rv2-gemma12b-ov_A_shared / 400M /trainer_state.json
rqadri's picture
Upload folder using huggingface_hub
5d91d48 verified
Raw History Blame Contribute Delete
70.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.5215274218349565,
"eval_steps": 500,
"global_step": 4070,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0012813941568426447,
"grad_norm": 48.60657259177163,
"learning_rate": 2.910112359550562e-06,
"loss": 1.136702537536621,
"step": 10
},
{
"epoch": 0.0025627883136852894,
"grad_norm": 91.91822214983361,
"learning_rate": 3.921348314606742e-06,
"loss": 0.7458788871765136,
"step": 20
},
{
"epoch": 0.0038441824705279346,
"grad_norm": 26.46552787500785,
"learning_rate": 4.932584269662922e-06,
"loss": 0.640007734298706,
"step": 30
},
{
"epoch": 0.005125576627370579,
"grad_norm": 22.809848983257012,
"learning_rate": 5.943820224719102e-06,
"loss": 0.6162827491760254,
"step": 40
},
{
"epoch": 0.006406970784213224,
"grad_norm": 29.62460709017976,
"learning_rate": 6.955056179775282e-06,
"loss": 0.5786398887634278,
"step": 50
},
{
"epoch": 0.007688364941055869,
"grad_norm": 16.691100627355937,
"learning_rate": 7.966292134831462e-06,
"loss": 0.5697914123535156,
"step": 60
},
{
"epoch": 0.008969759097898513,
"grad_norm": 8.744409243854706,
"learning_rate": 8.977528089887641e-06,
"loss": 0.5806342124938965,
"step": 70
},
{
"epoch": 0.010251153254741158,
"grad_norm": 7.582330015247281,
"learning_rate": 9.988764044943822e-06,
"loss": 0.5879335403442383,
"step": 80
},
{
"epoch": 0.011532547411583802,
"grad_norm": 79.23406804551544,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.5835853576660156,
"step": 90
},
{
"epoch": 0.012813941568426449,
"grad_norm": 310.9497155099939,
"learning_rate": 1.2011235955056182e-05,
"loss": 0.5906369686126709,
"step": 100
},
{
"epoch": 0.014095335725269094,
"grad_norm": 36.11222696404999,
"learning_rate": 1.3022471910112362e-05,
"loss": 0.5857430458068847,
"step": 110
},
{
"epoch": 0.015376729882111738,
"grad_norm": 6.481841818631169,
"learning_rate": 1.403370786516854e-05,
"loss": 0.5944482326507569,
"step": 120
},
{
"epoch": 0.01665812403895438,
"grad_norm": 21.67855790138833,
"learning_rate": 1.504494382022472e-05,
"loss": 0.5830103874206543,
"step": 130
},
{
"epoch": 0.017939518195797026,
"grad_norm": 8.15350712038059,
"learning_rate": 1.60561797752809e-05,
"loss": 0.6037096977233887,
"step": 140
},
{
"epoch": 0.01922091235263967,
"grad_norm": 33.51394138576384,
"learning_rate": 1.706741573033708e-05,
"loss": 0.5845141410827637,
"step": 150
},
{
"epoch": 0.020502306509482315,
"grad_norm": 5.233920603807281,
"learning_rate": 1.8078651685393256e-05,
"loss": 0.5999966621398926,
"step": 160
},
{
"epoch": 0.02178370066632496,
"grad_norm": 6.387467489565004,
"learning_rate": 1.908988764044944e-05,
"loss": 0.5866672039031983,
"step": 170
},
{
"epoch": 0.023065094823167605,
"grad_norm": 3.3507330659662147,
"learning_rate": 2e-05,
"loss": 0.5995364189147949,
"step": 180
},
{
"epoch": 0.02434648898001025,
"grad_norm": 7.947338320392441,
"learning_rate": 2e-05,
"loss": 0.5478545188903808,
"step": 190
},
{
"epoch": 0.025627883136852898,
"grad_norm": 16.041227689012175,
"learning_rate": 2e-05,
"loss": 0.585174560546875,
"step": 200
},
{
"epoch": 0.026909277293695542,
"grad_norm": 26.708451549556067,
"learning_rate": 2e-05,
"loss": 0.6306365966796875,
"step": 210
},
{
"epoch": 0.028190671450538187,
"grad_norm": 43.434970417427316,
"learning_rate": 2e-05,
"loss": 0.574945592880249,
"step": 220
},
{
"epoch": 0.029472065607380832,
"grad_norm": 7.745227881348918,
"learning_rate": 2e-05,
"loss": 0.6046634674072265,
"step": 230
},
{
"epoch": 0.030753459764223477,
"grad_norm": 5.494753079256009,
"learning_rate": 2e-05,
"loss": 0.5993133544921875,
"step": 240
},
{
"epoch": 0.03203485392106612,
"grad_norm": 10.94252347649759,
"learning_rate": 2e-05,
"loss": 0.6251495361328125,
"step": 250
},
{
"epoch": 0.03331624807790876,
"grad_norm": 3.8515593208507175,
"learning_rate": 2e-05,
"loss": 0.6059386730194092,
"step": 260
},
{
"epoch": 0.03459764223475141,
"grad_norm": 13.624895012129093,
"learning_rate": 2e-05,
"loss": 0.6095015525817871,
"step": 270
},
{
"epoch": 0.03587903639159405,
"grad_norm": 92.4144274779054,
"learning_rate": 2e-05,
"loss": 0.6274710655212402,
"step": 280
},
{
"epoch": 0.0371604305484367,
"grad_norm": 63.87797936144395,
"learning_rate": 2e-05,
"loss": 0.6483458995819091,
"step": 290
},
{
"epoch": 0.03844182470527934,
"grad_norm": 29.665399521499776,
"learning_rate": 2e-05,
"loss": 0.6508570671081543,
"step": 300
},
{
"epoch": 0.03972321886212199,
"grad_norm": 10.143942268074744,
"learning_rate": 2e-05,
"loss": 0.6228761196136474,
"step": 310
},
{
"epoch": 0.04100461301896463,
"grad_norm": 3.3645822727750283,
"learning_rate": 2e-05,
"loss": 0.5910223960876465,
"step": 320
},
{
"epoch": 0.04228600717580728,
"grad_norm": 2.326932296638607,
"learning_rate": 2e-05,
"loss": 0.5986968994140625,
"step": 330
},
{
"epoch": 0.04356740133264992,
"grad_norm": 19.774971839868684,
"learning_rate": 2e-05,
"loss": 0.5774398803710937,
"step": 340
},
{
"epoch": 0.04484879548949257,
"grad_norm": 8.625513047621965,
"learning_rate": 2e-05,
"loss": 0.5988794326782226,
"step": 350
},
{
"epoch": 0.04613018964633521,
"grad_norm": 2.6493447545498943,
"learning_rate": 2e-05,
"loss": 0.5667895793914794,
"step": 360
},
{
"epoch": 0.04741158380317786,
"grad_norm": 4.063128105930308,
"learning_rate": 2e-05,
"loss": 0.5583474636077881,
"step": 370
},
{
"epoch": 0.0486929779600205,
"grad_norm": 1.7332087084494128,
"learning_rate": 2e-05,
"loss": 0.5471595287322998,
"step": 380
},
{
"epoch": 0.04997437211686315,
"grad_norm": 3.3400257212079927,
"learning_rate": 2e-05,
"loss": 0.5477984428405762,
"step": 390
},
{
"epoch": 0.051255766273705795,
"grad_norm": 5.278064316114456,
"learning_rate": 2e-05,
"loss": 0.5472843170166015,
"step": 400
},
{
"epoch": 0.05253716043054844,
"grad_norm": 2.074410435293684,
"learning_rate": 2e-05,
"loss": 0.5686273097991943,
"step": 410
},
{
"epoch": 0.053818554587391085,
"grad_norm": 68.09645975955011,
"learning_rate": 2e-05,
"loss": 0.5544178009033203,
"step": 420
},
{
"epoch": 0.055099948744233726,
"grad_norm": 525.6925356969697,
"learning_rate": 2e-05,
"loss": 0.5655110359191895,
"step": 430
},
{
"epoch": 0.056381342901076374,
"grad_norm": 22.424679914814075,
"learning_rate": 2e-05,
"loss": 0.59327712059021,
"step": 440
},
{
"epoch": 0.057662737057919015,
"grad_norm": 5.444309007850252,
"learning_rate": 2e-05,
"loss": 0.5968106746673584,
"step": 450
},
{
"epoch": 0.058944131214761664,
"grad_norm": 2.903284182345692,
"learning_rate": 2e-05,
"loss": 0.5818487167358398,
"step": 460
},
{
"epoch": 0.060225525371604305,
"grad_norm": 4.400504161764957,
"learning_rate": 2e-05,
"loss": 0.5862027645111084,
"step": 470
},
{
"epoch": 0.06150691952844695,
