{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 8070,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006195786864931847,
      "grad_norm": 11.78878402709961,
      "learning_rate": 5.947955390334573e-06,
      "loss": 3.345,
      "step": 50
    },
    {
      "epoch": 0.012391573729863693,
      "grad_norm": 5.766342639923096,
      "learning_rate": 1.2019826517967782e-05,
      "loss": 0.8099,
      "step": 100
    },
    {
      "epoch": 0.01858736059479554,
      "grad_norm": 3.6993865966796875,
      "learning_rate": 1.821561338289963e-05,
      "loss": 0.3079,
      "step": 150
    },
    {
      "epoch": 0.024783147459727387,
      "grad_norm": 3.7327823638916016,
      "learning_rate": 2.4411400247831474e-05,
      "loss": 0.2089,
      "step": 200
    },
    {
      "epoch": 0.030978934324659233,
      "grad_norm": 3.176992416381836,
      "learning_rate": 3.0607187112763326e-05,
      "loss": 0.1961,
      "step": 250
    },
    {
      "epoch": 0.03717472118959108,
      "grad_norm": 3.009659767150879,
      "learning_rate": 3.667905824039653e-05,
      "loss": 0.186,
      "step": 300
    },
    {
      "epoch": 0.04337050805452292,
      "grad_norm": 3.961405038833618,
      "learning_rate": 4.287484510532838e-05,
      "loss": 0.1607,
      "step": 350
    },
    {
      "epoch": 0.04956629491945477,
      "grad_norm": 2.818254232406616,
      "learning_rate": 4.907063197026023e-05,
      "loss": 0.1483,
      "step": 400
    },
    {
      "epoch": 0.055762081784386616,
      "grad_norm": 1.836232304573059,
      "learning_rate": 5.526641883519207e-05,
      "loss": 0.1391,
      "step": 450
    },
    {
      "epoch": 0.061957868649318466,
      "grad_norm": 2.5643157958984375,
      "learning_rate": 6.146220570012391e-05,
      "loss": 0.1343,
      "step": 500
    },
    {
      "epoch": 0.06815365551425032,
      "grad_norm": 2.9712822437286377,
      "learning_rate": 6.765799256505576e-05,
      "loss": 0.1235,
      "step": 550
    },
    {
      "epoch": 0.07434944237918216,
      "grad_norm": 3.6296300888061523,
      "learning_rate": 7.385377942998762e-05,
      "loss": 0.1222,
      "step": 600
    },
    {
      "epoch": 0.080545229244114,
      "grad_norm": 10.417726516723633,
      "learning_rate": 8.004956629491945e-05,
      "loss": 0.1186,
      "step": 650
    },
    {
      "epoch": 0.08674101610904585,
      "grad_norm": 2.5631349086761475,
      "learning_rate": 8.624535315985131e-05,
      "loss": 0.128,
      "step": 700
    },
    {
      "epoch": 0.09293680297397769,
      "grad_norm": 1.6498531103134155,
      "learning_rate": 9.244114002478316e-05,
      "loss": 0.1145,
      "step": 750
    },
    {
      "epoch": 0.09913258983890955,
      "grad_norm": 2.1243278980255127,
      "learning_rate": 9.8636926889715e-05,
      "loss": 0.1153,
      "step": 800
    },
    {
      "epoch": 0.10532837670384139,
      "grad_norm": 1.869721531867981,
      "learning_rate": 9.946303180503924e-05,
      "loss": 0.1143,
      "step": 850
    },
    {
      "epoch": 0.11152416356877323,
      "grad_norm": 1.236275315284729,
      "learning_rate": 9.877461104226904e-05,
      "loss": 0.1099,
      "step": 900
    },
    {
      "epoch": 0.11771995043370508,
      "grad_norm": 2.4983303546905518,
      "learning_rate": 9.808619027949883e-05,
      "loss": 0.1018,
      "step": 950
    },
    {
      "epoch": 0.12391573729863693,
      "grad_norm": 1.4123425483703613,
      "learning_rate": 9.739776951672863e-05,
      "loss": 0.1079,
