{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9995872884853487,
  "eval_steps": 500,
  "global_step": 1211,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008254230293025176,
      "grad_norm": 24.685088561067097,
      "learning_rate": 8.19672131147541e-08,
      "loss": 1.4397,
      "step": 1
    },
    {
      "epoch": 0.004127115146512587,
      "grad_norm": 23.945153463040807,
      "learning_rate": 4.0983606557377047e-07,
      "loss": 1.4292,
      "step": 5
    },
    {
      "epoch": 0.008254230293025175,
      "grad_norm": 8.201914286275175,
      "learning_rate": 8.196721311475409e-07,
      "loss": 1.3321,
      "step": 10
    },
    {
      "epoch": 0.012381345439537762,
      "grad_norm": 11.180188226746075,
      "learning_rate": 1.2295081967213116e-06,
      "loss": 1.1764,
      "step": 15
    },
    {
      "epoch": 0.01650846058605035,
      "grad_norm": 3.0006985745643777,
      "learning_rate": 1.6393442622950819e-06,
      "loss": 1.0311,
      "step": 20
    },
    {
      "epoch": 0.020635575732562937,
      "grad_norm": 3.0231996454961116,
      "learning_rate": 2.0491803278688526e-06,
      "loss": 0.9824,
      "step": 25
    },
    {
      "epoch": 0.024762690879075525,
      "grad_norm": 2.394417208637789,
      "learning_rate": 2.459016393442623e-06,
      "loss": 0.9428,
      "step": 30
    },
    {
      "epoch": 0.028889806025588115,
      "grad_norm": 2.243762354492006,
      "learning_rate": 2.8688524590163937e-06,
      "loss": 0.929,
      "step": 35
    },
    {
      "epoch": 0.0330169211721007,
      "grad_norm": 2.31711751154106,
      "learning_rate": 3.2786885245901638e-06,
      "loss": 0.9036,
      "step": 40
    },
    {
      "epoch": 0.03714403631861329,
      "grad_norm": 2.717403904295213,
      "learning_rate": 3.6885245901639347e-06,
      "loss": 0.8884,
      "step": 45
    },
    {
      "epoch": 0.041271151465125874,
      "grad_norm": 2.274524441450386,
      "learning_rate": 4.098360655737705e-06,
      "loss": 0.885,
      "step": 50
    },
    {
      "epoch": 0.04539826661163846,
      "grad_norm": 2.457481290190043,
      "learning_rate": 4.508196721311476e-06,
      "loss": 0.8721,
      "step": 55
    },
    {
      "epoch": 0.04952538175815105,
      "grad_norm": 2.3198474831023175,
      "learning_rate": 4.918032786885246e-06,
      "loss": 0.8735,
      "step": 60
    },
    {
      "epoch": 0.053652496904663644,
      "grad_norm": 2.2677876278066753,
      "learning_rate": 5.327868852459017e-06,
      "loss": 0.8592,
      "step": 65
    },
    {
      "epoch": 0.05777961205117623,
      "grad_norm": 2.334132338411567,
      "learning_rate": 5.737704918032787e-06,
      "loss": 0.8573,
      "step": 70
    },
    {
      "epoch": 0.06190672719768882,
      "grad_norm": 2.3114481733901213,
      "learning_rate": 6.147540983606558e-06,
      "loss": 0.8511,
      "step": 75
    },
    {
      "epoch": 0.0660338423442014,
      "grad_norm": 2.445435314197215,
      "learning_rate": 6.5573770491803276e-06,
      "loss": 0.8427,
      "step": 80
    },
    {
      "epoch": 0.070160957490714,
      "grad_norm": 2.3320755043408483,
      "learning_rate": 6.967213114754099e-06,
      "loss": 0.8494,
      "step": 85
    },
    {
      "epoch": 0.07428807263722657,
      "grad_norm": 2.4947952053454916,
      "learning_rate": 7.3770491803278695e-06,
      "loss": 0.8334,
      "step": 90
    },
    {
      "epoch": 0.07841518778373917,
      "grad_norm": 2.4255315812672666,
      "learning_rate": 7.786885245901639e-06,
      "loss": 0.8292,
      "step": 95
    },
    {
      "epoch": 0.08254230293025175,
      "grad_norm": 2.5980903528360586,
      "learning_rate": 8.19672131147541e-06,
      "loss": 0.8182,
      "step": 100
    },
    {
      "epoch": 0.08666941807676434,
      "grad_norm": 2.557969631480717,
      "learning_rate": 8.60655737704918e-06,
      "loss": 0.83,
      "step": 105
    },
    {
      "epoch": 0.09079653322327692,
      "grad_norm": 2.3958016117230363,
      "learning_rate": 9.016393442622952e-06,
      "loss": 0.8132,
      "step": 110
    },
    {
      "epoch": 0.09492364836978952,
      "grad_norm": 2.47094805804508,
      "learning_rate": 9.426229508196723e-06,
      "loss": 0.8058,
      "step": 115
    },
    {
      "epoch": 0.0990507635163021,
      "grad_norm": 2.2173686833192234,
      "learning_rate": 9.836065573770493e-06,
      "loss": 0.8129,
      "step": 120
    },
    {
      "epoch": 0.10317787866281469,
      "grad_norm": 2.442299703694065,
      "learning_rate": 9.999812749151968e-06,
      "loss": 0.8043,
      "step": 125
    },
    {
      "epoch": 0.10730499380932729,
      "grad_norm": 2.70020809980298,
      "learning_rate": 9.998668489204265e-06,
      "loss": 0.8005,
      "step": 130
    },
    {
      "epoch": 0.11143210895583987,
      "grad_norm": 2.4881429411766014,
      "learning_rate": 9.99648423533726e-06,
      "loss": 0.7993,
      "step": 135
    },
    {
      "epoch": 0.11555922410235246,
      "grad_norm": 2.3569754221609953,
      "learning_rate": 9.993260441994116e-06,
      "loss": 0.7966,
      "step": 140
    },
    {
      "epoch": 0.11968633924886504,
      "grad_norm": 2.455448628360433,
      "learning_rate": 9.988997779898544e-06,
      "loss": 0.8019,
      "step": 145
    },
    {
