Files
Caracal_GPT/checkpoints/checkpoint-86520/trainer_state.json

12144 lines
277 KiB
JSON
Raw Permalink Normal View History

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.999988442045527,
"eval_steps": 500,
"global_step": 86520,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.001155795447321733,
"grad_norm": 14.8125,
"learning_rate": 2e-05,
"loss": 5.7862,
"step": 50
},
{
"epoch": 0.002311590894643466,
"grad_norm": 13.4375,
"learning_rate": 4e-05,
"loss": 5.2432,
"step": 100
},
{
"epoch": 0.003467386341965199,
"grad_norm": 11.5,
"learning_rate": 6e-05,
"loss": 4.905,
"step": 150
},
{
"epoch": 0.004623181789286932,
"grad_norm": 10.9375,
"learning_rate": 8e-05,
"loss": 4.6741,
"step": 200
},
{
"epoch": 0.005778977236608665,
"grad_norm": 11.0,
"learning_rate": 0.0001,
"loss": 4.5555,
"step": 250
},
{
"epoch": 0.006934772683930398,
"grad_norm": 10.1875,
"learning_rate": 0.00012,
"loss": 4.6028,
"step": 300
},
{
"epoch": 0.008090568131252132,
"grad_norm": 9.8125,
"learning_rate": 0.00014,
"loss": 4.704,
"step": 350
},
{
"epoch": 0.009246363578573865,
"grad_norm": 9.875,
"learning_rate": 0.00016,
"loss": 4.697,
"step": 400
},
{
"epoch": 0.010402159025895598,
"grad_norm": 9.4375,
"learning_rate": 0.00018,
"loss": 4.7582,
"step": 450
},
{
"epoch": 0.01155795447321733,
"grad_norm": 9.5,
"learning_rate": 0.0002,
"loss": 4.738,
"step": 500
},
{
"epoch": 0.012713749920539063,
"grad_norm": 9.4375,
"learning_rate": 0.00019999992618459312,
"loss": 4.7601,
"step": 550
},
{
"epoch": 0.013869545367860796,
"grad_norm": 9.4375,
"learning_rate": 0.0001999997047384814,
"loss": 4.7854,
"step": 600
},
{
"epoch": 0.015025340815182529,
"grad_norm": 8.75,
"learning_rate": 0.00019999933566199177,
"loss": 4.7152,
"step": 650
},
{
"epoch": 0.016181136262504264,
"grad_norm": 8.1875,
"learning_rate": 0.00019999881895566907,
"loss": 4.6014,
"step": 700
},
{
"epoch": 0.017336931709825996,
"grad_norm": 9.5625,
"learning_rate": 0.0001999981546202762,
"loss": 4.5855,
"step": 750
},
{
"epoch": 0.01849272715714773,
"grad_norm": 7.53125,
"learning_rate": 0.00019999734265679385,
"loss": 4.5282,
"step": 800
},
{
"epoch": 0.019648522604469462,
"grad_norm": 8.125,
"learning_rate": 0.0001999963830664208,
"loss": 4.5121,
"step": 850
},
{
"epoch": 0.020804318051791195,
"grad_norm": 7.8125,
"learning_rate": 0.0001999952758505736,
"loss": 4.4262,
"step": 900
},
{
"epoch": 0.021960113499112928,
"grad_norm": 7.84375,
"learning_rate": 0.00019999402101088693,
"loss": 4.3959,
"step": 950
},
{
"epoch": 0.02311590894643466,
"grad_norm": 7.1875,
"learning_rate": 0.00019999261854921328,
"loss": 4.3384,
"step": 1000
},
{
"epoch": 0.024271704393756394,
"grad_norm": 7.46875,
"learning_rate": 0.00019999106846762312,
"loss": 4.3487,
"step": 1050
},
{
"epoch": 0.025427499841078127,
"grad_norm": 7.125,
"learning_rate": 0.00019998937076840488,
"loss": 4.3253,
"step": 1100
},
{
"epoch": 0.02658329528839986,
"grad_norm": 7.75,
"learning_rate": 0.00019998752545406486,
"loss": 4.329,
"step": 1150
},
{
"epoch": 0.027739090735721592,
"grad_norm": 7.0625,
"learning_rate": 0.00019998553252732727,
"loss": 4.3254,
"step": 1200
},
{
"epoch": 0.028894886183043325,
"grad_norm": 8.375,
"learning_rate": 0.00019998339199113433,
"loss": 4.0993,
"step": 1250
},
{
"epoch": 0.030050681630365058,
"grad_norm": 7.40625,
"learning_rate": 0.00019998110384864614,
"loss": 4.2106,
"step": 1300
},
{
"epoch": 0.03120647707768679,
"grad_norm": 7.28125,
"learning_rate": 0.00019997866810324074,
"loss": 4.2057,
"step": 1350
},
{
"epoch": 0.03236227252500853,
"grad_norm": 7.375,
"learning_rate": 0.00019997608475851394,
"loss": 4.1795,
"step": 1400
},
{
"epoch": 0.03351806797233026,
"grad_norm": 7.34375,
"learning_rate": 0.00019997335381827957,
"loss": 4.0939,
"step": 1450
},
{
"epoch": 0.03467386341965199,
"grad_norm": 6.84375,
"learning_rate": 0.00019997047528656943,
"loss": 4.1456,
"step": 1500
},
{
"epoch": 0.03582965886697372,
"grad_norm": 7.59375,
"learning_rate": 0.00019996744916763305,
"loss": 4.1742,
"step": 1550
},
{
"epoch": 0.03698545431429546,
"grad_norm": 8.625,
"learning_rate": 0.00019996427546593791,
"loss": 4.0882,
"step": 1600
},
{
"epoch": 0.03814124976161719,
"grad_norm": 7.375,
"learning_rate": 0.00019996095418616943,
"loss": 4.0693,
"step": 1650
},
{
"epoch": 0.039297045208938924,
"grad_norm": 7.3125,
"learning_rate": 0.00019995748533323075,
"loss": 4.0422,
"step": 1700
},
{
"epoch": 0.040452840656260654,
"grad_norm": 7.9375,
"learning_rate": 0.00019995386891224305,
"loss": 4.134,
"step": 1750
},
{
"epoch": 0.04160863610358239,
"grad_norm": 7.875,
"learning_rate": 0.00019995010492854522,
"loss": 4.018,
"step": 1800
},
{
"epoch": 0.04276443155090412,
"grad_norm": 7.375,
"learning_rate": 0.00019994619338769412,
"loss": 4.0152,
"step": 1850
},
{
"epoch": 0.043920226998225856,
"grad_norm": 7.65625,
"learning_rate": 0.00019994213429546433,
"loss": 3.9747,
"step": 1900
},
{
"epoch": 0.045076022445547585,
"grad_norm": 8.125,
"learning_rate": 0.00019993792765784837,
"loss": 4.0306,
"step": 1950
},
{
"epoch": 0.04623181789286932,
"grad_norm": 8.1875,
"learning_rate": 0.00019993357348105652,
"loss": 3.9736,
"step": 2000
},
{
"epoch": 0.04738761334019105,
"grad_norm": 7.0,
"learning_rate": 0.00019992907177151685,
"loss": 3.9079,
"step": 2050
},
{
"epoch": 0.04854340878751279,
"grad_norm": 7.65625,
"learning_rate": 0.0001999244225358753,
"loss": 3.8878,
"step": 2100
},
{
"epoch": 0.04969920423483452,
"grad_norm": 7.90625,
"learning_rate": 0.0001999196257809956,
"loss": 3.988,
"step": 2150
},
{
"epoch": 0.05085499968215625,
"grad_norm": 7.28125,
"learning_rate": 0.0001999146815139592,
"loss": 3.9289,
"step": 2200
},
{
"epoch": 0.05201079512947798,
"grad_norm": 7.78125,
"learning_rate": 0.0001999095897420654,
"loss": 3.8404,
"step": 2250
},
{
"epoch": 0.05316659057679972,
"grad_norm": 7.4375,
"learning_rate": 0.00019990435047283117,
"loss": 3.885,
"step": 2300
},
{
"epoch": 0.05432238602412145,
"grad_norm": 7.90625,
"learning_rate": 0.0001998989637139913,
"loss": 3.8803,
"step": 2350
},
{
"epoch": 0.055478181471443185,
"grad_norm": 7.78125,
"learning_rate": 0.00019989342947349833,
"loss": 3.8668,
"step": 2400
},
{
"epoch": 0.056633976918764914,
"grad_norm": 7.84375,
"learning_rate": 0.0001998877477595225,
"loss": 3.872,
"step": 2450
},
{
"epoch": 0.05778977236608665,
"grad_norm": 6.90625,
"learning_rate": 0.00019988191858045178,
"loss": 3.8193,
"step": 2500
},
{
"epoch": 0.05894556781340838,
"grad_norm": 7.5625,
"learning_rate": 0.0001998759419448918,
"loss": 3.7731,
"step": 2550
},
{
"epoch": 0.060101363260730116,
"grad_norm": 7.96875,
"learning_rate": 0.00019986981786166593,
"loss": 3.7634,
"step": 2600
},
{
"epoch": 0.06125715870805185,
"grad_norm": 7.78125,
"learning_rate": 0.0001998635463398152,
"loss": 3.7882,
"step": 2650
},
{
"epoch": 0.06241295415537358,
"grad_norm": 7.375,
"learning_rate": 0.00019985712738859835,
"loss": 3.7735,
"step": 2700
},
{
"epoch": 0.06356874960269532,
"grad_norm": 7.59375,
"learning_rate": 0.00019985056101749166,
"loss": 3.7398,
"step": 2750
},
{
"epoch": 0.06472454505001705,
"grad_norm": 7.53125,
"learning_rate": 0.00019984384723618917,
"loss": 3.816,
"step": 2800
},
{
"epoch": 0.06588034049733878,
"grad_norm": 7.375,
"learning_rate": 0.00019983698605460247,
"loss": 3.7191,
"step": 2850
},
{
"epoch": 0.06703613594466051,
"grad_norm": 7.84375,
"learning_rate": 0.00019982997748286082,
"loss": 3.7606,
"step": 2900
},
{
"epoch": 0.06819193139198225,
"grad_norm": 8.0625,
"learning_rate": 0.00019982282153131097,
"loss": 3.7214,
"step": 2950
},
{
"epoch": 0.06934772683930399,
"grad_norm": 7.1875,
"learning_rate": 0.0001998155182105173,
"loss": 3.7548,
"step": 3000
},
{
"epoch": 0.07050352228662571,
"grad_norm": 7.59375,
"learning_rate": 0.00019980806753126183,
"loss": 3.7671,
"step": 3050
},
{
"epoch": 0.07165931773394744,
"grad_norm": 7.875,
"learning_rate": 0.00019980046950454403,
"loss": 3.6935,
"step": 3100
},
{
"epoch": 0.07281511318126918,
"grad_norm": 7.6875,
"learning_rate": 0.0001997927241415809,
"loss": 3.7336,
"step": 3150
},
{
"epoch": 0.07397090862859092,
"grad_norm": 7.125,
"learning_rate": 0.00019978483145380703,
"loss": 3.6674,
"step": 3200
},
{
"epoch": 0.07512670407591264,
"grad_norm": 7.28125,
"learning_rate": 0.0001997767914528744,
"loss": 3.7334,
"step": 3250
},
{
"epoch": 0.07628249952323438,
"grad_norm": 7.09375,
"learning_rate": 0.00019976860415065256,
"loss": 3.6719,
"step": 3300
},
{
"epoch": 0.07743829497055611,
"grad_norm": 6.90625,
"learning_rate": 0.00019976026955922853,
"loss": 3.7022,
"step": 3350
},
{
"epoch": 0.07859409041787785,
"grad_norm": 7.5625,
"learning_rate": 0.0001997517876909067,
"loss": 3.6312,
"step": 3400
},
{
"epoch": 0.07974988586519957,
"grad_norm": 7.5625,
"learning_rate": 0.00019974315855820888,
"loss": 3.6595,
"step": 3450
},
{
"epoch": 0.08090568131252131,
"grad_norm": 7.59375,
"learning_rate": 0.0001997343821738744,
"loss": 3.627,
"step": 3500
},
{
"epoch": 0.08206147675984304,
"grad_norm": 7.1875,
"learning_rate": 0.0001997254585508599,
"loss": 3.6572,
"step": 3550
},
{
"epoch": 0.08321727220716478,
"grad_norm": 7.3125,
"learning_rate": 0.00019971638770233932,
"loss": 3.6212,
"step": 3600
},
{
"epoch": 0.0843730676544865,
"grad_norm": 7.3125,
"learning_rate": 0.00019970716964170412,
"loss": 3.6135,
"step": 3650
},
{
"epoch": 0.08552886310180824,
"grad_norm": 8.1875,
"learning_rate": 0.00019969780438256293,
"loss": 3.5601,
"step": 3700
},
{
"epoch": 0.08668465854912998,
"grad_norm": 7.375,
"learning_rate": 0.00019968829193874182,
"loss": 3.6298,
"step": 3750
},
{
"epoch": 0.08784045399645171,
"grad_norm": 7.90625,
"learning_rate": 0.00019967863232428403,
"loss": 3.6493,
"step": 3800
},
{
"epoch": 0.08899624944377343,
"grad_norm": 7.96875,
"learning_rate": 0.00019966882555345013,
"loss": 3.6357,
"step": 3850
},
{
"epoch": 0.09015204489109517,
"grad_norm": 6.84375,
"learning_rate": 0.000199658871640718,
"loss": 3.6406,
"step": 3900
},
{
"epoch": 0.0913078403384169,
"grad_norm": 7.46875,
"learning_rate": 0.00019964877060078267,
"loss": 3.522,
"step": 3950
},
{
"epoch": 0.09246363578573864,
"grad_norm": 8.25,
"learning_rate": 0.0001996385224485563,
"loss": 3.6441,
"step": 4000
},
{
"epoch": 0.09361943123306038,
"grad_norm": 7.59375,
"learning_rate": 0.00019962812719916841,
"loss": 3.6021,
"step": 4050
},
{
"epoch": 0.0947752266803821,
"grad_norm": 6.96875,
"learning_rate": 0.00019961758486796555,
"loss": 3.5959,
"step": 4100
},
{
"epoch": 0.09593102212770384,
"grad_norm": 7.6875,
"learning_rate": 0.00019960689547051142,
"loss": 3.6222,
"step": 4150
},
{
"epoch": 0.09708681757502557,
"grad_norm": 7.3125,
"learning_rate": 0.00019959605902258697,
"loss": 3.5324,
"step": 4200
},
{
"epoch": 0.09824261302234731,
"grad_norm": 7.375,
"learning_rate": 0.00019958507554019005,
"loss": 3.5327,
"step": 4250
},
{
"epoch": 0.09939840846966903,
"grad_norm": 6.8125,
"learning_rate": 0.00019957394503953568,
"loss": 3.5488,
"step": 4300
},
{
"epoch": 0.10055420391699077,
"grad_norm": 6.96875,
"learning_rate": 0.00019956266753705593,
"loss": 3.5541,
"step": 4350
},
{
"epoch": 0.1017099993643125,
"grad_norm": 6.8125,
"learning_rate": 0.0001995512430493998,
"loss": 3.5331,
"step": 4400
},
{
"epoch": 0.10286579481163424,
"grad_norm": 7.34375,
"learning_rate": 0.00019953967159343344,
"loss": 3.5906,
"step": 4450
},
{
"epoch": 0.10402159025895596,
"grad_norm": 7.4375,
"learning_rate": 0.00019952795318623986,
"loss": 3.5671,
"step": 4500
},
{
"epoch": 0.1051773857062777,
"grad_norm": 7.71875,
"learning_rate": 0.00019951608784511904,
"loss": 3.505,
"step": 4550
},
{
"epoch": 0.10633318115359944,
"grad_norm": 7.03125,
"learning_rate": 0.00019950407558758784,
"loss": 3.4847,
"step": 4600
},
{
"epoch": 0.10748897660092117,
"grad_norm": 7.3125,
"learning_rate": 0.00019949191643138013,
"loss": 3.5246,
"step": 4650
},
{
"epoch": 0.1086447720482429,
"grad_norm": 7.0625,
"learning_rate": 0.00019947961039444646,
"loss": 3.4638,
"step": 4700
},
{
"epoch": 0.10980056749556463,
"grad_norm": 6.5625,
"learning_rate": 0.00019946715749495444,
"loss": 3.5223,
"step": 4750
},
{
"epoch": 0.11095636294288637,
"grad_norm": 7.34375,
"learning_rate": 0.00019945455775128833,
"loss": 3.5654,
"step": 4800
},
{
"epoch": 0.1121121583902081,
"grad_norm": 8.4375,
"learning_rate": 0.00019944181118204928,
"loss": 3.5376,
"step": 4850
},
{
"epoch": 0.11326795383752983,
"grad_norm": 6.75,
"learning_rate": 0.00019942891780605512,
"loss": 3.5008,
"step": 4900
},
{
"epoch": 0.11442374928485156,
"grad_norm": 7.875,
"learning_rate": 0.0001994158776423404,
"loss": 3.463,
"step": 4950
},
{
"epoch": 0.1155795447321733,
"grad_norm": 7.6875,
"learning_rate": 0.0001994026907101565,
"loss": 3.4199,
"step": 5000
},
{
"epoch": 0.11673534017949504,
"grad_norm": 7.46875,
"learning_rate": 0.00019938935702897137,
"loss": 3.416,
"step": 5050
},
{
"epoch": 0.11789113562681676,
"grad_norm": 6.4375,
"learning_rate": 0.00019937587661846958,
"loss": 3.4747,
"step": 5100
},
{
"epoch": 0.1190469310741385,
"grad_norm": 7.34375,
"learning_rate": 0.00019936224949855248,
"loss": 3.4771,
"step": 5150
},
{
"epoch": 0.12020272652146023,
"grad_norm": 7.59375,
"learning_rate": 0.0001993484756893378,
"loss": 3.4427,
"step": 5200
},
{
"epoch": 0.12135852196878197,
"grad_norm": 7.125,
"learning_rate": 0.00019933455521116,
"loss": 3.4939,
"step": 5250
},
{
"epoch": 0.1225143174161037,
"grad_norm": 7.53125,
"learning_rate": 0.0001993204880845699,
"loss": 3.4462,
"step": 5300
},
{
"epoch": 0.12367011286342543,
"grad_norm": 7.78125,
"learning_rate": 0.00019930627433033498,
"loss": 3.4467,
"step": 5350
},
{
"epoch": 0.12482590831074716,
"grad_norm": 8.25,
"learning_rate": 0.00019929191396943913,
"loss": 3.4436,
"step": 5400
},
{
"epoch": 0.12598170375806889,
"grad_norm": 7.84375,
"learning_rate": 0.00019927740702308265,
"loss": 3.4574,
"step": 5450
},
{
"epoch": 0.12713749920539064,
"grad_norm": 7.21875,
"learning_rate": 0.00019926275351268223,
"loss": 3.449,
"step": 5500
},
{
"epoch": 0.12829329465271236,
"grad_norm": 7.96875,
"learning_rate": 0.000199247953459871,
"loss": 3.4157,
"step": 5550
},
{
"epoch": 0.1294490901000341,
"grad_norm": 6.625,
"learning_rate": 0.00019923300688649842,
"loss": 3.4196,
"step": 5600
},
{
"epoch": 0.13060488554735583,
"grad_norm": 7.40625,
"learning_rate": 0.0001992179138146302,
"loss": 3.415,
"step": 5650
},
{
"epoch": 0.13176068099467755,
"grad_norm": 7.09375,
"learning_rate": 0.0001992026742665484,
"loss": 3.3605,
"step": 5700
},
{
"epoch": 0.1329164764419993,
"grad_norm": 7.46875,
"learning_rate": 0.00019918728826475124,
"loss": 3.4089,
"step": 5750
},
{
"epoch": 0.13407227188932103,
"grad_norm": 7.28125,
"learning_rate": 0.00019917175583195325,
"loss": 3.3466,
"step": 5800
},
{
"epoch": 0.13522806733664275,
"grad_norm": 7.15625,
"learning_rate": 0.00019915607699108506,
"loss": 3.3767,
"step": 5850
},
{
"epoch": 0.1363838627839645,
"grad_norm": 7.375,
"learning_rate": 0.00019914025176529342,
"loss": 3.4191,
"step": 5900
},
{
"epoch": 0.13753965823128622,
"grad_norm": 7.875,
"learning_rate": 0.00019912428017794135,
"loss": 3.4582,
"step": 5950
},
{
"epoch": 0.13869545367860797,
"grad_norm": 6.65625,
"learning_rate": 0.00019910816225260778,
"loss": 3.39,
"step": 6000
},
{
"epoch": 0.1398512491259297,
"grad_norm": 8.125,
"learning_rate": 0.00019909189801308773,
"loss": 3.3571,
"step": 6050
},
{
"epoch": 0.14100704457325142,
"grad_norm": 6.78125,
"learning_rate": 0.00019907548748339222,
"loss": 3.3865,
"step": 6100
},
{
"epoch": 0.14216284002057317,
"grad_norm": 8.5625,
"learning_rate": 0.00019905893068774825,
"loss": 3.4326,
"step": 6150
},
{
"epoch": 0.1433186354678949,
"grad_norm": 6.6875,
"learning_rate": 0.00019904222765059878,
"loss": 3.3557,
"step": 6200
},
{
"epoch": 0.1444744309152166,
"grad_norm": 6.6875,
"learning_rate": 0.0001990253783966026,
"loss": 3.3589,
"step": 6250
},
{
"epoch": 0.14563022636253836,
"grad_norm": 7.84375,
"learning_rate": 0.00019900838295063445,
"loss": 3.3481,
"step": 6300
},
{
"epoch": 0.14678602180986008,
"grad_norm": 7.5625,
"learning_rate": 0.0001989912413377848,
"loss": 3.3661,
"step": 6350
},
{
"epoch": 0.14794181725718183,
"grad_norm": 8.0,
"learning_rate": 0.00019897395358336,
"loss": 3.3895,
"step": 6400
},
{
"epoch": 0.14909761270450356,
"grad_norm": 7.4375,
"learning_rate": 0.00019895651971288202,
"loss": 3.4181,
"step": 6450
},
{
"epoch": 0.15025340815182528,
"grad_norm": 7.28125,
"learning_rate": 0.00019893893975208874,
"loss": 3.3521,
"step": 6500
},
{
"epoch": 0.15140920359914703,
"grad_norm": 7.125,
"learning_rate": 0.0001989212137269335,
"loss": 3.4058,
"step": 6550
},
{
"epoch": 0.15256499904646875,
"grad_norm": 7.3125,
"learning_rate": 0.00019890334166358542,
"loss": 3.3733,
"step": 6600
},
{
"epoch": 0.1537207944937905,
"grad_norm": 8.1875,
"learning_rate": 0.00019888532358842915,
"loss": 3.3731,
"step": 6650
},
{
"epoch": 0.15487658994111222,
"grad_norm": 7.71875,
"learning_rate": 0.00019886715952806497,
"loss": 3.4022,
"step": 6700
},
{
"epoch": 0.15603238538843395,
"grad_norm": 7.40625,
"learning_rate": 0.00019884884950930855,
"loss": 3.3928,
"step": 6750
},
{
"epoch": 0.1571881808357557,
"grad_norm": 6.96875,
"learning_rate": 0.00019883039355919122,
"loss": 3.3792,
"step": 6800
},
{
"epoch": 0.15834397628307742,
"grad_norm": 7.15625,
"learning_rate": 0.00019881179170495956,
"loss": 3.3139,
"step": 6850
},
{
"epoch": 0.15949977173039914,
"grad_norm": 7.78125,
"learning_rate": 0.0001987930439740757,
"loss": 3.2702,
"step": 6900
},
{
"epoch": 0.1606555671777209,
"grad_norm": 7.15625,
"learning_rate": 0.000198774150394217,
"loss": 3.3236,
"step": 6950
},
{
"epoch": 0.16181136262504262,
"grad_norm": 7.4375,
"learning_rate": 0.0001987551109932763,
"loss": 3.3258,
"step": 7000
},
{
"epoch": 0.16296715807236437,
"grad_norm": 7.125,
"learning_rate": 0.0001987359257993615,
"loss": 3.2937,
"step": 7050
},
{
"epoch": 0.1641229535196861,
"grad_norm": 7.28125,
"learning_rate": 0.000198716594840796,
"loss": 3.3124,
"step": 7100
},
{
"epoch": 0.1652787489670078,
"grad_norm": 6.6875,
"learning_rate": 0.00019869711814611814,
"loss": 3.31,
"step": 7150
},
{
"epoch": 0.16643454441432956,
"grad_norm": 7.5,
"learning_rate": 0.00019867749574408158,
"loss": 3.2669,
"step": 7200
},
{
"epoch": 0.16759033986165128,
"grad_norm": 7.53125,
"learning_rate": 0.00019865772766365498,
"loss": 3.319,
"step": 7250
},
{
"epoch": 0.168746135308973,
"grad_norm": 7.0625,
"learning_rate": 0.0001986378139340222,
"loss": 3.3054,
"step": 7300
},
{
"epoch": 0.16990193075629476,
"grad_norm": 6.8125,
"learning_rate": 0.00019861775458458194,
"loss": 3.3608,
"step": 7350
},
{
"epoch": 0.17105772620361648,
"grad_norm": 7.75,
"learning_rate": 0.00019859754964494805,
"loss": 3.3012,
"step": 7400
},
{
"epoch": 0.17221352165093823,
"grad_norm": 7.59375,
"learning_rate": 0.00019857719914494925,
"loss": 3.2881,
"step": 7450
},
{
"epoch": 0.17336931709825995,
"grad_norm": 6.9375,
"learning_rate": 0.00019855670311462912,
"loss": 3.2947,
"step": 7500
},
{
"epoch": 0.17452511254558167,
"grad_norm": 6.90625,
"learning_rate": 0.00019853606158424614,
"loss": 3.2918,
"step": 7550
},
{
"epoch": 0.17568090799290342,
"grad_norm": 7.0,
"learning_rate": 0.00019851527458427355,
"loss": 3.2817,
"step": 7600
},
{
"epoch": 0.17683670344022515,
"grad_norm": 6.71875,
"learning_rate": 0.00019849434214539935,
"loss": 3.2904,
"step": 7650
},
{
"epoch": 0.17799249888754687,
"grad_norm": 7.03125,
"learning_rate": 0.0001984732642985263,
"loss": 3.3213,
"step": 7700
},
{
"epoch": 0.17914829433486862,
"grad_norm": 8.0625,
"learning_rate": 0.0001984520410747718,
"loss": 3.2702,
"step": 7750
},
{
"epoch": 0.18030408978219034,
"grad_norm": 7.15625,
"learning_rate": 0.00019843067250546788,
"loss": 3.3034,
"step": 7800
},
{
"epoch": 0.1814598852295121,
"grad_norm": 6.46875,
"learning_rate": 0.00019840915862216106,
"loss": 3.2654,
"step": 7850
},
{
"epoch": 0.1826156806768338,
"grad_norm": 7.03125,
"learning_rate": 0.00019838749945661253,
"loss": 3.2997,
"step": 7900
},
{
"epoch": 0.18377147612415554,
"grad_norm": 6.90625,
"learning_rate": 0.00019836569504079788,
"loss": 3.2356,
"step": 7950
},
{
"epoch": 0.1849272715714773,
"grad_norm": 6.34375,
"learning_rate": 0.00019834374540690715,
"loss": 3.2827,
"step": 8000
},
{
"epoch": 0.186083067018799,
"grad_norm": 7.15625,
"learning_rate": 0.00019832165058734474,
"loss": 3.2872,
"step": 8050
},
{
"epoch": 0.18723886246612076,
"grad_norm": 7.25,
"learning_rate": 0.00019829941061472943,
"loss": 3.3042,
"step": 8100
},
{
"epoch": 0.18839465791344248,
"grad_norm": 7.09375,
"learning_rate": 0.00019827702552189424,
"loss": 3.1755,
"step": 8150
},
{
"epoch": 0.1895504533607642,
"grad_norm": 6.90625,
"learning_rate": 0.00019825449534188652,
"loss": 3.2881,
"step": 8200
},
{
"epoch": 0.19070624880808595,
"grad_norm": 6.8125,
"learning_rate": 0.00019823182010796772,
"loss": 3.2459,
"step": 8250
},
{
"epoch": 0.19186204425540768,
"grad_norm": 8.75,
"learning_rate": 0.0001982089998536135,
"loss": 3.2342,
"step": 8300
},
{
"epoch": 0.1930178397027294,
"grad_norm": 7.8125,
"learning_rate": 0.00019818603461251356,
"loss": 3.2382,
"step": 8350
},
{
"epoch": 0.19417363515005115,
"grad_norm": 7.5,
"learning_rate": 0.00019816292441857166,
"loss": 3.2235,
"step": 8400
},
{
"epoch": 0.19532943059737287,
"grad_norm": 6.53125,
"learning_rate": 0.00019813966930590563,
"loss": 3.2588,
"step": 8450
},
{
"epoch": 0.19648522604469462,
"grad_norm": 7.03125,
"learning_rate": 0.0001981162693088471,
"loss": 3.2921,
"step": 8500
},
{
"epoch": 0.19764102149201634,
"grad_norm": 7.0,
"learning_rate": 0.00019809272446194172,
"loss": 3.2513,
"step": 8550
},
{
"epoch": 0.19879681693933807,
"grad_norm": 6.75,
"learning_rate": 0.00019806903479994896,
"loss": 3.3163,
"step": 8600
},
{
"epoch": 0.19995261238665982,
"grad_norm": 7.21875,
"learning_rate": 0.00019804520035784204,
"loss": 3.1747,
"step": 8650
},
{
"epoch": 0.20110840783398154,
"grad_norm": 7.28125,
"learning_rate": 0.0001980212211708079,
"loss": 3.2237,
"step": 8700
},
{
"epoch": 0.20226420328130326,
"grad_norm": 6.65625,
"learning_rate": 0.0001979970972742473,
"loss": 3.2788,
"step": 8750
},
{
"epoch": 0.203419998728625,
"grad_norm": 6.75,
"learning_rate": 0.00019797282870377445,
"loss": 3.2688,
"step": 8800
},
{
"epoch": 0.20457579417594673,
"grad_norm": 6.9375,
"learning_rate": 0.0001979484154952173,
"loss": 3.2805,
"step": 8850
},
{
"epoch": 0.20573158962326848,
"grad_norm": 7.28125,
"learning_rate": 0.00019792385768461723,
"loss": 3.2826,
"step": 8900
},
{
"epoch": 0.2068873850705902,
"grad_norm": 7.1875,
"learning_rate": 0.00019789915530822916,
"loss": 3.2345,
"step": 8950
},
{
"epoch": 0.20804318051791193,
"grad_norm": 6.625,
"learning_rate": 0.00019787430840252144,
"loss": 3.2344,
"step": 9000
},
{
"epoch": 0.20919897596523368,
"grad_norm": 7.09375,
"learning_rate": 0.0001978493170041757,
"loss": 3.2483,
"step": 9050
},
{
"epoch": 0.2103547714125554,
"grad_norm": 7.15625,
"learning_rate": 0.00019782418115008694,
"loss": 3.2859,
"step": 9100
},
{
"epoch": 0.21151056685987715,
"grad_norm": 7.53125,
"learning_rate": 0.0001977989008773635,
"loss": 3.2379,
"step": 9150
},
{
"epoch": 0.21266636230719888,
"grad_norm": 6.6875,
"learning_rate": 0.00019777347622332677,
"loss": 3.2463,
"step": 9200
},
{
"epoch": 0.2138221577545206,
"grad_norm": 6.84375,
"learning_rate": 0.00019774790722551142,
"loss": 3.2408,
"step": 9250
},
{
"epoch": 0.21497795320184235,
"grad_norm": 7.03125,
"learning_rate": 0.00019772219392166519,
"loss": 3.2583,
"step": 9300
},
{
"epoch": 0.21613374864916407,
"grad_norm": 7.875,
"learning_rate": 0.00019769633634974875,
"loss": 3.1612,
"step": 9350
},
{
"epoch": 0.2172895440964858,
"grad_norm": 6.8125,
"learning_rate": 0.00019767033454793593,
"loss": 3.2153,
"step": 9400
},
{
"epoch": 0.21844533954380754,
"grad_norm": 7.125,
"learning_rate": 0.00019764418855461338,
"loss": 3.1688,
"step": 9450
},
{
"epoch": 0.21960113499112927,
"grad_norm": 7.53125,
"learning_rate": 0.00019761789840838062,
"loss": 3.2298,
"step": 9500
},
{
"epoch": 0.22075693043845102,
"grad_norm": 7.15625,
"learning_rate": 0.00019759146414805004,
"loss": 3.1326,
"step": 9550
},
{
"epoch": 0.22191272588577274,
"grad_norm": 7.21875,
"learning_rate": 0.00019756488581264674,
"loss": 3.25,
"step": 9600
},
{
"epoch": 0.22306852133309446,
"grad_norm": 7.34375,
"learning_rate": 0.00019753816344140852,
"loss": 3.1666,
"step": 9650
},
{
"epoch": 0.2242243167804162,
"grad_norm": 7.0,
"learning_rate": 0.00019751129707378583,
"loss": 3.2119,
"step": 9700
},
{
"epoch": 0.22538011222773793,
"grad_norm": 8.5625,
"learning_rate": 0.00019748428674944176,
"loss": 3.1728,
"step": 9750
},
{
"epoch": 0.22653590767505966,
"grad_norm": 7.125,
"learning_rate": 0.0001974571325082518,
"loss": 3.1875,
"step": 9800
},
{
"epoch": 0.2276917031223814,
"grad_norm": 7.09375,
"learning_rate": 0.00019742983439030403,
"loss": 3.1583,
"step": 9850
},
{
"epoch": 0.22884749856970313,
"grad_norm": 6.71875,
"learning_rate": 0.00019740239243589883,
"loss": 3.2148,
"step": 9900
},
{
"epoch": 0.23000329401702488,
"grad_norm": 6.75,
"learning_rate": 0.00019737480668554906,
"loss": 3.1629,
"step": 9950
},
{
"epoch": 0.2311590894643466,
"grad_norm": 7.0625,
"learning_rate": 0.00019734707717997973,
"loss": 3.1924,
"step": 10000
},
{
"epoch": 0.23231488491166832,
"grad_norm": 7.0,
"learning_rate": 0.00019731920396012816,
"loss": 3.2101,
"step": 10050
},
{
"epoch": 0.23347068035899007,
"grad_norm": 7.15625,
"learning_rate": 0.00019729118706714375,
"loss": 3.2028,
"step": 10100
},
{
"epoch": 0.2346264758063118,
"grad_norm": 7.375,
"learning_rate": 0.00019726302654238812,
"loss": 3.2211,
"step": 10150
},
{
"epoch": 0.23578227125363352,
"grad_norm": 6.5,
"learning_rate": 0.00019723472242743492,
"loss": 3.207,
"step": 10200
},
{
"epoch": 0.23693806670095527,
"grad_norm": 6.875,
"learning_rate": 0.0001972062747640697,
"loss": 3.2313,
"step": 10250
},
{
"epoch": 0.238093862148277,
"grad_norm": 6.9375,
"learning_rate": 0.00019717768359428997,
"loss": 3.1633,
"step": 10300
},
{
"epoch": 0.23924965759559874,
"grad_norm": 7.15625,
"learning_rate": 0.0001971489489603051,
"loss": 3.1803,
"step": 10350
},
{
"epoch": 0.24040545304292046,
"grad_norm": 6.3125,
"learning_rate": 0.00019712007090453627,
"loss": 3.1878,
"step": 10400
},
{
"epoch": 0.2415612484902422,
"grad_norm": 7.40625,
"learning_rate": 0.00019709104946961643,
"loss": 3.1921,
"step": 10450
},
{
"epoch": 0.24271704393756394,
"grad_norm": 6.53125,
"learning_rate": 0.00019706188469839012,
"loss": 3.1166,
"step": 10500
},
{
"epoch": 0.24387283938488566,
"grad_norm": 7.375,
"learning_rate": 0.00019703257663391353,
"loss": 3.1669,
"step": 10550
},
{
"epoch": 0.2450286348322074,
"grad_norm": 6.5,
"learning_rate": 0.00019700312531945442,
"loss": 3.1521,
"step": 10600
},
{
"epoch": 0.24618443027952913,
"grad_norm": 7.15625,
"learning_rate": 0.00019697353079849197,
"loss": 3.1369,
"step": 10650
},
{
"epoch": 0.24734022572685085,
"grad_norm": 6.8125,
"learning_rate": 0.00019694379311471685,
"loss": 3.2003,
"step": 10700
},
{
"epoch": 0.2484960211741726,
"grad_norm": 7.25,
"learning_rate": 0.000196913912312031,
"loss": 3.1819,
"step": 10750
},
{
"epoch": 0.24965181662149433,
"grad_norm": 7.875,
"learning_rate": 0.00019688388843454774,
"loss": 3.1791,
"step": 10800
},
{
