spatialstack_ablation_second_r2r / trainer_state.json
phamquandung's picture
Upload folder using huggingface_hub
610dc50 verified
Raw History Blame Contribute Delete
185 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 9864,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0010138710229324953,
"grad_norm": 62.021419525146484,
"learning_rate": 6.081081081081082e-07,
"loss": 1.5460986137390136,
"step": 10
},
{
"epoch": 0.0020277420458649905,
"grad_norm": 22.252260208129883,
"learning_rate": 1.2837837837837838e-06,
"loss": 0.2811699390411377,
"step": 20
},
{
"epoch": 0.003041613068797486,
"grad_norm": 14.530186653137207,
"learning_rate": 1.9594594594594595e-06,
"loss": 0.22973246574401857,
"step": 30
},
{
"epoch": 0.004055484091729981,
"grad_norm": 18.47085952758789,
"learning_rate": 2.6351351351351353e-06,
"loss": 0.2235541820526123,
"step": 40
},
{
"epoch": 0.005069355114662476,
"grad_norm": 14.398887634277344,
"learning_rate": 3.310810810810811e-06,
"loss": 0.2268673896789551,
"step": 50
},
{
"epoch": 0.006083226137594972,
"grad_norm": 10.402259826660156,
"learning_rate": 3.986486486486487e-06,
"loss": 0.24228920936584472,
"step": 60
},
{
"epoch": 0.0070970971605274664,
"grad_norm": 7.911172389984131,
"learning_rate": 4.6621621621621625e-06,
"loss": 0.21393949985504152,
"step": 70
},
{
"epoch": 0.008110968183459962,
"grad_norm": 9.686079025268555,
"learning_rate": 5.337837837837838e-06,
"loss": 0.23887848854064941,
"step": 80
},
{
"epoch": 0.009124839206392456,
"grad_norm": 6.521183490753174,
"learning_rate": 6.013513513513514e-06,
"loss": 0.22633512020111085,
"step": 90
},
{
"epoch": 0.010138710229324952,
"grad_norm": 7.011555194854736,
"learning_rate": 6.689189189189191e-06,
"loss": 0.2084352493286133,
"step": 100
},
{
"epoch": 0.011152581252257448,
"grad_norm": 8.008681297302246,
"learning_rate": 7.3648648648648655e-06,
"loss": 0.24362306594848632,
"step": 110
},
{
"epoch": 0.012166452275189943,
"grad_norm": 12.802906036376953,
"learning_rate": 8.040540540540541e-06,
"loss": 0.22571699619293212,
"step": 120
},
{
"epoch": 0.013180323298122437,
"grad_norm": 7.9114603996276855,
"learning_rate": 8.716216216216217e-06,
"loss": 0.22319207191467286,
"step": 130
},
{
"epoch": 0.014194194321054933,
"grad_norm": 8.737406730651855,
"learning_rate": 9.391891891891893e-06,
"loss": 0.22394332885742188,
"step": 140
},
{
"epoch": 0.015208065343987429,
"grad_norm": 13.713725090026855,
"learning_rate": 1.0067567567567569e-05,
"loss": 0.21357395648956298,
"step": 150
},
{
"epoch": 0.016221936366919924,
"grad_norm": 5.7237091064453125,
"learning_rate": 1.0743243243243244e-05,
"loss": 0.19049994945526122,
"step": 160
},
{
"epoch": 0.01723580738985242,
"grad_norm": 10.73676872253418,
"learning_rate": 1.141891891891892e-05,
"loss": 0.1988959789276123,
"step": 170
},
{
"epoch": 0.018249678412784912,
"grad_norm": 8.933271408081055,
"learning_rate": 1.2094594594594596e-05,
"loss": 0.19690818786621095,
"step": 180
},
{
"epoch": 0.019263549435717408,
"grad_norm": 10.613162994384766,
"learning_rate": 1.2770270270270272e-05,
"loss": 0.19739096164703368,
"step": 190
},
{
"epoch": 0.020277420458649904,
"grad_norm": 11.870220184326172,
"learning_rate": 1.3445945945945946e-05,
"loss": 0.23743460178375245,
"step": 200
},
{
"epoch": 0.0212912914815824,
"grad_norm": 5.415020942687988,
"learning_rate": 1.4121621621621621e-05,
"loss": 0.22554893493652345,
"step": 210
},
{
"epoch": 0.022305162504514895,
"grad_norm": 8.49407958984375,
"learning_rate": 1.4797297297297299e-05,
"loss": 0.1984415054321289,
"step": 220
},
{
"epoch": 0.02331903352744739,
"grad_norm": 5.5636067390441895,
"learning_rate": 1.5472972972972976e-05,
"loss": 0.19727907180786133,
"step": 230
},
{
"epoch": 0.024332904550379886,
"grad_norm": 4.73772668838501,
"learning_rate": 1.614864864864865e-05,
"loss": 0.1966691255569458,
"step": 240
},
{
"epoch": 0.02534677557331238,
"grad_norm": 6.54323673248291,
"learning_rate": 1.6824324324324324e-05,
"loss": 0.20856122970581054,
"step": 250
},
{
"epoch": 0.026360646596244874,
"grad_norm": 8.056952476501465,
"learning_rate": 1.7500000000000002e-05,
"loss": 0.21386234760284423,
"step": 260
},
{
"epoch": 0.02737451761917737,
"grad_norm": 7.149807929992676,
"learning_rate": 1.8175675675675676e-05,
"loss": 0.20478286743164062,
"step": 270
},
{
"epoch": 0.028388388642109866,
"grad_norm": 9.336917877197266,
"learning_rate": 1.8851351351351353e-05,
"loss": 0.21644811630249022,
"step": 280
},
{
"epoch": 0.02940225966504236,
"grad_norm": 5.670894145965576,
"learning_rate": 1.9527027027027027e-05,
"loss": 0.20888333320617675,
"step": 290
},
{
"epoch": 0.030416130687974857,
"grad_norm": 3.6504454612731934,
"learning_rate": 1.999999514856865e-05,
"loss": 0.18688127994537354,
"step": 300
},
{
"epoch": 0.03143000171090735,
"grad_norm": 3.965350866317749,
"learning_rate": 1.999990890103106e-05,
"loss": 0.19425785541534424,
"step": 310
},
{
"epoch": 0.03244387273383985,
"grad_norm": 6.121560573577881,
"learning_rate": 1.999971484497808e-05,
"loss": 0.19730303287506104,
"step": 320
},
{
"epoch": 0.033457743756772344,
"grad_norm": 4.194677829742432,
"learning_rate": 1.9999412982501806e-05,
"loss": 0.1787198305130005,
"step": 330
},
{
"epoch": 0.03447161477970484,
"grad_norm": 6.226432800292969,
"learning_rate": 1.9999003316856605e-05,
"loss": 0.18834079504013063,
"step": 340
},
{
"epoch": 0.03548548580263733,
"grad_norm": 6.038729667663574,
"learning_rate": 1.9998485852459068e-05,
"loss": 0.1778560161590576,
"step": 350
},
{
"epoch": 0.036499356825569824,
"grad_norm": 4.788501262664795,
"learning_rate": 1.9997860594887947e-05,
"loss": 0.20031824111938476,
"step": 360
},
{
"epoch": 0.03751322784850232,
"grad_norm": 10.611888885498047,
"learning_rate": 1.9997127550884116e-05,
"loss": 0.2026304244995117,
"step": 370
},
{
"epoch": 0.038527098871434816,
"grad_norm": 5.256176948547363,
"learning_rate": 1.9996286728350484e-05,
"loss": 0.18984835147857665,
"step": 380
},
{
"epoch": 0.03954096989436731,
"grad_norm": 5.486377716064453,
"learning_rate": 1.9995338136351913e-05,
"loss": 0.20085592269897462,
"step": 390
},
{
"epoch": 0.04055484091729981,
"grad_norm": 6.910491466522217,
"learning_rate": 1.9994281785115127e-05,
"loss": 0.18826465606689452,
"step": 400
},
{
"epoch": 0.0415687119402323,
"grad_norm": 4.588818073272705,
"learning_rate": 1.9993117686028603e-05,
"loss": 0.1818854570388794,
"step": 410
},
{
"epoch": 0.0425825829631648,
"grad_norm": 9.174452781677246,
"learning_rate": 1.9991845851642426e-05,
"loss": 0.2065051317214966,
"step": 420
},
{
"epoch": 0.043596453986097294,
"grad_norm": 4.2308454513549805,
"learning_rate": 1.999046629566818e-05,
"loss": 0.19427311420440674,
"step": 430
},
{
"epoch": 0.04461032500902979,
"grad_norm": 3.44735050201416,
"learning_rate": 1.998897903297879e-05,
"loss": 0.19772359132766723,
"step": 440
},
{
"epoch": 0.045624196031962286,
"grad_norm": 8.716513633728027,
"learning_rate": 1.9987384079608362e-05,
"loss": 0.19426032304763793,
"step": 450
},
{
"epoch": 0.04663806705489478,
"grad_norm": 6.285415172576904,
"learning_rate": 1.998568145275201e-05,
"loss": 0.20750689506530762,
"step": 460
},
{
"epoch": 0.04765193807782728,
"grad_norm": 4.095232963562012,
"learning_rate": 1.998387117076566e-05,
"loss": 0.2023838520050049,
"step": 470
},
{
"epoch": 0.04866580910075977,
"grad_norm": 6.3036322593688965,
"learning_rate": 1.998195325316588e-05,
"loss": 0.19709764719009398,
"step": 480
},
{
"epoch": 0.04967968012369226,
"grad_norm": 4.3454132080078125,
"learning_rate": 1.997992772062964e-05,
"loss": 0.211586332321167,
"step": 490
},
{
"epoch": 0.05069355114662476,
"grad_norm": 9.028620719909668,
"learning_rate": 1.9977794594994103e-05,
"loss": 0.20292437076568604,
"step": 500
},
{
"epoch": 0.05170742216955725,
"grad_norm": 2.2354586124420166,
"learning_rate": 1.9975553899256388e-05,
"loss": 0.20422961711883544,
"step": 510
},
{
"epoch": 0.05272129319248975,
"grad_norm": 3.614346504211426,
"learning_rate": 1.9973205657573328e-05,
"loss": 0.19386916160583495,
"step": 520
},
{
"epoch": 0.053735164215422244,
"grad_norm": 6.322831153869629,
"learning_rate": 1.99707498952612e-05,
"loss": 0.18057929277420043,
"step": 530
},
{
"epoch": 0.05474903523835474,
"grad_norm": 5.097690582275391,
"learning_rate": 1.9968186638795445e-05,
"loss": 0.17434091567993165,
"step": 540
},
{
"epoch": 0.055762906261287236,
"grad_norm": 4.673974514007568,
"learning_rate": 1.9965515915810408e-05,
"loss": 0.18120155334472657,
"step": 550
},
{
"epoch": 0.05677677728421973,
"grad_norm": 5.2899017333984375,
"learning_rate": 1.9962737755099023e-05,
"loss": 0.1942937970161438,
"step": 560
},
{
"epoch": 0.05779064830715223,
"grad_norm": 3.852194309234619,
"learning_rate": 1.9959852186612492e-05,
"loss": 0.19399497509002686,
"step": 570
},
{
"epoch": 0.05880451933008472,
"grad_norm": 5.761685848236084,
"learning_rate": 1.995685924145999e-05,
"loss": 0.18722137212753295,
"step": 580
},
{
"epoch": 0.05981839035301722,
"grad_norm": 6.441926002502441,
"learning_rate": 1.9953758951908305e-05,
"loss": 0.1709079146385193,
"step": 590
},
{
"epoch": 0.060832261375949714,
"grad_norm": 4.219082355499268,
"learning_rate": 1.995055135138151e-05,
"loss": 0.1924675226211548,
"step": 600
},
{
"epoch": 0.06184613239888221,
"grad_norm": 6.043975830078125,
"learning_rate": 1.994723647446058e-05,
"loss": 0.19664247035980226,
"step": 610
