{
  "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": 24.527768091219585,
      "learning_rate": 8.19672131147541e-08,
      "loss": 1.4526,
      "step": 1
    },
    {
      "epoch": 0.004127115146512587,
      "grad_norm": 24.06574033360543,
      "learning_rate": 4.0983606557377047e-07,
      "loss": 1.4262,
      "step": 5
    },
    {
      "epoch": 0.008254230293025175,
      "grad_norm": 8.408253799970256,
      "learning_rate": 8.196721311475409e-07,
      "loss": 1.3214,
      "step": 10
    },
    {
      "epoch": 0.012381345439537762,
      "grad_norm": 10.618280212141128,
      "learning_rate": 1.2295081967213116e-06,
      "loss": 1.1627,
      "step": 15
    },
    {
      "epoch": 0.01650846058605035,
      "grad_norm": 3.0056706196625753,
      "learning_rate": 1.6393442622950819e-06,
      "loss": 1.0211,
      "step": 20
    },
    {
      "epoch": 0.020635575732562937,
      "grad_norm": 2.80244952396409,
      "learning_rate": 2.0491803278688526e-06,
      "loss": 0.9766,
      "step": 25
    },
    {
      "epoch": 0.024762690879075525,
      "grad_norm": 2.4916214253566005,
      "learning_rate": 2.459016393442623e-06,
      "loss": 0.9463,
      "step": 30
    },
    {
      "epoch": 0.028889806025588115,
      "grad_norm": 2.332420416883666,
      "learning_rate": 2.8688524590163937e-06,
      "loss": 0.9086,
      "step": 35
    },
    {
      "epoch": 0.0330169211721007,
      "grad_norm": 2.324411296099785,
      "learning_rate": 3.2786885245901638e-06,
      "loss": 0.9237,
      "step": 40
    },
    {
      "epoch": 0.03714403631861329,
      "grad_norm": 2.2166466173210533,
      "learning_rate": 3.6885245901639347e-06,
      "loss": 0.8874,
      "step": 45
    },
    {
      "epoch": 0.041271151465125874,
      "grad_norm": 2.187061208635295,
      "learning_rate": 4.098360655737705e-06,
      "loss": 0.8853,
      "step": 50
    },
    {
      "epoch": 0.04539826661163846,
      "grad_norm": 2.3549138169493014,
      "learning_rate": 4.508196721311476e-06,
      "loss": 0.8741,
      "step": 55
    },
    {
      "epoch": 0.04952538175815105,
      "grad_norm": 2.50076860105583,
      "learning_rate": 4.918032786885246e-06,
      "loss": 0.8712,
      "step": 60
    },
    {
      "epoch": 0.053652496904663644,
      "grad_norm": 2.2937651316774357,
      "learning_rate": 5.327868852459017e-06,
      "loss": 0.8606,
      "step": 65
    },
    {
      "epoch": 0.05777961205117623,
      "grad_norm": 2.600551672083637,
      "learning_rate": 5.737704918032787e-06,
      "loss": 0.8579,
      "step": 70
    },
    {
      "epoch": 0.06190672719768882,
      "grad_norm": 2.3502153854983825,
      "learning_rate": 6.147540983606558e-06,
      "loss": 0.8596,
      "step": 75
    },
    {
      "epoch": 0.0660338423442014,
      "grad_norm": 2.5388298566841394,
      "learning_rate": 6.5573770491803276e-06,
      "loss": 0.8268,
      "step": 80
    },
    {
      "epoch": 0.070160957490714,
      "grad_norm": 2.3050474743750136,
      "learning_rate": 6.967213114754099e-06,
      "loss": 0.8391,
      "step": 85
    },
    {
      "epoch": 0.07428807263722657,
      "grad_norm": 2.371799609042535,
      "learning_rate": 7.3770491803278695e-06,
      "loss": 0.8255,
      "step": 90
    },
    {
      "epoch": 0.07841518778373917,
      "grad_norm": 2.4920625709896793,
      "learning_rate": 7.786885245901639e-06,
      "loss": 0.8287,
      "step": 95
    },
    {
      "epoch": 0.08254230293025175,
      "grad_norm": 2.638561657589798,
      "learning_rate": 8.19672131147541e-06,
      "loss": 0.8186,
      "step": 100
    },
    {
      "epoch": 0.08666941807676434,
      "grad_norm": 2.3901740015101884,
      "learning_rate": 8.60655737704918e-06,
      "loss": 0.8301,
      "step": 105
    },
    {
      "epoch": 0.09079653322327692,
      "grad_norm": 2.6300899584489925,
      "learning_rate": 9.016393442622952e-06,
      "loss": 0.8187,
      "step": 110
    },
    {
      "epoch": 0.09492364836978952,
      "grad_norm": 2.3616261176867965,
      "learning_rate": 9.426229508196723e-06,
      "loss": 0.8083,
      "step": 115
    },
    {
      "epoch": 0.0990507635163021,
      "grad_norm": 2.4752485733881198,
      "learning_rate": 9.836065573770493e-06,
      "loss": 0.8168,
      "step": 120
    },
    {
      "epoch": 0.10317787866281469,
      "grad_norm": 2.4101059264429496,
      "learning_rate": 9.999812749151968e-06,
      "loss": 0.8077,
      "step": 125
    },
    {
      "epoch": 0.10730499380932729,
      "grad_norm": 2.6712071601759857,
      "learning_rate": 9.998668489204265e-06,
      "loss": 0.8055,
      "step": 130
    },
    {
      "epoch": 0.11143210895583987,
      "grad_norm": 2.4087034620873835,
      "learning_rate": 9.99648423533726e-06,
      "loss": 0.7969,
      "step": 135
    },
    {
      "epoch": 0.11555922410235246,
      "grad_norm": 2.7170848710573785,
      "learning_rate": 9.993260441994116e-06,
      "loss": 0.7817,
      "step": 140
    },
    {
      "epoch": 0.11968633924886504,
      "grad_norm": 2.609457342521693,
      "learning_rate": 9.988997779898544e-06,
      "loss": 0.7889,
      "step": 145
    },
    {
      "epoch": 0.12381345439537764,
      "grad_norm": 2.4448007509369036,
      "learning_rate": 9.983697135915253e-06,