"epoch": 0.2508076120688161,
"grad_norm": 7.0,
"learning_rate": 0.0001968537215265915,
"loss": 3.1503,
"step": 10850
},
{
"epoch": 0.25196340751613777,
"grad_norm": 7.3125,
"learning_rate": 0.000196823411632698,
"loss": 3.1087,
"step": 10900
},
{
"epoch": 0.2531192029634595,
"grad_norm": 7.59375,
"learning_rate": 0.00019679295879761397,
"loss": 3.179,
"step": 10950
},
{
"epoch": 0.25427499841078127,
"grad_norm": 6.65625,
"learning_rate": 0.00019676236306629714,
"loss": 3.13,
"step": 11000
},
{
"epoch": 0.25543079385810297,
"grad_norm": 7.28125,
"learning_rate": 0.00019673162448391628,
"loss": 3.1229,
"step": 11050
},
{
"epoch": 0.2565865893054247,
"grad_norm": 7.21875,
"learning_rate": 0.00019670074309585098,
"loss": 3.1872,
"step": 11100
},
{
"epoch": 0.25774238475274647,
"grad_norm": 6.78125,
"learning_rate": 0.0001966697189476917,
"loss": 3.1442,
"step": 11150
},
{
"epoch": 0.2588981802000682,
"grad_norm": 6.5,
"learning_rate": 0.00019663855208523965,
"loss": 3.1522,
"step": 11200
},
{
"epoch": 0.2600539756473899,
"grad_norm": 6.96875,
"learning_rate": 0.00019660724255450668,
"loss": 3.1626,
"step": 11250
},
{
"epoch": 0.26120977109471166,
"grad_norm": 7.0625,
"learning_rate": 0.00019657579040171537,
"loss": 3.1859,
"step": 11300
},
{
"epoch": 0.2623655665420334,
"grad_norm": 6.8125,
"learning_rate": 0.00019654419567329874,
"loss": 3.1316,
"step": 11350
},
{
"epoch": 0.2635213619893551,
"grad_norm": 6.90625,
"learning_rate": 0.00019651245841590032,
"loss": 3.1144,
"step": 11400
},
{
"epoch": 0.26467715743667686,
"grad_norm": 6.6875,
"learning_rate": 0.0001964805786763742,
"loss": 3.2302,
"step": 11450
},
{
"epoch": 0.2658329528839986,
"grad_norm": 7.28125,
"learning_rate": 0.00019644855650178457,
"loss": 3.1583,
"step": 11500
},
{
"epoch": 0.2669887483313203,
"grad_norm": 6.8125,
"learning_rate": 0.00019641639193940607,
"loss": 3.1643,
"step": 11550
},
{
"epoch": 0.26814454377864205,
"grad_norm": 6.4375,
"learning_rate": 0.0001963840850367235,
"loss": 3.2049,
"step": 11600
},
{
"epoch": 0.2693003392259638,
"grad_norm": 6.78125,
"learning_rate": 0.00019635163584143188,
"loss": 3.1416,
"step": 11650
},
{
"epoch": 0.2704561346732855,
"grad_norm": 6.6875,
"learning_rate": 0.00019631904440143612,
"loss": 3.1389,
"step": 11700
},
{
"epoch": 0.27161193012060725,
"grad_norm": 7.8125,
"learning_rate": 0.00019628631076485124,
"loss": 3.1439,
"step": 11750
},
{
"epoch": 0.272767725567929,
"grad_norm": 6.78125,
"learning_rate": 0.0001962534349800022,
"loss": 3.1273,
"step": 11800
},
{
"epoch": 0.27392352101525075,
"grad_norm": 6.53125,
"learning_rate": 0.00019622041709542383,
"loss": 3.0798,
"step": 11850
},
{
"epoch": 0.27507931646257244,
"grad_norm": 6.40625,
"learning_rate": 0.00019618725715986064,
"loss": 3.1382,
"step": 11900
},
{
"epoch": 0.2762351119098942,
"grad_norm": 6.625,
"learning_rate": 0.00019615395522226692,
"loss": 3.1197,
"step": 11950
},
{
"epoch": 0.27739090735721594,
"grad_norm": 7.71875,
"learning_rate": 0.00019612051133180663,
"loss": 3.1386,
"step": 12000
},
{
"epoch": 0.27854670280453764,
"grad_norm": 7.03125,
"learning_rate": 0.0001960869255378532,
"loss": 3.1003,
"step": 12050
},
{
"epoch": 0.2797024982518594,
"grad_norm": 8.0625,
"learning_rate": 0.00019605319788998965,
"loss": 3.1696,
"step": 12100
},
{
"epoch": 0.28085829369918114,
"grad_norm": 6.71875,
"learning_rate": 0.00019601932843800837,
"loss": 3.1305,
"step": 12150
},
{
"epoch": 0.28201408914650283,
"grad_norm": 7.09375,
"learning_rate": 0.00019598531723191112,
"loss": 3.1251,
"step": 12200
},
{
"epoch": 0.2831698845938246,
"grad_norm": 7.5625,
"learning_rate": 0.0001959511643219089,
"loss": 3.1354,
"step": 12250
},
{
"epoch": 0.28432568004114633,
"grad_norm": 6.78125,
"learning_rate": 0.00019591686975842194,
"loss": 3.1014,
"step": 12300
},
{
"epoch": 0.28548147548846803,
"grad_norm": 7.0,
"learning_rate": 0.0001958824335920796,
"loss": 3.0746,
"step": 12350
},
{
"epoch": 0.2866372709357898,
"grad_norm": 7.84375,
"learning_rate": 0.00019584785587372023,
"loss": 3.12,
"step": 12400
},
{
"epoch": 0.28779306638311153,
"grad_norm": 7.09375,
"learning_rate": 0.00019581313665439123,
"loss": 3.1307,
"step": 12450
},
{
"epoch": 0.2889488618304332,
"grad_norm": 6.65625,
"learning_rate": 0.00019577827598534885,
"loss": 3.099,
"step": 12500
},
{
"epoch": 0.290104657277755,
"grad_norm": 6.9375,
"learning_rate": 0.0001957432739180582,
"loss": 3.1105,
"step": 12550
},
{
"epoch": 0.2912604527250767,
"grad_norm": 6.65625,
"learning_rate": 0.0001957081305041931,
"loss": 3.0801,
"step": 12600
},
{
"epoch": 0.2924162481723985,
"grad_norm": 6.625,
"learning_rate": 0.00019567284579563606,
"loss": 3.1182,
"step": 12650
},
{
"epoch": 0.29357204361972017,
"grad_norm": 7.75,
"learning_rate": 0.0001956374198444782,
"loss": 3.1097,
"step": 12700
},
{
"epoch": 0.2947278390670419,
"grad_norm": 6.96875,
"learning_rate": 0.00019560185270301912,
"loss": 3.0643,
"step": 12750
},
{
"epoch": 0.29588363451436367,
"grad_norm": 6.125,
"learning_rate": 0.00019556614442376685,
"loss": 3.0469,
"step": 12800
},
{
"epoch": 0.29703942996168536,
"grad_norm": 7.09375,
"learning_rate": 0.0001955302950594379,
"loss": 3.1142,
"step": 12850
},
{
"epoch": 0.2981952254090071,
"grad_norm": 6.78125,
"learning_rate": 0.0001954943046629569,
"loss": 3.029,
"step": 12900
},
{
"epoch": 0.29935102085632886,
"grad_norm": 5.9375,
"learning_rate": 0.00019545817328745678,
"loss": 3.0884,
"step": 12950
},
{
"epoch": 0.30050681630365056,
"grad_norm": 7.25,
"learning_rate": 0.00019542190098627865,
"loss": 3.0932,
"step": 13000
},
{
"epoch": 0.3016626117509723,
"grad_norm": 7.03125,
"learning_rate": 0.0001953854878129715,
"loss": 3.1015,
"step": 13050
},
{
"epoch": 0.30281840719829406,
"grad_norm": 6.65625,
"learning_rate": 0.00019534893382129246,
"loss": 3.0881,
"step": 13100
},
{
"epoch": 0.30397420264561575,
"grad_norm": 6.625,
"learning_rate": 0.00019531223906520646,
"loss": 3.0396,
"step": 13150
},
{
"epoch": 0.3051299980929375,
"grad_norm": 7.03125,
"learning_rate": 0.00019527540359888626,
"loss": 3.0928,
"step": 13200
},
{
"epoch": 0.30628579354025925,
"grad_norm": 7.1875,
"learning_rate": 0.00019523842747671242,
"loss": 3.0865,
"step": 13250
},
{
"epoch": 0.307441588987581,
"grad_norm": 6.65625,
"learning_rate": 0.00019520131075327298,
"loss": 3.1052,
"step": 13300
},
{
"epoch": 0.3085973844349027,
"grad_norm": 6.5625,
"learning_rate": 0.00019516405348336376,
"loss": 3.1072,
"step": 13350
},
{
"epoch": 0.30975317988222445,
"grad_norm": 6.3125,
"learning_rate": 0.00019512665572198797,
"loss": 3.1091,
"step": 13400
},
{
"epoch": 0.3109089753295462,
"grad_norm": 6.78125,
"learning_rate": 0.00019508911752435615,
"loss": 3.1306,
"step": 13450
},
{
"epoch": 0.3120647707768679,
"grad_norm": 6.65625,
"learning_rate": 0.0001950514389458863,
"loss": 3.068,
"step": 13500
},
{
"epoch": 0.31322056622418964,
"grad_norm": 6.75,
"learning_rate": 0.0001950136200422036,
"loss": 3.0686,
"step": 13550
},
{
"epoch": 0.3143763616715114,
"grad_norm": 7.53125,
"learning_rate": 0.00019497566086914043,
"loss": 3.0874,
"step": 13600
},
{
"epoch": 0.3155321571188331,
"grad_norm": 7.25,
"learning_rate": 0.00019493756148273618,
"loss": 3.1049,
"step": 13650
},
{
"epoch": 0.31668795256615484,
"grad_norm": 6.5,
"learning_rate": 0.00019489932193923735,
"loss": 3.0234,
"step": 13700
},
{
"epoch": 0.3178437480134766,
"grad_norm": 6.28125,
"learning_rate": 0.0001948609422950972,
"loss": 3.1039,
"step": 13750
},
{
"epoch": 0.3189995434607983,
"grad_norm": 7.15625,
"learning_rate": 0.00019482242260697598,
"loss": 3.0457,
"step": 13800
},
{
"epoch": 0.32015533890812004,
"grad_norm": 6.84375,
"learning_rate": 0.00019478376293174062,
"loss": 3.0817,
"step": 13850
},
{
"epoch": 0.3213111343554418,
"grad_norm": 6.96875,
"learning_rate": 0.00019474496332646466,
"loss": 3.0984,
"step": 13900
},
{
"epoch": 0.3224669298027635,
"grad_norm": 7.09375,
"learning_rate": 0.0001947060238484283,
"loss": 3.0505,
"step": 13950
},
{
"epoch": 0.32362272525008523,
"grad_norm": 7.0625,
"learning_rate": 0.00019466694455511825,
"loss": 3.1112,
"step": 14000
},
{
"epoch": 0.324778520697407,
"grad_norm": 6.96875,
"learning_rate": 0.00019462772550422753,
"loss": 3.0869,
"step": 14050
},
{
"epoch": 0.32593431614472873,
"grad_norm": 7.1875,
"learning_rate": 0.00019458836675365556,
"loss": 3.1541,
"step": 14100
},
{
"epoch": 0.3270901115920504,
"grad_norm": 6.96875,
"learning_rate": 0.00019454886836150798,
"loss": 3.1055,
"step": 14150
},
{
"epoch": 0.3282459070393722,
"grad_norm": 7.375,
"learning_rate": 0.00019450923038609667,
"loss": 3.0548,
"step": 14200
},
{
"epoch": 0.3294017024866939,
"grad_norm": 7.1875,
"learning_rate": 0.00019446945288593935,
"loss": 3.1027,
"step": 14250
},
{
"epoch": 0.3305574979340156,
"grad_norm": 6.71875,
"learning_rate": 0.00019442953591975994,
"loss": 3.1003,
"step": 14300
},
{
"epoch": 0.33171329338133737,
"grad_norm": 6.6875,
"learning_rate": 0.0001943894795464882,
"loss": 3.1062,
"step": 14350
},
{
"epoch": 0.3328690888286591,
"grad_norm": 6.46875,
"learning_rate": 0.00019434928382525967,
"loss": 3.0658,
"step": 14400
},
{
"epoch": 0.3340248842759808,
"grad_norm": 6.5625,
"learning_rate": 0.0001943089488154156,
"loss": 3.1125,
"step": 14450
},
{
"epoch": 0.33518067972330257,
"grad_norm": 6.9375,
"learning_rate": 0.00019426847457650292,
"loss": 3.0812,
"step": 14500
},
{
"epoch": 0.3363364751706243,
"grad_norm": 7.125,
"learning_rate": 0.00019422786116827406,
"loss": 3.0598,
"step": 14550
},
{
"epoch": 0.337492270617946,
"grad_norm": 6.65625,
"learning_rate": 0.00019418710865068693,
"loss": 3.1258,
"step": 14600
},
{
"epoch": 0.33864806606526776,
"grad_norm": 6.5625,
"learning_rate": 0.0001941462170839048,
"loss": 3.0724,
"step": 14650
},
{
"epoch": 0.3398038615125895,
"grad_norm": 7.09375,
"learning_rate": 0.0001941051865282962,
"loss": 3.0358,
"step": 14700
},
{
"epoch": 0.34095965695991126,
"grad_norm": 6.625,
"learning_rate": 0.00019406401704443496,
"loss": 3.128,
"step": 14750
},
{
"epoch": 0.34211545240723296,
"grad_norm": 7.4375,
"learning_rate": 0.00019402270869309983,
"loss": 3.119,
"step": 14800
},
{
"epoch": 0.3432712478545547,
"grad_norm": 7.0625,
"learning_rate": 0.00019398126153527466,
"loss": 3.1128,
"step": 14850
},
{
"epoch": 0.34442704330187646,
"grad_norm": 6.5,
"learning_rate": 0.00019393967563214833,
"loss": 3.0496,
"step": 14900
},
{
"epoch": 0.34558283874919815,
"grad_norm": 7.5625,
"learning_rate": 0.00019389795104511436,
"loss": 3.0841,
"step": 14950
},
{
"epoch": 0.3467386341965199,
"grad_norm": 6.71875,
"learning_rate": 0.00019385608783577108,
"loss": 3.059,
"step": 15000
},
{
"epoch": 0.34789442964384165,
"grad_norm": 7.375,
"learning_rate": 0.00019381408606592159,
"loss": 3.0653,
"step": 15050
},
{
"epoch": 0.34905022509116335,
"grad_norm": 6.90625,
"learning_rate": 0.00019377194579757331,
"loss": 3.0711,
"step": 15100
},
{
"epoch": 0.3502060205384851,
"grad_norm": 7.6875,
"learning_rate": 0.00019372966709293836,
"loss": 3.0586,
"step": 15150
},
{
"epoch": 0.35136181598580685,
"grad_norm": 6.34375,
"learning_rate": 0.00019368725001443312,
"loss": 3.1387,
"step": 15200
},
{
"epoch": 0.35251761143312854,
"grad_norm": 6.6875,
"learning_rate": 0.00019364469462467826,
"loss": 3.0329,
"step": 15250
},
{
"epoch": 0.3536734068804503,
"grad_norm": 6.90625,
"learning_rate": 0.00019360200098649864,
"loss": 3.0361,
"step": 15300
},
{
"epoch": 0.35482920232777204,
"grad_norm": 6.71875,
"learning_rate": 0.0001935591691629232,
"loss": 3.0029,
"step": 15350
},
{
"epoch": 0.35598499777509374,
"grad_norm": 6.40625,
"learning_rate": 0.00019351619921718495,
"loss": 3.0481,
"step": 15400
},
{
"epoch": 0.3571407932224155,
"grad_norm": 7.15625,
"learning_rate": 0.0001934730912127208,
"loss": 3.1042,
"step": 15450
},
{
"epoch": 0.35829658866973724,
"grad_norm": 6.8125,
"learning_rate": 0.00019342984521317135,
"loss": 3.0829,
"step": 15500
},
{
"epoch": 0.359452384117059,
"grad_norm": 7.4375,
"learning_rate": 0.00019338646128238112,
"loss": 2.9839,
"step": 15550
},
{
"epoch": 0.3606081795643807,
"grad_norm": 7.03125,
"learning_rate": 0.0001933429394843981,
"loss": 3.0557,
"step": 15600
},
{
"epoch": 0.36176397501170243,
"grad_norm": 7.25,
"learning_rate": 0.0001932992798834739,
"loss": 3.0046,
"step": 15650
},
{
"epoch": 0.3629197704590242,
"grad_norm": 6.625,
"learning_rate": 0.00019325548254406352,
"loss": 3.0484,
"step": 15700
},
{
"epoch": 0.3640755659063459,
"grad_norm": 6.8125,
"learning_rate": 0.00019321154753082538,
"loss": 2.9846,
"step": 15750
},
{
"epoch": 0.3652313613536676,
"grad_norm": 7.1875,
"learning_rate": 0.000193167474908621,
"loss": 3.0412,
"step": 15800
},
{
"epoch": 0.3663871568009894,
"grad_norm": 6.21875,
"learning_rate": 0.00019312326474251528,
"loss": 2.9781,
"step": 15850
},
{
"epoch": 0.3675429522483111,
"grad_norm": 6.71875,
"learning_rate": 0.00019307891709777596,
"loss": 3.0459,
"step": 15900
},
{
"epoch": 0.3686987476956328,
"grad_norm": 7.1875,
"learning_rate": 0.00019303443203987387,
"loss": 3.0314,
"step": 15950
},
{
"epoch": 0.3698545431429546,
"grad_norm": 6.90625,
"learning_rate": 0.0001929898096344826,
"loss": 3.0569,
"step": 16000
},
{
"epoch": 0.37101033859027627,
"grad_norm": 6.875,
"learning_rate": 0.00019294504994747864,
"loss": 2.9924,
"step": 16050
},
{
"epoch": 0.372166134037598,
"grad_norm": 6.90625,
"learning_rate": 0.00019290015304494103,
"loss": 3.0642,
"step": 16100
},
{
"epoch": 0.37332192948491977,
"grad_norm": 6.53125,
"learning_rate": 0.00019285511899315148,
"loss": 3.0618,
"step": 16150
},
{
"epoch": 0.3744777249322415,
"grad_norm": 7.90625,
"learning_rate": 0.0001928099478585941,
"loss": 3.0071,
"step": 16200
},
{
"epoch": 0.3756335203795632,
"grad_norm": 7.03125,
"learning_rate": 0.00019276463970795542,
"loss": 3.0529,
"step": 16250
},
{
"epoch": 0.37678931582688496,
"grad_norm": 6.5625,
"learning_rate": 0.0001927191946081242,
"loss": 3.0172,
"step": 16300
},
{
"epoch": 0.3779451112742067,
"grad_norm": 7.4375,
"learning_rate": 0.00019267361262619142,
"loss": 2.9783,
"step": 16350
},
{
"epoch": 0.3791009067215284,
"grad_norm": 7.125,
"learning_rate": 0.00019262789382945016,
"loss": 3.0373,
"step": 16400
},
{
"epoch": 0.38025670216885016,
"grad_norm": 7.15625,
"learning_rate": 0.00019258203828539545,
"loss": 3.0515,
"step": 16450
},
{
"epoch": 0.3814124976161719,
"grad_norm": 6.84375,
"learning_rate": 0.00019253604606172417,
"loss": 3.0341,
"step": 16500
},
{
"epoch": 0.3825682930634936,
"grad_norm": 6.5,
"learning_rate": 0.00019248991722633504,
"loss": 3.0231,
"step": 16550
},
{
"epoch": 0.38372408851081535,
"grad_norm": 6.8125,
"learning_rate": 0.0001924436518473284,
"loss": 3.0398,
"step": 16600
},
{
"epoch": 0.3848798839581371,
"grad_norm": 7.0625,
"learning_rate": 0.00019239724999300624,
"loss": 3.052,
"step": 16650
},
{
"epoch": 0.3860356794054588,
"grad_norm": 7.90625,
"learning_rate": 0.00019235071173187197,
"loss": 3.0088,
"step": 16700
},
{
"epoch": 0.38719147485278055,
"grad_norm": 6.84375,
"learning_rate": 0.00019230403713263043,
"loss": 3.0295,
"step": 16750
},
{
"epoch": 0.3883472703001023,
"grad_norm": 6.9375,
"learning_rate": 0.00019225722626418768,
"loss": 3.0033,
"step": 16800
},
{
"epoch": 0.389503065747424,
"grad_norm": 6.84375,
"learning_rate": 0.00019221027919565103,
"loss": 3.0053,
"step": 16850
},
{
"epoch": 0.39065886119474574,
"grad_norm": 6.4375,
"learning_rate": 0.0001921631959963288,
"loss": 2.9991,
"step": 16900
},
{
"epoch": 0.3918146566420675,
"grad_norm": 6.78125,
"learning_rate": 0.0001921159767357303,
"loss": 3.0567,
"step": 16950
},
{
"epoch": 0.39297045208938924,
"grad_norm": 7.15625,
"learning_rate": 0.0001920686214835657,
"loss": 3.0437,
"step": 17000
},
{
"epoch": 0.39412624753671094,
"grad_norm": 6.875,
"learning_rate": 0.00019202113030974591,
"loss": 3.0011,
"step": 17050
},
{
"epoch": 0.3952820429840327,
"grad_norm": 6.5625,
"learning_rate": 0.0001919735032843826,
"loss": 2.9839,
"step": 17100
},
{
"epoch": 0.39643783843135444,
"grad_norm": 7.25,
"learning_rate": 0.00019192574047778794,
"loss": 3.0655,
"step": 17150
},
{
"epoch": 0.39759363387867613,
"grad_norm": 6.46875,
"learning_rate": 0.00019187784196047448,
"loss": 2.9919,
"step": 17200
},
{
"epoch": 0.3987494293259979,
"grad_norm": 6.96875,
"learning_rate": 0.00019182980780315523,
"loss": 2.9839,
"step": 17250
},
{
"epoch": 0.39990522477331963,
"grad_norm": 6.875,
"learning_rate": 0.0001917816380767434,
"loss": 3.0017,
"step": 17300
},
{
"epoch": 0.40106102022064133,
"grad_norm": 6.78125,
"learning_rate": 0.00019173333285235237,
"loss": 2.9293,
"step": 17350
},
{
"epoch": 0.4022168156679631,
"grad_norm": 7.1875,
"learning_rate": 0.00019168489220129555,
"loss": 2.9956,
"step": 17400
},
{
"epoch": 0.40337261111528483,
"grad_norm": 6.625,
"learning_rate": 0.00019163631619508622,
"loss": 3.0152,
"step": 17450
},
{
"epoch": 0.4045284065626065,
"grad_norm": 7.09375,
"learning_rate": 0.00019158760490543754,
"loss": 3.0451,
"step": 17500
},
{
"epoch": 0.4056842020099283,
"grad_norm": 6.5,
"learning_rate": 0.0001915387584042624,
"loss": 2.9809,
"step": 17550
},
{
"epoch": 0.40683999745725,
"grad_norm": 6.90625,
"learning_rate": 0.0001914897767636733,
"loss": 2.9634,
"step": 17600
},
{
"epoch": 0.4079957929045718,
"grad_norm": 7.125,
"learning_rate": 0.0001914406600559822,
"loss": 3.0375,
"step": 17650
},
{
"epoch": 0.40915158835189347,
"grad_norm": 6.875,
"learning_rate": 0.00019139140835370053,
"loss": 3.022,
"step": 17700
},
{
"epoch": 0.4103073837992152,
"grad_norm": 6.25,
"learning_rate": 0.00019134202172953896,
"loss": 3.028,
"step": 17750
},
{
"epoch": 0.41146317924653697,
"grad_norm": 6.71875,
"learning_rate": 0.00019129250025640734,
"loss": 2.9621,
"step": 17800
},
{
"epoch": 0.41261897469385866,
"grad_norm": 6.625,
"learning_rate": 0.0001912428440074147,
"loss": 3.0232,
"step": 17850
},
{
"epoch": 0.4137747701411804,
"grad_norm": 7.6875,
"learning_rate": 0.0001911930530558689,
"loss": 3.0162,
"step": 17900
},
{
"epoch": 0.41493056558850216,
"grad_norm": 6.15625,
"learning_rate": 0.0001911431274752767,
"loss": 3.0113,
"step": 17950
},
{
"epoch": 0.41608636103582386,
"grad_norm": 6.96875,
"learning_rate": 0.00019109306733934374,
"loss": 2.9373,
"step": 18000
},
{
"epoch": 0.4172421564831456,
"grad_norm": 6.875,
"learning_rate": 0.00019104287272197413,
"loss": 3.0107,
"step": 18050
},
{
"epoch": 0.41839795193046736,
"grad_norm": 6.84375,
"learning_rate": 0.0001909925436972706,
"loss": 2.9478,
"step": 18100
},
{
"epoch": 0.41955374737778905,
"grad_norm": 7.15625,
"learning_rate": 0.00019094208033953433,
"loss": 2.9299,
"step": 18150
},
{
"epoch": 0.4207095428251108,
"grad_norm": 6.84375,
"learning_rate": 0.00019089148272326476,
"loss": 2.9888,
"step": 18200
},
{
"epoch": 0.42186533827243256,
"grad_norm": 6.59375,
"learning_rate": 0.00019084075092315957,
"loss": 2.971,
"step": 18250
},
{
"epoch": 0.4230211337197543,
"grad_norm": 6.75,
"learning_rate": 0.00019078988501411453,
"loss": 3.0031,
"step": 18300
},
{
"epoch": 0.424176929167076,
"grad_norm": 7.0625,
"learning_rate": 0.00019073888507122337,
"loss": 2.9433,
"step": 18350
},
{
"epoch": 0.42533272461439775,
"grad_norm": 7.3125,
"learning_rate": 0.00019068775116977779,
"loss": 2.9418,
"step": 18400
},
{
"epoch": 0.4264885200617195,
"grad_norm": 6.5625,
"learning_rate": 0.00019063648338526708,
"loss": 3.0341,
"step": 18450
},
{
"epoch": 0.4276443155090412,
"grad_norm": 6.78125,
"learning_rate": 0.0001905850817933784,
"loss": 3.0207,
"step": 18500
},
{
"epoch": 0.42880011095636295,
"grad_norm": 6.78125,
"learning_rate": 0.00019053354646999626,
"loss": 2.9638,
"step": 18550
},
{
"epoch": 0.4299559064036847,
"grad_norm": 6.6875,
"learning_rate": 0.00019048187749120271,
"loss": 2.9844,
"step": 18600
},
{
"epoch": 0.4311117018510064,
"grad_norm": 6.96875,
"learning_rate": 0.00019043007493327706,
"loss": 2.9668,
"step": 18650
},
{
"epoch": 0.43226749729832814,
"grad_norm": 6.65625,
"learning_rate": 0.00019037813887269592,
"loss": 2.9708,
"step": 18700
},
{
"epoch": 0.4334232927456499,
"grad_norm": 6.9375,
"learning_rate": 0.0001903260693861328,
"loss": 2.9387,
"step": 18750
},
{
"epoch": 0.4345790881929716,
"grad_norm": 6.46875,
"learning_rate": 0.0001902738665504584,
"loss": 2.9638,
"step": 18800
},
{
"epoch": 0.43573488364029334,
"grad_norm": 6.625,
"learning_rate": 0.00019022153044274015,
"loss": 2.9572,
"step": 18850
},
{
"epoch": 0.4368906790876151,
"grad_norm": 6.59375,
"learning_rate": 0.0001901690611402423,
"loss": 2.9324,
"step": 18900
},
{
"epoch": 0.4380464745349368,
"grad_norm": 6.375,
"learning_rate": 0.00019011645872042568,
"loss": 2.9814,
"step": 18950
},
{
"epoch": 0.43920226998225853,
"grad_norm": 6.90625,
"learning_rate": 0.0001900637232609477,
"loss": 2.9905,
"step": 19000
},
{
"epoch": 0.4403580654295803,
"grad_norm": 6.5,
"learning_rate": 0.0001900108548396621,
"loss": 2.9576,
"step": 19050
},
{
"epoch": 0.44151386087690203,
"grad_norm": 6.65625,
"learning_rate": 0.000189957853534619,
"loss": 2.9608,
"step": 19100
},
{
"epoch": 0.4426696563242237,
"grad_norm": 6.59375,
"learning_rate": 0.00018990471942406464,
"loss": 2.9023,
"step": 19150
},
{
"epoch": 0.4438254517715455,
"grad_norm": 6.96875,
"learning_rate": 0.00018985145258644136,
"loss": 2.9505,
"step": 19200
},
{
"epoch": 0.4449812472188672,
"grad_norm": 6.3125,
"learning_rate": 0.00018979805310038742,
"loss": 2.9685,
"step": 19250
},
{
"epoch": 0.4461370426661889,
"grad_norm": 7.34375,
"learning_rate": 0.0001897445210447369,
"loss": 2.9346,
"step": 19300
},
{
"epoch": 0.44729283811351067,
"grad_norm": 6.34375,
"learning_rate": 0.00018969085649851963,
"loss": 2.9651,
"step": 19350
},
{
"epoch": 0.4484486335608324,
"grad_norm": 7.71875,
"learning_rate": 0.000189637059540961,
"loss": 2.997,
"step": 19400
},
{
"epoch": 0.4496044290081541,
"grad_norm": 7.3125,
"learning_rate": 0.00018958313025148189,
"loss": 3.0097,
"step": 19450
},
{
"epoch": 0.45076022445547587,
"grad_norm": 7.34375,
"learning_rate": 0.00018952906870969857,
"loss": 2.9984,
"step": 19500
},
{
"epoch": 0.4519160199027976,
"grad_norm": 7.0625,
"learning_rate": 0.00018947487499542252,
"loss": 2.9843,
"step": 19550
},
{
"epoch": 0.4530718153501193,
"grad_norm": 6.34375,
"learning_rate": 0.00018942054918866038,
"loss": 2.9948,
"step": 19600
},
{
"epoch": 0.45422761079744106,
"grad_norm": 6.6875,
"learning_rate": 0.00018936609136961373,
"loss": 2.9407,
"step": 19650
},
{
"epoch": 0.4553834062447628,
"grad_norm": 6.9375,
"learning_rate": 0.00018931150161867916,
"loss": 2.9388,
"step": 19700
},
{
"epoch": 0.45653920169208456,
"grad_norm": 6.96875,
"learning_rate": 0.00018925678001644792,
"loss": 2.945,
"step": 19750
},
{
"epoch": 0.45769499713940626,
"grad_norm": 6.59375,
"learning_rate": 0.000189201926643706,
"loss": 2.9302,
"step": 19800
},
{
"epoch": 0.458850792586728,
"grad_norm": 7.0,
"learning_rate": 0.00018914694158143382,
"loss": 2.9637,
"step": 19850
},
{
"epoch": 0.46000658803404976,
"grad_norm": 6.9375,
"learning_rate": 0.0001890918249108063,
"loss": 2.947,
"step": 19900
},
{
"epoch": 0.46116238348137145,
"grad_norm": 6.625,
"learning_rate": 0.00018903657671319265,
"loss": 2.9906,
"step": 19950
},
{
"epoch": 0.4623181789286932,
"grad_norm": 6.90625,
"learning_rate": 0.0001889811970701562,
"loss": 2.9613,
"step": 20000
},
{
"epoch": 0.46347397437601495,
"grad_norm": 6.84375,
"learning_rate": 0.0001889256860634544,
"loss": 2.9332,
"step": 20050
},
{
"epoch": 0.46462976982333665,
"grad_norm": 6.5625,
"learning_rate": 0.00018887004377503858,
"loss": 2.9338,
"step": 20100
},
{
"epoch": 0.4657855652706584,
"grad_norm": 6.78125,
"learning_rate": 0.00018881427028705393,
"loss": 2.9767,
"step": 20150
},
{
"epoch": 0.46694136071798015,
"grad_norm": 6.53125,
"learning_rate": 0.00018875836568183925,
"loss": 2.9601,
"step": 20200
},
{
"epoch": 0.46809715616530184,
"grad_norm": 6.5625,
"learning_rate": 0.000188702330041927,
"loss": 2.9519,
"step": 20250
},
{
"epoch": 0.4692529516126236,
"grad_norm": 7.21875,
"learning_rate": 0.00018864616345004307,
"loss": 2.9398,
"step": 20300
},
{
"epoch": 0.47040874705994534,
"grad_norm": 6.59375,
"learning_rate": 0.00018858986598910662,
"loss": 2.9702,
"step": 20350
},
{
"epoch": 0.47156454250726704,
"grad_norm": 6.875,
"learning_rate": 0.00018853343774223006,
"loss": 2.9297,
"step": 20400
},
{
"epoch": 0.4727203379545888,
"grad_norm": 6.71875,
"learning_rate": 0.00018847687879271888,
"loss": 2.959,
"step": 20450
},
{
"epoch": 0.47387613340191054,
"grad_norm": 6.6875,
"learning_rate": 0.0001884201892240715,
"loss": 2.9404,
"step": 20500
},
{
"epoch": 0.4750319288492323,
"grad_norm": 6.53125,
"learning_rate": 0.00018836336911997922,
"loss": 2.9394,
"step": 20550
},
{
"epoch": 0.476187724296554,
"grad_norm": 6.5,
"learning_rate": 0.000188306418564326,
"loss": 2.9666,
"step": 20600
},
{
"epoch": 0.47734351974387573,
"grad_norm": 7.15625,
"learning_rate": 0.00018824933764118842,
"loss": 2.9523,
"step": 20650
},
{
"epoch": 0.4784993151911975,
"grad_norm": 6.71875,
"learning_rate": 0.0001881921264348355,
"loss": 2.9814,
"step": 20700
},
{
"epoch": 0.4796551106385192,
"grad_norm": 7.5,
"learning_rate": 0.00018813478502972863,
"loss": 2.9244,
"step": 20750
},
{
"epoch": 0.48081090608584093,
"grad_norm": 7.4375,
"learning_rate": 0.00018807731351052137,
"loss": 2.9277,
"step": 20800
},
{
"epoch": 0.4819667015331627,
"grad_norm": 6.40625,
"learning_rate": 0.0001880197119620594,
"loss": 2.9359,
"step": 20850
},
{
"epoch": 0.4831224969804844,
"grad_norm": 6.65625,
"learning_rate": 0.00018796198046938033,
"loss": 2.9466,
"step": 20900
},
{
"epoch": 0.4842782924278061,
"grad_norm": 7.53125,
"learning_rate": 0.00018790411911771368,
"loss": 2.9225,
"step": 20950
},
{
"epoch": 0.4854340878751279,
"grad_norm": 6.6875,
"learning_rate": 0.00018784612799248063,
"loss": 2.96,
"step": 21000
},
{
"epoch": 0.48658988332244957,
"grad_norm": 7.21875,
"learning_rate": 0.0001877880071792939,
"loss": 2.9368,
"step": 21050
},
{
"epoch": 0.4877456787697713,
"grad_norm": 6.84375,
"learning_rate": 0.00018772975676395776,
"loss": 2.9535,
"step": 21100
},
{
"epoch": 0.48890147421709307,
"grad_norm": 7.5,
"learning_rate": 0.00018767137683246775,
"loss": 2.9352,
"step": 21150
},
{
"epoch": 0.4900572696644148,
"grad_norm": 6.5625,
"learning_rate": 0.00018761286747101067,
"loss": 2.9526,
"step": 21200
},
{
"epoch": 0.4912130651117365,
"grad_norm": 6.84375,
"learning_rate": 0.00018755422876596436,
"loss": 2.9168,
"step": 21250
},
{
"epoch": 0.49236886055905826,
"grad_norm": 6.5625,
"learning_rate": 0.00018749546080389757,
"loss": 2.9413,
"step": 21300
},
{
"epoch": 0.49352465600638,
"grad_norm": 6.78125,
"learning_rate": 0.00018743656367156998,
"loss": 2.9969,
"step": 21350
},
{
"epoch": 0.4946804514537017,
"grad_norm": 6.90625,
"learning_rate": 0.00018737753745593188,
"loss": 2.9312,
"step": 21400
},
{
"epoch": 0.49583624690102346,
"grad_norm": 6.15625,
"learning_rate": 0.00018731838224412414,
"loss": 2.9545,
"step": 21450
},
{
"epoch": 0.4969920423483452,
"grad_norm": 6.3125,
"learning_rate": 0.0001872590981234781,
"loss": 2.936,
"step": 21500
},
{
"epoch": 0.4981478377956669,
"grad_norm": 6.5625,
"learning_rate": 0.00018719968518151537,
"loss": 2.9431,
"step": 21550
},
{
"epoch": 0.49930363324298865,
"grad_norm": 7.15625,
"learning_rate": 0.00018714014350594782,
"loss": 2.9495,
"step": 21600
},
{
