rqadri's picture
Upload folder using huggingface_hub
fe71c9a verified
Raw History Blame Contribute Delete
49.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.20053054201319184,
"eval_steps": 500,
"global_step": 2797,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0007169486664754804,
"grad_norm": 315.4236659535465,
"learning_rate": 1.1061797752808989e-06,
"loss": 1.848151397705078,
"step": 10
},
{
"epoch": 0.0014338973329509608,
"grad_norm": 672.973252055138,
"learning_rate": 2.2241573033707866e-06,
"loss": 1.1725639343261718,
"step": 20
},
{
"epoch": 0.002150845999426441,
"grad_norm": 126.65993818302222,
"learning_rate": 3.342134831460675e-06,
"loss": 0.7593138694763184,
"step": 30
},
{
"epoch": 0.0028677946659019216,
"grad_norm": 29.886751429599133,
"learning_rate": 4.4601123595505624e-06,
"loss": 0.6618130683898926,
"step": 40
},
{
"epoch": 0.0035847433323774017,
"grad_norm": 137.5005248660405,
"learning_rate": 5.5780898876404495e-06,
"loss": 0.6227120399475098,
"step": 50
},
{
"epoch": 0.004301691998852882,
"grad_norm": 38.139287491655594,
"learning_rate": 6.6960674157303374e-06,
"loss": 0.5859609127044678,
"step": 60
},
{
"epoch": 0.005018640665328362,
"grad_norm": 51.02157400174971,
"learning_rate": 7.814044943820226e-06,
"loss": 0.6018378734588623,
"step": 70
},
{
"epoch": 0.005735589331803843,
"grad_norm": 24.465251435228527,
"learning_rate": 8.932022471910112e-06,
"loss": 0.5624160289764404,
"step": 80
},
{
"epoch": 0.006452537998279323,
"grad_norm": 22.828409412609748,
"learning_rate": 1.005e-05,
"loss": 0.5475245475769043,
"step": 90
},
{
"epoch": 0.007169486664754803,
"grad_norm": 14.786662427088277,
"learning_rate": 1.1167977528089888e-05,
"loss": 0.5528249263763427,
"step": 100
},
{
"epoch": 0.007886435331230283,
"grad_norm": 75.92842280330609,
"learning_rate": 1.2285955056179776e-05,
"loss": 0.5614291667938233,
"step": 110
},
{
"epoch": 0.008603383997705764,
"grad_norm": 20.360896855493966,
"learning_rate": 1.3403932584269664e-05,
"loss": 0.6078104972839355,
"step": 120
},
{
"epoch": 0.009320332664181245,
"grad_norm": 31.43963052094272,
"learning_rate": 1.4521910112359552e-05,
"loss": 0.5890892028808594,
"step": 130
},
{
"epoch": 0.010037281330656725,
"grad_norm": 54.35895463776518,
"learning_rate": 1.563988764044944e-05,
"loss": 0.6670564174652099,
"step": 140
},
{
"epoch": 0.010754229997132205,
"grad_norm": 33.964610072430865,
"learning_rate": 1.6757865168539328e-05,
"loss": 0.7157962322235107,
"step": 150
},
{
"epoch": 0.011471178663607686,
"grad_norm": 110.65639964008429,
"learning_rate": 1.7875842696629214e-05,
"loss": 0.7455081462860107,
"step": 160
},
{
"epoch": 0.012188127330083166,
"grad_norm": 41.868609627942014,
"learning_rate": 1.89938202247191e-05,
"loss": 1.1094837188720703,
"step": 170
},
{
"epoch": 0.012905075996558647,
"grad_norm": 10.702193812872004,
"learning_rate": 2e-05,
"loss": 0.6436423778533935,
"step": 180
},
{
"epoch": 0.013622024663034128,
"grad_norm": 51.71265251185425,
"learning_rate": 2e-05,
"loss": 0.6243439674377441,
"step": 190
},
{
"epoch": 0.014338973329509607,
"grad_norm": 154.8177418260545,
"learning_rate": 2e-05,
"loss": 0.610740852355957,
"step": 200
},
{
"epoch": 0.015055921995985088,
"grad_norm": 52.282960752120125,
"learning_rate": 2e-05,
"loss": 0.6687478065490723,
"step": 210
},
{
"epoch": 0.015772870662460567,
"grad_norm": 3.8483090765907435,
"learning_rate": 2e-05,
"loss": 0.579859733581543,
"step": 220
},
{
"epoch": 0.016489819328936048,
"grad_norm": 3.6109561986511833,
"learning_rate": 2e-05,
"loss": 0.6074420928955078,
"step": 230
},
{
"epoch": 0.01720676799541153,
"grad_norm": 5.053213094247748,
"learning_rate": 2e-05,
"loss": 0.5664905071258545,
"step": 240
},
{
"epoch": 0.01792371666188701,
"grad_norm": 3.0138514562209093,
"learning_rate": 2e-05,
"loss": 0.5948857307434082,
"step": 250
},
{
"epoch": 0.01864066532836249,
"grad_norm": 13.449705648790312,
"learning_rate": 2e-05,
"loss": 0.6020577907562256,
"step": 260
},
{
"epoch": 0.019357613994837968,
"grad_norm": 3.6227046108363377,
"learning_rate": 2e-05,
"loss": 0.5548747062683106,
"step": 270
},
{
"epoch": 0.02007456266131345,
"grad_norm": 4.112840464071606,
"learning_rate": 2e-05,
"loss": 0.5805762767791748,
"step": 280
},
{
"epoch": 0.02079151132778893,
"grad_norm": 1.9209785496382135,
"learning_rate": 2e-05,
"loss": 0.5842366218566895,
"step": 290
},
{
"epoch": 0.02150845999426441,
"grad_norm": 3.0684228704996004,
"learning_rate": 2e-05,
"loss": 0.5580601692199707,
"step": 300
},
{
"epoch": 0.022225408660739892,
"grad_norm": 1.1909889223759265,
"learning_rate": 2e-05,
"loss": 0.5615853786468505,
"step": 310
},
{
"epoch": 0.022942357327215373,
"grad_norm": 2.644802472025187,
"learning_rate": 2e-05,
"loss": 0.5611689567565918,
"step": 320
},
{