},
{
"epoch": 0.0628600034218147,
"grad_norm": 2.7879490852355957,
"learning_rate": 1.994381435688303e-05,
"loss": 0.19006526470184326,
"step": 620
},
{
"epoch": 0.0638738744447472,
"grad_norm": 4.146441459655762,
"learning_rate": 1.994028503554255e-05,
"loss": 0.18191558122634888,
"step": 630
},
{
"epoch": 0.0648877454676797,
"grad_norm": 5.49281120300293,
"learning_rate": 1.9936648548488568e-05,
"loss": 0.19351388216018678,
"step": 640
},
{
"epoch": 0.06590161649061219,
"grad_norm": 8.451753616333008,
"learning_rate": 1.993290493492587e-05,
"loss": 0.18637858629226683,
"step": 650
},
{
"epoch": 0.06691548751354469,
"grad_norm": 8.582316398620605,
"learning_rate": 1.9929054235214168e-05,
"loss": 0.17686495780944825,
"step": 660
},
{
"epoch": 0.06792935853647718,
"grad_norm": 6.897122859954834,
"learning_rate": 1.992509649086766e-05,
"loss": 0.1525785207748413,
"step": 670
},
{
"epoch": 0.06894322955940968,
"grad_norm": 4.552595615386963,
"learning_rate": 1.9921031744554588e-05,
"loss": 0.20790276527404786,
"step": 680
},
{
"epoch": 0.06995710058234217,
"grad_norm": 6.036705017089844,
"learning_rate": 1.991686004009678e-05,
"loss": 0.18892626762390136,
"step": 690
},
{
"epoch": 0.07097097160527466,
"grad_norm": 3.9059786796569824,
"learning_rate": 1.991258142246917e-05,
"loss": 0.19136371612548828,
"step": 700
},
{
"epoch": 0.07198484262820716,
"grad_norm": 6.214681625366211,
"learning_rate": 1.990819593779932e-05,
"loss": 0.19012080430984496,
"step": 710
},
{
"epoch": 0.07299871365113965,
"grad_norm": 5.653169631958008,
"learning_rate": 1.9903703633366923e-05,
"loss": 0.17889082431793213,
"step": 720
},
{
"epoch": 0.07401258467407215,
"grad_norm": 9.918580055236816,
"learning_rate": 1.9899104557603285e-05,
"loss": 0.174575138092041,
"step": 730
},
{
"epoch": 0.07502645569700464,
"grad_norm": 8.099326133728027,
"learning_rate": 1.9894398760090807e-05,
"loss": 0.21620001792907714,
"step": 740
},
{
"epoch": 0.07604032671993714,
"grad_norm": 4.759199619293213,
"learning_rate": 1.988958629156246e-05,
"loss": 0.19416173696517944,
"step": 750
},
{
"epoch": 0.07705419774286963,
"grad_norm": 2.6554713249206543,
"learning_rate": 1.988466720390123e-05,
"loss": 0.1893139123916626,
"step": 760
},
{
"epoch": 0.07806806876580213,
"grad_norm": 4.75584602355957,
"learning_rate": 1.9879641550139545e-05,
"loss": 0.18220908641815187,
"step": 770
},
{
"epoch": 0.07908193978873462,
"grad_norm": 6.723847389221191,
"learning_rate": 1.9874509384458726e-05,
"loss": 0.18354787826538085,
"step": 780
},
{
"epoch": 0.08009581081166713,
"grad_norm": 2.430344581604004,
"learning_rate": 1.9869270762188402e-05,
"loss": 0.18160395622253417,
"step": 790
},
{
"epoch": 0.08110968183459961,
"grad_norm": 6.609601020812988,
"learning_rate": 1.986392573980589e-05,
"loss": 0.1999581813812256,
"step": 800
},
{
"epoch": 0.08212355285753212,
"grad_norm": 5.581293106079102,
"learning_rate": 1.9858474374935606e-05,
"loss": 0.1819005489349365,
"step": 810
},
{
"epoch": 0.0831374238804646,
"grad_norm": 3.27018141746521,
"learning_rate": 1.9852916726348446e-05,
"loss": 0.18695205450057983,
"step": 820
},
{
"epoch": 0.0841512949033971,
"grad_norm": 2.5501277446746826,
"learning_rate": 1.9847252853961136e-05,
"loss": 0.18115479946136476,
"step": 830
},
{
"epoch": 0.0851651659263296,
"grad_norm": 6.091857433319092,
"learning_rate": 1.98414828188356e-05,
"loss": 0.18884366750717163,
"step": 840
},
{
"epoch": 0.08617903694926209,
"grad_norm": 3.0005879402160645,
"learning_rate": 1.98356066831783e-05,
"loss": 0.1822347402572632,
"step": 850
},
{
"epoch": 0.08719290797219459,
"grad_norm": 4.181893825531006,
"learning_rate": 1.982962451033956e-05,
"loss": 0.17895636558532715,
"step": 860
},
{
"epoch": 0.08820677899512708,
"grad_norm": 4.0642828941345215,
"learning_rate": 1.9823536364812875e-05,
"loss": 0.17416672706604003,
"step": 870
},
{
"epoch": 0.08922065001805958,
"grad_norm": 4.161611557006836,
"learning_rate": 1.981734231223425e-05,
"loss": 0.16172165870666505,
"step": 880
},
{
"epoch": 0.09023452104099207,
"grad_norm": 6.35130500793457,
"learning_rate": 1.9811042419381444e-05,
"loss": 0.17692358493804933,
"step": 890
},
{
"epoch": 0.09124839206392457,
"grad_norm": 5.028285503387451,
"learning_rate": 1.980463675417329e-05,
"loss": 0.17761753797531127,
"step": 900
},
{
"epoch": 0.09226226308685706,
"grad_norm": 5.436389923095703,
"learning_rate": 1.9798125385668942e-05,
"loss": 0.18976407051086425,
"step": 910
},
{
"epoch": 0.09327613410978956,
"grad_norm": 2.9825220108032227,
"learning_rate": 1.9791508384067138e-05,
"loss": 0.1789491891860962,
"step": 920
},
{
"epoch": 0.09429000513272205,
"grad_norm": 3.166353225708008,
"learning_rate": 1.9784785820705435e-05,
"loss": 0.184602689743042,
"step": 930
},
{
"epoch": 0.09530387615565455,
"grad_norm": 5.379031658172607,
"learning_rate": 1.9777957768059456e-05,
"loss": 0.17579230070114135,
"step": 940
},
{
"epoch": 0.09631774717858704,
"grad_norm": 8.485708236694336,
"learning_rate": 1.977102429974209e-05,
"loss": 0.18115860223770142,
"step": 950
},
{
"epoch": 0.09733161820151955,
"grad_norm": 4.98340368270874,
"learning_rate": 1.9763985490502714e-05,
"loss": 0.15802576541900634,
"step": 960
},
{
"epoch": 0.09834548922445203,
"grad_norm": 5.106139183044434,
"learning_rate": 1.9756841416226372e-05,
"loss": 0.19273967742919923,
"step": 970
},
{
"epoch": 0.09935936024738452,
"grad_norm": 4.630636215209961,
"learning_rate": 1.9749592153932974e-05,
"loss": 0.18440072536468505,
"step": 980
},
{
"epoch": 0.10037323127031703,
"grad_norm": 6.212395668029785,
"learning_rate": 1.9742237781776455e-05,
"loss": 0.17631895542144777,
"step": 990
},
{
"epoch": 0.10138710229324951,
"grad_norm": 3.0858983993530273,
"learning_rate": 1.9734778379043923e-05,
"loss": 0.1957733988761902,
"step": 1000
},
{
"epoch": 0.10240097331618202,
"grad_norm": 3.5432631969451904,
"learning_rate": 1.9727214026154827e-05,
"loss": 0.17648632526397706,
"step": 1010
},
{
"epoch": 0.1034148443391145,
"grad_norm": 5.26745080947876,
"learning_rate": 1.9719544804660084e-05,
"loss": 0.18444701433181762,
"step": 1020
},
{
"epoch": 0.10442871536204701,
"grad_norm": 8.154745101928711,
"learning_rate": 1.9711770797241174e-05,
"loss": 0.19123973846435546,
"step": 1030
},
{
"epoch": 0.1054425863849795,
"grad_norm": 3.113598346710205,
"learning_rate": 1.9703892087709287e-05,
"loss": 0.15859501361846923,
"step": 1040
},
{
"epoch": 0.106456457407912,
"grad_norm": 2.0993430614471436,
"learning_rate": 1.9695908761004392e-05,
"loss": 0.16301543712615968,
"step": 1050
},
{
"epoch": 0.10747032843084449,
"grad_norm": 3.11907958984375,
"learning_rate": 1.9687820903194335e-05,
"loss": 0.15477687120437622,
"step": 1060
},
{
"epoch": 0.10848419945377699,
"grad_norm": 3.5661044120788574,
"learning_rate": 1.9679628601473905e-05,
"loss": 0.1985307216644287,
"step": 1070
},
{
"epoch": 0.10949807047670948,
"grad_norm": 5.560718536376953,
"learning_rate": 1.967133194416389e-05,
"loss": 0.17782819271087646,
"step": 1080
},
{
"epoch": 0.11051194149964198,
"grad_norm": 5.289105415344238,
"learning_rate": 1.9662931020710138e-05,
"loss": 0.18611276149749756,
"step": 1090
},
{
"epoch": 0.11152581252257447,
"grad_norm": 2.6720094680786133,
"learning_rate": 1.965442592168259e-05,
"loss": 0.19742172956466675,
"step": 1100
},
{
"epoch": 0.11253968354550696,
"grad_norm": 4.448987007141113,
"learning_rate": 1.964581673877428e-05,
"loss": 0.16697547435760499,
"step": 1110
},
{
"epoch": 0.11355355456843946,
"grad_norm": 2.618384838104248,
"learning_rate": 1.9637103564800384e-05,
"loss": 0.169250750541687,
"step": 1120
},
{
"epoch": 0.11456742559137195,
"grad_norm": 3.548800468444824,
"learning_rate": 1.962828649369719e-05,
"loss": 0.17893331050872802,
"step": 1130
},
{
"epoch": 0.11558129661430445,
"grad_norm": 5.874704360961914,
"learning_rate": 1.96193656205211e-05,
"loss": 0.17956098318099975,
"step": 1140
},
{
"epoch": 0.11659516763723694,
"grad_norm": 3.164788246154785,
"learning_rate": 1.9610341041447595e-05,
"loss": 0.17496427297592163,
"step": 1150
},
{
"epoch": 0.11760903866016945,
"grad_norm": 2.7442479133605957,
"learning_rate": 1.9601212853770214e-05,
"loss": 0.16691102981567382,
"step": 1160
},
{
"epoch": 0.11862290968310193,
"grad_norm": 6.0904412269592285,
"learning_rate": 1.959198115589949e-05,
"loss": 0.19420123100280762,
"step": 1170
},
{
"epoch": 0.11963678070603444,
"grad_norm": 3.2445404529571533,
"learning_rate": 1.9582646047361888e-05,
"loss": 0.1711505889892578,
"step": 1180
},
{
"epoch": 0.12065065172896693,
"grad_norm": 1.8408147096633911,
"learning_rate": 1.9573207628798747e-05,
"loss": 0.17035949230194092,
"step": 1190
},
{
"epoch": 0.12166452275189943,
"grad_norm": 2.598249912261963,
"learning_rate": 1.956366600196519e-05,
"loss": 0.19111280441284179,
"step": 1200
},
{
"epoch": 0.12267839377483192,
"grad_norm": 2.228766441345215,
"learning_rate": 1.9554021269729003e-05,
"loss": 0.17088277339935304,
"step": 1210
},
{
"epoch": 0.12369226479776442,
"grad_norm": 3.276712656021118,
"learning_rate": 1.9544273536069573e-05,
"loss": 0.1715364933013916,
"step": 1220
},
{
"epoch": 0.12470613582069691,
"grad_norm": 6.10109281539917,
"learning_rate": 1.9534422906076714e-05,
"loss": 0.18286828994750975,
"step": 1230
},
{
"epoch": 0.1257200068436294,
"grad_norm": 1.8711620569229126,
"learning_rate": 1.9524469485949586e-05,
"loss": 0.17996398210525513,
