SPLASH_1B / trainer_state.json
xxinzzi's picture
Add files using upload-large-folder tool
4acb56b verified
Raw
History Blame Contribute Delete
44.2 kB
{
"best_global_step": 1214,
"best_metric": 0.975861668586731,
"best_model_checkpoint": "/hddsda/team_VLT/experiments/splash_1B/outputs/60_2e5/checkpoint-1214",
"epoch": 2.0,
"eval_steps": 500,
"global_step": 2428,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.008237232289950576,
"grad_norm": 59.42716598510742,
"learning_rate": 5.47945205479452e-07,
"loss": 3.9203125,
"step": 10
},
{
"epoch": 0.016474464579901153,
"grad_norm": 38.90184020996094,
"learning_rate": 1.095890410958904e-06,
"loss": 3.6734375,
"step": 20
},
{
"epoch": 0.02471169686985173,
"grad_norm": 25.503145217895508,
"learning_rate": 1.6438356164383561e-06,
"loss": 2.7765625,
"step": 30
},
{
"epoch": 0.032948929159802305,
"grad_norm": 12.902335166931152,
"learning_rate": 2.191780821917808e-06,
"loss": 2.01171875,
"step": 40
},
{
"epoch": 0.04118616144975288,
"grad_norm": 8.680400848388672,
"learning_rate": 2.7397260273972604e-06,
"loss": 1.62421875,
"step": 50
},
{
"epoch": 0.04942339373970346,
"grad_norm": 9.069554328918457,
"learning_rate": 3.2876712328767123e-06,
"loss": 1.43828125,
"step": 60
},
{
"epoch": 0.057660626029654036,
"grad_norm": 12.622966766357422,
"learning_rate": 3.8356164383561645e-06,
"loss": 1.3921875,
"step": 70
},
{
"epoch": 0.06589785831960461,
"grad_norm": 8.354354858398438,
"learning_rate": 4.383561643835616e-06,
"loss": 1.371875,
"step": 80
},
{
"epoch": 0.07413509060955518,
"grad_norm": 7.582118511199951,
"learning_rate": 4.931506849315069e-06,
"loss": 1.31953125,
"step": 90
},
{
"epoch": 0.08237232289950576,
"grad_norm": 7.6723313331604,
"learning_rate": 5.479452054794521e-06,
"loss": 1.2421875,
"step": 100
},
{
"epoch": 0.09060955518945635,
"grad_norm": 6.57471227645874,
"learning_rate": 6.027397260273973e-06,
"loss": 1.22265625,
"step": 110
},
{
"epoch": 0.09884678747940692,
"grad_norm": 6.930084228515625,
"learning_rate": 6.5753424657534245e-06,
"loss": 1.178125,
"step": 120
},
{
"epoch": 0.1070840197693575,
"grad_norm": 6.962713718414307,
"learning_rate": 7.123287671232877e-06,
"loss": 1.2109375,
"step": 130
},
{
"epoch": 0.11532125205930807,
"grad_norm": 5.813303470611572,
"learning_rate": 7.671232876712329e-06,
"loss": 1.259375,
"step": 140
},
{
"epoch": 0.12355848434925865,
"grad_norm": 5.77918815612793,
"learning_rate": 8.219178082191782e-06,
"loss": 1.181640625,
"step": 150
},
{
"epoch": 0.13179571663920922,
"grad_norm": 6.510586738586426,
"learning_rate": 8.767123287671233e-06,
"loss": 1.1734375,
"step": 160
},
{
"epoch": 0.1400329489291598,
"grad_norm": 6.044201850891113,
"learning_rate": 9.315068493150685e-06,
"loss": 1.14765625,
"step": 170
},
{
"epoch": 0.14827018121911037,
"grad_norm": 7.2401041984558105,
"learning_rate": 9.863013698630138e-06,
"loss": 1.11796875,
"step": 180
},
{
"epoch": 0.15650741350906094,
"grad_norm": 5.330558776855469,
"learning_rate": 1.0410958904109589e-05,
"loss": 1.151953125,
"step": 190
},
{
"epoch": 0.16474464579901152,
"grad_norm": 5.636282444000244,
"learning_rate": 1.0958904109589042e-05,
"loss": 1.10859375,
"step": 200
},
{
"epoch": 0.17298187808896212,
"grad_norm": 6.467264652252197,
"learning_rate": 1.1506849315068493e-05,
"loss": 1.1375,
"step": 210
},
{
"epoch": 0.1812191103789127,
"grad_norm": 5.765719413757324,
"learning_rate": 1.2054794520547945e-05,
"loss": 1.14296875,
"step": 220
},
{
"epoch": 0.18945634266886327,
"grad_norm": 5.661046028137207,
"learning_rate": 1.2602739726027398e-05,
"loss": 1.15234375,
"step": 230
},
{
"epoch": 0.19769357495881384,
"grad_norm": 5.168608665466309,
"learning_rate": 1.3150684931506849e-05,
"loss": 1.1125,
"step": 240
},
{
"epoch": 0.20593080724876442,
"grad_norm": 4.618305683135986,
"learning_rate": 1.3698630136986302e-05,
"loss": 1.146875,
"step": 250
},
{
"epoch": 0.214168039538715,
"grad_norm": 4.392104148864746,
"learning_rate": 1.4246575342465754e-05,
"loss": 1.1328125,
"step": 260
},
{
"epoch": 0.22240527182866557,
"grad_norm": 5.369014263153076,
"learning_rate": 1.4794520547945205e-05,
"loss": 1.107421875,
"step": 270
},
{
"epoch": 0.23064250411861614,
"grad_norm": 4.20781946182251,
"learning_rate": 1.5342465753424658e-05,
"loss": 1.064453125,
"step": 280
},
{
"epoch": 0.23887973640856672,
