{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.9971722884265803,
  "eval_steps": 500,
  "global_step": 1236,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "grad_norm": 0.31760987639427185,
      "learning_rate": 5e-06,
      "loss": 3.2917,
      "step": 5
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.24959272146224976,
      "learning_rate": 1e-05,
      "loss": 3.3179,
      "step": 10
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.2924320101737976,
      "learning_rate": 1.5e-05,
      "loss": 3.3109,
      "step": 15
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.22757597267627716,
      "learning_rate": 2e-05,
      "loss": 3.251,
      "step": 20
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1803845316171646,
      "learning_rate": 2.5e-05,
      "loss": 3.2489,
      "step": 25
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.19006718695163727,
      "learning_rate": 3e-05,
      "loss": 3.2527,
      "step": 30
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.17421793937683105,
      "learning_rate": 3.5e-05,
      "loss": 3.225,
      "step": 35
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.16238895058631897,
      "learning_rate": 4e-05,
      "loss": 3.2243,
      "step": 40
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.16933956742286682,
      "learning_rate": 4.5e-05,
      "loss": 3.2182,
      "step": 45
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.16001147031784058,
      "learning_rate": 5e-05,
      "loss": 3.2524,
      "step": 50
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.16399356722831726,
      "learning_rate": 4.999599004988406e-05,
      "loss": 3.2055,
      "step": 55
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.17454692721366882,
      "learning_rate": 4.9983961485912235e-05,
      "loss": 3.2122,
      "step": 60
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.17302922904491425,
      "learning_rate": 4.9963918166799836e-05,
      "loss": 3.2483,
      "step": 65
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.16110074520111084,
      "learning_rate": 4.9935866522363665e-05,
      "loss": 3.2212,
      "step": 70
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.16342906653881073,
      "learning_rate": 4.989981555145928e-05,
      "loss": 3.2017,
      "step": 75
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.16569341719150543,
      "learning_rate": 4.985577681909431e-05,
      "loss": 3.1735,
      "step": 80
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.16199316084384918,
      "learning_rate": 4.9803764452718335e-05,
      "loss": 3.2362,
      "step": 85
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.163334921002388,
      "learning_rate": 4.974379513769093e-05,
      "loss": 3.1873,
      "step": 90
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.17866787314414978,
      "learning_rate": 4.967588811192902e-05,
      "loss": 3.2065,
      "step": 95
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.16487747430801392,
      "learning_rate": 4.960006515973548e-05,
      "loss": 3.2155,
      "step": 100
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1537981629371643,
      "learning_rate": 4.9516350604810793e-05,
      "loss": 3.2032,
      "step": 105
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.16236944496631622,
      "learning_rate": 4.9424771302450084e-05,
      "loss": 3.1732,
      "step": 110
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.18180155754089355,
      "learning_rate": 4.932535663092809e-05,
      "loss": 3.1764,
      "step": 115
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.21276631951332092,
      "learning_rate": 4.921813848207469e-05,
      "loss": 3.1902,
      "step": 120
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.18672068417072296,
      "learning_rate": 4.9103151251044174e-05,
      "loss": 3.1871,
      "step": 125
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.1597897857427597,
      "learning_rate": 4.898043182528136e-05,
      "loss": 3.1487,
      "step": 130
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.1650695651769638,
      "learning_rate": 4.88500195726883e-05,
      "loss": 3.1696,
      "step": 135
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.19672705233097076,
      "learning_rate": 4.871195632899518e-05,
      "loss": 3.1775,
      "step": 140
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.21495744585990906,
      "learning_rate": 4.856628638433962e-05,
      "loss": 3.1691,
      "step": 145
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2001628875732422,
      "learning_rate": 4.841305646905851e-05,
      "loss": 3.1456,
      "step": 150
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.18947800993919373,
      "learning_rate": 4.825231573869721e-05,
      "loss": 3.1269,
      "step": 155
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.17868556082248688,
