Gyaanchand-3B / trainer_state.json
Nikolester0's picture
Uploading fine-tuned Gyaanchand-3B model from Drive
62a69a6 verified
Raw
History Blame Contribute Delete
188 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.6521547003784895,
"eval_steps": 500,
"global_step": 10750,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0015370131991008472,
"grad_norm": 0.4083865284919739,
"learning_rate": 1.8e-05,
"loss": 1.7754,
"step": 10
},
{
"epoch": 0.0030740263982016944,
"grad_norm": 0.521465003490448,
"learning_rate": 3.8e-05,
"loss": 1.7381,
"step": 20
},
{
"epoch": 0.0046110395973025416,
"grad_norm": 0.8306611180305481,
"learning_rate": 5.8e-05,
"loss": 1.7298,
"step": 30
},
{
"epoch": 0.006148052796403389,
"grad_norm": 0.7454861402511597,
"learning_rate": 7.800000000000001e-05,
"loss": 1.2989,
"step": 40
},
{
"epoch": 0.007685065995504237,
"grad_norm": 2.564823627471924,
"learning_rate": 9.8e-05,
"loss": 1.2244,
"step": 50
},
{
"epoch": 0.009222079194605083,
"grad_norm": 0.2985040843486786,
"learning_rate": 0.000118,
"loss": 1.4184,
"step": 60
},
{
"epoch": 0.01075909239370593,
"grad_norm": 0.43939757347106934,
"learning_rate": 0.000138,
"loss": 1.3903,
"step": 70
},
{
"epoch": 0.012296105592806778,
"grad_norm": 0.4580193758010864,
"learning_rate": 0.00015800000000000002,
"loss": 1.316,
"step": 80
},
{
"epoch": 0.013833118791907625,
"grad_norm": 0.46390774846076965,
"learning_rate": 0.00017800000000000002,
"loss": 1.1456,
"step": 90
},
{
"epoch": 0.015370131991008474,
"grad_norm": 0.5774224996566772,
"learning_rate": 0.00019800000000000002,
"loss": 1.1181,
"step": 100
},
{
"epoch": 0.01690714519010932,
"grad_norm": 0.3326522409915924,
"learning_rate": 0.00019999737201033876,
"loss": 1.5177,
"step": 110
},
{
"epoch": 0.018444158389210166,
"grad_norm": 0.35869792103767395,
"learning_rate": 0.0001999882877789667,
"loss": 1.4151,
"step": 120
},
{
"epoch": 0.019981171588311013,
"grad_norm": 0.40248990058898926,
"learning_rate": 0.0001999727154508895,
"loss": 1.2115,
"step": 130
},
{
"epoch": 0.02151818478741186,
"grad_norm": 0.5765237808227539,
"learning_rate": 0.00019995065603657316,
"loss": 1.129,
"step": 140
},
{
"epoch": 0.023055197986512708,
"grad_norm": 0.6551015377044678,
"learning_rate": 0.00019992211096742148,
"loss": 1.0113,
"step": 150
},
{
"epoch": 0.024592211185613555,
"grad_norm": 0.2560618817806244,
"learning_rate": 0.00019988708209568308,
"loss": 1.4025,
"step": 160
},
{
"epoch": 0.026129224384714402,
"grad_norm": 0.33515942096710205,
"learning_rate": 0.00019984557169433126,
"loss": 1.3322,
"step": 170
},
{
"epoch": 0.02766623758381525,
"grad_norm": 0.3427557051181793,
"learning_rate": 0.00019979758245691648,
"loss": 1.2615,
"step": 180
},
{
"epoch": 0.029203250782916097,
"grad_norm": 0.430173397064209,
"learning_rate": 0.00019974311749739157,
"loss": 1.0835,
"step": 190
},
{
"epoch": 0.030740263982016947,
"grad_norm": 0.5968395471572876,
"learning_rate": 0.00019968218034990972,
"loss": 1.072,
"step": 200
},
{
"epoch": 0.032277277181117794,
"grad_norm": 0.3198504149913788,
"learning_rate": 0.00019961477496859508,
"loss": 1.434,
"step": 210
},
{
"epoch": 0.03381429038021864,
"grad_norm": 0.2965967655181885,
"learning_rate": 0.00019954090572728625,
"loss": 1.246,
"step": 220
},
{
"epoch": 0.03535130357931949,
"grad_norm": 0.48814648389816284,
"learning_rate": 0.00019946057741925249,
"loss": 1.1561,
"step": 230
},
{
"epoch": 0.03688831677842033,
"grad_norm": 0.39564383029937744,
"learning_rate": 0.0001993737952568825,
"loss": 1.0359,
"step": 240
},
{
"epoch": 0.03842532997752118,
"grad_norm": 0.7815542221069336,
"learning_rate": 0.00019928056487134653,
"loss": 1.0161,
"step": 250
},
{
"epoch": 0.03996234317662203,
"grad_norm": 0.27169305086135864,
"learning_rate": 0.00019918089231223064,
"loss": 1.5133,
"step": 260
},
{
"epoch": 0.04149935637572288,
"grad_norm": 0.2866431176662445,
"learning_rate": 0.00019907478404714436,
"loss": 1.3815,
"step": 270
},
{
"epoch": 0.04303636957482372,
"grad_norm": 0.2940818667411804,
"learning_rate": 0.00019896224696130097,
"loss": 1.2333,
"step": 280
},
{
"epoch": 0.04457338277392457,
"grad_norm": 0.4689043462276459,
"learning_rate": 0.00019884328835707068,
"loss": 1.1281,
"step": 290
},
{
"epoch": 0.046110395973025416,
"grad_norm": 0.5727034211158752,
"learning_rate": 0.00019871791595350684,
"loss": 1.0234,
"step": 300
},
{
"epoch": 0.047647409172126266,
"grad_norm": 0.26414376497268677,
"learning_rate": 0.00019858613788584507,
"loss": 1.3496,
"step": 310
},
{
"epoch": 0.04918442237122711,
"grad_norm": 0.3377169668674469,
"learning_rate": 0.00019844796270497531,
"loss": 1.1944,
"step": 320
},
{
"epoch": 0.05072143557032796,
"grad_norm": 0.3532752990722656,
"learning_rate": 0.00019830339937688707,
"loss": 1.1993,
"step": 330
},
{
"epoch": 0.052258448769428804,
"grad_norm": 0.4452616572380066,
"learning_rate": 0.00019815245728208752,
"loss": 1.1752,
"step": 340
},
{
"epoch": 0.053795461968529655,
"grad_norm": 0.5334734916687012,
"learning_rate": 0.00019799514621499288,
"loss": 1.0094,
"step": 350
},
{
"epoch": 0.0553324751676305,
"grad_norm": 0.26407647132873535,
"learning_rate": 0.00019783147638329288,
"loss": 1.3612,
"step": 360
},
{
"epoch": 0.05686948836673135,
"grad_norm": 0.2937082350254059,
"learning_rate": 0.0001976614584072883,
"loss": 1.2174,
"step": 370
},
{
"epoch": 0.05840650156583219,
"grad_norm": 0.3781548738479614,
"learning_rate": 0.00019748510331920205,
"loss": 1.1142,
"step": 380
},
{
"epoch": 0.059943514764933044,
"grad_norm": 0.38902518153190613,
"learning_rate": 0.00019730242256246297,
"loss": 0.9849,
"step": 390
},
{
"epoch": 0.061480527964033894,
"grad_norm": 0.5630488991737366,
"learning_rate": 0.00019711342799096361,
"loss": 0.9313,
"step": 400
},
{
"epoch": 0.06301754116313474,
"grad_norm": 0.27971673011779785,
"learning_rate": 0.0001969181318682909,
"loss": 1.3923,
"step": 410
},
{
"epoch": 0.06455455436223559,
"grad_norm": 0.33238485455513,
"learning_rate": 0.00019671654686693034,
"loss": 1.1636,
"step": 420
},
{
"epoch": 0.06609156756133644,
"grad_norm": 0.42027440667152405,
"learning_rate": 0.0001965086860674438,
"loss": 1.1959,
"step": 430
},
{
"epoch": 0.06762858076043728,
"grad_norm": 0.4032113552093506,
"learning_rate": 0.00019629456295762067,
"loss": 1.01,
"step": 440
},
{
"epoch": 0.06916559395953813,
"grad_norm": 0.5734425783157349,
"learning_rate": 0.0001960741914316027,
"loss": 0.9169,
"step": 450
},
{
"epoch": 0.07070260715863898,
"grad_norm": 0.3068268597126007,
"learning_rate": 0.00019584758578898245,
"loss": 1.3052,
"step": 460
},
{
"epoch": 0.07223962035773983,
"grad_norm": 0.3328419029712677,
"learning_rate": 0.00019561476073387527,
"loss": 1.2887,
"step": 470
},
{
"epoch": 0.07377663355684067,
"grad_norm": 0.35405659675598145,
"learning_rate": 0.00019537573137396534,
"loss": 1.1414,
"step": 480
},
{
"epoch": 0.07531364675594152,
"grad_norm": 0.488251656293869,
"learning_rate": 0.00019513051321952532,
"loss": 1.0648,
"step": 490
},
{
"epoch": 0.07685065995504237,
"grad_norm": 0.6530411839485168,
"learning_rate": 0.00019487912218240982,
"loss": 0.9655,
"step": 500
},
{
"epoch": 0.07838767315414322,
"grad_norm": 0.2529815435409546,
"learning_rate": 0.00019462157457502297,
"loss": 1.4195,
"step": 510
},
{
"epoch": 0.07992468635324405,
"grad_norm": 0.3448476493358612,
"learning_rate": 0.00019435788710925997,
"loss": 1.2444,
"step": 520
},
{
"epoch": 0.0814616995523449,
"grad_norm": 0.3466683328151703,
"learning_rate": 0.00019408807689542257,
"loss": 1.1288,
"step": 530
},
{
"epoch": 0.08299871275144576,
"grad_norm": 0.3787826597690582,
"learning_rate": 0.00019381216144110885,
"loss": 1.0535,
"step": 540
},
{
"epoch": 0.0845357259505466,
"grad_norm": 0.576309859752655,
"learning_rate": 0.00019353015865007734,
"loss": 1.0602,
"step": 550
},
{
"epoch": 0.08607273914964744,
"grad_norm": 0.25945737957954407,
"learning_rate": 0.0001932420868210849,
"loss": 1.3483,
"step": 560
},
{
"epoch": 0.0876097523487483,
"grad_norm": 0.3733840584754944,
"learning_rate": 0.0001929479646466998,
"loss": 1.2047,
"step": 570
},
{
"epoch": 0.08914676554784914,
"grad_norm": 0.4557248651981354,
"learning_rate": 0.00019264781121208842,
"loss": 1.1616,
"step": 580
},
{
"epoch": 0.09068377874695,
"grad_norm": 0.43104612827301025,
"learning_rate": 0.00019234164599377692,
"loss": 1.1464,
"step": 590
},
{
"epoch": 0.09222079194605083,
"grad_norm": 1.878812551498413,
"learning_rate": 0.00019202948885838766,
"loss": 0.8946,
"step": 600
},
{
"epoch": 0.09375780514515168,
"grad_norm": 0.2881523668766022,
"learning_rate": 0.00019171136006134972,
"loss": 1.3945,
"step": 610
},
{
"epoch": 0.09529481834425253,
"grad_norm": 0.2849006950855255,
"learning_rate": 0.00019138728024558493,
"loss": 1.3161,
"step": 620
},
{
"epoch": 0.09683183154335338,
"grad_norm": 0.39194759726524353,
"learning_rate": 0.00019105727044016807,
"loss": 1.0863,
"step": 630
},
{
"epoch": 0.09836884474245422,
"grad_norm": 0.4738089144229889,
"learning_rate": 0.00019072135205896253,
"loss": 1.0859,
"step": 640
},
{
"epoch": 0.09990585794155507,
"grad_norm": 0.5343679785728455,
"learning_rate": 0.00019037954689923062,
"loss": 0.9623,
"step": 650
},
{
"epoch": 0.10144287114065592,
"grad_norm": 0.2589882016181946,
"learning_rate": 0.00019003187714021938,
"loss": 1.4186,
"step": 660
},
{
"epoch": 0.10297988433975677,
"grad_norm": 0.33215847611427307,
"learning_rate": 0.00018967836534172116,
"loss": 1.1417,
"step": 670
},
{
"epoch": 0.10451689753885761,
"grad_norm": 0.3690186142921448,
"learning_rate": 0.00018931903444261008,
"loss": 1.1129,
"step": 680
},
{
"epoch": 0.10605391073795846,
"grad_norm": 0.4274938106536865,
"learning_rate": 0.00018895390775935312,
"loss": 1.0314,
"step": 690
},
{
"epoch": 0.10759092393705931,
"grad_norm": 0.5962989330291748,
"learning_rate": 0.00018858300898449764,
"loss": 0.9607,
"step": 700
},
{
"epoch": 0.10912793713616016,
"grad_norm": 0.27997785806655884,
"learning_rate": 0.00018820636218513354,
"loss": 1.3478,
"step": 710
},
{
"epoch": 0.110664950335261,
"grad_norm": 0.3026648759841919,
"learning_rate": 0.00018782399180133187,
"loss": 1.1,
"step": 720
},
{
"epoch": 0.11220196353436185,
"grad_norm": 0.3552872836589813,
"learning_rate": 0.00018743592264455888,
"loss": 1.0816,
"step": 730
},
{
"epoch": 0.1137389767334627,
"grad_norm": 0.5152252316474915,
"learning_rate": 0.00018704217989606604,
"loss": 1.1635,
"step": 740
},
{
"epoch": 0.11527598993256355,
"grad_norm": 0.49878814816474915,
"learning_rate": 0.00018664278910525607,
"loss": 0.9708,
"step": 750
},
{
"epoch": 0.11681300313166439,
"grad_norm": 0.36649784445762634,
"learning_rate": 0.00018466961068200864,
"loss": 1.3605,
"step": 760
},
{
"epoch": 0.11835001633076524,
"grad_norm": 0.3871256411075592,
"learning_rate": 0.00018427399080293285,
"loss": 1.2319,
"step": 770
},
{
"epoch": 0.11988702952986609,
"grad_norm": 0.3366815447807312,
"learning_rate": 0.0001838737660952719,
"loss": 1.0277,
"step": 780
},
{
"epoch": 0.12142404272896694,
"grad_norm": 0.5359935164451599,
"learning_rate": 0.0001834689584277679,
"loss": 1.1181,
"step": 790
},
{
"epoch": 0.12296105592806779,
"grad_norm": 0.6008575558662415,
"learning_rate": 0.00018305958991958127,
"loss": 0.9519,
"step": 800
},
{
"epoch": 0.12449806912716863,
"grad_norm": 0.2891458570957184,
"learning_rate": 0.000182645682939082,
"loss": 1.3973,
"step": 810
},
{
"epoch": 0.12603508232626948,
"grad_norm": 0.34036985039711,
"learning_rate": 0.00018222726010262754,
"loss": 1.3075,
"step": 820
},
{
"epoch": 0.12757209552537033,
"grad_norm": 0.3681768476963043,
"learning_rate": 0.00018180434427332682,
"loss": 1.1499,
"step": 830
},
{
"epoch": 0.12910910872447118,
"grad_norm": 0.43317529559135437,
"learning_rate": 0.0001813769585597912,
"loss": 1.0596,
"step": 840
},
{
"epoch": 0.13064612192357203,
"grad_norm": 0.5329003930091858,
"learning_rate": 0.00018094512631487174,
"loss": 0.9966,
"step": 850
},
{
"epoch": 0.13218313512267288,
"grad_norm": 0.2697789967060089,
"learning_rate": 0.00018050887113438299,
"loss": 1.3611,
"step": 860
},
{
"epoch": 0.1337201483217737,
"grad_norm": 0.30477920174598694,
"learning_rate": 0.00018006821685581396,
"loss": 1.2413,
"step": 870
},
{
"epoch": 0.13525716152087455,
"grad_norm": 0.36294734477996826,
"learning_rate": 0.00017962318755702545,
"loss": 1.1026,
"step": 880
},
{
"epoch": 0.1367941747199754,
"grad_norm": 0.39103931188583374,
"learning_rate": 0.00017917380755493438,
"loss": 1.0255,
"step": 890
},
{
"epoch": 0.13833118791907625,
"grad_norm": 0.4765736162662506,
"learning_rate": 0.00017872010140418523,
"loss": 1.0665,
"step": 900
},
{
"epoch": 0.1398682011181771,
"grad_norm": 0.3002453148365021,
"learning_rate": 0.00017826209389580817,
"loss": 1.3826,
"step": 910
},
{
"epoch": 0.14140521431727796,
"grad_norm": 0.46594709157943726,
"learning_rate": 0.00017779981005586456,
"loss": 1.1233,
"step": 920
},
{
"epoch": 0.1429422275163788,
"grad_norm": 0.3838465213775635,
"learning_rate": 0.0001773332751440796,
"loss": 1.1245,
"step": 930
},
{
"epoch": 0.14447924071547966,
"grad_norm": 0.4277382493019104,
"learning_rate": 0.0001768625146524618,
"loss": 1.0859,
"step": 940
},
{
"epoch": 0.14601625391458048,
"grad_norm": 0.5642426013946533,
"learning_rate": 0.00017638755430391032,
"loss": 0.987,
"step": 950
},
{
"epoch": 0.14755326711368133,
"grad_norm": 0.33334049582481384,
"learning_rate": 0.00017590842005080944,
"loss": 1.5208,
"step": 960
},
{
"epoch": 0.14909028031278218,
"grad_norm": 0.386211633682251,
"learning_rate": 0.00017542513807361037,
"loss": 1.3236,
"step": 970
},
{
"epoch": 0.15062729351188303,
"grad_norm": 0.47577378153800964,
"learning_rate": 0.0001749377347794008,
"loss": 1.1644,
"step": 980
},
{
"epoch": 0.15216430671098388,
"grad_norm": 0.4588184654712677,
"learning_rate": 0.00017444623680046186,
"loss": 1.0378,
"step": 990
},
{
"epoch": 0.15370131991008473,
"grad_norm": 0.5750458836555481,
"learning_rate": 0.00017395067099281315,
"loss": 0.9985,
"step": 1000
},
{
"epoch": 0.15523833310918558,
"grad_norm": 0.2784772515296936,
"learning_rate": 0.00017345106443474503,
"loss": 1.4056,
"step": 1010
},
{
"epoch": 0.15677534630828643,
"grad_norm": 0.33730465173721313,
"learning_rate": 0.00017294744442533926,
"loss": 1.2181,
"step": 1020
},
{
"epoch": 0.15831235950738726,
"grad_norm": 0.32729485630989075,
"learning_rate": 0.00017243983848297712,
"loss": 1.1402,
"step": 1030
},
{
"epoch": 0.1598493727064881,
"grad_norm": 0.5524439811706543,
"learning_rate": 0.00017192827434383596,
"loss": 1.0252,
"step": 1040
},
{
"epoch": 0.16138638590558896,
"grad_norm": 0.9348641037940979,
"learning_rate": 0.00017141277996037357,
"loss": 1.0467,
"step": 1050
},
{
"epoch": 0.1629233991046898,
"grad_norm": 0.3065289258956909,
"learning_rate": 0.00017089338349980085,
"loss": 1.2842,
"step": 1060
},
{
"epoch": 0.16446041230379066,
"grad_norm": 0.30789056420326233,
"learning_rate": 0.00017037011334254275,
"loss": 1.1952,
"step": 1070
},
{
"epoch": 0.1659974255028915,
"grad_norm": 0.3652806282043457,
"learning_rate": 0.00016984299808068745,
"loss": 1.1373,
"step": 1080
},
{
"epoch": 0.16753443870199236,
"grad_norm": 0.45899513363838196,
"learning_rate": 0.0001693120665164241,
"loss": 1.0658,
"step": 1090
},
{
"epoch": 0.1690714519010932,
"grad_norm": 0.4147142469882965,
"learning_rate": 0.00016877734766046904,
"loss": 0.939,
"step": 1100
},
{
"epoch": 0.17060846510019403,
"grad_norm": 0.35689154267311096,
"learning_rate": 0.00016823887073048072,
"loss": 1.4027,
"step": 1110
},
{
"epoch": 0.17214547829929489,
"grad_norm": 0.34796151518821716,
"learning_rate": 0.00016769666514946302,
"loss": 1.3322,
"step": 1120
},
{
"epoch": 0.17368249149839574,
"grad_norm": 0.3684556186199188,
"learning_rate": 0.0001671507605441576,
"loss": 1.0941,
"step": 1130
},
{
"epoch": 0.1752195046974966,
"grad_norm": 0.6131370067596436,
"learning_rate": 0.00016660118674342517,
"loss": 0.9889,
"step": 1140
},
{
"epoch": 0.17675651789659744,
"grad_norm": 0.40539348125457764,
"learning_rate": 0.00016604797377661553,
"loss": 0.9889,
"step": 1150
},
{
"epoch": 0.1782935310956983,
"grad_norm": 0.3540327847003937,
"learning_rate": 0.0001654911518719267,
"loss": 1.4491,
"step": 1160
},
{
"epoch": 0.17983054429479914,
"grad_norm": 0.30772891640663147,
"learning_rate": 0.00016493075145475328,
"loss": 1.2342,
"step": 1170
},
{
"epoch": 0.1813675574939,
"grad_norm": 0.4468260705471039,
"learning_rate": 0.00016436680314602388,
"loss": 1.1081,
"step": 1180
},
{
"epoch": 0.18290457069300084,
"grad_norm": 0.4621022641658783,
"learning_rate": 0.00016379933776052816,
"loss": 1.0314,
"step": 1190
},
{
"epoch": 0.18444158389210166,
"grad_norm": 0.6054849624633789,
"learning_rate": 0.00016322838630523275,
"loss": 1.0168,
"step": 1200
},
{
"epoch": 0.1859785970912025,
"grad_norm": 0.303941935300827,
"learning_rate": 0.0001626539799775873,
"loss": 1.3332,
"step": 1210
},
{
"epoch": 0.18751561029030336,
"grad_norm": 0.3241790533065796,
"learning_rate": 0.00016207615016381968,
"loss": 1.3425,
"step": 1220
},
{
"epoch": 0.18905262348940421,
"grad_norm": 0.3525269627571106,
"learning_rate": 0.00016149492843722097,
"loss": 1.0795,
"step": 1230
},
{
"epoch": 0.19058963668850507,
"grad_norm": 0.513918936252594,
"learning_rate": 0.0001609103465564203,
"loss": 1.0941,
"step": 1240
},
{
"epoch": 0.19212664988760592,
"grad_norm": 1.0173454284667969,
"learning_rate": 0.00016032243646364958,
"loss": 0.9835,
"step": 1250
},
{
"epoch": 0.19366366308670677,
"grad_norm": 0.3847029507160187,
"learning_rate": 0.000159731230282998,
"loss": 1.3678,
"step": 1260
},
{
"epoch": 0.19520067628580762,
"grad_norm": 0.35179686546325684,
"learning_rate": 0.00015913676031865693,
"loss": 1.2158,
"step": 1270
},
{
"epoch": 0.19673768948490844,
"grad_norm": 0.4118972420692444,
"learning_rate": 0.00015853905905315453,
"loss": 1.1549,
"step": 1280
},
{
"epoch": 0.1982747026840093,
"grad_norm": 0.4301435351371765,
"learning_rate": 0.0001579381591455811,
"loss": 1.1278,
"step": 1290
},
{
"epoch": 0.19981171588311014,
"grad_norm": 0.5036038756370544,
"learning_rate": 0.0001573340934298044,
"loss": 0.9755,
"step": 1300
},
{
"epoch": 0.201348729082211,
"grad_norm": 0.3321966230869293,
"learning_rate": 0.00015672689491267567,
"loss": 1.2173,
"step": 1310
},
{
"epoch": 0.20288574228131184,
"grad_norm": 0.30094143748283386,
"learning_rate": 0.00015611659677222596,
"loss": 1.1808,
"step": 1320
},
{
"epoch": 0.2044227554804127,
"grad_norm": 0.3850231170654297,
"learning_rate": 0.00015550323235585348,
