{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7535460992907801,
  "eval_steps": 500,
  "global_step": 8500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004432624113475178,
      "grad_norm": 0.009513450788341307,
      "learning_rate": 5.066502241635949e-06,
      "loss": 0.0353,
      "step": 50
    },
    {
      "epoch": 0.008865248226950355,
      "grad_norm": 0.12959164414811705,
      "learning_rate": 5.964204463534375e-06,
      "loss": 0.0209,
      "step": 100
    },
    {
      "epoch": 0.013297872340425532,
      "grad_norm": 0.03079899259156006,
      "learning_rate": 6.489326600159014e-06,
      "loss": 0.0114,
      "step": 150
    },
    {
      "epoch": 0.01773049645390071,
      "grad_norm": 1.904334998115993,
      "learning_rate": 6.855414889666249e-06,
      "loss": 0.0426,
      "step": 200
    },
    {
      "epoch": 0.022163120567375887,
      "grad_norm": 0.00029657300723797526,
      "learning_rate": 7.14571142073327e-06,
      "loss": 0.0092,
      "step": 250
    },
    {
      "epoch": 0.026595744680851064,
      "grad_norm": 0.5635451567630495,
      "learning_rate": 7.382704575824005e-06,
      "loss": 0.0207,
      "step": 300
    },
    {
      "epoch": 0.03102836879432624,
      "grad_norm": 17.174480153841266,
      "learning_rate": 7.582965380729319e-06,
      "loss": 0.0128,
      "step": 350
    },
    {
      "epoch": 0.03546099290780142,
      "grad_norm": 0.0029856714585610065,
      "learning_rate": 7.756367076993413e-06,
      "loss": 0.0087,
      "step": 400
    },
    {
      "epoch": 0.0398936170212766,
      "grad_norm": 46.59324697917607,
      "learning_rate": 7.909269732702365e-06,
      "loss": 0.0358,
      "step": 450
    },
    {
      "epoch": 0.044326241134751775,
      "grad_norm": 0.011038539916815544,
      "learning_rate": 8.046011658635775e-06,
      "loss": 0.0099,
      "step": 500
    },
    {
      "epoch": 0.04875886524822695,
      "grad_norm": 0.06122707900561642,
      "learning_rate": 8.169684802472282e-06,
      "loss": 0.0257,
      "step": 550
    },
    {
      "epoch": 0.05319148936170213,
      "grad_norm": 11.17528630173658,
      "learning_rate": 8.28257072561721e-06,
      "loss": 0.0399,
      "step": 600
    },
    {
      "epoch": 0.057624113475177305,
      "grad_norm": 0.004434274436946811,
      "learning_rate": 8.386401186603816e-06,
      "loss": 0.0106,
      "step": 650
    },
    {
      "epoch": 0.06205673758865248,
      "grad_norm": 0.00022486652288303773,
      "learning_rate": 8.482521734002021e-06,
      "loss": 0.0175,
      "step": 700
    },
    {
      "epoch": 0.06648936170212766,
      "grad_norm": 0.08252216933057055,
      "learning_rate": 8.571998643735778e-06,
      "loss": 0.0122,
      "step": 750
    },
    {
      "epoch": 0.07092198581560284,
      "grad_norm": 38.71346721678791,
      "learning_rate": 8.655691228400912e-06,
      "loss": 0.0107,
      "step": 800
    },
    {
      "epoch": 0.07535460992907801,
      "grad_norm": 0.016399532174579123,
      "learning_rate": 8.734302159716567e-06,
      "loss": 0.0272,
      "step": 850
    },
    {
      "epoch": 0.0797872340425532,
      "grad_norm": 13.96210364288869,
      "learning_rate": 8.808413368821617e-06,
      "loss": 0.0162,
      "step": 900
    },
    {
      "epoch": 0.08421985815602837,
      "grad_norm": 20.616657581291367,
      "learning_rate": 8.878512216644468e-06,
      "loss": 0.018,
      "step": 950
    },
    {
      "epoch": 0.08865248226950355,
      "grad_norm": 8.554096191036654,
      "learning_rate": 8.945010936770532e-06,
      "loss": 0.0122,
      "step": 1000
    },
    {
      "epoch": 0.09308510638297872,
      "grad_norm": 9.002995310481221,
      "learning_rate": 9.008261325009621e-06,