      "learning_rate": 4.808411575824051e-05,
      "loss": 3.181,
      "step": 160
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.16371792554855347,
      "learning_rate": 4.7908510485570925e-05,
      "loss": 3.145,
      "step": 165
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.16323748230934143,
      "learning_rate": 4.772555625415912e-05,
      "loss": 3.1727,
      "step": 170
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1625365912914276,
      "learning_rate": 4.753531175499243e-05,
      "loss": 3.147,
      "step": 175
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1947379857301712,
      "learning_rate": 4.733783801774696e-05,
      "loss": 3.1401,
      "step": 180
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.19056910276412964,
      "learning_rate": 4.7133198391209566e-05,
      "loss": 3.1175,
      "step": 185
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.184840589761734,
      "learning_rate": 4.692145852295576e-05,
      "loss": 3.1431,
      "step": 190
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.16869348287582397,
      "learning_rate": 4.670268633829031e-05,
      "loss": 3.1332,
      "step": 195
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.18127334117889404,
      "learning_rate": 4.6476952018456974e-05,
      "loss": 3.0935,
      "step": 200
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.19219477474689484,
      "learning_rate": 4.6244327978124734e-05,
      "loss": 3.1763,
      "step": 205
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.17717011272907257,
      "learning_rate": 4.600488884215737e-05,
      "loss": 3.1111,
      "step": 210
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.17769476771354675,
      "learning_rate": 4.5758711421674166e-05,
      "loss": 3.1324,
      "step": 215
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.1706475168466568,
      "learning_rate": 4.55058746894092e-05,
      "loss": 3.1194,
      "step": 220
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.17008131742477417,
      "learning_rate": 4.524645975437717e-05,
      "loss": 3.1486,
      "step": 225
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.19806550443172455,
      "learning_rate": 4.498054983585398e-05,
      "loss": 3.1491,
      "step": 230
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.17164206504821777,
      "learning_rate": 4.470823023668032e-05,
      "loss": 3.1492,
      "step": 235
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.16124162077903748,
      "learning_rate": 4.442958831589686e-05,
      "loss": 3.1251,
      "step": 240
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.17326141893863678,
      "learning_rate": 4.41447134607198e-05,
      "loss": 3.1005,
      "step": 245
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.18267503380775452,
      "learning_rate": 4.3853697057865816e-05,
      "loss": 3.1378,
      "step": 250
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.17902736365795135,
      "learning_rate": 4.3556632464235594e-05,
      "loss": 3.139,
      "step": 255
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.18006114661693573,
      "learning_rate": 4.3253614976965254e-05,
      "loss": 3.1213,
      "step": 260
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.18118861317634583,
      "learning_rate": 4.294474180285545e-05,
      "loss": 3.1044,
      "step": 265
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.17744861543178558,
      "learning_rate": 4.2630112027187816e-05,
      "loss": 3.1343,
      "step": 270
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.16853739321231842,
      "learning_rate": 4.230982658193879e-05,
      "loss": 3.1137,
      "step": 275
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.19506467878818512,
      "learning_rate": 4.1983988213401024e-05,
      "loss": 3.1064,
      "step": 280
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.17228995263576508,
      "learning_rate": 4.165270144922283e-05,
      "loss": 3.0868,
      "step": 285
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.17022188007831573,
      "learning_rate": 4.131607256487606e-05,
      "loss": 3.124,
      "step": 290
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.17359808087348938,
      "learning_rate": 4.097420954956345e-05,
      "loss": 3.0944,
      "step": 295
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.16375815868377686,
      "learning_rate": 4.0627222071576e-05,
      "loss": 3.1295,
      "step": 300
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.18526001274585724,
      "learning_rate": 4.027522144311193e-05,
      "loss": 3.1082,
      "step": 305
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.19401881098747253,
      "learning_rate": 3.991832058456812e-05,
      "loss": 3.0918,
      "step": 310
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.19234879314899445,
      "learning_rate": 3.955663398831569e-05,