"epoch": 0.02365930599369085,
"grad_norm": 20.263309516988937,
"learning_rate": 2e-05,
"loss": 0.5905077934265137,
"step": 330
},
{
"epoch": 0.02437625466016633,
"grad_norm": 5.118749842105238,
"learning_rate": 2e-05,
"loss": 0.5447088241577148,
"step": 340
},
{
"epoch": 0.025093203326641812,
"grad_norm": 14.83085444344932,
"learning_rate": 2e-05,
"loss": 0.5638551235198974,
"step": 350
},
{
"epoch": 0.025810151993117293,
"grad_norm": 32.69304087592208,
"learning_rate": 2e-05,
"loss": 0.6271306037902832,
"step": 360
},
{
"epoch": 0.026527100659592774,
"grad_norm": 7.062265311502185,
"learning_rate": 2e-05,
"loss": 0.5702651023864747,
"step": 370
},
{
"epoch": 0.027244049326068255,
"grad_norm": 30.12218728774149,
"learning_rate": 2e-05,
"loss": 0.6142964839935303,
"step": 380
},
{
"epoch": 0.027960997992543733,
"grad_norm": 4.161661997042373,
"learning_rate": 2e-05,
"loss": 0.6302419662475586,
"step": 390
},
{
"epoch": 0.028677946659019213,
"grad_norm": 3.3921348737529113,
"learning_rate": 2e-05,
"loss": 0.5987069129943847,
"step": 400
},
{
"epoch": 0.029394895325494694,
"grad_norm": 1.4838082708559646,
"learning_rate": 2e-05,
"loss": 0.602195930480957,
"step": 410
},
{
"epoch": 0.030111843991970175,
"grad_norm": 3.3128177340785627,
"learning_rate": 2e-05,
"loss": 0.5845789909362793,
"step": 420
},
{
"epoch": 0.030828792658445656,
"grad_norm": 2.1866628727610182,
"learning_rate": 2e-05,
"loss": 0.5644162654876709,
"step": 430
},
{
"epoch": 0.031545741324921134,
"grad_norm": 11.884198120767302,
"learning_rate": 2e-05,
"loss": 0.5629150390625,
"step": 440
},
{
"epoch": 0.032262689991396615,
"grad_norm": 18.16899260160586,
"learning_rate": 2e-05,
"loss": 0.5555632591247559,
"step": 450
},
{
"epoch": 0.032979638657872096,
"grad_norm": 5.462581933210383,
"learning_rate": 2e-05,
"loss": 0.5991993904113769,
"step": 460
},
{
"epoch": 0.03369658732434758,
"grad_norm": 3.045413742803253,
"learning_rate": 2e-05,
"loss": 0.5700560569763183,
"step": 470
},
{
"epoch": 0.03441353599082306,
"grad_norm": 4.246757908127417,
"learning_rate": 2e-05,
"loss": 0.5917448997497559,
"step": 480
},
{
"epoch": 0.03513048465729854,
"grad_norm": 5.424046887719749,
"learning_rate": 2e-05,
"loss": 0.5816346645355225,
"step": 490
},
{
"epoch": 0.03584743332377402,
"grad_norm": 6.583800218516907,
"learning_rate": 2e-05,
"loss": 0.5913246154785157,
"step": 500
},
{
"epoch": 0.0365643819902495,
"grad_norm": 3.482147587454769,
"learning_rate": 2e-05,
"loss": 0.5650829315185547,
"step": 510
},
{
"epoch": 0.03728133065672498,
"grad_norm": 2.808982864585292,
"learning_rate": 2e-05,
"loss": 0.5790875911712646,
"step": 520
},
{
"epoch": 0.03799827932320046,
"grad_norm": 4.388524538222509,
"learning_rate": 2e-05,
"loss": 0.5646647453308106,
"step": 530
},
{
"epoch": 0.038715227989675936,
"grad_norm": 3.929514313642325,
"learning_rate": 2e-05,
"loss": 0.5568113327026367,
"step": 540
},
{
"epoch": 0.03943217665615142,
"grad_norm": 1.928176676276484,
"learning_rate": 2e-05,
"loss": 0.5727216720581054,
"step": 550
},
{
"epoch": 0.0401491253226269,
"grad_norm": 6.68335008196921,
"learning_rate": 2e-05,
"loss": 0.5850999355316162,
"step": 560
},
{
"epoch": 0.04086607398910238,
"grad_norm": 15.837152933980025,
"learning_rate": 2e-05,
"loss": 0.5926525115966796,
"step": 570
},
{
"epoch": 0.04158302265557786,
"grad_norm": 3.6114569877104423,
"learning_rate": 2e-05,
"loss": 0.5768462657928467,
"step": 580
},
{
"epoch": 0.04229997132205334,
"grad_norm": 2.164511077764764,
"learning_rate": 2e-05,
"loss": 0.567426061630249,
"step": 590
},
{
"epoch": 0.04301691998852882,
"grad_norm": 2.948019819021623,
"learning_rate": 2e-05,
"loss": 0.5795512199401855,
"step": 600
},
{
"epoch": 0.0437338686550043,
"grad_norm": 39.2052027751844,
"learning_rate": 2e-05,
"loss": 0.5852450370788574,
"step": 610
},
{
"epoch": 0.044450817321479784,
"grad_norm": 11.115023206584201,
"learning_rate": 2e-05,
"loss": 0.5830046653747558,
"step": 620
},
{
"epoch": 0.045167765987955265,
"grad_norm": 11.130215580191274,
"learning_rate": 2e-05,
"loss": 0.55075101852417,
"step": 630
},
{
"epoch": 0.045884714654430746,
"grad_norm": 2.2401242450660765,
"learning_rate": 2e-05,
"loss": 0.5412608623504639,
"step": 640
},
{
"epoch": 0.04660166332090622,
"grad_norm": 25.8476496640342,
"learning_rate": 2e-05,
"loss": 0.5914183616638183,
"step": 650
},
{
"epoch": 0.0473186119873817,
"grad_norm": 3.955944237962116,
"learning_rate": 2e-05,
"loss": 0.5842460632324219,
"step": 660
},
{
"epoch": 0.04803556065385718,
"grad_norm": 2.509395729576886,
"learning_rate": 2e-05,
"loss": 0.5712038993835449,
"step": 670
},
{
"epoch": 0.04875250932033266,
"grad_norm": 2.4532942736851084,
"learning_rate": 2e-05,
"loss": 0.5865258693695068,
"step": 680
},
{
"epoch": 0.049469457986808144,
"grad_norm": 3.6555036797810754,