      "step": 1000
    },
    {
      "epoch": 0.13011152416356878,
      "grad_norm": 2.497483253479004,
      "learning_rate": 9.670934875395842e-05,
      "loss": 0.0982,
      "step": 1050
    },
    {
      "epoch": 0.13630731102850063,
      "grad_norm": 1.100753903388977,
      "learning_rate": 9.602092799118821e-05,
      "loss": 0.0949,
      "step": 1100
    },
    {
      "epoch": 0.14250309789343246,
      "grad_norm": 2.238828182220459,
      "learning_rate": 9.533250722841801e-05,
      "loss": 0.0945,
      "step": 1150
    },
    {
      "epoch": 0.14869888475836432,
      "grad_norm": 1.5351171493530273,
      "learning_rate": 9.46440864656478e-05,
      "loss": 0.0952,
      "step": 1200
    },
    {
      "epoch": 0.15489467162329615,
      "grad_norm": 3.6645586490631104,
      "learning_rate": 9.39556657028776e-05,
      "loss": 0.1162,
      "step": 1250
    },
    {
      "epoch": 0.161090458488228,
      "grad_norm": 2.153944730758667,
      "learning_rate": 9.32672449401074e-05,
      "loss": 0.0965,
      "step": 1300
    },
    {
      "epoch": 0.16728624535315986,
      "grad_norm": 5.332737922668457,
      "learning_rate": 9.25788241773372e-05,
      "loss": 0.0847,
      "step": 1350
    },
    {
      "epoch": 0.1734820322180917,
      "grad_norm": 1.9773648977279663,
      "learning_rate": 9.189040341456699e-05,
      "loss": 0.0916,
      "step": 1400
    },
    {
      "epoch": 0.17967781908302355,
      "grad_norm": 1.072949767112732,
      "learning_rate": 9.120198265179678e-05,
      "loss": 0.0821,
      "step": 1450
    },
    {
      "epoch": 0.18587360594795538,
      "grad_norm": 0.5937812924385071,
      "learning_rate": 9.051356188902658e-05,
      "loss": 0.0693,
      "step": 1500
    },
    {
      "epoch": 0.19206939281288724,
      "grad_norm": 3.5840229988098145,
      "learning_rate": 8.982514112625638e-05,
      "loss": 0.0753,
      "step": 1550
    },
    {
      "epoch": 0.1982651796778191,
      "grad_norm": 1.8539464473724365,
      "learning_rate": 8.913672036348617e-05,
      "loss": 0.072,
      "step": 1600
    },
    {
      "epoch": 0.20446096654275092,
      "grad_norm": 1.8574514389038086,
      "learning_rate": 8.844829960071597e-05,
      "loss": 0.0718,
      "step": 1650
    },
    {
      "epoch": 0.21065675340768278,
      "grad_norm": 2.465460777282715,
      "learning_rate": 8.775987883794575e-05,
      "loss": 0.0715,
      "step": 1700
    },
    {
      "epoch": 0.21685254027261464,
      "grad_norm": 1.191398024559021,
      "learning_rate": 8.707145807517555e-05,
      "loss": 0.0711,
      "step": 1750
    },
    {
      "epoch": 0.22304832713754646,
      "grad_norm": 1.4969195127487183,
      "learning_rate": 8.638303731240535e-05,
      "loss": 0.0731,
      "step": 1800
    },
    {
      "epoch": 0.22924411400247832,
      "grad_norm": 1.8418350219726562,
      "learning_rate": 8.569461654963514e-05,
      "loss": 0.0636,
      "step": 1850
    },
    {
      "epoch": 0.23543990086741015,
      "grad_norm": 2.362734794616699,
      "learning_rate": 8.500619578686494e-05,
      "loss": 0.076,
      "step": 1900
    },
    {
      "epoch": 0.241635687732342,
      "grad_norm": 1.3227559328079224,
      "learning_rate": 8.431777502409473e-05,
      "loss": 0.06,
      "step": 1950
    },
    {
      "epoch": 0.24783147459727387,
      "grad_norm": 2.1812291145324707,
      "learning_rate": 8.362935426132452e-05,
      "loss": 0.0582,
      "step": 2000
    },
    {
      "epoch": 0.2540272614622057,