      "loss": 3.0422,
      "step": 315
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.18378770351409912,
      "learning_rate": 3.919027768197132e-05,
      "loss": 3.0276,
      "step": 320
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.189378023147583,
      "learning_rate": 3.881936919117607e-05,
      "loss": 3.0004,
      "step": 325
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.2043304294347763,
      "learning_rate": 3.84440275018936e-05,
      "loss": 3.019,
      "step": 330
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.1825193613767624,
      "learning_rate": 3.806437302223992e-05,
      "loss": 3.0234,
      "step": 335
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.20000045001506805,
      "learning_rate": 3.768052754385703e-05,
      "loss": 3.0334,
      "step": 340
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.2001664638519287,
      "learning_rate": 3.7292614202842566e-05,
      "loss": 3.0101,
      "step": 345
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.18154805898666382,
      "learning_rate": 3.690075744024826e-05,
      "loss": 3.0098,
      "step": 350
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.19567397236824036,
      "learning_rate": 3.650508296215978e-05,
      "loss": 3.0064,
      "step": 355
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.19580692052841187,
      "learning_rate": 3.610571769937066e-05,
      "loss": 2.9863,
      "step": 360
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.1916712522506714,
      "learning_rate": 3.5702789766663435e-05,
      "loss": 3.0066,
      "step": 365
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.19139952957630157,
      "learning_rate": 3.529642842171096e-05,
      "loss": 3.0356,
      "step": 370
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.18503962457180023,
      "learning_rate": 3.488676402361101e-05,
      "loss": 2.9687,
      "step": 375
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.18717491626739502,
      "learning_rate": 3.4473927991067636e-05,
      "loss": 3.0223,
      "step": 380
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.19255763292312622,
      "learning_rate": 3.405805276023256e-05,
      "loss": 3.0031,
      "step": 385
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.1963498741388321,
      "learning_rate": 3.363927174222021e-05,
      "loss": 3.0106,
      "step": 390
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.19114218652248383,
      "learning_rate": 3.3217719280309905e-05,
      "loss": 2.9939,
      "step": 395
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.18220435082912445,
      "learning_rate": 3.279353060684914e-05,
      "loss": 2.9843,
      "step": 400
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.19181542098522186,
      "learning_rate": 3.236684179987152e-05,
      "loss": 2.967,
      "step": 405
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.188537135720253,
      "learning_rate": 3.193778973944354e-05,
      "loss": 2.9838,
      "step": 410
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.18813706934452057,
      "learning_rate": 3.150651206375397e-05,
      "loss": 3.0091,
      "step": 415
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.212008997797966,
      "learning_rate": 3.1073147124960034e-05,
      "loss": 2.9995,
      "step": 420
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.22185884416103363,
      "learning_rate": 3.063783394480467e-05,
      "loss": 3.0065,
      "step": 425
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.19503167271614075,
      "learning_rate": 3.0200712170018852e-05,
      "loss": 2.976,
      "step": 430
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.2213529348373413,
      "learning_rate": 2.9761922027523513e-05,
      "loss": 2.9819,
      "step": 435
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.1939724087715149,
      "learning_rate": 2.9321604279445258e-05,
      "loss": 3.0025,
      "step": 440
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.19026894867420197,
      "learning_rate": 2.8879900177960488e-05,
      "loss": 2.9868,
      "step": 445
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.2038724273443222,
      "learning_rate": 2.843695141998225e-05,
      "loss": 2.9699,
      "step": 450
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.19388242065906525,
      "learning_rate": 2.7992900101704406e-05,
      "loss": 3.0146,
      "step": 455
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.21257171034812927,
      "learning_rate": 2.7547888673017775e-05,
      "loss": 2.9556,
      "step": 460
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.204994797706604,
      "learning_rate": 2.7102059891812757e-05,
      "loss": 2.9929,
      "step": 465
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.21233408153057098,
      "learning_rate": 2.6655556778183198e-05,
      "loss": 2.969,
      "step": 470