      "epoch": 0.12381345439537764,
      "grad_norm": 2.5006306104937623,
      "learning_rate": 9.983697135915253e-06,
      "loss": 0.7921,
      "step": 150
    },
    {
      "epoch": 0.12794056954189023,
      "grad_norm": 2.5688606957093594,
      "learning_rate": 9.977359612865424e-06,
      "loss": 0.7825,
      "step": 155
    },
    {
      "epoch": 0.1320676846884028,
      "grad_norm": 2.416695522058028,
      "learning_rate": 9.96998652929728e-06,
      "loss": 0.7809,
      "step": 160
    },
    {
      "epoch": 0.1361947998349154,
      "grad_norm": 2.2779935869735732,
      "learning_rate": 9.961579419211741e-06,
      "loss": 0.7767,
      "step": 165
    },
    {
      "epoch": 0.140321914981428,
      "grad_norm": 2.405259985341037,
      "learning_rate": 9.952140031743282e-06,
      "loss": 0.7696,
      "step": 170
    },
    {
      "epoch": 0.14444903012794058,
      "grad_norm": 2.2359965979949186,
      "learning_rate": 9.941670330796007e-06,
      "loss": 0.7571,
      "step": 175
    },
    {
      "epoch": 0.14857614527445315,
      "grad_norm": 2.123914969069781,
      "learning_rate": 9.930172494635058e-06,
      "loss": 0.7545,
      "step": 180
    },
    {
      "epoch": 0.15270326042096574,
      "grad_norm": 2.3685141262823057,
      "learning_rate": 9.917648915433413e-06,
      "loss": 0.7577,
      "step": 185
    },
    {
      "epoch": 0.15683037556747834,
      "grad_norm": 2.288582295953061,
      "learning_rate": 9.904102198774188e-06,
      "loss": 0.752,
      "step": 190
    },
    {
      "epoch": 0.16095749071399093,
      "grad_norm": 2.237056392660144,
      "learning_rate": 9.889535163108537e-06,
      "loss": 0.7439,
      "step": 195
    },
    {
      "epoch": 0.1650846058605035,
      "grad_norm": 2.4463435064141312,
      "learning_rate": 9.873950839169248e-06,
      "loss": 0.7302,
      "step": 200
    },
    {
      "epoch": 0.1692117210070161,
      "grad_norm": 2.255727307108107,
      "learning_rate": 9.857352469340204e-06,
      "loss": 0.7332,
      "step": 205
    },
    {
      "epoch": 0.17333883615352869,
      "grad_norm": 2.315889036709674,
      "learning_rate": 9.839743506981783e-06,
      "loss": 0.7334,
      "step": 210
    },
    {
      "epoch": 0.17746595130004128,
      "grad_norm": 2.4721326878211363,
      "learning_rate": 9.821127615712365e-06,
      "loss": 0.7184,
      "step": 215
    },
    {
      "epoch": 0.18159306644655385,
      "grad_norm": 2.4415976276779277,
      "learning_rate": 9.801508668646117e-06,
      "loss": 0.7149,
      "step": 220
    },
    {
      "epoch": 0.18572018159306644,
      "grad_norm": 2.412349551631901,
      "learning_rate": 9.780890747587163e-06,
      "loss": 0.713,
      "step": 225
    },
    {
      "epoch": 0.18984729673957904,
      "grad_norm": 2.515848530060018,
      "learning_rate": 9.759278142180348e-06,
      "loss": 0.7172,
      "step": 230
    },
    {
      "epoch": 0.19397441188609163,
      "grad_norm": 2.55352651790107,
      "learning_rate": 9.736675349018768e-06,
      "loss": 0.7042,
      "step": 235
    },
    {
      "epoch": 0.1981015270326042,
      "grad_norm": 2.2895880245761515,
      "learning_rate": 9.713087070708225e-06,
      "loss": 0.7109,
      "step": 240
    },
    {
      "epoch": 0.2022286421791168,
      "grad_norm": 2.3671724385797566,
      "learning_rate": 9.688518214888836e-06,
      "loss": 0.7001,
      "step": 245
    },
    {
      "epoch": 0.20635575732562939,
      "grad_norm": 2.3617892749786074,
      "learning_rate": 9.662973893213976e-06,
      "loss": 0.6979,
      "step": 250
    },
    {
      "epoch": 0.21048287247214198,
      "grad_norm": 2.254544557833255,
      "learning_rate": 9.636459420286779e-06,
      "loss": 0.6914,
      "step": 255
    },
    {
      "epoch": 0.21460998761865457,
      "grad_norm": 2.219547321736457,
      "learning_rate": 9.60898031255441e-06,
      "loss": 0.6964,
      "step": 260
    },
    {
      "epoch": 0.21873710276516714,
      "grad_norm": 2.211511469572542,
      "learning_rate": 9.580542287160348e-06,
      "loss": 0.6695,
      "step": 265
    },
    {
      "epoch": 0.22286421791167974,
      "grad_norm": 2.5041061805226787,
      "learning_rate": 9.551151260754907e-06,
      "loss": 0.6864,
      "step": 270
    },
    {
      "epoch": 0.22699133305819233,
      "grad_norm": 2.3511386929231097,
      "learning_rate": 9.520813348264252e-06,
      "loss": 0.6809,
      "step": 275
    },
    {
      "epoch": 0.23111844820470492,
      "grad_norm": 2.960097513048209,
      "learning_rate": 9.489534861618168e-06,
      "loss": 0.6555,
      "step": 280
    },
    {
      "epoch": 0.2352455633512175,
      "grad_norm": 2.2082387429502712,
      "learning_rate": 9.45732230843683e-06,
      "loss": 0.6568,
      "step": 285
    },
    {
      "epoch": 0.23937267849773008,
      "grad_norm": 2.151410227639591,
      "learning_rate": 9.424182390676872e-06,
      "loss": 0.6645,
      "step": 290
    },
    {
      "epoch": 0.24349979364424268,
      "grad_norm": 2.2209736960884334,
      "learning_rate": 9.39012200323701e-06,
      "loss": 0.6705,
      "step": 295
    },
    {
      "epoch": 0.24762690879075527,
      "grad_norm": 2.414842272352699,
      "learning_rate": 9.355148232523538e-06,
      "loss": 0.6496,
      "step": 300
    },
    {