"loss": 0.9827,
"step": 1330
},
{
"epoch": 0.20595976867951354,
"grad_norm": 0.4884297251701355,
"learning_rate": 0.0001548868351785011,
"loss": 1.1398,
"step": 1340
},
{
"epoch": 0.2074967818786144,
"grad_norm": 0.5098474621772766,
"learning_rate": 0.00015426743892082553,
"loss": 0.9477,
"step": 1350
},
{
"epoch": 0.20903379507771522,
"grad_norm": 0.33422473073005676,
"learning_rate": 0.00015364507742735646,
"loss": 1.4652,
"step": 1360
},
{
"epoch": 0.21057080827681607,
"grad_norm": 0.32810479402542114,
"learning_rate": 0.00015301978470464768,
"loss": 1.1287,
"step": 1370
},
{
"epoch": 0.21210782147591692,
"grad_norm": 0.42300912737846375,
"learning_rate": 0.00015239159491941864,
"loss": 1.1596,
"step": 1380
},
{
"epoch": 0.21364483467501777,
"grad_norm": 0.5686854124069214,
"learning_rate": 0.00015176054239668767,
"loss": 1.1619,
"step": 1390
},
{
"epoch": 0.21518184787411862,
"grad_norm": 0.5695457458496094,
"learning_rate": 0.00015112666161789644,
"loss": 0.9554,
"step": 1400
},
{
"epoch": 0.21671886107321947,
"grad_norm": 0.2944013774394989,
"learning_rate": 0.00015048998721902575,
"loss": 1.325,
"step": 1410
},
{
"epoch": 0.21825587427232032,
"grad_norm": 0.3382372558116913,
"learning_rate": 0.0001498505539887031,
"loss": 1.235,
"step": 1420
},
{
"epoch": 0.21979288747142117,
"grad_norm": 0.3798046410083771,
"learning_rate": 0.00014920839686630166,
"loss": 1.1059,
"step": 1430
},
{
"epoch": 0.221329900670522,
"grad_norm": 0.4425491988658905,
"learning_rate": 0.00014856355094003133,
"loss": 1.0067,
"step": 1440
},
{
"epoch": 0.22286691386962285,
"grad_norm": 0.4461119771003723,
"learning_rate": 0.00014791605144502126,
"loss": 1.0163,
"step": 1450
},
{
"epoch": 0.2244039270687237,
"grad_norm": 0.30612340569496155,
"learning_rate": 0.00014726593376139478,
"loss": 1.3924,
"step": 1460
},
{
"epoch": 0.22594094026782455,
"grad_norm": 0.33173027634620667,
"learning_rate": 0.00014661323341233596,
"loss": 1.0982,
"step": 1470
},
{
"epoch": 0.2274779534669254,
"grad_norm": 0.3924216330051422,
"learning_rate": 0.00014595798606214882,
"loss": 1.1227,
"step": 1480
},
{
"epoch": 0.22901496666602625,
"grad_norm": 0.46780848503112793,
"learning_rate": 0.00014530022751430837,
"loss": 1.1159,
"step": 1490
},
{
"epoch": 0.2305519798651271,
"grad_norm": 0.47939273715019226,
"learning_rate": 0.00014463999370950453,
"loss": 0.9624,
"step": 1500
},
{
"epoch": 0.23208899306422795,
"grad_norm": 0.36672261357307434,
"learning_rate": 0.00014397732072367794,
"loss": 1.3798,
"step": 1510
},
{
"epoch": 0.23362600626332877,
"grad_norm": 0.31935515999794006,
"learning_rate": 0.00014331224476604908,
"loss": 1.0577,
"step": 1520
},
{
"epoch": 0.23516301946242962,
"grad_norm": 0.36450207233428955,
"learning_rate": 0.00014264480217713945,
"loss": 0.984,
"step": 1530
},
{
"epoch": 0.23670003266153047,
"grad_norm": 0.44956207275390625,
"learning_rate": 0.00014197502942678614,
"loss": 0.964,
"step": 1540
},
{
"epoch": 0.23823704586063132,
"grad_norm": 0.5394241809844971,
"learning_rate": 0.00014130296311214898,
"loss": 0.9702,
"step": 1550
},
{
"epoch": 0.23977405905973218,
"grad_norm": 0.32651957869529724,
"learning_rate": 0.00014062863995571064,
"loss": 1.369,
"step": 1560
},
{
"epoch": 0.24131107225883303,
"grad_norm": 0.4551924765110016,
"learning_rate": 0.00013995209680327047,
"loss": 1.2096,
"step": 1570
},
{
"epoch": 0.24284808545793388,
"grad_norm": 0.3988681733608246,
"learning_rate": 0.00013927337062193078,
"loss": 1.0385,
"step": 1580
},
{
"epoch": 0.24438509865703473,
"grad_norm": 0.46369948983192444,
"learning_rate": 0.00013859249849807725,
"loss": 1.0987,
"step": 1590
},
{
"epoch": 0.24592211185613558,
"grad_norm": 0.4631679952144623,
"learning_rate": 0.0001379095176353523,
"loss": 0.9197,
"step": 1600
},
{
"epoch": 0.2474591250552364,
"grad_norm": 0.2949797213077545,
"learning_rate": 0.00013722446535262227,
"loss": 1.4423,
"step": 1610
},
{
"epoch": 0.24899613825433725,
"grad_norm": 0.35625016689300537,
"learning_rate": 0.00013653737908193828,
"loss": 1.2224,
"step": 1620
},
{
"epoch": 0.25053315145343813,
"grad_norm": 0.6409035325050354,
"learning_rate": 0.00013584829636649096,
"loss": 1.1653,
"step": 1630
},
{
"epoch": 0.25207016465253895,
"grad_norm": 0.6155157089233398,
"learning_rate": 0.00013515725485855901,
"loss": 0.9812,
"step": 1640
},
{
"epoch": 0.2536071778516398,
"grad_norm": 0.5449208617210388,
"learning_rate": 0.0001344642923174517,
"loss": 0.9328,
"step": 1650
},
{
"epoch": 0.25514419105074065,
"grad_norm": 0.3098817467689514,
"learning_rate": 0.00013376944660744596,
"loss": 1.3232,
"step": 1660
},
{
"epoch": 0.2566812042498415,
"grad_norm": 0.38490045070648193,
"learning_rate": 0.00013307275569571708,
"loss": 1.2452,
"step": 1670
},
{
"epoch": 0.25821821744894236,
"grad_norm": 0.37269195914268494,
"learning_rate": 0.0001323742576502644,
"loss": 1.1327,
"step": 1680
},
{
"epoch": 0.2597552306480432,
"grad_norm": 0.5462695956230164,
"learning_rate": 0.00013167399063783114,
"loss": 1.0827,
"step": 1690
},
{
"epoch": 0.26129224384714406,
"grad_norm": 0.5448663234710693,
"learning_rate": 0.00013097199292181895,
"loss": 0.9588,
"step": 1700
},
{
"epoch": 0.2628292570462449,
"grad_norm": 0.30637407302856445,
"learning_rate": 0.00013026830286019708,
"loss": 1.456,
"step": 1710
},
{
"epoch": 0.26436627024534576,
"grad_norm": 0.4137035012245178,
"learning_rate": 0.00012956295890340647,
"loss": 1.167,
"step": 1720
},
{
"epoch": 0.2659032834444466,
"grad_norm": 0.34960073232650757,
"learning_rate": 0.00012885599959225897,
"loss": 1.0419,
"step": 1730
},
{
"epoch": 0.2674402966435474,
"grad_norm": 0.5856695175170898,
"learning_rate": 0.00012814746355583107,
"loss": 1.0183,
"step": 1740
},
{
"epoch": 0.2689773098426483,
"grad_norm": 0.5495254397392273,
"learning_rate": 0.00012743738950935353,
"loss": 0.9649,
"step": 1750
},
{
"epoch": 0.2705143230417491,
"grad_norm": 0.3238767385482788,
"learning_rate": 0.0001267258162520956,
"loss": 1.3423,
"step": 1760
},
{
"epoch": 0.27205133624085,
"grad_norm": 0.41960009932518005,
"learning_rate": 0.0001260127826652453,
"loss": 1.2678,
"step": 1770
},
{
"epoch": 0.2735883494399508,
"grad_norm": 0.4313008189201355,
"learning_rate": 0.0001252983277097846,
"loss": 1.0865,
"step": 1780
},
{
"epoch": 0.2751253626390517,
"grad_norm": 0.43165504932403564,
"learning_rate": 0.00012458249042436085,
"loss": 0.9728,
"step": 1790
},
{
"epoch": 0.2766623758381525,
"grad_norm": 0.46663326025009155,
"learning_rate": 0.0001238653099231534,
"loss": 0.921,
"step": 1800
},
{
"epoch": 0.27819938903725333,
"grad_norm": 0.3772903382778168,
"learning_rate": 0.00012314682539373654,
"loss": 1.5042,
"step": 1810
},
{
"epoch": 0.2797364022363542,
"grad_norm": 0.35220977663993835,
"learning_rate": 0.00012242707609493814,
"loss": 1.2307,
"step": 1820
},
{
"epoch": 0.28127341543545503,
"grad_norm": 0.33617788553237915,
"learning_rate": 0.0001217061013546945,
"loss": 1.0916,
"step": 1830
},
{
"epoch": 0.2828104286345559,
"grad_norm": 0.4361517131328583,
"learning_rate": 0.00012098394056790152,
"loss": 1.1055,
"step": 1840
},
{
"epoch": 0.28434744183365673,
"grad_norm": 0.47860532999038696,
"learning_rate": 0.00012026063319426204,
"loss": 0.9713,
"step": 1850
},
{
"epoch": 0.2858844550327576,
"grad_norm": 0.26366356015205383,
"learning_rate": 0.00011953621875612975,
"loss": 1.3679,
"step": 1860
},
{
"epoch": 0.28742146823185843,
"grad_norm": 0.35310980677604675,
"learning_rate": 0.0001188107368363496,
"loss": 1.1388,
"step": 1870
},
{
"epoch": 0.2889584814309593,
"grad_norm": 0.37847745418548584,
"learning_rate": 0.00011808422707609496,
"loss": 1.1046,
"step": 1880
},
{
"epoch": 0.29049549463006014,
"grad_norm": 0.3944340944290161,
"learning_rate": 0.00011735672917270166,
"loss": 1.0088,
"step": 1890
},
{
"epoch": 0.29203250782916096,
"grad_norm": 0.5353487730026245,
"learning_rate": 0.00011662828287749876,
"loss": 0.915,
"step": 1900
},
{
"epoch": 0.29356952102826184,
"grad_norm": 0.28208622336387634,
"learning_rate": 0.00011589892799363657,
"loss": 1.4331,
"step": 1910
},
{
"epoch": 0.29510653422736266,
"grad_norm": 0.3970017731189728,
"learning_rate": 0.00011516870437391174,
"loss": 1.1775,
"step": 1920
},
{
"epoch": 0.29664354742646354,
"grad_norm": 0.5350090861320496,
"learning_rate": 0.00011443765191858966,
"loss": 1.0827,
"step": 1930
},
{
"epoch": 0.29818056062556436,
"grad_norm": 0.4986719489097595,
"learning_rate": 0.0001137058105732242,
"loss": 1.0189,
"step": 1940
},
{
"epoch": 0.29971757382466524,
"grad_norm": 0.4850538969039917,
"learning_rate": 0.00011297322032647522,
"loss": 0.8724,
"step": 1950
},
{
"epoch": 0.30125458702376606,
"grad_norm": 0.408117413520813,
"learning_rate": 0.00011223992120792328,
"loss": 1.3742,
"step": 1960
},
{
"epoch": 0.30279160022286694,
"grad_norm": 0.2824944257736206,
"learning_rate": 0.0001115059532858826,
"loss": 1.2337,
"step": 1970
},
{
"epoch": 0.30432861342196776,
"grad_norm": 0.35323965549468994,
"learning_rate": 0.00011077135666521161,
"loss": 1.0118,
"step": 1980
},
{
"epoch": 0.3058656266210686,
"grad_norm": 0.6557132601737976,
"learning_rate": 0.00011003617148512149,
"loss": 1.0471,
"step": 1990
},
{
"epoch": 0.30740263982016947,
"grad_norm": 0.48450276255607605,
"learning_rate": 0.00010930043791698307,
"loss": 0.9396,
"step": 2000
},
{
"epoch": 0.3089396530192703,
"grad_norm": 0.32962167263031006,
"learning_rate": 0.00010856419616213168,
"loss": 1.3716,
"step": 2010
},
{
"epoch": 0.31047666621837117,
"grad_norm": 0.3237484395503998,
"learning_rate": 0.00010782748644967061,
"loss": 1.183,
"step": 2020
},
{
"epoch": 0.312013679417472,
"grad_norm": 0.41155266761779785,
"learning_rate": 0.00010709034903427278,
"loss": 1.1026,
"step": 2030
},
{
"epoch": 0.31355069261657287,
"grad_norm": 0.4315926730632782,
"learning_rate": 0.0001063528241939815,
"loss": 0.9435,
"step": 2040
},
{
"epoch": 0.3150877058156737,
"grad_norm": 0.5973414778709412,
"learning_rate": 0.00010561495222800923,
"loss": 0.9511,
"step": 2050
},
{
"epoch": 0.3166247190147745,
"grad_norm": 0.33315151929855347,
"learning_rate": 0.00010487677345453591,
"loss": 1.3807,
"step": 2060
},
{
"epoch": 0.3181617322138754,
"grad_norm": 0.338500440120697,
"learning_rate": 0.00010413832820850569,
"loss": 1.2349,
"step": 2070
},
{
"epoch": 0.3196987454129762,
"grad_norm": 0.37937551736831665,
"learning_rate": 0.00010339965683942319,
"loss": 1.0745,
"step": 2080
},
{
"epoch": 0.3212357586120771,
"grad_norm": 0.41661012172698975,
"learning_rate": 0.00010266079970914862,
"loss": 1.0241,
"step": 2090
},
{
"epoch": 0.3227727718111779,
"grad_norm": 0.5309539437294006,
"learning_rate": 0.00010192179718969233,
"loss": 0.9444,
"step": 2100
},
{
"epoch": 0.3243097850102788,
"grad_norm": 0.2757546603679657,
"learning_rate": 0.00010118268966100903,
"loss": 1.3945,
"step": 2110
},
{
"epoch": 0.3258467982093796,
"grad_norm": 0.381526380777359,
"learning_rate": 0.00010044351750879111,
"loss": 1.2218,
"step": 2120
},
{
"epoch": 0.3273838114084805,
"grad_norm": 0.36533376574516296,
"learning_rate": 9.970432112226225e-05,
"loss": 1.0668,
"step": 2130
},
{
"epoch": 0.3289208246075813,
"grad_norm": 0.5481194853782654,
"learning_rate": 9.896514089197011e-05,
"loss": 1.0275,
"step": 2140
},
{
"epoch": 0.33045783780668214,
"grad_norm": 0.6866357922554016,
"learning_rate": 9.822601720757969e-05,
"loss": 1.0401,
"step": 2150
},
{
"epoch": 0.331994851005783,
"grad_norm": 0.2852606475353241,
"learning_rate": 9.748699045566626e-05,
"loss": 1.3537,
"step": 2160
},
{
"epoch": 0.33353186420488384,
"grad_norm": 0.39954090118408203,
"learning_rate": 9.674810101750852e-05,
"loss": 1.1194,
"step": 2170
},
{
"epoch": 0.3350688774039847,
"grad_norm": 0.3414720892906189,
"learning_rate": 9.600938926688223e-05,
"loss": 1.0265,
"step": 2180
},
{
"epoch": 0.33660589060308554,
"grad_norm": 0.5632903575897217,
"learning_rate": 9.527089556785405e-05,
"loss": 1.0218,
"step": 2190
},
{
"epoch": 0.3381429038021864,
"grad_norm": 0.5569449663162231,
"learning_rate": 9.453266027257614e-05,
"loss": 1.0168,
"step": 2200
},
{
"epoch": 0.33967991700128725,
"grad_norm": 0.3332814574241638,
"learning_rate": 9.37947237190811e-05,
"loss": 1.2925,
"step": 2210
},
{
"epoch": 0.34121693020038807,
"grad_norm": 0.3622121214866638,
"learning_rate": 9.30571262290779e-05,
"loss": 1.2141,
"step": 2220
},
{
"epoch": 0.34275394339948895,
"grad_norm": 0.3554399311542511,
"learning_rate": 9.231990810574884e-05,
"loss": 1.0059,
"step": 2230
},
{
"epoch": 0.34429095659858977,
"grad_norm": 0.46376103162765503,
"learning_rate": 9.158310963154705e-05,
"loss": 0.9922,
"step": 2240
},
{
"epoch": 0.34582796979769065,
"grad_norm": 0.5186789035797119,
"learning_rate": 9.084677106599559e-05,
"loss": 0.9778,
"step": 2250
},
{
"epoch": 0.34736498299679147,
"grad_norm": 0.25012660026550293,
"learning_rate": 9.01109326434875e-05,
"loss": 1.3192,
"step": 2260
},
{
"epoch": 0.34890199619589235,
"grad_norm": 0.39215368032455444,
"learning_rate": 8.937563457108758e-05,
"loss": 1.2129,
"step": 2270
},
{
"epoch": 0.3504390093949932,
"grad_norm": 0.33098575472831726,
"learning_rate": 8.864091702633516e-05,
"loss": 1.1595,
"step": 2280
},
{
"epoch": 0.35197602259409405,
"grad_norm": 0.47959211468696594,
"learning_rate": 8.790682015504889e-05,
"loss": 1.0077,
"step": 2290
},
{
"epoch": 0.3535130357931949,
"grad_norm": 0.6163510084152222,
"learning_rate": 8.717338406913312e-05,
"loss": 0.875,
"step": 2300
},
{
"epoch": 0.3550500489922957,
"grad_norm": 0.29523080587387085,
"learning_rate": 8.64406488443861e-05,
"loss": 1.3584,
"step": 2310
},
{
"epoch": 0.3565870621913966,
"grad_norm": 0.3199009895324707,
"learning_rate": 8.570865451831022e-05,
"loss": 1.1183,
"step": 2320
},
{
"epoch": 0.3581240753904974,
"grad_norm": 0.5864169597625732,
"learning_rate": 8.497744108792429e-05,
"loss": 1.0751,
"step": 2330
},
{
"epoch": 0.3596610885895983,
"grad_norm": 0.46623820066452026,
"learning_rate": 8.424704850757816e-05,
"loss": 1.0338,
"step": 2340
},
{
"epoch": 0.3611981017886991,
"grad_norm": 0.5505573749542236,
"learning_rate": 8.351751668676933e-05,
"loss": 0.9063,
"step": 2350
},
{
"epoch": 0.3627351149878,
"grad_norm": 0.33322644233703613,
"learning_rate": 8.278888548796256e-05,
"loss": 1.4438,
"step": 2360
},
{
"epoch": 0.3642721281869008,
"grad_norm": 0.36330074071884155,
"learning_rate": 8.20611947244114e-05,
"loss": 1.0938,
"step": 2370
},
{
"epoch": 0.3658091413860017,
"grad_norm": 0.33430981636047363,
"learning_rate": 8.133448415798313e-05,
"loss": 1.1108,
"step": 2380
},
{
"epoch": 0.3673461545851025,
"grad_norm": 0.5140926241874695,
"learning_rate": 8.060879349698582e-05,
"loss": 1.0851,
"step": 2390
},
{
"epoch": 0.3688831677842033,
"grad_norm": 0.521024763584137,
"learning_rate": 7.988416239399865e-05,
"loss": 0.927,
"step": 2400
},
{
"epoch": 0.3704201809833042,
"grad_norm": 0.3185390532016754,
"learning_rate": 7.916063044370548e-05,
"loss": 1.4175,
"step": 2410
},
{
"epoch": 0.371957194182405,
"grad_norm": 0.3560965359210968,
"learning_rate": 7.84382371807311e-05,
"loss": 1.1595,
"step": 2420
},
{
"epoch": 0.3734942073815059,
"grad_norm": 0.5643250346183777,
"learning_rate": 7.771702207748108e-05,
"loss": 1.0753,
"step": 2430
},
{
"epoch": 0.37503122058060673,
"grad_norm": 0.43922996520996094,
"learning_rate": 7.699702454198502e-05,
"loss": 0.9515,
"step": 2440
},
{
"epoch": 0.3765682337797076,
"grad_norm": 0.5192488431930542,
"learning_rate": 7.627828391574324e-05,
"loss": 0.982,
"step": 2450
},
{
"epoch": 0.37810524697880843,
"grad_norm": 0.3009122312068939,
"learning_rate": 7.5560839471577e-05,
"loss": 1.3318,
"step": 2460
},
{
"epoch": 0.37964226017790925,
"grad_norm": 0.3585622012615204,
"learning_rate": 7.484473041148274e-05,
"loss": 1.0875,
"step": 2470
},
{
"epoch": 0.38117927337701013,
"grad_norm": 0.3755141496658325,
"learning_rate": 7.412999586448988e-05,
"loss": 1.0223,
"step": 2480
},
{
"epoch": 0.38271628657611095,
"grad_norm": 0.3720460534095764,
"learning_rate": 7.341667488452294e-05,
"loss": 0.9879,
"step": 2490
},
{
"epoch": 0.38425329977521183,
"grad_norm": 0.5014258027076721,
"learning_rate": 7.270480644826749e-05,
"loss": 0.8952,
"step": 2500
},
{
"epoch": 0.38579031297431265,
"grad_norm": 0.3038366734981537,
"learning_rate": 7.199442945304034e-05,
"loss": 1.2543,
"step": 2510
},
{
"epoch": 0.38732732617341353,
"grad_norm": 0.3395344913005829,
"learning_rate": 7.128558271466436e-05,
"loss": 1.1549,
"step": 2520
},
{
"epoch": 0.38886433937251436,
"grad_norm": 0.3712739646434784,
"learning_rate": 7.057830496534729e-05,
"loss": 1.1054,
"step": 2530
},
{
"epoch": 0.39040135257161523,
"grad_norm": 0.46756452322006226,
"learning_rate": 6.987263485156556e-05,
"loss": 1.026,
"step": 2540
},
{
"epoch": 0.39193836577071606,
"grad_norm": 0.4664836525917053,
"learning_rate": 6.916861093195249e-05,
"loss": 0.9236,
"step": 2550
},
{
"epoch": 0.3934753789698169,
"grad_norm": 0.35312631726264954,
"learning_rate": 6.84662716751915e-05,
"loss": 1.3595,
"step": 2560
},
{
"epoch": 0.39501239216891776,
"grad_norm": 0.41039901971817017,
"learning_rate": 6.776565545791406e-05,
"loss": 1.1796,
"step": 2570
},
{
"epoch": 0.3965494053680186,
"grad_norm": 0.39480772614479065,
"learning_rate": 6.706680056260268e-05,
"loss": 1.1143,
"step": 2580
},
{
"epoch": 0.39808641856711946,
"grad_norm": 0.46342942118644714,
"learning_rate": 6.636974517549928e-05,
"loss": 1.0041,
"step": 2590
},
{
"epoch": 0.3996234317662203,
"grad_norm": 0.48610740900039673,
"learning_rate": 6.567452738451859e-05,
"loss": 0.9284,
"step": 2600
},
{
"epoch": 0.40116044496532116,
"grad_norm": 0.357565701007843,
"learning_rate": 6.498118517716687e-05,
"loss": 1.3317,
"step": 2610
},
{
"epoch": 0.402697458164422,
"grad_norm": 0.31676915287971497,
"learning_rate": 6.428975643846636e-05,
"loss": 1.161,
"step": 2620
},
{
"epoch": 0.4042344713635228,
"grad_norm": 0.3662184476852417,
"learning_rate": 6.360027894888519e-05,
"loss": 1.0355,
"step": 2630
},
{
"epoch": 0.4057714845626237,
"grad_norm": 0.4919685125350952,
"learning_rate": 6.291279038227292e-05,
"loss": 0.9824,
"step": 2640
},
{
"epoch": 0.4073084977617245,
"grad_norm": 0.5702899098396301,
"learning_rate": 6.222732830380207e-05,
"loss": 0.9796,
"step": 2650
},
{
"epoch": 0.4088455109608254,
"grad_norm": 0.3060368299484253,
"learning_rate": 6.154393016791545e-05,
"loss": 1.3381,
"step": 2660
},
{
"epoch": 0.4103825241599262,
"grad_norm": 0.3788270354270935,
"learning_rate": 6.086263331627976e-05,
"loss": 1.2656,