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.20436938107013702,
      "learning_rate": 2.6208522568546063e-05,
      "loss": 2.9893,
      "step": 475
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.196218803524971,
      "learning_rate": 2.5761100669691817e-05,
      "loss": 2.9813,
      "step": 480
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.20421084761619568,
      "learning_rate": 2.5313434612780084e-05,
      "loss": 2.9651,
      "step": 485
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.18784867227077484,
      "learning_rate": 2.4865668007295394e-05,
      "loss": 2.9521,
      "step": 490
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.22264321148395538,
      "learning_rate": 2.4417944494977888e-05,
      "loss": 2.9846,
      "step": 495
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.20101739466190338,
      "learning_rate": 2.3970407703743568e-05,
      "loss": 2.9799,
      "step": 500
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.20549960434436798,
      "learning_rate": 2.352320120160907e-05,
      "loss": 2.9651,
      "step": 505
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.20113161206245422,
      "learning_rate": 2.3076468450635593e-05,
      "loss": 3.0041,
      "step": 510
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.19988353550434113,
      "learning_rate": 2.263035276090687e-05,
      "loss": 2.9815,
      "step": 515
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.20471665263175964,
      "learning_rate": 2.218499724455584e-05,
      "loss": 3.0102,
      "step": 520
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.1928517073392868,
      "learning_rate": 2.1740544769854852e-05,
      "loss": 2.9728,
      "step": 525
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.19472594559192657,
      "learning_rate": 2.129713791538411e-05,
      "loss": 2.9326,
      "step": 530
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.19601967930793762,
      "learning_rate": 2.0854918924293006e-05,
      "loss": 2.9677,
      "step": 535
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.21098604798316956,
      "learning_rate": 2.041402965866911e-05,
      "loss": 2.9511,
      "step": 540
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.2089085876941681,
      "learning_rate": 1.9974611554029374e-05,
      "loss": 2.9258,
      "step": 545
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.19997884333133698,
      "learning_rate": 1.953680557394816e-05,
      "loss": 2.9502,
      "step": 550
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.19092586636543274,
      "learning_rate": 1.910075216483672e-05,
      "loss": 2.9492,
      "step": 555
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.2013268768787384,
      "learning_rate": 1.8666591210888524e-05,
      "loss": 2.9583,
      "step": 560
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.19930323958396912,
      "learning_rate": 1.8234461989204984e-05,
      "loss": 2.9586,
      "step": 565
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.20546257495880127,
      "learning_rate": 1.7804503125115908e-05,
      "loss": 2.9489,
      "step": 570
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.2034834921360016,
      "learning_rate": 1.7376852547709042e-05,
      "loss": 2.9444,
      "step": 575
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.2054486870765686,
      "learning_rate": 1.6951647445582997e-05,
      "loss": 2.9398,
      "step": 580
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.2075914889574051,
      "learning_rate": 1.652902422283764e-05,
      "loss": 2.9243,
      "step": 585
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.20934872329235077,
      "learning_rate": 1.6109118455316273e-05,
      "loss": 2.9431,
      "step": 590
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.21601444482803345,
      "learning_rate": 1.5692064847113375e-05,
      "loss": 2.9359,
      "step": 595
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.19840656220912933,
      "learning_rate": 1.5277997187362113e-05,
      "loss": 2.9224,
      "step": 600
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.19977068901062012,
      "learning_rate": 1.4867048307315318e-05,
      "loss": 2.9303,
      "step": 605
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.20031294226646423,
      "learning_rate": 1.445935003773371e-05,
      "loss": 2.965,
      "step": 610
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.20860356092453003,
      "learning_rate": 1.4055033166595159e-05,
      "loss": 2.9572,
      "step": 615
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2589782774448395,
      "learning_rate": 1.3654227397138419e-05,
      "loss": 2.9187,
      "step": 620
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.22443072497844696,
      "learning_rate": 1.3257061306254825e-05,
      "loss": 2.9113,
      "step": 625
