Files
Supra2-100M-Base/checkpoint-84000/trainer_state.json
ModelHub XC 0e60131b4f 初始化项目,由ModelHub XC社区提供模型
Model: SupraLabs/Supra2-100M-Base
Source: Original Platform
2026-08-05 09:25:13 +08:00

12019 lines
288 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.7340030747301419,
"eval_steps": 3000,
"global_step": 84000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00043690659210127495,
"grad_norm": 1.8826903104782104,
"learning_rate": 2.4500000000000003e-05,
"loss": 9.762939453125,
"step": 50
},
{
"epoch": 0.0008738131842025499,
"grad_norm": 0.9995923042297363,
"learning_rate": 4.9500000000000004e-05,
"loss": 8.305597534179688,
"step": 100
},
{
"epoch": 0.001310719776303825,
"grad_norm": 0.9061496257781982,
"learning_rate": 7.45e-05,
"loss": 7.173495483398438,
"step": 150
},
{
"epoch": 0.0017476263684050998,
"grad_norm": 1.6024847030639648,
"learning_rate": 9.95e-05,
"loss": 6.624603271484375,
"step": 200
},
{
"epoch": 0.002184532960506375,
"grad_norm": 1.2123334407806396,
"learning_rate": 0.0001245,
"loss": 6.243399658203125,
"step": 250
},
{
"epoch": 0.00262143955260765,
"grad_norm": 1.0432322025299072,
"learning_rate": 0.0001495,
"loss": 5.9879296875,
"step": 300
},
{
"epoch": 0.0030583461447089245,
"grad_norm": 0.9280146360397339,
"learning_rate": 0.00017449999999999999,
"loss": 5.776607666015625,
"step": 350
},
{
"epoch": 0.0034952527368101996,
"grad_norm": 1.0289900302886963,
"learning_rate": 0.00019950000000000002,
"loss": 5.5920703125,
"step": 400
},
{
"epoch": 0.003932159328911475,
"grad_norm": 1.0593407154083252,
"learning_rate": 0.0002245,
"loss": 5.437271728515625,
"step": 450
},
{
"epoch": 0.00436906592101275,
"grad_norm": 1.0122662782669067,
"learning_rate": 0.0002495,
"loss": 5.287777709960937,
"step": 500
},
{
"epoch": 0.004805972513114025,
"grad_norm": 0.7662743926048279,
"learning_rate": 0.0002745,
"loss": 5.155453491210937,
"step": 550
},
{
"epoch": 0.0052428791052153,
"grad_norm": 0.7941539287567139,
"learning_rate": 0.0002995,
"loss": 5.030631713867187,
"step": 600
},
{
"epoch": 0.005679785697316574,
"grad_norm": 0.6195730566978455,
"learning_rate": 0.00032450000000000003,
"loss": 4.909389953613282,
"step": 650
},
{
"epoch": 0.006116692289417849,
"grad_norm": 0.6587674617767334,
"learning_rate": 0.0003495,
"loss": 4.786273498535156,
"step": 700
},
{
"epoch": 0.006553598881519124,
"grad_norm": 0.7233771681785583,
"learning_rate": 0.0003745,
"loss": 4.6648342895507815,
"step": 750
},
{
"epoch": 0.006990505473620399,
"grad_norm": 0.5489601492881775,
"learning_rate": 0.0003995,
"loss": 4.555048522949218,
"step": 800
},
{
"epoch": 0.007427412065721674,
"grad_norm": 0.6154086589813232,
"learning_rate": 0.0004245,
"loss": 4.461051330566407,
"step": 850
},
{
"epoch": 0.00786431865782295,
"grad_norm": 0.41333407163619995,
"learning_rate": 0.00044950000000000003,
"loss": 4.374936828613281,
"step": 900
},
{
"epoch": 0.008301225249924224,
"grad_norm": 0.5036063194274902,
"learning_rate": 0.0004745,
"loss": 4.3033209228515625,
"step": 950
},
{
"epoch": 0.0087381318420255,
"grad_norm": 0.3687891364097595,
"learning_rate": 0.0004995,
"loss": 4.233715209960938,
"step": 1000
},
{
"epoch": 0.009175038434126774,
"grad_norm": 0.5433846116065979,
"learning_rate": 0.0005245,
"loss": 4.186253967285157,
"step": 1050
},
{
"epoch": 0.00961194502622805,
"grad_norm": 0.439351350069046,
"learning_rate": 0.0005495,
"loss": 4.139383850097656,
"step": 1100
},
{
"epoch": 0.010048851618329325,
"grad_norm": 0.36215880513191223,
"learning_rate": 0.0005745,
"loss": 4.08899658203125,
"step": 1150
},
{
"epoch": 0.0104857582104306,
"grad_norm": 0.3970701992511749,
"learning_rate": 0.0005995000000000001,
"loss": 4.061238403320313,
"step": 1200
},
{
"epoch": 0.010922664802531873,
"grad_norm": 0.3410918712615967,
"learning_rate": 0.0006245000000000001,
"loss": 4.026336669921875,
"step": 1250
},
{
"epoch": 0.011359571394633148,
"grad_norm": 0.3047930598258972,
"learning_rate": 0.0006495,
"loss": 3.9928851318359375,
"step": 1300
},
{
"epoch": 0.011796477986734423,
"grad_norm": 0.2819700241088867,
"learning_rate": 0.0006745,
"loss": 3.9584146118164063,
"step": 1350
},
{
"epoch": 0.012233384578835698,
"grad_norm": 0.29095014929771423,
"learning_rate": 0.0006995,
"loss": 3.9291363525390626,
"step": 1400
},
{
"epoch": 0.012670291170936973,
"grad_norm": 0.2987957298755646,
"learning_rate": 0.0007245000000000001,
"loss": 3.9090225219726564,
"step": 1450
},
{
"epoch": 0.013107197763038248,
"grad_norm": 0.28419050574302673,
"learning_rate": 0.0007495000000000001,
"loss": 3.890744934082031,
"step": 1500
},
{
"epoch": 0.013544104355139523,
"grad_norm": 0.3039053976535797,
"learning_rate": 0.0007745,
"loss": 3.861073303222656,
"step": 1550
},
{
"epoch": 0.013981010947240798,
"grad_norm": 0.25676754117012024,
"learning_rate": 0.0007995,
"loss": 3.8428253173828124,
"step": 1600
},
{
"epoch": 0.014417917539342073,
"grad_norm": 0.2573896646499634,
"learning_rate": 0.0008245,
"loss": 3.8260653686523436,
"step": 1650
},
{
"epoch": 0.014854824131443348,
"grad_norm": 0.272127240896225,
"learning_rate": 0.0008495000000000001,
"loss": 3.8006341552734373,
"step": 1700
},
{
"epoch": 0.015291730723544623,
"grad_norm": 0.23205944895744324,
"learning_rate": 0.0008745000000000001,
"loss": 3.7862649536132813,
"step": 1750
},
{
"epoch": 0.0157286373156459,
"grad_norm": 0.22847425937652588,
"learning_rate": 0.0008995,
"loss": 3.771934814453125,
"step": 1800
},
{
"epoch": 0.016165543907747174,
"grad_norm": 0.2651786208152771,
"learning_rate": 0.0009245,
"loss": 3.76009521484375,
"step": 1850
},
{
"epoch": 0.01660245049984845,
"grad_norm": 0.24903950095176697,
"learning_rate": 0.0009495,
"loss": 3.7454238891601563,
"step": 1900
},
{
"epoch": 0.017039357091949724,
"grad_norm": 0.24524764716625214,
"learning_rate": 0.0009745000000000001,
"loss": 3.735928955078125,
"step": 1950
},
{
"epoch": 0.017476263684051,
"grad_norm": 0.22325558960437775,
"learning_rate": 0.0009995000000000002,
"loss": 3.7209417724609377,
"step": 2000
},
{
"epoch": 0.017913170276152274,
"grad_norm": 0.23860357701778412,
"learning_rate": 0.001,
"loss": 3.7122769165039062,
"step": 2050
},
{
"epoch": 0.01835007686825355,
"grad_norm": 0.2162218689918518,
"learning_rate": 0.001,
"loss": 3.690401611328125,
"step": 2100
},
{
"epoch": 0.018786983460354824,
"grad_norm": 0.2171812504529953,
"learning_rate": 0.001,
"loss": 3.6825027465820312,
"step": 2150
},
{
"epoch": 0.0192238900524561,
"grad_norm": 0.22677843272686005,
"learning_rate": 0.001,
"loss": 3.6679034423828125,
"step": 2200
},
{
"epoch": 0.019660796644557374,
"grad_norm": 0.23934350907802582,
"learning_rate": 0.001,
"loss": 3.6494000244140623,
"step": 2250
},
{
"epoch": 0.02009770323665865,
"grad_norm": 0.20819289982318878,
"learning_rate": 0.001,
"loss": 3.640907897949219,
"step": 2300
},
{
"epoch": 0.020534609828759924,
"grad_norm": 0.22890810668468475,
"learning_rate": 0.001,
"loss": 3.6309326171875,
"step": 2350
},
{
"epoch": 0.0209715164208612,
"grad_norm": 0.2085026353597641,
"learning_rate": 0.001,
"loss": 3.6261746215820314,
"step": 2400
},
{
"epoch": 0.021408423012962474,
"grad_norm": 0.22136099636554718,
"learning_rate": 0.001,
"loss": 3.614878845214844,
"step": 2450
},
{
"epoch": 0.021845329605063746,
"grad_norm": 0.21210630238056183,
"learning_rate": 0.001,
"loss": 3.599817810058594,
"step": 2500
},
{
"epoch": 0.02228223619716502,
"grad_norm": 0.24638274312019348,
"learning_rate": 0.001,
"loss": 3.591536865234375,
"step": 2550
},
{
"epoch": 0.022719142789266296,
"grad_norm": 0.20501184463500977,
"learning_rate": 0.001,
"loss": 3.58053955078125,
"step": 2600
},
{
"epoch": 0.02315604938136757,
"grad_norm": 0.2101815640926361,
"learning_rate": 0.001,
"loss": 3.5824459838867186,
"step": 2650
},
{
"epoch": 0.023592955973468846,
"grad_norm": 0.22906431555747986,
"learning_rate": 0.001,
"loss": 3.564821472167969,
"step": 2700
},
{
"epoch": 0.02402986256557012,
"grad_norm": 0.20708167552947998,
"learning_rate": 0.001,
"loss": 3.5566305541992187,
"step": 2750
},
{
"epoch": 0.024466769157671396,
"grad_norm": 0.21956811845302582,
"learning_rate": 0.001,
"loss": 3.5507748413085936,
"step": 2800
},
{
"epoch": 0.02490367574977267,
"grad_norm": 0.24124488234519958,
"learning_rate": 0.001,
"loss": 3.5426095581054686,
"step": 2850
},
{
"epoch": 0.025340582341873946,
"grad_norm": 0.20921742916107178,
"learning_rate": 0.001,
"loss": 3.531602783203125,
"step": 2900
},
{
"epoch": 0.02577748893397522,
"grad_norm": 0.19601280987262726,
"learning_rate": 0.001,
"loss": 3.5225894165039064,
"step": 2950
},
{
"epoch": 0.026214395526076496,
"grad_norm": 0.20243392884731293,
"learning_rate": 0.001,
"loss": 3.5183840942382814,
"step": 3000
},
{
"epoch": 0.026214395526076496,
"eval_loss": 3.518667221069336,
"eval_runtime": 5.8945,
"eval_samples_per_second": 165.579,
"eval_steps_per_second": 10.349,
"step": 3000
},
{
"epoch": 0.02665130211817777,
"grad_norm": 0.1983480155467987,
"learning_rate": 0.001,
"loss": 3.5175689697265624,
"step": 3050
},
{
"epoch": 0.027088208710279046,
"grad_norm": 0.20962044596672058,
"learning_rate": 0.001,
"loss": 3.501134033203125,
"step": 3100
},
{
"epoch": 0.02752511530238032,
"grad_norm": 0.20161673426628113,
"learning_rate": 0.001,
"loss": 3.508039855957031,
"step": 3150
},
{
"epoch": 0.027962021894481597,
"grad_norm": 0.21198803186416626,
"learning_rate": 0.001,
"loss": 3.493387451171875,
"step": 3200
},
{
"epoch": 0.02839892848658287,
"grad_norm": 0.21940743923187256,
"learning_rate": 0.001,
"loss": 3.4941644287109375,
"step": 3250
},
{
"epoch": 0.028835835078684147,
"grad_norm": 0.23343515396118164,
"learning_rate": 0.001,
"loss": 3.4835015869140626,
"step": 3300
},
{
"epoch": 0.02927274167078542,
"grad_norm": 0.19761498272418976,
"learning_rate": 0.001,
"loss": 3.482951965332031,
"step": 3350
},
{
"epoch": 0.029709648262886697,
"grad_norm": 0.22693106532096863,
"learning_rate": 0.001,
"loss": 3.475736389160156,
"step": 3400
},
{
"epoch": 0.030146554854987972,
"grad_norm": 0.19969327747821808,
"learning_rate": 0.001,
"loss": 3.4689141845703126,
"step": 3450
},
{
"epoch": 0.030583461447089247,
"grad_norm": 0.23987655341625214,
"learning_rate": 0.001,
"loss": 3.464185791015625,
"step": 3500
},
{
"epoch": 0.031020368039190522,
"grad_norm": 0.20894227921962738,
"learning_rate": 0.001,
"loss": 3.4571575927734375,
"step": 3550
},
{
"epoch": 0.0314572746312918,
"grad_norm": 0.22110916674137115,
"learning_rate": 0.001,
"loss": 3.4559121704101563,
"step": 3600
},
{
"epoch": 0.03189418122339307,
"grad_norm": 0.21295017004013062,
"learning_rate": 0.001,
"loss": 3.451357421875,
"step": 3650
},
{
"epoch": 0.03233108781549435,
"grad_norm": 0.19029361009597778,
"learning_rate": 0.001,
"loss": 3.4517440795898438,
"step": 3700
},
{
"epoch": 0.03276799440759562,
"grad_norm": 0.2307061105966568,
"learning_rate": 0.001,
"loss": 3.4454116821289062,
"step": 3750
},
{
"epoch": 0.0332049009996969,
"grad_norm": 0.2172439694404602,
"learning_rate": 0.001,
"loss": 3.4348458862304687,
"step": 3800
},
{
"epoch": 0.03364180759179817,
"grad_norm": 0.19329442083835602,
"learning_rate": 0.001,
"loss": 3.437733154296875,
"step": 3850
},
{
"epoch": 0.03407871418389945,
"grad_norm": 0.2241591513156891,
"learning_rate": 0.001,
"loss": 3.426455383300781,
"step": 3900
},
{
"epoch": 0.03451562077600072,
"grad_norm": 0.2319568693637848,
"learning_rate": 0.001,
"loss": 3.433238830566406,
"step": 3950
},
{
"epoch": 0.034952527368102,
"grad_norm": 0.1924877166748047,
"learning_rate": 0.001,
"loss": 3.4249606323242188,
"step": 4000
},
{
"epoch": 0.03538943396020327,
"grad_norm": 0.2120707631111145,
"learning_rate": 0.001,
"loss": 3.421075439453125,
"step": 4050
},
{
"epoch": 0.03582634055230455,
"grad_norm": 0.19918771088123322,
"learning_rate": 0.001,
"loss": 3.4203054809570315,
"step": 4100
},
{
"epoch": 0.03626324714440582,
"grad_norm": 0.2003307044506073,
"learning_rate": 0.001,
"loss": 3.411707763671875,
"step": 4150
},
{
"epoch": 0.0367001537365071,
"grad_norm": 0.2270653247833252,
"learning_rate": 0.001,
"loss": 3.4078097534179688,
"step": 4200
},
{
"epoch": 0.03713706032860837,
"grad_norm": 0.21652576327323914,
"learning_rate": 0.001,
"loss": 3.4062127685546875,
"step": 4250
},
{
"epoch": 0.03757396692070965,
"grad_norm": 0.23578432202339172,
"learning_rate": 0.001,
"loss": 3.400606689453125,
"step": 4300
},
{
"epoch": 0.03801087351281092,
"grad_norm": 0.20475627481937408,
"learning_rate": 0.001,
"loss": 3.395752258300781,
"step": 4350
},
{
"epoch": 0.0384477801049122,
"grad_norm": 0.2542702555656433,
"learning_rate": 0.001,
"loss": 3.396254577636719,
"step": 4400
},
{
"epoch": 0.03888468669701347,
"grad_norm": 0.18298493325710297,
"learning_rate": 0.001,
"loss": 3.393648986816406,
"step": 4450
},
{
"epoch": 0.03932159328911475,
"grad_norm": 0.22572273015975952,
"learning_rate": 0.001,
"loss": 3.3865216064453123,
"step": 4500
},
{
"epoch": 0.03975849988121602,
"grad_norm": 0.19484151899814606,
"learning_rate": 0.001,
"loss": 3.3849478149414063,
"step": 4550
},
{
"epoch": 0.0401954064733173,
"grad_norm": 0.19386635720729828,
"learning_rate": 0.001,
"loss": 3.38037109375,
"step": 4600
},
{
"epoch": 0.04063231306541857,
"grad_norm": 0.20124493539333344,
"learning_rate": 0.001,
"loss": 3.382728271484375,
"step": 4650
},
{
"epoch": 0.04106921965751985,
"grad_norm": 0.19963102042675018,
"learning_rate": 0.001,
"loss": 3.381324462890625,
"step": 4700
},
{
"epoch": 0.04150612624962112,
"grad_norm": 0.20611299574375153,
"learning_rate": 0.001,
"loss": 3.3788223266601562,
"step": 4750
},
{
"epoch": 0.0419430328417224,
"grad_norm": 0.19405238330364227,
"learning_rate": 0.001,
"loss": 3.3741259765625,
"step": 4800
},
{
"epoch": 0.04237993943382367,
"grad_norm": 0.20914608240127563,
"learning_rate": 0.001,
"loss": 3.370547180175781,
"step": 4850
},
{
"epoch": 0.04281684602592495,
"grad_norm": 0.19159191846847534,
"learning_rate": 0.001,
"loss": 3.3625115966796875,
"step": 4900
},
{
"epoch": 0.04325375261802622,
"grad_norm": 0.19163894653320312,
"learning_rate": 0.001,
"loss": 3.360504455566406,
"step": 4950
},
{
"epoch": 0.04369065921012749,
"grad_norm": 0.19763918220996857,
"learning_rate": 0.001,
"loss": 3.3601028442382814,
"step": 5000
},
{
"epoch": 0.04412756580222877,
"grad_norm": 0.18974512815475464,
"learning_rate": 0.001,
"loss": 3.3579150390625,
"step": 5050
},
{
"epoch": 0.04456447239433004,
"grad_norm": 0.21315094828605652,
"learning_rate": 0.001,
"loss": 3.355892333984375,
"step": 5100
},
{
"epoch": 0.04500137898643132,
"grad_norm": 0.1904531568288803,
"learning_rate": 0.001,
"loss": 3.360745849609375,
"step": 5150
},
{
"epoch": 0.04543828557853259,
"grad_norm": 0.19907866418361664,
"learning_rate": 0.001,
"loss": 3.357358703613281,
"step": 5200
},
{
"epoch": 0.04587519217063387,
"grad_norm": 0.22703590989112854,
"learning_rate": 0.001,
"loss": 3.356294860839844,
"step": 5250
},
{
"epoch": 0.04631209876273514,
"grad_norm": 0.1892520785331726,
"learning_rate": 0.001,
"loss": 3.3490374755859373,
"step": 5300
},
{
"epoch": 0.04674900535483642,
"grad_norm": 0.18953296542167664,
"learning_rate": 0.001,
"loss": 3.3442022705078127,
"step": 5350
},
{
"epoch": 0.04718591194693769,
"grad_norm": 0.1933782547712326,
"learning_rate": 0.001,
"loss": 3.3514590454101563,
"step": 5400
},
{
"epoch": 0.04762281853903897,
"grad_norm": 0.22573590278625488,
"learning_rate": 0.001,
"loss": 3.3468975830078125,
"step": 5450
},
{
"epoch": 0.04805972513114024,
"grad_norm": 0.2123604267835617,
"learning_rate": 0.001,
"loss": 3.3379788208007812,
"step": 5500
},
{
"epoch": 0.04849663172324152,
"grad_norm": 0.18217475712299347,
"learning_rate": 0.001,
"loss": 3.344074401855469,
"step": 5550
},
{
"epoch": 0.04893353831534279,
"grad_norm": 0.2128428965806961,
"learning_rate": 0.001,
"loss": 3.3349212646484374,
"step": 5600
},
{
"epoch": 0.04937044490744407,
"grad_norm": 0.19785532355308533,
"learning_rate": 0.001,
"loss": 3.327950744628906,
"step": 5650
},
{
"epoch": 0.04980735149954534,
"grad_norm": 0.18241338431835175,
"learning_rate": 0.001,
"loss": 3.334012451171875,
"step": 5700
},
{
"epoch": 0.05024425809164662,
"grad_norm": 0.2020372748374939,
"learning_rate": 0.001,
"loss": 3.3329043579101563,
"step": 5750
},
{
"epoch": 0.05068116468374789,
"grad_norm": 0.2071504294872284,
"learning_rate": 0.001,
"loss": 3.33412353515625,
"step": 5800
},
{
"epoch": 0.05111807127584917,
"grad_norm": 0.20365303754806519,
"learning_rate": 0.001,
"loss": 3.322493591308594,
"step": 5850
},
{
"epoch": 0.05155497786795044,
"grad_norm": 0.21742033958435059,
"learning_rate": 0.001,
"loss": 3.329978332519531,
"step": 5900
},
{
"epoch": 0.05199188446005172,
"grad_norm": 0.18907567858695984,
"learning_rate": 0.001,
"loss": 3.328219299316406,
"step": 5950
},
{
"epoch": 0.05242879105215299,
"grad_norm": 0.18905521929264069,
"learning_rate": 0.001,
"loss": 3.323075866699219,
"step": 6000
},
{
"epoch": 0.05242879105215299,
"eval_loss": 3.3261899948120117,
"eval_runtime": 6.1096,
"eval_samples_per_second": 159.748,
"eval_steps_per_second": 9.984,
"step": 6000
},
{
"epoch": 0.05286569764425427,
"grad_norm": 0.19901056587696075,
"learning_rate": 0.001,
"loss": 3.3208355712890625,
"step": 6050
},
{
"epoch": 0.05330260423635554,
"grad_norm": 0.19175691902637482,
"learning_rate": 0.001,
"loss": 3.3096588134765623,
"step": 6100
},
{
"epoch": 0.05373951082845682,
"grad_norm": 0.20367567241191864,
"learning_rate": 0.001,
"loss": 3.318222961425781,
"step": 6150
},
{
"epoch": 0.05417641742055809,
"grad_norm": 0.20697498321533203,
"learning_rate": 0.001,
"loss": 3.314552917480469,
"step": 6200
},
{
"epoch": 0.05461332401265937,
"grad_norm": 0.23248012363910675,
"learning_rate": 0.001,
"loss": 3.312541809082031,
"step": 6250
},
{
"epoch": 0.05505023060476064,
"grad_norm": 0.21030908823013306,
"learning_rate": 0.001,
"loss": 3.3117620849609377,
"step": 6300
},
{
"epoch": 0.05548713719686192,
"grad_norm": 0.19639821350574493,
"learning_rate": 0.001,
"loss": 3.314966125488281,
"step": 6350
},
{
"epoch": 0.05592404378896319,
"grad_norm": 0.1988273561000824,
"learning_rate": 0.001,
"loss": 3.308790588378906,
"step": 6400
},
{
"epoch": 0.05636095038106447,
"grad_norm": 0.20665325224399567,
"learning_rate": 0.001,
"loss": 3.3096978759765623,
"step": 6450
},
{
"epoch": 0.05679785697316574,
"grad_norm": 0.1974414736032486,
"learning_rate": 0.001,
"loss": 3.3091079711914064,
"step": 6500
},
{
"epoch": 0.057234763565267015,
"grad_norm": 0.19521687924861908,
"learning_rate": 0.001,
"loss": 3.3086837768554687,
"step": 6550
},
{
"epoch": 0.05767167015736829,
"grad_norm": 0.19938398897647858,
"learning_rate": 0.001,
"loss": 3.3038284301757814,
"step": 6600
},
{
"epoch": 0.058108576749469565,
"grad_norm": 0.18721790611743927,
"learning_rate": 0.001,
"loss": 3.3031985473632814,
"step": 6650
},
{
"epoch": 0.05854548334157084,
"grad_norm": 0.20039626955986023,
"learning_rate": 0.001,
"loss": 3.3011334228515623,
"step": 6700
},
{
"epoch": 0.058982389933672115,
"grad_norm": 0.22351515293121338,
"learning_rate": 0.001,
"loss": 3.3007638549804685,
"step": 6750
},
{
"epoch": 0.059419296525773393,
"grad_norm": 0.20300878584384918,
"learning_rate": 0.001,
"loss": 3.2940847778320315,
"step": 6800
},
{
"epoch": 0.059856203117874665,
"grad_norm": 0.20710642635822296,
"learning_rate": 0.001,
"loss": 3.297225036621094,
"step": 6850
},
{
"epoch": 0.060293109709975944,
"grad_norm": 0.1960466355085373,
"learning_rate": 0.001,
"loss": 3.2912692260742187,
"step": 6900
},
{
"epoch": 0.060730016302077215,
"grad_norm": 0.1825372278690338,
"learning_rate": 0.001,
"loss": 3.2928152465820313,
"step": 6950
},
{
"epoch": 0.061166922894178494,
"grad_norm": 0.19677802920341492,
"learning_rate": 0.001,
"loss": 3.2997564697265624,
"step": 7000
},
{
"epoch": 0.061603829486279765,
"grad_norm": 0.21398645639419556,
"learning_rate": 0.001,
"loss": 3.293714904785156,
"step": 7050
},
{
"epoch": 0.062040736078381044,
"grad_norm": 0.20046885311603546,
"learning_rate": 0.001,
"loss": 3.2939657592773437,
"step": 7100
},
{
"epoch": 0.062477642670482315,
"grad_norm": 0.1991434544324875,
"learning_rate": 0.001,
"loss": 3.284585266113281,
"step": 7150
},
{
"epoch": 0.0629145492625836,
"grad_norm": 0.20454499125480652,
"learning_rate": 0.001,
"loss": 3.2882635498046877,
"step": 7200
},
{
"epoch": 0.06335145585468487,
"grad_norm": 0.20718543231487274,
"learning_rate": 0.001,
"loss": 3.283497619628906,
"step": 7250
},
{
"epoch": 0.06378836244678614,
"grad_norm": 0.1971973031759262,
"learning_rate": 0.001,
"loss": 3.284901123046875,
"step": 7300
},
{
"epoch": 0.06422526903888742,
"grad_norm": 0.18264292180538177,
"learning_rate": 0.001,
"loss": 3.285682373046875,
"step": 7350
},
{
"epoch": 0.0646621756309887,
"grad_norm": 0.2100597620010376,
"learning_rate": 0.001,
"loss": 3.284263000488281,
"step": 7400
},
{
"epoch": 0.06509908222308997,
"grad_norm": 0.19299328327178955,
"learning_rate": 0.001,
"loss": 3.283685302734375,
"step": 7450
},
{
"epoch": 0.06553598881519124,
"grad_norm": 0.18712171912193298,
"learning_rate": 0.001,
"loss": 3.279105224609375,
"step": 7500
},
{
"epoch": 0.06597289540729252,
"grad_norm": 0.18977321684360504,
"learning_rate": 0.001,
"loss": 3.287223205566406,
"step": 7550
},
{
"epoch": 0.0664098019993938,
"grad_norm": 0.22269035875797272,
"learning_rate": 0.001,
"loss": 3.2770404052734374,
"step": 7600
},
{
"epoch": 0.06684670859149507,
"grad_norm": 0.18847301602363586,
"learning_rate": 0.001,
"loss": 3.2692807006835936,
"step": 7650
},
{
"epoch": 0.06728361518359634,
"grad_norm": 0.21769355237483978,
"learning_rate": 0.001,
"loss": 3.2741900634765626,
"step": 7700
},
{
"epoch": 0.06772052177569762,
"grad_norm": 0.18970917165279388,
"learning_rate": 0.001,
"loss": 3.2736630249023437,
"step": 7750
},
{
"epoch": 0.0681574283677989,
"grad_norm": 0.19735926389694214,
"learning_rate": 0.001,
"loss": 3.27821044921875,
"step": 7800
},
{
"epoch": 0.06859433495990017,
"grad_norm": 0.19194583594799042,
"learning_rate": 0.001,
"loss": 3.2705389404296876,
"step": 7850
},
{
"epoch": 0.06903124155200144,
"grad_norm": 0.2058054655790329,
"learning_rate": 0.001,
"loss": 3.2689944458007814,
"step": 7900
},
{
"epoch": 0.06946814814410272,
"grad_norm": 0.18322297930717468,
"learning_rate": 0.001,
"loss": 3.2653887939453123,
"step": 7950
},
{
"epoch": 0.069905054736204,
"grad_norm": 0.1952529400587082,
"learning_rate": 0.001,
"loss": 3.2640380859375,
"step": 8000
},
{
"epoch": 0.07034196132830527,
"grad_norm": 0.1916273683309555,
"learning_rate": 0.001,
"loss": 3.274902648925781,
"step": 8050
},
{
"epoch": 0.07077886792040654,
"grad_norm": 0.21100108325481415,
"learning_rate": 0.001,
"loss": 3.265535583496094,
"step": 8100
},
{
"epoch": 0.07121577451250782,
"grad_norm": 0.2012566179037094,
"learning_rate": 0.001,
"loss": 3.2651483154296876,
"step": 8150
},
{
"epoch": 0.0716526811046091,
"grad_norm": 0.1913347840309143,
"learning_rate": 0.001,
"loss": 3.2659671020507814,
"step": 8200
},
{
"epoch": 0.07208958769671037,
"grad_norm": 0.19437524676322937,
"learning_rate": 0.001,
"loss": 3.268343811035156,
"step": 8250
},
{
"epoch": 0.07252649428881164,
"grad_norm": 0.20302678644657135,
"learning_rate": 0.001,
"loss": 3.270460510253906,
"step": 8300
},
{
"epoch": 0.07296340088091291,
"grad_norm": 0.1874990612268448,
"learning_rate": 0.001,
"loss": 3.2629928588867188,
"step": 8350
},
{
"epoch": 0.0734003074730142,
"grad_norm": 0.19520102441310883,
"learning_rate": 0.001,
"loss": 3.265036315917969,
"step": 8400
},
{
"epoch": 0.07383721406511547,
"grad_norm": 0.18731839954853058,
"learning_rate": 0.001,
"loss": 3.2578021240234376,
"step": 8450
},
{
"epoch": 0.07427412065721674,
"grad_norm": 0.20026656985282898,
"learning_rate": 0.001,
"loss": 3.2632598876953125,
"step": 8500
},
{
"epoch": 0.07471102724931801,
"grad_norm": 0.20248965919017792,
"learning_rate": 0.001,
"loss": 3.260670471191406,
"step": 8550
},
{
"epoch": 0.0751479338414193,
"grad_norm": 0.20652979612350464,
"learning_rate": 0.001,
"loss": 3.2585479736328127,
"step": 8600
},
{
"epoch": 0.07558484043352057,
"grad_norm": 0.18648388981819153,
"learning_rate": 0.001,
"loss": 3.259482727050781,
"step": 8650
},
{
"epoch": 0.07602174702562184,
"grad_norm": 0.20123572647571564,
"learning_rate": 0.001,
"loss": 3.2517401123046876,
"step": 8700
},
{
"epoch": 0.07645865361772311,
"grad_norm": 0.20070253312587738,
"learning_rate": 0.001,
"loss": 3.2553228759765624,
"step": 8750
},
{
"epoch": 0.0768955602098244,
"grad_norm": 0.19734439253807068,
"learning_rate": 0.001,
"loss": 3.2518017578125,
"step": 8800
},
{
"epoch": 0.07733246680192567,
"grad_norm": 0.2070627510547638,
"learning_rate": 0.001,
"loss": 3.251908874511719,
"step": 8850
},
{
"epoch": 0.07776937339402694,
"grad_norm": 0.19075073301792145,
"learning_rate": 0.001,
"loss": 3.2540167236328124,
"step": 8900
},
{
"epoch": 0.07820627998612821,
"grad_norm": 0.2121705412864685,
"learning_rate": 0.001,
"loss": 3.253029479980469,
"step": 8950
},
{
"epoch": 0.0786431865782295,
"grad_norm": 0.21024636924266815,
"learning_rate": 0.001,
"loss": 3.2557443237304686,
"step": 9000
},
{
"epoch": 0.0786431865782295,
"eval_loss": 3.257852554321289,
"eval_runtime": 6.3455,
"eval_samples_per_second": 153.811,
"eval_steps_per_second": 9.613,
"step": 9000
},
{
"epoch": 0.07908009317033077,
"grad_norm": 0.18182849884033203,
"learning_rate": 0.001,
"loss": 3.255887451171875,
"step": 9050
},
{
"epoch": 0.07951699976243204,
"grad_norm": 0.20949381589889526,
"learning_rate": 0.001,
"loss": 3.2509515380859373,
"step": 9100
},
{
"epoch": 0.07995390635453331,
"grad_norm": 0.1916966736316681,
"learning_rate": 0.001,
"loss": 3.2475936889648436,
"step": 9150
},
{
"epoch": 0.0803908129466346,
"grad_norm": 0.195968896150589,
"learning_rate": 0.001,
"loss": 3.2473486328125,
"step": 9200
},
{
"epoch": 0.08082771953873587,
"grad_norm": 0.19571411609649658,
"learning_rate": 0.001,
"loss": 3.2515054321289063,
"step": 9250
},
{
"epoch": 0.08126462613083714,
"grad_norm": 0.1898670643568039,
"learning_rate": 0.001,
"loss": 3.248476257324219,
"step": 9300
},
{
"epoch": 0.08170153272293841,
"grad_norm": 0.1957128793001175,
"learning_rate": 0.001,
"loss": 3.2485791015625,
"step": 9350
},
{
"epoch": 0.0821384393150397,
"grad_norm": 0.1989963799715042,
"learning_rate": 0.001,
"loss": 3.2457614135742188,
"step": 9400
},
{
"epoch": 0.08257534590714097,
"grad_norm": 0.18394291400909424,
"learning_rate": 0.001,
"loss": 3.245361328125,
"step": 9450
},
{
"epoch": 0.08301225249924224,
"grad_norm": 0.17211784422397614,
"learning_rate": 0.001,
"loss": 3.2452508544921876,
"step": 9500
},
{
"epoch": 0.08344915909134351,
"grad_norm": 0.20545506477355957,
"learning_rate": 0.001,
"loss": 3.2405926513671877,
"step": 9550
},
{
"epoch": 0.0838860656834448,
"grad_norm": 0.20575234293937683,
"learning_rate": 0.001,
"loss": 3.2421484375,
"step": 9600
},
{
"epoch": 0.08432297227554607,
"grad_norm": 0.19145144522190094,
"learning_rate": 0.001,
"loss": 3.2430648803710938,
"step": 9650
},
{
"epoch": 0.08475987886764734,
"grad_norm": 0.20513130724430084,
"learning_rate": 0.001,
"loss": 3.241943664550781,
"step": 9700
},
{
"epoch": 0.08519678545974861,
"grad_norm": 0.18931901454925537,
"learning_rate": 0.001,
"loss": 3.242476806640625,
"step": 9750
},
{
"epoch": 0.0856336920518499,
"grad_norm": 0.2036847621202469,
"learning_rate": 0.001,
"loss": 3.2381185913085937,
"step": 9800
},
{
"epoch": 0.08607059864395117,
"grad_norm": 0.18806156516075134,
"learning_rate": 0.001,
"loss": 3.2394903564453124,
"step": 9850
},
{
"epoch": 0.08650750523605244,
"grad_norm": 0.18992380797863007,
"learning_rate": 0.001,
"loss": 3.2361404418945314,
"step": 9900
},
{
"epoch": 0.08694441182815371,
"grad_norm": 0.1844012290239334,
"learning_rate": 0.001,
"loss": 3.238492126464844,
"step": 9950
},
{
"epoch": 0.08738131842025498,
"grad_norm": 0.19030481576919556,
"learning_rate": 0.001,
"loss": 3.2360128784179687,
"step": 10000
},
{
"epoch": 0.08781822501235627,
"grad_norm": 0.20944413542747498,
"learning_rate": 0.001,
"loss": 3.23896728515625,
"step": 10050
},
{
"epoch": 0.08825513160445754,
"grad_norm": 0.1730359047651291,
"learning_rate": 0.001,
"loss": 3.233212890625,
"step": 10100
},
{
"epoch": 0.08869203819655881,
"grad_norm": 0.18298979103565216,
"learning_rate": 0.001,
"loss": 3.2336996459960936,
"step": 10150
},
{
"epoch": 0.08912894478866008,
"grad_norm": 0.1836487203836441,
"learning_rate": 0.001,
"loss": 3.239130859375,
"step": 10200
},
{
"epoch": 0.08956585138076137,
"grad_norm": 0.20185187458992004,
"learning_rate": 0.001,
"loss": 3.2354965209960938,
"step": 10250
},
{
"epoch": 0.09000275797286264,
"grad_norm": 0.20701956748962402,
"learning_rate": 0.001,
"loss": 3.2316860961914062,
"step": 10300
},
{
"epoch": 0.09043966456496391,
"grad_norm": 0.17721368372440338,
"learning_rate": 0.001,
"loss": 3.2350833129882814,