"epoch": 0.5004594286903103,
"grad_norm": 7.34375,
"learning_rate": 0.00018708047318467724,
"loss": 2.9408,
"step": 21650
},
{
"epoch": 0.5016152241376322,
"grad_norm": 6.8125,
"learning_rate": 0.00018702067430579543,
"loss": 2.8727,
"step": 21700
},
{
"epoch": 0.5027710195849538,
"grad_norm": 6.6875,
"learning_rate": 0.00018696074695758393,
"loss": 2.9192,
"step": 21750
},
{
"epoch": 0.5039268150322755,
"grad_norm": 7.25,
"learning_rate": 0.00018690069122851409,
"loss": 2.9637,
"step": 21800
},
{
"epoch": 0.5050826104795973,
"grad_norm": 6.84375,
"learning_rate": 0.00018684050720724655,
"loss": 2.8868,
"step": 21850
},
{
"epoch": 0.506238405926919,
"grad_norm": 6.3125,
"learning_rate": 0.0001867801949826315,
"loss": 2.8911,
"step": 21900
},
{
"epoch": 0.5073942013742407,
"grad_norm": 6.625,
"learning_rate": 0.00018671975464370838,
"loss": 2.8561,
"step": 21950
},
{
"epoch": 0.5085499968215625,
"grad_norm": 6.40625,
"learning_rate": 0.00018665918627970576,
"loss": 2.8978,
"step": 22000
},
{
"epoch": 0.5097057922688842,
"grad_norm": 6.6875,
"learning_rate": 0.0001865984899800412,
"loss": 2.9393,
"step": 22050
},
{
"epoch": 0.5108615877162059,
"grad_norm": 6.90625,
"learning_rate": 0.00018653766583432113,
"loss": 2.8422,
"step": 22100
},
{
"epoch": 0.5120173831635277,
"grad_norm": 6.40625,
"learning_rate": 0.00018647671393234077,
"loss": 2.9341,
"step": 22150
},
{
"epoch": 0.5131731786108494,
"grad_norm": 6.46875,
"learning_rate": 0.00018641563436408386,
"loss": 2.8421,
"step": 22200
},
{
"epoch": 0.5143289740581712,
"grad_norm": 6.21875,
"learning_rate": 0.0001863544272197227,
"loss": 2.8525,
"step": 22250
},
{
"epoch": 0.5154847695054929,
"grad_norm": 6.9375,
"learning_rate": 0.00018629309258961786,
"loss": 2.9711,
"step": 22300
},
{
"epoch": 0.5166405649528146,
"grad_norm": 6.71875,
"learning_rate": 0.00018623163056431818,
"loss": 2.8792,
"step": 22350
},
{
"epoch": 0.5177963604001364,
"grad_norm": 6.65625,
"learning_rate": 0.00018617004123456057,
"loss": 2.9393,
"step": 22400
},
{
"epoch": 0.5189521558474581,
"grad_norm": 6.1875,
"learning_rate": 0.00018610832469126982,
"loss": 2.9378,
"step": 22450
},
{
"epoch": 0.5201079512947798,
"grad_norm": 6.75,
"learning_rate": 0.00018604648102555856,
"loss": 2.8853,
"step": 22500
},
{
"epoch": 0.5212637467421016,
"grad_norm": 6.9375,
"learning_rate": 0.00018598451032872713,
"loss": 2.9423,
"step": 22550
},
{
"epoch": 0.5224195421894233,
"grad_norm": 6.8125,
"learning_rate": 0.00018592241269226335,
"loss": 2.8725,
"step": 22600
},
{
"epoch": 0.523575337636745,
"grad_norm": 6.28125,
"learning_rate": 0.0001858601882078425,
"loss": 2.8732,
"step": 22650
},
{
"epoch": 0.5247311330840668,
"grad_norm": 6.5625,
"learning_rate": 0.00018579783696732704,
"loss": 2.9288,
"step": 22700
},
{
"epoch": 0.5258869285313885,
"grad_norm": 7.40625,
"learning_rate": 0.00018573535906276666,
"loss": 2.925,
"step": 22750
},
{
"epoch": 0.5270427239787102,
"grad_norm": 6.5625,
"learning_rate": 0.00018567275458639796,
"loss": 2.9187,
"step": 22800
},
{
"epoch": 0.528198519426032,
"grad_norm": 6.4375,
"learning_rate": 0.00018561002363064449,
"loss": 2.9511,
"step": 22850
},
{
"epoch": 0.5293543148733537,
"grad_norm": 6.78125,
"learning_rate": 0.0001855471662881164,
"loss": 2.8815,
"step": 22900
},
{
"epoch": 0.5305101103206754,
"grad_norm": 6.65625,
"learning_rate": 0.00018548418265161053,
"loss": 2.9141,
"step": 22950
},
{
"epoch": 0.5316659057679972,
"grad_norm": 6.34375,
"learning_rate": 0.0001854210728141102,
"loss": 2.91,
"step": 23000
},
{
"epoch": 0.5328217012153189,
"grad_norm": 6.65625,
"learning_rate": 0.00018535783686878486,
"loss": 2.9041,
"step": 23050
},
{
"epoch": 0.5339774966626406,
"grad_norm": 6.75,
"learning_rate": 0.0001852944749089903,
"loss": 2.9082,
"step": 23100
},
{
"epoch": 0.5351332921099624,
"grad_norm": 6.1875,
"learning_rate": 0.00018523098702826832,
"loss": 2.9143,
"step": 23150
},
{
"epoch": 0.5362890875572841,
"grad_norm": 6.375,
"learning_rate": 0.00018516737332034658,
"loss": 2.9031,
"step": 23200
},
{
"epoch": 0.5374448830046058,
"grad_norm": 7.09375,
"learning_rate": 0.00018510363387913854,
"loss": 2.8829,
"step": 23250
},
{
"epoch": 0.5386006784519276,
"grad_norm": 6.6875,
"learning_rate": 0.00018503976879874322,
"loss": 2.8267,
"step": 23300
},
{
"epoch": 0.5397564738992493,
"grad_norm": 6.40625,
"learning_rate": 0.00018497577817344513,
"loss": 2.9073,
"step": 23350
},
{
"epoch": 0.540912269346571,
"grad_norm": 6.71875,
"learning_rate": 0.00018491166209771423,
"loss": 2.9388,
"step": 23400
},
{
"epoch": 0.5420680647938928,
"grad_norm": 7.0625,
"learning_rate": 0.00018484742066620554,
"loss": 2.8731,
"step": 23450
},
{
"epoch": 0.5432238602412145,
"grad_norm": 6.71875,
"learning_rate": 0.00018478305397375927,
"loss": 2.9173,
"step": 23500
},
{
"epoch": 0.5443796556885362,
"grad_norm": 7.15625,
"learning_rate": 0.0001847185621154004,
"loss": 2.8957,
"step": 23550
},
{
"epoch": 0.545535451135858,
"grad_norm": 7.15625,
"learning_rate": 0.00018465394518633885,
"loss": 2.9423,
"step": 23600
},
{
"epoch": 0.5466912465831797,
"grad_norm": 6.5,
"learning_rate": 0.00018458920328196912,
"loss": 2.9083,
"step": 23650
},
{
"epoch": 0.5478470420305015,
"grad_norm": 6.5,
"learning_rate": 0.0001845243364978702,
"loss": 2.9317,
"step": 23700
},
{
"epoch": 0.5490028374778232,
"grad_norm": 6.71875,
"learning_rate": 0.00018445934492980544,
"loss": 2.9003,
"step": 23750
},
{
"epoch": 0.5501586329251449,
"grad_norm": 6.65625,
"learning_rate": 0.00018439422867372242,
"loss": 2.8697,
"step": 23800
},
{
"epoch": 0.5513144283724667,
"grad_norm": 6.40625,
"learning_rate": 0.0001843289878257528,
"loss": 2.8957,
"step": 23850
},
{
"epoch": 0.5524702238197884,
"grad_norm": 6.53125,
"learning_rate": 0.0001842636224822122,
"loss": 2.9012,
"step": 23900
},
{
"epoch": 0.5536260192671101,
"grad_norm": 6.46875,
"learning_rate": 0.0001841981327396,
"loss": 2.9053,
"step": 23950
},
{
"epoch": 0.5547818147144319,
"grad_norm": 6.1875,
"learning_rate": 0.00018413251869459923,
"loss": 2.8824,
"step": 24000
},
{
"epoch": 0.5559376101617536,
"grad_norm": 7.0,
"learning_rate": 0.00018406678044407643,
"loss": 2.8666,
"step": 24050
},
{
"epoch": 0.5570934056090753,
"grad_norm": 6.78125,
"learning_rate": 0.0001840009180850815,
"loss": 2.8784,
"step": 24100
},
{
"epoch": 0.5582492010563971,
"grad_norm": 6.0625,
"learning_rate": 0.00018393493171484763,
"loss": 2.8296,
"step": 24150
},
{
"epoch": 0.5594049965037188,
"grad_norm": 6.6875,
"learning_rate": 0.00018386882143079102,
"loss": 2.8968,
"step": 24200
},
{
"epoch": 0.5605607919510405,
"grad_norm": 6.8125,
"learning_rate": 0.00018380258733051075,
"loss": 2.9383,
"step": 24250
},
{
"epoch": 0.5617165873983623,
"grad_norm": 6.65625,
"learning_rate": 0.00018373622951178885,
"loss": 2.8566,
"step": 24300
},
{
"epoch": 0.562872382845684,
"grad_norm": 6.96875,
"learning_rate": 0.00018366974807258987,
"loss": 2.887,
"step": 24350
},
{
"epoch": 0.5640281782930057,
"grad_norm": 6.46875,
"learning_rate": 0.00018360314311106094,
"loss": 2.8875,
"step": 24400
},
{
"epoch": 0.5651839737403275,
"grad_norm": 7.6875,
"learning_rate": 0.0001835364147255314,
"loss": 2.819,
"step": 24450
},
{
"epoch": 0.5663397691876492,
"grad_norm": 6.6875,
"learning_rate": 0.00018346956301451304,
"loss": 2.9575,
"step": 24500
},
{
"epoch": 0.5674955646349709,
"grad_norm": 6.90625,
"learning_rate": 0.0001834025880766995,
"loss": 2.899,
"step": 24550
},
{
"epoch": 0.5686513600822927,
"grad_norm": 6.8125,
"learning_rate": 0.00018333549001096645,
"loss": 2.8903,
"step": 24600
},
{
"epoch": 0.5698071555296144,
"grad_norm": 6.8125,
"learning_rate": 0.00018326826891637134,
"loss": 2.8618,
"step": 24650
},
{
"epoch": 0.5709629509769361,
"grad_norm": 6.90625,
"learning_rate": 0.00018320092489215316,
"loss": 2.9063,
"step": 24700
},
{
"epoch": 0.5721187464242579,
"grad_norm": 6.71875,
"learning_rate": 0.00018313345803773248,
"loss": 2.853,
"step": 24750
},
{
"epoch": 0.5732745418715796,
"grad_norm": 6.34375,
"learning_rate": 0.00018306586845271116,
"loss": 2.8366,
"step": 24800
},
{
"epoch": 0.5744303373189013,
"grad_norm": 6.78125,
"learning_rate": 0.00018299815623687223,
"loss": 2.8669,
"step": 24850
},
{
"epoch": 0.5755861327662231,
"grad_norm": 6.09375,
"learning_rate": 0.00018293032149017984,
"loss": 2.833,
"step": 24900
},
{
"epoch": 0.5767419282135448,
"grad_norm": 7.0625,
"learning_rate": 0.00018286236431277892,
"loss": 2.8762,
"step": 24950
},
{
"epoch": 0.5778977236608664,
"grad_norm": 6.875,
"learning_rate": 0.00018279428480499525,
"loss": 2.9457,
"step": 25000
},
{
"epoch": 0.5790535191081883,
"grad_norm": 6.5625,
"learning_rate": 0.00018272608306733513,
"loss": 2.8913,
"step": 25050
},
{
"epoch": 0.58020931455551,
"grad_norm": 6.875,
"learning_rate": 0.00018265775920048535,
"loss": 2.8868,
"step": 25100
},
{
"epoch": 0.5813651100028318,
"grad_norm": 7.875,
"learning_rate": 0.000182589313305313,
"loss": 2.9019,
"step": 25150
},
{
"epoch": 0.5825209054501534,
"grad_norm": 6.71875,
"learning_rate": 0.00018252074548286528,
"loss": 2.9064,
"step": 25200
},
{
"epoch": 0.5836767008974751,
"grad_norm": 6.46875,
"learning_rate": 0.00018245205583436948,
"loss": 2.8294,
"step": 25250
},
{
"epoch": 0.584832496344797,
"grad_norm": 6.59375,
"learning_rate": 0.00018238324446123266,
"loss": 2.8393,
"step": 25300
},
{
"epoch": 0.5859882917921186,
"grad_norm": 7.09375,
"learning_rate": 0.0001823143114650416,
"loss": 2.8714,
"step": 25350
},
{
"epoch": 0.5871440872394403,
"grad_norm": 6.875,
"learning_rate": 0.0001822452569475626,
"loss": 2.9236,
"step": 25400
},
{
"epoch": 0.5882998826867621,
"grad_norm": 6.1875,
"learning_rate": 0.0001821760810107415,
"loss": 2.8087,
"step": 25450
},
{
"epoch": 0.5894556781340838,
"grad_norm": 6.53125,
"learning_rate": 0.00018210678375670325,
"loss": 2.8775,
"step": 25500
},
{
"epoch": 0.5906114735814055,
"grad_norm": 6.90625,
"learning_rate": 0.00018203736528775197,
"loss": 2.833,
"step": 25550
},
{
"epoch": 0.5917672690287273,
"grad_norm": 6.84375,
"learning_rate": 0.00018196782570637066,
"loss": 2.8305,
"step": 25600
},
{
"epoch": 0.592923064476049,
"grad_norm": 7.03125,
"learning_rate": 0.00018189816511522122,
"loss": 2.8961,
"step": 25650
},
{
"epoch": 0.5940788599233707,
"grad_norm": 6.84375,
"learning_rate": 0.0001818283836171441,
"loss": 2.9032,
"step": 25700
},
{
"epoch": 0.5952346553706925,
"grad_norm": 6.6875,
"learning_rate": 0.00018175848131515837,
"loss": 2.8641,
"step": 25750
},
{
"epoch": 0.5963904508180142,
"grad_norm": 7.15625,
"learning_rate": 0.0001816884583124613,
"loss": 2.8839,
"step": 25800
},
{
"epoch": 0.5975462462653359,
"grad_norm": 6.78125,
"learning_rate": 0.00018161831471242846,
"loss": 2.8364,
"step": 25850
},
{
"epoch": 0.5987020417126577,
"grad_norm": 7.5625,
"learning_rate": 0.0001815480506186134,
"loss": 2.8101,
"step": 25900
},
{
"epoch": 0.5998578371599794,
"grad_norm": 6.9375,
"learning_rate": 0.00018147766613474758,
"loss": 2.8351,
"step": 25950
},
{
"epoch": 0.6010136326073011,
"grad_norm": 6.5625,
"learning_rate": 0.0001814071613647401,
"loss": 2.8322,
"step": 26000
},
{
"epoch": 0.6021694280546229,
"grad_norm": 6.90625,
"learning_rate": 0.0001813365364126779,
"loss": 2.8188,
"step": 26050
},
{
"epoch": 0.6033252235019446,
"grad_norm": 7.5625,
"learning_rate": 0.00018126579138282503,
"loss": 2.8133,
"step": 26100
},
{
"epoch": 0.6044810189492663,
"grad_norm": 7.09375,
"learning_rate": 0.000181194926379623,
"loss": 2.8583,
"step": 26150
},
{
"epoch": 0.6056368143965881,
"grad_norm": 6.9375,
"learning_rate": 0.0001811239415076904,
"loss": 2.849,
"step": 26200
},
{
"epoch": 0.6067926098439098,
"grad_norm": 6.40625,
"learning_rate": 0.00018105283687182278,
"loss": 2.8612,
"step": 26250
},
{
"epoch": 0.6079484052912315,
"grad_norm": 7.28125,
"learning_rate": 0.00018098161257699245,
"loss": 2.8544,
"step": 26300
},
{
"epoch": 0.6091042007385533,
"grad_norm": 6.90625,
"learning_rate": 0.00018091026872834847,
"loss": 2.8698,
"step": 26350
},
{
"epoch": 0.610259996185875,
"grad_norm": 7.4375,
"learning_rate": 0.00018083880543121628,
"loss": 2.847,
"step": 26400
},
{
"epoch": 0.6114157916331967,
"grad_norm": 6.5,
"learning_rate": 0.0001807672227910978,
"loss": 2.8813,
"step": 26450
},
{
"epoch": 0.6125715870805185,
"grad_norm": 6.6875,
"learning_rate": 0.000180695520913671,
"loss": 2.8616,
"step": 26500
},
{
"epoch": 0.6137273825278402,
"grad_norm": 6.5625,
"learning_rate": 0.00018062369990478998,
"loss": 2.8458,
"step": 26550
},
{
"epoch": 0.614883177975162,
"grad_norm": 7.40625,
"learning_rate": 0.00018055175987048466,
"loss": 2.8574,
"step": 26600
},
{
"epoch": 0.6160389734224837,
"grad_norm": 6.4375,
"learning_rate": 0.00018047970091696075,
"loss": 2.8885,
"step": 26650
},
{
"epoch": 0.6171947688698054,
"grad_norm": 7.125,
"learning_rate": 0.00018040752315059942,
"loss": 2.8808,
"step": 26700
},
{
"epoch": 0.6183505643171272,
"grad_norm": 7.1875,
"learning_rate": 0.0001803352266779573,
"loss": 2.8746,
"step": 26750
},
{
"epoch": 0.6195063597644489,
"grad_norm": 7.3125,
"learning_rate": 0.00018026281160576625,
"loss": 2.8616,
"step": 26800
},
{
"epoch": 0.6206621552117706,
"grad_norm": 6.78125,
"learning_rate": 0.00018019027804093326,
"loss": 2.8151,
"step": 26850
},
{
"epoch": 0.6218179506590924,
"grad_norm": 7.21875,
"learning_rate": 0.0001801176260905402,
"loss": 2.8687,
"step": 26900
},
{
"epoch": 0.6229737461064141,
"grad_norm": 7.0,
"learning_rate": 0.00018004485586184377,
"loss": 2.8713,
"step": 26950
},
{
"epoch": 0.6241295415537358,
"grad_norm": 6.5625,
"learning_rate": 0.0001799719674622752,
"loss": 2.8493,
"step": 27000
},
{
"epoch": 0.6252853370010576,
"grad_norm": 7.0625,
"learning_rate": 0.00017989896099944027,
"loss": 2.8604,
"step": 27050
},
{
"epoch": 0.6264411324483793,
"grad_norm": 6.625,
"learning_rate": 0.000179825836581119,
"loss": 2.831,
"step": 27100
},
{
"epoch": 0.627596927895701,
"grad_norm": 6.375,
"learning_rate": 0.00017975259431526552,
"loss": 2.8433,
"step": 27150
},
{
"epoch": 0.6287527233430228,
"grad_norm": 6.8125,
"learning_rate": 0.00017967923431000805,
"loss": 2.8051,
"step": 27200
},
{
"epoch": 0.6299085187903445,
"grad_norm": 6.28125,
"learning_rate": 0.00017960575667364857,
"loss": 2.8349,
"step": 27250
},
{
"epoch": 0.6310643142376662,
"grad_norm": 7.0,
"learning_rate": 0.00017953216151466263,
"loss": 2.8848,
"step": 27300
},
{
"epoch": 0.632220109684988,
"grad_norm": 6.34375,
"learning_rate": 0.00017945844894169942,
"loss": 2.8184,
"step": 27350
},
{
"epoch": 0.6333759051323097,
"grad_norm": 6.78125,
"learning_rate": 0.00017938461906358142,
"loss": 2.8404,
"step": 27400
},
{
"epoch": 0.6345317005796314,
"grad_norm": 7.34375,
"learning_rate": 0.00017931067198930425,
"loss": 2.8079,
"step": 27450
},
{
"epoch": 0.6356874960269532,
"grad_norm": 6.9375,
"learning_rate": 0.00017923660782803662,
"loss": 2.8663,
"step": 27500
},
{
"epoch": 0.6368432914742749,
"grad_norm": 6.4375,
"learning_rate": 0.00017916242668911998,
"loss": 2.8888,
"step": 27550
},
{
"epoch": 0.6379990869215966,
"grad_norm": 6.59375,
"learning_rate": 0.0001790881286820686,
"loss": 2.8957,
"step": 27600
},
{
"epoch": 0.6391548823689184,
"grad_norm": 7.28125,
"learning_rate": 0.00017901371391656928,
"loss": 2.8137,
"step": 27650
},
{
"epoch": 0.6403106778162401,
"grad_norm": 6.09375,
"learning_rate": 0.00017893918250248104,
"loss": 2.8416,
"step": 27700
},
{
"epoch": 0.6414664732635618,
"grad_norm": 6.375,
"learning_rate": 0.00017886453454983526,
"loss": 2.8992,
"step": 27750
},
{
"epoch": 0.6426222687108836,
"grad_norm": 7.125,
"learning_rate": 0.00017878977016883534,
"loss": 2.8262,
"step": 27800
},
{
"epoch": 0.6437780641582053,
"grad_norm": 6.84375,
"learning_rate": 0.00017871488946985653,
"loss": 2.8643,
"step": 27850
},
{
"epoch": 0.644933859605527,
"grad_norm": 6.53125,
"learning_rate": 0.0001786398925634458,
"loss": 2.8146,
"step": 27900
},
{
"epoch": 0.6460896550528488,
"grad_norm": 7.21875,
"learning_rate": 0.00017856477956032174,
"loss": 2.8061,
"step": 27950
},
{
"epoch": 0.6472454505001705,
"grad_norm": 6.5625,
"learning_rate": 0.0001784895505713742,
"loss": 2.841,
"step": 28000
},
{
"epoch": 0.6484012459474923,
"grad_norm": 6.65625,
"learning_rate": 0.00017841420570766445,
"loss": 2.8143,
"step": 28050
},
{
"epoch": 0.649557041394814,
"grad_norm": 6.75,
"learning_rate": 0.00017833874508042466,
"loss": 2.8273,
"step": 28100
},
{
"epoch": 0.6507128368421357,
"grad_norm": 6.46875,
"learning_rate": 0.00017826316880105797,
"loss": 2.8529,
"step": 28150
},
{
"epoch": 0.6518686322894575,
"grad_norm": 7.4375,
"learning_rate": 0.0001781874769811383,
"loss": 2.9175,
"step": 28200
},
{
"epoch": 0.6530244277367792,
"grad_norm": 6.46875,
"learning_rate": 0.00017811166973241005,
"loss": 2.8498,
"step": 28250
},
{
"epoch": 0.6541802231841009,
"grad_norm": 6.59375,
"learning_rate": 0.00017803574716678813,
"loss": 2.8645,
"step": 28300
},
{
"epoch": 0.6553360186314227,
"grad_norm": 7.09375,
"learning_rate": 0.00017795970939635754,
"loss": 2.8361,
"step": 28350
},
{
"epoch": 0.6564918140787444,
"grad_norm": 6.65625,
"learning_rate": 0.0001778835565333736,
"loss": 2.8582,
"step": 28400
},
{
"epoch": 0.657647609526066,
"grad_norm": 6.375,
"learning_rate": 0.0001778072886902613,
"loss": 2.8292,
"step": 28450
},
{
"epoch": 0.6588034049733879,
"grad_norm": 6.09375,
"learning_rate": 0.00017773090597961554,
"loss": 2.8296,
"step": 28500
},
{
"epoch": 0.6599592004207095,
"grad_norm": 6.78125,
"learning_rate": 0.00017765440851420067,
"loss": 2.8063,
"step": 28550
},
{
"epoch": 0.6611149958680312,
"grad_norm": 6.90625,
"learning_rate": 0.00017757779640695058,
"loss": 2.8427,
"step": 28600
},
{
"epoch": 0.662270791315353,
"grad_norm": 6.46875,
"learning_rate": 0.00017750106977096835,
"loss": 2.7857,
"step": 28650
},
{
"epoch": 0.6634265867626747,
"grad_norm": 7.03125,
"learning_rate": 0.00017742422871952607,
"loss": 2.8331,
"step": 28700
},
{
"epoch": 0.6645823822099964,
"grad_norm": 6.96875,
"learning_rate": 0.00017734727336606485,
"loss": 2.8215,
"step": 28750
},
{
"epoch": 0.6657381776573182,
"grad_norm": 6.625,
"learning_rate": 0.00017727020382419452,
"loss": 2.8128,
"step": 28800
},
{
"epoch": 0.6668939731046399,
"grad_norm": 6.875,
"learning_rate": 0.00017719302020769345,
"loss": 2.8049,
"step": 28850
},
{
"epoch": 0.6680497685519616,
"grad_norm": 6.875,
"learning_rate": 0.00017711572263050845,
"loss": 2.8384,
"step": 28900
},
{
"epoch": 0.6692055639992834,
"grad_norm": 6.15625,
"learning_rate": 0.0001770383112067546,
"loss": 2.8349,
"step": 28950
},
{
"epoch": 0.6703613594466051,
"grad_norm": 7.53125,
"learning_rate": 0.00017696078605071493,
"loss": 2.8273,
"step": 29000
},
{
"epoch": 0.6715171548939268,
"grad_norm": 6.4375,
"learning_rate": 0.0001768831472768405,
"loss": 2.847,
"step": 29050
},
{
"epoch": 0.6726729503412486,
"grad_norm": 6.96875,
"learning_rate": 0.0001768053949997501,
"loss": 2.871,
"step": 29100
},
{
"epoch": 0.6738287457885703,
"grad_norm": 6.78125,
"learning_rate": 0.00017672752933423002,
"loss": 2.8197,
"step": 29150
},
{
"epoch": 0.674984541235892,
"grad_norm": 6.25,
"learning_rate": 0.00017664955039523397,
"loss": 2.7989,
"step": 29200
},
{
"epoch": 0.6761403366832138,
"grad_norm": 6.40625,
"learning_rate": 0.0001765714582978829,
"loss": 2.7921,
"step": 29250
},
{
"epoch": 0.6772961321305355,
"grad_norm": 7.28125,
"learning_rate": 0.00017649325315746478,
"loss": 2.8165,
"step": 29300
},
{
"epoch": 0.6784519275778572,
"grad_norm": 6.21875,
"learning_rate": 0.00017641493508943455,
"loss": 2.7825,
"step": 29350
},
{
"epoch": 0.679607723025179,
"grad_norm": 6.75,
"learning_rate": 0.00017633650420941375,
"loss": 2.8214,
"step": 29400
},
{
"epoch": 0.6807635184725007,
"grad_norm": 6.96875,
"learning_rate": 0.00017625796063319056,
"loss": 2.8449,
"step": 29450
},
{
"epoch": 0.6819193139198225,
"grad_norm": 6.40625,
"learning_rate": 0.00017617930447671952,
"loss": 2.7756,
"step": 29500
},
{
"epoch": 0.6830751093671442,
"grad_norm": 7.21875,
"learning_rate": 0.00017610053585612133,
"loss": 2.8324,
"step": 29550
},
{
"epoch": 0.6842309048144659,
"grad_norm": 6.625,
"learning_rate": 0.00017602165488768274,
"loss": 2.7835,
"step": 29600
},
{
"epoch": 0.6853867002617877,
"grad_norm": 6.78125,
"learning_rate": 0.00017594266168785637,
"loss": 2.8095,
"step": 29650
},
{
"epoch": 0.6865424957091094,
"grad_norm": 6.6875,
"learning_rate": 0.00017586355637326054,
"loss": 2.7817,
"step": 29700
},
{
"epoch": 0.6876982911564311,
"grad_norm": 6.15625,
"learning_rate": 0.00017578433906067904,
"loss": 2.8561,
"step": 29750
},
{
"epoch": 0.6888540866037529,
"grad_norm": 6.71875,
"learning_rate": 0.00017570500986706107,
"loss": 2.8384,
"step": 29800
},
{
"epoch": 0.6900098820510746,
"grad_norm": 6.75,
"learning_rate": 0.00017562556890952092,
"loss": 2.7635,
"step": 29850
},
{
"epoch": 0.6911656774983963,
"grad_norm": 7.34375,
"learning_rate": 0.00017554601630533797,
"loss": 2.8574,
"step": 29900
},
{
"epoch": 0.6923214729457181,
"grad_norm": 7.0625,
"learning_rate": 0.00017546635217195635,
"loss": 2.8494,
"step": 29950
},
{
"epoch": 0.6934772683930398,
"grad_norm": 6.875,
"learning_rate": 0.00017538657662698488,
"loss": 2.8229,
"step": 30000
},
{
"epoch": 0.6946330638403615,
"grad_norm": 7.03125,
"learning_rate": 0.0001753066897881968,
"loss": 2.8555,
"step": 30050
},
{
"epoch": 0.6957888592876833,
"grad_norm": 6.65625,
"learning_rate": 0.00017522669177352977,
"loss": 2.8557,
"step": 30100
},
{
"epoch": 0.696944654735005,
"grad_norm": 6.5,
"learning_rate": 0.00017514658270108548,
"loss": 2.8082,
"step": 30150
},
{
"epoch": 0.6981004501823267,
"grad_norm": 7.03125,
"learning_rate": 0.0001750663626891296,
"loss": 2.8585,
"step": 30200
},
{
"epoch": 0.6992562456296485,
"grad_norm": 6.625,
"learning_rate": 0.0001749860318560916,
"loss": 2.7937,
"step": 30250
},
{
"epoch": 0.7004120410769702,
"grad_norm": 6.21875,
"learning_rate": 0.0001749055903205645,
"loss": 2.7666,
"step": 30300
},
{
"epoch": 0.7015678365242919,
"grad_norm": 6.53125,
"learning_rate": 0.00017482503820130484,
"loss": 2.801,
"step": 30350
},
{
"epoch": 0.7027236319716137,
"grad_norm": 7.0,
"learning_rate": 0.0001747443756172324,
"loss": 2.7861,
"step": 30400
},
{
"epoch": 0.7038794274189354,
"grad_norm": 6.09375,
"learning_rate": 0.0001746636026874299,
"loss": 2.7659,
"step": 30450
},
{
"epoch": 0.7050352228662571,
"grad_norm": 6.96875,
"learning_rate": 0.00017458271953114317,
"loss": 2.8292,
"step": 30500
},
{
"epoch": 0.7061910183135789,
"grad_norm": 7.15625,
"learning_rate": 0.00017450172626778064,
"loss": 2.8015,
"step": 30550
},
{
"epoch": 0.7073468137609006,
"grad_norm": 7.34375,
"learning_rate": 0.00017442062301691334,
"loss": 2.8231,
"step": 30600
},
{
"epoch": 0.7085026092082223,
"grad_norm": 6.40625,
"learning_rate": 0.0001743394098982746,
"loss": 2.7942,
"step": 30650
},
{
"epoch": 0.7096584046555441,
"grad_norm": 6.84375,
"learning_rate": 0.0001742580870317601,
"loss": 2.7623,
"step": 30700
},
{
"epoch": 0.7108142001028658,
"grad_norm": 6.4375,
"learning_rate": 0.00017417665453742739,
"loss": 2.8076,
"step": 30750
},
{
"epoch": 0.7119699955501875,
"grad_norm": 6.5,
"learning_rate": 0.00017409511253549593,
"loss": 2.8188,
"step": 30800
},
{
"epoch": 0.7131257909975093,
"grad_norm": 6.1875,
"learning_rate": 0.00017401346114634682,
"loss": 2.819,
"step": 30850
},
{
"epoch": 0.714281586444831,
"grad_norm": 6.65625,
"learning_rate": 0.0001739317004905227,
"loss": 2.8145,
"step": 30900
},
{
"epoch": 0.7154373818921528,
"grad_norm": 6.375,
"learning_rate": 0.00017384983068872756,
"loss": 2.7709,
"step": 30950
},
{
"epoch": 0.7165931773394745,
"grad_norm": 6.71875,
"learning_rate": 0.00017376785186182632,
"loss": 2.8343,
"step": 31000
},
{
"epoch": 0.7177489727867962,
"grad_norm": 6.65625,
"learning_rate": 0.00017368576413084505,
"loss": 2.8354,
"step": 31050
},
{
"epoch": 0.718904768234118,
"grad_norm": 7.0,
"learning_rate": 0.00017360356761697055,
"loss": 2.8315,
"step": 31100
},
{
"epoch": 0.7200605636814397,
"grad_norm": 6.78125,
"learning_rate": 0.0001735212624415502,
"loss": 2.7859,
"step": 31150
},
{
"epoch": 0.7212163591287614,
"grad_norm": 7.34375,
"learning_rate": 0.0001734388487260918,
"loss": 2.7939,
"step": 31200
},
{
"epoch": 0.7223721545760832,
"grad_norm": 6.40625,
"learning_rate": 0.00017335632659226335,
"loss": 2.7738,
"step": 31250
},
{
"epoch": 0.7235279500234049,
"grad_norm": 6.65625,
"learning_rate": 0.000173273696161893,
"loss": 2.7681,
"step": 31300
},
{
"epoch": 0.7246837454707266,
"grad_norm": 6.625,
"learning_rate": 0.00017319095755696865,
"loss": 2.7363,
"step": 31350
},
{
"epoch": 0.7258395409180484,
"grad_norm": 6.75,
"learning_rate": 0.0001731081108996381,
"loss": 2.8358,
"step": 31400
},
{
"epoch": 0.7269953363653701,
"grad_norm": 6.8125,
"learning_rate": 0.00017302515631220845,
"loss": 2.8379,
"step": 31450
},
{
"epoch": 0.7281511318126918,
"grad_norm": 7.09375,
"learning_rate": 0.00017294209391714623,
"loss": 2.8118,
"step": 31500
},
{
"epoch": 0.7293069272600136,
"grad_norm": 6.78125,
"learning_rate": 0.0001728589238370772,
"loss": 2.7696,
"step": 31550
},
{
"epoch": 0.7304627227073353,
"grad_norm": 7.03125,
"learning_rate": 0.00017277564619478595,
"loss": 2.7988,
"step": 31600
},
{
"epoch": 0.731618518154657,
"grad_norm": 6.4375,
"learning_rate": 0.00017269226111321596,
"loss": 2.7845,
"step": 31650
},
{
"epoch": 0.7327743136019788,
"grad_norm": 6.40625,
"learning_rate": 0.00017260876871546936,
"loss": 2.7806,
"step": 31700
},
{
"epoch": 0.7339301090493004,
"grad_norm": 6.5,
"learning_rate": 0.0001725251691248066,
"loss": 2.7502,
"step": 31750
},
{
"epoch": 0.7350859044966221,
"grad_norm": 6.28125,
"learning_rate": 0.00017244146246464642,
"loss": 2.7601,
"step": 31800
},
{
"epoch": 0.736241699943944,
"grad_norm": 6.375,
"learning_rate": 0.00017235764885856572,
"loss": 2.847,
"step": 31850
},
{
"epoch": 0.7373974953912656,
"grad_norm": 6.46875,
"learning_rate": 0.00017227372843029915,
"loss": 2.7661,
"step": 31900
},
{
"epoch": 0.7385532908385873,
"grad_norm": 6.75,
"learning_rate": 0.0001721897013037391,
"loss": 2.8107,
"step": 31950
},
{
"epoch": 0.7397090862859091,
"grad_norm": 6.46875,
"learning_rate": 0.00017210556760293558,
"loss": 2.8154,
"step": 32000
},
{
"epoch": 0.7408648817332308,
"grad_norm": 5.8125,
"learning_rate": 0.00017202132745209578,
"loss": 2.7842,
"step": 32050
},
{
"epoch": 0.7420206771805525,
"grad_norm": 7.15625,
"learning_rate": 0.00017193698097558416,
"loss": 2.765,
"step": 32100
},
{
"epoch": 0.7431764726278743,
"grad_norm": 6.84375,
"learning_rate": 0.0001718525282979221,
"loss": 2.888,
"step": 32150
},
{
"epoch": 0.744332268075196,
"grad_norm": 6.375,
"learning_rate": 0.00017176796954378778,
"loss": 2.8301,
"step": 32200
},
{
"epoch": 0.7454880635225177,
"grad_norm": 7.53125,
"learning_rate": 0.00017168330483801597,
"loss": 2.8191,
"step": 32250
},
{
"epoch": 0.7466438589698395,
"grad_norm": 6.71875,
"learning_rate": 0.00017159853430559786,
"loss": 2.7855,
"step": 32300
},
{
"epoch": 0.7477996544171612,
"grad_norm": 7.15625,
"learning_rate": 0.00017151365807168092,
"loss": 2.7921,
"step": 32350
},
{
"epoch": 0.748955449864483,
"grad_norm": 6.71875,
"learning_rate": 0.00017142867626156855,
"loss": 2.8085,
"step": 32400
},
{
"epoch": 0.7501112453118047,
"grad_norm": 6.65625,
"learning_rate": 0.00017134358900072015,