      "loss": 0.004,
      "step": 1050
    },
    {
      "epoch": 0.0975177304964539,
      "grad_norm": 0.04332552525052799,
      "learning_rate": 9.068566005374956e-06,
      "loss": 0.0049,
      "step": 1100
    },
    {
      "epoch": 0.10195035460992907,
      "grad_norm": 0.566818534495066,
      "learning_rate": 9.126187187473155e-06,
      "loss": 0.0348,
      "step": 1150
    },
    {
      "epoch": 0.10638297872340426,
      "grad_norm": 2.2102759352121244,
      "learning_rate": 9.181353557126703e-06,
      "loss": 0.0157,
      "step": 1200
    },
    {
      "epoch": 0.11081560283687943,
      "grad_norm": 0.0007883033335850071,
      "learning_rate": 9.233228424991629e-06,
      "loss": 0.0153,
      "step": 1250
    },
    {
      "epoch": 0.11524822695035461,
      "grad_norm": 8.962657174789879,
      "learning_rate": 9.28410340850224e-06,
      "loss": 0.011,
      "step": 1300
    },
    {
      "epoch": 0.1196808510638298,
      "grad_norm": 0.08478258585114734,
      "learning_rate": 9.333055212266742e-06,
      "loss": 0.0193,
      "step": 1350
    },
    {
      "epoch": 0.12411347517730496,
      "grad_norm": 5.126269299634291,
      "learning_rate": 9.380223955900445e-06,
      "loss": 0.0224,
      "step": 1400
    },
    {
      "epoch": 0.12854609929078015,
      "grad_norm": 0.001630326745394445,
      "learning_rate": 9.425734982425631e-06,
      "loss": 0.0198,
      "step": 1450
    },
    {
      "epoch": 0.13297872340425532,
      "grad_norm": 31.3251284414765,
      "learning_rate": 9.469700865634203e-06,
      "loss": 0.0201,
      "step": 1500
    },
    {
      "epoch": 0.1374113475177305,
      "grad_norm": 0.003720201610220148,
      "learning_rate": 9.512223087748166e-06,
      "loss": 0.0386,
      "step": 1550
    },
    {
      "epoch": 0.14184397163120568,
      "grad_norm": 1.4908536371789232,
      "learning_rate": 9.553393450299339e-06,
      "loss": 0.0155,
      "step": 1600
    },
    {
      "epoch": 0.14627659574468085,
      "grad_norm": 0.18557111682138577,
      "learning_rate": 9.593295267571064e-06,
      "loss": 0.0132,
      "step": 1650
    },
    {
      "epoch": 0.15070921985815602,
      "grad_norm": 0.0012351164899243579,
      "learning_rate": 9.632004381614992e-06,
      "loss": 0.0169,
      "step": 1700
    },
    {
      "epoch": 0.1551418439716312,
      "grad_norm": 4.390950581164127,
      "learning_rate": 9.669590029926173e-06,
      "loss": 0.0188,
      "step": 1750
    },
    {
      "epoch": 0.1595744680851064,
      "grad_norm": 1.5823201233659507,
      "learning_rate": 9.706115590720043e-06,
      "loss": 0.0176,
      "step": 1800
    },
    {
      "epoch": 0.16400709219858156,
      "grad_norm": 0.6041784257950227,
      "learning_rate": 9.741639225963296e-06,
      "loss": 0.0076,
      "step": 1850
    },
    {
      "epoch": 0.16843971631205673,
      "grad_norm": 8.096633927567137,
      "learning_rate": 9.776214438542894e-06,
      "loss": 0.0254,
      "step": 1900
    },
    {
      "epoch": 0.17287234042553193,
      "grad_norm": 0.001381384244099819,
      "learning_rate": 9.80989055697396e-06,
      "loss": 0.021,
      "step": 1950
    },
    {
      "epoch": 0.1773049645390071,
      "grad_norm": 0.6173202999510908,
      "learning_rate": 9.842713158668956e-06,
      "loss": 0.0153,
      "step": 2000
    },
    {
      "epoch": 0.18173758865248227,
      "grad_norm": 0.4702304667158159,
      "learning_rate": 9.874724440882673e-06,
      "loss": 0.0322,
      "step": 2050
    },
    {
      "epoch": 0.18617021276595744,
      "grad_norm": 0.007418267836546937,
      "learning_rate": 9.905963546908046e-06,
      "loss": 0.0189,
      "step": 2100
    },
    {