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.22083857655525208,
      "learning_rate": 1.2863662303241342e-05,
      "loss": 2.8691,
      "step": 630
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.22610703110694885,
      "learning_rate": 1.2474156588928201e-05,
      "loss": 2.8508,
      "step": 635
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.2190137356519699,
      "learning_rate": 1.208866911519414e-05,
      "loss": 2.8914,
      "step": 640
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.22293512523174286,
      "learning_rate": 1.1707323544882348e-05,
      "loss": 2.8692,
      "step": 645
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.21358615159988403,
      "learning_rate": 1.1330242212129946e-05,
      "loss": 2.8271,
      "step": 650
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.21398493647575378,
      "learning_rate": 1.095754608312366e-05,
      "loss": 2.8766,
      "step": 655
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.22630859911441803,
      "learning_rate": 1.0589354717294339e-05,
      "loss": 2.8624,
      "step": 660
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.20600877702236176,
      "learning_rate": 1.0225786228962794e-05,
      "loss": 2.8448,
      "step": 665
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.2191879153251648,
      "learning_rate": 9.866957249449168e-06,
      "loss": 2.8518,
      "step": 670
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.22143206000328064,
      "learning_rate": 9.512982889658118e-06,
      "loss": 2.8477,
      "step": 675
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.2108563482761383,
      "learning_rate": 9.163976703151645e-06,
      "loss": 2.856,
      "step": 680
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.21850666403770447,
      "learning_rate": 8.820050649721581e-06,
      "loss": 2.8622,
      "step": 685
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.2310490608215332,
      "learning_rate": 8.481315059473372e-06,
      "loss": 2.8614,
      "step": 690
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.22149670124053955,
      "learning_rate": 8.14787859743256e-06,
      "loss": 2.8515,
      "step": 695
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.2082003504037857,
      "learning_rate": 7.819848228685511e-06,
      "loss": 2.8315,
      "step": 700
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.23224076628684998,
      "learning_rate": 7.497329184065438e-06,
      "loss": 2.8405,
      "step": 705
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.2128145843744278,
      "learning_rate": 7.1804249263947825e-06,
      "loss": 2.8578,
      "step": 710
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.21818694472312927,
      "learning_rate": 6.869237117294722e-06,
      "loss": 2.8221,
      "step": 715
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.21370700001716614,
      "learning_rate": 6.56386558457254e-06,
      "loss": 2.8116,
      "step": 720
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.2164902687072754,
      "learning_rate": 6.264408290197296e-06,
      "loss": 2.8525,
      "step": 725
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.22312401235103607,
      "learning_rate": 5.9709612988739695e-06,
      "loss": 2.8241,
      "step": 730
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.22373555600643158,
      "learning_rate": 5.683618747226302e-06,
      "loss": 2.8624,
      "step": 735
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.22297130525112152,
      "learning_rate": 5.4024728135981695e-06,
      "loss": 2.8759,
      "step": 740
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.20991022884845734,
      "learning_rate": 5.127613688483099e-06,
      "loss": 2.8545,
      "step": 745
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.23325349390506744,
      "learning_rate": 4.8591295455915396e-06,
      "loss": 2.8478,
      "step": 750
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.20747652649879456,
      "learning_rate": 4.5971065135650775e-06,
      "loss": 2.8373,
      "step": 755
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.2196613997220993,
      "learning_rate": 4.341628648346738e-06,
      "loss": 2.8495,
      "step": 760
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.21464018523693085,
      "learning_rate": 4.092777906216136e-06,
      "loss": 2.8437,
      "step": 765
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.21386373043060303,
      "learning_rate": 3.850634117498248e-06,
      "loss": 2.822,
      "step": 770
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.22314497828483582,
      "learning_rate": 3.615274960954171e-06,
      "loss": 2.8418,
      "step": 775
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.21419712901115417,
      "learning_rate": 3.3867759388620717e-06,
      "loss": 2.8767,
      "step": 780
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.2160976380109787,
