-05,
      "loss": 0.9302,
      "step": 3655
    },
    {
      "epoch": 0.8399311531841652,
      "grad_norm": 0.13577106595039368,
      "learning_rate": 1.5192359951113755e-05,
      "loss": 0.9682,
      "step": 3660
    },
    {
      "epoch": 0.8410786001147447,
      "grad_norm": 0.1369991898536682,
      "learning_rate": 1.4980767634351877e-05,
      "loss": 0.9267,
      "step": 3665
    },
    {
      "epoch": 0.8422260470453241,
      "grad_norm": 0.12964287400245667,
      "learning_rate": 1.4770539783432113e-05,
      "loss": 0.9279,
      "step": 3670
    },
    {
      "epoch": 0.8433734939759037,
      "grad_norm": 0.1294836550951004,
      "learning_rate": 1.4561679772281877e-05,
      "loss": 0.9176,
      "step": 3675
    },
    {
      "epoch": 0.8445209409064831,
      "grad_norm": 0.1307843178510666,
      "learning_rate": 1.4354190952876334e-05,
      "loss": 0.944,
      "step": 3680
    },
    {
      "epoch": 0.8456683878370626,
      "grad_norm": 0.1265002340078354,
      "learning_rate": 1.4148076655184373e-05,
      "loss": 0.9434,
      "step": 3685
    },
    {
      "epoch": 0.846815834767642,
      "grad_norm": 0.13340280950069427,
      "learning_rate": 1.3943340187115494e-05,
      "loss": 0.9606,
      "step": 3690
    },
    {
      "epoch": 0.8479632816982214,
      "grad_norm": 0.12884503602981567,
      "learning_rate": 1.373998483446638e-05,
      "loss": 0.9117,
      "step": 3695
    },
    {
      "epoch": 0.8491107286288009,
      "grad_norm": 0.13453038036823273,
      "learning_rate": 1.3538013860868436e-05,
      "loss": 0.9357,
      "step": 3700
    },
    {
      "epoch": 0.8502581755593803,
      "grad_norm": 0.12499819695949554,
      "learning_rate": 1.3337430507735205e-05,
      "loss": 0.949,
      "step": 3705
    },
    {
      "epoch": 0.8514056224899599,
      "grad_norm": 0.12760350108146667,
      "learning_rate": 1.313823799421051e-05,
      "loss": 0.9245,
      "step": 3710
    },
    {
      "epoch": 0.8525530694205393,
      "grad_norm": 0.12545782327651978,
      "learning_rate": 1.2940439517116676e-05,
      "loss": 0.9163,
      "step": 3715
    },
    {
      "epoch": 0.8537005163511188,
      "grad_norm": 0.14146262407302856,
      "learning_rate": 1.2744038250903267e-05,
      "loss": 0.9344,
      "step": 3720
    },
    {
      "epoch": 0.8548479632816982,
      "grad_norm": 0.1278744488954544,
      "learning_rate": 1.2549037347596115e-05,
      "loss": 0.9252,
      "step": 3725
    },
    {
      "epoch": 0.8559954102122777,
      "grad_norm": 0.1335706263780594,
      "learning_rate": 1.2355439936746827e-05,
      "loss": 0.9753,
      "step": 3730
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.12698203325271606,
      "learning_rate": 1.2163249125382426e-05,
      "loss": 0.9356,
      "step": 3735
    },
    {
      "epoch": 0.8582903040734366,
      "grad_norm": 0.129505917429924,
      "learning_rate": 1.1972467997955595e-05,
      "loss": 0.9424,
      "step": 3740
    },
    {
      "epoch": 0.8594377510040161,
      "grad_norm": 0.13229629397392273,
      "learning_rate": 1.1783099616295056e-05,
      "loss": 0.9503,
      "step": 3745
    },
    {
      "epoch": 0.8605851979345955,
      "grad_norm": 0.13013221323490143,