      "loss": 0.7884,
      "step": 150
    },
    {
      "epoch": 0.12794056954189023,
      "grad_norm": 2.4835789788836413,
      "learning_rate": 9.977359612865424e-06,
      "loss": 0.7846,
      "step": 155
    },
    {
      "epoch": 0.1320676846884028,
      "grad_norm": 2.5497182573160715,
      "learning_rate": 9.96998652929728e-06,
      "loss": 0.7745,
      "step": 160
    },
    {
      "epoch": 0.1361947998349154,
      "grad_norm": 2.4134195891584707,
      "learning_rate": 9.961579419211741e-06,
      "loss": 0.7662,
      "step": 165
    },
    {
      "epoch": 0.140321914981428,
      "grad_norm": 2.363551295048404,
      "learning_rate": 9.952140031743282e-06,
      "loss": 0.7689,
      "step": 170
    },
    {
      "epoch": 0.14444903012794058,
      "grad_norm": 2.5912204484961725,
      "learning_rate": 9.941670330796007e-06,
      "loss": 0.7631,
      "step": 175
    },
    {
      "epoch": 0.14857614527445315,
      "grad_norm": 2.9224664902673156,
      "learning_rate": 9.930172494635058e-06,
      "loss": 0.7474,
      "step": 180
    },
    {
      "epoch": 0.15270326042096574,
      "grad_norm": 2.413069765526069,
      "learning_rate": 9.917648915433413e-06,
      "loss": 0.7564,
      "step": 185
    },
    {
      "epoch": 0.15683037556747834,
      "grad_norm": 2.287951655167943,
      "learning_rate": 9.904102198774188e-06,
      "loss": 0.7608,
      "step": 190
    },
    {
      "epoch": 0.16095749071399093,
      "grad_norm": 2.3478183977261233,
      "learning_rate": 9.889535163108537e-06,
      "loss": 0.7492,
      "step": 195
    },
    {
      "epoch": 0.1650846058605035,
      "grad_norm": 2.335853369295283,
      "learning_rate": 9.873950839169248e-06,
      "loss": 0.7424,
      "step": 200
    },
    {
      "epoch": 0.1692117210070161,
      "grad_norm": 2.431644019701366,
      "learning_rate": 9.857352469340204e-06,
      "loss": 0.7339,
      "step": 205
    },
    {
      "epoch": 0.17333883615352869,
      "grad_norm": 2.467518852941353,
      "learning_rate": 9.839743506981783e-06,
      "loss": 0.7199,
      "step": 210
    },
    {
      "epoch": 0.17746595130004128,
      "grad_norm": 2.4545799822273233,
      "learning_rate": 9.821127615712365e-06,
      "loss": 0.7276,
      "step": 215
    },
    {
      "epoch": 0.18159306644655385,
      "grad_norm": 2.2908915072144405,
      "learning_rate": 9.801508668646117e-06,
      "loss": 0.7186,
      "step": 220
    },
    {
      "epoch": 0.18572018159306644,
      "grad_norm": 2.429940544373184,
      "learning_rate": 9.780890747587163e-06,
      "loss": 0.7123,
      "step": 225
    },
    {
      "epoch": 0.18984729673957904,
      "grad_norm": 2.566519603485226,
      "learning_rate": 9.759278142180348e-06,
      "loss": 0.7126,
      "step": 230
    },
    {
      "epoch": 0.19397441188609163,
      "grad_norm": 2.259124198746028,
      "learning_rate": 9.736675349018768e-06,
      "loss": 0.7129,
      "step": 235
    },
    {
      "epoch": 0.1981015270326042,
      "grad_norm": 2.388987469748086,
      "learning_rate": 9.713087070708225e-06,
      "loss": 0.723,
      "step": 240
    },
    {
      "epoch": 0.2022286421791168,
      "grad_norm": 2.26678703688603,
      "learning_rate": 9.688518214888836e-06,
      "loss": 0.7193,
      "step": 245
    },
    {
      "epoch": 0.20635575732562939,
      "grad_norm": 2.118252251899789,
      "learning_rate": 9.662973893213976e-06,
      "loss": 0.6954,
      "step": 250
    },
    {
      "epoch": 0.21048287247214198,
      "grad_norm": 2.3331873161331877,
      "learning_rate": 9.636459420286779e-06,
      "loss": 0.6849,
      "step": 255
    },
    {
      "epoch": 0.21460998761865457,
      "grad_norm": 2.3076302273355567,
      "learning_rate": 9.60898031255441e-06,
      "loss": 0.6825,
      "step": 260
    },
    {
      "epoch": 0.21873710276516714,
      "grad_norm": 2.6088703694048405,
      "learning_rate": 9.580542287160348e-06,
      "loss": 0.6869,
      "step": 265
    },
    {
      "epoch": 0.22286421791167974,
      "grad_norm": 2.364361000804096,
      "learning_rate": 9.551151260754907e-06,
      "loss": 0.683,
      "step": 270
    },
    {
      "epoch": 0.22699133305819233,
      "grad_norm": 2.232900651700554,
      "learning_rate": 9.520813348264252e-06,
      "loss": 0.6828,
      "step": 275
    },
    {
      "epoch": 0.23111844820470492,
      "grad_norm": 2.35722592750872,
      "learning_rate": 9.489534861618168e-06,
      "loss": 0.6567,
      "step": 280
    },
    {
      "epoch": 0.2352455633512175,
      "grad_norm": 2.319853004406621,
      "learning_rate": 9.45732230843683e-06,
      "loss": 0.6694,
      "step": 285
    },
    {
      "epoch": 0.23937267849773008,
      "grad_norm": 2.358145153885717,
      "learning_rate": 9.424182390676872e-06,
      "loss": 0.6716,
      "step": 290
    },
    {
      "epoch": 0.24349979364424268,
      "grad_norm": 2.6691908854406736,
      "learning_rate": 9.39012200323701e-06,
      "loss": 0.668,
      "step": 295
    },
    {
      "epoch": 0.24762690879075527,
      "grad_norm": 2.7341705151937896,
      "learning_rate": 9.355148232523538e-06,
      "loss": 0.6609,
      "step": 300
    },
    {
      "epoch": 0.25175402393726787,
      "grad_norm": 2.386539335699224,