"loss": 2.8073,
"step": 32450
},
{
"epoch": 0.7512670407591264,
"grad_norm": 7.0,
"learning_rate": 0.00017125839641475072,
"loss": 2.832,
"step": 32500
},
{
"epoch": 0.7524228362064482,
"grad_norm": 6.8125,
"learning_rate": 0.00017117309862943074,
"loss": 2.7696,
"step": 32550
},
{
"epoch": 0.7535786316537699,
"grad_norm": 7.0,
"learning_rate": 0.00017108769577068604,
"loss": 2.8044,
"step": 32600
},
{
"epoch": 0.7547344271010916,
"grad_norm": 6.9375,
"learning_rate": 0.00017100218796459757,
"loss": 2.7827,
"step": 32650
},
{
"epoch": 0.7558902225484134,
"grad_norm": 7.84375,
"learning_rate": 0.0001709165753374012,
"loss": 2.8036,
"step": 32700
},
{
"epoch": 0.7570460179957351,
"grad_norm": 5.78125,
"learning_rate": 0.00017083085801548753,
"loss": 2.7204,
"step": 32750
},
{
"epoch": 0.7582018134430568,
"grad_norm": 6.625,
"learning_rate": 0.00017074503612540175,
"loss": 2.7612,
"step": 32800
},
{
"epoch": 0.7593576088903786,
"grad_norm": 6.5,
"learning_rate": 0.0001706591097938434,
"loss": 2.7828,
"step": 32850
},
{
"epoch": 0.7605134043377003,
"grad_norm": 6.625,
"learning_rate": 0.00017057307914766624,
"loss": 2.6761,
"step": 32900
},
{
"epoch": 0.761669199785022,
"grad_norm": 6.96875,
"learning_rate": 0.00017048694431387804,
"loss": 2.8327,
"step": 32950
},
{
"epoch": 0.7628249952323438,
"grad_norm": 6.9375,
"learning_rate": 0.00017040070541964034,
"loss": 2.7114,
"step": 33000
},
{
"epoch": 0.7639807906796655,
"grad_norm": 6.15625,
"learning_rate": 0.00017031436259226825,
"loss": 2.7744,
"step": 33050
},
{
"epoch": 0.7651365861269872,
"grad_norm": 6.5625,
"learning_rate": 0.00017022791595923046,
"loss": 2.7879,
"step": 33100
},
{
"epoch": 0.766292381574309,
"grad_norm": 6.84375,
"learning_rate": 0.0001701413656481488,
"loss": 2.7954,
"step": 33150
},
{
"epoch": 0.7674481770216307,
"grad_norm": 6.375,
"learning_rate": 0.00017005471178679828,
"loss": 2.8084,
"step": 33200
},
{
"epoch": 0.7686039724689524,
"grad_norm": 6.34375,
"learning_rate": 0.00016996795450310663,
"loss": 2.7583,
"step": 33250
},
{
"epoch": 0.7697597679162742,
"grad_norm": 6.25,
"learning_rate": 0.0001698810939251543,
"loss": 2.7605,
"step": 33300
},
{
"epoch": 0.7709155633635959,
"grad_norm": 6.15625,
"learning_rate": 0.00016979413018117437,
"loss": 2.7707,
"step": 33350
},
{
"epoch": 0.7720713588109176,
"grad_norm": 6.78125,
"learning_rate": 0.00016970706339955208,
"loss": 2.751,
"step": 33400
},
{
"epoch": 0.7732271542582394,
"grad_norm": 6.6875,
"learning_rate": 0.00016961989370882475,
"loss": 2.7815,
"step": 33450
},
{
"epoch": 0.7743829497055611,
"grad_norm": 6.28125,
"learning_rate": 0.0001695326212376818,
"loss": 2.7803,
"step": 33500
},
{
"epoch": 0.7755387451528828,
"grad_norm": 6.28125,
"learning_rate": 0.00016944524611496427,
"loss": 2.7057,
"step": 33550
},
{
"epoch": 0.7766945406002046,
"grad_norm": 6.6875,
"learning_rate": 0.00016935776846966475,
"loss": 2.7923,
"step": 33600
},
{
"epoch": 0.7778503360475263,
"grad_norm": 6.84375,
"learning_rate": 0.00016927018843092718,
"loss": 2.724,
"step": 33650
},
{
"epoch": 0.779006131494848,
"grad_norm": 6.59375,
"learning_rate": 0.00016918250612804673,
"loss": 2.7862,
"step": 33700
},
{
"epoch": 0.7801619269421698,
"grad_norm": 6.78125,
"learning_rate": 0.00016909472169046946,
"loss": 2.7961,
"step": 33750
},
{
"epoch": 0.7813177223894915,
"grad_norm": 6.84375,
"learning_rate": 0.00016900683524779232,
"loss": 2.8042,
"step": 33800
},
{
"epoch": 0.7824735178368133,
"grad_norm": 6.59375,
"learning_rate": 0.00016891884692976265,
"loss": 2.7276,
"step": 33850
},
{
"epoch": 0.783629313284135,
"grad_norm": 6.03125,
"learning_rate": 0.00016883075686627844,
"loss": 2.7939,
"step": 33900
},
{
"epoch": 0.7847851087314567,
"grad_norm": 7.0625,
"learning_rate": 0.0001687425651873877,
"loss": 2.7348,
"step": 33950
},
{
"epoch": 0.7859409041787785,
"grad_norm": 6.5,
"learning_rate": 0.00016865427202328854,
"loss": 2.7605,
"step": 34000
},
{
"epoch": 0.7870966996261002,
"grad_norm": 6.34375,
"learning_rate": 0.00016856587750432888,
"loss": 2.731,
"step": 34050
},
{
"epoch": 0.7882524950734219,
"grad_norm": 6.625,
"learning_rate": 0.00016847738176100632,
"loss": 2.7715,
"step": 34100
},
{
"epoch": 0.7894082905207437,
"grad_norm": 6.65625,
"learning_rate": 0.00016838878492396776,
"loss": 2.7068,
"step": 34150
},
{
"epoch": 0.7905640859680654,
"grad_norm": 6.5,
"learning_rate": 0.00016830008712400948,
"loss": 2.7513,
"step": 34200
},
{
"epoch": 0.7917198814153871,
"grad_norm": 6.6875,
"learning_rate": 0.0001682112884920767,
"loss": 2.709,
"step": 34250
},
{
"epoch": 0.7928756768627089,
"grad_norm": 6.4375,
"learning_rate": 0.0001681223891592637,
"loss": 2.7329,
"step": 34300
},
{
"epoch": 0.7940314723100306,
"grad_norm": 6.8125,
"learning_rate": 0.00016803338925681314,
"loss": 2.7743,
"step": 34350
},
{
"epoch": 0.7951872677573523,
"grad_norm": 6.25,
"learning_rate": 0.00016794428891611638,
"loss": 2.7581,
"step": 34400
},
{
"epoch": 0.7963430632046741,
"grad_norm": 6.75,
"learning_rate": 0.00016785508826871293,
"loss": 2.8453,
"step": 34450
},
{
"epoch": 0.7974988586519958,
"grad_norm": 7.46875,
"learning_rate": 0.00016776578744629052,
"loss": 2.8231,
"step": 34500
},
{
"epoch": 0.7986546540993175,
"grad_norm": 6.75,
"learning_rate": 0.0001676763865806846,
"loss": 2.7601,
"step": 34550
},
{
"epoch": 0.7998104495466393,
"grad_norm": 5.96875,
"learning_rate": 0.0001675868858038784,
"loss": 2.7725,
"step": 34600
},
{
"epoch": 0.800966244993961,
"grad_norm": 6.5625,
"learning_rate": 0.0001674972852480027,
"loss": 2.7787,
"step": 34650
},
{
"epoch": 0.8021220404412827,
"grad_norm": 6.65625,
"learning_rate": 0.00016740758504533548,
"loss": 2.7418,
"step": 34700
},
{
"epoch": 0.8032778358886045,
"grad_norm": 6.5,
"learning_rate": 0.0001673177853283019,
"loss": 2.7893,
"step": 34750
},
{
"epoch": 0.8044336313359262,
"grad_norm": 6.8125,
"learning_rate": 0.000167227886229474,
"loss": 2.7196,
"step": 34800
},
{
"epoch": 0.8055894267832479,
"grad_norm": 7.34375,
"learning_rate": 0.00016713788788157055,
"loss": 2.7814,
"step": 34850
},
{
"epoch": 0.8067452222305697,
"grad_norm": 6.59375,
"learning_rate": 0.00016704779041745686,
"loss": 2.7562,
"step": 34900
},
{
"epoch": 0.8079010176778914,
"grad_norm": 6.21875,
"learning_rate": 0.00016695759397014452,
"loss": 2.7703,
"step": 34950
},
{
"epoch": 0.809056813125213,
"grad_norm": 6.59375,
"learning_rate": 0.00016686729867279132,
"loss": 2.7196,
"step": 35000
},
{
"epoch": 0.8102126085725349,
"grad_norm": 6.40625,
"learning_rate": 0.00016677690465870091,
"loss": 2.7431,
"step": 35050
},
{
"epoch": 0.8113684040198565,
"grad_norm": 6.71875,
"learning_rate": 0.00016668641206132274,
"loss": 2.7328,
"step": 35100
},
{
"epoch": 0.8125241994671784,
"grad_norm": 7.0,
"learning_rate": 0.00016659582101425174,
"loss": 2.7782,
"step": 35150
},
{
"epoch": 0.8136799949145,
"grad_norm": 6.375,
"learning_rate": 0.00016650513165122823,
"loss": 2.7481,
"step": 35200
},
{
"epoch": 0.8148357903618217,
"grad_norm": 6.125,
"learning_rate": 0.0001664143441061376,
"loss": 2.7854,
"step": 35250
},
{
"epoch": 0.8159915858091435,
"grad_norm": 6.25,
"learning_rate": 0.00016632345851301031,
"loss": 2.7391,
"step": 35300
},
{
"epoch": 0.8171473812564652,
"grad_norm": 6.65625,
"learning_rate": 0.00016623247500602154,
"loss": 2.7753,
"step": 35350
},
{
"epoch": 0.8183031767037869,
"grad_norm": 6.65625,
"learning_rate": 0.00016614139371949084,
"loss": 2.703,
"step": 35400
},
{
"epoch": 0.8194589721511087,
"grad_norm": 6.71875,
"learning_rate": 0.00016605021478788233,
"loss": 2.7986,
"step": 35450
},
{
"epoch": 0.8206147675984304,
"grad_norm": 6.6875,
"learning_rate": 0.00016595893834580426,
"loss": 2.7398,
"step": 35500
},
{
"epoch": 0.8217705630457521,
"grad_norm": 6.125,
"learning_rate": 0.0001658675645280087,
"loss": 2.7266,
"step": 35550
},
{
"epoch": 0.8229263584930739,
"grad_norm": 6.6875,
"learning_rate": 0.0001657760934693916,
"loss": 2.7777,
"step": 35600
},
{
"epoch": 0.8240821539403956,
"grad_norm": 7.375,
"learning_rate": 0.00016568452530499244,
"loss": 2.7649,
"step": 35650
},
{
"epoch": 0.8252379493877173,
"grad_norm": 6.8125,
"learning_rate": 0.000165592860169994,
"loss": 2.6939,
"step": 35700
},
{
"epoch": 0.8263937448350391,
"grad_norm": 6.84375,
"learning_rate": 0.0001655010981997223,
"loss": 2.7536,
"step": 35750
},
{
"epoch": 0.8275495402823608,
"grad_norm": 6.1875,
"learning_rate": 0.00016540923952964632,
"loss": 2.7,
"step": 35800
},
{
"epoch": 0.8287053357296825,
"grad_norm": 6.6875,
"learning_rate": 0.00016531728429537766,
"loss": 2.6672,
"step": 35850
},
{
"epoch": 0.8298611311770043,
"grad_norm": 7.3125,
"learning_rate": 0.00016522523263267069,
"loss": 2.7885,
"step": 35900
},
{
"epoch": 0.831016926624326,
"grad_norm": 6.71875,
"learning_rate": 0.00016513308467742198,
"loss": 2.7541,
"step": 35950
},
{
"epoch": 0.8321727220716477,
"grad_norm": 7.125,
"learning_rate": 0.0001650408405656703,
"loss": 2.7671,
"step": 36000
},
{
"epoch": 0.8333285175189695,
"grad_norm": 7.15625,
"learning_rate": 0.00016494850043359638,
"loss": 2.8096,
"step": 36050
},
{
"epoch": 0.8344843129662912,
"grad_norm": 7.28125,
"learning_rate": 0.0001648560644175227,
"loss": 2.794,
"step": 36100
},
{
"epoch": 0.8356401084136129,
"grad_norm": 6.4375,
"learning_rate": 0.00016476353265391333,
"loss": 2.7825,
"step": 36150
},
{
"epoch": 0.8367959038609347,
"grad_norm": 6.9375,
"learning_rate": 0.0001646709052793737,
"loss": 2.7561,
"step": 36200
},
{
"epoch": 0.8379516993082564,
"grad_norm": 6.03125,
"learning_rate": 0.00016457818243065025,
"loss": 2.7079,
"step": 36250
},
{
"epoch": 0.8391074947555781,
"grad_norm": 6.375,
"learning_rate": 0.0001644853642446306,
"loss": 2.727,
"step": 36300
},
{
"epoch": 0.8402632902028999,
"grad_norm": 7.1875,
"learning_rate": 0.0001643924508583429,
"loss": 2.7446,
"step": 36350
},
{
"epoch": 0.8414190856502216,
"grad_norm": 6.625,
"learning_rate": 0.00016429944240895595,
"loss": 2.7231,
"step": 36400
},
{
"epoch": 0.8425748810975433,
"grad_norm": 6.15625,
"learning_rate": 0.00016420633903377896,
"loss": 2.7515,
"step": 36450
},
{
"epoch": 0.8437306765448651,
"grad_norm": 6.65625,
"learning_rate": 0.00016411314087026106,
"loss": 2.7612,
"step": 36500
},
{
"epoch": 0.8448864719921868,
"grad_norm": 6.65625,
"learning_rate": 0.0001640198480559916,
"loss": 2.7066,
"step": 36550
},
{
"epoch": 0.8460422674395086,
"grad_norm": 6.21875,
"learning_rate": 0.00016392646072869944,
"loss": 2.68,
"step": 36600
},
{
"epoch": 0.8471980628868303,
"grad_norm": 6.46875,
"learning_rate": 0.00016383297902625307,
"loss": 2.7536,
"step": 36650
},
{
"epoch": 0.848353858334152,
"grad_norm": 6.4375,
"learning_rate": 0.0001637394030866603,
"loss": 2.7675,
"step": 36700
},
{
"epoch": 0.8495096537814738,
"grad_norm": 7.0,
"learning_rate": 0.00016364573304806804,
"loss": 2.758,
"step": 36750
},
{
"epoch": 0.8506654492287955,
"grad_norm": 6.03125,
"learning_rate": 0.00016355196904876217,
"loss": 2.7143,
"step": 36800
},
{
"epoch": 0.8518212446761172,
"grad_norm": 6.65625,
"learning_rate": 0.00016345811122716713,
"loss": 2.81,
"step": 36850
},
{
"epoch": 0.852977040123439,
"grad_norm": 6.46875,
"learning_rate": 0.00016336415972184612,
"loss": 2.7638,
"step": 36900
},
{
"epoch": 0.8541328355707607,
"grad_norm": 7.15625,
"learning_rate": 0.00016327011467150043,
"loss": 2.7598,
"step": 36950
},
{
"epoch": 0.8552886310180824,
"grad_norm": 6.75,
"learning_rate": 0.00016317597621496953,
"loss": 2.757,
"step": 37000
},
{
"epoch": 0.8564444264654042,
"grad_norm": 7.125,
"learning_rate": 0.00016308174449123085,
"loss": 2.7136,
"step": 37050
},
{
"epoch": 0.8576002219127259,
"grad_norm": 6.5625,
"learning_rate": 0.00016298741963939942,
"loss": 2.739,
"step": 37100
},
{
"epoch": 0.8587560173600476,
"grad_norm": 6.46875,
"learning_rate": 0.0001628930017987278,
"loss": 2.7185,
"step": 37150
},
{
"epoch": 0.8599118128073694,
"grad_norm": 7.0625,
"learning_rate": 0.00016279849110860576,
"loss": 2.6844,
"step": 37200
},
{
"epoch": 0.8610676082546911,
"grad_norm": 6.6875,
"learning_rate": 0.00016270388770856027,
"loss": 2.7351,
"step": 37250
},
{
"epoch": 0.8622234037020128,
"grad_norm": 6.875,
"learning_rate": 0.00016260919173825508,
"loss": 2.7284,
"step": 37300
},
{
"epoch": 0.8633791991493346,
"grad_norm": 6.75,
"learning_rate": 0.00016251440333749055,
"loss": 2.702,
"step": 37350
},
{
"epoch": 0.8645349945966563,
"grad_norm": 6.625,
"learning_rate": 0.00016241952264620365,
"loss": 2.7096,
"step": 37400
},
{
"epoch": 0.865690790043978,
"grad_norm": 6.96875,
"learning_rate": 0.00016232454980446754,
"loss": 2.7184,
"step": 37450
},
{
"epoch": 0.8668465854912998,
"grad_norm": 6.59375,
"learning_rate": 0.00016222948495249134,
"loss": 2.7414,
"step": 37500
},
{
"epoch": 0.8680023809386215,
"grad_norm": 7.28125,
"learning_rate": 0.00016213432823062007,
"loss": 2.7313,
"step": 37550
},
{
"epoch": 0.8691581763859432,
"grad_norm": 6.78125,
"learning_rate": 0.0001620390797793344,
"loss": 2.7671,
"step": 37600
},
{
"epoch": 0.870313971833265,
"grad_norm": 6.84375,
"learning_rate": 0.00016194373973925036,
"loss": 2.7509,
"step": 37650
},
{
"epoch": 0.8714697672805867,
"grad_norm": 6.28125,
"learning_rate": 0.00016184830825111924,
"loss": 2.7784,
"step": 37700
},
{
"epoch": 0.8726255627279084,
"grad_norm": 7.40625,
"learning_rate": 0.0001617527854558273,
"loss": 2.8124,
"step": 37750
},
{
"epoch": 0.8737813581752302,
"grad_norm": 7.0,
"learning_rate": 0.0001616571714943957,
"loss": 2.7399,
"step": 37800
},
{
"epoch": 0.8749371536225519,
"grad_norm": 6.28125,
"learning_rate": 0.00016156146650798004,
"loss": 2.7199,
"step": 37850
},
{
"epoch": 0.8760929490698736,
"grad_norm": 6.625,
"learning_rate": 0.00016146567063787036,
"loss": 2.7486,
"step": 37900
},
{
"epoch": 0.8772487445171954,
"grad_norm": 6.96875,
"learning_rate": 0.00016136978402549095,
"loss": 2.7681,
"step": 37950
},
{
"epoch": 0.8784045399645171,
"grad_norm": 6.875,
"learning_rate": 0.00016127380681239993,
"loss": 2.7472,
"step": 38000
},
{
"epoch": 0.8795603354118389,
"grad_norm": 6.5,
"learning_rate": 0.00016117773914028926,
"loss": 2.7367,
"step": 38050
},
{
"epoch": 0.8807161308591606,
"grad_norm": 6.21875,
"learning_rate": 0.00016108158115098444,
"loss": 2.739,
"step": 38100
},
{
"epoch": 0.8818719263064823,
"grad_norm": 6.78125,
"learning_rate": 0.0001609853329864443,
"loss": 2.7475,
"step": 38150
},
{
"epoch": 0.8830277217538041,
"grad_norm": 7.90625,
"learning_rate": 0.00016088899478876073,
"loss": 2.8016,
"step": 38200
},
{
"epoch": 0.8841835172011258,
"grad_norm": 7.03125,
"learning_rate": 0.00016079256670015868,
"loss": 2.7396,
"step": 38250
},
{
"epoch": 0.8853393126484475,
"grad_norm": 6.34375,
"learning_rate": 0.00016069604886299565,
"loss": 2.7901,
"step": 38300
},
{
"epoch": 0.8864951080957693,
"grad_norm": 6.40625,
"learning_rate": 0.00016059944141976176,
"loss": 2.6889,
"step": 38350
},
{
"epoch": 0.887650903543091,
"grad_norm": 6.5,
"learning_rate": 0.00016050274451307934,
"loss": 2.7298,
"step": 38400
},
{
"epoch": 0.8888066989904126,
"grad_norm": 7.03125,
"learning_rate": 0.00016040595828570282,
"loss": 2.7234,
"step": 38450
},
{
"epoch": 0.8899624944377345,
"grad_norm": 7.15625,
"learning_rate": 0.0001603090828805185,
"loss": 2.7399,
"step": 38500
},
{
"epoch": 0.8911182898850561,
"grad_norm": 6.6875,
"learning_rate": 0.00016021211844054434,
"loss": 2.7144,
"step": 38550
},
{
"epoch": 0.8922740853323778,
"grad_norm": 6.5625,
"learning_rate": 0.00016011506510892968,
"loss": 2.7206,
"step": 38600
},
{
"epoch": 0.8934298807796996,
"grad_norm": 6.59375,
"learning_rate": 0.0001600179230289552,
"loss": 2.7212,
"step": 38650
},
{
"epoch": 0.8945856762270213,
"grad_norm": 6.4375,
"learning_rate": 0.00015992069234403253,
"loss": 2.6791,
"step": 38700
},
{
"epoch": 0.895741471674343,
"grad_norm": 7.0,
"learning_rate": 0.00015982337319770412,
"loss": 2.7405,
"step": 38750
},
{
"epoch": 0.8968972671216648,
"grad_norm": 6.75,
"learning_rate": 0.000159725965733643,
"loss": 2.7146,
"step": 38800
},
{
"epoch": 0.8980530625689865,
"grad_norm": 6.65625,
"learning_rate": 0.00015962847009565266,
"loss": 2.7248,
"step": 38850
},
{
"epoch": 0.8992088580163082,
"grad_norm": 6.625,
"learning_rate": 0.0001595308864276666,
"loss": 2.7034,
"step": 38900
},
{
"epoch": 0.90036465346363,
"grad_norm": 6.21875,
"learning_rate": 0.0001594332148737485,
"loss": 2.7422,
"step": 38950
},
{
"epoch": 0.9015204489109517,
"grad_norm": 7.03125,
"learning_rate": 0.0001593354555780916,
"loss": 2.7502,
"step": 39000
},
{
"epoch": 0.9026762443582734,
"grad_norm": 6.53125,
"learning_rate": 0.0001592376086850187,
"loss": 2.7072,
"step": 39050
},
{
"epoch": 0.9038320398055952,
"grad_norm": 6.4375,
"learning_rate": 0.00015913967433898207,
"loss": 2.6995,
"step": 39100
},
{
"epoch": 0.9049878352529169,
"grad_norm": 6.25,
"learning_rate": 0.00015904165268456293,
"loss": 2.7144,
"step": 39150
},
{
"epoch": 0.9061436307002386,
"grad_norm": 7.21875,
"learning_rate": 0.00015894354386647146,
"loss": 2.7083,
"step": 39200
},
{
"epoch": 0.9072994261475604,
"grad_norm": 6.28125,
"learning_rate": 0.00015884534802954647,
"loss": 2.7375,
"step": 39250
},
{
"epoch": 0.9084552215948821,
"grad_norm": 6.8125,
"learning_rate": 0.0001587470653187553,
"loss": 2.6863,
"step": 39300
},
{
"epoch": 0.9096110170422038,
"grad_norm": 6.375,
"learning_rate": 0.00015864869587919355,
"loss": 2.7086,
"step": 39350
},
{
"epoch": 0.9107668124895256,
"grad_norm": 6.3125,
"learning_rate": 0.00015855023985608476,
"loss": 2.6703,
"step": 39400
},
{
"epoch": 0.9119226079368473,
"grad_norm": 6.5,
"learning_rate": 0.0001584516973947804,
"loss": 2.7426,
"step": 39450
},
{
"epoch": 0.9130784033841691,
"grad_norm": 6.5625,
"learning_rate": 0.0001583530686407595,
"loss": 2.7318,
"step": 39500
},
{
"epoch": 0.9142341988314908,
"grad_norm": 6.5625,
"learning_rate": 0.00015825435373962847,
"loss": 2.6807,
"step": 39550
},
{
"epoch": 0.9153899942788125,
"grad_norm": 7.1875,
"learning_rate": 0.00015815555283712094,
"loss": 2.7323,
"step": 39600
},
{
"epoch": 0.9165457897261343,
"grad_norm": 6.15625,
"learning_rate": 0.0001580566660790975,
"loss": 2.6974,
"step": 39650
},
{
"epoch": 0.917701585173456,
"grad_norm": 7.375,
"learning_rate": 0.00015795769361154547,
"loss": 2.7352,
"step": 39700
},
{
"epoch": 0.9188573806207777,
"grad_norm": 6.1875,
"learning_rate": 0.00015785863558057868,
"loss": 2.6928,
"step": 39750
},
{
"epoch": 0.9200131760680995,
"grad_norm": 6.71875,
"learning_rate": 0.00015775949213243733,
"loss": 2.7731,
"step": 39800
},
{
"epoch": 0.9211689715154212,
"grad_norm": 6.625,
"learning_rate": 0.0001576602634134877,
"loss": 2.728,
"step": 39850
},
{
"epoch": 0.9223247669627429,
"grad_norm": 6.375,
"learning_rate": 0.00015756094957022193,
"loss": 2.7062,
"step": 39900
},
{
"epoch": 0.9234805624100647,
"grad_norm": 6.59375,
"learning_rate": 0.00015746155074925793,
"loss": 2.727,
"step": 39950
},
{
"epoch": 0.9246363578573864,
"grad_norm": 6.65625,
"learning_rate": 0.0001573620670973389,
"loss": 2.7162,
"step": 40000
},
{
"epoch": 0.9257921533047081,
"grad_norm": 7.03125,
"learning_rate": 0.00015726249876133342,
"loss": 2.6423,
"step": 40050
},
{
"epoch": 0.9269479487520299,
"grad_norm": 6.84375,
"learning_rate": 0.000157162845888235,
"loss": 2.6936,
"step": 40100
},
{
"epoch": 0.9281037441993516,
"grad_norm": 6.34375,
"learning_rate": 0.000157063108625162,
"loss": 2.7024,
"step": 40150
},
{
"epoch": 0.9292595396466733,
"grad_norm": 6.4375,
"learning_rate": 0.00015696328711935738,
"loss": 2.7654,
"step": 40200
},
{
"epoch": 0.9304153350939951,
"grad_norm": 6.15625,
"learning_rate": 0.00015686338151818839,
"loss": 2.6738,
"step": 40250
},
{
"epoch": 0.9315711305413168,
"grad_norm": 6.53125,
"learning_rate": 0.0001567633919691465,
"loss": 2.6937,
"step": 40300
},
{
"epoch": 0.9327269259886385,
"grad_norm": 7.0,
"learning_rate": 0.00015666331861984714,
"loss": 2.7191,
"step": 40350
},
{
"epoch": 0.9338827214359603,
"grad_norm": 6.5,
"learning_rate": 0.00015656316161802934,
"loss": 2.7406,
"step": 40400
},
{
"epoch": 0.935038516883282,
"grad_norm": 6.3125,
"learning_rate": 0.00015646292111155577,
"loss": 2.7148,
"step": 40450
},
{
"epoch": 0.9361943123306037,
"grad_norm": 6.625,
"learning_rate": 0.00015636259724841222,
"loss": 2.7191,
"step": 40500
},
{
"epoch": 0.9373501077779255,
"grad_norm": 6.28125,
"learning_rate": 0.00015626219017670772,
"loss": 2.7118,
"step": 40550
},
{
"epoch": 0.9385059032252472,
"grad_norm": 6.875,
"learning_rate": 0.00015616170004467396,
"loss": 2.7047,
"step": 40600
},
{
"epoch": 0.9396616986725689,
"grad_norm": 6.9375,
"learning_rate": 0.00015606112700066542,
"loss": 2.6555,
"step": 40650
},
{
"epoch": 0.9408174941198907,
"grad_norm": 6.96875,
"learning_rate": 0.0001559604711931588,
"loss": 2.6832,
"step": 40700
},
{
"epoch": 0.9419732895672124,
"grad_norm": 7.03125,
"learning_rate": 0.0001558597327707532,
"loss": 2.6998,
"step": 40750
},
{
"epoch": 0.9431290850145341,
"grad_norm": 6.78125,
"learning_rate": 0.0001557589118821695,
"loss": 2.7634,
"step": 40800
},
{
"epoch": 0.9442848804618559,
"grad_norm": 6.53125,
"learning_rate": 0.00015565800867625048,
"loss": 2.7463,
"step": 40850
},
{
"epoch": 0.9454406759091776,
"grad_norm": 7.0,
"learning_rate": 0.00015555702330196023,
"loss": 2.7142,
"step": 40900
},
{
"epoch": 0.9465964713564994,
"grad_norm": 6.75,
"learning_rate": 0.0001554559559083844,
"loss": 2.6963,
"step": 40950
},
{
"epoch": 0.9477522668038211,
"grad_norm": 6.15625,
"learning_rate": 0.00015535480664472954,
"loss": 2.7453,
"step": 41000
},
{
"epoch": 0.9489080622511428,
"grad_norm": 6.75,
"learning_rate": 0.00015525357566032315,
"loss": 2.72,
"step": 41050
},
{
"epoch": 0.9500638576984646,
"grad_norm": 6.59375,
"learning_rate": 0.00015515226310461338,
"loss": 2.6938,
"step": 41100
},
{
"epoch": 0.9512196531457863,
"grad_norm": 6.6875,
"learning_rate": 0.00015505086912716873,
"loss": 2.757,
"step": 41150
},
{
"epoch": 0.952375448593108,
"grad_norm": 6.375,
"learning_rate": 0.000154949393877678,
"loss": 2.7046,
"step": 41200
},
{
"epoch": 0.9535312440404298,
"grad_norm": 6.5625,
"learning_rate": 0.00015484783750594988,
"loss": 2.6413,
"step": 41250
},
{
"epoch": 0.9546870394877515,
"grad_norm": 7.125,
"learning_rate": 0.00015474620016191294,
"loss": 2.7611,
"step": 41300
},
{
"epoch": 0.9558428349350732,
"grad_norm": 6.875,
"learning_rate": 0.00015464448199561517,
"loss": 2.7541,
"step": 41350
},
{
"epoch": 0.956998630382395,
"grad_norm": 6.71875,
"learning_rate": 0.0001545426831572239,
"loss": 2.6785,
"step": 41400
},
{
"epoch": 0.9581544258297167,
"grad_norm": 6.5625,
"learning_rate": 0.00015444080379702565,
"loss": 2.679,
"step": 41450
},
{
"epoch": 0.9593102212770384,
"grad_norm": 6.25,
"learning_rate": 0.00015433884406542572,
"loss": 2.6634,
"step": 41500
},
{
"epoch": 0.9604660167243602,
"grad_norm": 6.21875,
"learning_rate": 0.00015423680411294805,
"loss": 2.7108,
"step": 41550
},
{
"epoch": 0.9616218121716819,
"grad_norm": 6.46875,
"learning_rate": 0.00015413468409023511,
"loss": 2.7187,
"step": 41600
},
{
"epoch": 0.9627776076190036,
"grad_norm": 6.875,
"learning_rate": 0.00015403248414804746,
"loss": 2.6921,
"step": 41650
},
{
"epoch": 0.9639334030663254,
"grad_norm": 6.0625,
"learning_rate": 0.00015393020443726381,
"loss": 2.6772,
"step": 41700
},
{
"epoch": 0.965089198513647,
"grad_norm": 6.625,
"learning_rate": 0.00015382784510888043,
"loss": 2.712,
"step": 41750
},
{
"epoch": 0.9662449939609687,
"grad_norm": 6.0,
"learning_rate": 0.00015372540631401125,
"loss": 2.6733,
"step": 41800
},
{
"epoch": 0.9674007894082906,
"grad_norm": 6.8125,
"learning_rate": 0.00015362288820388757,
"loss": 2.6919,
"step": 41850
},
{
"epoch": 0.9685565848556122,
"grad_norm": 6.625,
"learning_rate": 0.0001535202909298576,
"loss": 2.6868,
"step": 41900
},
{
"epoch": 0.9697123803029339,
"grad_norm": 6.9375,
"learning_rate": 0.0001534176146433866,
"loss": 2.7386,
"step": 41950
},
{
"epoch": 0.9708681757502557,
"grad_norm": 6.375,
"learning_rate": 0.00015331485949605637,
"loss": 2.7163,
"step": 42000
},
{
"epoch": 0.9720239711975774,
"grad_norm": 6.9375,
"learning_rate": 0.0001532120256395652,
"loss": 2.7232,
"step": 42050
},
{
"epoch": 0.9731797666448991,
"grad_norm": 6.9375,
"learning_rate": 0.00015310911322572753,
"loss": 2.6654,
"step": 42100
},
{
"epoch": 0.9743355620922209,
"grad_norm": 6.3125,
"learning_rate": 0.00015300612240647378,
"loss": 2.7098,
"step": 42150
},
{
"epoch": 0.9754913575395426,
"grad_norm": 6.90625,
"learning_rate": 0.0001529030533338502,
"loss": 2.7065,
"step": 42200
},
{
"epoch": 0.9766471529868643,
"grad_norm": 6.5625,
"learning_rate": 0.00015279990616001845,
"loss": 2.7226,
"step": 42250
},
{
"epoch": 0.9778029484341861,
"grad_norm": 6.78125,
"learning_rate": 0.00015269668103725555,
"loss": 2.6874,
"step": 42300
},
{
"epoch": 0.9789587438815078,
"grad_norm": 6.6875,
"learning_rate": 0.00015259337811795355,
"loss": 2.7032,
"step": 42350
},
{
"epoch": 0.9801145393288296,
"grad_norm": 7.03125,
"learning_rate": 0.00015248999755461946,
"loss": 2.7252,
"step": 42400
},
{
"epoch": 0.9812703347761513,
"grad_norm": 6.90625,
"learning_rate": 0.00015238653949987483,
"loss": 2.6823,
"step": 42450
},
{
"epoch": 0.982426130223473,
"grad_norm": 6.75,
"learning_rate": 0.00015228300410645556,
"loss": 2.6763,
"step": 42500
},
{
"epoch": 0.9835819256707948,
"grad_norm": 6.78125,
"learning_rate": 0.00015217939152721192,
"loss": 2.6563,
"step": 42550
},
{
"epoch": 0.9847377211181165,
"grad_norm": 6.75,
"learning_rate": 0.00015207570191510787,
"loss": 2.7296,
"step": 42600
},
{
"epoch": 0.9858935165654382,
"grad_norm": 6.4375,
"learning_rate": 0.00015197193542322132,
"loss": 2.6661,
"step": 42650
},
{
"epoch": 0.98704931201276,
"grad_norm": 6.53125,
"learning_rate": 0.00015186809220474356,
"loss": 2.7088,
"step": 42700
},
{
"epoch": 0.9882051074600817,
"grad_norm": 6.9375,
"learning_rate": 0.0001517641724129792,
"loss": 2.7169,
"step": 42750
},
{
"epoch": 0.9893609029074034,
"grad_norm": 6.53125,
"learning_rate": 0.00015166017620134577,
"loss": 2.7288,
"step": 42800
},
{
"epoch": 0.9905166983547252,
"grad_norm": 6.625,
"learning_rate": 0.00015155610372337384,
"loss": 2.7166,
"step": 42850
},
{
"epoch": 0.9916724938020469,
"grad_norm": 7.34375,
"learning_rate": 0.00015145195513270644,
"loss": 2.7265,
"step": 42900
},
{
"epoch": 0.9928282892493686,
"grad_norm": 6.71875,
"learning_rate": 0.00015134773058309895,
"loss": 2.6791,
"step": 42950
},
{
"epoch": 0.9939840846966904,
"grad_norm": 5.78125,
"learning_rate": 0.00015124343022841893,
"loss": 2.7288,
"step": 43000
},
{
"epoch": 0.9951398801440121,
"grad_norm": 7.0,
"learning_rate": 0.00015113905422264585,
"loss": 2.6671,
"step": 43050
},
{
"epoch": 0.9962956755913338,
"grad_norm": 6.53125,
"learning_rate": 0.00015103460271987084,
"loss": 2.7216,
"step": 43100
},
{
"epoch": 0.9974514710386556,
"grad_norm": 6.90625,
"learning_rate": 0.0001509300758742965,
"loss": 2.6319,
"step": 43150
},
{
"epoch": 0.9986072664859773,
"grad_norm": 6.75,
"learning_rate": 0.00015082547384023674,
"loss": 2.7319,
"step": 43200
},
{
"epoch": 0.999763061933299,
"grad_norm": 7.0625,
"learning_rate": 0.00015072079677211627,
"loss": 2.6939,
"step": 43250
},
{
"epoch": 1.0009188573806207,
"grad_norm": 3.453125,
"learning_rate": 0.00015061604482447075,
"loss": 2.4489,
"step": 43300
},
