{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.941747572815534,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009708737864077669,
      "grad_norm": 3.141545295715332,
      "learning_rate": 1.5e-06,
      "loss": 0.3406,
      "step": 10
    },
    {
      "epoch": 0.019417475728155338,
      "grad_norm": 3.8397164344787598,
      "learning_rate": 3.166666666666667e-06,
      "loss": 0.3078,
      "step": 20
    },
    {
      "epoch": 0.02912621359223301,
      "grad_norm": 1.7836706638336182,
      "learning_rate": 4.833333333333333e-06,
      "loss": 0.1926,
      "step": 30
    },
    {
      "epoch": 0.038834951456310676,
      "grad_norm": 1.257520079612732,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.1476,
      "step": 40
    },
    {
      "epoch": 0.04854368932038835,
      "grad_norm": 1.0694407224655151,
      "learning_rate": 8.166666666666668e-06,
      "loss": 0.135,
      "step": 50
    },
    {
      "epoch": 0.05825242718446602,
      "grad_norm": 0.9249787926673889,
      "learning_rate": 9.833333333333333e-06,
      "loss": 0.1516,
      "step": 60
    },
    {
      "epoch": 0.06796116504854369,
      "grad_norm": 1.0610548257827759,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 0.1073,
      "step": 70
    },
    {
      "epoch": 0.07766990291262135,
      "grad_norm": 0.9249606132507324,
      "learning_rate": 1.3166666666666665e-05,
      "loss": 0.0894,
      "step": 80
    },
    {
      "epoch": 0.08737864077669903,
      "grad_norm": 0.6027079224586487,
      "learning_rate": 1.4833333333333336e-05,
      "loss": 0.0814,
      "step": 90
    },
    {
      "epoch": 0.0970873786407767,
      "grad_norm": 0.8849421739578247,
      "learning_rate": 1.65e-05,
      "loss": 0.0963,
      "step": 100
    },
    {
      "epoch": 0.10679611650485436,
      "grad_norm": 1.2120862007141113,
      "learning_rate": 1.8166666666666667e-05,
      "loss": 0.0817,
      "step": 110
    },
    {
      "epoch": 0.11650485436893204,
      "grad_norm": 0.781768262386322,
      "learning_rate": 1.9833333333333335e-05,
      "loss": 0.087,
      "step": 120
    },
    {
      "epoch": 0.1262135922330097,
      "grad_norm": 0.5335855484008789,
      "learning_rate": 2.15e-05,
      "loss": 0.0502,
      "step": 130
    },
    {
      "epoch": 0.13592233009708737,
      "grad_norm": 0.6909231543540955,
      "learning_rate": 2.3166666666666666e-05,
      "loss": 0.0813,
      "step": 140
    },
    {
      "epoch": 0.14563106796116504,
      "grad_norm": 0.8511887192726135,
      "learning_rate": 2.4833333333333335e-05,
      "loss": 0.0823,
      "step": 150
    },
    {
      "epoch": 0.1553398058252427,
      "grad_norm": 1.1109554767608643,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 0.0638,
      "step": 160
    },
    {
      "epoch": 0.1650485436893204,
      "grad_norm": 0.7318443655967712,
      "learning_rate": 2.816666666666667e-05,
      "loss": 0.0649,
      "step": 170
    },
    {
      "epoch": 0.17475728155339806,
      "grad_norm": 0.7571411728858948,
      "learning_rate": 2.9833333333333335e-05,
      "loss": 0.064,
      "step": 180
    },
    {
      "epoch": 0.18446601941747573,
      "grad_norm": 0.8281847834587097,
      "learning_rate": 3.15e-05,
      "loss": 0.0731,
      "step": 190
    },
    {
      "epoch": 0.1941747572815534,
      "grad_norm": 0.7228130102157593,
      "learning_rate": 3.316666666666667e-05,
      "loss": 0.0674,
      "step": 200
    },
    {
      "epoch": 0.20388349514563106,
      "grad_norm": 0.9129646420478821,
      "learning_rate": 3.483333333333334e-05,
      "loss": 0.0674,
      "step": 210
    },
    {
      "epoch": 0.21359223300970873,
      "grad_norm": 0.7561999559402466,
      "learning_rate": 3.65e-05,
      "loss": 0.0632,
      "step": 220
    },
    {
      "epoch": 0.22330097087378642,
      "grad_norm": 0.6391080617904663,
      "learning_rate": 3.816666666666667e-05,
      "loss": 0.0748,
      "step": 230
    },
    {
      "epoch": 0.23300970873786409,
      "grad_norm": 0.8094485402107239,
      "learning_rate": 3.983333333333333e-05,
      "loss": 0.0822,
      "step": 240
    },
    {
      "epoch": 0.24271844660194175,
      "grad_norm": 0.5360986590385437,
      "learning_rate": 4.15e-05,
      "loss": 0.0617,
      "step": 250
    },
    {