"grad_norm": 5.343926429748535,
"learning_rate": 1.589041095890411e-05,
"loss": 1.09453125,
"step": 290
},
{
"epoch": 0.2471169686985173,
"grad_norm": 4.698163986206055,
"learning_rate": 1.6438356164383563e-05,
"loss": 1.109375,
"step": 300
},
{
"epoch": 0.2553542009884679,
"grad_norm": 4.694039821624756,
"learning_rate": 1.6986301369863014e-05,
"loss": 1.06796875,
"step": 310
},
{
"epoch": 0.26359143327841844,
"grad_norm": 4.443817138671875,
"learning_rate": 1.7534246575342465e-05,
"loss": 1.096484375,
"step": 320
},
{
"epoch": 0.27182866556836904,
"grad_norm": 4.604736804962158,
"learning_rate": 1.808219178082192e-05,
"loss": 1.04609375,
"step": 330
},
{
"epoch": 0.2800658978583196,
"grad_norm": 4.275271415710449,
"learning_rate": 1.863013698630137e-05,
"loss": 1.054296875,
"step": 340
},
{
"epoch": 0.2883031301482702,
"grad_norm": 4.036405086517334,
"learning_rate": 1.9178082191780822e-05,
"loss": 1.0875,
"step": 350
},
{
"epoch": 0.29654036243822074,
"grad_norm": 4.500349521636963,
"learning_rate": 1.9726027397260276e-05,
"loss": 1.05078125,
"step": 360
},
{
"epoch": 0.30477759472817134,
"grad_norm": 4.127949237823486,
"learning_rate": 1.9999885116926973e-05,
"loss": 1.040625,
"step": 370
},
{
"epoch": 0.3130148270181219,
"grad_norm": 3.660872220993042,
"learning_rate": 1.9998966068180426e-05,
"loss": 1.01484375,
"step": 380
},
{
"epoch": 0.3212520593080725,
"grad_norm": 4.046405792236328,
"learning_rate": 1.999712805515336e-05,
"loss": 1.045703125,
"step": 390
},
{
"epoch": 0.32948929159802304,
"grad_norm": 3.8753719329833984,
"learning_rate": 1.9994371246770076e-05,
"loss": 1.080859375,
"step": 400
},
{
"epoch": 0.33772652388797364,
"grad_norm": 4.130710124969482,
"learning_rate": 1.9990695896397612e-05,
"loss": 0.983203125,
"step": 410
},
{
"epoch": 0.34596375617792424,
"grad_norm": 3.3500330448150635,
"learning_rate": 1.998610234182246e-05,
"loss": 1.07265625,
"step": 420
},
{
"epoch": 0.3542009884678748,
"grad_norm": 3.836678981781006,
"learning_rate": 1.9980591005219537e-05,
"loss": 1.004296875,
"step": 430
},
{
"epoch": 0.3624382207578254,
"grad_norm": 3.467137098312378,
"learning_rate": 1.9974162393113353e-05,
"loss": 1.018359375,
"step": 440
},
{
"epoch": 0.37067545304777594,
"grad_norm": 3.424943685531616,
"learning_rate": 1.9966817096331485e-05,
"loss": 1.03359375,
"step": 450
},
{
"epoch": 0.37891268533772654,
"grad_norm": 4.344949245452881,
"learning_rate": 1.9958555789950278e-05,
"loss": 1.069140625,
"step": 460
},
{
"epoch": 0.3871499176276771,
"grad_norm": 3.3675737380981445,
"learning_rate": 1.9949379233232773e-05,
"loss": 1.0390625,
"step": 470
},
{
"epoch": 0.3953871499176277,
"grad_norm": 4.10872220993042,
"learning_rate": 1.993928826955896e-05,
"loss": 0.98671875,
"step": 480
},
{
"epoch": 0.40362438220757824,
"grad_norm": 2.9576351642608643,
"learning_rate": 1.9928283826348237e-05,
"loss": 1.01796875,
"step": 490
},
{
"epoch": 0.41186161449752884,
"grad_norm": 3.2495648860931396,
"learning_rate": 1.9916366914974207e-05,
"loss": 1.0703125,
"step": 500
},
{
"epoch": 0.4200988467874794,
"grad_norm": 3.462428569793701,
"learning_rate": 1.9903538630671687e-05,
"loss": 1.02109375,
"step": 510
},
{
"epoch": 0.42833607907743,
"grad_norm": 3.6364905834198,
"learning_rate": 1.9889800152436095e-05,
"loss": 0.97734375,
"step": 520
},
{
"epoch": 0.43657331136738053,
"grad_norm": 2.8674442768096924,
"learning_rate": 1.987515274291505e-05,
"loss": 1.0671875,
"step": 530
},
{
"epoch": 0.44481054365733114,
"grad_norm": 3.5747509002685547,
"learning_rate": 1.9859597748292355e-05,
"loss": 0.992578125,
"step": 540
},
{
"epoch": 0.45304777594728174,
"grad_norm": 2.901777982711792,
"learning_rate": 1.9843136598164257e-05,
"loss": 1.01875,
"step": 550
},
{
"epoch": 0.4612850082372323,
"grad_norm": 3.292612075805664,
"learning_rate": 1.982577080540809e-05,
"loss": 0.984765625,
"step": 560
},
{
"epoch": 0.4695222405271829,
"grad_norm": 3.03844952583313,
"learning_rate": 1.9807501966043186e-05,
"loss": 0.9421875,
"step": 570
},
{
"epoch": 0.47775947281713343,
"grad_norm": 3.5532376766204834,
"learning_rate": 1.978833175908423e-05,
"loss": 1.0203125,
"step": 580
},
{
"epoch": 0.48599670510708404,
"grad_norm": 3.1076343059539795,
"learning_rate": 1.976826194638693e-05,
"loss": 1.01484375,
"step": 590
},
{
"epoch": 0.4942339373970346,
"grad_norm": 3.1734344959259033,