"step": 1240
},
{
"epoch": 0.1267338778665619,
"grad_norm": 4.709359645843506,
"learning_rate": 1.9514413382995498e-05,
"loss": 0.1972990870475769,
"step": 1250
},
{
"epoch": 0.1277477488894944,
"grad_norm": 5.174152374267578,
"learning_rate": 1.9504254705628806e-05,
"loss": 0.17949414253234863,
"step": 1260
},
{
"epoch": 0.12876161991242688,
"grad_norm": 2.4198756217956543,
"learning_rate": 1.9493993563369688e-05,
"loss": 0.19394837617874144,
"step": 1270
},
{
"epoch": 0.1297754909353594,
"grad_norm": 5.560269832611084,
"learning_rate": 1.9483630066843005e-05,
"loss": 0.17067973613739013,
"step": 1280
},
{
"epoch": 0.13078936195829188,
"grad_norm": 2.370176076889038,
"learning_rate": 1.9473164327777095e-05,
"loss": 0.15654314756393434,
"step": 1290
},
{
"epoch": 0.13180323298122437,
"grad_norm": 4.748249530792236,
"learning_rate": 1.9462596459002565e-05,
"loss": 0.17893182039260863,
"step": 1300
},
{
"epoch": 0.13281710400415686,
"grad_norm": 1.2652174234390259,
"learning_rate": 1.9451926574451074e-05,
"loss": 0.1640818238258362,
"step": 1310
},
{
"epoch": 0.13383097502708938,
"grad_norm": 2.3822202682495117,
"learning_rate": 1.944115478915412e-05,
"loss": 0.17195615768432618,
"step": 1320
},
{
"epoch": 0.13484484605002187,
"grad_norm": 3.901524543762207,
"learning_rate": 1.9430281219241767e-05,
"loss": 0.17268164157867433,
"step": 1330
},
{
"epoch": 0.13585871707295435,
"grad_norm": 1.5545343160629272,
"learning_rate": 1.9419305981941436e-05,
"loss": 0.1842245101928711,
"step": 1340
},
{
"epoch": 0.13687258809588684,
"grad_norm": 4.9156694412231445,
"learning_rate": 1.94082291955766e-05,
"loss": 0.16630892753601073,
"step": 1350
},
{
"epoch": 0.13788645911881936,
"grad_norm": 4.4813008308410645,
"learning_rate": 1.939705097956554e-05,
"loss": 0.17221004962921144,
"step": 1360
},
{
"epoch": 0.13890033014175185,
"grad_norm": 2.5484750270843506,
"learning_rate": 1.9385771454420048e-05,
"loss": 0.18418803215026855,
"step": 1370
},
{
"epoch": 0.13991420116468434,
"grad_norm": 4.803271770477295,
"learning_rate": 1.9374390741744107e-05,
"loss": 0.16605162620544434,
"step": 1380
},
{
"epoch": 0.14092807218761683,
"grad_norm": 7.078513145446777,
"learning_rate": 1.9362908964232616e-05,
"loss": 0.19816160202026367,
"step": 1390
},
{
"epoch": 0.14194194321054932,
"grad_norm": 2.1742444038391113,
"learning_rate": 1.9351326245670038e-05,
"loss": 0.1833820939064026,
"step": 1400
},
{
"epoch": 0.14295581423348183,
"grad_norm": 3.455594301223755,
"learning_rate": 1.933964271092908e-05,
"loss": 0.17028840780258178,
"step": 1410
},
{
"epoch": 0.14396968525641432,
"grad_norm": 2.8004674911499023,
"learning_rate": 1.932785848596934e-05,
"loss": 0.19512341022491456,
"step": 1420
},
{
"epoch": 0.1449835562793468,
"grad_norm": 3.2355940341949463,
"learning_rate": 1.9315973697835966e-05,
"loss": 0.18598527908325196,
"step": 1430
},
{
"epoch": 0.1459974273022793,
"grad_norm": 3.1999714374542236,
"learning_rate": 1.930398847465825e-05,
"loss": 0.17063626050949096,
"step": 1440
},
{
"epoch": 0.14701129832521181,
"grad_norm": 3.2557685375213623,
"learning_rate": 1.9291902945648283e-05,
"loss": 0.1639224648475647,
"step": 1450
},
{
"epoch": 0.1480251693481443,
"grad_norm": 2.863807201385498,
"learning_rate": 1.9279717241099554e-05,
"loss": 0.1876792073249817,
"step": 1460
},
{
"epoch": 0.1490390403710768,
"grad_norm": 1.8403249979019165,
"learning_rate": 1.9267431492385524e-05,
"loss": 0.1790088891983032,
"step": 1470
},
{
"epoch": 0.15005291139400928,
"grad_norm": 2.4912803173065186,
"learning_rate": 1.925504583195823e-05,
"loss": 0.17622541189193724,
"step": 1480
},
{
"epoch": 0.1510667824169418,
"grad_norm": 2.085783004760742,
"learning_rate": 1.924256039334686e-05,
"loss": 0.17695412635803223,
"step": 1490
},
{
"epoch": 0.15208065343987429,
"grad_norm": 2.157817840576172,
"learning_rate": 1.9229975311156305e-05,
"loss": 0.18331342935562134,
"step": 1500
},
{
"epoch": 0.15309452446280677,
"grad_norm": 3.835862159729004,
"learning_rate": 1.9217290721065693e-05,
"loss": 0.19229450225830078,
"step": 1510
},
{
"epoch": 0.15410839548573926,
"grad_norm": 4.298842430114746,
"learning_rate": 1.9204506759826957e-05,
"loss": 0.17591114044189454,
"step": 1520
},
{
"epoch": 0.15512226650867175,
"grad_norm": 1.3041647672653198,
"learning_rate": 1.9191623565263342e-05,
"loss": 0.1793145179748535,
"step": 1530
},
{
"epoch": 0.15613613753160427,
"grad_norm": 2.686422109603882,
"learning_rate": 1.9178641276267922e-05,
"loss": 0.17791481018066407,
"step": 1540
},
{
"epoch": 0.15715000855453676,
"grad_norm": 2.192145347595215,
"learning_rate": 1.91655600328021e-05,
"loss": 0.1834151864051819,
"step": 1550
},
{
"epoch": 0.15816387957746925,
"grad_norm": 7.047032833099365,
"learning_rate": 1.91523799758941e-05,
"loss": 0.16839590072631835,
"step": 1560
},
{
"epoch": 0.15917775060040174,
"grad_norm": 7.3560099601745605,
"learning_rate": 1.9139101247637465e-05,
"loss": 0.16786422729492187,
"step": 1570
},
{
"epoch": 0.16019162162333425,
"grad_norm": 5.999558448791504,
"learning_rate": 1.912572399118949e-05,
"loss": 0.17829041481018065,
"step": 1580
},
{
"epoch": 0.16120549264626674,
"grad_norm": 3.8529789447784424,
"learning_rate": 1.9112248350769704e-05,
"loss": 0.18974729776382446,
"step": 1590
},
{
"epoch": 0.16221936366919923,
"grad_norm": 4.099862098693848,
"learning_rate": 1.909867447165831e-05,
"loss": 0.1661280632019043,
"step": 1600
},
{
"epoch": 0.16323323469213172,
"grad_norm": 1.5326069593429565,
"learning_rate": 1.908500250019462e-05,
"loss": 0.16829389333724976,
"step": 1610
},
{
"epoch": 0.16424710571506423,
"grad_norm": 5.206752300262451,
"learning_rate": 1.9071232583775465e-05,
"loss": 0.16341688632965087,
"step": 1620
},
{
"epoch": 0.16526097673799672,
"grad_norm": 2.8415088653564453,
"learning_rate": 1.9057364870853628e-05,
"loss": 0.15261949300765992,
"step": 1630
},
{
"epoch": 0.1662748477609292,
"grad_norm": 3.5473005771636963,
"learning_rate": 1.904339951093622e-05,
"loss": 0.17281086444854737,
"step": 1640
},
{
"epoch": 0.1672887187838617,
"grad_norm": 2.8960368633270264,
"learning_rate": 1.902933665458309e-05,
"loss": 0.17578482627868652,
"step": 1650
},
{
"epoch": 0.1683025898067942,
"grad_norm": 2.692998170852661,
"learning_rate": 1.901517645340518e-05,
"loss": 0.18700896501541137,
"step": 1660
},
{
"epoch": 0.1693164608297267,
"grad_norm": 3.7117505073547363,
"learning_rate": 1.900091906006291e-05,
"loss": 0.18755427598953248,
"step": 1670
},
{
"epoch": 0.1703303318526592,
"grad_norm": 2.7558751106262207,
"learning_rate": 1.8986564628264527e-05,
"loss": 0.1871054768562317,
"step": 1680
},
{
"epoch": 0.17134420287559168,
"grad_norm": 2.3950297832489014,
"learning_rate": 1.8972113312764432e-05,
"loss": 0.16128084659576417,
"step": 1690
},
{
"epoch": 0.17235807389852417,
"grad_norm": 2.5185062885284424,
"learning_rate": 1.895756526936153e-05,
"loss": 0.16611018180847167,
"step": 1700
},
{
"epoch": 0.1733719449214567,
"grad_norm": 3.197969913482666,
"learning_rate": 1.8942920654897553e-05,
"loss": 0.17853533029556273,
"step": 1710
},
{
"epoch": 0.17438581594438918,
"grad_norm": 1.1210142374038696,
"learning_rate": 1.8928179627255347e-05,
"loss": 0.18526017665863037,
"step": 1720
},
{
"epoch": 0.17539968696732167,
"grad_norm": 3.008004903793335,
"learning_rate": 1.891334234535719e-05,
"loss": 0.1761615514755249,
"step": 1730
},
{
"epoch": 0.17641355799025416,
"grad_norm": 2.8438479900360107,
"learning_rate": 1.8898408969163078e-05,
"loss": 0.157371985912323,
"step": 1740
},
{
"epoch": 0.17742742901318667,
"grad_norm": 2.329352855682373,
"learning_rate": 1.8883379659668987e-05,
"loss": 0.18568003177642822,
"step": 1750
},
{
"epoch": 0.17844130003611916,
"grad_norm": 4.572145938873291,
"learning_rate": 1.8868254578905147e-05,
"loss": 0.18587981462478637,
"step": 1760
},
{
"epoch": 0.17945517105905165,
"grad_norm": 3.3880224227905273,
"learning_rate": 1.8853033889934293e-05,
"loss": 0.14560937881469727,
"step": 1770
},
{
"epoch": 0.18046904208198414,
"grad_norm": 2.0204989910125732,
"learning_rate": 1.883771775684991e-05,
"loss": 0.1610896348953247,
"step": 1780
},
{
"epoch": 0.18148291310491665,
"grad_norm": 4.179126739501953,
"learning_rate": 1.882230634477446e-05,
"loss": 0.17705745697021485,
"step": 1790
},
{
"epoch": 0.18249678412784914,
"grad_norm": 5.712968349456787,
"learning_rate": 1.8806799819857598e-05,
"loss": 0.1875457763671875,
"step": 1800
},
{
"epoch": 0.18351065515078163,
"grad_norm": 2.0398459434509277,
"learning_rate": 1.8791198349274393e-05,
"loss": 0.16037026643753052,
"step": 1810
},
{
"epoch": 0.18452452617371412,
"grad_norm": 4.262085914611816,
"learning_rate": 1.8775502101223516e-05,
"loss": 0.17650781869888305,
"step": 1820
},
{
"epoch": 0.1855383971966466,
"grad_norm": 3.581883668899536,
"learning_rate": 1.8759711244925424e-05,
"loss": 0.17241846323013305,
"step": 1830
},
{
"epoch": 0.18655226821957913,
"grad_norm": 2.296926259994507,
"learning_rate": 1.874382595062055e-05,
"loss": 0.15162642002105714,
"step": 1840
},
{
"epoch": 0.18756613924251161,
"grad_norm": 1.2849929332733154,
"learning_rate": 1.8727846389567457e-05,
"loss": 0.17084654569625854,
"step": 1850
},
{
"epoch": 0.1885800102654441,
"grad_norm": 3.266038417816162,
"learning_rate": 1.871177273404098e-05,
"loss": 0.18423705101013182,
"step": 1860
},
{
"epoch": 0.1895938812883766,