      "grad_norm": 0.8610977530479431,
      "learning_rate": 8.294093349855432e-05,
      "loss": 0.056,
      "step": 2050
    },
    {
      "epoch": 0.26022304832713755,
      "grad_norm": 1.2188324928283691,
      "learning_rate": 8.225251273578411e-05,
      "loss": 0.0539,
      "step": 2100
    },
    {
      "epoch": 0.2664188351920694,
      "grad_norm": 0.6571711301803589,
      "learning_rate": 8.156409197301392e-05,
      "loss": 0.0548,
      "step": 2150
    },
    {
      "epoch": 0.27261462205700127,
      "grad_norm": 9.587064743041992,
      "learning_rate": 8.08756712102437e-05,
      "loss": 0.0522,
      "step": 2200
    },
    {
      "epoch": 0.2788104089219331,
      "grad_norm": 1.4897613525390625,
      "learning_rate": 8.018725044747349e-05,
      "loss": 0.0429,
      "step": 2250
    },
    {
      "epoch": 0.2850061957868649,
      "grad_norm": 1.3524194955825806,
      "learning_rate": 7.94988296847033e-05,
      "loss": 0.0543,
      "step": 2300
    },
    {
      "epoch": 0.29120198265179675,
      "grad_norm": 0.49432888627052307,
      "learning_rate": 7.88241773371885e-05,
      "loss": 0.055,
      "step": 2350
    },
    {
      "epoch": 0.29739776951672864,
      "grad_norm": 1.082529902458191,
      "learning_rate": 7.813575657441828e-05,
      "loss": 0.0449,
      "step": 2400
    },
    {
      "epoch": 0.30359355638166047,
      "grad_norm": 3.580268144607544,
      "learning_rate": 7.744733581164808e-05,
      "loss": 0.0526,
      "step": 2450
    },
    {
      "epoch": 0.3097893432465923,
      "grad_norm": 3.7611606121063232,
      "learning_rate": 7.675891504887788e-05,
      "loss": 0.0432,
      "step": 2500
    },
    {
      "epoch": 0.3159851301115242,
      "grad_norm": 3.605032205581665,
      "learning_rate": 7.607049428610767e-05,
      "loss": 0.0513,
      "step": 2550
    },
    {
      "epoch": 0.322180916976456,
      "grad_norm": 0.8407074809074402,
      "learning_rate": 7.538207352333747e-05,
      "loss": 0.0478,
      "step": 2600
    },
    {
      "epoch": 0.32837670384138784,
      "grad_norm": 1.2463380098342896,
      "learning_rate": 7.469365276056726e-05,
      "loss": 0.041,
      "step": 2650
    },
    {
      "epoch": 0.3345724907063197,
      "grad_norm": 1.2079633474349976,
      "learning_rate": 7.400523199779705e-05,
      "loss": 0.0455,
      "step": 2700
    },
    {
      "epoch": 0.34076827757125155,
      "grad_norm": 1.7008190155029297,
      "learning_rate": 7.331681123502685e-05,
      "loss": 0.0515,
      "step": 2750
    },
    {
      "epoch": 0.3469640644361834,
      "grad_norm": 0.6003010272979736,
      "learning_rate": 7.262839047225664e-05,
      "loss": 0.0455,
      "step": 2800
    },
    {
      "epoch": 0.35315985130111527,
      "grad_norm": 0.7806116342544556,
      "learning_rate": 7.193996970948645e-05,
      "loss": 0.0429,
      "step": 2850
    },
    {
      "epoch": 0.3593556381660471,
      "grad_norm": 1.190233826637268,
      "learning_rate": 7.125154894671623e-05,
      "loss": 0.037,
      "step": 2900
    },
    {
      "epoch": 0.3655514250309789,
      "grad_norm": 0.49826720356941223,
      "learning_rate": 7.056312818394602e-05,
      "loss": 0.0329,
      "step": 2950
    },
    {
      "epoch": 0.37174721189591076,
      "grad_norm": 1.38163423538208,
      "learning_rate": 6.987470742117583e-05,
      "loss": 0.0381,
      "step": 3000
    },
    {
      "epoch": 0.37794299876084264,
      "grad_norm": 0.6930022835731506,
