moldir_new_exp / checkpoint-5000 /trainer_state.json
Sarim-Hash's picture
Upload folder using huggingface_hub
5cd5e10 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.1221065626665918,
"eval_steps": 500,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0022446059313711735,
"grad_norm": 7.085300922393799,
"learning_rate": 1.3462976813762156e-07,
"loss": 0.7522,
"step": 10
},
{
"epoch": 0.004489211862742347,
"grad_norm": 6.747836112976074,
"learning_rate": 2.842183994016455e-07,
"loss": 0.7426,
"step": 20
},
{
"epoch": 0.006733817794113521,
"grad_norm": 6.484082221984863,
"learning_rate": 4.3380703066566945e-07,
"loss": 0.7352,
"step": 30
},
{
"epoch": 0.008978423725484694,
"grad_norm": 5.182391166687012,
"learning_rate": 5.833956619296934e-07,
"loss": 0.6797,
"step": 40
},
{
"epoch": 0.011223029656855869,
"grad_norm": 3.4766058921813965,
"learning_rate": 7.329842931937173e-07,
"loss": 0.5879,
"step": 50
},
{
"epoch": 0.013467635588227042,
"grad_norm": 2.9406728744506836,
"learning_rate": 8.825729244577413e-07,
"loss": 0.4917,
"step": 60
},
{
"epoch": 0.015712241519598217,
"grad_norm": 2.0085201263427734,
"learning_rate": 1.0321615557217653e-06,
"loss": 0.362,
"step": 70
},
{
"epoch": 0.01795684745096939,
"grad_norm": 1.118497610092163,
"learning_rate": 1.1817501869857892e-06,
"loss": 0.2813,
"step": 80
},
{
"epoch": 0.020201453382340563,
"grad_norm": 0.764472484588623,
"learning_rate": 1.3313388182498132e-06,
"loss": 0.2272,
"step": 90
},
{
"epoch": 0.022446059313711738,
"grad_norm": 0.5994025468826294,
"learning_rate": 1.480927449513837e-06,
"loss": 0.1925,
"step": 100
},
{
"epoch": 0.02469066524508291,
"grad_norm": 0.5585537552833557,
"learning_rate": 1.630516080777861e-06,
"loss": 0.1728,
"step": 110
},
{
"epoch": 0.026935271176454084,
"grad_norm": 0.4982088804244995,
"learning_rate": 1.780104712041885e-06,
"loss": 0.1535,
"step": 120
},
{
"epoch": 0.02917987710782526,
"grad_norm": 0.5461462736129761,
"learning_rate": 1.929693343305909e-06,
"loss": 0.142,
"step": 130
},
{
"epoch": 0.031424483039196434,
"grad_norm": 0.4731317460536957,
"learning_rate": 2.0792819745699327e-06,
"loss": 0.1325,
"step": 140
},
{
"epoch": 0.033669088970567605,
"grad_norm": 0.4937710165977478,
"learning_rate": 2.228870605833957e-06,
"loss": 0.1237,
"step": 150
},
{
"epoch": 0.03591369490193878,
"grad_norm": 0.4619077742099762,
"learning_rate": 2.378459237097981e-06,
"loss": 0.1134,
"step": 160
},
{
"epoch": 0.038158300833309955,
"grad_norm": 0.4655718207359314,
"learning_rate": 2.5280478683620047e-06,
"loss": 0.1082,
"step": 170
},
{
"epoch": 0.040402906764681126,
"grad_norm": 0.6071754097938538,
"learning_rate": 2.6776364996260284e-06,
"loss": 0.103,
"step": 180
},
{
"epoch": 0.0426475126960523,
"grad_norm": 0.44958844780921936,
"learning_rate": 2.8272251308900526e-06,
"loss": 0.1004,
"step": 190
},
{
"epoch": 0.044892118627423476,
"grad_norm": 0.4919949471950531,
"learning_rate": 2.9768137621540767e-06,
"loss": 0.0967,
"step": 200
},
{
"epoch": 0.04713672455879465,
"grad_norm": 0.4707821011543274,
"learning_rate": 3.1264023934181005e-06,
"loss": 0.0912,
"step": 210
},
{
"epoch": 0.04938133049016582,
"grad_norm": 0.4630533456802368,
"learning_rate": 3.2759910246821242e-06,
"loss": 0.0895,
"step": 220
},
{
"epoch": 0.051625936421537,
"grad_norm": 0.5952326059341431,
"learning_rate": 3.4255796559461484e-06,
"loss": 0.0877,
"step": 230
},
{
"epoch": 0.05387054235290817,
"grad_norm": 0.47812679409980774,
"learning_rate": 3.575168287210172e-06,
"loss": 0.0851,
"step": 240
},
{
"epoch": 0.05611514828427934,
"grad_norm": 0.5016306638717651,
"learning_rate": 3.7247569184741963e-06,
"loss": 0.0808,
"step": 250
},
{
"epoch": 0.05835975421565052,
"grad_norm": 0.49919837713241577,
"learning_rate": 3.874345549738221e-06,
"loss": 0.0803,
"step": 260
},
{
"epoch": 0.06060436014702169,
"grad_norm": 0.4146517813205719,
"learning_rate": 4.0239341810022446e-06,
"loss": 0.0777,
"step": 270
},
{
"epoch": 0.06284896607839287,
"grad_norm": 0.4712880849838257,
"learning_rate": 4.173522812266268e-06,
"loss": 0.0788,
"step": 280
},
{
"epoch": 0.06509357200976404,
"grad_norm": 0.45096713304519653,
"learning_rate": 4.323111443530292e-06,
"loss": 0.074,
"step": 290
},
{
"epoch": 0.06733817794113521,
"grad_norm": 0.4806267023086548,
"learning_rate": 4.472700074794316e-06,
"loss": 0.072,
"step": 300
},
{
"epoch": 0.06958278387250638,
"grad_norm": 0.43456050753593445,
"learning_rate": 4.6222887060583395e-06,
"loss": 0.0716,
"step": 310
},
{
"epoch": 0.07182738980387755,
"grad_norm": 0.3963911533355713,
"learning_rate": 4.771877337322364e-06,
"loss": 0.0713,
"step": 320
},
{
"epoch": 0.07407199573524872,
"grad_norm": 0.36991456151008606,
"learning_rate": 4.921465968586388e-06,
"loss": 0.067,
"step": 330
},
{
"epoch": 0.07631660166661991,
"grad_norm": 0.5009816288948059,
"learning_rate": 5.0710545998504115e-06,
"loss": 0.068,
"step": 340
},
{
"epoch": 0.07856120759799108,
"grad_norm": 0.46386927366256714,
"learning_rate": 5.220643231114436e-06,
"loss": 0.0683,
"step": 350
},
{
"epoch": 0.08080581352936225,
"grad_norm": 0.508665144443512,
"learning_rate": 5.37023186237846e-06,
"loss": 0.0672,
"step": 360
},
{
"epoch": 0.08305041946073342,
"grad_norm": 0.41486406326293945,
"learning_rate": 5.5198204936424836e-06,
"loss": 0.0649,
"step": 370
},
{
"epoch": 0.0852950253921046,
"grad_norm": 0.4998893737792969,
"learning_rate": 5.669409124906507e-06,
"loss": 0.0658,
"step": 380
},
{
"epoch": 0.08753963132347577,
"grad_norm": 0.5872018933296204,
"learning_rate": 5.818997756170531e-06,
"loss": 0.0626,
"step": 390
},
{
"epoch": 0.08978423725484695,
"grad_norm": 0.5181050896644592,
"learning_rate": 5.968586387434555e-06,
"loss": 0.0623,
"step": 400
},
{
"epoch": 0.09202884318621812,
"grad_norm": 0.4561860263347626,
"learning_rate": 6.1181750186985785e-06,
"loss": 0.0637,
"step": 410
},
{
"epoch": 0.0942734491175893,
"grad_norm": 0.4229351878166199,
"learning_rate": 6.267763649962604e-06,
"loss": 0.0626,
"step": 420
},
{
"epoch": 0.09651805504896047,
"grad_norm": 0.43529367446899414,
"learning_rate": 6.417352281226628e-06,
"loss": 0.0623,
"step": 430
},
{
"epoch": 0.09876266098033164,
"grad_norm": 0.44300997257232666,
"learning_rate": 6.566940912490651e-06,
"loss": 0.062,
"step": 440
},
{
"epoch": 0.10100726691170281,
"grad_norm": 0.415102481842041,
"learning_rate": 6.716529543754675e-06,
"loss": 0.0583,
"step": 450
},
{
"epoch": 0.103251872843074,
"grad_norm": 0.39811965823173523,
"learning_rate": 6.866118175018699e-06,
"loss": 0.0569,
"step": 460
},
{
"epoch": 0.10549647877444517,
"grad_norm": 0.4333321750164032,
"learning_rate": 7.015706806282723e-06,
"loss": 0.0615,
"step": 470
},
{
"epoch": 0.10774108470581634,
"grad_norm": 0.379936546087265,
"learning_rate": 7.165295437546746e-06,
"loss": 0.0552,
"step": 480
},
{
"epoch": 0.10998569063718751,
"grad_norm": 0.3810630142688751,
"learning_rate": 7.314884068810772e-06,
"loss": 0.0564,
"step": 490
},
{
"epoch": 0.11223029656855868,
"grad_norm": 0.38735371828079224,
"learning_rate": 7.4644727000747955e-06,
"loss": 0.0553,
"step": 500
},
{
"epoch": 0.11447490249992985,
"grad_norm": 0.41105860471725464,
"learning_rate": 7.614061331338819e-06,
"loss": 0.0552,
"step": 510
},
{
"epoch": 0.11671950843130104,
"grad_norm": 0.41426989436149597,
"learning_rate": 7.763649962602843e-06,
"loss": 0.0538,
"step": 520
},
{
"epoch": 0.11896411436267221,
"grad_norm": 0.3534553647041321,
"learning_rate": 7.913238593866866e-06,
"loss": 0.0529,
"step": 530
},
{
"epoch": 0.12120872029404338,
"grad_norm": 0.4285803437232971,
"learning_rate": 8.06282722513089e-06,
"loss": 0.0509,
"step": 540
},
{
"epoch": 0.12345332622541455,
"grad_norm": 0.40142738819122314,
"learning_rate": 8.212415856394913e-06,
"loss": 0.0544,
"step": 550
},
{
"epoch": 0.12569793215678574,
"grad_norm": 0.434146910905838,
"learning_rate": 8.36200448765894e-06,
"loss": 0.0514,
"step": 560
},
{
"epoch": 0.1279425380881569,
"grad_norm": 0.421758770942688,
"learning_rate": 8.511593118922962e-06,
"loss": 0.0491,
"step": 570
},
{
"epoch": 0.13018714401952808,
"grad_norm": 0.41949188709259033,
"learning_rate": 8.661181750186987e-06,
"loss": 0.051,
"step": 580
},
{
"epoch": 0.13243174995089924,
"grad_norm": 0.34015727043151855,
"learning_rate": 8.81077038145101e-06,
"loss": 0.0491,
"step": 590
},
{
"epoch": 0.13467635588227042,
"grad_norm": 0.3929171562194824,
"learning_rate": 8.960359012715034e-06,
"loss": 0.0477,
"step": 600
},
{
"epoch": 0.1369209618136416,
"grad_norm": 0.41220733523368835,
"learning_rate": 9.109947643979057e-06,
"loss": 0.0483,
"step": 610
},
{
"epoch": 0.13916556774501276,
"grad_norm": 0.38868510723114014,