{
"epoch": 1.0020746528279425,
"grad_norm": 3.3125,
"learning_rate": 0.00015051121815194636,
"loss": 2.4593,
"step": 43350
},
{
"epoch": 1.0032304482752643,
"grad_norm": 3.484375,
"learning_rate": 0.00015040631690929957,
"loss": 2.5265,
"step": 43400
},
{
"epoch": 1.004386243722586,
"grad_norm": 3.1875,
"learning_rate": 0.0001503013412513969,
"loss": 2.4751,
"step": 43450
},
{
"epoch": 1.0055420391699077,
"grad_norm": 3.25,
"learning_rate": 0.0001501962913332148,
"loss": 2.4424,
"step": 43500
},
{
"epoch": 1.0066978346172295,
"grad_norm": 3.15625,
"learning_rate": 0.0001500911673098393,
"loss": 2.4635,
"step": 43550
},
{
"epoch": 1.007853630064551,
"grad_norm": 3.359375,
"learning_rate": 0.00014998596933646585,
"loss": 2.5076,
"step": 43600
},
{
"epoch": 1.009009425511873,
"grad_norm": 3.5,
"learning_rate": 0.00014988069756839912,
"loss": 2.4836,
"step": 43650
},
{
"epoch": 1.0101652209591947,
"grad_norm": 3.390625,
"learning_rate": 0.0001497753521610526,
"loss": 2.5108,
"step": 43700
},
{
"epoch": 1.0113210164065163,
"grad_norm": 3.34375,
"learning_rate": 0.0001496699332699486,
"loss": 2.5315,
"step": 43750
},
{
"epoch": 1.012476811853838,
"grad_norm": 3.375,
"learning_rate": 0.00014956444105071793,
"loss": 2.4555,
"step": 43800
},
{
"epoch": 1.01363260730116,
"grad_norm": 3.390625,
"learning_rate": 0.00014945887565909956,
"loss": 2.5043,
"step": 43850
},
{
"epoch": 1.0147884027484815,
"grad_norm": 3.390625,
"learning_rate": 0.0001493532372509406,
"loss": 2.4642,
"step": 43900
},
{
"epoch": 1.0159441981958033,
"grad_norm": 3.390625,
"learning_rate": 0.0001492475259821958,
"loss": 2.5043,
"step": 43950
},
{
"epoch": 1.017099993643125,
"grad_norm": 3.1875,
"learning_rate": 0.00014914174200892762,
"loss": 2.485,
"step": 44000
},
{
"epoch": 1.0182557890904467,
"grad_norm": 3.515625,
"learning_rate": 0.00014903588548730582,
"loss": 2.5028,
"step": 44050
},
{
"epoch": 1.0194115845377685,
"grad_norm": 3.515625,
"learning_rate": 0.00014892995657360717,
"loss": 2.4697,
"step": 44100
},
{
"epoch": 1.0205673799850903,
"grad_norm": 3.515625,
"learning_rate": 0.0001488239554242155,
"loss": 2.4635,
"step": 44150
},
{
"epoch": 1.0217231754324119,
"grad_norm": 3.328125,
"learning_rate": 0.00014871788219562109,
"loss": 2.4455,
"step": 44200
},
{
"epoch": 1.0228789708797337,
"grad_norm": 3.5,
"learning_rate": 0.00014861173704442074,
"loss": 2.5389,
"step": 44250
},
{
"epoch": 1.0240347663270555,
"grad_norm": 3.296875,
"learning_rate": 0.00014850552012731736,
"loss": 2.4803,
"step": 44300
},
{
"epoch": 1.025190561774377,
"grad_norm": 3.296875,
"learning_rate": 0.0001483992316011199,
"loss": 2.4699,
"step": 44350
},
{
"epoch": 1.0263463572216989,
"grad_norm": 3.328125,
"learning_rate": 0.00014829287162274293,
"loss": 2.4825,
"step": 44400
},
{
"epoch": 1.0275021526690207,
"grad_norm": 3.421875,
"learning_rate": 0.00014818644034920658,
"loss": 2.5092,
"step": 44450
},
{
"epoch": 1.0286579481163423,
"grad_norm": 3.34375,
"learning_rate": 0.00014807993793763619,
"loss": 2.4922,
"step": 44500
},
{
"epoch": 1.029813743563664,
"grad_norm": 3.25,
"learning_rate": 0.00014797336454526217,
"loss": 2.49,
"step": 44550
},
{
"epoch": 1.0309695390109859,
"grad_norm": 3.421875,
"learning_rate": 0.00014786672032941967,
"loss": 2.464,
"step": 44600
},
{
"epoch": 1.0321253344583077,
"grad_norm": 3.375,
"learning_rate": 0.00014776000544754837,
"loss": 2.4816,
"step": 44650
},
{
"epoch": 1.0332811299056293,
"grad_norm": 3.46875,
"learning_rate": 0.0001476532200571924,
"loss": 2.4897,
"step": 44700
},
{
"epoch": 1.034436925352951,
"grad_norm": 3.484375,
"learning_rate": 0.0001475463643159998,
"loss": 2.5117,
"step": 44750
},
{
"epoch": 1.0355927208002729,
"grad_norm": 3.21875,
"learning_rate": 0.00014743943838172268,
"loss": 2.4808,
"step": 44800
},
{
"epoch": 1.0367485162475945,
"grad_norm": 3.59375,
"learning_rate": 0.0001473324424122166,
"loss": 2.4938,
"step": 44850
},
{
"epoch": 1.0379043116949163,
"grad_norm": 3.421875,
"learning_rate": 0.0001472253765654406,
"loss": 2.4875,
"step": 44900
},
{
"epoch": 1.039060107142238,
"grad_norm": 3.328125,
"learning_rate": 0.00014711824099945684,
"loss": 2.4874,
"step": 44950
},
{
"epoch": 1.0402159025895596,
"grad_norm": 3.484375,
"learning_rate": 0.00014701103587243047,
"loss": 2.4644,
"step": 45000
},
{
"epoch": 1.0413716980368815,
"grad_norm": 3.46875,
"learning_rate": 0.00014690376134262923,
"loss": 2.5093,
"step": 45050
},
{
"epoch": 1.0425274934842033,
"grad_norm": 3.421875,
"learning_rate": 0.00014679641756842345,
"loss": 2.4923,
"step": 45100
},
{
"epoch": 1.0436832889315248,
"grad_norm": 3.421875,
"learning_rate": 0.00014668900470828558,
"loss": 2.5029,
"step": 45150
},
{
"epoch": 1.0448390843788466,
"grad_norm": 3.515625,
"learning_rate": 0.00014658152292079007,
"loss": 2.5055,
"step": 45200
},
{
"epoch": 1.0459948798261685,
"grad_norm": 3.265625,
"learning_rate": 0.00014647397236461322,
"loss": 2.5005,
"step": 45250
},
{
"epoch": 1.04715067527349,
"grad_norm": 3.453125,
"learning_rate": 0.00014636635319853275,
"loss": 2.5161,
"step": 45300
},
{
"epoch": 1.0483064707208118,
"grad_norm": 3.609375,
"learning_rate": 0.00014625866558142772,
"loss": 2.4453,
"step": 45350
},
{
"epoch": 1.0494622661681337,
"grad_norm": 3.546875,
"learning_rate": 0.00014615090967227823,
"loss": 2.5309,
"step": 45400
},
{
"epoch": 1.0506180616154552,
"grad_norm": 3.484375,
"learning_rate": 0.00014604308563016526,
"loss": 2.4851,
"step": 45450
},
{
"epoch": 1.051773857062777,
"grad_norm": 3.4375,
"learning_rate": 0.0001459351936142702,
"loss": 2.498,
"step": 45500
},
{
"epoch": 1.0529296525100988,
"grad_norm": 3.25,
"learning_rate": 0.00014582723378387503,
"loss": 2.4626,
"step": 45550
},
{
"epoch": 1.0540854479574204,
"grad_norm": 3.5,
"learning_rate": 0.0001457192062983617,
"loss": 2.4847,
"step": 45600
},
{
"epoch": 1.0552412434047422,
"grad_norm": 3.359375,
"learning_rate": 0.00014561111131721204,
"loss": 2.5068,
"step": 45650
},
{
"epoch": 1.056397038852064,
"grad_norm": 3.53125,
"learning_rate": 0.00014550294900000753,
"loss": 2.4519,
"step": 45700
},
{
"epoch": 1.0575528342993856,
"grad_norm": 3.3125,
"learning_rate": 0.00014539471950642912,
"loss": 2.503,
"step": 45750
},
{
"epoch": 1.0587086297467074,
"grad_norm": 3.4375,
"learning_rate": 0.0001452864229962569,
"loss": 2.5127,
"step": 45800
},
{
"epoch": 1.0598644251940292,
"grad_norm": 3.234375,
"learning_rate": 0.00014517805962936987,
"loss": 2.4373,
"step": 45850
},
{
"epoch": 1.0610202206413508,
"grad_norm": 3.375,
"learning_rate": 0.0001450696295657457,
"loss": 2.4911,
"step": 45900
},
{
"epoch": 1.0621760160886726,
"grad_norm": 3.703125,
"learning_rate": 0.00014496113296546067,
"loss": 2.5432,
"step": 45950
},
{
"epoch": 1.0633318115359944,
"grad_norm": 3.34375,
"learning_rate": 0.00014485256998868912,
"loss": 2.4635,
"step": 46000
},
{
"epoch": 1.064487606983316,
"grad_norm": 3.171875,
"learning_rate": 0.00014474394079570351,
"loss": 2.466,
"step": 46050
},
{
"epoch": 1.0656434024306378,
"grad_norm": 3.5,
"learning_rate": 0.00014463524554687399,
"loss": 2.4885,
"step": 46100
},
{
"epoch": 1.0667991978779596,
"grad_norm": 3.390625,
"learning_rate": 0.0001445264844026682,
"loss": 2.4917,
"step": 46150
},
{
"epoch": 1.0679549933252812,
"grad_norm": 3.46875,
"learning_rate": 0.0001444176575236511,
"loss": 2.4329,
"step": 46200
},
{
"epoch": 1.069110788772603,
"grad_norm": 3.390625,
"learning_rate": 0.00014430876507048478,
"loss": 2.4896,
"step": 46250
},
{
"epoch": 1.0702665842199248,
"grad_norm": 3.375,
"learning_rate": 0.000144199807203928,
"loss": 2.5178,
"step": 46300
},
{
"epoch": 1.0714223796672464,
"grad_norm": 3.3125,
"learning_rate": 0.0001440907840848361,
"loss": 2.4531,
"step": 46350
},
{
"epoch": 1.0725781751145682,
"grad_norm": 3.46875,
"learning_rate": 0.0001439816958741609,
"loss": 2.4841,
"step": 46400
},
{
"epoch": 1.07373397056189,
"grad_norm": 3.59375,
"learning_rate": 0.0001438725427329501,
"loss": 2.5327,
"step": 46450
},
{
"epoch": 1.0748897660092116,
"grad_norm": 3.234375,
"learning_rate": 0.00014376332482234747,
"loss": 2.4471,
"step": 46500
},
{
"epoch": 1.0760455614565334,
"grad_norm": 3.5,
"learning_rate": 0.00014365404230359225,
"loss": 2.4805,
"step": 46550
},
{
"epoch": 1.0772013569038552,
"grad_norm": 3.546875,
"learning_rate": 0.00014354469533801908,
"loss": 2.5377,
"step": 46600
},
{
"epoch": 1.0783571523511768,
"grad_norm": 3.71875,
"learning_rate": 0.00014343528408705786,
"loss": 2.4854,
"step": 46650
},
{
"epoch": 1.0795129477984986,
"grad_norm": 3.71875,
"learning_rate": 0.00014332580871223325,
"loss": 2.5109,
"step": 46700
},
{
"epoch": 1.0806687432458204,
"grad_norm": 3.34375,
"learning_rate": 0.00014321626937516466,
"loss": 2.5114,
"step": 46750
},
{
"epoch": 1.081824538693142,
"grad_norm": 3.4375,
"learning_rate": 0.0001431066662375659,
"loss": 2.5599,
"step": 46800
},
{
"epoch": 1.0829803341404638,
"grad_norm": 3.390625,
"learning_rate": 0.00014299699946124496,
"loss": 2.4898,
"step": 46850
},
{
"epoch": 1.0841361295877856,
"grad_norm": 3.5,
"learning_rate": 0.0001428872692081038,
"loss": 2.4909,
"step": 46900
},
{
"epoch": 1.0852919250351072,
"grad_norm": 3.375,
"learning_rate": 0.0001427774756401381,
"loss": 2.4714,
"step": 46950
},
{
"epoch": 1.086447720482429,
"grad_norm": 3.453125,
"learning_rate": 0.000142667618919437,
"loss": 2.5363,
"step": 47000
},
{
"epoch": 1.0876035159297508,
"grad_norm": 3.46875,
"learning_rate": 0.00014255769920818283,
"loss": 2.4297,
"step": 47050
},
{
"epoch": 1.0887593113770726,
"grad_norm": 3.375,
"learning_rate": 0.000142447716668651,
"loss": 2.4871,
"step": 47100
},
{
"epoch": 1.0899151068243942,
"grad_norm": 3.546875,
"learning_rate": 0.0001423376714632096,
"loss": 2.4647,
"step": 47150
},
{
"epoch": 1.091070902271716,
"grad_norm": 3.40625,
"learning_rate": 0.0001422275637543193,
"loss": 2.5155,
"step": 47200
},
{
"epoch": 1.0922266977190378,
"grad_norm": 3.4375,
"learning_rate": 0.00014211739370453299,
"loss": 2.481,
"step": 47250
},
{
"epoch": 1.0933824931663594,
"grad_norm": 3.359375,
"learning_rate": 0.00014200716147649557,
"loss": 2.4693,
"step": 47300
},
{
"epoch": 1.0945382886136812,
"grad_norm": 3.46875,
"learning_rate": 0.00014189686723294382,
"loss": 2.5398,
"step": 47350
},
{
"epoch": 1.095694084061003,
"grad_norm": 3.421875,
"learning_rate": 0.00014178651113670601,
"loss": 2.4566,
"step": 47400
},
{
"epoch": 1.0968498795083246,
"grad_norm": 3.28125,
"learning_rate": 0.00014167609335070174,
"loss": 2.4797,
"step": 47450
},
{
"epoch": 1.0980056749556464,
"grad_norm": 3.296875,
"learning_rate": 0.00014156561403794174,
"loss": 2.4582,
"step": 47500
},
{
"epoch": 1.0991614704029682,
"grad_norm": 3.421875,
"learning_rate": 0.0001414550733615274,
"loss": 2.4542,
"step": 47550
},
{
"epoch": 1.1003172658502898,
"grad_norm": 3.265625,
"learning_rate": 0.00014134447148465094,
"loss": 2.4871,
"step": 47600
},
{
"epoch": 1.1014730612976116,
"grad_norm": 3.3125,
"learning_rate": 0.00014123380857059478,
"loss": 2.459,
"step": 47650
},
{
"epoch": 1.1026288567449334,
"grad_norm": 3.59375,
"learning_rate": 0.00014112308478273145,
"loss": 2.471,
"step": 47700
},
{
"epoch": 1.103784652192255,
"grad_norm": 3.3125,
"learning_rate": 0.0001410123002845234,
"loss": 2.4934,
"step": 47750
},
{
"epoch": 1.1049404476395768,
"grad_norm": 3.328125,
"learning_rate": 0.00014090145523952267,
"loss": 2.4716,
"step": 47800
},
{
"epoch": 1.1060962430868986,
"grad_norm": 3.46875,
"learning_rate": 0.0001407905498113707,
"loss": 2.5015,
"step": 47850
},
{
"epoch": 1.1072520385342202,
"grad_norm": 3.53125,
"learning_rate": 0.0001406795841637981,
"loss": 2.4897,
"step": 47900
},
{
"epoch": 1.108407833981542,
"grad_norm": 3.46875,
"learning_rate": 0.00014056855846062438,
"loss": 2.5183,
"step": 47950
},
{
"epoch": 1.1095636294288638,
"grad_norm": 3.453125,
"learning_rate": 0.00014045747286575765,
"loss": 2.5172,
"step": 48000
},
{
"epoch": 1.1107194248761854,
"grad_norm": 3.4375,
"learning_rate": 0.00014034632754319443,
"loss": 2.4897,
"step": 48050
},
{
"epoch": 1.1118752203235072,
"grad_norm": 3.515625,
"learning_rate": 0.00014023512265701955,
"loss": 2.4529,
"step": 48100
},
{
"epoch": 1.113031015770829,
"grad_norm": 3.390625,
"learning_rate": 0.00014012385837140567,
"loss": 2.4687,
"step": 48150
},
{
"epoch": 1.1141868112181506,
"grad_norm": 3.421875,
"learning_rate": 0.00014001253485061314,
"loss": 2.5028,
"step": 48200
},
{
"epoch": 1.1153426066654724,
"grad_norm": 3.4375,
"learning_rate": 0.00013990115225898981,
"loss": 2.5051,
"step": 48250
},
{
"epoch": 1.1164984021127942,
"grad_norm": 3.453125,
"learning_rate": 0.0001397897107609706,
"loss": 2.5101,
"step": 48300
},
{
"epoch": 1.1176541975601157,
"grad_norm": 3.3125,
"learning_rate": 0.0001396782105210777,
"loss": 2.4319,
"step": 48350
},
{
"epoch": 1.1188099930074376,
"grad_norm": 3.203125,
"learning_rate": 0.00013956665170391963,
"loss": 2.4638,
"step": 48400
},
{
"epoch": 1.1199657884547594,
"grad_norm": 3.4375,
"learning_rate": 0.00013945503447419168,
"loss": 2.482,
"step": 48450
},
{
"epoch": 1.121121583902081,
"grad_norm": 3.5625,
"learning_rate": 0.00013934335899667527,
"loss": 2.4771,
"step": 48500
},
{
"epoch": 1.1222773793494027,
"grad_norm": 3.46875,
"learning_rate": 0.00013923162543623776,
"loss": 2.5156,
"step": 48550
},
{
"epoch": 1.1234331747967246,
"grad_norm": 3.703125,
"learning_rate": 0.00013911983395783234,
"loss": 2.5293,
"step": 48600
},
{
"epoch": 1.1245889702440461,
"grad_norm": 3.46875,
"learning_rate": 0.0001390079847264977,
"loss": 2.4806,
"step": 48650
},
{
"epoch": 1.125744765691368,
"grad_norm": 3.40625,
"learning_rate": 0.0001388960779073578,
"loss": 2.5185,
"step": 48700
},
{
"epoch": 1.1269005611386897,
"grad_norm": 3.53125,
"learning_rate": 0.00013878411366562152,
"loss": 2.4888,
"step": 48750
},
{
"epoch": 1.1280563565860113,
"grad_norm": 3.390625,
"learning_rate": 0.0001386720921665826,
"loss": 2.4726,
"step": 48800
},
{
"epoch": 1.1292121520333331,
"grad_norm": 3.546875,
"learning_rate": 0.0001385600135756193,
"loss": 2.498,
"step": 48850
},
{
"epoch": 1.130367947480655,
"grad_norm": 3.34375,
"learning_rate": 0.0001384478780581942,
"loss": 2.4179,
"step": 48900
},
{
"epoch": 1.1315237429279765,
"grad_norm": 3.453125,
"learning_rate": 0.0001383356857798538,
"loss": 2.4694,
"step": 48950
},
{
"epoch": 1.1326795383752983,
"grad_norm": 3.40625,
"learning_rate": 0.00013822343690622852,
"loss": 2.4334,
"step": 49000
},
{
"epoch": 1.1338353338226201,
"grad_norm": 3.46875,
"learning_rate": 0.00013811113160303227,
"loss": 2.4452,
"step": 49050
},
{
"epoch": 1.1349911292699417,
"grad_norm": 3.390625,
"learning_rate": 0.00013799877003606225,
"loss": 2.5213,
"step": 49100
},
{
"epoch": 1.1361469247172635,
"grad_norm": 3.390625,
"learning_rate": 0.00013788635237119883,
"loss": 2.4899,
"step": 49150
},
{
"epoch": 1.1373027201645853,
"grad_norm": 3.3125,
"learning_rate": 0.00013777387877440508,
"loss": 2.4481,
"step": 49200
},
{
"epoch": 1.138458515611907,
"grad_norm": 3.296875,
"learning_rate": 0.00013766134941172665,
"loss": 2.4779,
"step": 49250
},
{
"epoch": 1.1396143110592287,
"grad_norm": 3.484375,
"learning_rate": 0.00013754876444929166,
"loss": 2.5089,
"step": 49300
},
{
"epoch": 1.1407701065065505,
"grad_norm": 3.53125,
"learning_rate": 0.0001374361240533101,
"loss": 2.4753,
"step": 49350
},
{
"epoch": 1.1419259019538721,
"grad_norm": 3.5625,
"learning_rate": 0.00013732342839007394,
"loss": 2.4319,
"step": 49400
},
{
"epoch": 1.143081697401194,
"grad_norm": 3.375,
"learning_rate": 0.00013721067762595675,
"loss": 2.4881,
"step": 49450
},
{
"epoch": 1.1442374928485157,
"grad_norm": 3.375,
"learning_rate": 0.00013709787192741333,
"loss": 2.4892,
"step": 49500
},
{
"epoch": 1.1453932882958373,
"grad_norm": 3.46875,
"learning_rate": 0.00013698501146097963,
"loss": 2.4686,
"step": 49550
},
{
"epoch": 1.1465490837431591,
"grad_norm": 3.6875,
"learning_rate": 0.00013687209639327258,
"loss": 2.4744,
"step": 49600
},
{
"epoch": 1.147704879190481,
"grad_norm": 3.578125,
"learning_rate": 0.0001367591268909896,
"loss": 2.4712,
"step": 49650
},
{
"epoch": 1.1488606746378025,
"grad_norm": 3.609375,
"learning_rate": 0.00013664610312090838,
"loss": 2.4969,
"step": 49700
},
{
"epoch": 1.1500164700851243,
"grad_norm": 3.53125,
"learning_rate": 0.00013653302524988694,
"loss": 2.4409,
"step": 49750
},
{
"epoch": 1.1511722655324461,
"grad_norm": 3.265625,
"learning_rate": 0.000136419893444863,
"loss": 2.4727,
"step": 49800
},
{
"epoch": 1.1523280609797677,
"grad_norm": 3.53125,
"learning_rate": 0.00013630670787285397,
"loss": 2.4807,
"step": 49850
},
{
"epoch": 1.1534838564270895,
"grad_norm": 3.453125,
"learning_rate": 0.00013619346870095666,
"loss": 2.4666,
"step": 49900
},
{
"epoch": 1.1546396518744113,
"grad_norm": 3.59375,
"learning_rate": 0.0001360801760963469,
"loss": 2.5184,
"step": 49950
},
{
"epoch": 1.155795447321733,
"grad_norm": 3.625,
"learning_rate": 0.00013596683022627963,
"loss": 2.4725,
"step": 50000
},
{
"epoch": 1.1569512427690547,
"grad_norm": 3.359375,
"learning_rate": 0.00013585343125808815,
"loss": 2.3922,
"step": 50050
},
{
"epoch": 1.1581070382163765,
"grad_norm": 3.640625,
"learning_rate": 0.0001357399793591844,
"loss": 2.519,
"step": 50100
},
{
"epoch": 1.159262833663698,
"grad_norm": 3.5,
"learning_rate": 0.00013562647469705818,
"loss": 2.5099,
"step": 50150
},
{
"epoch": 1.16041862911102,
"grad_norm": 3.21875,
"learning_rate": 0.00013551291743927747,
"loss": 2.4294,
"step": 50200
},
{
"epoch": 1.1615744245583417,
"grad_norm": 3.4375,
"learning_rate": 0.00013539930775348776,
"loss": 2.5493,
"step": 50250
},
{
"epoch": 1.1627302200056633,
"grad_norm": 3.421875,
"learning_rate": 0.0001352856458074119,
"loss": 2.4688,
"step": 50300
},
{
"epoch": 1.163886015452985,
"grad_norm": 3.375,
"learning_rate": 0.00013517193176885,
"loss": 2.4601,
"step": 50350
},
{
"epoch": 1.165041810900307,
"grad_norm": 3.859375,
"learning_rate": 0.000135058165805679,
"loss": 2.4417,
"step": 50400
},
{
"epoch": 1.1661976063476285,
"grad_norm": 3.4375,
"learning_rate": 0.0001349443480858525,
"loss": 2.4708,
"step": 50450
},
{
"epoch": 1.1673534017949503,
"grad_norm": 3.40625,
"learning_rate": 0.00013483047877740055,
"loss": 2.4765,
"step": 50500
},
{
"epoch": 1.168509197242272,
"grad_norm": 3.734375,
"learning_rate": 0.00013471655804842933,
"loss": 2.507,
"step": 50550
},
{
"epoch": 1.1696649926895937,
"grad_norm": 3.484375,
"learning_rate": 0.00013460258606712096,
"loss": 2.4378,
"step": 50600
},
{
"epoch": 1.1708207881369155,
"grad_norm": 3.484375,
"learning_rate": 0.00013448856300173317,
"loss": 2.4559,
"step": 50650
},
{
"epoch": 1.1719765835842373,
"grad_norm": 3.28125,
"learning_rate": 0.00013437448902059915,
"loss": 2.4808,
"step": 50700
},
{
"epoch": 1.173132379031559,
"grad_norm": 3.234375,
"learning_rate": 0.00013426036429212726,
"loss": 2.5196,
"step": 50750
},
{
"epoch": 1.1742881744788807,
"grad_norm": 3.46875,
"learning_rate": 0.0001341461889848007,
"loss": 2.4844,
"step": 50800
},
{
"epoch": 1.1754439699262025,
"grad_norm": 3.453125,
"learning_rate": 0.0001340319632671775,
"loss": 2.4552,
"step": 50850
},
{
"epoch": 1.1765997653735243,
"grad_norm": 3.546875,
"learning_rate": 0.00013391768730789002,
"loss": 2.4945,
"step": 50900
},
{
"epoch": 1.1777555608208459,
"grad_norm": 3.46875,
"learning_rate": 0.00013380336127564467,
"loss": 2.408,
"step": 50950
},
{
"epoch": 1.1789113562681677,
"grad_norm": 3.40625,
"learning_rate": 0.000133688985339222,
"loss": 2.4645,
"step": 51000
},
{
"epoch": 1.1800671517154895,
"grad_norm": 3.5625,
"learning_rate": 0.00013357455966747615,
"loss": 2.4897,
"step": 51050
},
{
"epoch": 1.181222947162811,
"grad_norm": 3.53125,
"learning_rate": 0.0001334600844293346,
"loss": 2.4561,
"step": 51100
},
{
"epoch": 1.1823787426101329,
"grad_norm": 3.53125,
"learning_rate": 0.0001333455597937981,
"loss": 2.4419,
"step": 51150
},
{
"epoch": 1.1835345380574547,
"grad_norm": 3.3125,
"learning_rate": 0.00013323098592994033,
"loss": 2.4806,
"step": 51200
},
{
"epoch": 1.1846903335047763,
"grad_norm": 3.46875,
"learning_rate": 0.00013311636300690756,
"loss": 2.4532,
"step": 51250
},
{
"epoch": 1.185846128952098,
"grad_norm": 3.546875,
"learning_rate": 0.00013300169119391864,
"loss": 2.4768,
"step": 51300
},
{
"epoch": 1.1870019243994199,
"grad_norm": 3.40625,
"learning_rate": 0.0001328869706602644,
"loss": 2.4738,
"step": 51350
},
{
"epoch": 1.1881577198467415,
"grad_norm": 3.546875,
"learning_rate": 0.00013277220157530774,
"loss": 2.4579,
"step": 51400
},
{
"epoch": 1.1893135152940633,
"grad_norm": 3.328125,
"learning_rate": 0.0001326573841084832,
"loss": 2.5154,
"step": 51450
},
{
"epoch": 1.190469310741385,
"grad_norm": 3.40625,
"learning_rate": 0.00013254251842929676,
"loss": 2.4843,
"step": 51500
},
{
"epoch": 1.1916251061887067,
"grad_norm": 3.484375,
"learning_rate": 0.00013242760470732548,
"loss": 2.453,
"step": 51550
},
{
"epoch": 1.1927809016360285,
"grad_norm": 3.40625,
"learning_rate": 0.00013231264311221755,
"loss": 2.4359,
"step": 51600
},
{
"epoch": 1.1939366970833503,
"grad_norm": 3.453125,
"learning_rate": 0.00013219763381369155,
"loss": 2.4442,
"step": 51650
},
{
"epoch": 1.1950924925306718,
"grad_norm": 3.296875,
"learning_rate": 0.00013208257698153677,
"loss": 2.4893,
"step": 51700
},
{
"epoch": 1.1962482879779937,
"grad_norm": 3.390625,
"learning_rate": 0.00013196747278561248,
"loss": 2.4486,
"step": 51750
},
{
"epoch": 1.1974040834253155,
"grad_norm": 3.625,
"learning_rate": 0.00013185232139584802,
"loss": 2.4553,
"step": 51800
},
{
"epoch": 1.198559878872637,
"grad_norm": 3.3125,
"learning_rate": 0.00013173712298224222,
"loss": 2.4743,
"step": 51850
},
{
"epoch": 1.1997156743199588,
"grad_norm": 3.5,
"learning_rate": 0.0001316218777148635,
"loss": 2.469,
"step": 51900
},
{
"epoch": 1.2008714697672807,
"grad_norm": 3.53125,
"learning_rate": 0.00013150658576384936,
"loss": 2.4755,
"step": 51950
},
{
"epoch": 1.2020272652146022,
"grad_norm": 3.421875,
"learning_rate": 0.00013139124729940623,
"loss": 2.4775,
"step": 52000
},
{
"epoch": 1.203183060661924,
"grad_norm": 3.515625,
"learning_rate": 0.00013127586249180932,
"loss": 2.4589,
"step": 52050
},
{
"epoch": 1.2043388561092458,
"grad_norm": 3.359375,
"learning_rate": 0.00013116043151140203,
"loss": 2.4222,
"step": 52100
},
{
"epoch": 1.2054946515565674,
"grad_norm": 3.265625,
"learning_rate": 0.00013104495452859614,
"loss": 2.4191,
"step": 52150
},
{
"epoch": 1.2066504470038892,
"grad_norm": 3.625,
"learning_rate": 0.00013092943171387123,
"loss": 2.4591,
"step": 52200
},
{
"epoch": 1.207806242451211,
"grad_norm": 3.578125,
"learning_rate": 0.00013081386323777455,
"loss": 2.4685,
"step": 52250
},
{
"epoch": 1.2089620378985326,
"grad_norm": 3.34375,
"learning_rate": 0.00013069824927092088,
"loss": 2.4781,
"step": 52300
},
{
"epoch": 1.2101178333458544,
"grad_norm": 3.578125,
"learning_rate": 0.0001305825899839919,
"loss": 2.4643,
"step": 52350
},
{
"epoch": 1.2112736287931762,
"grad_norm": 3.46875,
"learning_rate": 0.0001304668855477365,
"loss": 2.4096,
"step": 52400
},
{
"epoch": 1.2124294242404978,
"grad_norm": 3.375,
"learning_rate": 0.00013035113613297004,
"loss": 2.4343,
"step": 52450
},
{
"epoch": 1.2135852196878196,
"grad_norm": 3.5,
"learning_rate": 0.00013023534191057426,
"loss": 2.4775,
"step": 52500
},
{
"epoch": 1.2147410151351414,
"grad_norm": 3.625,
"learning_rate": 0.00013011950305149718,
"loss": 2.4449,
"step": 52550
},
{
"epoch": 1.2158968105824632,
"grad_norm": 3.5625,
"learning_rate": 0.00013000361972675264,
"loss": 2.4343,
"step": 52600
},
{
"epoch": 1.2170526060297848,
"grad_norm": 3.453125,
"learning_rate": 0.00012988769210742012,
"loss": 2.45,
"step": 52650
},
{
"epoch": 1.2182084014771066,
"grad_norm": 3.421875,
"learning_rate": 0.00012977172036464452,
"loss": 2.4507,
"step": 52700
},
{
"epoch": 1.2193641969244284,
"grad_norm": 3.53125,
"learning_rate": 0.00012965570466963588,
"loss": 2.4783,
"step": 52750
},
{
"epoch": 1.22051999237175,
"grad_norm": 3.484375,
"learning_rate": 0.00012953964519366907,
"loss": 2.4781,
"step": 52800
},
{
"epoch": 1.2216757878190718,
"grad_norm": 3.546875,
"learning_rate": 0.00012942354210808367,
"loss": 2.4388,
"step": 52850
},
{
"epoch": 1.2228315832663936,
"grad_norm": 3.40625,
"learning_rate": 0.0001293073955842836,
"loss": 2.4692,
"step": 52900
},
{
"epoch": 1.2239873787137152,
"grad_norm": 3.5625,
"learning_rate": 0.00012919120579373694,
"loss": 2.4692,
"step": 52950
},
{
"epoch": 1.225143174161037,
"grad_norm": 3.4375,
"learning_rate": 0.0001290749729079756,
"loss": 2.4548,
"step": 53000
},
{
"epoch": 1.2262989696083588,
"grad_norm": 3.625,
"learning_rate": 0.00012895869709859516,
"loss": 2.4838,
"step": 53050
},
{
"epoch": 1.2274547650556804,
"grad_norm": 3.453125,
"learning_rate": 0.00012884237853725452,
"loss": 2.4552,
"step": 53100
},
{
"epoch": 1.2286105605030022,
"grad_norm": 3.375,
"learning_rate": 0.0001287260173956757,
"loss": 2.474,
"step": 53150
},
{
"epoch": 1.229766355950324,
"grad_norm": 3.40625,
"learning_rate": 0.00012860961384564368,
"loss": 2.4277,
"step": 53200
},
{
"epoch": 1.2309221513976456,
"grad_norm": 3.453125,
"learning_rate": 0.0001284931680590059,
"loss": 2.4488,
"step": 53250
},
{
"epoch": 1.2320779468449674,
"grad_norm": 3.546875,
"learning_rate": 0.0001283766802076722,
"loss": 2.4537,
"step": 53300
},
{
"epoch": 1.2332337422922892,
"grad_norm": 3.3125,
"learning_rate": 0.0001282601504636146,
"loss": 2.4625,
"step": 53350
},
{
"epoch": 1.2343895377396108,
"grad_norm": 3.375,
"learning_rate": 0.00012814357899886686,
"loss": 2.4071,
"step": 53400
},
{
"epoch": 1.2355453331869326,
"grad_norm": 3.515625,
"learning_rate": 0.00012802696598552445,
"loss": 2.4627,
"step": 53450
},
{
"epoch": 1.2367011286342544,
"grad_norm": 3.5625,
"learning_rate": 0.00012791031159574407,
"loss": 2.4558,
"step": 53500
},
{
"epoch": 1.237856924081576,
"grad_norm": 3.46875,
"learning_rate": 0.00012779361600174353,
"loss": 2.4141,
"step": 53550
},
{
"epoch": 1.2390127195288978,
"grad_norm": 3.5,
"learning_rate": 0.0001276768793758015,
"loss": 2.4601,
"step": 53600
},
{
"epoch": 1.2401685149762196,
"grad_norm": 3.5,
"learning_rate": 0.00012756010189025722,
"loss": 2.4312,
"step": 53650
},
{
"epoch": 1.2413243104235412,
"grad_norm": 3.421875,
"learning_rate": 0.00012744328371751024,
"loss": 2.424,
"step": 53700
},
{
"epoch": 1.242480105870863,
"grad_norm": 3.46875,
"learning_rate": 0.00012732642503002017,
"loss": 2.4648,
"step": 53750
},
{
"epoch": 1.2436359013181848,
"grad_norm": 3.5,
"learning_rate": 0.00012720952600030644,
"loss": 2.3961,
"step": 53800
},
{
"epoch": 1.2447916967655064,
"grad_norm": 3.46875,
"learning_rate": 0.00012709258680094804,
"loss": 2.4423,
"step": 53850
},
{
"epoch": 1.2459474922128282,
"grad_norm": 3.609375,
"learning_rate": 0.0001269756076045833,
"loss": 2.4409,
"step": 53900
},
{
"epoch": 1.24710328766015,
"grad_norm": 3.421875,
"learning_rate": 0.0001268585885839095,
"loss": 2.4091,
"step": 53950
},
{
"epoch": 1.2482590831074716,
"grad_norm": 3.453125,
"learning_rate": 0.0001267415299116828,
"loss": 2.4516,
"step": 54000
},
{
"epoch": 1.2494148785547934,
"grad_norm": 3.578125,
"learning_rate": 0.00012662443176071785,
"loss": 2.4707,
"step": 54050
},
{
"epoch": 1.2505706740021152,
"grad_norm": 3.609375,
"learning_rate": 0.00012650729430388764,
"loss": 2.4177,