"step": 10350
},
{
"epoch": 0.09087657115706518,
"grad_norm": 0.1922338306903839,
"learning_rate": 0.001,
"loss": 3.2318270874023436,
"step": 10400
},
{
"epoch": 0.09131347774916647,
"grad_norm": 0.1857219636440277,
"learning_rate": 0.001,
"loss": 3.2357696533203124,
"step": 10450
},
{
"epoch": 0.09175038434126774,
"grad_norm": 0.18690387904644012,
"learning_rate": 0.001,
"loss": 3.233714599609375,
"step": 10500
},
{
"epoch": 0.09218729093336901,
"grad_norm": 0.22552669048309326,
"learning_rate": 0.001,
"loss": 3.2236834716796876,
"step": 10550
},
{
"epoch": 0.09262419752547028,
"grad_norm": 0.2104293406009674,
"learning_rate": 0.001,
"loss": 3.2278411865234373,
"step": 10600
},
{
"epoch": 0.09306110411757157,
"grad_norm": 0.1924365907907486,
"learning_rate": 0.001,
"loss": 3.2275039672851564,
"step": 10650
},
{
"epoch": 0.09349801070967284,
"grad_norm": 0.17614150047302246,
"learning_rate": 0.001,
"loss": 3.2285580444335937,
"step": 10700
},
{
"epoch": 0.09393491730177411,
"grad_norm": 0.1965877264738083,
"learning_rate": 0.001,
"loss": 3.23059814453125,
"step": 10750
},
{
"epoch": 0.09437182389387538,
"grad_norm": 0.17845283448696136,
"learning_rate": 0.001,
"loss": 3.225852355957031,
"step": 10800
},
{
"epoch": 0.09480873048597667,
"grad_norm": 0.1709090620279312,
"learning_rate": 0.001,
"loss": 3.2312142944335935,
"step": 10850
},
{
"epoch": 0.09524563707807794,
"grad_norm": 0.17048615217208862,
"learning_rate": 0.001,
"loss": 3.22577392578125,
"step": 10900
},
{
"epoch": 0.09568254367017921,
"grad_norm": 0.2272968739271164,
"learning_rate": 0.001,
"loss": 3.2276971435546873,
"step": 10950
},
{
"epoch": 0.09611945026228048,
"grad_norm": 0.19360628724098206,
"learning_rate": 0.001,
"loss": 3.2247369384765623,
"step": 11000
},
{
"epoch": 0.09655635685438177,
"grad_norm": 0.20023608207702637,
"learning_rate": 0.001,
"loss": 3.2287957763671873,
"step": 11050
},
{
"epoch": 0.09699326344648304,
"grad_norm": 0.19376908242702484,
"learning_rate": 0.001,
"loss": 3.2265338134765624,
"step": 11100
},
{
"epoch": 0.09743017003858431,
"grad_norm": 0.196820467710495,
"learning_rate": 0.001,
"loss": 3.218832702636719,
"step": 11150
},
{
"epoch": 0.09786707663068558,
"grad_norm": 0.1909375935792923,
"learning_rate": 0.001,
"loss": 3.2203463745117187,
"step": 11200
},
{
"epoch": 0.09830398322278687,
"grad_norm": 0.18288679420948029,
"learning_rate": 0.001,
"loss": 3.218720703125,
"step": 11250
},
{
"epoch": 0.09874088981488814,
"grad_norm": 0.18179824948310852,
"learning_rate": 0.001,
"loss": 3.217382507324219,
"step": 11300
},
{
"epoch": 0.09917779640698941,
"grad_norm": 0.18757817149162292,
"learning_rate": 0.001,
"loss": 3.222406311035156,
"step": 11350
},
{
"epoch": 0.09961470299909068,
"grad_norm": 0.19598425924777985,
"learning_rate": 0.001,
"loss": 3.221529235839844,
"step": 11400
},
{
"epoch": 0.10005160959119196,
"grad_norm": 0.1995646357536316,
"learning_rate": 0.001,
"loss": 3.212632141113281,
"step": 11450
},
{
"epoch": 0.10048851618329324,
"grad_norm": 0.1981515884399414,
"learning_rate": 0.001,
"loss": 3.2169342041015625,
"step": 11500
},
{
"epoch": 0.10092542277539451,
"grad_norm": 0.19607605040073395,
"learning_rate": 0.001,
"loss": 3.212418212890625,
"step": 11550
},
{
"epoch": 0.10136232936749578,
"grad_norm": 0.1733725219964981,
"learning_rate": 0.001,
"loss": 3.221158447265625,
"step": 11600
},
{
"epoch": 0.10179923595959706,
"grad_norm": 0.18705779314041138,
"learning_rate": 0.001,
"loss": 3.2119464111328124,
"step": 11650
},
{
"epoch": 0.10223614255169834,
"grad_norm": 0.19549337029457092,
"learning_rate": 0.001,
"loss": 3.2163186645507813,
"step": 11700
},
{
"epoch": 0.10267304914379961,
"grad_norm": 0.21174098551273346,
"learning_rate": 0.001,
"loss": 3.22341552734375,
"step": 11750
},
{
"epoch": 0.10310995573590089,
"grad_norm": 0.2125338762998581,
"learning_rate": 0.001,
"loss": 3.215631408691406,
"step": 11800
},
{
"epoch": 0.10354686232800216,
"grad_norm": 0.1913818120956421,
"learning_rate": 0.001,
"loss": 3.215351867675781,
"step": 11850
},
{
"epoch": 0.10398376892010344,
"grad_norm": 0.1795399934053421,
"learning_rate": 0.001,
"loss": 3.209778137207031,
"step": 11900
},
{
"epoch": 0.10442067551220471,
"grad_norm": 0.18420544266700745,
"learning_rate": 0.001,
"loss": 3.214957275390625,
"step": 11950
},
{
"epoch": 0.10485758210430599,
"grad_norm": 0.1968565434217453,
"learning_rate": 0.001,
"loss": 3.2145254516601565,
"step": 12000
},
{
"epoch": 0.10485758210430599,
"eval_loss": 3.2212815284729004,
"eval_runtime": 6.0309,
"eval_samples_per_second": 161.835,
"eval_steps_per_second": 10.115,
"step": 12000
},
{
"epoch": 0.10529448869640726,
"grad_norm": 0.2099006026983261,
"learning_rate": 0.001,
"loss": 3.2097091674804688,
"step": 12050
},
{
"epoch": 0.10573139528850854,
"grad_norm": 0.20223559439182281,
"learning_rate": 0.001,
"loss": 3.2140032958984377,
"step": 12100
},
{
"epoch": 0.10616830188060981,
"grad_norm": 0.2003611922264099,
"learning_rate": 0.001,
"loss": 3.2128176879882813,
"step": 12150
},
{
"epoch": 0.10660520847271109,
"grad_norm": 0.17242281138896942,
"learning_rate": 0.001,
"loss": 3.21211669921875,
"step": 12200
},
{
"epoch": 0.10704211506481236,
"grad_norm": 0.18861070275306702,
"learning_rate": 0.001,
"loss": 3.2088800048828126,
"step": 12250
},
{
"epoch": 0.10747902165691364,
"grad_norm": 0.1851765513420105,
"learning_rate": 0.001,
"loss": 3.2097900390625,
"step": 12300
},
{
"epoch": 0.10791592824901491,
"grad_norm": 0.18498055636882782,
"learning_rate": 0.001,
"loss": 3.2144100952148436,
"step": 12350
},
{
"epoch": 0.10835283484111619,
"grad_norm": 0.18698084354400635,
"learning_rate": 0.001,
"loss": 3.20849609375,
"step": 12400
},
{
"epoch": 0.10878974143321746,
"grad_norm": 0.19474317133426666,
"learning_rate": 0.001,
"loss": 3.211873779296875,
"step": 12450
},
{
"epoch": 0.10922664802531874,
"grad_norm": 0.17884086072444916,
"learning_rate": 0.001,
"loss": 3.2136822509765626,
"step": 12500
},
{
"epoch": 0.10966355461742001,
"grad_norm": 0.19080394506454468,
"learning_rate": 0.001,
"loss": 3.2123370361328125,
"step": 12550
},
{
"epoch": 0.11010046120952129,
"grad_norm": 0.1898597925901413,
"learning_rate": 0.001,
"loss": 3.2091488647460937,
"step": 12600
},
{
"epoch": 0.11053736780162256,
"grad_norm": 0.18075662851333618,
"learning_rate": 0.001,
"loss": 3.2083639526367187,
"step": 12650
},
{
"epoch": 0.11097427439372384,
"grad_norm": 0.1833634078502655,
"learning_rate": 0.001,
"loss": 3.2076034545898438,
"step": 12700
},
{
"epoch": 0.11141118098582511,
"grad_norm": 0.19021724164485931,
"learning_rate": 0.001,
"loss": 3.203535461425781,
"step": 12750
},
{
"epoch": 0.11184808757792639,
"grad_norm": 0.19512109458446503,
"learning_rate": 0.001,
"loss": 3.2149392700195314,
"step": 12800
},
{
"epoch": 0.11228499417002766,
"grad_norm": 0.19543170928955078,
"learning_rate": 0.001,
"loss": 3.2061972045898437,
"step": 12850
},
{
"epoch": 0.11272190076212894,
"grad_norm": 0.18879614770412445,
"learning_rate": 0.001,
"loss": 3.20185302734375,
"step": 12900
},
{
"epoch": 0.11315880735423021,
"grad_norm": 0.17561078071594238,
"learning_rate": 0.001,
"loss": 3.19679443359375,
"step": 12950
},
{
"epoch": 0.11359571394633149,
"grad_norm": 0.17411254346370697,
"learning_rate": 0.001,
"loss": 3.206636657714844,
"step": 13000
},
{
"epoch": 0.11403262053843276,
"grad_norm": 0.19477669894695282,
"learning_rate": 0.001,
"loss": 3.2047265625,
"step": 13050
},
{
"epoch": 0.11446952713053403,
"grad_norm": 0.2075432687997818,
"learning_rate": 0.001,
"loss": 3.2073565673828126,
"step": 13100
},
{
"epoch": 0.11490643372263531,
"grad_norm": 0.19364356994628906,
"learning_rate": 0.001,
"loss": 3.2007147216796876,
"step": 13150
},
{
"epoch": 0.11534334031473659,
"grad_norm": 0.19227240979671478,
"learning_rate": 0.001,
"loss": 3.2039303588867187,
"step": 13200
},
{
"epoch": 0.11578024690683786,
"grad_norm": 0.1965664029121399,
"learning_rate": 0.001,
"loss": 3.204370422363281,
"step": 13250
},
{
"epoch": 0.11621715349893913,
"grad_norm": 0.19625689089298248,
"learning_rate": 0.001,
"loss": 3.2018893432617186,
"step": 13300
},
{
"epoch": 0.11665406009104042,
"grad_norm": 0.1872740089893341,
"learning_rate": 0.001,
"loss": 3.1984957885742187,
"step": 13350
},
{
"epoch": 0.11709096668314169,
"grad_norm": 0.17404606938362122,
"learning_rate": 0.001,
"loss": 3.2010247802734373,
"step": 13400
},
{
"epoch": 0.11752787327524296,
"grad_norm": 0.20951177179813385,
"learning_rate": 0.001,
"loss": 3.2027737426757814,
"step": 13450
},
{
"epoch": 0.11796477986734423,
"grad_norm": 0.20374071598052979,
"learning_rate": 0.001,
"loss": 3.1948858642578126,
"step": 13500
},
{
"epoch": 0.11840168645944552,
"grad_norm": 0.21140068769454956,
"learning_rate": 0.001,
"loss": 3.206798095703125,
"step": 13550
},
{
"epoch": 0.11883859305154679,
"grad_norm": 0.17180323600769043,
"learning_rate": 0.001,
"loss": 3.197232666015625,
"step": 13600
},
{
"epoch": 0.11927549964364806,
"grad_norm": 0.1982830911874771,
"learning_rate": 0.001,
"loss": 3.196986083984375,
"step": 13650
},
{
"epoch": 0.11971240623574933,
"grad_norm": 0.19306933879852295,
"learning_rate": 0.001,
"loss": 3.1933523559570314,
"step": 13700
},
{
"epoch": 0.12014931282785062,
"grad_norm": 0.21267230808734894,
"learning_rate": 0.001,
"loss": 3.2039981079101563,
"step": 13750
},
{
"epoch": 0.12058621941995189,
"grad_norm": 0.2193242907524109,
"learning_rate": 0.001,
"loss": 3.1955245971679687,
"step": 13800
},
{
"epoch": 0.12102312601205316,
"grad_norm": 0.23339790105819702,
"learning_rate": 0.001,
"loss": 3.1981427001953127,
"step": 13850
},
{
"epoch": 0.12146003260415443,
"grad_norm": 0.22573474049568176,
"learning_rate": 0.001,
"loss": 3.196231689453125,
"step": 13900
},
{
"epoch": 0.12189693919625572,
"grad_norm": 0.19238780438899994,
"learning_rate": 0.001,
"loss": 3.1946224975585937,
"step": 13950
},
{
"epoch": 0.12233384578835699,
"grad_norm": 0.202445387840271,
"learning_rate": 0.001,
"loss": 3.194338073730469,
"step": 14000
},
{
"epoch": 0.12277075238045826,
"grad_norm": 0.18492762744426727,
"learning_rate": 0.001,
"loss": 3.2063372802734373,
"step": 14050
},
{
"epoch": 0.12320765897255953,
"grad_norm": 0.20408518612384796,
"learning_rate": 0.001,
"loss": 3.1926824951171877,
"step": 14100
},
{
"epoch": 0.12364456556466082,
"grad_norm": 0.19603824615478516,
"learning_rate": 0.001,
"loss": 3.1970639038085937,
"step": 14150
},
{
"epoch": 0.12408147215676209,
"grad_norm": 0.18211990594863892,
"learning_rate": 0.001,
"loss": 3.193045654296875,
"step": 14200
},
{
"epoch": 0.12451837874886336,
"grad_norm": 0.21678605675697327,
"learning_rate": 0.001,
"loss": 3.1917852783203124,
"step": 14250
},
{
"epoch": 0.12495528534096463,
"grad_norm": 0.19074733555316925,
"learning_rate": 0.001,
"loss": 3.1935299682617186,
"step": 14300
},
{
"epoch": 0.12539219193306592,
"grad_norm": 0.17810741066932678,
"learning_rate": 0.001,
"loss": 3.1906802368164064,
"step": 14350
},
{
"epoch": 0.1258290985251672,
"grad_norm": 0.19822879135608673,
"learning_rate": 0.001,
"loss": 3.1980941772460936,
"step": 14400
},
{
"epoch": 0.12626600511726846,
"grad_norm": 0.20151387155056,
"learning_rate": 0.001,
"loss": 3.1888507080078123,
"step": 14450
},
{
"epoch": 0.12670291170936973,
"grad_norm": 0.1979341208934784,
"learning_rate": 0.001,
"loss": 3.1934823608398437,
"step": 14500
},
{
"epoch": 0.127139818301471,
"grad_norm": 0.1844247728586197,
"learning_rate": 0.001,
"loss": 3.1945904541015624,
"step": 14550
},
{
"epoch": 0.12757672489357227,
"grad_norm": 0.1843341737985611,
"learning_rate": 0.001,
"loss": 3.1908944702148436,
"step": 14600
},
{
"epoch": 0.12801363148567357,
"grad_norm": 0.19181329011917114,
"learning_rate": 0.001,
"loss": 3.186395263671875,
"step": 14650
},
{
"epoch": 0.12845053807777485,
"grad_norm": 0.1925257444381714,
"learning_rate": 0.001,
"loss": 3.187642822265625,
"step": 14700
},
{
"epoch": 0.12888744466987612,
"grad_norm": 0.17801596224308014,
"learning_rate": 0.001,
"loss": 3.18369140625,
"step": 14750
},
{
"epoch": 0.1293243512619774,
"grad_norm": 0.20504137873649597,
"learning_rate": 0.001,
"loss": 3.1881332397460938,
"step": 14800
},
{
"epoch": 0.12976125785407866,
"grad_norm": 0.18415763974189758,
"learning_rate": 0.001,
"loss": 3.1859930419921874,
"step": 14850
},
{
"epoch": 0.13019816444617993,
"grad_norm": 0.2078758180141449,
"learning_rate": 0.001,
"loss": 3.19252197265625,
"step": 14900
},
{
"epoch": 0.1306350710382812,
"grad_norm": 0.1912207156419754,
"learning_rate": 0.001,
"loss": 3.1930859375,
"step": 14950
},
{
"epoch": 0.13107197763038247,
"grad_norm": 0.18659965693950653,
"learning_rate": 0.001,
"loss": 3.1894403076171876,
"step": 15000
},
{
"epoch": 0.13107197763038247,
"eval_loss": 3.1984639167785645,
"eval_runtime": 9.5838,
"eval_samples_per_second": 101.838,
"eval_steps_per_second": 6.365,
"step": 15000
},
{
"epoch": 0.13150888422248375,
"grad_norm": 0.18150147795677185,
"learning_rate": 0.001,
"loss": 3.1855987548828124,
"step": 15050
},
{
"epoch": 0.13194579081458505,
"grad_norm": 0.1850251406431198,
"learning_rate": 0.001,
"loss": 3.18405517578125,
"step": 15100
},
{
"epoch": 0.13238269740668632,
"grad_norm": 0.18930986523628235,
"learning_rate": 0.001,
"loss": 3.181584777832031,
"step": 15150
},
{
"epoch": 0.1328196039987876,
"grad_norm": 0.19745807349681854,
"learning_rate": 0.001,
"loss": 3.1843905639648438,
"step": 15200
},
{
"epoch": 0.13325651059088886,
"grad_norm": 0.2121584564447403,
"learning_rate": 0.001,
"loss": 3.194527893066406,
"step": 15250
},
{
"epoch": 0.13369341718299013,
"grad_norm": 0.1978662610054016,
"learning_rate": 0.001,
"loss": 3.185810241699219,
"step": 15300
},
{
"epoch": 0.1341303237750914,
"grad_norm": 0.19299598038196564,
"learning_rate": 0.001,
"loss": 3.178111877441406,
"step": 15350
},
{
"epoch": 0.13456723036719268,
"grad_norm": 0.1843535304069519,
"learning_rate": 0.001,
"loss": 3.185999450683594,
"step": 15400
},
{
"epoch": 0.13500413695929395,
"grad_norm": 0.17675578594207764,
"learning_rate": 0.001,
"loss": 3.1846661376953125,
"step": 15450
},
{
"epoch": 0.13544104355139525,
"grad_norm": 0.19691866636276245,
"learning_rate": 0.001,
"loss": 3.184285888671875,
"step": 15500
},
{
"epoch": 0.13587795014349652,
"grad_norm": 0.23492923378944397,
"learning_rate": 0.001,
"loss": 3.1907785034179685,
"step": 15550
},
{
"epoch": 0.1363148567355978,
"grad_norm": 0.1797175109386444,
"learning_rate": 0.001,
"loss": 3.184339599609375,
"step": 15600
},
{
"epoch": 0.13675176332769906,
"grad_norm": 0.18342596292495728,
"learning_rate": 0.001,
"loss": 3.1823294067382815,
"step": 15650
},
{
"epoch": 0.13718866991980033,
"grad_norm": 0.20209701359272003,
"learning_rate": 0.001,
"loss": 3.187270202636719,
"step": 15700
},
{
"epoch": 0.1376255765119016,
"grad_norm": 0.17159967124462128,
"learning_rate": 0.001,
"loss": 3.187642822265625,
"step": 15750
},
{
"epoch": 0.13806248310400288,
"grad_norm": 0.18459446728229523,
"learning_rate": 0.001,
"loss": 3.196175537109375,
"step": 15800
},
{
"epoch": 0.13849938969610415,
"grad_norm": 0.21580736339092255,
"learning_rate": 0.001,
"loss": 3.179403076171875,
"step": 15850
},
{
"epoch": 0.13893629628820545,
"grad_norm": 0.20760612189769745,
"learning_rate": 0.001,
"loss": 3.1841162109375,
"step": 15900
},
{
"epoch": 0.13937320288030672,
"grad_norm": 0.19794167578220367,
"learning_rate": 0.001,
"loss": 3.1841552734375,
"step": 15950
},
{
"epoch": 0.139810109472408,
"grad_norm": 0.19085007905960083,
"learning_rate": 0.001,
"loss": 3.179669189453125,
"step": 16000
},
{
"epoch": 0.14024701606450926,
"grad_norm": 0.18777291476726532,
"learning_rate": 0.001,
"loss": 3.1862728881835936,
"step": 16050
},
{
"epoch": 0.14068392265661053,
"grad_norm": 0.1883874386548996,
"learning_rate": 0.001,
"loss": 3.1826882934570313,
"step": 16100
},
{
"epoch": 0.1411208292487118,
"grad_norm": 0.1788993775844574,
"learning_rate": 0.001,
"loss": 3.1862246704101564,
"step": 16150
},
{
"epoch": 0.14155773584081308,
"grad_norm": 0.19650617241859436,
"learning_rate": 0.001,
"loss": 3.1779071044921876,
"step": 16200
},
{
"epoch": 0.14199464243291435,
"grad_norm": 0.17883451282978058,
"learning_rate": 0.001,
"loss": 3.1806765747070314,
"step": 16250
},
{
"epoch": 0.14243154902501565,
"grad_norm": 0.22816094756126404,
"learning_rate": 0.001,
"loss": 3.179928894042969,
"step": 16300
},
{
"epoch": 0.14286845561711692,
"grad_norm": 0.18868671357631683,
"learning_rate": 0.001,
"loss": 3.180938720703125,
"step": 16350
},
{
"epoch": 0.1433053622092182,
"grad_norm": 0.18312901258468628,
"learning_rate": 0.001,
"loss": 3.1809161376953123,
"step": 16400
},
{
"epoch": 0.14374226880131946,
"grad_norm": 0.18817898631095886,
"learning_rate": 0.001,
"loss": 3.1803634643554686,
"step": 16450
},
{
"epoch": 0.14417917539342073,
"grad_norm": 0.1878128945827484,
"learning_rate": 0.001,
"loss": 3.184964294433594,
"step": 16500
},
{
"epoch": 0.144616081985522,
"grad_norm": 0.19661137461662292,
"learning_rate": 0.001,
"loss": 3.1799737548828126,
"step": 16550
},
{
"epoch": 0.14505298857762328,
"grad_norm": 0.19748558104038239,
"learning_rate": 0.001,
"loss": 3.174441223144531,
"step": 16600
},
{
"epoch": 0.14548989516972455,
"grad_norm": 0.18286362290382385,
"learning_rate": 0.001,
"loss": 3.1855096435546875,
"step": 16650
},
{
"epoch": 0.14592680176182582,
"grad_norm": 0.20769278705120087,
"learning_rate": 0.001,
"loss": 3.1789508056640625,
"step": 16700
},
{
"epoch": 0.14636370835392712,
"grad_norm": 0.19547678530216217,
"learning_rate": 0.001,
"loss": 3.180931396484375,
"step": 16750
},
{
"epoch": 0.1468006149460284,
"grad_norm": 0.18609896302223206,
"learning_rate": 0.001,
"loss": 3.1838433837890623,
"step": 16800
},
{
"epoch": 0.14723752153812966,
"grad_norm": 0.19330042600631714,
"learning_rate": 0.001,
"loss": 3.1765182495117186,
"step": 16850
},
{
"epoch": 0.14767442813023093,
"grad_norm": 0.18436136841773987,
"learning_rate": 0.001,
"loss": 3.17610595703125,
"step": 16900
},
{
"epoch": 0.1481113347223322,
"grad_norm": 0.18651925027370453,
"learning_rate": 0.001,
"loss": 3.1826718139648436,
"step": 16950
},
{
"epoch": 0.14854824131443348,
"grad_norm": 0.1850690394639969,
"learning_rate": 0.001,
"loss": 3.172263488769531,
"step": 17000
},
{
"epoch": 0.14898514790653475,
"grad_norm": 0.19027315080165863,
"learning_rate": 0.001,
"loss": 3.174286193847656,
"step": 17050
},
{
"epoch": 0.14942205449863602,
"grad_norm": 0.1988609880208969,
"learning_rate": 0.001,
"loss": 3.1703976440429686,
"step": 17100
},
{
"epoch": 0.14985896109073732,
"grad_norm": 0.1764635443687439,
"learning_rate": 0.001,
"loss": 3.1757376098632815,
"step": 17150
},
{
"epoch": 0.1502958676828386,
"grad_norm": 0.17891408503055573,
"learning_rate": 0.001,
"loss": 3.179452209472656,
"step": 17200
},
{
"epoch": 0.15073277427493986,
"grad_norm": 0.18927693367004395,
"learning_rate": 0.001,
"loss": 3.1811477661132814,
"step": 17250
},
{
"epoch": 0.15116968086704113,
"grad_norm": 0.19046278297901154,
"learning_rate": 0.001,
"loss": 3.1781185913085936,
"step": 17300
},
{
"epoch": 0.1516065874591424,
"grad_norm": 0.2253022938966751,
"learning_rate": 0.001,
"loss": 3.1700704956054686,
"step": 17350
},
{
"epoch": 0.15204349405124368,
"grad_norm": 0.18004675209522247,
"learning_rate": 0.001,
"loss": 3.174186096191406,
"step": 17400
},
{
"epoch": 0.15248040064334495,
"grad_norm": 0.18288516998291016,
"learning_rate": 0.001,
"loss": 3.1777203369140623,
"step": 17450
},
{
"epoch": 0.15291730723544622,
"grad_norm": 0.18469488620758057,
"learning_rate": 0.001,
"loss": 3.1725430297851562,
"step": 17500
},
{
"epoch": 0.15335421382754752,
"grad_norm": 0.19650670886039734,
"learning_rate": 0.001,
"loss": 3.180466613769531,
"step": 17550
},
{
"epoch": 0.1537911204196488,
"grad_norm": 0.18393175303936005,
"learning_rate": 0.001,
"loss": 3.1719656372070313,
"step": 17600
},
{
"epoch": 0.15422802701175006,
"grad_norm": 0.22304636240005493,
"learning_rate": 0.001,
"loss": 3.176797180175781,
"step": 17650
},
{
"epoch": 0.15466493360385133,
"grad_norm": 0.19842109084129333,
"learning_rate": 0.001,
"loss": 3.1707284545898435,
"step": 17700
},
{
"epoch": 0.1551018401959526,
"grad_norm": 0.17771974205970764,
"learning_rate": 0.001,
"loss": 3.1734857177734375,
"step": 17750
},
{
"epoch": 0.15553874678805388,
"grad_norm": 0.17204603552818298,
"learning_rate": 0.001,
"loss": 3.170624084472656,
"step": 17800
},
{
"epoch": 0.15597565338015515,
"grad_norm": 0.22022388875484467,
"learning_rate": 0.001,
"loss": 3.180273132324219,
"step": 17850
},
{
"epoch": 0.15641255997225642,
"grad_norm": 0.183755025267601,
"learning_rate": 0.001,
"loss": 3.177940673828125,
"step": 17900
},
{
"epoch": 0.15684946656435772,
"grad_norm": 0.20767413079738617,
"learning_rate": 0.001,
"loss": 3.17085205078125,
"step": 17950
},
{
"epoch": 0.157286373156459,
"grad_norm": 0.1834259182214737,
"learning_rate": 0.001,
"loss": 3.168044128417969,
"step": 18000
},
{
"epoch": 0.157286373156459,
"eval_loss": 3.180673837661743,
"eval_runtime": 6.2659,
"eval_samples_per_second": 155.763,
"eval_steps_per_second": 9.735,
"step": 18000
},
{
"epoch": 0.15772327974856026,
"grad_norm": 0.20340833067893982,
"learning_rate": 0.001,
"loss": 3.1730722045898436,
"step": 18050
},
{
"epoch": 0.15816018634066153,
"grad_norm": 0.19746743142604828,
"learning_rate": 0.001,
"loss": 3.1712353515625,
"step": 18100
},
{
"epoch": 0.1585970929327628,
"grad_norm": 0.2155403047800064,
"learning_rate": 0.001,
"loss": 3.1726614379882814,
"step": 18150
},
{
"epoch": 0.15903399952486408,
"grad_norm": 0.187425434589386,
"learning_rate": 0.001,
"loss": 3.1722906494140624,
"step": 18200
},
{
"epoch": 0.15947090611696535,
"grad_norm": 0.19845345616340637,
"learning_rate": 0.001,
"loss": 3.176360778808594,
"step": 18250
},
{
"epoch": 0.15990781270906662,
"grad_norm": 0.19284312427043915,
"learning_rate": 0.001,
"loss": 3.168287658691406,
"step": 18300
},
{
"epoch": 0.1603447193011679,
"grad_norm": 0.21521936357021332,
"learning_rate": 0.001,
"loss": 3.16669921875,
"step": 18350
},
{
"epoch": 0.1607816258932692,
"grad_norm": 0.18518303334712982,
"learning_rate": 0.001,
"loss": 3.1698013305664063,
"step": 18400
},
{
"epoch": 0.16121853248537046,
"grad_norm": 0.1973997950553894,
"learning_rate": 0.001,
"loss": 3.175386962890625,
"step": 18450
},
{
"epoch": 0.16165543907747174,
"grad_norm": 0.18872328102588654,
"learning_rate": 0.001,
"loss": 3.1612030029296876,
"step": 18500
},
{
"epoch": 0.162092345669573,
"grad_norm": 0.19865576922893524,
"learning_rate": 0.001,
"loss": 3.166875,
"step": 18550
},
{
"epoch": 0.16252925226167428,
"grad_norm": 0.19910448789596558,
"learning_rate": 0.001,
"loss": 3.1679531860351564,
"step": 18600
},
{
"epoch": 0.16296615885377555,
"grad_norm": 0.1925744265317917,
"learning_rate": 0.001,
"loss": 3.1705352783203127,
"step": 18650
},
{
"epoch": 0.16340306544587682,
"grad_norm": 0.19496862590312958,
"learning_rate": 0.001,
"loss": 3.1727005004882813,
"step": 18700
},
{
"epoch": 0.1638399720379781,
"grad_norm": 0.18822863698005676,
"learning_rate": 0.001,
"loss": 3.1648239135742187,
"step": 18750
},
{
"epoch": 0.1642768786300794,
"grad_norm": 0.20176562666893005,
"learning_rate": 0.001,
"loss": 3.1724688720703127,
"step": 18800
},
{
"epoch": 0.16471378522218066,
"grad_norm": 0.19679468870162964,
"learning_rate": 0.001,
"loss": 3.1685910034179687,
"step": 18850
},
{
"epoch": 0.16515069181428194,
"grad_norm": 0.20385122299194336,
"learning_rate": 0.001,
"loss": 3.1655545043945312,
"step": 18900
},
{
"epoch": 0.1655875984063832,
"grad_norm": 0.1873558759689331,
"learning_rate": 0.001,
"loss": 3.1621163940429686,
"step": 18950
},
{
"epoch": 0.16602450499848448,
"grad_norm": 0.1796867698431015,
"learning_rate": 0.001,
"loss": 3.161924133300781,
"step": 19000
},
{
"epoch": 0.16646141159058575,
"grad_norm": 0.17817625403404236,
"learning_rate": 0.001,
"loss": 3.1670086669921873,
"step": 19050
},
{
"epoch": 0.16689831818268702,
"grad_norm": 0.19963178038597107,
"learning_rate": 0.001,
"loss": 3.165671691894531,
"step": 19100
},
{
"epoch": 0.1673352247747883,
"grad_norm": 0.2087634652853012,
"learning_rate": 0.001,
"loss": 3.1683724975585936,
"step": 19150
},
{
"epoch": 0.1677721313668896,
"grad_norm": 0.1969711184501648,
"learning_rate": 0.001,
"loss": 3.168401184082031,
"step": 19200
},
{
"epoch": 0.16820903795899086,
"grad_norm": 0.2119653820991516,
"learning_rate": 0.001,
"loss": 3.1596734619140623,
"step": 19250
},
{
"epoch": 0.16864594455109214,
"grad_norm": 0.1788693368434906,
"learning_rate": 0.001,
"loss": 3.1643597412109377,
"step": 19300
},
{
"epoch": 0.1690828511431934,
"grad_norm": 0.19928136467933655,
"learning_rate": 0.001,
"loss": 3.1645272827148436,
"step": 19350
},
{
"epoch": 0.16951975773529468,
"grad_norm": 0.20922206342220306,
"learning_rate": 0.001,
"loss": 3.1599459838867188,
"step": 19400
},
{
"epoch": 0.16995666432739595,
"grad_norm": 0.1890125721693039,
"learning_rate": 0.001,
"loss": 3.164220886230469,
"step": 19450
},
{
"epoch": 0.17039357091949722,
"grad_norm": 0.1845017820596695,
"learning_rate": 0.001,
"loss": 3.1608465576171874,
"step": 19500
},
{
"epoch": 0.1708304775115985,
"grad_norm": 0.18918895721435547,
"learning_rate": 0.001,
"loss": 3.1621929931640627,
"step": 19550
},
{
"epoch": 0.1712673841036998,
"grad_norm": 0.2020469605922699,
"learning_rate": 0.001,
"loss": 3.16183349609375,
"step": 19600
},
{
"epoch": 0.17170429069580107,
"grad_norm": 0.19557921588420868,
"learning_rate": 0.001,
"loss": 3.16815673828125,
"step": 19650
},
{
"epoch": 0.17214119728790234,
"grad_norm": 0.1984385997056961,
"learning_rate": 0.001,
"loss": 3.1694110107421873,
"step": 19700
},
{
"epoch": 0.1725781038800036,
"grad_norm": 0.18434768915176392,
"learning_rate": 0.001,
"loss": 3.1603521728515624,
"step": 19750
},
{
"epoch": 0.17301501047210488,
"grad_norm": 0.19901612401008606,
"learning_rate": 0.001,
"loss": 3.1662472534179686,
"step": 19800
},
{
"epoch": 0.17345191706420615,
"grad_norm": 0.18059001863002777,
"learning_rate": 0.001,
"loss": 3.162430725097656,
"step": 19850
},
{
"epoch": 0.17388882365630742,
"grad_norm": 0.18952184915542603,
"learning_rate": 0.001,
"loss": 3.162498779296875,
"step": 19900
},
{
"epoch": 0.1743257302484087,
"grad_norm": 0.1816520243883133,
"learning_rate": 0.001,
"loss": 3.164564208984375,
"step": 19950
},
{
"epoch": 0.17476263684050997,
"grad_norm": 0.22983364760875702,
"learning_rate": 0.001,
"loss": 3.1721530151367188,
"step": 20000
},
{
"epoch": 0.17519954343261127,
"grad_norm": 0.1910918802022934,
"learning_rate": 0.001,
"loss": 3.1639764404296873,
"step": 20050
},
{
"epoch": 0.17563645002471254,
"grad_norm": 0.20957250893115997,
"learning_rate": 0.001,
"loss": 3.1658724975585937,
"step": 20100
},
{
"epoch": 0.1760733566168138,
"grad_norm": 0.1926119029521942,
"learning_rate": 0.001,
"loss": 3.1673150634765626,
"step": 20150
},
{
"epoch": 0.17651026320891508,
"grad_norm": 0.19113746285438538,
"learning_rate": 0.001,
"loss": 3.161557922363281,
"step": 20200
},
{
"epoch": 0.17694716980101635,
"grad_norm": 0.1946914941072464,
"learning_rate": 0.001,
"loss": 3.159283447265625,
"step": 20250
},
{
"epoch": 0.17738407639311762,
"grad_norm": 0.19165056943893433,
"learning_rate": 0.001,
"loss": 3.1629107666015623,
"step": 20300
},
{
"epoch": 0.1778209829852189,
"grad_norm": 0.19362075626850128,
"learning_rate": 0.001,
"loss": 3.1580712890625,
"step": 20350
},
{
"epoch": 0.17825788957732017,
"grad_norm": 0.19913926720619202,
"learning_rate": 0.001,
"loss": 3.1604278564453123,
"step": 20400
},
{
"epoch": 0.17869479616942147,
"grad_norm": 0.19049301743507385,
"learning_rate": 0.001,
"loss": 3.1671746826171874,
"step": 20450
},
{
"epoch": 0.17913170276152274,
"grad_norm": 0.19689318537712097,
"learning_rate": 0.001,
"loss": 3.1572509765625,
"step": 20500
},
{
"epoch": 0.179568609353624,
"grad_norm": 0.20480966567993164,
"learning_rate": 0.001,
"loss": 3.1548611450195314,
"step": 20550
},
{
"epoch": 0.18000551594572528,
"grad_norm": 0.18391041457653046,
"learning_rate": 0.001,
"loss": 3.157296447753906,
"step": 20600
},
{
"epoch": 0.18044242253782655,
"grad_norm": 0.2023339867591858,
"learning_rate": 0.001,
"loss": 3.155291748046875,
"step": 20650
},
{
"epoch": 0.18087932912992782,
"grad_norm": 0.20646877586841583,
"learning_rate": 0.001,
"loss": 3.162528381347656,
"step": 20700
},
{
"epoch": 0.1813162357220291,
"grad_norm": 0.20429447293281555,
"learning_rate": 0.001,
"loss": 3.1612124633789063,
"step": 20750
},
{
"epoch": 0.18175314231413037,
"grad_norm": 0.18544623255729675,
"learning_rate": 0.001,
"loss": 3.1625381469726563,
"step": 20800
},
{
"epoch": 0.18219004890623167,
"grad_norm": 0.19926291704177856,
"learning_rate": 0.001,
"loss": 3.158988037109375,
"step": 20850
},
{
"epoch": 0.18262695549833294,
"grad_norm": 0.20882956683635712,