"learning_rate": 9.259536275243082e-06,
"loss": 0.0482,
"step": 620
},
{
"epoch": 0.14141017367638395,
"grad_norm": 0.35203954577445984,
"learning_rate": 9.409124906507107e-06,
"loss": 0.0467,
"step": 630
},
{
"epoch": 0.1436547796077551,
"grad_norm": 0.35794001817703247,
"learning_rate": 9.558713537771131e-06,
"loss": 0.0467,
"step": 640
},
{
"epoch": 0.1458993855391263,
"grad_norm": 0.33909112215042114,
"learning_rate": 9.708302169035154e-06,
"loss": 0.0453,
"step": 650
},
{
"epoch": 0.14814399147049745,
"grad_norm": 0.3420808017253876,
"learning_rate": 9.857890800299179e-06,
"loss": 0.0473,
"step": 660
},
{
"epoch": 0.15038859740186863,
"grad_norm": 0.3780367970466614,
"learning_rate": 1.0007479431563203e-05,
"loss": 0.046,
"step": 670
},
{
"epoch": 0.15263320333323982,
"grad_norm": 0.3484839200973511,
"learning_rate": 1.0157068062827226e-05,
"loss": 0.0443,
"step": 680
},
{
"epoch": 0.15487780926461098,
"grad_norm": 0.3256385624408722,
"learning_rate": 1.030665669409125e-05,
"loss": 0.0468,
"step": 690
},
{
"epoch": 0.15712241519598216,
"grad_norm": 0.3509197235107422,
"learning_rate": 1.0456245325355273e-05,
"loss": 0.0467,
"step": 700
},
{
"epoch": 0.15936702112735332,
"grad_norm": 0.3882471024990082,
"learning_rate": 1.0605833956619298e-05,
"loss": 0.0441,
"step": 710
},
{
"epoch": 0.1616116270587245,
"grad_norm": 0.41281041502952576,
"learning_rate": 1.0755422587883323e-05,
"loss": 0.045,
"step": 720
},
{
"epoch": 0.16385623299009566,
"grad_norm": 0.34965190291404724,
"learning_rate": 1.0905011219147346e-05,
"loss": 0.0472,
"step": 730
},
{
"epoch": 0.16610083892146685,
"grad_norm": 0.3685629367828369,
"learning_rate": 1.105459985041137e-05,
"loss": 0.043,
"step": 740
},
{
"epoch": 0.16834544485283803,
"grad_norm": 0.3236459791660309,
"learning_rate": 1.1204188481675393e-05,
"loss": 0.0446,
"step": 750
},
{
"epoch": 0.1705900507842092,
"grad_norm": 0.4009409546852112,
"learning_rate": 1.1353777112939418e-05,
"loss": 0.0432,
"step": 760
},
{
"epoch": 0.17283465671558038,
"grad_norm": 0.3602420389652252,
"learning_rate": 1.150336574420344e-05,
"loss": 0.0421,
"step": 770
},
{
"epoch": 0.17507926264695153,
"grad_norm": 0.30885422229766846,
"learning_rate": 1.1652954375467465e-05,
"loss": 0.0428,
"step": 780
},
{
"epoch": 0.17732386857832272,
"grad_norm": 0.3470623791217804,
"learning_rate": 1.1802543006731488e-05,
"loss": 0.0405,
"step": 790
},
{
"epoch": 0.1795684745096939,
"grad_norm": 0.31664037704467773,
"learning_rate": 1.1952131637995514e-05,
"loss": 0.0414,
"step": 800
},
{
"epoch": 0.18181308044106506,
"grad_norm": 0.28882402181625366,
"learning_rate": 1.2101720269259539e-05,
"loss": 0.042,
"step": 810
},
{
"epoch": 0.18405768637243625,
"grad_norm": 0.341634064912796,
"learning_rate": 1.2251308900523562e-05,
"loss": 0.0419,
"step": 820
},
{
"epoch": 0.1863022923038074,
"grad_norm": 0.32544925808906555,
"learning_rate": 1.2400897531787586e-05,
"loss": 0.0386,
"step": 830
},
{
"epoch": 0.1885468982351786,
"grad_norm": 0.32963812351226807,
"learning_rate": 1.2550486163051609e-05,
"loss": 0.0433,
"step": 840
},
{
"epoch": 0.19079150416654975,
"grad_norm": 0.34846410155296326,
"learning_rate": 1.2700074794315634e-05,
"loss": 0.0425,
"step": 850
},
{
"epoch": 0.19303611009792093,
"grad_norm": 0.4119662642478943,
"learning_rate": 1.2849663425579657e-05,
"loss": 0.0411,
"step": 860
},
{
"epoch": 0.19528071602929212,
"grad_norm": 0.36676207184791565,
"learning_rate": 1.2999252056843681e-05,
"loss": 0.0407,
"step": 870
},
{
"epoch": 0.19752532196066327,
"grad_norm": 0.320718914270401,
"learning_rate": 1.3148840688107704e-05,
"loss": 0.0408,
"step": 880
},
{
"epoch": 0.19976992789203446,
"grad_norm": 0.297377347946167,
"learning_rate": 1.3298429319371729e-05,
"loss": 0.0393,
"step": 890
},
{
"epoch": 0.20201453382340562,
"grad_norm": 0.29334571957588196,
"learning_rate": 1.3448017950635752e-05,
"loss": 0.0404,
"step": 900
},
{
"epoch": 0.2042591397547768,
"grad_norm": 0.3002082407474518,
"learning_rate": 1.3597606581899776e-05,
"loss": 0.0381,
"step": 910
},
{
"epoch": 0.206503745686148,
"grad_norm": 0.3855285942554474,
"learning_rate": 1.3747195213163799e-05,
"loss": 0.0395,
"step": 920
},
{
"epoch": 0.20874835161751915,
"grad_norm": 0.37728843092918396,
"learning_rate": 1.3896783844427824e-05,
"loss": 0.0387,
"step": 930
},
{
"epoch": 0.21099295754889033,
"grad_norm": 0.3107747733592987,
"learning_rate": 1.404637247569185e-05,
"loss": 0.0365,
"step": 940
},
{
"epoch": 0.2132375634802615,
"grad_norm": 0.31299862265586853,
"learning_rate": 1.4195961106955873e-05,
"loss": 0.0386,
"step": 950
},
{
"epoch": 0.21548216941163267,
"grad_norm": 0.2866477072238922,
"learning_rate": 1.4345549738219897e-05,
"loss": 0.0384,
"step": 960
},
{
"epoch": 0.21772677534300383,
"grad_norm": 0.2896515429019928,
"learning_rate": 1.449513836948392e-05,
"loss": 0.0368,
"step": 970
},
{
"epoch": 0.21997138127437502,
"grad_norm": 0.3208376169204712,
"learning_rate": 1.4644727000747945e-05,
"loss": 0.035,
"step": 980
},
{
"epoch": 0.2222159872057462,
"grad_norm": 0.26684877276420593,
"learning_rate": 1.479431563201197e-05,
"loss": 0.0386,
"step": 990
},
{
"epoch": 0.22446059313711736,
"grad_norm": 0.2770873010158539,
"learning_rate": 1.4943904263275992e-05,
"loss": 0.0359,
"step": 1000
},
{
"epoch": 0.22670519906848854,
"grad_norm": 0.33455002307891846,
"learning_rate": 1.5093492894540017e-05,
"loss": 0.0386,
"step": 1010
},
{
"epoch": 0.2289498049998597,
"grad_norm": 0.2847146689891815,
"learning_rate": 1.524308152580404e-05,
"loss": 0.036,
"step": 1020
},
{
"epoch": 0.2311944109312309,
"grad_norm": 0.3237374722957611,
"learning_rate": 1.5392670157068064e-05,
"loss": 0.0383,
"step": 1030
},
{
"epoch": 0.23343901686260207,
"grad_norm": 0.26867830753326416,
"learning_rate": 1.5542258788332087e-05,
"loss": 0.0336,
"step": 1040
},
{
"epoch": 0.23568362279397323,
"grad_norm": 0.3390122056007385,
"learning_rate": 1.569184741959611e-05,
"loss": 0.037,
"step": 1050
},
{
"epoch": 0.23792822872534442,
"grad_norm": 0.3243669867515564,
"learning_rate": 1.5841436050860136e-05,
"loss": 0.0369,
"step": 1060
},
{
"epoch": 0.24017283465671557,
"grad_norm": 0.2567446827888489,
"learning_rate": 1.599102468212416e-05,
"loss": 0.0343,
"step": 1070
},
{
"epoch": 0.24241744058808676,
"grad_norm": 0.3240516781806946,
"learning_rate": 1.6140613313388185e-05,
"loss": 0.0361,
"step": 1080
},
{
"epoch": 0.24466204651945792,
"grad_norm": 0.30721980333328247,
"learning_rate": 1.629020194465221e-05,
"loss": 0.0351,
"step": 1090
},
{
"epoch": 0.2469066524508291,
"grad_norm": 0.2947623133659363,
"learning_rate": 1.643979057591623e-05,
"loss": 0.0349,
"step": 1100
},
{
"epoch": 0.24915125838220029,
"grad_norm": 0.3040514290332794,
"learning_rate": 1.6589379207180258e-05,
"loss": 0.0347,
"step": 1110
},
{
"epoch": 0.25139586431357147,
"grad_norm": 0.2889814078807831,
"learning_rate": 1.673896783844428e-05,
"loss": 0.0336,
"step": 1120
},
{
"epoch": 0.25364047024494263,
"grad_norm": 0.27692848443984985,
"learning_rate": 1.6888556469708303e-05,
"loss": 0.0358,
"step": 1130
},
{
"epoch": 0.2558850761763138,
"grad_norm": 0.344008207321167,
"learning_rate": 1.7038145100972326e-05,
"loss": 0.0356,
"step": 1140
},
{
"epoch": 0.25812968210768494,
"grad_norm": 0.2950272858142853,
"learning_rate": 1.7187733732236352e-05,
"loss": 0.0351,
"step": 1150
},
{
"epoch": 0.26037428803905616,
"grad_norm": 0.28355881571769714,
"learning_rate": 1.7337322363500375e-05,
"loss": 0.0328,
"step": 1160
},
{
"epoch": 0.2626188939704273,
"grad_norm": 0.3065374493598938,
"learning_rate": 1.7486910994764398e-05,
"loss": 0.0354,
"step": 1170
},
{
"epoch": 0.26486349990179847,
"grad_norm": 0.249955415725708,
"learning_rate": 1.763649962602842e-05,
"loss": 0.0342,
"step": 1180
},
{
"epoch": 0.2671081058331697,
"grad_norm": 0.26293373107910156,
"learning_rate": 1.7786088257292447e-05,
"loss": 0.0334,
"step": 1190
},
{
"epoch": 0.26935271176454084,
"grad_norm": 0.27912405133247375,
"learning_rate": 1.793567688855647e-05,
"loss": 0.0341,
"step": 1200
},
{
"epoch": 0.271597317695912,
"grad_norm": 0.2818824052810669,
"learning_rate": 1.8085265519820493e-05,
"loss": 0.0334,
"step": 1210
},
{
"epoch": 0.2738419236272832,
"grad_norm": 0.3036912977695465,
"learning_rate": 1.823485415108452e-05,
"loss": 0.0321,
"step": 1220
},
{
"epoch": 0.27608652955865437,
"grad_norm": 0.24145382642745972,
"learning_rate": 1.8384442782348542e-05,
"loss": 0.0335,
"step": 1230
},
{
"epoch": 0.2783311354900255,
"grad_norm": 0.3205563724040985,
"learning_rate": 1.853403141361257e-05,
"loss": 0.0337,
"step": 1240
},
{
"epoch": 0.2805757414213967,
"grad_norm": 0.24150335788726807,