"grad_norm": 2.583120584487915,
"learning_rate": 1.8695605157330398e-05,
"loss": 0.15420024394989013,
"step": 1870
},
{
"epoch": 0.1906077523113091,
"grad_norm": 3.659569501876831,
"learning_rate": 1.8679343833737546e-05,
"loss": 0.18107463121414186,
"step": 1880
},
{
"epoch": 0.1916216233342416,
"grad_norm": 3.2595953941345215,
"learning_rate": 1.8662988938574926e-05,
"loss": 0.1679612874984741,
"step": 1890
},
{
"epoch": 0.1926354943571741,
"grad_norm": 2.94905686378479,
"learning_rate": 1.8646540648163853e-05,
"loss": 0.16105222702026367,
"step": 1900
},
{
"epoch": 0.19364936538010658,
"grad_norm": 3.533355474472046,
"learning_rate": 1.862999913983251e-05,
"loss": 0.17724395990371705,
"step": 1910
},
{
"epoch": 0.1946632364030391,
"grad_norm": 4.647031784057617,
"learning_rate": 1.8613364591914073e-05,
"loss": 0.15909557342529296,
"step": 1920
},
{
"epoch": 0.19567710742597158,
"grad_norm": 3.989941358566284,
"learning_rate": 1.8596637183744762e-05,
"loss": 0.16189029216766357,
"step": 1930
},
{
"epoch": 0.19669097844890407,
"grad_norm": 4.648385524749756,
"learning_rate": 1.857981709566193e-05,
"loss": 0.18520007133483887,
"step": 1940
},
{
"epoch": 0.19770484947183656,
"grad_norm": 3.3418097496032715,
"learning_rate": 1.85629045090021e-05,
"loss": 0.16784424781799318,
"step": 1950
},
{
"epoch": 0.19871872049476905,
"grad_norm": 1.822830080986023,
"learning_rate": 1.854589960609902e-05,
"loss": 0.14671528339385986,
"step": 1960
},
{
"epoch": 0.19973259151770156,
"grad_norm": 2.8054540157318115,
"learning_rate": 1.8528802570281686e-05,
"loss": 0.18483035564422606,
"step": 1970
},
{
"epoch": 0.20074646254063405,
"grad_norm": 5.281755447387695,
"learning_rate": 1.851161358587239e-05,
"loss": 0.16921157836914064,
"step": 1980
},
{
"epoch": 0.20176033356356654,
"grad_norm": 6.076530456542969,
"learning_rate": 1.8494332838184707e-05,
"loss": 0.18245284557342528,
"step": 1990
},
{
"epoch": 0.20277420458649903,
"grad_norm": 2.553358554840088,
"learning_rate": 1.847696051352151e-05,
"loss": 0.16620737314224243,
"step": 2000
},
{
"epoch": 0.20378807560943155,
"grad_norm": 2.087684154510498,
"learning_rate": 1.8459496799172957e-05,
"loss": 0.1580868363380432,
"step": 2010
},
{
"epoch": 0.20480194663236403,
"grad_norm": 7.9493727684021,
"learning_rate": 1.8441941883414475e-05,
"loss": 0.1946648359298706,
"step": 2020
},
{
"epoch": 0.20581581765529652,
"grad_norm": 2.4337525367736816,
"learning_rate": 1.8424295955504727e-05,
"loss": 0.16351776123046874,
"step": 2030
},
{
"epoch": 0.206829688678229,
"grad_norm": 2.257033586502075,
"learning_rate": 1.840655920568358e-05,
"loss": 0.18212809562683105,
"step": 2040
},
{
"epoch": 0.20784355970116153,
"grad_norm": 3.3089919090270996,
"learning_rate": 1.838873182517004e-05,
"loss": 0.17051252126693725,
"step": 2050
},
{
"epoch": 0.20885743072409402,
"grad_norm": 3.546356678009033,
"learning_rate": 1.83708140061602e-05,
"loss": 0.17348551750183105,
"step": 2060
},
{
"epoch": 0.2098713017470265,
"grad_norm": 3.6019411087036133,
"learning_rate": 1.835280594182517e-05,
"loss": 0.1629140257835388,
"step": 2070
},
{
"epoch": 0.210885172769959,
"grad_norm": 2.3701884746551514,
"learning_rate": 1.833470782630898e-05,
"loss": 0.17219257354736328,
"step": 2080
},
{
"epoch": 0.21189904379289148,
"grad_norm": 4.043793678283691,
"learning_rate": 1.831651985472651e-05,
"loss": 0.1713582158088684,
"step": 2090
},
{
"epoch": 0.212912914815824,
"grad_norm": 4.716768264770508,
"learning_rate": 1.8298242223161362e-05,
"loss": 0.18315128087997437,
"step": 2100
},
{
"epoch": 0.2139267858387565,
"grad_norm": 2.2644801139831543,
"learning_rate": 1.827987512866377e-05,
"loss": 0.16830618381500245,
"step": 2110
},
{
"epoch": 0.21494065686168898,
"grad_norm": 3.893977403640747,
"learning_rate": 1.826141876924845e-05,
"loss": 0.16109193563461305,
"step": 2120
},
{
"epoch": 0.21595452788462147,
"grad_norm": 8.111363410949707,
"learning_rate": 1.8242873343892494e-05,
"loss": 0.19096704721450805,
"step": 2130
},
{
"epoch": 0.21696839890755398,
"grad_norm": 3.086493968963623,
"learning_rate": 1.8224239052533194e-05,
"loss": 0.17559614181518554,
"step": 2140
},
{
"epoch": 0.21798226993048647,
"grad_norm": 3.5360262393951416,
"learning_rate": 1.820551609606591e-05,
"loss": 0.17029857635498047,
"step": 2150
},
{
"epoch": 0.21899614095341896,
"grad_norm": 2.3676207065582275,
"learning_rate": 1.81867046763419e-05,
"loss": 0.1838122844696045,
"step": 2160
},
{
"epoch": 0.22001001197635145,
"grad_norm": 2.7566721439361572,
"learning_rate": 1.816780499616613e-05,
"loss": 0.17880046367645264,
"step": 2170
},
{
"epoch": 0.22102388299928397,
"grad_norm": 2.6265792846679688,
"learning_rate": 1.81488172592951e-05,
"loss": 0.16293754577636718,
"step": 2180
},
{
"epoch": 0.22203775402221645,
"grad_norm": 3.1044182777404785,
"learning_rate": 1.812974167043465e-05,
"loss": 0.16721816062927247,
"step": 2190
},
{
"epoch": 0.22305162504514894,
"grad_norm": 2.447758436203003,
"learning_rate": 1.8110578435237748e-05,
"loss": 0.16733355522155763,
"step": 2200
},
{
"epoch": 0.22406549606808143,
"grad_norm": 2.908238649368286,
"learning_rate": 1.8091327760302267e-05,
"loss": 0.1737284779548645,
"step": 2210
},
{
"epoch": 0.22507936709101392,
"grad_norm": 9.919211387634277,
"learning_rate": 1.8071989853168765e-05,
"loss": 0.2197141170501709,
"step": 2220
},
{
"epoch": 0.22609323811394644,
"grad_norm": 1.3959416151046753,
"learning_rate": 1.8052564922318247e-05,
"loss": 0.17591532468795776,
"step": 2230
},
{
"epoch": 0.22710710913687893,
"grad_norm": 6.850743770599365,
"learning_rate": 1.803305317716992e-05,
"loss": 0.18894954919815063,
"step": 2240
},
{
"epoch": 0.22812098015981142,
"grad_norm": 2.7058162689208984,
"learning_rate": 1.8013454828078925e-05,
"loss": 0.16529448032379152,
"step": 2250
},
{
"epoch": 0.2291348511827439,
"grad_norm": 2.9855010509490967,
"learning_rate": 1.7993770086334082e-05,
"loss": 0.16036938428878783,
"step": 2260
},
{
"epoch": 0.23014872220567642,
"grad_norm": 3.212131977081299,
"learning_rate": 1.7973999164155597e-05,
"loss": 0.15762698650360107,
"step": 2270
},
{
"epoch": 0.2311625932286089,
"grad_norm": 3.2607977390289307,
"learning_rate": 1.795414227469279e-05,
"loss": 0.1579272985458374,
"step": 2280
},
{
"epoch": 0.2321764642515414,
"grad_norm": 4.722792625427246,
"learning_rate": 1.7934199632021795e-05,
"loss": 0.17293379306793213,
"step": 2290
},
{
"epoch": 0.2331903352744739,
"grad_norm": 2.316347360610962,
"learning_rate": 1.7914171451143235e-05,
"loss": 0.18338587284088134,
"step": 2300
},
{
"epoch": 0.2342042062974064,
"grad_norm": 2.3632454872131348,
"learning_rate": 1.7894057947979923e-05,
"loss": 0.16329991817474365,
"step": 2310
},
{
"epoch": 0.2352180773203389,
"grad_norm": 4.149104595184326,
"learning_rate": 1.7873859339374528e-05,
"loss": 0.19164077043533326,
"step": 2320
},
{
"epoch": 0.23623194834327138,
"grad_norm": 6.297323703765869,
"learning_rate": 1.7853575843087227e-05,
"loss": 0.17194660902023315,
"step": 2330
},
{
"epoch": 0.23724581936620387,
"grad_norm": 2.3634324073791504,
"learning_rate": 1.7833207677793382e-05,
"loss": 0.16985080242156983,
"step": 2340
},
{
"epoch": 0.23825969038913636,
"grad_norm": 1.603291630744934,
"learning_rate": 1.7812755063081156e-05,
"loss": 0.1687072515487671,
"step": 2350
},
{
"epoch": 0.23927356141206887,
"grad_norm": 3.6724653244018555,
"learning_rate": 1.7792218219449165e-05,
"loss": 0.1644115447998047,
"step": 2360
},
{
"epoch": 0.24028743243500136,
"grad_norm": 3.5850541591644287,
"learning_rate": 1.777159736830408e-05,
"loss": 0.16028075218200682,
"step": 2370
},
{
"epoch": 0.24130130345793385,
"grad_norm": 3.9248809814453125,
"learning_rate": 1.7750892731958268e-05,
"loss": 0.16430915594100953,
"step": 2380
},
{
"epoch": 0.24231517448086634,
"grad_norm": 2.8146421909332275,
"learning_rate": 1.773010453362737e-05,
"loss": 0.1613592743873596,
"step": 2390
},
{
"epoch": 0.24332904550379886,
"grad_norm": 3.774242401123047,
"learning_rate": 1.770923299742791e-05,
"loss": 0.16295711994171141,
"step": 2400
},
{
"epoch": 0.24434291652673135,
"grad_norm": 2.820051431655884,
"learning_rate": 1.768827834837487e-05,
"loss": 0.16539347171783447,
"step": 2410
},
{
"epoch": 0.24535678754966384,
"grad_norm": 4.351449489593506,
"learning_rate": 1.7667240812379266e-05,
"loss": 0.14830682277679444,
"step": 2420
},
{
"epoch": 0.24637065857259632,
"grad_norm": 2.9413020610809326,
"learning_rate": 1.7646120616245724e-05,
"loss": 0.16948883533477782,
"step": 2430
},
{
"epoch": 0.24738452959552884,
"grad_norm": 1.9874354600906372,
"learning_rate": 1.7624917987670006e-05,
"loss": 0.17281630039215087,
"step": 2440
},
{
"epoch": 0.24839840061846133,
"grad_norm": 2.802910566329956,
"learning_rate": 1.7603633155236595e-05,
"loss": 0.1651758670806885,
"step": 2450
},
{
"epoch": 0.24941227164139382,
"grad_norm": 3.2484872341156006,
"learning_rate": 1.7582266348416194e-05,
"loss": 0.17678223848342894,
"step": 2460
},
{
"epoch": 0.25042614266432633,
"grad_norm": 1.956612229347229,
"learning_rate": 1.7560817797563275e-05,
"loss": 0.1642727255821228,
"step": 2470
},
{
"epoch": 0.2514400136872588,
"grad_norm": 2.7539384365081787,
"learning_rate": 1.7539287733913592e-05,
"loss": 0.17392456531524658,
"step": 2480
},
{
"epoch": 0.2524538847101913,
"grad_norm": 4.354172706604004,
"learning_rate": 1.7517676389581675e-05,
"loss": 0.16802728176116943,