"learning_rate": 2e-05,
"loss": 0.5788607597351074,
"step": 690
},
{
"epoch": 0.050186406653283624,
"grad_norm": 5.581489874452689,
"learning_rate": 2e-05,
"loss": 0.5549694061279297,
"step": 700
},
{
"epoch": 0.050903355319759105,
"grad_norm": 7.448830832646055,
"learning_rate": 2e-05,
"loss": 0.5698171138763428,
"step": 710
},
{
"epoch": 0.051620303986234586,
"grad_norm": 3.6608686716559906,
"learning_rate": 2e-05,
"loss": 0.554749584197998,
"step": 720
},
{
"epoch": 0.05233725265271007,
"grad_norm": 4.307177970225383,
"learning_rate": 2e-05,
"loss": 0.5440076351165771,
"step": 730
},
{
"epoch": 0.05305420131918555,
"grad_norm": 4.418311679242279,
"learning_rate": 2e-05,
"loss": 0.5399693489074707,
"step": 740
},
{
"epoch": 0.05377114998566103,
"grad_norm": 5.044806271145424,
"learning_rate": 2e-05,
"loss": 0.5785567283630371,
"step": 750
},
{
"epoch": 0.05448809865213651,
"grad_norm": 3.4544327147592595,
"learning_rate": 2e-05,
"loss": 0.6100111961364746,
"step": 760
},
{
"epoch": 0.055205047318611984,
"grad_norm": 2.9036817715442393,
"learning_rate": 2e-05,
"loss": 0.5828133583068847,
"step": 770
},
{
"epoch": 0.055921995985087465,
"grad_norm": 1.7522351082267351,
"learning_rate": 2e-05,
"loss": 0.5638906478881835,
"step": 780
},
{
"epoch": 0.056638944651562946,
"grad_norm": 2.209946265416314,
"learning_rate": 2e-05,
"loss": 0.590320348739624,
"step": 790
},
{
"epoch": 0.05735589331803843,
"grad_norm": 2.6684001731598195,
"learning_rate": 2e-05,
"loss": 0.5646738052368164,
"step": 800
},
{
"epoch": 0.05807284198451391,
"grad_norm": 5.223266428619974,
"learning_rate": 2e-05,
"loss": 0.5637637138366699,
"step": 810
},
{
"epoch": 0.05878979065098939,
"grad_norm": 1.9142769139433,
"learning_rate": 2e-05,
"loss": 0.5934517860412598,
"step": 820
},
{
"epoch": 0.05950673931746487,
"grad_norm": 2.6149464203468504,
"learning_rate": 2e-05,
"loss": 0.5700265884399414,
"step": 830
},
{
"epoch": 0.06022368798394035,
"grad_norm": 3.96105780832147,
"learning_rate": 2e-05,
"loss": 0.5675205230712891,
"step": 840
},
{
"epoch": 0.06094063665041583,
"grad_norm": 2.5400423051210974,
"learning_rate": 2e-05,
"loss": 0.5908157348632812,
"step": 850
},
{
"epoch": 0.06165758531689131,
"grad_norm": 2.6344860048789127,
"learning_rate": 2e-05,
"loss": 0.5718094825744628,
"step": 860
},
{
"epoch": 0.062374533983366794,
"grad_norm": 2.600050881708375,
"learning_rate": 2e-05,
"loss": 0.5780810832977294,
"step": 870
},
{
"epoch": 0.06309148264984227,
"grad_norm": 3.134985069799007,
"learning_rate": 2e-05,
"loss": 0.567197322845459,
"step": 880
},
{
"epoch": 0.06380843131631775,
"grad_norm": 2.0285438198733443,
"learning_rate": 2e-05,
"loss": 0.5852311611175537,
"step": 890
},
{
"epoch": 0.06452537998279323,
"grad_norm": 5.248736255854688,
"learning_rate": 2e-05,
"loss": 0.5891732215881348,
"step": 900
},
{
"epoch": 0.06524232864926871,
"grad_norm": 1.8727144020099094,
"learning_rate": 2e-05,
"loss": 0.5862767696380615,
"step": 910
},
{
"epoch": 0.06595927731574419,
"grad_norm": 1.711765938123259,
"learning_rate": 2e-05,
"loss": 0.6182688713073731,
"step": 920
},
{
"epoch": 0.06667622598221967,
"grad_norm": 1.763371319742271,
"learning_rate": 2e-05,
"loss": 0.5868721008300781,
"step": 930
},
{
"epoch": 0.06739317464869515,
"grad_norm": 1.2813315472585687,
"learning_rate": 2e-05,
"loss": 0.5589268684387207,
"step": 940
},
{
"epoch": 0.06811012331517063,
"grad_norm": 2.751663887567051,
"learning_rate": 2e-05,
"loss": 0.5824700355529785,
"step": 950
},
{
"epoch": 0.06882707198164612,
"grad_norm": 4.993093109327952,
"learning_rate": 2e-05,
"loss": 0.5773461818695068,
"step": 960
},
{
"epoch": 0.0695440206481216,
"grad_norm": 6.264445826159941,
"learning_rate": 2e-05,
"loss": 0.5533229827880859,
"step": 970
},
{
"epoch": 0.07026096931459708,
"grad_norm": 3.2290134324615365,
"learning_rate": 2e-05,
"loss": 0.5897500514984131,
"step": 980
},
{
"epoch": 0.07097791798107256,
"grad_norm": 3.0414609504263934,
"learning_rate": 2e-05,
"loss": 0.5628997802734375,
"step": 990
},
{
"epoch": 0.07169486664754804,
"grad_norm": 2.0651674928177504,
"learning_rate": 2e-05,
"loss": 0.5647363662719727,
"step": 1000
},
{
"epoch": 0.07241181531402352,
"grad_norm": 2.325752137633854,
"learning_rate": 2e-05,
"loss": 0.574032211303711,
"step": 1010
},
{
"epoch": 0.073128763980499,
"grad_norm": 2.1263355903408008,
"learning_rate": 2e-05,
"loss": 0.5692534446716309,
"step": 1020
},
{
"epoch": 0.07384571264697448,
"grad_norm": 24.27180244689631,
"learning_rate": 2e-05,
"loss": 0.5639516353607178,
"step": 1030
},
{
"epoch": 0.07456266131344996,
"grad_norm": 1.0790226331900736,
"learning_rate": 2e-05,
"loss": 0.5968218803405761,
"step": 1040
},
{
"epoch": 0.07527960997992544,
"grad_norm": 1.0623420342326966,
"learning_rate": 2e-05,