      "epoch": 0.2524271844660194,
      "grad_norm": 0.7553185224533081,
      "learning_rate": 4.316666666666667e-05,
      "loss": 0.0666,
      "step": 260
    },
    {
      "epoch": 0.2621359223300971,
      "grad_norm": 0.6815693378448486,
      "learning_rate": 4.483333333333333e-05,
      "loss": 0.061,
      "step": 270
    },
    {
      "epoch": 0.27184466019417475,
      "grad_norm": 0.5073156356811523,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 0.0473,
      "step": 280
    },
    {
      "epoch": 0.2815533980582524,
      "grad_norm": 0.8138278722763062,
      "learning_rate": 4.8166666666666674e-05,
      "loss": 0.0652,
      "step": 290
    },
    {
      "epoch": 0.2912621359223301,
      "grad_norm": 0.8348583579063416,
      "learning_rate": 4.9833333333333336e-05,
      "loss": 0.0746,
      "step": 300
    },
    {
      "epoch": 0.30097087378640774,
      "grad_norm": 1.0741016864776611,
      "learning_rate": 4.999969242985639e-05,
      "loss": 0.0718,
      "step": 310
    },
    {
      "epoch": 0.3106796116504854,
      "grad_norm": 1.182652473449707,
      "learning_rate": 4.999862923413781e-05,
      "loss": 0.06,
      "step": 320
    },
    {
      "epoch": 0.32038834951456313,
      "grad_norm": 0.835841953754425,
      "learning_rate": 4.999680664797127e-05,
      "loss": 0.0607,
      "step": 330
    },
    {
      "epoch": 0.3300970873786408,
      "grad_norm": 0.6821876764297485,
      "learning_rate": 4.999422472672202e-05,
      "loss": 0.0554,
      "step": 340
    },
    {
      "epoch": 0.33980582524271846,
      "grad_norm": 0.9348113536834717,
      "learning_rate": 4.99908835488218e-05,
      "loss": 0.0629,
      "step": 350
    },
    {
      "epoch": 0.34951456310679613,
      "grad_norm": 0.5926679372787476,
      "learning_rate": 4.998678321576651e-05,
      "loss": 0.0506,
      "step": 360
    },
    {
      "epoch": 0.3592233009708738,
      "grad_norm": 1.0646535158157349,
      "learning_rate": 4.9981923852113145e-05,
      "loss": 0.0782,
      "step": 370
    },
    {
      "epoch": 0.36893203883495146,
      "grad_norm": 1.0149277448654175,
      "learning_rate": 4.997630560547597e-05,
      "loss": 0.0665,
      "step": 380
    },
    {
      "epoch": 0.3786407766990291,
      "grad_norm": 0.9700633883476257,
      "learning_rate": 4.996992864652204e-05,
      "loss": 0.067,
      "step": 390
    },
    {
      "epoch": 0.3883495145631068,
      "grad_norm": 1.1201013326644897,
      "learning_rate": 4.996279316896606e-05,
      "loss": 0.0656,
      "step": 400
    },
    {
      "epoch": 0.39805825242718446,
      "grad_norm": 0.7025376558303833,
      "learning_rate": 4.9954899389564455e-05,
      "loss": 0.0711,
      "step": 410
    },
    {
      "epoch": 0.4077669902912621,
      "grad_norm": 0.6064212918281555,
      "learning_rate": 4.9946247548108794e-05,
      "loss": 0.0572,
      "step": 420
    },
    {
      "epoch": 0.4174757281553398,
      "grad_norm": 0.5238557457923889,
      "learning_rate": 4.993683790741852e-05,
      "loss": 0.0633,
      "step": 430
    },
    {
      "epoch": 0.42718446601941745,
      "grad_norm": 0.599403977394104,
      "learning_rate": 4.992667075333296e-05,
      "loss": 0.0508,
      "step": 440
    },
    {
      "epoch": 0.4368932038834951,
      "grad_norm": 0.55897057056427,
      "learning_rate": 4.991574639470263e-05,
      "loss": 0.0539,
      "step": 450
    },
    {
      "epoch": 0.44660194174757284,
      "grad_norm": 0.628349244594574,
      "learning_rate": 4.990406516337987e-05,
      "loss": 0.0496,
      "step": 460
    },
    {
      "epoch": 0.4563106796116505,
      "grad_norm": 0.7520539164543152,
      "learning_rate": 4.989162741420876e-05,
      "loss": 0.0504,
      "step": 470
    },
    {
      "epoch": 0.46601941747572817,
      "grad_norm": 1.0347453355789185,
      "learning_rate": 4.9878433525014335e-05,
      "loss": 0.0646,
      "step": 480
    },
    {
      "epoch": 0.47572815533980584,
      "grad_norm": 1.04426109790802,
      "learning_rate": 4.9864483896591094e-05,
      "loss": 0.0679,
      "step": 490
    },
    {
      "epoch": 0.4854368932038835,
      "grad_norm": 0.9814009070396423,
      "learning_rate": 4.984977895269087e-05,
      "loss": 0.0607,
      "step": 500
    },
    {
      "epoch": 0.49514563106796117,