      "epoch": 0.19060283687943264,
      "grad_norm": 0.040611583128721894,
      "learning_rate": 9.936466853848526e-06,
      "loss": 0.004,
      "step": 2150
    },
    {
      "epoch": 0.1950354609929078,
      "grad_norm": 35.80087991837194,
      "learning_rate": 9.96626822727338e-06,
      "loss": 0.022,
      "step": 2200
    },
    {
      "epoch": 0.19946808510638298,
      "grad_norm": 0.44489231698014525,
      "learning_rate": 9.995399247226628e-06,
      "loss": 0.0165,
      "step": 2250
    },
    {
      "epoch": 0.20390070921985815,
      "grad_norm": 0.0002357955980587179,
      "learning_rate": 9.980299448384557e-06,
      "loss": 0.0091,
      "step": 2300
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 0.011156109382490509,
      "learning_rate": 9.955673758865249e-06,
      "loss": 0.0293,
      "step": 2350
    },
    {
      "epoch": 0.2127659574468085,
      "grad_norm": 0.010237311021590336,
      "learning_rate": 9.931048069345942e-06,
      "loss": 0.0267,
      "step": 2400
    },
    {
      "epoch": 0.21719858156028368,
      "grad_norm": 0.1907181340319684,
      "learning_rate": 9.906422379826635e-06,
      "loss": 0.0035,
      "step": 2450
    },
    {
      "epoch": 0.22163120567375885,
      "grad_norm": 7.244482111036572,
      "learning_rate": 9.881796690307329e-06,
      "loss": 0.0129,
      "step": 2500
    },
    {
      "epoch": 0.22606382978723405,
      "grad_norm": 2.614333142067009,
      "learning_rate": 9.857171000788024e-06,
      "loss": 0.0176,
      "step": 2550
    },
    {
      "epoch": 0.23049645390070922,
      "grad_norm": 0.19026758190049964,
      "learning_rate": 9.832545311268717e-06,
      "loss": 0.0118,
      "step": 2600
    },
    {
      "epoch": 0.2349290780141844,
      "grad_norm": 1.926354870277716,
      "learning_rate": 9.80791962174941e-06,
      "loss": 0.0113,
      "step": 2650
    },
    {
      "epoch": 0.2393617021276596,
      "grad_norm": 1.3390029353713434e-05,
      "learning_rate": 9.783293932230104e-06,
      "loss": 0.0095,
      "step": 2700
    },
    {
      "epoch": 0.24379432624113476,
      "grad_norm": 0.25506782394102356,
      "learning_rate": 9.758668242710797e-06,
      "loss": 0.0235,
      "step": 2750
    },
    {
      "epoch": 0.24822695035460993,
      "grad_norm": 0.0011667263533171556,
      "learning_rate": 9.73404255319149e-06,
      "loss": 0.01,
      "step": 2800
    },
    {
      "epoch": 0.2526595744680851,
      "grad_norm": 2.676783820787547,
      "learning_rate": 9.709416863672184e-06,
      "loss": 0.0305,
      "step": 2850
    },
    {
      "epoch": 0.2570921985815603,
      "grad_norm": 0.0019099837220493063,
      "learning_rate": 9.684791174152877e-06,
      "loss": 0.0246,
      "step": 2900
    },
    {
      "epoch": 0.26152482269503546,
      "grad_norm": 0.6898702380926142,
      "learning_rate": 9.66016548463357e-06,
      "loss": 0.0211,
      "step": 2950
    },
    {
      "epoch": 0.26595744680851063,
      "grad_norm": 0.0006438124789484562,
      "learning_rate": 9.635539795114264e-06,
      "loss": 0.008,
      "step": 3000
    },
    {
      "epoch": 0.2703900709219858,
      "grad_norm": 3.819700570925564e-05,
      "learning_rate": 9.610914105594957e-06,
      "loss": 0.0199,
      "step": 3050
    },
    {
      "epoch": 0.274822695035461,
      "grad_norm": 0.019487722143147885,
      "learning_rate": 9.58628841607565e-06,
      "loss": 0.0126,
      "step": 3100
    },
    {
      "epoch": 0.27925531914893614,
      "grad_norm": 0.49075883570668455,
      "learning_rate": 9.561662726556344e-06,
      "loss": 0.0114,
      "step": 3150
    },
    {
      "epoch": 0.28368794326241137,
      "grad_norm": 0.661703227107662,