"loss": 0.5545104026794434,
"step": 1050
},
{
"epoch": 0.07599655864640092,
"grad_norm": 2.342778745540576,
"learning_rate": 2e-05,
"loss": 0.5392826080322266,
"step": 1060
},
{
"epoch": 0.07671350731287639,
"grad_norm": 2.003241443878693,
"learning_rate": 2e-05,
"loss": 0.5342343807220459,
"step": 1070
},
{
"epoch": 0.07743045597935187,
"grad_norm": 2.027419316411222,
"learning_rate": 2e-05,
"loss": 0.5778005123138428,
"step": 1080
},
{
"epoch": 0.07814740464582735,
"grad_norm": 1.896028667711285,
"learning_rate": 2e-05,
"loss": 0.5735631465911866,
"step": 1090
},
{
"epoch": 0.07886435331230283,
"grad_norm": 2.0470674801420214,
"learning_rate": 2e-05,
"loss": 0.580590009689331,
"step": 1100
},
{
"epoch": 0.07958130197877832,
"grad_norm": 2.7415867683239825,
"learning_rate": 2e-05,
"loss": 0.5596016883850098,
"step": 1110
},
{
"epoch": 0.0802982506452538,
"grad_norm": 1.6531033173482321,
"learning_rate": 2e-05,
"loss": 0.5827084541320801,
"step": 1120
},
{
"epoch": 0.08101519931172928,
"grad_norm": 1.2823997995083143,
"learning_rate": 2e-05,
"loss": 0.5864274024963378,
"step": 1130
},
{
"epoch": 0.08173214797820476,
"grad_norm": 1.3642460710168849,
"learning_rate": 2e-05,
"loss": 0.5656659603118896,
"step": 1140
},
{
"epoch": 0.08244909664468024,
"grad_norm": 2.6194064364228318,
"learning_rate": 2e-05,
"loss": 0.5763952255249023,
"step": 1150
},
{
"epoch": 0.08316604531115572,
"grad_norm": 1.2781451995694622,
"learning_rate": 2e-05,
"loss": 0.5566621780395508,
"step": 1160
},
{
"epoch": 0.0838829939776312,
"grad_norm": 1.5589065178330486,
"learning_rate": 2e-05,
"loss": 0.53390474319458,
"step": 1170
},
{
"epoch": 0.08459994264410668,
"grad_norm": 7.459041859593309,
"learning_rate": 2e-05,
"loss": 0.5653298854827881,
"step": 1180
},
{
"epoch": 0.08531689131058216,
"grad_norm": 2.9904694002471066,
"learning_rate": 2e-05,
"loss": 0.560931396484375,
"step": 1190
},
{
"epoch": 0.08603383997705764,
"grad_norm": 1.5979927096992035,
"learning_rate": 2e-05,
"loss": 0.530519676208496,
"step": 1200
},
{
"epoch": 0.08675078864353312,
"grad_norm": 8.353200078910842,
"learning_rate": 2e-05,
"loss": 0.5533581733703613,
"step": 1210
},
{
"epoch": 0.0874677373100086,
"grad_norm": 37.21010749725826,
"learning_rate": 2e-05,
"loss": 0.5901515007019043,
"step": 1220
},
{
"epoch": 0.08818468597648409,
"grad_norm": 5.122367920805441,
"learning_rate": 2e-05,
"loss": 0.5856198310852051,
"step": 1230
},
{
"epoch": 0.08890163464295957,
"grad_norm": 3.3489073569661105,
"learning_rate": 2e-05,
"loss": 0.5682595729827881,
"step": 1240
},
{
"epoch": 0.08961858330943505,
"grad_norm": 2.327849536892137,
"learning_rate": 2e-05,
"loss": 0.5584495067596436,
"step": 1250
},
{
"epoch": 0.09033553197591053,
"grad_norm": 22.72509490275124,
"learning_rate": 2e-05,
"loss": 0.5762526512145996,
"step": 1260
},
{
"epoch": 0.09105248064238601,
"grad_norm": 7.3344483954259285,
"learning_rate": 2e-05,
"loss": 0.5725028991699219,
"step": 1270
},
{
"epoch": 0.09176942930886149,
"grad_norm": 3.9521562517230238,
"learning_rate": 2e-05,
"loss": 0.5834849357604981,
"step": 1280
},
{
"epoch": 0.09248637797533697,
"grad_norm": 6.43496222827271,
"learning_rate": 2e-05,
"loss": 0.5840559005737305,
"step": 1290
},
{
"epoch": 0.09320332664181244,
"grad_norm": 3.8438872068840104,
"learning_rate": 2e-05,
"loss": 0.6151006698608399,
"step": 1300
},
{
"epoch": 0.09392027530828792,
"grad_norm": 1.4662403573164653,
"learning_rate": 2e-05,
"loss": 0.5667572021484375,
"step": 1310
},
{
"epoch": 0.0946372239747634,
"grad_norm": 17.21523102720178,
"learning_rate": 2e-05,
"loss": 0.5740190505981445,
"step": 1320
},
{
"epoch": 0.09535417264123888,
"grad_norm": 1.766413981347864,
"learning_rate": 2e-05,
"loss": 0.5626147270202637,
"step": 1330
},
{
"epoch": 0.09607112130771436,
"grad_norm": 2.6800514021980004,
"learning_rate": 2e-05,
"loss": 0.5712190628051758,
"step": 1340
},
{
"epoch": 0.09678806997418984,
"grad_norm": 2.961182111187328,
"learning_rate": 2e-05,
"loss": 0.5644628047943115,
"step": 1350
},
{
"epoch": 0.09750501864066533,
"grad_norm": 2.454375644899256,
"learning_rate": 2e-05,
"loss": 0.5722562313079834,
"step": 1360
},
{
"epoch": 0.0982219673071408,
"grad_norm": 1.6873399339759625,
"learning_rate": 2e-05,
"loss": 0.5544299125671387,
"step": 1370
},
{
"epoch": 0.09893891597361629,
"grad_norm": 1.1555872203445525,
"learning_rate": 2e-05,
"loss": 0.5385549068450928,
"step": 1380
},
{
"epoch": 0.09965586464009177,
"grad_norm": 1.5548337351700245,
"learning_rate": 2e-05,
"loss": 0.5589668273925781,
"step": 1390
},
{
"epoch": 0.10037281330656725,
"grad_norm": 1.2505995954976357,
"learning_rate": 2e-05,
"loss": 0.57919921875,
"step": 1400
},
{
"epoch": 0.10108976197304273,
"grad_norm": 5.206526378064766,
"learning_rate": 2e-05,
"loss": 0.5539649963378906,