"learning_rate": 1.974729437248609e-05,
"loss": 0.98515625,
"step": 600
},
{
"epoch": 0.5024711696869851,
"grad_norm": 3.0382258892059326,
"learning_rate": 1.972543096442611e-05,
"loss": 0.96953125,
"step": 610
},
{
"epoch": 0.5107084019769358,
"grad_norm": 3.5049564838409424,
"learning_rate": 1.9702673731583847e-05,
"loss": 0.99765625,
"step": 620
},
{
"epoch": 0.5189456342668863,
"grad_norm": 3.0992934703826904,
"learning_rate": 1.967902476548396e-05,
"loss": 0.992578125,
"step": 630
},
{
"epoch": 0.5271828665568369,
"grad_norm": 2.7654359340667725,
"learning_rate": 1.9654486239606686e-05,
"loss": 0.955859375,
"step": 640
},
{
"epoch": 0.5354200988467874,
"grad_norm": 3.8260226249694824,
"learning_rate": 1.962906040918808e-05,
"loss": 0.940625,
"step": 650
},
{
"epoch": 0.5436573311367381,
"grad_norm": 3.1025912761688232,
"learning_rate": 1.9602749611012738e-05,
"loss": 0.943359375,
"step": 660
},
{
"epoch": 0.5518945634266886,
"grad_norm": 3.2073378562927246,
"learning_rate": 1.9575556263199055e-05,
"loss": 0.919921875,
"step": 670
},
{
"epoch": 0.5601317957166392,
"grad_norm": 3.0796053409576416,
"learning_rate": 1.9547482864976964e-05,
"loss": 0.9953125,
"step": 680
},
{
"epoch": 0.5683690280065898,
"grad_norm": 3.2200849056243896,
"learning_rate": 1.951853199645825e-05,
"loss": 0.971484375,
"step": 690
},
{
"epoch": 0.5766062602965404,
"grad_norm": 2.985048770904541,
"learning_rate": 1.9488706318399426e-05,
"loss": 0.921484375,
"step": 700
},
{
"epoch": 0.5848434925864909,
"grad_norm": 3.4840943813323975,
"learning_rate": 1.945800857195718e-05,
"loss": 0.9640625,
"step": 710
},
{
"epoch": 0.5930807248764415,
"grad_norm": 3.4614648818969727,
"learning_rate": 1.942644157843647e-05,
"loss": 1.0015625,
"step": 720
},
{
"epoch": 0.6013179571663921,
"grad_norm": 3.204954147338867,
"learning_rate": 1.9394008239031203e-05,
"loss": 0.958984375,
"step": 730
},
{
"epoch": 0.6095551894563427,
"grad_norm": 3.055288314819336,
"learning_rate": 1.936071153455762e-05,
"loss": 0.95546875,
"step": 740
},
{
"epoch": 0.6177924217462932,
"grad_norm": 2.852473258972168,
"learning_rate": 1.932655452518033e-05,
"loss": 0.945703125,
"step": 750
},
{
"epoch": 0.6260296540362438,
"grad_norm": 3.1580588817596436,
"learning_rate": 1.9291540350131056e-05,
"loss": 0.917578125,
"step": 760
},
{
"epoch": 0.6342668863261944,
"grad_norm": 2.799454689025879,
"learning_rate": 1.925567222742014e-05,
"loss": 0.941015625,
"step": 770
},
{
"epoch": 0.642504118616145,
"grad_norm": 2.688756227493286,
"learning_rate": 1.9218953453540778e-05,
"loss": 0.909375,
"step": 780
},
{
"epoch": 0.6507413509060955,
"grad_norm": 3.1734654903411865,
"learning_rate": 1.9181387403166046e-05,
"loss": 0.916796875,
"step": 790
},
{
"epoch": 0.6589785831960461,
"grad_norm": 3.331912040710449,
"learning_rate": 1.9142977528838763e-05,
"loss": 0.912109375,
"step": 800
},
{
"epoch": 0.6672158154859967,
"grad_norm": 2.7767021656036377,
"learning_rate": 1.9103727360654166e-05,
"loss": 0.961328125,
"step": 810
},
{
"epoch": 0.6754530477759473,
"grad_norm": 2.9284727573394775,
"learning_rate": 1.906364050593548e-05,
"loss": 0.94296875,
"step": 820
},
{
"epoch": 0.6836902800658978,
"grad_norm": 3.007228136062622,
"learning_rate": 1.90227206489024e-05,
"loss": 0.9671875,
"step": 830
},
{
"epoch": 0.6919275123558485,
"grad_norm": 3.078132390975952,
"learning_rate": 1.8980971550332455e-05,
"loss": 0.920703125,
"step": 840
},
{
"epoch": 0.700164744645799,
"grad_norm": 3.112370491027832,
"learning_rate": 1.8938397047215392e-05,
"loss": 0.9328125,
"step": 850
},
{
"epoch": 0.7084019769357496,
"grad_norm": 2.957212209701538,
"learning_rate": 1.889500105240055e-05,
"loss": 0.904296875,
"step": 860
},
{
"epoch": 0.7166392092257001,
"grad_norm": 3.200221538543701,
"learning_rate": 1.88507875542372e-05,
"loss": 0.92265625,
"step": 870
},
{
"epoch": 0.7248764415156508,
"grad_norm": 2.9260823726654053,
"learning_rate": 1.8805760616208032e-05,
"loss": 0.957421875,
"step": 880
},
{
"epoch": 0.7331136738056013,
"grad_norm": 2.978455066680908,
"learning_rate": 1.875992437655569e-05,
"loss": 0.966796875,
"step": 890
},
{
"epoch": 0.7413509060955519,
"grad_norm": 2.8367488384246826,
"learning_rate": 1.871328304790242e-05,
"loss": 0.94375,
"step": 900
},
{
"epoch": 0.7495881383855024,
"grad_norm": 2.8874731063842773,