      "learning_rate": 9.319268354975958e-06,
      "loss": 0.6484,
      "step": 305
    },
    {
      "epoch": 0.25588113908378046,
      "grad_norm": 2.279413488060828,
      "learning_rate": 9.282489835553106e-06,
      "loss": 0.6424,
      "step": 310
    },
    {
      "epoch": 0.260008254230293,
      "grad_norm": 2.188135665187205,
      "learning_rate": 9.244820326180013e-06,
      "loss": 0.6524,
      "step": 315
    },
    {
      "epoch": 0.2641353693768056,
      "grad_norm": 2.2684748409078423,
      "learning_rate": 9.206267664155906e-06,
      "loss": 0.6604,
      "step": 320
    },
    {
      "epoch": 0.2682624845233182,
      "grad_norm": 2.262865794776347,
      "learning_rate": 9.166839870523627e-06,
      "loss": 0.6468,
      "step": 325
    },
    {
      "epoch": 0.2723895996698308,
      "grad_norm": 2.4581062840025942,
      "learning_rate": 9.126545148400807e-06,
      "loss": 0.6411,
      "step": 330
    },
    {
      "epoch": 0.2765167148163434,
      "grad_norm": 2.450742436164245,
      "learning_rate": 9.085391881273182e-06,
      "loss": 0.6469,
      "step": 335
    },
    {
      "epoch": 0.280643829962856,
      "grad_norm": 2.43800822057093,
      "learning_rate": 9.04338863125038e-06,
      "loss": 0.6333,
      "step": 340
    },
    {
      "epoch": 0.28477094510936857,
      "grad_norm": 2.391479100614559,
      "learning_rate": 9.00054413728452e-06,
      "loss": 0.6281,
      "step": 345
    },
    {
      "epoch": 0.28889806025588116,
      "grad_norm": 2.407756154639032,
      "learning_rate": 8.956867313352055e-06,
      "loss": 0.6166,
      "step": 350
    },
    {
      "epoch": 0.2930251754023937,
      "grad_norm": 2.256108783602758,
      "learning_rate": 8.912367246599175e-06,
      "loss": 0.6205,
      "step": 355
    },
    {
      "epoch": 0.2971522905489063,
      "grad_norm": 2.234657616830803,
      "learning_rate": 8.867053195451182e-06,
      "loss": 0.6222,
      "step": 360
    },
    {
      "epoch": 0.3012794056954189,
      "grad_norm": 2.2577550570512988,
      "learning_rate": 8.820934587686247e-06,
      "loss": 0.6055,
      "step": 365
    },
    {
      "epoch": 0.3054065208419315,
      "grad_norm": 2.6360921307136613,
      "learning_rate": 8.77402101847391e-06,
      "loss": 0.6077,
      "step": 370
    },
    {
      "epoch": 0.3095336359884441,
      "grad_norm": 2.183430116006734,
      "learning_rate": 8.726322248378775e-06,
      "loss": 0.6127,
      "step": 375
    },
    {
      "epoch": 0.3136607511349567,
      "grad_norm": 2.3206875632337853,
      "learning_rate": 8.677848201329775e-06,
      "loss": 0.6076,
      "step": 380
    },
    {
      "epoch": 0.31778786628146927,
      "grad_norm": 2.1734495418432185,
      "learning_rate": 8.628608962555467e-06,
      "loss": 0.5993,
      "step": 385
    },
    {
      "epoch": 0.32191498142798186,
      "grad_norm": 2.314380886468653,
      "learning_rate": 8.578614776485744e-06,
      "loss": 0.6096,
      "step": 390
    },
    {
      "epoch": 0.32604209657449446,
      "grad_norm": 2.2736697333193527,
      "learning_rate": 8.527876044620453e-06,
      "loss": 0.5877,
      "step": 395
    },
    {
      "epoch": 0.330169211721007,
      "grad_norm": 2.29671056878197,
      "learning_rate": 8.47640332336531e-06,
      "loss": 0.5941,
      "step": 400
    },
    {
      "epoch": 0.3342963268675196,
      "grad_norm": 2.405544598077206,
      "learning_rate": 8.424207321835598e-06,
      "loss": 0.6039,
      "step": 405
    },
    {
      "epoch": 0.3384234420140322,
      "grad_norm": 2.2242053189632043,
      "learning_rate": 8.371298899628091e-06,
      "loss": 0.5818,
      "step": 410
    },
    {
      "epoch": 0.3425505571605448,
      "grad_norm": 2.5103451771347287,
      "learning_rate": 8.317689064561671e-06,
      "loss": 0.5821,
      "step": 415
    },
    {
      "epoch": 0.34667767230705737,
      "grad_norm": 2.1638232343460535,
      "learning_rate": 8.263388970387102e-06,
      "loss": 0.5675,
      "step": 420
    },
    {
      "epoch": 0.35080478745356997,
      "grad_norm": 2.3380030293780716,
      "learning_rate": 8.20840991446645e-06,
      "loss": 0.5747,
      "step": 425
    },
    {
      "epoch": 0.35493190260008256,
      "grad_norm": 2.303084137873228,
      "learning_rate": 8.152763335422612e-06,
      "loss": 0.5586,
      "step": 430
    },
    {
      "epoch": 0.35905901774659515,
      "grad_norm": 2.229950988354625,
      "learning_rate": 8.096460810759473e-06,
      "loss": 0.5677,
      "step": 435
    },
    {
      "epoch": 0.3631861328931077,
      "grad_norm": 2.571672304368483,
      "learning_rate": 8.03951405445314e-06,
      "loss": 0.5715,
      "step": 440
    },
    {
      "epoch": 0.3673132480396203,
      "grad_norm": 2.1478754361718977,
      "learning_rate": 7.98193491451483e-06,
      "loss": 0.5622,
      "step": 445
    },
    {
      "epoch": 0.3714403631861329,
      "grad_norm": 2.190588527928839,
      "learning_rate": 7.923735370525809e-06,
      "loss": 0.5702,
      "step": 450
    },
    {
      "epoch": 0.3755674783326455,
      "grad_norm": 2.0752352437472554,
      "learning_rate": 7.864927531145012e-06,
      "loss": 0.552,
      "step": 455
    },
    {
      "epoch": 0.37969459347915807,
      "grad_norm": 2.170688708600695,