      "learning_rate": 3.165210352796355e-06,
      "loss": 2.8508,
      "step": 785
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.22127215564250946,
      "learning_rate": 2.9506492801128216e-06,
      "loss": 2.864,
      "step": 790
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.22035780549049377,
      "learning_rate": 2.743161551147344e-06,
      "loss": 2.8496,
      "step": 795
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.21932028234004974,
      "learning_rate": 2.542813727135343e-06,
      "loss": 2.8862,
      "step": 800
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.22386135160923004,
      "learning_rate": 2.349670078859226e-06,
      "loss": 2.8799,
      "step": 805
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.2204558104276657,
      "learning_rate": 2.1637925660305756e-06,
      "loss": 2.8744,
      "step": 810
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.2307337075471878,
      "learning_rate": 1.9852408174137305e-06,
      "loss": 2.8959,
      "step": 815
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.21677350997924805,
      "learning_rate": 1.8140721116970904e-06,
      "loss": 2.8413,
      "step": 820
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.20689088106155396,
      "learning_rate": 1.6503413591183598e-06,
      "loss": 2.8479,
      "step": 825
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.21706977486610413,
      "learning_rate": 1.494101083849564e-06,
      "loss": 2.8886,
      "step": 830
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.20675326883792877,
      "learning_rate": 1.3454014071475008e-06,
      "loss": 2.8745,
      "step": 835
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.232208713889122,
      "learning_rate": 1.2042900312750265e-06,
      "loss": 2.8619,
      "step": 840
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.22064746916294098,
      "learning_rate": 1.0708122241983949e-06,
      "loss": 2.8327,
      "step": 845
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.22476071119308472,
      "learning_rate": 9.45010805065441e-07,
      "loss": 2.8462,
      "step": 850
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.21655958890914917,
      "learning_rate": 8.269261304693798e-07,
      "loss": 2.858,
      "step": 855
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.2096630185842514,
      "learning_rate": 7.165960815025769e-07,
      "loss": 2.8449,
      "step": 860
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.20629090070724487,
      "learning_rate": 6.14056051604453e-07,
      "loss": 2.8543,
      "step": 865
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.2086755633354187,
      "learning_rate": 5.193389352073836e-07,
      "loss": 2.8563,
      "step": 870
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.2304910272359848,
      "learning_rate": 4.3247511718431663e-07,
      "loss": 2.8676,
      "step": 875
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.21316254138946533,
      "learning_rate": 3.5349246310142914e-07,
      "loss": 2.849,
      "step": 880
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.22124738991260529,
      "learning_rate": 2.8241631027895454e-07,
      "loss": 2.8742,
      "step": 885
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.21196460723876953,
      "learning_rate": 2.1926945966306568e-07,
      "loss": 2.8629,
      "step": 890
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.21042677760124207,
      "learning_rate": 1.6407216851144491e-07,
      "loss": 2.8601,
      "step": 895
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.2190413475036621,
      "learning_rate": 1.1684214389481307e-07,
      "loss": 2.8737,
      "step": 900
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.2288431078195572,
      "learning_rate": 9.009358251619367e-06,
      "loss": 2.8718,
      "step": 905
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.2152530401945114,
      "learning_rate": 8.756246682582565e-06,
      "loss": 2.89,
      "step": 910
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.21837005019187927,
      "learning_rate": 8.505984493817182e-06,
      "loss": 2.8403,
      "step": 915
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.21821348369121552,
      "learning_rate": 8.258615584792079e-06,
      "loss": 2.8578,
      "step": 920
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.22208793461322784,
      "learning_rate": 8.014183347454662e-06,
      "loss": 2.8394,
      "step": 925
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.21861128509044647,
      "learning_rate": 7.77273065861924e-06,
      "loss": 2.8518,
      "step": 930
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.21865125000476837,
      "learning_rate": 7.534299872445855e-06,
      "loss": 2.8264,
      "step": 935
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.22050639986991882,
      "learning_rate": 7.298932813010834e-06,