"learning_rate": 1.866584091686294e-05,
"loss": 0.9046875,
"step": 910
},
{
"epoch": 0.7578253706754531,
"grad_norm": 3.263575553894043,
"learning_rate": 1.8617602343650447e-05,
"loss": 0.8765625,
"step": 920
},
{
"epoch": 0.7660626029654036,
"grad_norm": 3.1561167240142822,
"learning_rate": 1.8568571761675893e-05,
"loss": 0.901171875,
"step": 930
},
{
"epoch": 0.7742998352553542,
"grad_norm": 2.7282521724700928,
"learning_rate": 1.851875367714054e-05,
"loss": 0.912890625,
"step": 940
},
{
"epoch": 0.7825370675453048,
"grad_norm": 2.757443428039551,
"learning_rate": 1.8468152668621807e-05,
"loss": 0.915625,
"step": 950
},
{
"epoch": 0.7907742998352554,
"grad_norm": 4.307620525360107,
"learning_rate": 1.8416773386652464e-05,
"loss": 0.9203125,
"step": 960
},
{
"epoch": 0.7990115321252059,
"grad_norm": 2.970149517059326,
"learning_rate": 1.8364620553293222e-05,
"loss": 0.870703125,
"step": 970
},
{
"epoch": 0.8072487644151565,
"grad_norm": 2.706763505935669,
"learning_rate": 1.8311698961698766e-05,
"loss": 0.916015625,
"step": 980
},
{
"epoch": 0.8154859967051071,
"grad_norm": 2.9329113960266113,
"learning_rate": 1.8258013475677216e-05,
"loss": 0.90859375,
"step": 990
},
{
"epoch": 0.8237232289950577,
"grad_norm": 2.8903682231903076,
"learning_rate": 1.8203569029243113e-05,
"loss": 0.96328125,
"step": 1000
},
{
"epoch": 0.8319604612850082,
"grad_norm": 2.7756187915802,
"learning_rate": 1.814837062616397e-05,
"loss": 0.91953125,
"step": 1010
},
{
"epoch": 0.8401976935749588,
"grad_norm": 2.7421905994415283,
"learning_rate": 1.8092423339500383e-05,
"loss": 0.89375,
"step": 1020
},
{
"epoch": 0.8484349258649094,
"grad_norm": 3.0162296295166016,
"learning_rate": 1.8035732311139798e-05,
"loss": 0.925,
"step": 1030
},
{
"epoch": 0.85667215815486,
"grad_norm": 2.7499682903289795,
"learning_rate": 1.797830275132392e-05,
"loss": 0.925390625,
"step": 1040
},
{
"epoch": 0.8649093904448105,
"grad_norm": 2.886056423187256,
"learning_rate": 1.792013993816988e-05,
"loss": 0.9,
"step": 1050
},
{
"epoch": 0.8731466227347611,
"grad_norm": 2.8466620445251465,
"learning_rate": 1.786124921718515e-05,
"loss": 0.917578125,
"step": 1060
},
{
"epoch": 0.8813838550247117,
"grad_norm": 3.1304376125335693,
"learning_rate": 1.7801636000776232e-05,
"loss": 0.912109375,
"step": 1070
},
{
"epoch": 0.8896210873146623,
"grad_norm": 2.630829334259033,
"learning_rate": 1.7741305767751256e-05,
"loss": 0.905078125,
"step": 1080
},
{
"epoch": 0.8978583196046128,
"grad_norm": 2.7537589073181152,
"learning_rate": 1.768026406281642e-05,
"loss": 0.899609375,
"step": 1090
},
{
"epoch": 0.9060955518945635,
"grad_norm": 2.9089252948760986,
"learning_rate": 1.7618516496066422e-05,
"loss": 0.8953125,
"step": 1100
},
{
"epoch": 0.914332784184514,
"grad_norm": 2.397491455078125,
"learning_rate": 1.7556068742468836e-05,
"loss": 0.8875,
"step": 1110
},
{
"epoch": 0.9225700164744646,
"grad_norm": 2.698484182357788,
"learning_rate": 1.7492926541342562e-05,
"loss": 0.87265625,
"step": 1120
},
{
"epoch": 0.9308072487644151,
"grad_norm": 2.7854061126708984,
"learning_rate": 1.7429095695830348e-05,
"loss": 0.867578125,
"step": 1130
},
{
"epoch": 0.9390444810543658,
"grad_norm": 2.9530394077301025,
"learning_rate": 1.7364582072365454e-05,
"loss": 0.912890625,
"step": 1140
},
{
"epoch": 0.9472817133443163,
"grad_norm": 2.7242469787597656,
"learning_rate": 1.7299391600132463e-05,
"loss": 0.91328125,
"step": 1150
},
{
"epoch": 0.9555189456342669,
"grad_norm": 2.798717975616455,
"learning_rate": 1.723353027052239e-05,
"loss": 0.9046875,
"step": 1160
},
{
"epoch": 0.9637561779242174,
"grad_norm": 2.965654134750366,
"learning_rate": 1.716700413658202e-05,
"loss": 0.891015625,
"step": 1170
},
{
"epoch": 0.9719934102141681,
"grad_norm": 2.8522415161132812,
"learning_rate": 1.709981931245759e-05,
"loss": 0.8828125,
"step": 1180
},
{
"epoch": 0.9802306425041186,
"grad_norm": 2.54638671875,
"learning_rate": 1.7031981972832875e-05,
"loss": 0.881640625,
"step": 1190
},
{
"epoch": 0.9884678747940692,
"grad_norm": 2.872195243835449,
"learning_rate": 1.6963498352361704e-05,
"loss": 0.884765625,
"step": 1200
},
{
"epoch": 0.9967051070840197,
"grad_norm": 3.168564558029175,
"learning_rate": 1.6894374745094935e-05,
"loss": 0.910546875,
"step": 1210
},
{
"epoch": 1.0,
"eval_loss": 0.975861668586731,
"eval_runtime": 275.0235,