      "learning_rate": 9.537037037037037e-06,
      "loss": 0.0081,
      "step": 3200
    },
    {
      "epoch": 0.28812056737588654,
      "grad_norm": 0.00012406562558187035,
      "learning_rate": 9.512411347517732e-06,
      "loss": 0.0154,
      "step": 3250
    },
    {
      "epoch": 0.2925531914893617,
      "grad_norm": 10.456271349514031,
      "learning_rate": 9.487785657998426e-06,
      "loss": 0.0056,
      "step": 3300
    },
    {
      "epoch": 0.2969858156028369,
      "grad_norm": 34.88954914653437,
      "learning_rate": 9.463159968479119e-06,
      "loss": 0.0215,
      "step": 3350
    },
    {
      "epoch": 0.30141843971631205,
      "grad_norm": 0.0037401951660654617,
      "learning_rate": 9.43853427895981e-06,
      "loss": 0.0073,
      "step": 3400
    },
    {
      "epoch": 0.3058510638297872,
      "grad_norm": 1.9770145427070573,
      "learning_rate": 9.413908589440504e-06,
      "loss": 0.026,
      "step": 3450
    },
    {
      "epoch": 0.3102836879432624,
      "grad_norm": 0.0005171843885202088,
      "learning_rate": 9.389282899921197e-06,
      "loss": 0.0054,
      "step": 3500
    },
    {
      "epoch": 0.3147163120567376,
      "grad_norm": 6.671945039975605,
      "learning_rate": 9.364657210401892e-06,
      "loss": 0.0193,
      "step": 3550
    },
    {
      "epoch": 0.3191489361702128,
      "grad_norm": 24.199369765973504,
      "learning_rate": 9.340031520882586e-06,
      "loss": 0.0253,
      "step": 3600
    },
    {
      "epoch": 0.32358156028368795,
      "grad_norm": 0.03237450123162466,
      "learning_rate": 9.31540583136328e-06,
      "loss": 0.0108,
      "step": 3650
    },
    {
      "epoch": 0.3280141843971631,
      "grad_norm": 0.06582091922650166,
      "learning_rate": 9.290780141843973e-06,
      "loss": 0.0052,
      "step": 3700
    },
    {
      "epoch": 0.3324468085106383,
      "grad_norm": 1.1046639468259907,
      "learning_rate": 9.266154452324666e-06,
      "loss": 0.0077,
      "step": 3750
    },
    {
      "epoch": 0.33687943262411346,
      "grad_norm": 0.12685978465854558,
      "learning_rate": 9.24152876280536e-06,
      "loss": 0.0055,
      "step": 3800
    },
    {
      "epoch": 0.34131205673758863,
      "grad_norm": 0.001176875548815465,
      "learning_rate": 9.216903073286053e-06,
      "loss": 0.0116,
      "step": 3850
    },
    {
      "epoch": 0.34574468085106386,
      "grad_norm": 19.93923722594081,
      "learning_rate": 9.192277383766746e-06,
      "loss": 0.0339,
      "step": 3900
    },
    {
      "epoch": 0.350177304964539,
      "grad_norm": 0.5136555426144171,
      "learning_rate": 9.16765169424744e-06,
      "loss": 0.0109,
      "step": 3950
    },
    {
      "epoch": 0.3546099290780142,
      "grad_norm": 29.577148937276096,
      "learning_rate": 9.143026004728133e-06,
      "loss": 0.0195,
      "step": 4000
    },
    {
      "epoch": 0.35904255319148937,
      "grad_norm": 5.471652062023114,
      "learning_rate": 9.118400315208826e-06,
      "loss": 0.0075,
      "step": 4050
    },
    {
      "epoch": 0.36347517730496454,
      "grad_norm": 0.0039179745195827426,
      "learning_rate": 9.09377462568952e-06,
      "loss": 0.0203,
      "step": 4100
    },
    {
      "epoch": 0.3679078014184397,
      "grad_norm": 30.214203138409303,
      "learning_rate": 9.069148936170213e-06,
      "loss": 0.0323,
      "step": 4150
    },
    {
      "epoch": 0.3723404255319149,
      "grad_norm": 1.0344365911211837,
      "learning_rate": 9.044523246650908e-06,
      "loss": 0.0252,
      "step": 4200
    },
    {
      "epoch": 0.37677304964539005,
      "grad_norm": 0.6080095692433664,
      "learning_rate": 9.019897557131601e-06,
      "loss": 0.0065,
      "step": 4250