"grad_norm": 2.5317111608646763,
"learning_rate": 2e-05,
"loss": 0.5865765571594238,
"step": 480
},
{
"epoch": 0.0627883136852896,
"grad_norm": 7.6823324809313185,
"learning_rate": 2e-05,
"loss": 0.5774165153503418,
"step": 490
},
{
"epoch": 0.06406970784213224,
"grad_norm": 3.4342003064347866,
"learning_rate": 2e-05,
"loss": 0.5581344604492188,
"step": 500
},
{
"epoch": 0.06535110199897488,
"grad_norm": 3.4940869527673684,
"learning_rate": 2e-05,
"loss": 0.5384406566619873,
"step": 510
},
{
"epoch": 0.06663249615581752,
"grad_norm": 22.622831777103364,
"learning_rate": 2e-05,
"loss": 0.5864431381225585,
"step": 520
},
{
"epoch": 0.06791389031266018,
"grad_norm": 8.952483502466984,
"learning_rate": 2e-05,
"loss": 0.5497615814208985,
"step": 530
},
{
"epoch": 0.06919528446950282,
"grad_norm": 14.798724017806745,
"learning_rate": 2e-05,
"loss": 0.591427230834961,
"step": 540
},
{
"epoch": 0.07047667862634546,
"grad_norm": 6.240194593748034,
"learning_rate": 2e-05,
"loss": 0.5721195220947266,
"step": 550
},
{
"epoch": 0.0717580727831881,
"grad_norm": 1.3836185042814684,
"learning_rate": 2e-05,
"loss": 0.5673149108886719,
"step": 560
},
{
"epoch": 0.07303946694003076,
"grad_norm": 2.3358117255287776,
"learning_rate": 2e-05,
"loss": 0.5625670433044434,
"step": 570
},
{
"epoch": 0.0743208610968734,
"grad_norm": 1.625469477847444,
"learning_rate": 2e-05,
"loss": 0.5684123992919922,
"step": 580
},
{
"epoch": 0.07560225525371604,
"grad_norm": 1.133401896469163,
"learning_rate": 2e-05,
"loss": 0.5579514503479004,
"step": 590
},
{
"epoch": 0.07688364941055868,
"grad_norm": 1.6347916075043618,
"learning_rate": 2e-05,
"loss": 0.5497482299804688,
"step": 600
},
{
"epoch": 0.07816504356740134,
"grad_norm": 4.31001001862141,
"learning_rate": 2e-05,
"loss": 0.5504285335540772,
"step": 610
},
{
"epoch": 0.07944643772424398,
"grad_norm": 1.6923695835210164,
"learning_rate": 2e-05,
"loss": 0.5389340877532959,
"step": 620
},
{
"epoch": 0.08072783188108662,
"grad_norm": 20.37920258837912,
"learning_rate": 2e-05,
"loss": 0.5571833610534668,
"step": 630
},
{
"epoch": 0.08200922603792926,
"grad_norm": 2.531475008780508,
"learning_rate": 2e-05,
"loss": 0.5565690994262695,
"step": 640
},
{
"epoch": 0.08329062019477192,
"grad_norm": 3.605544322163944,
"learning_rate": 2e-05,
"loss": 0.5658109188079834,
"step": 650
},
{
"epoch": 0.08457201435161456,
"grad_norm": 6.10159024054078,
"learning_rate": 2e-05,
"loss": 0.5663240432739258,
"step": 660
},
{
"epoch": 0.0858534085084572,
"grad_norm": 6.192844710780108,
"learning_rate": 2e-05,
"loss": 0.5819952964782715,
"step": 670
},
{
"epoch": 0.08713480266529984,
"grad_norm": 3.474212678080138,
"learning_rate": 2e-05,
"loss": 0.5608247756958008,
"step": 680
},
{
"epoch": 0.0884161968221425,
"grad_norm": 31.397865343442433,
"learning_rate": 2e-05,
"loss": 0.55595703125,
"step": 690
},
{
"epoch": 0.08969759097898514,
"grad_norm": 3.3902647864069437,
"learning_rate": 2e-05,
"loss": 0.5607826232910156,
"step": 700
},
{
"epoch": 0.09097898513582778,
"grad_norm": 2.6130800788890887,
"learning_rate": 2e-05,
"loss": 0.5772881984710694,
"step": 710
},
{
"epoch": 0.09226037929267042,
"grad_norm": 1.8412885445002003,
"learning_rate": 2e-05,
"loss": 0.5467967033386231,
"step": 720
},
{
"epoch": 0.09354177344951307,
"grad_norm": 13.878342002550012,
"learning_rate": 2e-05,
"loss": 0.5741470336914063,
"step": 730
},
{
"epoch": 0.09482316760635572,
"grad_norm": 7.91637564641484,
"learning_rate": 2e-05,
"loss": 0.5568782806396484,
"step": 740
},
{
"epoch": 0.09610456176319836,
"grad_norm": 14.994930545043719,
"learning_rate": 2e-05,
"loss": 0.5728705883026123,
"step": 750
},
{
"epoch": 0.097385955920041,
"grad_norm": 3.2399122356353556,
"learning_rate": 2e-05,
"loss": 0.5602639198303223,
"step": 760
},
{
"epoch": 0.09866735007688365,
"grad_norm": 8.435306389890595,
"learning_rate": 2e-05,
"loss": 0.5648305416107178,
"step": 770
},
{
"epoch": 0.0999487442337263,
"grad_norm": 2.7229485397202824,
"learning_rate": 2e-05,
"loss": 0.5501457214355469,
"step": 780
},
{
"epoch": 0.10123013839056894,
"grad_norm": 1.7845496569761272,
"learning_rate": 2e-05,
"loss": 0.5729188442230224,
"step": 790
},
{
"epoch": 0.10251153254741159,
"grad_norm": 6.7858062416328515,
"learning_rate": 2e-05,
"loss": 0.5560100555419922,
"step": 800
},
{
"epoch": 0.10379292670425423,
"grad_norm": 2.8769397871268003,
"learning_rate": 2e-05,
"loss": 0.5690321922302246,
"step": 810
},
{
"epoch": 0.10507432086109687,
"grad_norm": 20.514686157387473,
"learning_rate": 2e-05,
"loss": 0.5718609809875488,
"step": 820
},
{
"epoch": 0.10635571501793951,
"grad_norm": 3.2148909747572105,
"learning_rate": 2e-05,
"loss": 0.5593263149261475,
"step": 830
},
{
"epoch": 0.10763710917478217,
"grad_norm": 3.3909934965013178,
"learning_rate": 2e-05,
"loss": 0.5652659416198731,
"step": 840
},
{
"epoch": 0.10891850333162481,
"grad_norm": 2.8601258221724177,
"learning_rate": 2e-05,
"loss": 0.5684373855590821,
"step": 850
},
{
"epoch": 0.11019989748846745,
"grad_norm": 3.430890987769173,
"learning_rate": 2e-05,
"loss": 0.5800694465637207,
"step": 860
},
{
"epoch": 0.1114812916453101,
"grad_norm": 6.8521625837959785,
"learning_rate": 2e-05,
"loss": 0.5646146774291992,
"step": 870
},
{
"epoch": 0.11276268580215275,
"grad_norm": 2.167299561040544,
"learning_rate": 2e-05,
"loss": 0.5554808616638184,
"step": 880
},
{
"epoch": 0.11404407995899539,
"grad_norm": 12.81539845275287,
"learning_rate": 2e-05,
"loss": 0.5791922569274902,
"step": 890
},
{
"epoch": 0.11532547411583803,
"grad_norm": 2.976710005657373,
"learning_rate": 2e-05,
"loss": 0.6072131156921386,
"step": 900
},
{
"epoch": 0.11660686827268067,
"grad_norm": 10.849710017677005,
"learning_rate": 2e-05,
"loss": 0.5678472518920898,
"step": 910
},
{
"epoch": 0.11788826242952333,
"grad_norm": 2.4416306998839428,
"learning_rate": 2e-05,
"loss": 0.5731144905090332,
"step": 920
},
{
"epoch": 0.11916965658636597,
"grad_norm": 7.038648849245178,
"learning_rate": 2e-05,
"loss": 0.5526491165161133,
"step": 930
},
{
"epoch": 0.12045105074320861,
"grad_norm": 2.9470404844909095,
"learning_rate": 2e-05,
"loss": 0.5424953460693359,
"step": 940
},
{
"epoch": 0.12173244490005125,
"grad_norm": 9.454322797818554,
"learning_rate": 2e-05,
"loss": 0.548456335067749,
"step": 950
},
{
"epoch": 0.1230138390568939,
"grad_norm": 4.222464982835741,
"learning_rate": 2e-05,
"loss": 0.5253325462341308,
"step": 960
},
{
"epoch": 0.12429523321373655,
"grad_norm": 2.4647201408100443,
"learning_rate": 2e-05,
"loss": 0.5565387725830078,
"step": 970
},
{
"epoch": 0.1255766273705792,
"grad_norm": 5.646285894060971,
"learning_rate": 2e-05,
"loss": 1.1217697143554688,
"step": 980
},
{
"epoch": 0.12685802152742184,
"grad_norm": 2.964157436056934,
"learning_rate": 2e-05,
"loss": 0.5679878234863281,
"step": 990
},
{
"epoch": 0.12813941568426448,