"step": 54100
},
{
"epoch": 1.2517264694494368,
"grad_norm": 3.46875,
"learning_rate": 0.00012639011771412313,
"loss": 2.3546,
"step": 54150
},
{
"epoch": 1.2528822648967586,
"grad_norm": 3.4375,
"learning_rate": 0.00012627290216441308,
"loss": 2.4482,
"step": 54200
},
{
"epoch": 1.2540380603440804,
"grad_norm": 3.59375,
"learning_rate": 0.00012615564782780374,
"loss": 2.4512,
"step": 54250
},
{
"epoch": 1.255193855791402,
"grad_norm": 3.484375,
"learning_rate": 0.00012603835487739868,
"loss": 2.461,
"step": 54300
},
{
"epoch": 1.2563496512387238,
"grad_norm": 3.390625,
"learning_rate": 0.0001259210234863584,
"loss": 2.4268,
"step": 54350
},
{
"epoch": 1.2575054466860456,
"grad_norm": 3.46875,
"learning_rate": 0.0001258036538279002,
"loss": 2.4977,
"step": 54400
},
{
"epoch": 1.2586612421333672,
"grad_norm": 3.296875,
"learning_rate": 0.00012568624607529788,
"loss": 2.4258,
"step": 54450
},
{
"epoch": 1.259817037580689,
"grad_norm": 3.515625,
"learning_rate": 0.00012556880040188144,
"loss": 2.4501,
"step": 54500
},
{
"epoch": 1.2609728330280108,
"grad_norm": 3.484375,
"learning_rate": 0.00012545131698103688,
"loss": 2.4272,
"step": 54550
},
{
"epoch": 1.2621286284753324,
"grad_norm": 3.484375,
"learning_rate": 0.0001253337959862059,
"loss": 2.4533,
"step": 54600
},
{
"epoch": 1.2632844239226542,
"grad_norm": 3.65625,
"learning_rate": 0.00012521623759088578,
"loss": 2.4377,
"step": 54650
},
{
"epoch": 1.264440219369976,
"grad_norm": 3.46875,
"learning_rate": 0.0001250986419686289,
"loss": 2.4597,
"step": 54700
},
{
"epoch": 1.2655960148172976,
"grad_norm": 3.5625,
"learning_rate": 0.00012498100929304263,
"loss": 2.4537,
"step": 54750
},
{
"epoch": 1.2667518102646194,
"grad_norm": 3.609375,
"learning_rate": 0.00012486333973778902,
"loss": 2.3968,
"step": 54800
},
{
"epoch": 1.2679076057119412,
"grad_norm": 3.421875,
"learning_rate": 0.0001247456334765846,
"loss": 2.4655,
"step": 54850
},
{
"epoch": 1.2690634011592627,
"grad_norm": 3.5,
"learning_rate": 0.00012462789068320017,
"loss": 2.4666,
"step": 54900
},
{
"epoch": 1.2702191966065846,
"grad_norm": 3.59375,
"learning_rate": 0.00012451011153146024,
"loss": 2.4857,
"step": 54950
},
{
"epoch": 1.2713749920539064,
"grad_norm": 3.59375,
"learning_rate": 0.0001243922961952432,
"loss": 2.4564,
"step": 55000
},
{
"epoch": 1.272530787501228,
"grad_norm": 3.875,
"learning_rate": 0.0001242744448484808,
"loss": 2.4368,
"step": 55050
},
{
"epoch": 1.2736865829485498,
"grad_norm": 3.546875,
"learning_rate": 0.0001241565576651579,
"loss": 2.4119,
"step": 55100
},
{
"epoch": 1.2748423783958716,
"grad_norm": 3.25,
"learning_rate": 0.00012403863481931236,
"loss": 2.4204,
"step": 55150
},
{
"epoch": 1.2759981738431931,
"grad_norm": 3.3125,
"learning_rate": 0.0001239206764850346,
"loss": 2.4494,
"step": 55200
},
{
"epoch": 1.277153969290515,
"grad_norm": 3.5625,
"learning_rate": 0.00012380268283646747,
"loss": 2.4398,
"step": 55250
},
{
"epoch": 1.2783097647378368,
"grad_norm": 3.390625,
"learning_rate": 0.00012368465404780598,
"loss": 2.4288,
"step": 55300
},
{
"epoch": 1.2794655601851583,
"grad_norm": 3.421875,
"learning_rate": 0.00012356659029329697,
"loss": 2.4292,
"step": 55350
},
{
"epoch": 1.2806213556324801,
"grad_norm": 3.34375,
"learning_rate": 0.00012344849174723895,
"loss": 2.4426,
"step": 55400
},
{
"epoch": 1.281777151079802,
"grad_norm": 3.484375,
"learning_rate": 0.00012333035858398168,
"loss": 2.4796,
"step": 55450
},
{
"epoch": 1.2829329465271235,
"grad_norm": 3.546875,
"learning_rate": 0.00012321219097792617,
"loss": 2.4156,
"step": 55500
},
{
"epoch": 1.2840887419744453,
"grad_norm": 3.53125,
"learning_rate": 0.00012309398910352425,
"loss": 2.4757,
"step": 55550
},
{
"epoch": 1.2852445374217671,
"grad_norm": 3.390625,
"learning_rate": 0.00012297575313527828,
"loss": 2.532,
"step": 55600
},
{
"epoch": 1.2864003328690887,
"grad_norm": 3.5,
"learning_rate": 0.00012285748324774098,
"loss": 2.4094,
"step": 55650
},
{
"epoch": 1.2875561283164105,
"grad_norm": 3.53125,
"learning_rate": 0.00012273917961551513,
"loss": 2.423,
"step": 55700
},
{
"epoch": 1.2887119237637323,
"grad_norm": 3.421875,
"learning_rate": 0.00012262084241325337,
"loss": 2.4914,
"step": 55750
},
{
"epoch": 1.289867719211054,
"grad_norm": 3.546875,
"learning_rate": 0.00012250247181565786,
"loss": 2.4259,
"step": 55800
},
{
"epoch": 1.2910235146583757,
"grad_norm": 3.375,
"learning_rate": 0.0001223840679974801,
"loss": 2.4359,
"step": 55850
},
{
"epoch": 1.2921793101056975,
"grad_norm": 3.65625,
"learning_rate": 0.0001222656311335206,
"loss": 2.3998,
"step": 55900
},
{
"epoch": 1.2933351055530191,
"grad_norm": 3.5,
"learning_rate": 0.00012214716139862864,
"loss": 2.4365,
"step": 55950
},
{
"epoch": 1.294490901000341,
"grad_norm": 3.40625,
"learning_rate": 0.00012202865896770207,
"loss": 2.4325,
"step": 56000
},
{
"epoch": 1.2956466964476627,
"grad_norm": 3.390625,
"learning_rate": 0.00012191012401568698,
"loss": 2.4289,
"step": 56050
},
{
"epoch": 1.2968024918949843,
"grad_norm": 3.59375,
"learning_rate": 0.00012179155671757754,
"loss": 2.397,
"step": 56100
},
{
"epoch": 1.2979582873423061,
"grad_norm": 3.421875,
"learning_rate": 0.00012167295724841557,
"loss": 2.4829,
"step": 56150
},
{
"epoch": 1.299114082789628,
"grad_norm": 3.6875,
"learning_rate": 0.00012155432578329042,
"loss": 2.5402,
"step": 56200
},
{
"epoch": 1.3002698782369495,
"grad_norm": 3.578125,
"learning_rate": 0.0001214356624973387,
"loss": 2.3915,
"step": 56250
},
{
"epoch": 1.3014256736842713,
"grad_norm": 3.4375,
"learning_rate": 0.00012131696756574403,
"loss": 2.4756,
"step": 56300
},
{
"epoch": 1.3025814691315931,
"grad_norm": 3.421875,
"learning_rate": 0.00012119824116373663,
"loss": 2.4938,
"step": 56350
},
{
"epoch": 1.3037372645789147,
"grad_norm": 3.5625,
"learning_rate": 0.00012107948346659331,
"loss": 2.4553,
"step": 56400
},
{
"epoch": 1.3048930600262365,
"grad_norm": 3.578125,
"learning_rate": 0.00012096069464963697,
"loss": 2.446,
"step": 56450
},
{
"epoch": 1.3060488554735583,
"grad_norm": 3.484375,
"learning_rate": 0.00012084187488823657,
"loss": 2.4595,
"step": 56500
},
{
"epoch": 1.30720465092088,
"grad_norm": 3.3125,
"learning_rate": 0.00012072302435780664,
"loss": 2.4635,
"step": 56550
},
{
"epoch": 1.3083604463682017,
"grad_norm": 3.59375,
"learning_rate": 0.00012060414323380726,
"loss": 2.3684,
"step": 56600
},
{
"epoch": 1.3095162418155235,
"grad_norm": 3.6875,
"learning_rate": 0.0001204852316917435,
"loss": 2.446,
"step": 56650
},
{
"epoch": 1.310672037262845,
"grad_norm": 3.46875,
"learning_rate": 0.00012036628990716547,
"loss": 2.4636,
"step": 56700
},
{
"epoch": 1.311827832710167,
"grad_norm": 3.46875,
"learning_rate": 0.00012024731805566793,
"loss": 2.4151,
"step": 56750
},
{
"epoch": 1.3129836281574887,
"grad_norm": 3.359375,
"learning_rate": 0.00012012831631288998,
"loss": 2.4755,
"step": 56800
},
{
"epoch": 1.3141394236048103,
"grad_norm": 3.59375,
"learning_rate": 0.00012000928485451486,
"loss": 2.4238,
"step": 56850
},
{
"epoch": 1.315295219052132,
"grad_norm": 3.515625,
"learning_rate": 0.00011989022385626967,
"loss": 2.4486,
"step": 56900
},
{
"epoch": 1.316451014499454,
"grad_norm": 3.40625,
"learning_rate": 0.0001197711334939251,
"loss": 2.4172,
"step": 56950
},
{
"epoch": 1.3176068099467755,
"grad_norm": 3.34375,
"learning_rate": 0.00011965201394329529,
"loss": 2.4352,
"step": 57000
},
{
"epoch": 1.3187626053940973,
"grad_norm": 3.546875,
"learning_rate": 0.00011953286538023735,
"loss": 2.442,
"step": 57050
},
{
"epoch": 1.319918400841419,
"grad_norm": 3.46875,
"learning_rate": 0.0001194136879806513,
"loss": 2.4202,
"step": 57100
},
{
"epoch": 1.321074196288741,
"grad_norm": 3.484375,
"learning_rate": 0.00011929448192047968,
"loss": 2.4186,
"step": 57150
},
{
"epoch": 1.3222299917360625,
"grad_norm": 3.453125,
"learning_rate": 0.00011917524737570737,
"loss": 2.3965,
"step": 57200
},
{
"epoch": 1.3233857871833843,
"grad_norm": 3.421875,
"learning_rate": 0.00011905598452236136,
"loss": 2.4067,
"step": 57250
},
{
"epoch": 1.324541582630706,
"grad_norm": 3.484375,
"learning_rate": 0.00011893669353651031,
"loss": 2.4234,
"step": 57300
},
{
"epoch": 1.3256973780780277,
"grad_norm": 3.453125,
"learning_rate": 0.00011881737459426446,
"loss": 2.4617,
"step": 57350
},
{
"epoch": 1.3268531735253495,
"grad_norm": 3.453125,
"learning_rate": 0.00011869802787177537,
"loss": 2.3895,
"step": 57400
},
{
"epoch": 1.3280089689726713,
"grad_norm": 3.546875,
"learning_rate": 0.00011857865354523556,
"loss": 2.4561,
"step": 57450
},
{
"epoch": 1.3291647644199929,
"grad_norm": 3.46875,
"learning_rate": 0.00011845925179087833,
"loss": 2.4377,
"step": 57500
},
{
"epoch": 1.3303205598673147,
"grad_norm": 3.375,
"learning_rate": 0.00011833982278497749,
"loss": 2.4407,
"step": 57550
},
{
"epoch": 1.3314763553146365,
"grad_norm": 3.4375,
"learning_rate": 0.00011822036670384698,
"loss": 2.4198,
"step": 57600
},
{
"epoch": 1.332632150761958,
"grad_norm": 3.203125,
"learning_rate": 0.00011810088372384085,
"loss": 2.3881,
"step": 57650
},
{
"epoch": 1.3337879462092799,
"grad_norm": 3.53125,
"learning_rate": 0.00011798137402135277,
"loss": 2.4491,
"step": 57700
},
{
"epoch": 1.3349437416566017,
"grad_norm": 3.625,
"learning_rate": 0.00011786183777281587,
"loss": 2.4336,
"step": 57750
},
{
"epoch": 1.3360995371039235,
"grad_norm": 3.640625,
"learning_rate": 0.00011774227515470256,
"loss": 2.4258,
"step": 57800
},
{
"epoch": 1.337255332551245,
"grad_norm": 3.28125,
"learning_rate": 0.000117622686343524,
"loss": 2.3757,
"step": 57850
},
{
"epoch": 1.3384111279985669,
"grad_norm": 3.453125,
"learning_rate": 0.0001175030715158302,
"loss": 2.4153,
"step": 57900
},
{
"epoch": 1.3395669234458887,
"grad_norm": 3.390625,
"learning_rate": 0.00011738343084820947,
"loss": 2.451,
"step": 57950
},
{
"epoch": 1.3407227188932103,
"grad_norm": 3.546875,
"learning_rate": 0.00011726376451728831,
"loss": 2.4501,
"step": 58000
},
{
"epoch": 1.341878514340532,
"grad_norm": 3.484375,
"learning_rate": 0.00011714407269973111,
"loss": 2.4135,
"step": 58050
},
{
"epoch": 1.3430343097878539,
"grad_norm": 3.5,
"learning_rate": 0.00011702435557223987,
"loss": 2.4268,
"step": 58100
},
{
"epoch": 1.3441901052351755,
"grad_norm": 3.3125,
"learning_rate": 0.00011690461331155393,
"loss": 2.4109,
"step": 58150
},
{
"epoch": 1.3453459006824973,
"grad_norm": 3.6875,
"learning_rate": 0.00011678484609444982,
"loss": 2.4105,
"step": 58200
},
{
"epoch": 1.346501696129819,
"grad_norm": 3.53125,
"learning_rate": 0.00011666505409774082,
"loss": 2.4375,
"step": 58250
},
{
"epoch": 1.3476574915771407,
"grad_norm": 3.578125,
"learning_rate": 0.00011654523749827684,
"loss": 2.4052,
"step": 58300
},
{
"epoch": 1.3488132870244625,
"grad_norm": 3.296875,
"learning_rate": 0.00011642539647294406,
"loss": 2.4454,
"step": 58350
},
{
"epoch": 1.3499690824717843,
"grad_norm": 3.515625,
"learning_rate": 0.00011630553119866481,
"loss": 2.3899,
"step": 58400
},
{
"epoch": 1.3511248779191058,
"grad_norm": 3.578125,
"learning_rate": 0.00011618564185239716,
"loss": 2.4592,
"step": 58450
},
{
"epoch": 1.3522806733664277,
"grad_norm": 3.421875,
"learning_rate": 0.00011606572861113474,
"loss": 2.4272,
"step": 58500
},
{
"epoch": 1.3534364688137495,
"grad_norm": 3.5,
"learning_rate": 0.00011594579165190643,
"loss": 2.4118,
"step": 58550
},
{
"epoch": 1.354592264261071,
"grad_norm": 3.40625,
"learning_rate": 0.0001158258311517761,
"loss": 2.4527,
"step": 58600
},
{
"epoch": 1.3557480597083928,
"grad_norm": 3.734375,
"learning_rate": 0.00011570584728784248,
"loss": 2.4012,
"step": 58650
},
{
"epoch": 1.3569038551557147,
"grad_norm": 3.640625,
"learning_rate": 0.00011558584023723869,
"loss": 2.4338,
"step": 58700
},
{
"epoch": 1.3580596506030362,
"grad_norm": 3.515625,
"learning_rate": 0.00011546581017713214,
"loss": 2.4052,
"step": 58750
},
{
"epoch": 1.359215446050358,
"grad_norm": 3.546875,
"learning_rate": 0.00011534575728472412,
"loss": 2.3941,
"step": 58800
},
{
"epoch": 1.3603712414976799,
"grad_norm": 3.578125,
"learning_rate": 0.00011522568173724975,
"loss": 2.4312,
"step": 58850
},
{
"epoch": 1.3615270369450014,
"grad_norm": 3.453125,
"learning_rate": 0.00011510558371197753,
"loss": 2.3817,
"step": 58900
},
{
"epoch": 1.3626828323923232,
"grad_norm": 3.640625,
"learning_rate": 0.00011498546338620916,
"loss": 2.4281,
"step": 58950
},
{
"epoch": 1.363838627839645,
"grad_norm": 3.390625,
"learning_rate": 0.00011486532093727924,
"loss": 2.4262,
"step": 59000
},
{
"epoch": 1.3649944232869666,
"grad_norm": 3.625,
"learning_rate": 0.00011474515654255504,
"loss": 2.43,
"step": 59050
},
{
"epoch": 1.3661502187342884,
"grad_norm": 3.421875,
"learning_rate": 0.00011462497037943622,
"loss": 2.4498,
"step": 59100
},
{
"epoch": 1.3673060141816102,
"grad_norm": 3.4375,
"learning_rate": 0.00011450476262535463,
"loss": 2.4732,
"step": 59150
},
{
"epoch": 1.3684618096289318,
"grad_norm": 3.328125,
"learning_rate": 0.00011438453345777393,
"loss": 2.4574,
"step": 59200
},
{
"epoch": 1.3696176050762536,
"grad_norm": 3.296875,
"learning_rate": 0.00011426428305418944,
"loss": 2.4564,
"step": 59250
},
{
"epoch": 1.3707734005235754,
"grad_norm": 3.515625,
"learning_rate": 0.00011414401159212778,
"loss": 2.3652,
"step": 59300
},
{
"epoch": 1.371929195970897,
"grad_norm": 3.5625,
"learning_rate": 0.0001140237192491467,
"loss": 2.3653,
"step": 59350
},
{
"epoch": 1.3730849914182188,
"grad_norm": 3.5625,
"learning_rate": 0.00011390340620283477,
"loss": 2.4714,
"step": 59400
},
{
"epoch": 1.3742407868655406,
"grad_norm": 3.59375,
"learning_rate": 0.00011378307263081113,
"loss": 2.4355,
"step": 59450
},
{
"epoch": 1.3753965823128622,
"grad_norm": 3.3125,
"learning_rate": 0.00011366271871072521,
"loss": 2.3665,
"step": 59500
},
{
"epoch": 1.376552377760184,
"grad_norm": 3.359375,
"learning_rate": 0.00011354234462025644,
"loss": 2.3788,
"step": 59550
},
{
"epoch": 1.3777081732075058,
"grad_norm": 3.5,
"learning_rate": 0.00011342195053711409,
"loss": 2.4185,
"step": 59600
},
{
"epoch": 1.3788639686548274,
"grad_norm": 3.296875,
"learning_rate": 0.00011330153663903695,
"loss": 2.4278,
"step": 59650
},
{
"epoch": 1.3800197641021492,
"grad_norm": 3.421875,
"learning_rate": 0.00011318110310379301,
"loss": 2.4273,
"step": 59700
},
{
"epoch": 1.381175559549471,
"grad_norm": 3.53125,
"learning_rate": 0.00011306065010917932,
"loss": 2.4447,
"step": 59750
},
{
"epoch": 1.3823313549967926,
"grad_norm": 3.4375,
"learning_rate": 0.00011294017783302156,
"loss": 2.4411,
"step": 59800
},
{
"epoch": 1.3834871504441144,
"grad_norm": 3.40625,
"learning_rate": 0.00011281968645317396,
"loss": 2.421,
"step": 59850
},
{
"epoch": 1.3846429458914362,
"grad_norm": 3.5,
"learning_rate": 0.00011269917614751892,
"loss": 2.3975,
"step": 59900
},
{
"epoch": 1.3857987413387578,
"grad_norm": 3.453125,
"learning_rate": 0.00011257864709396684,
"loss": 2.4024,
"step": 59950
},
{
"epoch": 1.3869545367860796,
"grad_norm": 3.78125,
"learning_rate": 0.00011245809947045562,
"loss": 2.3894,
"step": 60000
},
{
"epoch": 1.3881103322334014,
"grad_norm": 3.390625,
"learning_rate": 0.00011233753345495078,
"loss": 2.4111,
"step": 60050
},
{
"epoch": 1.389266127680723,
"grad_norm": 3.296875,
"learning_rate": 0.00011221694922544491,
"loss": 2.4587,
"step": 60100
},
{
"epoch": 1.3904219231280448,
"grad_norm": 3.34375,
"learning_rate": 0.0001120963469599575,
"loss": 2.4135,
"step": 60150
},
{
"epoch": 1.3915777185753666,
"grad_norm": 3.53125,
"learning_rate": 0.00011197572683653461,
"loss": 2.4611,
"step": 60200
},
{
"epoch": 1.3927335140226882,
"grad_norm": 3.390625,
"learning_rate": 0.00011185508903324873,
"loss": 2.4433,
"step": 60250
},
{
"epoch": 1.39388930947001,
"grad_norm": 3.5,
"learning_rate": 0.00011173443372819843,
"loss": 2.4424,
"step": 60300
},
{
"epoch": 1.3950451049173318,
"grad_norm": 3.578125,
"learning_rate": 0.00011161376109950811,
"loss": 2.4081,
"step": 60350
},
{
"epoch": 1.3962009003646534,
"grad_norm": 3.53125,
"learning_rate": 0.0001114930713253278,
"loss": 2.398,
"step": 60400
},
{
"epoch": 1.3973566958119752,
"grad_norm": 3.59375,
"learning_rate": 0.00011137236458383276,
"loss": 2.411,
"step": 60450
},
{
"epoch": 1.398512491259297,
"grad_norm": 3.296875,
"learning_rate": 0.0001112516410532233,
"loss": 2.4367,
"step": 60500
},
{
"epoch": 1.3996682867066186,
"grad_norm": 3.53125,
"learning_rate": 0.0001111309009117246,
"loss": 2.3649,
"step": 60550
},
{
"epoch": 1.4008240821539404,
"grad_norm": 3.71875,
"learning_rate": 0.00011101014433758631,
"loss": 2.3737,
"step": 60600
},
{
"epoch": 1.4019798776012622,
"grad_norm": 3.375,
"learning_rate": 0.00011088937150908236,
"loss": 2.3795,
"step": 60650
},
{
"epoch": 1.4031356730485838,
"grad_norm": 3.53125,
"learning_rate": 0.0001107685826045106,
"loss": 2.4014,
"step": 60700
},
{
"epoch": 1.4042914684959056,
"grad_norm": 3.5,
"learning_rate": 0.00011064777780219272,
"loss": 2.4026,
"step": 60750
},
{
"epoch": 1.4054472639432274,
"grad_norm": 3.71875,
"learning_rate": 0.0001105269572804738,
"loss": 2.4324,
"step": 60800
},
{
"epoch": 1.406603059390549,
"grad_norm": 3.375,
"learning_rate": 0.00011040612121772218,
"loss": 2.4012,
"step": 60850
},
{
"epoch": 1.4077588548378708,
"grad_norm": 3.515625,
"learning_rate": 0.00011028526979232913,
"loss": 2.4071,
"step": 60900
},
{
"epoch": 1.4089146502851926,
"grad_norm": 3.453125,
"learning_rate": 0.00011016440318270857,
"loss": 2.4244,
"step": 60950
},
{
"epoch": 1.4100704457325142,
"grad_norm": 3.34375,
"learning_rate": 0.0001100435215672969,
"loss": 2.4308,
"step": 61000
},
{
"epoch": 1.411226241179836,
"grad_norm": 3.46875,
"learning_rate": 0.00010992262512455258,
"loss": 2.439,
"step": 61050
},
{
"epoch": 1.4123820366271578,
"grad_norm": 3.484375,
"learning_rate": 0.0001098017140329561,
"loss": 2.442,
"step": 61100
},
{
"epoch": 1.4135378320744794,
"grad_norm": 3.484375,
"learning_rate": 0.00010968078847100936,
"loss": 2.3854,
"step": 61150
},
{
"epoch": 1.4146936275218012,
"grad_norm": 3.375,
"learning_rate": 0.00010955984861723582,
"loss": 2.3899,
"step": 61200
},
{
"epoch": 1.415849422969123,
"grad_norm": 3.453125,
"learning_rate": 0.00010943889465018001,
"loss": 2.402,
"step": 61250
},
{
"epoch": 1.4170052184164446,
"grad_norm": 3.53125,
"learning_rate": 0.00010931792674840718,
"loss": 2.3878,
"step": 61300
},
{
"epoch": 1.4181610138637664,
"grad_norm": 3.65625,
"learning_rate": 0.00010919694509050327,
"loss": 2.3943,
"step": 61350
},
{
"epoch": 1.4193168093110882,
"grad_norm": 3.53125,
"learning_rate": 0.00010907594985507446,
"loss": 2.3948,
"step": 61400
},
{
"epoch": 1.4204726047584098,
"grad_norm": 3.453125,
"learning_rate": 0.00010895494122074703,
"loss": 2.4131,
"step": 61450
},
{
"epoch": 1.4216284002057316,
"grad_norm": 3.4375,
"learning_rate": 0.00010883391936616702,
"loss": 2.42,
"step": 61500
},
{
"epoch": 1.4227841956530534,
"grad_norm": 3.625,
"learning_rate": 0.00010871288446999994,
"loss": 2.3698,
"step": 61550
},
{
"epoch": 1.423939991100375,
"grad_norm": 3.375,
"learning_rate": 0.00010859183671093063,
"loss": 2.4076,
"step": 61600
},
{
"epoch": 1.4250957865476968,
"grad_norm": 3.546875,
"learning_rate": 0.00010847077626766284,
"loss": 2.4204,
"step": 61650
},
{
"epoch": 1.4262515819950186,
"grad_norm": 3.484375,
"learning_rate": 0.00010834970331891914,
"loss": 2.417,
"step": 61700
},
{
"epoch": 1.4274073774423401,
"grad_norm": 3.609375,
"learning_rate": 0.00010822861804344042,
"loss": 2.4246,
"step": 61750
},
{
"epoch": 1.428563172889662,
"grad_norm": 3.46875,
"learning_rate": 0.00010810752061998599,
"loss": 2.3776,
"step": 61800
},
{
"epoch": 1.4297189683369838,
"grad_norm": 3.390625,
"learning_rate": 0.00010798641122733286,
"loss": 2.4248,
"step": 61850
},
{
"epoch": 1.4308747637843053,
"grad_norm": 3.28125,
"learning_rate": 0.00010786529004427584,
"loss": 2.382,
"step": 61900
},
{
"epoch": 1.4320305592316271,
"grad_norm": 3.453125,
"learning_rate": 0.00010774415724962711,
"loss": 2.3833,
"step": 61950
},
{
"epoch": 1.433186354678949,
"grad_norm": 3.640625,
"learning_rate": 0.00010762301302221602,
"loss": 2.3907,
"step": 62000
},
{
"epoch": 1.4343421501262705,
"grad_norm": 3.625,
"learning_rate": 0.00010750185754088878,
"loss": 2.404,
"step": 62050
},
{
"epoch": 1.4354979455735923,
"grad_norm": 3.46875,
"learning_rate": 0.0001073806909845082,
"loss": 2.4161,
"step": 62100
},
{
"epoch": 1.4366537410209141,
"grad_norm": 3.6875,
"learning_rate": 0.00010725951353195345,
"loss": 2.4353,
"step": 62150
},
{
"epoch": 1.4378095364682357,
"grad_norm": 3.5,
"learning_rate": 0.00010713832536211976,
"loss": 2.3838,
"step": 62200
},
{
"epoch": 1.4389653319155575,
"grad_norm": 3.65625,
"learning_rate": 0.00010701712665391828,
"loss": 2.3691,
"step": 62250
},
{
"epoch": 1.4401211273628793,
"grad_norm": 3.515625,
"learning_rate": 0.00010689591758627564,
"loss": 2.447,
"step": 62300
},
{
"epoch": 1.441276922810201,
"grad_norm": 3.40625,
"learning_rate": 0.00010677469833813372,
"loss": 2.4055,
"step": 62350
},
{
"epoch": 1.4424327182575227,
"grad_norm": 3.703125,
"learning_rate": 0.0001066534690884495,
"loss": 2.389,
"step": 62400
},
{
"epoch": 1.4435885137048445,
"grad_norm": 3.625,
"learning_rate": 0.00010653223001619475,
"loss": 2.4264,
"step": 62450
},
{
"epoch": 1.4447443091521661,
"grad_norm": 3.578125,
"learning_rate": 0.00010641098130035562,
"loss": 2.3731,
"step": 62500
},
{
"epoch": 1.445900104599488,
"grad_norm": 3.359375,
"learning_rate": 0.00010628972311993266,
"loss": 2.3837,
"step": 62550
},
{
"epoch": 1.4470559000468097,
"grad_norm": 3.421875,
"learning_rate": 0.00010616845565394027,
"loss": 2.4124,
"step": 62600
},
{
"epoch": 1.4482116954941313,
"grad_norm": 3.453125,
"learning_rate": 0.0001060471790814066,
"loss": 2.4056,
"step": 62650
},
{
"epoch": 1.4493674909414531,
"grad_norm": 3.515625,
"learning_rate": 0.00010592589358137322,
"loss": 2.3662,
"step": 62700
},
{
"epoch": 1.450523286388775,
"grad_norm": 3.390625,
"learning_rate": 0.00010580459933289492,
"loss": 2.4258,
"step": 62750
},
{
"epoch": 1.4516790818360967,
"grad_norm": 3.328125,
"learning_rate": 0.00010568329651503943,
"loss": 2.3697,
"step": 62800
},
{
"epoch": 1.4528348772834183,
"grad_norm": 3.390625,
"learning_rate": 0.00010556198530688701,
"loss": 2.4139,
"step": 62850
},
{
"epoch": 1.4539906727307401,
"grad_norm": 3.484375,
"learning_rate": 0.00010544066588753044,
"loss": 2.4143,
"step": 62900
},
{
"epoch": 1.455146468178062,
"grad_norm": 3.6875,
"learning_rate": 0.0001053193384360745,
"loss": 2.3537,
"step": 62950
},
{
"epoch": 1.4563022636253835,
"grad_norm": 3.53125,
"learning_rate": 0.00010519800313163597,
"loss": 2.4131,
"step": 63000
},
{
"epoch": 1.4574580590727053,
"grad_norm": 3.765625,
"learning_rate": 0.00010507666015334312,
"loss": 2.4109,
"step": 63050
},
{
"epoch": 1.4586138545200271,
"grad_norm": 3.5,
"learning_rate": 0.00010495530968033556,
"loss": 2.3809,
"step": 63100
},
{
"epoch": 1.4597696499673487,
"grad_norm": 3.21875,
"learning_rate": 0.000104833951891764,
"loss": 2.4746,
"step": 63150
},
{
"epoch": 1.4609254454146705,
"grad_norm": 3.5,
"learning_rate": 0.00010471258696678992,
"loss": 2.3919,
"step": 63200
},
{
"epoch": 1.4620812408619923,
"grad_norm": 3.53125,
"learning_rate": 0.00010459121508458533,
"loss": 2.4055,
"step": 63250
},
{
"epoch": 1.463237036309314,
"grad_norm": 3.46875,
"learning_rate": 0.00010446983642433259,
"loss": 2.3537,
"step": 63300
},
{
"epoch": 1.4643928317566357,
"grad_norm": 3.328125,
"learning_rate": 0.00010434845116522394,
"loss": 2.4049,
"step": 63350
},
{
"epoch": 1.4655486272039575,
"grad_norm": 3.34375,
"learning_rate": 0.00010422705948646143,
"loss": 2.3458,
"step": 63400
},
{
"epoch": 1.466704422651279,
"grad_norm": 3.609375,
"learning_rate": 0.0001041056615672566,
"loss": 2.4055,
"step": 63450
},
{
"epoch": 1.467860218098601,
"grad_norm": 3.546875,
"learning_rate": 0.00010398425758683022,
"loss": 2.4059,
"step": 63500
},
{
"epoch": 1.4690160135459227,
"grad_norm": 3.46875,
"learning_rate": 0.00010386284772441192,
"loss": 2.3789,
"step": 63550
},
{
"epoch": 1.4701718089932445,
"grad_norm": 3.65625,
"learning_rate": 0.00010374143215924008,
"loss": 2.3911,
"step": 63600
},
{
"epoch": 1.471327604440566,
"grad_norm": 3.625,
"learning_rate": 0.0001036200110705615,
"loss": 2.337,
"step": 63650
},
{
"epoch": 1.472483399887888,
"grad_norm": 3.5,
"learning_rate": 0.00010349858463763113,
"loss": 2.3958,
"step": 63700
},
{
"epoch": 1.4736391953352097,
"grad_norm": 3.5,
"learning_rate": 0.0001033771530397118,
"loss": 2.418,
"step": 63750
},
{
"epoch": 1.4747949907825313,
"grad_norm": 3.390625,
"learning_rate": 0.00010325571645607392,
"loss": 2.3813,
"step": 63800
},
{
"epoch": 1.475950786229853,
"grad_norm": 3.671875,
"learning_rate": 0.00010313427506599534,
"loss": 2.3715,
"step": 63850
},
{
"epoch": 1.477106581677175,
"grad_norm": 3.65625,
"learning_rate": 0.000103012829048761,
"loss": 2.3773,
"step": 63900
},
{
"epoch": 1.4782623771244965,
"grad_norm": 3.734375,
"learning_rate": 0.00010289137858366258,
"loss": 2.3401,
"step": 63950
},
{
"epoch": 1.4794181725718183,
"grad_norm": 3.40625,
"learning_rate": 0.00010276992384999849,
"loss": 2.4401,
"step": 64000
},
{
"epoch": 1.48057396801914,
"grad_norm": 3.578125,
"learning_rate": 0.00010264846502707324,
"loss": 2.3876,
"step": 64050
},
{
"epoch": 1.4817297634664617,
"grad_norm": 3.484375,
"learning_rate": 0.0001025270022941975,
"loss": 2.3975,
"step": 64100
},
{
"epoch": 1.4828855589137835,
"grad_norm": 3.5,
"learning_rate": 0.00010240553583068773,
"loss": 2.3809,
"step": 64150
},
{
"epoch": 1.4840413543611053,
"grad_norm": 3.28125,
"learning_rate": 0.00010228406581586583,
"loss": 2.4176,
"step": 64200
},
{
"epoch": 1.4851971498084269,
"grad_norm": 3.34375,
"learning_rate": 0.00010216259242905899,
"loss": 2.4046,
"step": 64250
},
{
"epoch": 1.4863529452557487,
"grad_norm": 3.296875,
"learning_rate": 0.00010204111584959932,
"loss": 2.3288,
"step": 64300
},
{
"epoch": 1.4875087407030705,
"grad_norm": 3.625,
"learning_rate": 0.00010191963625682373,
"loss": 2.382,
"step": 64350
},
{
"epoch": 1.488664536150392,
"grad_norm": 3.234375,
"learning_rate": 0.00010179815383007352,
"loss": 2.3512,
"step": 64400
},
{
"epoch": 1.4898203315977139,
"grad_norm": 3.4375,
"learning_rate": 0.00010167666874869416,
"loss": 2.3529,
"step": 64450
},
{
"epoch": 1.4909761270450357,
"grad_norm": 3.546875,
"learning_rate": 0.0001015551811920351,
"loss": 2.4369,
"step": 64500
},
{
"epoch": 1.4921319224923573,
"grad_norm": 3.71875,
"learning_rate": 0.00010143369133944938,
"loss": 2.4239,
"step": 64550
},
{
"epoch": 1.493287717939679,
"grad_norm": 3.640625,
"learning_rate": 0.00010131219937029348,
"loss": 2.4369,
"step": 64600
},
{
"epoch": 1.4944435133870009,
"grad_norm": 3.5625,
"learning_rate": 0.00010119070546392695,
"loss": 2.4352,
"step": 64650
},
{
"epoch": 1.4955993088343225,
"grad_norm": 3.546875,
"learning_rate": 0.00010106920979971226,
"loss": 2.354,
"step": 64700
},
{
"epoch": 1.4967551042816443,
"grad_norm": 3.5625,
"learning_rate": 0.00010094771255701444,
"loss": 2.3864,
"step": 64750
},
{
"epoch": 1.497910899728966,
"grad_norm": 3.703125,
"learning_rate": 0.00010082621391520086,
"loss": 2.4267,
"step": 64800
},
{
"epoch": 1.4990666951762877,
"grad_norm": 3.65625,
"learning_rate": 0.00010070471405364097,
"loss": 2.3831,
"step": 64850
},
{
"epoch": 1.5002224906236095,
"grad_norm": 3.515625,