"learning_rate": 1.868362004487659e-05,
"loss": 0.0339,
"step": 1250
},
{
"epoch": 0.2828203473527679,
"grad_norm": 0.25747033953666687,
"learning_rate": 1.8833208676140614e-05,
"loss": 0.0323,
"step": 1260
},
{
"epoch": 0.28506495328413906,
"grad_norm": 0.2524860203266144,
"learning_rate": 1.898279730740464e-05,
"loss": 0.0333,
"step": 1270
},
{
"epoch": 0.2873095592155102,
"grad_norm": 0.20733146369457245,
"learning_rate": 1.9132385938668663e-05,
"loss": 0.032,
"step": 1280
},
{
"epoch": 0.2895541651468814,
"grad_norm": 0.24969425797462463,
"learning_rate": 1.9281974569932686e-05,
"loss": 0.0334,
"step": 1290
},
{
"epoch": 0.2917987710782526,
"grad_norm": 0.26867517828941345,
"learning_rate": 1.943156320119671e-05,
"loss": 0.0317,
"step": 1300
},
{
"epoch": 0.29404337700962374,
"grad_norm": 0.23903688788414001,
"learning_rate": 1.9581151832460736e-05,
"loss": 0.0324,
"step": 1310
},
{
"epoch": 0.2962879829409949,
"grad_norm": 0.24634772539138794,
"learning_rate": 1.973074046372476e-05,
"loss": 0.0315,
"step": 1320
},
{
"epoch": 0.2985325888723661,
"grad_norm": 0.24410435557365417,
"learning_rate": 1.988032909498878e-05,
"loss": 0.0307,
"step": 1330
},
{
"epoch": 0.30077719480373727,
"grad_norm": 0.24449145793914795,
"learning_rate": 1.9999998636276648e-05,
"loss": 0.0311,
"step": 1340
},
{
"epoch": 0.3030218007351084,
"grad_norm": 0.2650293707847595,
"learning_rate": 1.999995090599835e-05,
"loss": 0.0327,
"step": 1350
},
{
"epoch": 0.30526640666647964,
"grad_norm": 0.2856460511684418,
"learning_rate": 1.9999834989924354e-05,
"loss": 0.0305,
"step": 1360
},
{
"epoch": 0.3075110125978508,
"grad_norm": 0.23453539609909058,
"learning_rate": 1.9999650888845042e-05,
"loss": 0.0314,
"step": 1370
},
{
"epoch": 0.30975561852922195,
"grad_norm": 0.25706619024276733,
"learning_rate": 1.9999398604015737e-05,
"loss": 0.0305,
"step": 1380
},
{
"epoch": 0.3120002244605931,
"grad_norm": 0.26156002283096313,
"learning_rate": 1.9999078137156663e-05,
"loss": 0.031,
"step": 1390
},
{
"epoch": 0.3142448303919643,
"grad_norm": 0.3038729131221771,
"learning_rate": 1.9998689490452966e-05,
"loss": 0.0312,
"step": 1400
},
{
"epoch": 0.3164894363233355,
"grad_norm": 0.393939346075058,
"learning_rate": 1.9998232666554674e-05,
"loss": 0.0361,
"step": 1410
},
{
"epoch": 0.31873404225470664,
"grad_norm": 0.3251310884952545,
"learning_rate": 1.999770766857669e-05,
"loss": 0.0338,
"step": 1420
},
{
"epoch": 0.32097864818607785,
"grad_norm": 0.2506806552410126,
"learning_rate": 1.999711450009878e-05,
"loss": 0.0311,
"step": 1430
},
{
"epoch": 0.323223254117449,
"grad_norm": 0.274758517742157,
"learning_rate": 1.999645316516552e-05,
"loss": 0.0309,
"step": 1440
},
{
"epoch": 0.32546786004882017,
"grad_norm": 0.2681165337562561,
"learning_rate": 1.9995723668286304e-05,
"loss": 0.0299,
"step": 1450
},
{
"epoch": 0.3277124659801913,
"grad_norm": 0.2143799513578415,
"learning_rate": 1.999492601443529e-05,
"loss": 0.03,
"step": 1460
},
{
"epoch": 0.32995707191156254,
"grad_norm": 0.2572579085826874,
"learning_rate": 1.9994060209051366e-05,
"loss": 0.0303,
"step": 1470
},
{
"epoch": 0.3322016778429337,
"grad_norm": 0.21482343971729279,
"learning_rate": 1.9993126258038126e-05,
"loss": 0.0301,
"step": 1480
},
{
"epoch": 0.33444628377430485,
"grad_norm": 0.2568591237068176,
"learning_rate": 1.9992124167763823e-05,
"loss": 0.0315,
"step": 1490
},
{
"epoch": 0.33669088970567607,
"grad_norm": 0.2510339617729187,
"learning_rate": 1.9991053945061318e-05,
"loss": 0.0299,
"step": 1500
},
{
"epoch": 0.3389354956370472,
"grad_norm": 0.2693713903427124,
"learning_rate": 1.9989915597228048e-05,
"loss": 0.0302,
"step": 1510
},
{
"epoch": 0.3411801015684184,
"grad_norm": 0.2134774774312973,
"learning_rate": 1.9988709132025967e-05,
"loss": 0.0286,
"step": 1520
},
{
"epoch": 0.3434247074997896,
"grad_norm": 0.21737788617610931,
"learning_rate": 1.9987434557681492e-05,
"loss": 0.029,
"step": 1530
},
{
"epoch": 0.34566931343116075,
"grad_norm": 0.21042650938034058,
"learning_rate": 1.998609188288546e-05,
"loss": 0.0298,
"step": 1540
},
{
"epoch": 0.3479139193625319,
"grad_norm": 0.24526509642601013,
"learning_rate": 1.998468111679304e-05,
"loss": 0.0285,
"step": 1550
},
{
"epoch": 0.35015852529390307,
"grad_norm": 0.2452835738658905,
"learning_rate": 1.9983202269023706e-05,
"loss": 0.0285,
"step": 1560
},
{
"epoch": 0.3524031312252743,
"grad_norm": 0.21953906118869781,
"learning_rate": 1.9981655349661146e-05,
"loss": 0.0288,
"step": 1570
},
{
"epoch": 0.35464773715664544,
"grad_norm": 0.2213865965604782,
"learning_rate": 1.9980040369253206e-05,
"loss": 0.0294,
"step": 1580
},
{
"epoch": 0.3568923430880166,
"grad_norm": 0.24777103960514069,
"learning_rate": 1.9978357338811814e-05,
"loss": 0.0302,
"step": 1590
},
{
"epoch": 0.3591369490193878,
"grad_norm": 0.23945562541484833,
"learning_rate": 1.9976606269812904e-05,
"loss": 0.029,
"step": 1600
},
{
"epoch": 0.36138155495075897,
"grad_norm": 0.1871986836194992,
"learning_rate": 1.9974787174196332e-05,
"loss": 0.0279,
"step": 1610
},
{
"epoch": 0.3636261608821301,
"grad_norm": 0.22911369800567627,
"learning_rate": 1.9972900064365817e-05,
"loss": 0.0299,
"step": 1620
},
{
"epoch": 0.3658707668135013,
"grad_norm": 0.21428261697292328,
"learning_rate": 1.9970944953188818e-05,
"loss": 0.0273,
"step": 1630
},
{
"epoch": 0.3681153727448725,
"grad_norm": 0.240596204996109,
"learning_rate": 1.9968921853996492e-05,
"loss": 0.0286,
"step": 1640
},
{
"epoch": 0.37035997867624365,
"grad_norm": 0.2429644763469696,
"learning_rate": 1.9966830780583564e-05,
"loss": 0.0284,
"step": 1650
},
{
"epoch": 0.3726045846076148,
"grad_norm": 0.2194654494524002,
"learning_rate": 1.996467174720826e-05,
"loss": 0.0288,
"step": 1660
},
{
"epoch": 0.374849190538986,
"grad_norm": 0.2488565593957901,
"learning_rate": 1.9962444768592185e-05,
"loss": 0.0279,
"step": 1670
},
{
"epoch": 0.3770937964703572,
"grad_norm": 0.22848640382289886,
"learning_rate": 1.996014985992025e-05,
"loss": 0.0281,
"step": 1680
},
{
"epoch": 0.37933840240172834,
"grad_norm": 0.20066608488559723,
"learning_rate": 1.9957787036840548e-05,
"loss": 0.0289,
"step": 1690
},
{
"epoch": 0.3815830083330995,
"grad_norm": 0.22604599595069885,
"learning_rate": 1.9955356315464257e-05,
"loss": 0.0295,
"step": 1700
},
{
"epoch": 0.3838276142644707,
"grad_norm": 0.24926242232322693,
"learning_rate": 1.9952857712365523e-05,
"loss": 0.0293,
"step": 1710
},
{
"epoch": 0.38607222019584186,
"grad_norm": 0.2565000653266907,
"learning_rate": 1.9950291244581356e-05,
"loss": 0.0292,
"step": 1720
},
{
"epoch": 0.388316826127213,
"grad_norm": 0.18487876653671265,
"learning_rate": 1.9947656929611502e-05,
"loss": 0.0268,
"step": 1730
},
{
"epoch": 0.39056143205858423,
"grad_norm": 0.21695543825626373,
"learning_rate": 1.994495478541834e-05,
"loss": 0.0278,
"step": 1740
},
{
"epoch": 0.3928060379899554,
"grad_norm": 0.18348629772663116,
"learning_rate": 1.9942184830426744e-05,
"loss": 0.0274,
"step": 1750
},
{
"epoch": 0.39505064392132655,
"grad_norm": 0.235065758228302,
"learning_rate": 1.9939347083523967e-05,
"loss": 0.0284,
"step": 1760
},
{
"epoch": 0.39729524985269776,
"grad_norm": 0.2515522539615631,
"learning_rate": 1.9936441564059504e-05,
"loss": 0.0277,
"step": 1770
},
{
"epoch": 0.3995398557840689,
"grad_norm": 0.22060731053352356,
"learning_rate": 1.9933468291844972e-05,
"loss": 0.0263,
"step": 1780
},
{
"epoch": 0.4017844617154401,
"grad_norm": 0.21448932588100433,
"learning_rate": 1.993042728715396e-05,
"loss": 0.0277,
"step": 1790
},
{
"epoch": 0.40402906764681124,
"grad_norm": 0.20610381662845612,
"learning_rate": 1.9927318570721903e-05,
"loss": 0.0284,
"step": 1800
},
{
"epoch": 0.40627367357818245,
"grad_norm": 0.21657615900039673,
"learning_rate": 1.9924142163745935e-05,
"loss": 0.0263,
"step": 1810
},
{
"epoch": 0.4085182795095536,
"grad_norm": 0.20956206321716309,
"learning_rate": 1.992089808788475e-05,
"loss": 0.027,
"step": 1820
},
{
"epoch": 0.41076288544092476,
"grad_norm": 0.20613710582256317,
"learning_rate": 1.991758636525844e-05,
"loss": 0.0262,
"step": 1830
},
{
"epoch": 0.413007491372296,
"grad_norm": 0.18393391370773315,
"learning_rate": 1.9914207018448363e-05,
"loss": 0.0248,
"step": 1840
},
{
"epoch": 0.41525209730366713,
"grad_norm": 0.19099575281143188,
"learning_rate": 1.991076007049698e-05,
"loss": 0.0252,
"step": 1850
},
{
"epoch": 0.4174967032350383,
"grad_norm": 0.20824317634105682,
"learning_rate": 1.990724554490769e-05,
"loss": 0.0265,
"step": 1860
},
{
"epoch": 0.41974130916640945,
"grad_norm": 0.219533771276474,
"learning_rate": 1.9903663465644687e-05,
"loss": 0.0241,
"step": 1870
},
{
"epoch": 0.42198591509778066,