    },
    {
      "epoch": 0.38120567375886527,
      "grad_norm": 1.132001494443736,
      "learning_rate": 8.995271867612294e-06,
      "loss": 0.0137,
      "step": 4300
    },
    {
      "epoch": 0.38563829787234044,
      "grad_norm": 0.6281310332658542,
      "learning_rate": 8.971138691883373e-06,
      "loss": 0.0086,
      "step": 4350
    },
    {
      "epoch": 0.3900709219858156,
      "grad_norm": 2.5248461085388904e-05,
      "learning_rate": 8.946513002364066e-06,
      "loss": 0.0015,
      "step": 4400
    },
    {
      "epoch": 0.3945035460992908,
      "grad_norm": 0.010001162867900915,
      "learning_rate": 8.921887312844761e-06,
      "loss": 0.0174,
      "step": 4450
    },
    {
      "epoch": 0.39893617021276595,
      "grad_norm": 0.3111641106006767,
      "learning_rate": 8.897261623325454e-06,
      "loss": 0.0174,
      "step": 4500
    },
    {
      "epoch": 0.4033687943262411,
      "grad_norm": 0.022375831871911778,
      "learning_rate": 8.872635933806148e-06,
      "loss": 0.0115,
      "step": 4550
    },
    {
      "epoch": 0.4078014184397163,
      "grad_norm": 0.08396303659869327,
      "learning_rate": 8.848010244286841e-06,
      "loss": 0.0136,
      "step": 4600
    },
    {
      "epoch": 0.4122340425531915,
      "grad_norm": 5.854304446661381,
      "learning_rate": 8.823384554767534e-06,
      "loss": 0.0042,
      "step": 4650
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 16.679488984474407,
      "learning_rate": 8.798758865248228e-06,
      "loss": 0.0102,
      "step": 4700
    },
    {
      "epoch": 0.42109929078014185,
      "grad_norm": 0.6379686502927935,
      "learning_rate": 8.774625689519308e-06,
      "loss": 0.019,
      "step": 4750
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 0.05658909698246918,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.0178,
      "step": 4800
    },
    {
      "epoch": 0.4299645390070922,
      "grad_norm": 0.36425987836236606,
      "learning_rate": 8.725374310480694e-06,
      "loss": 0.0212,
      "step": 4850
    },
    {
      "epoch": 0.43439716312056736,
      "grad_norm": 2.0754475949590696,
      "learning_rate": 8.700748620961388e-06,
      "loss": 0.0077,
      "step": 4900
    },
    {
      "epoch": 0.43882978723404253,
      "grad_norm": 0.012181818674097341,
      "learning_rate": 8.676122931442081e-06,
      "loss": 0.0125,
      "step": 4950
    },
    {
      "epoch": 0.4432624113475177,
      "grad_norm": 0.00018403879761046281,
      "learning_rate": 8.651497241922774e-06,
      "loss": 0.0188,
      "step": 5000
    },
    {
      "epoch": 0.44769503546099293,
      "grad_norm": 0.004407047372502831,
      "learning_rate": 8.626871552403468e-06,
      "loss": 0.0073,
      "step": 5050
    },
    {
      "epoch": 0.4521276595744681,
      "grad_norm": 0.012230911143282817,
      "learning_rate": 8.602245862884161e-06,
      "loss": 0.008,
      "step": 5100
    },
    {
      "epoch": 0.45656028368794327,
      "grad_norm": 0.9921871395560984,
      "learning_rate": 8.577620173364854e-06,
      "loss": 0.0168,
      "step": 5150
    },
    {
      "epoch": 0.46099290780141844,
      "grad_norm": 0.1285429154497177,
      "learning_rate": 8.552994483845548e-06,
      "loss": 0.0122,
      "step": 5200
    },
    {
      "epoch": 0.4654255319148936,
      "grad_norm": 3.308588561647349,
      "learning_rate": 8.528861308116628e-06,
      "loss": 0.0158,
      "step": 5250
    },
    {
      "epoch": 0.4698581560283688,
      "grad_norm": 2.8184778898385976,
      "learning_rate": 8.504235618597323e-06,
      "loss": 0.0237,
      "step": 5300
    },
    {
      "epoch": 0.47429078014184395,
      "grad_norm": 0.47692861720548735,