"learning_rate": 0.001,
"loss": 3.16245361328125,
"step": 20900
},
{
"epoch": 0.1830638620904342,
"grad_norm": 0.17148427665233612,
"learning_rate": 0.001,
"loss": 3.158699645996094,
"step": 20950
},
{
"epoch": 0.18350076868253548,
"grad_norm": 0.18300697207450867,
"learning_rate": 0.001,
"loss": 3.1594781494140625,
"step": 21000
},
{
"epoch": 0.18350076868253548,
"eval_loss": 3.168538808822632,
"eval_runtime": 6.2184,
"eval_samples_per_second": 156.954,
"eval_steps_per_second": 9.81,
"step": 21000
},
{
"epoch": 0.18393767527463675,
"grad_norm": 0.18563342094421387,
"learning_rate": 0.001,
"loss": 3.1598519897460937,
"step": 21050
},
{
"epoch": 0.18437458186673802,
"grad_norm": 0.18384075164794922,
"learning_rate": 0.001,
"loss": 3.1611053466796877,
"step": 21100
},
{
"epoch": 0.1848114884588393,
"grad_norm": 0.20437927544116974,
"learning_rate": 0.001,
"loss": 3.1592279052734376,
"step": 21150
},
{
"epoch": 0.18524839505094057,
"grad_norm": 0.19333499670028687,
"learning_rate": 0.001,
"loss": 3.1564932250976563,
"step": 21200
},
{
"epoch": 0.18568530164304184,
"grad_norm": 0.18931256234645844,
"learning_rate": 0.001,
"loss": 3.1634091186523436,
"step": 21250
},
{
"epoch": 0.18612220823514314,
"grad_norm": 0.1811954826116562,
"learning_rate": 0.001,
"loss": 3.157086181640625,
"step": 21300
},
{
"epoch": 0.1865591148272444,
"grad_norm": 0.20481161773204803,
"learning_rate": 0.001,
"loss": 3.157492370605469,
"step": 21350
},
{
"epoch": 0.18699602141934568,
"grad_norm": 0.19608381390571594,
"learning_rate": 0.001,
"loss": 3.1567718505859377,
"step": 21400
},
{
"epoch": 0.18743292801144695,
"grad_norm": 0.19365058839321136,
"learning_rate": 0.001,
"loss": 3.1550326538085938,
"step": 21450
},
{
"epoch": 0.18786983460354822,
"grad_norm": 0.1760626882314682,
"learning_rate": 0.001,
"loss": 3.156153259277344,
"step": 21500
},
{
"epoch": 0.1883067411956495,
"grad_norm": 0.20002944767475128,
"learning_rate": 0.001,
"loss": 3.156208801269531,
"step": 21550
},
{
"epoch": 0.18874364778775077,
"grad_norm": 0.21102434396743774,
"learning_rate": 0.001,
"loss": 3.1604791259765626,
"step": 21600
},
{
"epoch": 0.18918055437985204,
"grad_norm": 0.20257405936717987,
"learning_rate": 0.001,
"loss": 3.156624450683594,
"step": 21650
},
{
"epoch": 0.18961746097195334,
"grad_norm": 0.1889481395483017,
"learning_rate": 0.001,
"loss": 3.1555862426757812,
"step": 21700
},
{
"epoch": 0.1900543675640546,
"grad_norm": 0.1976545751094818,
"learning_rate": 0.001,
"loss": 3.164063720703125,
"step": 21750
},
{
"epoch": 0.19049127415615588,
"grad_norm": 0.18660299479961395,
"learning_rate": 0.001,
"loss": 3.1602642822265623,
"step": 21800
},
{
"epoch": 0.19092818074825715,
"grad_norm": 0.1809668391942978,
"learning_rate": 0.001,
"loss": 3.1550946044921875,
"step": 21850
},
{
"epoch": 0.19136508734035843,
"grad_norm": 0.20604996383190155,
"learning_rate": 0.001,
"loss": 3.159581298828125,
"step": 21900
},
{
"epoch": 0.1918019939324597,
"grad_norm": 0.18194423615932465,
"learning_rate": 0.001,
"loss": 3.1616864013671875,
"step": 21950
},
{
"epoch": 0.19223890052456097,
"grad_norm": 0.20112648606300354,
"learning_rate": 0.001,
"loss": 3.1593780517578125,
"step": 22000
},
{
"epoch": 0.19267580711666224,
"grad_norm": 0.18732091784477234,
"learning_rate": 0.001,
"loss": 3.15038818359375,
"step": 22050
},
{
"epoch": 0.19311271370876354,
"grad_norm": 0.19900913536548615,
"learning_rate": 0.001,
"loss": 3.157307434082031,
"step": 22100
},
{
"epoch": 0.1935496203008648,
"grad_norm": 0.19695959985256195,
"learning_rate": 0.001,
"loss": 3.1564794921875,
"step": 22150
},
{
"epoch": 0.19398652689296608,
"grad_norm": 0.18376977741718292,
"learning_rate": 0.001,
"loss": 3.1521331787109377,
"step": 22200
},
{
"epoch": 0.19442343348506735,
"grad_norm": 0.20794112980365753,
"learning_rate": 0.001,
"loss": 3.153861389160156,
"step": 22250
},
{
"epoch": 0.19486034007716863,
"grad_norm": 0.22864745557308197,
"learning_rate": 0.001,
"loss": 3.156719055175781,
"step": 22300
},
{
"epoch": 0.1952972466692699,
"grad_norm": 0.2035977840423584,
"learning_rate": 0.001,
"loss": 3.155831604003906,
"step": 22350
},
{
"epoch": 0.19573415326137117,
"grad_norm": 0.1758171021938324,
"learning_rate": 0.001,
"loss": 3.1510830688476563,
"step": 22400
},
{
"epoch": 0.19617105985347244,
"grad_norm": 0.19761677086353302,
"learning_rate": 0.001,
"loss": 3.1547085571289064,
"step": 22450
},
{
"epoch": 0.19660796644557374,
"grad_norm": 0.19923441112041473,
"learning_rate": 0.001,
"loss": 3.1492523193359374,
"step": 22500
},
{
"epoch": 0.197044873037675,
"grad_norm": 0.18077634274959564,
"learning_rate": 0.001,
"loss": 3.1544876098632812,
"step": 22550
},
{
"epoch": 0.19748177962977628,
"grad_norm": 0.18283407390117645,
"learning_rate": 0.001,
"loss": 3.1544061279296876,
"step": 22600
},
{
"epoch": 0.19791868622187755,
"grad_norm": 0.186768040060997,
"learning_rate": 0.001,
"loss": 3.1512969970703124,
"step": 22650
},
{
"epoch": 0.19835559281397883,
"grad_norm": 0.21104729175567627,
"learning_rate": 0.001,
"loss": 3.1526315307617185,
"step": 22700
},
{
"epoch": 0.1987924994060801,
"grad_norm": 0.20985271036624908,
"learning_rate": 0.001,
"loss": 3.1552078247070314,
"step": 22750
},
{
"epoch": 0.19922940599818137,
"grad_norm": 0.19860677421092987,
"learning_rate": 0.001,
"loss": 3.1497076416015624,
"step": 22800
},
{
"epoch": 0.19966631259028264,
"grad_norm": 0.19701924920082092,
"learning_rate": 0.001,
"loss": 3.151370849609375,
"step": 22850
},
{
"epoch": 0.2001032191823839,
"grad_norm": 0.18555696308612823,
"learning_rate": 0.001,
"loss": 3.1431695556640626,
"step": 22900
},
{
"epoch": 0.2005401257744852,
"grad_norm": 0.17813950777053833,
"learning_rate": 0.001,
"loss": 3.1471441650390624,
"step": 22950
},
{
"epoch": 0.20097703236658648,
"grad_norm": 0.19566522538661957,
"learning_rate": 0.001,
"loss": 3.1563845825195314,
"step": 23000
},
{
"epoch": 0.20141393895868775,
"grad_norm": 0.21006260812282562,
"learning_rate": 0.001,
"loss": 3.151533508300781,
"step": 23050
},
{
"epoch": 0.20185084555078903,
"grad_norm": 0.1936446726322174,
"learning_rate": 0.001,
"loss": 3.15564208984375,
"step": 23100
},
{
"epoch": 0.2022877521428903,
"grad_norm": 0.21060171723365784,
"learning_rate": 0.001,
"loss": 3.15681884765625,
"step": 23150
},
{
"epoch": 0.20272465873499157,
"grad_norm": 0.1832597851753235,
"learning_rate": 0.001,
"loss": 3.1566827392578123,
"step": 23200
},
{
"epoch": 0.20316156532709284,
"grad_norm": 0.21573704481124878,
"learning_rate": 0.001,
"loss": 3.1494873046875,
"step": 23250
},
{
"epoch": 0.2035984719191941,
"grad_norm": 0.20478427410125732,
"learning_rate": 0.001,
"loss": 3.150159912109375,
"step": 23300
},
{
"epoch": 0.2040353785112954,
"grad_norm": 0.19913440942764282,
"learning_rate": 0.001,
"loss": 3.1586724853515626,
"step": 23350
},
{
"epoch": 0.20447228510339668,
"grad_norm": 0.17924748361110687,
"learning_rate": 0.001,
"loss": 3.1538009643554688,
"step": 23400
},
{
"epoch": 0.20490919169549796,
"grad_norm": 0.20912882685661316,
"learning_rate": 0.001,
"loss": 3.15191650390625,
"step": 23450
},
{
"epoch": 0.20534609828759923,
"grad_norm": 0.20407678186893463,
"learning_rate": 0.001,
"loss": 3.1518075561523435,
"step": 23500
},
{
"epoch": 0.2057830048797005,
"grad_norm": 0.20418840646743774,
"learning_rate": 0.001,
"loss": 3.14880859375,
"step": 23550
},
{
"epoch": 0.20621991147180177,
"grad_norm": 0.20117631554603577,
"learning_rate": 0.001,
"loss": 3.14843994140625,
"step": 23600
},
{
"epoch": 0.20665681806390304,
"grad_norm": 0.19418101012706757,
"learning_rate": 0.001,
"loss": 3.1554824829101564,
"step": 23650
},
{
"epoch": 0.2070937246560043,
"grad_norm": 0.2021266669034958,
"learning_rate": 0.001,
"loss": 3.154755859375,
"step": 23700
},
{
"epoch": 0.2075306312481056,
"grad_norm": 0.19447430968284607,
"learning_rate": 0.001,
"loss": 3.1476751708984376,
"step": 23750
},
{
"epoch": 0.20796753784020688,
"grad_norm": 0.18747089803218842,
"learning_rate": 0.001,
"loss": 3.1445318603515626,
"step": 23800
},
{
"epoch": 0.20840444443230816,
"grad_norm": 0.20383097231388092,
"learning_rate": 0.001,
"loss": 3.1506866455078124,
"step": 23850
},
{
"epoch": 0.20884135102440943,
"grad_norm": 0.200015589594841,
"learning_rate": 0.001,
"loss": 3.1511416625976563,
"step": 23900
},
{
"epoch": 0.2092782576165107,
"grad_norm": 0.2223626971244812,
"learning_rate": 0.001,
"loss": 3.1494268798828124,
"step": 23950
},
{
"epoch": 0.20971516420861197,
"grad_norm": 0.19981783628463745,
"learning_rate": 0.001,
"loss": 3.1447100830078125,
"step": 24000
},
{
"epoch": 0.20971516420861197,
"eval_loss": 3.158378839492798,
"eval_runtime": 6.4304,
"eval_samples_per_second": 151.779,
"eval_steps_per_second": 9.486,
"step": 24000
},
{
"epoch": 0.21015207080071324,
"grad_norm": 0.19517436623573303,
"learning_rate": 0.001,
"loss": 3.15059814453125,
"step": 24050
},
{
"epoch": 0.2105889773928145,
"grad_norm": 0.19617164134979248,
"learning_rate": 0.001,
"loss": 3.145147399902344,
"step": 24100
},
{
"epoch": 0.2110258839849158,
"grad_norm": 0.2006959617137909,
"learning_rate": 0.001,
"loss": 3.1506362915039063,
"step": 24150
},
{
"epoch": 0.21146279057701708,
"grad_norm": 0.19889812171459198,
"learning_rate": 0.001,
"loss": 3.146689758300781,
"step": 24200
},
{
"epoch": 0.21189969716911836,
"grad_norm": 0.19370006024837494,
"learning_rate": 0.001,
"loss": 3.1550631713867188,
"step": 24250
},
{
"epoch": 0.21233660376121963,
"grad_norm": 0.22065851092338562,
"learning_rate": 0.001,
"loss": 3.14722900390625,
"step": 24300
},
{
"epoch": 0.2127735103533209,
"grad_norm": 0.21452419459819794,
"learning_rate": 0.001,
"loss": 3.1473452758789064,
"step": 24350
},
{
"epoch": 0.21321041694542217,
"grad_norm": 0.1960809975862503,
"learning_rate": 0.001,
"loss": 3.1492404174804687,
"step": 24400
},
{
"epoch": 0.21364732353752344,
"grad_norm": 0.2180708944797516,
"learning_rate": 0.001,
"loss": 3.148568115234375,
"step": 24450
},
{
"epoch": 0.21408423012962471,
"grad_norm": 0.20061273872852325,
"learning_rate": 0.001,
"loss": 3.1441583251953125,
"step": 24500
},
{
"epoch": 0.21452113672172599,
"grad_norm": 0.20494696497917175,
"learning_rate": 0.001,
"loss": 3.1491708374023437,
"step": 24550
},
{
"epoch": 0.21495804331382729,
"grad_norm": 0.18533244729042053,
"learning_rate": 0.001,
"loss": 3.1487017822265626,
"step": 24600
},
{
"epoch": 0.21539494990592856,
"grad_norm": 0.20078839361667633,
"learning_rate": 0.001,
"loss": 3.143594055175781,
"step": 24650
},
{
"epoch": 0.21583185649802983,
"grad_norm": 0.2131863683462143,
"learning_rate": 0.001,
"loss": 3.1505694580078125,
"step": 24700
},
{
"epoch": 0.2162687630901311,
"grad_norm": 0.2060062438249588,
"learning_rate": 0.001,
"loss": 3.1452517700195313,
"step": 24750
},
{
"epoch": 0.21670566968223237,
"grad_norm": 0.17908109724521637,
"learning_rate": 0.001,
"loss": 3.145859375,
"step": 24800
},
{
"epoch": 0.21714257627433364,
"grad_norm": 0.18847830593585968,
"learning_rate": 0.001,
"loss": 3.1476791381835936,
"step": 24850
},
{
"epoch": 0.21757948286643491,
"grad_norm": 0.22164477407932281,
"learning_rate": 0.001,
"loss": 3.14805908203125,
"step": 24900
},
{
"epoch": 0.21801638945853619,
"grad_norm": 0.2043878138065338,
"learning_rate": 0.001,
"loss": 3.1431939697265623,
"step": 24950
},
{
"epoch": 0.21845329605063749,
"grad_norm": 0.1971021145582199,
"learning_rate": 0.001,
"loss": 3.142523193359375,
"step": 25000
},
{
"epoch": 0.21889020264273876,
"grad_norm": 0.17631393671035767,
"learning_rate": 0.001,
"loss": 3.1423294067382814,
"step": 25050
},
{
"epoch": 0.21932710923484003,
"grad_norm": 0.21117663383483887,
"learning_rate": 0.001,
"loss": 3.1507272338867187,
"step": 25100
},
{
"epoch": 0.2197640158269413,
"grad_norm": 0.1925990879535675,
"learning_rate": 0.001,
"loss": 3.150421142578125,
"step": 25150
},
{
"epoch": 0.22020092241904257,
"grad_norm": 0.2095244973897934,
"learning_rate": 0.001,
"loss": 3.1468170166015623,
"step": 25200
},
{
"epoch": 0.22063782901114384,
"grad_norm": 0.17956186830997467,
"learning_rate": 0.001,
"loss": 3.141022033691406,
"step": 25250
},
{
"epoch": 0.22107473560324511,
"grad_norm": 0.18447719514369965,
"learning_rate": 0.001,
"loss": 3.148127136230469,
"step": 25300
},
{
"epoch": 0.2215116421953464,
"grad_norm": 0.1958089917898178,
"learning_rate": 0.001,
"loss": 3.147033996582031,
"step": 25350
},
{
"epoch": 0.22194854878744769,
"grad_norm": 0.20372235774993896,
"learning_rate": 0.001,
"loss": 3.1438693237304687,
"step": 25400
},
{
"epoch": 0.22238545537954896,
"grad_norm": 0.18154527246952057,
"learning_rate": 0.001,
"loss": 3.146240234375,
"step": 25450
},
{
"epoch": 0.22282236197165023,
"grad_norm": 0.1875283271074295,
"learning_rate": 0.001,
"loss": 3.1459466552734376,
"step": 25500
},
{
"epoch": 0.2232592685637515,
"grad_norm": 0.18850144743919373,
"learning_rate": 0.001,
"loss": 3.144921875,
"step": 25550
},
{
"epoch": 0.22369617515585277,
"grad_norm": 0.20837713778018951,
"learning_rate": 0.001,
"loss": 3.144251708984375,
"step": 25600
},
{
"epoch": 0.22413308174795404,
"grad_norm": 0.19441227614879608,
"learning_rate": 0.001,
"loss": 3.1461331176757814,
"step": 25650
},
{
"epoch": 0.22456998834005532,
"grad_norm": 0.20006810128688812,
"learning_rate": 0.001,
"loss": 3.1491522216796874,
"step": 25700
},
{
"epoch": 0.2250068949321566,
"grad_norm": 0.1997118890285492,
"learning_rate": 0.001,
"loss": 3.1450509643554687,
"step": 25750
},
{
"epoch": 0.2254438015242579,
"grad_norm": 0.1937110722064972,
"learning_rate": 0.001,
"loss": 3.1459671020507813,
"step": 25800
},
{
"epoch": 0.22588070811635916,
"grad_norm": 0.20349667966365814,
"learning_rate": 0.001,
"loss": 3.1398193359375,
"step": 25850
},
{
"epoch": 0.22631761470846043,
"grad_norm": 0.21692262589931488,
"learning_rate": 0.001,
"loss": 3.146085510253906,
"step": 25900
},
{
"epoch": 0.2267545213005617,
"grad_norm": 0.2061356157064438,
"learning_rate": 0.001,
"loss": 3.14493408203125,
"step": 25950
},
{
"epoch": 0.22719142789266297,
"grad_norm": 0.18468759953975677,
"learning_rate": 0.001,
"loss": 3.141280517578125,
"step": 26000
},
{
"epoch": 0.22762833448476424,
"grad_norm": 0.19995801150798798,
"learning_rate": 0.001,
"loss": 3.1449270629882813,
"step": 26050
},
{
"epoch": 0.22806524107686552,
"grad_norm": 0.1986706703901291,
"learning_rate": 0.001,
"loss": 3.1400848388671876,
"step": 26100
},
{
"epoch": 0.2285021476689668,
"grad_norm": 0.1953519582748413,
"learning_rate": 0.001,
"loss": 3.140890808105469,
"step": 26150
},
{
"epoch": 0.22893905426106806,
"grad_norm": 0.19395847618579865,
"learning_rate": 0.001,
"loss": 3.140115966796875,
"step": 26200
},
{
"epoch": 0.22937596085316936,
"grad_norm": 0.1840379536151886,
"learning_rate": 0.001,
"loss": 3.142007141113281,
"step": 26250
},
{
"epoch": 0.22981286744527063,
"grad_norm": 0.18377895653247833,
"learning_rate": 0.001,
"loss": 3.1446292114257814,
"step": 26300
},
{
"epoch": 0.2302497740373719,
"grad_norm": 0.2181359976530075,
"learning_rate": 0.001,
"loss": 3.1472604370117185,
"step": 26350
},
{
"epoch": 0.23068668062947317,
"grad_norm": 0.1907571703195572,
"learning_rate": 0.001,
"loss": 3.1401361083984374,
"step": 26400
},
{
"epoch": 0.23112358722157444,
"grad_norm": 0.20355018973350525,
"learning_rate": 0.001,
"loss": 3.1461245727539064,
"step": 26450
},
{
"epoch": 0.23156049381367572,
"grad_norm": 0.18987947702407837,
"learning_rate": 0.001,
"loss": 3.133076477050781,
"step": 26500
},
{
"epoch": 0.231997400405777,
"grad_norm": 0.20657390356063843,
"learning_rate": 0.001,
"loss": 3.1453286743164064,
"step": 26550
},
{
"epoch": 0.23243430699787826,
"grad_norm": 0.20074376463890076,
"learning_rate": 0.001,
"loss": 3.1440554809570314,
"step": 26600
},
{
"epoch": 0.23287121358997956,
"grad_norm": 0.21610620617866516,
"learning_rate": 0.001,
"loss": 3.1415972900390625,
"step": 26650
},
{
"epoch": 0.23330812018208083,
"grad_norm": 0.17672847211360931,
"learning_rate": 0.001,
"loss": 3.143729553222656,
"step": 26700
},
{
"epoch": 0.2337450267741821,
"grad_norm": 0.2039860486984253,
"learning_rate": 0.001,
"loss": 3.139731750488281,
"step": 26750
},
{
"epoch": 0.23418193336628337,
"grad_norm": 0.1961374282836914,
"learning_rate": 0.001,
"loss": 3.1416336059570313,
"step": 26800
},
{
"epoch": 0.23461883995838465,
"grad_norm": 0.2058544158935547,
"learning_rate": 0.001,
"loss": 3.134196472167969,
"step": 26850
},
{
"epoch": 0.23505574655048592,
"grad_norm": 0.18912115693092346,
"learning_rate": 0.001,
"loss": 3.140076904296875,
"step": 26900
},
{
"epoch": 0.2354926531425872,
"grad_norm": 0.19993124902248383,
"learning_rate": 0.001,
"loss": 3.1428677368164064,
"step": 26950
},
{
"epoch": 0.23592955973468846,
"grad_norm": 0.2177840620279312,
"learning_rate": 0.001,
"loss": 3.139360046386719,
"step": 27000
},
{
"epoch": 0.23592955973468846,
"eval_loss": 3.151303768157959,
"eval_runtime": 6.3933,
"eval_samples_per_second": 152.661,
"eval_steps_per_second": 9.541,
"step": 27000
},
{
"epoch": 0.23636646632678976,
"grad_norm": 0.20935207605361938,
"learning_rate": 0.001,
"loss": 3.1434817504882813,
"step": 27050
},
{
"epoch": 0.23680337291889103,
"grad_norm": 0.21745233237743378,
"learning_rate": 0.001,
"loss": 3.138890686035156,
"step": 27100
},
{
"epoch": 0.2372402795109923,
"grad_norm": 0.1814340204000473,
"learning_rate": 0.001,
"loss": 3.1400631713867186,
"step": 27150
},
{
"epoch": 0.23767718610309357,
"grad_norm": 0.21654419600963593,
"learning_rate": 0.001,
"loss": 3.1386923217773437,
"step": 27200
},
{
"epoch": 0.23811409269519485,
"grad_norm": 0.20901715755462646,
"learning_rate": 0.001,
"loss": 3.143807678222656,
"step": 27250
},
{
"epoch": 0.23855099928729612,
"grad_norm": 0.18465693295001984,
"learning_rate": 0.001,
"loss": 3.1438729858398435,
"step": 27300
},
{
"epoch": 0.2389879058793974,
"grad_norm": 0.20159631967544556,
"learning_rate": 0.001,
"loss": 3.1333203125,
"step": 27350
},
{
"epoch": 0.23942481247149866,
"grad_norm": 0.1925830841064453,
"learning_rate": 0.001,
"loss": 3.1410565185546875,
"step": 27400
},
{
"epoch": 0.23986171906359996,
"grad_norm": 0.2061873972415924,
"learning_rate": 0.001,
"loss": 3.1350579833984376,
"step": 27450
},
{
"epoch": 0.24029862565570123,
"grad_norm": 0.2098609209060669,
"learning_rate": 0.001,
"loss": 3.1358438110351563,
"step": 27500
},
{
"epoch": 0.2407355322478025,
"grad_norm": 0.19218602776527405,
"learning_rate": 0.001,
"loss": 3.1425982666015626,
"step": 27550
},
{
"epoch": 0.24117243883990377,
"grad_norm": 0.20362652838230133,
"learning_rate": 0.001,
"loss": 3.137518310546875,
"step": 27600
},
{
"epoch": 0.24160934543200505,
"grad_norm": 0.1742187738418579,
"learning_rate": 0.001,
"loss": 3.1371829223632814,
"step": 27650
},
{
"epoch": 0.24204625202410632,
"grad_norm": 0.208481103181839,
"learning_rate": 0.001,
"loss": 3.1426657104492186,
"step": 27700
},
{
"epoch": 0.2424831586162076,
"grad_norm": 0.21637871861457825,
"learning_rate": 0.001,
"loss": 3.1345880126953123,
"step": 27750
},
{
"epoch": 0.24292006520830886,
"grad_norm": 0.2094481885433197,
"learning_rate": 0.001,
"loss": 3.1394207763671873,
"step": 27800
},
{
"epoch": 0.24335697180041013,
"grad_norm": 0.18196998536586761,
"learning_rate": 0.001,
"loss": 3.1421343994140627,
"step": 27850
},
{
"epoch": 0.24379387839251143,
"grad_norm": 0.19949769973754883,
"learning_rate": 0.001,
"loss": 3.129438781738281,
"step": 27900
},
{
"epoch": 0.2442307849846127,
"grad_norm": 0.19427987933158875,
"learning_rate": 0.001,
"loss": 3.146867370605469,
"step": 27950
},
{
"epoch": 0.24466769157671397,
"grad_norm": 0.19375596940517426,
"learning_rate": 0.001,
"loss": 3.1379046630859375,
"step": 28000
},
{
"epoch": 0.24510459816881525,
"grad_norm": 0.19712986052036285,
"learning_rate": 0.001,
"loss": 3.138131408691406,
"step": 28050
},
{
"epoch": 0.24554150476091652,
"grad_norm": 0.20073772966861725,
"learning_rate": 0.001,
"loss": 3.1416021728515626,
"step": 28100
},
{
"epoch": 0.2459784113530178,
"grad_norm": 0.18688786029815674,
"learning_rate": 0.001,
"loss": 3.1308734130859377,
"step": 28150
},
{
"epoch": 0.24641531794511906,
"grad_norm": 0.20580218732357025,
"learning_rate": 0.001,
"loss": 3.142344970703125,
"step": 28200
},
{
"epoch": 0.24685222453722033,
"grad_norm": 0.1983879953622818,
"learning_rate": 0.001,
"loss": 3.146060791015625,
"step": 28250
},
{
"epoch": 0.24728913112932163,
"grad_norm": 0.19333505630493164,
"learning_rate": 0.001,
"loss": 3.1410641479492187,
"step": 28300
},
{
"epoch": 0.2477260377214229,
"grad_norm": 0.1915540099143982,
"learning_rate": 0.001,
"loss": 3.1385321044921874,
"step": 28350
},
{
"epoch": 0.24816294431352418,
"grad_norm": 0.20334628224372864,
"learning_rate": 0.001,
"loss": 3.1348602294921877,
"step": 28400
},
{
"epoch": 0.24859985090562545,
"grad_norm": 0.20253990590572357,
"learning_rate": 0.001,
"loss": 3.1378768920898437,
"step": 28450
},
{
"epoch": 0.24903675749772672,
"grad_norm": 0.20241330564022064,
"learning_rate": 0.001,
"loss": 3.131033630371094,
"step": 28500
},
{
"epoch": 0.249473664089828,
"grad_norm": 0.19613344967365265,
"learning_rate": 0.001,
"loss": 3.1422305297851563,
"step": 28550
},
{
"epoch": 0.24991057068192926,
"grad_norm": 0.20870091021060944,
"learning_rate": 0.001,
"loss": 3.1335931396484376,
"step": 28600
},
{
"epoch": 0.25034747727403056,
"grad_norm": 0.1986633539199829,
"learning_rate": 0.001,
"loss": 3.1338870239257814,
"step": 28650
},
{
"epoch": 0.25078438386613183,
"grad_norm": 0.1975926011800766,
"learning_rate": 0.001,
"loss": 3.1350277709960936,
"step": 28700
},
{
"epoch": 0.2512212904582331,
"grad_norm": 0.20917613804340363,
"learning_rate": 0.001,
"loss": 3.1365911865234377,
"step": 28750
},
{
"epoch": 0.2516581970503344,
"grad_norm": 0.17652350664138794,
"learning_rate": 0.001,
"loss": 3.1322979736328125,
"step": 28800
},
{
"epoch": 0.25209510364243565,
"grad_norm": 0.18789704144001007,
"learning_rate": 0.001,
"loss": 3.135655822753906,
"step": 28850
},
{
"epoch": 0.2525320102345369,
"grad_norm": 0.2171935886144638,
"learning_rate": 0.001,
"loss": 3.1341122436523436,
"step": 28900
},
{
"epoch": 0.2529689168266382,
"grad_norm": 0.2077326476573944,
"learning_rate": 0.001,
"loss": 3.13816162109375,
"step": 28950
},
{
"epoch": 0.25340582341873946,
"grad_norm": 0.20034188032150269,
"learning_rate": 0.001,
"loss": 3.13508544921875,
"step": 29000
},
{
"epoch": 0.25384273001084073,
"grad_norm": 0.18811163306236267,
"learning_rate": 0.001,
"loss": 3.134895935058594,
"step": 29050
},
{
"epoch": 0.254279636602942,
"grad_norm": 0.19685913622379303,
"learning_rate": 0.001,
"loss": 3.1354241943359376,
"step": 29100
},
{
"epoch": 0.2547165431950433,
"grad_norm": 0.2031325399875641,
"learning_rate": 0.001,
"loss": 3.1332156372070314,
"step": 29150
},
{
"epoch": 0.25515344978714455,
"grad_norm": 0.20226146280765533,
"learning_rate": 0.001,
"loss": 3.1414865112304686,
"step": 29200
},
{
"epoch": 0.2555903563792458,
"grad_norm": 0.23303231596946716,
"learning_rate": 0.001,
"loss": 3.1362576293945312,
"step": 29250
},
{
"epoch": 0.25602726297134715,
"grad_norm": 0.18115435540676117,
"learning_rate": 0.001,
"loss": 3.1362594604492187,
"step": 29300
},
{
"epoch": 0.2564641695634484,
"grad_norm": 0.1966370791196823,
"learning_rate": 0.001,
"loss": 3.135086975097656,
"step": 29350
},
{
"epoch": 0.2569010761555497,
"grad_norm": 0.19436295330524445,
"learning_rate": 0.001,
"loss": 3.1318792724609374,
"step": 29400
},
{
"epoch": 0.25733798274765096,
"grad_norm": 0.21984174847602844,
"learning_rate": 0.001,
"loss": 3.138869323730469,
"step": 29450
},
{
"epoch": 0.25777488933975223,
"grad_norm": 0.2019258290529251,
"learning_rate": 0.001,
"loss": 3.1350537109375,
"step": 29500
},
{
"epoch": 0.2582117959318535,
"grad_norm": 0.21706347167491913,
"learning_rate": 0.001,
"loss": 3.1332632446289064,
"step": 29550
},
{
"epoch": 0.2586487025239548,
"grad_norm": 0.18457020819187164,
"learning_rate": 0.001,
"loss": 3.1358929443359376,
"step": 29600
},
{
"epoch": 0.25908560911605605,
"grad_norm": 0.20873647928237915,
"learning_rate": 0.001,
"loss": 3.134085998535156,
"step": 29650
},
{
"epoch": 0.2595225157081573,
"grad_norm": 0.19496671855449677,
"learning_rate": 0.001,
"loss": 3.128114013671875,
"step": 29700
},
{
"epoch": 0.2599594223002586,
"grad_norm": 0.19370408356189728,
"learning_rate": 0.001,
"loss": 3.1351947021484374,
"step": 29750
},
{
"epoch": 0.26039632889235986,
"grad_norm": 0.1864786595106125,
"learning_rate": 0.001,
"loss": 3.1307614135742186,
"step": 29800
},
{
"epoch": 0.26083323548446113,
"grad_norm": 0.18954840302467346,
"learning_rate": 0.001,
"loss": 3.128426208496094,
"step": 29850
},
{
"epoch": 0.2612701420765624,
"grad_norm": 0.20375385880470276,
"learning_rate": 0.001,
"loss": 3.128997802734375,
"step": 29900
},
{
"epoch": 0.2617070486686637,
"grad_norm": 0.22339008748531342,
"learning_rate": 0.001,
"loss": 3.1360702514648438,
"step": 29950
},
{
"epoch": 0.26214395526076495,
"grad_norm": 0.18949832022190094,
"learning_rate": 0.001,
"loss": 3.1370416259765626,
"step": 30000
},
{
"epoch": 0.26214395526076495,
"eval_loss": 3.144541025161743,
"eval_runtime": 7.9931,
"eval_samples_per_second": 122.106,
"eval_steps_per_second": 7.632,
"step": 30000
},
{
"epoch": 0.2625808618528662,
"grad_norm": 0.20105642080307007,
"learning_rate": 0.001,
"loss": 3.139212646484375,
"step": 30050
},
{
"epoch": 0.2630177684449675,
"grad_norm": 0.20688988268375397,
"learning_rate": 0.001,
"loss": 3.1380105590820313,
"step": 30100
},
{
"epoch": 0.2634546750370688,
"grad_norm": 0.2038249522447586,
"learning_rate": 0.001,
"loss": 3.1261517333984377,
"step": 30150
},
{
"epoch": 0.2638915816291701,
"grad_norm": 0.18218083679676056,
"learning_rate": 0.001,
"loss": 3.132615051269531,
"step": 30200
},
{
"epoch": 0.26432848822127136,
"grad_norm": 0.18844331800937653,
"learning_rate": 0.001,
"loss": 3.136449890136719,
"step": 30250
},
{
"epoch": 0.26476539481337263,
"grad_norm": 0.18804620206356049,
"learning_rate": 0.001,
"loss": 3.1299261474609374,
"step": 30300
},
{
"epoch": 0.2652023014054739,
"grad_norm": 0.22502121329307556,
"learning_rate": 0.001,
"loss": 3.129739990234375,
"step": 30350
},
{
"epoch": 0.2656392079975752,
"grad_norm": 0.19493575394153595,
"learning_rate": 0.001,
"loss": 3.134158020019531,
"step": 30400
},
{
"epoch": 0.26607611458967645,
"grad_norm": 0.19536198675632477,
"learning_rate": 0.001,
"loss": 3.1372747802734375,
"step": 30450
},
{
"epoch": 0.2665130211817777,
"grad_norm": 0.19196318089962006,
"learning_rate": 0.001,
"loss": 3.1322476196289064,
"step": 30500
},
{
"epoch": 0.266949927773879,
"grad_norm": 0.19632644951343536,
"learning_rate": 0.001,
"loss": 3.130311279296875,
"step": 30550
},
{
"epoch": 0.26738683436598026,
"grad_norm": 0.20887307822704315,
"learning_rate": 0.001,
"loss": 3.1273138427734377,
"step": 30600
},
{
"epoch": 0.26782374095808154,
"grad_norm": 0.21327802538871765,
"learning_rate": 0.001,
"loss": 3.1299261474609374,
"step": 30650
},
{
"epoch": 0.2682606475501828,
"grad_norm": 0.1895800083875656,
"learning_rate": 0.001,
"loss": 3.1285821533203126,
"step": 30700
},
{
"epoch": 0.2686975541422841,
"grad_norm": 0.18221773207187653,
"learning_rate": 0.001,
"loss": 3.1297210693359374,
"step": 30750
},
{
"epoch": 0.26913446073438535,
"grad_norm": 0.2368168830871582,
"learning_rate": 0.001,
"loss": 3.1300042724609374,
"step": 30800
},
{
"epoch": 0.2695713673264866,
"grad_norm": 0.1990697830915451,
"learning_rate": 0.001,
"loss": 3.1298654174804685,
"step": 30850
},
{
"epoch": 0.2700082739185879,
"grad_norm": 0.19808709621429443,
"learning_rate": 0.001,
"loss": 3.130029296875,
"step": 30900
},
{
"epoch": 0.2704451805106892,
"grad_norm": 0.1890052706003189,
"learning_rate": 0.001,
"loss": 3.1365090942382814,
"step": 30950
},
{
"epoch": 0.2708820871027905,
"grad_norm": 0.2171575129032135,
"learning_rate": 0.001,
"loss": 3.1323233032226563,
"step": 31000
},
{
"epoch": 0.27131899369489176,
"grad_norm": 0.1941414177417755,
"learning_rate": 0.001,
"loss": 3.1407867431640626,
"step": 31050
},
{
"epoch": 0.27175590028699304,
"grad_norm": 0.18840859830379486,
"learning_rate": 0.001,
"loss": 3.1423263549804688,
"step": 31100
},
{
"epoch": 0.2721928068790943,
"grad_norm": 0.2036226987838745,
"learning_rate": 0.001,
"loss": 3.131452941894531,
"step": 31150
},
{
"epoch": 0.2726297134711956,
"grad_norm": 0.2012850046157837,
"learning_rate": 0.001,
"loss": 3.1307635498046875,
"step": 31200
},
{
"epoch": 0.27306662006329685,
"grad_norm": 0.20540261268615723,
"learning_rate": 0.001,
"loss": 3.1357354736328125,
"step": 31250
},
{
"epoch": 0.2735035266553981,
"grad_norm": 0.21789686381816864,
"learning_rate": 0.001,
"loss": 3.130373840332031,
"step": 31300
},
{
"epoch": 0.2739404332474994,
"grad_norm": 0.19087694585323334,
"learning_rate": 0.001,
"loss": 3.13119384765625,
"step": 31350
},
{