"grad_norm": 0.22310471534729004,
"learning_rate": 1.990001385713278e-05,
"loss": 0.0261,
"step": 1880
},
{
"epoch": 0.4242305210291518,
"grad_norm": 0.18399077653884888,
"learning_rate": 1.9896296744257243e-05,
"loss": 0.0243,
"step": 1890
},
{
"epoch": 0.426475126960523,
"grad_norm": 0.17510205507278442,
"learning_rate": 1.989251215236362e-05,
"loss": 0.0248,
"step": 1900
},
{
"epoch": 0.4287197328918942,
"grad_norm": 0.20756864547729492,
"learning_rate": 1.988866010725759e-05,
"loss": 0.0266,
"step": 1910
},
{
"epoch": 0.43096433882326535,
"grad_norm": 0.20950593054294586,
"learning_rate": 1.9884740635204755e-05,
"loss": 0.0265,
"step": 1920
},
{
"epoch": 0.4332089447546365,
"grad_norm": 0.17630548775196075,
"learning_rate": 1.9880753762930473e-05,
"loss": 0.0251,
"step": 1930
},
{
"epoch": 0.43545355068600766,
"grad_norm": 0.17798230051994324,
"learning_rate": 1.9876699517619686e-05,
"loss": 0.0272,
"step": 1940
},
{
"epoch": 0.4376981566173789,
"grad_norm": 0.17199009656906128,
"learning_rate": 1.9872577926916727e-05,
"loss": 0.0249,
"step": 1950
},
{
"epoch": 0.43994276254875003,
"grad_norm": 0.19935180246829987,
"learning_rate": 1.9868389018925126e-05,
"loss": 0.0262,
"step": 1960
},
{
"epoch": 0.4421873684801212,
"grad_norm": 0.2195315659046173,
"learning_rate": 1.9864132822207424e-05,
"loss": 0.026,
"step": 1970
},
{
"epoch": 0.4444319744114924,
"grad_norm": 0.2122250646352768,
"learning_rate": 1.9859809365784982e-05,
"loss": 0.0263,
"step": 1980
},
{
"epoch": 0.44667658034286356,
"grad_norm": 0.21738848090171814,
"learning_rate": 1.9855418679137775e-05,
"loss": 0.0263,
"step": 1990
},
{
"epoch": 0.4489211862742347,
"grad_norm": 0.21562571823596954,
"learning_rate": 1.9850960792204197e-05,
"loss": 0.0261,
"step": 2000
},
{
"epoch": 0.4511657922056059,
"grad_norm": 0.1967390477657318,
"learning_rate": 1.9846435735380855e-05,
"loss": 0.025,
"step": 2010
},
{
"epoch": 0.4534103981369771,
"grad_norm": 0.1860983669757843,
"learning_rate": 1.9841843539522353e-05,
"loss": 0.0234,
"step": 2020
},
{
"epoch": 0.45565500406834825,
"grad_norm": 0.20370668172836304,
"learning_rate": 1.983718423594111e-05,
"loss": 0.0246,
"step": 2030
},
{
"epoch": 0.4578996099997194,
"grad_norm": 0.20051617920398712,
"learning_rate": 1.9832457856407104e-05,
"loss": 0.0254,
"step": 2040
},
{
"epoch": 0.4601442159310906,
"grad_norm": 0.19836241006851196,
"learning_rate": 1.9827664433147694e-05,
"loss": 0.0259,
"step": 2050
},
{
"epoch": 0.4623888218624618,
"grad_norm": 0.21288608014583588,
"learning_rate": 1.9822803998847378e-05,
"loss": 0.026,
"step": 2060
},
{
"epoch": 0.46463342779383293,
"grad_norm": 0.21260212361812592,
"learning_rate": 1.9817876586647572e-05,
"loss": 0.0254,
"step": 2070
},
{
"epoch": 0.46687803372520414,
"grad_norm": 0.19190289080142975,
"learning_rate": 1.98128822301464e-05,
"loss": 0.0245,
"step": 2080
},
{
"epoch": 0.4691226396565753,
"grad_norm": 0.17896448075771332,
"learning_rate": 1.980782096339844e-05,
"loss": 0.0254,
"step": 2090
},
{
"epoch": 0.47136724558794646,
"grad_norm": 0.20132991671562195,
"learning_rate": 1.9802692820914514e-05,
"loss": 0.0245,
"step": 2100
},
{
"epoch": 0.4736118515193176,
"grad_norm": 0.18621641397476196,
"learning_rate": 1.9797497837661442e-05,
"loss": 0.0244,
"step": 2110
},
{
"epoch": 0.47585645745068883,
"grad_norm": 0.19119122624397278,
"learning_rate": 1.97922360490618e-05,
"loss": 0.0252,
"step": 2120
},
{
"epoch": 0.47810106338206,
"grad_norm": 0.21065153181552887,
"learning_rate": 1.9786907490993698e-05,
"loss": 0.0241,
"step": 2130
},
{
"epoch": 0.48034566931343115,
"grad_norm": 0.20101730525493622,
"learning_rate": 1.97815121997905e-05,
"loss": 0.025,
"step": 2140
},
{
"epoch": 0.48259027524480236,
"grad_norm": 0.19666339457035065,
"learning_rate": 1.9776050212240624e-05,
"loss": 0.0247,
"step": 2150
},
{
"epoch": 0.4848348811761735,
"grad_norm": 0.19246205687522888,
"learning_rate": 1.9770521565587234e-05,
"loss": 0.0243,
"step": 2160
},
{
"epoch": 0.4870794871075447,
"grad_norm": 0.1777629405260086,
"learning_rate": 1.9764926297528036e-05,
"loss": 0.025,
"step": 2170
},
{
"epoch": 0.48932409303891583,
"grad_norm": 0.16932068765163422,
"learning_rate": 1.9759264446215006e-05,
"loss": 0.0232,
"step": 2180
},
{
"epoch": 0.49156869897028704,
"grad_norm": 0.20206604897975922,
"learning_rate": 1.975353605025411e-05,
"loss": 0.0248,
"step": 2190
},
{
"epoch": 0.4938133049016582,
"grad_norm": 0.2303825467824936,
"learning_rate": 1.9747741148705063e-05,
"loss": 0.0247,
"step": 2200
},
{
"epoch": 0.49605791083302936,
"grad_norm": 0.1818767488002777,
"learning_rate": 1.974187978108106e-05,
"loss": 0.0242,
"step": 2210
},
{
"epoch": 0.49830251676440057,
"grad_norm": 0.1894429624080658,
"learning_rate": 1.9735951987348496e-05,
"loss": 0.0244,
"step": 2220
},
{
"epoch": 0.5005471226957717,
"grad_norm": 0.17991645634174347,
"learning_rate": 1.9729957807926705e-05,
"loss": 0.0249,
"step": 2230
},
{
"epoch": 0.5027917286271429,
"grad_norm": 0.214262455701828,
"learning_rate": 1.9723897283687673e-05,
"loss": 0.0231,
"step": 2240
},
{
"epoch": 0.505036334558514,
"grad_norm": 0.19449850916862488,
"learning_rate": 1.9717770455955776e-05,
"loss": 0.0237,
"step": 2250
},
{
"epoch": 0.5072809404898853,
"grad_norm": 0.1926821768283844,
"learning_rate": 1.9711577366507477e-05,
"loss": 0.0248,
"step": 2260
},
{
"epoch": 0.5095255464212565,
"grad_norm": 0.2037525773048401,
"learning_rate": 1.9705318057571056e-05,
"loss": 0.023,
"step": 2270
},
{
"epoch": 0.5117701523526276,
"grad_norm": 0.1556444764137268,
"learning_rate": 1.9698992571826322e-05,
"loss": 0.0234,
"step": 2280
},
{
"epoch": 0.5140147582839988,
"grad_norm": 0.20549353957176208,
"learning_rate": 1.9692600952404312e-05,
"loss": 0.0232,
"step": 2290
},
{
"epoch": 0.5162593642153699,
"grad_norm": 0.1754326969385147,
"learning_rate": 1.9686143242887e-05,
"loss": 0.0237,
"step": 2300
},
{
"epoch": 0.5185039701467411,
"grad_norm": 0.1987963318824768,
"learning_rate": 1.967961948730702e-05,
"loss": 0.0251,
"step": 2310
},
{
"epoch": 0.5207485760781123,
"grad_norm": 0.19438199698925018,
"learning_rate": 1.9673029730147332e-05,
"loss": 0.0239,
"step": 2320
},
{
"epoch": 0.5229931820094834,
"grad_norm": 0.19073012471199036,
"learning_rate": 1.966637401634094e-05,
"loss": 0.0227,
"step": 2330
},
{
"epoch": 0.5252377879408546,
"grad_norm": 0.1820707619190216,
"learning_rate": 1.9659652391270576e-05,
"loss": 0.023,
"step": 2340
},
{
"epoch": 0.5274823938722258,
"grad_norm": 0.17488014698028564,
"learning_rate": 1.965286490076841e-05,
"loss": 0.0241,
"step": 2350
},
{
"epoch": 0.5297269998035969,
"grad_norm": 0.20880946516990662,
"learning_rate": 1.9646011591115706e-05,
"loss": 0.024,
"step": 2360
},
{
"epoch": 0.5319716057349682,
"grad_norm": 0.18289141356945038,
"learning_rate": 1.9639092509042533e-05,
"loss": 0.0241,
"step": 2370
},
{
"epoch": 0.5342162116663394,
"grad_norm": 0.20128047466278076,
"learning_rate": 1.963210770172743e-05,
"loss": 0.0237,
"step": 2380
},
{
"epoch": 0.5364608175977105,
"grad_norm": 0.17308512330055237,
"learning_rate": 1.9625057216797106e-05,
"loss": 0.025,
"step": 2390
},
{
"epoch": 0.5387054235290817,
"grad_norm": 0.19591808319091797,
"learning_rate": 1.961794110232608e-05,
"loss": 0.022,
"step": 2400
},
{
"epoch": 0.5409500294604529,
"grad_norm": 0.16618050634860992,
"learning_rate": 1.9610759406836382e-05,
"loss": 0.0227,
"step": 2410
},
{
"epoch": 0.543194635391824,
"grad_norm": 0.17903420329093933,
"learning_rate": 1.9603512179297226e-05,
"loss": 0.0236,
"step": 2420
},
{
"epoch": 0.5454392413231952,
"grad_norm": 0.20347997546195984,
"learning_rate": 1.959619946912464e-05,
"loss": 0.0231,
"step": 2430
},
{
"epoch": 0.5476838472545664,
"grad_norm": 0.22455480694770813,
"learning_rate": 1.9588821326181172e-05,
"loss": 0.0248,
"step": 2440
},
{
"epoch": 0.5499284531859375,
"grad_norm": 0.16458284854888916,
"learning_rate": 1.9581377800775518e-05,
"loss": 0.0224,
"step": 2450
},
{
"epoch": 0.5521730591173087,
"grad_norm": 0.1932162195444107,
"learning_rate": 1.95738689436622e-05,
"loss": 0.0235,
"step": 2460
},
{
"epoch": 0.5544176650486798,
"grad_norm": 0.1769714504480362,
"learning_rate": 1.9566294806041214e-05,
"loss": 0.0215,
"step": 2470
},
{
"epoch": 0.556662270980051,
"grad_norm": 0.20865550637245178,
"learning_rate": 1.9558655439557665e-05,
"loss": 0.0223,
"step": 2480
},
{
"epoch": 0.5589068769114223,
"grad_norm": 0.2173549085855484,
"learning_rate": 1.9550950896301445e-05,
"loss": 0.0234,
"step": 2490
},
{
"epoch": 0.5611514828427934,
"grad_norm": 0.18162918090820312,
"learning_rate": 1.954318122880685e-05,
"loss": 0.0243,
"step": 2500
},
{