      "learning_rate": 8.479609929078016e-06,
      "loss": 0.0181,
      "step": 5350
    },
    {
      "epoch": 0.4787234042553192,
      "grad_norm": 0.002116990082837995,
      "learning_rate": 8.454984239558708e-06,
      "loss": 0.0235,
      "step": 5400
    },
    {
      "epoch": 0.48315602836879434,
      "grad_norm": 0.016230203235315883,
      "learning_rate": 8.430358550039401e-06,
      "loss": 0.0123,
      "step": 5450
    },
    {
      "epoch": 0.4875886524822695,
      "grad_norm": 0.10798463365752625,
      "learning_rate": 8.405732860520094e-06,
      "loss": 0.0202,
      "step": 5500
    },
    {
      "epoch": 0.4920212765957447,
      "grad_norm": 0.00038212180284718085,
      "learning_rate": 8.381107171000788e-06,
      "loss": 0.0049,
      "step": 5550
    },
    {
      "epoch": 0.49645390070921985,
      "grad_norm": 0.0022870323047176084,
      "learning_rate": 8.356481481481483e-06,
      "loss": 0.0061,
      "step": 5600
    },
    {
      "epoch": 0.500886524822695,
      "grad_norm": 0.4205002106769229,
      "learning_rate": 8.331855791962176e-06,
      "loss": 0.0018,
      "step": 5650
    },
    {
      "epoch": 0.5053191489361702,
      "grad_norm": 0.0003576863161645832,
      "learning_rate": 8.30723010244287e-06,
      "loss": 0.01,
      "step": 5700
    },
    {
      "epoch": 0.5097517730496454,
      "grad_norm": 27.715734272602834,
      "learning_rate": 8.282604412923563e-06,
      "loss": 0.0173,
      "step": 5750
    },
    {
      "epoch": 0.5141843971631206,
      "grad_norm": 0.0029419354875175013,
      "learning_rate": 8.257978723404256e-06,
      "loss": 0.0067,
      "step": 5800
    },
    {
      "epoch": 0.5186170212765957,
      "grad_norm": 27.610931266268892,
      "learning_rate": 8.23335303388495e-06,
      "loss": 0.0116,
      "step": 5850
    },
    {
      "epoch": 0.5230496453900709,
      "grad_norm": 0.004511848611750885,
      "learning_rate": 8.208727344365643e-06,
      "loss": 0.0115,
      "step": 5900
    },
    {
      "epoch": 0.5274822695035462,
      "grad_norm": 0.5244401751220326,
      "learning_rate": 8.184101654846336e-06,
      "loss": 0.0025,
      "step": 5950
    },
    {
      "epoch": 0.5319148936170213,
      "grad_norm": 0.0075670950741103474,
      "learning_rate": 8.15947596532703e-06,
      "loss": 0.0164,
      "step": 6000
    },
    {
      "epoch": 0.5363475177304965,
      "grad_norm": 0.06038816285079991,
      "learning_rate": 8.134850275807723e-06,
      "loss": 0.0079,
      "step": 6050
    },
    {
      "epoch": 0.5407801418439716,
      "grad_norm": 0.0007768875893699433,
      "learning_rate": 8.110224586288416e-06,
      "loss": 0.0052,
      "step": 6100
    },
    {
      "epoch": 0.5452127659574468,
      "grad_norm": 0.2909181418775317,
      "learning_rate": 8.08559889676911e-06,
      "loss": 0.0032,
      "step": 6150
    },
    {
      "epoch": 0.549645390070922,
      "grad_norm": 6.463507993140611,
      "learning_rate": 8.060973207249803e-06,
      "loss": 0.0098,
      "step": 6200
    },
    {
      "epoch": 0.5540780141843972,
      "grad_norm": 0.26740235451618943,
      "learning_rate": 8.036347517730498e-06,
      "loss": 0.0144,
      "step": 6250
    },
    {
      "epoch": 0.5585106382978723,
      "grad_norm": 0.00010577380374890247,
      "learning_rate": 8.011721828211191e-06,
      "loss": 0.0156,
      "step": 6300
    },
    {
      "epoch": 0.5629432624113475,
      "grad_norm": 0.14691208386477084,
      "learning_rate": 7.987096138691885e-06,
      "loss": 0.0152,
      "step": 6350
    },
    {
      "epoch": 0.5673758865248227,
      "grad_norm": 0.07499554653157833,
      "learning_rate": 7.962470449172578e-06,
      "loss": 0.0282,
      "step": 6400