"grad_norm": 6.288479464982876,
"learning_rate": 2e-05,
"loss": 0.5491109371185303,
"step": 1000
},
{
"epoch": 0.12942080984110713,
"grad_norm": 2.6763665966892654,
"learning_rate": 2e-05,
"loss": 0.5552764892578125,
"step": 1010
},
{
"epoch": 0.13070220399794977,
"grad_norm": 2.6351932907118973,
"learning_rate": 2e-05,
"loss": 0.5406902313232422,
"step": 1020
},
{
"epoch": 0.1319835981547924,
"grad_norm": 6.352666744995283,
"learning_rate": 2e-05,
"loss": 0.536877965927124,
"step": 1030
},
{
"epoch": 0.13326499231163505,
"grad_norm": 1.5421364933619635,
"learning_rate": 2e-05,
"loss": 0.5313173294067383,
"step": 1040
},
{
"epoch": 0.1345463864684777,
"grad_norm": 2.254141771785945,
"learning_rate": 2e-05,
"loss": 0.5454859733581543,
"step": 1050
},
{
"epoch": 0.13582778062532036,
"grad_norm": 1.6832085581211533,
"learning_rate": 2e-05,
"loss": 0.5299513816833497,
"step": 1060
},
{
"epoch": 0.137109174782163,
"grad_norm": 1.3101115317307876,
"learning_rate": 2e-05,
"loss": 0.5533650875091553,
"step": 1070
},
{
"epoch": 0.13839056893900564,
"grad_norm": 1.4250183862871697,
"learning_rate": 2e-05,
"loss": 0.518002986907959,
"step": 1080
},
{
"epoch": 0.13967196309584828,
"grad_norm": 3.106371141787246,
"learning_rate": 2e-05,
"loss": 0.540026044845581,
"step": 1090
},
{
"epoch": 0.14095335725269093,
"grad_norm": 1.2197064413226348,
"learning_rate": 2e-05,
"loss": 0.5478499889373779,
"step": 1100
},
{
"epoch": 0.14223475140953357,
"grad_norm": 6.522414224439747,
"learning_rate": 2e-05,
"loss": 0.5612347602844239,
"step": 1110
},
{
"epoch": 0.1435161455663762,
"grad_norm": 6.14600010534463,
"learning_rate": 2e-05,
"loss": 0.5392224788665771,
"step": 1120
},
{
"epoch": 0.14479753972321885,
"grad_norm": 1.3283775549672316,
"learning_rate": 2e-05,
"loss": 0.533447265625,
"step": 1130
},
{
"epoch": 0.14607893388006152,
"grad_norm": 1.5628798128196701,
"learning_rate": 2e-05,
"loss": 0.5403986930847168,
"step": 1140
},
{
"epoch": 0.14736032803690416,
"grad_norm": 5.557852174967083,
"learning_rate": 2e-05,
"loss": 0.5566314697265625,
"step": 1150
},
{
"epoch": 0.1486417221937468,
"grad_norm": 1.1771969176803327,
"learning_rate": 2e-05,
"loss": 0.5445056915283203,
"step": 1160
},
{
"epoch": 0.14992311635058944,
"grad_norm": 1.4048844272020151,
"learning_rate": 2e-05,
"loss": 0.5374322414398194,
"step": 1170
},
{
"epoch": 0.15120451050743208,
"grad_norm": 5.830084565815051,
"learning_rate": 2e-05,
"loss": 0.5391404628753662,
"step": 1180
},
{
"epoch": 0.15248590466427472,
"grad_norm": 1.4259555425294457,
"learning_rate": 2e-05,
"loss": 0.5523967742919922,
"step": 1190
},
{
"epoch": 0.15376729882111737,
"grad_norm": 2.3737683690736557,
"learning_rate": 2e-05,
"loss": 0.5384624481201172,
"step": 1200
},
{
"epoch": 0.15504869297796003,
"grad_norm": 2.628818495134982,
"learning_rate": 2e-05,
"loss": 0.5284938812255859,
"step": 1210
},
{
"epoch": 0.15633008713480268,
"grad_norm": 5.650114558533271,
"learning_rate": 2e-05,
"loss": 0.5476373195648193,
"step": 1220
},
{
"epoch": 0.15761148129164532,
"grad_norm": 2.344057641504692,
"learning_rate": 2e-05,
"loss": 0.5365348815917969,
"step": 1230
},
{
"epoch": 0.15889287544848796,
"grad_norm": 1.5391525983798033,
"learning_rate": 2e-05,
"loss": 0.5522804260253906,
"step": 1240
},
{
"epoch": 0.1601742696053306,
"grad_norm": 6.278578329314706,
"learning_rate": 2e-05,
"loss": 0.522200632095337,
"step": 1250
},
{
"epoch": 0.16145566376217324,
"grad_norm": 1.5191792185780197,
"learning_rate": 2e-05,
"loss": 0.5446555614471436,
"step": 1260
},
{
"epoch": 0.16273705791901588,
"grad_norm": 1.9845277263415022,
"learning_rate": 2e-05,
"loss": 0.5265367507934571,
"step": 1270
},
{
"epoch": 0.16401845207585852,
"grad_norm": 1.5282936674074756,
"learning_rate": 2e-05,
"loss": 0.5399628639221191,
"step": 1280
},
{
"epoch": 0.1652998462327012,
"grad_norm": 0.9895416727679237,
"learning_rate": 2e-05,
"loss": 0.5551435470581054,
"step": 1290
},
{
"epoch": 0.16658124038954383,
"grad_norm": 1.1540340702355871,
"learning_rate": 2e-05,
"loss": 0.5648162364959717,
"step": 1300
},
{
"epoch": 0.16786263454638647,
"grad_norm": 0.9095807077668226,
"learning_rate": 2e-05,
"loss": 0.5284963130950928,
"step": 1310
},
{
"epoch": 0.16914402870322912,
"grad_norm": 1.1906828334008543,
"learning_rate": 2e-05,
"loss": 0.5369954586029053,
"step": 1320
},
{
"epoch": 0.17042542286007176,
"grad_norm": 0.8909637934451952,
"learning_rate": 2e-05,
"loss": 0.5323962211608887,
"step": 1330
},
{
"epoch": 0.1717068170169144,
"grad_norm": 1.1346565661010557,
"learning_rate": 2e-05,
"loss": 0.5515767574310303,
"step": 1340
},
{
"epoch": 0.17298821117375704,
"grad_norm": 3.289182515818016,
"learning_rate": 2e-05,
"loss": 0.52701735496521,
"step": 1350
},
{
"epoch": 0.17426960533059968,
"grad_norm": 2.0880455303011325,
"learning_rate": 2e-05,
"loss": 0.5358563899993897,
"step": 1360
},
{
"epoch": 0.17555099948744235,
"grad_norm": 1.566832308941952,
"learning_rate": 2e-05,
"loss": 0.5285274028778076,
"step": 1370
},
{
"epoch": 0.176832393644285,
"grad_norm": 2.2990344953249022,
"learning_rate": 2e-05,
"loss": 0.5297482490539551,
"step": 1380
},
{
"epoch": 0.17811378780112763,
"grad_norm": 4.145571798823395,
"learning_rate": 2e-05,
"loss": 0.5191768169403076,
"step": 1390
},
{
"epoch": 0.17939518195797027,
"grad_norm": 1.7441796691857296,
"learning_rate": 2e-05,
"loss": 0.5435975074768067,
"step": 1400
},
{
"epoch": 0.18067657611481291,
"grad_norm": 2.6707491629206,
"learning_rate": 2e-05,
"loss": 0.5274767875671387,
"step": 1410
},
{
"epoch": 0.18195797027165556,
"grad_norm": 1.4710114789936155,
"learning_rate": 2e-05,
"loss": 0.5248068809509278,
"step": 1420
},
{
"epoch": 0.1832393644284982,
"grad_norm": 1.1661674872735948,
"learning_rate": 2e-05,
"loss": 0.5320616722106933,
"step": 1430
},
{
"epoch": 0.18452075858534084,
"grad_norm": 1.0201093941062644,
"learning_rate": 2e-05,
"loss": 0.5391462326049805,
"step": 1440
},
{
"epoch": 0.1858021527421835,
"grad_norm": 0.9609940685409813,
"learning_rate": 2e-05,
"loss": 0.51759934425354,
"step": 1450
},
{
"epoch": 0.18708354689902615,
"grad_norm": 4.4099586870345115,
"learning_rate": 2e-05,
"loss": 0.5610156059265137,
"step": 1460
},
{
"epoch": 0.1883649410558688,
"grad_norm": 1.2797704309832572,
"learning_rate": 2e-05,
"loss": 0.5134520530700684,
"step": 1470
},
{
"epoch": 0.18964633521271143,
"grad_norm": 1.604061991941183,
"learning_rate": 2e-05,
"loss": 0.5380812644958496,
"step": 1480
},
{
"epoch": 0.19092772936955407,
"grad_norm": 4.380080993954105,
"learning_rate": 2e-05,
"loss": 0.5286383152008056,
"step": 1490
},
{
"epoch": 0.1922091235263967,
"grad_norm": 0.9652036579027636,
"learning_rate": 2e-05,
"loss": 0.5347155570983887,
"step": 1500
},
{
"epoch": 0.19349051768323935,
"grad_norm": 1.0767282091811887,
"learning_rate": 2e-05,
"loss": 0.5506364822387695,