      "epoch": 0.25175402393726787,
      "grad_norm": 2.5739956654811236,
      "learning_rate": 9.319268354975958e-06,
      "loss": 0.657,
      "step": 305
    },
    {
      "epoch": 0.25588113908378046,
      "grad_norm": 2.618163978298313,
      "learning_rate": 9.282489835553106e-06,
      "loss": 0.6496,
      "step": 310
    },
    {
      "epoch": 0.260008254230293,
      "grad_norm": 2.281248176869453,
      "learning_rate": 9.244820326180013e-06,
      "loss": 0.6563,
      "step": 315
    },
    {
      "epoch": 0.2641353693768056,
      "grad_norm": 2.297254363064407,
      "learning_rate": 9.206267664155906e-06,
      "loss": 0.6432,
      "step": 320
    },
    {
      "epoch": 0.2682624845233182,
      "grad_norm": 2.3343424077594306,
      "learning_rate": 9.166839870523627e-06,
      "loss": 0.6352,
      "step": 325
    },
    {
      "epoch": 0.2723895996698308,
      "grad_norm": 2.3001895352129904,
      "learning_rate": 9.126545148400807e-06,
      "loss": 0.6346,
      "step": 330
    },
    {
      "epoch": 0.2765167148163434,
      "grad_norm": 2.335491244315102,
      "learning_rate": 9.085391881273182e-06,
      "loss": 0.6417,
      "step": 335
    },
    {
      "epoch": 0.280643829962856,
      "grad_norm": 2.8711945935898986,
      "learning_rate": 9.04338863125038e-06,
      "loss": 0.6203,
      "step": 340
    },
    {
      "epoch": 0.28477094510936857,
      "grad_norm": 2.4481124095885964,
      "learning_rate": 9.00054413728452e-06,
      "loss": 0.6253,
      "step": 345
    },
    {
      "epoch": 0.28889806025588116,
      "grad_norm": 2.208108640856046,
      "learning_rate": 8.956867313352055e-06,
      "loss": 0.6246,
      "step": 350
    },
    {
      "epoch": 0.2930251754023937,
      "grad_norm": 2.2861750757235995,
      "learning_rate": 8.912367246599175e-06,
      "loss": 0.6152,
      "step": 355
    },
    {
      "epoch": 0.2971522905489063,
      "grad_norm": 2.548396902849064,
      "learning_rate": 8.867053195451182e-06,
      "loss": 0.6262,
      "step": 360
    },
    {
      "epoch": 0.3012794056954189,
      "grad_norm": 2.305963742549307,
      "learning_rate": 8.820934587686247e-06,
      "loss": 0.6151,
      "step": 365
    },
    {
      "epoch": 0.3054065208419315,
      "grad_norm": 2.4126232699277157,
      "learning_rate": 8.77402101847391e-06,
      "loss": 0.6071,
      "step": 370
    },
    {
      "epoch": 0.3095336359884441,
      "grad_norm": 2.2092911857967867,
      "learning_rate": 8.726322248378775e-06,
      "loss": 0.6144,
      "step": 375
    },
    {
      "epoch": 0.3136607511349567,
      "grad_norm": 2.2414757536483125,
      "learning_rate": 8.677848201329775e-06,
      "loss": 0.5971,
      "step": 380
    },
    {
      "epoch": 0.31778786628146927,
      "grad_norm": 2.1301302453717423,
      "learning_rate": 8.628608962555467e-06,
      "loss": 0.6171,
      "step": 385
    },
    {
      "epoch": 0.32191498142798186,
      "grad_norm": 2.1138762246320373,
      "learning_rate": 8.578614776485744e-06,
      "loss": 0.5913,
      "step": 390
    },
    {
      "epoch": 0.32604209657449446,
      "grad_norm": 2.2822356894428277,
      "learning_rate": 8.527876044620453e-06,
      "loss": 0.5967,
      "step": 395
    },
    {
      "epoch": 0.330169211721007,
      "grad_norm": 2.241849162079401,
      "learning_rate": 8.47640332336531e-06,
      "loss": 0.6042,
      "step": 400
    },
    {
      "epoch": 0.3342963268675196,
      "grad_norm": 2.2294939629300097,
      "learning_rate": 8.424207321835598e-06,
      "loss": 0.5824,
      "step": 405
    },
    {
      "epoch": 0.3384234420140322,
      "grad_norm": 2.2031216423678543,
      "learning_rate": 8.371298899628091e-06,
      "loss": 0.5778,
      "step": 410
    },
    {
      "epoch": 0.3425505571605448,
      "grad_norm": 2.33426131831025,
      "learning_rate": 8.317689064561671e-06,
      "loss": 0.5846,
      "step": 415
    },
    {
      "epoch": 0.34667767230705737,
      "grad_norm": 2.3437558474733895,
      "learning_rate": 8.263388970387102e-06,
      "loss": 0.5752,
      "step": 420
    },
    {
      "epoch": 0.35080478745356997,
      "grad_norm": 2.194788845362847,
      "learning_rate": 8.20840991446645e-06,
      "loss": 0.5796,
      "step": 425
    },
    {
      "epoch": 0.35493190260008256,
      "grad_norm": 2.3039527642284776,
      "learning_rate": 8.152763335422612e-06,
      "loss": 0.5768,
      "step": 430
    },
    {
      "epoch": 0.35905901774659515,
      "grad_norm": 2.271949627752587,
      "learning_rate": 8.096460810759473e-06,
      "loss": 0.5745,
      "step": 435
    },
    {
      "epoch": 0.3631861328931077,
      "grad_norm": 2.299369126451729,
      "learning_rate": 8.03951405445314e-06,
      "loss": 0.5629,
      "step": 440
    },
    {
      "epoch": 0.3673132480396203,
      "grad_norm": 2.701487019844598,
      "learning_rate": 7.98193491451483e-06,
      "loss": 0.5682,
      "step": 445
    },
    {
      "epoch": 0.3714403631861329,
      "grad_norm": 2.329511707999701,
      "learning_rate": 7.923735370525809e-06,
      "loss": 0.5741,
      "step": 450
    },
    {
      "epoch": 0.3755674783326455,
      "grad_norm": 2.2666378605955155,
      "learning_rate": 7.864927531145012e-06,
      "loss": 0.5613,
      "step": 455
    },
    {