      "learning_rate": 6.918628665840562e-05,
      "loss": 0.0381,
      "step": 3050
    },
    {
      "epoch": 0.38413878562577447,
      "grad_norm": 0.3004520535469055,
      "learning_rate": 6.849786589563542e-05,
      "loss": 0.0373,
      "step": 3100
    },
    {
      "epoch": 0.3903345724907063,
      "grad_norm": 1.0762969255447388,
      "learning_rate": 6.78094451328652e-05,
      "loss": 0.0405,
      "step": 3150
    },
    {
      "epoch": 0.3965303593556382,
      "grad_norm": 0.9407225251197815,
      "learning_rate": 6.712102437009501e-05,
      "loss": 0.0344,
      "step": 3200
    },
    {
      "epoch": 0.40272614622057,
      "grad_norm": 1.1790508031845093,
      "learning_rate": 6.64326036073248e-05,
      "loss": 0.0345,
      "step": 3250
    },
    {
      "epoch": 0.40892193308550184,
      "grad_norm": 0.6774219274520874,
      "learning_rate": 6.574418284455459e-05,
      "loss": 0.0327,
      "step": 3300
    },
    {
      "epoch": 0.41511771995043373,
      "grad_norm": 0.5361367464065552,
      "learning_rate": 6.505576208178439e-05,
      "loss": 0.0271,
      "step": 3350
    },
    {
      "epoch": 0.42131350681536556,
      "grad_norm": 0.2760148048400879,
      "learning_rate": 6.436734131901418e-05,
      "loss": 0.0314,
      "step": 3400
    },
    {
      "epoch": 0.4275092936802974,
      "grad_norm": 0.7887628674507141,
      "learning_rate": 6.367892055624398e-05,
      "loss": 0.0352,
      "step": 3450
    },
    {
      "epoch": 0.43370508054522927,
      "grad_norm": 1.0329231023788452,
      "learning_rate": 6.299049979347378e-05,
      "loss": 0.0262,
      "step": 3500
    },
    {
      "epoch": 0.4399008674101611,
      "grad_norm": 0.9285022020339966,
      "learning_rate": 6.230207903070356e-05,
      "loss": 0.0295,
      "step": 3550
    },
    {
      "epoch": 0.44609665427509293,
      "grad_norm": 0.23095445334911346,
      "learning_rate": 6.161365826793336e-05,
      "loss": 0.0284,
      "step": 3600
    },
    {
      "epoch": 0.45229244114002476,
      "grad_norm": 0.42350977659225464,
      "learning_rate": 6.092523750516316e-05,
      "loss": 0.0248,
      "step": 3650
    },
    {
      "epoch": 0.45848822800495664,
      "grad_norm": 0.9503557085990906,
      "learning_rate": 6.023681674239295e-05,
      "loss": 0.0239,
      "step": 3700
    },
    {
      "epoch": 0.4646840148698885,
      "grad_norm": 1.4695123434066772,
      "learning_rate": 5.954839597962275e-05,
      "loss": 0.0272,
      "step": 3750
    },
    {
      "epoch": 0.4708798017348203,
      "grad_norm": 0.835435688495636,
      "learning_rate": 5.8859975216852544e-05,
      "loss": 0.0253,
      "step": 3800
    },
    {
      "epoch": 0.4770755885997522,
      "grad_norm": 0.8063173294067383,
      "learning_rate": 5.817155445408233e-05,
      "loss": 0.0288,
      "step": 3850
    },
    {
      "epoch": 0.483271375464684,
      "grad_norm": 0.5583626627922058,
      "learning_rate": 5.748313369131213e-05,
      "loss": 0.0267,
      "step": 3900
    },
    {
      "epoch": 0.48946716232961585,
      "grad_norm": 0.3900113105773926,
      "learning_rate": 5.6794712928541925e-05,
      "loss": 0.027,
      "step": 3950
    },
    {
      "epoch": 0.49566294919454773,
      "grad_norm": 0.684743344783783,
      "learning_rate": 5.610629216577172e-05,
      "loss": 0.0233,
      "step": 4000
    },
    {
      "epoch": 0.5018587360594795,
      "grad_norm": 1.3466298580169678,
      "learning_rate": 5.541787140300152e-05,