"step": 2670
},
{
"epoch": 0.4119195373590271,
"grad_norm": 0.40077683329582214,
"learning_rate": 6.0183474975745044e-05,
"loss": 1.1281,
"step": 2680
},
{
"epoch": 0.4134565505581279,
"grad_norm": 0.3977091610431671,
"learning_rate": 5.950649225631055e-05,
"loss": 1.0375,
"step": 2690
},
{
"epoch": 0.4149935637572288,
"grad_norm": 0.5339929461479187,
"learning_rate": 5.883172214909718e-05,
"loss": 0.962,
"step": 2700
},
{
"epoch": 0.4165305769563296,
"grad_norm": 0.30091047286987305,
"learning_rate": 5.815920152432608e-05,
"loss": 1.3766,
"step": 2710
},
{
"epoch": 0.41806759015543044,
"grad_norm": 0.33363446593284607,
"learning_rate": 5.748896712930407e-05,
"loss": 1.2019,
"step": 2720
},
{
"epoch": 0.4196046033545313,
"grad_norm": 0.3920862078666687,
"learning_rate": 5.682105558641566e-05,
"loss": 1.1197,
"step": 2730
},
{
"epoch": 0.42114161655363214,
"grad_norm": 0.44188392162323,
"learning_rate": 5.615550339112214e-05,
"loss": 1.0676,
"step": 2740
},
{
"epoch": 0.422678629752733,
"grad_norm": 0.8947547674179077,
"learning_rate": 5.5492346909967315e-05,
"loss": 0.9482,
"step": 2750
},
{
"epoch": 0.42421564295183384,
"grad_norm": 0.3606780767440796,
"learning_rate": 5.48316223785903e-05,
"loss": 1.3776,
"step": 2760
},
{
"epoch": 0.4257526561509347,
"grad_norm": 0.4248056709766388,
"learning_rate": 5.417336589974564e-05,
"loss": 1.1134,
"step": 2770
},
{
"epoch": 0.42728966935003554,
"grad_norm": 0.410755455493927,
"learning_rate": 5.351761344133077e-05,
"loss": 1.0821,
"step": 2780
},
{
"epoch": 0.4288266825491364,
"grad_norm": 0.47652509808540344,
"learning_rate": 5.2864400834420405e-05,
"loss": 0.9666,
"step": 2790
},
{
"epoch": 0.43036369574823724,
"grad_norm": 0.6972517371177673,
"learning_rate": 5.221376377130884e-05,
"loss": 1.0056,
"step": 2800
},
{
"epoch": 0.43190070894733806,
"grad_norm": 0.34944674372673035,
"learning_rate": 5.1565737803559775e-05,
"loss": 1.3874,
"step": 2810
},
{
"epoch": 0.43343772214643894,
"grad_norm": 0.3123984932899475,
"learning_rate": 5.092035834006359e-05,
"loss": 1.1813,
"step": 2820
},
{
"epoch": 0.43497473534553976,
"grad_norm": 0.4548565745353699,
"learning_rate": 5.0277660645102534e-05,
"loss": 1.1569,
"step": 2830
},
{
"epoch": 0.43651174854464064,
"grad_norm": 0.51619553565979,
"learning_rate": 4.9637679836423924e-05,
"loss": 0.9336,
"step": 2840
},
{
"epoch": 0.43804876174374147,
"grad_norm": 0.48604270815849304,
"learning_rate": 4.9000450883321366e-05,
"loss": 0.8799,
"step": 2850
},
{
"epoch": 0.43958577494284234,
"grad_norm": 0.35553327202796936,
"learning_rate": 4.8366008604723756e-05,
"loss": 1.2362,
"step": 2860
},
{
"epoch": 0.44112278814194317,
"grad_norm": 0.3687051236629486,
"learning_rate": 4.773438766729299e-05,
"loss": 1.2282,
"step": 2870
},
{
"epoch": 0.442659801341044,
"grad_norm": 0.37749341130256653,
"learning_rate": 4.7105622583529466e-05,
"loss": 1.1588,
"step": 2880
},
{
"epoch": 0.44419681454014487,
"grad_norm": 0.46902376413345337,
"learning_rate": 4.647974770988658e-05,
"loss": 0.9736,
"step": 2890
},
{
"epoch": 0.4457338277392457,
"grad_norm": 0.5046992897987366,
"learning_rate": 4.585679724489318e-05,
"loss": 0.9313,
"step": 2900
},
{
"epoch": 0.44727084093834657,
"grad_norm": 0.27727243304252625,
"learning_rate": 4.523680522728503e-05,
"loss": 1.3174,
"step": 2910
},
{
"epoch": 0.4488078541374474,
"grad_norm": 0.71333247423172,
"learning_rate": 4.4619805534145e-05,
"loss": 1.2534,
"step": 2920
},
{
"epoch": 0.45034486733654827,
"grad_norm": 0.38468223810195923,
"learning_rate": 4.4005831879051826e-05,
"loss": 1.1244,
"step": 2930
},
{
"epoch": 0.4518818805356491,
"grad_norm": 0.4765458106994629,
"learning_rate": 4.339491781023801e-05,
"loss": 1.0577,
"step": 2940
},
{
"epoch": 0.45341889373475,
"grad_norm": 0.6454850435256958,
"learning_rate": 4.278709670875666e-05,
"loss": 0.9595,
"step": 2950
},
{
"epoch": 0.4549559069338508,
"grad_norm": 0.3479834198951721,
"learning_rate": 4.218240178665768e-05,
"loss": 1.4543,
"step": 2960
},
{
"epoch": 0.4564929201329516,
"grad_norm": 0.28524839878082275,
"learning_rate": 4.158086608517279e-05,
"loss": 1.1712,
"step": 2970
},
{
"epoch": 0.4580299333320525,
"grad_norm": 0.4688021242618561,
"learning_rate": 4.0982522472910315e-05,
"loss": 1.1151,
"step": 2980
},
{
"epoch": 0.4595669465311533,
"grad_norm": 0.46672606468200684,
"learning_rate": 4.038740364405911e-05,
"loss": 1.0985,
"step": 2990
},
{
"epoch": 0.4611039597302542,
"grad_norm": 0.5262246131896973,
"learning_rate": 3.9795542116602105e-05,
"loss": 0.9157,
"step": 3000
},
{
"epoch": 0.462640972929355,
"grad_norm": 0.5141198039054871,
"learning_rate": 3.920697023053949e-05,
"loss": 1.3735,
"step": 3010
},
{
"epoch": 0.4641779861284559,
"grad_norm": 0.37512609362602234,
"learning_rate": 3.862172014612162e-05,
"loss": 1.0864,
"step": 3020
},
{
"epoch": 0.4657149993275567,
"grad_norm": 0.5075603127479553,
"learning_rate": 3.8039823842091815e-05,
"loss": 1.0951,
"step": 3030
},
{
"epoch": 0.46725201252665755,
"grad_norm": 0.5750014781951904,
"learning_rate": 3.746131311393898e-05,
"loss": 0.9517,
"step": 3040
},
{
"epoch": 0.4687890257257584,
"grad_norm": 0.5644025802612305,
"learning_rate": 3.6886219572160095e-05,
"loss": 0.8794,
"step": 3050
},
{
"epoch": 0.47032603892485925,
"grad_norm": 0.3495708107948303,
"learning_rate": 3.631457464053318e-05,
"loss": 1.359,
"step": 3060
},
{
"epoch": 0.4718630521239601,
"grad_norm": 0.35652947425842285,
"learning_rate": 3.574640955440026e-05,
"loss": 1.1372,
"step": 3070
},
{
"epoch": 0.47340006532306095,
"grad_norm": 0.38782593607902527,
"learning_rate": 3.518175535896048e-05,
"loss": 1.0315,
"step": 3080
},
{
"epoch": 0.4749370785221618,
"grad_norm": 0.49639537930488586,
"learning_rate": 3.462064290757384e-05,
"loss": 1.0334,
"step": 3090
},
{
"epoch": 0.47647409172126265,
"grad_norm": 0.5320082306861877,
"learning_rate": 3.40631028600754e-05,
"loss": 0.9042,
"step": 3100
},
{
"epoch": 0.4780111049203635,
"grad_norm": 0.35209575295448303,
"learning_rate": 3.350916568109995e-05,
"loss": 1.3482,
"step": 3110
},
{
"epoch": 0.47954811811946435,
"grad_norm": 0.36500558257102966,
"learning_rate": 3.2958861638417284e-05,
"loss": 1.1504,
"step": 3120
},
{
"epoch": 0.4810851313185652,
"grad_norm": 0.5091226696968079,
"learning_rate": 3.241222080127839e-05,
"loss": 1.1243,
"step": 3130
},
{
"epoch": 0.48262214451766605,
"grad_norm": 0.4954063892364502,
"learning_rate": 3.186927303877262e-05,
"loss": 1.0066,
"step": 3140
},
{
"epoch": 0.4841591577167669,
"grad_norm": 0.7741270661354065,
"learning_rate": 3.1330048018195255e-05,
"loss": 0.8641,
"step": 3150
},
{
"epoch": 0.48569617091586775,
"grad_norm": 0.3022279143333435,
"learning_rate": 3.0794575203426804e-05,
"loss": 1.4419,
"step": 3160
},
{
"epoch": 0.4872331841149686,
"grad_norm": 0.28615984320640564,
"learning_rate": 3.0262883853322744e-05,
"loss": 1.1142,
"step": 3170
},
{
"epoch": 0.48877019731406945,
"grad_norm": 0.3599777817726135,
"learning_rate": 2.9735003020115092e-05,
"loss": 1.1776,
"step": 3180
},
{
"epoch": 0.4903072105131703,
"grad_norm": 0.6157532930374146,
"learning_rate": 2.921096154782469e-05,
"loss": 1.0352,
"step": 3190
},
{
"epoch": 0.49184422371227116,
"grad_norm": 0.4587867856025696,
"learning_rate": 2.869078807068526e-05,
"loss": 0.9127,
"step": 3200
},
{
"epoch": 0.493381236911372,
"grad_norm": 0.2948139011859894,
"learning_rate": 2.817451101157884e-05,
"loss": 1.2389,
"step": 3210
},
{
"epoch": 0.4949182501104728,
"grad_norm": 0.34423375129699707,
"learning_rate": 2.7662158580482656e-05,
"loss": 1.1716,
"step": 3220
},
{
"epoch": 0.4964552633095737,
"grad_norm": 0.4484558701515198,
"learning_rate": 2.7153758772927664e-05,
"loss": 1.1127,
"step": 3230
},
{
"epoch": 0.4979922765086745,
"grad_norm": 0.4787762463092804,
"learning_rate": 2.6649339368468886e-05,
"loss": 1.0071,
"step": 3240
},
{
"epoch": 0.4995292897077754,
"grad_norm": 0.5165082812309265,
"learning_rate": 2.6148927929167588e-05,
"loss": 0.8908,
"step": 3250
},
{
"epoch": 0.5010663029068763,
"grad_norm": 0.3385765850543976,
"learning_rate": 2.5652551798085032e-05,
"loss": 1.3068,
"step": 3260
},
{
"epoch": 0.5026033161059771,
"grad_norm": 0.29017236828804016,
"learning_rate": 2.5160238097788725e-05,
"loss": 1.182,
"step": 3270
},
{
"epoch": 0.5041403293050779,
"grad_norm": 0.5140998959541321,
"learning_rate": 2.4672013728870092e-05,
"loss": 1.1406,
"step": 3280
},
{
"epoch": 0.5056773425041787,
"grad_norm": 0.4207160472869873,
"learning_rate": 2.4187905368474863e-05,
"loss": 1.0186,
"step": 3290
},
{
"epoch": 0.5072143557032796,
"grad_norm": 0.6964996457099915,
"learning_rate": 2.370793946884522e-05,
"loss": 0.895,
"step": 3300
},
{
"epoch": 0.5087513689023805,
"grad_norm": 0.35131826996803284,
"learning_rate": 2.323214225587449e-05,
"loss": 1.3802,
"step": 3310
},
{
"epoch": 0.5102883821014813,
"grad_norm": 0.2624411880970001,
"learning_rate": 2.276053972767419e-05,
"loss": 1.1691,
"step": 3320
},
{
"epoch": 0.5118253953005821,
"grad_norm": 0.49757492542266846,
"learning_rate": 2.2293157653153297e-05,
"loss": 1.1201,
"step": 3330
},
{
"epoch": 0.513362408499683,
"grad_norm": 0.46858152747154236,
"learning_rate": 2.1830021570610416e-05,
"loss": 0.9846,
"step": 3340
},
{
"epoch": 0.5148994216987839,
"grad_norm": 0.9624618291854858,
"learning_rate": 2.137115678633811e-05,
"loss": 0.9536,
"step": 3350
},
{
"epoch": 0.5164364348978847,
"grad_norm": 0.32574358582496643,
"learning_rate": 2.0916588373240365e-05,
"loss": 1.2527,
"step": 3360
},
{
"epoch": 0.5179734480969855,
"grad_norm": 0.36960333585739136,
"learning_rate": 2.04663411694624e-05,
"loss": 1.2249,
"step": 3370
},
{
"epoch": 0.5195104612960864,
"grad_norm": 0.5395803451538086,
"learning_rate": 2.0020439777033506e-05,
"loss": 1.0808,
"step": 3380
},
{
"epoch": 0.5210474744951872,
"grad_norm": 0.5174748301506042,
"learning_rate": 1.957890856052288e-05,
"loss": 1.0051,
"step": 3390
},
{
"epoch": 0.5225844876942881,
"grad_norm": 0.45128193497657776,
"learning_rate": 1.914177164570825e-05,
"loss": 0.8491,
"step": 3400
},
{
"epoch": 0.5241215008933889,
"grad_norm": 0.32283082604408264,
"learning_rate": 1.870905291825753e-05,
"loss": 1.2458,
"step": 3410
},
{
"epoch": 0.5256585140924898,
"grad_norm": 0.34791356325149536,
"learning_rate": 1.828077602242376e-05,
"loss": 1.2013,
"step": 3420
},
{
"epoch": 0.5271955272915906,
"grad_norm": 0.3557356894016266,
"learning_rate": 1.7856964359753236e-05,
"loss": 1.0688,
"step": 3430
},
{
"epoch": 0.5287325404906915,
"grad_norm": 0.4580517113208771,
"learning_rate": 1.743764108780661e-05,
"loss": 1.0174,
"step": 3440
},
{
"epoch": 0.5302695536897923,
"grad_norm": 0.4637156128883362,
"learning_rate": 1.7022829118893802e-05,
"loss": 0.8894,
"step": 3450
},
{
"epoch": 0.5318065668888932,
"grad_norm": 0.3728739023208618,
"learning_rate": 1.661255111882174e-05,
"loss": 1.3953,
"step": 3460
},
{
"epoch": 0.533343580087994,
"grad_norm": 0.4086034893989563,
"learning_rate": 1.6206829505656208e-05,
"loss": 1.2781,
"step": 3470
},
{
"epoch": 0.5348805932870948,
"grad_norm": 0.32859915494918823,
"learning_rate": 1.5805686448496605e-05,
"loss": 1.0734,
"step": 3480
},
{
"epoch": 0.5364176064861957,
"grad_norm": 0.5218992233276367,
"learning_rate": 1.540914386626472e-05,
"loss": 1.066,
"step": 3490
},
{
"epoch": 0.5379546196852966,
"grad_norm": 0.5934432148933411,
"learning_rate": 1.5017223426507132e-05,
"loss": 0.8607,
"step": 3500
},
{
"epoch": 0.5394916328843974,
"grad_norm": 0.2949748933315277,
"learning_rate": 1.4629946544211182e-05,
"loss": 1.3593,
"step": 3510
},
{
"epoch": 0.5410286460834982,
"grad_norm": 0.4337247610092163,
"learning_rate": 1.4247334380634792e-05,
"loss": 1.2056,
"step": 3520
},
{
"epoch": 0.5425656592825991,
"grad_norm": 0.4211966097354889,
"learning_rate": 1.386940784215024e-05,
"loss": 1.0682,
"step": 3530
},
{
"epoch": 0.5441026724817,
"grad_norm": 0.48186883330345154,
"learning_rate": 1.3496187579101915e-05,
"loss": 1.04,
"step": 3540
},
{
"epoch": 0.5456396856808008,
"grad_norm": 0.5528340339660645,
"learning_rate": 1.3127693984677736e-05,
"loss": 0.9161,
"step": 3550
},
{
"epoch": 0.5471766988799016,
"grad_norm": 0.26369714736938477,
"learning_rate": 1.2763947193795067e-05,
"loss": 1.2259,
"step": 3560
},
{
"epoch": 0.5487137120790024,
"grad_norm": 0.40054404735565186,
"learning_rate": 1.240496708200033e-05,
"loss": 1.2161,
"step": 3570
},
{
"epoch": 0.5502507252781034,
"grad_norm": 0.5079723000526428,
"learning_rate": 1.2050773264383175e-05,
"loss": 1.1272,
"step": 3580
},
{
"epoch": 0.5517877384772042,
"grad_norm": 0.6676374077796936,
"learning_rate": 1.1701385094504513e-05,
"loss": 0.949,
"step": 3590
},
{
"epoch": 0.553324751676305,
"grad_norm": 0.7548049688339233,
"learning_rate": 1.135682166333909e-05,
"loss": 0.92,
"step": 3600
},
{
"epoch": 0.5548617648754058,
"grad_norm": 0.32646673917770386,
"learning_rate": 1.1017101798232388e-05,
"loss": 1.364,
"step": 3610
},
{
"epoch": 0.5563987780745067,
"grad_norm": 0.37211284041404724,
"learning_rate": 1.0682244061871748e-05,
"loss": 1.1881,
"step": 3620
},
{
"epoch": 0.5579357912736076,
"grad_norm": 0.3693225085735321,
"learning_rate": 1.0352266751272255e-05,
"loss": 1.08,
"step": 3630
},
{
"epoch": 0.5594728044727084,
"grad_norm": 0.5169116258621216,
"learning_rate": 1.0027187896776769e-05,
"loss": 0.9328,
"step": 3640
},
{
"epoch": 0.5610098176718092,
"grad_norm": 0.517723023891449,
"learning_rate": 9.707025261070912e-06,
"loss": 0.951,
"step": 3650
},
{
"epoch": 0.5625468308709101,
"grad_norm": 0.28930410742759705,
"learning_rate": 9.391796338212356e-06,
"loss": 1.3214,
"step": 3660
},
{
"epoch": 0.564083844070011,
"grad_norm": 0.3643380403518677,
"learning_rate": 9.081518352674967e-06,
"loss": 1.1138,
"step": 3670
},
{
"epoch": 0.5656208572691118,
"grad_norm": 0.5238456130027771,
"learning_rate": 8.776208258407725e-06,
"loss": 0.9906,
"step": 3680
},
{
"epoch": 0.5671578704682126,
"grad_norm": 0.46154043078422546,
"learning_rate": 8.475882737908248e-06,
"loss": 1.0237,
"step": 3690
},
{
"epoch": 0.5686948836673135,
"grad_norm": 0.5606771111488342,
"learning_rate": 8.180558201311216e-06,
"loss": 0.9836,
"step": 3700
},
{
"epoch": 0.5702318968664143,
"grad_norm": 0.2615475058555603,
"learning_rate": 7.890250785491771e-06,
"loss": 1.2654,
"step": 3710
},
{
"epoch": 0.5717689100655152,
"grad_norm": 0.3426634669303894,
"learning_rate": 7.60497635318379e-06,
"loss": 1.1903,
"step": 3720
},
{
"epoch": 0.573305923264616,
"grad_norm": 0.33861038088798523,
"learning_rate": 7.324750492113053e-06,
"loss": 1.0488,
"step": 3730
},
{
"epoch": 0.5748429364637169,
"grad_norm": 0.43294966220855713,
"learning_rate": 7.049588514145589e-06,
"loss": 1.1214,
"step": 3740
},
{
"epoch": 0.5763799496628177,
"grad_norm": 0.5775050520896912,
"learning_rate": 6.7795054544509296e-06,
"loss": 0.9913,
"step": 3750
},
{
"epoch": 0.5779169628619186,
"grad_norm": 0.30302977561950684,
"learning_rate": 9.419673244218923e-05,
"loss": 1.2611,
"step": 3760
},
{
"epoch": 0.5794539760610194,
"grad_norm": 0.32641565799713135,
"learning_rate": 9.376419544126368e-05,
"loss": 1.0785,
"step": 3770
},
{
"epoch": 0.5809909892601203,
"grad_norm": 0.39564621448516846,
"learning_rate": 9.333177552919009e-05,
"loss": 1.0693,
"step": 3780
},
{
"epoch": 0.5825280024592211,
"grad_norm": 0.5899733901023865,
"learning_rate": 9.289948082545823e-05,
"loss": 1.0189,
"step": 3790
},
{
"epoch": 0.5840650156583219,
"grad_norm": 0.6120781898498535,
"learning_rate": 9.246731944720675e-05,
"loss": 0.97,
"step": 3800
},
{
"epoch": 0.5856020288574229,
"grad_norm": 0.31813132762908936,
"learning_rate": 9.203529950907096e-05,
"loss": 1.2171,
"step": 3810
},
{
"epoch": 0.5871390420565237,
"grad_norm": 0.4008195698261261,
"learning_rate": 9.16034291230303e-05,
"loss": 1.1715,
"step": 3820
},
{
"epoch": 0.5886760552556245,
"grad_norm": 0.45466354489326477,
"learning_rate": 9.117171639825616e-05,
"loss": 1.0245,
"step": 3830
},
{
"epoch": 0.5902130684547253,
"grad_norm": 0.4874997138977051,
"learning_rate": 9.074016944095954e-05,
"loss": 1.0824,
"step": 3840
},
{
"epoch": 0.5917500816538263,
"grad_norm": 0.47545289993286133,
"learning_rate": 9.030879635423874e-05,
"loss": 0.9017,
"step": 3850
},
{
"epoch": 0.5932870948529271,
"grad_norm": 0.3246556520462036,
"learning_rate": 8.987760523792746e-05,
"loss": 1.4075,
"step": 3860
},
{
"epoch": 0.5948241080520279,
"grad_norm": 0.32509294152259827,
"learning_rate": 8.944660418844245e-05,
"loss": 1.2578,
"step": 3870
},
{
"epoch": 0.5963611212511287,
"grad_norm": 0.34634068608283997,
"learning_rate": 8.901580129863171e-05,
"loss": 1.1101,
"step": 3880
},
{
"epoch": 0.5978981344502295,
"grad_norm": 0.5410276651382446,
"learning_rate": 8.858520465762228e-05,
"loss": 1.0809,
"step": 3890
},
{
"epoch": 0.5994351476493305,
"grad_norm": 0.6559542417526245,
"learning_rate": 8.815482235066862e-05,
"loss": 0.9017,
"step": 3900
},
{
"epoch": 0.6009721608484313,
"grad_norm": 0.3090398609638214,
"learning_rate": 8.772466245900066e-05,
"loss": 1.3324,
"step": 3910
},
{
"epoch": 0.6025091740475321,
"grad_norm": 0.342724472284317,
"learning_rate": 8.729473305967197e-05,
"loss": 1.1054,
"step": 3920
},
{
"epoch": 0.604046187246633,
"grad_norm": 0.4747468829154968,
"learning_rate": 8.686504222540835e-05,
"loss": 1.0227,
"step": 3930
},
{
"epoch": 0.6055832004457339,
"grad_norm": 0.3615957796573639,
"learning_rate": 8.643559802445592e-05,
"loss": 1.0123,
"step": 3940
},
{
"epoch": 0.6071202136448347,
"grad_norm": 0.5801814198493958,
"learning_rate": 8.600640852042994e-05,
"loss": 0.9734,
"step": 3950
},
{
"epoch": 0.6086572268439355,
"grad_norm": 0.343477725982666,
"learning_rate": 8.557748177216316e-05,
"loss": 1.3499,
"step": 3960
},
{
"epoch": 0.6101942400430364,
"grad_norm": 0.3365076780319214,
"learning_rate": 8.514882583355468e-05,
"loss": 1.1898,
"step": 3970
},
{
"epoch": 0.6117312532421372,
"grad_norm": 0.4051002860069275,
"learning_rate": 8.472044875341866e-05,
"loss": 1.0753,
"step": 3980
},
{
"epoch": 0.6132682664412381,
"grad_norm": 0.5787518620491028,
"learning_rate": 8.429235857533303e-05,
"loss": 1.1869,
"step": 3990
},
{
"epoch": 0.6148052796403389,
"grad_norm": 0.6468502283096313,
"learning_rate": 8.386456333748878e-05,
"loss": 0.9503,
"step": 4000
},
{
"epoch": 0.6163422928394398,
"grad_norm": 0.30203506350517273,
"learning_rate": 8.34370710725387e-05,
"loss": 1.3229,
"step": 4010
},
{
"epoch": 0.6178793060385406,
"grad_norm": 0.3489297926425934,
"learning_rate": 8.300988980744682e-05,