"learning_rate": 0.00010058321315170596,
"loss": 2.3986,
"step": 64900
},
{
"epoch": 1.5013782860709313,
"grad_norm": 3.3125,
"learning_rate": 0.00010046171138876865,
"loss": 2.3412,
"step": 64950
},
{
"epoch": 1.5025340815182529,
"grad_norm": 3.40625,
"learning_rate": 0.00010034020894420306,
"loss": 2.364,
"step": 65000
},
{
"epoch": 1.5036898769655747,
"grad_norm": 3.578125,
"learning_rate": 0.00010021870599738423,
"loss": 2.4282,
"step": 65050
},
{
"epoch": 1.5048456724128965,
"grad_norm": 3.46875,
"learning_rate": 0.00010009720272768792,
"loss": 2.3844,
"step": 65100
},
{
"epoch": 1.506001467860218,
"grad_norm": 3.421875,
"learning_rate": 9.99756993144905e-05,
"loss": 2.4311,
"step": 65150
},
{
"epoch": 1.5071572633075399,
"grad_norm": 3.546875,
"learning_rate": 9.985419593716836e-05,
"loss": 2.37,
"step": 65200
},
{
"epoch": 1.5083130587548617,
"grad_norm": 3.46875,
"learning_rate": 9.973269277509791e-05,
"loss": 2.3961,
"step": 65250
},
{
"epoch": 1.5094688542021832,
"grad_norm": 3.65625,
"learning_rate": 9.961119000765531e-05,
"loss": 2.4096,
"step": 65300
},
{
"epoch": 1.510624649649505,
"grad_norm": 3.546875,
"learning_rate": 9.948968781421607e-05,
"loss": 2.3637,
"step": 65350
},
{
"epoch": 1.5117804450968269,
"grad_norm": 3.578125,
"learning_rate": 9.936818637415485e-05,
"loss": 2.363,
"step": 65400
},
{
"epoch": 1.5129362405441484,
"grad_norm": 3.703125,
"learning_rate": 9.924668586684527e-05,
"loss": 2.3965,
"step": 65450
},
{
"epoch": 1.5140920359914702,
"grad_norm": 3.640625,
"learning_rate": 9.912518647165945e-05,
"loss": 2.368,
"step": 65500
},
{
"epoch": 1.515247831438792,
"grad_norm": 3.6875,
"learning_rate": 9.900368836796795e-05,
"loss": 2.4017,
"step": 65550
},
{
"epoch": 1.5164036268861136,
"grad_norm": 3.484375,
"learning_rate": 9.888219173513942e-05,
"loss": 2.3933,
"step": 65600
},
{
"epoch": 1.5175594223334354,
"grad_norm": 3.46875,
"learning_rate": 9.876069675254034e-05,
"loss": 2.3664,
"step": 65650
},
{
"epoch": 1.5187152177807572,
"grad_norm": 3.4375,
"learning_rate": 9.863920359953474e-05,
"loss": 2.3548,
"step": 65700
},
{
"epoch": 1.5198710132280788,
"grad_norm": 3.4375,
"learning_rate": 9.851771245548388e-05,
"loss": 2.4018,
"step": 65750
},
{
"epoch": 1.5210268086754006,
"grad_norm": 3.484375,
"learning_rate": 9.839622349974621e-05,
"loss": 2.373,
"step": 65800
},
{
"epoch": 1.5221826041227224,
"grad_norm": 3.515625,
"learning_rate": 9.827473691167689e-05,
"loss": 2.3623,
"step": 65850
},
{
"epoch": 1.523338399570044,
"grad_norm": 3.265625,
"learning_rate": 9.815325287062748e-05,
"loss": 2.3677,
"step": 65900
},
{
"epoch": 1.5244941950173658,
"grad_norm": 3.265625,
"learning_rate": 9.803177155594587e-05,
"loss": 2.417,
"step": 65950
},
{
"epoch": 1.5256499904646876,
"grad_norm": 3.40625,
"learning_rate": 9.791029314697593e-05,
"loss": 2.3894,
"step": 66000
},
{
"epoch": 1.5268057859120092,
"grad_norm": 3.484375,
"learning_rate": 9.778881782305725e-05,
"loss": 2.4083,
"step": 66050
},
{
"epoch": 1.527961581359331,
"grad_norm": 3.53125,
"learning_rate": 9.766734576352478e-05,
"loss": 2.3732,
"step": 66100
},
{
"epoch": 1.5291173768066528,
"grad_norm": 3.796875,
"learning_rate": 9.754587714770878e-05,
"loss": 2.402,
"step": 66150
},
{
"epoch": 1.5302731722539744,
"grad_norm": 3.53125,
"learning_rate": 9.742441215493431e-05,
"loss": 2.351,
"step": 66200
},
{
"epoch": 1.5314289677012962,
"grad_norm": 3.484375,
"learning_rate": 9.730295096452112e-05,
"loss": 2.4011,
"step": 66250
},
{
"epoch": 1.532584763148618,
"grad_norm": 3.515625,
"learning_rate": 9.718149375578337e-05,
"loss": 2.3992,
"step": 66300
},
{
"epoch": 1.5337405585959396,
"grad_norm": 3.578125,
"learning_rate": 9.706004070802932e-05,
"loss": 2.4035,
"step": 66350
},
{
"epoch": 1.5348963540432614,
"grad_norm": 3.359375,
"learning_rate": 9.693859200056113e-05,
"loss": 2.3704,
"step": 66400
},
{
"epoch": 1.5360521494905832,
"grad_norm": 3.515625,
"learning_rate": 9.681714781267446e-05,
"loss": 2.3361,
"step": 66450
},
{
"epoch": 1.5372079449379048,
"grad_norm": 3.703125,
"learning_rate": 9.669570832365838e-05,
"loss": 2.3592,
"step": 66500
},
{
"epoch": 1.5383637403852266,
"grad_norm": 3.46875,
"learning_rate": 9.657427371279498e-05,
"loss": 2.4444,
"step": 66550
},
{
"epoch": 1.5395195358325484,
"grad_norm": 3.640625,
"learning_rate": 9.645284415935916e-05,
"loss": 2.3763,
"step": 66600
},
{
"epoch": 1.54067533127987,
"grad_norm": 3.46875,
"learning_rate": 9.63314198426184e-05,
"loss": 2.3437,
"step": 66650
},
{
"epoch": 1.5418311267271918,
"grad_norm": 3.375,
"learning_rate": 9.621000094183238e-05,
"loss": 2.3798,
"step": 66700
},
{
"epoch": 1.5429869221745136,
"grad_norm": 3.640625,
"learning_rate": 9.60885876362528e-05,
"loss": 2.4158,
"step": 66750
},
{
"epoch": 1.5441427176218352,
"grad_norm": 3.703125,
"learning_rate": 9.596718010512314e-05,
"loss": 2.3443,
"step": 66800
},
{
"epoch": 1.545298513069157,
"grad_norm": 3.734375,
"learning_rate": 9.584577852767833e-05,
"loss": 2.3497,
"step": 66850
},
{
"epoch": 1.5464543085164788,
"grad_norm": 3.609375,
"learning_rate": 9.572438308314446e-05,
"loss": 2.3927,
"step": 66900
},
{
"epoch": 1.5476101039638004,
"grad_norm": 3.65625,
"learning_rate": 9.560299395073863e-05,
"loss": 2.3941,
"step": 66950
},
{
"epoch": 1.5487658994111222,
"grad_norm": 3.53125,
"learning_rate": 9.548161130966863e-05,
"loss": 2.3322,
"step": 67000
},
{
"epoch": 1.549921694858444,
"grad_norm": 3.640625,
"learning_rate": 9.536023533913263e-05,
"loss": 2.365,
"step": 67050
},
{
"epoch": 1.5510774903057656,
"grad_norm": 3.5,
"learning_rate": 9.523886621831897e-05,
"loss": 2.4273,
"step": 67100
},
{
"epoch": 1.5522332857530874,
"grad_norm": 3.578125,
"learning_rate": 9.511750412640583e-05,
"loss": 2.4188,
"step": 67150
},
{
"epoch": 1.5533890812004092,
"grad_norm": 3.671875,
"learning_rate": 9.499614924256108e-05,
"loss": 2.3481,
"step": 67200
},
{
"epoch": 1.5545448766477308,
"grad_norm": 3.421875,
"learning_rate": 9.487480174594193e-05,
"loss": 2.3783,
"step": 67250
},
{
"epoch": 1.5557006720950526,
"grad_norm": 3.671875,
"learning_rate": 9.475346181569467e-05,
"loss": 2.4128,
"step": 67300
},
{
"epoch": 1.5568564675423744,
"grad_norm": 3.421875,
"learning_rate": 9.463212963095443e-05,
"loss": 2.3658,
"step": 67350
},
{
"epoch": 1.558012262989696,
"grad_norm": 3.484375,
"learning_rate": 9.451080537084493e-05,
"loss": 2.3462,
"step": 67400
},
{
"epoch": 1.5591680584370178,
"grad_norm": 3.625,
"learning_rate": 9.438948921447806e-05,
"loss": 2.3614,
"step": 67450
},
{
"epoch": 1.5603238538843396,
"grad_norm": 3.453125,
"learning_rate": 9.426818134095396e-05,
"loss": 2.3541,
"step": 67500
},
{
"epoch": 1.5614796493316612,
"grad_norm": 3.296875,
"learning_rate": 9.414688192936042e-05,
"loss": 2.3368,
"step": 67550
},
{
"epoch": 1.562635444778983,
"grad_norm": 3.375,
"learning_rate": 9.402559115877269e-05,
"loss": 2.3589,
"step": 67600
},
{
"epoch": 1.5637912402263048,
"grad_norm": 3.5,
"learning_rate": 9.390430920825336e-05,
"loss": 2.3951,
"step": 67650
},
{
"epoch": 1.5649470356736264,
"grad_norm": 3.375,
"learning_rate": 9.378303625685195e-05,
"loss": 2.4081,
"step": 67700
},
{
"epoch": 1.5661028311209482,
"grad_norm": 3.625,
"learning_rate": 9.366177248360471e-05,
"loss": 2.3511,
"step": 67750
},
{
"epoch": 1.56725862656827,
"grad_norm": 3.4375,
"learning_rate": 9.354051806753435e-05,
"loss": 2.4319,
"step": 67800
},
{
"epoch": 1.5684144220155916,
"grad_norm": 3.65625,
"learning_rate": 9.341927318764972e-05,
"loss": 2.3825,
"step": 67850
},
{
"epoch": 1.5695702174629134,
"grad_norm": 3.40625,
"learning_rate": 9.329803802294563e-05,
"loss": 2.3715,
"step": 67900
},
{
"epoch": 1.5707260129102352,
"grad_norm": 3.53125,
"learning_rate": 9.317681275240255e-05,
"loss": 2.4246,
"step": 67950
},
{
"epoch": 1.5718818083575568,
"grad_norm": 3.5625,
"learning_rate": 9.305559755498632e-05,
"loss": 2.3936,
"step": 68000
},
{
"epoch": 1.5730376038048786,
"grad_norm": 3.609375,
"learning_rate": 9.293439260964796e-05,
"loss": 2.3718,
"step": 68050
},
{
"epoch": 1.5741933992522004,
"grad_norm": 3.734375,
"learning_rate": 9.281319809532329e-05,
"loss": 2.3967,
"step": 68100
},
{
"epoch": 1.575349194699522,
"grad_norm": 3.390625,
"learning_rate": 9.269201419093274e-05,
"loss": 2.3803,
"step": 68150
},
{
"epoch": 1.576504990146844,
"grad_norm": 3.40625,
"learning_rate": 9.257084107538112e-05,
"loss": 2.3673,
"step": 68200
},
{
"epoch": 1.5776607855941656,
"grad_norm": 3.484375,
"learning_rate": 9.244967892755728e-05,
"loss": 2.3929,
"step": 68250
},
{
"epoch": 1.5788165810414871,
"grad_norm": 3.625,
"learning_rate": 9.232852792633383e-05,
"loss": 2.3989,
"step": 68300
},
{
"epoch": 1.5799723764888092,
"grad_norm": 3.625,
"learning_rate": 9.220738825056709e-05,
"loss": 2.3882,
"step": 68350
},
{
"epoch": 1.5811281719361308,
"grad_norm": 3.5,
"learning_rate": 9.208626007909648e-05,
"loss": 2.3859,
"step": 68400
},
{
"epoch": 1.5822839673834523,
"grad_norm": 3.59375,
"learning_rate": 9.196514359074452e-05,
"loss": 2.3475,
"step": 68450
},
{
"epoch": 1.5834397628307744,
"grad_norm": 3.53125,
"learning_rate": 9.18440389643165e-05,
"loss": 2.3829,
"step": 68500
},
{
"epoch": 1.584595558278096,
"grad_norm": 3.453125,
"learning_rate": 9.17229463786001e-05,
"loss": 2.3564,
"step": 68550
},
{
"epoch": 1.5857513537254175,
"grad_norm": 3.453125,
"learning_rate": 9.160186601236531e-05,
"loss": 2.3884,
"step": 68600
},
{
"epoch": 1.5869071491727396,
"grad_norm": 3.5625,
"learning_rate": 9.148079804436407e-05,
"loss": 2.3816,
"step": 68650
},
{
"epoch": 1.5880629446200611,
"grad_norm": 3.5,
"learning_rate": 9.135974265333002e-05,
"loss": 2.3455,
"step": 68700
},
{
"epoch": 1.5892187400673827,
"grad_norm": 3.421875,
"learning_rate": 9.123870001797819e-05,
"loss": 2.3697,
"step": 68750
},
{
"epoch": 1.5903745355147048,
"grad_norm": 3.4375,
"learning_rate": 9.111767031700485e-05,
"loss": 2.3497,
"step": 68800
},
{
"epoch": 1.5915303309620263,
"grad_norm": 3.453125,
"learning_rate": 9.099665372908709e-05,
"loss": 2.3952,
"step": 68850
},
{
"epoch": 1.592686126409348,
"grad_norm": 3.84375,
"learning_rate": 9.08756504328827e-05,
"loss": 2.3956,
"step": 68900
},
{
"epoch": 1.59384192185667,
"grad_norm": 3.390625,
"learning_rate": 9.075466060702982e-05,
"loss": 2.3925,
"step": 68950
},
{
"epoch": 1.5949977173039915,
"grad_norm": 3.671875,
"learning_rate": 9.063368443014672e-05,
"loss": 2.3874,
"step": 69000
},
{
"epoch": 1.5961535127513131,
"grad_norm": 3.53125,
"learning_rate": 9.051272208083157e-05,
"loss": 2.367,
"step": 69050
},
{
"epoch": 1.5973093081986351,
"grad_norm": 3.4375,
"learning_rate": 9.039177373766192e-05,
"loss": 2.3654,
"step": 69100
},
{
"epoch": 1.5984651036459567,
"grad_norm": 3.359375,
"learning_rate": 9.027083957919496e-05,
"loss": 2.3618,
"step": 69150
},
{
"epoch": 1.5996208990932783,
"grad_norm": 3.609375,
"learning_rate": 9.014991978396672e-05,
"loss": 2.3453,
"step": 69200
},
{
"epoch": 1.6007766945406003,
"grad_norm": 3.578125,
"learning_rate": 9.002901453049205e-05,
"loss": 2.3818,
"step": 69250
},
{
"epoch": 1.601932489987922,
"grad_norm": 3.5625,
"learning_rate": 8.990812399726435e-05,
"loss": 2.3424,
"step": 69300
},
{
"epoch": 1.6030882854352435,
"grad_norm": 3.625,
"learning_rate": 8.978724836275534e-05,
"loss": 2.3545,
"step": 69350
},
{
"epoch": 1.6042440808825655,
"grad_norm": 3.765625,
"learning_rate": 8.966638780541467e-05,
"loss": 2.4298,
"step": 69400
},
{
"epoch": 1.6053998763298871,
"grad_norm": 3.53125,
"learning_rate": 8.954554250366979e-05,
"loss": 2.3284,
"step": 69450
},
{
"epoch": 1.6065556717772087,
"grad_norm": 3.453125,
"learning_rate": 8.94247126359256e-05,
"loss": 2.419,
"step": 69500
},
{
"epoch": 1.6077114672245307,
"grad_norm": 3.625,
"learning_rate": 8.93038983805642e-05,
"loss": 2.4106,
"step": 69550
},
{
"epoch": 1.6088672626718523,
"grad_norm": 3.4375,
"learning_rate": 8.918309991594466e-05,
"loss": 2.3503,
"step": 69600
},
{
"epoch": 1.6100230581191741,
"grad_norm": 3.5,
"learning_rate": 8.906231742040274e-05,
"loss": 2.3438,
"step": 69650
},
{
"epoch": 1.611178853566496,
"grad_norm": 3.390625,
"learning_rate": 8.894155107225062e-05,
"loss": 2.3683,
"step": 69700
},
{
"epoch": 1.6123346490138175,
"grad_norm": 3.609375,
"learning_rate": 8.882080104977667e-05,
"loss": 2.4076,
"step": 69750
},
{
"epoch": 1.6134904444611393,
"grad_norm": 3.375,
"learning_rate": 8.87000675312451e-05,
"loss": 2.4051,
"step": 69800
},
{
"epoch": 1.6146462399084611,
"grad_norm": 3.421875,
"learning_rate": 8.857935069489576e-05,
"loss": 2.4103,
"step": 69850
},
{
"epoch": 1.6158020353557827,
"grad_norm": 3.609375,
"learning_rate": 8.845865071894396e-05,
"loss": 2.3655,
"step": 69900
},
{
"epoch": 1.6169578308031045,
"grad_norm": 3.5,
"learning_rate": 8.833796778157999e-05,
"loss": 2.3782,
"step": 69950
},
{
"epoch": 1.6181136262504263,
"grad_norm": 3.578125,
"learning_rate": 8.821730206096912e-05,
"loss": 2.3839,
"step": 70000
},
{
"epoch": 1.619269421697748,
"grad_norm": 3.421875,
"learning_rate": 8.80966537352511e-05,
"loss": 2.3511,
"step": 70050
},
{
"epoch": 1.6204252171450697,
"grad_norm": 3.640625,
"learning_rate": 8.797602298254004e-05,
"loss": 2.3517,
"step": 70100
},
{
"epoch": 1.6215810125923915,
"grad_norm": 3.59375,
"learning_rate": 8.78554099809241e-05,
"loss": 2.4164,
"step": 70150
},
{
"epoch": 1.622736808039713,
"grad_norm": 3.53125,
"learning_rate": 8.773481490846526e-05,
"loss": 2.308,
"step": 70200
},
{
"epoch": 1.623892603487035,
"grad_norm": 3.453125,
"learning_rate": 8.761423794319899e-05,
"loss": 2.4021,
"step": 70250
},
{
"epoch": 1.6250483989343567,
"grad_norm": 3.578125,
"learning_rate": 8.7493679263134e-05,
"loss": 2.3936,
"step": 70300
},
{
"epoch": 1.6262041943816783,
"grad_norm": 3.453125,
"learning_rate": 8.737313904625211e-05,
"loss": 2.3224,
"step": 70350
},
{
"epoch": 1.627359989829,
"grad_norm": 3.671875,
"learning_rate": 8.725261747050781e-05,
"loss": 2.3444,
"step": 70400
},
{
"epoch": 1.628515785276322,
"grad_norm": 3.4375,
"learning_rate": 8.71321147138281e-05,
"loss": 2.3564,
"step": 70450
},
{
"epoch": 1.6296715807236435,
"grad_norm": 3.546875,
"learning_rate": 8.701163095411212e-05,
"loss": 2.3245,
"step": 70500
},
{
"epoch": 1.6308273761709653,
"grad_norm": 3.46875,
"learning_rate": 8.689116636923107e-05,
"loss": 2.3569,
"step": 70550
},
{
"epoch": 1.631983171618287,
"grad_norm": 3.515625,
"learning_rate": 8.677072113702778e-05,
"loss": 2.3293,
"step": 70600
},
{
"epoch": 1.6331389670656087,
"grad_norm": 3.65625,
"learning_rate": 8.665029543531654e-05,
"loss": 2.3691,
"step": 70650
},
{
"epoch": 1.6342947625129305,
"grad_norm": 3.453125,
"learning_rate": 8.65298894418828e-05,
"loss": 2.407,
"step": 70700
},
{
"epoch": 1.6354505579602523,
"grad_norm": 3.390625,
"learning_rate": 8.64095033344829e-05,
"loss": 2.35,
"step": 70750
},
{
"epoch": 1.6366063534075739,
"grad_norm": 3.484375,
"learning_rate": 8.628913729084377e-05,
"loss": 2.42,
"step": 70800
},
{
"epoch": 1.6377621488548957,
"grad_norm": 3.546875,
"learning_rate": 8.616879148866288e-05,
"loss": 2.3618,
"step": 70850
},
{
"epoch": 1.6389179443022175,
"grad_norm": 3.453125,
"learning_rate": 8.604846610560771e-05,
"loss": 2.3378,
"step": 70900
},
{
"epoch": 1.640073739749539,
"grad_norm": 3.546875,
"learning_rate": 8.592816131931553e-05,
"loss": 2.3254,
"step": 70950
},
{
"epoch": 1.6412295351968609,
"grad_norm": 3.546875,
"learning_rate": 8.58078773073933e-05,
"loss": 2.3384,
"step": 71000
},
{
"epoch": 1.6423853306441827,
"grad_norm": 3.671875,
"learning_rate": 8.568761424741732e-05,
"loss": 2.3542,
"step": 71050
},
{
"epoch": 1.6435411260915043,
"grad_norm": 3.671875,
"learning_rate": 8.556737231693289e-05,
"loss": 2.3588,
"step": 71100
},
{
"epoch": 1.644696921538826,
"grad_norm": 3.359375,
"learning_rate": 8.544715169345418e-05,
"loss": 2.3145,
"step": 71150
},
{
"epoch": 1.6458527169861479,
"grad_norm": 3.5,
"learning_rate": 8.532695255446383e-05,
"loss": 2.359,
"step": 71200
},
{
"epoch": 1.6470085124334695,
"grad_norm": 3.46875,
"learning_rate": 8.520677507741287e-05,
"loss": 2.3523,
"step": 71250
},
{
"epoch": 1.6481643078807913,
"grad_norm": 3.5,
"learning_rate": 8.508661943972021e-05,
"loss": 2.3953,
"step": 71300
},
{
"epoch": 1.649320103328113,
"grad_norm": 3.71875,
"learning_rate": 8.496648581877265e-05,
"loss": 2.3972,
"step": 71350
},
{
"epoch": 1.6504758987754347,
"grad_norm": 3.515625,
"learning_rate": 8.484637439192441e-05,
"loss": 2.32,
"step": 71400
},
{
"epoch": 1.6516316942227565,
"grad_norm": 3.875,
"learning_rate": 8.4726285336497e-05,
"loss": 2.3099,
"step": 71450
},
{
"epoch": 1.6527874896700783,
"grad_norm": 3.359375,
"learning_rate": 8.460621882977884e-05,
"loss": 2.3937,
"step": 71500
},
{
"epoch": 1.6539432851173999,
"grad_norm": 3.359375,
"learning_rate": 8.448617504902508e-05,
"loss": 2.3531,
"step": 71550
},
{
"epoch": 1.6550990805647217,
"grad_norm": 3.609375,
"learning_rate": 8.436615417145735e-05,
"loss": 2.3576,
"step": 71600
},
{
"epoch": 1.6562548760120435,
"grad_norm": 3.515625,
"learning_rate": 8.424615637426342e-05,
"loss": 2.3524,
"step": 71650
},
{
"epoch": 1.657410671459365,
"grad_norm": 3.34375,
"learning_rate": 8.412618183459708e-05,
"loss": 2.3923,
"step": 71700
},
{
"epoch": 1.6585664669066869,
"grad_norm": 3.46875,
"learning_rate": 8.400623072957766e-05,
"loss": 2.2978,
"step": 71750
},
{
"epoch": 1.6597222623540087,
"grad_norm": 3.546875,
"learning_rate": 8.388630323628999e-05,
"loss": 2.3488,
"step": 71800
},
{
"epoch": 1.6608780578013302,
"grad_norm": 3.375,
"learning_rate": 8.376639953178398e-05,
"loss": 2.4018,
"step": 71850
},
{
"epoch": 1.662033853248652,
"grad_norm": 3.4375,
"learning_rate": 8.364651979307445e-05,
"loss": 2.3534,
"step": 71900
},
{
"epoch": 1.6631896486959739,
"grad_norm": 3.390625,
"learning_rate": 8.352666419714082e-05,
"loss": 2.3224,
"step": 71950
},
{
"epoch": 1.6643454441432954,
"grad_norm": 3.5,
"learning_rate": 8.340683292092691e-05,
"loss": 2.3638,
"step": 72000
},
{
"epoch": 1.6655012395906172,
"grad_norm": 3.578125,
"learning_rate": 8.328702614134059e-05,
"loss": 2.3693,
"step": 72050
},
{
"epoch": 1.666657035037939,
"grad_norm": 3.4375,
"learning_rate": 8.316724403525359e-05,
"loss": 2.367,
"step": 72100
},
{
"epoch": 1.6678128304852606,
"grad_norm": 3.65625,
"learning_rate": 8.304748677950122e-05,
"loss": 2.3799,
"step": 72150
},
{
"epoch": 1.6689686259325824,
"grad_norm": 3.46875,
"learning_rate": 8.292775455088206e-05,
"loss": 2.3588,
"step": 72200
},
{
"epoch": 1.6701244213799042,
"grad_norm": 3.59375,
"learning_rate": 8.28080475261578e-05,
"loss": 2.3729,
"step": 72250
},
{
"epoch": 1.6712802168272258,
"grad_norm": 3.46875,
"learning_rate": 8.268836588205286e-05,
"loss": 2.3371,
"step": 72300
},
{
"epoch": 1.6724360122745476,
"grad_norm": 3.71875,
"learning_rate": 8.256870979525427e-05,
"loss": 2.4258,
"step": 72350
},
{
"epoch": 1.6735918077218694,
"grad_norm": 3.34375,
"learning_rate": 8.244907944241129e-05,
"loss": 2.2985,
"step": 72400
},
{
"epoch": 1.674747603169191,
"grad_norm": 3.5625,
"learning_rate": 8.232947500013512e-05,
"loss": 2.3356,
"step": 72450
},
{
"epoch": 1.6759033986165128,
"grad_norm": 3.5,
"learning_rate": 8.220989664499878e-05,
"loss": 2.3574,
"step": 72500
},
{
"epoch": 1.6770591940638346,
"grad_norm": 3.609375,
"learning_rate": 8.209034455353687e-05,
"loss": 2.3434,
"step": 72550
},
{
"epoch": 1.6782149895111562,
"grad_norm": 3.4375,
"learning_rate": 8.197081890224502e-05,
"loss": 2.363,
"step": 72600
},
{
"epoch": 1.679370784958478,
"grad_norm": 3.5625,
"learning_rate": 8.185131986757995e-05,
"loss": 2.3444,
"step": 72650
},
{
"epoch": 1.6805265804057998,
"grad_norm": 3.546875,
"learning_rate": 8.173184762595904e-05,
"loss": 2.343,
"step": 72700
},
{
"epoch": 1.6816823758531214,
"grad_norm": 3.71875,
"learning_rate": 8.161240235376015e-05,
"loss": 2.3581,
"step": 72750
},
{
"epoch": 1.6828381713004432,
"grad_norm": 3.75,
"learning_rate": 8.14929842273213e-05,
"loss": 2.4132,
"step": 72800
},
{
"epoch": 1.683993966747765,
"grad_norm": 3.625,
"learning_rate": 8.137359342294047e-05,
"loss": 2.3281,
"step": 72850
},
{
"epoch": 1.6851497621950866,
"grad_norm": 3.453125,
"learning_rate": 8.125423011687524e-05,
"loss": 2.3742,
"step": 72900
},
{
"epoch": 1.6863055576424084,
"grad_norm": 3.4375,
"learning_rate": 8.113489448534263e-05,
"loss": 2.3318,
"step": 72950
},
{
"epoch": 1.6874613530897302,
"grad_norm": 3.515625,
"learning_rate": 8.101558670451883e-05,
"loss": 2.3178,
"step": 73000
},
{
"epoch": 1.6886171485370518,
"grad_norm": 3.6875,
"learning_rate": 8.089630695053886e-05,
"loss": 2.386,
"step": 73050
},
{
"epoch": 1.6897729439843736,
"grad_norm": 3.515625,
"learning_rate": 8.077705539949642e-05,
"loss": 2.3654,
"step": 73100
},
{
"epoch": 1.6909287394316954,
"grad_norm": 3.5,
"learning_rate": 8.065783222744353e-05,
"loss": 2.318,
"step": 73150
},
{
"epoch": 1.692084534879017,
"grad_norm": 3.640625,
"learning_rate": 8.053863761039032e-05,
"loss": 2.3351,
"step": 73200
},
{
"epoch": 1.6932403303263388,
"grad_norm": 3.578125,
"learning_rate": 8.041947172430478e-05,
"loss": 2.3598,
"step": 73250
},
{
"epoch": 1.6943961257736606,
"grad_norm": 3.5,
"learning_rate": 8.030033474511249e-05,
"loss": 2.3599,
"step": 73300
},
{
"epoch": 1.6955519212209822,
"grad_norm": 3.46875,
"learning_rate": 8.018122684869629e-05,
"loss": 2.337,
"step": 73350
},
{
"epoch": 1.696707716668304,
"grad_norm": 3.53125,
"learning_rate": 8.006214821089621e-05,
"loss": 2.3799,
"step": 73400
},
{
"epoch": 1.6978635121156258,
"grad_norm": 3.40625,
"learning_rate": 7.9943099007509e-05,
"loss": 2.3678,
"step": 73450
},
{
"epoch": 1.6990193075629474,
"grad_norm": 3.234375,
"learning_rate": 7.982407941428792e-05,
"loss": 2.3366,
"step": 73500
},
{
"epoch": 1.7001751030102692,
"grad_norm": 3.46875,
"learning_rate": 7.970508960694263e-05,
"loss": 2.3648,
"step": 73550
},
{
"epoch": 1.701330898457591,
"grad_norm": 3.546875,
"learning_rate": 7.958612976113869e-05,
"loss": 2.3437,
"step": 73600
},
{
"epoch": 1.7024866939049126,
"grad_norm": 3.546875,
"learning_rate": 7.94672000524975e-05,
"loss": 2.3439,
"step": 73650
},
{
"epoch": 1.7036424893522344,
"grad_norm": 3.625,
"learning_rate": 7.934830065659599e-05,
"loss": 2.3137,
"step": 73700
},
{
"epoch": 1.7047982847995562,
"grad_norm": 3.359375,
"learning_rate": 7.922943174896626e-05,
"loss": 2.3589,
"step": 73750
},
{
"epoch": 1.7059540802468778,
"grad_norm": 3.40625,
"learning_rate": 7.911059350509551e-05,
"loss": 2.3306,
"step": 73800
},
{
"epoch": 1.7071098756941996,
"grad_norm": 3.59375,
"learning_rate": 7.899178610042555e-05,
"loss": 2.3529,
"step": 73850
},
{
"epoch": 1.7082656711415214,
"grad_norm": 3.59375,
"learning_rate": 7.887300971035272e-05,
"loss": 2.3207,
"step": 73900
},
{
"epoch": 1.709421466588843,
"grad_norm": 3.75,
"learning_rate": 7.875426451022759e-05,
"loss": 2.3448,
"step": 73950
},
{
"epoch": 1.710577262036165,
"grad_norm": 3.484375,
"learning_rate": 7.863555067535466e-05,
"loss": 2.2927,
"step": 74000
},
{
"epoch": 1.7117330574834866,
"grad_norm": 3.78125,
"learning_rate": 7.851686838099217e-05,
"loss": 2.3835,
"step": 74050
},
{
"epoch": 1.7128888529308082,
"grad_norm": 3.59375,
"learning_rate": 7.839821780235168e-05,
"loss": 2.3362,
"step": 74100
},
{
"epoch": 1.7140446483781302,
"grad_norm": 3.453125,
"learning_rate": 7.827959911459801e-05,
"loss": 2.3735,
"step": 74150
},
{
"epoch": 1.7152004438254518,
"grad_norm": 3.453125,
"learning_rate": 7.816101249284897e-05,
"loss": 2.344,
"step": 74200
},
{
"epoch": 1.7163562392727734,
"grad_norm": 3.59375,
"learning_rate": 7.804245811217493e-05,
"loss": 2.3369,
"step": 74250
},
{
"epoch": 1.7175120347200954,
"grad_norm": 3.46875,
"learning_rate": 7.792393614759866e-05,
"loss": 2.322,
"step": 74300
},
{
"epoch": 1.718667830167417,
"grad_norm": 3.625,
"learning_rate": 7.780544677409509e-05,
"loss": 2.3679,
"step": 74350
},
{
"epoch": 1.7198236256147386,
"grad_norm": 3.46875,
"learning_rate": 7.768699016659108e-05,
"loss": 2.3657,
"step": 74400
},
{
"epoch": 1.7209794210620606,
"grad_norm": 3.609375,
"learning_rate": 7.756856649996506e-05,
"loss": 2.4038,
"step": 74450
},
{
"epoch": 1.7221352165093822,
"grad_norm": 3.6875,
"learning_rate": 7.74501759490469e-05,
"loss": 2.3459,
"step": 74500
},
{
"epoch": 1.7232910119567038,
"grad_norm": 3.65625,
"learning_rate": 7.733181868861748e-05,
"loss": 2.3432,
"step": 74550
},
{
"epoch": 1.7244468074040258,
"grad_norm": 3.609375,
"learning_rate": 7.721349489340858e-05,
"loss": 2.3485,
"step": 74600
},
{
"epoch": 1.7256026028513474,
"grad_norm": 3.515625,
"learning_rate": 7.709520473810262e-05,
"loss": 2.3445,
"step": 74650
},
{
"epoch": 1.726758398298669,
"grad_norm": 3.625,
"learning_rate": 7.697694839733231e-05,
"loss": 2.3523,
"step": 74700
},
{
"epoch": 1.727914193745991,
"grad_norm": 3.5625,
"learning_rate": 7.685872604568045e-05,
"loss": 2.3686,
"step": 74750
},
{
"epoch": 1.7290699891933126,
"grad_norm": 3.671875,
"learning_rate": 7.674053785767965e-05,
"loss": 2.3387,
"step": 74800
},
{
"epoch": 1.7302257846406341,
"grad_norm": 3.578125,
"learning_rate": 7.662238400781208e-05,
"loss": 2.3435,
"step": 74850
},
{
"epoch": 1.7313815800879562,
"grad_norm": 3.34375,
"learning_rate": 7.650426467050926e-05,
"loss": 2.337,
"step": 74900
},
{
"epoch": 1.7325373755352778,
"grad_norm": 3.671875,
"learning_rate": 7.63861800201517e-05,
"loss": 2.3129,
"step": 74950
},
{
"epoch": 1.7336931709825993,
"grad_norm": 3.421875,
"learning_rate": 7.626813023106874e-05,
"loss": 2.3496,
"step": 75000
},
{
"epoch": 1.7348489664299214,
"grad_norm": 3.546875,
"learning_rate": 7.615011547753827e-05,
"loss": 2.3382,
"step": 75050
},
{
"epoch": 1.736004761877243,
"grad_norm": 3.578125,
"learning_rate": 7.60321359337864e-05,
"loss": 2.3324,
"step": 75100
},
{
"epoch": 1.7371605573245645,
"grad_norm": 3.734375,
"learning_rate": 7.591419177398731e-05,
"loss": 2.3096,
"step": 75150
},
{
"epoch": 1.7383163527718866,
"grad_norm": 3.75,
"learning_rate": 7.579628317226293e-05,
"loss": 2.346,
"step": 75200
},
{
"epoch": 1.7394721482192081,
"grad_norm": 3.359375,
"learning_rate": 7.567841030268266e-05,
"loss": 2.332,
"step": 75250
},
{
"epoch": 1.7406279436665297,
"grad_norm": 3.640625,
"learning_rate": 7.556057333926318e-05,
"loss": 2.3707,
"step": 75300
},
{
"epoch": 1.7417837391138518,
"grad_norm": 3.6875,
"learning_rate": 7.544277245596818e-05,
"loss": 2.3969,
"step": 75350
},
{
"epoch": 1.7429395345611733,
"grad_norm": 3.6875,
"learning_rate": 7.532500782670803e-05,
"loss": 2.402,
"step": 75400
},
{
"epoch": 1.7440953300084951,
"grad_norm": 3.4375,
"learning_rate": 7.520727962533967e-05,
"loss": 2.3619,
"step": 75450
},
{
"epoch": 1.745251125455817,
"grad_norm": 3.265625,
"learning_rate": 7.508958802566615e-05,
"loss": 2.3499,
"step": 75500
},
{
"epoch": 1.7464069209031385,
"grad_norm": 3.75,
"learning_rate": 7.497193320143653e-05,
"loss": 2.3189,
"step": 75550
},
{
"epoch": 1.7475627163504603,
"grad_norm": 3.40625,
"learning_rate": 7.485431532634561e-05,
"loss": 2.3587,
"step": 75600
},
{
"epoch": 1.7487185117977821,
"grad_norm": 3.609375,
"learning_rate": 7.473673457403361e-05,
"loss": 2.409,
"step": 75650
},
{
"epoch": 1.7498743072451037,
"grad_norm": 3.640625,
"learning_rate": 7.461919111808595e-05,
"loss": 2.3365,
"step": 75700
},
{
"epoch": 1.7510301026924255,
"grad_norm": 3.609375,