      "learning_rate": 7.805523631589775e-06,
      "loss": 0.5705,
      "step": 460
    },
    {
      "epoch": 0.38382170862567067,
      "grad_norm": 2.320105443139919,
      "learning_rate": 7.745536031090252e-06,
      "loss": 0.5506,
      "step": 465
    },
    {
      "epoch": 0.38794882377218326,
      "grad_norm": 2.4389141970585952,
      "learning_rate": 7.684977210318024e-06,
      "loss": 0.5522,
      "step": 470
    },
    {
      "epoch": 0.39207593891869585,
      "grad_norm": 2.153706916707687,
      "learning_rate": 7.623859768789441e-06,
      "loss": 0.537,
      "step": 475
    },
    {
      "epoch": 0.3962030540652084,
      "grad_norm": 2.272410344361673,
      "learning_rate": 7.5621964222442455e-06,
      "loss": 0.5533,
      "step": 480
    },
    {
      "epoch": 0.400330169211721,
      "grad_norm": 2.269121379963274,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.5339,
      "step": 485
    },
    {
      "epoch": 0.4044572843582336,
      "grad_norm": 2.1641159438102497,
      "learning_rate": 7.437283442282903e-06,
      "loss": 0.5308,
      "step": 490
    },
    {
      "epoch": 0.4085843995047462,
      "grad_norm": 2.2333206506903176,
      "learning_rate": 7.374059797535516e-06,
      "loss": 0.5215,
      "step": 495
    },
    {
      "epoch": 0.41271151465125877,
      "grad_norm": 2.3350190406293514,
      "learning_rate": 7.310342219701981e-06,
      "loss": 0.5357,
      "step": 500
    },
    {
      "epoch": 0.41683862979777137,
      "grad_norm": 2.1111507498174142,
      "learning_rate": 7.246143965491288e-06,
      "loss": 0.5398,
      "step": 505
    },
    {
      "epoch": 0.42096574494428396,
      "grad_norm": 2.179999263691962,
      "learning_rate": 7.181478391619162e-06,
      "loss": 0.526,
      "step": 510
    },
    {
      "epoch": 0.42509286009079655,
      "grad_norm": 2.234723286133957,
      "learning_rate": 7.11635895202914e-06,
      "loss": 0.5319,
      "step": 515
    },
    {
      "epoch": 0.42921997523730915,
      "grad_norm": 2.3625762744028105,
      "learning_rate": 7.050799195093421e-06,
      "loss": 0.5232,
      "step": 520
    },
    {
      "epoch": 0.4333470903838217,
      "grad_norm": 2.151720448324523,
      "learning_rate": 6.984812760794078e-06,
      "loss": 0.5184,
      "step": 525
    },
    {
      "epoch": 0.4374742055303343,
      "grad_norm": 2.1956000381712433,
      "learning_rate": 6.918413377885193e-06,
      "loss": 0.5345,
      "step": 530
    },
    {
      "epoch": 0.4416013206768469,
      "grad_norm": 2.2229124245862364,
      "learning_rate": 6.851614861036533e-06,
      "loss": 0.5171,
      "step": 535
    },
    {
      "epoch": 0.44572843582335947,
      "grad_norm": 2.1531210729445336,
      "learning_rate": 6.78443110795936e-06,
      "loss": 0.5097,
      "step": 540
    },
    {
      "epoch": 0.44985555096987206,
      "grad_norm": 2.14693065420584,
      "learning_rate": 6.716876096514944e-06,
      "loss": 0.5129,
      "step": 545
    },
    {
      "epoch": 0.45398266611638466,
      "grad_norm": 2.163715272133522,
      "learning_rate": 6.648963881806411e-06,
      "loss": 0.5124,
      "step": 550
    },
    {
      "epoch": 0.45810978126289725,
      "grad_norm": 2.1010364723302506,
      "learning_rate": 6.580708593254526e-06,
      "loss": 0.5186,
      "step": 555
    },
    {
      "epoch": 0.46223689640940985,
      "grad_norm": 2.2467642835081683,
      "learning_rate": 6.512124431658006e-06,
      "loss": 0.5088,
      "step": 560
    },
    {
      "epoch": 0.4663640115559224,
      "grad_norm": 2.1609546827041055,
      "learning_rate": 6.443225666238976e-06,
      "loss": 0.4934,
      "step": 565
    },
    {
      "epoch": 0.470491126702435,
      "grad_norm": 2.236319324239601,
      "learning_rate": 6.37402663167421e-06,
      "loss": 0.5129,
      "step": 570
    },
    {
      "epoch": 0.4746182418489476,
      "grad_norm": 2.117151693329715,
      "learning_rate": 6.304541725112734e-06,
      "loss": 0.5006,
      "step": 575
    },
    {
      "epoch": 0.47874535699546017,
      "grad_norm": 2.2422969978564895,
      "learning_rate": 6.234785403180438e-06,
      "loss": 0.4976,
      "step": 580
    },
    {
      "epoch": 0.48287247214197276,
      "grad_norm": 2.2994111726673174,
      "learning_rate": 6.164772178972321e-06,
      "loss": 0.4992,
      "step": 585
    },
    {
      "epoch": 0.48699958728848536,
      "grad_norm": 2.095094004562891,
      "learning_rate": 6.094516619032975e-06,
      "loss": 0.4918,
      "step": 590
    },
    {
      "epoch": 0.49112670243499795,
      "grad_norm": 2.1619801251290194,
      "learning_rate": 6.024033340325954e-06,
      "loss": 0.4925,
      "step": 595
    },
    {
      "epoch": 0.49525381758151055,
      "grad_norm": 2.212661104115349,
      "learning_rate": 5.953337007192659e-06,
      "loss": 0.4905,
      "step": 600
    },
    {
      "epoch": 0.4993809327280231,
      "grad_norm": 2.178233561063734,
      "learning_rate": 5.882442328301356e-06,
      "loss": 0.4749,
      "step": 605
    },
    {
      "epoch": 0.5035080478745357,
      "grad_norm": 2.065630506818835,
      "learning_rate": 5.811364053586973e-06,
      "loss": 0.4776,
      "step": 610
    },
    {
      "epoch": 0.5076351630210483,