"epoch": 0.5633960887741646,
"grad_norm": 0.17128832638263702,
"learning_rate": 1.9535346490052235e-05,
"loss": 0.0239,
"step": 2510
},
{
"epoch": 0.5656406947055358,
"grad_norm": 0.18167993426322937,
"learning_rate": 1.952744673345965e-05,
"loss": 0.0241,
"step": 2520
},
{
"epoch": 0.5678853006369069,
"grad_norm": 0.20374241471290588,
"learning_rate": 1.9519482012894483e-05,
"loss": 0.0235,
"step": 2530
},
{
"epoch": 0.5701299065682781,
"grad_norm": 0.17413835227489471,
"learning_rate": 1.9511452382665077e-05,
"loss": 0.0226,
"step": 2540
},
{
"epoch": 0.5723745124996493,
"grad_norm": 0.20678037405014038,
"learning_rate": 1.9503357897522378e-05,
"loss": 0.0238,
"step": 2550
},
{
"epoch": 0.5746191184310204,
"grad_norm": 0.17220425605773926,
"learning_rate": 1.949519861265954e-05,
"loss": 0.0226,
"step": 2560
},
{
"epoch": 0.5768637243623916,
"grad_norm": 0.19188691675662994,
"learning_rate": 1.9486974583711584e-05,
"loss": 0.0223,
"step": 2570
},
{
"epoch": 0.5791083302937629,
"grad_norm": 0.18967117369174957,
"learning_rate": 1.947868586675497e-05,
"loss": 0.0234,
"step": 2580
},
{
"epoch": 0.581352936225134,
"grad_norm": 0.16889534890651703,
"learning_rate": 1.947033251830725e-05,
"loss": 0.0223,
"step": 2590
},
{
"epoch": 0.5835975421565052,
"grad_norm": 0.23073066771030426,
"learning_rate": 1.9461914595326684e-05,
"loss": 0.0228,
"step": 2600
},
{
"epoch": 0.5858421480878763,
"grad_norm": 0.19199195504188538,
"learning_rate": 1.945343215521182e-05,
"loss": 0.0217,
"step": 2610
},
{
"epoch": 0.5880867540192475,
"grad_norm": 0.1747967153787613,
"learning_rate": 1.944488525580114e-05,
"loss": 0.0218,
"step": 2620
},
{
"epoch": 0.5903313599506187,
"grad_norm": 0.16960597038269043,
"learning_rate": 1.943627395537265e-05,
"loss": 0.0232,
"step": 2630
},
{
"epoch": 0.5925759658819898,
"grad_norm": 1.8774281740188599,
"learning_rate": 1.9427598312643467e-05,
"loss": 0.0268,
"step": 2640
},
{
"epoch": 0.594820571813361,
"grad_norm": 0.26028814911842346,
"learning_rate": 1.9418858386769447e-05,
"loss": 0.0327,
"step": 2650
},
{
"epoch": 0.5970651777447322,
"grad_norm": 0.22514741122722626,
"learning_rate": 1.9410054237344757e-05,
"loss": 0.0288,
"step": 2660
},
{
"epoch": 0.5993097836761033,
"grad_norm": 0.21931113302707672,
"learning_rate": 1.9401185924401488e-05,
"loss": 0.0251,
"step": 2670
},
{
"epoch": 0.6015543896074745,
"grad_norm": 0.1910109966993332,
"learning_rate": 1.9392253508409235e-05,
"loss": 0.0242,
"step": 2680
},
{
"epoch": 0.6037989955388458,
"grad_norm": 0.17131438851356506,
"learning_rate": 1.938325705027469e-05,
"loss": 0.0234,
"step": 2690
},
{
"epoch": 0.6060436014702169,
"grad_norm": 0.1875181645154953,
"learning_rate": 1.9374196611341212e-05,
"loss": 0.0241,
"step": 2700
},
{
"epoch": 0.6082882074015881,
"grad_norm": 0.18994557857513428,
"learning_rate": 1.936507225338843e-05,
"loss": 0.0226,
"step": 2710
},
{
"epoch": 0.6105328133329593,
"grad_norm": 0.19114020466804504,
"learning_rate": 1.9355884038631812e-05,
"loss": 0.023,
"step": 2720
},
{
"epoch": 0.6127774192643304,
"grad_norm": 0.18819326162338257,
"learning_rate": 1.9346632029722243e-05,
"loss": 0.023,
"step": 2730
},
{
"epoch": 0.6150220251957016,
"grad_norm": 0.2011801302433014,
"learning_rate": 1.9337316289745586e-05,
"loss": 0.0233,
"step": 2740
},
{
"epoch": 0.6172666311270728,
"grad_norm": 0.176308736205101,
"learning_rate": 1.9327936882222267e-05,
"loss": 0.0227,
"step": 2750
},
{
"epoch": 0.6195112370584439,
"grad_norm": 0.1711064577102661,
"learning_rate": 1.931849387110684e-05,
"loss": 0.0238,
"step": 2760
},
{
"epoch": 0.6217558429898151,
"grad_norm": 0.1591506451368332,
"learning_rate": 1.930898732078754e-05,
"loss": 0.0218,
"step": 2770
},
{
"epoch": 0.6240004489211862,
"grad_norm": 0.17407435178756714,
"learning_rate": 1.929941729608586e-05,
"loss": 0.0235,
"step": 2780
},
{
"epoch": 0.6262450548525574,
"grad_norm": 0.16112394630908966,
"learning_rate": 1.9289783862256087e-05,
"loss": 0.0213,
"step": 2790
},
{
"epoch": 0.6284896607839286,
"grad_norm": 0.15335716307163239,
"learning_rate": 1.9280087084984892e-05,
"loss": 0.0228,
"step": 2800
},
{
"epoch": 0.6307342667152998,
"grad_norm": 0.18014346063137054,
"learning_rate": 1.9270327030390832e-05,
"loss": 0.0233,
"step": 2810
},
{
"epoch": 0.632978872646671,
"grad_norm": 0.16869892179965973,
"learning_rate": 1.9260503765023954e-05,
"loss": 0.022,
"step": 2820
},
{
"epoch": 0.6352234785780422,
"grad_norm": 0.1692594587802887,
"learning_rate": 1.9250617355865292e-05,
"loss": 0.0226,
"step": 2830
},
{
"epoch": 0.6374680845094133,
"grad_norm": 0.19598528742790222,
"learning_rate": 1.924066787032646e-05,
"loss": 0.0214,
"step": 2840
},
{
"epoch": 0.6397126904407845,
"grad_norm": 0.17125625908374786,
"learning_rate": 1.9230655376249134e-05,
"loss": 0.022,
"step": 2850
},
{
"epoch": 0.6419572963721557,
"grad_norm": 0.1635292023420334,
"learning_rate": 1.9220579941904645e-05,
"loss": 0.0217,
"step": 2860
},
{
"epoch": 0.6442019023035268,
"grad_norm": 0.18467473983764648,
"learning_rate": 1.9210441635993483e-05,
"loss": 0.0226,
"step": 2870
},
{
"epoch": 0.646446508234898,
"grad_norm": 0.17099305987358093,
"learning_rate": 1.9200240527644828e-05,
"loss": 0.021,
"step": 2880
},
{
"epoch": 0.6486911141662692,
"grad_norm": 0.15335629880428314,
"learning_rate": 1.918997668641609e-05,
"loss": 0.0207,
"step": 2890
},
{
"epoch": 0.6509357200976403,
"grad_norm": 0.15978077054023743,
"learning_rate": 1.9179650182292436e-05,
"loss": 0.0214,
"step": 2900
},
{
"epoch": 0.6531803260290115,
"grad_norm": 0.1606997549533844,
"learning_rate": 1.91692610856863e-05,
"loss": 0.021,
"step": 2910
},
{
"epoch": 0.6554249319603827,
"grad_norm": 0.16981342434883118,
"learning_rate": 1.915880946743691e-05,
"loss": 0.0218,
"step": 2920
},
{
"epoch": 0.6576695378917539,
"grad_norm": 0.17529399693012238,
"learning_rate": 1.91482953988098e-05,
"loss": 0.0217,
"step": 2930
},
{
"epoch": 0.6599141438231251,
"grad_norm": 0.1750255823135376,
"learning_rate": 1.9137718951496352e-05,
"loss": 0.0221,
"step": 2940
},
{
"epoch": 0.6621587497544962,
"grad_norm": 0.17034149169921875,
"learning_rate": 1.912708019761325e-05,
"loss": 0.021,
"step": 2950
},
{
"epoch": 0.6644033556858674,
"grad_norm": 0.1882370561361313,
"learning_rate": 1.9116379209702056e-05,
"loss": 0.0208,
"step": 2960
},
{
"epoch": 0.6666479616172386,
"grad_norm": 0.17587997019290924,
"learning_rate": 1.910561606072865e-05,
"loss": 0.0208,
"step": 2970
},
{
"epoch": 0.6688925675486097,
"grad_norm": 0.18303704261779785,
"learning_rate": 1.909479082408279e-05,
"loss": 0.0196,
"step": 2980
},
{
"epoch": 0.6711371734799809,
"grad_norm": 0.1817975491285324,
"learning_rate": 1.908390357357758e-05,
"loss": 0.0209,
"step": 2990
},
{
"epoch": 0.6733817794113521,
"grad_norm": 0.194996640086174,
"learning_rate": 1.9072954383448955e-05,
"loss": 0.0215,
"step": 3000
},
{
"epoch": 0.6756263853427232,
"grad_norm": 0.20948557555675507,
"learning_rate": 1.9061943328355214e-05,
"loss": 0.02,
"step": 3010
},
{
"epoch": 0.6778709912740944,
"grad_norm": 0.17736251652240753,
"learning_rate": 1.905087048337648e-05,
"loss": 0.0202,
"step": 3020
},
{
"epoch": 0.6801155972054657,
"grad_norm": 0.14099887013435364,
"learning_rate": 1.9039735924014197e-05,
"loss": 0.0203,
"step": 3030
},
{
"epoch": 0.6823602031368368,
"grad_norm": 0.16595132648944855,
"learning_rate": 1.902853972619062e-05,
"loss": 0.0213,
"step": 3040
},
{
"epoch": 0.684604809068208,
"grad_norm": 0.14856795966625214,
"learning_rate": 1.9017281966248284e-05,
"loss": 0.02,
"step": 3050
},
{
"epoch": 0.6868494149995792,
"grad_norm": 0.16550858318805695,
"learning_rate": 1.90059627209495e-05,
"loss": 0.0214,
"step": 3060
},
{
"epoch": 0.6890940209309503,
"grad_norm": 0.20143896341323853,
"learning_rate": 1.8994582067475827e-05,
"loss": 0.0208,
"step": 3070
},
{
"epoch": 0.6913386268623215,
"grad_norm": 0.16624027490615845,
"learning_rate": 1.8983140083427528e-05,
"loss": 0.0202,
"step": 3080
},
{
"epoch": 0.6935832327936926,
"grad_norm": 0.17451374232769012,
"learning_rate": 1.8971636846823074e-05,
"loss": 0.0215,
"step": 3090
},
{
"epoch": 0.6958278387250638,
"grad_norm": 0.18585652112960815,
"learning_rate": 1.8960072436098577e-05,
"loss": 0.0194,
"step": 3100
},
{
"epoch": 0.698072444656435,
"grad_norm": 0.1638859361410141,
"learning_rate": 1.894844693010728e-05,
"loss": 0.0203,
"step": 3110
},
{
"epoch": 0.7003170505878061,
"grad_norm": 0.18684209883213043,
"learning_rate": 1.8936760408119012e-05,
"loss": 0.0223,
"step": 3120
},
{
"epoch": 0.7025616565191773,
"grad_norm": 0.1838609278202057,
"learning_rate": 1.8925012949819645e-05,
"loss": 0.0216,
"step": 3130