      "grad_norm": 0.9229623079299927,
      "learning_rate": 4.983431914000991e-05,
      "loss": 0.0667,
      "step": 510
    },
    {
      "epoch": 0.5048543689320388,
      "grad_norm": 1.0306048393249512,
      "learning_rate": 4.981810492817532e-05,
      "loss": 0.0658,
      "step": 520
    },
    {
      "epoch": 0.5145631067961165,
      "grad_norm": 0.9046841859817505,
      "learning_rate": 4.980113680973082e-05,
      "loss": 0.0746,
      "step": 530
    },
    {
      "epoch": 0.5242718446601942,
      "grad_norm": 0.6024751663208008,
      "learning_rate": 4.978341530012175e-05,
      "loss": 0.0525,
      "step": 540
    },
    {
      "epoch": 0.5339805825242718,
      "grad_norm": 1.0224015712738037,
      "learning_rate": 4.976494093767943e-05,
      "loss": 0.0725,
      "step": 550
    },
    {
      "epoch": 0.5436893203883495,
      "grad_norm": 1.0598809719085693,
      "learning_rate": 4.9745714283604803e-05,
      "loss": 0.0853,
      "step": 560
    },
    {
      "epoch": 0.5533980582524272,
      "grad_norm": 0.6167446970939636,
      "learning_rate": 4.972573592195139e-05,
      "loss": 0.0612,
      "step": 570
    },
    {
      "epoch": 0.5631067961165048,
      "grad_norm": 0.6139333844184875,
      "learning_rate": 4.970500645960756e-05,
      "loss": 0.0525,
      "step": 580
    },
    {
      "epoch": 0.5728155339805825,
      "grad_norm": 0.7935957908630371,
      "learning_rate": 4.968352652627806e-05,
      "loss": 0.0526,
      "step": 590
    },
    {
      "epoch": 0.5825242718446602,
      "grad_norm": 0.5704646706581116,
      "learning_rate": 4.966129677446492e-05,
      "loss": 0.0614,
      "step": 600
    },
    {
      "epoch": 0.5922330097087378,
      "grad_norm": 0.5081838369369507,
      "learning_rate": 4.9638317879447606e-05,
      "loss": 0.0549,
      "step": 610
    },
    {
      "epoch": 0.6019417475728155,
      "grad_norm": 0.5273293256759644,
      "learning_rate": 4.961459053926252e-05,
      "loss": 0.0523,
      "step": 620
    },
    {
      "epoch": 0.6116504854368932,
      "grad_norm": 0.423949271440506,
      "learning_rate": 4.9590115474681816e-05,
      "loss": 0.0571,
      "step": 630
    },
    {
      "epoch": 0.6213592233009708,
      "grad_norm": 0.8299949169158936,
      "learning_rate": 4.956489342919147e-05,
      "loss": 0.0548,
      "step": 640
    },
    {
      "epoch": 0.6310679611650486,
      "grad_norm": 0.8283652067184448,
      "learning_rate": 4.95389251689687e-05,
      "loss": 0.0495,
      "step": 650
    },
    {
      "epoch": 0.6407766990291263,
      "grad_norm": 0.828658938407898,
      "learning_rate": 4.9512211482858714e-05,
      "loss": 0.0642,
      "step": 660
    },
    {
      "epoch": 0.6504854368932039,
      "grad_norm": 0.7801154851913452,
      "learning_rate": 4.948475318235073e-05,
      "loss": 0.0467,
      "step": 670
    },
    {
      "epoch": 0.6601941747572816,
      "grad_norm": 0.7108743786811829,
      "learning_rate": 4.945655110155333e-05,
      "loss": 0.0624,
      "step": 680
    },
    {
      "epoch": 0.6699029126213593,
      "grad_norm": 0.5172573328018188,
      "learning_rate": 4.9427606097169117e-05,
      "loss": 0.0601,
      "step": 690
    },
    {
      "epoch": 0.6796116504854369,
      "grad_norm": 0.7694447636604309,
      "learning_rate": 4.939791904846869e-05,
      "loss": 0.0585,
      "step": 700
    },
    {
      "epoch": 0.6893203883495146,
      "grad_norm": 0.651713490486145,
      "learning_rate": 4.9367490857263944e-05,
      "loss": 0.0521,
      "step": 710
    },
    {
      "epoch": 0.6990291262135923,
      "grad_norm": 1.2775520086288452,
      "learning_rate": 4.9336322447880676e-05,
      "loss": 0.0687,
      "step": 720
    },
    {
      "epoch": 0.7087378640776699,
      "grad_norm": 0.6819589138031006,
      "learning_rate": 4.930441476713049e-05,
      "loss": 0.047,
      "step": 730
    },
    {
      "epoch": 0.7184466019417476,
      "grad_norm": 0.8026332855224609,
      "learning_rate": 4.927176878428206e-05,
      "loss": 0.0419,
      "step": 740
    },
    {
      "epoch": 0.7281553398058253,
      "grad_norm": 0.7379328608512878,
      "learning_rate": 4.923838549103166e-05,
      "loss": 0.0547,
      "step": 750
    },
    {
      "epoch": 0.7378640776699029,