      "learning_rate": 1.1595147019556607e-05,
      "loss": 0.9307,
      "step": 3750
    },
    {
      "epoch": 0.861732644865175,
      "grad_norm": 0.1345113217830658,
      "learning_rate": 1.140861322417417e-05,
      "loss": 0.9276,
      "step": 3755
    },
    {
      "epoch": 0.8628800917957544,
      "grad_norm": 0.1310335397720337,
      "learning_rate": 1.1223501223811451e-05,
      "loss": 0.9732,
      "step": 3760
    },
    {
      "epoch": 0.8640275387263339,
      "grad_norm": 0.12777724862098694,
      "learning_rate": 1.1039813989313951e-05,
      "loss": 0.9675,
      "step": 3765
    },
    {
      "epoch": 0.8651749856569134,
      "grad_norm": 0.13015612959861755,
      "learning_rate": 1.085755446866119e-05,
      "loss": 0.9165,
      "step": 3770
    },
    {
      "epoch": 0.8663224325874929,
      "grad_norm": 0.1256059855222702,
      "learning_rate": 1.0676725586919457e-05,
      "loss": 0.9824,
      "step": 3775
    },
    {
      "epoch": 0.8674698795180723,
      "grad_norm": 0.13032986223697662,
      "learning_rate": 1.0497330246194848e-05,
      "loss": 0.9506,
      "step": 3780
    },
    {
      "epoch": 0.8686173264486517,
      "grad_norm": 0.12943556904792786,
      "learning_rate": 1.031937132558668e-05,
      "loss": 0.9085,
      "step": 3785
    },
    {
      "epoch": 0.8697647733792312,
      "grad_norm": 0.13594165444374084,
      "learning_rate": 1.014285168114133e-05,
      "loss": 0.9401,
      "step": 3790
    },
    {
      "epoch": 0.8709122203098106,
      "grad_norm": 0.12672476470470428,
      "learning_rate": 9.96777414580633e-06,
      "loss": 0.9781,
      "step": 3795
    },
    {
      "epoch": 0.8720596672403901,
      "grad_norm": 0.13382181525230408,
      "learning_rate": 9.794141529384915e-06,
      "loss": 0.9456,
      "step": 3800
    },
    {
      "epoch": 0.8732071141709696,
      "grad_norm": 0.12753736972808838,
      "learning_rate": 9.621956618491024e-06,
      "loss": 0.9561,
      "step": 3805
    },
    {
      "epoch": 0.8743545611015491,
      "grad_norm": 0.1312447190284729,
      "learning_rate": 9.451222176504414e-06,
      "loss": 0.966,
      "step": 3810
    },
    {
      "epoch": 0.8755020080321285,
      "grad_norm": 0.1302998960018158,
      "learning_rate": 9.281940943526491e-06,
      "loss": 0.936,
      "step": 3815
    },
    {
      "epoch": 0.876649454962708,
      "grad_norm": 0.12688572704792023,
      "learning_rate": 9.114115636336152e-06,
      "loss": 0.9623,
      "step": 3820
    },
    {
      "epoch": 0.8777969018932874,
      "grad_norm": 0.1292993426322937,
      "learning_rate": 8.947748948346357e-06,
      "loss": 0.9251,
      "step": 3825
    },
    {
      "epoch": 0.8789443488238669,
      "grad_norm": 0.12286306172609329,
      "learning_rate": 8.782843549560771e-06,
      "loss": 0.9388,
      "step": 3830
    },
    {
      "epoch": 0.8800917957544464,
      "grad_norm": 0.12918008863925934,
      "learning_rate": 8.61940208653097e-06,
      "loss": 0.9499,
      "step": 3835
    },
    {
      "epoch": 0.8812392426850258,
      "grad_norm": 0.13565728068351746,
      "learning_rate": 8.457427182313937e-06,
      "loss": 0.9287,
      "step": 3840
    },
    {
      "epoch": 0.8823866896156053,