      "epoch": 0.37969459347915807,
      "grad_norm": 2.44812863797748,
      "learning_rate": 7.805523631589775e-06,
      "loss": 0.5654,
      "step": 460
    },
    {
      "epoch": 0.38382170862567067,
      "grad_norm": 2.346335768826544,
      "learning_rate": 7.745536031090252e-06,
      "loss": 0.5577,
      "step": 465
    },
    {
      "epoch": 0.38794882377218326,
      "grad_norm": 2.340018514786721,
      "learning_rate": 7.684977210318024e-06,
      "loss": 0.5575,
      "step": 470
    },
    {
      "epoch": 0.39207593891869585,
      "grad_norm": 2.250185737483932,
      "learning_rate": 7.623859768789441e-06,
      "loss": 0.5456,
      "step": 475
    },
    {
      "epoch": 0.3962030540652084,
      "grad_norm": 2.1535843417391938,
      "learning_rate": 7.5621964222442455e-06,
      "loss": 0.5504,
      "step": 480
    },
    {
      "epoch": 0.400330169211721,
      "grad_norm": 2.279014812989976,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.5418,
      "step": 485
    },
    {
      "epoch": 0.4044572843582336,
      "grad_norm": 2.248120170776638,
      "learning_rate": 7.437283442282903e-06,
      "loss": 0.5299,
      "step": 490
    },
    {
      "epoch": 0.4085843995047462,
      "grad_norm": 2.1290912998771443,
      "learning_rate": 7.374059797535516e-06,
      "loss": 0.5382,
      "step": 495
    },
    {
      "epoch": 0.41271151465125877,
      "grad_norm": 2.2318811155096325,
      "learning_rate": 7.310342219701981e-06,
      "loss": 0.5386,
      "step": 500
    },
    {
      "epoch": 0.41683862979777137,
      "grad_norm": 2.0172840744038445,
      "learning_rate": 7.246143965491288e-06,
      "loss": 0.532,
      "step": 505
    },
    {
      "epoch": 0.42096574494428396,
      "grad_norm": 2.0888969433680953,
      "learning_rate": 7.181478391619162e-06,
      "loss": 0.5205,
      "step": 510
    },
    {
      "epoch": 0.42509286009079655,
      "grad_norm": 2.15061141359648,
      "learning_rate": 7.11635895202914e-06,
      "loss": 0.5311,
      "step": 515
    },
    {
      "epoch": 0.42921997523730915,
      "grad_norm": 2.2686115013098704,
      "learning_rate": 7.050799195093421e-06,
      "loss": 0.5301,
      "step": 520
    },
    {
      "epoch": 0.4333470903838217,
      "grad_norm": 2.341272332235732,
      "learning_rate": 6.984812760794078e-06,
      "loss": 0.5362,
      "step": 525
    },
    {
      "epoch": 0.4374742055303343,
      "grad_norm": 2.166891534024358,
      "learning_rate": 6.918413377885193e-06,
      "loss": 0.5293,
      "step": 530
    },
    {
      "epoch": 0.4416013206768469,
      "grad_norm": 2.2412320223529134,
      "learning_rate": 6.851614861036533e-06,
      "loss": 0.5283,
      "step": 535
    },
    {
      "epoch": 0.44572843582335947,
      "grad_norm": 2.2162169091736166,
      "learning_rate": 6.78443110795936e-06,
      "loss": 0.5187,
      "step": 540
    },
    {
      "epoch": 0.44985555096987206,
      "grad_norm": 2.1687285736037447,
      "learning_rate": 6.716876096514944e-06,
      "loss": 0.5208,
      "step": 545
    },
    {
      "epoch": 0.45398266611638466,
      "grad_norm": 2.1587915614775297,
      "learning_rate": 6.648963881806411e-06,
      "loss": 0.5058,
      "step": 550
    },
    {
      "epoch": 0.45810978126289725,
      "grad_norm": 2.05653346463343,
      "learning_rate": 6.580708593254526e-06,
      "loss": 0.5103,
      "step": 555
    },
    {
      "epoch": 0.46223689640940985,
      "grad_norm": 2.1531395501548642,
      "learning_rate": 6.512124431658006e-06,
      "loss": 0.5103,
      "step": 560
    },
    {
      "epoch": 0.4663640115559224,
      "grad_norm": 2.176090356825283,
      "learning_rate": 6.443225666238976e-06,
      "loss": 0.5062,
      "step": 565
    },
    {
      "epoch": 0.470491126702435,
      "grad_norm": 2.209786359023677,
      "learning_rate": 6.37402663167421e-06,
      "loss": 0.5027,
      "step": 570
    },
    {
      "epoch": 0.4746182418489476,
      "grad_norm": 2.056272842336906,
      "learning_rate": 6.304541725112734e-06,
      "loss": 0.5056,
      "step": 575
    },
    {
      "epoch": 0.47874535699546017,
      "grad_norm": 2.162812572951533,
      "learning_rate": 6.234785403180438e-06,
      "loss": 0.4928,
      "step": 580
    },
    {
      "epoch": 0.48287247214197276,
      "grad_norm": 2.211681052073838,
      "learning_rate": 6.164772178972321e-06,
      "loss": 0.4938,
      "step": 585
    },
    {
      "epoch": 0.48699958728848536,
      "grad_norm": 2.0594553618227613,
      "learning_rate": 6.094516619032975e-06,
      "loss": 0.5024,
      "step": 590
    },
    {
      "epoch": 0.49112670243499795,
      "grad_norm": 2.1196964005402803,
      "learning_rate": 6.024033340325954e-06,
      "loss": 0.4877,
      "step": 595
    },
    {
      "epoch": 0.49525381758151055,
      "grad_norm": 2.0731282875409787,
      "learning_rate": 5.953337007192659e-06,
      "loss": 0.4903,
      "step": 600
    },
    {
      "epoch": 0.4993809327280231,
      "grad_norm": 2.072327686896433,
      "learning_rate": 5.882442328301356e-06,
      "loss": 0.4814,
      "step": 605
    },
    {
      "epoch": 0.5035080478745357,
      "grad_norm": 2.118695720658721,
      "learning_rate": 5.811364053586973e-06,
      "loss": 0.4766,
      "step": 610
    },
    {