"eval_samples_per_second": 15.646,
"eval_steps_per_second": 1.956,
"step": 1214
},
{
"epoch": 1.0049423393739703,
"grad_norm": 2.7183890342712402,
"learning_rate": 1.6824617503902012e-05,
"loss": 0.851171875,
"step": 1220
},
{
"epoch": 1.013179571663921,
"grad_norm": 2.986445665359497,
"learning_rate": 1.67542330398871e-05,
"loss": 0.8515625,
"step": 1230
},
{
"epoch": 1.0214168039538716,
"grad_norm": 3.4267704486846924,
"learning_rate": 1.6683227821799853e-05,
"loss": 0.840625,
"step": 1240
},
{
"epoch": 1.029654036243822,
"grad_norm": 3.226541757583618,
"learning_rate": 1.6611608375440914e-05,
"loss": 0.815625,
"step": 1250
},
{
"epoch": 1.0378912685337727,
"grad_norm": 3.0246481895446777,
"learning_rate": 1.6539381283062135e-05,
"loss": 0.823046875,
"step": 1260
},
{
"epoch": 1.0461285008237233,
"grad_norm": 3.1954500675201416,
"learning_rate": 1.6466553182761657e-05,
"loss": 0.794921875,
"step": 1270
},
{
"epoch": 1.0543657331136738,
"grad_norm": 3.0190796852111816,
"learning_rate": 1.6393130767873798e-05,
"loss": 0.836328125,
"step": 1280
},
{
"epoch": 1.0626029654036244,
"grad_norm": 3.312737226486206,
"learning_rate": 1.6319120786353918e-05,
"loss": 0.848046875,
"step": 1290
},
{
"epoch": 1.0708401976935749,
"grad_norm": 3.154492139816284,
"learning_rate": 1.624453004015824e-05,
"loss": 0.834375,
"step": 1300
},
{
"epoch": 1.0790774299835255,
"grad_norm": 3.3488335609436035,
"learning_rate": 1.6169365384618697e-05,
"loss": 0.8265625,
"step": 1310
},
{
"epoch": 1.0873146622734762,
"grad_norm": 3.1863484382629395,
"learning_rate": 1.6093633727812895e-05,
"loss": 0.8234375,
"step": 1320
},
{
"epoch": 1.0955518945634266,
"grad_norm": 3.4723682403564453,
"learning_rate": 1.6017342029929223e-05,
"loss": 0.833203125,
"step": 1330
},
{
"epoch": 1.1037891268533773,
"grad_norm": 3.222951650619507,
"learning_rate": 1.5940497302627165e-05,
"loss": 0.819140625,
"step": 1340
},
{
"epoch": 1.112026359143328,
"grad_norm": 3.2630727291107178,
"learning_rate": 1.5863106608392884e-05,
"loss": 0.8984375,
"step": 1350
},
{
"epoch": 1.1202635914332784,
"grad_norm": 3.097550868988037,
"learning_rate": 1.578517705989015e-05,
"loss": 0.823828125,
"step": 1360
},
{
"epoch": 1.128500823723229,
"grad_norm": 3.1272077560424805,
"learning_rate": 1.5706715819306622e-05,
"loss": 0.810546875,
"step": 1370
},
{
"epoch": 1.1367380560131797,
"grad_norm": 3.524237871170044,
"learning_rate": 1.562773009769564e-05,
"loss": 0.844921875,
"step": 1380
},
{
"epoch": 1.1449752883031301,
"grad_norm": 3.30476450920105,
"learning_rate": 1.5548227154313433e-05,
"loss": 0.8484375,
"step": 1390
},
{
"epoch": 1.1532125205930808,
"grad_norm": 3.4365389347076416,
"learning_rate": 1.546821429595199e-05,
"loss": 0.791796875,
"step": 1400
},
{
"epoch": 1.1614497528830312,
"grad_norm": 3.0440518856048584,
"learning_rate": 1.5387698876267517e-05,
"loss": 0.7921875,
"step": 1410
},
{
"epoch": 1.1696869851729819,
"grad_norm": 2.8585503101348877,
"learning_rate": 1.5306688295104575e-05,
"loss": 0.82734375,
"step": 1420
},
{
"epoch": 1.1779242174629325,
"grad_norm": 3.4312729835510254,
"learning_rate": 1.5225189997816006e-05,
"loss": 0.823828125,
"step": 1430
},
{
"epoch": 1.186161449752883,
"grad_norm": 3.015860080718994,
"learning_rate": 1.5143211474578658e-05,
"loss": 0.880859375,
"step": 1440
},
{
"epoch": 1.1943986820428336,
"grad_norm": 3.2433841228485107,
"learning_rate": 1.5060760259704988e-05,
"loss": 0.841015625,
"step": 1450
},
{
"epoch": 1.2026359143327843,
"grad_norm": 2.8961446285247803,
"learning_rate": 1.4977843930950622e-05,
"loss": 0.809375,
"step": 1460
},
{
"epoch": 1.2108731466227347,
"grad_norm": 3.448920249938965,
"learning_rate": 1.4894470108817905e-05,
"loss": 0.813671875,
"step": 1470
},
{
"epoch": 1.2191103789126854,
"grad_norm": 3.053058624267578,
"learning_rate": 1.4810646455855542e-05,
"loss": 0.82265625,
"step": 1480
},
{
"epoch": 1.227347611202636,
"grad_norm": 3.519761085510254,
"learning_rate": 1.4726380675954352e-05,
"loss": 0.8296875,
"step": 1490
},
{
"epoch": 1.2355848434925865,
"grad_norm": 3.0173768997192383,
"learning_rate": 1.4641680513639254e-05,
"loss": 0.76796875,
"step": 1500
},
{
"epoch": 1.2438220757825371,
"grad_norm": 3.2784814834594727,
"learning_rate": 1.4556553753357472e-05,
"loss": 0.84140625,
"step": 1510
},
{
"epoch": 1.2520593080724876,
"grad_norm": 2.9391214847564697,