      "grad_norm": 0.5019276738166809,
      "learning_rate": 4.920426590147304e-05,
      "loss": 0.0492,
      "step": 760
    },
    {
      "epoch": 0.7475728155339806,
      "grad_norm": 0.8244302272796631,
      "learning_rate": 4.916941105206666e-05,
      "loss": 0.0542,
      "step": 770
    },
    {
      "epoch": 0.7572815533980582,
      "grad_norm": 0.4618038535118103,
      "learning_rate": 4.9133822001608164e-05,
      "loss": 0.0676,
      "step": 780
    },
    {
      "epoch": 0.7669902912621359,
      "grad_norm": 0.5262795686721802,
      "learning_rate": 4.9097499831196216e-05,
      "loss": 0.0415,
      "step": 790
    },
    {
      "epoch": 0.7766990291262136,
      "grad_norm": 0.537287175655365,
      "learning_rate": 4.906044564419969e-05,
      "loss": 0.0511,
      "step": 800
    },
    {
      "epoch": 0.7864077669902912,
      "grad_norm": 0.9852665066719055,
      "learning_rate": 4.902266056622414e-05,
      "loss": 0.055,
      "step": 810
    },
    {
      "epoch": 0.7961165048543689,
      "grad_norm": 0.6320510506629944,
      "learning_rate": 4.8984145745077584e-05,
      "loss": 0.0554,
      "step": 820
    },
    {
      "epoch": 0.8058252427184466,
      "grad_norm": 0.6188237071037292,
      "learning_rate": 4.894490235073566e-05,
      "loss": 0.0783,
      "step": 830
    },
    {
      "epoch": 0.8155339805825242,
      "grad_norm": 0.6136628985404968,
      "learning_rate": 4.890493157530609e-05,
      "loss": 0.0532,
      "step": 840
    },
    {
      "epoch": 0.8252427184466019,
      "grad_norm": 0.6563002467155457,
      "learning_rate": 4.8864234632992457e-05,
      "loss": 0.0583,
      "step": 850
    },
    {
      "epoch": 0.8349514563106796,
      "grad_norm": 0.779821515083313,
      "learning_rate": 4.88228127600573e-05,
      "loss": 0.0608,
      "step": 860
    },
    {
      "epoch": 0.8446601941747572,
      "grad_norm": 0.3783950209617615,
      "learning_rate": 4.878066721478461e-05,
      "loss": 0.0599,
      "step": 870
    },
    {
      "epoch": 0.8543689320388349,
      "grad_norm": 0.7626412510871887,
      "learning_rate": 4.8737799277441566e-05,
      "loss": 0.0526,
      "step": 880
    },
    {
      "epoch": 0.8640776699029126,
      "grad_norm": 0.5785266160964966,
      "learning_rate": 4.869421025023965e-05,
      "loss": 0.0623,
      "step": 890
    },
    {
      "epoch": 0.8737864077669902,
      "grad_norm": 0.959281861782074,
      "learning_rate": 4.8649901457295096e-05,
      "loss": 0.0509,
      "step": 900
    },
    {
      "epoch": 0.883495145631068,
      "grad_norm": 0.6331967115402222,
      "learning_rate": 4.860487424458867e-05,
      "loss": 0.0498,
      "step": 910
    },
    {
      "epoch": 0.8932038834951457,
      "grad_norm": 0.7830513119697571,
      "learning_rate": 4.8559129979924787e-05,
      "loss": 0.0457,
      "step": 920
    },
    {
      "epoch": 0.9029126213592233,
      "grad_norm": 0.8294515013694763,
      "learning_rate": 4.8512670052889955e-05,
      "loss": 0.0528,
      "step": 930
    },
    {
      "epoch": 0.912621359223301,
      "grad_norm": 0.7075780034065247,
      "learning_rate": 4.846549587481052e-05,
      "loss": 0.0474,
      "step": 940
    },
    {
      "epoch": 0.9223300970873787,
      "grad_norm": 0.4232042729854584,
      "learning_rate": 4.841760887870988e-05,
      "loss": 0.042,
      "step": 950
    },
    {
      "epoch": 0.9320388349514563,
      "grad_norm": 0.6930996179580688,
      "learning_rate": 4.836901051926489e-05,
      "loss": 0.0616,
      "step": 960
    },
    {
      "epoch": 0.941747572815534,
      "grad_norm": 1.1047744750976562,
      "learning_rate": 4.831970227276171e-05,
      "loss": 0.0606,
      "step": 970
    },
    {
      "epoch": 0.9514563106796117,
      "grad_norm": 0.6250813603401184,
      "learning_rate": 4.82696856370509e-05,
      "loss": 0.0419,
      "step": 980
    },
    {
      "epoch": 0.9611650485436893,
      "grad_norm": 0.5649757385253906,
      "learning_rate": 4.8218962131502e-05,
      "loss": 0.0434,
      "step": 990
    },
    {
      "epoch": 0.970873786407767,
      "grad_norm": 0.5142524242401123,
      "learning_rate": 4.81675332969573e-05,
      "loss": 0.0455,
      "step": 1000
    },
    {
      "epoch": 0.9805825242718447,
      "grad_norm": 0.437963604927063,