"loss": 1.156,
"step": 4020
},
{
"epoch": 0.6194163192376414,
"grad_norm": 0.41906577348709106,
"learning_rate": 8.258302756333744e-05,
"loss": 1.1545,
"step": 4030
},
{
"epoch": 0.6209533324367423,
"grad_norm": 0.40264812111854553,
"learning_rate": 8.215649235534477e-05,
"loss": 1.1117,
"step": 4040
},
{
"epoch": 0.6224903456358432,
"grad_norm": 0.6303793787956238,
"learning_rate": 8.173029219246226e-05,
"loss": 0.918,
"step": 4050
},
{
"epoch": 0.624027358834944,
"grad_norm": 0.3673880398273468,
"learning_rate": 8.13044350773922e-05,
"loss": 1.3564,
"step": 4060
},
{
"epoch": 0.6255643720340448,
"grad_norm": 0.33907923102378845,
"learning_rate": 8.087892900639565e-05,
"loss": 1.215,
"step": 4070
},
{
"epoch": 0.6271013852331457,
"grad_norm": 0.5690033435821533,
"learning_rate": 8.045378196914202e-05,
"loss": 1.1178,
"step": 4080
},
{
"epoch": 0.6286383984322466,
"grad_norm": 0.43504247069358826,
"learning_rate": 8.002900194855932e-05,
"loss": 1.042,
"step": 4090
},
{
"epoch": 0.6301754116313474,
"grad_norm": 0.6302089095115662,
"learning_rate": 7.960459692068402e-05,
"loss": 0.8747,
"step": 4100
},
{
"epoch": 0.6317124248304482,
"grad_norm": 0.3127936124801636,
"learning_rate": 7.91805748545115e-05,
"loss": 1.279,
"step": 4110
},
{
"epoch": 0.633249438029549,
"grad_norm": 0.3348170816898346,
"learning_rate": 7.87569437118463e-05,
"loss": 1.1193,
"step": 4120
},
{
"epoch": 0.63478645122865,
"grad_norm": 0.4612022042274475,
"learning_rate": 7.833371144715257e-05,
"loss": 0.988,
"step": 4130
},
{
"epoch": 0.6363234644277508,
"grad_norm": 0.42834344506263733,
"learning_rate": 7.791088600740489e-05,
"loss": 1.0456,
"step": 4140
},
{
"epoch": 0.6378604776268516,
"grad_norm": 0.5120622515678406,
"learning_rate": 7.748847533193884e-05,
"loss": 0.8692,
"step": 4150
},
{
"epoch": 0.6393974908259524,
"grad_norm": 0.35314714908599854,
"learning_rate": 7.706648735230212e-05,
"loss": 1.3143,
"step": 4160
},
{
"epoch": 0.6409345040250534,
"grad_norm": 0.41545453667640686,
"learning_rate": 7.664492999210551e-05,
"loss": 1.1569,
"step": 4170
},
{
"epoch": 0.6424715172241542,
"grad_norm": 0.3695985972881317,
"learning_rate": 7.6223811166874e-05,
"loss": 1.1036,
"step": 4180
},
{
"epoch": 0.644008530423255,
"grad_norm": 0.6917260885238647,
"learning_rate": 7.580313878389847e-05,
"loss": 1.0497,
"step": 4190
},
{
"epoch": 0.6455455436223558,
"grad_norm": 0.567042887210846,
"learning_rate": 7.538292074208677e-05,
"loss": 0.9362,
"step": 4200
},
{
"epoch": 0.6470825568214567,
"grad_norm": 0.2965785562992096,
"learning_rate": 7.496316493181585e-05,
"loss": 1.4894,
"step": 4210
},
{
"epoch": 0.6486195700205576,
"grad_norm": 0.4221814274787903,
"learning_rate": 7.454387923478331e-05,
"loss": 1.1457,
"step": 4220
},
{
"epoch": 0.6501565832196584,
"grad_norm": 0.5893160104751587,
"learning_rate": 7.412507152385951e-05,
"loss": 1.2551,
"step": 4230
},
{
"epoch": 0.6516935964187592,
"grad_norm": 0.5759446024894714,
"learning_rate": 7.370674966293977e-05,
"loss": 1.0301,
"step": 4240
},
{
"epoch": 0.6532306096178601,
"grad_norm": 0.557947039604187,
"learning_rate": 7.32889215067966e-05,
"loss": 0.9201,
"step": 4250
},
{
"epoch": 0.654767622816961,
"grad_norm": 0.36502039432525635,
"learning_rate": 7.287159490093234e-05,
"loss": 1.3313,
"step": 4260
},
{
"epoch": 0.6563046360160618,
"grad_norm": 0.34715965390205383,
"learning_rate": 7.245477768143172e-05,
"loss": 1.1441,
"step": 4270
},
{
"epoch": 0.6578416492151626,
"grad_norm": 0.34881067276000977,
"learning_rate": 7.203847767481491e-05,
"loss": 1.2417,
"step": 4280
},
{
"epoch": 0.6593786624142635,
"grad_norm": 0.5187172889709473,
"learning_rate": 7.162270269789031e-05,
"loss": 1.0478,
"step": 4290
},
{
"epoch": 0.6609156756133643,
"grad_norm": 0.7966432571411133,
"learning_rate": 7.120746055760799e-05,
"loss": 0.9468,
"step": 4300
},
{
"epoch": 0.6624526888124652,
"grad_norm": 0.2697109878063202,
"learning_rate": 7.079275905091303e-05,
"loss": 1.304,
"step": 4310
},
{
"epoch": 0.663989702011566,
"grad_norm": 0.4051036834716797,
"learning_rate": 7.037860596459904e-05,
"loss": 1.197,
"step": 4320
},
{
"epoch": 0.6655267152106669,
"grad_norm": 0.34982064366340637,
"learning_rate": 6.996500907516213e-05,
"loss": 1.0326,
"step": 4330
},
{
"epoch": 0.6670637284097677,
"grad_norm": 0.48661354184150696,
"learning_rate": 6.955197614865457e-05,
"loss": 1.0212,
"step": 4340
},
{
"epoch": 0.6686007416088686,
"grad_norm": 0.504176139831543,
"learning_rate": 6.913951494053942e-05,
"loss": 0.8828,
"step": 4350
},
{
"epoch": 0.6701377548079694,
"grad_norm": 0.33689260482788086,
"learning_rate": 6.872763319554447e-05,
"loss": 1.4458,
"step": 4360
},
{
"epoch": 0.6716747680070703,
"grad_norm": 0.38656774163246155,
"learning_rate": 6.831633864751712e-05,
"loss": 1.1991,
"step": 4370
},
{
"epoch": 0.6732117812061711,
"grad_norm": 0.339083731174469,
"learning_rate": 6.790563901927907e-05,
"loss": 1.0625,
"step": 4380
},
{
"epoch": 0.6747487944052719,
"grad_norm": 0.40219244360923767,
"learning_rate": 6.749554202248116e-05,
"loss": 0.9733,
"step": 4390
},
{
"epoch": 0.6762858076043728,
"grad_norm": 0.9205519556999207,
"learning_rate": 6.708605535745881e-05,
"loss": 0.9578,
"step": 4400
},
{
"epoch": 0.6778228208034737,
"grad_norm": 0.29063740372657776,
"learning_rate": 6.667718671308729e-05,
"loss": 1.2317,
"step": 4410
},
{
"epoch": 0.6793598340025745,
"grad_norm": 0.35966232419013977,
"learning_rate": 6.626894376663745e-05,
"loss": 1.2436,
"step": 4420
},
{
"epoch": 0.6808968472016753,
"grad_norm": 0.476040780544281,
"learning_rate": 6.586133418363134e-05,
"loss": 1.1253,
"step": 4430
},
{
"epoch": 0.6824338604007761,
"grad_norm": 0.35419192910194397,
"learning_rate": 6.545436561769858e-05,
"loss": 0.9924,
"step": 4440
},
{
"epoch": 0.6839708735998771,
"grad_norm": 0.47478023171424866,
"learning_rate": 6.504804571043242e-05,
"loss": 0.9418,
"step": 4450
},
{
"epoch": 0.6855078867989779,
"grad_norm": 0.31886205077171326,
"learning_rate": 6.464238209124637e-05,
"loss": 1.3339,
"step": 4460
},
{
"epoch": 0.6870448999980787,
"grad_norm": 0.3391808867454529,
"learning_rate": 6.423738237723094e-05,
"loss": 1.1208,
"step": 4470
},
{
"epoch": 0.6885819131971795,
"grad_norm": 0.34816595911979675,
"learning_rate": 6.383305417301045e-05,
"loss": 1.0332,
"step": 4480
},
{
"epoch": 0.6901189263962805,
"grad_norm": 0.8269426822662354,
"learning_rate": 6.342940507060051e-05,
"loss": 0.9906,
"step": 4490
},
{
"epoch": 0.6916559395953813,
"grad_norm": 0.6578885912895203,
"learning_rate": 6.302644264926527e-05,
"loss": 1.0467,
"step": 4500
},
{
"epoch": 0.6931929527944821,
"grad_norm": 0.27568280696868896,
"learning_rate": 6.262417447537518e-05,
"loss": 1.2157,
"step": 4510
},
{
"epoch": 0.6947299659935829,
"grad_norm": 0.37618783116340637,
"learning_rate": 6.222260810226493e-05,
"loss": 1.1461,
"step": 4520
},
{
"epoch": 0.6962669791926838,
"grad_norm": 0.40639522671699524,
"learning_rate": 6.182175107009147e-05,
"loss": 1.1524,
"step": 4530
},
{
"epoch": 0.6978039923917847,
"grad_norm": 0.5950527787208557,
"learning_rate": 6.142161090569271e-05,
"loss": 0.97,
"step": 4540
},
{
"epoch": 0.6993410055908855,
"grad_norm": 0.6210660338401794,
"learning_rate": 6.1022195122445924e-05,
"loss": 0.9543,
"step": 4550
},
{
"epoch": 0.7008780187899863,
"grad_norm": 0.27577099204063416,
"learning_rate": 6.062351122012682e-05,
"loss": 1.3338,
"step": 4560
},
{
"epoch": 0.7024150319890872,
"grad_norm": 0.43212491273880005,
"learning_rate": 6.022556668476872e-05,
"loss": 1.1098,
"step": 4570
},
{
"epoch": 0.7039520451881881,
"grad_norm": 0.3773362934589386,
"learning_rate": 5.982836898852182e-05,
"loss": 1.1991,
"step": 4580
},
{
"epoch": 0.7054890583872889,
"grad_norm": 0.44134753942489624,
"learning_rate": 5.943192558951317e-05,
"loss": 1.0419,
"step": 4590
},
{
"epoch": 0.7070260715863897,
"grad_norm": 0.5980659127235413,
"learning_rate": 5.903624393170638e-05,
"loss": 0.8815,
"step": 4600
},
{
"epoch": 0.7085630847854906,
"grad_norm": 0.4459092319011688,
"learning_rate": 5.864133144476205e-05,
"loss": 1.3986,
"step": 4610
},
{
"epoch": 0.7101000979845914,
"grad_norm": 0.3655239939689636,
"learning_rate": 5.824719554389802e-05,
"loss": 1.1383,
"step": 4620
},
{
"epoch": 0.7116371111836923,
"grad_norm": 0.3964081108570099,
"learning_rate": 5.7853843629750405e-05,
"loss": 1.1092,
"step": 4630
},
{
"epoch": 0.7131741243827932,
"grad_norm": 0.542210042476654,
"learning_rate": 5.74612830882345e-05,
"loss": 0.9898,
"step": 4640
},
{
"epoch": 0.714711137581894,
"grad_norm": 0.5573174953460693,
"learning_rate": 5.706952129040599e-05,
"loss": 0.9413,
"step": 4650
},
{
"epoch": 0.7162481507809948,
"grad_norm": 0.39869463443756104,
"learning_rate": 5.667856559232281e-05,
"loss": 1.3621,
"step": 4660
},
{
"epoch": 0.7177851639800957,
"grad_norm": 0.427641898393631,
"learning_rate": 5.6288423334906735e-05,
"loss": 1.1562,
"step": 4670
},
{
"epoch": 0.7193221771791966,
"grad_norm": 0.6465682983398438,
"learning_rate": 5.5899101843805756e-05,
"loss": 1.0832,
"step": 4680
},
{
"epoch": 0.7208591903782974,
"grad_norm": 0.5046477913856506,
"learning_rate": 5.551060842925646e-05,
"loss": 1.1515,
"step": 4690
},
{
"epoch": 0.7223962035773982,
"grad_norm": 0.5882601737976074,
"learning_rate": 5.5122950385946684e-05,
"loss": 0.9227,
"step": 4700
},
{
"epoch": 0.723933216776499,
"grad_norm": 0.310811311006546,
"learning_rate": 5.473613499287872e-05,
"loss": 1.2842,
"step": 4710
},
{
"epoch": 0.7254702299756,
"grad_norm": 0.38150686025619507,
"learning_rate": 5.435016951323236e-05,
"loss": 1.2493,
"step": 4720
},
{
"epoch": 0.7270072431747008,
"grad_norm": 0.4792502522468567,
"learning_rate": 5.396506119422889e-05,
"loss": 1.0693,
"step": 4730
},
{
"epoch": 0.7285442563738016,
"grad_norm": 0.46154388785362244,
"learning_rate": 5.358081726699462e-05,
"loss": 1.0313,
"step": 4740
},
{
"epoch": 0.7300812695729024,
"grad_norm": 0.5081371068954468,
"learning_rate": 5.319744494642547e-05,
"loss": 0.9025,
"step": 4750
},
{
"epoch": 0.7316182827720034,
"grad_norm": 0.29481634497642517,
"learning_rate": 5.281495143105113e-05,
"loss": 1.3368,
"step": 4760
},
{
"epoch": 0.7331552959711042,
"grad_norm": 0.3294944167137146,
"learning_rate": 5.2433343902900255e-05,
"loss": 1.164,
"step": 4770
},
{
"epoch": 0.734692309170205,
"grad_norm": 0.523278534412384,
"learning_rate": 5.205262952736534e-05,
"loss": 1.0735,
"step": 4780
},
{
"epoch": 0.7362293223693058,
"grad_norm": 0.4263080060482025,
"learning_rate": 5.167281545306833e-05,
"loss": 1.138,
"step": 4790
},
{
"epoch": 0.7377663355684067,
"grad_norm": 0.450485497713089,
"learning_rate": 5.129390881172633e-05,
"loss": 0.8463,
"step": 4800
},
{
"epoch": 0.7393033487675076,
"grad_norm": 0.3693994879722595,
"learning_rate": 5.0915916718017634e-05,
"loss": 1.3698,
"step": 4810
},
{
"epoch": 0.7408403619666084,
"grad_norm": 0.508443295955658,
"learning_rate": 5.053884626944829e-05,
"loss": 1.1441,
"step": 4820
},
{
"epoch": 0.7423773751657092,
"grad_norm": 0.3785404562950134,
"learning_rate": 5.016270454621864e-05,
"loss": 1.1044,
"step": 4830
},
{
"epoch": 0.74391438836481,
"grad_norm": 0.6221136450767517,
"learning_rate": 4.978749861109056e-05,
"loss": 0.9437,
"step": 4840
},
{
"epoch": 0.7454514015639109,
"grad_norm": 0.6176669597625732,
"learning_rate": 4.941323550925481e-05,
"loss": 0.9218,
"step": 4850
},
{
"epoch": 0.7469884147630118,
"grad_norm": 0.3306792974472046,
"learning_rate": 4.90399222681985e-05,
"loss": 1.2402,
"step": 4860
},
{
"epoch": 0.7485254279621126,
"grad_norm": 0.4066197872161865,
"learning_rate": 4.866756589757353e-05,
"loss": 1.1279,
"step": 4870
},
{
"epoch": 0.7500624411612135,
"grad_norm": 0.4889521598815918,
"learning_rate": 4.829617338906468e-05,
"loss": 1.0597,
"step": 4880
},
{
"epoch": 0.7515994543603143,
"grad_norm": 0.514279842376709,
"learning_rate": 4.792575171625846e-05,
"loss": 0.9773,
"step": 4890
},
{
"epoch": 0.7531364675594152,
"grad_norm": 0.4709707796573639,
"learning_rate": 4.755630783451215e-05,
"loss": 0.9716,
"step": 4900
},
{
"epoch": 0.754673480758516,
"grad_norm": 0.2940506637096405,
"learning_rate": 4.7187848680823054e-05,
"loss": 1.3863,
"step": 4910
},
{
"epoch": 0.7562104939576169,
"grad_norm": 0.4398675858974457,
"learning_rate": 4.682038117369856e-05,
"loss": 1.1936,
"step": 4920
},
{
"epoch": 0.7577475071567177,
"grad_norm": 0.4706341326236725,
"learning_rate": 4.645391221302585e-05,
"loss": 0.9639,
"step": 4930
},
{
"epoch": 0.7592845203558185,
"grad_norm": 0.49938830733299255,
"learning_rate": 4.608844867994272e-05,
"loss": 0.9407,
"step": 4940
},
{
"epoch": 0.7608215335549194,
"grad_norm": 0.5872446298599243,
"learning_rate": 4.572399743670802e-05,
"loss": 0.9904,
"step": 4950
},
{
"epoch": 0.7623585467540203,
"grad_norm": 0.3020681142807007,
"learning_rate": 4.5360565326573104e-05,
"loss": 1.3422,
"step": 4960
},
{
"epoch": 0.7638955599531211,
"grad_norm": 0.31915101408958435,
"learning_rate": 4.499815917365313e-05,
"loss": 1.1206,
"step": 4970
},
{
"epoch": 0.7654325731522219,
"grad_norm": 0.3373018205165863,
"learning_rate": 4.463678578279903e-05,
"loss": 0.9979,
"step": 4980
},
{
"epoch": 0.7669695863513228,
"grad_norm": 0.556707501411438,
"learning_rate": 4.427645193946972e-05,
"loss": 1.0649,
"step": 4990
},
{
"epoch": 0.7685065995504237,
"grad_norm": 0.48355329036712646,
"learning_rate": 4.3917164409604596e-05,
"loss": 0.8827,
"step": 5000
},
{
"epoch": 0.7700436127495245,
"grad_norm": 0.3347935974597931,
"learning_rate": 4.355892993949666e-05,
"loss": 1.3363,
"step": 5010
},
{
"epoch": 0.7715806259486253,
"grad_norm": 0.4474698305130005,
"learning_rate": 4.320175525566567e-05,
"loss": 1.1031,
"step": 5020
},
{
"epoch": 0.7731176391477261,
"grad_norm": 0.4563886821269989,
"learning_rate": 4.2845647064731976e-05,
"loss": 1.0726,
"step": 5030
},
{
"epoch": 0.7746546523468271,
"grad_norm": 0.38811904191970825,
"learning_rate": 4.2490612053290565e-05,
"loss": 0.9596,
"step": 5040
},
{
"epoch": 0.7761916655459279,
"grad_norm": 0.6792904138565063,
"learning_rate": 4.213665688778539e-05,
"loss": 0.8853,
"step": 5050
},
{
"epoch": 0.7777286787450287,
"grad_norm": 0.3306553065776825,
"learning_rate": 4.1783788214384376e-05,
"loss": 1.3408,
"step": 5060
},
{
"epoch": 0.7792656919441295,
"grad_norm": 0.3388022482395172,
"learning_rate": 4.1432012658854405e-05,
"loss": 1.1149,
"step": 5070
},
{
"epoch": 0.7808027051432305,
"grad_norm": 0.3981464207172394,
"learning_rate": 4.1081336826437254e-05,
"loss": 1.0999,
"step": 5080
},
{
"epoch": 0.7823397183423313,
"grad_norm": 0.530524492263794,
"learning_rate": 4.07317673017251e-05,
"loss": 1.0388,
"step": 5090
},
{
"epoch": 0.7838767315414321,
"grad_norm": 0.5063654184341431,
"learning_rate": 4.0383310648537285e-05,
"loss": 0.9344,
"step": 5100
},
{
"epoch": 0.7854137447405329,
"grad_norm": 0.35088545083999634,
"learning_rate": 4.0035973409796925e-05,
"loss": 1.3162,
"step": 5110
},
{
"epoch": 0.7869507579396338,
"grad_norm": 0.3408149480819702,
"learning_rate": 3.9689762107407915e-05,
"loss": 1.1665,
"step": 5120
},
{
"epoch": 0.7884877711387347,
"grad_norm": 0.410655677318573,
"learning_rate": 3.9344683242132726e-05,
"loss": 1.1202,
"step": 5130
},
{
"epoch": 0.7900247843378355,
"grad_norm": 0.4364660084247589,
"learning_rate": 3.900074329347012e-05,
"loss": 0.9835,
"step": 5140
},
{
"epoch": 0.7915617975369363,
"grad_norm": 0.5834974646568298,
"learning_rate": 3.865794871953368e-05,
"loss": 0.9289,
"step": 5150
},
{
"epoch": 0.7930988107360372,
"grad_norm": 0.3499378561973572,
"learning_rate": 3.8316305956930286e-05,
"loss": 1.399,
"step": 5160
},
{
"epoch": 0.7946358239351381,
"grad_norm": 0.3590654730796814,
"learning_rate": 3.797582142063958e-05,
"loss": 1.2532,
"step": 5170
},
{
"epoch": 0.7961728371342389,
"grad_norm": 0.3866276741027832,
"learning_rate": 3.763650150389326e-05,
"loss": 1.1097,
"step": 5180
},
{
"epoch": 0.7977098503333397,
"grad_norm": 0.5099689364433289,
"learning_rate": 3.729835257805514e-05,
"loss": 0.9948,
"step": 5190
},
{
"epoch": 0.7992468635324406,
"grad_norm": 0.8499717116355896,
"learning_rate": 3.696138099250154e-05,
"loss": 1.0089,
"step": 5200
},
{
"epoch": 0.8007838767315414,
"grad_norm": 0.3350473940372467,
"learning_rate": 3.662559307450187e-05,
"loss": 1.3356,
"step": 5210
},
{
"epoch": 0.8023208899306423,
"grad_norm": 0.37701916694641113,
"learning_rate": 3.62909951291002e-05,
"loss": 1.1198,
"step": 5220
},
{
"epoch": 0.8038579031297431,
"grad_norm": 0.4176217317581177,
"learning_rate": 3.595759343899643e-05,
"loss": 1.0938,
"step": 5230
},
{
"epoch": 0.805394916328844,
"grad_norm": 0.3889678120613098,
"learning_rate": 3.562539426442868e-05,
"loss": 1.0096,
"step": 5240
},
{
"epoch": 0.8069319295279448,
"grad_norm": 0.5030645728111267,
"learning_rate": 3.52944038430556e-05,
"loss": 0.891,
"step": 5250
},
{
"epoch": 0.8084689427270456,
"grad_norm": 0.30685076117515564,
"learning_rate": 3.496462838983912e-05,
"loss": 1.2576,
"step": 5260
},
{
"epoch": 0.8100059559261465,
"grad_norm": 0.34047237038612366,
"learning_rate": 3.4636074096927994e-05,
"loss": 1.1709,
"step": 5270
},
{
"epoch": 0.8115429691252474,
"grad_norm": 0.44897961616516113,
"learning_rate": 3.4308747133541386e-05,
"loss": 1.076,
"step": 5280
},
{
"epoch": 0.8130799823243482,
"grad_norm": 0.49718570709228516,
"learning_rate": 3.398265364585307e-05,
"loss": 1.1013,
"step": 5290
},
{
"epoch": 0.814616995523449,
"grad_norm": 0.5700457096099854,
"learning_rate": 3.3657799756876026e-05,
"loss": 0.85,
"step": 5300
},
{
"epoch": 0.81615400872255,
"grad_norm": 0.2544783055782318,
"learning_rate": 3.333419156634737e-05,
"loss": 1.2266,
"step": 5310
},
{
"epoch": 0.8176910219216508,
"grad_norm": 0.34246590733528137,
"learning_rate": 3.3011835150614036e-05,
"loss": 1.2871,
"step": 5320
},
{
"epoch": 0.8192280351207516,
"grad_norm": 0.35369962453842163,
"learning_rate": 3.269073656251842e-05,
"loss": 1.0619,
"step": 5330
},
{
"epoch": 0.8207650483198524,
"grad_norm": 0.6653052568435669,
"learning_rate": 3.237090183128501e-05,
"loss": 1.0234,
"step": 5340
},
{
"epoch": 0.8223020615189532,
"grad_norm": 0.5207810401916504,
"learning_rate": 3.205233696240686e-05,
"loss": 0.9258,
"step": 5350
},
{
"epoch": 0.8238390747180542,
"grad_norm": 0.3241281807422638,
"learning_rate": 3.1735047937533134e-05,
"loss": 1.3819,
"step": 5360
},
{
"epoch": 0.825376087917155,
"grad_norm": 0.3154536485671997,
"learning_rate": 3.1419040714356574e-05,
"loss": 1.1996,
"step": 5370
},