"step": 2490
},
{
"epoch": 0.2534677557331238,
"grad_norm": 4.203573703765869,
"learning_rate": 1.7495983997558342e-05,
"loss": 0.16166112422943116,
"step": 2500
},
{
"epoch": 0.2544816267560563,
"grad_norm": 2.9904534816741943,
"learning_rate": 1.7474210791708177e-05,
"loss": 0.16875858306884767,
"step": 2510
},
{
"epoch": 0.2554954977789888,
"grad_norm": 3.5539586544036865,
"learning_rate": 1.7452357006767026e-05,
"loss": 0.1583446741104126,
"step": 2520
},
{
"epoch": 0.25650936880192127,
"grad_norm": 1.9296183586120605,
"learning_rate": 1.7430422878339443e-05,
"loss": 0.16881377696990968,
"step": 2530
},
{
"epoch": 0.25752323982485376,
"grad_norm": 2.1659669876098633,
"learning_rate": 1.740840864289616e-05,
"loss": 0.15659983158111573,
"step": 2540
},
{
"epoch": 0.2585371108477863,
"grad_norm": 2.580319404602051,
"learning_rate": 1.7386314537771553e-05,
"loss": 0.16108183860778807,
"step": 2550
},
{
"epoch": 0.2595509818707188,
"grad_norm": 3.7652580738067627,
"learning_rate": 1.736414080116105e-05,
"loss": 0.18085507154464722,
"step": 2560
},
{
"epoch": 0.2605648528936513,
"grad_norm": 3.2972421646118164,
"learning_rate": 1.7341887672118596e-05,
"loss": 0.1669074773788452,
"step": 2570
},
{
"epoch": 0.26157872391658377,
"grad_norm": 3.26116943359375,
"learning_rate": 1.7319555390554054e-05,
"loss": 0.19582275152206421,
"step": 2580
},
{
"epoch": 0.26259259493951626,
"grad_norm": 2.813002586364746,
"learning_rate": 1.7297144197230637e-05,
"loss": 0.1766123056411743,
"step": 2590
},
{
"epoch": 0.26360646596244874,
"grad_norm": 3.350167989730835,
"learning_rate": 1.7274654333762284e-05,
"loss": 0.15856661796569824,
"step": 2600
},
{
"epoch": 0.26462033698538123,
"grad_norm": 4.5556745529174805,
"learning_rate": 1.7252086042611086e-05,
"loss": 0.17615453004837037,
"step": 2610
},
{
"epoch": 0.2656342080083137,
"grad_norm": 2.4725983142852783,
"learning_rate": 1.722943956708466e-05,
"loss": 0.16040667295455932,
"step": 2620
},
{
"epoch": 0.2666480790312462,
"grad_norm": 3.6839599609375,
"learning_rate": 1.720671515133352e-05,
"loss": 0.16277241706848145,
"step": 2630
},
{
"epoch": 0.26766195005417875,
"grad_norm": 3.7754201889038086,
"learning_rate": 1.7183913040348447e-05,
"loss": 0.16113916635513306,
"step": 2640
},
{
"epoch": 0.26867582107711124,
"grad_norm": 2.9745471477508545,
"learning_rate": 1.716103347995785e-05,
"loss": 0.15952463150024415,
"step": 2650
},
{
"epoch": 0.26968969210004373,
"grad_norm": 2.0326273441314697,
"learning_rate": 1.713807671682513e-05,
"loss": 0.14709432125091554,
"step": 2660
},
{
"epoch": 0.2707035631229762,
"grad_norm": 3.365922451019287,
"learning_rate": 1.711504299844599e-05,
"loss": 0.16798888444900512,
"step": 2670
},
{
"epoch": 0.2717174341459087,
"grad_norm": 2.283329725265503,
"learning_rate": 1.7091932573145798e-05,
"loss": 0.17366816997528076,
"step": 2680
},
{
"epoch": 0.2727313051688412,
"grad_norm": 3.9605870246887207,
"learning_rate": 1.706874569007688e-05,
"loss": 0.16557005643844605,
"step": 2690
},
{
"epoch": 0.2737451761917737,
"grad_norm": 1.7530375719070435,
"learning_rate": 1.7045482599215856e-05,
"loss": 0.1802292585372925,
"step": 2700
},
{
"epoch": 0.2747590472147062,
"grad_norm": 3.836883544921875,
"learning_rate": 1.7022143551360947e-05,
"loss": 0.15431183576583862,
"step": 2710
},
{
"epoch": 0.2757729182376387,
"grad_norm": 4.177915573120117,
"learning_rate": 1.699872879812925e-05,
"loss": 0.18119832277297973,
"step": 2720
},
{
"epoch": 0.2767867892605712,
"grad_norm": 3.9304263591766357,
"learning_rate": 1.6975238591954048e-05,
"loss": 0.18387283086776735,
"step": 2730
},
{
"epoch": 0.2778006602835037,
"grad_norm": 3.2422659397125244,
"learning_rate": 1.695167318608206e-05,
"loss": 0.1660290241241455,
"step": 2740
},
{
"epoch": 0.2788145313064362,
"grad_norm": 1.841316819190979,
"learning_rate": 1.6928032834570767e-05,
"loss": 0.1722469449043274,
"step": 2750
},
{
"epoch": 0.2798284023293687,
"grad_norm": 3.4637413024902344,
"learning_rate": 1.690431779228559e-05,
"loss": 0.17210263013839722,
"step": 2760
},
{
"epoch": 0.28084227335230116,
"grad_norm": 4.952032566070557,
"learning_rate": 1.6880528314897212e-05,
"loss": 0.16909925937652587,
"step": 2770
},
{
"epoch": 0.28185614437523365,
"grad_norm": 4.525207996368408,
"learning_rate": 1.6856664658878797e-05,
"loss": 0.16290751695632935,
"step": 2780
},
{
"epoch": 0.28287001539816614,
"grad_norm": 2.7825348377227783,
"learning_rate": 1.6832727081503227e-05,
"loss": 0.17031203508377074,
"step": 2790
},
{
"epoch": 0.28388388642109863,
"grad_norm": 3.0043795108795166,
"learning_rate": 1.6808715840840308e-05,
"loss": 0.15307979583740233,
"step": 2800
},
{
"epoch": 0.2848977574440312,
"grad_norm": 4.287590503692627,
"learning_rate": 1.6784631195754027e-05,
"loss": 0.17513206005096435,
"step": 2810
},
{
"epoch": 0.28591162846696366,
"grad_norm": 2.5217339992523193,
"learning_rate": 1.6760473405899733e-05,
"loss": 0.14664981365203858,
"step": 2820
},
{
"epoch": 0.28692549948989615,
"grad_norm": 3.296112060546875,
"learning_rate": 1.673624273172134e-05,
"loss": 0.16826030015945434,
"step": 2830
},
{
"epoch": 0.28793937051282864,
"grad_norm": 2.0429632663726807,
"learning_rate": 1.6711939434448537e-05,
"loss": 0.15776162147521972,
"step": 2840
},
{
"epoch": 0.28895324153576113,
"grad_norm": 4.890894889831543,
"learning_rate": 1.6687563776093943e-05,
"loss": 0.17978334426879883,
"step": 2850
},
{
"epoch": 0.2899671125586936,
"grad_norm": 2.8318073749542236,
"learning_rate": 1.6663116019450306e-05,
"loss": 0.15746296644210817,
"step": 2860
},
{
"epoch": 0.2909809835816261,
"grad_norm": 1.5715001821517944,
"learning_rate": 1.6638596428087662e-05,
"loss": 0.18041276931762695,
"step": 2870
},
{
"epoch": 0.2919948546045586,
"grad_norm": 4.541372776031494,
"learning_rate": 1.66140052663505e-05,
"loss": 0.16454099416732787,
"step": 2880
},
{
"epoch": 0.29300872562749114,
"grad_norm": 2.800013303756714,
"learning_rate": 1.6589342799354895e-05,
"loss": 0.16230067014694213,
"step": 2890
},
{
"epoch": 0.29402259665042363,
"grad_norm": 1.9327799081802368,
"learning_rate": 1.6564609292985663e-05,
"loss": 0.16799471378326417,
"step": 2900
},
{
"epoch": 0.2950364676733561,
"grad_norm": 3.9879050254821777,
"learning_rate": 1.6539805013893493e-05,
"loss": 0.1707507610321045,
"step": 2910
},
{
"epoch": 0.2960503386962886,
"grad_norm": 5.380219459533691,
"learning_rate": 1.6514930229492073e-05,
"loss": 0.1742720365524292,
"step": 2920
},
{
"epoch": 0.2970642097192211,
"grad_norm": 4.503504753112793,
"learning_rate": 1.6489985207955207e-05,
"loss": 0.16898276805877685,
"step": 2930
},
{
"epoch": 0.2980780807421536,
"grad_norm": 3.2431960105895996,
"learning_rate": 1.6464970218213912e-05,
"loss": 0.169741690158844,
"step": 2940
},
{
"epoch": 0.2990919517650861,
"grad_norm": 2.992523670196533,
"learning_rate": 1.643988552995354e-05,
"loss": 0.17430849075317384,
"step": 2950
},
{
"epoch": 0.30010582278801856,
"grad_norm": 5.103341102600098,
"learning_rate": 1.6414731413610856e-05,
"loss": 0.15804864168167115,
"step": 2960
},
{
"epoch": 0.30111969381095105,
"grad_norm": 2.466834783554077,
"learning_rate": 1.638950814037112e-05,
"loss": 0.16053221225738526,
"step": 2970
},
{
"epoch": 0.3021335648338836,
"grad_norm": 2.8250789642333984,
"learning_rate": 1.6364215982165178e-05,
"loss": 0.1593393087387085,
"step": 2980
},
{
"epoch": 0.3031474358568161,
"grad_norm": 1.742450475692749,
"learning_rate": 1.633885521166651e-05,
"loss": 0.15977132320404053,
"step": 2990
},
{
"epoch": 0.30416130687974857,
"grad_norm": 2.384411096572876,
"learning_rate": 1.6313426102288315e-05,
"loss": 0.160738742351532,
"step": 3000
},
{
"epoch": 0.30517517790268106,
"grad_norm": 4.3694257736206055,
"learning_rate": 1.6287928928180544e-05,
"loss": 0.1773722290992737,
"step": 3010
},
{
"epoch": 0.30618904892561355,
"grad_norm": 1.8558675050735474,
"learning_rate": 1.6262363964226943e-05,
"loss": 0.14960556030273436,
"step": 3020
},
{
"epoch": 0.30720291994854604,
"grad_norm": 3.946931838989258,
"learning_rate": 1.623673148604211e-05,
"loss": 0.15237779617309571,
"step": 3030
},
{
"epoch": 0.3082167909714785,
"grad_norm": 2.6963064670562744,
"learning_rate": 1.6211031769968503e-05,
"loss": 0.1592946767807007,
"step": 3040
},
{
"epoch": 0.309230661994411,
"grad_norm": 3.9974443912506104,
"learning_rate": 1.6185265093073468e-05,
"loss": 0.15044374465942384,
"step": 3050
},
{
"epoch": 0.3102445330173435,
"grad_norm": 2.7180349826812744,
"learning_rate": 1.6159431733146255e-05,
"loss": 0.16534498929977418,
"step": 3060
},
{
"epoch": 0.31125840404027605,
"grad_norm": 2.1132118701934814,
"learning_rate": 1.613353196869502e-05,
"loss": 0.17043709754943848,
"step": 3070
},
{
"epoch": 0.31227227506320854,
"grad_norm": 2.8977210521698,
"learning_rate": 1.6107566078943818e-05,
"loss": 0.16933920383453369,
"step": 3080
},
{
"epoch": 0.313286146086141,
"grad_norm": 2.481511354446411,
"learning_rate": 1.608153434382961e-05,
"loss": 0.16899768114089966,
"step": 3090
},
{
"epoch": 0.3143000171090735,
"grad_norm": 1.3814888000488281,
"learning_rate": 1.6055437043999215e-05,
"loss": 0.1491153597831726,
"step": 3100
},
{
"epoch": 0.315313888132006,
"grad_norm": 2.6816048622131348,
"learning_rate": 1.6029274460806318e-05,
"loss": 0.16550488471984864,
"step": 3110
},
{
"epoch": 0.3163277591549385,
"grad_norm": 2.7753071784973145,