      "epoch": 0.5076351630210483,
      "grad_norm": 2.119885860879392,
      "learning_rate": 5.740116971182322e-06,
      "loss": 0.4777,
      "step": 615
    },
    {
      "epoch": 0.5117622781675609,
      "grad_norm": 2.073375813482556,
      "learning_rate": 5.668715904341365e-06,
      "loss": 0.4734,
      "step": 620
    },
    {
      "epoch": 0.5158893933140735,
      "grad_norm": 2.0709791028178572,
      "learning_rate": 5.5971757083551625e-06,
      "loss": 0.4707,
      "step": 625
    },
    {
      "epoch": 0.520016508460586,
      "grad_norm": 2.0492205452436396,
      "learning_rate": 5.525511267461186e-06,
      "loss": 0.4736,
      "step": 630
    },
    {
      "epoch": 0.5241436236070987,
      "grad_norm": 2.0319046135786953,
      "learning_rate": 5.453737491746572e-06,
      "loss": 0.4713,
      "step": 635
    },
    {
      "epoch": 0.5282707387536112,
      "grad_norm": 2.0892666680641327,
      "learning_rate": 5.381869314046031e-06,
      "loss": 0.4764,
      "step": 640
    },
    {
      "epoch": 0.5323978539001238,
      "grad_norm": 2.078568518806719,
      "learning_rate": 5.3099216868349965e-06,
      "loss": 0.4671,
      "step": 645
    },
    {
      "epoch": 0.5365249690466364,
      "grad_norm": 2.0410694490990724,
      "learning_rate": 5.237909579118713e-06,
      "loss": 0.4778,
      "step": 650
    },
    {
      "epoch": 0.540652084193149,
      "grad_norm": 2.0761769744117444,
      "learning_rate": 5.165847973317854e-06,
      "loss": 0.4678,
      "step": 655
    },
    {
      "epoch": 0.5447791993396616,
      "grad_norm": 2.130099579144156,
      "learning_rate": 5.093751862151388e-06,
      "loss": 0.4649,
      "step": 660
    },
    {
      "epoch": 0.5489063144861742,
      "grad_norm": 2.078379103493414,
      "learning_rate": 5.021636245517261e-06,
      "loss": 0.4732,
      "step": 665
    },
    {
      "epoch": 0.5530334296326868,
      "grad_norm": 2.0058829746015507,
      "learning_rate": 4.9495161273716174e-06,
      "loss": 0.4566,
      "step": 670
    },
    {
      "epoch": 0.5571605447791993,
      "grad_norm": 2.22400636300283,
      "learning_rate": 4.877406512607159e-06,
      "loss": 0.4572,
      "step": 675
    },
    {
      "epoch": 0.561287659925712,
      "grad_norm": 2.115628462513928,
      "learning_rate": 4.805322403931312e-06,
      "loss": 0.4599,
      "step": 680
    },
    {
      "epoch": 0.5654147750722245,
      "grad_norm": 2.0907681450444353,
      "learning_rate": 4.73327879874486e-06,
      "loss": 0.4647,
      "step": 685
    },
    {
      "epoch": 0.5695418902187371,
      "grad_norm": 2.203495989612946,
      "learning_rate": 4.661290686021661e-06,
      "loss": 0.4631,
      "step": 690
    },
    {
      "epoch": 0.5736690053652497,
      "grad_norm": 2.0602084077093235,
      "learning_rate": 4.589373043190137e-06,
      "loss": 0.4596,
      "step": 695
    },
    {
      "epoch": 0.5777961205117623,
      "grad_norm": 2.08958806455361,
      "learning_rate": 4.517540833017152e-06,
      "loss": 0.4481,
      "step": 700
    },
    {
      "epoch": 0.5819232356582749,
      "grad_norm": 2.0835614126353112,
      "learning_rate": 4.445809000494945e-06,
      "loss": 0.4457,
      "step": 705
    },
    {
      "epoch": 0.5860503508047874,
      "grad_norm": 2.1121573908487057,
      "learning_rate": 4.374192469731771e-06,
      "loss": 0.4485,
      "step": 710
    },
    {
      "epoch": 0.5901774659513,
      "grad_norm": 2.0348765238388307,
      "learning_rate": 4.302706140846864e-06,
      "loss": 0.4531,
      "step": 715
    },
    {
      "epoch": 0.5943045810978126,
      "grad_norm": 2.097167725695452,
      "learning_rate": 4.231364886870417e-06,
      "loss": 0.432,
      "step": 720
    },
    {
      "epoch": 0.5984316962443252,
      "grad_norm": 2.035634145828901,
      "learning_rate": 4.160183550649176e-06,
      "loss": 0.4542,
      "step": 725
    },
    {
      "epoch": 0.6025588113908378,
      "grad_norm": 2.083465204457464,
      "learning_rate": 4.089176941758332e-06,
      "loss": 0.4373,
      "step": 730
    },
    {
      "epoch": 0.6066859265373504,
      "grad_norm": 2.302682902657171,
      "learning_rate": 4.018359833420323e-06,
      "loss": 0.4355,
      "step": 735
    },
    {
      "epoch": 0.610813041683863,
      "grad_norm": 2.1209322204861167,
      "learning_rate": 3.9477469594311975e-06,
      "loss": 0.4382,
      "step": 740
    },
    {
      "epoch": 0.6149401568303756,
      "grad_norm": 2.1130799454748694,
      "learning_rate": 3.8773530110952e-06,
      "loss": 0.4303,
      "step": 745
    },
    {
      "epoch": 0.6190672719768882,
      "grad_norm": 2.067330224712638,
      "learning_rate": 3.807192634168168e-06,
      "loss": 0.4282,
      "step": 750
    },
    {
      "epoch": 0.6231943871234007,
      "grad_norm": 2.121507686505271,
      "learning_rate": 3.7372804258104367e-06,
      "loss": 0.4398,
      "step": 755
    },
    {
      "epoch": 0.6273215022699133,
      "grad_norm": 2.081955763632376,
      "learning_rate": 3.667630931549826e-06,
      "loss": 0.4321,
      "step": 760
    },
    {
      "epoch": 0.6314486174164259,
      "grad_norm": 2.0671194356029936,
      "learning_rate": 3.598258642255387e-06,
      "loss": 0.4297,
      "step": 765
    },
    {