      "loss": 0.0237,
      "step": 4050
    },
    {
      "epoch": 0.5080545229244114,
      "grad_norm": 0.814662754535675,
      "learning_rate": 5.472945064023132e-05,
      "loss": 0.0224,
      "step": 4100
    },
    {
      "epoch": 0.5142503097893433,
      "grad_norm": 0.20325958728790283,
      "learning_rate": 5.40410298774611e-05,
      "loss": 0.0252,
      "step": 4150
    },
    {
      "epoch": 0.5204460966542751,
      "grad_norm": 0.9722644686698914,
      "learning_rate": 5.3352609114690896e-05,
      "loss": 0.0228,
      "step": 4200
    },
    {
      "epoch": 0.5266418835192069,
      "grad_norm": 0.8962277770042419,
      "learning_rate": 5.26641883519207e-05,
      "loss": 0.0243,
      "step": 4250
    },
    {
      "epoch": 0.5328376703841388,
      "grad_norm": 0.8505231738090515,
      "learning_rate": 5.1975767589150495e-05,
      "loss": 0.0269,
      "step": 4300
    },
    {
      "epoch": 0.5390334572490706,
      "grad_norm": 1.4101253747940063,
      "learning_rate": 5.128734682638029e-05,
      "loss": 0.0228,
      "step": 4350
    },
    {
      "epoch": 0.5452292441140025,
      "grad_norm": 1.0223798751831055,
      "learning_rate": 5.0598926063610086e-05,
      "loss": 0.0219,
      "step": 4400
    },
    {
      "epoch": 0.5514250309789344,
      "grad_norm": 0.3649487793445587,
      "learning_rate": 4.9910505300839875e-05,
      "loss": 0.0215,
      "step": 4450
    },
    {
      "epoch": 0.5576208178438662,
      "grad_norm": 0.467814564704895,
      "learning_rate": 4.922208453806967e-05,
      "loss": 0.0242,
      "step": 4500
    },
    {
      "epoch": 0.563816604708798,
      "grad_norm": 1.218153953552246,
      "learning_rate": 4.8533663775299466e-05,
      "loss": 0.0199,
      "step": 4550
    },
    {
      "epoch": 0.5700123915737298,
      "grad_norm": 0.2747304141521454,
      "learning_rate": 4.784524301252926e-05,
      "loss": 0.0196,
      "step": 4600
    },
    {
      "epoch": 0.5762081784386617,
      "grad_norm": 0.572721004486084,
      "learning_rate": 4.715682224975905e-05,
      "loss": 0.0171,
      "step": 4650
    },
    {
      "epoch": 0.5824039653035935,
      "grad_norm": 0.48146089911460876,
      "learning_rate": 4.646840148698885e-05,
      "loss": 0.0161,
      "step": 4700
    },
    {
      "epoch": 0.5885997521685254,
      "grad_norm": 0.3745860457420349,
      "learning_rate": 4.577998072421865e-05,
      "loss": 0.0182,
      "step": 4750
    },
    {
      "epoch": 0.5947955390334573,
      "grad_norm": 0.5441644191741943,
      "learning_rate": 4.509155996144844e-05,
      "loss": 0.019,
      "step": 4800
    },
    {
      "epoch": 0.6009913258983891,
      "grad_norm": 0.39052465558052063,
      "learning_rate": 4.4403139198678234e-05,
      "loss": 0.0192,
      "step": 4850
    },
    {
      "epoch": 0.6071871127633209,
      "grad_norm": 0.2999328672885895,
      "learning_rate": 4.371471843590803e-05,
      "loss": 0.018,
      "step": 4900
    },
    {
      "epoch": 0.6133828996282528,
      "grad_norm": 0.5317378044128418,
      "learning_rate": 4.3026297673137825e-05,
      "loss": 0.016,
      "step": 4950
    },
    {
      "epoch": 0.6195786864931846,
      "grad_norm": 1.1451416015625,
      "learning_rate": 4.233787691036762e-05,
      "loss": 0.0167,
      "step": 5000
    },
    {
      "epoch": 0.6257744733581165,
      "grad_norm": 1.846800684928894,
      "learning_rate": 4.164945614759742e-05,
      "loss": 0.0155,
      "step": 5050
    },
    {