"epoch": 0.27437733983960066,
"grad_norm": 0.19062580168247223,
"learning_rate": 0.001,
"loss": 3.13743896484375,
"step": 31400
},
{
"epoch": 0.27481424643170194,
"grad_norm": 0.19447751343250275,
"learning_rate": 0.001,
"loss": 3.129958190917969,
"step": 31450
},
{
"epoch": 0.2752511530238032,
"grad_norm": 0.21276557445526123,
"learning_rate": 0.001,
"loss": 3.130648193359375,
"step": 31500
},
{
"epoch": 0.2756880596159045,
"grad_norm": 0.20020796358585358,
"learning_rate": 0.001,
"loss": 3.1287417602539063,
"step": 31550
},
{
"epoch": 0.27612496620800575,
"grad_norm": 0.1890374720096588,
"learning_rate": 0.001,
"loss": 3.128742980957031,
"step": 31600
},
{
"epoch": 0.276561872800107,
"grad_norm": 0.2186155617237091,
"learning_rate": 0.001,
"loss": 3.1354425048828123,
"step": 31650
},
{
"epoch": 0.2769987793922083,
"grad_norm": 0.18568390607833862,
"learning_rate": 0.001,
"loss": 3.1281448364257813,
"step": 31700
},
{
"epoch": 0.27743568598430957,
"grad_norm": 0.21665509045124054,
"learning_rate": 0.001,
"loss": 3.125045166015625,
"step": 31750
},
{
"epoch": 0.2778725925764109,
"grad_norm": 0.19424442946910858,
"learning_rate": 0.001,
"loss": 3.1284844970703123,
"step": 31800
},
{
"epoch": 0.27830949916851216,
"grad_norm": 0.18474118411540985,
"learning_rate": 0.001,
"loss": 3.1286361694335936,
"step": 31850
},
{
"epoch": 0.27874640576061344,
"grad_norm": 0.2170608937740326,
"learning_rate": 0.001,
"loss": 3.123915100097656,
"step": 31900
},
{
"epoch": 0.2791833123527147,
"grad_norm": 0.19326017796993256,
"learning_rate": 0.001,
"loss": 3.1343429565429686,
"step": 31950
},
{
"epoch": 0.279620218944816,
"grad_norm": 0.19592459499835968,
"learning_rate": 0.001,
"loss": 3.130262451171875,
"step": 32000
},
{
"epoch": 0.28005712553691725,
"grad_norm": 0.211247980594635,
"learning_rate": 0.001,
"loss": 3.1321234130859374,
"step": 32050
},
{
"epoch": 0.2804940321290185,
"grad_norm": 0.17618435621261597,
"learning_rate": 0.001,
"loss": 3.1377389526367185,
"step": 32100
},
{
"epoch": 0.2809309387211198,
"grad_norm": 0.19517415761947632,
"learning_rate": 0.001,
"loss": 3.12404296875,
"step": 32150
},
{
"epoch": 0.28136784531322107,
"grad_norm": 0.18064315617084503,
"learning_rate": 0.001,
"loss": 3.1274789428710936,
"step": 32200
},
{
"epoch": 0.28180475190532234,
"grad_norm": 0.20532603561878204,
"learning_rate": 0.001,
"loss": 3.1279559326171875,
"step": 32250
},
{
"epoch": 0.2822416584974236,
"grad_norm": 0.20819665491580963,
"learning_rate": 0.001,
"loss": 3.127613830566406,
"step": 32300
},
{
"epoch": 0.2826785650895249,
"grad_norm": 0.1905108392238617,
"learning_rate": 0.001,
"loss": 3.12306884765625,
"step": 32350
},
{
"epoch": 0.28311547168162615,
"grad_norm": 0.20049357414245605,
"learning_rate": 0.001,
"loss": 3.1273574829101562,
"step": 32400
},
{
"epoch": 0.2835523782737274,
"grad_norm": 0.20894324779510498,
"learning_rate": 0.001,
"loss": 3.1294757080078126,
"step": 32450
},
{
"epoch": 0.2839892848658287,
"grad_norm": 0.1933911293745041,
"learning_rate": 0.001,
"loss": 3.128467712402344,
"step": 32500
},
{
"epoch": 0.28442619145792997,
"grad_norm": 0.22167061269283295,
"learning_rate": 0.001,
"loss": 3.1268344116210938,
"step": 32550
},
{
"epoch": 0.2848630980500313,
"grad_norm": 0.22398404777050018,
"learning_rate": 0.001,
"loss": 3.1302618408203124,
"step": 32600
},
{
"epoch": 0.28530000464213257,
"grad_norm": 0.199432373046875,
"learning_rate": 0.001,
"loss": 3.1282308959960936,
"step": 32650
},
{
"epoch": 0.28573691123423384,
"grad_norm": 0.20887622237205505,
"learning_rate": 0.001,
"loss": 3.12262451171875,
"step": 32700
},
{
"epoch": 0.2861738178263351,
"grad_norm": 0.21258816123008728,
"learning_rate": 0.001,
"loss": 3.1272879028320313,
"step": 32750
},
{
"epoch": 0.2866107244184364,
"grad_norm": 0.19245383143424988,
"learning_rate": 0.001,
"loss": 3.1249609375,
"step": 32800
},
{
"epoch": 0.28704763101053765,
"grad_norm": 0.19634206593036652,
"learning_rate": 0.001,
"loss": 3.1282113647460936,
"step": 32850
},
{
"epoch": 0.2874845376026389,
"grad_norm": 0.20635859668254852,
"learning_rate": 0.001,
"loss": 3.1323126220703124,
"step": 32900
},
{
"epoch": 0.2879214441947402,
"grad_norm": 0.1980540156364441,
"learning_rate": 0.001,
"loss": 3.12607177734375,
"step": 32950
},
{
"epoch": 0.28835835078684147,
"grad_norm": 0.2247803509235382,
"learning_rate": 0.001,
"loss": 3.1273507690429687,
"step": 33000
},
{
"epoch": 0.28835835078684147,
"eval_loss": 3.13966965675354,
"eval_runtime": 6.6578,
"eval_samples_per_second": 146.596,
"eval_steps_per_second": 9.162,
"step": 33000
},
{
"epoch": 0.28879525737894274,
"grad_norm": 0.19877365231513977,
"learning_rate": 0.001,
"loss": 3.1261257934570312,
"step": 33050
},
{
"epoch": 0.289232163971044,
"grad_norm": 0.188657745718956,
"learning_rate": 0.001,
"loss": 3.1309219360351563,
"step": 33100
},
{
"epoch": 0.2896690705631453,
"grad_norm": 0.19768841564655304,
"learning_rate": 0.001,
"loss": 3.1297171020507815,
"step": 33150
},
{
"epoch": 0.29010597715524655,
"grad_norm": 0.19437547028064728,
"learning_rate": 0.001,
"loss": 3.1266049194335936,
"step": 33200
},
{
"epoch": 0.2905428837473478,
"grad_norm": 0.17969079315662384,
"learning_rate": 0.001,
"loss": 3.121018981933594,
"step": 33250
},
{
"epoch": 0.2909797903394491,
"grad_norm": 0.21679022908210754,
"learning_rate": 0.001,
"loss": 3.1391571044921873,
"step": 33300
},
{
"epoch": 0.29141669693155037,
"grad_norm": 0.21835945546627045,
"learning_rate": 0.001,
"loss": 3.128807373046875,
"step": 33350
},
{
"epoch": 0.29185360352365164,
"grad_norm": 0.21215882897377014,
"learning_rate": 0.001,
"loss": 3.131269836425781,
"step": 33400
},
{
"epoch": 0.29229051011575297,
"grad_norm": 0.21180956065654755,
"learning_rate": 0.001,
"loss": 3.129888000488281,
"step": 33450
},
{
"epoch": 0.29272741670785424,
"grad_norm": 0.19976063072681427,
"learning_rate": 0.001,
"loss": 3.1300149536132813,
"step": 33500
},
{
"epoch": 0.2931643232999555,
"grad_norm": 0.2134390026330948,
"learning_rate": 0.001,
"loss": 3.1304666137695314,
"step": 33550
},
{
"epoch": 0.2936012298920568,
"grad_norm": 0.2210865467786789,
"learning_rate": 0.001,
"loss": 3.1274459838867186,
"step": 33600
},
{
"epoch": 0.29403813648415805,
"grad_norm": 0.24642080068588257,
"learning_rate": 0.001,
"loss": 3.1177212524414064,
"step": 33650
},
{
"epoch": 0.2944750430762593,
"grad_norm": 0.19083425402641296,
"learning_rate": 0.001,
"loss": 3.1270721435546873,
"step": 33700
},
{
"epoch": 0.2949119496683606,
"grad_norm": 0.21025504171848297,
"learning_rate": 0.001,
"loss": 3.1216671752929686,
"step": 33750
},
{
"epoch": 0.29534885626046187,
"grad_norm": 0.20673063397407532,
"learning_rate": 0.001,
"loss": 3.12328125,
"step": 33800
},
{
"epoch": 0.29578576285256314,
"grad_norm": 0.2012501209974289,
"learning_rate": 0.001,
"loss": 3.1250482177734376,
"step": 33850
},
{
"epoch": 0.2962226694446644,
"grad_norm": 0.19993802905082703,
"learning_rate": 0.001,
"loss": 3.127768859863281,
"step": 33900
},
{
"epoch": 0.2966595760367657,
"grad_norm": 0.20680220425128937,
"learning_rate": 0.001,
"loss": 3.1216485595703123,
"step": 33950
},
{
"epoch": 0.29709648262886695,
"grad_norm": 0.19726578891277313,
"learning_rate": 0.001,
"loss": 3.1306277465820314,
"step": 34000
},
{
"epoch": 0.2975333892209682,
"grad_norm": 0.18965868651866913,
"learning_rate": 0.001,
"loss": 3.1217327880859376,
"step": 34050
},
{
"epoch": 0.2979702958130695,
"grad_norm": 0.20208507776260376,
"learning_rate": 0.001,
"loss": 3.1228936767578124,
"step": 34100
},
{
"epoch": 0.29840720240517077,
"grad_norm": 0.1896025836467743,
"learning_rate": 0.001,
"loss": 3.127978515625,
"step": 34150
},
{
"epoch": 0.29884410899727204,
"grad_norm": 0.22235505282878876,
"learning_rate": 0.001,
"loss": 3.1210400390625,
"step": 34200
},
{
"epoch": 0.29928101558937337,
"grad_norm": 0.19582481682300568,
"learning_rate": 0.001,
"loss": 3.126922302246094,
"step": 34250
},
{
"epoch": 0.29971792218147464,
"grad_norm": 0.23264071345329285,
"learning_rate": 0.001,
"loss": 3.1247817993164064,
"step": 34300
},
{
"epoch": 0.3001548287735759,
"grad_norm": 0.22010396420955658,
"learning_rate": 0.001,
"loss": 3.128842468261719,
"step": 34350
},
{
"epoch": 0.3005917353656772,
"grad_norm": 0.20431144535541534,
"learning_rate": 0.001,
"loss": 3.1259225463867186,
"step": 34400
},
{
"epoch": 0.30102864195777845,
"grad_norm": 0.19362811744213104,
"learning_rate": 0.001,
"loss": 3.1228598022460936,
"step": 34450
},
{
"epoch": 0.3014655485498797,
"grad_norm": 0.2126990109682083,
"learning_rate": 0.001,
"loss": 3.1284494018554687,
"step": 34500
},
{
"epoch": 0.301902455141981,
"grad_norm": 0.19638939201831818,
"learning_rate": 0.001,
"loss": 3.127447204589844,
"step": 34550
},
{
"epoch": 0.30233936173408227,
"grad_norm": 0.19893930852413177,
"learning_rate": 0.001,
"loss": 3.1273974609375,
"step": 34600
},
{
"epoch": 0.30277626832618354,
"grad_norm": 0.1789814680814743,
"learning_rate": 0.001,
"loss": 3.1320419311523438,
"step": 34650
},
{
"epoch": 0.3032131749182848,
"grad_norm": 0.18921281397342682,
"learning_rate": 0.001,
"loss": 3.1326229858398436,
"step": 34700
},
{
"epoch": 0.3036500815103861,
"grad_norm": 0.1902114301919937,
"learning_rate": 0.001,
"loss": 3.1266995239257813,
"step": 34750
},
{
"epoch": 0.30408698810248735,
"grad_norm": 0.19261230528354645,
"learning_rate": 0.001,
"loss": 3.1286856079101564,
"step": 34800
},
{
"epoch": 0.3045238946945886,
"grad_norm": 0.19680652022361755,
"learning_rate": 0.001,
"loss": 3.1249435424804686,
"step": 34850
},
{
"epoch": 0.3049608012866899,
"grad_norm": 0.19080501794815063,
"learning_rate": 0.001,
"loss": 3.1208596801757813,
"step": 34900
},
{
"epoch": 0.30539770787879117,
"grad_norm": 0.19262269139289856,
"learning_rate": 0.001,
"loss": 3.1231289672851563,
"step": 34950
},
{
"epoch": 0.30583461447089244,
"grad_norm": 0.2081262618303299,
"learning_rate": 0.001,
"loss": 3.1275579833984377,
"step": 35000
},
{
"epoch": 0.3062715210629937,
"grad_norm": 0.19446617364883423,
"learning_rate": 0.001,
"loss": 3.1230514526367186,
"step": 35050
},
{
"epoch": 0.30670842765509504,
"grad_norm": 0.2236691564321518,
"learning_rate": 0.001,
"loss": 3.1231533813476564,
"step": 35100
},
{
"epoch": 0.3071453342471963,
"grad_norm": 0.20570874214172363,
"learning_rate": 0.001,
"loss": 3.1256478881835936,
"step": 35150
},
{
"epoch": 0.3075822408392976,
"grad_norm": 0.19653649628162384,
"learning_rate": 0.001,
"loss": 3.1202517700195314,
"step": 35200
},
{
"epoch": 0.30801914743139885,
"grad_norm": 0.19038353860378265,
"learning_rate": 0.001,
"loss": 3.122431640625,
"step": 35250
},
{
"epoch": 0.3084560540235001,
"grad_norm": 0.20607247948646545,
"learning_rate": 0.001,
"loss": 3.125845947265625,
"step": 35300
},
{
"epoch": 0.3088929606156014,
"grad_norm": 0.19749155640602112,
"learning_rate": 0.001,
"loss": 3.119671630859375,
"step": 35350
},
{
"epoch": 0.30932986720770267,
"grad_norm": 0.19711557030677795,
"learning_rate": 0.001,
"loss": 3.121778869628906,
"step": 35400
},
{
"epoch": 0.30976677379980394,
"grad_norm": 0.1881953328847885,
"learning_rate": 0.001,
"loss": 3.1199746704101563,
"step": 35450
},
{
"epoch": 0.3102036803919052,
"grad_norm": 0.18396177887916565,
"learning_rate": 0.001,
"loss": 3.129500732421875,
"step": 35500
},
{
"epoch": 0.3106405869840065,
"grad_norm": 0.1972985565662384,
"learning_rate": 0.001,
"loss": 3.1283114624023436,
"step": 35550
},
{
"epoch": 0.31107749357610776,
"grad_norm": 0.19514159858226776,
"learning_rate": 0.001,
"loss": 3.121815185546875,
"step": 35600
},
{
"epoch": 0.311514400168209,
"grad_norm": 0.18686498701572418,
"learning_rate": 0.001,
"loss": 3.1208340454101564,
"step": 35650
},
{
"epoch": 0.3119513067603103,
"grad_norm": 0.2287224531173706,
"learning_rate": 0.001,
"loss": 3.1249725341796877,
"step": 35700
},
{
"epoch": 0.31238821335241157,
"grad_norm": 0.2007932960987091,
"learning_rate": 0.001,
"loss": 3.1244778442382812,
"step": 35750
},
{
"epoch": 0.31282511994451284,
"grad_norm": 0.19419297575950623,
"learning_rate": 0.001,
"loss": 3.122419738769531,
"step": 35800
},
{
"epoch": 0.3132620265366141,
"grad_norm": 0.20398344099521637,
"learning_rate": 0.001,
"loss": 3.124461669921875,
"step": 35850
},
{
"epoch": 0.31369893312871544,
"grad_norm": 0.2154412418603897,
"learning_rate": 0.001,
"loss": 3.133614807128906,
"step": 35900
},
{
"epoch": 0.3141358397208167,
"grad_norm": 0.20187996327877045,
"learning_rate": 0.001,
"loss": 3.1194140625,
"step": 35950
},
{
"epoch": 0.314572746312918,
"grad_norm": 0.20458835363388062,
"learning_rate": 0.001,
"loss": 3.1240689086914064,
"step": 36000
},
{
"epoch": 0.314572746312918,
"eval_loss": 3.134185791015625,
"eval_runtime": 6.6286,
"eval_samples_per_second": 147.241,
"eval_steps_per_second": 9.203,
"step": 36000
},
{
"epoch": 0.31500965290501926,
"grad_norm": 0.2001228779554367,
"learning_rate": 0.001,
"loss": 3.1255401611328124,
"step": 36050
},
{
"epoch": 0.3154465594971205,
"grad_norm": 0.20988516509532928,
"learning_rate": 0.001,
"loss": 3.12226806640625,
"step": 36100
},
{
"epoch": 0.3158834660892218,
"grad_norm": 0.19366823136806488,
"learning_rate": 0.001,
"loss": 3.1212237548828123,
"step": 36150
},
{
"epoch": 0.31632037268132307,
"grad_norm": 0.2106378674507141,
"learning_rate": 0.001,
"loss": 3.1272991943359374,
"step": 36200
},
{
"epoch": 0.31675727927342434,
"grad_norm": 0.19957683980464935,
"learning_rate": 0.001,
"loss": 3.1302420043945314,
"step": 36250
},
{
"epoch": 0.3171941858655256,
"grad_norm": 0.21629318594932556,
"learning_rate": 0.001,
"loss": 3.11791259765625,
"step": 36300
},
{
"epoch": 0.3176310924576269,
"grad_norm": 0.21578121185302734,
"learning_rate": 0.001,
"loss": 3.1275088500976564,
"step": 36350
},
{
"epoch": 0.31806799904972816,
"grad_norm": 0.2041950672864914,
"learning_rate": 0.001,
"loss": 3.1243646240234373,
"step": 36400
},
{
"epoch": 0.31850490564182943,
"grad_norm": 0.18400739133358002,
"learning_rate": 0.001,
"loss": 3.1152313232421873,
"step": 36450
},
{
"epoch": 0.3189418122339307,
"grad_norm": 0.215887650847435,
"learning_rate": 0.001,
"loss": 3.1258389282226564,
"step": 36500
},
{
"epoch": 0.31937871882603197,
"grad_norm": 0.228815495967865,
"learning_rate": 0.001,
"loss": 3.1239584350585936,
"step": 36550
},
{
"epoch": 0.31981562541813324,
"grad_norm": 0.19070304930210114,
"learning_rate": 0.001,
"loss": 3.126492919921875,
"step": 36600
},
{
"epoch": 0.3202525320102345,
"grad_norm": 0.21217049658298492,
"learning_rate": 0.001,
"loss": 3.113551025390625,
"step": 36650
},
{
"epoch": 0.3206894386023358,
"grad_norm": 0.2227277010679245,
"learning_rate": 0.001,
"loss": 3.1215505981445313,
"step": 36700
},
{
"epoch": 0.3211263451944371,
"grad_norm": 0.20254331827163696,
"learning_rate": 0.001,
"loss": 3.124097900390625,
"step": 36750
},
{
"epoch": 0.3215632517865384,
"grad_norm": 0.19885151088237762,
"learning_rate": 0.001,
"loss": 3.1168997192382815,
"step": 36800
},
{
"epoch": 0.32200015837863966,
"grad_norm": 0.20904117822647095,
"learning_rate": 0.001,
"loss": 3.117975158691406,
"step": 36850
},
{
"epoch": 0.3224370649707409,
"grad_norm": 0.18927255272865295,
"learning_rate": 0.001,
"loss": 3.118792724609375,
"step": 36900
},
{
"epoch": 0.3228739715628422,
"grad_norm": 0.1917814016342163,
"learning_rate": 0.001,
"loss": 3.1174697875976562,
"step": 36950
},
{
"epoch": 0.32331087815494347,
"grad_norm": 0.19405600428581238,
"learning_rate": 0.001,
"loss": 3.1238449096679686,
"step": 37000
},
{
"epoch": 0.32374778474704474,
"grad_norm": 0.20563170313835144,
"learning_rate": 0.001,
"loss": 3.1181976318359377,
"step": 37050
},
{
"epoch": 0.324184691339146,
"grad_norm": 0.18454208970069885,
"learning_rate": 0.001,
"loss": 3.1169464111328127,
"step": 37100
},
{
"epoch": 0.3246215979312473,
"grad_norm": 0.21267572045326233,
"learning_rate": 0.001,
"loss": 3.1171444702148436,
"step": 37150
},
{
"epoch": 0.32505850452334856,
"grad_norm": 0.20087648928165436,
"learning_rate": 0.001,
"loss": 3.1206906127929686,
"step": 37200
},
{
"epoch": 0.32549541111544983,
"grad_norm": 0.2119586318731308,
"learning_rate": 0.001,
"loss": 3.122110595703125,
"step": 37250
},
{
"epoch": 0.3259323177075511,
"grad_norm": 0.1976786106824875,
"learning_rate": 0.001,
"loss": 3.1239993286132814,
"step": 37300
},
{
"epoch": 0.32636922429965237,
"grad_norm": 0.20974650979042053,
"learning_rate": 0.001,
"loss": 3.1207305908203127,
"step": 37350
},
{
"epoch": 0.32680613089175364,
"grad_norm": 0.18441034853458405,
"learning_rate": 0.001,
"loss": 3.1210308837890626,
"step": 37400
},
{
"epoch": 0.3272430374838549,
"grad_norm": 0.19625826179981232,
"learning_rate": 0.001,
"loss": 3.1168194580078126,
"step": 37450
},
{
"epoch": 0.3276799440759562,
"grad_norm": 0.20468471944332123,
"learning_rate": 0.001,
"loss": 3.122637939453125,
"step": 37500
},
{
"epoch": 0.3281168506680575,
"grad_norm": 0.20104554295539856,
"learning_rate": 0.001,
"loss": 3.122271728515625,
"step": 37550
},
{
"epoch": 0.3285537572601588,
"grad_norm": 0.2072601318359375,
"learning_rate": 0.001,
"loss": 3.124110107421875,
"step": 37600
},
{
"epoch": 0.32899066385226006,
"grad_norm": 0.21337221562862396,
"learning_rate": 0.001,
"loss": 3.1205178833007814,
"step": 37650
},
{
"epoch": 0.32942757044436133,
"grad_norm": 0.21294334530830383,
"learning_rate": 0.001,
"loss": 3.120369873046875,
"step": 37700
},
{
"epoch": 0.3298644770364626,
"grad_norm": 0.20325663685798645,
"learning_rate": 0.001,
"loss": 3.1193276977539064,
"step": 37750
},
{
"epoch": 0.33030138362856387,
"grad_norm": 0.213812455534935,
"learning_rate": 0.001,
"loss": 3.118055419921875,
"step": 37800
},
{
"epoch": 0.33073829022066514,
"grad_norm": 0.20142120122909546,
"learning_rate": 0.001,
"loss": 3.1187542724609374,
"step": 37850
},
{
"epoch": 0.3311751968127664,
"grad_norm": 0.21969522535800934,
"learning_rate": 0.001,
"loss": 3.113778076171875,
"step": 37900
},
{
"epoch": 0.3316121034048677,
"grad_norm": 0.2110193967819214,
"learning_rate": 0.001,
"loss": 3.119208984375,
"step": 37950
},
{
"epoch": 0.33204900999696896,
"grad_norm": 0.18195150792598724,
"learning_rate": 0.001,
"loss": 3.1189395141601564,
"step": 38000
},
{
"epoch": 0.33248591658907023,
"grad_norm": 0.22660240530967712,
"learning_rate": 0.001,
"loss": 3.1194659423828126,
"step": 38050
},
{
"epoch": 0.3329228231811715,
"grad_norm": 0.21752111613750458,
"learning_rate": 0.001,
"loss": 3.1144259643554686,
"step": 38100
},
{
"epoch": 0.3333597297732728,
"grad_norm": 0.19816753268241882,
"learning_rate": 0.001,
"loss": 3.118554992675781,
"step": 38150
},
{
"epoch": 0.33379663636537404,
"grad_norm": 0.1945333182811737,
"learning_rate": 0.001,
"loss": 3.126115417480469,
"step": 38200
},
{
"epoch": 0.3342335429574753,
"grad_norm": 0.19264337420463562,
"learning_rate": 0.001,
"loss": 3.126806640625,
"step": 38250
},
{
"epoch": 0.3346704495495766,
"grad_norm": 0.20601359009742737,
"learning_rate": 0.001,
"loss": 3.121160583496094,
"step": 38300
},
{
"epoch": 0.33510735614167786,
"grad_norm": 0.20243000984191895,
"learning_rate": 0.001,
"loss": 3.1194607543945314,
"step": 38350
},
{
"epoch": 0.3355442627337792,
"grad_norm": 0.2247641682624817,
"learning_rate": 0.001,
"loss": 3.1182537841796876,
"step": 38400
},
{
"epoch": 0.33598116932588046,
"grad_norm": 0.19684353470802307,
"learning_rate": 0.001,
"loss": 3.1167251586914064,
"step": 38450
},
{
"epoch": 0.33641807591798173,
"grad_norm": 0.220078244805336,
"learning_rate": 0.001,
"loss": 3.1210321044921874,
"step": 38500
},
{
"epoch": 0.336854982510083,
"grad_norm": 0.22685925662517548,
"learning_rate": 0.001,
"loss": 3.1189035034179686,
"step": 38550
},
{
"epoch": 0.3372918891021843,
"grad_norm": 0.2115422934293747,
"learning_rate": 0.001,
"loss": 3.113541259765625,
"step": 38600
},
{
"epoch": 0.33772879569428554,
"grad_norm": 0.22083063423633575,
"learning_rate": 0.001,
"loss": 3.1152862548828124,
"step": 38650
},
{
"epoch": 0.3381657022863868,
"grad_norm": 0.24635091423988342,
"learning_rate": 0.001,
"loss": 3.1192333984375,
"step": 38700
},
{
"epoch": 0.3386026088784881,
"grad_norm": 0.1958378255367279,
"learning_rate": 0.001,
"loss": 3.1148300170898438,
"step": 38750
},
{
"epoch": 0.33903951547058936,
"grad_norm": 0.18606902658939362,
"learning_rate": 0.001,
"loss": 3.116632995605469,
"step": 38800
},
{
"epoch": 0.33947642206269063,
"grad_norm": 0.24798093736171722,
"learning_rate": 0.001,
"loss": 3.121573486328125,
"step": 38850
},
{
"epoch": 0.3399133286547919,
"grad_norm": 0.1906382292509079,
"learning_rate": 0.001,
"loss": 3.1170053100585937,
"step": 38900
},
{
"epoch": 0.3403502352468932,
"grad_norm": 0.1849077194929123,
"learning_rate": 0.001,
"loss": 3.1188088989257814,
"step": 38950
},
{
"epoch": 0.34078714183899445,
"grad_norm": 0.19308972358703613,
"learning_rate": 0.001,
"loss": 3.12092529296875,
"step": 39000
},
{
"epoch": 0.34078714183899445,
"eval_loss": 3.1283721923828125,
"eval_runtime": 6.2376,
"eval_samples_per_second": 156.472,
"eval_steps_per_second": 9.779,
"step": 39000
},
{
"epoch": 0.3412240484310957,
"grad_norm": 0.22229205071926117,
"learning_rate": 0.001,
"loss": 3.1210977172851564,
"step": 39050
},
{
"epoch": 0.341660955023197,
"grad_norm": 0.20119309425354004,
"learning_rate": 0.001,
"loss": 3.1165362548828126,
"step": 39100
},
{
"epoch": 0.34209786161529826,
"grad_norm": 0.1843200922012329,
"learning_rate": 0.001,
"loss": 3.1190365600585936,
"step": 39150
},
{
"epoch": 0.3425347682073996,
"grad_norm": 0.2104836255311966,
"learning_rate": 0.001,
"loss": 3.117159423828125,
"step": 39200
},
{
"epoch": 0.34297167479950086,
"grad_norm": 0.20398400723934174,
"learning_rate": 0.001,
"loss": 3.114034423828125,
"step": 39250
},
{
"epoch": 0.34340858139160213,
"grad_norm": 0.22703580558300018,
"learning_rate": 0.001,
"loss": 3.1191915893554687,
"step": 39300
},
{
"epoch": 0.3438454879837034,
"grad_norm": 0.2472376674413681,
"learning_rate": 0.001,
"loss": 3.1246774291992185,
"step": 39350
},
{
"epoch": 0.3442823945758047,
"grad_norm": 0.21263231337070465,
"learning_rate": 0.001,
"loss": 3.1177713012695314,
"step": 39400
},
{
"epoch": 0.34471930116790594,
"grad_norm": 0.19162337481975555,
"learning_rate": 0.001,
"loss": 3.117065734863281,
"step": 39450
},
{
"epoch": 0.3451562077600072,
"grad_norm": 0.2151661068201065,
"learning_rate": 0.001,
"loss": 3.1196600341796876,
"step": 39500
},
{
"epoch": 0.3455931143521085,
"grad_norm": 0.2298618108034134,
"learning_rate": 0.001,
"loss": 3.1223019409179686,
"step": 39550
},
{
"epoch": 0.34603002094420976,
"grad_norm": 0.2149757444858551,
"learning_rate": 0.001,
"loss": 3.116190185546875,
"step": 39600
},
{
"epoch": 0.34646692753631103,
"grad_norm": 0.19931890070438385,
"learning_rate": 0.001,
"loss": 3.1219137573242186,
"step": 39650
},
{
"epoch": 0.3469038341284123,
"grad_norm": 0.21460124850273132,
"learning_rate": 0.001,
"loss": 3.119295654296875,
"step": 39700
},
{
"epoch": 0.3473407407205136,
"grad_norm": 0.21581821143627167,
"learning_rate": 0.001,
"loss": 3.117487487792969,
"step": 39750
},
{
"epoch": 0.34777764731261485,
"grad_norm": 0.19807817041873932,
"learning_rate": 0.001,
"loss": 3.1186810302734376,
"step": 39800
},
{
"epoch": 0.3482145539047161,
"grad_norm": 0.21846213936805725,
"learning_rate": 0.001,
"loss": 3.12263427734375,
"step": 39850
},
{
"epoch": 0.3486514604968174,
"grad_norm": 0.20867601037025452,
"learning_rate": 0.001,
"loss": 3.11615478515625,
"step": 39900
},
{
"epoch": 0.34908836708891866,
"grad_norm": 0.1970822513103485,
"learning_rate": 0.001,
"loss": 3.115738830566406,
"step": 39950
},
{
"epoch": 0.34952527368101993,
"grad_norm": 0.219162717461586,
"learning_rate": 0.001,
"loss": 3.119095458984375,
"step": 40000
},
{
"epoch": 0.34996218027312126,
"grad_norm": 0.18753738701343536,
"learning_rate": 0.001,
"loss": 3.114410095214844,
"step": 40050
},
{
"epoch": 0.35039908686522253,
"grad_norm": 0.2209300994873047,
"learning_rate": 0.001,
"loss": 3.1168057250976564,
"step": 40100
},
{
"epoch": 0.3508359934573238,
"grad_norm": 0.237261101603508,
"learning_rate": 0.001,
"loss": 3.1176416015625,
"step": 40150
},
{
"epoch": 0.3512729000494251,
"grad_norm": 0.1959320306777954,
"learning_rate": 0.001,
"loss": 3.1155279541015624,
"step": 40200
},
{
"epoch": 0.35170980664152635,
"grad_norm": 0.1914614588022232,
"learning_rate": 0.001,
"loss": 3.1104986572265627,
"step": 40250
},
{
"epoch": 0.3521467132336276,
"grad_norm": 0.22222118079662323,
"learning_rate": 0.001,
"loss": 3.11650634765625,
"step": 40300
},
{
"epoch": 0.3525836198257289,
"grad_norm": 0.1963379681110382,
"learning_rate": 0.001,
"loss": 3.1162673950195314,
"step": 40350
},
{
"epoch": 0.35302052641783016,
"grad_norm": 0.19033868610858917,
"learning_rate": 0.001,
"loss": 3.11138671875,
"step": 40400
},
{
"epoch": 0.35345743300993143,
"grad_norm": 0.19313368201255798,
"learning_rate": 0.001,
"loss": 3.11771484375,
"step": 40450
},
{
"epoch": 0.3538943396020327,
"grad_norm": 0.2146463394165039,
"learning_rate": 0.001,
"loss": 3.1086221313476563,
"step": 40500
},
{
"epoch": 0.354331246194134,
"grad_norm": 0.22047048807144165,
"learning_rate": 0.001,
"loss": 3.1154776000976563,
"step": 40550
},
{
"epoch": 0.35476815278623525,
"grad_norm": 0.19261007010936737,
"learning_rate": 0.001,
"loss": 3.118349304199219,
"step": 40600
},
{
"epoch": 0.3552050593783365,
"grad_norm": 0.21026641130447388,
"learning_rate": 0.001,
"loss": 3.115614929199219,
"step": 40650
},
{
"epoch": 0.3556419659704378,
"grad_norm": 0.19967412948608398,
"learning_rate": 0.001,
"loss": 3.1102642822265625,
"step": 40700
},
{
"epoch": 0.35607887256253906,
"grad_norm": 0.19184772670269012,
"learning_rate": 0.001,
"loss": 3.1165069580078124,
"step": 40750
},
{
"epoch": 0.35651577915464033,
"grad_norm": 0.20440271496772766,
"learning_rate": 0.001,
"loss": 3.1100430297851562,
"step": 40800
},
{
"epoch": 0.35695268574674166,
"grad_norm": 0.21017535030841827,
"learning_rate": 0.001,
"loss": 3.1197610473632813,
"step": 40850
},
{
"epoch": 0.35738959233884293,
"grad_norm": 0.19996143877506256,
"learning_rate": 0.001,
"loss": 3.116007080078125,
"step": 40900
},
{
"epoch": 0.3578264989309442,
"grad_norm": 0.22619201242923737,
"learning_rate": 0.001,
"loss": 3.1122076416015627,
"step": 40950
},
{
"epoch": 0.3582634055230455,
"grad_norm": 0.19171087443828583,
"learning_rate": 0.001,
"loss": 3.115628356933594,
"step": 41000
},
{
"epoch": 0.35870031211514675,
"grad_norm": 0.2097511887550354,
"learning_rate": 0.001,
"loss": 3.1177496337890624,
"step": 41050
},
{
"epoch": 0.359137218707248,
"grad_norm": 0.19585464894771576,
"learning_rate": 0.001,
"loss": 3.11097412109375,
"step": 41100
},
{
"epoch": 0.3595741252993493,
"grad_norm": 0.1915462464094162,
"learning_rate": 0.001,
"loss": 3.1149215698242188,
"step": 41150
},
{
"epoch": 0.36001103189145056,
"grad_norm": 0.21087348461151123,
"learning_rate": 0.001,
"loss": 3.1177166748046874,
"step": 41200
},
{
"epoch": 0.36044793848355183,
"grad_norm": 0.20831310749053955,
"learning_rate": 0.001,
"loss": 3.1163970947265627,
"step": 41250
},
{
"epoch": 0.3608848450756531,
"grad_norm": 0.22942714393138885,
"learning_rate": 0.001,
"loss": 3.1201040649414065,
"step": 41300
},
{
"epoch": 0.3613217516677544,
"grad_norm": 0.19121773540973663,
"learning_rate": 0.001,
"loss": 3.114645690917969,
"step": 41350
},
{
"epoch": 0.36175865825985565,
"grad_norm": 0.23914961516857147,
"learning_rate": 0.001,
"loss": 3.118790588378906,
"step": 41400
},
{
"epoch": 0.3621955648519569,
"grad_norm": 0.2047552764415741,
"learning_rate": 0.001,
"loss": 3.1183242797851562,
"step": 41450
},
{
"epoch": 0.3626324714440582,
"grad_norm": 0.18905483186244965,
"learning_rate": 0.001,
"loss": 3.1132696533203124,
"step": 41500
},
{
"epoch": 0.36306937803615946,
"grad_norm": 0.21784336864948273,
"learning_rate": 0.001,
"loss": 3.1163897705078125,
"step": 41550
},
{
"epoch": 0.36350628462826073,
"grad_norm": 0.2144789695739746,
"learning_rate": 0.001,
"loss": 3.115078125,
"step": 41600
},
{
"epoch": 0.363943191220362,
"grad_norm": 0.21054674685001373,
"learning_rate": 0.001,
"loss": 3.1157470703125,
"step": 41650
},
{
"epoch": 0.36438009781246333,
"grad_norm": 0.19948728382587433,
"learning_rate": 0.001,
"loss": 3.11491455078125,
"step": 41700
},
{
"epoch": 0.3648170044045646,
"grad_norm": 0.20507873594760895,
"learning_rate": 0.001,
"loss": 3.1132489013671876,
"step": 41750
},
{
"epoch": 0.3652539109966659,
"grad_norm": 0.2141726315021515,
"learning_rate": 0.001,
"loss": 3.1157525634765624,
"step": 41800
},
{
"epoch": 0.36569081758876715,
"grad_norm": 0.2137220799922943,
"learning_rate": 0.001,
"loss": 3.114307861328125,
"step": 41850
},
{
"epoch": 0.3661277241808684,
"grad_norm": 0.21856769919395447,
"learning_rate": 0.001,
"loss": 3.11392822265625,
"step": 41900
},
{
"epoch": 0.3665646307729697,
"grad_norm": 0.20412378013134003,