"step": 1410
},
{
"epoch": 0.10180671063951821,
"grad_norm": 2.610432174521263,
"learning_rate": 2e-05,
"loss": 0.592057180404663,
"step": 1420
},
{
"epoch": 0.10252365930599369,
"grad_norm": 3.3133808987579276,
"learning_rate": 2e-05,
"loss": 0.5443053245544434,
"step": 1430
},
{
"epoch": 0.10324060797246917,
"grad_norm": 2.0455832858183682,
"learning_rate": 2e-05,
"loss": 0.546113920211792,
"step": 1440
},
{
"epoch": 0.10395755663894465,
"grad_norm": 1.7592817899937072,
"learning_rate": 2e-05,
"loss": 0.551348066329956,
"step": 1450
},
{
"epoch": 0.10467450530542013,
"grad_norm": 2.943775668256852,
"learning_rate": 2e-05,
"loss": 0.5249824523925781,
"step": 1460
},
{
"epoch": 0.10539145397189562,
"grad_norm": 2.140971602034373,
"learning_rate": 2e-05,
"loss": 0.5666528224945069,
"step": 1470
},
{
"epoch": 0.1061084026383711,
"grad_norm": 4.331861493161884,
"learning_rate": 2e-05,
"loss": 0.5543347835540772,
"step": 1480
},
{
"epoch": 0.10682535130484658,
"grad_norm": 6.353962200216883,
"learning_rate": 2e-05,
"loss": 0.5454977035522461,
"step": 1490
},
{
"epoch": 0.10754229997132206,
"grad_norm": 41.865001318361585,
"learning_rate": 2e-05,
"loss": 0.5601842880249024,
"step": 1500
},
{
"epoch": 0.10825924863779754,
"grad_norm": 1.7010891103793604,
"learning_rate": 2e-05,
"loss": 0.5699163436889648,
"step": 1510
},
{
"epoch": 0.10897619730427302,
"grad_norm": 2.8044652038050373,
"learning_rate": 2e-05,
"loss": 0.5804720878601074,
"step": 1520
},
{
"epoch": 0.1096931459707485,
"grad_norm": 2.8044499219399186,
"learning_rate": 2e-05,
"loss": 0.5642358779907226,
"step": 1530
},
{
"epoch": 0.11041009463722397,
"grad_norm": 1.6404567243971169,
"learning_rate": 2e-05,
"loss": 0.547740125656128,
"step": 1540
},
{
"epoch": 0.11112704330369945,
"grad_norm": 5.8842210253367595,
"learning_rate": 2e-05,
"loss": 0.5724470138549804,
"step": 1550
},
{
"epoch": 0.11184399197017493,
"grad_norm": 2.0263887469590465,
"learning_rate": 2e-05,
"loss": 0.5523801803588867,
"step": 1560
},
{
"epoch": 0.11256094063665041,
"grad_norm": 1.3512109762818472,
"learning_rate": 2e-05,
"loss": 0.5701728343963623,
"step": 1570
},
{
"epoch": 0.11327788930312589,
"grad_norm": 2.961754694676749,
"learning_rate": 2e-05,
"loss": 0.5410641670227051,
"step": 1580
},
{
"epoch": 0.11399483796960137,
"grad_norm": 9.38177531956118,
"learning_rate": 2e-05,
"loss": 0.5361968994140625,
"step": 1590
},
{
"epoch": 0.11471178663607685,
"grad_norm": 3.1019434885984083,
"learning_rate": 2e-05,
"loss": 0.5702048301696777,
"step": 1600
},
{
"epoch": 0.11542873530255233,
"grad_norm": 1.8141527074697572,
"learning_rate": 2e-05,
"loss": 0.5201643466949463,
"step": 1610
},
{
"epoch": 0.11614568396902782,
"grad_norm": 2.81574948493408,
"learning_rate": 2e-05,
"loss": 0.5578659057617188,
"step": 1620
},
{
"epoch": 0.1168626326355033,
"grad_norm": 5.40288732260851,
"learning_rate": 2e-05,
"loss": 0.5688968658447265,
"step": 1630
},
{
"epoch": 0.11757958130197878,
"grad_norm": 2.2012274195157153,
"learning_rate": 2e-05,
"loss": 0.5558701515197754,
"step": 1640
},
{
"epoch": 0.11829652996845426,
"grad_norm": 1.730570521948697,
"learning_rate": 2e-05,
"loss": 0.5515780448913574,
"step": 1650
},
{
"epoch": 0.11901347863492974,
"grad_norm": 3.083940659262359,
"learning_rate": 2e-05,
"loss": 0.5680615901947021,
"step": 1660
},
{
"epoch": 0.11973042730140522,
"grad_norm": 1.8293486250460052,
"learning_rate": 2e-05,
"loss": 0.5579127311706543,
"step": 1670
},
{
"epoch": 0.1204473759678807,
"grad_norm": 2.691528194596175,
"learning_rate": 2e-05,
"loss": 0.5666030406951904,
"step": 1680
},
{
"epoch": 0.12116432463435618,
"grad_norm": 2.3017548820591376,
"learning_rate": 2e-05,
"loss": 0.5542361259460449,
"step": 1690
},
{
"epoch": 0.12188127330083166,
"grad_norm": 1.3982396451314711,
"learning_rate": 2e-05,
"loss": 0.5412776470184326,
"step": 1700
},
{
"epoch": 0.12259822196730714,
"grad_norm": 2.8633211849346836,
"learning_rate": 2e-05,
"loss": 0.5596530914306641,
"step": 1710
},
{
"epoch": 0.12331517063378263,
"grad_norm": 1.2756481909481026,
"learning_rate": 2e-05,
"loss": 0.5272120475769043,
"step": 1720
},
{
"epoch": 0.1240321193002581,
"grad_norm": 1.884108459767808,
"learning_rate": 2e-05,
"loss": 0.5613333702087402,
"step": 1730
},
{
"epoch": 0.12474906796673359,
"grad_norm": 5.273226308180893,
"learning_rate": 2e-05,
"loss": 0.5628758430480957,
"step": 1740
},
{
"epoch": 0.12546601663320905,
"grad_norm": 3.0995531259526348,
"learning_rate": 2e-05,
"loss": 0.5465875148773194,
"step": 1750
},
{
"epoch": 0.12618296529968454,
"grad_norm": 3.2626826548420476,
"learning_rate": 2e-05,
"loss": 0.5627561092376709,
"step": 1760
},
{
"epoch": 0.12689991396616002,
"grad_norm": 2.364412227942818,
"learning_rate": 2e-05,
"loss": 0.5472114562988282,
"step": 1770
},
{