      "epoch": 0.6319702602230484,
      "grad_norm": 0.2182386815547943,
      "learning_rate": 4.0961035384827206e-05,
      "loss": 0.0151,
      "step": 5100
    },
    {
      "epoch": 0.6381660470879802,
      "grad_norm": 0.6467189788818359,
      "learning_rate": 4.0272614622057e-05,
      "loss": 0.0169,
      "step": 5150
    },
    {
      "epoch": 0.644361833952912,
      "grad_norm": 1.209778070449829,
      "learning_rate": 3.95841938592868e-05,
      "loss": 0.0151,
      "step": 5200
    },
    {
      "epoch": 0.6505576208178439,
      "grad_norm": 0.8446183204650879,
      "learning_rate": 3.889577309651659e-05,
      "loss": 0.0152,
      "step": 5250
    },
    {
      "epoch": 0.6567534076827757,
      "grad_norm": 1.0092103481292725,
      "learning_rate": 3.820735233374639e-05,
      "loss": 0.0141,
      "step": 5300
    },
    {
      "epoch": 0.6629491945477075,
      "grad_norm": 0.4524877369403839,
      "learning_rate": 3.7518931570976184e-05,
      "loss": 0.0167,
      "step": 5350
    },
    {
      "epoch": 0.6691449814126395,
      "grad_norm": 0.09342479705810547,
      "learning_rate": 3.683051080820597e-05,
      "loss": 0.0159,
      "step": 5400
    },
    {
      "epoch": 0.6753407682775713,
      "grad_norm": 0.5114253759384155,
      "learning_rate": 3.6142090045435776e-05,
      "loss": 0.0163,
      "step": 5450
    },
    {
      "epoch": 0.6815365551425031,
      "grad_norm": 0.1552111655473709,
      "learning_rate": 3.545366928266557e-05,
      "loss": 0.0153,
      "step": 5500
    },
    {
      "epoch": 0.6877323420074349,
      "grad_norm": 0.7425574064254761,
      "learning_rate": 3.476524851989536e-05,
      "loss": 0.0122,
      "step": 5550
    },
    {
      "epoch": 0.6939281288723668,
      "grad_norm": 0.43186843395233154,
      "learning_rate": 3.4076827757125156e-05,
      "loss": 0.016,
      "step": 5600
    },
    {
      "epoch": 0.7001239157372986,
      "grad_norm": 0.7145938277244568,
      "learning_rate": 3.338840699435495e-05,
      "loss": 0.0169,
      "step": 5650
    },
    {
      "epoch": 0.7063197026022305,
      "grad_norm": 0.09214766323566437,
      "learning_rate": 3.269998623158475e-05,
      "loss": 0.0132,
      "step": 5700
    },
    {
      "epoch": 0.7125154894671624,
      "grad_norm": 1.138299584388733,
      "learning_rate": 3.201156546881454e-05,
      "loss": 0.0129,
      "step": 5750
    },
    {
      "epoch": 0.7187112763320942,
      "grad_norm": 1.3126403093338013,
      "learning_rate": 3.132314470604434e-05,
      "loss": 0.0128,
      "step": 5800
    },
    {
      "epoch": 0.724907063197026,
      "grad_norm": 1.0886186361312866,
      "learning_rate": 3.063472394327413e-05,
      "loss": 0.0119,
      "step": 5850
    },
    {
      "epoch": 0.7311028500619579,
      "grad_norm": 0.23192736506462097,
      "learning_rate": 2.9946303180503927e-05,
      "loss": 0.0118,
      "step": 5900
    },
    {
      "epoch": 0.7372986369268897,
      "grad_norm": 0.9914634227752686,
      "learning_rate": 2.9257882417733723e-05,
      "loss": 0.0138,
      "step": 5950
    },
    {
      "epoch": 0.7434944237918215,
      "grad_norm": 0.29283687472343445,
      "learning_rate": 2.8569461654963515e-05,
      "loss": 0.0123,
      "step": 6000
    },
    {
      "epoch": 0.7496902106567535,
      "grad_norm": 0.9173604249954224,
      "learning_rate": 2.788104089219331e-05,
      "loss": 0.0104,
      "step": 6050
    },
    {
      "epoch": 0.7558859975216853,
      "grad_norm": 0.35434839129447937,