"learning_rate": 1.6003046876308413e-05,
"loss": 0.16282572746276855,
"step": 3120
},
{
"epoch": 0.317341630177871,
"grad_norm": 2.936276912689209,
"learning_rate": 1.5976754573263767e-05,
"loss": 0.13820860385894776,
"step": 3130
},
{
"epoch": 0.31835550120080347,
"grad_norm": 2.173248529434204,
"learning_rate": 1.5950397835128378e-05,
"loss": 0.16149394512176513,
"step": 3140
},
{
"epoch": 0.319369372223736,
"grad_norm": 2.2020862102508545,
"learning_rate": 1.5923976946052918e-05,
"loss": 0.14146852493286133,
"step": 3150
},
{
"epoch": 0.3203832432466685,
"grad_norm": 3.7949137687683105,
"learning_rate": 1.5897492190879667e-05,
"loss": 0.1732304334640503,
"step": 3160
},
{
"epoch": 0.321397114269601,
"grad_norm": 2.2579259872436523,
"learning_rate": 1.5870943855139437e-05,
"loss": 0.17511886358261108,
"step": 3170
},
{
"epoch": 0.3224109852925335,
"grad_norm": 2.280219078063965,
"learning_rate": 1.5844332225048496e-05,
"loss": 0.1429037094116211,
"step": 3180
},
{
"epoch": 0.32342485631546597,
"grad_norm": 3.9050657749176025,
"learning_rate": 1.5817657587505503e-05,
"loss": 0.16374385356903076,
"step": 3190
},
{
"epoch": 0.32443872733839846,
"grad_norm": 3.321664571762085,
"learning_rate": 1.5790920230088375e-05,
"loss": 0.16279804706573486,
"step": 3200
},
{
"epoch": 0.32545259836133095,
"grad_norm": 5.999819755554199,
"learning_rate": 1.576412044105122e-05,
"loss": 0.1595212697982788,
"step": 3210
},
{
"epoch": 0.32646646938426344,
"grad_norm": 4.3053154945373535,
"learning_rate": 1.5737258509321224e-05,
"loss": 0.15952061414718627,
"step": 3220
},
{
"epoch": 0.3274803404071959,
"grad_norm": 5.148134231567383,
"learning_rate": 1.5710334724495514e-05,
"loss": 0.16083049774169922,
"step": 3230
},
{
"epoch": 0.32849421143012847,
"grad_norm": 1.6683813333511353,
"learning_rate": 1.5683349376838067e-05,
"loss": 0.16355595588684083,
"step": 3240
},
{
"epoch": 0.32950808245306096,
"grad_norm": 2.354039192199707,
"learning_rate": 1.565630275727656e-05,
"loss": 0.15580629110336303,
"step": 3250
},
{
"epoch": 0.33052195347599345,
"grad_norm": 2.911437511444092,
"learning_rate": 1.5629195157399236e-05,
"loss": 0.17624651193618773,
"step": 3260
},
{
"epoch": 0.33153582449892594,
"grad_norm": 3.621835231781006,
"learning_rate": 1.5602026869451773e-05,
"loss": 0.16255561113357545,
"step": 3270
},
{
"epoch": 0.3325496955218584,
"grad_norm": 2.216557741165161,
"learning_rate": 1.5574798186334112e-05,
"loss": 0.14939769506454467,
"step": 3280
},
{
"epoch": 0.3335635665447909,
"grad_norm": 3.1235218048095703,
"learning_rate": 1.5547509401597322e-05,
"loss": 0.16122963428497314,
"step": 3290
},
{
"epoch": 0.3345774375677234,
"grad_norm": 1.705169916152954,
"learning_rate": 1.552016080944042e-05,
"loss": 0.15093945264816283,
"step": 3300
},
{
"epoch": 0.3355913085906559,
"grad_norm": 2.4070117473602295,
"learning_rate": 1.5492752704707198e-05,
"loss": 0.16157331466674804,
"step": 3310
},
{
"epoch": 0.3366051796135884,
"grad_norm": 4.289289474487305,
"learning_rate": 1.5465285382883065e-05,
"loss": 0.16152210235595704,
"step": 3320
},
{
"epoch": 0.3376190506365209,
"grad_norm": 5.066343784332275,
"learning_rate": 1.5437759140091827e-05,
"loss": 0.17490711212158203,
"step": 3330
},
{
"epoch": 0.3386329216594534,
"grad_norm": 2.9590840339660645,
"learning_rate": 1.541017427309254e-05,
"loss": 0.16888394355773925,
"step": 3340
},
{
"epoch": 0.3396467926823859,
"grad_norm": 3.971317768096924,
"learning_rate": 1.538253107927625e-05,
"loss": 0.1551954746246338,
"step": 3350
},
{
"epoch": 0.3406606637053184,
"grad_norm": 3.507760524749756,
"learning_rate": 1.535482985666285e-05,
"loss": 0.1505804181098938,
"step": 3360
},
{
"epoch": 0.3416745347282509,
"grad_norm": 5.321365833282471,
"learning_rate": 1.5327070903897832e-05,
"loss": 0.17013548612594603,
"step": 3370
},
{
"epoch": 0.34268840575118337,
"grad_norm": 6.4275431632995605,
"learning_rate": 1.529925452024907e-05,
"loss": 0.16357724666595458,
"step": 3380
},
{
"epoch": 0.34370227677411586,
"grad_norm": 3.4442403316497803,
"learning_rate": 1.5271381005603593e-05,
"loss": 0.16363760232925414,
"step": 3390
},
{
"epoch": 0.34471614779704834,
"grad_norm": 5.135153770446777,
"learning_rate": 1.5243450660464368e-05,
"loss": 0.14957830905914307,
"step": 3400
},
{
"epoch": 0.3457300188199809,
"grad_norm": 2.7656168937683105,
"learning_rate": 1.5215463785947036e-05,
"loss": 0.16679515838623046,
"step": 3410
},
{
"epoch": 0.3467438898429134,
"grad_norm": 2.6100821495056152,
"learning_rate": 1.518742068377669e-05,
"loss": 0.1571561098098755,
"step": 3420
},
{
"epoch": 0.34775776086584587,
"grad_norm": 1.5830525159835815,
"learning_rate": 1.5159321656284602e-05,
"loss": 0.1536739468574524,
"step": 3430
},
{
"epoch": 0.34877163188877836,
"grad_norm": 2.8285255432128906,
"learning_rate": 1.5131167006404978e-05,
"loss": 0.17519924640655518,
"step": 3440
},
{
"epoch": 0.34978550291171084,
"grad_norm": 5.235306262969971,
"learning_rate": 1.5102957037671689e-05,
"loss": 0.14956181049346923,
"step": 3450
},
{
"epoch": 0.35079937393464333,
"grad_norm": 3.1237504482269287,
"learning_rate": 1.5074692054214977e-05,
"loss": 0.1577211618423462,
"step": 3460
},
{
"epoch": 0.3518132449575758,
"grad_norm": 2.362372398376465,
"learning_rate": 1.504637236075822e-05,
"loss": 0.14937528371810913,
"step": 3470
},
{
"epoch": 0.3528271159805083,
"grad_norm": 3.1074187755584717,
"learning_rate": 1.5017998262614609e-05,
"loss": 0.16449918746948242,
"step": 3480
},
{
"epoch": 0.3538409870034408,
"grad_norm": 1.3769546747207642,
"learning_rate": 1.498957006568386e-05,
"loss": 0.15591460466384888,
"step": 3490
},
{
"epoch": 0.35485485802637334,
"grad_norm": 6.061954021453857,
"learning_rate": 1.4961088076448947e-05,
"loss": 0.15338077545166015,
"step": 3500
},
{
"epoch": 0.35586872904930583,
"grad_norm": 2.387238025665283,
"learning_rate": 1.4932552601972763e-05,
"loss": 0.16947635412216186,
"step": 3510
},
{
"epoch": 0.3568826000722383,
"grad_norm": 4.8353142738342285,
"learning_rate": 1.490396394989482e-05,
"loss": 0.1691033124923706,
"step": 3520
},
{
"epoch": 0.3578964710951708,
"grad_norm": 3.576732873916626,
"learning_rate": 1.4875322428427943e-05,
"loss": 0.1579538106918335,
"step": 3530
},
{
"epoch": 0.3589103421181033,
"grad_norm": 2.1625771522521973,
"learning_rate": 1.4846628346354934e-05,
"loss": 0.15312987565994263,
"step": 3540
},
{
"epoch": 0.3599242131410358,
"grad_norm": 4.348502159118652,
"learning_rate": 1.4817882013025253e-05,
"loss": 0.172758150100708,
"step": 3550
},
{
"epoch": 0.3609380841639683,
"grad_norm": 3.007671356201172,
"learning_rate": 1.478908373835167e-05,
"loss": 0.16237627267837523,
"step": 3560
},
{
"epoch": 0.36195195518690076,
"grad_norm": 2.955519676208496,
"learning_rate": 1.476023383280694e-05,
"loss": 0.16364006996154784,
"step": 3570
},
{
"epoch": 0.3629658262098333,
"grad_norm": 3.6302084922790527,
"learning_rate": 1.473133260742045e-05,
"loss": 0.15317835807800292,
"step": 3580
},
{
"epoch": 0.3639796972327658,
"grad_norm": 4.2745280265808105,
"learning_rate": 1.4702380373774852e-05,
"loss": 0.16490899324417113,
"step": 3590
},
{
"epoch": 0.3649935682556983,
"grad_norm": 3.547313928604126,
"learning_rate": 1.4673377444002724e-05,
"loss": 0.14667212963104248,
"step": 3600
},
{
"epoch": 0.3660074392786308,
"grad_norm": 4.827533721923828,
"learning_rate": 1.4644324130783197e-05,
"loss": 0.14877151250839232,
"step": 3610
},
{
"epoch": 0.36702131030156326,
"grad_norm": 2.5564522743225098,
"learning_rate": 1.4615220747338574e-05,
"loss": 0.16967167854309081,
"step": 3620
},
{
"epoch": 0.36803518132449575,
"grad_norm": 3.0046467781066895,
"learning_rate": 1.4586067607430976e-05,
"loss": 0.15763437747955322,
"step": 3630
},
{
"epoch": 0.36904905234742824,
"grad_norm": 4.62322998046875,
"learning_rate": 1.4556865025358942e-05,
"loss": 0.16663315296173095,
"step": 3640
},
{
"epoch": 0.37006292337036073,
"grad_norm": 3.036996603012085,
"learning_rate": 1.4527613315954035e-05,
"loss": 0.1531865954399109,
"step": 3650
},
{
"epoch": 0.3710767943932932,
"grad_norm": 2.701925039291382,
"learning_rate": 1.4498312794577465e-05,
"loss": 0.15420522689819335,
"step": 3660
},
{
"epoch": 0.37209066541622576,
"grad_norm": 2.0809624195098877,
"learning_rate": 1.4468963777116686e-05,
"loss": 0.18379952907562255,
"step": 3670
},
{
"epoch": 0.37310453643915825,
"grad_norm": 4.319645881652832,
"learning_rate": 1.4439566579981979e-05,
"loss": 0.16864385604858398,
"step": 3680
},
{
"epoch": 0.37411840746209074,
"grad_norm": 2.5941433906555176,
"learning_rate": 1.4410121520103045e-05,
"loss": 0.15414513349533082,
"step": 3690
},
{
"epoch": 0.37513227848502323,
"grad_norm": 2.1557583808898926,
"learning_rate": 1.4380628914925597e-05,
"loss": 0.14596984386444092,
"step": 3700
},
{
"epoch": 0.3761461495079557,
"grad_norm": 3.4211366176605225,
"learning_rate": 1.4351089082407929e-05,
"loss": 0.16002888679504396,
"step": 3710
},
{
"epoch": 0.3771600205308882,
"grad_norm": 2.767240524291992,
"learning_rate": 1.4321502341017496e-05,
"loss": 0.16459736824035645,
"step": 3720
},
{
"epoch": 0.3781738915538207,
"grad_norm": 3.7257165908813477,
"learning_rate": 1.4291869009727463e-05,
"loss": 0.15439112186431886,
"step": 3730
},
{
"epoch": 0.3791877625767532,
"grad_norm": 2.944345474243164,
"learning_rate": 1.4262189408013294e-05,
"loss": 0.15097581148147582,
"step": 3740
},
{