"step": 1510
},
{
"epoch": 0.194771911840082,
"grad_norm": 1.0660692222184038,
"learning_rate": 2e-05,
"loss": 0.5429599761962891,
"step": 1520
},
{
"epoch": 0.19605330599692466,
"grad_norm": 2.4245876116071563,
"learning_rate": 2e-05,
"loss": 0.530169677734375,
"step": 1530
},
{
"epoch": 0.1973347001537673,
"grad_norm": 1.6076349330268236,
"learning_rate": 2e-05,
"loss": 0.5128543853759766,
"step": 1540
},
{
"epoch": 0.19861609431060995,
"grad_norm": 3.2175817598128313,
"learning_rate": 2e-05,
"loss": 0.516771650314331,
"step": 1550
},
{
"epoch": 0.1998974884674526,
"grad_norm": 1.8454501672212902,
"learning_rate": 2e-05,
"loss": 0.555311918258667,
"step": 1560
},
{
"epoch": 0.20117888262429523,
"grad_norm": 1.1853740718254484,
"learning_rate": 2e-05,
"loss": 0.5157443046569824,
"step": 1570
},
{
"epoch": 0.20246027678113787,
"grad_norm": 1.482074792455346,
"learning_rate": 2e-05,
"loss": 0.554429006576538,
"step": 1580
},
{
"epoch": 0.2037416709379805,
"grad_norm": 0.9961893253106735,
"learning_rate": 2e-05,
"loss": 0.521328067779541,
"step": 1590
},
{
"epoch": 0.20502306509482318,
"grad_norm": 1.0692681388680267,
"learning_rate": 2e-05,
"loss": 0.5231988906860352,
"step": 1600
},
{
"epoch": 0.20630445925166582,
"grad_norm": 1.0289431510643638,
"learning_rate": 2e-05,
"loss": 0.5237535953521728,
"step": 1610
},
{
"epoch": 0.20758585340850846,
"grad_norm": 1.3099254730060057,
"learning_rate": 2e-05,
"loss": 0.5360143661499024,
"step": 1620
},
{
"epoch": 0.2088672475653511,
"grad_norm": 0.9703289709546258,
"learning_rate": 2e-05,
"loss": 0.5047122955322265,
"step": 1630
},
{
"epoch": 0.21014864172219375,
"grad_norm": 1.1996803050731473,
"learning_rate": 2e-05,
"loss": 0.5276361465454101,
"step": 1640
},
{
"epoch": 0.2114300358790364,
"grad_norm": 2.544046802772408,
"learning_rate": 2e-05,
"loss": 0.5131739139556885,
"step": 1650
},
{
"epoch": 0.21271143003587903,
"grad_norm": 3.025156971008793,
"learning_rate": 2e-05,
"loss": 0.5437355041503906,
"step": 1660
},
{
"epoch": 0.21399282419272167,
"grad_norm": 3.1281738525788203,
"learning_rate": 2e-05,
"loss": 0.5330821990966796,
"step": 1670
},
{
"epoch": 0.21527421834956434,
"grad_norm": 1.2180430188451103,
"learning_rate": 2e-05,
"loss": 0.5416089057922363,
"step": 1680
},
{
"epoch": 0.21655561250640698,
"grad_norm": 1.1470813594962164,
"learning_rate": 2e-05,
"loss": 0.539018201828003,
"step": 1690
},
{
"epoch": 0.21783700666324962,
"grad_norm": 3.6500797473607127,
"learning_rate": 2e-05,
"loss": 0.5521119594573974,
"step": 1700
},
{
"epoch": 0.21911840082009226,
"grad_norm": 2.067078070329206,
"learning_rate": 2e-05,
"loss": 0.5423431396484375,
"step": 1710
},
{
"epoch": 0.2203997949769349,
"grad_norm": 1.920293781550725,
"learning_rate": 2e-05,
"loss": 0.5424245357513428,
"step": 1720
},
{
"epoch": 0.22168118913377755,
"grad_norm": 14.544191040774015,
"learning_rate": 2e-05,
"loss": 0.5101382255554199,
"step": 1730
},
{
"epoch": 0.2229625832906202,
"grad_norm": 3.318513247980214,
"learning_rate": 2e-05,
"loss": 0.5420653343200683,
"step": 1740
},
{
"epoch": 0.22424397744746283,
"grad_norm": 1.1920262845160348,
"learning_rate": 2e-05,
"loss": 0.5346701622009278,
"step": 1750
},
{
"epoch": 0.2255253716043055,
"grad_norm": 1.3946653414472736,
"learning_rate": 2e-05,
"loss": 0.5356001853942871,
"step": 1760
},
{
"epoch": 0.22680676576114814,
"grad_norm": 1.0234843167991925,
"learning_rate": 2e-05,
"loss": 0.5313162326812744,
"step": 1770
},
{
"epoch": 0.22808815991799078,
"grad_norm": 2.2716514775718673,
"learning_rate": 2e-05,
"loss": 0.5260788917541503,
"step": 1780
},
{
"epoch": 0.22936955407483342,
"grad_norm": 1.1027690656722524,
"learning_rate": 2e-05,
"loss": 0.5265604019165039,
"step": 1790
},
{
"epoch": 0.23065094823167606,
"grad_norm": 0.8199711771574363,
"learning_rate": 2e-05,
"loss": 0.546507740020752,
"step": 1800
},
{
"epoch": 0.2319323423885187,
"grad_norm": 1.2314213846991644,
"learning_rate": 2e-05,
"loss": 0.5251349449157715,
"step": 1810
},
{
"epoch": 0.23321373654536134,
"grad_norm": 0.7783596277645343,
"learning_rate": 2e-05,
"loss": 0.5430421829223633,
"step": 1820
},
{
"epoch": 0.23449513070220399,
"grad_norm": 1.1146414292968791,
"learning_rate": 2e-05,
"loss": 0.5292141437530518,
"step": 1830
},
{
"epoch": 0.23577652485904665,
"grad_norm": 2.0137003356097765,
"learning_rate": 2e-05,
"loss": 0.5194851875305175,
"step": 1840
},
{
"epoch": 0.2370579190158893,
"grad_norm": 2.3397726475469534,
"learning_rate": 2e-05,
"loss": 0.5346482753753662,
"step": 1850
},
{
"epoch": 0.23833931317273194,
"grad_norm": 1.5663430221147905,
"learning_rate": 2e-05,
"loss": 0.5378023147583008,
"step": 1860
},
{
"epoch": 0.23962070732957458,
"grad_norm": 1.201294851008554,
"learning_rate": 2e-05,
"loss": 0.5159610271453857,
"step": 1870
},
{
"epoch": 0.24090210148641722,
"grad_norm": 1.0301545491772452,
"learning_rate": 2e-05,
"loss": 0.5299476623535156,
"step": 1880
},
{
"epoch": 0.24218349564325986,
"grad_norm": 0.8432389867329458,
"learning_rate": 2e-05,
"loss": 0.5182618618011474,
"step": 1890
},
{
"epoch": 0.2434648898001025,
"grad_norm": 0.8795423911191317,
"learning_rate": 2e-05,
"loss": 0.5129746437072754,
"step": 1900
},
{
"epoch": 0.24474628395694514,
"grad_norm": 0.8721039881044989,
"learning_rate": 2e-05,
"loss": 0.5283527374267578,
"step": 1910
},
{
"epoch": 0.2460276781137878,
"grad_norm": 3.1412433155534645,
"learning_rate": 2e-05,
"loss": 0.5297590732574463,
"step": 1920
},
{
"epoch": 0.24730907227063045,
"grad_norm": 1.0305392597354273,
"learning_rate": 2e-05,
"loss": 0.5080532550811767,
"step": 1930
},
{
"epoch": 0.2485904664274731,
"grad_norm": 1.0281646954720554,
"learning_rate": 2e-05,
"loss": 0.51605224609375,
"step": 1940
},
{
"epoch": 0.24987186058431574,
"grad_norm": 0.9629752110376899,
"learning_rate": 2e-05,
"loss": 0.5215667724609375,
"step": 1950
},
{
"epoch": 0.2511532547411584,
"grad_norm": 1.140557545820514,
"learning_rate": 2e-05,
"loss": 0.5421156883239746,
"step": 1960
},
{
"epoch": 0.25243464889800105,
"grad_norm": 1.0068158023619374,
"learning_rate": 2e-05,
"loss": 0.5088915348052978,
"step": 1970
},
{
"epoch": 0.2537160430548437,
"grad_norm": 0.9745366191237504,
"learning_rate": 2e-05,
"loss": 0.5463441371917724,
"step": 1980
},
{
"epoch": 0.25499743721168633,
"grad_norm": 1.2588237002272227,
"learning_rate": 2e-05,
"loss": 0.516777229309082,
"step": 1990
},
{
"epoch": 0.25627883136852897,
"grad_norm": 1.4490073257903515,
"learning_rate": 2e-05,
"loss": 0.51793212890625,
"step": 2000
},
{
"epoch": 0.2575602255253716,
"grad_norm": 3.963921685704575,
"learning_rate": 2e-05,
"loss": 0.5346158981323242,
"step": 2010
},
{
"epoch": 0.25884161968221425,
"grad_norm": 2.1713007316186417,
"learning_rate": 2e-05,
"loss": 0.5081641674041748,
"step": 2020
},
{
"epoch": 0.2601230138390569,
"grad_norm": 1.824815205654626,