      "learning_rate": 4.811540069568512e-05,
      "loss": 0.0477,
      "step": 1010
    },
    {
      "epoch": 0.9902912621359223,
      "grad_norm": 0.9865050315856934,
      "learning_rate": 4.8062565911332235e-05,
      "loss": 0.0636,
      "step": 1020
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.4108821749687195,
      "learning_rate": 4.8009030548875896e-05,
      "loss": 0.0386,
      "step": 1030
    },
    {
      "epoch": 1.0097087378640777,
      "grad_norm": 0.40395405888557434,
      "learning_rate": 4.795479623457497e-05,
      "loss": 0.0406,
      "step": 1040
    },
    {
      "epoch": 1.0194174757281553,
      "grad_norm": 0.6007186770439148,
      "learning_rate": 4.789986461592061e-05,
      "loss": 0.05,
      "step": 1050
    },
    {
      "epoch": 1.029126213592233,
      "grad_norm": 0.7217828631401062,
      "learning_rate": 4.784423736158616e-05,
      "loss": 0.051,
      "step": 1060
    },
    {
      "epoch": 1.0388349514563107,
      "grad_norm": 0.6370289325714111,
      "learning_rate": 4.7787916161376515e-05,
      "loss": 0.0641,
      "step": 1070
    },
    {
      "epoch": 1.0485436893203883,
      "grad_norm": 0.6220316886901855,
      "learning_rate": 4.773090272617672e-05,
      "loss": 0.0624,
      "step": 1080
    },
    {
      "epoch": 1.058252427184466,
      "grad_norm": 0.7248903512954712,
      "learning_rate": 4.7673198787900063e-05,
      "loss": 0.0567,
      "step": 1090
    },
    {
      "epoch": 1.0679611650485437,
      "grad_norm": 0.7164710164070129,
      "learning_rate": 4.761480609943546e-05,
      "loss": 0.0424,
      "step": 1100
    },
    {
      "epoch": 1.0776699029126213,
      "grad_norm": 0.6365995407104492,
      "learning_rate": 4.755572643459414e-05,
      "loss": 0.062,
      "step": 1110
    },
    {
      "epoch": 1.087378640776699,
      "grad_norm": 0.5840019583702087,
      "learning_rate": 4.7495961588055836e-05,
      "loss": 0.0344,
      "step": 1120
    },
    {
      "epoch": 1.0970873786407767,
      "grad_norm": 0.6124367713928223,
      "learning_rate": 4.7435513375314253e-05,
      "loss": 0.0524,
      "step": 1130
    },
    {
      "epoch": 1.1067961165048543,
      "grad_norm": 0.6590617299079895,
      "learning_rate": 4.737438363262187e-05,
      "loss": 0.0615,
      "step": 1140
    },
    {
      "epoch": 1.116504854368932,
      "grad_norm": 0.6731277108192444,
      "learning_rate": 4.7312574216934225e-05,
      "loss": 0.0536,
      "step": 1150
    },
    {
      "epoch": 1.1262135922330097,
      "grad_norm": 0.4859127700328827,
      "learning_rate": 4.7250087005853446e-05,
      "loss": 0.0456,
      "step": 1160
    },
    {
      "epoch": 1.1359223300970873,
      "grad_norm": 0.48708751797676086,
      "learning_rate": 4.718692389757128e-05,
      "loss": 0.0499,
      "step": 1170
    },
    {
      "epoch": 1.145631067961165,
      "grad_norm": 0.7826804518699646,
      "learning_rate": 4.7123086810811356e-05,
      "loss": 0.0437,
      "step": 1180
    },
    {
      "epoch": 1.1553398058252426,
      "grad_norm": 0.550841748714447,
      "learning_rate": 4.705857768477098e-05,
      "loss": 0.0507,
      "step": 1190
    },
    {
      "epoch": 1.1650485436893203,
      "grad_norm": 0.7365569472312927,
      "learning_rate": 4.699339847906215e-05,
      "loss": 0.0624,
      "step": 1200
    },
    {
      "epoch": 1.174757281553398,
      "grad_norm": 1.0098744630813599,
      "learning_rate": 4.6927551173652075e-05,
      "loss": 0.0619,
      "step": 1210
    },
    {
      "epoch": 1.1844660194174756,
      "grad_norm": 0.738837718963623,
      "learning_rate": 4.6861037768803016e-05,
      "loss": 0.0503,
      "step": 1220
    },
    {
      "epoch": 1.1941747572815533,
      "grad_norm": 0.5268725752830505,
      "learning_rate": 4.679386028501156e-05,
      "loss": 0.0489,
      "step": 1230
    },
    {
      "epoch": 1.203883495145631,
      "grad_norm": 0.8031307458877563,
      "learning_rate": 4.672602076294714e-05,
      "loss": 0.0595,
      "step": 1240
    },
    {
      "epoch": 1.2135922330097086,
      "grad_norm": 0.6441630721092224,
      "learning_rate": 4.665752126339018e-05,
      "loss": 0.0425,
      "step": 1250
    },
    {
      "epoch": 1.2233009708737863,
      "grad_norm": 0.5169068574905396,