"learning_rate": 7.450168513203301e-05,
"loss": 2.359,
"step": 75750
},
{
"epoch": 1.7521858981397473,
"grad_norm": 3.8125,
"learning_rate": 7.43842167893498e-05,
"loss": 2.3351,
"step": 75800
},
{
"epoch": 1.753341693587069,
"grad_norm": 3.453125,
"learning_rate": 7.426678626345576e-05,
"loss": 2.3143,
"step": 75850
},
{
"epoch": 1.7544974890343907,
"grad_norm": 3.421875,
"learning_rate": 7.414939372771464e-05,
"loss": 2.3293,
"step": 75900
},
{
"epoch": 1.7556532844817125,
"grad_norm": 3.5625,
"learning_rate": 7.403203935543392e-05,
"loss": 2.4164,
"step": 75950
},
{
"epoch": 1.7568090799290341,
"grad_norm": 3.421875,
"learning_rate": 7.391472331986482e-05,
"loss": 2.3686,
"step": 76000
},
{
"epoch": 1.757964875376356,
"grad_norm": 3.578125,
"learning_rate": 7.379744579420195e-05,
"loss": 2.319,
"step": 76050
},
{
"epoch": 1.7591206708236777,
"grad_norm": 3.5625,
"learning_rate": 7.368020695158312e-05,
"loss": 2.3215,
"step": 76100
},
{
"epoch": 1.7602764662709993,
"grad_norm": 3.40625,
"learning_rate": 7.356300696508896e-05,
"loss": 2.3489,
"step": 76150
},
{
"epoch": 1.7614322617183211,
"grad_norm": 3.46875,
"learning_rate": 7.344584600774276e-05,
"loss": 2.3142,
"step": 76200
},
{
"epoch": 1.762588057165643,
"grad_norm": 3.6875,
"learning_rate": 7.332872425251018e-05,
"loss": 2.3346,
"step": 76250
},
{
"epoch": 1.7637438526129645,
"grad_norm": 3.390625,
"learning_rate": 7.321164187229903e-05,
"loss": 2.3361,
"step": 76300
},
{
"epoch": 1.7648996480602863,
"grad_norm": 3.625,
"learning_rate": 7.309459903995901e-05,
"loss": 2.3299,
"step": 76350
},
{
"epoch": 1.7660554435076081,
"grad_norm": 3.375,
"learning_rate": 7.297759592828137e-05,
"loss": 2.353,
"step": 76400
},
{
"epoch": 1.7672112389549297,
"grad_norm": 3.34375,
"learning_rate": 7.286063270999879e-05,
"loss": 2.3314,
"step": 76450
},
{
"epoch": 1.7683670344022515,
"grad_norm": 3.734375,
"learning_rate": 7.274370955778498e-05,
"loss": 2.3352,
"step": 76500
},
{
"epoch": 1.7695228298495733,
"grad_norm": 3.5,
"learning_rate": 7.262682664425457e-05,
"loss": 2.4151,
"step": 76550
},
{
"epoch": 1.770678625296895,
"grad_norm": 3.65625,
"learning_rate": 7.250998414196275e-05,
"loss": 2.3475,
"step": 76600
},
{
"epoch": 1.7718344207442167,
"grad_norm": 3.6875,
"learning_rate": 7.239318222340505e-05,
"loss": 2.3381,
"step": 76650
},
{
"epoch": 1.7729902161915385,
"grad_norm": 3.359375,
"learning_rate": 7.227642106101706e-05,
"loss": 2.3555,
"step": 76700
},
{
"epoch": 1.77414601163886,
"grad_norm": 3.671875,
"learning_rate": 7.215970082717431e-05,
"loss": 2.4074,
"step": 76750
},
{
"epoch": 1.775301807086182,
"grad_norm": 3.578125,
"learning_rate": 7.20430216941918e-05,
"loss": 2.3924,
"step": 76800
},
{
"epoch": 1.7764576025335037,
"grad_norm": 3.53125,
"learning_rate": 7.192638383432389e-05,
"loss": 2.3347,
"step": 76850
},
{
"epoch": 1.7776133979808253,
"grad_norm": 3.59375,
"learning_rate": 7.180978741976397e-05,
"loss": 2.3133,
"step": 76900
},
{
"epoch": 1.778769193428147,
"grad_norm": 3.53125,
"learning_rate": 7.16932326226443e-05,
"loss": 2.3038,
"step": 76950
},
{
"epoch": 1.779924988875469,
"grad_norm": 3.5,
"learning_rate": 7.157671961503567e-05,
"loss": 2.3299,
"step": 77000
},
{
"epoch": 1.7810807843227905,
"grad_norm": 3.46875,
"learning_rate": 7.146024856894717e-05,
"loss": 2.3291,
"step": 77050
},
{
"epoch": 1.7822365797701123,
"grad_norm": 3.625,
"learning_rate": 7.134381965632598e-05,
"loss": 2.3855,
"step": 77100
},
{
"epoch": 1.783392375217434,
"grad_norm": 3.515625,
"learning_rate": 7.122743304905704e-05,
"loss": 2.3659,
"step": 77150
},
{
"epoch": 1.7845481706647557,
"grad_norm": 3.609375,
"learning_rate": 7.111108891896283e-05,
"loss": 2.3236,
"step": 77200
},
{
"epoch": 1.7857039661120775,
"grad_norm": 3.484375,
"learning_rate": 7.099478743780315e-05,
"loss": 2.3383,
"step": 77250
},
{
"epoch": 1.7868597615593993,
"grad_norm": 3.65625,
"learning_rate": 7.087852877727481e-05,
"loss": 2.3519,
"step": 77300
},
{
"epoch": 1.7880155570067209,
"grad_norm": 3.703125,
"learning_rate": 7.076231310901143e-05,
"loss": 2.3189,
"step": 77350
},
{
"epoch": 1.7891713524540427,
"grad_norm": 3.65625,
"learning_rate": 7.064614060458317e-05,
"loss": 2.3331,
"step": 77400
},
{
"epoch": 1.7903271479013645,
"grad_norm": 3.671875,
"learning_rate": 7.053001143549641e-05,
"loss": 2.3231,
"step": 77450
},
{
"epoch": 1.791482943348686,
"grad_norm": 3.703125,
"learning_rate": 7.041392577319357e-05,
"loss": 2.3418,
"step": 77500
},
{
"epoch": 1.7926387387960079,
"grad_norm": 3.484375,
"learning_rate": 7.029788378905287e-05,
"loss": 2.3424,
"step": 77550
},
{
"epoch": 1.7937945342433297,
"grad_norm": 3.46875,
"learning_rate": 7.018188565438807e-05,
"loss": 2.3442,
"step": 77600
},
{
"epoch": 1.7949503296906513,
"grad_norm": 3.3125,
"learning_rate": 7.006593154044813e-05,
"loss": 2.3057,
"step": 77650
},
{
"epoch": 1.796106125137973,
"grad_norm": 3.71875,
"learning_rate": 6.995002161841708e-05,
"loss": 2.3021,
"step": 77700
},
{
"epoch": 1.7972619205852949,
"grad_norm": 3.53125,
"learning_rate": 6.983415605941365e-05,
"loss": 2.3256,
"step": 77750
},
{
"epoch": 1.7984177160326165,
"grad_norm": 3.796875,
"learning_rate": 6.971833503449112e-05,
"loss": 2.3083,
"step": 77800
},
{
"epoch": 1.7995735114799383,
"grad_norm": 3.578125,
"learning_rate": 6.960255871463703e-05,
"loss": 2.3508,
"step": 77850
},
{
"epoch": 1.80072930692726,
"grad_norm": 3.5,
"learning_rate": 6.948682727077287e-05,
"loss": 2.2898,
"step": 77900
},
{
"epoch": 1.8018851023745817,
"grad_norm": 3.65625,
"learning_rate": 6.937114087375394e-05,
"loss": 2.3363,
"step": 77950
},
{
"epoch": 1.8030408978219035,
"grad_norm": 3.46875,
"learning_rate": 6.925549969436897e-05,
"loss": 2.2997,
"step": 78000
},
{
"epoch": 1.8041966932692253,
"grad_norm": 3.390625,
"learning_rate": 6.913990390334002e-05,
"loss": 2.333,
"step": 78050
},
{
"epoch": 1.8053524887165469,
"grad_norm": 3.546875,
"learning_rate": 6.902435367132208e-05,
"loss": 2.3,
"step": 78100
},
{
"epoch": 1.8065082841638687,
"grad_norm": 3.40625,
"learning_rate": 6.890884916890289e-05,
"loss": 2.3727,
"step": 78150
},
{
"epoch": 1.8076640796111905,
"grad_norm": 3.546875,
"learning_rate": 6.879339056660268e-05,
"loss": 2.3568,
"step": 78200
},
{
"epoch": 1.808819875058512,
"grad_norm": 3.53125,
"learning_rate": 6.867797803487395e-05,
"loss": 2.3478,
"step": 78250
},
{
"epoch": 1.8099756705058339,
"grad_norm": 3.734375,
"learning_rate": 6.856261174410113e-05,
"loss": 2.3582,
"step": 78300
},
{
"epoch": 1.8111314659531557,
"grad_norm": 3.609375,
"learning_rate": 6.844729186460044e-05,
"loss": 2.3293,
"step": 78350
},
{
"epoch": 1.8122872614004772,
"grad_norm": 3.25,
"learning_rate": 6.833201856661952e-05,
"loss": 2.3159,
"step": 78400
},
{
"epoch": 1.813443056847799,
"grad_norm": 3.5,
"learning_rate": 6.821679202033735e-05,
"loss": 2.3759,
"step": 78450
},
{
"epoch": 1.8145988522951209,
"grad_norm": 3.546875,
"learning_rate": 6.810161239586375e-05,
"loss": 2.3388,
"step": 78500
},
{
"epoch": 1.8157546477424424,
"grad_norm": 3.75,
"learning_rate": 6.798647986323938e-05,
"loss": 2.3151,
"step": 78550
},
{
"epoch": 1.8169104431897642,
"grad_norm": 3.6875,
"learning_rate": 6.78713945924353e-05,
"loss": 2.3193,
"step": 78600
},
{
"epoch": 1.818066238637086,
"grad_norm": 3.609375,
"learning_rate": 6.775635675335283e-05,
"loss": 2.3643,
"step": 78650
},
{
"epoch": 1.8192220340844076,
"grad_norm": 3.609375,
"learning_rate": 6.764136651582332e-05,
"loss": 2.3335,
"step": 78700
},
{
"epoch": 1.8203778295317294,
"grad_norm": 3.734375,
"learning_rate": 6.752642404960773e-05,
"loss": 2.3114,
"step": 78750
},
{
"epoch": 1.8215336249790512,
"grad_norm": 3.65625,
"learning_rate": 6.741152952439662e-05,
"loss": 2.3353,
"step": 78800
},
{
"epoch": 1.8226894204263728,
"grad_norm": 3.625,
"learning_rate": 6.729668310980963e-05,
"loss": 2.3007,
"step": 78850
},
{
"epoch": 1.8238452158736946,
"grad_norm": 3.65625,
"learning_rate": 6.718188497539554e-05,
"loss": 2.3404,
"step": 78900
},
{
"epoch": 1.8250010113210164,
"grad_norm": 3.59375,
"learning_rate": 6.706713529063171e-05,
"loss": 2.3629,
"step": 78950
},
{
"epoch": 1.826156806768338,
"grad_norm": 3.65625,
"learning_rate": 6.695243422492408e-05,
"loss": 2.3434,
"step": 79000
},
{
"epoch": 1.8273126022156598,
"grad_norm": 3.765625,
"learning_rate": 6.683778194760672e-05,
"loss": 2.3845,
"step": 79050
},
{
"epoch": 1.8284683976629816,
"grad_norm": 3.578125,
"learning_rate": 6.67231786279418e-05,
"loss": 2.3274,
"step": 79100
},
{
"epoch": 1.8296241931103032,
"grad_norm": 3.40625,
"learning_rate": 6.660862443511903e-05,
"loss": 2.3606,
"step": 79150
},
{
"epoch": 1.830779988557625,
"grad_norm": 3.59375,
"learning_rate": 6.649411953825572e-05,
"loss": 2.331,
"step": 79200
},
{
"epoch": 1.8319357840049468,
"grad_norm": 3.546875,
"learning_rate": 6.637966410639647e-05,
"loss": 2.2986,
"step": 79250
},
{
"epoch": 1.8330915794522684,
"grad_norm": 3.71875,
"learning_rate": 6.626525830851267e-05,
"loss": 2.313,
"step": 79300
},
{
"epoch": 1.8342473748995902,
"grad_norm": 3.546875,
"learning_rate": 6.615090231350256e-05,
"loss": 2.3333,
"step": 79350
},
{
"epoch": 1.835403170346912,
"grad_norm": 3.390625,
"learning_rate": 6.603659629019084e-05,
"loss": 2.3164,
"step": 79400
},
{
"epoch": 1.8365589657942336,
"grad_norm": 3.546875,
"learning_rate": 6.592234040732841e-05,
"loss": 2.3404,
"step": 79450
},
{
"epoch": 1.8377147612415554,
"grad_norm": 3.578125,
"learning_rate": 6.580813483359219e-05,
"loss": 2.2948,
"step": 79500
},
{
"epoch": 1.8388705566888772,
"grad_norm": 3.78125,
"learning_rate": 6.569397973758473e-05,
"loss": 2.3654,
"step": 79550
},
{
"epoch": 1.8400263521361988,
"grad_norm": 3.640625,
"learning_rate": 6.557987528783419e-05,
"loss": 2.3123,
"step": 79600
},
{
"epoch": 1.8411821475835208,
"grad_norm": 3.625,
"learning_rate": 6.546582165279384e-05,
"loss": 2.3959,
"step": 79650
},
{
"epoch": 1.8423379430308424,
"grad_norm": 3.453125,
"learning_rate": 6.535181900084206e-05,
"loss": 2.2866,
"step": 79700
},
{
"epoch": 1.843493738478164,
"grad_norm": 3.46875,
"learning_rate": 6.523786750028184e-05,
"loss": 2.2898,
"step": 79750
},
{
"epoch": 1.844649533925486,
"grad_norm": 3.671875,
"learning_rate": 6.512396731934074e-05,
"loss": 2.3159,
"step": 79800
},
{
"epoch": 1.8458053293728076,
"grad_norm": 3.625,
"learning_rate": 6.50101186261705e-05,
"loss": 2.3085,
"step": 79850
},
{
"epoch": 1.8469611248201292,
"grad_norm": 3.59375,
"learning_rate": 6.489632158884687e-05,
"loss": 2.3059,
"step": 79900
},
{
"epoch": 1.8481169202674512,
"grad_norm": 3.3125,
"learning_rate": 6.478257637536935e-05,
"loss": 2.3274,
"step": 79950
},
{
"epoch": 1.8492727157147728,
"grad_norm": 3.609375,
"learning_rate": 6.466888315366093e-05,
"loss": 2.2651,
"step": 80000
},
{
"epoch": 1.8504285111620944,
"grad_norm": 3.625,
"learning_rate": 6.455524209156783e-05,
"loss": 2.3017,
"step": 80050
},
{
"epoch": 1.8515843066094164,
"grad_norm": 3.8125,
"learning_rate": 6.444165335685927e-05,
"loss": 2.311,
"step": 80100
},
{
"epoch": 1.852740102056738,
"grad_norm": 3.25,
"learning_rate": 6.432811711722726e-05,
"loss": 2.3129,
"step": 80150
},
{
"epoch": 1.8538958975040596,
"grad_norm": 3.625,
"learning_rate": 6.421463354028626e-05,
"loss": 2.3188,
"step": 80200
},
{
"epoch": 1.8550516929513816,
"grad_norm": 3.421875,
"learning_rate": 6.410120279357297e-05,
"loss": 2.3482,
"step": 80250
},
{
"epoch": 1.8562074883987032,
"grad_norm": 3.625,
"learning_rate": 6.398782504454613e-05,
"loss": 2.3435,
"step": 80300
},
{
"epoch": 1.8573632838460248,
"grad_norm": 3.46875,
"learning_rate": 6.387450046058625e-05,
"loss": 2.3344,
"step": 80350
},
{
"epoch": 1.8585190792933468,
"grad_norm": 3.59375,
"learning_rate": 6.376122920899535e-05,
"loss": 2.2848,
"step": 80400
},
{
"epoch": 1.8596748747406684,
"grad_norm": 3.640625,
"learning_rate": 6.364801145699666e-05,
"loss": 2.33,
"step": 80450
},
{
"epoch": 1.86083067018799,
"grad_norm": 3.78125,
"learning_rate": 6.35348473717345e-05,
"loss": 2.3495,
"step": 80500
},
{
"epoch": 1.861986465635312,
"grad_norm": 3.609375,
"learning_rate": 6.342173712027392e-05,
"loss": 2.3133,
"step": 80550
},
{
"epoch": 1.8631422610826336,
"grad_norm": 3.359375,
"learning_rate": 6.330868086960048e-05,
"loss": 2.3087,
"step": 80600
},
{
"epoch": 1.8642980565299552,
"grad_norm": 3.671875,
"learning_rate": 6.31956787866201e-05,
"loss": 2.3214,
"step": 80650
},
{
"epoch": 1.8654538519772772,
"grad_norm": 3.625,
"learning_rate": 6.30827310381586e-05,
"loss": 2.3794,
"step": 80700
},
{
"epoch": 1.8666096474245988,
"grad_norm": 3.546875,
"learning_rate": 6.296983779096175e-05,
"loss": 2.3508,
"step": 80750
},
{
"epoch": 1.8677654428719204,
"grad_norm": 3.359375,
"learning_rate": 6.28569992116947e-05,
"loss": 2.3406,
"step": 80800
},
{
"epoch": 1.8689212383192424,
"grad_norm": 3.640625,
"learning_rate": 6.274421546694199e-05,
"loss": 2.3368,
"step": 80850
},
{
"epoch": 1.870077033766564,
"grad_norm": 3.5,
"learning_rate": 6.263148672320714e-05,
"loss": 2.2902,
"step": 80900
},
{
"epoch": 1.8712328292138856,
"grad_norm": 3.53125,
"learning_rate": 6.251881314691257e-05,
"loss": 2.2905,
"step": 80950
},
{
"epoch": 1.8723886246612076,
"grad_norm": 3.75,
"learning_rate": 6.240619490439918e-05,
"loss": 2.3571,
"step": 81000
},
{
"epoch": 1.8735444201085292,
"grad_norm": 3.5625,
"learning_rate": 6.229363216192618e-05,
"loss": 2.334,
"step": 81050
},
{
"epoch": 1.874700215555851,
"grad_norm": 3.515625,
"learning_rate": 6.218112508567087e-05,
"loss": 2.3428,
"step": 81100
},
{
"epoch": 1.8758560110031728,
"grad_norm": 3.640625,
"learning_rate": 6.206867384172839e-05,
"loss": 2.3274,
"step": 81150
},
{
"epoch": 1.8770118064504944,
"grad_norm": 3.71875,
"learning_rate": 6.19562785961114e-05,
"loss": 2.2869,
"step": 81200
},
{
"epoch": 1.8781676018978162,
"grad_norm": 3.546875,
"learning_rate": 6.184393951474992e-05,
"loss": 2.3361,
"step": 81250
},
{
"epoch": 1.879323397345138,
"grad_norm": 3.546875,
"learning_rate": 6.173165676349103e-05,
"loss": 2.3494,
"step": 81300
},
{
"epoch": 1.8804791927924596,
"grad_norm": 3.78125,
"learning_rate": 6.161943050809871e-05,
"loss": 2.3464,
"step": 81350
},
{
"epoch": 1.8816349882397814,
"grad_norm": 3.484375,
"learning_rate": 6.150726091425345e-05,
"loss": 2.3376,
"step": 81400
},
{
"epoch": 1.8827907836871032,
"grad_norm": 3.65625,
"learning_rate": 6.139514814755219e-05,
"loss": 2.3667,
"step": 81450
},
{
"epoch": 1.8839465791344248,
"grad_norm": 3.6875,
"learning_rate": 6.128309237350787e-05,
"loss": 2.3352,
"step": 81500
},
{
"epoch": 1.8851023745817466,
"grad_norm": 3.640625,
"learning_rate": 6.117109375754935e-05,
"loss": 2.3088,
"step": 81550
},
{
"epoch": 1.8862581700290684,
"grad_norm": 3.59375,
"learning_rate": 6.105915246502109e-05,
"loss": 2.34,
"step": 81600
},
{
"epoch": 1.88741396547639,
"grad_norm": 3.78125,
"learning_rate": 6.0947268661182966e-05,
"loss": 2.3509,
"step": 81650
},
{
"epoch": 1.8885697609237118,
"grad_norm": 3.484375,
"learning_rate": 6.083544251120993e-05,
"loss": 2.3051,
"step": 81700
},
{
"epoch": 1.8897255563710336,
"grad_norm": 3.484375,
"learning_rate": 6.072367418019181e-05,
"loss": 2.3669,
"step": 81750
},
{
"epoch": 1.8908813518183551,
"grad_norm": 3.640625,
"learning_rate": 6.061196383313316e-05,
"loss": 2.3132,
"step": 81800
},
{
"epoch": 1.892037147265677,
"grad_norm": 3.65625,
"learning_rate": 6.0500311634952835e-05,
"loss": 2.3286,
"step": 81850
},
{
"epoch": 1.8931929427129988,
"grad_norm": 3.515625,
"learning_rate": 6.038871775048392e-05,
"loss": 2.3482,
"step": 81900
},
{
"epoch": 1.8943487381603203,
"grad_norm": 3.578125,
"learning_rate": 6.0277182344473314e-05,
"loss": 2.3202,
"step": 81950
},
{
"epoch": 1.8955045336076422,
"grad_norm": 3.609375,
"learning_rate": 6.016570558158168e-05,
"loss": 2.2831,
"step": 82000
},
{
"epoch": 1.896660329054964,
"grad_norm": 3.515625,
"learning_rate": 6.005428762638308e-05,
"loss": 2.3176,
"step": 82050
},
{
"epoch": 1.8978161245022855,
"grad_norm": 3.5625,
"learning_rate": 5.9942928643364724e-05,
"loss": 2.2974,
"step": 82100
},
{
"epoch": 1.8989719199496073,
"grad_norm": 3.390625,
"learning_rate": 5.983162879692683e-05,
"loss": 2.2814,
"step": 82150
},
{
"epoch": 1.9001277153969292,
"grad_norm": 3.46875,
"learning_rate": 5.97203882513822e-05,
"loss": 2.2715,
"step": 82200
},
{
"epoch": 1.9012835108442507,
"grad_norm": 3.484375,
"learning_rate": 5.9609207170956197e-05,
"loss": 2.2981,
"step": 82250
},
{
"epoch": 1.9024393062915725,
"grad_norm": 3.84375,
"learning_rate": 5.949808571978634e-05,
"loss": 2.3206,
"step": 82300
},
{
"epoch": 1.9035951017388943,
"grad_norm": 3.671875,
"learning_rate": 5.938702406192216e-05,
"loss": 2.3207,
"step": 82350
},
{
"epoch": 1.904750897186216,
"grad_norm": 3.53125,
"learning_rate": 5.9276022361324855e-05,
"loss": 2.3097,
"step": 82400
},
{
"epoch": 1.9059066926335377,
"grad_norm": 3.71875,
"learning_rate": 5.9165080781867164e-05,
"loss": 2.3367,
"step": 82450
},
{
"epoch": 1.9070624880808595,
"grad_norm": 3.640625,
"learning_rate": 5.905419948733302e-05,
"loss": 2.2123,
"step": 82500
},
{
"epoch": 1.9082182835281811,
"grad_norm": 3.171875,
"learning_rate": 5.894337864141739e-05,
"loss": 2.336,
"step": 82550
},
{
"epoch": 1.909374078975503,
"grad_norm": 3.59375,
"learning_rate": 5.8832618407725956e-05,
"loss": 2.2765,
"step": 82600
},
{
"epoch": 1.9105298744228247,
"grad_norm": 3.328125,
"learning_rate": 5.8721918949775035e-05,
"loss": 2.2918,
"step": 82650
},
{
"epoch": 1.9116856698701463,
"grad_norm": 3.59375,
"learning_rate": 5.861128043099107e-05,
"loss": 2.3286,
"step": 82700
},
{
"epoch": 1.9128414653174681,
"grad_norm": 3.578125,
"learning_rate": 5.850070301471062e-05,
"loss": 2.3351,
"step": 82750
},
{
"epoch": 1.91399726076479,
"grad_norm": 3.421875,
"learning_rate": 5.839018686418005e-05,
"loss": 2.3013,
"step": 82800
},
{
"epoch": 1.9151530562121115,
"grad_norm": 3.828125,
"learning_rate": 5.827973214255524e-05,
"loss": 2.2878,
"step": 82850
},
{
"epoch": 1.9163088516594333,
"grad_norm": 3.484375,
"learning_rate": 5.816933901290136e-05,
"loss": 2.3412,
"step": 82900
},
{
"epoch": 1.9174646471067551,
"grad_norm": 3.546875,
"learning_rate": 5.805900763819275e-05,
"loss": 2.3268,
"step": 82950
},
{
"epoch": 1.9186204425540767,
"grad_norm": 3.609375,
"learning_rate": 5.794873818131248e-05,
"loss": 2.3166,
"step": 83000
},
{
"epoch": 1.9197762380013985,
"grad_norm": 3.390625,
"learning_rate": 5.7838530805052206e-05,
"loss": 2.3676,
"step": 83050
},
{
"epoch": 1.9209320334487203,
"grad_norm": 3.40625,
"learning_rate": 5.772838567211205e-05,
"loss": 2.3025,
"step": 83100
},
{
"epoch": 1.922087828896042,
"grad_norm": 3.625,
"learning_rate": 5.76183029451001e-05,
"loss": 2.2938,
"step": 83150
},
{
"epoch": 1.9232436243433637,
"grad_norm": 3.640625,
"learning_rate": 5.750828278653245e-05,
"loss": 2.2764,
"step": 83200
},
{
"epoch": 1.9243994197906855,
"grad_norm": 3.6875,
"learning_rate": 5.7398325358832725e-05,
"loss": 2.3326,
"step": 83250
},
{
"epoch": 1.925555215238007,
"grad_norm": 3.6875,
"learning_rate": 5.728843082433193e-05,
"loss": 2.3352,
"step": 83300
},
{
"epoch": 1.926711010685329,
"grad_norm": 3.53125,
"learning_rate": 5.717859934526832e-05,
"loss": 2.2833,
"step": 83350
},
{
"epoch": 1.9278668061326507,
"grad_norm": 3.5,
"learning_rate": 5.706883108378696e-05,
"loss": 2.347,
"step": 83400
},
{
"epoch": 1.9290226015799723,
"grad_norm": 3.40625,
"learning_rate": 5.695912620193964e-05,
"loss": 2.3827,
"step": 83450
},
{
"epoch": 1.930178397027294,
"grad_norm": 3.40625,
"learning_rate": 5.68494848616846e-05,
"loss": 2.2967,
"step": 83500
},
{
"epoch": 1.931334192474616,
"grad_norm": 3.734375,
"learning_rate": 5.673990722488619e-05,
"loss": 2.3418,
"step": 83550
},
{
"epoch": 1.9324899879219375,
"grad_norm": 3.640625,
"learning_rate": 5.6630393453314835e-05,
"loss": 2.2952,
"step": 83600
},
{
"epoch": 1.9336457833692593,
"grad_norm": 3.765625,
"learning_rate": 5.6520943708646534e-05,
"loss": 2.3201,
"step": 83650
},
{
"epoch": 1.934801578816581,
"grad_norm": 3.5,
"learning_rate": 5.6411558152462894e-05,
"loss": 2.32,
"step": 83700
},
{
"epoch": 1.9359573742639027,
"grad_norm": 3.734375,
"learning_rate": 5.630223694625068e-05,
"loss": 2.3526,
"step": 83750
},
{
"epoch": 1.9371131697112245,
"grad_norm": 3.734375,
"learning_rate": 5.619298025140165e-05,
"loss": 2.3447,
"step": 83800
},
{
"epoch": 1.9382689651585463,
"grad_norm": 3.484375,
"learning_rate": 5.608378822921241e-05,
"loss": 2.3232,
"step": 83850
},
{
"epoch": 1.9394247606058679,
"grad_norm": 3.28125,
"learning_rate": 5.597466104088395e-05,
"loss": 2.3302,
"step": 83900
},
{
"epoch": 1.9405805560531897,
"grad_norm": 3.546875,
"learning_rate": 5.5865598847521716e-05,
"loss": 2.2756,
"step": 83950
},
{
"epoch": 1.9417363515005115,
"grad_norm": 3.375,
"learning_rate": 5.5756601810135064e-05,
"loss": 2.3229,
"step": 84000
},
{
"epoch": 1.942892146947833,
"grad_norm": 3.578125,
"learning_rate": 5.5647670089637184e-05,
"loss": 2.3346,
"step": 84050
},
{
"epoch": 1.9440479423951549,
"grad_norm": 3.640625,
"learning_rate": 5.553880384684493e-05,
"loss": 2.3146,
"step": 84100
},
{
"epoch": 1.9452037378424767,
"grad_norm": 3.5,
"learning_rate": 5.5430003242478344e-05,
"loss": 2.293,
"step": 84150
},
{
"epoch": 1.9463595332897983,
"grad_norm": 3.46875,
"learning_rate": 5.5321268437160725e-05,
"loss": 2.329,
"step": 84200
},
{
"epoch": 1.94751532873712,
"grad_norm": 3.53125,
"learning_rate": 5.5212599591418066e-05,
"loss": 2.306,
"step": 84250
},
{
"epoch": 1.9486711241844419,
"grad_norm": 3.5,
"learning_rate": 5.5103996865679106e-05,
"loss": 2.3212,
"step": 84300
},
{
"epoch": 1.9498269196317635,
"grad_norm": 3.515625,
"learning_rate": 5.499546042027498e-05,
"loss": 2.2866,
"step": 84350
},
{
"epoch": 1.9509827150790853,
"grad_norm": 3.4375,
"learning_rate": 5.488699041543888e-05,
"loss": 2.3109,
"step": 84400
},
{
"epoch": 1.952138510526407,
"grad_norm": 3.40625,
"learning_rate": 5.477858701130593e-05,
"loss": 2.2872,
"step": 84450
},
{
"epoch": 1.9532943059737287,
"grad_norm": 3.5625,
"learning_rate": 5.4670250367913023e-05,
"loss": 2.329,
"step": 84500
},
{
"epoch": 1.9544501014210505,
"grad_norm": 3.671875,
"learning_rate": 5.4561980645198354e-05,
"loss": 2.3018,
"step": 84550
},
{
"epoch": 1.9556058968683723,
"grad_norm": 3.46875,
"learning_rate": 5.445377800300147e-05,
"loss": 2.373,
"step": 84600
},
{
"epoch": 1.9567616923156939,
"grad_norm": 3.53125,
"learning_rate": 5.4345642601062764e-05,
"loss": 2.2973,
"step": 84650
},
{
"epoch": 1.9579174877630157,
"grad_norm": 3.46875,
"learning_rate": 5.42375745990234e-05,
"loss": 2.2149,
"step": 84700
},
{
"epoch": 1.9590732832103375,
"grad_norm": 3.546875,
"learning_rate": 5.41295741564251e-05,
"loss": 2.285,
"step": 84750
},
{
"epoch": 1.960229078657659,
"grad_norm": 3.40625,
"learning_rate": 5.402164143270973e-05,
"loss": 2.3215,
"step": 84800
},
{
"epoch": 1.9613848741049809,
"grad_norm": 3.453125,
"learning_rate": 5.391377658721933e-05,
"loss": 2.3408,
"step": 84850
},
{
"epoch": 1.9625406695523027,
"grad_norm": 3.65625,
"learning_rate": 5.380597977919557e-05,
"loss": 2.3186,
"step": 84900
},
{
"epoch": 1.9636964649996242,
"grad_norm": 3.4375,
"learning_rate": 5.3698251167779825e-05,
"loss": 2.3511,
"step": 84950
},
{
"epoch": 1.964852260446946,
"grad_norm": 3.46875,
"learning_rate": 5.359059091201268e-05,
"loss": 2.2835,
"step": 85000
},
{
"epoch": 1.9660080558942679,
"grad_norm": 3.578125,
"learning_rate": 5.348299917083381e-05,
"loss": 2.2941,
"step": 85050
},
{
"epoch": 1.9671638513415894,
"grad_norm": 3.71875,
"learning_rate": 5.337547610308182e-05,
"loss": 2.3058,
"step": 85100
},
{
"epoch": 1.9683196467889112,
"grad_norm": 3.453125,
"learning_rate": 5.3268021867493946e-05,
"loss": 2.352,
"step": 85150
},
{
"epoch": 1.969475442236233,
"grad_norm": 3.46875,
"learning_rate": 5.316063662270564e-05,
"loss": 2.3484,
"step": 85200
},
{
"epoch": 1.9706312376835546,
"grad_norm": 3.46875,
"learning_rate": 5.305332052725072e-05,
"loss": 2.3278,
"step": 85250
},
{
"epoch": 1.9717870331308764,
"grad_norm": 3.390625,
"learning_rate": 5.2946073739560706e-05,
"loss": 2.3079,
"step": 85300
},
{
"epoch": 1.9729428285781982,
"grad_norm": 3.53125,
"learning_rate": 5.2838896417965e-05,
"loss": 2.3203,
"step": 85350
},
{
"epoch": 1.9740986240255198,
"grad_norm": 3.515625,
"learning_rate": 5.273178872069031e-05,
"loss": 2.3308,
"step": 85400
},
{
"epoch": 1.9752544194728419,
"grad_norm": 3.53125,
"learning_rate": 5.2624750805860554e-05,
"loss": 2.2916,
"step": 85450
},
{
"epoch": 1.9764102149201634,
"grad_norm": 3.640625,
"learning_rate": 5.2517782831496775e-05,
"loss": 2.2903,
"step": 85500
},
{
"epoch": 1.977566010367485,
"grad_norm": 3.75,
"learning_rate": 5.241088495551657e-05,
"loss": 2.327,
"step": 85550
},
{
"epoch": 1.978721805814807,
"grad_norm": 3.640625,
"learning_rate": 5.230405733573422e-05,
"loss": 2.2966,
"step": 85600
},
{
"epoch": 1.9798776012621286,
"grad_norm": 3.6875,
"learning_rate": 5.219730012986018e-05,
"loss": 2.3182,
"step": 85650
},
{
"epoch": 1.9810333967094502,
"grad_norm": 3.640625,
"learning_rate": 5.209061349550095e-05,
"loss": 2.2843,
"step": 85700
},
{
"epoch": 1.9821891921567723,
"grad_norm": 3.65625,
"learning_rate": 5.198399759015893e-05,
"loss": 2.3555,
"step": 85750
},
{
"epoch": 1.9833449876040938,
"grad_norm": 3.53125,
"learning_rate": 5.1877452571231974e-05,
"loss": 2.3035,
"step": 85800
},
{
"epoch": 1.9845007830514154,
"grad_norm": 3.578125,
"learning_rate": 5.177097859601346e-05,
"loss": 2.3221,
"step": 85850
},
{
"epoch": 1.9856565784987374,
"grad_norm": 3.53125,
"learning_rate": 5.166457582169174e-05,
"loss": 2.3303,
"step": 85900
},
{
"epoch": 1.986812373946059,
"grad_norm": 3.625,
"learning_rate": 5.155824440535e-05,
"loss": 2.333,
"step": 85950
},
{
"epoch": 1.9879681693933806,
"grad_norm": 3.578125,
"learning_rate": 5.145198450396635e-05,
"loss": 2.3202,
"step": 86000
},
{
"epoch": 1.9891239648407026,
"grad_norm": 3.59375,
"learning_rate": 5.134579627441305e-05,
"loss": 2.3209,
"step": 86050
},
{
"epoch": 1.9902797602880242,
"grad_norm": 3.3125,
"learning_rate": 5.1239679873456634e-05,
"loss": 2.282,
"step": 86100
},
{
"epoch": 1.9914355557353458,
"grad_norm": 3.625,
"learning_rate": 5.113363545775767e-05,
"loss": 2.2858,
"step": 86150
},
{
"epoch": 1.9925913511826678,
"grad_norm": 3.625,
"learning_rate": 5.102766318387031e-05,
"loss": 2.3453,
"step": 86200
},
{
"epoch": 1.9937471466299894,
"grad_norm": 3.6875,
"learning_rate": 5.092176320824237e-05,
"loss": 2.3067,
"step": 86250
},
{
"epoch": 1.994902942077311,
"grad_norm": 3.640625,
"learning_rate": 5.081593568721477e-05,
"loss": 2.2703,
"step": 86300
},
{
"epoch": 1.996058737524633,
"grad_norm": 3.34375,
"learning_rate": 5.071018077702161e-05,
"loss": 2.3038,
"step": 86350
},
{
"epoch": 1.9972145329719546,
"grad_norm": 3.65625,
"learning_rate": 5.0604498633789686e-05,
"loss": 2.3159,
"step": 86400
},
{
"epoch": 1.9983703284192762,
"grad_norm": 3.421875,
"learning_rate": 5.049888941353837e-05,
"loss": 2.3605,
"step": 86450
},
{
"epoch": 1.9995261238665982,
"grad_norm": 3.53125,
"learning_rate": 5.039335327217951e-05,
"loss": 2.3139,
"step": 86500
}
],
"logging_steps": 50,
"max_steps": 129780,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.1231618419465667e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}