"epoch": 0.1276168626326355,
"grad_norm": 2.4263037850994746,
"learning_rate": 2e-05,
"loss": 0.5574633598327636,
"step": 1780
},
{
"epoch": 0.12833381129911098,
"grad_norm": 2.309082260471337,
"learning_rate": 2e-05,
"loss": 0.5340644836425781,
"step": 1790
},
{
"epoch": 0.12905075996558646,
"grad_norm": 2.426509513022813,
"learning_rate": 2e-05,
"loss": 0.5799418926239014,
"step": 1800
},
{
"epoch": 0.12976770863206194,
"grad_norm": 1.8704066062154519,
"learning_rate": 2e-05,
"loss": 0.5632325649261475,
"step": 1810
},
{
"epoch": 0.13048465729853742,
"grad_norm": 1.406035471805342,
"learning_rate": 2e-05,
"loss": 0.5656003952026367,
"step": 1820
},
{
"epoch": 0.1312016059650129,
"grad_norm": 1.544336567362451,
"learning_rate": 2e-05,
"loss": 0.5451941967010498,
"step": 1830
},
{
"epoch": 0.13191855463148838,
"grad_norm": 2.0799351359110485,
"learning_rate": 2e-05,
"loss": 0.5621081352233886,
"step": 1840
},
{
"epoch": 0.13263550329796386,
"grad_norm": 1.7299640932134708,
"learning_rate": 2e-05,
"loss": 0.5848733425140381,
"step": 1850
},
{
"epoch": 0.13335245196443934,
"grad_norm": 1.1961375728783252,
"learning_rate": 2e-05,
"loss": 0.5659356117248535,
"step": 1860
},
{
"epoch": 0.13406940063091483,
"grad_norm": 2.9075264586093783,
"learning_rate": 2e-05,
"loss": 0.5509964942932128,
"step": 1870
},
{
"epoch": 0.1347863492973903,
"grad_norm": 1.3125243339991093,
"learning_rate": 2e-05,
"loss": 0.5528433799743653,
"step": 1880
},
{
"epoch": 0.1355032979638658,
"grad_norm": 1.3727193915645148,
"learning_rate": 2e-05,
"loss": 0.564556884765625,
"step": 1890
},
{
"epoch": 0.13622024663034127,
"grad_norm": 3.7953721625845933,
"learning_rate": 2e-05,
"loss": 0.5450845718383789,
"step": 1900
},
{
"epoch": 0.13693719529681675,
"grad_norm": 0.9639735389876055,
"learning_rate": 2e-05,
"loss": 0.5541686534881591,
"step": 1910
},
{
"epoch": 0.13765414396329223,
"grad_norm": 1.003569244043582,
"learning_rate": 2e-05,
"loss": 0.5625730514526367,
"step": 1920
},
{
"epoch": 0.1383710926297677,
"grad_norm": 0.8880655840375095,
"learning_rate": 2e-05,
"loss": 0.5609973907470703,
"step": 1930
},
{
"epoch": 0.1390880412962432,
"grad_norm": 1.8013979122929717,
"learning_rate": 2e-05,
"loss": 0.5645060539245605,
"step": 1940
},
{
"epoch": 0.13980498996271867,
"grad_norm": 1.534112281642532,
"learning_rate": 2e-05,
"loss": 0.5428046226501465,
"step": 1950
},
{
"epoch": 0.14052193862919415,
"grad_norm": 3.630130753040599,
"learning_rate": 2e-05,
"loss": 0.5524575233459472,
"step": 1960
},
{
"epoch": 0.14123888729566964,
"grad_norm": 1.334941948733717,
"learning_rate": 2e-05,
"loss": 0.5768626213073731,
"step": 1970
},
{
"epoch": 0.14195583596214512,
"grad_norm": 6.284844482295266,
"learning_rate": 2e-05,
"loss": 0.5756556987762451,
"step": 1980
},
{
"epoch": 0.1426727846286206,
"grad_norm": 2.7603549710378648,
"learning_rate": 2e-05,
"loss": 0.5478507995605468,
"step": 1990
},
{
"epoch": 0.14338973329509608,
"grad_norm": 2.711164302375169,
"learning_rate": 2e-05,
"loss": 0.5454816341400146,
"step": 2000
},
{
"epoch": 0.14410668196157156,
"grad_norm": 1.6373091053111855,
"learning_rate": 2e-05,
"loss": 0.56094331741333,
"step": 2010
},
{
"epoch": 0.14482363062804704,
"grad_norm": 1.3231963543618217,
"learning_rate": 2e-05,
"loss": 0.5709875106811524,
"step": 2020
},
{
"epoch": 0.14554057929452252,
"grad_norm": 2.6013946026230177,
"learning_rate": 2e-05,
"loss": 0.5760745048522949,
"step": 2030
},
{
"epoch": 0.146257527960998,
"grad_norm": 1.0760891963639665,
"learning_rate": 2e-05,
"loss": 0.540398359298706,
"step": 2040
},
{
"epoch": 0.14697447662747348,
"grad_norm": 3.806990207982052,
"learning_rate": 2e-05,
"loss": 0.5452116966247559,
"step": 2050
},
{
"epoch": 0.14769142529394896,
"grad_norm": 0.8770094294050765,
"learning_rate": 2e-05,
"loss": 0.516285514831543,
"step": 2060
},
{
"epoch": 0.14840837396042444,
"grad_norm": 1.64509283352414,
"learning_rate": 2e-05,
"loss": 0.5785709381103515,
"step": 2070
},
{
"epoch": 0.14912532262689993,
"grad_norm": 1.0501298033510167,
"learning_rate": 2e-05,
"loss": 0.5674317836761474,
"step": 2080
},
{
"epoch": 0.1498422712933754,
"grad_norm": 1.0810242187793186,
"learning_rate": 2e-05,
"loss": 0.5573361873626709,
"step": 2090
},
{
"epoch": 0.1505592199598509,
"grad_norm": 1.3235649158062135,
"learning_rate": 2e-05,
"loss": 0.5721589088439941,
"step": 2100
},
{
"epoch": 0.15127616862632637,
"grad_norm": 2.61075713737253,
"learning_rate": 2e-05,
"loss": 0.5381807327270508,
"step": 2110
},
{
"epoch": 0.15199311729280185,
"grad_norm": 2.0071819238060655,
"learning_rate": 2e-05,
"loss": 0.5428504943847656,
"step": 2120
},
{
"epoch": 0.1527100659592773,
"grad_norm": 1.9874329432228202,
"learning_rate": 2e-05,
"loss": 0.5540834426879883,
"step": 2130
},
{
"epoch": 0.15342701462575278,