{
"epoch": 0.8269131011162558,
"grad_norm": 0.8289194703102112,
"learning_rate": 3.110432122650172e-05,
"loss": 1.0843,
"step": 5380
},
{
"epoch": 0.8284501143153566,
"grad_norm": 0.4755181670188904,
"learning_rate": 3.079089538341353e-05,
"loss": 0.9432,
"step": 5390
},
{
"epoch": 0.8299871275144576,
"grad_norm": 0.6212325692176819,
"learning_rate": 3.047876907024626e-05,
"loss": 0.9162,
"step": 5400
},
{
"epoch": 0.8315241407135584,
"grad_norm": 0.30075520277023315,
"learning_rate": 3.016794814775319e-05,
"loss": 1.3175,
"step": 5410
},
{
"epoch": 0.8330611539126592,
"grad_norm": 0.38691189885139465,
"learning_rate": 2.985843845217634e-05,
"loss": 1.1845,
"step": 5420
},
{
"epoch": 0.83459816711176,
"grad_norm": 0.4434199035167694,
"learning_rate": 2.95502457951371e-05,
"loss": 1.1708,
"step": 5430
},
{
"epoch": 0.8361351803108609,
"grad_norm": 0.4436211585998535,
"learning_rate": 2.9243375963526998e-05,
"loss": 1.021,
"step": 5440
},
{
"epoch": 0.8376721935099618,
"grad_norm": 0.4411422908306122,
"learning_rate": 2.893783471939896e-05,
"loss": 0.8866,
"step": 5450
},
{
"epoch": 0.8392092067090626,
"grad_norm": 0.43314051628112793,
"learning_rate": 2.8633627799859297e-05,
"loss": 1.3161,
"step": 5460
},
{
"epoch": 0.8407462199081635,
"grad_norm": 0.42803534865379333,
"learning_rate": 2.8330760916959864e-05,
"loss": 1.2033,
"step": 5470
},
{
"epoch": 0.8422832331072643,
"grad_norm": 0.2930775582790375,
"learning_rate": 2.802923975759085e-05,
"loss": 1.0419,
"step": 5480
},
{
"epoch": 0.8438202463063652,
"grad_norm": 0.42763224244117737,
"learning_rate": 2.7729069983373902e-05,
"loss": 1.1682,
"step": 5490
},
{
"epoch": 0.845357259505466,
"grad_norm": 0.5239199995994568,
"learning_rate": 2.7430257230555954e-05,
"loss": 0.8944,
"step": 5500
},
{
"epoch": 0.8468942727045669,
"grad_norm": 0.4464668929576874,
"learning_rate": 2.713280710990338e-05,
"loss": 1.3969,
"step": 5510
},
{
"epoch": 0.8484312859036677,
"grad_norm": 0.37915849685668945,
"learning_rate": 2.6836725206596457e-05,
"loss": 1.1227,
"step": 5520
},
{
"epoch": 0.8499682991027685,
"grad_norm": 0.4141576886177063,
"learning_rate": 2.654201708012479e-05,
"loss": 1.0685,
"step": 5530
},
{
"epoch": 0.8515053123018694,
"grad_norm": 0.45722368359565735,
"learning_rate": 2.624868826418262e-05,
"loss": 1.0001,
"step": 5540
},
{
"epoch": 0.8530423255009703,
"grad_norm": 0.545322060585022,
"learning_rate": 2.595674426656519e-05,
"loss": 0.9519,
"step": 5550
},
{
"epoch": 0.8545793387000711,
"grad_norm": 0.3177925646305084,
"learning_rate": 2.5666190569065073e-05,
"loss": 1.3333,
"step": 5560
},
{
"epoch": 0.8561163518991719,
"grad_norm": 0.32911601662635803,
"learning_rate": 2.537703262736947e-05,
"loss": 1.1953,
"step": 5570
},
{
"epoch": 0.8576533650982728,
"grad_norm": 0.4170762598514557,
"learning_rate": 2.508927587095762e-05,
"loss": 1.073,
"step": 5580
},
{
"epoch": 0.8591903782973737,
"grad_norm": 0.3976740837097168,
"learning_rate": 2.4802925702998902e-05,
"loss": 0.9053,
"step": 5590
},
{
"epoch": 0.8607273914964745,
"grad_norm": 0.4783420264720917,
"learning_rate": 2.4517987500251417e-05,
"loss": 0.9275,
"step": 5600
},
{
"epoch": 0.8622644046955753,
"grad_norm": 0.34218695759773254,
"learning_rate": 2.423446661296087e-05,
"loss": 1.3844,
"step": 5610
},
{
"epoch": 0.8638014178946761,
"grad_norm": 0.3341657221317291,
"learning_rate": 2.395236836476038e-05,
"loss": 1.2727,
"step": 5620
},
{
"epoch": 0.8653384310937771,
"grad_norm": 0.37647122144699097,
"learning_rate": 2.3671698052570223e-05,
"loss": 1.0715,
"step": 5630
},
{
"epoch": 0.8668754442928779,
"grad_norm": 0.47447216510772705,
"learning_rate": 2.3392460946498617e-05,
"loss": 0.9315,
"step": 5640
},
{
"epoch": 0.8684124574919787,
"grad_norm": 0.5783852934837341,
"learning_rate": 2.311466228974266e-05,
"loss": 0.885,
"step": 5650
},
{
"epoch": 0.8699494706910795,
"grad_norm": 0.3584993779659271,
"learning_rate": 2.283830729848978e-05,
"loss": 1.326,
"step": 5660
},
{
"epoch": 0.8714864838901804,
"grad_norm": 0.35056164860725403,
"learning_rate": 2.2563401161820018e-05,
"loss": 1.1802,
"step": 5670
},
{
"epoch": 0.8730234970892813,
"grad_norm": 0.3188401162624359,
"learning_rate": 2.2289949041608415e-05,
"loss": 1.011,
"step": 5680
},
{
"epoch": 0.8745605102883821,
"grad_norm": 0.4745340347290039,
"learning_rate": 2.2017956072428125e-05,
"loss": 1.043,
"step": 5690
},
{
"epoch": 0.8760975234874829,
"grad_norm": 0.5221130847930908,
"learning_rate": 2.174742736145411e-05,
"loss": 0.9263,
"step": 5700
},
{
"epoch": 0.8776345366865838,
"grad_norm": 0.37105482816696167,
"learning_rate": 2.1478367988367e-05,
"loss": 1.4659,
"step": 5710
},
{
"epoch": 0.8791715498856847,
"grad_norm": 0.3961479961872101,
"learning_rate": 2.121078300525803e-05,
"loss": 1.0952,
"step": 5720
},
{
"epoch": 0.8807085630847855,
"grad_norm": 0.5004900693893433,
"learning_rate": 2.094467743653389e-05,
"loss": 1.1854,
"step": 5730
},
{
"epoch": 0.8822455762838863,
"grad_norm": 0.4058009386062622,
"learning_rate": 2.0680056278822614e-05,
"loss": 0.9656,
"step": 5740
},
{
"epoch": 0.8837825894829872,
"grad_norm": 0.5562613606452942,
"learning_rate": 2.0416924500879543e-05,
"loss": 0.92,
"step": 5750
},
{
"epoch": 0.885319602682088,
"grad_norm": 0.31459611654281616,
"learning_rate": 2.015528704349422e-05,
"loss": 1.4171,
"step": 5760
},
{
"epoch": 0.8868566158811889,
"grad_norm": 0.36349666118621826,
"learning_rate": 1.9895148819397525e-05,
"loss": 1.111,
"step": 5770
},
{
"epoch": 0.8883936290802897,
"grad_norm": 0.37721896171569824,
"learning_rate": 1.9636514713169398e-05,
"loss": 1.194,
"step": 5780
},
{
"epoch": 0.8899306422793906,
"grad_norm": 0.5621936321258545,
"learning_rate": 1.9379389581147224e-05,
"loss": 1.0126,
"step": 5790
},
{
"epoch": 0.8914676554784914,
"grad_norm": 0.48834091424942017,
"learning_rate": 1.91237782513345e-05,
"loss": 0.9043,
"step": 5800
},
{
"epoch": 0.8930046686775923,
"grad_norm": 0.29431217908859253,
"learning_rate": 1.8869685523310364e-05,
"loss": 1.3211,
"step": 5810
},
{
"epoch": 0.8945416818766931,
"grad_norm": 0.36356300115585327,
"learning_rate": 1.861711616813928e-05,
"loss": 1.1987,
"step": 5820
},
{
"epoch": 0.896078695075794,
"grad_norm": 0.4829373359680176,
"learning_rate": 1.8366074928281607e-05,
"loss": 1.0968,
"step": 5830
},
{
"epoch": 0.8976157082748948,
"grad_norm": 0.506184458732605,
"learning_rate": 1.811656651750451e-05,
"loss": 1.0478,
"step": 5840
},
{
"epoch": 0.8991527214739956,
"grad_norm": 0.742889940738678,
"learning_rate": 1.7868595620793337e-05,
"loss": 1.0448,
"step": 5850
},
{
"epoch": 0.9006897346730965,
"grad_norm": 0.30385154485702515,
"learning_rate": 1.762216689426386e-05,
"loss": 1.2941,
"step": 5860
},
{
"epoch": 0.9022267478721974,
"grad_norm": 0.3338213264942169,
"learning_rate": 1.7377284965074635e-05,
"loss": 1.2072,
"step": 5870
},
{
"epoch": 0.9037637610712982,
"grad_norm": 0.3148481547832489,
"learning_rate": 1.7133954431340317e-05,
"loss": 1.0213,
"step": 5880
},
{
"epoch": 0.905300774270399,
"grad_norm": 0.4497423768043518,
"learning_rate": 1.68921798620451e-05,
"loss": 1.0517,
"step": 5890
},
{
"epoch": 0.9068377874695,
"grad_norm": 0.5314846038818359,
"learning_rate": 1.6651965796957125e-05,
"loss": 0.8729,
"step": 5900
},
{
"epoch": 0.9083748006686008,
"grad_norm": 0.3650830090045929,
"learning_rate": 1.6413316746543162e-05,
"loss": 1.3745,
"step": 5910
},
{
"epoch": 0.9099118138677016,
"grad_norm": 0.35135117173194885,
"learning_rate": 1.617623719188386e-05,
"loss": 1.1486,
"step": 5920
},
{
"epoch": 0.9114488270668024,
"grad_norm": 0.4427955448627472,
"learning_rate": 1.594073158458973e-05,
"loss": 0.9874,
"step": 5930
},
{
"epoch": 0.9129858402659032,
"grad_norm": 0.41412049531936646,
"learning_rate": 1.5706804346717397e-05,
"loss": 1.0228,
"step": 5940
},
{
"epoch": 0.9145228534650042,
"grad_norm": 0.4491603374481201,
"learning_rate": 1.5474459870686742e-05,
"loss": 0.8531,
"step": 5950
},
{
"epoch": 0.916059866664105,
"grad_norm": 0.3620060384273529,
"learning_rate": 1.5243702519198311e-05,
"loss": 1.4768,
"step": 5960
},
{
"epoch": 0.9175968798632058,
"grad_norm": 0.44465261697769165,
"learning_rate": 1.501453662515142e-05,
"loss": 1.1917,
"step": 5970
},
{
"epoch": 0.9191338930623066,
"grad_norm": 0.34438657760620117,
"learning_rate": 1.478696649156286e-05,
"loss": 1.1452,
"step": 5980
},
{
"epoch": 0.9206709062614076,
"grad_norm": 0.6627624034881592,
"learning_rate": 1.4560996391485937e-05,
"loss": 0.9758,
"step": 5990
},
{
"epoch": 0.9222079194605084,
"grad_norm": 0.6186857223510742,
"learning_rate": 1.4336630567930465e-05,
"loss": 0.886,
"step": 6000
},
{
"epoch": 0.9237449326596092,
"grad_norm": 0.34359419345855713,
"learning_rate": 1.4113873233782859e-05,
"loss": 1.2881,
"step": 6010
},
{
"epoch": 0.92528194585871,
"grad_norm": 0.3291650116443634,
"learning_rate": 1.3892728571727276e-05,
"loss": 1.138,
"step": 6020
},
{
"epoch": 0.9268189590578109,
"grad_norm": 0.36718741059303284,
"learning_rate": 1.3673200734166835e-05,
"loss": 1.1353,
"step": 6030
},
{
"epoch": 0.9283559722569118,
"grad_norm": 0.45962631702423096,
"learning_rate": 1.3477011472803913e-05,
"loss": 1.0381,
"step": 6040
},
{
"epoch": 0.9298929854560126,
"grad_norm": 0.7081500887870789,
"learning_rate": 1.3260566932862262e-05,
"loss": 0.935,
"step": 6050
},
{
"epoch": 0.9314299986551134,
"grad_norm": 0.2997978627681732,
"learning_rate": 1.3045751087428648e-05,
"loss": 1.2845,
"step": 6060
},
{
"epoch": 0.9329670118542143,
"grad_norm": 0.34809163212776184,
"learning_rate": 1.2832567970070942e-05,
"loss": 1.1698,
"step": 6070
},
{
"epoch": 0.9345040250533151,
"grad_norm": 0.4251902997493744,
"learning_rate": 1.2621021583699476e-05,
"loss": 1.0401,
"step": 6080
},
{
"epoch": 0.936041038252416,
"grad_norm": 0.5546366572380066,
"learning_rate": 1.2411115900491865e-05,
"loss": 1.0531,
"step": 6090
},
{
"epoch": 0.9375780514515168,
"grad_norm": 0.535227358341217,
"learning_rate": 1.2202854861818558e-05,
"loss": 0.874,
"step": 6100
},
{
"epoch": 0.9391150646506177,
"grad_norm": 0.31744185090065,
"learning_rate": 1.199624237816862e-05,
"loss": 1.3501,
"step": 6110
},
{
"epoch": 0.9406520778497185,
"grad_norm": 0.30914106965065,
"learning_rate": 1.1791282329076526e-05,
"loss": 1.2025,
"step": 6120
},
{
"epoch": 0.9421890910488194,
"grad_norm": 0.522088348865509,
"learning_rate": 1.1587978563049163e-05,
"loss": 1.0336,
"step": 6130
},
{
"epoch": 0.9437261042479202,
"grad_norm": 0.3826617896556854,
"learning_rate": 1.1386334897493633e-05,
"loss": 1.0287,
"step": 6140
},
{
"epoch": 0.9452631174470211,
"grad_norm": 0.5151504874229431,
"learning_rate": 1.1186355118645554e-05,
"loss": 0.9354,
"step": 6150
},
{
"epoch": 0.9468001306461219,
"grad_norm": 0.3254314064979553,
"learning_rate": 1.0988042981497947e-05,
"loss": 1.2842,
"step": 6160
},
{
"epoch": 0.9483371438452227,
"grad_norm": 0.33801066875457764,
"learning_rate": 1.0791402209730794e-05,
"loss": 1.2039,
"step": 6170
},
{
"epoch": 0.9498741570443237,
"grad_norm": 0.4988054931163788,
"learning_rate": 1.0596436495641027e-05,
"loss": 1.0537,
"step": 6180
},
{
"epoch": 0.9514111702434245,
"grad_norm": 0.607498288154602,
"learning_rate": 1.0403149500073295e-05,
"loss": 1.0255,
"step": 6190
},
{
"epoch": 0.9529481834425253,
"grad_norm": 0.42610228061676025,
"learning_rate": 1.0211544852351185e-05,
"loss": 0.8488,
"step": 6200
},
{
"epoch": 0.9544851966416261,
"grad_norm": 0.3109932541847229,
"learning_rate": 1.0021626150208984e-05,
"loss": 1.4041,
"step": 6210
},
{
"epoch": 0.956022209840727,
"grad_norm": 0.3864147961139679,
"learning_rate": 9.833396959724307e-06,
"loss": 1.0973,
"step": 6220
},
{
"epoch": 0.9575592230398279,
"grad_norm": 0.39751794934272766,
"learning_rate": 9.64686081525098e-06,
"loss": 1.0633,
"step": 6230
},
{
"epoch": 0.9590962362389287,
"grad_norm": 0.3791200518608093,
"learning_rate": 9.462021219352801e-06,
"loss": 0.9775,
"step": 6240
},
{
"epoch": 0.9606332494380295,
"grad_norm": 0.7502495646476746,
"learning_rate": 9.27888164273759e-06,
"loss": 0.89,
"step": 6250
},
{
"epoch": 0.9621702626371303,
"grad_norm": 0.35471296310424805,
"learning_rate": 9.09744552419225e-06,
"loss": 1.3647,
"step": 6260
},
{
"epoch": 0.9637072758362313,
"grad_norm": 0.3802105188369751,
"learning_rate": 8.91771627051805e-06,
"loss": 1.214,
"step": 6270
},
{
"epoch": 0.9652442890353321,
"grad_norm": 0.4803863763809204,
"learning_rate": 8.739697256466639e-06,
"loss": 1.1009,
"step": 6280
},
{
"epoch": 0.9667813022344329,
"grad_norm": 0.5591475963592529,
"learning_rate": 8.563391824676814e-06,
"loss": 0.9481,
"step": 6290
},
{
"epoch": 0.9683183154335337,
"grad_norm": 0.6896303296089172,
"learning_rate": 8.388803285611602e-06,
"loss": 0.9008,
"step": 6300
},
{
"epoch": 0.9698553286326347,
"grad_norm": 0.3024981617927551,
"learning_rate": 8.215934917496193e-06,
"loss": 1.3053,
"step": 6310
},
{
"epoch": 0.9713923418317355,
"grad_norm": 0.3303580582141876,
"learning_rate": 8.044789966256383e-06,
"loss": 1.1818,
"step": 6320
},
{
"epoch": 0.9729293550308363,
"grad_norm": 0.5090763568878174,
"learning_rate": 7.875371645457574e-06,
"loss": 1.1062,
"step": 6330
},
{
"epoch": 0.9744663682299372,
"grad_norm": 0.4125332534313202,
"learning_rate": 7.707683136244503e-06,
"loss": 1.0009,
"step": 6340
},
{
"epoch": 0.976003381429038,
"grad_norm": 0.5410733222961426,
"learning_rate": 7.5417275872814e-06,
"loss": 0.8963,
"step": 6350
},
{
"epoch": 0.9775403946281389,
"grad_norm": 0.29235604405403137,
"learning_rate": 7.377508114693022e-06,
"loss": 1.2773,
"step": 6360
},
{
"epoch": 0.9790774078272397,
"grad_norm": 0.33701273798942566,
"learning_rate": 7.21502780200598e-06,
"loss": 1.2048,
"step": 6370
},
{
"epoch": 0.9806144210263406,
"grad_norm": 0.4447634816169739,
"learning_rate": 7.054289700090988e-06,
"loss": 1.0046,
"step": 6380
},
{
"epoch": 0.9821514342254414,
"grad_norm": 0.457303911447525,
"learning_rate": 6.895296827105424e-06,
"loss": 1.0234,
"step": 6390
},
{
"epoch": 0.9836884474245423,
"grad_norm": 0.6247261762619019,
"learning_rate": 6.738052168436815e-06,
"loss": 0.8845,
"step": 6400
},
{
"epoch": 0.9852254606236431,
"grad_norm": 0.32420042157173157,
"learning_rate": 6.582558676646677e-06,
"loss": 1.2638,
"step": 6410
},
{
"epoch": 0.986762473822744,
"grad_norm": 0.4311647415161133,
"learning_rate": 6.428819271415098e-06,
"loss": 1.1503,
"step": 6420
},
{
"epoch": 0.9882994870218448,
"grad_norm": 0.3983921408653259,
"learning_rate": 6.276836839485945e-06,
"loss": 1.0527,
"step": 6430
},
{
"epoch": 0.9898365002209456,
"grad_norm": 0.4446724057197571,
"learning_rate": 6.126614234612593e-06,
"loss": 0.9741,
"step": 6440
},
{
"epoch": 0.9913735134200465,
"grad_norm": 0.6508687734603882,
"learning_rate": 5.978154277504433e-06,
"loss": 0.8693,
"step": 6450
},
{
"epoch": 0.9929105266191474,
"grad_norm": 0.4542991518974304,
"learning_rate": 5.831459755773816e-06,
"loss": 1.2544,
"step": 6460
},
{
"epoch": 0.9944475398182482,
"grad_norm": 0.3671010434627533,
"learning_rate": 5.6865334238837886e-06,
"loss": 1.137,
"step": 6470
},
{
"epoch": 0.995984553017349,
"grad_norm": 0.3959558606147766,
"learning_rate": 5.543378003096344e-06,
"loss": 0.9852,
"step": 6480
},
{
"epoch": 0.9975215662164498,
"grad_norm": 0.4711992144584656,
"learning_rate": 5.401996181421254e-06,
"loss": 0.9684,
"step": 6490
},
{
"epoch": 0.9990585794155508,
"grad_norm": 0.5218202471733093,
"learning_rate": 5.262390613565737e-06,
"loss": 0.8946,
"step": 6500
},
{
"epoch": 1.0004611039597302,
"grad_norm": 0.33302509784698486,
"learning_rate": 5.124563920884495e-06,
"loss": 1.2683,
"step": 6510
},
{
"epoch": 1.001998117158831,
"grad_norm": 0.36543136835098267,
"learning_rate": 4.988518691330579e-06,
"loss": 1.3433,
"step": 6520
},
{
"epoch": 1.0035351303579318,
"grad_norm": 0.3370875120162964,
"learning_rate": 4.8542574794066544e-06,
"loss": 1.0514,
"step": 6530
},
{
"epoch": 1.005072143557033,
"grad_norm": 0.3809349238872528,
"learning_rate": 4.721782806117236e-06,
"loss": 1.0427,
"step": 6540
},
{
"epoch": 1.0066091567561337,
"grad_norm": 0.3858659267425537,
"learning_rate": 4.591097158921198e-06,
"loss": 0.8425,
"step": 6550
},
{
"epoch": 1.0081461699552345,
"grad_norm": 0.2548363208770752,
"learning_rate": 4.4622029916850936e-06,
"loss": 0.9544,
"step": 6560
},
{
"epoch": 1.0096831831543354,
"grad_norm": 0.35040903091430664,
"learning_rate": 4.335102724637163e-06,
"loss": 1.2002,
"step": 6570
},
{
"epoch": 1.0112201963534362,
"grad_norm": 0.41010960936546326,
"learning_rate": 4.209798744321758e-06,
"loss": 1.029,
"step": 6580
},
{
"epoch": 1.012757209552537,
"grad_norm": 0.43580350279808044,
"learning_rate": 4.0862934035546415e-06,
"loss": 1.0397,
"step": 6590
},
{
"epoch": 1.0142942227516378,
"grad_norm": 0.5700265765190125,
"learning_rate": 3.964589021378773e-06,
"loss": 0.9584,
"step": 6600
},
{
"epoch": 1.0158312359507387,
"grad_norm": 0.2861765921115875,
"learning_rate": 3.844687883020725e-06,
"loss": 0.9829,
"step": 6610
},
{
"epoch": 1.0173682491498395,
"grad_norm": 0.34072452783584595,
"learning_rate": 3.7265922398478257e-06,
"loss": 1.3269,
"step": 6620
},
{
"epoch": 1.0189052623489405,
"grad_norm": 0.3608275055885315,
"learning_rate": 3.6103043093258627e-06,
"loss": 1.0098,
"step": 6630
},
{
"epoch": 1.0204422755480413,
"grad_norm": 0.36581897735595703,
"learning_rate": 3.4958262749774873e-06,
"loss": 0.9918,
"step": 6640
},
{
"epoch": 1.0219792887471422,
"grad_norm": 0.4597557485103607,
"learning_rate": 3.3831602863410915e-06,
"loss": 0.9656,
"step": 6650
},
{
"epoch": 1.023516301946243,
"grad_norm": 0.3232296109199524,
"learning_rate": 3.2723084589306063e-06,
"loss": 0.9914,
"step": 6660
},
{
"epoch": 1.0250533151453438,
"grad_norm": 0.37136781215667725,
"learning_rate": 3.1632728741956887e-06,
"loss": 1.2288,
"step": 6670
},
{
"epoch": 1.0265903283444446,
"grad_norm": 0.33485859632492065,
"learning_rate": 3.0560555794826196e-06,
"loss": 1.1068,
"step": 6680
},
{
"epoch": 1.0281273415435455,
"grad_norm": 0.38648903369903564,
"learning_rate": 2.950658587995958e-06,
"loss": 1.0232,
"step": 6690
},
{
"epoch": 1.0296643547426463,
"grad_norm": 0.4292406141757965,
"learning_rate": 2.8470838787606034e-06,
"loss": 0.9137,
"step": 6700
},
{
"epoch": 1.031201367941747,
"grad_norm": 0.2354525476694107,
"learning_rate": 2.7453333965847816e-06,
"loss": 1.0016,
"step": 6710
},
{
"epoch": 1.0327383811408481,
"grad_norm": 0.3365088403224945,
"learning_rate": 2.6454090520234064e-06,
"loss": 1.2594,
"step": 6720
},
{