      "learning_rate": 4.658836386716938e-05,
      "loss": 0.0536,
      "step": 1260
    },
    {
      "epoch": 1.233009708737864,
      "grad_norm": 0.6084913015365601,
      "learning_rate": 4.65185506750986e-05,
      "loss": 0.0349,
      "step": 1270
    },
    {
      "epoch": 1.2427184466019416,
      "grad_norm": 0.7514814734458923,
      "learning_rate": 4.6448083807912934e-05,
      "loss": 0.053,
      "step": 1280
    },
    {
      "epoch": 1.2524271844660193,
      "grad_norm": 0.49506863951683044,
      "learning_rate": 4.637696540620441e-05,
      "loss": 0.0497,
      "step": 1290
    },
    {
      "epoch": 1.262135922330097,
      "grad_norm": 0.6199485659599304,
      "learning_rate": 4.630519763035687e-05,
      "loss": 0.0517,
      "step": 1300
    },
    {
      "epoch": 1.2718446601941746,
      "grad_norm": 0.6022672653198242,
      "learning_rate": 4.623278266048039e-05,
      "loss": 0.0439,
      "step": 1310
    },
    {
      "epoch": 1.2815533980582523,
      "grad_norm": 0.7462307810783386,
      "learning_rate": 4.6159722696345045e-05,
      "loss": 0.0454,
      "step": 1320
    },
    {
      "epoch": 1.29126213592233,
      "grad_norm": 0.40218812227249146,
      "learning_rate": 4.608601995731407e-05,
      "loss": 0.0326,
      "step": 1330
    },
    {
      "epoch": 1.3009708737864076,
      "grad_norm": 0.7571138739585876,
      "learning_rate": 4.601167668227648e-05,
      "loss": 0.0364,
      "step": 1340
    },
    {
      "epoch": 1.3106796116504853,
      "grad_norm": 0.77698814868927,
      "learning_rate": 4.593669512957901e-05,
      "loss": 0.058,
      "step": 1350
    },
    {
      "epoch": 1.3203883495145632,
      "grad_norm": 0.7244229316711426,
      "learning_rate": 4.586107757695755e-05,
      "loss": 0.056,
      "step": 1360
    },
    {
      "epoch": 1.3300970873786409,
      "grad_norm": 0.5994431972503662,
      "learning_rate": 4.578482632146793e-05,
      "loss": 0.038,
      "step": 1370
    },
    {
      "epoch": 1.3398058252427185,
      "grad_norm": 0.7894623279571533,
      "learning_rate": 4.570794367941616e-05,
      "loss": 0.0425,
      "step": 1380
    },
    {
      "epoch": 1.3495145631067962,
      "grad_norm": 0.7884004712104797,
      "learning_rate": 4.563043198628806e-05,
      "loss": 0.0495,
      "step": 1390
    },
    {
      "epoch": 1.3592233009708738,
      "grad_norm": 0.6530587673187256,
      "learning_rate": 4.5552293596678294e-05,
      "loss": 0.0414,
      "step": 1400
    },
    {
      "epoch": 1.3689320388349515,
      "grad_norm": 0.8399489521980286,
      "learning_rate": 4.5473530884218886e-05,
      "loss": 0.0493,
      "step": 1410
    },
    {
      "epoch": 1.3786407766990292,
      "grad_norm": 0.6782814860343933,
      "learning_rate": 4.539414624150708e-05,
      "loss": 0.0472,
      "step": 1420
    },
    {
      "epoch": 1.3883495145631068,
      "grad_norm": 0.6846131682395935,
      "learning_rate": 4.5314142080032696e-05,
      "loss": 0.0335,
      "step": 1430
    },
    {
      "epoch": 1.3980582524271845,
      "grad_norm": 0.4258232116699219,
      "learning_rate": 4.5233520830104805e-05,
      "loss": 0.0576,
      "step": 1440
    },
    {
      "epoch": 1.4077669902912622,
      "grad_norm": 0.6838470101356506,
      "learning_rate": 4.515228494077798e-05,
      "loss": 0.0606,
      "step": 1450
    },
    {
      "epoch": 1.4174757281553398,
      "grad_norm": 0.5958576798439026,
      "learning_rate": 4.5070436879777865e-05,
      "loss": 0.0444,
      "step": 1460
    },
    {
      "epoch": 1.4271844660194175,
      "grad_norm": 0.48623862862586975,
      "learning_rate": 4.4987979133426215e-05,
      "loss": 0.0449,
      "step": 1470
    },
    {
      "epoch": 1.4368932038834952,
      "grad_norm": 0.5873615741729736,
      "learning_rate": 4.490491420656537e-05,
      "loss": 0.0442,
      "step": 1480
    },
    {
      "epoch": 1.4466019417475728,
      "grad_norm": 0.4631931781768799,
      "learning_rate": 4.482124462248217e-05,
      "loss": 0.0423,
      "step": 1490
    },
    {
      "epoch": 1.4563106796116505,
      "grad_norm": 0.5900189280509949,
      "learning_rate": 4.473697292283129e-05,
      "loss": 0.0485,
      "step": 1500
    },
    {
      "epoch": 1.4660194174757282,
      "grad_norm": 0.36097192764282227,