"learning_rate": 1.4471008218763122e-05,
"loss": 0.796484375,
"step": 1520
},
{
"epoch": 1.2602965403624382,
"grad_norm": 3.34924578666687,
"learning_rate": 1.4385051771998159e-05,
"loss": 0.816015625,
"step": 1530
},
{
"epoch": 1.2685337726523889,
"grad_norm": 3.1663498878479004,
"learning_rate": 1.4298692312969808e-05,
"loss": 0.79375,
"step": 1540
},
{
"epoch": 1.2767710049423393,
"grad_norm": 3.537677526473999,
"learning_rate": 1.421193777862451e-05,
"loss": 0.810546875,
"step": 1550
},
{
"epoch": 1.28500823723229,
"grad_norm": 3.195664167404175,
"learning_rate": 1.4124796142218461e-05,
"loss": 0.81328125,
"step": 1560
},
{
"epoch": 1.2932454695222404,
"grad_norm": 3.368626117706299,
"learning_rate": 1.4037275412584851e-05,
"loss": 0.816015625,
"step": 1570
},
{
"epoch": 1.301482701812191,
"grad_norm": 2.9551095962524414,
"learning_rate": 1.3949383633397768e-05,
"loss": 0.78984375,
"step": 1580
},
{
"epoch": 1.3097199341021417,
"grad_norm": 3.085491418838501,
"learning_rate": 1.3861128882432961e-05,
"loss": 0.83671875,
"step": 1590
},
{
"epoch": 1.3179571663920924,
"grad_norm": 4.083651065826416,
"learning_rate": 1.3772519270825443e-05,
"loss": 0.788671875,
"step": 1600
},
{
"epoch": 1.3261943986820428,
"grad_norm": 3.4720098972320557,
"learning_rate": 1.3683562942324017e-05,
"loss": 0.816015625,
"step": 1610
},
{
"epoch": 1.3344316309719935,
"grad_norm": 3.243168354034424,
"learning_rate": 1.3594268072542824e-05,
"loss": 0.81875,
"step": 1620
},
{
"epoch": 1.342668863261944,
"grad_norm": 3.299379348754883,
"learning_rate": 1.3504642868209965e-05,
"loss": 0.81875,
"step": 1630
},
{
"epoch": 1.3509060955518946,
"grad_norm": 3.0587539672851562,
"learning_rate": 1.3414695566413241e-05,
"loss": 0.851953125,
"step": 1640
},
{
"epoch": 1.3591433278418452,
"grad_norm": 3.3016016483306885,
"learning_rate": 1.3324434433843112e-05,
"loss": 0.830078125,
"step": 1650
},
{
"epoch": 1.3673805601317957,
"grad_norm": 3.3171045780181885,
"learning_rate": 1.3233867766032965e-05,
"loss": 0.81875,
"step": 1660
},
{
"epoch": 1.3756177924217463,
"grad_norm": 3.0956857204437256,
"learning_rate": 1.314300388659667e-05,
"loss": 0.7796875,
"step": 1670
},
{
"epoch": 1.3838550247116967,
"grad_norm": 3.117342948913574,
"learning_rate": 1.3051851146463618e-05,
"loss": 0.78359375,
"step": 1680
},
{
"epoch": 1.3920922570016474,
"grad_norm": 3.4312281608581543,
"learning_rate": 1.2960417923111209e-05,
"loss": 0.804296875,
"step": 1690
},
{
"epoch": 1.400329489291598,
"grad_norm": 2.9448163509368896,
"learning_rate": 1.2868712619794913e-05,
"loss": 0.806640625,
"step": 1700
},
{
"epoch": 1.4085667215815487,
"grad_norm": 2.9322917461395264,
"learning_rate": 1.2776743664775958e-05,
"loss": 0.8125,
"step": 1710
},
{
"epoch": 1.4168039538714992,
"grad_norm": 2.9572620391845703,
"learning_rate": 1.268451951054674e-05,
"loss": 0.76796875,
"step": 1720
},
{
"epoch": 1.4250411861614498,
"grad_norm": 3.5945940017700195,
"learning_rate": 1.2592048633053963e-05,
"loss": 0.824609375,
"step": 1730
},
{
"epoch": 1.4332784184514002,
"grad_norm": 3.008906364440918,
"learning_rate": 1.2499339530919671e-05,
"loss": 0.760546875,
"step": 1740
},
{
"epoch": 1.441515650741351,
"grad_norm": 2.9414122104644775,
"learning_rate": 1.2406400724660155e-05,
"loss": 0.785546875,
"step": 1750
},
{
"epoch": 1.4497528830313016,
"grad_norm": 3.6647346019744873,
"learning_rate": 1.231324075590289e-05,
"loss": 0.831640625,
"step": 1760
},
{
"epoch": 1.457990115321252,
"grad_norm": 3.328450918197632,
"learning_rate": 1.2219868186601484e-05,
"loss": 0.807421875,
"step": 1770
},
{
"epoch": 1.4662273476112027,
"grad_norm": 3.4296746253967285,
"learning_rate": 1.2126291598248808e-05,
"loss": 0.837890625,
"step": 1780
},
{
"epoch": 1.474464579901153,
"grad_norm": 3.1492526531219482,
"learning_rate": 1.2032519591088284e-05,
"loss": 0.790234375,
"step": 1790
},
{
"epoch": 1.4827018121911038,
"grad_norm": 3.2090094089508057,
"learning_rate": 1.1938560783323483e-05,
"loss": 0.761328125,
"step": 1800
},
{
"epoch": 1.4909390444810544,
"grad_norm": 3.2001869678497314,
"learning_rate": 1.1844423810326062e-05,
"loss": 0.795703125,
"step": 1810
},
{
"epoch": 1.499176276771005,
"grad_norm": 3.036942958831787,
"learning_rate": 1.1750117323842123e-05,
"loss": 0.771484375,
"step": 1820
},
{
"epoch": 1.5074135090609555,