    },
    {
      "epoch": 0.5718085106382979,
      "grad_norm": 0.005398079692845853,
      "learning_rate": 7.93784475965327e-06,
      "loss": 0.0193,
      "step": 6450
    },
    {
      "epoch": 0.5762411347517731,
      "grad_norm": 0.8004218807047324,
      "learning_rate": 7.913219070133963e-06,
      "loss": 0.0059,
      "step": 6500
    },
    {
      "epoch": 0.5806737588652482,
      "grad_norm": 0.0018961263364736887,
      "learning_rate": 7.888593380614658e-06,
      "loss": 0.0192,
      "step": 6550
    },
    {
      "epoch": 0.5851063829787234,
      "grad_norm": 7.167523295933813,
      "learning_rate": 7.863967691095352e-06,
      "loss": 0.0026,
      "step": 6600
    },
    {
      "epoch": 0.5895390070921985,
      "grad_norm": 0.2545664640109103,
      "learning_rate": 7.839342001576045e-06,
      "loss": 0.0178,
      "step": 6650
    },
    {
      "epoch": 0.5939716312056738,
      "grad_norm": 0.0003677116767933612,
      "learning_rate": 7.814716312056738e-06,
      "loss": 0.0127,
      "step": 6700
    },
    {
      "epoch": 0.598404255319149,
      "grad_norm": 1.449028640163906,
      "learning_rate": 7.790090622537432e-06,
      "loss": 0.0353,
      "step": 6750
    },
    {
      "epoch": 0.6028368794326241,
      "grad_norm": 0.018076926628577003,
      "learning_rate": 7.765464933018125e-06,
      "loss": 0.0022,
      "step": 6800
    },
    {
      "epoch": 0.6072695035460993,
      "grad_norm": 1.314844716480283,
      "learning_rate": 7.740839243498818e-06,
      "loss": 0.0125,
      "step": 6850
    },
    {
      "epoch": 0.6117021276595744,
      "grad_norm": 0.05713668183450516,
      "learning_rate": 7.716213553979512e-06,
      "loss": 0.0103,
      "step": 6900
    },
    {
      "epoch": 0.6161347517730497,
      "grad_norm": 0.00012171886896694071,
      "learning_rate": 7.691587864460207e-06,
      "loss": 0.0411,
      "step": 6950
    },
    {
      "epoch": 0.6205673758865248,
      "grad_norm": 0.005050200509222459,
      "learning_rate": 7.666962174940898e-06,
      "loss": 0.0088,
      "step": 7000
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.044231465953618015,
      "learning_rate": 7.642336485421592e-06,
      "loss": 0.0121,
      "step": 7050
    },
    {
      "epoch": 0.6294326241134752,
      "grad_norm": 0.002970509653416846,
      "learning_rate": 7.617710795902286e-06,
      "loss": 0.0115,
      "step": 7100
    },
    {
      "epoch": 0.6338652482269503,
      "grad_norm": 0.061580769377440274,
      "learning_rate": 7.593085106382979e-06,
      "loss": 0.0074,
      "step": 7150
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 0.0030036827619670152,
      "learning_rate": 7.568459416863673e-06,
      "loss": 0.0075,
      "step": 7200
    },
    {
      "epoch": 0.6427304964539007,
      "grad_norm": 0.12485293215547061,
      "learning_rate": 7.543833727344366e-06,
      "loss": 0.0112,
      "step": 7250
    },
    {
      "epoch": 0.6471631205673759,
      "grad_norm": 0.001231009728533709,
      "learning_rate": 7.519208037825059e-06,
      "loss": 0.0068,
      "step": 7300
    },
    {
      "epoch": 0.651595744680851,
      "grad_norm": 1.066825644497329,
      "learning_rate": 7.494582348305754e-06,
      "loss": 0.0219,
      "step": 7350
    },
    {
      "epoch": 0.6560283687943262,
      "grad_norm": 0.18183803085941427,
      "learning_rate": 7.469956658786447e-06,
      "loss": 0.0052,
      "step": 7400
    },
    {
      "epoch": 0.6604609929078015,
      "grad_norm": 7.176261541067922,
      "learning_rate": 7.44533096926714e-06,
      "loss": 0.012,
      "step": 7450
    },
    {
      "epoch": 0.6648936170212766,
      "grad_norm": 0.08470951895549413,