},
{
"epoch": 0.7048062624505486,
"grad_norm": 0.17233972251415253,
"learning_rate": 1.8913204635310548e-05,
"loss": 0.0201,
"step": 3140
},
{
"epoch": 0.7070508683819197,
"grad_norm": 0.18251539766788483,
"learning_rate": 1.8901335545108054e-05,
"loss": 0.0207,
"step": 3150
},
{
"epoch": 0.7092954743132909,
"grad_norm": 0.1725347936153412,
"learning_rate": 1.8889405760142887e-05,
"loss": 0.0209,
"step": 3160
},
{
"epoch": 0.7115400802446621,
"grad_norm": 0.17256340384483337,
"learning_rate": 1.8877415361759645e-05,
"loss": 0.0209,
"step": 3170
},
{
"epoch": 0.7137846861760332,
"grad_norm": 0.1629970371723175,
"learning_rate": 1.8865364431716205e-05,
"loss": 0.0208,
"step": 3180
},
{
"epoch": 0.7160292921074044,
"grad_norm": 0.20335057377815247,
"learning_rate": 1.8853253052183206e-05,
"loss": 0.0206,
"step": 3190
},
{
"epoch": 0.7182738980387756,
"grad_norm": 0.15751416981220245,
"learning_rate": 1.8841081305743447e-05,
"loss": 0.0202,
"step": 3200
},
{
"epoch": 0.7205185039701467,
"grad_norm": 0.15457507967948914,
"learning_rate": 1.8828849275391366e-05,
"loss": 0.0214,
"step": 3210
},
{
"epoch": 0.7227631099015179,
"grad_norm": 0.1803683042526245,
"learning_rate": 1.8816557044532443e-05,
"loss": 0.0196,
"step": 3220
},
{
"epoch": 0.7250077158328891,
"grad_norm": 0.18161389231681824,
"learning_rate": 1.880420469698264e-05,
"loss": 0.0212,
"step": 3230
},
{
"epoch": 0.7272523217642602,
"grad_norm": 0.17406050860881805,
"learning_rate": 1.879179231696784e-05,
"loss": 0.02,
"step": 3240
},
{
"epoch": 0.7294969276956315,
"grad_norm": 0.15951761603355408,
"learning_rate": 1.8779319989123253e-05,
"loss": 0.0202,
"step": 3250
},
{
"epoch": 0.7317415336270026,
"grad_norm": 0.17296704649925232,
"learning_rate": 1.8766787798492863e-05,
"loss": 0.0195,
"step": 3260
},
{
"epoch": 0.7339861395583738,
"grad_norm": 0.16530241072177887,
"learning_rate": 1.8754195830528826e-05,
"loss": 0.0208,
"step": 3270
},
{
"epoch": 0.736230745489745,
"grad_norm": 0.16106492280960083,
"learning_rate": 1.8741544171090898e-05,
"loss": 0.0209,
"step": 3280
},
{
"epoch": 0.7384753514211161,
"grad_norm": 0.17415162920951843,
"learning_rate": 1.872883290644585e-05,
"loss": 0.021,
"step": 3290
},
{
"epoch": 0.7407199573524873,
"grad_norm": 0.16503530740737915,
"learning_rate": 1.8716062123266877e-05,
"loss": 0.0199,
"step": 3300
},
{
"epoch": 0.7429645632838585,
"grad_norm": 0.17592734098434448,
"learning_rate": 1.8703231908633007e-05,
"loss": 0.0199,
"step": 3310
},
{
"epoch": 0.7452091692152296,
"grad_norm": 0.16938206553459167,
"learning_rate": 1.8690342350028513e-05,
"loss": 0.0195,
"step": 3320
},
{
"epoch": 0.7474537751466008,
"grad_norm": 0.16405285894870758,
"learning_rate": 1.8677393535342298e-05,
"loss": 0.021,
"step": 3330
},
{
"epoch": 0.749698381077972,
"grad_norm": 0.16769342124462128,
"learning_rate": 1.8664385552867332e-05,
"loss": 0.0199,
"step": 3340
},
{
"epoch": 0.7519429870093431,
"grad_norm": 0.17460818588733673,
"learning_rate": 1.8651318491300002e-05,
"loss": 0.0206,
"step": 3350
},
{
"epoch": 0.7541875929407144,
"grad_norm": 0.1495262235403061,
"learning_rate": 1.863819243973955e-05,
"loss": 0.0194,
"step": 3360
},
{
"epoch": 0.7564321988720856,
"grad_norm": 0.15108689665794373,
"learning_rate": 1.862500748768744e-05,
"loss": 0.0198,
"step": 3370
},
{
"epoch": 0.7586768048034567,
"grad_norm": 0.15738345682621002,
"learning_rate": 1.8611763725046758e-05,
"loss": 0.0191,
"step": 3380
},
{
"epoch": 0.7609214107348279,
"grad_norm": 0.14987443387508392,
"learning_rate": 1.85984612421216e-05,
"loss": 0.0186,
"step": 3390
},
{
"epoch": 0.763166016666199,
"grad_norm": 0.1464933604001999,
"learning_rate": 1.858510012961645e-05,
"loss": 0.0209,
"step": 3400
},
{
"epoch": 0.7654106225975702,
"grad_norm": 0.14822939038276672,
"learning_rate": 1.857168047863556e-05,
"loss": 0.02,
"step": 3410
},
{
"epoch": 0.7676552285289414,
"grad_norm": 0.1794120967388153,
"learning_rate": 1.8558202380682343e-05,
"loss": 0.0211,
"step": 3420
},
{
"epoch": 0.7698998344603125,
"grad_norm": 0.15938477218151093,
"learning_rate": 1.8544665927658723e-05,
"loss": 0.0197,
"step": 3430
},
{
"epoch": 0.7721444403916837,
"grad_norm": 0.1710355430841446,
"learning_rate": 1.853107121186455e-05,
"loss": 0.0196,
"step": 3440
},
{
"epoch": 0.7743890463230549,
"grad_norm": 0.15917006134986877,
"learning_rate": 1.8517418325996916e-05,
"loss": 0.0197,
"step": 3450
},
{
"epoch": 0.776633652254426,
"grad_norm": 0.15458229184150696,
"learning_rate": 1.8503707363149573e-05,
"loss": 0.0193,
"step": 3460
},
{
"epoch": 0.7788782581857973,
"grad_norm": 0.1885480433702469,
"learning_rate": 1.8489938416812273e-05,
"loss": 0.0194,
"step": 3470
},
{
"epoch": 0.7811228641171685,
"grad_norm": 0.1443496197462082,
"learning_rate": 1.847611158087013e-05,
"loss": 0.0196,
"step": 3480
},
{
"epoch": 0.7833674700485396,
"grad_norm": 0.1544029414653778,
"learning_rate": 1.846222694960299e-05,
"loss": 0.0198,
"step": 3490
},
{
"epoch": 0.7856120759799108,
"grad_norm": 0.17099037766456604,
"learning_rate": 1.8448284617684784e-05,
"loss": 0.0193,
"step": 3500
},
{
"epoch": 0.787856681911282,
"grad_norm": 0.1704227477312088,
"learning_rate": 1.843428468018287e-05,
"loss": 0.019,
"step": 3510
},
{
"epoch": 0.7901012878426531,
"grad_norm": 0.1799788922071457,
"learning_rate": 1.842022723255741e-05,
"loss": 0.0199,
"step": 3520
},
{
"epoch": 0.7923458937740243,
"grad_norm": 0.16611269116401672,
"learning_rate": 1.8406112370660695e-05,
"loss": 0.0206,
"step": 3530
},
{
"epoch": 0.7945904997053955,
"grad_norm": 0.16963708400726318,
"learning_rate": 1.8391940190736514e-05,
"loss": 0.0186,
"step": 3540
},
{
"epoch": 0.7968351056367666,
"grad_norm": 0.16803216934204102,
"learning_rate": 1.8377710789419473e-05,
"loss": 0.0185,
"step": 3550
},
{
"epoch": 0.7990797115681378,
"grad_norm": 0.15741732716560364,
"learning_rate": 1.8363424263734354e-05,
"loss": 0.0201,
"step": 3560
},
{
"epoch": 0.8013243174995089,
"grad_norm": 0.1796884834766388,
"learning_rate": 1.834908071109545e-05,
"loss": 0.019,
"step": 3570
},
{
"epoch": 0.8035689234308802,
"grad_norm": 0.15816806256771088,
"learning_rate": 1.8334680229305894e-05,
"loss": 0.0193,
"step": 3580
},
{
"epoch": 0.8058135293622514,
"grad_norm": 0.15980146825313568,
"learning_rate": 1.8320222916556996e-05,
"loss": 0.0198,
"step": 3590
},
{
"epoch": 0.8080581352936225,
"grad_norm": 0.14864562451839447,
"learning_rate": 1.8305708871427583e-05,
"loss": 0.0188,
"step": 3600
},
{
"epoch": 0.8103027412249937,
"grad_norm": 0.145126610994339,
"learning_rate": 1.829113819288331e-05,
"loss": 0.0187,
"step": 3610
},
{
"epoch": 0.8125473471563649,
"grad_norm": 0.14894086122512817,
"learning_rate": 1.827651098027599e-05,
"loss": 0.0198,
"step": 3620
},
{
"epoch": 0.814791953087736,
"grad_norm": 0.1573464423418045,
"learning_rate": 1.8261827333342938e-05,
"loss": 0.0192,
"step": 3630
},
{
"epoch": 0.8170365590191072,
"grad_norm": 0.16083981096744537,
"learning_rate": 1.824708735220625e-05,
"loss": 0.0194,
"step": 3640
},
{
"epoch": 0.8192811649504784,
"grad_norm": 0.15828320384025574,
"learning_rate": 1.823229113737216e-05,
"loss": 0.0189,
"step": 3650
},
{
"epoch": 0.8215257708818495,
"grad_norm": 0.13140316307544708,
"learning_rate": 1.8217438789730325e-05,
"loss": 0.0194,
"step": 3660
},
{
"epoch": 0.8237703768132207,
"grad_norm": 0.16682444512844086,
"learning_rate": 1.8202530410553162e-05,
"loss": 0.0199,
"step": 3670
},
{
"epoch": 0.826014982744592,
"grad_norm": 0.15172256529331207,
"learning_rate": 1.818756610149514e-05,
"loss": 0.0189,
"step": 3680
},
{
"epoch": 0.828259588675963,
"grad_norm": 0.1695002317428589,
"learning_rate": 1.8172545964592087e-05,
"loss": 0.0197,
"step": 3690
},
{
"epoch": 0.8305041946073343,
"grad_norm": 0.16010014712810516,
"learning_rate": 1.815747010226051e-05,
"loss": 0.0182,
"step": 3700
},
{
"epoch": 0.8327488005387054,
"grad_norm": 0.15437164902687073,
"learning_rate": 1.8142338617296878e-05,
"loss": 0.0198,
"step": 3710
},
{
"epoch": 0.8349934064700766,
"grad_norm": 0.17187710106372833,
"learning_rate": 1.812715161287693e-05,
"loss": 0.0195,
"step": 3720
},
{
"epoch": 0.8372380124014478,
"grad_norm": 0.16505780816078186,
"learning_rate": 1.8111909192554976e-05,
"loss": 0.0177,
"step": 3730
},
{
"epoch": 0.8394826183328189,
"grad_norm": 0.161300927400589,
"learning_rate": 1.8096611460263178e-05,
"loss": 0.0189,
"step": 3740
},
{
"epoch": 0.8417272242641901,
"grad_norm": 0.16342850029468536,
"learning_rate": 1.8081258520310856e-05,
"loss": 0.0183,
"step": 3750
},
{
"epoch": 0.8439718301955613,
"grad_norm": 0.15704095363616943,
"learning_rate": 1.8065850477383767e-05,
"loss": 0.0183,