      "grad_norm": 0.137693852186203,
      "learning_rate": 8.296921436430071e-06,
      "loss": 0.966,
      "step": 3845
    },
    {
      "epoch": 0.8835341365461847,
      "grad_norm": 0.12873981893062592,
      "learning_rate": 8.137887424821277e-06,
      "loss": 0.9276,
      "step": 3850
    },
    {
      "epoch": 0.8846815834767642,
      "grad_norm": 0.13087016344070435,
      "learning_rate": 7.980327699809832e-06,
      "loss": 0.9506,
      "step": 3855
    },
    {
      "epoch": 0.8858290304073436,
      "grad_norm": 0.13923919200897217,
      "learning_rate": 7.824244790057223e-06,
      "loss": 0.9048,
      "step": 3860
    },
    {
      "epoch": 0.8869764773379232,
      "grad_norm": 0.12863433361053467,
      "learning_rate": 7.66964120052377e-06,
      "loss": 0.9634,
      "step": 3865
    },
    {
      "epoch": 0.8881239242685026,
      "grad_norm": 0.12779028713703156,
      "learning_rate": 7.516519412428203e-06,
      "loss": 0.9463,
      "step": 3870
    },
    {
      "epoch": 0.8892713711990821,
      "grad_norm": 0.13034799695014954,
      "learning_rate": 7.3648818832080745e-06,
      "loss": 0.979,
      "step": 3875
    },
    {
      "epoch": 0.8904188181296615,
      "grad_norm": 0.13406170904636383,
      "learning_rate": 7.214731046480094e-06,
      "loss": 0.9234,
      "step": 3880
    },
    {
      "epoch": 0.891566265060241,
      "grad_norm": 0.13078339397907257,
      "learning_rate": 7.066069312001289e-06,
      "loss": 0.9578,
      "step": 3885
    },
    {
      "epoch": 0.8927137119908204,
      "grad_norm": 0.13037657737731934,
      "learning_rate": 6.918899065630113e-06,
      "loss": 0.933,
      "step": 3890
    },
    {
      "epoch": 0.8938611589213998,
      "grad_norm": 0.12971769273281097,
      "learning_rate": 6.773222669288359e-06,
      "loss": 0.9478,
      "step": 3895
    },
    {
      "epoch": 0.8950086058519794,
      "grad_norm": 0.13167798519134521,
      "learning_rate": 6.629042460923096e-06,
      "loss": 0.9555,
      "step": 3900
    },
    {
      "epoch": 0.8961560527825588,
      "grad_norm": 0.13504882156848907,
      "learning_rate": 6.486360754469234e-06,
      "loss": 0.9192,
      "step": 3905
    },
    {
      "epoch": 0.8973034997131383,
      "grad_norm": 0.1290195882320404,
      "learning_rate": 6.345179839812343e-06,
      "loss": 0.9263,
      "step": 3910
    },
    {
      "epoch": 0.8984509466437177,
      "grad_norm": 0.127256378531456,
      "learning_rate": 6.205501982751971e-06,
      "loss": 0.9283,
      "step": 3915
    },
    {
      "epoch": 0.8995983935742972,
      "grad_norm": 0.1287613958120346,
      "learning_rate": 6.067329424965162e-06,
      "loss": 0.9485,
      "step": 3920
    },
    {
      "epoch": 0.9007458405048766,
      "grad_norm": 0.13389694690704346,
      "learning_rate": 5.930664383970641e-06,
      "loss": 0.9235,
      "step": 3925
    },
    {
      "epoch": 0.9018932874354562,
      "grad_norm": 0.13568086922168732,
      "learning_rate": 5.795509053093029e-06,
      "loss": 0.9484,
      "step": 3930
    },
    {
      "epoch": 0.9030407343660356,
      "grad_norm": 0.12701798975467682,
      "learning_rate": 5.6618656014278406e-06,
      "loss": 0.9161,
      "step": 3935
    },
    {
      "epoch": 0.904188181296615,