      "learning_rate": 2.7192620129423103e-05,
      "loss": 0.0147,
      "step": 6100
    },
    {
      "epoch": 0.7620817843866171,
      "grad_norm": 0.18677830696105957,
      "learning_rate": 2.65041993666529e-05,
      "loss": 0.0102,
      "step": 6150
    },
    {
      "epoch": 0.7682775712515489,
      "grad_norm": 0.28843411803245544,
      "learning_rate": 2.5815778603882695e-05,
      "loss": 0.0125,
      "step": 6200
    },
    {
      "epoch": 0.7744733581164808,
      "grad_norm": 0.5970668792724609,
      "learning_rate": 2.5127357841112487e-05,
      "loss": 0.0133,
      "step": 6250
    },
    {
      "epoch": 0.7806691449814126,
      "grad_norm": 0.4063868224620819,
      "learning_rate": 2.4438937078342283e-05,
      "loss": 0.0134,
      "step": 6300
    },
    {
      "epoch": 0.7868649318463445,
      "grad_norm": 0.7077623605728149,
      "learning_rate": 2.375051631557208e-05,
      "loss": 0.0125,
      "step": 6350
    },
    {
      "epoch": 0.7930607187112764,
      "grad_norm": 0.056457001715898514,
      "learning_rate": 2.3062095552801874e-05,
      "loss": 0.0119,
      "step": 6400
    },
    {
      "epoch": 0.7992565055762082,
      "grad_norm": 0.0479193776845932,
      "learning_rate": 2.2373674790031666e-05,
      "loss": 0.0134,
      "step": 6450
    },
    {
      "epoch": 0.80545229244114,
      "grad_norm": 0.2201872020959854,
      "learning_rate": 2.1685254027261465e-05,
      "loss": 0.0144,
      "step": 6500
    },
    {
      "epoch": 0.8116480793060719,
      "grad_norm": 0.24738062918186188,
      "learning_rate": 2.0996833264491258e-05,
      "loss": 0.0119,
      "step": 6550
    },
    {
      "epoch": 0.8178438661710037,
      "grad_norm": 0.38614606857299805,
      "learning_rate": 2.0308412501721053e-05,
      "loss": 0.0109,
      "step": 6600
    },
    {
      "epoch": 0.8240396530359355,
      "grad_norm": 0.05887860804796219,
      "learning_rate": 1.9619991738950846e-05,
      "loss": 0.0098,
      "step": 6650
    },
    {
      "epoch": 0.8302354399008675,
      "grad_norm": 0.7075309753417969,
      "learning_rate": 1.893157097618064e-05,
      "loss": 0.0115,
      "step": 6700
    },
    {
      "epoch": 0.8364312267657993,
      "grad_norm": 0.39252975583076477,
      "learning_rate": 1.8243150213410437e-05,
      "loss": 0.0127,
      "step": 6750
    },
    {
      "epoch": 0.8426270136307311,
      "grad_norm": 0.4794786870479584,
      "learning_rate": 1.755472945064023e-05,
      "loss": 0.0102,
      "step": 6800
    },
    {
      "epoch": 0.8488228004956629,
      "grad_norm": 0.39937812089920044,
      "learning_rate": 1.686630868787003e-05,
      "loss": 0.011,
      "step": 6850
    },
    {
      "epoch": 0.8550185873605948,
      "grad_norm": 0.7926356792449951,
      "learning_rate": 1.617788792509982e-05,
      "loss": 0.0114,
      "step": 6900
    },
    {
      "epoch": 0.8612143742255266,
      "grad_norm": 1.0466240644454956,
      "learning_rate": 1.5489467162329617e-05,
      "loss": 0.0116,
      "step": 6950
    },
    {
      "epoch": 0.8674101610904585,
      "grad_norm": 0.6414132714271545,
      "learning_rate": 1.4801046399559412e-05,
      "loss": 0.0102,
      "step": 7000
    },
    {
      "epoch": 0.8736059479553904,
      "grad_norm": 0.8947381377220154,
      "learning_rate": 1.4112625636789206e-05,
      "loss": 0.0104,
      "step": 7050
    },
    {
      "epoch": 0.8798017348203222,
      "grad_norm": 0.8255850076675415,
      "learning_rate": 1.3424204874019e-05,