      "grad_norm": 2.092353975022163,
      "learning_rate": 5.740116971182322e-06,
      "loss": 0.485,
      "step": 615
    },
    {
      "epoch": 0.5117622781675609,
      "grad_norm": 2.1253939738705916,
      "learning_rate": 5.668715904341365e-06,
      "loss": 0.4893,
      "step": 620
    },
    {
      "epoch": 0.5158893933140735,
      "grad_norm": 2.1054886510378865,
      "learning_rate": 5.5971757083551625e-06,
      "loss": 0.4762,
      "step": 625
    },
    {
      "epoch": 0.520016508460586,
      "grad_norm": 2.1261261707153176,
      "learning_rate": 5.525511267461186e-06,
      "loss": 0.474,
      "step": 630
    },
    {
      "epoch": 0.5241436236070987,
      "grad_norm": 2.0961240074666403,
      "learning_rate": 5.453737491746572e-06,
      "loss": 0.4737,
      "step": 635
    },
    {
      "epoch": 0.5282707387536112,
      "grad_norm": 2.245074234522742,
      "learning_rate": 5.381869314046031e-06,
      "loss": 0.4754,
      "step": 640
    },
    {
      "epoch": 0.5323978539001238,
      "grad_norm": 2.0710026861411928,
      "learning_rate": 5.3099216868349965e-06,
      "loss": 0.4583,
      "step": 645
    },
    {
      "epoch": 0.5365249690466364,
      "grad_norm": 2.0545375231065455,
      "learning_rate": 5.237909579118713e-06,
      "loss": 0.4663,
      "step": 650
    },
    {
      "epoch": 0.540652084193149,
      "grad_norm": 2.1403499315140944,
      "learning_rate": 5.165847973317854e-06,
      "loss": 0.4669,
      "step": 655
    },
    {
      "epoch": 0.5447791993396616,
      "grad_norm": 1.9566981568041115,
      "learning_rate": 5.093751862151388e-06,
      "loss": 0.469,
      "step": 660
    },
    {
      "epoch": 0.5489063144861742,
      "grad_norm": 2.018373173481524,
      "learning_rate": 5.021636245517261e-06,
      "loss": 0.4631,
      "step": 665
    },
    {
      "epoch": 0.5530334296326868,
      "grad_norm": 2.0287210921461525,
      "learning_rate": 4.9495161273716174e-06,
      "loss": 0.4523,
      "step": 670
    },
    {
      "epoch": 0.5571605447791993,
      "grad_norm": 2.0979432063643717,
      "learning_rate": 4.877406512607159e-06,
      "loss": 0.4659,
      "step": 675
    },
    {
      "epoch": 0.561287659925712,
      "grad_norm": 2.109641878105565,
      "learning_rate": 4.805322403931312e-06,
      "loss": 0.4565,
      "step": 680
    },
    {
      "epoch": 0.5654147750722245,
      "grad_norm": 2.042972869631033,
      "learning_rate": 4.73327879874486e-06,
      "loss": 0.4604,
      "step": 685
    },
    {
      "epoch": 0.5695418902187371,
      "grad_norm": 2.077048659643493,
      "learning_rate": 4.661290686021661e-06,
      "loss": 0.4532,
      "step": 690
    },
    {
      "epoch": 0.5736690053652497,
      "grad_norm": 2.190754559278812,
      "learning_rate": 4.589373043190137e-06,
      "loss": 0.4544,
      "step": 695
    },
    {
      "epoch": 0.5777961205117623,
      "grad_norm": 2.0600760514379983,
      "learning_rate": 4.517540833017152e-06,
      "loss": 0.4388,
      "step": 700
    },
    {
      "epoch": 0.5819232356582749,
      "grad_norm": 2.085650574224392,
      "learning_rate": 4.445809000494945e-06,
      "loss": 0.4571,
      "step": 705
    },
    {
      "epoch": 0.5860503508047874,
      "grad_norm": 2.0328851509507935,
      "learning_rate": 4.374192469731771e-06,
      "loss": 0.4468,
      "step": 710
    },
    {
      "epoch": 0.5901774659513,
      "grad_norm": 2.1834463600319087,
      "learning_rate": 4.302706140846864e-06,
      "loss": 0.4438,
      "step": 715
    },
    {
      "epoch": 0.5943045810978126,
      "grad_norm": 2.068868362244341,
      "learning_rate": 4.231364886870417e-06,
      "loss": 0.4358,
      "step": 720
    },
    {
      "epoch": 0.5984316962443252,
      "grad_norm": 2.071607001638305,
      "learning_rate": 4.160183550649176e-06,
      "loss": 0.4414,
      "step": 725
    },
    {
      "epoch": 0.6025588113908378,
      "grad_norm": 2.070373251758235,
      "learning_rate": 4.089176941758332e-06,
      "loss": 0.4334,
      "step": 730
    },
    {
      "epoch": 0.6066859265373504,
      "grad_norm": 1.929229950335449,
      "learning_rate": 4.018359833420323e-06,
      "loss": 0.4408,
      "step": 735
    },
    {
      "epoch": 0.610813041683863,
      "grad_norm": 1.9813661913181657,
      "learning_rate": 3.9477469594311975e-06,
      "loss": 0.4307,
      "step": 740
    },
    {
      "epoch": 0.6149401568303756,
      "grad_norm": 2.017699438378818,
      "learning_rate": 3.8773530110952e-06,
      "loss": 0.4407,
      "step": 745
    },
    {
      "epoch": 0.6190672719768882,
      "grad_norm": 1.9968368675785926,
      "learning_rate": 3.807192634168168e-06,
      "loss": 0.4355,
      "step": 750
    },
    {
      "epoch": 0.6231943871234007,
      "grad_norm": 2.297629412730533,
      "learning_rate": 3.7372804258104367e-06,
      "loss": 0.4344,
      "step": 755
    },
    {
      "epoch": 0.6273215022699133,
      "grad_norm": 2.083065196032756,
      "learning_rate": 3.667630931549826e-06,
      "loss": 0.4281,
      "step": 760
    },
    {
      "epoch": 0.6314486174164259,
      "grad_norm": 2.1990182920796144,
      "learning_rate": 3.598258642255387e-06,
      "loss": 0.4321,
      "step": 765
    },
    {
      "epoch": 0.6355757325629385,