      "grad_norm": 0.13050422072410583,
      "learning_rate": 5.5297361738065325e-06,
      "loss": 0.9688,
      "step": 3940
    },
    {
      "epoch": 0.9053356282271945,
      "grad_norm": 0.13256116211414337,
      "learning_rate": 5.399122890762143e-06,
      "loss": 0.9124,
      "step": 3945
    },
    {
      "epoch": 0.9064830751577739,
      "grad_norm": 0.1276332139968872,
      "learning_rate": 5.270027848495207e-06,
      "loss": 0.9692,
      "step": 3950
    },
    {
      "epoch": 0.9076305220883534,
      "grad_norm": 0.140411376953125,
      "learning_rate": 5.1424531188402405e-06,
      "loss": 0.9739,
      "step": 3955
    },
    {
      "epoch": 0.9087779690189329,
      "grad_norm": 0.12793396413326263,
      "learning_rate": 5.016400749232297e-06,
      "loss": 0.9482,
      "step": 3960
    },
    {
      "epoch": 0.9099254159495124,
      "grad_norm": 0.12200927734375,
      "learning_rate": 4.89187276267431e-06,
      "loss": 0.95,
      "step": 3965
    },
    {
      "epoch": 0.9110728628800918,
      "grad_norm": 0.13150322437286377,
      "learning_rate": 4.7688711577044354e-06,
      "loss": 0.9651,
      "step": 3970
    },
    {
      "epoch": 0.9122203098106713,
      "grad_norm": 0.13082684576511383,
      "learning_rate": 4.647397908364182e-06,
      "loss": 0.9028,
      "step": 3975
    },
    {
      "epoch": 0.9133677567412507,
      "grad_norm": 0.12200944870710373,
      "learning_rate": 4.5274549641665105e-06,
      "loss": 0.9135,
      "step": 3980
    },
    {
      "epoch": 0.9145152036718301,
      "grad_norm": 0.1254609227180481,
      "learning_rate": 4.40904425006472e-06,
      "loss": 0.9437,
      "step": 3985
    },
    {
      "epoch": 0.9156626506024096,
      "grad_norm": 0.12539255619049072,
      "learning_rate": 4.2921676664214535e-06,
      "loss": 0.8961,
      "step": 3990
    },
    {
      "epoch": 0.9168100975329891,
      "grad_norm": 0.12549173831939697,
      "learning_rate": 4.176827088978297e-06,
      "loss": 0.9377,
      "step": 3995
    },
    {
      "epoch": 0.9179575444635686,
      "grad_norm": 0.1272364854812622,
      "learning_rate": 4.0630243688255185e-06,
      "loss": 0.952,
      "step": 4000
    },
    {
      "epoch": 0.919104991394148,
      "grad_norm": 0.12669628858566284,
      "learning_rate": 3.950761332372543e-06,
      "loss": 0.9593,
      "step": 4005
    },
    {
      "epoch": 0.9202524383247275,
      "grad_norm": 0.13297000527381897,
      "learning_rate": 3.8400397813185054e-06,
      "loss": 0.9559,
      "step": 4010
    },
    {
      "epoch": 0.9213998852553069,
      "grad_norm": 0.1260041445493698,
      "learning_rate": 3.7308614926234165e-06,
      "loss": 0.9277,
      "step": 4015
    },
    {
      "epoch": 0.9225473321858864,
      "grad_norm": 0.12770813703536987,
      "learning_rate": 3.6232282184795794e-06,
      "loss": 0.9448,
      "step": 4020
    },
    {
      "epoch": 0.9236947791164659,
      "grad_norm": 0.13022121787071228,
      "learning_rate": 3.517141686283498e-06,
      "loss": 0.9415,
      "step": 4025
    },
    {
      "epoch": 0.9248422260470454,
      "grad_norm": 0.12520062923431396,
      "learning_rate": 3.412603598608188e-06,
      "loss": 0.9453,
      "step": 4030