      "loss": 2.8247,
      "step": 940
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.21507680416107178,
      "learning_rate": 7.066670766970201e-06,
      "loss": 2.8686,
      "step": 945
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.22617888450622559,
      "learning_rate": 6.837554476317412e-06,
      "loss": 2.8339,
      "step": 950
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.22186200320720673,
      "learning_rate": 6.611624131236724e-06,
      "loss": 2.8544,
      "step": 955
    },
    {
      "epoch": 3.11,
      "grad_norm": 0.22832706570625305,
      "learning_rate": 6.388919363053197e-06,
      "loss": 2.8552,
      "step": 960
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.22354218363761902,
      "learning_rate": 6.169479237280823e-06,
      "loss": 2.8238,
      "step": 965
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.2282090187072754,
      "learning_rate": 5.95334224676995e-06,
      "loss": 2.8314,
      "step": 970
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.24048872292041779,
      "learning_rate": 5.740546304955024e-06,
      "loss": 2.8377,
      "step": 975
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.22570078074932098,
      "learning_rate": 5.531128739204072e-06,
      "loss": 2.8314,
      "step": 980
    },
    {
      "epoch": 3.19,
      "grad_norm": 0.22778679430484772,
      "learning_rate": 5.325126284270998e-06,
      "loss": 2.8072,
      "step": 985
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.2253977507352829,
      "learning_rate": 5.1225750758517545e-06,
      "loss": 2.8477,
      "step": 990
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.21220557391643524,
      "learning_rate": 4.923510644245632e-06,
      "loss": 2.8143,
      "step": 995
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.2333599179983139,
      "learning_rate": 4.727967908122788e-06,
      "loss": 2.8508,
      "step": 1000
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.2198331207036972,
      "learning_rate": 4.5359811683990175e-06,
      "loss": 2.8267,
      "step": 1005
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.22090409696102142,
      "learning_rate": 4.3475841022188604e-06,
      "loss": 2.8183,
      "step": 1010
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.21586580574512482,
      "learning_rate": 4.162809757048197e-06,
      "loss": 2.799,
      "step": 1015
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.22138214111328125,
      "learning_rate": 3.981690544877273e-06,
      "loss": 2.867,
      "step": 1020
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.23438501358032227,
      "learning_rate": 3.80425823653513e-06,
      "loss": 2.8302,
      "step": 1025
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.2235979288816452,
      "learning_rate": 3.630543956116614e-06,
      "loss": 2.8301,
      "step": 1030
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.2205328643321991,
      "learning_rate": 3.4605781755227774e-06,
      "loss": 2.8322,
      "step": 1035
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.226404070854187,
      "learning_rate": 3.294390709115622e-06,
      "loss": 2.8633,
      "step": 1040
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.22632290422916412,
      "learning_rate": 3.132010708488317e-06,
      "loss": 2.8269,
      "step": 1045
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.22171452641487122,
      "learning_rate": 2.9734666573515784e-06,
      "loss": 2.8466,
      "step": 1050
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.2134045958518982,
      "learning_rate": 2.8187863665372017e-06,
      "loss": 2.8398,
      "step": 1055
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.22553056478500366,
      "learning_rate": 2.6679969691196904e-06,
      "loss": 2.801,
      "step": 1060
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.23389177024364471,
      "learning_rate": 2.52112491565675e-06,
      "loss": 2.8276,
      "step": 1065
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.22804100811481476,
      "learning_rate": 2.3781959695494267e-06,
      "loss": 2.8394,
      "step": 1070
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.2190062403678894,
      "learning_rate": 2.2392352025229184e-06,
      "loss": 2.831,
      "step": 1075
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.21799859404563904,
      "learning_rate": 2.1042669902286204e-06,
      "loss": 2.8119,
      "step": 1080
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.22122910618782043,
      "learning_rate": 1.973315007968274e-06,
      "loss": 2.8602,
      "step": 1085
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.2231130748987198,
      "learning_rate": 1.846402226541019e-06,
      "loss": 2.8384,
      "step": 1090
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.23154446482658386,
      "learning_rate": 1.723550908214e-06,