      "epoch": 0.6355757325629385,
      "grad_norm": 2.0745096928290274,
      "learning_rate": 3.529177991122519e-06,
      "loss": 0.4291,
      "step": 770
    },
    {
      "epoch": 0.6397028477094511,
      "grad_norm": 2.04047679569905,
      "learning_rate": 3.460403350670077e-06,
      "loss": 0.4234,
      "step": 775
    },
    {
      "epoch": 0.6438299628559637,
      "grad_norm": 2.0184632129343254,
      "learning_rate": 3.3919490297501167e-06,
      "loss": 0.4322,
      "step": 780
    },
    {
      "epoch": 0.6479570780024763,
      "grad_norm": 2.0467555064446983,
      "learning_rate": 3.3238292705708675e-06,
      "loss": 0.423,
      "step": 785
    },
    {
      "epoch": 0.6520841931489889,
      "grad_norm": 2.15202224763677,
      "learning_rate": 3.256058245733592e-06,
      "loss": 0.4189,
      "step": 790
    },
    {
      "epoch": 0.6562113082955014,
      "grad_norm": 1.9966574895688705,
      "learning_rate": 3.1886500552839105e-06,
      "loss": 0.4199,
      "step": 795
    },
    {
      "epoch": 0.660338423442014,
      "grad_norm": 2.0485705146360753,
      "learning_rate": 3.121618723778225e-06,
      "loss": 0.4106,
      "step": 800
    },
    {
      "epoch": 0.6644655385885266,
      "grad_norm": 1.939157035039348,
      "learning_rate": 3.054978197365861e-06,
      "loss": 0.4164,
      "step": 805
    },
    {
      "epoch": 0.6685926537350392,
      "grad_norm": 1.9248864355584618,
      "learning_rate": 2.9887423408875056e-06,
      "loss": 0.4082,
      "step": 810
    },
    {
      "epoch": 0.6727197688815518,
      "grad_norm": 2.0698662946290423,
      "learning_rate": 2.9229249349905686e-06,
      "loss": 0.415,
      "step": 815
    },
    {
      "epoch": 0.6768468840280644,
      "grad_norm": 2.1520358829070094,
      "learning_rate": 2.8575396732620673e-06,
      "loss": 0.4003,
      "step": 820
    },
    {
      "epoch": 0.680973999174577,
      "grad_norm": 1.9676814885778169,
      "learning_rate": 2.7926001593796114e-06,
      "loss": 0.4173,
      "step": 825
    },
    {
      "epoch": 0.6851011143210896,
      "grad_norm": 2.0886545446660363,
      "learning_rate": 2.728119904281105e-06,
      "loss": 0.4174,
      "step": 830
    },
    {
      "epoch": 0.6892282294676021,
      "grad_norm": 2.1021768729202637,
      "learning_rate": 2.6641123233537395e-06,
      "loss": 0.4009,
      "step": 835
    },
    {
      "epoch": 0.6933553446141147,
      "grad_norm": 2.164281085927058,
      "learning_rate": 2.6005907336428615e-06,
      "loss": 0.4017,
      "step": 840
    },
    {
      "epoch": 0.6974824597606273,
      "grad_norm": 2.092014051056356,
      "learning_rate": 2.537568351081311e-06,
      "loss": 0.4046,
      "step": 845
    },
    {
      "epoch": 0.7016095749071399,
      "grad_norm": 1.9368454491782257,
      "learning_rate": 2.4750582877397933e-06,
      "loss": 0.4073,
      "step": 850
    },
    {
      "epoch": 0.7057366900536525,
      "grad_norm": 2.019204414322489,
      "learning_rate": 2.4130735490988457e-06,
      "loss": 0.4019,
      "step": 855
    },
    {
      "epoch": 0.7098638052001651,
      "grad_norm": 1.9404031254190708,
      "learning_rate": 2.3516270313430085e-06,
      "loss": 0.3965,
      "step": 860
    },
    {
      "epoch": 0.7139909203466777,
      "grad_norm": 1.9803275733019916,
      "learning_rate": 2.2907315186777057e-06,
      "loss": 0.4009,
      "step": 865
    },
    {
      "epoch": 0.7181180354931903,
      "grad_norm": 1.9105419735279907,
      "learning_rate": 2.230399680669449e-06,
      "loss": 0.3925,
      "step": 870
    },
    {
      "epoch": 0.7222451506397028,
      "grad_norm": 2.11701521254791,
      "learning_rate": 2.170644069609876e-06,
      "loss": 0.3878,
      "step": 875
    },
    {
      "epoch": 0.7263722657862154,
      "grad_norm": 2.009828108641354,
      "learning_rate": 2.1114771179041894e-06,
      "loss": 0.3939,
      "step": 880
    },
    {
      "epoch": 0.730499380932728,
      "grad_norm": 2.0525311494906417,
      "learning_rate": 2.052911135484551e-06,
      "loss": 0.3983,
      "step": 885
    },
    {
      "epoch": 0.7346264960792406,
      "grad_norm": 2.0024645682646214,
      "learning_rate": 1.9949583072489455e-06,
      "loss": 0.3991,
      "step": 890
    },
    {
      "epoch": 0.7387536112257532,
      "grad_norm": 1.983708310096234,
      "learning_rate": 1.9376306905260604e-06,
      "loss": 0.3908,
      "step": 895
    },
    {
      "epoch": 0.7428807263722658,
      "grad_norm": 1.9201326193885593,
      "learning_rate": 1.8809402125667064e-06,
      "loss": 0.3894,
      "step": 900
    },
    {
      "epoch": 0.7470078415187784,
      "grad_norm": 2.0461993333856068,
      "learning_rate": 1.8248986680623077e-06,
      "loss": 0.3845,
      "step": 905
    },
    {
      "epoch": 0.751134956665291,
      "grad_norm": 1.9895671478806258,
      "learning_rate": 1.7695177166909672e-06,
      "loss": 0.3924,
      "step": 910
    },
    {
      "epoch": 0.7552620718118036,
      "grad_norm": 2.057619939386755,
      "learning_rate": 1.7148088806916114e-06,
      "loss": 0.3895,
      "step": 915
    },
    {
      "epoch": 0.7593891869583161,
      "grad_norm": 2.0151906155411354,
      "learning_rate": 1.6607835424667578e-06,
      "loss": 0.3865,
      "step": 920
    },