    },
    {
      "epoch": 0.9259896729776248,
      "grad_norm": 0.12913860380649567,
      "learning_rate": 3.3096156331758e-06,
      "loss": 0.9492,
      "step": 4035
    },
    {
      "epoch": 0.9271371199082042,
      "grad_norm": 0.12940272688865662,
      "learning_rate": 3.2081794428307278e-06,
      "loss": 0.949,
      "step": 4040
    },
    {
      "epoch": 0.9282845668387837,
      "grad_norm": 0.12925906479358673,
      "learning_rate": 3.1082966555130654e-06,
      "loss": 0.9143,
      "step": 4045
    },
    {
      "epoch": 0.9294320137693631,
      "grad_norm": 0.1461576372385025,
      "learning_rate": 3.0099688742324715e-06,
      "loss": 0.9194,
      "step": 4050
    },
    {
      "epoch": 0.9305794606999426,
      "grad_norm": 0.1329105645418167,
      "learning_rate": 2.913197677042456e-06,
      "loss": 0.9007,
      "step": 4055
    },
    {
      "epoch": 0.9317269076305221,
      "grad_norm": 0.13174231350421906,
      "learning_rate": 2.8179846170150903e-06,
      "loss": 0.9247,
      "step": 4060
    },
    {
      "epoch": 0.9328743545611016,
      "grad_norm": 0.12422098964452744,
      "learning_rate": 2.7243312222159924e-06,
      "loss": 0.9283,
      "step": 4065
    },
    {
      "epoch": 0.934021801491681,
      "grad_norm": 0.13059760630130768,
      "learning_rate": 2.6322389956799143e-06,
      "loss": 0.9783,
      "step": 4070
    },
    {
      "epoch": 0.9351692484222605,
      "grad_norm": 0.1337437480688095,
      "learning_rate": 2.541709415386495e-06,
      "loss": 0.9643,
      "step": 4075
    },
    {
      "epoch": 0.9363166953528399,
      "grad_norm": 0.13037900626659393,
      "learning_rate": 2.4527439342366785e-06,
      "loss": 0.9529,
      "step": 4080
    },
    {
      "epoch": 0.9374641422834193,
      "grad_norm": 0.13122588396072388,
      "learning_rate": 2.3653439800292556e-06,
      "loss": 0.9504,
      "step": 4085
    },
    {
      "epoch": 0.9386115892139989,
      "grad_norm": 0.12561407685279846,
      "learning_rate": 2.2795109554381024e-06,
      "loss": 0.9457,
      "step": 4090
    },
    {
      "epoch": 0.9397590361445783,
      "grad_norm": 0.13426658511161804,
      "learning_rate": 2.195246237989479e-06,
      "loss": 0.9089,
      "step": 4095
    },
    {
      "epoch": 0.9409064830751578,
      "grad_norm": 0.12589140236377716,
      "learning_rate": 2.1125511800401234e-06,
      "loss": 0.8939,
      "step": 4100
    },
    {
      "epoch": 0.9420539300057372,
      "grad_norm": 0.13555549085140228,
      "learning_rate": 2.0314271087554126e-06,
      "loss": 0.9647,
      "step": 4105
    },
    {
      "epoch": 0.9432013769363167,
      "grad_norm": 0.1271609365940094,
      "learning_rate": 1.951875326088104e-06,
      "loss": 0.9331,
      "step": 4110
    },
    {
      "epoch": 0.9443488238668961,
      "grad_norm": 0.14045216143131256,
      "learning_rate": 1.8738971087574275e-06,
      "loss": 0.9597,
      "step": 4115
    },
    {
      "epoch": 0.9454962707974757,
      "grad_norm": 0.12654034793376923,
      "learning_rate": 1.797493708228659e-06,
      "loss": 0.9267,
      "step": 4120
    },
    {
      "epoch": 0.9466437177280551,
      "grad_norm": 0.1343478262424469,
      "learning_rate": 1.7226663506929142e-06,