      "grad_norm": 2.034767533939522,
      "learning_rate": 3.529177991122519e-06,
      "loss": 0.4249,
      "step": 770
    },
    {
      "epoch": 0.6397028477094511,
      "grad_norm": 2.0893128290504217,
      "learning_rate": 3.460403350670077e-06,
      "loss": 0.425,
      "step": 775
    },
    {
      "epoch": 0.6438299628559637,
      "grad_norm": 2.2774223056638503,
      "learning_rate": 3.3919490297501167e-06,
      "loss": 0.43,
      "step": 780
    },
    {
      "epoch": 0.6479570780024763,
      "grad_norm": 2.0215299051371574,
      "learning_rate": 3.3238292705708675e-06,
      "loss": 0.4236,
      "step": 785
    },
    {
      "epoch": 0.6520841931489889,
      "grad_norm": 2.1346863821345483,
      "learning_rate": 3.256058245733592e-06,
      "loss": 0.4283,
      "step": 790
    },
    {
      "epoch": 0.6562113082955014,
      "grad_norm": 1.9952609517532462,
      "learning_rate": 3.1886500552839105e-06,
      "loss": 0.4258,
      "step": 795
    },
    {
      "epoch": 0.660338423442014,
      "grad_norm": 2.13096013979772,
      "learning_rate": 3.121618723778225e-06,
      "loss": 0.4203,
      "step": 800
    },
    {
      "epoch": 0.6644655385885266,
      "grad_norm": 1.9003371358709635,
      "learning_rate": 3.054978197365861e-06,
      "loss": 0.4205,
      "step": 805
    },
    {
      "epoch": 0.6685926537350392,
      "grad_norm": 2.1054234107701455,
      "learning_rate": 2.9887423408875056e-06,
      "loss": 0.4102,
      "step": 810
    },
    {
      "epoch": 0.6727197688815518,
      "grad_norm": 2.0086799082910813,
      "learning_rate": 2.9229249349905686e-06,
      "loss": 0.4132,
      "step": 815
    },
    {
      "epoch": 0.6768468840280644,
      "grad_norm": 2.053993947170071,
      "learning_rate": 2.8575396732620673e-06,
      "loss": 0.4209,
      "step": 820
    },
    {
      "epoch": 0.680973999174577,
      "grad_norm": 1.978130155419886,
      "learning_rate": 2.7926001593796114e-06,
      "loss": 0.411,
      "step": 825
    },
    {
      "epoch": 0.6851011143210896,
      "grad_norm": 2.029741499090481,
      "learning_rate": 2.728119904281105e-06,
      "loss": 0.4132,
      "step": 830
    },
    {
      "epoch": 0.6892282294676021,
      "grad_norm": 2.003813411220096,
      "learning_rate": 2.6641123233537395e-06,
      "loss": 0.4084,
      "step": 835
    },
    {
      "epoch": 0.6933553446141147,
      "grad_norm": 1.8978236980737733,
      "learning_rate": 2.6005907336428615e-06,
      "loss": 0.4067,
      "step": 840
    },
    {
      "epoch": 0.6974824597606273,
      "grad_norm": 2.0102189681652356,
      "learning_rate": 2.537568351081311e-06,
      "loss": 0.4042,
      "step": 845
    },
    {
      "epoch": 0.7016095749071399,
      "grad_norm": 1.9833503306404032,
      "learning_rate": 2.4750582877397933e-06,
      "loss": 0.4033,
      "step": 850
    },
    {
      "epoch": 0.7057366900536525,
      "grad_norm": 2.0412127058121063,
      "learning_rate": 2.4130735490988457e-06,
      "loss": 0.4008,
      "step": 855
    },
    {
      "epoch": 0.7098638052001651,
      "grad_norm": 2.156188349951087,
      "learning_rate": 2.3516270313430085e-06,
      "loss": 0.3942,
      "step": 860
    },
    {
      "epoch": 0.7139909203466777,
      "grad_norm": 2.1350064162632187,
      "learning_rate": 2.2907315186777057e-06,
      "loss": 0.4044,
      "step": 865
    },
    {
      "epoch": 0.7181180354931903,
      "grad_norm": 2.0279876481819534,
      "learning_rate": 2.230399680669449e-06,
      "loss": 0.3975,
      "step": 870
    },
    {
      "epoch": 0.7222451506397028,
      "grad_norm": 2.0026732743887647,
      "learning_rate": 2.170644069609876e-06,
      "loss": 0.4022,
      "step": 875
    },
    {
      "epoch": 0.7263722657862154,
      "grad_norm": 2.1697007212318495,
      "learning_rate": 2.1114771179041894e-06,
      "loss": 0.4013,
      "step": 880
    },
    {
      "epoch": 0.730499380932728,
      "grad_norm": 1.9716176275869417,
      "learning_rate": 2.052911135484551e-06,
      "loss": 0.3994,
      "step": 885
    },
    {
      "epoch": 0.7346264960792406,
      "grad_norm": 2.034092895423435,
      "learning_rate": 1.9949583072489455e-06,
      "loss": 0.4061,
      "step": 890
    },
    {
      "epoch": 0.7387536112257532,
      "grad_norm": 2.005053459378774,
      "learning_rate": 1.9376306905260604e-06,
      "loss": 0.3914,
      "step": 895
    },
    {
      "epoch": 0.7428807263722658,
      "grad_norm": 2.0574703701368944,
      "learning_rate": 1.8809402125667064e-06,
      "loss": 0.3968,
      "step": 900
    },
    {
      "epoch": 0.7470078415187784,
      "grad_norm": 2.0316413222622876,
      "learning_rate": 1.8248986680623077e-06,
      "loss": 0.3974,
      "step": 905
    },
    {
      "epoch": 0.751134956665291,
      "grad_norm": 2.1497596921656723,
      "learning_rate": 1.7695177166909672e-06,
      "loss": 0.3854,
      "step": 910
    },
    {
      "epoch": 0.7552620718118036,
      "grad_norm": 2.018266195220802,
      "learning_rate": 1.7148088806916114e-06,
      "loss": 0.3869,
      "step": 915
    },
    {
      "epoch": 0.7593891869583161,
      "grad_norm": 1.9829612702874728,
      "learning_rate": 1.6607835424667578e-06,
      "loss": 0.3958,
      "step": 920