      "loss": 0.0098,
      "step": 7100
    },
    {
      "epoch": 0.885997521685254,
      "grad_norm": 0.16911354660987854,
      "learning_rate": 1.2735784111248796e-05,
      "loss": 0.01,
      "step": 7150
    },
    {
      "epoch": 0.8921933085501859,
      "grad_norm": 0.04949663579463959,
      "learning_rate": 1.204736334847859e-05,
      "loss": 0.01,
      "step": 7200
    },
    {
      "epoch": 0.8983890954151177,
      "grad_norm": 0.3117857277393341,
      "learning_rate": 1.1358942585708386e-05,
      "loss": 0.0085,
      "step": 7250
    },
    {
      "epoch": 0.9045848822800495,
      "grad_norm": 0.38259008526802063,
      "learning_rate": 1.067052182293818e-05,
      "loss": 0.0106,
      "step": 7300
    },
    {
      "epoch": 0.9107806691449815,
      "grad_norm": 0.423921138048172,
      "learning_rate": 9.982101060167976e-06,
      "loss": 0.0085,
      "step": 7350
    },
    {
      "epoch": 0.9169764560099133,
      "grad_norm": 0.2425990253686905,
      "learning_rate": 9.29368029739777e-06,
      "loss": 0.0092,
      "step": 7400
    },
    {
      "epoch": 0.9231722428748451,
      "grad_norm": 0.17992499470710754,
      "learning_rate": 8.605259534627565e-06,
      "loss": 0.0095,
      "step": 7450
    },
    {
      "epoch": 0.929368029739777,
      "grad_norm": 0.634684681892395,
      "learning_rate": 7.91683877185736e-06,
      "loss": 0.0099,
      "step": 7500
    },
    {
      "epoch": 0.9355638166047088,
      "grad_norm": 0.92573481798172,
      "learning_rate": 7.228418009087154e-06,
      "loss": 0.0113,
      "step": 7550
    },
    {
      "epoch": 0.9417596034696406,
      "grad_norm": 0.07547607272863388,
      "learning_rate": 6.539997246316949e-06,
      "loss": 0.0077,
      "step": 7600
    },
    {
      "epoch": 0.9479553903345725,
      "grad_norm": 0.8026949167251587,
      "learning_rate": 5.851576483546744e-06,
      "loss": 0.0095,
      "step": 7650
    },
    {
      "epoch": 0.9541511771995044,
      "grad_norm": 0.40003785490989685,
      "learning_rate": 5.163155720776539e-06,
      "loss": 0.0077,
      "step": 7700
    },
    {
      "epoch": 0.9603469640644362,
      "grad_norm": 0.1786423921585083,
      "learning_rate": 4.474734958006334e-06,
      "loss": 0.0089,
      "step": 7750
    },
    {
      "epoch": 0.966542750929368,
      "grad_norm": 1.094724416732788,
      "learning_rate": 3.7863141952361286e-06,
      "loss": 0.009,
      "step": 7800
    },
    {
      "epoch": 0.9727385377942999,
      "grad_norm": 0.7275987863540649,
      "learning_rate": 3.0978934324659235e-06,
      "loss": 0.0077,
      "step": 7850
    },
    {
      "epoch": 0.9789343246592317,
      "grad_norm": 0.17183449864387512,
      "learning_rate": 2.409472669695718e-06,
      "loss": 0.0101,
      "step": 7900
    },
    {
      "epoch": 0.9851301115241635,
      "grad_norm": 0.138813316822052,
      "learning_rate": 1.721051906925513e-06,
      "loss": 0.0088,
      "step": 7950
    },
    {
      "epoch": 0.9913258983890955,
      "grad_norm": 0.39123526215553284,
      "learning_rate": 1.0326311441553077e-06,
      "loss": 0.0098,
      "step": 8000
    },
    {
      "epoch": 0.9975216852540273,
      "grad_norm": 0.0659351795911789,
      "learning_rate": 3.442103813851026e-07,
      "loss": 0.0102,
      "step": 8050
    }
  ],
  "logging_steps": 50,
  "max_steps": 8070,
  "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": 0.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}