      "loss": 2.8427,
      "step": 1095
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.21399781107902527,
      "learning_rate": 1.6047826028172202e-06,
      "loss": 2.8229,
      "step": 1100
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.2319013625383377,
      "learning_rate": 1.4901181439634288e-06,
      "loss": 2.8229,
      "step": 1105
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.22346845269203186,
      "learning_rate": 1.3795776453936287e-06,
      "loss": 2.8075,
      "step": 1110
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.22235344350337982,
      "learning_rate": 1.2731804974487977e-06,
      "loss": 2.8388,
      "step": 1115
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.22748830914497375,
      "learning_rate": 1.1709453636685964e-06,
      "loss": 2.8469,
      "step": 1120
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.2320627123117447,
      "learning_rate": 1.0728901775175272e-06,
      "loss": 2.8171,
      "step": 1125
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.2156900018453598,
      "learning_rate": 9.790321392391027e-07,
      "loss": 2.8332,
      "step": 1130
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.21801438927650452,
      "learning_rate": 8.893877128387462e-07,
      "loss": 2.836,
      "step": 1135
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.23140712082386017,
      "learning_rate": 8.039726231957479e-07,
      "loss": 2.804,
      "step": 1140
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.22428776323795319,
      "learning_rate": 7.228018533048886e-07,
      "loss": 2.8004,
      "step": 1145
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.22678926587104797,
      "learning_rate": 6.45889641648234e-07,
      "loss": 2.8458,
      "step": 1150
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.23222048580646515,
      "learning_rate": 5.732494796975157e-07,
      "loss": 2.853,
      "step": 1155
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.20704413950443268,
      "learning_rate": 5.048941095475074e-07,
      "loss": 2.8043,
      "step": 1160
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.2371576577425003,
      "learning_rate": 4.408355216809129e-07,
      "loss": 2.867,
      "step": 1165
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.21293425559997559,
      "learning_rate": 3.810849528650545e-07,
      "loss": 2.8478,
      "step": 1170
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.25140973925590515,
      "learning_rate": 3.2565288418079096e-07,
      "loss": 2.8324,
      "step": 1175
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.21989835798740387,
      "learning_rate": 2.7454903918399386e-07,
      "loss": 2.8174,
      "step": 1180
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.22354339063167572,
      "learning_rate": 2.2778238219989533e-07,
      "loss": 2.8396,
      "step": 1185
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.23933887481689453,
      "learning_rate": 1.8536111675062084e-07,
      "loss": 2.822,
      "step": 1190
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.21597786247730255,
      "learning_rate": 1.4729268411618202e-07,
      "loss": 2.8396,
      "step": 1195
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.2182658612728119,
      "learning_rate": 1.1358376202916798e-07,
      "loss": 2.8289,
      "step": 1200
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.23224902153015137,
      "learning_rate": 8.42402635033851e-08,
      "loss": 2.8171,
      "step": 1205
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.2257353514432907,
      "learning_rate": 5.926733579661725e-08,
      "loss": 2.7882,
      "step": 1210
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.2192804217338562,
      "learning_rate": 3.866935950774253e-08,
      "loss": 2.8312,
      "step": 1215
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.2221554070711136,
      "learning_rate": 2.2449947808309002e-08,
      "loss": 2.8118,
      "step": 1220
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.2162608802318573,
      "learning_rate": 1.0611945808725043e-08,
      "loss": 2.8038,
      "step": 1225
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.2273118495941162,
      "learning_rate": 3.157430059200217e-09,
      "loss": 2.8124,
      "step": 1230
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.22424130141735077,
      "learning_rate": 8.770818548942128e-11,
      "loss": 2.8155,
      "step": 1235
    },
    {
      "epoch": 4.0,
      "step": 1236,
      "total_flos": 1.3106595430434079e+19,
      "train_loss": 0.7700594385850776,
      "train_runtime": 36129.1454,
      "train_samples_per_second": 4.385,
      "train_steps_per_second": 0.034
    }
  ],
  "logging_steps": 5,
  "max_steps": 1236,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 100,
  "total_flos": 1.3106595430434079e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}