    },
    {
      "epoch": 0.7635163021048287,
      "grad_norm": 2.0944985053497804,
      "learning_rate": 1.6074529422143398e-06,
      "loss": 0.3903,
      "step": 925
    },
    {
      "epoch": 0.7676434172513413,
      "grad_norm": 2.0288825311492373,
      "learning_rate": 1.554828175589151e-06,
      "loss": 0.387,
      "step": 930
    },
    {
      "epoch": 0.7717705323978539,
      "grad_norm": 2.0044313136237575,
      "learning_rate": 1.5029201913943425e-06,
      "loss": 0.3877,
      "step": 935
    },
    {
      "epoch": 0.7758976475443665,
      "grad_norm": 2.0080374408263557,
      "learning_rate": 1.4517397893034718e-06,
      "loss": 0.3796,
      "step": 940
    },
    {
      "epoch": 0.7800247626908791,
      "grad_norm": 2.0587700368615356,
      "learning_rate": 1.4012976176135978e-06,
      "loss": 0.3748,
      "step": 945
    },
    {
      "epoch": 0.7841518778373917,
      "grad_norm": 2.0898100439782614,
      "learning_rate": 1.35160417102985e-06,
      "loss": 0.3886,
      "step": 950
    },
    {
      "epoch": 0.7882789929839042,
      "grad_norm": 1.9808019328492945,
      "learning_rate": 1.3026697884819644e-06,
      "loss": 0.3863,
      "step": 955
    },
    {
      "epoch": 0.7924061081304168,
      "grad_norm": 2.1132444631377347,
      "learning_rate": 1.25450465097322e-06,
      "loss": 0.3759,
      "step": 960
    },
    {
      "epoch": 0.7965332232769294,
      "grad_norm": 1.9937976759731069,
      "learning_rate": 1.207118779462248e-06,
      "loss": 0.3815,
      "step": 965
    },
    {
      "epoch": 0.800660338423442,
      "grad_norm": 1.9815880927129814,
      "learning_rate": 1.160522032778123e-06,
      "loss": 0.3799,
      "step": 970
    },
    {
      "epoch": 0.8047874535699546,
      "grad_norm": 1.954589085358194,
      "learning_rate": 1.1147241055691909e-06,
      "loss": 0.3729,
      "step": 975
    },
    {
      "epoch": 0.8089145687164672,
      "grad_norm": 2.0090926554473483,
      "learning_rate": 1.0697345262860638e-06,
      "loss": 0.3762,
      "step": 980
    },
    {
      "epoch": 0.8130416838629798,
      "grad_norm": 2.0548702162748804,
      "learning_rate": 1.0255626551991765e-06,
      "loss": 0.3789,
      "step": 985
    },
    {
      "epoch": 0.8171687990094924,
      "grad_norm": 2.0560582305639166,
      "learning_rate": 9.822176824513524e-07,
      "loss": 0.3788,
      "step": 990
    },
    {
      "epoch": 0.821295914156005,
      "grad_norm": 2.0889270318940585,
      "learning_rate": 9.397086261457511e-07,
      "loss": 0.37,
      "step": 995
    },
    {
      "epoch": 0.8254230293025175,
      "grad_norm": 2.0446625577354047,
      "learning_rate": 8.980443304696107e-07,
      "loss": 0.3777,
      "step": 1000
    },
    {
      "epoch": 0.8295501444490301,
      "grad_norm": 2.0486212476420818,
      "learning_rate": 8.572334638541857e-07,
      "loss": 0.3756,
      "step": 1005
    },
    {
      "epoch": 0.8336772595955427,
      "grad_norm": 2.002401159823963,
      "learning_rate": 8.172845171712379e-07,
      "loss": 0.3708,
      "step": 1010
    },
    {
      "epoch": 0.8378043747420553,
      "grad_norm": 2.1384693815400286,
      "learning_rate": 7.782058019664773e-07,
      "loss": 0.3818,
      "step": 1015
    },
    {
      "epoch": 0.8419314898885679,
      "grad_norm": 1.9145047936397264,
      "learning_rate": 7.40005448730306e-07,
      "loss": 0.3722,
      "step": 1020
    },
    {
      "epoch": 0.8460586050350805,
      "grad_norm": 1.9971688248700292,
      "learning_rate": 7.026914052062433e-07,
      "loss": 0.3759,
      "step": 1025
    },
    {
      "epoch": 0.8501857201815931,
      "grad_norm": 2.1314313481782445,
      "learning_rate": 6.662714347373589e-07,
      "loss": 0.3794,
      "step": 1030
    },
    {
      "epoch": 0.8543128353281056,
      "grad_norm": 2.0749616131915083,
      "learning_rate": 6.307531146510754e-07,
      "loss": 0.3683,
      "step": 1035
    },
    {
      "epoch": 0.8584399504746183,
      "grad_norm": 2.064414306275117,
      "learning_rate": 5.961438346826792e-07,
      "loss": 0.3679,
      "step": 1040
    },
    {
      "epoch": 0.8625670656211308,
      "grad_norm": 2.175753609552035,
      "learning_rate": 5.6245079543785e-07,
      "loss": 0.376,
      "step": 1045
    },
    {
      "epoch": 0.8666941807676434,
      "grad_norm": 2.001643753214543,
      "learning_rate": 5.296810068945474e-07,
      "loss": 0.3684,
      "step": 1050
    },
    {
      "epoch": 0.870821295914156,
      "grad_norm": 1.9714345242801743,
      "learning_rate": 4.97841286944557e-07,
      "loss": 0.3653,
      "step": 1055
    },
    {
      "epoch": 0.8749484110606686,
      "grad_norm": 2.0426482960795505,
      "learning_rate": 4.6693825997499274e-07,
      "loss": 0.3685,
      "step": 1060
    },
    {
      "epoch": 0.8790755262071812,
      "grad_norm": 2.065168614269635,
      "learning_rate": 4.3697835549007474e-07,
      "loss": 0.3698,
      "step": 1065
    },
    {
      "epoch": 0.8832026413536938,
      "grad_norm": 2.0536856821090117,
      "learning_rate": 4.0796780677343606e-07,
      "loss": 0.3765,
      "step": 1070
    },
    {
      "epoch": 0.8873297565002064,
      "grad_norm": 1.9520606705536954,
      "learning_rate": 3.799126495912686e-07,