"epoch": 0.3802016335996857,
"grad_norm": 4.861728191375732,
"learning_rate": 1.4232463855849275e-05,
"loss": 0.1664339303970337,
"step": 3750
},
{
"epoch": 0.3812155046226182,
"grad_norm": 1.7162894010543823,
"learning_rate": 1.4202692673705097e-05,
"loss": 0.1732870817184448,
"step": 3760
},
{
"epoch": 0.3822293756455507,
"grad_norm": 2.547882318496704,
"learning_rate": 1.4172876182542372e-05,
"loss": 0.15242340564727783,
"step": 3770
},
{
"epoch": 0.3832432466684832,
"grad_norm": 2.9089040756225586,
"learning_rate": 1.41430147038112e-05,
"loss": 0.14958879947662354,
"step": 3780
},
{
"epoch": 0.3842571176914157,
"grad_norm": 2.8593204021453857,
"learning_rate": 1.411310855944668e-05,
"loss": 0.1654530167579651,
"step": 3790
},
{
"epoch": 0.3852709887143482,
"grad_norm": 3.1222939491271973,
"learning_rate": 1.4083158071865451e-05,
"loss": 0.1672975778579712,
"step": 3800
},
{
"epoch": 0.38628485973728066,
"grad_norm": 2.3869423866271973,
"learning_rate": 1.4053163563962215e-05,
"loss": 0.15945762395858765,
"step": 3810
},
{
"epoch": 0.38729873076021315,
"grad_norm": 3.978175163269043,
"learning_rate": 1.4023125359106253e-05,
"loss": 0.15605721473693848,
"step": 3820
},
{
"epoch": 0.38831260178314564,
"grad_norm": 3.0087058544158936,
"learning_rate": 1.3993043781137945e-05,
"loss": 0.16197741031646729,
"step": 3830
},
{
"epoch": 0.3893264728060782,
"grad_norm": 2.251584053039551,
"learning_rate": 1.3962919154365273e-05,
"loss": 0.1675764799118042,
"step": 3840
},
{
"epoch": 0.39034034382901067,
"grad_norm": 2.9911038875579834,
"learning_rate": 1.393275180356032e-05,
"loss": 0.1670203447341919,
"step": 3850
},
{
"epoch": 0.39135421485194316,
"grad_norm": 4.101426601409912,
"learning_rate": 1.3902542053955781e-05,
"loss": 0.16584452390670776,
"step": 3860
},
{
"epoch": 0.39236808587487565,
"grad_norm": 2.6066462993621826,
"learning_rate": 1.387229023124145e-05,
"loss": 0.1601638078689575,
"step": 3870
},
{
"epoch": 0.39338195689780814,
"grad_norm": 2.611488103866577,
"learning_rate": 1.3841996661560706e-05,
"loss": 0.15126419067382812,
"step": 3880
},
{
"epoch": 0.3943958279207406,
"grad_norm": 2.9255313873291016,
"learning_rate": 1.3811661671507004e-05,
"loss": 0.14636518955230712,
"step": 3890
},
{
"epoch": 0.3954096989436731,
"grad_norm": 5.417571067810059,
"learning_rate": 1.3781285588120343e-05,
"loss": 0.15263845920562744,
"step": 3900
},
{
"epoch": 0.3964235699666056,
"grad_norm": 1.8141567707061768,
"learning_rate": 1.3750868738883755e-05,
"loss": 0.15445313453674317,
"step": 3910
},
{
"epoch": 0.3974374409895381,
"grad_norm": 3.246340274810791,
"learning_rate": 1.3720411451719762e-05,
"loss": 0.16212179660797119,
"step": 3920
},
{
"epoch": 0.39845131201247064,
"grad_norm": 2.589411973953247,
"learning_rate": 1.3689914054986849e-05,
"loss": 0.15581687688827514,
"step": 3930
},
{
"epoch": 0.3994651830354031,
"grad_norm": 3.060903549194336,
"learning_rate": 1.3659376877475916e-05,
"loss": 0.15856468677520752,
"step": 3940
},
{
"epoch": 0.4004790540583356,
"grad_norm": 6.090756416320801,
"learning_rate": 1.3628800248406738e-05,
"loss": 0.15875492095947266,
"step": 3950
},
{
"epoch": 0.4014929250812681,
"grad_norm": 3.7103734016418457,
"learning_rate": 1.3598184497424416e-05,
"loss": 0.15350689888000488,
"step": 3960
},
{
"epoch": 0.4025067961042006,
"grad_norm": 3.7416274547576904,
"learning_rate": 1.3567529954595826e-05,
"loss": 0.1611272931098938,
"step": 3970
},
{
"epoch": 0.4035206671271331,
"grad_norm": 2.947030782699585,
"learning_rate": 1.3536836950406057e-05,
"loss": 0.15202665328979492,
"step": 3980
},
{
"epoch": 0.40453453815006557,
"grad_norm": 3.031306743621826,
"learning_rate": 1.3506105815754836e-05,
"loss": 0.14707481861114502,
"step": 3990
},
{
"epoch": 0.40554840917299806,
"grad_norm": 3.9497714042663574,
"learning_rate": 1.3475336881952988e-05,
"loss": 0.1840651035308838,
"step": 4000
},
{
"epoch": 0.40656228019593055,
"grad_norm": 1.471725583076477,
"learning_rate": 1.3444530480718838e-05,
"loss": 0.16138653755187987,
"step": 4010
},
{
"epoch": 0.4075761512188631,
"grad_norm": 2.276521682739258,
"learning_rate": 1.3413686944174655e-05,
"loss": 0.14237165451049805,
"step": 4020
},
{
"epoch": 0.4085900222417956,
"grad_norm": 3.100368022918701,
"learning_rate": 1.3382806604843053e-05,
"loss": 0.1655198812484741,
"step": 4030
},
{
"epoch": 0.40960389326472807,
"grad_norm": 2.40413498878479,
"learning_rate": 1.3351889795643423e-05,
"loss": 0.14299442768096923,
"step": 4040
},
{
"epoch": 0.41061776428766056,
"grad_norm": 4.591117858886719,
"learning_rate": 1.3320936849888322e-05,
"loss": 0.1557908296585083,
"step": 4050
},
{
"epoch": 0.41163163531059305,
"grad_norm": 1.8203728199005127,
"learning_rate": 1.3289948101279912e-05,
"loss": 0.16674418449401857,
"step": 4060
},
{
"epoch": 0.41264550633352554,
"grad_norm": 5.783075332641602,
"learning_rate": 1.3258923883906337e-05,
"loss": 0.1582515001296997,
"step": 4070
},
{
"epoch": 0.413659377356458,
"grad_norm": 2.6356709003448486,
"learning_rate": 1.3227864532238113e-05,
"loss": 0.16452465057373047,
"step": 4080
},
{
"epoch": 0.4146732483793905,
"grad_norm": 1.3169995546340942,
"learning_rate": 1.3196770381124557e-05,
"loss": 0.1535449743270874,
"step": 4090
},
{
"epoch": 0.41568711940232306,
"grad_norm": 2.9951820373535156,
"learning_rate": 1.3165641765790153e-05,
"loss": 0.13714048862457276,
"step": 4100
},
{
"epoch": 0.41670099042525555,
"grad_norm": 1.4830864667892456,
"learning_rate": 1.3134479021830938e-05,
"loss": 0.16309331655502318,
"step": 4110
},
{
"epoch": 0.41771486144818804,
"grad_norm": 4.9110941886901855,
"learning_rate": 1.3103282485210882e-05,
"loss": 0.15220718383789061,
"step": 4120
},
{
"epoch": 0.4187287324711205,
"grad_norm": 1.9279332160949707,
"learning_rate": 1.3072052492258286e-05,
"loss": 0.14481024742126464,
"step": 4130
},
{
"epoch": 0.419742603494053,
"grad_norm": 3.0089051723480225,
"learning_rate": 1.3040789379662131e-05,
"loss": 0.13670690059661866,
"step": 4140
},
{
"epoch": 0.4207564745169855,
"grad_norm": 2.1598100662231445,
"learning_rate": 1.3009493484468464e-05,
"loss": 0.1507465124130249,
"step": 4150
},
{
"epoch": 0.421770345539918,
"grad_norm": 1.4499905109405518,
"learning_rate": 1.2978165144076757e-05,
"loss": 0.15479912757873535,
"step": 4160
},
{
"epoch": 0.4227842165628505,
"grad_norm": 3.2272274494171143,
"learning_rate": 1.2946804696236277e-05,
"loss": 0.1462334156036377,
"step": 4170
},
{
"epoch": 0.42379808758578297,
"grad_norm": 3.0212464332580566,
"learning_rate": 1.2915412479042436e-05,
"loss": 0.1434977889060974,
"step": 4180
},
{
"epoch": 0.4248119586087155,
"grad_norm": 1.569291591644287,
"learning_rate": 1.2883988830933148e-05,
"loss": 0.15275589227676392,
"step": 4190
},
{
"epoch": 0.425825829631648,
"grad_norm": 2.6395115852355957,
"learning_rate": 1.2852534090685184e-05,
"loss": 0.1492305278778076,
"step": 4200
},
{
"epoch": 0.4268397006545805,
"grad_norm": 2.2366533279418945,
"learning_rate": 1.282104859741052e-05,
"loss": 0.16312043666839598,
"step": 4210
},
{
"epoch": 0.427853571677513,
"grad_norm": 1.252985954284668,
"learning_rate": 1.2789532690552673e-05,
"loss": 0.15947911739349366,
"step": 4220
},
{
"epoch": 0.42886744270044547,
"grad_norm": 7.337235450744629,
"learning_rate": 1.2757986709883059e-05,
"loss": 0.15317082405090332,
"step": 4230
},
{
"epoch": 0.42988131372337796,
"grad_norm": 3.649880886077881,
"learning_rate": 1.2726410995497304e-05,
"loss": 0.1558199405670166,
"step": 4240
},
{
"epoch": 0.43089518474631044,
"grad_norm": 1.8817944526672363,
"learning_rate": 1.2694805887811601e-05,
"loss": 0.12999532222747803,
"step": 4250
},
{
"epoch": 0.43190905576924293,
"grad_norm": 2.1384754180908203,
"learning_rate": 1.2663171727559025e-05,
"loss": 0.1560016393661499,
"step": 4260
},
{
"epoch": 0.4329229267921755,
"grad_norm": 2.1979405879974365,
"learning_rate": 1.2631508855785877e-05,
"loss": 0.1494905471801758,
"step": 4270
},
{
"epoch": 0.43393679781510797,
"grad_norm": 3.199312448501587,
"learning_rate": 1.2599817613847973e-05,
"loss": 0.16957972049713135,
"step": 4280
},
{
"epoch": 0.43495066883804046,
"grad_norm": 2.563324451446533,
"learning_rate": 1.2568098343407007e-05,
"loss": 0.1424316167831421,
"step": 4290
},
{
"epoch": 0.43596453986097294,
"grad_norm": 3.91619873046875,
"learning_rate": 1.2536351386426844e-05,
"loss": 0.15048011541366577,
"step": 4300
},
{
"epoch": 0.43697841088390543,
"grad_norm": 2.458712100982666,
"learning_rate": 1.2504577085169826e-05,
"loss": 0.16417486667633058,
"step": 4310
},
{
"epoch": 0.4379922819068379,
"grad_norm": 2.7342424392700195,
"learning_rate": 1.2472775782193099e-05,
"loss": 0.15604608058929442,
"step": 4320
},
{
"epoch": 0.4390061529297704,
"grad_norm": 3.289560079574585,
"learning_rate": 1.2440947820344916e-05,
"loss": 0.1540846824645996,
"step": 4330
},
{
"epoch": 0.4400200239527029,
"grad_norm": 2.5234906673431396,
"learning_rate": 1.2409093542760925e-05,
"loss": 0.148193097114563,
"step": 4340
},
{
"epoch": 0.4410338949756354,
"grad_norm": 4.650540828704834,
"learning_rate": 1.2377213292860506e-05,
"loss": 0.17957816123962403,
"step": 4350
},
{
"epoch": 0.44204776599856793,
"grad_norm": 5.297595500946045,
"learning_rate": 1.2345307414343021e-05,
"loss": 0.15656602382659912,
"step": 4360
},
{
"epoch": 0.4430616370215004,
"grad_norm": 3.004148483276367,
"learning_rate": 1.2313376251184148e-05,