      "learning_rate": 4.465210166755803e-05,
      "loss": 0.0364,
      "step": 1510
    },
    {
      "epoch": 1.4757281553398058,
      "grad_norm": 0.6405830383300781,
      "learning_rate": 4.456663343482059e-05,
      "loss": 0.0416,
      "step": 1520
    },
    {
      "epoch": 1.4854368932038835,
      "grad_norm": 0.5862817168235779,
      "learning_rate": 4.44805708209117e-05,
      "loss": 0.0475,
      "step": 1530
    },
    {
      "epoch": 1.4951456310679612,
      "grad_norm": 0.7803816199302673,
      "learning_rate": 4.4393916440179786e-05,
      "loss": 0.0472,
      "step": 1540
    },
    {
      "epoch": 1.5048543689320388,
      "grad_norm": 0.5544291138648987,
      "learning_rate": 4.430667292494955e-05,
      "loss": 0.0361,
      "step": 1550
    },
    {
      "epoch": 1.5145631067961165,
      "grad_norm": 0.5777389407157898,
      "learning_rate": 4.4218842925441966e-05,
      "loss": 0.0439,
      "step": 1560
    },
    {
      "epoch": 1.5242718446601942,
      "grad_norm": 0.6668662428855896,
      "learning_rate": 4.413042910969385e-05,
      "loss": 0.0312,
      "step": 1570
    },
    {
      "epoch": 1.5339805825242718,
      "grad_norm": 0.8618906736373901,
      "learning_rate": 4.404143416347675e-05,
      "loss": 0.0481,
      "step": 1580
    },
    {
      "epoch": 1.5436893203883495,
      "grad_norm": 0.5840690732002258,
      "learning_rate": 4.395186079021537e-05,
      "loss": 0.0485,
      "step": 1590
    },
    {
      "epoch": 1.5533980582524272,
      "grad_norm": 0.3441118001937866,
      "learning_rate": 4.386171171090547e-05,
      "loss": 0.0401,
      "step": 1600
    },
    {
      "epoch": 1.5631067961165048,
      "grad_norm": 0.27396291494369507,
      "learning_rate": 4.37709896640312e-05,
      "loss": 0.0379,
      "step": 1610
    },
    {
      "epoch": 1.5728155339805825,
      "grad_norm": 0.4788952171802521,
      "learning_rate": 4.367969740548189e-05,
      "loss": 0.0398,
      "step": 1620
    },
    {
      "epoch": 1.5825242718446602,
      "grad_norm": 0.6418083906173706,
      "learning_rate": 4.358783770846836e-05,
      "loss": 0.0657,
      "step": 1630
    },
    {
      "epoch": 1.5922330097087378,
      "grad_norm": 0.6678286790847778,
      "learning_rate": 4.349541336343867e-05,
      "loss": 0.0466,
      "step": 1640
    },
    {
      "epoch": 1.6019417475728155,
      "grad_norm": 1.0702563524246216,
      "learning_rate": 4.3402427177993366e-05,
      "loss": 0.0452,
      "step": 1650
    },
    {
      "epoch": 1.6116504854368932,
      "grad_norm": 0.6941045522689819,
      "learning_rate": 4.3308881976800146e-05,
      "loss": 0.0521,
      "step": 1660
    },
    {
      "epoch": 1.6213592233009708,
      "grad_norm": 0.5572572350502014,
      "learning_rate": 4.321478060150813e-05,
      "loss": 0.0336,
      "step": 1670
    },
    {
      "epoch": 1.6310679611650487,
      "grad_norm": 0.5476228594779968,
      "learning_rate": 4.312012591066146e-05,
      "loss": 0.0375,
      "step": 1680
    },
    {
      "epoch": 1.6407766990291264,
      "grad_norm": 0.6187303066253662,
      "learning_rate": 4.302492077961253e-05,
      "loss": 0.041,
      "step": 1690
    },
    {
      "epoch": 1.650485436893204,
      "grad_norm": 0.8198221921920776,
      "learning_rate": 4.292916810043459e-05,
      "loss": 0.0463,
      "step": 1700
    },
    {
      "epoch": 1.6601941747572817,
      "grad_norm": 0.6056732535362244,
      "learning_rate": 4.283287078183392e-05,
      "loss": 0.0494,
      "step": 1710
    },
    {
      "epoch": 1.6699029126213594,
      "grad_norm": 0.6936320662498474,
      "learning_rate": 4.273603174906149e-05,
      "loss": 0.0413,
      "step": 1720
    },
    {
      "epoch": 1.679611650485437,
      "grad_norm": 0.6517147421836853,
      "learning_rate": 4.2638653943824026e-05,
      "loss": 0.0462,
      "step": 1730
    },
    {
      "epoch": 1.6893203883495147,
      "grad_norm": 0.5960354804992676,
      "learning_rate": 4.254074032419474e-05,
      "loss": 0.0403,
      "step": 1740
    },
    {
      "epoch": 1.6990291262135924,
      "grad_norm": 0.5907045602798462,
      "learning_rate": 4.244229386452342e-05,
      "loss": 0.0378,
      "step": 1750
    },
    {
      "epoch": 1.70873786407767,
      "grad_norm": 0.5899525284767151,