      "loss": 0.3696,
      "step": 1075
    },
    {
      "epoch": 0.8914568716467189,
      "grad_norm": 2.0256345371223423,
      "learning_rate": 3.528187209365486e-07,
      "loss": 0.3632,
      "step": 1080
    },
    {
      "epoch": 0.8955839867932315,
      "grad_norm": 2.0341797163670803,
      "learning_rate": 3.26691657814634e-07,
      "loss": 0.3697,
      "step": 1085
    },
    {
      "epoch": 0.8997111019397441,
      "grad_norm": 2.0238570768757214,
      "learning_rate": 3.015368960704584e-07,
      "loss": 0.3697,
      "step": 1090
    },
    {
      "epoch": 0.9038382170862567,
      "grad_norm": 2.0482723959973494,
      "learning_rate": 2.7735966925757807e-07,
      "loss": 0.3743,
      "step": 1095
    },
    {
      "epoch": 0.9079653322327693,
      "grad_norm": 2.070484598348767,
      "learning_rate": 2.5416500754931294e-07,
      "loss": 0.3584,
      "step": 1100
    },
    {
      "epoch": 0.9120924473792819,
      "grad_norm": 2.0139930549422944,
      "learning_rate": 2.3195773669219767e-07,
      "loss": 0.3709,
      "step": 1105
    },
    {
      "epoch": 0.9162195625257945,
      "grad_norm": 1.992128335347791,
      "learning_rate": 2.1074247700196337e-07,
      "loss": 0.3627,
      "step": 1110
    },
    {
      "epoch": 0.920346677672307,
      "grad_norm": 2.0866051408396222,
      "learning_rate": 1.905236424022633e-07,
      "loss": 0.3592,
      "step": 1115
    },
    {
      "epoch": 0.9244737928188197,
      "grad_norm": 2.12346665841604,
      "learning_rate": 1.7130543950633405e-07,
      "loss": 0.3653,
      "step": 1120
    },
    {
      "epoch": 0.9286009079653322,
      "grad_norm": 2.047113547519804,
      "learning_rate": 1.5309186674179633e-07,
      "loss": 0.3699,
      "step": 1125
    },
    {
      "epoch": 0.9327280231118448,
      "grad_norm": 1.9950316722566084,
      "learning_rate": 1.358867135187636e-07,
      "loss": 0.3554,
      "step": 1130
    },
    {
      "epoch": 0.9368551382583574,
      "grad_norm": 1.96686387365348,
      "learning_rate": 1.1969355944143834e-07,
      "loss": 0.3491,
      "step": 1135
    },
    {
      "epoch": 0.94098225340487,
      "grad_norm": 2.0771745550819682,
      "learning_rate": 1.0451577356336118e-07,
      "loss": 0.3627,
      "step": 1140
    },
    {
      "epoch": 0.9451093685513826,
      "grad_norm": 1.9361758127483037,
      "learning_rate": 9.035651368646647e-08,
      "loss": 0.3607,
      "step": 1145
    },
    {
      "epoch": 0.9492364836978952,
      "grad_norm": 2.1055645801286715,
      "learning_rate": 7.721872570408684e-08,
      "loss": 0.3661,
      "step": 1150
    },
    {
      "epoch": 0.9533635988444078,
      "grad_norm": 2.0208444370681153,
      "learning_rate": 6.510514298804838e-08,
      "loss": 0.3661,
      "step": 1155
    },
    {
      "epoch": 0.9574907139909203,
      "grad_norm": 2.0397131358725753,
      "learning_rate": 5.401828581997948e-08,
      "loss": 0.3534,
      "step": 1160
    },
    {
      "epoch": 0.961617829137433,
      "grad_norm": 2.020631083047579,
      "learning_rate": 4.396046086695915e-08,
      "loss": 0.3654,
      "step": 1165
    },
    {
      "epoch": 0.9657449442839455,
      "grad_norm": 2.073342231781071,
      "learning_rate": 3.4933760701602595e-08,
      "loss": 0.3591,
      "step": 1170
    },
    {
      "epoch": 0.9698720594304581,
      "grad_norm": 2.0188173197726558,
      "learning_rate": 2.6940063366693303e-08,
      "loss": 0.3584,
      "step": 1175
    },
    {
      "epoch": 0.9739991745769707,
      "grad_norm": 2.0720645114918743,
      "learning_rate": 1.998103198444845e-08,
      "loss": 0.3618,
      "step": 1180
    },
    {
      "epoch": 0.9781262897234833,
      "grad_norm": 2.0436662884162753,
      "learning_rate": 1.405811441049898e-08,
      "loss": 0.3555,
      "step": 1185
    },
    {
      "epoch": 0.9822534048699959,
      "grad_norm": 1.9703598755915417,
      "learning_rate": 9.1725429326589e-09,
      "loss": 0.3639,
      "step": 1190
    },
    {
      "epoch": 0.9863805200165084,
      "grad_norm": 2.0602528259644552,
      "learning_rate": 5.325334014539829e-09,
      "loss": 0.3615,
      "step": 1195
    },
    {
      "epoch": 0.9905076351630211,
      "grad_norm": 2.0964063828388024,
      "learning_rate": 2.5172880840745873e-09,
      "loss": 0.3733,
      "step": 1200
    },
    {
      "epoch": 0.9946347503095336,
      "grad_norm": 1.994585654952302,
      "learning_rate": 7.48989366980979e-10,
      "loss": 0.361,
      "step": 1205
    },
    {
      "epoch": 0.9987618654560462,
      "grad_norm": 1.8509542578033722,
      "learning_rate": 2.080576521568123e-11,
      "loss": 0.3659,
      "step": 1210
    },
    {
      "epoch": 0.9995872884853487,
      "eval_runtime": 3.4177,
      "eval_samples_per_second": 2.926,
      "eval_steps_per_second": 0.878,
      "step": 1211
    },
    {
      "epoch": 0.9995872884853487,
      "step": 1211,
      "total_flos": 6.266791881867264e+16,
      "train_loss": 0.5434653141299719,
      "train_runtime": 21156.3929,
      "train_samples_per_second": 1.832,
      "train_steps_per_second": 0.057
    }
  ],
  "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": 6.266791881867264e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}