      "loss": 0.9373,
      "step": 4125
    },
    {
      "epoch": 0.9477911646586346,
      "grad_norm": 0.1329105794429779,
      "learning_rate": 1.6494162370475852e-06,
      "loss": 0.9255,
      "step": 4130
    },
    {
      "epoch": 0.948938611589214,
      "grad_norm": 0.1293703317642212,
      "learning_rate": 1.5777445428770022e-06,
      "loss": 0.96,
      "step": 4135
    },
    {
      "epoch": 0.9500860585197934,
      "grad_norm": 0.13163906335830688,
      "learning_rate": 1.5076524184336027e-06,
      "loss": 0.924,
      "step": 4140
    },
    {
      "epoch": 0.9512335054503729,
      "grad_norm": 0.12938107550144196,
      "learning_rate": 1.4391409886194474e-06,
      "loss": 0.9338,
      "step": 4145
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 0.13095106184482574,
      "learning_rate": 1.3722113529681668e-06,
      "loss": 0.9538,
      "step": 4150
    },
    {
      "epoch": 0.9535283993115319,
      "grad_norm": 0.13413451611995697,
      "learning_rate": 1.306864585627332e-06,
      "loss": 0.9812,
      "step": 4155
    },
    {
      "epoch": 0.9546758462421113,
      "grad_norm": 0.1287609338760376,
      "learning_rate": 1.2431017353412233e-06,
      "loss": 0.9374,
      "step": 4160
    },
    {
      "epoch": 0.9558232931726908,
      "grad_norm": 0.12516798079013824,
      "learning_rate": 1.1809238254339105e-06,
      "loss": 0.9607,
      "step": 4165
    },
    {
      "epoch": 0.9569707401032702,
      "grad_norm": 0.13837283849716187,
      "learning_rate": 1.1203318537929996e-06,
      "loss": 0.9585,
      "step": 4170
    },
    {
      "epoch": 0.9581181870338497,
      "grad_norm": 0.13161715865135193,
      "learning_rate": 1.0613267928534453e-06,
      "loss": 0.9289,
      "step": 4175
    },
    {
      "epoch": 0.9592656339644291,
      "grad_norm": 0.12652412056922913,
      "learning_rate": 1.0039095895820639e-06,
      "loss": 0.9477,
      "step": 4180
    },
    {
      "epoch": 0.9604130808950087,
      "grad_norm": 0.12420324981212616,
      "learning_rate": 9.480811654622557e-07,
      "loss": 0.9562,
      "step": 4185
    },
    {
      "epoch": 0.9615605278255881,
      "grad_norm": 0.12834912538528442,
      "learning_rate": 8.938424164792736e-07,
      "loss": 0.9724,
      "step": 4190
    },
    {
      "epoch": 0.9627079747561675,
      "grad_norm": 0.12531572580337524,
      "learning_rate": 8.411942131058115e-07,
      "loss": 0.9464,
      "step": 4195
    },
    {
      "epoch": 0.963855421686747,
      "grad_norm": 0.12882524728775024,
      "learning_rate": 7.90137400288049e-07,
      "loss": 0.9373,
      "step": 4200
    },
    {
      "epoch": 0.9650028686173264,
      "grad_norm": 0.13834188878536224,
      "learning_rate": 7.406727974320627e-07,
      "loss": 0.937,
      "step": 4205
    },
    {
      "epoch": 0.9661503155479059,
      "grad_norm": 0.13418422639369965,
      "learning_rate": 6.928011983907245e-07,
      "loss": 0.9551,
      "step": 4210
    },
    {
      "epoch": 0.9672977624784854,
      "grad_norm": 0.1264306902885437,
      "learning_rate": 6.465233714509245e-07,
      "loss": 0.9582,
      "step": 4215
    },
    {
      "epoch": 0.9684452094090649,
      "grad_norm": 0.1280186027288437,