      "learning_rate": 4.2343317555346084e-05,
      "loss": 0.0457,
      "step": 1760
    },
    {
      "epoch": 1.7184466019417477,
      "grad_norm": 0.4403168261051178,
      "learning_rate": 4.2243814403294126e-05,
      "loss": 0.0317,
      "step": 1770
    },
    {
      "epoch": 1.7281553398058254,
      "grad_norm": 0.4933370351791382,
      "learning_rate": 4.214378743100302e-05,
      "loss": 0.0497,
      "step": 1780
    },
    {
      "epoch": 1.737864077669903,
      "grad_norm": 0.6774694919586182,
      "learning_rate": 4.204323967702045e-05,
      "loss": 0.0383,
      "step": 1790
    },
    {
      "epoch": 1.7475728155339807,
      "grad_norm": 0.5610571503639221,
      "learning_rate": 4.1942174195714066e-05,
      "loss": 0.0469,
      "step": 1800
    },
    {
      "epoch": 1.7572815533980584,
      "grad_norm": 0.8438130617141724,
      "learning_rate": 4.184059405717863e-05,
      "loss": 0.0478,
      "step": 1810
    },
    {
      "epoch": 1.766990291262136,
      "grad_norm": 0.8638678789138794,
      "learning_rate": 4.173850234714282e-05,
      "loss": 0.0489,
      "step": 1820
    },
    {
      "epoch": 1.7766990291262137,
      "grad_norm": 0.8042917251586914,
      "learning_rate": 4.1635902166875456e-05,
      "loss": 0.0527,
      "step": 1830
    },
    {
      "epoch": 1.7864077669902914,
      "grad_norm": 1.2712494134902954,
      "learning_rate": 4.1532796633091296e-05,
      "loss": 0.0512,
      "step": 1840
    },
    {
      "epoch": 1.796116504854369,
      "grad_norm": 0.5284692645072937,
      "learning_rate": 4.142918887785638e-05,
      "loss": 0.061,
      "step": 1850
    },
    {
      "epoch": 1.8058252427184467,
      "grad_norm": 0.6059673428535461,
      "learning_rate": 4.1325082048492866e-05,
      "loss": 0.0513,
      "step": 1860
    },
    {
      "epoch": 1.8155339805825244,
      "grad_norm": 0.6848878860473633,
      "learning_rate": 4.122047930748343e-05,
      "loss": 0.0395,
      "step": 1870
    },
    {
      "epoch": 1.825242718446602,
      "grad_norm": 0.8983065485954285,
      "learning_rate": 4.1115383832375174e-05,
      "loss": 0.0334,
      "step": 1880
    },
    {
      "epoch": 1.8349514563106797,
      "grad_norm": 0.5672741532325745,
      "learning_rate": 4.100979881568316e-05,
      "loss": 0.0384,
      "step": 1890
    },
    {
      "epoch": 1.8446601941747574,
      "grad_norm": 0.3233070969581604,
      "learning_rate": 4.090372746479337e-05,
      "loss": 0.0416,
      "step": 1900
    },
    {
      "epoch": 1.854368932038835,
      "grad_norm": 0.6278699040412903,
      "learning_rate": 4.0797173001865305e-05,
      "loss": 0.0528,
      "step": 1910
    },
    {
      "epoch": 1.8640776699029127,
      "grad_norm": 0.7250891923904419,
      "learning_rate": 4.069013866373409e-05,
      "loss": 0.0361,
      "step": 1920
    },
    {
      "epoch": 1.8737864077669903,
      "grad_norm": 1.1977667808532715,
      "learning_rate": 4.058262770181217e-05,
      "loss": 0.0455,
      "step": 1930
    },
    {
      "epoch": 1.883495145631068,
      "grad_norm": 0.7614680528640747,
      "learning_rate": 4.0474643381990505e-05,
      "loss": 0.0453,
      "step": 1940
    },
    {
      "epoch": 1.8932038834951457,
      "grad_norm": 0.7502397298812866,
      "learning_rate": 4.036618898453941e-05,
      "loss": 0.0366,
      "step": 1950
    },
    {
      "epoch": 1.9029126213592233,
      "grad_norm": 1.1674546003341675,
      "learning_rate": 4.025726780400886e-05,
      "loss": 0.0536,
      "step": 1960
    },
    {
      "epoch": 1.912621359223301,
      "grad_norm": 0.6499914526939392,
      "learning_rate": 4.0147883149128433e-05,
      "loss": 0.0355,
      "step": 1970
    },
    {
      "epoch": 1.9223300970873787,
      "grad_norm": 0.5907668471336365,
      "learning_rate": 4.003803834270681e-05,
      "loss": 0.0436,
      "step": 1980
    },
    {
      "epoch": 1.9320388349514563,
      "grad_norm": 0.6319796442985535,
      "learning_rate": 3.9927736721530805e-05,
      "loss": 0.0356,
      "step": 1990
    },
    {
      "epoch": 1.941747572815534,
      "grad_norm": 0.49555259943008423,
      "learning_rate": 3.981698163626406e-05,
      "loss": 0.0394,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 6000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}