"learning_rate": 0.001,
"loss": 3.1116738891601563,
"step": 41950
},
{
"epoch": 0.36700153736507096,
"grad_norm": 0.20900988578796387,
"learning_rate": 0.001,
"loss": 3.113911437988281,
"step": 42000
},
{
"epoch": 0.36700153736507096,
"eval_loss": 3.126250982284546,
"eval_runtime": 5.4777,
"eval_samples_per_second": 178.177,
"eval_steps_per_second": 11.136,
"step": 42000
},
{
"epoch": 0.36743844395717223,
"grad_norm": 0.2194087654352188,
"learning_rate": 0.001,
"loss": 3.1196575927734376,
"step": 42050
},
{
"epoch": 0.3678753505492735,
"grad_norm": 0.22563667595386505,
"learning_rate": 0.001,
"loss": 3.1091033935546877,
"step": 42100
},
{
"epoch": 0.3683122571413748,
"grad_norm": 0.1888885796070099,
"learning_rate": 0.001,
"loss": 3.1200460815429687,
"step": 42150
},
{
"epoch": 0.36874916373347605,
"grad_norm": 0.22187890112400055,
"learning_rate": 0.001,
"loss": 3.1165335083007815,
"step": 42200
},
{
"epoch": 0.3691860703255773,
"grad_norm": 0.21313756704330444,
"learning_rate": 0.001,
"loss": 3.1126593017578124,
"step": 42250
},
{
"epoch": 0.3696229769176786,
"grad_norm": 0.19937962293624878,
"learning_rate": 0.001,
"loss": 3.118894348144531,
"step": 42300
},
{
"epoch": 0.37005988350977986,
"grad_norm": 0.22994975745677948,
"learning_rate": 0.001,
"loss": 3.1161279296875,
"step": 42350
},
{
"epoch": 0.37049679010188113,
"grad_norm": 0.184345543384552,
"learning_rate": 0.001,
"loss": 3.116735534667969,
"step": 42400
},
{
"epoch": 0.3709336966939824,
"grad_norm": 0.1899576485157013,
"learning_rate": 0.001,
"loss": 3.1158157348632813,
"step": 42450
},
{
"epoch": 0.3713706032860837,
"grad_norm": 0.2235243171453476,
"learning_rate": 0.001,
"loss": 3.105382080078125,
"step": 42500
},
{
"epoch": 0.371807509878185,
"grad_norm": 0.18403592705726624,
"learning_rate": 0.001,
"loss": 3.1148175048828124,
"step": 42550
},
{
"epoch": 0.3722444164702863,
"grad_norm": 0.19629758596420288,
"learning_rate": 0.001,
"loss": 3.1145794677734373,
"step": 42600
},
{
"epoch": 0.37268132306238755,
"grad_norm": 0.20445330440998077,
"learning_rate": 0.001,
"loss": 3.1126394653320313,
"step": 42650
},
{
"epoch": 0.3731182296544888,
"grad_norm": 0.21803605556488037,
"learning_rate": 0.001,
"loss": 3.1135186767578125,
"step": 42700
},
{
"epoch": 0.3735551362465901,
"grad_norm": 0.20326778292655945,
"learning_rate": 0.001,
"loss": 3.1120184326171874,
"step": 42750
},
{
"epoch": 0.37399204283869136,
"grad_norm": 0.19862455129623413,
"learning_rate": 0.001,
"loss": 3.1159231567382815,
"step": 42800
},
{
"epoch": 0.37442894943079263,
"grad_norm": 0.21534425020217896,
"learning_rate": 0.001,
"loss": 3.110506896972656,
"step": 42850
},
{
"epoch": 0.3748658560228939,
"grad_norm": 0.20181404054164886,
"learning_rate": 0.001,
"loss": 3.1135235595703126,
"step": 42900
},
{
"epoch": 0.3753027626149952,
"grad_norm": 0.21086396276950836,
"learning_rate": 0.001,
"loss": 3.1066864013671873,
"step": 42950
},
{
"epoch": 0.37573966920709645,
"grad_norm": 0.2272285521030426,
"learning_rate": 0.001,
"loss": 3.1153875732421876,
"step": 43000
},
{
"epoch": 0.3761765757991977,
"grad_norm": 0.21504978835582733,
"learning_rate": 0.001,
"loss": 3.109436340332031,
"step": 43050
},
{
"epoch": 0.376613482391299,
"grad_norm": 0.20367594063282013,
"learning_rate": 0.001,
"loss": 3.1120089721679687,
"step": 43100
},
{
"epoch": 0.37705038898340026,
"grad_norm": 0.2112228125333786,
"learning_rate": 0.001,
"loss": 3.112843017578125,
"step": 43150
},
{
"epoch": 0.37748729557550154,
"grad_norm": 0.195306658744812,
"learning_rate": 0.001,
"loss": 3.112016296386719,
"step": 43200
},
{
"epoch": 0.3779242021676028,
"grad_norm": 0.21097637712955475,
"learning_rate": 0.001,
"loss": 3.1130935668945314,
"step": 43250
},
{
"epoch": 0.3783611087597041,
"grad_norm": 0.21255624294281006,
"learning_rate": 0.001,
"loss": 3.1110089111328123,
"step": 43300
},
{
"epoch": 0.3787980153518054,
"grad_norm": 0.225327730178833,
"learning_rate": 0.001,
"loss": 3.105743713378906,
"step": 43350
},
{
"epoch": 0.3792349219439067,
"grad_norm": 0.19936905801296234,
"learning_rate": 0.001,
"loss": 3.1101507568359374,
"step": 43400
},
{
"epoch": 0.37967182853600795,
"grad_norm": 0.19566135108470917,
"learning_rate": 0.001,
"loss": 3.1115786743164064,
"step": 43450
},
{
"epoch": 0.3801087351281092,
"grad_norm": 0.22114816308021545,
"learning_rate": 0.001,
"loss": 3.1026132202148435,
"step": 43500
},
{
"epoch": 0.3805456417202105,
"grad_norm": 0.2205081284046173,
"learning_rate": 0.001,
"loss": 3.1094985961914063,
"step": 43550
},
{
"epoch": 0.38098254831231176,
"grad_norm": 0.18787072598934174,
"learning_rate": 0.001,
"loss": 3.1099993896484377,
"step": 43600
},
{
"epoch": 0.38141945490441304,
"grad_norm": 0.19435258209705353,
"learning_rate": 0.001,
"loss": 3.1116567993164064,
"step": 43650
},
{
"epoch": 0.3818563614965143,
"grad_norm": 0.2072121649980545,
"learning_rate": 0.001,
"loss": 3.1085601806640626,
"step": 43700
},
{
"epoch": 0.3822932680886156,
"grad_norm": 0.21253183484077454,
"learning_rate": 0.001,
"loss": 3.119143371582031,
"step": 43750
},
{
"epoch": 0.38273017468071685,
"grad_norm": 0.19218683242797852,
"learning_rate": 0.001,
"loss": 3.1107879638671876,
"step": 43800
},
{
"epoch": 0.3831670812728181,
"grad_norm": 0.19902510941028595,
"learning_rate": 0.001,
"loss": 3.1106719970703125,
"step": 43850
},
{
"epoch": 0.3836039878649194,
"grad_norm": 0.19710908830165863,
"learning_rate": 0.001,
"loss": 3.113710632324219,
"step": 43900
},
{
"epoch": 0.38404089445702067,
"grad_norm": 0.2198529839515686,
"learning_rate": 0.001,
"loss": 3.1120013427734374,
"step": 43950
},
{
"epoch": 0.38447780104912194,
"grad_norm": 0.21689631044864655,
"learning_rate": 0.001,
"loss": 3.1084619140625,
"step": 44000
},
{
"epoch": 0.3849147076412232,
"grad_norm": 0.1787080466747284,
"learning_rate": 0.001,
"loss": 3.108823547363281,
"step": 44050
},
{
"epoch": 0.3853516142333245,
"grad_norm": 0.19698339700698853,
"learning_rate": 0.001,
"loss": 3.1080697631835936,
"step": 44100
},
{
"epoch": 0.38578852082542575,
"grad_norm": 0.2172456830739975,
"learning_rate": 0.001,
"loss": 3.1126702880859374,
"step": 44150
},
{
"epoch": 0.3862254274175271,
"grad_norm": 0.2018110752105713,
"learning_rate": 0.001,
"loss": 3.108742370605469,
"step": 44200
},
{
"epoch": 0.38666233400962835,
"grad_norm": 0.21206806600093842,
"learning_rate": 0.001,
"loss": 3.109871520996094,
"step": 44250
},
{
"epoch": 0.3870992406017296,
"grad_norm": 0.20941662788391113,
"learning_rate": 0.001,
"loss": 3.109721984863281,
"step": 44300
},
{
"epoch": 0.3875361471938309,
"grad_norm": 0.20126648247241974,
"learning_rate": 0.001,
"loss": 3.1068310546875,
"step": 44350
},
{
"epoch": 0.38797305378593216,
"grad_norm": 0.18548788130283356,
"learning_rate": 0.001,
"loss": 3.11313720703125,
"step": 44400
},
{
"epoch": 0.38840996037803344,
"grad_norm": 0.19674333930015564,
"learning_rate": 0.001,
"loss": 3.1099951171875,
"step": 44450
},
{
"epoch": 0.3888468669701347,
"grad_norm": 0.2064523547887802,
"learning_rate": 0.001,
"loss": 3.110168762207031,
"step": 44500
},
{
"epoch": 0.389283773562236,
"grad_norm": 0.21665744483470917,
"learning_rate": 0.001,
"loss": 3.112205810546875,
"step": 44550
},
{
"epoch": 0.38972068015433725,
"grad_norm": 0.2189055234193802,
"learning_rate": 0.001,
"loss": 3.1112820434570314,
"step": 44600
},
{
"epoch": 0.3901575867464385,
"grad_norm": 0.2260696142911911,
"learning_rate": 0.001,
"loss": 3.11107177734375,
"step": 44650
},
{
"epoch": 0.3905944933385398,
"grad_norm": 0.1923733949661255,
"learning_rate": 0.001,
"loss": 3.109862060546875,
"step": 44700
},
{
"epoch": 0.39103139993064107,
"grad_norm": 0.2178075909614563,
"learning_rate": 0.001,
"loss": 3.1175469970703125,
"step": 44750
},
{
"epoch": 0.39146830652274234,
"grad_norm": 0.20600248873233795,
"learning_rate": 0.001,
"loss": 3.1112896728515627,
"step": 44800
},
{
"epoch": 0.3919052131148436,
"grad_norm": 0.1963505744934082,
"learning_rate": 0.001,
"loss": 3.111578369140625,
"step": 44850
},
{
"epoch": 0.3923421197069449,
"grad_norm": 0.2002233862876892,
"learning_rate": 0.001,
"loss": 3.1122171020507814,
"step": 44900
},
{
"epoch": 0.39277902629904615,
"grad_norm": 0.18052497506141663,
"learning_rate": 0.001,
"loss": 3.1110784912109377,
"step": 44950
},
{
"epoch": 0.3932159328911475,
"grad_norm": 0.20493514835834503,
"learning_rate": 0.001,
"loss": 3.1110995483398436,
"step": 45000
},
{
"epoch": 0.3932159328911475,
"eval_loss": 3.1220333576202393,
"eval_runtime": 7.0319,
"eval_samples_per_second": 138.796,
"eval_steps_per_second": 8.675,
"step": 45000
},
{
"epoch": 0.39365283948324875,
"grad_norm": 0.22265596687793732,
"learning_rate": 0.001,
"loss": 3.1090939331054686,
"step": 45050
},
{
"epoch": 0.39408974607535,
"grad_norm": 0.1966310739517212,
"learning_rate": 0.001,
"loss": 3.104522705078125,
"step": 45100
},
{
"epoch": 0.3945266526674513,
"grad_norm": 0.20936229825019836,
"learning_rate": 0.001,
"loss": 3.103654479980469,
"step": 45150
},
{
"epoch": 0.39496355925955257,
"grad_norm": 0.227894589304924,
"learning_rate": 0.001,
"loss": 3.10695068359375,
"step": 45200
},
{
"epoch": 0.39540046585165384,
"grad_norm": 0.2056993544101715,
"learning_rate": 0.001,
"loss": 3.108395690917969,
"step": 45250
},
{
"epoch": 0.3958373724437551,
"grad_norm": 0.21064303815364838,
"learning_rate": 0.001,
"loss": 3.1041717529296875,
"step": 45300
},
{
"epoch": 0.3962742790358564,
"grad_norm": 0.2195715755224228,
"learning_rate": 0.001,
"loss": 3.10718505859375,
"step": 45350
},
{
"epoch": 0.39671118562795765,
"grad_norm": 0.18564003705978394,
"learning_rate": 0.001,
"loss": 3.110213623046875,
"step": 45400
},
{
"epoch": 0.3971480922200589,
"grad_norm": 0.22178950905799866,
"learning_rate": 0.001,
"loss": 3.1148031616210936,
"step": 45450
},
{
"epoch": 0.3975849988121602,
"grad_norm": 0.19489014148712158,
"learning_rate": 0.001,
"loss": 3.1107083129882813,
"step": 45500
},
{
"epoch": 0.39802190540426147,
"grad_norm": 0.20859771966934204,
"learning_rate": 0.001,
"loss": 3.1107061767578124,
"step": 45550
},
{
"epoch": 0.39845881199636274,
"grad_norm": 0.20409439504146576,
"learning_rate": 0.001,
"loss": 3.1071646118164065,
"step": 45600
},
{
"epoch": 0.398895718588464,
"grad_norm": 0.18957524001598358,
"learning_rate": 0.001,
"loss": 3.1135198974609377,
"step": 45650
},
{
"epoch": 0.3993326251805653,
"grad_norm": 0.20108622312545776,
"learning_rate": 0.001,
"loss": 3.11476806640625,
"step": 45700
},
{
"epoch": 0.39976953177266655,
"grad_norm": 0.18374258279800415,
"learning_rate": 0.001,
"loss": 3.115598449707031,
"step": 45750
},
{
"epoch": 0.4002064383647678,
"grad_norm": 0.235465407371521,
"learning_rate": 0.001,
"loss": 3.1099920654296875,
"step": 45800
},
{
"epoch": 0.40064334495686915,
"grad_norm": 0.19902831315994263,
"learning_rate": 0.001,
"loss": 3.106435546875,
"step": 45850
},
{
"epoch": 0.4010802515489704,
"grad_norm": 0.2039678692817688,
"learning_rate": 0.001,
"loss": 3.11017578125,
"step": 45900
},
{
"epoch": 0.4015171581410717,
"grad_norm": 0.19842007756233215,
"learning_rate": 0.001,
"loss": 3.1094546508789063,
"step": 45950
},
{
"epoch": 0.40195406473317297,
"grad_norm": 0.23383654654026031,
"learning_rate": 0.001,
"loss": 3.1192886352539064,
"step": 46000
},
{
"epoch": 0.40239097132527424,
"grad_norm": 0.2018861025571823,
"learning_rate": 0.001,
"loss": 3.106986389160156,
"step": 46050
},
{
"epoch": 0.4028278779173755,
"grad_norm": 0.2274814248085022,
"learning_rate": 0.001,
"loss": 3.1122772216796877,
"step": 46100
},
{
"epoch": 0.4032647845094768,
"grad_norm": 0.22933192551136017,
"learning_rate": 0.001,
"loss": 3.1104696655273436,
"step": 46150
},
{
"epoch": 0.40370169110157805,
"grad_norm": 0.2070489227771759,
"learning_rate": 0.001,
"loss": 3.10775634765625,
"step": 46200
},
{
"epoch": 0.4041385976936793,
"grad_norm": 0.18977881968021393,
"learning_rate": 0.001,
"loss": 3.1118939208984373,
"step": 46250
},
{
"epoch": 0.4045755042857806,
"grad_norm": 0.2149508148431778,
"learning_rate": 0.001,
"loss": 3.1053485107421874,
"step": 46300
},
{
"epoch": 0.40501241087788187,
"grad_norm": 0.1903674602508545,
"learning_rate": 0.001,
"loss": 3.1119683837890624,
"step": 46350
},
{
"epoch": 0.40544931746998314,
"grad_norm": 0.206802636384964,
"learning_rate": 0.001,
"loss": 3.1122677612304686,
"step": 46400
},
{
"epoch": 0.4058862240620844,
"grad_norm": 0.202114537358284,
"learning_rate": 0.001,
"loss": 3.110292663574219,
"step": 46450
},
{
"epoch": 0.4063231306541857,
"grad_norm": 0.19268907606601715,
"learning_rate": 0.001,
"loss": 3.1127801513671876,
"step": 46500
},
{
"epoch": 0.40676003724628695,
"grad_norm": 0.18836362659931183,
"learning_rate": 0.001,
"loss": 3.11020751953125,
"step": 46550
},
{
"epoch": 0.4071969438383882,
"grad_norm": 0.23016463220119476,
"learning_rate": 0.001,
"loss": 3.1144329833984377,
"step": 46600
},
{
"epoch": 0.40763385043048955,
"grad_norm": 0.22168609499931335,
"learning_rate": 0.001,
"loss": 3.11146484375,
"step": 46650
},
{
"epoch": 0.4080707570225908,
"grad_norm": 0.21283961832523346,
"learning_rate": 0.001,
"loss": 3.110495300292969,
"step": 46700
},
{
"epoch": 0.4085076636146921,
"grad_norm": 0.21619445085525513,
"learning_rate": 0.001,
"loss": 3.1094964599609374,
"step": 46750
},
{
"epoch": 0.40894457020679337,
"grad_norm": 0.19885626435279846,
"learning_rate": 0.001,
"loss": 3.1066973876953123,
"step": 46800
},
{
"epoch": 0.40938147679889464,
"grad_norm": 0.21709154546260834,
"learning_rate": 0.001,
"loss": 3.113372497558594,
"step": 46850
},
{
"epoch": 0.4098183833909959,
"grad_norm": 0.1988174319267273,
"learning_rate": 0.001,
"loss": 3.108941955566406,
"step": 46900
},
{
"epoch": 0.4102552899830972,
"grad_norm": 0.2376369833946228,
"learning_rate": 0.001,
"loss": 3.1140142822265626,
"step": 46950
},
{
"epoch": 0.41069219657519845,
"grad_norm": 0.20672737061977386,
"learning_rate": 0.001,
"loss": 3.110223693847656,
"step": 47000
},
{
"epoch": 0.4111291031672997,
"grad_norm": 0.1998968869447708,
"learning_rate": 0.001,
"loss": 3.1042745971679686,
"step": 47050
},
{
"epoch": 0.411566009759401,
"grad_norm": 0.21469420194625854,
"learning_rate": 0.001,
"loss": 3.1072637939453127,
"step": 47100
},
{
"epoch": 0.41200291635150227,
"grad_norm": 0.22460858523845673,
"learning_rate": 0.001,
"loss": 3.1115594482421876,
"step": 47150
},
{
"epoch": 0.41243982294360354,
"grad_norm": 0.21756698191165924,
"learning_rate": 0.001,
"loss": 3.1059274291992187,
"step": 47200
},
{
"epoch": 0.4128767295357048,
"grad_norm": 0.19789716601371765,
"learning_rate": 0.001,
"loss": 3.1101675415039063,
"step": 47250
},
{
"epoch": 0.4133136361278061,
"grad_norm": 0.19859153032302856,
"learning_rate": 0.001,
"loss": 3.108814697265625,
"step": 47300
},
{
"epoch": 0.41375054271990735,
"grad_norm": 0.1878359615802765,
"learning_rate": 0.001,
"loss": 3.1100103759765627,
"step": 47350
},
{
"epoch": 0.4141874493120086,
"grad_norm": 0.2045094221830368,
"learning_rate": 0.001,
"loss": 3.106664123535156,
"step": 47400
},
{
"epoch": 0.4146243559041099,
"grad_norm": 0.2144695371389389,
"learning_rate": 0.001,
"loss": 3.105127868652344,
"step": 47450
},
{
"epoch": 0.4150612624962112,
"grad_norm": 0.20183008909225464,
"learning_rate": 0.001,
"loss": 3.11333984375,
"step": 47500
},
{
"epoch": 0.4154981690883125,
"grad_norm": 0.19365358352661133,
"learning_rate": 0.001,
"loss": 3.111649169921875,
"step": 47550
},
{
"epoch": 0.41593507568041377,
"grad_norm": 0.19153247773647308,
"learning_rate": 0.001,
"loss": 3.1091769409179686,
"step": 47600
},
{
"epoch": 0.41637198227251504,
"grad_norm": 0.21135690808296204,
"learning_rate": 0.001,
"loss": 3.106083984375,
"step": 47650
},
{
"epoch": 0.4168088888646163,
"grad_norm": 0.2160654217004776,
"learning_rate": 0.001,
"loss": 3.1125274658203126,
"step": 47700
},
{
"epoch": 0.4172457954567176,
"grad_norm": 0.23028786480426788,
"learning_rate": 0.001,
"loss": 3.109255676269531,
"step": 47750
},
{
"epoch": 0.41768270204881885,
"grad_norm": 0.19282853603363037,
"learning_rate": 0.001,
"loss": 3.106436462402344,
"step": 47800
},
{
"epoch": 0.4181196086409201,
"grad_norm": 0.2034446895122528,
"learning_rate": 0.001,
"loss": 3.104494934082031,
"step": 47850
},
{
"epoch": 0.4185565152330214,
"grad_norm": 0.22074636816978455,
"learning_rate": 0.001,
"loss": 3.108076171875,
"step": 47900
},
{
"epoch": 0.41899342182512267,
"grad_norm": 0.2043931931257248,
"learning_rate": 0.001,
"loss": 3.106430969238281,
"step": 47950
},
{
"epoch": 0.41943032841722394,
"grad_norm": 0.1944197416305542,
"learning_rate": 0.001,
"loss": 3.105413513183594,
"step": 48000
},
{
"epoch": 0.41943032841722394,
"eval_loss": 3.117443561553955,
"eval_runtime": 5.5436,
"eval_samples_per_second": 176.057,
"eval_steps_per_second": 11.004,
"step": 48000
},
{
"epoch": 0.4198672350093252,
"grad_norm": 0.2274683564901352,
"learning_rate": 0.001,
"loss": 3.1015142822265624,
"step": 48050
},
{
"epoch": 0.4203041416014265,
"grad_norm": 0.21068282425403595,
"learning_rate": 0.001,
"loss": 3.1132937622070314,
"step": 48100
},
{
"epoch": 0.42074104819352776,
"grad_norm": 0.2048313319683075,
"learning_rate": 0.001,
"loss": 3.1093524169921873,
"step": 48150
},
{
"epoch": 0.421177954785629,
"grad_norm": 0.2029421031475067,
"learning_rate": 0.001,
"loss": 3.1043911743164063,
"step": 48200
},
{
"epoch": 0.4216148613777303,
"grad_norm": 0.2242981195449829,
"learning_rate": 0.001,
"loss": 3.106868896484375,
"step": 48250
},
{
"epoch": 0.4220517679698316,
"grad_norm": 0.2007884681224823,
"learning_rate": 0.001,
"loss": 3.1069216918945313,
"step": 48300
},
{
"epoch": 0.4224886745619329,
"grad_norm": 0.20348778367042542,
"learning_rate": 0.001,
"loss": 3.108001708984375,
"step": 48350
},
{
"epoch": 0.42292558115403417,
"grad_norm": 0.18854443728923798,
"learning_rate": 0.001,
"loss": 3.1086880493164064,
"step": 48400
},
{
"epoch": 0.42336248774613544,
"grad_norm": 0.19267909228801727,
"learning_rate": 0.001,
"loss": 3.1042105102539064,
"step": 48450
},
{
"epoch": 0.4237993943382367,
"grad_norm": 0.20865902304649353,
"learning_rate": 0.001,
"loss": 3.1016375732421877,
"step": 48500
},
{
"epoch": 0.424236300930338,
"grad_norm": 0.20658360421657562,
"learning_rate": 0.001,
"loss": 3.103634033203125,
"step": 48550
},
{
"epoch": 0.42467320752243926,
"grad_norm": 0.2347349226474762,
"learning_rate": 0.001,
"loss": 3.1078448486328125,
"step": 48600
},
{
"epoch": 0.4251101141145405,
"grad_norm": 0.21502061188220978,
"learning_rate": 0.001,
"loss": 3.0999810791015623,
"step": 48650
},
{
"epoch": 0.4255470207066418,
"grad_norm": 0.24060916900634766,
"learning_rate": 0.001,
"loss": 3.1052035522460937,
"step": 48700
},
{
"epoch": 0.42598392729874307,
"grad_norm": 0.19379866123199463,
"learning_rate": 0.001,
"loss": 3.1052151489257813,
"step": 48750
},
{
"epoch": 0.42642083389084434,
"grad_norm": 0.20438747107982635,
"learning_rate": 0.001,
"loss": 3.113894348144531,
"step": 48800
},
{
"epoch": 0.4268577404829456,
"grad_norm": 0.20153990387916565,
"learning_rate": 0.001,
"loss": 3.103798522949219,
"step": 48850
},
{
"epoch": 0.4272946470750469,
"grad_norm": 0.19886857271194458,
"learning_rate": 0.001,
"loss": 3.108213195800781,
"step": 48900
},
{
"epoch": 0.42773155366714816,
"grad_norm": 0.19363293051719666,
"learning_rate": 0.001,
"loss": 3.1007220458984377,
"step": 48950
},
{
"epoch": 0.42816846025924943,
"grad_norm": 0.209380641579628,
"learning_rate": 0.001,
"loss": 3.1065097045898438,
"step": 49000
},
{
"epoch": 0.4286053668513507,
"grad_norm": 0.21404701471328735,
"learning_rate": 0.001,
"loss": 3.1068161010742186,
"step": 49050
},
{
"epoch": 0.42904227344345197,
"grad_norm": 0.19078534841537476,
"learning_rate": 0.001,
"loss": 3.1129681396484377,
"step": 49100
},
{
"epoch": 0.4294791800355533,
"grad_norm": 0.20382896065711975,
"learning_rate": 0.001,
"loss": 3.1087237548828126,
"step": 49150
},
{
"epoch": 0.42991608662765457,
"grad_norm": 0.23335450887680054,
"learning_rate": 0.001,
"loss": 3.104140625,
"step": 49200
},
{
"epoch": 0.43035299321975584,
"grad_norm": 0.20891313254833221,
"learning_rate": 0.001,
"loss": 3.1034475708007814,
"step": 49250
},
{
"epoch": 0.4307898998118571,
"grad_norm": 0.1953035295009613,
"learning_rate": 0.001,
"loss": 3.111490478515625,
"step": 49300
},
{
"epoch": 0.4312268064039584,
"grad_norm": 0.23456130921840668,
"learning_rate": 0.001,
"loss": 3.104456787109375,
"step": 49350
},
{
"epoch": 0.43166371299605966,
"grad_norm": 0.19096161425113678,
"learning_rate": 0.001,
"loss": 3.1005657958984374,
"step": 49400
},
{
"epoch": 0.43210061958816093,
"grad_norm": 0.19206157326698303,
"learning_rate": 0.001,
"loss": 3.1070101928710936,
"step": 49450
},
{
"epoch": 0.4325375261802622,
"grad_norm": 0.23981696367263794,
"learning_rate": 0.001,
"loss": 3.1099197387695314,
"step": 49500
},
{
"epoch": 0.43297443277236347,
"grad_norm": 0.21515274047851562,
"learning_rate": 0.001,
"loss": 3.106592712402344,
"step": 49550
},
{
"epoch": 0.43341133936446474,
"grad_norm": 0.23269905149936676,
"learning_rate": 0.001,
"loss": 3.1037396240234374,
"step": 49600
},
{
"epoch": 0.433848245956566,
"grad_norm": 0.18905073404312134,
"learning_rate": 0.001,
"loss": 3.1058563232421874,
"step": 49650
},
{
"epoch": 0.4342851525486673,
"grad_norm": 0.18905054032802582,
"learning_rate": 0.001,
"loss": 3.105947265625,
"step": 49700
},
{
"epoch": 0.43472205914076856,
"grad_norm": 0.2018064558506012,
"learning_rate": 0.001,
"loss": 3.102055969238281,
"step": 49750
},
{
"epoch": 0.43515896573286983,
"grad_norm": 0.2202407568693161,
"learning_rate": 0.001,
"loss": 3.1106491088867188,
"step": 49800
},
{
"epoch": 0.4355958723249711,
"grad_norm": 0.23667579889297485,
"learning_rate": 0.001,
"loss": 3.1040682983398438,
"step": 49850
},
{
"epoch": 0.43603277891707237,
"grad_norm": 0.2142457216978073,
"learning_rate": 0.001,
"loss": 3.1000152587890626,
"step": 49900
},
{
"epoch": 0.4364696855091737,
"grad_norm": 0.21064357459545135,
"learning_rate": 0.001,
"loss": 3.1074310302734376,
"step": 49950
},
{
"epoch": 0.43690659210127497,
"grad_norm": 0.21424436569213867,
"learning_rate": 0.001,
"loss": 3.106015319824219,
"step": 50000
},
{
"epoch": 0.43734349869337624,
"grad_norm": 0.18170633912086487,
"learning_rate": 0.001,
"loss": 3.1060635375976564,
"step": 50050
},
{
"epoch": 0.4377804052854775,
"grad_norm": 0.20768190920352936,
"learning_rate": 0.001,
"loss": 3.10955078125,
"step": 50100
},
{
"epoch": 0.4382173118775788,
"grad_norm": 0.22088074684143066,
"learning_rate": 0.001,
"loss": 3.110235290527344,
"step": 50150
},
{
"epoch": 0.43865421846968006,
"grad_norm": 0.18902453780174255,
"learning_rate": 0.001,
"loss": 3.10544677734375,
"step": 50200
},
{
"epoch": 0.43909112506178133,
"grad_norm": 0.20212435722351074,
"learning_rate": 0.001,
"loss": 3.102640380859375,
"step": 50250
},
{
"epoch": 0.4395280316538826,
"grad_norm": 0.21360079944133759,
"learning_rate": 0.001,
"loss": 3.0996890258789063,
"step": 50300
},
{
"epoch": 0.43996493824598387,
"grad_norm": 0.19784964621067047,
"learning_rate": 0.001,
"loss": 3.1004718017578123,
"step": 50350
},
{
"epoch": 0.44040184483808514,
"grad_norm": 0.19656217098236084,
"learning_rate": 0.001,
"loss": 3.1105706787109373,
"step": 50400
},
{
"epoch": 0.4408387514301864,
"grad_norm": 0.22763794660568237,
"learning_rate": 0.001,
"loss": 3.10812255859375,
"step": 50450
},
{
"epoch": 0.4412756580222877,
"grad_norm": 0.19452765583992004,
"learning_rate": 0.001,
"loss": 3.102926025390625,
"step": 50500
},
{
"epoch": 0.44171256461438896,
"grad_norm": 0.23871538043022156,
"learning_rate": 0.001,
"loss": 3.1065164184570313,
"step": 50550
},
{
"epoch": 0.44214947120649023,
"grad_norm": 0.21007570624351501,
"learning_rate": 0.001,
"loss": 3.10342041015625,
"step": 50600
},
{
"epoch": 0.4425863777985915,
"grad_norm": 0.19869360327720642,
"learning_rate": 0.001,
"loss": 3.1000320434570314,
"step": 50650
},
{
"epoch": 0.4430232843906928,
"grad_norm": 0.18830987811088562,
"learning_rate": 0.001,
"loss": 3.100297546386719,
"step": 50700
},
{
"epoch": 0.44346019098279404,
"grad_norm": 0.2078581154346466,
"learning_rate": 0.001,
"loss": 3.1070379638671874,
"step": 50750
},
{
"epoch": 0.44389709757489537,
"grad_norm": 0.19879338145256042,
"learning_rate": 0.001,
"loss": 3.1094097900390625,
"step": 50800
},
{
"epoch": 0.44433400416699664,
"grad_norm": 0.20449118316173553,
"learning_rate": 0.001,
"loss": 3.107032470703125,
"step": 50850
},
{
"epoch": 0.4447709107590979,
"grad_norm": 0.22050796449184418,
"learning_rate": 0.001,
"loss": 3.103486328125,
"step": 50900
},
{
"epoch": 0.4452078173511992,
"grad_norm": 0.19746118783950806,
"learning_rate": 0.001,
"loss": 3.1099722290039065,
"step": 50950
},
{
"epoch": 0.44564472394330046,
"grad_norm": 0.21872414648532867,
"learning_rate": 0.001,
"loss": 3.109808044433594,
"step": 51000
},
{
"epoch": 0.44564472394330046,
"eval_loss": 3.1145269870758057,
"eval_runtime": 6.7323,
"eval_samples_per_second": 144.973,
"eval_steps_per_second": 9.061,
"step": 51000
},
{
"epoch": 0.44608163053540173,
"grad_norm": 0.2116106152534485,
"learning_rate": 0.001,
"loss": 3.0972409057617187,
"step": 51050
},
{
"epoch": 0.446518537127503,
"grad_norm": 0.1994769126176834,
"learning_rate": 0.001,
"loss": 3.10564208984375,
"step": 51100
},
{
"epoch": 0.4469554437196043,
"grad_norm": 0.21499839425086975,
"learning_rate": 0.001,
"loss": 3.1062249755859375,
"step": 51150
},
{
"epoch": 0.44739235031170554,
"grad_norm": 0.20958735048770905,
"learning_rate": 0.001,
"loss": 3.1007257080078126,
"step": 51200
},
{
"epoch": 0.4478292569038068,
"grad_norm": 0.21780237555503845,
"learning_rate": 0.001,
"loss": 3.100411682128906,
"step": 51250
},
{
"epoch": 0.4482661634959081,
"grad_norm": 0.19220604002475739,
"learning_rate": 0.001,
"loss": 3.100971984863281,
"step": 51300
},
{
"epoch": 0.44870307008800936,
"grad_norm": 0.20440301299095154,
"learning_rate": 0.001,
"loss": 3.1032171630859375,
"step": 51350
},
{
"epoch": 0.44913997668011063,
"grad_norm": 0.2154955267906189,
"learning_rate": 0.001,
"loss": 3.1047589111328127,
"step": 51400
},
{
"epoch": 0.4495768832722119,
"grad_norm": 0.1911495178937912,
"learning_rate": 0.001,
"loss": 3.1060125732421877,
"step": 51450
},
{
"epoch": 0.4500137898643132,
"grad_norm": 0.20830364525318146,
"learning_rate": 0.001,
"loss": 3.1013253784179686,
"step": 51500
},
{
"epoch": 0.45045069645641445,
"grad_norm": 0.19092939794063568,
"learning_rate": 0.001,
"loss": 3.104780578613281,
"step": 51550
},
{
"epoch": 0.4508876030485158,
"grad_norm": 0.1889987587928772,
"learning_rate": 0.001,
"loss": 3.106145324707031,
"step": 51600
},
{
"epoch": 0.45132450964061704,
"grad_norm": 0.20764483511447906,
"learning_rate": 0.001,
"loss": 3.105035705566406,
"step": 51650
},
{
"epoch": 0.4517614162327183,
"grad_norm": 0.2057601511478424,
"learning_rate": 0.001,
"loss": 3.1114022827148435,
"step": 51700
},
{
"epoch": 0.4521983228248196,
"grad_norm": 0.20924074947834015,
"learning_rate": 0.001,
"loss": 3.107779541015625,
"step": 51750
},
{
"epoch": 0.45263522941692086,
"grad_norm": 0.2036653757095337,
"learning_rate": 0.001,
"loss": 3.110738525390625,
"step": 51800
},
{
"epoch": 0.45307213600902213,
"grad_norm": 0.19672690331935883,
"learning_rate": 0.001,
"loss": 3.1042144775390623,
"step": 51850
},
{
"epoch": 0.4535090426011234,
"grad_norm": 0.2197086066007614,
"learning_rate": 0.001,
"loss": 3.0994357299804687,
"step": 51900
},
{
"epoch": 0.4539459491932247,
"grad_norm": 0.20545099675655365,
"learning_rate": 0.001,
"loss": 3.10298095703125,
"step": 51950
},
{
"epoch": 0.45438285578532595,
"grad_norm": 0.20683728158473969,
"learning_rate": 0.001,
"loss": 3.1031430053710936,
"step": 52000
},
{
"epoch": 0.4548197623774272,
"grad_norm": 0.24851621687412262,
"learning_rate": 0.001,
"loss": 3.105254821777344,
"step": 52050
},
{
"epoch": 0.4552566689695285,
"grad_norm": 0.2017493098974228,
"learning_rate": 0.001,
"loss": 3.1050048828125,
"step": 52100
},
{
"epoch": 0.45569357556162976,
"grad_norm": 0.2170027494430542,
"learning_rate": 0.001,
"loss": 3.102432861328125,
"step": 52150
},
{
"epoch": 0.45613048215373103,
"grad_norm": 0.20583300292491913,
"learning_rate": 0.001,
"loss": 3.1081842041015624,
"step": 52200
},
{
"epoch": 0.4565673887458323,
"grad_norm": 0.20641958713531494,
"learning_rate": 0.001,
"loss": 3.1093170166015627,
"step": 52250
},
{
"epoch": 0.4570042953379336,
"grad_norm": 0.19930274784564972,
"learning_rate": 0.001,
"loss": 3.1027716064453124,
"step": 52300
},
{
"epoch": 0.45744120193003485,
"grad_norm": 0.20604479312896729,
"learning_rate": 0.001,
"loss": 3.09675537109375,
"step": 52350
},
{
"epoch": 0.4578781085221361,
"grad_norm": 0.1930854171514511,
"learning_rate": 0.001,
"loss": 3.0996658325195314,
"step": 52400
},
{
"epoch": 0.45831501511423745,
"grad_norm": 0.20069192349910736,
"learning_rate": 0.001,