"learning_rate": 2e-05,
"loss": 0.49935317039489746,
"step": 2030
},
{
"epoch": 0.26140440799589953,
"grad_norm": 0.9952189357081682,
"learning_rate": 2e-05,
"loss": 0.5178800582885742,
"step": 2040
},
{
"epoch": 0.2626858021527422,
"grad_norm": 4.16667101386579,
"learning_rate": 2e-05,
"loss": 0.529605770111084,
"step": 2050
},
{
"epoch": 0.2639671963095848,
"grad_norm": 1.1593812302159279,
"learning_rate": 2e-05,
"loss": 0.5175158977508545,
"step": 2060
},
{
"epoch": 0.26524859046642746,
"grad_norm": 0.9986285441340256,
"learning_rate": 2e-05,
"loss": 0.5173258781433105,
"step": 2070
},
{
"epoch": 0.2665299846232701,
"grad_norm": 1.004277740925964,
"learning_rate": 2e-05,
"loss": 0.5400328636169434,
"step": 2080
},
{
"epoch": 0.26781137878011274,
"grad_norm": 1.3589203570380393,
"learning_rate": 2e-05,
"loss": 0.5316391944885254,
"step": 2090
},
{
"epoch": 0.2690927729369554,
"grad_norm": 0.8970110458994068,
"learning_rate": 2e-05,
"loss": 0.49395999908447263,
"step": 2100
},
{
"epoch": 0.2703741670937981,
"grad_norm": 1.1138162233477493,
"learning_rate": 2e-05,
"loss": 0.489614200592041,
"step": 2110
},
{
"epoch": 0.2716555612506407,
"grad_norm": 0.91512055980476,
"learning_rate": 2e-05,
"loss": 0.5149580955505371,
"step": 2120
},
{
"epoch": 0.27293695540748336,
"grad_norm": 1.5596834044607704,
"learning_rate": 2e-05,
"loss": 0.51571626663208,
"step": 2130
},
{
"epoch": 0.274218349564326,
"grad_norm": 1.0561043035288422,
"learning_rate": 2e-05,
"loss": 0.5328054428100586,
"step": 2140
},
{
"epoch": 0.27549974372116864,
"grad_norm": 2.402498421541665,
"learning_rate": 2e-05,
"loss": 0.5326231956481934,
"step": 2150
},
{
"epoch": 0.2767811378780113,
"grad_norm": 1.2037813793115688,
"learning_rate": 2e-05,
"loss": 0.505500602722168,
"step": 2160
},
{
"epoch": 0.2780625320348539,
"grad_norm": 1.0836941909614977,
"learning_rate": 2e-05,
"loss": 0.5271368026733398,
"step": 2170
},
{
"epoch": 0.27934392619169657,
"grad_norm": 1.2415723963004586,
"learning_rate": 2e-05,
"loss": 0.5206530570983887,
"step": 2180
},
{
"epoch": 0.2806253203485392,
"grad_norm": 0.8375194634626388,
"learning_rate": 2e-05,
"loss": 0.5228913307189942,
"step": 2190
},
{
"epoch": 0.28190671450538185,
"grad_norm": 1.2768322284110671,
"learning_rate": 2e-05,
"loss": 0.5209582328796387,
"step": 2200
},
{
"epoch": 0.2831881086622245,
"grad_norm": 1.6086372977694905,
"learning_rate": 2e-05,
"loss": 0.5226896286010743,
"step": 2210
},
{
"epoch": 0.28446950281906713,
"grad_norm": 0.8494825087695085,
"learning_rate": 2e-05,
"loss": 0.5286993980407715,
"step": 2220
},
{
"epoch": 0.2857508969759098,
"grad_norm": 0.8156403401531495,
"learning_rate": 2e-05,
"loss": 0.5194881439208985,
"step": 2230
},
{
"epoch": 0.2870322911327524,
"grad_norm": 0.7830599753443489,
"learning_rate": 2e-05,
"loss": 0.5140536308288575,
"step": 2240
},
{
"epoch": 0.28831368528959506,
"grad_norm": 1.2965204600356965,
"learning_rate": 2e-05,
"loss": 0.5102174758911133,
"step": 2250
},
{
"epoch": 0.2895950794464377,
"grad_norm": 0.8820692738334768,
"learning_rate": 2e-05,
"loss": 0.5211565971374512,
"step": 2260
},
{
"epoch": 0.2908764736032804,
"grad_norm": 0.8962857487120582,
"learning_rate": 2e-05,
"loss": 0.525968599319458,
"step": 2270
},
{
"epoch": 0.29215786776012304,
"grad_norm": 0.9216657277454023,
"learning_rate": 2e-05,
"loss": 0.5215781211853028,
"step": 2280
},
{
"epoch": 0.2934392619169657,
"grad_norm": 0.9168444508408475,
"learning_rate": 2e-05,
"loss": 0.5295525550842285,
"step": 2290
},
{
"epoch": 0.2947206560738083,
"grad_norm": 0.9061188949639305,
"learning_rate": 2e-05,
"loss": 0.5105137825012207,
"step": 2300
},
{
"epoch": 0.29600205023065096,
"grad_norm": 1.445538952302157,
"learning_rate": 2e-05,
"loss": 0.5191007137298584,
"step": 2310
},
{
"epoch": 0.2972834443874936,
"grad_norm": 1.092777399918295,
"learning_rate": 2e-05,
"loss": 0.5119923114776611,
"step": 2320
},
{
"epoch": 0.29856483854433624,
"grad_norm": 0.7950934799480767,
"learning_rate": 2e-05,
"loss": 0.5029843330383301,
"step": 2330
},
{
"epoch": 0.2998462327011789,
"grad_norm": 0.8713953426463603,
"learning_rate": 2e-05,
"loss": 0.5177433013916015,
"step": 2340
},
{
"epoch": 0.3011276268580215,
"grad_norm": 1.1910769048168222,
"learning_rate": 2e-05,
"loss": 0.5175111293792725,
"step": 2350
},
{
"epoch": 0.30240902101486417,
"grad_norm": 0.7369439079734567,
"learning_rate": 2e-05,
"loss": 0.5052539348602295,
"step": 2360
},
{
"epoch": 0.3036904151717068,
"grad_norm": 0.8535611111992116,
"learning_rate": 2e-05,
"loss": 0.526234245300293,
"step": 2370
},
{
"epoch": 0.30497180932854945,
"grad_norm": 1.2619477944558055,
"learning_rate": 2e-05,
"loss": 0.5063686370849609,
"step": 2380
},
{
"epoch": 0.3062532034853921,
"grad_norm": 1.1581396474736574,
"learning_rate": 2e-05,
"loss": 0.5509318351745606,
"step": 2390
},
{
"epoch": 0.30753459764223473,
"grad_norm": 1.1993842399213077,
"learning_rate": 2e-05,
"loss": 0.5182885646820068,
"step": 2400
},
{
"epoch": 0.30881599179907737,
"grad_norm": 0.8928877365517348,
"learning_rate": 2e-05,
"loss": 0.5191317558288574,
"step": 2410
},
{
"epoch": 0.31009738595592007,
"grad_norm": 0.7740858441808175,
"learning_rate": 2e-05,
"loss": 0.5120766162872314,
"step": 2420
},
{
"epoch": 0.3113787801127627,
"grad_norm": 1.2066466568461025,
"learning_rate": 2e-05,
"loss": 0.5130606651306152,
"step": 2430
},
{
"epoch": 0.31266017426960535,
"grad_norm": 0.8936466558034992,
"learning_rate": 2e-05,
"loss": 0.498398494720459,
"step": 2440
},
{
"epoch": 0.313941568426448,
"grad_norm": 2.590789205476504,
"learning_rate": 2e-05,
"loss": 0.5014350891113282,
"step": 2450
},
{
"epoch": 0.31522296258329063,
"grad_norm": 0.9839532180576057,
"learning_rate": 2e-05,
"loss": 0.509852409362793,
"step": 2460
},
{
"epoch": 0.3165043567401333,
"grad_norm": 2.7901336826289196,
"learning_rate": 2e-05,
"loss": 0.5187605381011963,
"step": 2470
},
{
"epoch": 0.3177857508969759,
"grad_norm": 0.8477814700442736,
"learning_rate": 2e-05,
"loss": 0.5065323352813721,
"step": 2480
},
{
"epoch": 0.31906714505381856,
"grad_norm": 0.9127582158833917,
"learning_rate": 2e-05,
"loss": 0.5161973476409912,
"step": 2490
},
{
"epoch": 0.3203485392106612,
"grad_norm": 1.047329752302117,
"learning_rate": 2e-05,
"loss": 0.5033672332763672,
"step": 2500
},
{
"epoch": 0.32162993336750384,
"grad_norm": 1.0667824375743964,
"learning_rate": 2e-05,
"loss": 0.5326580047607422,
"step": 2510
},
{
"epoch": 0.3229113275243465,
"grad_norm": 1.2880018997570044,
"learning_rate": 2e-05,
"loss": 0.5263604640960693,
"step": 2520
},
{
"epoch": 0.3241927216811891,
"grad_norm": 1.27594129391459,
"learning_rate": 2e-05,
"loss": 0.523559284210205,
"step": 2530
},
{
"epoch": 0.32547411583803176,
"grad_norm": 2.163613965706386,
"learning_rate": 2e-05,
"loss": 0.5245936870574951,
"step": 2540
},
{