"step": 3760
},
{
"epoch": 0.8462164361269324,
"grad_norm": 0.1607922911643982,
"learning_rate": 1.805038743654339e-05,
"loss": 0.019,
"step": 3770
},
{
"epoch": 0.8484610420583036,
"grad_norm": 0.16337986290454865,
"learning_rate": 1.803486950322622e-05,
"loss": 0.0193,
"step": 3780
},
{
"epoch": 0.8507056479896749,
"grad_norm": 0.19210092723369598,
"learning_rate": 1.801929678324304e-05,
"loss": 0.0179,
"step": 3790
},
{
"epoch": 0.852950253921046,
"grad_norm": 0.1809729039669037,
"learning_rate": 1.80036693827782e-05,
"loss": 0.0193,
"step": 3800
},
{
"epoch": 0.8551948598524172,
"grad_norm": 0.14291028678417206,
"learning_rate": 1.79879874083889e-05,
"loss": 0.0189,
"step": 3810
},
{
"epoch": 0.8574394657837884,
"grad_norm": 0.2091740220785141,
"learning_rate": 1.7972250967004448e-05,
"loss": 0.0182,
"step": 3820
},
{
"epoch": 0.8596840717151595,
"grad_norm": 0.15339423716068268,
"learning_rate": 1.7956460165925556e-05,
"loss": 0.0207,
"step": 3830
},
{
"epoch": 0.8619286776465307,
"grad_norm": 0.1679808646440506,
"learning_rate": 1.794061511282358e-05,
"loss": 0.0178,
"step": 3840
},
{
"epoch": 0.8641732835779019,
"grad_norm": 0.1681501418352127,
"learning_rate": 1.792471591573981e-05,
"loss": 0.0182,
"step": 3850
},
{
"epoch": 0.866417889509273,
"grad_norm": 0.14759108424186707,
"learning_rate": 1.790876268308472e-05,
"loss": 0.0184,
"step": 3860
},
{
"epoch": 0.8686624954406442,
"grad_norm": 0.15728703141212463,
"learning_rate": 1.7892755523637224e-05,
"loss": 0.0178,
"step": 3870
},
{
"epoch": 0.8709071013720153,
"grad_norm": 0.14030325412750244,
"learning_rate": 1.7876694546543955e-05,
"loss": 0.0189,
"step": 3880
},
{
"epoch": 0.8731517073033865,
"grad_norm": 0.13951782882213593,
"learning_rate": 1.7860579861318496e-05,
"loss": 0.0191,
"step": 3890
},
{
"epoch": 0.8753963132347578,
"grad_norm": 0.16399997472763062,
"learning_rate": 1.7844411577840654e-05,
"loss": 0.0188,
"step": 3900
},
{
"epoch": 0.8776409191661289,
"grad_norm": 0.14510847628116608,
"learning_rate": 1.7828189806355694e-05,
"loss": 0.0176,
"step": 3910
},
{
"epoch": 0.8798855250975001,
"grad_norm": 0.1528862714767456,
"learning_rate": 1.78119146574736e-05,
"loss": 0.0198,
"step": 3920
},
{
"epoch": 0.8821301310288713,
"grad_norm": 0.2043169140815735,
"learning_rate": 1.7795586242168313e-05,
"loss": 0.0195,
"step": 3930
},
{
"epoch": 0.8843747369602424,
"grad_norm": 0.15285201370716095,
"learning_rate": 1.7779204671776984e-05,
"loss": 0.0184,
"step": 3940
},
{
"epoch": 0.8866193428916136,
"grad_norm": 0.1563611477613449,
"learning_rate": 1.7762770057999195e-05,
"loss": 0.0187,
"step": 3950
},
{
"epoch": 0.8888639488229848,
"grad_norm": 0.15826715528964996,
"learning_rate": 1.7746282512896224e-05,
"loss": 0.0179,
"step": 3960
},
{
"epoch": 0.8911085547543559,
"grad_norm": 0.14216351509094238,
"learning_rate": 1.7729742148890258e-05,
"loss": 0.0185,
"step": 3970
},
{
"epoch": 0.8933531606857271,
"grad_norm": 0.17049328982830048,
"learning_rate": 1.7713149078763644e-05,
"loss": 0.0183,
"step": 3980
},
{
"epoch": 0.8955977666170983,
"grad_norm": 0.14374832808971405,
"learning_rate": 1.7696503415658096e-05,
"loss": 0.0192,
"step": 3990
},
{
"epoch": 0.8978423725484694,
"grad_norm": 0.18668319284915924,
"learning_rate": 1.767980527307396e-05,
"loss": 0.0198,
"step": 4000
},
{
"epoch": 0.9000869784798406,
"grad_norm": 0.17893734574317932,
"learning_rate": 1.76630547648694e-05,
"loss": 0.0198,
"step": 4010
},
{
"epoch": 0.9023315844112118,
"grad_norm": 0.1579989641904831,
"learning_rate": 1.7646252005259657e-05,
"loss": 0.0182,
"step": 4020
},
{
"epoch": 0.904576190342583,
"grad_norm": 0.17703531682491302,
"learning_rate": 1.7629397108816242e-05,
"loss": 0.0177,
"step": 4030
},
{
"epoch": 0.9068207962739542,
"grad_norm": 0.14626428484916687,
"learning_rate": 1.7612490190466177e-05,
"loss": 0.0195,
"step": 4040
},
{
"epoch": 0.9090654022053253,
"grad_norm": 0.15867730975151062,
"learning_rate": 1.7595531365491192e-05,
"loss": 0.0174,
"step": 4050
},
{
"epoch": 0.9113100081366965,
"grad_norm": 0.14312785863876343,
"learning_rate": 1.7578520749526952e-05,
"loss": 0.0178,
"step": 4060
},
{
"epoch": 0.9135546140680677,
"grad_norm": 0.15664878487586975,
"learning_rate": 1.7561458458562265e-05,
"loss": 0.0186,
"step": 4070
},
{
"epoch": 0.9157992199994388,
"grad_norm": 0.1552259773015976,
"learning_rate": 1.7544344608938296e-05,
"loss": 0.0184,
"step": 4080
},
{
"epoch": 0.91804382593081,
"grad_norm": 0.1600690335035324,
"learning_rate": 1.7527179317347754e-05,
"loss": 0.0187,
"step": 4090
},
{
"epoch": 0.9202884318621812,
"grad_norm": 0.17803123593330383,
"learning_rate": 1.7509962700834128e-05,
"loss": 0.0175,
"step": 4100
},
{
"epoch": 0.9225330377935523,
"grad_norm": 0.17079535126686096,
"learning_rate": 1.749269487679086e-05,
"loss": 0.0182,
"step": 4110
},
{
"epoch": 0.9247776437249235,
"grad_norm": 0.14618434011936188,
"learning_rate": 1.7475375962960564e-05,
"loss": 0.018,
"step": 4120
},
{
"epoch": 0.9270222496562948,
"grad_norm": 0.140808567404747,
"learning_rate": 1.7458006077434208e-05,
"loss": 0.0183,
"step": 4130
},
{
"epoch": 0.9292668555876659,
"grad_norm": 0.15096674859523773,
"learning_rate": 1.744058533865032e-05,
"loss": 0.0177,
"step": 4140
},
{
"epoch": 0.9315114615190371,
"grad_norm": 0.16479936242103577,
"learning_rate": 1.742311386539418e-05,
"loss": 0.0182,
"step": 4150
},
{
"epoch": 0.9337560674504083,
"grad_norm": 0.14364103972911835,
"learning_rate": 1.7405591776797e-05,
"loss": 0.0171,
"step": 4160
},
{
"epoch": 0.9360006733817794,
"grad_norm": 0.16919943690299988,
"learning_rate": 1.7388019192335123e-05,
"loss": 0.0184,
"step": 4170
},
{
"epoch": 0.9382452793131506,
"grad_norm": 0.1467256247997284,
"learning_rate": 1.7370396231829202e-05,
"loss": 0.0173,
"step": 4180
},
{
"epoch": 0.9404898852445217,
"grad_norm": 0.1647726446390152,
"learning_rate": 1.735272301544339e-05,
"loss": 0.0178,
"step": 4190
},
{
"epoch": 0.9427344911758929,
"grad_norm": 0.16072970628738403,
"learning_rate": 1.7334999663684504e-05,
"loss": 0.0173,
"step": 4200
},
{
"epoch": 0.9449790971072641,
"grad_norm": 0.15171808004379272,
"learning_rate": 1.731722629740123e-05,
"loss": 0.0167,
"step": 4210
},
{
"epoch": 0.9472237030386352,
"grad_norm": 0.13393864035606384,
"learning_rate": 1.7299403037783264e-05,
"loss": 0.0178,
"step": 4220
},
{
"epoch": 0.9494683089700064,
"grad_norm": 0.18740731477737427,
"learning_rate": 1.7281530006360525e-05,
"loss": 0.0184,
"step": 4230
},
{
"epoch": 0.9517129149013777,
"grad_norm": 0.13795392215251923,
"learning_rate": 1.7263607325002298e-05,
"loss": 0.018,
"step": 4240
},
{
"epoch": 0.9539575208327488,
"grad_norm": 0.16634628176689148,
"learning_rate": 1.7245635115916403e-05,
"loss": 0.0186,
"step": 4250
},
{
"epoch": 0.95620212676412,
"grad_norm": 0.15677855908870697,
"learning_rate": 1.7227613501648388e-05,
"loss": 0.0179,
"step": 4260
},
{
"epoch": 0.9584467326954912,
"grad_norm": 0.13720932602882385,
"learning_rate": 1.720954260508067e-05,
"loss": 0.0185,
"step": 4270
},
{
"epoch": 0.9606913386268623,
"grad_norm": 0.15041740238666534,
"learning_rate": 1.7191422549431692e-05,
"loss": 0.0176,
"step": 4280
},
{
"epoch": 0.9629359445582335,
"grad_norm": 0.1521487534046173,
"learning_rate": 1.7173253458255103e-05,
"loss": 0.0179,
"step": 4290
},
{
"epoch": 0.9651805504896047,
"grad_norm": 0.15372952818870544,
"learning_rate": 1.715503545543891e-05,
"loss": 0.0177,
"step": 4300
},
{
"epoch": 0.9674251564209758,
"grad_norm": 0.16285191476345062,
"learning_rate": 1.713676866520462e-05,
"loss": 0.0179,
"step": 4310
},
{
"epoch": 0.969669762352347,
"grad_norm": 0.13474644720554352,
"learning_rate": 1.7118453212106408e-05,
"loss": 0.0175,
"step": 4320
},
{
"epoch": 0.9719143682837182,
"grad_norm": 0.15485605597496033,
"learning_rate": 1.710008922103027e-05,
"loss": 0.0172,
"step": 4330
},
{
"epoch": 0.9741589742150893,
"grad_norm": 0.18369507789611816,
"learning_rate": 1.7081676817193136e-05,
"loss": 0.0177,
"step": 4340
},
{
"epoch": 0.9764035801464606,
"grad_norm": 0.15704116225242615,
"learning_rate": 1.7063216126142078e-05,
"loss": 0.0181,
"step": 4350
},
{
"epoch": 0.9786481860778317,
"grad_norm": 0.14740833640098572,
"learning_rate": 1.7044707273753407e-05,
"loss": 0.0163,
"step": 4360
},
{
"epoch": 0.9808927920092029,
"grad_norm": 0.13891027867794037,
"learning_rate": 1.7026150386231814e-05,
"loss": 0.0178,
"step": 4370
},
{
"epoch": 0.9831373979405741,
"grad_norm": 0.15418022871017456,
"learning_rate": 1.700754559010954e-05,
"loss": 0.0183,
"step": 4380
},
{
"epoch": 0.9853820038719452,
"grad_norm": 0.14405600726604462,
"learning_rate": 1.6988893012245494e-05,