"grad_norm": 1.0751634954880047,
"learning_rate": 2e-05,
"loss": 0.5777225494384766,
"step": 2140
},
{
"epoch": 0.15414396329222826,
"grad_norm": 2.247684458013884,
"learning_rate": 2e-05,
"loss": 0.5361541748046875,
"step": 2150
},
{
"epoch": 0.15486091195870375,
"grad_norm": 1.5069655343366213,
"learning_rate": 2e-05,
"loss": 0.5514012336730957,
"step": 2160
},
{
"epoch": 0.15557786062517923,
"grad_norm": 1.4677351079218146,
"learning_rate": 2e-05,
"loss": 0.5583604335784912,
"step": 2170
},
{
"epoch": 0.1562948092916547,
"grad_norm": 1.1533300901460213,
"learning_rate": 2e-05,
"loss": 0.5542792797088623,
"step": 2180
},
{
"epoch": 0.1570117579581302,
"grad_norm": 1.7028225571829634,
"learning_rate": 2e-05,
"loss": 0.5386368751525878,
"step": 2190
},
{
"epoch": 0.15772870662460567,
"grad_norm": 1.3309927566018185,
"learning_rate": 2e-05,
"loss": 0.5358123302459716,
"step": 2200
},
{
"epoch": 0.15844565529108115,
"grad_norm": 8.920239112234068,
"learning_rate": 2e-05,
"loss": 0.5610211849212646,
"step": 2210
},
{
"epoch": 0.15916260395755663,
"grad_norm": 0.9909451029281341,
"learning_rate": 2e-05,
"loss": 0.5076250076293946,
"step": 2220
},
{
"epoch": 0.1598795526240321,
"grad_norm": 1.2993026276344535,
"learning_rate": 2e-05,
"loss": 0.540493392944336,
"step": 2230
},
{
"epoch": 0.1605965012905076,
"grad_norm": 2.3512466519648587,
"learning_rate": 2e-05,
"loss": 0.5421253204345703,
"step": 2240
},
{
"epoch": 0.16131344995698307,
"grad_norm": 3.215608826413565,
"learning_rate": 2e-05,
"loss": 0.5742346763610839,
"step": 2250
},
{
"epoch": 0.16203039862345855,
"grad_norm": 2.1540313753980445,
"learning_rate": 2e-05,
"loss": 0.5484731674194336,
"step": 2260
},
{
"epoch": 0.16274734728993404,
"grad_norm": 1.844563278680931,
"learning_rate": 2e-05,
"loss": 0.5414963722229004,
"step": 2270
},
{
"epoch": 0.16346429595640952,
"grad_norm": 1.2178930369073198,
"learning_rate": 2e-05,
"loss": 0.5368363380432128,
"step": 2280
},
{
"epoch": 0.164181244622885,
"grad_norm": 0.8834306050985979,
"learning_rate": 2e-05,
"loss": 0.5298836708068848,
"step": 2290
},
{
"epoch": 0.16489819328936048,
"grad_norm": 1.5799606162459594,
"learning_rate": 2e-05,
"loss": 0.5445155143737793,
"step": 2300
},
{
"epoch": 0.16561514195583596,
"grad_norm": 1.106839067943099,
"learning_rate": 2e-05,
"loss": 0.5452072143554687,
"step": 2310
},
{
"epoch": 0.16633209062231144,
"grad_norm": 2.598513025278755,
"learning_rate": 2e-05,
"loss": 0.5509998321533203,
"step": 2320
},
{
"epoch": 0.16704903928878692,
"grad_norm": 2.825348671573568,
"learning_rate": 2e-05,
"loss": 0.5631167411804199,
"step": 2330
},
{
"epoch": 0.1677659879552624,
"grad_norm": 1.7145872571500615,
"learning_rate": 2e-05,
"loss": 0.5552741050720215,
"step": 2340
},
{
"epoch": 0.16848293662173788,
"grad_norm": 14.03401035026617,
"learning_rate": 2e-05,
"loss": 0.5463265895843505,
"step": 2350
},
{
"epoch": 0.16919988528821336,
"grad_norm": 2.0847220323880165,
"learning_rate": 2e-05,
"loss": 0.5252552986145019,
"step": 2360
},
{
"epoch": 0.16991683395468885,
"grad_norm": 1.011338519703196,
"learning_rate": 2e-05,
"loss": 0.5438662528991699,
"step": 2370
},
{
"epoch": 0.17063378262116433,
"grad_norm": 1.1683565008517496,
"learning_rate": 2e-05,
"loss": 0.5535166740417481,
"step": 2380
},
{
"epoch": 0.1713507312876398,
"grad_norm": 5.08246929812119,
"learning_rate": 2e-05,
"loss": 0.5406145095825196,
"step": 2390
},
{
"epoch": 0.1720676799541153,
"grad_norm": 1.088285394309193,
"learning_rate": 2e-05,
"loss": 0.5489974021911621,
"step": 2400
},
{
"epoch": 0.17278462862059077,
"grad_norm": 0.7764690731022792,
"learning_rate": 2e-05,
"loss": 0.5712192058563232,
"step": 2410
},
{
"epoch": 0.17350157728706625,
"grad_norm": 1.4063769617297237,
"learning_rate": 2e-05,
"loss": 0.5649506568908691,
"step": 2420
},
{
"epoch": 0.17421852595354173,
"grad_norm": 0.8841185813513404,
"learning_rate": 2e-05,
"loss": 0.5460720062255859,
"step": 2430
},
{
"epoch": 0.1749354746200172,
"grad_norm": 1.0771999567027115,
"learning_rate": 2e-05,
"loss": 0.568940544128418,
"step": 2440
},
{
"epoch": 0.1756524232864927,
"grad_norm": 1.2155770616956478,
"learning_rate": 2e-05,
"loss": 0.5199066638946533,
"step": 2450
},
{
"epoch": 0.17636937195296817,
"grad_norm": 1.677688151590649,
"learning_rate": 2e-05,
"loss": 0.556132459640503,
"step": 2460
},
{
"epoch": 0.17708632061944365,
"grad_norm": 0.9244792122316684,
"learning_rate": 2e-05,
"loss": 0.5273470878601074,
"step": 2470
},
{
"epoch": 0.17780326928591914,
"grad_norm": 4.24012694050476,
"learning_rate": 2e-05,
"loss": 0.5406831741333008,
"step": 2480
},
{
"epoch": 0.17852021795239462,
"grad_norm": 0.9989434456238647,
"learning_rate": 2e-05,
"loss": 0.5287195682525635,
"step": 2490
},
{
"epoch": 0.1792371666188701,