"epoch": 1.034275394339949,
"grad_norm": 0.35403972864151,
"learning_rate": 2.5473127213422763e-06,
"loss": 1.0669,
"step": 6730
},
{
"epoch": 1.0358124075390498,
"grad_norm": 0.3797454535961151,
"learning_rate": 2.4510462464828356e-06,
"loss": 0.9636,
"step": 6740
},
{
"epoch": 1.0373494207381506,
"grad_norm": 0.4822129011154175,
"learning_rate": 2.3566114350275227e-06,
"loss": 0.9996,
"step": 6750
},
{
"epoch": 1.0388864339372514,
"grad_norm": 0.3022718131542206,
"learning_rate": 2.2640100601659443e-06,
"loss": 1.0637,
"step": 6760
},
{
"epoch": 1.0404234471363523,
"grad_norm": 0.31982535123825073,
"learning_rate": 2.1732438606614667e-06,
"loss": 1.3094,
"step": 6770
},
{
"epoch": 1.041960460335453,
"grad_norm": 0.38044679164886475,
"learning_rate": 2.0843145408186547e-06,
"loss": 1.0644,
"step": 6780
},
{
"epoch": 1.043497473534554,
"grad_norm": 0.38083702325820923,
"learning_rate": 1.9972237704512286e-06,
"loss": 0.9808,
"step": 6790
},
{
"epoch": 1.0450344867336547,
"grad_norm": 0.37573423981666565,
"learning_rate": 1.9119731848506906e-06,
"loss": 0.8321,
"step": 6800
},
{
"epoch": 1.0465714999327558,
"grad_norm": 0.2685695290565491,
"learning_rate": 1.828564384755682e-06,
"loss": 0.9654,
"step": 6810
},
{
"epoch": 1.0481085131318566,
"grad_norm": 0.26364433765411377,
"learning_rate": 1.7469989363218532e-06,
"loss": 1.1894,
"step": 6820
},
{
"epoch": 1.0496455263309574,
"grad_norm": 0.32037970423698425,
"learning_rate": 1.6672783710925287e-06,
"loss": 1.1689,
"step": 6830
},
{
"epoch": 1.0511825395300582,
"grad_norm": 0.5007265210151672,
"learning_rate": 1.5894041859698784e-06,
"loss": 1.0002,
"step": 6840
},
{
"epoch": 1.052719552729159,
"grad_norm": 0.49157097935676575,
"learning_rate": 1.5133778431868584e-06,
"loss": 0.8826,
"step": 6850
},
{
"epoch": 1.05425656592826,
"grad_norm": 0.2990802526473999,
"learning_rate": 1.4392007702797362e-06,
"loss": 0.9696,
"step": 6860
},
{
"epoch": 1.0557935791273607,
"grad_norm": 0.3625422716140747,
"learning_rate": 1.3668743600613099e-06,
"loss": 1.259,
"step": 6870
},
{
"epoch": 1.0573305923264615,
"grad_norm": 0.342822402715683,
"learning_rate": 1.2963999705947195e-06,
"loss": 1.1933,
"step": 6880
},
{
"epoch": 1.0588676055255624,
"grad_norm": 0.3823775351047516,
"learning_rate": 1.227778925167955e-06,
"loss": 1.0306,
"step": 6890
},
{
"epoch": 1.0604046187246632,
"grad_norm": 0.5350134968757629,
"learning_rate": 1.161012512269033e-06,
"loss": 0.9263,
"step": 6900
},
{
"epoch": 1.0619416319237642,
"grad_norm": 0.31349486112594604,
"learning_rate": 1.0961019855618038e-06,
"loss": 0.9779,
"step": 6910
},
{
"epoch": 1.063478645122865,
"grad_norm": 0.33665308356285095,
"learning_rate": 1.0330485638623488e-06,
"loss": 1.1529,
"step": 6920
},
{
"epoch": 1.0650156583219659,
"grad_norm": 0.44944867491722107,
"learning_rate": 9.718534311161986e-07,
"loss": 1.0706,
"step": 6930
},
{
"epoch": 1.0665526715210667,
"grad_norm": 0.3681994378566742,
"learning_rate": 9.125177363759952e-07,
"loss": 0.9934,
"step": 6940
},
{
"epoch": 1.0680896847201675,
"grad_norm": 0.5265238881111145,
"learning_rate": 8.550425937800088e-07,
"loss": 0.8407,
"step": 6950
},
{
"epoch": 1.0696266979192683,
"grad_norm": 0.29450106620788574,
"learning_rate": 7.994290825311334e-07,
"loss": 0.9489,
"step": 6960
},
{
"epoch": 1.0711637111183692,
"grad_norm": 0.43040338158607483,
"learning_rate": 7.456782468767021e-07,
"loss": 1.1235,
"step": 6970
},
{
"epoch": 1.07270072431747,
"grad_norm": 0.34595033526420593,
"learning_rate": 6.937910960888139e-07,
"loss": 1.0821,
"step": 6980
},
{
"epoch": 1.074237737516571,
"grad_norm": 0.44076165556907654,
"learning_rate": 6.437686044454383e-07,
"loss": 1.0215,
"step": 6990
},
{
"epoch": 1.0757747507156719,
"grad_norm": 0.4509797990322113,
"learning_rate": 5.956117112120741e-07,
"loss": 0.7941,
"step": 7000
},
{
"epoch": 1.0773117639147727,
"grad_norm": 0.32351309061050415,
"learning_rate": 5.49321320624141e-07,
"loss": 1.0092,
"step": 7010
},
{
"epoch": 1.0788487771138735,
"grad_norm": 0.31016451120376587,
"learning_rate": 5.048983018699827e-07,
"loss": 1.1812,
"step": 7020
},
{
"epoch": 1.0803857903129743,
"grad_norm": 0.3519907593727112,
"learning_rate": 4.6234348907457924e-07,
"loss": 1.0151,
"step": 7030
},
{
"epoch": 1.0819228035120751,
"grad_norm": 0.41296255588531494,
"learning_rate": 4.216576812838491e-07,
"loss": 1.0136,
"step": 7040
},
{
"epoch": 1.083459816711176,
"grad_norm": 0.44014647603034973,
"learning_rate": 3.8284164244963837e-07,
"loss": 0.9012,
"step": 7050
},
{
"epoch": 1.0849968299102768,
"grad_norm": 0.29607173800468445,
"learning_rate": 3.458961014154327e-07,
"loss": 0.9432,
"step": 7060
},
{
"epoch": 1.0865338431093776,
"grad_norm": 0.3175375163555145,
"learning_rate": 3.1082175190262353e-07,
"loss": 1.2191,
"step": 7070
},
{
"epoch": 1.0880708563084784,
"grad_norm": 0.4205903112888336,
"learning_rate": 2.776192524974741e-07,
"loss": 1.0987,
"step": 7080
},
{
"epoch": 1.0896078695075795,
"grad_norm": 0.3771834373474121,
"learning_rate": 2.4628922663879616e-07,
"loss": 1.0152,
"step": 7090
},
{
"epoch": 1.0911448827066803,
"grad_norm": 0.4023333489894867,
"learning_rate": 2.168322626062147e-07,
"loss": 0.8014,
"step": 7100
},
{
"epoch": 1.0926818959057811,
"grad_norm": 0.29351213574409485,
"learning_rate": 1.8924891350911022e-07,
"loss": 0.9837,
"step": 7110
},
{
"epoch": 1.094218909104882,
"grad_norm": 0.36660757660865784,
"learning_rate": 1.635396972762937e-07,
"loss": 1.2037,
"step": 7120
},
{
"epoch": 1.0957559223039828,
"grad_norm": 0.32501330971717834,
"learning_rate": 1.3970509664620323e-07,
"loss": 1.0518,
"step": 7130
},
{
"epoch": 1.0972929355030836,
"grad_norm": 0.503971517086029,
"learning_rate": 1.17745559157878e-07,
"loss": 1.0858,
"step": 7140
},
{
"epoch": 1.0988299487021844,
"grad_norm": 0.5006624460220337,
"learning_rate": 9.7661497142576e-08,
"loss": 0.872,
"step": 7150
},
{
"epoch": 1.1003669619012852,
"grad_norm": 0.2728460431098938,
"learning_rate": 7.945328771596928e-08,
"loss": 0.9752,
"step": 7160
},
{
"epoch": 1.1019039751003863,
"grad_norm": 0.3342999219894409,
"learning_rate": 6.312127277113833e-08,
"loss": 1.2468,
"step": 7170
},
{
"epoch": 1.1034409882994871,
"grad_norm": 0.327546089887619,
"learning_rate": 4.866575897208847e-08,
"loss": 0.9932,
"step": 7180
},
{
"epoch": 1.104978001498588,
"grad_norm": 0.49405235052108765,
"learning_rate": 3.608701774804324e-08,
"loss": 1.0088,
"step": 7190
},
{
"epoch": 1.1065150146976888,
"grad_norm": 0.41683152318000793,
"learning_rate": 2.5385285288315186e-08,
"loss": 0.8976,
"step": 7200
},
{
"epoch": 1.1080520278967896,
"grad_norm": 0.3056623041629791,
"learning_rate": 1.656076253786498e-08,
"loss": 1.0254,
"step": 7210
},
{
"epoch": 1.1095890410958904,
"grad_norm": 0.41433843970298767,
"learning_rate": 9.613615193548864e-09,
"loss": 1.2263,
"step": 7220
},
{
"epoch": 1.1111260542949912,
"grad_norm": 0.3642580211162567,
"learning_rate": 4.543973701021109e-09,
"loss": 1.1267,
"step": 7230
},
{
"epoch": 1.112663067494092,
"grad_norm": 0.44879820942878723,
"learning_rate": 1.3519332522471394e-09,
"loss": 1.0014,
"step": 7240
},
{
"epoch": 1.1142000806931929,
"grad_norm": 0.47233569622039795,
"learning_rate": 3.755378371606e-11,
"loss": 0.9992,
"step": 7250
},
{
"epoch": 1.1157370938922937,
"grad_norm": 0.2350589781999588,
"learning_rate": 4.770654446442715e-05,
"loss": 0.9208,
"step": 7260
},
{
"epoch": 1.1172741070913947,
"grad_norm": 0.39865681529045105,
"learning_rate": 4.745766968394424e-05,
"loss": 1.1291,
"step": 7270
},
{
"epoch": 1.1188111202904956,
"grad_norm": 0.36644208431243896,
"learning_rate": 4.720924364041792e-05,
"loss": 1.0064,
"step": 7280
},
{
"epoch": 1.1203481334895964,
"grad_norm": 0.4023211896419525,
"learning_rate": 4.6961268455526993e-05,
"loss": 0.945,
"step": 7290
},
{
"epoch": 1.1218851466886972,
"grad_norm": 0.45005688071250916,
"learning_rate": 4.671374624709961e-05,
"loss": 0.8435,
"step": 7300
},
{
"epoch": 1.123422159887798,
"grad_norm": 0.29800528287887573,
"learning_rate": 4.646667912909538e-05,
"loss": 0.9143,
"step": 7310
},
{
"epoch": 1.1249591730868989,
"grad_norm": 0.4643053710460663,
"learning_rate": 4.62200692115872e-05,
"loss": 1.2059,
"step": 7320
},
{
"epoch": 1.1264961862859997,
"grad_norm": 0.32091811299324036,
"learning_rate": 4.597391860074327e-05,
"loss": 1.0411,
"step": 7330
},
{
"epoch": 1.1280331994851005,
"grad_norm": 0.41880083084106445,
"learning_rate": 4.572822939880898e-05,
"loss": 0.9205,
"step": 7340
},
{
"epoch": 1.1295702126842015,
"grad_norm": 0.42431172728538513,
"learning_rate": 4.5483003704089246e-05,
"loss": 0.8339,
"step": 7350
},
{
"epoch": 1.1311072258833024,
"grad_norm": 0.305320680141449,
"learning_rate": 4.523824361093018e-05,
"loss": 0.9576,
"step": 7360
},
{
"epoch": 1.1326442390824032,
"grad_norm": 0.3122274577617645,
"learning_rate": 4.4993951209701704e-05,
"loss": 1.1862,
"step": 7370
},
{
"epoch": 1.134181252281504,
"grad_norm": 0.4227127134799957,
"learning_rate": 4.4750128586779174e-05,
"loss": 1.0013,
"step": 7380
},
{
"epoch": 1.1357182654806048,
"grad_norm": 0.3456993103027344,
"learning_rate": 4.4506777824525983e-05,
"loss": 0.9097,
"step": 7390
},
{
"epoch": 1.1372552786797057,
"grad_norm": 0.539394199848175,
"learning_rate": 4.426390100127547e-05,
"loss": 0.8957,
"step": 7400
},
{
"epoch": 1.1387922918788065,
"grad_norm": 0.3111777603626251,
"learning_rate": 4.402150019131343e-05,
"loss": 0.9709,
"step": 7410
},
{
"epoch": 1.1403293050779073,
"grad_norm": 0.35398274660110474,
"learning_rate": 4.377957746486022e-05,
"loss": 1.0995,
"step": 7420
},
{
"epoch": 1.1418663182770081,
"grad_norm": 0.38963282108306885,
"learning_rate": 4.353813488805308e-05,
"loss": 0.9935,
"step": 7430
},
{
"epoch": 1.143403331476109,
"grad_norm": 0.4354744851589203,
"learning_rate": 4.329717452292866e-05,
"loss": 0.9146,
"step": 7440
},
{
"epoch": 1.1449403446752098,
"grad_norm": 0.4899888336658478,
"learning_rate": 4.3056698427405105e-05,
"loss": 0.8607,
"step": 7450
},
{
"epoch": 1.1464773578743108,
"grad_norm": 0.29648080468177795,
"learning_rate": 4.281670865526493e-05,
"loss": 0.941,
"step": 7460
},
{
"epoch": 1.1480143710734116,
"grad_norm": 0.4165453612804413,
"learning_rate": 4.257720725613695e-05,
"loss": 1.2787,
"step": 7470
},
{
"epoch": 1.1495513842725125,
"grad_norm": 0.40669748187065125,
"learning_rate": 4.23381962754792e-05,
"loss": 1.1264,
"step": 7480
},
{
"epoch": 1.1510883974716133,
"grad_norm": 0.4564281105995178,
"learning_rate": 4.209967775456117e-05,
"loss": 0.9682,
"step": 7490
},
{
"epoch": 1.1526254106707141,
"grad_norm": 0.49203336238861084,
"learning_rate": 4.186165373044671e-05,
"loss": 0.8325,
"step": 7500
},
{
"epoch": 1.154162423869815,
"grad_norm": 0.32847046852111816,
"learning_rate": 4.16241262359762e-05,
"loss": 0.8895,
"step": 7510
},
{
"epoch": 1.1556994370689158,
"grad_norm": 0.33343231678009033,
"learning_rate": 4.138709729974961e-05,
"loss": 1.1938,
"step": 7520
},
{
"epoch": 1.1572364502680166,
"grad_norm": 0.4513290226459503,
"learning_rate": 4.115056894610879e-05,
"loss": 1.0468,
"step": 7530
},
{
"epoch": 1.1587734634671176,
"grad_norm": 0.3978137969970703,
"learning_rate": 4.0914543195120525e-05,
"loss": 0.957,
"step": 7540
},
{
"epoch": 1.1603104766662184,
"grad_norm": 0.5279019474983215,
"learning_rate": 4.067902206255911e-05,
"loss": 0.83,
"step": 7550
},
{
"epoch": 1.1618474898653193,
"grad_norm": 0.3176910877227783,
"learning_rate": 4.044400755988906e-05,
"loss": 0.9436,
"step": 7560
},
{
"epoch": 1.16338450306442,
"grad_norm": 0.3535185754299164,
"learning_rate": 4.020950169424815e-05,
"loss": 1.0437,
"step": 7570
},
{
"epoch": 1.164921516263521,
"grad_norm": 0.44217225909233093,
"learning_rate": 3.997550646842998e-05,
"loss": 1.0199,
"step": 7580
},
{
"epoch": 1.1664585294626217,
"grad_norm": 0.448397159576416,
"learning_rate": 3.974202388086729e-05,
"loss": 0.9873,
"step": 7590
},
{
"epoch": 1.1679955426617226,
"grad_norm": 0.5027180314064026,
"learning_rate": 3.950905592561438e-05,
"loss": 0.8109,
"step": 7600
},
{
"epoch": 1.1695325558608234,
"grad_norm": 0.33573320508003235,
"learning_rate": 3.927660459233053e-05,
"loss": 0.9064,
"step": 7610
},
{
"epoch": 1.1710695690599242,
"grad_norm": 0.4378392696380615,
"learning_rate": 3.9044671866262685e-05,
"loss": 1.1899,
"step": 7620
},
{
"epoch": 1.172606582259025,
"grad_norm": 0.4187834560871124,
"learning_rate": 3.881325972822872e-05,
"loss": 1.0873,
"step": 7630
},
{
"epoch": 1.174143595458126,
"grad_norm": 0.36880719661712646,
"learning_rate": 3.85823701546004e-05,
"loss": 0.8929,
"step": 7640
},
{
"epoch": 1.175680608657227,
"grad_norm": 0.5537612438201904,
"learning_rate": 3.8352005117286594e-05,
"loss": 0.7997,
"step": 7650
},
{
"epoch": 1.1772176218563277,
"grad_norm": 0.36902502179145813,
"learning_rate": 3.8122166583716255e-05,
"loss": 0.9056,
"step": 7660
},
{
"epoch": 1.1787546350554285,
"grad_norm": 0.3818625509738922,
"learning_rate": 3.789285651682184e-05,
"loss": 1.2545,
"step": 7670
},
{
"epoch": 1.1802916482545294,
"grad_norm": 0.3932090997695923,
"learning_rate": 3.7664076875022445e-05,
"loss": 0.9871,
"step": 7680
},
{
"epoch": 1.1818286614536302,
"grad_norm": 0.4951578974723816,
"learning_rate": 3.7435829612207006e-05,
"loss": 0.9259,
"step": 7690
},
{
"epoch": 1.183365674652731,
"grad_norm": 0.47290998697280884,
"learning_rate": 3.7208116677717774e-05,
"loss": 0.8285,
"step": 7700
},
{
"epoch": 1.1849026878518318,
"grad_norm": 0.2940005362033844,
"learning_rate": 3.6980940016333444e-05,
"loss": 0.8995,
"step": 7710
},
{
"epoch": 1.1864397010509329,
"grad_norm": 0.425828218460083,
"learning_rate": 3.675430156825288e-05,
"loss": 1.1385,
"step": 7720
},
{
"epoch": 1.1879767142500337,
"grad_norm": 0.5028491020202637,
"learning_rate": 3.652820326907815e-05,
"loss": 1.1308,
"step": 7730
},
{
"epoch": 1.1895137274491345,
"grad_norm": 0.4647107422351837,
"learning_rate": 3.630264704979834e-05,
"loss": 0.9242,
"step": 7740
},
{
"epoch": 1.1910507406482354,
"grad_norm": 0.6053696870803833,
"learning_rate": 3.607763483677276e-05,
"loss": 0.8548,
"step": 7750
},
{
"epoch": 1.1925877538473362,
"grad_norm": 0.3277945816516876,
"learning_rate": 3.585316855171476e-05,
"loss": 0.8647,
"step": 7760
},
{
"epoch": 1.194124767046437,
"grad_norm": 0.38459452986717224,
"learning_rate": 3.562925011167516e-05,
"loss": 1.2039,
"step": 7770
},
{
"epoch": 1.1956617802455378,
"grad_norm": 0.4018721878528595,
"learning_rate": 3.540588142902597e-05,
"loss": 1.0324,
"step": 7780
},
{
"epoch": 1.1971987934446386,
"grad_norm": 0.47276759147644043,
"learning_rate": 3.518306441144387e-05,
"loss": 0.9599,
"step": 7790
},
{
"epoch": 1.1987358066437395,
"grad_norm": 0.5612033605575562,
"learning_rate": 3.496080096189418e-05,
"loss": 0.895,
"step": 7800
},
{
"epoch": 1.2002728198428403,
"grad_norm": 0.2704327702522278,
"learning_rate": 3.4739092978614487e-05,
"loss": 0.8086,
"step": 7810
},
{
"epoch": 1.2018098330419413,
"grad_norm": 0.34935832023620605,
"learning_rate": 3.451794235509831e-05,
"loss": 1.1258,
"step": 7820
},
{
"epoch": 1.2033468462410422,
"grad_norm": 0.4486200511455536,
"learning_rate": 3.429735098007917e-05,
"loss": 0.9846,
"step": 7830
},
{
"epoch": 1.204883859440143,
"grad_norm": 0.4711863398551941,
"learning_rate": 3.407732073751423e-05,
"loss": 0.8391,
"step": 7840
},
{
"epoch": 1.2064208726392438,
"grad_norm": 0.43690264225006104,
"learning_rate": 3.3857853506568384e-05,
"loss": 0.8572,
"step": 7850
},
{
"epoch": 1.2079578858383446,
"grad_norm": 0.39491182565689087,
"learning_rate": 3.363895116159812e-05,
"loss": 0.8875,
"step": 7860
},
{
"epoch": 1.2094948990374454,
"grad_norm": 0.4491494297981262,
"learning_rate": 3.342061557213555e-05,
"loss": 1.262,
"step": 7870
},
{
"epoch": 1.2110319122365463,
"grad_norm": 0.4749878942966461,
"learning_rate": 3.320284860287232e-05,
"loss": 0.9851,
"step": 7880
},
{
"epoch": 1.212568925435647,
"grad_norm": 0.4113333821296692,
"learning_rate": 3.298565211364388e-05,
"loss": 0.9006,
"step": 7890
},
{
"epoch": 1.2141059386347481,
"grad_norm": 0.5665573477745056,
"learning_rate": 3.2769027959413475e-05,
"loss": 0.9201,
"step": 7900
},
{
"epoch": 1.215642951833849,
"grad_norm": 0.31964221596717834,
"learning_rate": 3.255297799025638e-05,
"loss": 0.9425,
"step": 7910
},
{
"epoch": 1.2171799650329498,
"grad_norm": 0.40363678336143494,
"learning_rate": 3.233750405134392e-05,
"loss": 1.1258,
"step": 7920
},
{
"epoch": 1.2187169782320506,
"grad_norm": 0.43074437975883484,
"learning_rate": 3.2122607982928e-05,
"loss": 0.9956,
"step": 7930
},
{
"epoch": 1.2202539914311514,
"grad_norm": 0.45079270005226135,
"learning_rate": 3.190829162032509e-05,
"loss": 0.9253,
"step": 7940
},
{
"epoch": 1.2217910046302523,
"grad_norm": 0.5983675122261047,
"learning_rate": 3.1694556793900786e-05,
"loss": 0.8134,
"step": 7950
},
{
"epoch": 1.223328017829353,
"grad_norm": 0.32823607325553894,
"learning_rate": 3.148140532905409e-05,
"loss": 0.9768,
"step": 7960
},
{
"epoch": 1.224865031028454,
"grad_norm": 0.2925056219100952,
"learning_rate": 3.1268839046201725e-05,
"loss": 1.1468,
"step": 7970
},
{
"epoch": 1.2264020442275547,
"grad_norm": 0.4182628095149994,
"learning_rate": 3.1056859760762755e-05,
"loss": 0.9531,
"step": 7980
},
{
"epoch": 1.2279390574266555,
"grad_norm": 0.4837040305137634,
"learning_rate": 3.084546928314298e-05,
"loss": 0.9282,
"step": 7990
},
{
"epoch": 1.2294760706257566,
"grad_norm": 0.6217512488365173,
"learning_rate": 3.063466941871952e-05,
"loss": 0.915,
"step": 8000
},
{
"epoch": 1.2310130838248574,
"grad_norm": 0.32692593336105347,
"learning_rate": 3.042446196782529e-05,
"loss": 0.9192,
"step": 8010
},
{
"epoch": 1.2325500970239582,
"grad_norm": 0.4098741114139557,
"learning_rate": 3.0214848725733813e-05,
"loss": 1.1674,
"step": 8020
},
{
"epoch": 1.234087110223059,
"grad_norm": 0.3959311544895172,
"learning_rate": 3.0005831482643644e-05,
"loss": 0.894,
"step": 8030
},
{
"epoch": 1.2356241234221599,
"grad_norm": 0.4396664798259735,
"learning_rate": 2.979741202366343e-05,
"loss": 0.8167,
"step": 8040
},
{
"epoch": 1.2371611366212607,
"grad_norm": 0.48536422848701477,
"learning_rate": 2.958959212879623e-05,
"loss": 0.8007,
"step": 8050
},
{
"epoch": 1.2386981498203615,
"grad_norm": 0.3064787983894348,
"learning_rate": 2.9382373572924694e-05,
"loss": 0.9055,
"step": 8060
},
{
"epoch": 1.2402351630194623,
"grad_norm": 0.3551320731639862,
"learning_rate": 2.9175758125795605e-05,
"loss": 1.1758,
"step": 8070
},
{