"grad_norm": 2.7320806980133057,
"learning_rate": 1.1655649991197052e-05,
"loss": 0.770703125,
"step": 1830
},
{
"epoch": 1.515650741350906,
"grad_norm": 3.2744085788726807,
"learning_rate": 1.1561030494498963e-05,
"loss": 0.808984375,
"step": 1840
},
{
"epoch": 1.5238879736408566,
"grad_norm": 3.292689561843872,
"learning_rate": 1.1466267529840737e-05,
"loss": 0.799609375,
"step": 1850
},
{
"epoch": 1.5321252059308073,
"grad_norm": 3.570671796798706,
"learning_rate": 1.1371369806500828e-05,
"loss": 0.780078125,
"step": 1860
},
{
"epoch": 1.540362438220758,
"grad_norm": 3.117856740951538,
"learning_rate": 1.127634604614279e-05,
"loss": 0.79140625,
"step": 1870
},
{
"epoch": 1.5485996705107083,
"grad_norm": 3.2196526527404785,
"learning_rate": 1.1181204982013744e-05,
"loss": 0.801171875,
"step": 1880
},
{
"epoch": 1.556836902800659,
"grad_norm": 3.4039859771728516,
"learning_rate": 1.1085955358141717e-05,
"loss": 0.796484375,
"step": 1890
},
{
"epoch": 1.5650741350906094,
"grad_norm": 2.81375789642334,
"learning_rate": 1.0990605928532023e-05,
"loss": 0.790625,
"step": 1900
},
{
"epoch": 1.57331136738056,
"grad_norm": 3.5346741676330566,
"learning_rate": 1.0895165456362707e-05,
"loss": 0.77734375,
"step": 1910
},
{
"epoch": 1.5815485996705108,
"grad_norm": 3.2716338634490967,
"learning_rate": 1.0799642713179177e-05,
"loss": 0.781640625,
"step": 1920
},
{
"epoch": 1.5897858319604614,
"grad_norm": 3.0632574558258057,
"learning_rate": 1.0704046478088022e-05,
"loss": 0.81640625,
"step": 1930
},
{
"epoch": 1.5980230642504119,
"grad_norm": 3.2588024139404297,
"learning_rate": 1.0608385536950188e-05,
"loss": 0.768359375,
"step": 1940
},
{
"epoch": 1.6062602965403623,
"grad_norm": 2.9469897747039795,
"learning_rate": 1.0512668681573476e-05,
"loss": 0.796484375,
"step": 1950
},
{
"epoch": 1.614497528830313,
"grad_norm": 3.055677890777588,
"learning_rate": 1.041690470890455e-05,
"loss": 0.76171875,
"step": 1960
},
{
"epoch": 1.6227347611202636,
"grad_norm": 3.449039936065674,
"learning_rate": 1.0321102420220418e-05,
"loss": 0.812890625,
"step": 1970
},
{
"epoch": 1.6309719934102143,
"grad_norm": 3.1603500843048096,
"learning_rate": 1.0225270620319576e-05,
"loss": 0.7859375,
"step": 1980
},
{
"epoch": 1.6392092257001647,
"grad_norm": 3.036633253097534,
"learning_rate": 1.0129418116712755e-05,
"loss": 0.8078125,
"step": 1990
},
{
"epoch": 1.6474464579901154,
"grad_norm": 3.318446397781372,
"learning_rate": 1.0033553718813497e-05,
"loss": 0.7828125,
"step": 2000
},
{
"epoch": 1.6556836902800658,
"grad_norm": 3.854395866394043,
"learning_rate": 9.937686237128488e-06,
"loss": 0.79375,
"step": 2010
},
{
"epoch": 1.6639209225700164,
"grad_norm": 3.301496744155884,
"learning_rate": 9.841824482447837e-06,
"loss": 0.773046875,
"step": 2020
},
{
"epoch": 1.672158154859967,
"grad_norm": 3.150346040725708,
"learning_rate": 9.745977265035313e-06,
"loss": 0.815234375,
"step": 2030
},
{
"epoch": 1.6803953871499178,
"grad_norm": 3.222334146499634,
"learning_rate": 9.65015339381861e-06,
"loss": 0.798046875,
"step": 2040
},
{
"epoch": 1.6886326194398682,
"grad_norm": 3.303046226501465,
"learning_rate": 9.554361675579783e-06,
"loss": 0.791015625,
"step": 2050
},
{
"epoch": 1.6968698517298186,
"grad_norm": 3.5736613273620605,
"learning_rate": 9.458610914145826e-06,
"loss": 0.78515625,
"step": 2060
},
{
"epoch": 1.7051070840197693,
"grad_norm": 3.3751983642578125,
"learning_rate": 9.362909909579567e-06,
"loss": 0.8265625,
"step": 2070
},
{
"epoch": 1.71334431630972,
"grad_norm": 2.9090216159820557,
"learning_rate": 9.267267457370878e-06,
"loss": 0.7953125,
"step": 2080
},
{
"epoch": 1.7215815485996706,
"grad_norm": 3.2832682132720947,
"learning_rate": 9.171692347628324e-06,
"loss": 0.786328125,
"step": 2090
},
{
"epoch": 1.729818780889621,
"grad_norm": 3.343050718307495,
"learning_rate": 9.076193364271303e-06,
"loss": 0.759765625,
"step": 2100
},
{
"epoch": 1.7380560131795717,
"grad_norm": 2.9476571083068848,
"learning_rate": 8.98077928422274e-06,
"loss": 0.7828125,
"step": 2110
},
{
"epoch": 1.7462932454695221,
"grad_norm": 2.922978401184082,
"learning_rate": 8.885458876602442e-06,
"loss": 0.787890625,
"step": 2120
},
{
"epoch": 1.7545304777594728,
"grad_norm": 3.898466110229492,
"learning_rate": 8.790240901921167e-06,
"loss": 0.78828125,
"step": 2130
},
{
"epoch": 1.7627677100494235,