"epoch": 0.3267555099948744,
"grad_norm": 1.4988595252237185,
"learning_rate": 2e-05,
"loss": 0.5442821025848389,
"step": 2550
},
{
"epoch": 0.32803690415171705,
"grad_norm": 1.3434721205594993,
"learning_rate": 2e-05,
"loss": 0.5150107383728028,
"step": 2560
},
{
"epoch": 0.3293182983085597,
"grad_norm": 0.9061066110318906,
"learning_rate": 2e-05,
"loss": 0.5170779228210449,
"step": 2570
},
{
"epoch": 0.3305996924654024,
"grad_norm": 0.9298858186488728,
"learning_rate": 2e-05,
"loss": 0.5022658824920654,
"step": 2580
},
{
"epoch": 0.331881086622245,
"grad_norm": 0.8407741186459021,
"learning_rate": 2e-05,
"loss": 0.5079696178436279,
"step": 2590
},
{
"epoch": 0.33316248077908767,
"grad_norm": 1.055024158702933,
"learning_rate": 2e-05,
"loss": 0.4967644691467285,
"step": 2600
},
{
"epoch": 0.3344438749359303,
"grad_norm": 0.9249574841418488,
"learning_rate": 2e-05,
"loss": 0.5244412899017334,
"step": 2610
},
{
"epoch": 0.33572526909277295,
"grad_norm": 0.8312050241116705,
"learning_rate": 2e-05,
"loss": 0.49590363502502444,
"step": 2620
},
{
"epoch": 0.3370066632496156,
"grad_norm": 0.876550792707944,
"learning_rate": 2e-05,
"loss": 0.49512629508972167,
"step": 2630
},
{
"epoch": 0.33828805740645823,
"grad_norm": 1.0403845880599576,
"learning_rate": 2e-05,
"loss": 0.49561777114868166,
"step": 2640
},
{
"epoch": 0.3395694515633009,
"grad_norm": 4.746418607012456,
"learning_rate": 2e-05,
"loss": 0.5274491786956788,
"step": 2650
},
{
"epoch": 0.3408508457201435,
"grad_norm": 0.8464624782392126,
"learning_rate": 2e-05,
"loss": 0.5022400856018067,
"step": 2660
},
{
"epoch": 0.34213223987698616,
"grad_norm": 0.8292129365138868,
"learning_rate": 2e-05,
"loss": 0.5213953018188476,
"step": 2670
},
{
"epoch": 0.3434136340338288,
"grad_norm": 0.9269412386375918,
"learning_rate": 2e-05,
"loss": 0.5191431045532227,
"step": 2680
},
{
"epoch": 0.34469502819067144,
"grad_norm": 1.0042837272832486,
"learning_rate": 2e-05,
"loss": 0.5267402172088623,
"step": 2690
},
{
"epoch": 0.3459764223475141,
"grad_norm": 1.739635011944283,
"learning_rate": 2e-05,
"loss": 0.5184992790222168,
"step": 2700
},
{
"epoch": 0.3472578165043567,
"grad_norm": 12.669625948002842,
"learning_rate": 2e-05,
"loss": 0.5027872562408447,
"step": 2710
},
{
"epoch": 0.34853921066119936,
"grad_norm": 0.9919174315727555,
"learning_rate": 2e-05,
"loss": 0.4882629871368408,
"step": 2720
},
{
"epoch": 0.349820604818042,
"grad_norm": 1.1700680937804502,
"learning_rate": 2e-05,
"loss": 0.5146281719207764,
"step": 2730
},
{
"epoch": 0.3511019989748847,
"grad_norm": 1.1404289562707015,
"learning_rate": 2e-05,
"loss": 0.5151095390319824,
"step": 2740
},
{
"epoch": 0.35238339313172734,
"grad_norm": 1.215916629687437,
"learning_rate": 2e-05,
"loss": 0.4892124176025391,
"step": 2750
},
{
"epoch": 0.35366478728857,
"grad_norm": 1.5321752360132277,
"learning_rate": 2e-05,
"loss": 0.49410400390625,
"step": 2760
},
{
"epoch": 0.3549461814454126,
"grad_norm": 2.836224836676823,
"learning_rate": 2e-05,
"loss": 0.5124226093292237,
"step": 2770
},
{
"epoch": 0.35622757560225526,
"grad_norm": 0.8267447902560755,
"learning_rate": 2e-05,
"loss": 0.5107037544250488,
"step": 2780
},
{
"epoch": 0.3575089697590979,
"grad_norm": 0.8763583030075933,
"learning_rate": 2e-05,
"loss": 0.5248641490936279,
"step": 2790
},
{
"epoch": 0.35879036391594055,
"grad_norm": 1.4195757557352582,
"learning_rate": 2e-05,
"loss": 0.5127915382385254,
"step": 2800
},
{
"epoch": 0.3600717580727832,
"grad_norm": 1.1333727395391804,
"learning_rate": 2e-05,
"loss": 0.49205379486083983,
"step": 2810
},
{
"epoch": 0.36135315222962583,
"grad_norm": 1.0377576334127312,
"learning_rate": 2e-05,
"loss": 0.49173612594604493,
"step": 2820
},
{
"epoch": 0.36263454638646847,
"grad_norm": 1.5452289102806698,
"learning_rate": 2e-05,
"loss": 0.511233139038086,
"step": 2830
},
{
"epoch": 0.3639159405433111,
"grad_norm": 0.8603560735233966,
"learning_rate": 2e-05,
"loss": 0.4994975566864014,
"step": 2840
},
{
"epoch": 0.36519733470015375,
"grad_norm": 0.7917461806509101,
"learning_rate": 2e-05,
"loss": 0.49323644638061526,
"step": 2850
},
{
"epoch": 0.3664787288569964,
"grad_norm": 1.0309639678750901,
"learning_rate": 2e-05,
"loss": 0.4953965663909912,
"step": 2860
},
{
"epoch": 0.36776012301383904,
"grad_norm": 0.7064695173012207,
"learning_rate": 2e-05,
"loss": 0.4873600959777832,
"step": 2870
},
{
"epoch": 0.3690415171706817,
"grad_norm": 1.2732892833171587,
"learning_rate": 2e-05,
"loss": 0.5117098331451416,
"step": 2880
},
{
"epoch": 0.3703229113275244,
"grad_norm": 0.7013993388393664,
"learning_rate": 2e-05,
"loss": 0.5019349098205567,
"step": 2890
},
{
"epoch": 0.371604305484367,
"grad_norm": 0.846139733033606,
"learning_rate": 2e-05,
"loss": 0.4878437042236328,
"step": 2900
},
{
"epoch": 0.37288569964120966,
"grad_norm": 1.064153789050579,
"learning_rate": 2e-05,
"loss": 0.5087510108947754,
"step": 2910
},
{
"epoch": 0.3741670937980523,
"grad_norm": 0.7262051519448712,
"learning_rate": 2e-05,
"loss": 0.506193733215332,
"step": 2920
},
{
"epoch": 0.37544848795489494,
"grad_norm": 0.9120045935081182,
"learning_rate": 2e-05,
"loss": 0.49692955017089846,
"step": 2930
},
{
"epoch": 0.3767298821117376,
"grad_norm": 0.7936517947365992,
"learning_rate": 2e-05,
"loss": 0.4955914974212646,
"step": 2940
},
{
"epoch": 0.3780112762685802,
"grad_norm": 1.2828153517374288,
"learning_rate": 2e-05,
"loss": 0.5013689994812012,
"step": 2950
},
{
"epoch": 0.37929267042542286,
"grad_norm": 0.9518433356223106,
"learning_rate": 2e-05,
"loss": 0.49332561492919924,
"step": 2960
},
{
"epoch": 0.3805740645822655,
"grad_norm": 0.8342944677215532,
"learning_rate": 2e-05,
"loss": 0.5117284297943115,
"step": 2970
},
{
"epoch": 0.38185545873910814,
"grad_norm": 0.9025711024479366,
"learning_rate": 2e-05,
"loss": 0.48902301788330077,
"step": 2980
},
{
"epoch": 0.3831368528959508,
"grad_norm": 0.7332870017865983,
"learning_rate": 2e-05,
"loss": 0.491532039642334,
"step": 2990
},
{
"epoch": 0.3844182470527934,
"grad_norm": 0.7846446603166206,
"learning_rate": 2e-05,
"loss": 0.5058146476745605,
"step": 3000
},
{
"epoch": 0.38569964120963607,
"grad_norm": 0.8542058335541718,
"learning_rate": 2e-05,
"loss": 0.4892392158508301,
"step": 3010
},
{
"epoch": 0.3869810353664787,
"grad_norm": 0.8334146471423989,
"learning_rate": 2e-05,
"loss": 0.48704729080200193,
"step": 3020
},
{
"epoch": 0.38826242952332135,
"grad_norm": 0.7752037638841061,
"learning_rate": 2e-05,
"loss": 0.4776434421539307,
"step": 3030
},
{
"epoch": 0.389543823680164,
"grad_norm": 1.1980007481008972,
"learning_rate": 2e-05,
"loss": 0.5104766368865967,
"step": 3040
},
{
"epoch": 0.3908252178370067,
"grad_norm": 0.9371345842036125,
"learning_rate": 2e-05,
"loss": 0.4924893379211426,
"step": 3050
},
{
"epoch": 0.39210661199384933,
"grad_norm": 0.8360826741026628,
"learning_rate": 2e-05,