"epoch": 1.2417721762185634,
"grad_norm": 0.4403756260871887,
"learning_rate": 2.8969747552005043e-05,
"loss": 1.0292,
"step": 8080
},
{
"epoch": 1.2433091894176642,
"grad_norm": 0.5475226044654846,
"learning_rate": 2.8764343610983123e-05,
"loss": 0.9173,
"step": 8090
},
{
"epoch": 1.244846202616765,
"grad_norm": 0.415691614151001,
"learning_rate": 2.855954805697897e-05,
"loss": 0.7966,
"step": 8100
},
{
"epoch": 1.2463832158158659,
"grad_norm": 0.3230118751525879,
"learning_rate": 2.8355362639045846e-05,
"loss": 0.9523,
"step": 8110
},
{
"epoch": 1.2479202290149667,
"grad_norm": 0.4181591272354126,
"learning_rate": 2.815178910102616e-05,
"loss": 1.241,
"step": 8120
},
{
"epoch": 1.2494572422140675,
"grad_norm": 0.49596700072288513,
"learning_rate": 2.7948829181536584e-05,
"loss": 1.0833,
"step": 8130
},
{
"epoch": 1.2509942554131683,
"grad_norm": 0.43868303298950195,
"learning_rate": 2.7746484613953084e-05,
"loss": 0.9118,
"step": 8140
},
{
"epoch": 1.2525312686122692,
"grad_norm": 0.5722730159759521,
"learning_rate": 2.754475712639638e-05,
"loss": 0.8306,
"step": 8150
},
{
"epoch": 1.25406828181137,
"grad_norm": 0.34712284803390503,
"learning_rate": 2.734364844171684e-05,
"loss": 0.8655,
"step": 8160
},
{
"epoch": 1.2556052950104708,
"grad_norm": 0.46307608485221863,
"learning_rate": 2.7143160277480196e-05,
"loss": 1.1314,
"step": 8170
},
{
"epoch": 1.2571423082095716,
"grad_norm": 0.4576514959335327,
"learning_rate": 2.694329434595242e-05,
"loss": 1.055,
"step": 8180
},
{
"epoch": 1.2586793214086727,
"grad_norm": 0.626433253288269,
"learning_rate": 2.6744052354085436e-05,
"loss": 1.0039,
"step": 8190
},
{
"epoch": 1.2602163346077735,
"grad_norm": 0.5256068706512451,
"learning_rate": 2.6545436003502346e-05,
"loss": 0.832,
"step": 8200
},
{
"epoch": 1.2617533478068743,
"grad_norm": 0.3357698917388916,
"learning_rate": 2.6347446990483015e-05,
"loss": 0.9247,
"step": 8210
},
{
"epoch": 1.2632903610059751,
"grad_norm": 0.4270179569721222,
"learning_rate": 2.6150087005949563e-05,
"loss": 1.27,
"step": 8220
},
{
"epoch": 1.264827374205076,
"grad_norm": 0.4550669491291046,
"learning_rate": 2.5953357735451888e-05,
"loss": 0.9822,
"step": 8230
},
{
"epoch": 1.2663643874041768,
"grad_norm": 0.46991702914237976,
"learning_rate": 2.5757260859153208e-05,
"loss": 0.8595,
"step": 8240
},
{
"epoch": 1.2679014006032776,
"grad_norm": 0.7008426785469055,
"learning_rate": 2.55617980518159e-05,
"loss": 0.8841,
"step": 8250
},
{
"epoch": 1.2694384138023787,
"grad_norm": 0.3327344059944153,
"learning_rate": 2.536697098278704e-05,
"loss": 0.9015,
"step": 8260
},
{
"epoch": 1.2709754270014795,
"grad_norm": 0.3551974296569824,
"learning_rate": 2.5172781315984152e-05,
"loss": 1.1629,
"step": 8270
},
{
"epoch": 1.2725124402005803,
"grad_norm": 0.42040205001831055,
"learning_rate": 2.497923070988113e-05,
"loss": 1.0752,
"step": 8280
},
{
"epoch": 1.2740494533996811,
"grad_norm": 0.44223856925964355,
"learning_rate": 2.4786320817493813e-05,
"loss": 0.9307,
"step": 8290
},
{
"epoch": 1.275586466598782,
"grad_norm": 0.49881550669670105,
"learning_rate": 2.459405328636627e-05,
"loss": 0.803,
"step": 8300
},
{
"epoch": 1.2771234797978828,
"grad_norm": 0.29017916321754456,
"learning_rate": 2.4402429758556288e-05,
"loss": 0.916,
"step": 8310
},
{
"epoch": 1.2786604929969836,
"grad_norm": 0.42794203758239746,
"learning_rate": 2.4211451870621683e-05,
"loss": 1.303,
"step": 8320
},
{
"epoch": 1.2801975061960844,
"grad_norm": 0.43322041630744934,
"learning_rate": 2.4021121253606073e-05,
"loss": 1.0491,
"step": 8330
},
{
"epoch": 1.2817345193951852,
"grad_norm": 0.5845361351966858,
"learning_rate": 2.383143953302519e-05,
"loss": 0.9791,
"step": 8340
},
{
"epoch": 1.283271532594286,
"grad_norm": 0.5210367441177368,
"learning_rate": 2.3642408328852818e-05,
"loss": 0.8738,
"step": 8350
},
{
"epoch": 1.2848085457933869,
"grad_norm": 0.296173095703125,
"learning_rate": 2.3454029255507037e-05,
"loss": 0.9036,
"step": 8360
},
{
"epoch": 1.286345558992488,
"grad_norm": 0.35363104939460754,
"learning_rate": 2.326630392183635e-05,
"loss": 1.1842,
"step": 8370
},
{
"epoch": 1.2878825721915887,
"grad_norm": 0.45439237356185913,
"learning_rate": 2.307923393110609e-05,
"loss": 0.9963,
"step": 8380
},
{
"epoch": 1.2894195853906896,
"grad_norm": 0.4049324095249176,
"learning_rate": 2.2892820880984644e-05,
"loss": 0.9165,
"step": 8390
},
{
"epoch": 1.2909565985897904,
"grad_norm": 0.5537766814231873,
"learning_rate": 2.2707066363529694e-05,
"loss": 0.8448,
"step": 8400
},
{
"epoch": 1.2924936117888912,
"grad_norm": 0.3173496723175049,
"learning_rate": 2.2521971965174893e-05,
"loss": 0.9748,
"step": 8410
},
{
"epoch": 1.294030624987992,
"grad_norm": 0.42437151074409485,
"learning_rate": 2.2337539266715978e-05,
"loss": 1.2417,
"step": 8420
},
{
"epoch": 1.2955676381870929,
"grad_norm": 0.39316272735595703,
"learning_rate": 2.2153769843297667e-05,
"loss": 0.9587,
"step": 8430
},
{
"epoch": 1.297104651386194,
"grad_norm": 0.5042495727539062,
"learning_rate": 2.197066526439978e-05,
"loss": 0.9727,
"step": 8440
},
{
"epoch": 1.2986416645852947,
"grad_norm": 0.4606406092643738,
"learning_rate": 2.1788227093824186e-05,
"loss": 0.7925,
"step": 8450
},
{
"epoch": 1.3001786777843956,
"grad_norm": 0.32572638988494873,
"learning_rate": 2.1606456889681202e-05,
"loss": 0.8891,
"step": 8460
},
{
"epoch": 1.3017156909834964,
"grad_norm": 0.4118897020816803,
"learning_rate": 2.1425356204376456e-05,
"loss": 1.2133,
"step": 8470
},
{
"epoch": 1.3032527041825972,
"grad_norm": 0.46717625856399536,
"learning_rate": 2.1244926584597534e-05,
"loss": 1.0474,
"step": 8480
},
{
"epoch": 1.304789717381698,
"grad_norm": 0.4406215250492096,
"learning_rate": 2.1065169571300848e-05,
"loss": 0.8944,
"step": 8490
},
{
"epoch": 1.3063267305807988,
"grad_norm": 0.5081271529197693,
"learning_rate": 2.0886086699698317e-05,
"loss": 0.8721,
"step": 8500
},
{
"epoch": 1.3078637437798997,
"grad_norm": 0.3198259174823761,
"learning_rate": 2.070767949924446e-05,
"loss": 0.9137,
"step": 8510
},
{
"epoch": 1.3094007569790005,
"grad_norm": 0.3732990324497223,
"learning_rate": 2.0529949493623214e-05,
"loss": 1.2355,
"step": 8520
},
{
"epoch": 1.3109377701781013,
"grad_norm": 0.37645435333251953,
"learning_rate": 2.0352898200734883e-05,
"loss": 1.0286,
"step": 8530
},
{
"epoch": 1.3124747833772021,
"grad_norm": 0.48258766531944275,
"learning_rate": 2.0176527132683354e-05,
"loss": 0.9299,
"step": 8540
},
{
"epoch": 1.3140117965763032,
"grad_norm": 0.6357638239860535,
"learning_rate": 2.000083779576294e-05,
"loss": 0.787,
"step": 8550
},
{
"epoch": 1.315548809775404,
"grad_norm": 0.310835063457489,
"learning_rate": 1.9825831690445718e-05,
"loss": 0.9903,
"step": 8560
},
{
"epoch": 1.3170858229745048,
"grad_norm": 0.41710585355758667,
"learning_rate": 1.9651510311368604e-05,
"loss": 1.1655,
"step": 8570
},
{
"epoch": 1.3186228361736057,
"grad_norm": 0.39091309905052185,
"learning_rate": 1.9477875147320666e-05,
"loss": 1.0828,
"step": 8580
},
{
"epoch": 1.3201598493727065,
"grad_norm": 0.4138554334640503,
"learning_rate": 1.930492768123029e-05,
"loss": 0.9182,
"step": 8590
},
{
"epoch": 1.3216968625718073,
"grad_norm": 0.7107694745063782,
"learning_rate": 1.9132669390152634e-05,
"loss": 0.8715,
"step": 8600
},
{
"epoch": 1.3232338757709081,
"grad_norm": 0.3953629732131958,
"learning_rate": 1.8961101745256958e-05,
"loss": 0.8735,
"step": 8610
},
{
"epoch": 1.3247708889700092,
"grad_norm": 0.4464874267578125,
"learning_rate": 1.87902262118141e-05,
"loss": 1.2451,
"step": 8620
},
{
"epoch": 1.32630790216911,
"grad_norm": 0.4526148736476898,
"learning_rate": 1.8620044249183864e-05,
"loss": 1.082,
"step": 8630
},
{
"epoch": 1.3278449153682108,
"grad_norm": 0.5987499356269836,
"learning_rate": 1.8450557310802685e-05,
"loss": 0.8965,
"step": 8640
},
{
"epoch": 1.3293819285673116,
"grad_norm": 0.5726785659790039,
"learning_rate": 1.828176684417111e-05,
"loss": 0.8766,
"step": 8650
},
{
"epoch": 1.3309189417664125,
"grad_norm": 0.3449862003326416,
"learning_rate": 1.81136742908415e-05,
"loss": 0.9887,
"step": 8660
},
{
"epoch": 1.3324559549655133,
"grad_norm": 0.466362327337265,
"learning_rate": 1.7946281086405725e-05,
"loss": 1.1588,
"step": 8670
},
{
"epoch": 1.333992968164614,
"grad_norm": 0.4630391299724579,
"learning_rate": 1.7779588660482816e-05,
"loss": 0.9766,
"step": 8680
},
{
"epoch": 1.335529981363715,
"grad_norm": 0.6068161129951477,
"learning_rate": 1.761359843670686e-05,
"loss": 0.9039,
"step": 8690
},
{
"epoch": 1.3370669945628157,
"grad_norm": 0.6729193925857544,
"learning_rate": 1.744831183271479e-05,
"loss": 0.8534,
"step": 8700
},
{
"epoch": 1.3386040077619166,
"grad_norm": 0.31600821018218994,
"learning_rate": 1.7283730260134312e-05,
"loss": 0.9597,
"step": 8710
},
{
"epoch": 1.3401410209610174,
"grad_norm": 0.39009931683540344,
"learning_rate": 1.7119855124571725e-05,
"loss": 1.1718,
"step": 8720
},
{
"epoch": 1.3416780341601182,
"grad_norm": 0.4249347746372223,
"learning_rate": 1.695668782560016e-05,
"loss": 1.0047,
"step": 8730
},
{
"epoch": 1.3432150473592193,
"grad_norm": 0.486248254776001,
"learning_rate": 1.6794229756747304e-05,
"loss": 0.9427,
"step": 8740
},
{
"epoch": 1.34475206055832,
"grad_norm": 0.526313066482544,
"learning_rate": 1.6632482305483887e-05,
"loss": 0.7866,
"step": 8750
},
{
"epoch": 1.346289073757421,
"grad_norm": 0.32623177766799927,
"learning_rate": 1.6471446853211414e-05,
"loss": 1.0429,
"step": 8760
},
{
"epoch": 1.3478260869565217,
"grad_norm": 0.41753414273262024,
"learning_rate": 1.631112477525073e-05,
"loss": 1.1435,
"step": 8770
},
{
"epoch": 1.3493631001556226,
"grad_norm": 0.39541134238243103,
"learning_rate": 1.615151744083e-05,
"loss": 1.0731,
"step": 8780
},
{
"epoch": 1.3509001133547234,
"grad_norm": 0.47436606884002686,
"learning_rate": 1.5992626213073213e-05,
"loss": 1.0457,
"step": 8790
},
{
"epoch": 1.3524371265538242,
"grad_norm": 0.5544174313545227,
"learning_rate": 1.5834452448988456e-05,
"loss": 0.7688,
"step": 8800
},
{
"epoch": 1.3539741397529252,
"grad_norm": 0.3159116208553314,
"learning_rate": 1.5676997499456248e-05,
"loss": 0.9248,
"step": 8810
},
{
"epoch": 1.355511152952026,
"grad_norm": 0.4216868579387665,
"learning_rate": 1.552026270921818e-05,
"loss": 1.1765,
"step": 8820
},
{
"epoch": 1.3570481661511269,
"grad_norm": 0.3739314377307892,
"learning_rate": 1.5364249416865295e-05,
"loss": 0.9642,
"step": 8830
},
{
"epoch": 1.3585851793502277,
"grad_norm": 0.5721040964126587,
"learning_rate": 1.5208958954826714e-05,
"loss": 0.9671,
"step": 8840
},
{
"epoch": 1.3601221925493285,
"grad_norm": 0.46936526894569397,
"learning_rate": 1.5054392649358162e-05,
"loss": 0.8551,
"step": 8850
},
{
"epoch": 1.3616592057484294,
"grad_norm": 0.3435158431529999,
"learning_rate": 1.4900551820530828e-05,
"loss": 1.0253,
"step": 8860
},
{
"epoch": 1.3631962189475302,
"grad_norm": 0.3444412648677826,
"learning_rate": 1.4747437782219863e-05,
"loss": 1.1999,
"step": 8870
},
{
"epoch": 1.364733232146631,
"grad_norm": 0.46727800369262695,
"learning_rate": 1.459505184209341e-05,
"loss": 0.9987,
"step": 8880
},
{
"epoch": 1.3662702453457318,
"grad_norm": 0.48410764336586,
"learning_rate": 1.4443395301601171e-05,
"loss": 0.9729,
"step": 8890
},
{
"epoch": 1.3678072585448326,
"grad_norm": 0.5941210389137268,
"learning_rate": 1.4292469455963508e-05,
"loss": 0.9108,
"step": 8900
},
{
"epoch": 1.3693442717439335,
"grad_norm": 0.34704068303108215,
"learning_rate": 1.4142275594160226e-05,
"loss": 0.9584,
"step": 8910
},
{
"epoch": 1.3708812849430345,
"grad_norm": 0.37097346782684326,
"learning_rate": 1.399281499891969e-05,
"loss": 1.2058,
"step": 8920
},
{
"epoch": 1.3724182981421353,
"grad_norm": 0.4367105960845947,
"learning_rate": 1.3844088946707823e-05,
"loss": 1.0855,
"step": 8930
},
{
"epoch": 1.3739553113412362,
"grad_norm": 0.46947771310806274,
"learning_rate": 1.3696098707717109e-05,
"loss": 0.9315,
"step": 8940
},
{
"epoch": 1.375492324540337,
"grad_norm": 0.573675811290741,
"learning_rate": 1.3548845545855949e-05,
"loss": 0.8597,
"step": 8950
},
{
"epoch": 1.3770293377394378,
"grad_norm": 0.39916229248046875,
"learning_rate": 1.3402330718737599e-05,
"loss": 1.033,
"step": 8960
},
{
"epoch": 1.3785663509385386,
"grad_norm": 0.4043828845024109,
"learning_rate": 1.3256555477669775e-05,
"loss": 1.0683,
"step": 8970
},
{
"epoch": 1.3801033641376395,
"grad_norm": 0.44349074363708496,
"learning_rate": 1.3111521067643584e-05,
"loss": 0.9997,
"step": 8980
},
{
"epoch": 1.3816403773367405,
"grad_norm": 0.49961259961128235,
"learning_rate": 1.2967228727323199e-05,
"loss": 0.9263,
"step": 8990
},
{
"epoch": 1.3831773905358413,
"grad_norm": 0.4839262068271637,
"learning_rate": 1.282367968903504e-05,
"loss": 0.8291,
"step": 9000
},
{
"epoch": 1.3847144037349421,
"grad_norm": 0.3456418812274933,
"learning_rate": 1.2680875178757546e-05,
"loss": 0.9368,
"step": 9010
},
{
"epoch": 1.386251416934043,
"grad_norm": 0.29126814007759094,
"learning_rate": 1.2538816416110332e-05,
"loss": 1.1148,
"step": 9020
},
{
"epoch": 1.3877884301331438,
"grad_norm": 0.41824641823768616,
"learning_rate": 1.2397504614344113e-05,
"loss": 1.0474,
"step": 9030
},
{
"epoch": 1.3893254433322446,
"grad_norm": 0.5979284048080444,
"learning_rate": 1.2256940980330101e-05,
"loss": 0.9918,
"step": 9040
},
{
"epoch": 1.3908624565313454,
"grad_norm": 0.5245364308357239,
"learning_rate": 1.2117126714549843e-05,
"loss": 0.888,
"step": 9050
},
{
"epoch": 1.3923994697304463,
"grad_norm": 0.3799736201763153,
"learning_rate": 1.1978063011084927e-05,
"loss": 0.9632,
"step": 9060
},
{
"epoch": 1.393936482929547,
"grad_norm": 0.38089805841445923,
"learning_rate": 1.1839751057606718e-05,
"loss": 1.1867,
"step": 9070
},
{
"epoch": 1.395473496128648,
"grad_norm": 0.45045486092567444,
"learning_rate": 1.1702192035366344e-05,
"loss": 1.0758,
"step": 9080
},
{
"epoch": 1.3970105093277487,
"grad_norm": 0.5165355801582336,
"learning_rate": 1.1565387119184445e-05,
"loss": 0.9752,
"step": 9090
},
{
"epoch": 1.3985475225268498,
"grad_norm": 0.6082902550697327,
"learning_rate": 1.142933747744136e-05,
"loss": 0.8811,
"step": 9100
},
{
"epoch": 1.4000845357259506,
"grad_norm": 0.30874478816986084,
"learning_rate": 1.1294044272066872e-05,
"loss": 0.9209,
"step": 9110
},
{
"epoch": 1.4016215489250514,
"grad_norm": 0.41922399401664734,
"learning_rate": 1.1159508658530572e-05,
"loss": 1.2367,
"step": 9120
},
{
"epoch": 1.4031585621241522,
"grad_norm": 0.5856130123138428,
"learning_rate": 1.1025731785831705e-05,
"loss": 1.0612,
"step": 9130
},
{
"epoch": 1.404695575323253,
"grad_norm": 0.5159668326377869,
"learning_rate": 1.0892714796489612e-05,
"loss": 0.8883,
"step": 9140
},
{
"epoch": 1.4062325885223539,
"grad_norm": 0.6138080358505249,
"learning_rate": 1.0760458826533826e-05,
"loss": 0.9013,
"step": 9150
},
{
"epoch": 1.4077696017214547,
"grad_norm": 0.3192656338214874,
"learning_rate": 1.0628965005494429e-05,
"loss": 1.0084,
"step": 9160
},
{
"epoch": 1.4093066149205558,
"grad_norm": 0.4122283458709717,
"learning_rate": 1.0498234456392308e-05,
"loss": 1.2208,
"step": 9170
},
{
"epoch": 1.4108436281196566,
"grad_norm": 0.41741204261779785,
"learning_rate": 1.0368268295729722e-05,
"loss": 1.135,
"step": 9180
},
{
"epoch": 1.4123806413187574,
"grad_norm": 0.5270050764083862,
"learning_rate": 1.0239067633480659e-05,
"loss": 1.0629,
"step": 9190
},
{
"epoch": 1.4139176545178582,
"grad_norm": 0.6621428728103638,
"learning_rate": 1.0110633573081341e-05,
"loss": 0.8784,
"step": 9200
},
{
"epoch": 1.415454667716959,
"grad_norm": 0.3246777653694153,
"learning_rate": 9.982967211420891e-06,
"loss": 0.9843,
"step": 9210
},
{
"epoch": 1.4169916809160599,
"grad_norm": 0.41875627636909485,
"learning_rate": 9.856069638831845e-06,
"loss": 1.3286,
"step": 9220
},
{
"epoch": 1.4185286941151607,
"grad_norm": 0.45773589611053467,
"learning_rate": 9.72994193908101e-06,
"loss": 1.0736,
"step": 9230
},
{
"epoch": 1.4200657073142615,
"grad_norm": 0.5518283247947693,
"learning_rate": 9.604585189360016e-06,
"loss": 0.9831,
"step": 9240
},
{
"epoch": 1.4216027205133623,
"grad_norm": 0.5497329235076904,
"learning_rate": 9.480000460276261e-06,
"loss": 0.9625,
"step": 9250
},
{
"epoch": 1.4231397337124632,
"grad_norm": 0.332206666469574,
"learning_rate": 9.356188815843658e-06,
"loss": 1.0367,
"step": 9260
},
{
"epoch": 1.424676746911564,
"grad_norm": 0.4140862226486206,
"learning_rate": 9.233151313473654e-06,
"loss": 1.1779,
"step": 9270
},
{
"epoch": 1.426213760110665,
"grad_norm": 0.4356088936328888,
"learning_rate": 9.110889003966116e-06,
"loss": 1.0517,
"step": 9280
},
{
"epoch": 1.4277507733097659,
"grad_norm": 0.46055853366851807,
"learning_rate": 8.989402931500434e-06,
"loss": 0.944,
"step": 9290
},
{
"epoch": 1.4292877865088667,
"grad_norm": 0.6648761630058289,
"learning_rate": 8.868694133626487e-06,
"loss": 0.8789,
"step": 9300
},
{
"epoch": 1.4308247997079675,
"grad_norm": 0.4215174913406372,
"learning_rate": 8.748763641255909e-06,
"loss": 1.0096,
"step": 9310
},
{
"epoch": 1.4323618129070683,
"grad_norm": 0.4242957830429077,
"learning_rate": 8.629612478653205e-06,
"loss": 1.3106,
"step": 9320
},
{
"epoch": 1.4338988261061691,
"grad_norm": 0.5748409628868103,
"learning_rate": 8.511241663427027e-06,
"loss": 1.0679,
"step": 9330
},
{
"epoch": 1.43543583930527,
"grad_norm": 0.44487103819847107,
"learning_rate": 8.393652206521496e-06,
"loss": 0.9941,
"step": 9340
},
{
"epoch": 1.436972852504371,
"grad_norm": 0.5120466351509094,
"learning_rate": 8.27684511220751e-06,
"loss": 0.866,
"step": 9350
},
{
"epoch": 1.4385098657034718,
"grad_norm": 0.36236733198165894,
"learning_rate": 8.16082137807428e-06,
"loss": 1.0111,
"step": 9360
},
{
"epoch": 1.4400468789025727,
"grad_norm": 0.45448875427246094,
"learning_rate": 8.045581995020702e-06,
"loss": 1.0407,
"step": 9370
},
{
"epoch": 1.4415838921016735,
"grad_norm": 0.47790491580963135,
"learning_rate": 7.93112794724693e-06,
"loss": 1.1454,
"step": 9380
},
{
"epoch": 1.4431209053007743,
"grad_norm": 0.445593923330307,
"learning_rate": 7.817460212245985e-06,
"loss": 1.0428,
"step": 9390
},
{
"epoch": 1.4446579184998751,
"grad_norm": 0.5327218174934387,
"learning_rate": 7.704579760795395e-06,
"loss": 0.8962,
"step": 9400
},
{
"epoch": 1.446194931698976,
"grad_norm": 0.31549760699272156,
"learning_rate": 7.592487556948891e-06,
"loss": 0.9868,
"step": 9410
},
{
"epoch": 1.4477319448980768,
"grad_norm": 0.42698511481285095,
"learning_rate": 7.481184558028231e-06,
"loss": 1.2191,
"step": 9420