      "learning_rate": 7.420705279747834e-06,
      "loss": 0.005,
      "step": 7500
    },
    {
      "epoch": 0.6693262411347518,
      "grad_norm": 0.43564519071924557,
      "learning_rate": 7.396079590228526e-06,
      "loss": 0.0214,
      "step": 7550
    },
    {
      "epoch": 0.6737588652482269,
      "grad_norm": 0.006827701243280031,
      "learning_rate": 7.37145390070922e-06,
      "loss": 0.0382,
      "step": 7600
    },
    {
      "epoch": 0.6781914893617021,
      "grad_norm": 0.0006360372069413289,
      "learning_rate": 7.346828211189914e-06,
      "loss": 0.0058,
      "step": 7650
    },
    {
      "epoch": 0.6826241134751773,
      "grad_norm": 0.011063982123596255,
      "learning_rate": 7.322202521670607e-06,
      "loss": 0.0074,
      "step": 7700
    },
    {
      "epoch": 0.6870567375886525,
      "grad_norm": 0.8991727722769103,
      "learning_rate": 7.2975768321513005e-06,
      "loss": 0.0108,
      "step": 7750
    },
    {
      "epoch": 0.6914893617021277,
      "grad_norm": 0.012017251234112784,
      "learning_rate": 7.272951142631995e-06,
      "loss": 0.0127,
      "step": 7800
    },
    {
      "epoch": 0.6959219858156028,
      "grad_norm": 0.19374738160794408,
      "learning_rate": 7.248325453112688e-06,
      "loss": 0.0035,
      "step": 7850
    },
    {
      "epoch": 0.700354609929078,
      "grad_norm": 2.341510058377949,
      "learning_rate": 7.223699763593381e-06,
      "loss": 0.0317,
      "step": 7900
    },
    {
      "epoch": 0.7047872340425532,
      "grad_norm": 5.84240259775226e-06,
      "learning_rate": 7.199074074074075e-06,
      "loss": 0.0205,
      "step": 7950
    },
    {
      "epoch": 0.7092198581560284,
      "grad_norm": 0.0022186038540356347,
      "learning_rate": 7.174448384554769e-06,
      "loss": 0.0206,
      "step": 8000
    },
    {
      "epoch": 0.7136524822695035,
      "grad_norm": 2.3575055124131414,
      "learning_rate": 7.149822695035462e-06,
      "loss": 0.0047,
      "step": 8050
    },
    {
      "epoch": 0.7180851063829787,
      "grad_norm": 0.01023308841827913,
      "learning_rate": 7.125197005516155e-06,
      "loss": 0.0293,
      "step": 8100
    },
    {
      "epoch": 0.7225177304964538,
      "grad_norm": 0.006391574218526128,
      "learning_rate": 7.100571315996848e-06,
      "loss": 0.0137,
      "step": 8150
    },
    {
      "epoch": 0.7269503546099291,
      "grad_norm": 1.7478784234914295,
      "learning_rate": 7.0759456264775415e-06,
      "loss": 0.0144,
      "step": 8200
    },
    {
      "epoch": 0.7313829787234043,
      "grad_norm": 0.4480580469017403,
      "learning_rate": 7.051319936958235e-06,
      "loss": 0.0236,
      "step": 8250
    },
    {
      "epoch": 0.7358156028368794,
      "grad_norm": 0.0001810108372971621,
      "learning_rate": 7.0271867612293155e-06,
      "loss": 0.0162,
      "step": 8300
    },
    {
      "epoch": 0.7402482269503546,
      "grad_norm": 0.004944937344540902,
      "learning_rate": 7.002561071710009e-06,
      "loss": 0.0045,
      "step": 8350
    },
    {
      "epoch": 0.7446808510638298,
      "grad_norm": 0.009033810427587684,
      "learning_rate": 6.977935382190701e-06,
      "loss": 0.0021,
      "step": 8400
    },
    {
      "epoch": 0.749113475177305,
      "grad_norm": 0.0007266417865651897,
      "learning_rate": 6.953309692671395e-06,
      "loss": 0.0028,
      "step": 8450
    },
    {
      "epoch": 0.7535460992907801,
      "grad_norm": 0.5472428310965101,
      "learning_rate": 6.928684003152089e-06,
      "loss": 0.0072,
      "step": 8500
    }
  ],
  "logging_steps": 50,
  "max_steps": 22560,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 4250,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}