"grad_norm": 3.062774419784546,
"learning_rate": 8.695134111275481e-06,
"loss": 0.7859375,
"step": 2140
},
{
"epoch": 1.771004942339374,
"grad_norm": 3.3497931957244873,
"learning_rate": 8.600147245543477e-06,
"loss": 0.801171875,
"step": 2150
},
{
"epoch": 1.7792421746293245,
"grad_norm": 3.4660985469818115,
"learning_rate": 8.50528903458143e-06,
"loss": 0.78671875,
"step": 2160
},
{
"epoch": 1.787479406919275,
"grad_norm": 3.3449573516845703,
"learning_rate": 8.410568196421482e-06,
"loss": 0.773046875,
"step": 2170
},
{
"epoch": 1.7957166392092256,
"grad_norm": 3.444862127304077,
"learning_rate": 8.315993436470408e-06,
"loss": 0.809375,
"step": 2180
},
{
"epoch": 1.8039538714991763,
"grad_norm": 3.3574137687683105,
"learning_rate": 8.221573446709524e-06,
"loss": 0.756640625,
"step": 2190
},
{
"epoch": 1.812191103789127,
"grad_norm": 3.180471658706665,
"learning_rate": 8.12731690489584e-06,
"loss": 0.789453125,
"step": 2200
},
{
"epoch": 1.8204283360790774,
"grad_norm": 2.790720224380493,
"learning_rate": 8.033232473764539e-06,
"loss": 0.730859375,
"step": 2210
},
{
"epoch": 1.828665568369028,
"grad_norm": 3.5747971534729004,
"learning_rate": 7.93932880023281e-06,
"loss": 0.79140625,
"step": 2220
},
{
"epoch": 1.8369028006589785,
"grad_norm": 3.3052978515625,
"learning_rate": 7.845614514605151e-06,
"loss": 0.783203125,
"step": 2230
},
{
"epoch": 1.8451400329489291,
"grad_norm": 3.0872039794921875,
"learning_rate": 7.752098229780177e-06,
"loss": 0.804296875,
"step": 2240
},
{
"epoch": 1.8533772652388798,
"grad_norm": 3.7297873497009277,
"learning_rate": 7.658788540459063e-06,
"loss": 0.7921875,
"step": 2250
},
{
"epoch": 1.8616144975288305,
"grad_norm": 3.423245668411255,
"learning_rate": 7.565694022355629e-06,
"loss": 0.75625,
"step": 2260
},
{
"epoch": 1.869851729818781,
"grad_norm": 3.377229690551758,
"learning_rate": 7.472823231408195e-06,
"loss": 0.790234375,
"step": 2270
},
{
"epoch": 1.8780889621087313,
"grad_norm": 3.188108205795288,
"learning_rate": 7.380184702993207e-06,
"loss": 0.773828125,
"step": 2280
},
{
"epoch": 1.886326194398682,
"grad_norm": 3.0406744480133057,
"learning_rate": 7.287786951140825e-06,
"loss": 0.775,
"step": 2290
},
{
"epoch": 1.8945634266886326,
"grad_norm": 3.003800868988037,
"learning_rate": 7.1956384677524084e-06,
"loss": 0.742578125,
"step": 2300
},
{
"epoch": 1.9028006589785833,
"grad_norm": 3.3765439987182617,
"learning_rate": 7.103747721820069e-06,
"loss": 0.787109375,
"step": 2310
},
{
"epoch": 1.9110378912685337,
"grad_norm": 3.0775868892669678,
"learning_rate": 7.012123158648312e-06,
"loss": 0.749609375,
"step": 2320
},
{
"epoch": 1.9192751235584844,
"grad_norm": 3.8198721408843994,
"learning_rate": 6.920773199077872e-06,
"loss": 0.74765625,
"step": 2330
},
{
"epoch": 1.9275123558484348,
"grad_norm": 3.314674139022827,
"learning_rate": 6.829706238711786e-06,
"loss": 0.81015625,
"step": 2340
},
{
"epoch": 1.9357495881383855,
"grad_norm": 3.489034652709961,
"learning_rate": 6.738930647143789e-06,
"loss": 0.8,
"step": 2350
},
{
"epoch": 1.9439868204283361,
"grad_norm": 3.2915072441101074,
"learning_rate": 6.648454767189082e-06,
"loss": 0.74453125,
"step": 2360
},
{
"epoch": 1.9522240527182868,
"grad_norm": 3.420825958251953,
"learning_rate": 6.558286914117603e-06,
"loss": 0.7484375,
"step": 2370
},
{
"epoch": 1.9604612850082372,
"grad_norm": 3.3860113620758057,
"learning_rate": 6.468435374889791e-06,
"loss": 0.75703125,
"step": 2380
},
{
"epoch": 1.9686985172981877,
"grad_norm": 3.3130035400390625,
"learning_rate": 6.378908407394968e-06,
"loss": 0.76015625,
"step": 2390
},
{
"epoch": 1.9769357495881383,
"grad_norm": 3.723069667816162,
"learning_rate": 6.28971423969238e-06,
"loss": 0.780859375,
"step": 2400
},
{
"epoch": 1.985172981878089,
"grad_norm": 3.632185220718384,
"learning_rate": 6.200861069255006e-06,
"loss": 0.77578125,
"step": 2410
},
{
"epoch": 1.9934102141680397,
"grad_norm": 3.76944899559021,
"learning_rate": 6.1123570622161565e-06,
"loss": 0.742578125,
"step": 2420
},
{
"epoch": 2.0,
"eval_loss": 0.9797534346580505,
"eval_runtime": 189.672,
"eval_samples_per_second": 22.687,
"eval_steps_per_second": 2.836,
"step": 2428
}
],
"logging_steps": 10,
"max_steps": 3642,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.2699260600150655e+20,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}