},
{
"epoch": 1.4492689580971776,
"grad_norm": 0.5377118587493896,
"learning_rate": 7.370671714614918e-06,
"loss": 1.1371,
"step": 9430
},
{
"epoch": 1.4508059712962784,
"grad_norm": 0.4495163857936859,
"learning_rate": 7.260949970542208e-06,
"loss": 1.0493,
"step": 9440
},
{
"epoch": 1.4523429844953792,
"grad_norm": 0.6538254618644714,
"learning_rate": 7.152020262886916e-06,
"loss": 0.974,
"step": 9450
},
{
"epoch": 1.45387999769448,
"grad_norm": 0.34346768260002136,
"learning_rate": 7.043883521961536e-06,
"loss": 0.9963,
"step": 9460
},
{
"epoch": 1.455417010893581,
"grad_norm": 0.3803898096084595,
"learning_rate": 6.936540671306236e-06,
"loss": 1.3377,
"step": 9470
},
{
"epoch": 1.456954024092682,
"grad_norm": 0.4858042299747467,
"learning_rate": 6.829992627680948e-06,
"loss": 1.1107,
"step": 9480
},
{
"epoch": 1.4584910372917828,
"grad_norm": 0.5924073457717896,
"learning_rate": 6.7242403010575935e-06,
"loss": 1.0602,
"step": 9490
},
{
"epoch": 1.4600280504908836,
"grad_norm": 0.6468076109886169,
"learning_rate": 6.619284594612285e-06,
"loss": 0.9625,
"step": 9500
},
{
"epoch": 1.4615650636899844,
"grad_norm": 0.35481786727905273,
"learning_rate": 6.515126404717597e-06,
"loss": 0.9665,
"step": 9510
},
{
"epoch": 1.4631020768890852,
"grad_norm": 0.40993350744247437,
"learning_rate": 6.411766620934934e-06,
"loss": 1.2373,
"step": 9520
},
{
"epoch": 1.4646390900881863,
"grad_norm": 0.4032086133956909,
"learning_rate": 6.309206126006928e-06,
"loss": 1.0128,
"step": 9530
},
{
"epoch": 1.466176103287287,
"grad_norm": 0.4416462481021881,
"learning_rate": 6.207445795849864e-06,
"loss": 1.0567,
"step": 9540
},
{
"epoch": 1.467713116486388,
"grad_norm": 0.6962372660636902,
"learning_rate": 6.106486499546315e-06,
"loss": 0.8453,
"step": 9550
},
{
"epoch": 1.4692501296854887,
"grad_norm": 0.30278781056404114,
"learning_rate": 6.006329099337549e-06,
"loss": 0.943,
"step": 9560
},
{
"epoch": 1.4707871428845896,
"grad_norm": 0.39832136034965515,
"learning_rate": 5.9069744506163205e-06,
"loss": 1.2566,
"step": 9570
},
{
"epoch": 1.4723241560836904,
"grad_norm": 0.4463650584220886,
"learning_rate": 5.808423401919416e-06,
"loss": 1.0678,
"step": 9580
},
{
"epoch": 1.4738611692827912,
"grad_norm": 0.49209052324295044,
"learning_rate": 5.710676794920566e-06,
"loss": 0.9347,
"step": 9590
},
{
"epoch": 1.475398182481892,
"grad_norm": 0.561610221862793,
"learning_rate": 5.613735464423164e-06,
"loss": 0.9614,
"step": 9600
},
{
"epoch": 1.4769351956809929,
"grad_norm": 0.33109334111213684,
"learning_rate": 5.517600238353105e-06,
"loss": 0.9882,
"step": 9610
},
{
"epoch": 1.4784722088800937,
"grad_norm": 0.3682621419429779,
"learning_rate": 5.42227193775181e-06,
"loss": 1.1971,
"step": 9620
},
{
"epoch": 1.4800092220791945,
"grad_norm": 0.39119163155555725,
"learning_rate": 5.327751376769152e-06,
"loss": 1.1044,
"step": 9630
},
{
"epoch": 1.4815462352782953,
"grad_norm": 0.4651740789413452,
"learning_rate": 5.23403936265654e-06,
"loss": 1.0666,
"step": 9640
},
{
"epoch": 1.4830832484773964,
"grad_norm": 0.5594034194946289,
"learning_rate": 5.14113669575993e-06,
"loss": 0.9062,
"step": 9650
},
{
"epoch": 1.4846202616764972,
"grad_norm": 0.34486231207847595,
"learning_rate": 5.049044169513139e-06,
"loss": 1.0089,
"step": 9660
},
{
"epoch": 1.486157274875598,
"grad_norm": 0.4222332537174225,
"learning_rate": 4.9577625704309345e-06,
"loss": 1.2758,
"step": 9670
},
{
"epoch": 1.4876942880746988,
"grad_norm": 0.45434704422950745,
"learning_rate": 4.867292678102431e-06,
"loss": 1.1106,
"step": 9680
},
{
"epoch": 1.4892313012737997,
"grad_norm": 0.5504967570304871,
"learning_rate": 4.777635265184299e-06,
"loss": 1.0471,
"step": 9690
},
{
"epoch": 1.4907683144729005,
"grad_norm": 0.4923297166824341,
"learning_rate": 4.6887910973943185e-06,
"loss": 0.9356,
"step": 9700
},
{
"epoch": 1.4923053276720013,
"grad_norm": 0.34652334451675415,
"learning_rate": 4.6007609335047e-06,
"loss": 1.005,
"step": 9710
},
{
"epoch": 1.4938423408711023,
"grad_norm": 0.47138267755508423,
"learning_rate": 4.513545525335705e-06,
"loss": 1.174,
"step": 9720
},
{
"epoch": 1.4953793540702032,
"grad_norm": 0.4908987283706665,
"learning_rate": 4.427145617749174e-06,
"loss": 1.0486,
"step": 9730
},
{
"epoch": 1.496916367269304,
"grad_norm": 0.5694043040275574,
"learning_rate": 4.34156194864217e-06,
"loss": 1.0884,
"step": 9740
},
{
"epoch": 1.4984533804684048,
"grad_norm": 0.6227392554283142,
"learning_rate": 4.256795248940693e-06,
"loss": 0.9209,
"step": 9750
},
{
"epoch": 1.4999903936675056,
"grad_norm": 0.2886059284210205,
"learning_rate": 4.172846242593409e-06,
"loss": 0.9583,
"step": 9760
},
{
"epoch": 1.5015274068666065,
"grad_norm": 0.4085761308670044,
"learning_rate": 4.0897156465655506e-06,
"loss": 1.2192,
"step": 9770
},
{
"epoch": 1.5030644200657073,
"grad_norm": 0.46467334032058716,
"learning_rate": 4.00740417083263e-06,
"loss": 1.0993,
"step": 9780
},
{
"epoch": 1.504601433264808,
"grad_norm": 0.5951000452041626,
"learning_rate": 3.925912518374575e-06,
"loss": 1.1404,
"step": 9790
},
{
"epoch": 1.506138446463909,
"grad_norm": 0.5107768177986145,
"learning_rate": 3.845241385169518e-06,
"loss": 0.8793,
"step": 9800
},
{
"epoch": 1.5076754596630098,
"grad_norm": 0.3446359932422638,
"learning_rate": 3.7653914601880946e-06,
"loss": 0.9637,
"step": 9810
},
{
"epoch": 1.5092124728621106,
"grad_norm": 0.4146728813648224,
"learning_rate": 3.686363425387296e-06,
"loss": 1.3483,
"step": 9820
},
{
"epoch": 1.5107494860612114,
"grad_norm": 0.3822716176509857,
"learning_rate": 3.6081579557048497e-06,
"loss": 1.0807,
"step": 9830
},
{
"epoch": 1.5122864992603124,
"grad_norm": 0.4824408292770386,
"learning_rate": 3.5307757190533477e-06,
"loss": 1.0135,
"step": 9840
},
{
"epoch": 1.5138235124594133,
"grad_norm": 0.618729829788208,
"learning_rate": 3.4542173763145726e-06,
"loss": 0.9876,
"step": 9850
},
{
"epoch": 1.515360525658514,
"grad_norm": 0.3057554066181183,
"learning_rate": 3.378483581333891e-06,
"loss": 1.015,
"step": 9860
},
{
"epoch": 1.516897538857615,
"grad_norm": 0.38738444447517395,
"learning_rate": 3.30357498091457e-06,
"loss": 1.1708,
"step": 9870
},
{
"epoch": 1.5184345520567157,
"grad_norm": 0.48995938897132874,
"learning_rate": 3.2294922148123796e-06,
"loss": 1.1423,
"step": 9880
},
{
"epoch": 1.5199715652558168,
"grad_norm": 0.4809519052505493,
"learning_rate": 3.156235915729999e-06,
"loss": 1.0731,
"step": 9890
},
{
"epoch": 1.5215085784549176,
"grad_norm": 0.5960240960121155,
"learning_rate": 3.083806709311743e-06,
"loss": 0.9146,
"step": 9900
},
{
"epoch": 1.5230455916540184,
"grad_norm": 0.3351467549800873,
"learning_rate": 3.0122052141381085e-06,
"loss": 0.9161,
"step": 9910
},
{
"epoch": 1.5245826048531192,
"grad_norm": 0.37443968653678894,
"learning_rate": 2.941432041720582e-06,
"loss": 1.1657,
"step": 9920
},
{
"epoch": 1.52611961805222,
"grad_norm": 0.32315656542778015,
"learning_rate": 2.871487796496297e-06,
"loss": 1.1565,
"step": 9930
},
{
"epoch": 1.527656631251321,
"grad_norm": 0.5246838927268982,
"learning_rate": 2.8023730758230284e-06,
"loss": 1.0277,
"step": 9940
},
{
"epoch": 1.5291936444504217,
"grad_norm": 0.6615852117538452,
"learning_rate": 2.734088469973961e-06,
"loss": 0.9811,
"step": 9950
},
{
"epoch": 1.5307306576495225,
"grad_norm": 0.37553471326828003,
"learning_rate": 2.6666345621327303e-06,
"loss": 1.056,
"step": 9960
},
{
"epoch": 1.5322676708486234,
"grad_norm": 0.4169570803642273,
"learning_rate": 2.6000119283883686e-06,
"loss": 1.2816,
"step": 9970
},
{
"epoch": 1.5338046840477242,
"grad_norm": 0.4750652611255646,
"learning_rate": 2.5342211377304526e-06,
"loss": 1.1477,
"step": 9980
},
{
"epoch": 1.535341697246825,
"grad_norm": 0.568051278591156,
"learning_rate": 2.4692627520442102e-06,
"loss": 1.0597,
"step": 9990
},
{
"epoch": 1.5368787104459258,
"grad_norm": 0.4986318051815033,
"learning_rate": 2.405137326105722e-06,
"loss": 0.9715,
"step": 10000
},
{
"epoch": 1.5384157236450267,
"grad_norm": 0.4139535129070282,
"learning_rate": 2.3418454075772032e-06,
"loss": 1.0189,
"step": 10010
},
{
"epoch": 1.5399527368441277,
"grad_norm": 0.37886664271354675,
"learning_rate": 2.2793875370022753e-06,
"loss": 1.2023,
"step": 10020
},
{
"epoch": 1.5414897500432285,
"grad_norm": 0.5900385975837708,
"learning_rate": 2.2177642478014573e-06,
"loss": 1.1934,
"step": 10030
},
{
"epoch": 1.5430267632423293,
"grad_norm": 0.484694242477417,
"learning_rate": 2.1569760662674465e-06,
"loss": 1.0326,
"step": 10040
},
{
"epoch": 1.5445637764414302,
"grad_norm": 0.5154560804367065,
"learning_rate": 2.102981149175598e-06,
"loss": 0.95,
"step": 10050
},
{
"epoch": 1.546100789640531,
"grad_norm": 0.34481093287467957,
"learning_rate": 2.0437810965974524e-06,
"loss": 1.0188,
"step": 10060
},
{
"epoch": 1.547637802839632,
"grad_norm": 0.41280755400657654,
"learning_rate": 1.98541763758654e-06,
"loss": 1.16,
"step": 10070
},
{
"epoch": 1.5491748160387329,
"grad_norm": 0.5078760981559753,
"learning_rate": 1.9278912705950615e-06,
"loss": 1.1809,
"step": 10080
},
{
"epoch": 1.5507118292378337,
"grad_norm": 0.4625634551048279,
"learning_rate": 1.8712024869260824e-06,
"loss": 1.0411,
"step": 10090
},
{
"epoch": 1.5522488424369345,
"grad_norm": 0.5834779143333435,
"learning_rate": 1.8153517707292788e-06,
"loss": 0.8995,
"step": 10100
},
{
"epoch": 1.5537858556360353,
"grad_norm": 0.41031381487846375,
"learning_rate": 1.7603395989968186e-06,
"loss": 1.0536,
"step": 10110
},
{
"epoch": 1.5553228688351362,
"grad_norm": 0.3177848160266876,
"learning_rate": 1.7061664415592982e-06,
"loss": 1.2401,
"step": 10120
},
{
"epoch": 1.556859882034237,
"grad_norm": 0.6637346744537354,
"learning_rate": 1.6581283396284508e-06,
"loss": 1.1639,
"step": 10130
},
{
"epoch": 1.5583968952333378,
"grad_norm": 0.5097059607505798,
"learning_rate": 1.6055505780678937e-06,
"loss": 0.9695,
"step": 10140
},
{
"epoch": 1.5599339084324386,
"grad_norm": 0.6397558450698853,
"learning_rate": 1.5538131527754963e-06,
"loss": 0.9681,
"step": 10150
},
{
"epoch": 1.5614709216315394,
"grad_norm": 0.28344351053237915,
"learning_rate": 1.5029165056139493e-06,
"loss": 1.0747,
"step": 10160
},
{
"epoch": 1.5630079348306403,
"grad_norm": 0.384098082780838,
"learning_rate": 1.4528610712652768e-06,
"loss": 1.2219,
"step": 10170
},
{
"epoch": 1.564544948029741,
"grad_norm": 0.4621298909187317,
"learning_rate": 1.4036472772271491e-06,
"loss": 1.0414,
"step": 10180
},
{
"epoch": 1.566081961228842,
"grad_norm": 0.5704950094223022,
"learning_rate": 1.355275543809198e-06,
"loss": 0.9753,
"step": 10190
},
{
"epoch": 1.5676189744279427,
"grad_norm": 1.008958339691162,
"learning_rate": 1.307746284129463e-06,
"loss": 0.9536,
"step": 10200
},
{
"epoch": 1.5691559876270438,
"grad_norm": 0.2939854860305786,
"learning_rate": 1.2610599041108617e-06,
"loss": 1.064,
"step": 10210
},
{
"epoch": 1.5706930008261446,
"grad_norm": 0.3789184093475342,
"learning_rate": 1.215216802477681e-06,
"loss": 1.2413,
"step": 10220
},
{
"epoch": 1.5722300140252454,
"grad_norm": 0.5255075693130493,
"learning_rate": 1.1702173707522357e-06,
"loss": 1.1318,
"step": 10230
},
{
"epoch": 1.5737670272243462,
"grad_norm": 0.45387375354766846,
"learning_rate": 1.1260619932514483e-06,
"loss": 1.007,
"step": 10240
},
{
"epoch": 1.5753040404234473,
"grad_norm": 0.5807976126670837,
"learning_rate": 1.0827510470836634e-06,
"loss": 1.016,
"step": 10250
},
{
"epoch": 1.5768410536225481,
"grad_norm": 0.2939109206199646,
"learning_rate": 1.0402849021453277e-06,
"loss": 0.9903,
"step": 10260
},
{
"epoch": 1.578378066821649,
"grad_norm": 0.36469635367393494,
"learning_rate": 9.986639211179038e-07,
"loss": 1.1298,
"step": 10270
},
{
"epoch": 1.5799150800207498,
"grad_norm": 0.41160863637924194,
"learning_rate": 9.578884594647176e-07,
"loss": 0.9701,
"step": 10280
},
{
"epoch": 1.5814520932198506,
"grad_norm": 0.42246487736701965,
"learning_rate": 9.179588654279814e-07,
"loss": 0.9105,
"step": 10290
},
{
"epoch": 1.5829891064189514,
"grad_norm": 0.5103077292442322,
"learning_rate": 8.827457296077479e-07,
"loss": 0.9148,
"step": 10300
},
{
"epoch": 1.5845261196180522,
"grad_norm": 0.30235615372657776,
"learning_rate": 8.444242175739803e-07,
"loss": 0.924,
"step": 10310
},
{
"epoch": 1.586063132817153,
"grad_norm": 0.3630199432373047,
"learning_rate": 8.069495421962558e-07,
"loss": 1.057,
"step": 10320
},
{
"epoch": 1.5876001460162539,
"grad_norm": 0.475341260433197,
"learning_rate": 7.703220235264708e-07,
"loss": 1.1083,
"step": 10330
},
{
"epoch": 1.5891371592153547,
"grad_norm": 0.45893293619155884,
"learning_rate": 7.34541974381342e-07,
"loss": 0.8202,
"step": 10340
},
{
"epoch": 1.5906741724144555,
"grad_norm": 0.5396574139595032,
"learning_rate": 6.99609700339865e-07,
"loss": 0.9057,
"step": 10350
},
{
"epoch": 1.5922111856135563,
"grad_norm": 0.31328675150871277,
"learning_rate": 6.655254997405714e-07,
"loss": 0.9475,
"step": 10360
},
{
"epoch": 1.5937481988126572,
"grad_norm": 0.3822092115879059,
"learning_rate": 6.322896636790421e-07,
"loss": 1.3043,
"step": 10370
},
{
"epoch": 1.595285212011758,
"grad_norm": 0.4274253845214844,
"learning_rate": 5.999024760054095e-07,
"loss": 1.1086,
"step": 10380
},
{
"epoch": 1.596822225210859,
"grad_norm": 0.6360399723052979,
"learning_rate": 5.683642133219592e-07,
"loss": 0.9799,
"step": 10390
},
{
"epoch": 1.5983592384099599,
"grad_norm": 0.42891958355903625,
"learning_rate": 5.376751449807427e-07,
"loss": 0.8691,
"step": 10400
},
{
"epoch": 1.5998962516090607,
"grad_norm": 0.31562769412994385,
"learning_rate": 5.078355330812579e-07,
"loss": 0.8803,
"step": 10410
},
{
"epoch": 1.6014332648081615,
"grad_norm": 0.3709647059440613,
"learning_rate": 4.7884563246825e-07,
"loss": 1.2021,
"step": 10420
},
{
"epoch": 1.6029702780072626,
"grad_norm": 0.4071331024169922,
"learning_rate": 4.5070569072952483e-07,
"loss": 0.9899,
"step": 10430
},
{
"epoch": 1.6045072912063634,
"grad_norm": 0.5559951066970825,
"learning_rate": 4.2341594819380604e-07,
"loss": 0.9337,
"step": 10440
},
{
"epoch": 1.6060443044054642,
"grad_norm": 0.600069522857666,
"learning_rate": 3.969766379287143e-07,
"loss": 0.805,
"step": 10450
},
{
"epoch": 1.607581317604565,
"grad_norm": 0.31316104531288147,
"learning_rate": 3.713879857387803e-07,
"loss": 0.9853,
"step": 10460
},
{
"epoch": 1.6091183308036658,
"grad_norm": 0.438211590051651,
"learning_rate": 3.4665021016347944e-07,
"loss": 1.2802,
"step": 10470
},
{
"epoch": 1.6106553440027667,
"grad_norm": 0.4018309712409973,
"learning_rate": 3.227635224754111e-07,
"loss": 1.0073,
"step": 10480
},
{
"epoch": 1.6121923572018675,
"grad_norm": 0.5478571057319641,
"learning_rate": 2.9972812667845574e-07,
"loss": 0.9769,
"step": 10490
},
{
"epoch": 1.6137293704009683,
"grad_norm": 0.5530366897583008,
"learning_rate": 2.7754421950605405e-07,
"loss": 0.951,
"step": 10500
},
{
"epoch": 1.6152663836000691,
"grad_norm": 0.3348609209060669,
"learning_rate": 2.562119904195193e-07,
"loss": 0.962,
"step": 10510
},
{
"epoch": 1.61680339679917,
"grad_norm": 0.3660121262073517,
"learning_rate": 2.3573162160642759e-07,
"loss": 1.1655,
"step": 10520
},
{
"epoch": 1.6183404099982708,
"grad_norm": 0.41799792647361755,
"learning_rate": 2.1610328797904144e-07,
"loss": 1.0419,
"step": 10530
},
{
"epoch": 1.6198774231973716,
"grad_norm": 0.48620516061782837,
"learning_rate": 1.973271571728441e-07,
"loss": 1.0567,
"step": 10540
},
{
"epoch": 1.6214144363964724,
"grad_norm": 0.5304862856864929,
"learning_rate": 1.794033895450964e-07,
"loss": 0.924,
"step": 10550
},
{
"epoch": 1.6229514495955732,
"grad_norm": 0.28448235988616943,
"learning_rate": 1.6233213817344884e-07,
"loss": 0.9019,
"step": 10560
},
{
"epoch": 1.6244884627946743,
"grad_norm": 0.3161237835884094,
"learning_rate": 1.46113548854665e-07,
"loss": 1.2519,
"step": 10570
},
{
"epoch": 1.6260254759937751,
"grad_norm": 0.43361780047416687,
"learning_rate": 1.3074776010334467e-07,
"loss": 1.0793,
"step": 10580
},
{
"epoch": 1.627562489192876,
"grad_norm": 0.4638166129589081,
"learning_rate": 1.1623490315078034e-07,
"loss": 0.9881,
"step": 10590
},
{
"epoch": 1.6290995023919768,
"grad_norm": 0.5153199434280396,
"learning_rate": 1.0257510194379149e-07,
"loss": 0.8847,
"step": 10600
},
{
"epoch": 1.6306365155910776,
"grad_norm": 0.256044864654541,
"learning_rate": 8.976847314371428e-08,
"loss": 0.893,
"step": 10610
},
{
"epoch": 1.6321735287901786,
"grad_norm": 0.3465503752231598,
"learning_rate": 7.78151261253357e-08,
"loss": 1.1732,
"step": 10620
},
{
"epoch": 1.6337105419892795,
"grad_norm": 0.4150409698486328,
"learning_rate": 6.671516297606095e-08,
"loss": 0.9656,
"step": 10630
},
{
"epoch": 1.6352475551883803,
"grad_norm": 0.6488197445869446,
"learning_rate": 5.646867849493642e-08,
"loss": 0.8801,
"step": 10640
},
{
"epoch": 1.636784568387481,
"grad_norm": 0.4544541835784912,
"learning_rate": 4.707576019190585e-08,
"loss": 0.882,
"step": 10650
},
{
"epoch": 1.638321581586582,
"grad_norm": 0.31461796164512634,
"learning_rate": 3.853648828704426e-08,
"loss": 0.9086,
"step": 10660
},
{
"epoch": 1.6398585947856827,
"grad_norm": 0.39737704396247864,
"learning_rate": 3.085093570986963e-08,
"loss": 1.2193,
"step": 10670
},
{
"epoch": 1.6413956079847836,
"grad_norm": 0.4205061197280884,
"learning_rate": 2.401916809872118e-08,
"loss": 1.0293,
"step": 10680
},
{
"epoch": 1.6429326211838844,
"grad_norm": 0.5001878142356873,
"learning_rate": 1.8041243800215323e-08,
"loss": 0.9788,
"step": 10690
},
{
"epoch": 1.6444696343829852,
"grad_norm": 0.4962889850139618,
"learning_rate": 1.2917213868701705e-08,
"loss": 0.8945,
"step": 10700
},
{
"epoch": 1.646006647582086,
"grad_norm": 0.31304067373275757,
"learning_rate": 8.647122065896795e-09,
"loss": 0.9525,
"step": 10710
},
{
"epoch": 1.6475436607811869,
"grad_norm": 0.4079444110393524,
"learning_rate": 5.231004860450916e-09,
"loss": 1.3002,
"step": 10720
},
{
"epoch": 1.6490806739802877,
"grad_norm": 0.4398459196090698,
"learning_rate": 2.6688914276595812e-09,
"loss": 1.1095,
"step": 10730
},
{
"epoch": 1.6506176871793885,
"grad_norm": 0.5031657814979553,
"learning_rate": 9.608036491970396e-10,
"loss": 1.0557,
"step": 10740
},
{
"epoch": 1.6521547003784895,
"grad_norm": 0.45387646555900574,
"learning_rate": 1.06756112983053e-10,
"loss": 0.8578,
"step": 10750
}
],
"logging_steps": 10,
"max_steps": 10750,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 250,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.283096169185792e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}