"loss": 0.0167,
"step": 4390
},
{
"epoch": 0.9876266098033164,
"grad_norm": 0.18409493565559387,
"learning_rate": 1.6970192779824383e-05,
"loss": 0.019,
"step": 4400
},
{
"epoch": 0.9898712157346876,
"grad_norm": 0.15557466447353363,
"learning_rate": 1.6951445020355853e-05,
"loss": 0.0171,
"step": 4410
},
{
"epoch": 0.9921158216660587,
"grad_norm": 0.1519443243741989,
"learning_rate": 1.6932649861673626e-05,
"loss": 0.0162,
"step": 4420
},
{
"epoch": 0.9943604275974299,
"grad_norm": 0.14214234054088593,
"learning_rate": 1.6913807431934612e-05,
"loss": 0.0179,
"step": 4430
},
{
"epoch": 0.9966050335288011,
"grad_norm": 0.14018166065216064,
"learning_rate": 1.6894917859618054e-05,
"loss": 0.0174,
"step": 4440
},
{
"epoch": 0.9988496394601722,
"grad_norm": 0.15058715641498566,
"learning_rate": 1.6875981273524632e-05,
"loss": 0.0174,
"step": 4450
},
{
"epoch": 1.0008978423725485,
"grad_norm": 0.17037492990493774,
"learning_rate": 1.68569978027756e-05,
"loss": 0.017,
"step": 4460
},
{
"epoch": 1.0031424483039197,
"grad_norm": 0.1476113498210907,
"learning_rate": 1.68379675768119e-05,
"loss": 0.0172,
"step": 4470
},
{
"epoch": 1.0053870542352907,
"grad_norm": 0.13747438788414001,
"learning_rate": 1.6818890725393275e-05,
"loss": 0.0157,
"step": 4480
},
{
"epoch": 1.007631660166662,
"grad_norm": 0.18724720180034637,
"learning_rate": 1.67997673785974e-05,
"loss": 0.0172,
"step": 4490
},
{
"epoch": 1.0098762660980332,
"grad_norm": 0.16652311384677887,
"learning_rate": 1.678059766681897e-05,
"loss": 0.0166,
"step": 4500
},
{
"epoch": 1.0121208720294044,
"grad_norm": 0.14732131361961365,
"learning_rate": 1.6761381720768836e-05,
"loss": 0.0167,
"step": 4510
},
{
"epoch": 1.0143654779607756,
"grad_norm": 0.1441793441772461,
"learning_rate": 1.6742119671473096e-05,
"loss": 0.0165,
"step": 4520
},
{
"epoch": 1.0166100838921466,
"grad_norm": 0.14003914594650269,
"learning_rate": 1.6722811650272213e-05,
"loss": 0.0178,
"step": 4530
},
{
"epoch": 1.0188546898235178,
"grad_norm": 0.1431935429573059,
"learning_rate": 1.6703457788820113e-05,
"loss": 0.0173,
"step": 4540
},
{
"epoch": 1.021099295754889,
"grad_norm": 0.13941150903701782,
"learning_rate": 1.6684058219083283e-05,
"loss": 0.0161,
"step": 4550
},
{
"epoch": 1.0233439016862602,
"grad_norm": 0.1487084925174713,
"learning_rate": 1.6664613073339885e-05,
"loss": 0.0159,
"step": 4560
},
{
"epoch": 1.0255885076176314,
"grad_norm": 0.1264648735523224,
"learning_rate": 1.6645122484178847e-05,
"loss": 0.0163,
"step": 4570
},
{
"epoch": 1.0278331135490026,
"grad_norm": 0.1608554720878601,
"learning_rate": 1.662558658449895e-05,
"loss": 0.0159,
"step": 4580
},
{
"epoch": 1.0300777194803736,
"grad_norm": 0.15731705725193024,
"learning_rate": 1.6606005507507936e-05,
"loss": 0.0171,
"step": 4590
},
{
"epoch": 1.0323223254117448,
"grad_norm": 0.1623576283454895,
"learning_rate": 1.6586379386721593e-05,
"loss": 0.0167,
"step": 4600
},
{
"epoch": 1.034566931343116,
"grad_norm": 0.17850559949874878,
"learning_rate": 1.6566708355962845e-05,
"loss": 0.0167,
"step": 4610
},
{
"epoch": 1.0368115372744873,
"grad_norm": 0.13232485949993134,
"learning_rate": 1.654699254936083e-05,
"loss": 0.018,
"step": 4620
},
{
"epoch": 1.0390561432058585,
"grad_norm": 0.13724569976329803,
"learning_rate": 1.6527232101350017e-05,
"loss": 0.0159,
"step": 4630
},
{
"epoch": 1.0413007491372297,
"grad_norm": 0.14290913939476013,
"learning_rate": 1.6507427146669248e-05,
"loss": 0.0169,
"step": 4640
},
{
"epoch": 1.0435453550686007,
"grad_norm": 0.1413612961769104,
"learning_rate": 1.6487577820360844e-05,
"loss": 0.0168,
"step": 4650
},
{
"epoch": 1.045789960999972,
"grad_norm": 0.15013279020786285,
"learning_rate": 1.6467684257769684e-05,
"loss": 0.0163,
"step": 4660
},
{
"epoch": 1.0480345669313431,
"grad_norm": 0.142470121383667,
"learning_rate": 1.6447746594542273e-05,
"loss": 0.0166,
"step": 4670
},
{
"epoch": 1.0502791728627143,
"grad_norm": 0.15328320860862732,
"learning_rate": 1.6427764966625817e-05,
"loss": 0.0161,
"step": 4680
},
{
"epoch": 1.0525237787940855,
"grad_norm": 0.13435743749141693,
"learning_rate": 1.640773951026731e-05,
"loss": 0.0154,
"step": 4690
},
{
"epoch": 1.0547683847254565,
"grad_norm": 0.12938353419303894,
"learning_rate": 1.638767036201259e-05,
"loss": 0.016,
"step": 4700
},
{
"epoch": 1.0570129906568277,
"grad_norm": 0.13953883945941925,
"learning_rate": 1.636755765870541e-05,
"loss": 0.017,
"step": 4710
},
{
"epoch": 1.059257596588199,
"grad_norm": 0.14840665459632874,
"learning_rate": 1.634740153748651e-05,
"loss": 0.0156,
"step": 4720
},
{
"epoch": 1.0615022025195702,
"grad_norm": 0.14732645452022552,
"learning_rate": 1.6327202135792687e-05,
"loss": 0.0153,
"step": 4730
},
{
"epoch": 1.0637468084509414,
"grad_norm": 0.15273860096931458,
"learning_rate": 1.6306959591355838e-05,
"loss": 0.0162,
"step": 4740
},
{
"epoch": 1.0659914143823126,
"grad_norm": 0.1429833620786667,
"learning_rate": 1.6286674042202044e-05,
"loss": 0.0147,
"step": 4750
},
{
"epoch": 1.0682360203136836,
"grad_norm": 0.14377611875534058,
"learning_rate": 1.6266345626650618e-05,
"loss": 0.0153,
"step": 4760
},
{
"epoch": 1.0704806262450548,
"grad_norm": 0.1336628645658493,
"learning_rate": 1.6245974483313153e-05,
"loss": 0.0152,
"step": 4770
},
{
"epoch": 1.072725232176426,
"grad_norm": 0.1435423344373703,
"learning_rate": 1.6225560751092598e-05,
"loss": 0.0163,
"step": 4780
},
{
"epoch": 1.0749698381077972,
"grad_norm": 0.14871567487716675,
"learning_rate": 1.6205104569182294e-05,
"loss": 0.0161,
"step": 4790
},
{
"epoch": 1.0772144440391684,
"grad_norm": 0.17508377134799957,
"learning_rate": 1.6184606077065023e-05,
"loss": 0.0154,
"step": 4800
},
{
"epoch": 1.0794590499705397,
"grad_norm": 0.18037168681621552,
"learning_rate": 1.616406541451208e-05,
"loss": 0.0162,
"step": 4810
},
{
"epoch": 1.0817036559019106,
"grad_norm": 0.12987300753593445,
"learning_rate": 1.6143482721582285e-05,
"loss": 0.0156,
"step": 4820
},
{
"epoch": 1.0839482618332819,
"grad_norm": 0.135226309299469,
"learning_rate": 1.6122858138621068e-05,
"loss": 0.0154,
"step": 4830
},
{
"epoch": 1.086192867764653,
"grad_norm": 0.1384623944759369,
"learning_rate": 1.610219180625947e-05,
"loss": 0.0162,
"step": 4840
},
{
"epoch": 1.0884374736960243,
"grad_norm": 0.2058112919330597,
"learning_rate": 1.6081483865413215e-05,
"loss": 0.0162,
"step": 4850
},
{
"epoch": 1.0906820796273955,
"grad_norm": 0.14403843879699707,
"learning_rate": 1.6060734457281745e-05,
"loss": 0.015,
"step": 4860
},
{
"epoch": 1.0929266855587665,
"grad_norm": 0.11267949640750885,
"learning_rate": 1.603994372334724e-05,
"loss": 0.0147,
"step": 4870
},
{
"epoch": 1.0951712914901377,
"grad_norm": 0.1622733771800995,
"learning_rate": 1.6019111805373672e-05,
"loss": 0.0156,
"step": 4880
},
{
"epoch": 1.097415897421509,
"grad_norm": 0.16310657560825348,
"learning_rate": 1.5998238845405823e-05,
"loss": 0.0157,
"step": 4890
},
{
"epoch": 1.0996605033528801,
"grad_norm": 0.11895309388637543,
"learning_rate": 1.5977324985768344e-05,
"loss": 0.0159,
"step": 4900
},
{
"epoch": 1.1019051092842513,
"grad_norm": 0.12787587940692902,
"learning_rate": 1.5956370369064733e-05,
"loss": 0.0157,
"step": 4910
},
{
"epoch": 1.1041497152156226,
"grad_norm": 0.1474785953760147,
"learning_rate": 1.5935375138176422e-05,
"loss": 0.0162,
"step": 4920
},
{
"epoch": 1.1063943211469935,
"grad_norm": 0.13858622312545776,
"learning_rate": 1.591433943626177e-05,
"loss": 0.0164,
"step": 4930
},
{
"epoch": 1.1086389270783648,
"grad_norm": 0.15037216246128082,
"learning_rate": 1.5893263406755082e-05,
"loss": 0.0166,
"step": 4940
},
{
"epoch": 1.110883533009736,
"grad_norm": 0.1463032215833664,
"learning_rate": 1.587214719336565e-05,
"loss": 0.0152,
"step": 4950
},
{
"epoch": 1.1131281389411072,
"grad_norm": 0.14497928321361542,
"learning_rate": 1.585099094007676e-05,
"loss": 0.0151,
"step": 4960
},
{
"epoch": 1.1153727448724784,
"grad_norm": 0.1287575364112854,
"learning_rate": 1.5829794791144723e-05,
"loss": 0.0157,
"step": 4970
},
{
"epoch": 1.1176173508038496,
"grad_norm": 0.1426306962966919,
"learning_rate": 1.580855889109787e-05,
"loss": 0.0155,
"step": 4980
},
{
"epoch": 1.1198619567352206,
"grad_norm": 0.14880262315273285,
"learning_rate": 1.578728338473559e-05,
"loss": 0.0167,
"step": 4990
},
{
"epoch": 1.1221065626665918,
"grad_norm": 0.14157725870609283,
"learning_rate": 1.5765968417127325e-05,
"loss": 0.0167,
"step": 5000
}
],
"logging_steps": 10,
"max_steps": 13368,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 2500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.4745131970702868e+19,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}