"loss": 0.4908696174621582,
"step": 3060
},
{
"epoch": 0.39338800615069197,
"grad_norm": 7.58317297113024,
"learning_rate": 2e-05,
"loss": 0.5224600791931152,
"step": 3070
},
{
"epoch": 0.3946694003075346,
"grad_norm": 1.029436179140336,
"learning_rate": 2e-05,
"loss": 0.4978078842163086,
"step": 3080
},
{
"epoch": 0.39595079446437725,
"grad_norm": 1.0141630734542044,
"learning_rate": 2e-05,
"loss": 0.48975610733032227,
"step": 3090
},
{
"epoch": 0.3972321886212199,
"grad_norm": 0.8006267385254168,
"learning_rate": 2e-05,
"loss": 0.5065206527709961,
"step": 3100
},
{
"epoch": 0.39851358277806254,
"grad_norm": 0.7195878101917671,
"learning_rate": 2e-05,
"loss": 0.5112967967987061,
"step": 3110
},
{
"epoch": 0.3997949769349052,
"grad_norm": 1.2769233676311862,
"learning_rate": 2e-05,
"loss": 0.49534831047058103,
"step": 3120
},
{
"epoch": 0.4010763710917478,
"grad_norm": 7.360056009194376,
"learning_rate": 2e-05,
"loss": 0.47747159004211426,
"step": 3130
},
{
"epoch": 0.40235776524859046,
"grad_norm": 1.1009926775915881,
"learning_rate": 2e-05,
"loss": 0.5052808284759521,
"step": 3140
},
{
"epoch": 0.4036391594054331,
"grad_norm": 2.8272173828290197,
"learning_rate": 2e-05,
"loss": 0.4913668632507324,
"step": 3150
},
{
"epoch": 0.40492055356227574,
"grad_norm": 0.8123154440597995,
"learning_rate": 2e-05,
"loss": 0.5172830104827881,
"step": 3160
},
{
"epoch": 0.4062019477191184,
"grad_norm": 0.8554492507667826,
"learning_rate": 2e-05,
"loss": 0.48242883682250975,
"step": 3170
},
{
"epoch": 0.407483341875961,
"grad_norm": 0.9535672964450382,
"learning_rate": 2e-05,
"loss": 0.49997806549072266,
"step": 3180
},
{
"epoch": 0.40876473603280367,
"grad_norm": 2.456569423696086,
"learning_rate": 2e-05,
"loss": 0.49902682304382323,
"step": 3190
},
{
"epoch": 0.41004613018964636,
"grad_norm": 1.5917591243860325,
"learning_rate": 2e-05,
"loss": 0.49785447120666504,
"step": 3200
},
{
"epoch": 0.411327524346489,
"grad_norm": 1.310481377000517,
"learning_rate": 2e-05,
"loss": 0.5000613689422607,
"step": 3210
},
{
"epoch": 0.41260891850333165,
"grad_norm": 1.2804578386215548,
"learning_rate": 2e-05,
"loss": 0.49518542289733886,
"step": 3220
},
{
"epoch": 0.4138903126601743,
"grad_norm": 0.9930738760252762,
"learning_rate": 2e-05,
"loss": 0.5024672508239746,
"step": 3230
},
{
"epoch": 0.41517170681701693,
"grad_norm": 1.0223244616537686,
"learning_rate": 2e-05,
"loss": 0.5310104846954345,
"step": 3240
},
{
"epoch": 0.41645310097385957,
"grad_norm": 1.0130909531271475,
"learning_rate": 2e-05,
"loss": 0.49877562522888186,
"step": 3250
},
{
"epoch": 0.4177344951307022,
"grad_norm": 1.074676579642214,
"learning_rate": 2e-05,
"loss": 0.5068572044372559,
"step": 3260
},
{
"epoch": 0.41901588928754485,
"grad_norm": 0.9598475321636567,
"learning_rate": 2e-05,
"loss": 0.5096299648284912,
"step": 3270
},
{
"epoch": 0.4202972834443875,
"grad_norm": 1.2710462397519864,
"learning_rate": 2e-05,
"loss": 0.5282601833343505,
"step": 3280
},
{
"epoch": 0.42157867760123013,
"grad_norm": 1.2100028303019783,
"learning_rate": 2e-05,
"loss": 0.5017249584197998,
"step": 3290
},
{
"epoch": 0.4228600717580728,
"grad_norm": 1.2867520717660021,
"learning_rate": 2e-05,
"loss": 0.5081565380096436,
"step": 3300
},
{
"epoch": 0.4241414659149154,
"grad_norm": 3.0922536581498585,
"learning_rate": 2e-05,
"loss": 0.5076882839202881,
"step": 3310
},
{
"epoch": 0.42542286007175806,
"grad_norm": 1.915973582234196,
"learning_rate": 2e-05,
"loss": 0.4949382781982422,
"step": 3320
},
{
"epoch": 0.4267042542286007,
"grad_norm": 4.0038824467871565,
"learning_rate": 2e-05,
"loss": 0.4881223201751709,
"step": 3330
},
{
"epoch": 0.42798564838544334,
"grad_norm": 1.1996959856504303,
"learning_rate": 2e-05,
"loss": 0.5322386741638183,
"step": 3340
},
{
"epoch": 0.429267042542286,
"grad_norm": 1.0915056454763494,
"learning_rate": 2e-05,
"loss": 0.48554315567016604,
"step": 3350
},
{
"epoch": 0.4305484366991287,
"grad_norm": 1.5460208821091537,
"learning_rate": 2e-05,
"loss": 0.4900236129760742,
"step": 3360
},
{
"epoch": 0.4318298308559713,
"grad_norm": 1.1235983634369164,
"learning_rate": 2e-05,
"loss": 0.5000977039337158,
"step": 3370
},
{
"epoch": 0.43311122501281396,
"grad_norm": 1.012628712188859,
"learning_rate": 2e-05,
"loss": 0.5125413417816163,
"step": 3380
},
{
"epoch": 0.4343926191696566,
"grad_norm": 1.1171514127156708,
"learning_rate": 2e-05,
"loss": 0.5019603729248047,
"step": 3390
},
{
"epoch": 0.43567401332649924,
"grad_norm": 1.495315083098467,
"learning_rate": 2e-05,
"loss": 0.5123660087585449,
"step": 3400
},
{
"epoch": 0.4369554074833419,
"grad_norm": 0.9781907252807691,
"learning_rate": 2e-05,
"loss": 0.5015677452087403,
"step": 3410
},
{
"epoch": 0.4382368016401845,
"grad_norm": 0.95638885706939,
"learning_rate": 2e-05,
"loss": 0.49541263580322265,
"step": 3420
},
{
"epoch": 0.43951819579702717,
"grad_norm": 1.3181042951511683,
"learning_rate": 2e-05,
"loss": 0.46662297248840334,
"step": 3430
},
{
"epoch": 0.4407995899538698,
"grad_norm": 0.9151418466415979,
"learning_rate": 2e-05,
"loss": 0.4814878463745117,
"step": 3440
},
{
"epoch": 0.44208098411071245,
"grad_norm": 1.4700077860583944,
"learning_rate": 2e-05,
"loss": 0.5188211917877197,
"step": 3450
},
{
"epoch": 0.4433623782675551,
"grad_norm": 3.2369869685219963,
"learning_rate": 2e-05,
"loss": 0.4897316932678223,
"step": 3460
},
{
"epoch": 0.44464377242439773,
"grad_norm": 0.8315053103258366,
"learning_rate": 2e-05,
"loss": 0.4942572593688965,
"step": 3470
},
{
"epoch": 0.4459251665812404,
"grad_norm": 1.5721673322201397,
"learning_rate": 2e-05,
"loss": 0.4921001434326172,
"step": 3480
},
{
"epoch": 0.447206560738083,
"grad_norm": 1.4070240761277857,
"learning_rate": 2e-05,
"loss": 0.4868588924407959,
"step": 3490
},
{
"epoch": 0.44848795489492566,
"grad_norm": 5.843928638043908,
"learning_rate": 2e-05,
"loss": 0.4839268684387207,
"step": 3500
},
{
"epoch": 0.4497693490517683,
"grad_norm": 1.2364710242381716,
"learning_rate": 2e-05,
"loss": 0.5122146606445312,
"step": 3510
},
{
"epoch": 0.451050743208611,
"grad_norm": 1.0813148554333136,
"learning_rate": 2e-05,
"loss": 0.5095366477966309,
"step": 3520
},
{
"epoch": 0.45233213736545363,
"grad_norm": 1.2781091467549035,
"learning_rate": 2e-05,
"loss": 0.5002841949462891,
"step": 3530
},
{
"epoch": 0.4536135315222963,
"grad_norm": 3.992034934307398,
"learning_rate": 2e-05,
"loss": 0.5027037143707276,
"step": 3540
},
{
"epoch": 0.4548949256791389,
"grad_norm": 1.3710952421455993,
"learning_rate": 2e-05,
"loss": 0.479889965057373,
"step": 3550
},
{
"epoch": 0.45617631983598156,
"grad_norm": 0.8453396855429802,
"learning_rate": 2e-05,
"loss": 0.48880681991577146,
"step": 3560
},
{
"epoch": 0.4574577139928242,
"grad_norm": 0.7372477920202382,
"learning_rate": 2e-05,
"loss": 0.48568525314331057,
"step": 3570
},
{
"epoch": 0.45873910814966684,