    {
      "epoch": 0.7635163021048287,
      "grad_norm": 1.925068927133916,
      "learning_rate": 1.6074529422143398e-06,
      "loss": 0.3852,
      "step": 925
    },
    {
      "epoch": 0.7676434172513413,
      "grad_norm": 2.0136796449917256,
      "learning_rate": 1.554828175589151e-06,
      "loss": 0.3787,
      "step": 930
    },
    {
      "epoch": 0.7717705323978539,
      "grad_norm": 1.9559990017194175,
      "learning_rate": 1.5029201913943425e-06,
      "loss": 0.3796,
      "step": 935
    },
    {
      "epoch": 0.7758976475443665,
      "grad_norm": 2.1521890680893585,
      "learning_rate": 1.4517397893034718e-06,
      "loss": 0.3842,
      "step": 940
    },
    {
      "epoch": 0.7800247626908791,
      "grad_norm": 2.0029729095922213,
      "learning_rate": 1.4012976176135978e-06,
      "loss": 0.393,
      "step": 945
    },
    {
      "epoch": 0.7841518778373917,
      "grad_norm": 2.021018024866405,
      "learning_rate": 1.35160417102985e-06,
      "loss": 0.3873,
      "step": 950
    },
    {
      "epoch": 0.7882789929839042,
      "grad_norm": 1.9835697101743557,
      "learning_rate": 1.3026697884819644e-06,
      "loss": 0.3861,
      "step": 955
    },
    {
      "epoch": 0.7924061081304168,
      "grad_norm": 1.9340723815291019,
      "learning_rate": 1.25450465097322e-06,
      "loss": 0.3903,
      "step": 960
    },
    {
      "epoch": 0.7965332232769294,
      "grad_norm": 2.1191758392362035,
      "learning_rate": 1.207118779462248e-06,
      "loss": 0.3749,
      "step": 965
    },
    {
      "epoch": 0.800660338423442,
      "grad_norm": 2.132380601853603,
      "learning_rate": 1.160522032778123e-06,
      "loss": 0.3782,
      "step": 970
    },
    {
      "epoch": 0.8047874535699546,
      "grad_norm": 1.9708201131489755,
      "learning_rate": 1.1147241055691909e-06,
      "loss": 0.3705,
      "step": 975
    },
    {
      "epoch": 0.8089145687164672,
      "grad_norm": 2.01028167948197,
      "learning_rate": 1.0697345262860638e-06,
      "loss": 0.3742,
      "step": 980
    },
    {
      "epoch": 0.8130416838629798,
      "grad_norm": 1.9991004397644925,
      "learning_rate": 1.0255626551991765e-06,
      "loss": 0.3806,
      "step": 985
    },
    {
      "epoch": 0.8171687990094924,
      "grad_norm": 1.9398586585837028,
      "learning_rate": 9.822176824513524e-07,
      "loss": 0.3731,
      "step": 990
    },
    {
      "epoch": 0.821295914156005,
      "grad_norm": 1.96994225142757,
      "learning_rate": 9.397086261457511e-07,
      "loss": 0.3776,
      "step": 995
    },
    {
      "epoch": 0.8254230293025175,
      "grad_norm": 2.003600819188701,
      "learning_rate": 8.980443304696107e-07,
      "loss": 0.3819,
      "step": 1000
    },
    {
      "epoch": 0.8295501444490301,
      "grad_norm": 2.0351798859032484,
      "learning_rate": 8.572334638541857e-07,
      "loss": 0.3774,
      "step": 1005
    },
    {
      "epoch": 0.8336772595955427,
      "grad_norm": 2.1430516379993403,
      "learning_rate": 8.172845171712379e-07,
      "loss": 0.3737,
      "step": 1010
    },
    {
      "epoch": 0.8378043747420553,
      "grad_norm": 2.03671544389429,
      "learning_rate": 7.782058019664773e-07,
      "loss": 0.3709,
      "step": 1015
    },
    {
      "epoch": 0.8419314898885679,
      "grad_norm": 2.014237343029339,
      "learning_rate": 7.40005448730306e-07,
      "loss": 0.3717,
      "step": 1020
    },
    {
      "epoch": 0.8460586050350805,
      "grad_norm": 2.001733960692795,
      "learning_rate": 7.026914052062433e-07,
      "loss": 0.378,
      "step": 1025
    },
    {
      "epoch": 0.8501857201815931,
      "grad_norm": 2.0314797478672864,
      "learning_rate": 6.662714347373589e-07,
      "loss": 0.3718,
      "step": 1030
    },
    {
      "epoch": 0.8543128353281056,
      "grad_norm": 2.0398788512991985,
      "learning_rate": 6.307531146510754e-07,
      "loss": 0.3802,
      "step": 1035
    },
    {
      "epoch": 0.8584399504746183,
      "grad_norm": 2.0401593697279616,
      "learning_rate": 5.961438346826792e-07,
      "loss": 0.3723,
      "step": 1040
    },
    {
      "epoch": 0.8625670656211308,
      "grad_norm": 2.105641702439893,
      "learning_rate": 5.6245079543785e-07,
      "loss": 0.3674,
      "step": 1045
    },
    {
      "epoch": 0.8666941807676434,
      "grad_norm": 2.0475120583137962,
      "learning_rate": 5.296810068945474e-07,
      "loss": 0.3767,
      "step": 1050
    },
    {
      "epoch": 0.870821295914156,
      "grad_norm": 2.037325684321563,
      "learning_rate": 4.97841286944557e-07,
      "loss": 0.3717,
      "step": 1055
    },
    {
      "epoch": 0.8749484110606686,
      "grad_norm": 1.8826839605258934,
      "learning_rate": 4.6693825997499274e-07,
      "loss": 0.366,
      "step": 1060
    },
    {
      "epoch": 0.8790755262071812,
      "grad_norm": 2.1270152173670294,
      "learning_rate": 4.3697835549007474e-07,
      "loss": 0.3657,
      "step": 1065
    },
    {
      "epoch": 0.8832026413536938,
      "grad_norm": 1.986884523560406,
      "learning_rate": 4.0796780677343606e-07,
      "loss": 0.3687,
      "step": 1070
    },
    {
      "epoch": 0.8873297565002064,
      "grad_norm": 2.0992498730608453,
      "learning_rate": 3.799126495912686e-07,
      "loss": 0.3638,
      "step": 1075