"loss": 3.09954345703125,
"step": 52450
},
{
"epoch": 0.4587519217063387,
"grad_norm": 0.20842620730400085,
"learning_rate": 0.001,
"loss": 3.0996414184570313,
"step": 52500
},
{
"epoch": 0.45918882829844,
"grad_norm": 0.2344353348016739,
"learning_rate": 0.001,
"loss": 3.106051025390625,
"step": 52550
},
{
"epoch": 0.45962573489054126,
"grad_norm": 0.24192386865615845,
"learning_rate": 0.001,
"loss": 3.1109957885742188,
"step": 52600
},
{
"epoch": 0.46006264148264253,
"grad_norm": 0.19947467744350433,
"learning_rate": 0.001,
"loss": 3.1000485229492187,
"step": 52650
},
{
"epoch": 0.4604995480747438,
"grad_norm": 0.20350565016269684,
"learning_rate": 0.001,
"loss": 3.103514404296875,
"step": 52700
},
{
"epoch": 0.4609364546668451,
"grad_norm": 0.189019113779068,
"learning_rate": 0.001,
"loss": 3.1035525512695314,
"step": 52750
},
{
"epoch": 0.46137336125894635,
"grad_norm": 0.19230777025222778,
"learning_rate": 0.001,
"loss": 3.1025357055664062,
"step": 52800
},
{
"epoch": 0.4618102678510476,
"grad_norm": 0.20442283153533936,
"learning_rate": 0.001,
"loss": 3.099986572265625,
"step": 52850
},
{
"epoch": 0.4622471744431489,
"grad_norm": 0.20625442266464233,
"learning_rate": 0.001,
"loss": 3.096138610839844,
"step": 52900
},
{
"epoch": 0.46268408103525016,
"grad_norm": 0.2018747627735138,
"learning_rate": 0.001,
"loss": 3.101961975097656,
"step": 52950
},
{
"epoch": 0.46312098762735143,
"grad_norm": 0.23321203887462616,
"learning_rate": 0.001,
"loss": 3.100405578613281,
"step": 53000
},
{
"epoch": 0.4635578942194527,
"grad_norm": 0.1996694654226303,
"learning_rate": 0.001,
"loss": 3.1041162109375,
"step": 53050
},
{
"epoch": 0.463994800811554,
"grad_norm": 0.21219325065612793,
"learning_rate": 0.001,
"loss": 3.0952743530273437,
"step": 53100
},
{
"epoch": 0.46443170740365525,
"grad_norm": 0.19142311811447144,
"learning_rate": 0.001,
"loss": 3.1060162353515626,
"step": 53150
},
{
"epoch": 0.4648686139957565,
"grad_norm": 0.19829395413398743,
"learning_rate": 0.001,
"loss": 3.101197509765625,
"step": 53200
},
{
"epoch": 0.46530552058785785,
"grad_norm": 0.1917588859796524,
"learning_rate": 0.001,
"loss": 3.09935302734375,
"step": 53250
},
{
"epoch": 0.4657424271799591,
"grad_norm": 0.20187485218048096,
"learning_rate": 0.001,
"loss": 3.0969952392578124,
"step": 53300
},
{
"epoch": 0.4661793337720604,
"grad_norm": 0.21702250838279724,
"learning_rate": 0.001,
"loss": 3.1092071533203125,
"step": 53350
},
{
"epoch": 0.46661624036416166,
"grad_norm": 0.19403064250946045,
"learning_rate": 0.001,
"loss": 3.10686767578125,
"step": 53400
},
{
"epoch": 0.46705314695626293,
"grad_norm": 0.21579718589782715,
"learning_rate": 0.001,
"loss": 3.1040939331054687,
"step": 53450
},
{
"epoch": 0.4674900535483642,
"grad_norm": 0.19928251206874847,
"learning_rate": 0.001,
"loss": 3.0986978149414064,
"step": 53500
},
{
"epoch": 0.4679269601404655,
"grad_norm": 0.19755201041698456,
"learning_rate": 0.001,
"loss": 3.102489318847656,
"step": 53550
},
{
"epoch": 0.46836386673256675,
"grad_norm": 0.2118004411458969,
"learning_rate": 0.001,
"loss": 3.1006509399414064,
"step": 53600
},
{
"epoch": 0.468800773324668,
"grad_norm": 0.19246506690979004,
"learning_rate": 0.001,
"loss": 3.0940277099609377,
"step": 53650
},
{
"epoch": 0.4692376799167693,
"grad_norm": 0.18687696754932404,
"learning_rate": 0.001,
"loss": 3.106478576660156,
"step": 53700
},
{
"epoch": 0.46967458650887056,
"grad_norm": 0.20355890691280365,
"learning_rate": 0.001,
"loss": 3.098047180175781,
"step": 53750
},
{
"epoch": 0.47011149310097183,
"grad_norm": 0.20526985824108124,
"learning_rate": 0.001,
"loss": 3.1039959716796877,
"step": 53800
},
{
"epoch": 0.4705483996930731,
"grad_norm": 0.22559934854507446,
"learning_rate": 0.001,
"loss": 3.09285888671875,
"step": 53850
},
{
"epoch": 0.4709853062851744,
"grad_norm": 0.20821084082126617,
"learning_rate": 0.001,
"loss": 3.103165588378906,
"step": 53900
},
{
"epoch": 0.47142221287727565,
"grad_norm": 0.21997879445552826,
"learning_rate": 0.001,
"loss": 3.10185791015625,
"step": 53950
},
{
"epoch": 0.4718591194693769,
"grad_norm": 0.23838217556476593,
"learning_rate": 0.001,
"loss": 3.1012548828125,
"step": 54000
},
{
"epoch": 0.4718591194693769,
"eval_loss": 3.114837646484375,
"eval_runtime": 6.6274,
"eval_samples_per_second": 147.267,
"eval_steps_per_second": 9.204,
"step": 54000
},
{
"epoch": 0.4722960260614782,
"grad_norm": 0.2186373919248581,
"learning_rate": 0.001,
"loss": 3.1035476684570313,
"step": 54050
},
{
"epoch": 0.4727329326535795,
"grad_norm": 0.20394057035446167,
"learning_rate": 0.001,
"loss": 3.0960726928710938,
"step": 54100
},
{
"epoch": 0.4731698392456808,
"grad_norm": 0.19906754791736603,
"learning_rate": 0.001,
"loss": 3.0953274536132813,
"step": 54150
},
{
"epoch": 0.47360674583778206,
"grad_norm": 0.20089800655841827,
"learning_rate": 0.001,
"loss": 3.1064468383789063,
"step": 54200
},
{
"epoch": 0.47404365242988333,
"grad_norm": 0.19790150225162506,
"learning_rate": 0.001,
"loss": 3.096270751953125,
"step": 54250
},
{
"epoch": 0.4744805590219846,
"grad_norm": 0.2000056356191635,
"learning_rate": 0.001,
"loss": 3.1015286254882812,
"step": 54300
},
{
"epoch": 0.4749174656140859,
"grad_norm": 0.21339593827724457,
"learning_rate": 0.001,
"loss": 3.0957427978515626,
"step": 54350
},
{
"epoch": 0.47535437220618715,
"grad_norm": 0.21689261496067047,
"learning_rate": 0.001,
"loss": 3.0947918701171875,
"step": 54400
},
{
"epoch": 0.4757912787982884,
"grad_norm": 0.18498677015304565,
"learning_rate": 0.001,
"loss": 3.103760070800781,
"step": 54450
},
{
"epoch": 0.4762281853903897,
"grad_norm": 0.2074810266494751,
"learning_rate": 0.001,
"loss": 3.1025021362304686,
"step": 54500
},
{
"epoch": 0.47666509198249096,
"grad_norm": 0.2032674103975296,
"learning_rate": 0.001,
"loss": 3.1042074584960937,
"step": 54550
},
{
"epoch": 0.47710199857459223,
"grad_norm": 0.22849521040916443,
"learning_rate": 0.001,
"loss": 3.1026931762695313,
"step": 54600
},
{
"epoch": 0.4775389051666935,
"grad_norm": 0.22741909325122833,
"learning_rate": 0.001,
"loss": 3.0991290283203123,
"step": 54650
},
{
"epoch": 0.4779758117587948,
"grad_norm": 0.1985906958580017,
"learning_rate": 0.001,
"loss": 3.1045431518554687,
"step": 54700
},
{
"epoch": 0.47841271835089605,
"grad_norm": 0.20626330375671387,
"learning_rate": 0.001,
"loss": 3.1019976806640623,
"step": 54750
},
{
"epoch": 0.4788496249429973,
"grad_norm": 0.23172344267368317,
"learning_rate": 0.001,
"loss": 3.1045050048828124,
"step": 54800
},
{
"epoch": 0.4792865315350986,
"grad_norm": 0.20122891664505005,
"learning_rate": 0.001,
"loss": 3.105765380859375,
"step": 54850
},
{
"epoch": 0.4797234381271999,
"grad_norm": 0.21384865045547485,
"learning_rate": 0.001,
"loss": 3.103521423339844,
"step": 54900
},
{
"epoch": 0.4801603447193012,
"grad_norm": 0.21676206588745117,
"learning_rate": 0.001,
"loss": 3.1025885009765624,
"step": 54950
},
{
"epoch": 0.48059725131140246,
"grad_norm": 0.20777349174022675,
"learning_rate": 0.001,
"loss": 3.09767333984375,
"step": 55000
},
{
"epoch": 0.48103415790350373,
"grad_norm": 0.20779912173748016,
"learning_rate": 0.001,
"loss": 3.097218017578125,
"step": 55050
},
{
"epoch": 0.481471064495605,
"grad_norm": 0.21260333061218262,
"learning_rate": 0.001,
"loss": 3.0979034423828127,
"step": 55100
},
{
"epoch": 0.4819079710877063,
"grad_norm": 0.2428203672170639,
"learning_rate": 0.001,
"loss": 3.0946414184570314,
"step": 55150
},
{
"epoch": 0.48234487767980755,
"grad_norm": 0.19761182367801666,
"learning_rate": 0.001,
"loss": 3.1069851684570313,
"step": 55200
},
{
"epoch": 0.4827817842719088,
"grad_norm": 0.20102514326572418,
"learning_rate": 0.001,
"loss": 3.1008648681640625,
"step": 55250
},
{
"epoch": 0.4832186908640101,
"grad_norm": 0.21611596643924713,
"learning_rate": 0.001,
"loss": 3.0994418334960936,
"step": 55300
},
{
"epoch": 0.48365559745611136,
"grad_norm": 0.225076362490654,
"learning_rate": 0.001,
"loss": 3.1004150390625,
"step": 55350
},
{
"epoch": 0.48409250404821264,
"grad_norm": 0.21668097376823425,
"learning_rate": 0.001,
"loss": 3.1017755126953124,
"step": 55400
},
{
"epoch": 0.4845294106403139,
"grad_norm": 0.21668079495429993,
"learning_rate": 0.001,
"loss": 3.10286865234375,
"step": 55450
},
{
"epoch": 0.4849663172324152,
"grad_norm": 0.20780622959136963,
"learning_rate": 0.001,
"loss": 3.1003192138671873,
"step": 55500
},
{
"epoch": 0.48540322382451645,
"grad_norm": 0.2046990692615509,
"learning_rate": 0.001,
"loss": 3.095623779296875,
"step": 55550
},
{
"epoch": 0.4858401304166177,
"grad_norm": 0.21325261890888214,
"learning_rate": 0.001,
"loss": 3.098072814941406,
"step": 55600
},
{
"epoch": 0.486277037008719,
"grad_norm": 0.20987661182880402,
"learning_rate": 0.001,
"loss": 3.097833557128906,
"step": 55650
},
{
"epoch": 0.48671394360082026,
"grad_norm": 0.20375686883926392,
"learning_rate": 0.001,
"loss": 3.0966152954101562,
"step": 55700
},
{
"epoch": 0.4871508501929216,
"grad_norm": 0.21019193530082703,
"learning_rate": 0.001,
"loss": 3.097320251464844,
"step": 55750
},
{
"epoch": 0.48758775678502286,
"grad_norm": 0.2124178260564804,
"learning_rate": 0.001,
"loss": 3.1000653076171876,
"step": 55800
},
{
"epoch": 0.48802466337712413,
"grad_norm": 0.20298762619495392,
"learning_rate": 0.001,
"loss": 3.10153564453125,
"step": 55850
},
{
"epoch": 0.4884615699692254,
"grad_norm": 0.20471744239330292,
"learning_rate": 0.001,
"loss": 3.0964224243164065,
"step": 55900
},
{
"epoch": 0.4888984765613267,
"grad_norm": 0.20811215043067932,
"learning_rate": 0.001,
"loss": 3.103257141113281,
"step": 55950
},
{
"epoch": 0.48933538315342795,
"grad_norm": 0.18320778012275696,
"learning_rate": 0.001,
"loss": 3.1000506591796877,
"step": 56000
},
{
"epoch": 0.4897722897455292,
"grad_norm": 0.21062225103378296,
"learning_rate": 0.001,
"loss": 3.099864196777344,
"step": 56050
},
{
"epoch": 0.4902091963376305,
"grad_norm": 0.22070066630840302,
"learning_rate": 0.001,
"loss": 3.0956866455078127,
"step": 56100
},
{
"epoch": 0.49064610292973176,
"grad_norm": 0.20703208446502686,
"learning_rate": 0.001,
"loss": 3.097848205566406,
"step": 56150
},
{
"epoch": 0.49108300952183304,
"grad_norm": 0.20900507271289825,
"learning_rate": 0.001,
"loss": 3.0981231689453126,
"step": 56200
},
{
"epoch": 0.4915199161139343,
"grad_norm": 0.19017143547534943,
"learning_rate": 0.001,
"loss": 3.10042724609375,
"step": 56250
},
{
"epoch": 0.4919568227060356,
"grad_norm": 0.2100110501050949,
"learning_rate": 0.001,
"loss": 3.1031845092773436,
"step": 56300
},
{
"epoch": 0.49239372929813685,
"grad_norm": 0.199602872133255,
"learning_rate": 0.001,
"loss": 3.0943890380859376,
"step": 56350
},
{
"epoch": 0.4928306358902381,
"grad_norm": 0.21281187236309052,
"learning_rate": 0.001,
"loss": 3.101715393066406,
"step": 56400
},
{
"epoch": 0.4932675424823394,
"grad_norm": 0.2238076776266098,
"learning_rate": 0.001,
"loss": 3.098333435058594,
"step": 56450
},
{
"epoch": 0.49370444907444067,
"grad_norm": 0.18522411584854126,
"learning_rate": 0.001,
"loss": 3.099473876953125,
"step": 56500
},
{
"epoch": 0.494141355666542,
"grad_norm": 0.20802699029445648,
"learning_rate": 0.001,
"loss": 3.0976776123046874,
"step": 56550
},
{
"epoch": 0.49457826225864326,
"grad_norm": 0.19908156991004944,
"learning_rate": 0.001,
"loss": 3.0991046142578127,
"step": 56600
},
{
"epoch": 0.49501516885074454,
"grad_norm": 0.23973093926906586,
"learning_rate": 0.001,
"loss": 3.09651611328125,
"step": 56650
},
{
"epoch": 0.4954520754428458,
"grad_norm": 0.20567065477371216,
"learning_rate": 0.001,
"loss": 3.0990509033203124,
"step": 56700
},
{
"epoch": 0.4958889820349471,
"grad_norm": 0.19983196258544922,
"learning_rate": 0.001,
"loss": 3.1039718627929687,
"step": 56750
},
{
"epoch": 0.49632588862704835,
"grad_norm": 0.1869107484817505,
"learning_rate": 0.001,
"loss": 3.095478820800781,
"step": 56800
},
{
"epoch": 0.4967627952191496,
"grad_norm": 0.2134186327457428,
"learning_rate": 0.001,
"loss": 3.1055059814453125,
"step": 56850
},
{
"epoch": 0.4971997018112509,
"grad_norm": 0.20569279789924622,
"learning_rate": 0.001,
"loss": 3.1002142333984377,
"step": 56900
},
{
"epoch": 0.49763660840335217,
"grad_norm": 0.2199387401342392,
"learning_rate": 0.001,
"loss": 3.1022509765625,
"step": 56950
},
{
"epoch": 0.49807351499545344,
"grad_norm": 0.18851952254772186,
"learning_rate": 0.001,
"loss": 3.0978851318359375,
"step": 57000
},
{
"epoch": 0.49807351499545344,
"eval_loss": 3.111524820327759,
"eval_runtime": 6.6797,
"eval_samples_per_second": 146.115,
"eval_steps_per_second": 9.132,
"step": 57000
},
{
"epoch": 0.4985104215875547,
"grad_norm": 0.20856814086437225,
"learning_rate": 0.001,
"loss": 3.101390380859375,
"step": 57050
},
{
"epoch": 0.498947328179656,
"grad_norm": 0.20059722661972046,
"learning_rate": 0.001,
"loss": 3.0984121704101564,
"step": 57100
},
{
"epoch": 0.49938423477175725,
"grad_norm": 0.23056519031524658,
"learning_rate": 0.001,
"loss": 3.10336669921875,
"step": 57150
},
{
"epoch": 0.4998211413638585,
"grad_norm": 0.2020101249217987,
"learning_rate": 0.001,
"loss": 3.098371887207031,
"step": 57200
},
{
"epoch": 0.5002580479559599,
"grad_norm": 0.21829089522361755,
"learning_rate": 0.001,
"loss": 3.0959234619140625,
"step": 57250
},
{
"epoch": 0.5006949545480611,
"grad_norm": 0.20384812355041504,
"learning_rate": 0.001,
"loss": 3.0959716796875,
"step": 57300
},
{
"epoch": 0.5011318611401624,
"grad_norm": 0.1988534778356552,
"learning_rate": 0.001,
"loss": 3.1031439208984377,
"step": 57350
},
{
"epoch": 0.5015687677322637,
"grad_norm": 0.2948338985443115,
"learning_rate": 0.001,
"loss": 3.1001226806640627,
"step": 57400
},
{
"epoch": 0.5020056743243649,
"grad_norm": 0.2154141217470169,
"learning_rate": 0.001,
"loss": 3.10098388671875,
"step": 57450
},
{
"epoch": 0.5024425809164662,
"grad_norm": 0.20172695815563202,
"learning_rate": 0.001,
"loss": 3.0943447875976564,
"step": 57500
},
{
"epoch": 0.5028794875085675,
"grad_norm": 0.1951432079076767,
"learning_rate": 0.001,
"loss": 3.095655822753906,
"step": 57550
},
{
"epoch": 0.5033163941006688,
"grad_norm": 0.21197514235973358,
"learning_rate": 0.001,
"loss": 3.0945904541015623,
"step": 57600
},
{
"epoch": 0.50375330069277,
"grad_norm": 0.20073898136615753,
"learning_rate": 0.001,
"loss": 3.105246276855469,
"step": 57650
},
{
"epoch": 0.5041902072848713,
"grad_norm": 0.23221160471439362,
"learning_rate": 0.001,
"loss": 3.0944268798828123,
"step": 57700
},
{
"epoch": 0.5046271138769726,
"grad_norm": 0.2109694480895996,
"learning_rate": 0.001,
"loss": 3.09965576171875,
"step": 57750
},
{
"epoch": 0.5050640204690738,
"grad_norm": 0.20435838401317596,
"learning_rate": 0.001,
"loss": 3.100567626953125,
"step": 57800
},
{
"epoch": 0.5055009270611751,
"grad_norm": 0.2099217176437378,
"learning_rate": 0.001,
"loss": 3.1028372192382814,
"step": 57850
},
{
"epoch": 0.5059378336532764,
"grad_norm": 0.20463615655899048,
"learning_rate": 0.001,
"loss": 3.0954830932617186,
"step": 57900
},
{
"epoch": 0.5063747402453777,
"grad_norm": 0.2138853818178177,
"learning_rate": 0.001,
"loss": 3.10104736328125,
"step": 57950
},
{
"epoch": 0.5068116468374789,
"grad_norm": 0.21230050921440125,
"learning_rate": 0.001,
"loss": 3.1030227661132814,
"step": 58000
},
{
"epoch": 0.5072485534295802,
"grad_norm": 0.21621040999889374,
"learning_rate": 0.001,
"loss": 3.0994677734375,
"step": 58050
},
{
"epoch": 0.5076854600216815,
"grad_norm": 0.1944791078567505,
"learning_rate": 0.001,
"loss": 3.093312072753906,
"step": 58100
},
{
"epoch": 0.5081223666137827,
"grad_norm": 0.19555747509002686,
"learning_rate": 0.001,
"loss": 3.099140319824219,
"step": 58150
},
{
"epoch": 0.508559273205884,
"grad_norm": 0.2322762906551361,
"learning_rate": 0.001,
"loss": 3.0927813720703123,
"step": 58200
},
{
"epoch": 0.5089961797979853,
"grad_norm": 0.2077183872461319,
"learning_rate": 0.001,
"loss": 3.1027117919921876,
"step": 58250
},
{
"epoch": 0.5094330863900866,
"grad_norm": 0.2282315343618393,
"learning_rate": 0.001,
"loss": 3.093887939453125,
"step": 58300
},
{
"epoch": 0.5098699929821878,
"grad_norm": 0.19174745678901672,
"learning_rate": 0.001,
"loss": 3.1009097290039063,
"step": 58350
},
{
"epoch": 0.5103068995742891,
"grad_norm": 0.23863931000232697,
"learning_rate": 0.001,
"loss": 3.0973739624023438,
"step": 58400
},
{
"epoch": 0.5107438061663904,
"grad_norm": 0.2001272737979889,
"learning_rate": 0.001,
"loss": 3.0964923095703125,
"step": 58450
},
{
"epoch": 0.5111807127584916,
"grad_norm": 0.21236871182918549,
"learning_rate": 0.001,
"loss": 3.09669189453125,
"step": 58500
},
{
"epoch": 0.5116176193505929,
"grad_norm": 0.20710663497447968,
"learning_rate": 0.001,
"loss": 3.10250244140625,
"step": 58550
},
{
"epoch": 0.5120545259426943,
"grad_norm": 0.2294129580259323,
"learning_rate": 0.001,
"loss": 3.1013858032226564,
"step": 58600
},
{
"epoch": 0.5124914325347956,
"grad_norm": 0.2197817862033844,
"learning_rate": 0.001,
"loss": 3.098335876464844,
"step": 58650
},
{
"epoch": 0.5129283391268968,
"grad_norm": 0.20704595744609833,
"learning_rate": 0.001,
"loss": 3.098610534667969,
"step": 58700
},
{
"epoch": 0.5133652457189981,
"grad_norm": 0.19215127825737,
"learning_rate": 0.001,
"loss": 3.092769775390625,
"step": 58750
},
{
"epoch": 0.5138021523110994,
"grad_norm": 0.1911986768245697,
"learning_rate": 0.001,
"loss": 3.0982513427734375,
"step": 58800
},
{
"epoch": 0.5142390589032007,
"grad_norm": 0.21684405207633972,
"learning_rate": 0.001,
"loss": 3.099425048828125,
"step": 58850
},
{
"epoch": 0.5146759654953019,
"grad_norm": 0.21217948198318481,
"learning_rate": 0.001,
"loss": 3.0997525024414063,
"step": 58900
},
{
"epoch": 0.5151128720874032,
"grad_norm": 0.20972223579883575,
"learning_rate": 0.001,
"loss": 3.100913391113281,
"step": 58950
},
{
"epoch": 0.5155497786795045,
"grad_norm": 0.18779510259628296,
"learning_rate": 0.001,
"loss": 3.1015945434570313,
"step": 59000
},
{
"epoch": 0.5159866852716057,
"grad_norm": 0.21088241040706635,
"learning_rate": 0.001,
"loss": 3.09939697265625,
"step": 59050
},
{
"epoch": 0.516423591863707,
"grad_norm": 0.22213022410869598,
"learning_rate": 0.001,
"loss": 3.098623962402344,
"step": 59100
},
{
"epoch": 0.5168604984558083,
"grad_norm": 0.22304199635982513,
"learning_rate": 0.001,
"loss": 3.0950753784179685,
"step": 59150
},
{
"epoch": 0.5172974050479096,
"grad_norm": 0.22656190395355225,
"learning_rate": 0.001,
"loss": 3.098634033203125,
"step": 59200
},
{
"epoch": 0.5177343116400108,
"grad_norm": 0.19612087309360504,
"learning_rate": 0.001,
"loss": 3.0953594970703127,
"step": 59250
},
{
"epoch": 0.5181712182321121,
"grad_norm": 0.19617019593715668,
"learning_rate": 0.001,
"loss": 3.101268005371094,
"step": 59300
},
{
"epoch": 0.5186081248242134,
"grad_norm": 0.19519132375717163,
"learning_rate": 0.001,
"loss": 3.0935598754882814,
"step": 59350
},
{
"epoch": 0.5190450314163146,
"grad_norm": 0.2071872502565384,
"learning_rate": 0.001,
"loss": 3.0997381591796875,
"step": 59400
},
{
"epoch": 0.5194819380084159,
"grad_norm": 0.20153042674064636,
"learning_rate": 0.001,
"loss": 3.0961618041992187,
"step": 59450
},
{
"epoch": 0.5199188446005172,
"grad_norm": 0.2103044092655182,
"learning_rate": 0.001,
"loss": 3.0931768798828125,
"step": 59500
},
{
"epoch": 0.5203557511926185,
"grad_norm": 0.20313984155654907,
"learning_rate": 0.001,
"loss": 3.0989199829101564,
"step": 59550
},
{
"epoch": 0.5207926577847197,
"grad_norm": 0.19373391568660736,
"learning_rate": 0.001,
"loss": 3.0947311401367186,
"step": 59600
},
{
"epoch": 0.521229564376821,
"grad_norm": 0.20085763931274414,
"learning_rate": 0.001,
"loss": 3.09599365234375,
"step": 59650
},
{
"epoch": 0.5216664709689223,
"grad_norm": 0.2221629023551941,
"learning_rate": 0.001,
"loss": 3.0930169677734374,
"step": 59700
},
{
"epoch": 0.5221033775610235,
"grad_norm": 0.20920270681381226,
"learning_rate": 0.001,
"loss": 3.0992108154296876,
"step": 59750
},
{
"epoch": 0.5225402841531248,
"grad_norm": 0.21595247089862823,
"learning_rate": 0.001,
"loss": 3.0957595825195314,
"step": 59800
},
{
"epoch": 0.5229771907452261,
"grad_norm": 0.20243686437606812,
"learning_rate": 0.001,
"loss": 3.0995834350585936,
"step": 59850
},
{
"epoch": 0.5234140973373274,
"grad_norm": 0.21057631075382233,
"learning_rate": 0.001,
"loss": 3.0983642578125,
"step": 59900
},
{
"epoch": 0.5238510039294286,
"grad_norm": 0.19537286460399628,
"learning_rate": 0.001,
"loss": 3.0940924072265625,
"step": 59950
},
{
"epoch": 0.5242879105215299,
"grad_norm": 0.20319755375385284,
"learning_rate": 0.001,
"loss": 3.099693603515625,
"step": 60000
},
{
"epoch": 0.5242879105215299,
"eval_loss": 3.109640121459961,
"eval_runtime": 6.6253,
"eval_samples_per_second": 147.314,
"eval_steps_per_second": 9.207,
"step": 60000
},
{
"epoch": 0.5247248171136312,
"grad_norm": 0.2013198882341385,
"learning_rate": 0.001,
"loss": 3.0987652587890624,
"step": 60050
},
{
"epoch": 0.5251617237057324,
"grad_norm": 0.20471902191638947,
"learning_rate": 0.001,
"loss": 3.0976702880859377,
"step": 60100
},
{
"epoch": 0.5255986302978337,
"grad_norm": 0.2103605568408966,
"learning_rate": 0.001,
"loss": 3.09927734375,
"step": 60150
},
{
"epoch": 0.526035536889935,
"grad_norm": 0.19664128124713898,
"learning_rate": 0.001,
"loss": 3.0999896240234377,
"step": 60200
},
{
"epoch": 0.5264724434820364,
"grad_norm": 0.20885683596134186,
"learning_rate": 0.001,
"loss": 3.097658386230469,
"step": 60250
},
{
"epoch": 0.5269093500741376,
"grad_norm": 0.20266753435134888,
"learning_rate": 0.001,
"loss": 3.097698059082031,
"step": 60300
},
{
"epoch": 0.5273462566662389,
"grad_norm": 0.19238939881324768,
"learning_rate": 0.001,
"loss": 3.0966748046875,
"step": 60350
},
{
"epoch": 0.5277831632583402,
"grad_norm": 0.20401431620121002,
"learning_rate": 0.001,
"loss": 3.0938726806640626,
"step": 60400
},
{
"epoch": 0.5282200698504415,
"grad_norm": 0.19009321928024292,
"learning_rate": 0.001,
"loss": 3.0921676635742186,
"step": 60450
},
{
"epoch": 0.5286569764425427,
"grad_norm": 0.20416900515556335,
"learning_rate": 0.001,
"loss": 3.0846258544921876,
"step": 60500
},
{
"epoch": 0.529093883034644,
"grad_norm": 0.21365386247634888,
"learning_rate": 0.001,
"loss": 3.0948336791992186,
"step": 60550
},
{
"epoch": 0.5295307896267453,
"grad_norm": 0.1897747665643692,
"learning_rate": 0.001,
"loss": 3.0963815307617186,
"step": 60600
},
{
"epoch": 0.5299676962188465,
"grad_norm": 0.20467257499694824,
"learning_rate": 0.001,
"loss": 3.0999154663085937,
"step": 60650
},
{
"epoch": 0.5304046028109478,
"grad_norm": 0.18985669314861298,
"learning_rate": 0.001,
"loss": 3.096724548339844,
"step": 60700
},
{
"epoch": 0.5308415094030491,
"grad_norm": 0.2125224471092224,
"learning_rate": 0.001,
"loss": 3.097410888671875,
"step": 60750
},
{
"epoch": 0.5312784159951504,
"grad_norm": 0.21208299696445465,
"learning_rate": 0.001,
"loss": 3.10163818359375,
"step": 60800
},
{
"epoch": 0.5317153225872516,
"grad_norm": 0.19807051122188568,
"learning_rate": 0.001,
"loss": 3.0963229370117187,
"step": 60850
},
{
"epoch": 0.5321522291793529,
"grad_norm": 0.1940542757511139,
"learning_rate": 0.001,
"loss": 3.0924533081054686,
"step": 60900
},
{
"epoch": 0.5325891357714542,
"grad_norm": 0.2073223888874054,
"learning_rate": 0.001,
"loss": 3.0993002319335936,
"step": 60950
},
{
"epoch": 0.5330260423635554,
"grad_norm": 0.2114952653646469,
"learning_rate": 0.001,
"loss": 3.0935577392578124,
"step": 61000
},
{
"epoch": 0.5334629489556567,
"grad_norm": 0.20891067385673523,
"learning_rate": 0.001,
"loss": 3.0988369750976563,
"step": 61050
},
{
"epoch": 0.533899855547758,
"grad_norm": 0.19985494017601013,
"learning_rate": 0.001,
"loss": 3.1023428344726565,
"step": 61100
},
{
"epoch": 0.5343367621398593,
"grad_norm": 0.2000042200088501,
"learning_rate": 0.001,
"loss": 3.0963589477539064,
"step": 61150
},
{
"epoch": 0.5347736687319605,
"grad_norm": 0.22281034290790558,
"learning_rate": 0.001,
"loss": 3.096812438964844,
"step": 61200
},
{
"epoch": 0.5352105753240618,
"grad_norm": 0.18756063282489777,
"learning_rate": 0.001,
"loss": 3.0939346313476563,
"step": 61250
},
{
"epoch": 0.5356474819161631,
"grad_norm": 0.1990208625793457,
"learning_rate": 0.001,
"loss": 3.0929550170898437,
"step": 61300
},
{
"epoch": 0.5360843885082643,
"grad_norm": 0.199998140335083,
"learning_rate": 0.001,
"loss": 3.09830810546875,
"step": 61350
},
{
"epoch": 0.5365212951003656,
"grad_norm": 0.22184330224990845,
"learning_rate": 0.001,
"loss": 3.1024102783203125,
"step": 61400
},
{
"epoch": 0.5369582016924669,
"grad_norm": 0.21995766460895538,
"learning_rate": 0.001,
"loss": 3.0950997924804686,
"step": 61450
},
{
"epoch": 0.5373951082845682,
"grad_norm": 0.21542876958847046,
"learning_rate": 0.001,
"loss": 3.099861755371094,
"step": 61500
},
{
"epoch": 0.5378320148766694,
"grad_norm": 0.2013842910528183,
"learning_rate": 0.001,
"loss": 3.09324951171875,
"step": 61550
},
{
"epoch": 0.5382689214687707,
"grad_norm": 0.21761702001094818,
"learning_rate": 0.001,
"loss": 3.099310302734375,
"step": 61600
},
{
"epoch": 0.538705828060872,
"grad_norm": 0.1965852677822113,
"learning_rate": 0.001,
"loss": 3.100264892578125,
"step": 61650
},
{
"epoch": 0.5391427346529732,
"grad_norm": 0.21733810007572174,
"learning_rate": 0.001,
"loss": 3.0921200561523436,
"step": 61700
},
{
"epoch": 0.5395796412450745,
"grad_norm": 0.20797546207904816,
"learning_rate": 0.001,
"loss": 3.09889404296875,
"step": 61750
},
{
"epoch": 0.5400165478371758,
"grad_norm": 0.21520453691482544,
"learning_rate": 0.001,
"loss": 3.0957980346679688,
"step": 61800
},
{
"epoch": 0.5404534544292771,
"grad_norm": 0.22498275339603424,
"learning_rate": 0.001,
"loss": 3.0924002075195314,
"step": 61850
},
{
"epoch": 0.5408903610213784,
"grad_norm": 0.21598829329013824,
"learning_rate": 0.001,
"loss": 3.0910488891601564,
"step": 61900
},
{
"epoch": 0.5413272676134797,
"grad_norm": 0.2177823781967163,
"learning_rate": 0.001,
"loss": 3.0904299926757814,
"step": 61950
},
{
"epoch": 0.541764174205581,
"grad_norm": 0.22696982324123383,
"learning_rate": 0.001,
"loss": 3.0921951293945313,
"step": 62000
},
{
"epoch": 0.5422010807976823,
"grad_norm": 0.209177166223526,
"learning_rate": 0.001,
"loss": 3.0979205322265626,
"step": 62050
},
{
"epoch": 0.5426379873897835,
"grad_norm": 0.22123073041439056,
"learning_rate": 0.001,
"loss": 3.0893319702148436,
"step": 62100
},
{
"epoch": 0.5430748939818848,
"grad_norm": 0.20946907997131348,
"learning_rate": 0.001,
"loss": 3.093170166015625,
"step": 62150
},
{
"epoch": 0.5435118005739861,
"grad_norm": 0.18812572956085205,
"learning_rate": 0.001,
"loss": 3.0903616333007813,
"step": 62200
},
{
"epoch": 0.5439487071660873,
"grad_norm": 0.20360390841960907,
"learning_rate": 0.001,
"loss": 3.0912567138671876,
"step": 62250
},
{
"epoch": 0.5443856137581886,
"grad_norm": 0.23311933875083923,
"learning_rate": 0.001,
"loss": 3.0972406005859376,
"step": 62300
},
{
"epoch": 0.5448225203502899,
"grad_norm": 0.20024672150611877,
"learning_rate": 0.001,
"loss": 3.0917813110351564,
"step": 62350
},
{
"epoch": 0.5452594269423912,
"grad_norm": 0.22117973864078522,
"learning_rate": 0.001,
"loss": 3.095211181640625,
"step": 62400
},
{
"epoch": 0.5456963335344924,
"grad_norm": 0.23353464901447296,
"learning_rate": 0.001,
"loss": 3.0964874267578124,
"step": 62450
},
{
"epoch": 0.5461332401265937,
"grad_norm": 0.20897679030895233,
"learning_rate": 0.001,
"loss": 3.093771667480469,
"step": 62500
},
{
"epoch": 0.546570146718695,
"grad_norm": 0.2139606922864914,
"learning_rate": 0.001,
"loss": 3.100740661621094,
"step": 62550
},
{
"epoch": 0.5470070533107962,
"grad_norm": 0.21718043088912964,
"learning_rate": 0.001,
"loss": 3.096734619140625,
"step": 62600
},
{
"epoch": 0.5474439599028975,
"grad_norm": 0.2036345899105072,
"learning_rate": 0.001,
"loss": 3.092578125,
"step": 62650
},
{
"epoch": 0.5478808664949988,
"grad_norm": 0.21168434619903564,
"learning_rate": 0.001,
"loss": 3.0923907470703127,
"step": 62700
},
{
"epoch": 0.5483177730871001,
"grad_norm": 0.2323308289051056,
"learning_rate": 0.001,
"loss": 3.092884521484375,
"step": 62750
},
{
"epoch": 0.5487546796792013,
"grad_norm": 0.22221334278583527,
"learning_rate": 0.001,
"loss": 3.0940200805664064,
"step": 62800
},
{
"epoch": 0.5491915862713026,
"grad_norm": 0.2459583431482315,
"learning_rate": 0.001,
"loss": 3.0891375732421875,
"step": 62850
},
{
"epoch": 0.5496284928634039,
"grad_norm": 0.20646947622299194,
"learning_rate": 0.001,
"loss": 3.0943463134765623,
"step": 62900
},
{
"epoch": 0.5500653994555051,
"grad_norm": 0.19100476801395416,
"learning_rate": 0.001,
"loss": 3.0980868530273438,
"step": 62950
},
{
"epoch": 0.5505023060476064,
"grad_norm": 0.20407864451408386,
"learning_rate": 0.001,
"loss": 3.0936489868164063,
"step": 63000
},
{
"epoch": 0.5505023060476064,