      "learning_rate": 6.01840059321257e-07,
      "loss": 0.9523,
      "step": 4220
    },
    {
      "epoch": 0.9695926563396443,
      "grad_norm": 0.12469854950904846,
      "learning_rate": 5.587519791200869e-07,
      "loss": 0.9511,
      "step": 4225
    },
    {
      "epoch": 0.9707401032702238,
      "grad_norm": 0.13365860283374786,
      "learning_rate": 5.172598223640468e-07,
      "loss": 0.9195,
      "step": 4230
    },
    {
      "epoch": 0.9718875502008032,
      "grad_norm": 0.1352444291114807,
      "learning_rate": 4.773642549569579e-07,
      "loss": 0.9651,
      "step": 4235
    },
    {
      "epoch": 0.9730349971313826,
      "grad_norm": 0.12593533098697662,
      "learning_rate": 4.390659171790934e-07,
      "loss": 0.8991,
      "step": 4240
    },
    {
      "epoch": 0.9741824440619621,
      "grad_norm": 0.13229170441627502,
      "learning_rate": 4.023654236769647e-07,
      "loss": 0.9233,
      "step": 4245
    },
    {
      "epoch": 0.9753298909925416,
      "grad_norm": 0.13042078912258148,
      "learning_rate": 3.672633634534295e-07,
      "loss": 0.9276,
      "step": 4250
    },
    {
      "epoch": 0.9764773379231211,
      "grad_norm": 0.12525303661823273,
      "learning_rate": 3.3376029985819903e-07,
      "loss": 0.9373,
      "step": 4255
    },
    {
      "epoch": 0.9776247848537005,
      "grad_norm": 0.12787853181362152,
      "learning_rate": 3.0185677057887885e-07,
      "loss": 0.9313,
      "step": 4260
    },
    {
      "epoch": 0.97877223178428,
      "grad_norm": 0.12966030836105347,
      "learning_rate": 2.715532876322646e-07,
      "loss": 0.9288,
      "step": 4265
    },
    {
      "epoch": 0.9799196787148594,
      "grad_norm": 0.1332397758960724,
      "learning_rate": 2.428503373561708e-07,
      "loss": 0.9362,
      "step": 4270
    },
    {
      "epoch": 0.9810671256454389,
      "grad_norm": 0.13498224318027496,
      "learning_rate": 2.1574838040161473e-07,
      "loss": 0.9411,
      "step": 4275
    },
    {
      "epoch": 0.9822145725760184,
      "grad_norm": 0.12744614481925964,
      "learning_rate": 1.9024785172541136e-07,
      "loss": 0.9343,
      "step": 4280
    },
    {
      "epoch": 0.9833620195065979,
      "grad_norm": 0.1275186389684677,
      "learning_rate": 1.6634916058319018e-07,
      "loss": 0.9256,
      "step": 4285
    },
    {
      "epoch": 0.9845094664371773,
      "grad_norm": 0.12919220328330994,
      "learning_rate": 1.4405269052284455e-07,
      "loss": 0.9333,
      "step": 4290
    },
    {
      "epoch": 0.9856569133677567,
      "grad_norm": 0.13401302695274353,
      "learning_rate": 1.2335879937839246e-07,
      "loss": 0.8878,
      "step": 4295
    },
    {
      "epoch": 0.9868043602983362,
      "grad_norm": 0.1339481770992279,
      "learning_rate": 1.0426781926416996e-07,
      "loss": 0.9369,
      "step": 4300
    },
    {
      "epoch": 0.99,
      "step": 4300,
      "total_flos": 1.209382783849123e+19,
      "train_loss": 0.0,
      "train_runtime": 0.0433,
      "train_samples_per_second": 31910.696,
      "train_steps_per_second": 993.599
    }
  ],
  "logging_steps": 5,
  "max_steps": 43,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 1.209382783849123e+19,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}