    },
    {
      "epoch": 0.8914568716467189,
      "grad_norm": 2.028752109036286,
      "learning_rate": 3.528187209365486e-07,
      "loss": 0.3699,
      "step": 1080
    },
    {
      "epoch": 0.8955839867932315,
      "grad_norm": 2.0346601882201707,
      "learning_rate": 3.26691657814634e-07,
      "loss": 0.3703,
      "step": 1085
    },
    {
      "epoch": 0.8997111019397441,
      "grad_norm": 2.0217798718878717,
      "learning_rate": 3.015368960704584e-07,
      "loss": 0.3604,
      "step": 1090
    },
    {
      "epoch": 0.9038382170862567,
      "grad_norm": 1.9484068285089462,
      "learning_rate": 2.7735966925757807e-07,
      "loss": 0.3582,
      "step": 1095
    },
    {
      "epoch": 0.9079653322327693,
      "grad_norm": 1.9903170219755175,
      "learning_rate": 2.5416500754931294e-07,
      "loss": 0.3641,
      "step": 1100
    },
    {
      "epoch": 0.9120924473792819,
      "grad_norm": 2.025285970039861,
      "learning_rate": 2.3195773669219767e-07,
      "loss": 0.3651,
      "step": 1105
    },
    {
      "epoch": 0.9162195625257945,
      "grad_norm": 1.9954116267972222,
      "learning_rate": 2.1074247700196337e-07,
      "loss": 0.3665,
      "step": 1110
    },
    {
      "epoch": 0.920346677672307,
      "grad_norm": 2.0132907410321588,
      "learning_rate": 1.905236424022633e-07,
      "loss": 0.3586,
      "step": 1115
    },
    {
      "epoch": 0.9244737928188197,
      "grad_norm": 1.9966146901368165,
      "learning_rate": 1.7130543950633405e-07,
      "loss": 0.3588,
      "step": 1120
    },
    {
      "epoch": 0.9286009079653322,
      "grad_norm": 2.0300570451418314,
      "learning_rate": 1.5309186674179633e-07,
      "loss": 0.3634,
      "step": 1125
    },
    {
      "epoch": 0.9327280231118448,
      "grad_norm": 2.1175207366778146,
      "learning_rate": 1.358867135187636e-07,
      "loss": 0.3734,
      "step": 1130
    },
    {
      "epoch": 0.9368551382583574,
      "grad_norm": 2.0296632257291956,
      "learning_rate": 1.1969355944143834e-07,
      "loss": 0.3632,
      "step": 1135
    },
    {
      "epoch": 0.94098225340487,
      "grad_norm": 1.9716530471763118,
      "learning_rate": 1.0451577356336118e-07,
      "loss": 0.3556,
      "step": 1140
    },
    {
      "epoch": 0.9451093685513826,
      "grad_norm": 2.079524156836045,
      "learning_rate": 9.035651368646647e-08,
      "loss": 0.3627,
      "step": 1145
    },
    {
      "epoch": 0.9492364836978952,
      "grad_norm": 2.063707259064606,
      "learning_rate": 7.721872570408684e-08,
      "loss": 0.3633,
      "step": 1150
    },
    {
      "epoch": 0.9533635988444078,
      "grad_norm": 2.007674392778877,
      "learning_rate": 6.510514298804838e-08,
      "loss": 0.3622,
      "step": 1155
    },
    {
      "epoch": 0.9574907139909203,
      "grad_norm": 2.0337463475612534,
      "learning_rate": 5.401828581997948e-08,
      "loss": 0.3576,
      "step": 1160
    },
    {
      "epoch": 0.961617829137433,
      "grad_norm": 1.994551313475225,
      "learning_rate": 4.396046086695915e-08,
      "loss": 0.3569,
      "step": 1165
    },
    {
      "epoch": 0.9657449442839455,
      "grad_norm": 2.016111248082697,
      "learning_rate": 3.4933760701602595e-08,
      "loss": 0.3578,
      "step": 1170
    },
    {
      "epoch": 0.9698720594304581,
      "grad_norm": 2.061506062561967,
      "learning_rate": 2.6940063366693303e-08,
      "loss": 0.3642,
      "step": 1175
    },
    {
      "epoch": 0.9739991745769707,
      "grad_norm": 1.9380243614271475,
      "learning_rate": 1.998103198444845e-08,
      "loss": 0.3682,
      "step": 1180
    },
    {
      "epoch": 0.9781262897234833,
      "grad_norm": 2.0523022237333506,
      "learning_rate": 1.405811441049898e-08,
      "loss": 0.3558,
      "step": 1185
    },
    {
      "epoch": 0.9822534048699959,
      "grad_norm": 1.9847092252520235,
      "learning_rate": 9.1725429326589e-09,
      "loss": 0.3513,
      "step": 1190
    },
    {
      "epoch": 0.9863805200165084,
      "grad_norm": 1.9944372302390396,
      "learning_rate": 5.325334014539829e-09,
      "loss": 0.3682,
      "step": 1195
    },
    {
      "epoch": 0.9905076351630211,
      "grad_norm": 2.0557893820556847,
      "learning_rate": 2.5172880840745873e-09,
      "loss": 0.3642,
      "step": 1200
    },
    {
      "epoch": 0.9946347503095336,
      "grad_norm": 1.9769612777056,
      "learning_rate": 7.48989366980979e-10,
      "loss": 0.368,
      "step": 1205
    },
    {
      "epoch": 0.9987618654560462,
      "grad_norm": 2.1452580734926343,
      "learning_rate": 2.080576521568123e-11,
      "loss": 0.3657,
      "step": 1210
    },
    {
      "epoch": 0.9995872884853487,
      "eval_loss": 0.3757357895374298,
      "eval_runtime": 0.963,
      "eval_samples_per_second": 3.115,
      "eval_steps_per_second": 1.038,
      "step": 1211
    },
    {
      "epoch": 0.9995872884853487,
      "step": 1211,
      "total_flos": 253506418114560.0,
      "train_loss": 0.5435279840826496,
      "train_runtime": 29545.3245,
      "train_samples_per_second": 1.312,
      "train_steps_per_second": 0.041
    }
  ],
  "logging_steps": 5,
  "max_steps": 1211,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 253506418114560.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}