"grad_norm": 1.2294548845023252,
"learning_rate": 2e-05,
"loss": 0.5238923072814942,
"step": 2500
},
{
"epoch": 0.17995411528534558,
"grad_norm": 1.1680245691471733,
"learning_rate": 2e-05,
"loss": 0.5360855102539063,
"step": 2510
},
{
"epoch": 0.18067106395182106,
"grad_norm": 3.259902952148628,
"learning_rate": 2e-05,
"loss": 0.5570301055908203,
"step": 2520
},
{
"epoch": 0.18138801261829654,
"grad_norm": 2.790265316658619,
"learning_rate": 2e-05,
"loss": 0.532185411453247,
"step": 2530
},
{
"epoch": 0.18210496128477202,
"grad_norm": 1.2744671754626793,
"learning_rate": 2e-05,
"loss": 0.5266745567321778,
"step": 2540
},
{
"epoch": 0.1828219099512475,
"grad_norm": 0.8706250695378648,
"learning_rate": 1.997261400182901e-05,
"loss": 0.5322741780962262,
"step": 2550
},
{
"epoch": 0.18353885861772298,
"grad_norm": 0.8937536584310218,
"learning_rate": 1.9805800746810423e-05,
"loss": 0.5236109256744385,
"step": 2560
},
{
"epoch": 0.18425580728419846,
"grad_norm": 1.5009208180411513,
"learning_rate": 1.9489934220110274e-05,
"loss": 0.5552047729492188,
"step": 2570
},
{
"epoch": 0.18497275595067394,
"grad_norm": 1.067878655582292,
"learning_rate": 1.9029840364820392e-05,
"loss": 0.5710633754730224,
"step": 2580
},
{
"epoch": 0.18568970461714943,
"grad_norm": 1.7367892947296495,
"learning_rate": 1.8432548690479908e-05,
"loss": 0.5327499866485595,
"step": 2590
},
{
"epoch": 0.18640665328362488,
"grad_norm": 1.0611595778640688,
"learning_rate": 1.7707184873239772e-05,
"loss": 0.5528162956237793,
"step": 2600
},
{
"epoch": 0.18712360195010036,
"grad_norm": 1.1654287237288545,
"learning_rate": 1.6864831329902565e-05,
"loss": 0.565640926361084,
"step": 2610
},
{
"epoch": 0.18784055061657584,
"grad_norm": 1.0389015244604005,
"learning_rate": 1.591835789604667e-05,
"loss": 0.5381675243377686,
"step": 2620
},
{
"epoch": 0.18855749928305132,
"grad_norm": 1.5200728952859153,
"learning_rate": 1.4882225195206459e-05,
"loss": 0.5736266136169433,
"step": 2630
},
{
"epoch": 0.1892744479495268,
"grad_norm": 7.39050153722471,
"learning_rate": 1.3772263703317695e-05,
"loss": 0.5384951591491699,
"step": 2640
},
{
"epoch": 0.18999139661600228,
"grad_norm": 0.970839880455912,
"learning_rate": 1.2605431883975547e-05,
"loss": 0.5098603725433349,
"step": 2650
},
{
"epoch": 0.19070834528247776,
"grad_norm": 1.1576885219177286,
"learning_rate": 1.1399557089817574e-05,
"loss": 0.5423776149749756,
"step": 2660
},
{
"epoch": 0.19142529394895325,
"grad_norm": 0.8740188779473159,
"learning_rate": 1.0173063188650463e-05,
"loss": 0.5217263221740722,
"step": 2670
},
{
"epoch": 0.19214224261542873,
"grad_norm": 1.5148734130782993,
"learning_rate": 8.944689075764396e-06,
"loss": 0.5449352264404297,
"step": 2680
},
{
"epoch": 0.1928591912819042,
"grad_norm": 0.7627398016125817,
"learning_rate": 7.73320237312362e-06,
"loss": 0.5438205242156983,
"step": 2690
},
{
"epoch": 0.1935761399483797,
"grad_norm": 1.7268579058495153,
"learning_rate": 6.557112689658629e-06,
"loss": 0.5547382354736328,
"step": 2700
},
{
"epoch": 0.19429308861485517,
"grad_norm": 1.02765232265623,
"learning_rate": 5.434388823591213e-06,
"loss": 0.5238811492919921,
"step": 2710
},
{
"epoch": 0.19501003728133065,
"grad_norm": 1.0506951513608977,
"learning_rate": 4.382184227495424e-06,
"loss": 0.5501854896545411,
"step": 2720
},
{
"epoch": 0.19572698594780613,
"grad_norm": 0.6349682788251014,
"learning_rate": 3.4165749305560203e-06,
"loss": 0.5193760871887207,
"step": 2730
},
{
"epoch": 0.1964439346142816,
"grad_norm": 1.4296042716832305,
"learning_rate": 2.5523139221595556e-06,
"loss": 0.5402659416198731,
"step": 2740
},
{
"epoch": 0.1971608832807571,
"grad_norm": 1.2605831697909553,
"learning_rate": 1.8026057494499821e-06,
"loss": 0.5175878047943115,
"step": 2750
},
{
"epoch": 0.19787783194723257,
"grad_norm": 0.8273072980221112,
"learning_rate": 1.178904772643223e-06,
"loss": 0.530870771408081,
"step": 2760
},
{
"epoch": 0.19859478061370806,
"grad_norm": 0.8054812926865942,
"learning_rate": 6.90740160441754e-07,
"loss": 0.5352423667907715,
"step": 2770
},
{
"epoch": 0.19931172928018354,
"grad_norm": 1.6163829981140836,
"learning_rate": 3.45570299343632e-07,
"loss": 0.5283924102783203,
"step": 2780
},
{
"epoch": 0.20002867794665902,
"grad_norm": 1.202354355046297,
"learning_rate": 1.486688412425152e-07,
"loss": 0.5187806129455567,
"step": 2790
},
{
"epoch": 0.20053054201319184,
"step": 2797,
"total_flos": 1324435843842048.0,
"train_loss": 0.04913015948647809,
"train_runtime": 3664.7518,
"train_samples_per_second": 9.159,
"train_steps_per_second": 0.763
}
],
"logging_steps": 10,
"max_steps": 2797,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2543,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1324435843842048.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}