"grad_norm": 2.6243938972426917,
"learning_rate": 2e-05,
"loss": 0.4868001937866211,
"step": 3580
},
{
"epoch": 0.4600205023065095,
"grad_norm": 1.128110180365949,
"learning_rate": 2e-05,
"loss": 0.5153658866882325,
"step": 3590
},
{
"epoch": 0.4613018964633521,
"grad_norm": 1.0584593901939467,
"learning_rate": 2e-05,
"loss": 0.497204065322876,
"step": 3600
},
{
"epoch": 0.46258329062019476,
"grad_norm": 0.7268198080715501,
"learning_rate": 2e-05,
"loss": 0.5044869422912598,
"step": 3610
},
{
"epoch": 0.4638646847770374,
"grad_norm": 0.9345520899035163,
"learning_rate": 2e-05,
"loss": 0.49730720520019533,
"step": 3620
},
{
"epoch": 0.46514607893388005,
"grad_norm": 1.0009534927509294,
"learning_rate": 2e-05,
"loss": 0.5047263145446778,
"step": 3630
},
{
"epoch": 0.4664274730907227,
"grad_norm": 0.8326680425374008,
"learning_rate": 2e-05,
"loss": 0.5106517791748046,
"step": 3640
},
{
"epoch": 0.46770886724756533,
"grad_norm": 0.8222513837613713,
"learning_rate": 2e-05,
"loss": 0.5010424137115479,
"step": 3650
},
{
"epoch": 0.46899026140440797,
"grad_norm": 0.8473441482228149,
"learning_rate": 2e-05,
"loss": 0.4817332744598389,
"step": 3660
},
{
"epoch": 0.47027165556125067,
"grad_norm": 0.8447162393963341,
"learning_rate": 2e-05,
"loss": 0.4859748363494873,
"step": 3670
},
{
"epoch": 0.4715530497180933,
"grad_norm": 1.0133994166534215,
"learning_rate": 2e-05,
"loss": 0.4846841335296631,
"step": 3680
},
{
"epoch": 0.47283444387493595,
"grad_norm": 0.9614585475386802,
"learning_rate": 2e-05,
"loss": 0.49245414733886717,
"step": 3690
},
{
"epoch": 0.4741158380317786,
"grad_norm": 0.8807822311485736,
"learning_rate": 2e-05,
"loss": 0.48321146965026857,
"step": 3700
},
{
"epoch": 0.47539723218862123,
"grad_norm": 0.7439293038840538,
"learning_rate": 2e-05,
"loss": 0.4913172245025635,
"step": 3710
},
{
"epoch": 0.4766786263454639,
"grad_norm": 0.9073232870316841,
"learning_rate": 2e-05,
"loss": 0.5013354778289795,
"step": 3720
},
{
"epoch": 0.4779600205023065,
"grad_norm": 0.717406566656741,
"learning_rate": 2e-05,
"loss": 0.4782130241394043,
"step": 3730
},
{
"epoch": 0.47924141465914916,
"grad_norm": 0.6889423753617563,
"learning_rate": 2e-05,
"loss": 0.4947871208190918,
"step": 3740
},
{
"epoch": 0.4805228088159918,
"grad_norm": 1.1741454591156335,
"learning_rate": 2e-05,
"loss": 0.5143609523773194,
"step": 3750
},
{
"epoch": 0.48180420297283444,
"grad_norm": 5.241802980368391,
"learning_rate": 2e-05,
"loss": 0.501248550415039,
"step": 3760
},
{
"epoch": 0.4830855971296771,
"grad_norm": 0.8537426317992355,
"learning_rate": 2e-05,
"loss": 0.49720053672790526,
"step": 3770
},
{
"epoch": 0.4843669912865197,
"grad_norm": 1.0161740820072611,
"learning_rate": 2e-05,
"loss": 0.49223899841308594,
"step": 3780
},
{
"epoch": 0.48564838544336236,
"grad_norm": 0.8624434092976485,
"learning_rate": 2e-05,
"loss": 0.49568853378295896,
"step": 3790
},
{
"epoch": 0.486929779600205,
"grad_norm": 1.3875121273607534,
"learning_rate": 2e-05,
"loss": 0.49059505462646485,
"step": 3800
},
{
"epoch": 0.48821117375704765,
"grad_norm": 0.7146486747369168,
"learning_rate": 2e-05,
"loss": 0.4859918117523193,
"step": 3810
},
{
"epoch": 0.4894925679138903,
"grad_norm": 1.0405111527222635,
"learning_rate": 2e-05,
"loss": 0.5005202293395996,
"step": 3820
},
{
"epoch": 0.490773962070733,
"grad_norm": 1.2940545654383988,
"learning_rate": 2e-05,
"loss": 0.5019304275512695,
"step": 3830
},
{
"epoch": 0.4920553562275756,
"grad_norm": 0.735338242033843,
"learning_rate": 2e-05,
"loss": 0.48978242874145506,
"step": 3840
},
{
"epoch": 0.49333675038441827,
"grad_norm": 0.8314677609234699,
"learning_rate": 2e-05,
"loss": 0.5010252475738526,
"step": 3850
},
{
"epoch": 0.4946181445412609,
"grad_norm": 1.0169558643590326,
"learning_rate": 2e-05,
"loss": 0.48139405250549316,
"step": 3860
},
{
"epoch": 0.49589953869810355,
"grad_norm": 1.1255717041868347,
"learning_rate": 2e-05,
"loss": 0.4902916431427002,
"step": 3870
},
{
"epoch": 0.4971809328549462,
"grad_norm": 1.1733471424717912,
"learning_rate": 2e-05,
"loss": 0.4886775970458984,
"step": 3880
},
{
"epoch": 0.49846232701178883,
"grad_norm": 0.988006911560625,
"learning_rate": 2e-05,
"loss": 0.4853158950805664,
"step": 3890
},
{
"epoch": 0.49974372116863147,
"grad_norm": 0.8939034033774742,
"learning_rate": 2e-05,
"loss": 0.5042220115661621,
"step": 3900
},
{
"epoch": 0.5010251153254741,
"grad_norm": 1.2292781612758479,
"learning_rate": 2e-05,
"loss": 0.49028773307800294,
"step": 3910
},
{
"epoch": 0.5023065094823168,
"grad_norm": 1.7342532061855536,
"learning_rate": 2e-05,
"loss": 0.4965537071228027,
"step": 3920
},
{
"epoch": 0.5035879036391594,
"grad_norm": 0.9254919633386374,
"learning_rate": 2e-05,
"loss": 0.5004670143127441,
"step": 3930
},
{
"epoch": 0.5048692977960021,
"grad_norm": 1.0467122261841153,
"learning_rate": 2e-05,
"loss": 0.4774936676025391,
"step": 3940
},
{
"epoch": 0.5061506919528447,
"grad_norm": 0.9470838920635308,
"learning_rate": 2e-05,
"loss": 0.49421100616455077,
"step": 3950
},
{
"epoch": 0.5074320861096874,
"grad_norm": 2.222077818875719,
"learning_rate": 2e-05,
"loss": 0.5057079792022705,
"step": 3960
},
{
"epoch": 0.50871348026653,
"grad_norm": 0.871943207898756,
"learning_rate": 2e-05,
"loss": 0.48962907791137694,
"step": 3970
},
{
"epoch": 0.5099948744233727,
"grad_norm": 1.3908625371678747,
"learning_rate": 2e-05,
"loss": 0.4938501834869385,
"step": 3980
},
{
"epoch": 0.5112762685802152,
"grad_norm": 0.9759414900189028,
"learning_rate": 2e-05,
"loss": 0.4821597099304199,
"step": 3990
},
{
"epoch": 0.5125576627370579,
"grad_norm": 2.2702587838018222,
"learning_rate": 2e-05,
"loss": 0.5201299667358399,
"step": 4000
},
{
"epoch": 0.5138390568939005,
"grad_norm": 1.4897789199185976,
"learning_rate": 2e-05,
"loss": 0.5009584426879883,
"step": 4010
},
{
"epoch": 0.5151204510507432,
"grad_norm": 0.8503562901882815,
"learning_rate": 2e-05,
"loss": 0.4796182155609131,
"step": 4020
},
{
"epoch": 0.5164018452075858,
"grad_norm": 1.0514642837531456,
"learning_rate": 2e-05,
"loss": 0.4897005558013916,
"step": 4030
},
{
"epoch": 0.5176832393644285,
"grad_norm": 0.9846485431552143,
"learning_rate": 2e-05,
"loss": 0.5230988502502442,
"step": 4040
},
{
"epoch": 0.5189646335212711,
"grad_norm": 1.0223799600913301,
"learning_rate": 2e-05,
"loss": 0.4999506950378418,
"step": 4050
},
{
"epoch": 0.5202460276781138,
"grad_norm": 0.9313972404641525,
"learning_rate": 2e-05,
"loss": 0.49692330360412595,
"step": 4060
},
{
"epoch": 0.5215274218349565,
"grad_norm": 0.8943452451988742,
"learning_rate": 2e-05,
"loss": 0.4879587173461914,
"step": 4070
}
],
"logging_steps": 10,
"max_steps": 5935,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2035,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1927226987642880.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}