"eval_loss": 3.106282949447632,
"eval_runtime": 5.7603,
"eval_samples_per_second": 169.437,
"eval_steps_per_second": 10.59,
"step": 63000
},
{
"epoch": 0.5509392126397077,
"grad_norm": 0.2221687287092209,
"learning_rate": 0.001,
"loss": 3.0896145629882814,
"step": 63050
},
{
"epoch": 0.551376119231809,
"grad_norm": 0.21832983195781708,
"learning_rate": 0.001,
"loss": 3.0954666137695312,
"step": 63100
},
{
"epoch": 0.5518130258239102,
"grad_norm": 0.2022419273853302,
"learning_rate": 0.001,
"loss": 3.097255859375,
"step": 63150
},
{
"epoch": 0.5522499324160115,
"grad_norm": 0.22601716220378876,
"learning_rate": 0.001,
"loss": 3.094032897949219,
"step": 63200
},
{
"epoch": 0.5526868390081128,
"grad_norm": 0.20275865495204926,
"learning_rate": 0.001,
"loss": 3.095334777832031,
"step": 63250
},
{
"epoch": 0.553123745600214,
"grad_norm": 0.190517857670784,
"learning_rate": 0.001,
"loss": 3.094215087890625,
"step": 63300
},
{
"epoch": 0.5535606521923153,
"grad_norm": 0.20814111828804016,
"learning_rate": 0.001,
"loss": 3.0909783935546873,
"step": 63350
},
{
"epoch": 0.5539975587844166,
"grad_norm": 0.22653110325336456,
"learning_rate": 0.001,
"loss": 3.0945684814453127,
"step": 63400
},
{
"epoch": 0.5544344653765179,
"grad_norm": 0.25072765350341797,
"learning_rate": 0.001,
"loss": 3.101626281738281,
"step": 63450
},
{
"epoch": 0.5548713719686191,
"grad_norm": 0.20806042850017548,
"learning_rate": 0.001,
"loss": 3.0940213012695312,
"step": 63500
},
{
"epoch": 0.5553082785607205,
"grad_norm": 0.19432547688484192,
"learning_rate": 0.001,
"loss": 3.0937533569335938,
"step": 63550
},
{
"epoch": 0.5557451851528218,
"grad_norm": 0.22655421495437622,
"learning_rate": 0.001,
"loss": 3.0912106323242186,
"step": 63600
},
{
"epoch": 0.5561820917449231,
"grad_norm": 0.19302913546562195,
"learning_rate": 0.001,
"loss": 3.0894622802734375,
"step": 63650
},
{
"epoch": 0.5566189983370243,
"grad_norm": 0.2202378809452057,
"learning_rate": 0.001,
"loss": 3.0923782348632813,
"step": 63700
},
{
"epoch": 0.5570559049291256,
"grad_norm": 0.19396860897541046,
"learning_rate": 0.001,
"loss": 3.08629150390625,
"step": 63750
},
{
"epoch": 0.5574928115212269,
"grad_norm": 0.21728308498859406,
"learning_rate": 0.001,
"loss": 3.09364990234375,
"step": 63800
},
{
"epoch": 0.5579297181133281,
"grad_norm": 0.21487531065940857,
"learning_rate": 0.001,
"loss": 3.0956390380859373,
"step": 63850
},
{
"epoch": 0.5583666247054294,
"grad_norm": 0.2267727553844452,
"learning_rate": 0.001,
"loss": 3.1005755615234376,
"step": 63900
},
{
"epoch": 0.5588035312975307,
"grad_norm": 0.1991662234067917,
"learning_rate": 0.001,
"loss": 3.0944158935546877,
"step": 63950
},
{
"epoch": 0.559240437889632,
"grad_norm": 0.20589323341846466,
"learning_rate": 0.001,
"loss": 3.089896545410156,
"step": 64000
},
{
"epoch": 0.5596773444817332,
"grad_norm": 0.22006580233573914,
"learning_rate": 0.001,
"loss": 3.0944476318359375,
"step": 64050
},
{
"epoch": 0.5601142510738345,
"grad_norm": 0.2285172939300537,
"learning_rate": 0.001,
"loss": 3.0997689819335936,
"step": 64100
},
{
"epoch": 0.5605511576659358,
"grad_norm": 0.19842574000358582,
"learning_rate": 0.001,
"loss": 3.0965536499023436,
"step": 64150
},
{
"epoch": 0.560988064258037,
"grad_norm": 0.19299310445785522,
"learning_rate": 0.001,
"loss": 3.0941842651367186,
"step": 64200
},
{
"epoch": 0.5614249708501383,
"grad_norm": 0.20051312446594238,
"learning_rate": 0.001,
"loss": 3.0957354736328124,
"step": 64250
},
{
"epoch": 0.5618618774422396,
"grad_norm": 0.21229594945907593,
"learning_rate": 0.001,
"loss": 3.088150939941406,
"step": 64300
},
{
"epoch": 0.5622987840343409,
"grad_norm": 0.19384075701236725,
"learning_rate": 0.001,
"loss": 3.09663818359375,
"step": 64350
},
{
"epoch": 0.5627356906264421,
"grad_norm": 0.19096499681472778,
"learning_rate": 0.001,
"loss": 3.08644775390625,
"step": 64400
},
{
"epoch": 0.5631725972185434,
"grad_norm": 0.22296911478042603,
"learning_rate": 0.001,
"loss": 3.0915518188476563,
"step": 64450
},
{
"epoch": 0.5636095038106447,
"grad_norm": 0.22386468946933746,
"learning_rate": 0.001,
"loss": 3.095849914550781,
"step": 64500
},
{
"epoch": 0.564046410402746,
"grad_norm": 0.22815942764282227,
"learning_rate": 0.001,
"loss": 3.0928961181640626,
"step": 64550
},
{
"epoch": 0.5644833169948472,
"grad_norm": 0.2164069265127182,
"learning_rate": 0.001,
"loss": 3.0987203979492186,
"step": 64600
},
{
"epoch": 0.5649202235869485,
"grad_norm": 0.20548774302005768,
"learning_rate": 0.001,
"loss": 3.0992388916015625,
"step": 64650
},
{
"epoch": 0.5653571301790498,
"grad_norm": 0.26139140129089355,
"learning_rate": 0.001,
"loss": 3.089132080078125,
"step": 64700
},
{
"epoch": 0.565794036771151,
"grad_norm": 0.2258266806602478,
"learning_rate": 0.001,
"loss": 3.1002362060546873,
"step": 64750
},
{
"epoch": 0.5662309433632523,
"grad_norm": 0.19188599288463593,
"learning_rate": 0.001,
"loss": 3.08794677734375,
"step": 64800
},
{
"epoch": 0.5666678499553536,
"grad_norm": 0.23600107431411743,
"learning_rate": 0.001,
"loss": 3.0905966186523437,
"step": 64850
},
{
"epoch": 0.5671047565474548,
"grad_norm": 0.20012392103672028,
"learning_rate": 0.001,
"loss": 3.097625732421875,
"step": 64900
},
{
"epoch": 0.5675416631395561,
"grad_norm": 0.2047959864139557,
"learning_rate": 0.001,
"loss": 3.0898602294921873,
"step": 64950
},
{
"epoch": 0.5679785697316574,
"grad_norm": 0.20636017620563507,
"learning_rate": 0.001,
"loss": 3.0943109130859376,
"step": 65000
},
{
"epoch": 0.5684154763237587,
"grad_norm": 0.2176947295665741,
"learning_rate": 0.001,
"loss": 3.100743408203125,
"step": 65050
},
{
"epoch": 0.5688523829158599,
"grad_norm": 0.19603633880615234,
"learning_rate": 0.001,
"loss": 3.0980572509765625,
"step": 65100
},
{
"epoch": 0.5692892895079612,
"grad_norm": 0.2076655924320221,
"learning_rate": 0.001,
"loss": 3.0956036376953127,
"step": 65150
},
{
"epoch": 0.5697261961000626,
"grad_norm": 0.19063986837863922,
"learning_rate": 0.001,
"loss": 3.093685302734375,
"step": 65200
},
{
"epoch": 0.5701631026921639,
"grad_norm": 0.20425820350646973,
"learning_rate": 0.001,
"loss": 3.0948052978515626,
"step": 65250
},
{
"epoch": 0.5706000092842651,
"grad_norm": 0.20817136764526367,
"learning_rate": 0.001,
"loss": 3.089617919921875,
"step": 65300
},
{
"epoch": 0.5710369158763664,
"grad_norm": 0.21174578368663788,
"learning_rate": 0.001,
"loss": 3.1015667724609375,
"step": 65350
},
{
"epoch": 0.5714738224684677,
"grad_norm": 0.18564113974571228,
"learning_rate": 0.001,
"loss": 3.0898806762695314,
"step": 65400
},
{
"epoch": 0.571910729060569,
"grad_norm": 0.1915711909532547,
"learning_rate": 0.001,
"loss": 3.0940789794921875,
"step": 65450
},
{
"epoch": 0.5723476356526702,
"grad_norm": 0.2092977911233902,
"learning_rate": 0.001,
"loss": 3.0888333129882812,
"step": 65500
},
{
"epoch": 0.5727845422447715,
"grad_norm": 0.21278750896453857,
"learning_rate": 0.001,
"loss": 3.0880804443359375,
"step": 65550
},
{
"epoch": 0.5732214488368728,
"grad_norm": 0.20913387835025787,
"learning_rate": 0.001,
"loss": 3.0950588989257812,
"step": 65600
},
{
"epoch": 0.573658355428974,
"grad_norm": 0.2045316845178604,
"learning_rate": 0.001,
"loss": 3.092087707519531,
"step": 65650
},
{
"epoch": 0.5740952620210753,
"grad_norm": 0.2124064713716507,
"learning_rate": 0.001,
"loss": 3.0842990112304687,
"step": 65700
},
{
"epoch": 0.5745321686131766,
"grad_norm": 0.1893104612827301,
"learning_rate": 0.001,
"loss": 3.086448974609375,
"step": 65750
},
{
"epoch": 0.5749690752052778,
"grad_norm": 0.2089659869670868,
"learning_rate": 0.001,
"loss": 3.090556640625,
"step": 65800
},
{
"epoch": 0.5754059817973791,
"grad_norm": 0.20486311614513397,
"learning_rate": 0.001,
"loss": 3.0897393798828126,
"step": 65850
},
{
"epoch": 0.5758428883894804,
"grad_norm": 0.23565922677516937,
"learning_rate": 0.001,
"loss": 3.097444763183594,
"step": 65900
},
{
"epoch": 0.5762797949815817,
"grad_norm": 0.21557746827602386,
"learning_rate": 0.001,
"loss": 3.094757080078125,
"step": 65950
},
{
"epoch": 0.5767167015736829,
"grad_norm": 0.2065187245607376,
"learning_rate": 0.001,
"loss": 3.0929449462890624,
"step": 66000
},
{
"epoch": 0.5767167015736829,
"eval_loss": 3.104686737060547,
"eval_runtime": 6.6794,
"eval_samples_per_second": 146.121,
"eval_steps_per_second": 9.133,
"step": 66000
},
{
"epoch": 0.5771536081657842,
"grad_norm": 0.20960427820682526,
"learning_rate": 0.001,
"loss": 3.0989083862304687,
"step": 66050
},
{
"epoch": 0.5775905147578855,
"grad_norm": 0.21226616203784943,
"learning_rate": 0.001,
"loss": 3.0913629150390625,
"step": 66100
},
{
"epoch": 0.5780274213499867,
"grad_norm": 0.20174987614154816,
"learning_rate": 0.001,
"loss": 3.093486022949219,
"step": 66150
},
{
"epoch": 0.578464327942088,
"grad_norm": 0.1935935914516449,
"learning_rate": 0.001,
"loss": 3.089583435058594,
"step": 66200
},
{
"epoch": 0.5789012345341893,
"grad_norm": 0.20626269280910492,
"learning_rate": 0.001,
"loss": 3.094388122558594,
"step": 66250
},
{
"epoch": 0.5793381411262906,
"grad_norm": 0.19522081315517426,
"learning_rate": 0.001,
"loss": 3.094794921875,
"step": 66300
},
{
"epoch": 0.5797750477183918,
"grad_norm": 0.20273537933826447,
"learning_rate": 0.001,
"loss": 3.092439880371094,
"step": 66350
},
{
"epoch": 0.5802119543104931,
"grad_norm": 0.19178280234336853,
"learning_rate": 0.001,
"loss": 3.0910601806640625,
"step": 66400
},
{
"epoch": 0.5806488609025944,
"grad_norm": 0.20277613401412964,
"learning_rate": 0.001,
"loss": 3.0902255249023436,
"step": 66450
},
{
"epoch": 0.5810857674946956,
"grad_norm": 0.23700758814811707,
"learning_rate": 0.001,
"loss": 3.0934002685546873,
"step": 66500
},
{
"epoch": 0.5815226740867969,
"grad_norm": 0.21032118797302246,
"learning_rate": 0.001,
"loss": 3.0954425048828127,
"step": 66550
},
{
"epoch": 0.5819595806788982,
"grad_norm": 0.2345978319644928,
"learning_rate": 0.001,
"loss": 3.093941650390625,
"step": 66600
},
{
"epoch": 0.5823964872709995,
"grad_norm": 0.21336917579174042,
"learning_rate": 0.001,
"loss": 3.0917340087890626,
"step": 66650
},
{
"epoch": 0.5828333938631007,
"grad_norm": 0.20741680264472961,
"learning_rate": 0.001,
"loss": 3.092748107910156,
"step": 66700
},
{
"epoch": 0.583270300455202,
"grad_norm": 0.21132901310920715,
"learning_rate": 0.001,
"loss": 3.089022521972656,
"step": 66750
},
{
"epoch": 0.5837072070473033,
"grad_norm": 0.21097774803638458,
"learning_rate": 0.001,
"loss": 3.090925598144531,
"step": 66800
},
{
"epoch": 0.5841441136394047,
"grad_norm": 0.19547730684280396,
"learning_rate": 0.001,
"loss": 3.084955139160156,
"step": 66850
},
{
"epoch": 0.5845810202315059,
"grad_norm": 0.20946630835533142,
"learning_rate": 0.001,
"loss": 3.0949261474609373,
"step": 66900
},
{
"epoch": 0.5850179268236072,
"grad_norm": 0.23481234908103943,
"learning_rate": 0.001,
"loss": 3.096903991699219,
"step": 66950
},
{
"epoch": 0.5854548334157085,
"grad_norm": 0.20156437158584595,
"learning_rate": 0.001,
"loss": 3.0910491943359375,
"step": 67000
},
{
"epoch": 0.5858917400078097,
"grad_norm": 0.1909584105014801,
"learning_rate": 0.001,
"loss": 3.094480285644531,
"step": 67050
},
{
"epoch": 0.586328646599911,
"grad_norm": 0.20470935106277466,
"learning_rate": 0.001,
"loss": 3.088348693847656,
"step": 67100
},
{
"epoch": 0.5867655531920123,
"grad_norm": 0.22307629883289337,
"learning_rate": 0.001,
"loss": 3.09716552734375,
"step": 67150
},
{
"epoch": 0.5872024597841136,
"grad_norm": 0.2180834710597992,
"learning_rate": 0.001,
"loss": 3.087926330566406,
"step": 67200
},
{
"epoch": 0.5876393663762148,
"grad_norm": 0.2073758989572525,
"learning_rate": 0.001,
"loss": 3.095177001953125,
"step": 67250
},
{
"epoch": 0.5880762729683161,
"grad_norm": 0.21006031334400177,
"learning_rate": 0.001,
"loss": 3.09031494140625,
"step": 67300
},
{
"epoch": 0.5885131795604174,
"grad_norm": 0.1914648562669754,
"learning_rate": 0.001,
"loss": 3.0981390380859377,
"step": 67350
},
{
"epoch": 0.5889500861525186,
"grad_norm": 0.23198439180850983,
"learning_rate": 0.001,
"loss": 3.093330078125,
"step": 67400
},
{
"epoch": 0.5893869927446199,
"grad_norm": 0.20516900718212128,
"learning_rate": 0.001,
"loss": 3.093861389160156,
"step": 67450
},
{
"epoch": 0.5898238993367212,
"grad_norm": 0.20441021025180817,
"learning_rate": 0.001,
"loss": 3.0905535888671873,
"step": 67500
},
{
"epoch": 0.5902608059288225,
"grad_norm": 0.19688834249973297,
"learning_rate": 0.001,
"loss": 3.0924383544921876,
"step": 67550
},
{
"epoch": 0.5906977125209237,
"grad_norm": 0.22052137553691864,
"learning_rate": 0.001,
"loss": 3.088451232910156,
"step": 67600
},
{
"epoch": 0.591134619113025,
"grad_norm": 0.19194139540195465,
"learning_rate": 0.001,
"loss": 3.0940597534179686,
"step": 67650
},
{
"epoch": 0.5915715257051263,
"grad_norm": 0.2074732780456543,
"learning_rate": 0.001,
"loss": 3.097034606933594,
"step": 67700
},
{
"epoch": 0.5920084322972275,
"grad_norm": 0.21024189889431,
"learning_rate": 0.001,
"loss": 3.0904226684570313,
"step": 67750
},
{
"epoch": 0.5924453388893288,
"grad_norm": 0.21435067057609558,
"learning_rate": 0.001,
"loss": 3.091639099121094,
"step": 67800
},
{
"epoch": 0.5928822454814301,
"grad_norm": 0.199874147772789,
"learning_rate": 0.001,
"loss": 3.0975311279296873,
"step": 67850
},
{
"epoch": 0.5933191520735314,
"grad_norm": 0.2070246934890747,
"learning_rate": 0.001,
"loss": 3.09368896484375,
"step": 67900
},
{
"epoch": 0.5937560586656326,
"grad_norm": 0.20806828141212463,
"learning_rate": 0.001,
"loss": 3.0886526489257813,
"step": 67950
},
{
"epoch": 0.5941929652577339,
"grad_norm": 0.20247897505760193,
"learning_rate": 0.001,
"loss": 3.0894659423828124,
"step": 68000
},
{
"epoch": 0.5946298718498352,
"grad_norm": 0.2285785973072052,
"learning_rate": 0.001,
"loss": 3.09499755859375,
"step": 68050
},
{
"epoch": 0.5950667784419365,
"grad_norm": 0.209172323346138,
"learning_rate": 0.001,
"loss": 3.087811279296875,
"step": 68100
},
{
"epoch": 0.5955036850340377,
"grad_norm": 0.23774203658103943,
"learning_rate": 0.001,
"loss": 3.092928161621094,
"step": 68150
},
{
"epoch": 0.595940591626139,
"grad_norm": 0.17968431115150452,
"learning_rate": 0.001,
"loss": 3.0919284057617187,
"step": 68200
},
{
"epoch": 0.5963774982182403,
"grad_norm": 0.20794937014579773,
"learning_rate": 0.001,
"loss": 3.0950167846679686,
"step": 68250
},
{
"epoch": 0.5968144048103415,
"grad_norm": 0.19683150947093964,
"learning_rate": 0.001,
"loss": 3.0839337158203124,
"step": 68300
},
{
"epoch": 0.5972513114024428,
"grad_norm": 0.22212012112140656,
"learning_rate": 0.001,
"loss": 3.0870761108398437,
"step": 68350
},
{
"epoch": 0.5976882179945441,
"grad_norm": 0.19362536072731018,
"learning_rate": 0.001,
"loss": 3.0865191650390624,
"step": 68400
},
{
"epoch": 0.5981251245866454,
"grad_norm": 0.20000749826431274,
"learning_rate": 0.001,
"loss": 3.0844793701171875,
"step": 68450
},
{
"epoch": 0.5985620311787467,
"grad_norm": 0.22240105271339417,
"learning_rate": 0.001,
"loss": 3.084669494628906,
"step": 68500
},
{
"epoch": 0.598998937770848,
"grad_norm": 0.20492695271968842,
"learning_rate": 0.001,
"loss": 3.0854861450195314,
"step": 68550
},
{
"epoch": 0.5994358443629493,
"grad_norm": 0.20981861650943756,
"learning_rate": 0.001,
"loss": 3.086083984375,
"step": 68600
},
{
"epoch": 0.5998727509550505,
"grad_norm": 0.2117357701063156,
"learning_rate": 0.001,
"loss": 3.097230224609375,
"step": 68650
},
{
"epoch": 0.6003096575471518,
"grad_norm": 0.23792456090450287,
"learning_rate": 0.001,
"loss": 3.0937704467773437,
"step": 68700
},
{
"epoch": 0.6007465641392531,
"grad_norm": 0.2122933715581894,
"learning_rate": 0.001,
"loss": 3.1016470336914064,
"step": 68750
},
{
"epoch": 0.6011834707313544,
"grad_norm": 0.19460844993591309,
"learning_rate": 0.001,
"loss": 3.0967706298828124,
"step": 68800
},
{
"epoch": 0.6016203773234556,
"grad_norm": 0.22884444892406464,
"learning_rate": 0.001,
"loss": 3.089122314453125,
"step": 68850
},
{
"epoch": 0.6020572839155569,
"grad_norm": 0.21679161489009857,
"learning_rate": 0.001,
"loss": 3.08767333984375,
"step": 68900
},
{
"epoch": 0.6024941905076582,
"grad_norm": 0.23170873522758484,
"learning_rate": 0.001,
"loss": 3.092586669921875,
"step": 68950
},
{
"epoch": 0.6029310970997594,
"grad_norm": 0.21384017169475555,
"learning_rate": 0.001,
"loss": 3.092149658203125,
"step": 69000
},
{
"epoch": 0.6029310970997594,
"eval_loss": 3.1037347316741943,
"eval_runtime": 6.5992,
"eval_samples_per_second": 147.897,
"eval_steps_per_second": 9.244,
"step": 69000
},
{
"epoch": 0.6033680036918607,
"grad_norm": 0.22783087193965912,
"learning_rate": 0.001,
"loss": 3.0928695678710936,
"step": 69050
},
{
"epoch": 0.603804910283962,
"grad_norm": 0.19492192566394806,
"learning_rate": 0.001,
"loss": 3.090107727050781,
"step": 69100
},
{
"epoch": 0.6042418168760633,
"grad_norm": 0.21013516187667847,
"learning_rate": 0.001,
"loss": 3.097440490722656,
"step": 69150
},
{
"epoch": 0.6046787234681645,
"grad_norm": 0.21594007313251495,
"learning_rate": 0.001,
"loss": 3.091844482421875,
"step": 69200
},
{
"epoch": 0.6051156300602658,
"grad_norm": 0.22112025320529938,
"learning_rate": 0.001,
"loss": 3.0902169799804686,
"step": 69250
},
{
"epoch": 0.6055525366523671,
"grad_norm": 0.20222006738185883,
"learning_rate": 0.001,
"loss": 3.0878677368164062,
"step": 69300
},
{
"epoch": 0.6059894432444684,
"grad_norm": 0.2167014479637146,
"learning_rate": 0.001,
"loss": 3.094149475097656,
"step": 69350
},
{
"epoch": 0.6064263498365696,
"grad_norm": 0.20349176228046417,
"learning_rate": 0.001,
"loss": 3.091081237792969,
"step": 69400
},
{
"epoch": 0.6068632564286709,
"grad_norm": 0.20823749899864197,
"learning_rate": 0.001,
"loss": 3.0878890991210937,
"step": 69450
},
{
"epoch": 0.6073001630207722,
"grad_norm": 0.21119347214698792,
"learning_rate": 0.001,
"loss": 3.085202941894531,
"step": 69500
},
{
"epoch": 0.6077370696128734,
"grad_norm": 0.21648874878883362,
"learning_rate": 0.001,
"loss": 3.082725830078125,
"step": 69550
},
{
"epoch": 0.6081739762049747,
"grad_norm": 0.23378799855709076,
"learning_rate": 0.001,
"loss": 3.092845458984375,
"step": 69600
},
{
"epoch": 0.608610882797076,
"grad_norm": 0.22747142612934113,
"learning_rate": 0.001,
"loss": 3.087052001953125,
"step": 69650
},
{
"epoch": 0.6090477893891773,
"grad_norm": 0.19154012203216553,
"learning_rate": 0.001,
"loss": 3.088550720214844,
"step": 69700
},
{
"epoch": 0.6094846959812785,
"grad_norm": 0.2168135941028595,
"learning_rate": 0.001,
"loss": 3.0888552856445313,
"step": 69750
},
{
"epoch": 0.6099216025733798,
"grad_norm": 0.19610479474067688,
"learning_rate": 0.001,
"loss": 3.092249755859375,
"step": 69800
},
{
"epoch": 0.6103585091654811,
"grad_norm": 0.19346152245998383,
"learning_rate": 0.001,
"loss": 3.091408996582031,
"step": 69850
},
{
"epoch": 0.6107954157575823,
"grad_norm": 0.2086789458990097,
"learning_rate": 0.001,
"loss": 3.0925308227539063,
"step": 69900
},
{
"epoch": 0.6112323223496836,
"grad_norm": 0.2246716022491455,
"learning_rate": 0.001,
"loss": 3.0862396240234373,
"step": 69950
},
{
"epoch": 0.6116692289417849,
"grad_norm": 0.21956804394721985,
"learning_rate": 0.001,
"loss": 3.090303649902344,
"step": 70000
},
{
"epoch": 0.6121061355338862,
"grad_norm": 0.2048935443162918,
"learning_rate": 0.001,
"loss": 3.09474853515625,
"step": 70050
},
{
"epoch": 0.6125430421259874,
"grad_norm": 0.19252470135688782,
"learning_rate": 0.001,
"loss": 3.0852786254882814,
"step": 70100
},
{
"epoch": 0.6129799487180888,
"grad_norm": 0.22725819051265717,
"learning_rate": 0.001,
"loss": 3.0892864990234377,
"step": 70150
},
{
"epoch": 0.6134168553101901,
"grad_norm": 0.19779269397258759,
"learning_rate": 0.001,
"loss": 3.086781005859375,
"step": 70200
},
{
"epoch": 0.6138537619022914,
"grad_norm": 0.2138117551803589,
"learning_rate": 0.001,
"loss": 3.088963928222656,
"step": 70250
},
{
"epoch": 0.6142906684943926,
"grad_norm": 0.24275323748588562,
"learning_rate": 0.001,
"loss": 3.0880886840820314,
"step": 70300
},
{
"epoch": 0.6147275750864939,
"grad_norm": 0.1906837373971939,
"learning_rate": 0.001,
"loss": 3.088238525390625,
"step": 70350
},
{
"epoch": 0.6151644816785952,
"grad_norm": 0.20280472934246063,
"learning_rate": 0.001,
"loss": 3.0959124755859375,
"step": 70400
},
{
"epoch": 0.6156013882706964,
"grad_norm": 0.21057118475437164,
"learning_rate": 0.001,
"loss": 3.088071594238281,
"step": 70450
},
{
"epoch": 0.6160382948627977,
"grad_norm": 0.20867182314395905,
"learning_rate": 0.001,
"loss": 3.0873257446289064,
"step": 70500
},
{
"epoch": 0.616475201454899,
"grad_norm": 0.18485228717327118,
"learning_rate": 0.001,
"loss": 3.090894775390625,
"step": 70550
},
{
"epoch": 0.6169121080470003,
"grad_norm": 0.20439258217811584,
"learning_rate": 0.001,
"loss": 3.0873367309570314,
"step": 70600
},
{
"epoch": 0.6173490146391015,
"grad_norm": 0.19232890009880066,
"learning_rate": 0.001,
"loss": 3.086202392578125,
"step": 70650
},
{
"epoch": 0.6177859212312028,
"grad_norm": 0.20417119562625885,
"learning_rate": 0.001,
"loss": 3.0844235229492187,
"step": 70700
},
{
"epoch": 0.6182228278233041,
"grad_norm": 0.21768328547477722,
"learning_rate": 0.001,
"loss": 3.0910305786132812,
"step": 70750
},
{
"epoch": 0.6186597344154053,
"grad_norm": 0.20101118087768555,
"learning_rate": 0.001,
"loss": 3.086376953125,
"step": 70800
},
{
"epoch": 0.6190966410075066,
"grad_norm": 0.22932034730911255,
"learning_rate": 0.001,
"loss": 3.092792663574219,
"step": 70850
},
{
"epoch": 0.6195335475996079,
"grad_norm": 0.1919311285018921,
"learning_rate": 0.001,
"loss": 3.095375671386719,
"step": 70900
},
{
"epoch": 0.6199704541917092,
"grad_norm": 0.20381779968738556,
"learning_rate": 0.001,
"loss": 3.0846466064453124,
"step": 70950
},
{
"epoch": 0.6204073607838104,
"grad_norm": 0.19156542420387268,
"learning_rate": 0.001,
"loss": 3.0865545654296875,
"step": 71000
},
{
"epoch": 0.6208442673759117,
"grad_norm": 0.210403174161911,
"learning_rate": 0.001,
"loss": 3.0861190795898437,
"step": 71050
},
{
"epoch": 0.621281173968013,
"grad_norm": 0.21578097343444824,
"learning_rate": 0.001,
"loss": 3.089746398925781,
"step": 71100
},
{
"epoch": 0.6217180805601142,
"grad_norm": 0.21439765393733978,
"learning_rate": 0.001,
"loss": 3.090135192871094,
"step": 71150
},
{
"epoch": 0.6221549871522155,
"grad_norm": 0.2040693610906601,
"learning_rate": 0.001,
"loss": 3.0858038330078124,
"step": 71200
},
{
"epoch": 0.6225918937443168,
"grad_norm": 0.22903653979301453,
"learning_rate": 0.001,
"loss": 3.087599792480469,
"step": 71250
},
{
"epoch": 0.623028800336418,
"grad_norm": 0.20632053911685944,
"learning_rate": 0.001,
"loss": 3.0889736938476564,
"step": 71300
},
{
"epoch": 0.6234657069285193,
"grad_norm": 0.19149228930473328,
"learning_rate": 0.001,
"loss": 3.0878219604492188,
"step": 71350
},
{
"epoch": 0.6239026135206206,
"grad_norm": 0.1927986443042755,
"learning_rate": 0.001,
"loss": 3.0972030639648436,
"step": 71400
},
{
"epoch": 0.6243395201127219,
"grad_norm": 0.19735023379325867,
"learning_rate": 0.001,
"loss": 3.081488037109375,
"step": 71450
},
{
"epoch": 0.6247764267048231,
"grad_norm": 0.226057231426239,
"learning_rate": 0.001,
"loss": 3.0849783325195315,
"step": 71500
},
{
"epoch": 0.6252133332969244,
"grad_norm": 0.2008025199174881,
"learning_rate": 0.001,
"loss": 3.091714172363281,
"step": 71550
},
{
"epoch": 0.6256502398890257,
"grad_norm": 0.22580116987228394,
"learning_rate": 0.001,
"loss": 3.0940582275390627,
"step": 71600
},
{
"epoch": 0.626087146481127,
"grad_norm": 0.21546880900859833,
"learning_rate": 0.001,
"loss": 3.084023742675781,
"step": 71650
},
{
"epoch": 0.6265240530732282,
"grad_norm": 0.21075797080993652,
"learning_rate": 0.001,
"loss": 3.0934027099609374,
"step": 71700
},
{
"epoch": 0.6269609596653295,
"grad_norm": 0.21317969262599945,
"learning_rate": 0.001,
"loss": 3.0822622680664065,
"step": 71750
},
{
"epoch": 0.6273978662574309,
"grad_norm": 0.22095735371112823,
"learning_rate": 0.001,
"loss": 3.09416015625,
"step": 71800
},
{
"epoch": 0.6278347728495322,
"grad_norm": 0.23459088802337646,
"learning_rate": 0.001,
"loss": 3.0877099609375,
"step": 71850
},
{
"epoch": 0.6282716794416334,
"grad_norm": 0.20578697323799133,
"learning_rate": 0.001,
"loss": 3.0866461181640625,
"step": 71900
},
{
"epoch": 0.6287085860337347,
"grad_norm": 0.21029488742351532,
"learning_rate": 0.001,
"loss": 3.0918569946289063,
"step": 71950
},
{
"epoch": 0.629145492625836,
"grad_norm": 0.22013671696186066,
"learning_rate": 0.001,
"loss": 3.0859359741210937,
"step": 72000
},
{
"epoch": 0.629145492625836,
"eval_loss": 3.1005258560180664,
"eval_runtime": 5.9193,
"eval_samples_per_second": 164.883,
"eval_steps_per_second": 10.305,
"step": 72000
},
{
"epoch": 0.6295823992179372,
"grad_norm": 0.22188881039619446,
"learning_rate": 0.001,
"loss": 3.0876104736328127,
"step": 72050
},
{
"epoch": 0.6300193058100385,
"grad_norm": 0.19578015804290771,
"learning_rate": 0.001,
"loss": 3.0866427612304688,
"step": 72100
},
{
"epoch": 0.6304562124021398,
"grad_norm": 0.2074483186006546,
"learning_rate": 0.001,
"loss": 3.09542236328125,
"step": 72150
},
{
"epoch": 0.630893118994241,
"grad_norm": 0.19705738127231598,
"learning_rate": 0.001,
"loss": 3.085140380859375,
"step": 72200
},
{
"epoch": 0.6313300255863423,
"grad_norm": 0.22753173112869263,
"learning_rate": 0.001,
"loss": 3.090760498046875,
"step": 72250
},
{
"epoch": 0.6317669321784436,
"grad_norm": 0.21299652755260468,
"learning_rate": 0.001,
"loss": 3.0883963012695315,
"step": 72300
},
{
"epoch": 0.6322038387705449,
"grad_norm": 0.201585054397583,
"learning_rate": 0.001,
"loss": 3.085474853515625,
"step": 72350
},
{
"epoch": 0.6326407453626461,
"grad_norm": 0.2314334511756897,
"learning_rate": 0.001,
"loss": 3.0891351318359375,
"step": 72400
},
{
"epoch": 0.6330776519547474,
"grad_norm": 0.23124907910823822,
"learning_rate": 0.001,
"loss": 3.0904833984375,
"step": 72450
},
{
"epoch": 0.6335145585468487,
"grad_norm": 0.20642049610614777,
"learning_rate": 0.001,
"loss": 3.0963681030273436,
"step": 72500
},
{
"epoch": 0.63395146513895,
"grad_norm": 0.2025928646326065,
"learning_rate": 0.001,
"loss": 3.0896505737304687,
"step": 72550
},
{
"epoch": 0.6343883717310512,
"grad_norm": 0.201346755027771,
"learning_rate": 0.001,
"loss": 3.0857489013671877,
"step": 72600
},
{
"epoch": 0.6348252783231525,
"grad_norm": 0.20704689621925354,
"learning_rate": 0.001,
"loss": 3.09096435546875,
"step": 72650
},
{
"epoch": 0.6352621849152538,
"grad_norm": 0.234156996011734,
"learning_rate": 0.001,
"loss": 3.0863861083984374,
"step": 72700
},
{
"epoch": 0.635699091507355,
"grad_norm": 0.22558419406414032,
"learning_rate": 0.001,
"loss": 3.087818603515625,
"step": 72750
},
{
"epoch": 0.6361359980994563,
"grad_norm": 0.20652826130390167,
"learning_rate": 0.001,
"loss": 3.0841445922851562,
"step": 72800
},
{
"epoch": 0.6365729046915576,
"grad_norm": 0.2017735093832016,
"learning_rate": 0.001,
"loss": 3.0874560546875,
"step": 72850
},
{
"epoch": 0.6370098112836589,
"grad_norm": 0.219964399933815,
"learning_rate": 0.001,
"loss": 3.086517333984375,
"step": 72900
},
{
"epoch": 0.6374467178757601,
"grad_norm": 0.19547931849956512,
"learning_rate": 0.001,
"loss": 3.08653076171875,
"step": 72950
},
{
"epoch": 0.6378836244678614,
"grad_norm": 0.2237868309020996,
"learning_rate": 0.001,
"loss": 3.089991455078125,
"step": 73000
},
{
"epoch": 0.6383205310599627,
"grad_norm": 0.21151629090309143,
"learning_rate": 0.001,
"loss": 3.0868984985351564,
"step": 73050
},
{
"epoch": 0.6387574376520639,
"grad_norm": 0.20942968130111694,
"learning_rate": 0.001,
"loss": 3.0847763061523437,
"step": 73100
},
{
"epoch": 0.6391943442441652,
"grad_norm": 0.20986200869083405,
"learning_rate": 0.001,
"loss": 3.0854318237304685,
"step": 73150
},
{
"epoch": 0.6396312508362665,
"grad_norm": 0.19512173533439636,
"learning_rate": 0.001,
"loss": 3.083329162597656,
"step": 73200
},
{
"epoch": 0.6400681574283678,
"grad_norm": 0.20538833737373352,
"learning_rate": 0.001,
"loss": 3.092357482910156,
"step": 73250
},
{
"epoch": 0.640505064020469,
"grad_norm": 0.22105765342712402,
"learning_rate": 0.001,
"loss": 3.0875225830078126,
"step": 73300
},
{
"epoch": 0.6409419706125703,
"grad_norm": 0.2285480499267578,
"learning_rate": 0.001,
"loss": 3.0876531982421875,
"step": 73350
},
{
"epoch": 0.6413788772046716,
"grad_norm": 0.19559548795223236,
"learning_rate": 0.001,
"loss": 3.0894158935546874,
"step": 73400
},
{
"epoch": 0.641815783796773,
"grad_norm": 0.20379120111465454,
"learning_rate": 0.001,
"loss": 3.096446838378906,
"step": 73450
},
{
"epoch": 0.6422526903888742,
"grad_norm": 0.22814227640628815,
"learning_rate": 0.001,
"loss": 3.0895004272460938,
"step": 73500
},
{
"epoch": 0.6426895969809755,
"grad_norm": 0.2117760330438614,
"learning_rate": 0.001,
"loss": 3.086138916015625,
"step": 73550
},
{
"epoch": 0.6431265035730768,
