4315 lines
104 KiB
JSON
4315 lines
104 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.26214395526076495,
|
|
"eval_steps": 3000,
|
|
"global_step": 30000,
|
|
"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
|
|
}
|
|
],
|
|
"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": 3.56339612123136e+18,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|