"grad_norm": 0.21658946573734283,
"learning_rate": 0.001,
"loss": 3.0890579223632812,
"step": 73600
},
{
"epoch": 0.643563410165178,
"grad_norm": 0.19912591576576233,
"learning_rate": 0.001,
"loss": 3.0884158325195314,
"step": 73650
},
{
"epoch": 0.6440003167572793,
"grad_norm": 0.20438186824321747,
"learning_rate": 0.001,
"loss": 3.0847012329101564,
"step": 73700
},
{
"epoch": 0.6444372233493806,
"grad_norm": 0.21380993723869324,
"learning_rate": 0.001,
"loss": 3.090211181640625,
"step": 73750
},
{
"epoch": 0.6448741299414819,
"grad_norm": 0.21418142318725586,
"learning_rate": 0.001,
"loss": 3.089119567871094,
"step": 73800
},
{
"epoch": 0.6453110365335831,
"grad_norm": 0.20147238671779633,
"learning_rate": 0.001,
"loss": 3.0948516845703127,
"step": 73850
},
{
"epoch": 0.6457479431256844,
"grad_norm": 0.2023879885673523,
"learning_rate": 0.001,
"loss": 3.088772888183594,
"step": 73900
},
{
"epoch": 0.6461848497177857,
"grad_norm": 0.22049619257450104,
"learning_rate": 0.001,
"loss": 3.0841995239257813,
"step": 73950
},
{
"epoch": 0.6466217563098869,
"grad_norm": 0.23035335540771484,
"learning_rate": 0.001,
"loss": 3.086770324707031,
"step": 74000
},
{
"epoch": 0.6470586629019882,
"grad_norm": 0.20004555583000183,
"learning_rate": 0.001,
"loss": 3.0854559326171875,
"step": 74050
},
{
"epoch": 0.6474955694940895,
"grad_norm": 0.2120899260044098,
"learning_rate": 0.001,
"loss": 3.0889752197265623,
"step": 74100
},
{
"epoch": 0.6479324760861908,
"grad_norm": 0.1969432234764099,
"learning_rate": 0.001,
"loss": 3.0912216186523436,
"step": 74150
},
{
"epoch": 0.648369382678292,
"grad_norm": 0.22493019700050354,
"learning_rate": 0.001,
"loss": 3.0913217163085935,
"step": 74200
},
{
"epoch": 0.6488062892703933,
"grad_norm": 0.21642231941223145,
"learning_rate": 0.001,
"loss": 3.088470458984375,
"step": 74250
},
{
"epoch": 0.6492431958624946,
"grad_norm": 0.20363709330558777,
"learning_rate": 0.001,
"loss": 3.07613525390625,
"step": 74300
},
{
"epoch": 0.6496801024545958,
"grad_norm": 0.2133510559797287,
"learning_rate": 0.001,
"loss": 3.08767333984375,
"step": 74350
},
{
"epoch": 0.6501170090466971,
"grad_norm": 0.21559849381446838,
"learning_rate": 0.001,
"loss": 3.0837835693359374,
"step": 74400
},
{
"epoch": 0.6505539156387984,
"grad_norm": 0.18649795651435852,
"learning_rate": 0.001,
"loss": 3.0857568359375,
"step": 74450
},
{
"epoch": 0.6509908222308997,
"grad_norm": 0.17934051156044006,
"learning_rate": 0.001,
"loss": 3.08024658203125,
"step": 74500
},
{
"epoch": 0.6514277288230009,
"grad_norm": 0.2268570065498352,
"learning_rate": 0.001,
"loss": 3.0841656494140626,
"step": 74550
},
{
"epoch": 0.6518646354151022,
"grad_norm": 0.2021469622850418,
"learning_rate": 0.001,
"loss": 3.0838607788085937,
"step": 74600
},
{
"epoch": 0.6523015420072035,
"grad_norm": 0.21180258691310883,
"learning_rate": 0.001,
"loss": 3.084789123535156,
"step": 74650
},
{
"epoch": 0.6527384485993047,
"grad_norm": 0.19279135763645172,
"learning_rate": 0.001,
"loss": 3.0896685791015623,
"step": 74700
},
{
"epoch": 0.653175355191406,
"grad_norm": 0.20593208074569702,
"learning_rate": 0.001,
"loss": 3.087872314453125,
"step": 74750
},
{
"epoch": 0.6536122617835073,
"grad_norm": 0.21789884567260742,
"learning_rate": 0.001,
"loss": 3.0874728393554687,
"step": 74800
},
{
"epoch": 0.6540491683756086,
"grad_norm": 0.20391666889190674,
"learning_rate": 0.001,
"loss": 3.0856362915039064,
"step": 74850
},
{
"epoch": 0.6544860749677098,
"grad_norm": 0.21403469145298004,
"learning_rate": 0.001,
"loss": 3.087380065917969,
"step": 74900
},
{
"epoch": 0.6549229815598111,
"grad_norm": 0.20805788040161133,
"learning_rate": 0.001,
"loss": 3.0871127319335936,
"step": 74950
},
{
"epoch": 0.6553598881519124,
"grad_norm": 0.20461535453796387,
"learning_rate": 0.001,
"loss": 3.0870095825195314,
"step": 75000
},
{
"epoch": 0.6553598881519124,
"eval_loss": 3.0998106002807617,
"eval_runtime": 5.454,
"eval_samples_per_second": 178.953,
"eval_steps_per_second": 11.185,
"step": 75000
},
{
"epoch": 0.6557967947440136,
"grad_norm": 0.2140463888645172,
"learning_rate": 0.001,
"loss": 3.084339294433594,
"step": 75050
},
{
"epoch": 0.656233701336115,
"grad_norm": 0.22241680324077606,
"learning_rate": 0.001,
"loss": 3.0868234252929687,
"step": 75100
},
{
"epoch": 0.6566706079282163,
"grad_norm": 0.23080740869045258,
"learning_rate": 0.001,
"loss": 3.0898861694335937,
"step": 75150
},
{
"epoch": 0.6571075145203176,
"grad_norm": 0.22261208295822144,
"learning_rate": 0.001,
"loss": 3.0879562377929686,
"step": 75200
},
{
"epoch": 0.6575444211124188,
"grad_norm": 0.19588351249694824,
"learning_rate": 0.001,
"loss": 3.0872787475585937,
"step": 75250
},
{
"epoch": 0.6579813277045201,
"grad_norm": 0.19991254806518555,
"learning_rate": 0.001,
"loss": 3.0886505126953123,
"step": 75300
},
{
"epoch": 0.6584182342966214,
"grad_norm": 0.21594689786434174,
"learning_rate": 0.001,
"loss": 3.0816162109375,
"step": 75350
},
{
"epoch": 0.6588551408887227,
"grad_norm": 0.1952524036169052,
"learning_rate": 0.001,
"loss": 3.090860595703125,
"step": 75400
},
{
"epoch": 0.6592920474808239,
"grad_norm": 0.22160783410072327,
"learning_rate": 0.001,
"loss": 3.08528564453125,
"step": 75450
},
{
"epoch": 0.6597289540729252,
"grad_norm": 0.2093716561794281,
"learning_rate": 0.001,
"loss": 3.0845626831054687,
"step": 75500
},
{
"epoch": 0.6601658606650265,
"grad_norm": 0.20646870136260986,
"learning_rate": 0.001,
"loss": 3.086051025390625,
"step": 75550
},
{
"epoch": 0.6606027672571277,
"grad_norm": 0.2125961184501648,
"learning_rate": 0.001,
"loss": 3.08560546875,
"step": 75600
},
{
"epoch": 0.661039673849229,
"grad_norm": 0.20337621867656708,
"learning_rate": 0.001,
"loss": 3.087316589355469,
"step": 75650
},
{
"epoch": 0.6614765804413303,
"grad_norm": 0.2097005546092987,
"learning_rate": 0.001,
"loss": 3.0843130493164064,
"step": 75700
},
{
"epoch": 0.6619134870334316,
"grad_norm": 0.19445250928401947,
"learning_rate": 0.001,
"loss": 3.0862109375,
"step": 75750
},
{
"epoch": 0.6623503936255328,
"grad_norm": 0.23915830254554749,
"learning_rate": 0.001,
"loss": 3.0884664916992186,
"step": 75800
},
{
"epoch": 0.6627873002176341,
"grad_norm": 0.19137521088123322,
"learning_rate": 0.001,
"loss": 3.0809967041015627,
"step": 75850
},
{
"epoch": 0.6632242068097354,
"grad_norm": 0.20678843557834625,
"learning_rate": 0.001,
"loss": 3.0879534912109374,
"step": 75900
},
{
"epoch": 0.6636611134018366,
"grad_norm": 0.204592764377594,
"learning_rate": 0.001,
"loss": 3.087080993652344,
"step": 75950
},
{
"epoch": 0.6640980199939379,
"grad_norm": 0.19713979959487915,
"learning_rate": 0.001,
"loss": 3.0853662109375,
"step": 76000
},
{
"epoch": 0.6645349265860392,
"grad_norm": 0.19893234968185425,
"learning_rate": 0.001,
"loss": 3.083724060058594,
"step": 76050
},
{
"epoch": 0.6649718331781405,
"grad_norm": 0.23445264995098114,
"learning_rate": 0.001,
"loss": 3.0858554077148437,
"step": 76100
},
{
"epoch": 0.6654087397702417,
"grad_norm": 0.2330544888973236,
"learning_rate": 0.001,
"loss": 3.0829168701171876,
"step": 76150
},
{
"epoch": 0.665845646362343,
"grad_norm": 0.24096150696277618,
"learning_rate": 0.001,
"loss": 3.0872625732421874,
"step": 76200
},
{
"epoch": 0.6662825529544443,
"grad_norm": 0.20782263576984406,
"learning_rate": 0.001,
"loss": 3.091466979980469,
"step": 76250
},
{
"epoch": 0.6667194595465455,
"grad_norm": 0.203996941447258,
"learning_rate": 0.001,
"loss": 3.083094177246094,
"step": 76300
},
{
"epoch": 0.6671563661386468,
"grad_norm": 0.21673306822776794,
"learning_rate": 0.001,
"loss": 3.0734014892578125,
"step": 76350
},
{
"epoch": 0.6675932727307481,
"grad_norm": 0.22725573182106018,
"learning_rate": 0.001,
"loss": 3.0902645874023436,
"step": 76400
},
{
"epoch": 0.6680301793228494,
"grad_norm": 0.2029077708721161,
"learning_rate": 0.001,
"loss": 3.086488952636719,
"step": 76450
},
{
"epoch": 0.6684670859149506,
"grad_norm": 0.2107028216123581,
"learning_rate": 0.001,
"loss": 3.0911825561523436,
"step": 76500
},
{
"epoch": 0.6689039925070519,
"grad_norm": 0.21247823536396027,
"learning_rate": 0.001,
"loss": 3.0857855224609376,
"step": 76550
},
{
"epoch": 0.6693408990991532,
"grad_norm": 0.2248055636882782,
"learning_rate": 0.001,
"loss": 3.083785400390625,
"step": 76600
},
{
"epoch": 0.6697778056912544,
"grad_norm": 0.2240576446056366,
"learning_rate": 0.001,
"loss": 3.0927459716796877,
"step": 76650
},
{
"epoch": 0.6702147122833557,
"grad_norm": 0.2254248857498169,
"learning_rate": 0.001,
"loss": 3.087362060546875,
"step": 76700
},
{
"epoch": 0.6706516188754571,
"grad_norm": 0.1980004608631134,
"learning_rate": 0.001,
"loss": 3.0886932373046876,
"step": 76750
},
{
"epoch": 0.6710885254675584,
"grad_norm": 0.18997803330421448,
"learning_rate": 0.001,
"loss": 3.082783508300781,
"step": 76800
},
{
"epoch": 0.6715254320596596,
"grad_norm": 0.22984033823013306,
"learning_rate": 0.001,
"loss": 3.085504455566406,
"step": 76850
},
{
"epoch": 0.6719623386517609,
"grad_norm": 0.20839153230190277,
"learning_rate": 0.001,
"loss": 3.083135070800781,
"step": 76900
},
{
"epoch": 0.6723992452438622,
"grad_norm": 0.21274180710315704,
"learning_rate": 0.001,
"loss": 3.0884359741210936,
"step": 76950
},
{
"epoch": 0.6728361518359635,
"grad_norm": 0.2012646496295929,
"learning_rate": 0.001,
"loss": 3.0849029541015627,
"step": 77000
},
{
"epoch": 0.6732730584280647,
"grad_norm": 0.22452327609062195,
"learning_rate": 0.001,
"loss": 3.0958441162109374,
"step": 77050
},
{
"epoch": 0.673709965020166,
"grad_norm": 0.19732549786567688,
"learning_rate": 0.001,
"loss": 3.0778488159179687,
"step": 77100
},
{
"epoch": 0.6741468716122673,
"grad_norm": 0.21262523531913757,
"learning_rate": 0.001,
"loss": 3.0909249877929685,
"step": 77150
},
{
"epoch": 0.6745837782043685,
"grad_norm": 0.19116179645061493,
"learning_rate": 0.001,
"loss": 3.0825714111328124,
"step": 77200
},
{
"epoch": 0.6750206847964698,
"grad_norm": 0.19181202352046967,
"learning_rate": 0.001,
"loss": 3.0850335693359376,
"step": 77250
},
{
"epoch": 0.6754575913885711,
"grad_norm": 0.21367040276527405,
"learning_rate": 0.001,
"loss": 3.086380615234375,
"step": 77300
},
{
"epoch": 0.6758944979806724,
"grad_norm": 0.21532468497753143,
"learning_rate": 0.001,
"loss": 3.0895809936523437,
"step": 77350
},
{
"epoch": 0.6763314045727736,
"grad_norm": 0.20175503194332123,
"learning_rate": 0.001,
"loss": 3.081783447265625,
"step": 77400
},
{
"epoch": 0.6767683111648749,
"grad_norm": 0.20455457270145416,
"learning_rate": 0.001,
"loss": 3.0885769653320314,
"step": 77450
},
{
"epoch": 0.6772052177569762,
"grad_norm": 0.2001691460609436,
"learning_rate": 0.001,
"loss": 3.086563720703125,
"step": 77500
},
{
"epoch": 0.6776421243490774,
"grad_norm": 0.1886105239391327,
"learning_rate": 0.001,
"loss": 3.0872772216796873,
"step": 77550
},
{
"epoch": 0.6780790309411787,
"grad_norm": 0.20902469754219055,
"learning_rate": 0.001,
"loss": 3.085574951171875,
"step": 77600
},
{
"epoch": 0.67851593753328,
"grad_norm": 0.22364191710948944,
"learning_rate": 0.001,
"loss": 3.0925259399414062,
"step": 77650
},
{
"epoch": 0.6789528441253813,
"grad_norm": 0.203672856092453,
"learning_rate": 0.001,
"loss": 3.0917709350585936,
"step": 77700
},
{
"epoch": 0.6793897507174825,
"grad_norm": 0.219965860247612,
"learning_rate": 0.001,
"loss": 3.0855865478515625,
"step": 77750
},
{
"epoch": 0.6798266573095838,
"grad_norm": 0.23391039669513702,
"learning_rate": 0.001,
"loss": 3.0876434326171873,
"step": 77800
},
{
"epoch": 0.6802635639016851,
"grad_norm": 0.20248816907405853,
"learning_rate": 0.001,
"loss": 3.0855438232421877,
"step": 77850
},
{
"epoch": 0.6807004704937863,
"grad_norm": 0.18939025700092316,
"learning_rate": 0.001,
"loss": 3.0742877197265623,
"step": 77900
},
{
"epoch": 0.6811373770858876,
"grad_norm": 0.20688220858573914,
"learning_rate": 0.001,
"loss": 3.08621826171875,
"step": 77950
},
{
"epoch": 0.6815742836779889,
"grad_norm": 0.22130656242370605,
"learning_rate": 0.001,
"loss": 3.0854898071289063,
"step": 78000
},
{
"epoch": 0.6815742836779889,
"eval_loss": 3.0959784984588623,
"eval_runtime": 6.4509,
"eval_samples_per_second": 151.296,
"eval_steps_per_second": 9.456,
"step": 78000
},
{
"epoch": 0.6820111902700902,
"grad_norm": 0.2075755000114441,
"learning_rate": 0.001,
"loss": 3.0806759643554686,
"step": 78050
},
{
"epoch": 0.6824480968621914,
"grad_norm": 0.20226220786571503,
"learning_rate": 0.001,
"loss": 3.084062194824219,
"step": 78100
},
{
"epoch": 0.6828850034542927,
"grad_norm": 0.24825316667556763,
"learning_rate": 0.001,
"loss": 3.0854867553710936,
"step": 78150
},
{
"epoch": 0.683321910046394,
"grad_norm": 0.22582001984119415,
"learning_rate": 0.001,
"loss": 3.0803402709960936,
"step": 78200
},
{
"epoch": 0.6837588166384952,
"grad_norm": 0.19570297002792358,
"learning_rate": 0.001,
"loss": 3.0861212158203126,
"step": 78250
},
{
"epoch": 0.6841957232305965,
"grad_norm": 0.20353887975215912,
"learning_rate": 0.001,
"loss": 3.088966979980469,
"step": 78300
},
{
"epoch": 0.6846326298226978,
"grad_norm": 0.21870636940002441,
"learning_rate": 0.001,
"loss": 3.0893423461914065,
"step": 78350
},
{
"epoch": 0.6850695364147992,
"grad_norm": 0.20556245744228363,
"learning_rate": 0.001,
"loss": 3.0843768310546875,
"step": 78400
},
{
"epoch": 0.6855064430069004,
"grad_norm": 0.21255306899547577,
"learning_rate": 0.001,
"loss": 3.0870245361328124,
"step": 78450
},
{
"epoch": 0.6859433495990017,
"grad_norm": 0.2075134813785553,
"learning_rate": 0.001,
"loss": 3.088641357421875,
"step": 78500
},
{
"epoch": 0.686380256191103,
"grad_norm": 0.21092812716960907,
"learning_rate": 0.001,
"loss": 3.0882333374023436,
"step": 78550
},
{
"epoch": 0.6868171627832043,
"grad_norm": 0.21339748799800873,
"learning_rate": 0.001,
"loss": 3.0819876098632815,
"step": 78600
},
{
"epoch": 0.6872540693753055,
"grad_norm": 0.22137713432312012,
"learning_rate": 0.001,
"loss": 3.0866964721679686,
"step": 78650
},
{
"epoch": 0.6876909759674068,
"grad_norm": 0.19947455823421478,
"learning_rate": 0.001,
"loss": 3.0873593139648436,
"step": 78700
},
{
"epoch": 0.6881278825595081,
"grad_norm": 0.24083974957466125,
"learning_rate": 0.001,
"loss": 3.082724914550781,
"step": 78750
},
{
"epoch": 0.6885647891516093,
"grad_norm": 0.19919462502002716,
"learning_rate": 0.001,
"loss": 3.0806884765625,
"step": 78800
},
{
"epoch": 0.6890016957437106,
"grad_norm": 0.20141293108463287,
"learning_rate": 0.001,
"loss": 3.08558349609375,
"step": 78850
},
{
"epoch": 0.6894386023358119,
"grad_norm": 0.22001996636390686,
"learning_rate": 0.001,
"loss": 3.083612976074219,
"step": 78900
},
{
"epoch": 0.6898755089279132,
"grad_norm": 0.18413425981998444,
"learning_rate": 0.001,
"loss": 3.0830377197265624,
"step": 78950
},
{
"epoch": 0.6903124155200144,
"grad_norm": 0.21403107047080994,
"learning_rate": 0.001,
"loss": 3.0788504028320314,
"step": 79000
},
{
"epoch": 0.6907493221121157,
"grad_norm": 0.1979919970035553,
"learning_rate": 0.001,
"loss": 3.0873709106445313,
"step": 79050
},
{
"epoch": 0.691186228704217,
"grad_norm": 0.2356664389371872,
"learning_rate": 0.001,
"loss": 3.0828213500976562,
"step": 79100
},
{
"epoch": 0.6916231352963182,
"grad_norm": 0.1995965838432312,
"learning_rate": 0.001,
"loss": 3.0796002197265624,
"step": 79150
},
{
"epoch": 0.6920600418884195,
"grad_norm": 0.20483635365962982,
"learning_rate": 0.001,
"loss": 3.0888031005859373,
"step": 79200
},
{
"epoch": 0.6924969484805208,
"grad_norm": 0.20581825077533722,
"learning_rate": 0.001,
"loss": 3.0827951049804687,
"step": 79250
},
{
"epoch": 0.6929338550726221,
"grad_norm": 0.20429953932762146,
"learning_rate": 0.001,
"loss": 3.08158935546875,
"step": 79300
},
{
"epoch": 0.6933707616647233,
"grad_norm": 0.21808390319347382,
"learning_rate": 0.001,
"loss": 3.083011474609375,
"step": 79350
},
{
"epoch": 0.6938076682568246,
"grad_norm": 0.2212456613779068,
"learning_rate": 0.001,
"loss": 3.082039489746094,
"step": 79400
},
{
"epoch": 0.6942445748489259,
"grad_norm": 0.19717669486999512,
"learning_rate": 0.001,
"loss": 3.086158447265625,
"step": 79450
},
{
"epoch": 0.6946814814410271,
"grad_norm": 0.21572242677211761,
"learning_rate": 0.001,
"loss": 3.0856365966796875,
"step": 79500
},
{
"epoch": 0.6951183880331284,
"grad_norm": 0.1930527538061142,
"learning_rate": 0.001,
"loss": 3.0857278442382814,
"step": 79550
},
{
"epoch": 0.6955552946252297,
"grad_norm": 0.2112765610218048,
"learning_rate": 0.001,
"loss": 3.0875479125976564,
"step": 79600
},
{
"epoch": 0.695992201217331,
"grad_norm": 0.23278924822807312,
"learning_rate": 0.001,
"loss": 3.0846603393554686,
"step": 79650
},
{
"epoch": 0.6964291078094322,
"grad_norm": 0.18906286358833313,
"learning_rate": 0.001,
"loss": 3.0797198486328123,
"step": 79700
},
{
"epoch": 0.6968660144015335,
"grad_norm": 0.22060179710388184,
"learning_rate": 0.001,
"loss": 3.083451843261719,
"step": 79750
},
{
"epoch": 0.6973029209936348,
"grad_norm": 0.20691688358783722,
"learning_rate": 0.001,
"loss": 3.0809426879882813,
"step": 79800
},
{
"epoch": 0.697739827585736,
"grad_norm": 0.23333010077476501,
"learning_rate": 0.001,
"loss": 3.0865985107421876,
"step": 79850
},
{
"epoch": 0.6981767341778373,
"grad_norm": 0.21768400073051453,
"learning_rate": 0.001,
"loss": 3.0818386840820313,
"step": 79900
},
{
"epoch": 0.6986136407699386,
"grad_norm": 0.20732533931732178,
"learning_rate": 0.001,
"loss": 3.0867465209960936,
"step": 79950
},
{
"epoch": 0.6990505473620399,
"grad_norm": 0.2105298489332199,
"learning_rate": 0.001,
"loss": 3.0867990112304686,
"step": 80000
},
{
"epoch": 0.6994874539541412,
"grad_norm": 0.22331522405147552,
"learning_rate": 0.001,
"loss": 3.089678955078125,
"step": 80050
},
{
"epoch": 0.6999243605462425,
"grad_norm": 0.20896968245506287,
"learning_rate": 0.001,
"loss": 3.0842153930664065,
"step": 80100
},
{
"epoch": 0.7003612671383438,
"grad_norm": 0.19243353605270386,
"learning_rate": 0.001,
"loss": 3.0897247314453127,
"step": 80150
},
{
"epoch": 0.7007981737304451,
"grad_norm": 0.24495665729045868,
"learning_rate": 0.001,
"loss": 3.085721435546875,
"step": 80200
},
{
"epoch": 0.7012350803225463,
"grad_norm": 0.21539324522018433,
"learning_rate": 0.001,
"loss": 3.0914495849609374,
"step": 80250
},
{
"epoch": 0.7016719869146476,
"grad_norm": 0.23662789165973663,
"learning_rate": 0.001,
"loss": 3.0820404052734376,
"step": 80300
},
{
"epoch": 0.7021088935067489,
"grad_norm": 0.22640351951122284,
"learning_rate": 0.001,
"loss": 3.0834283447265625,
"step": 80350
},
{
"epoch": 0.7025458000988501,
"grad_norm": 0.2433343380689621,
"learning_rate": 0.001,
"loss": 3.0828851318359374,
"step": 80400
},
{
"epoch": 0.7029827066909514,
"grad_norm": 0.2037147581577301,
"learning_rate": 0.001,
"loss": 3.0873077392578123,
"step": 80450
},
{
"epoch": 0.7034196132830527,
"grad_norm": 0.20801591873168945,
"learning_rate": 0.001,
"loss": 3.0811553955078126,
"step": 80500
},
{
"epoch": 0.703856519875154,
"grad_norm": 0.21396999061107635,
"learning_rate": 0.001,
"loss": 3.0776361083984374,
"step": 80550
},
{
"epoch": 0.7042934264672552,
"grad_norm": 0.24520927667617798,
"learning_rate": 0.001,
"loss": 3.0767333984375,
"step": 80600
},
{
"epoch": 0.7047303330593565,
"grad_norm": 0.20743826031684875,
"learning_rate": 0.001,
"loss": 3.0833837890625,
"step": 80650
},
{
"epoch": 0.7051672396514578,
"grad_norm": 0.21662896871566772,
"learning_rate": 0.001,
"loss": 3.0757296752929686,
"step": 80700
},
{
"epoch": 0.705604146243559,
"grad_norm": 0.20927999913692474,
"learning_rate": 0.001,
"loss": 3.081734313964844,
"step": 80750
},
{
"epoch": 0.7060410528356603,
"grad_norm": 0.1927129179239273,
"learning_rate": 0.001,
"loss": 3.084912109375,
"step": 80800
},
{
"epoch": 0.7064779594277616,
"grad_norm": 0.21988406777381897,
"learning_rate": 0.001,
"loss": 3.078038635253906,
"step": 80850
},
{
"epoch": 0.7069148660198629,
"grad_norm": 0.20275036990642548,
"learning_rate": 0.001,
"loss": 3.0820074462890625,
"step": 80900
},
{
"epoch": 0.7073517726119641,
"grad_norm": 0.19036343693733215,
"learning_rate": 0.001,
"loss": 3.0888934326171875,
"step": 80950
},
{
"epoch": 0.7077886792040654,
"grad_norm": 0.22706688940525055,
"learning_rate": 0.001,
"loss": 3.0831027221679688,
"step": 81000
},
{
"epoch": 0.7077886792040654,
"eval_loss": 3.0957136154174805,
"eval_runtime": 5.5301,
"eval_samples_per_second": 176.487,
"eval_steps_per_second": 11.03,
"step": 81000
},
{
"epoch": 0.7082255857961667,
"grad_norm": 0.2148069590330124,
"learning_rate": 0.001,
"loss": 3.0751715087890625,
"step": 81050
},
{
"epoch": 0.708662492388268,
"grad_norm": 0.20097805559635162,
"learning_rate": 0.001,
"loss": 3.086511535644531,
"step": 81100
},
{
"epoch": 0.7090993989803692,
"grad_norm": 0.2104441225528717,
"learning_rate": 0.001,
"loss": 3.0861026000976564,
"step": 81150
},
{
"epoch": 0.7095363055724705,
"grad_norm": 0.18840894103050232,
"learning_rate": 0.001,
"loss": 3.0872503662109376,
"step": 81200
},
{
"epoch": 0.7099732121645718,
"grad_norm": 0.20789970457553864,
"learning_rate": 0.001,
"loss": 3.0822943115234374,
"step": 81250
},
{
"epoch": 0.710410118756673,
"grad_norm": 0.21030528843402863,
"learning_rate": 0.001,
"loss": 3.085260009765625,
"step": 81300
},
{
"epoch": 0.7108470253487743,
"grad_norm": 0.22340913116931915,
"learning_rate": 0.001,
"loss": 3.0874005126953126,
"step": 81350
},
{
"epoch": 0.7112839319408756,
"grad_norm": 0.2062855213880539,
"learning_rate": 0.001,
"loss": 3.0757476806640627,
"step": 81400
},
{
"epoch": 0.7117208385329769,
"grad_norm": 0.24663305282592773,
"learning_rate": 0.001,
"loss": 3.082265625,
"step": 81450
},
{
"epoch": 0.7121577451250781,
"grad_norm": 0.21674543619155884,
"learning_rate": 0.001,
"loss": 3.084146728515625,
"step": 81500
},
{
"epoch": 0.7125946517171794,
"grad_norm": 0.22249560058116913,
"learning_rate": 0.001,
"loss": 3.0836370849609374,
"step": 81550
},
{
"epoch": 0.7130315583092807,
"grad_norm": 0.204384446144104,
"learning_rate": 0.001,
"loss": 3.0854071044921874,
"step": 81600
},
{
"epoch": 0.7134684649013819,
"grad_norm": 0.23136423528194427,
"learning_rate": 0.001,
"loss": 3.0866470336914062,
"step": 81650
},
{
"epoch": 0.7139053714934833,
"grad_norm": 0.2122517228126526,
"learning_rate": 0.001,
"loss": 3.0826141357421877,
"step": 81700
},
{
"epoch": 0.7143422780855846,
"grad_norm": 0.261821448802948,
"learning_rate": 0.001,
"loss": 3.0813162231445315,
"step": 81750
},
{
"epoch": 0.7147791846776859,
"grad_norm": 0.21322719752788544,
"learning_rate": 0.001,
"loss": 3.0806265258789063,
"step": 81800
},
{
"epoch": 0.7152160912697871,
"grad_norm": 0.22185467183589935,
"learning_rate": 0.001,
"loss": 3.084389953613281,
"step": 81850
},
{
"epoch": 0.7156529978618884,
"grad_norm": 0.20002132654190063,
"learning_rate": 0.001,
"loss": 3.0820858764648436,
"step": 81900
},
{
"epoch": 0.7160899044539897,
"grad_norm": 0.1880277395248413,
"learning_rate": 0.001,
"loss": 3.078172302246094,
"step": 81950
},
{
"epoch": 0.716526811046091,
"grad_norm": 0.21600885689258575,
"learning_rate": 0.001,
"loss": 3.0811639404296876,
"step": 82000
},
{
"epoch": 0.7169637176381922,
"grad_norm": 0.21108900010585785,
"learning_rate": 0.001,
"loss": 3.086716613769531,
"step": 82050
},
{
"epoch": 0.7174006242302935,
"grad_norm": 0.22212134301662445,
"learning_rate": 0.001,
"loss": 3.0821343994140626,
"step": 82100
},
{
"epoch": 0.7178375308223948,
"grad_norm": 0.1969502866268158,
"learning_rate": 0.001,
"loss": 3.0883306884765624,
"step": 82150
},
{
"epoch": 0.718274437414496,
"grad_norm": 0.20928046107292175,
"learning_rate": 0.001,
"loss": 3.08385498046875,
"step": 82200
},
{
"epoch": 0.7187113440065973,
"grad_norm": 0.202848881483078,
"learning_rate": 0.001,
"loss": 3.0766571044921873,
"step": 82250
},
{
"epoch": 0.7191482505986986,
"grad_norm": 0.208697110414505,
"learning_rate": 0.001,
"loss": 3.0784698486328126,
"step": 82300
},
{
"epoch": 0.7195851571907999,
"grad_norm": 0.22407536208629608,
"learning_rate": 0.001,
"loss": 3.0895095825195313,
"step": 82350
},
{
"epoch": 0.7200220637829011,
"grad_norm": 0.21066206693649292,
"learning_rate": 0.001,
"loss": 3.0842730712890627,
"step": 82400
},
{
"epoch": 0.7204589703750024,
"grad_norm": 0.21734683215618134,
"learning_rate": 0.001,
"loss": 3.0775485229492188,
"step": 82450
},
{
"epoch": 0.7208958769671037,
"grad_norm": 0.19409386813640594,
"learning_rate": 0.001,
"loss": 3.0808538818359374,
"step": 82500
},
{
"epoch": 0.7213327835592049,
"grad_norm": 0.19656510651111603,
"learning_rate": 0.001,
"loss": 3.07595458984375,
"step": 82550
},
{
"epoch": 0.7217696901513062,
"grad_norm": 0.195686474442482,
"learning_rate": 0.001,
"loss": 3.0859234619140623,
"step": 82600
},
{
"epoch": 0.7222065967434075,
"grad_norm": 0.19736018776893616,
"learning_rate": 0.001,
"loss": 3.0833880615234377,
"step": 82650
},
{
"epoch": 0.7226435033355088,
"grad_norm": 0.1794862300157547,
"learning_rate": 0.001,
"loss": 3.0814175415039062,
"step": 82700
},
{
"epoch": 0.72308040992761,
"grad_norm": 0.199075847864151,
"learning_rate": 0.001,
"loss": 3.0851104736328123,
"step": 82750
},
{
"epoch": 0.7235173165197113,
"grad_norm": 0.2098972499370575,
"learning_rate": 0.001,
"loss": 3.091060791015625,
"step": 82800
},
{
"epoch": 0.7239542231118126,
"grad_norm": 0.18432964384555817,
"learning_rate": 0.001,
"loss": 3.086722106933594,
"step": 82850
},
{
"epoch": 0.7243911297039138,
"grad_norm": 0.21974876523017883,
"learning_rate": 0.001,
"loss": 3.0850701904296876,
"step": 82900
},
{
"epoch": 0.7248280362960151,
"grad_norm": 0.21936289966106415,
"learning_rate": 0.001,
"loss": 3.0796194458007813,
"step": 82950
},
{
"epoch": 0.7252649428881164,
"grad_norm": 0.21287301182746887,
"learning_rate": 0.001,
"loss": 3.085557556152344,
"step": 83000
},
{
"epoch": 0.7257018494802177,
"grad_norm": 0.1973222941160202,
"learning_rate": 0.001,
"loss": 3.0812753295898436,
"step": 83050
},
{
"epoch": 0.7261387560723189,
"grad_norm": 0.22154083847999573,
"learning_rate": 0.001,
"loss": 3.075750427246094,
"step": 83100
},
{
"epoch": 0.7265756626644202,
"grad_norm": 0.20486560463905334,
"learning_rate": 0.001,
"loss": 3.083943176269531,
"step": 83150
},
{
"epoch": 0.7270125692565215,
"grad_norm": 0.20191390812397003,
"learning_rate": 0.001,
"loss": 3.0908135986328125,
"step": 83200
},
{
"epoch": 0.7274494758486227,
"grad_norm": 0.1976170390844345,
"learning_rate": 0.001,
"loss": 3.08345947265625,
"step": 83250
},
{
"epoch": 0.727886382440724,
"grad_norm": 0.20339162647724152,
"learning_rate": 0.001,
"loss": 3.0777151489257815,
"step": 83300
},
{
"epoch": 0.7283232890328254,
"grad_norm": 0.21782030165195465,
"learning_rate": 0.001,
"loss": 3.0854791259765624,
"step": 83350
},
{
"epoch": 0.7287601956249267,
"grad_norm": 0.20211881399154663,
"learning_rate": 0.001,
"loss": 3.0840704345703127,
"step": 83400
},
{
"epoch": 0.7291971022170279,
"grad_norm": 0.21823638677597046,
"learning_rate": 0.001,
"loss": 3.0815493774414064,
"step": 83450
},
{
"epoch": 0.7296340088091292,
"grad_norm": 0.2063329517841339,
"learning_rate": 0.001,
"loss": 3.0879898071289062,
"step": 83500
},
{
"epoch": 0.7300709154012305,
"grad_norm": 0.19135266542434692,
"learning_rate": 0.001,
"loss": 3.082502746582031,
"step": 83550
},
{
"epoch": 0.7305078219933318,
"grad_norm": 0.22470586001873016,
"learning_rate": 0.001,
"loss": 3.080158996582031,
"step": 83600
},
{
"epoch": 0.730944728585433,
"grad_norm": 0.21016673743724823,
"learning_rate": 0.001,
"loss": 3.0833810424804686,
"step": 83650
},
{
"epoch": 0.7313816351775343,
"grad_norm": 0.2054825723171234,
"learning_rate": 0.001,
"loss": 3.0875494384765627,
"step": 83700
},
{
"epoch": 0.7318185417696356,
"grad_norm": 0.2236102670431137,
"learning_rate": 0.001,
"loss": 3.080935363769531,
"step": 83750
},
{
"epoch": 0.7322554483617368,
"grad_norm": 0.19892160594463348,
"learning_rate": 0.001,
"loss": 3.0884017944335938,
"step": 83800
},
{
"epoch": 0.7326923549538381,
"grad_norm": 0.21159380674362183,
"learning_rate": 0.001,
"loss": 3.0795269775390626,
"step": 83850
},
{
"epoch": 0.7331292615459394,
"grad_norm": 0.21392063796520233,
"learning_rate": 0.001,
"loss": 3.091121826171875,
"step": 83900
},
{
"epoch": 0.7335661681380407,
"grad_norm": 0.18876802921295166,
"learning_rate": 0.001,
"loss": 3.0859036254882812,
"step": 83950
},
{
"epoch": 0.7340030747301419,
"grad_norm": 0.23087337613105774,
"learning_rate": 0.001,
"loss": 3.082529296875,
"step": 84000
},
{
"epoch": 0.7340030747301419,
"eval_loss": 3.094377279281616,
"eval_runtime": 6.0242,
"eval_samples_per_second": 162.014,
"eval_steps_per_second": 10.126,
"step": 84000
}
],
"logging_steps": 50,
"max_steps": 114440,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 3000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 9.977509139447808e+18,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}