"loss": 0.15197930335998536,
"step": 4370
},
{
"epoch": 0.4440755080444329,
"grad_norm": 1.8785146474838257,
"learning_rate": 1.2281420147632148e-05,
"loss": 0.14419232606887816,
"step": 4380
},
{
"epoch": 0.4450893790673654,
"grad_norm": 3.805147886276245,
"learning_rate": 1.2249439448204172e-05,
"loss": 0.13514206409454346,
"step": 4390
},
{
"epoch": 0.4461032500902979,
"grad_norm": 2.720613956451416,
"learning_rate": 1.2217434497682537e-05,
"loss": 0.15017153024673463,
"step": 4400
},
{
"epoch": 0.4471171211132304,
"grad_norm": 4.880934238433838,
"learning_rate": 1.2185405641110997e-05,
"loss": 0.16089274883270263,
"step": 4410
},
{
"epoch": 0.44813099213616286,
"grad_norm": 3.6104581356048584,
"learning_rate": 1.2153353223791049e-05,
"loss": 0.15338767766952516,
"step": 4420
},
{
"epoch": 0.44914486315909535,
"grad_norm": 5.1818084716796875,
"learning_rate": 1.2121277591278197e-05,
"loss": 0.16855554580688475,
"step": 4430
},
{
"epoch": 0.45015873418202784,
"grad_norm": 4.87166690826416,
"learning_rate": 1.2089179089378222e-05,
"loss": 0.14823087453842163,
"step": 4440
},
{
"epoch": 0.4511726052049604,
"grad_norm": 3.8663687705993652,
"learning_rate": 1.2057058064143456e-05,
"loss": 0.1476778507232666,
"step": 4450
},
{
"epoch": 0.4521864762278929,
"grad_norm": 7.155592441558838,
"learning_rate": 1.2024914861869064e-05,
"loss": 0.15448462963104248,
"step": 4460
},
{
"epoch": 0.45320034725082536,
"grad_norm": 1.9186924695968628,
"learning_rate": 1.199274982908929e-05,
"loss": 0.16103228330612182,
"step": 4470
},
{
"epoch": 0.45421421827375785,
"grad_norm": 2.420443534851074,
"learning_rate": 1.196056331257374e-05,
"loss": 0.1477586269378662,
"step": 4480
},
{
"epoch": 0.45522808929669034,
"grad_norm": 2.959292411804199,
"learning_rate": 1.1928355659323627e-05,
"loss": 0.14802198410034179,
"step": 4490
},
{
"epoch": 0.45624196031962283,
"grad_norm": 2.303163528442383,
"learning_rate": 1.1896127216568047e-05,
"loss": 0.1706514835357666,
"step": 4500
},
{
"epoch": 0.4572558313425553,
"grad_norm": 2.5037248134613037,
"learning_rate": 1.1863878331760217e-05,
"loss": 0.15173400640487672,
"step": 4510
},
{
"epoch": 0.4582697023654878,
"grad_norm": 2.160269021987915,
"learning_rate": 1.183160935257375e-05,
"loss": 0.1506716012954712,
"step": 4520
},
{
"epoch": 0.45928357338842035,
"grad_norm": 2.8853442668914795,
"learning_rate": 1.1799320626898883e-05,
"loss": 0.1413341522216797,
"step": 4530
},
{
"epoch": 0.46029744441135284,
"grad_norm": 2.7343077659606934,
"learning_rate": 1.1767012502838746e-05,
"loss": 0.14676620960235595,
"step": 4540
},
{
"epoch": 0.46131131543428533,
"grad_norm": 5.13657283782959,
"learning_rate": 1.1734685328705604e-05,
"loss": 0.15392801761627198,
"step": 4550
},
{
"epoch": 0.4623251864572178,
"grad_norm": 2.3126180171966553,
"learning_rate": 1.17023394530171e-05,
"loss": 0.16811704635620117,
"step": 4560
},
{
"epoch": 0.4633390574801503,
"grad_norm": 3.027582883834839,
"learning_rate": 1.1669975224492486e-05,
"loss": 0.15279700756072997,
"step": 4570
},
{
"epoch": 0.4643529285030828,
"grad_norm": 4.189201354980469,
"learning_rate": 1.1637592992048888e-05,
"loss": 0.13826762437820433,
"step": 4580
},
{
"epoch": 0.4653667995260153,
"grad_norm": 4.2519145011901855,
"learning_rate": 1.160519310479753e-05,
"loss": 0.1671229600906372,
"step": 4590
},
{
"epoch": 0.4663806705489478,
"grad_norm": 1.856067180633545,
"learning_rate": 1.157277591203996e-05,
"loss": 0.15294448137283326,
"step": 4600
},
{
"epoch": 0.46739454157188026,
"grad_norm": 3.9634714126586914,
"learning_rate": 1.1540341763264314e-05,
"loss": 0.14562082290649414,
"step": 4610
},
{
"epoch": 0.4684084125948128,
"grad_norm": 2.4699907302856445,
"learning_rate": 1.1507891008141512e-05,
"loss": 0.14978381395339965,
"step": 4620
},
{
"epoch": 0.4694222836177453,
"grad_norm": 1.3213014602661133,
"learning_rate": 1.147542399652152e-05,
"loss": 0.16067898273468018,
"step": 4630
},
{
"epoch": 0.4704361546406778,
"grad_norm": 1.558398723602295,
"learning_rate": 1.1442941078429554e-05,
"loss": 0.13742054700851442,
"step": 4640
},
{
"epoch": 0.4714500256636103,
"grad_norm": 2.88055419921875,
"learning_rate": 1.141044260406233e-05,
"loss": 0.14868855476379395,
"step": 4650
},
{
"epoch": 0.47246389668654276,
"grad_norm": 3.9311447143554688,
"learning_rate": 1.1377928923784261e-05,
"loss": 0.14402248859405517,
"step": 4660
},
{
"epoch": 0.47347776770947525,
"grad_norm": 3.0685408115386963,
"learning_rate": 1.1345400388123698e-05,
"loss": 0.14492884874343873,
"step": 4670
},
{
"epoch": 0.47449163873240774,
"grad_norm": 2.5264904499053955,
"learning_rate": 1.1312857347769153e-05,
"loss": 0.14493656158447266,
"step": 4680
},
{
"epoch": 0.4755055097553402,
"grad_norm": 3.5055086612701416,
"learning_rate": 1.128030015356551e-05,
"loss": 0.15279898643493653,
"step": 4690
},
{
"epoch": 0.4765193807782727,
"grad_norm": 5.654348373413086,
"learning_rate": 1.1247729156510245e-05,
"loss": 0.15157454013824462,
"step": 4700
},
{
"epoch": 0.47753325180120526,
"grad_norm": 2.1037402153015137,
"learning_rate": 1.1215144707749635e-05,
"loss": 0.1466853141784668,
"step": 4710
},
{
"epoch": 0.47854712282413775,
"grad_norm": 3.1714541912078857,
"learning_rate": 1.1182547158574988e-05,
"loss": 0.13457432985305787,
"step": 4720
},
{
"epoch": 0.47956099384707024,
"grad_norm": 2.5036449432373047,
"learning_rate": 1.1149936860418846e-05,
"loss": 0.16383061408996583,
"step": 4730
},
{
"epoch": 0.4805748648700027,
"grad_norm": 1.225236415863037,
"learning_rate": 1.1117314164851191e-05,
"loss": 0.15031228065490723,
"step": 4740
},
{
"epoch": 0.4815887358929352,
"grad_norm": 4.2901105880737305,
"learning_rate": 1.1084679423575666e-05,
"loss": 0.1442958116531372,
"step": 4750
},
{
"epoch": 0.4826026069158677,
"grad_norm": 2.919139862060547,
"learning_rate": 1.1052032988425778e-05,
"loss": 0.12567509412765504,
"step": 4760
},
{
"epoch": 0.4836164779388002,
"grad_norm": 5.952666759490967,
"learning_rate": 1.1019375211361104e-05,
"loss": 0.13490984439849854,
"step": 4770
},
{
"epoch": 0.4846303489617327,
"grad_norm": 1.9709986448287964,
"learning_rate": 1.0986706444463496e-05,
"loss": 0.15800846815109254,
"step": 4780
},
{
"epoch": 0.4856442199846652,
"grad_norm": 3.495464324951172,
"learning_rate": 1.0954027039933288e-05,
"loss": 0.13826513290405273,
"step": 4790
},
{
"epoch": 0.4866580910075977,
"grad_norm": 5.193592548370361,
"learning_rate": 1.09213373500855e-05,
"loss": 0.15181844234466552,
"step": 4800
},
{
"epoch": 0.4876719620305302,
"grad_norm": 1.5989762544631958,
"learning_rate": 1.0888637727346032e-05,
"loss": 0.15787261724472046,
"step": 4810
},
{
"epoch": 0.4886858330534627,
"grad_norm": 4.218422889709473,
"learning_rate": 1.0855928524247884e-05,
"loss": 0.14581600427627564,
"step": 4820
},
{
"epoch": 0.4896997040763952,
"grad_norm": 2.8877553939819336,
"learning_rate": 1.0823210093427318e-05,
"loss": 0.13564257621765136,
"step": 4830
},
{
"epoch": 0.49071357509932767,
"grad_norm": 2.362994432449341,
"learning_rate": 1.07904827876201e-05,
"loss": 0.15038161277770995,
"step": 4840
},
{
"epoch": 0.49172744612226016,
"grad_norm": 3.39477801322937,
"learning_rate": 1.0757746959657666e-05,
"loss": 0.14491000175476074,
"step": 4850
},
{
"epoch": 0.49274131714519265,
"grad_norm": 4.99651575088501,
"learning_rate": 1.072500296246334e-05,
"loss": 0.1409554123878479,
"step": 4860
},
{
"epoch": 0.49375518816812514,
"grad_norm": 4.166104793548584,
"learning_rate": 1.06922511490485e-05,
"loss": 0.13930387496948243,
"step": 4870
},
{
"epoch": 0.4947690591910577,
"grad_norm": 2.7220804691314697,
"learning_rate": 1.0659491872508807e-05,
"loss": 0.14525125026702881,
"step": 4880
},
{
"epoch": 0.49578293021399017,
"grad_norm": 6.024302959442139,
"learning_rate": 1.0626725486020374e-05,
"loss": 0.14442189931869506,
"step": 4890
},
{
"epoch": 0.49679680123692266,
"grad_norm": 3.2836461067199707,
"learning_rate": 1.059395234283597e-05,
"loss": 0.13796718120574952,
"step": 4900
},
{
"epoch": 0.49781067225985515,
"grad_norm": 3.531688928604126,
"learning_rate": 1.0561172796281196e-05,
"loss": 0.15348341464996337,
"step": 4910
},
{
"epoch": 0.49882454328278764,
"grad_norm": 2.809380531311035,
"learning_rate": 1.0528387199750706e-05,
"loss": 0.1487370252609253,
"step": 4920
},
{
"epoch": 0.4998384143057201,
"grad_norm": 3.301713228225708,
"learning_rate": 1.0495595906704365e-05,
"loss": 0.13770787715911864,
"step": 4930
},
{
"epoch": 0.5008522853286527,
"grad_norm": 2.5095901489257812,
"learning_rate": 1.0462799270663457e-05,
"loss": 0.1323918581008911,
"step": 4940
},
{
"epoch": 0.5018661563515852,
"grad_norm": 2.0731892585754395,
"learning_rate": 1.0429997645206865e-05,
"loss": 0.12777047157287597,
"step": 4950
},
{
"epoch": 0.5028800273745176,
"grad_norm": 2.480684518814087,
"learning_rate": 1.0397191383967267e-05,
"loss": 0.13495092391967772,
"step": 4960
},
{
"epoch": 0.5038938983974501,
"grad_norm": 3.8417718410491943,
"learning_rate": 1.0364380840627315e-05,
"loss": 0.14414069652557374,
"step": 4970
},
{
"epoch": 0.5049077694203826,
"grad_norm": 3.9812800884246826,
"learning_rate": 1.033156636891583e-05,
"loss": 0.15525320768356324,
"step": 4980
},
{
"epoch": 0.5059216404433151,
"grad_norm": 2.1667683124542236,
"learning_rate": 1.0298748322603982e-05,
"loss": 0.13929738998413085,
"step": 4990
},
{
"epoch": 0.5069355114662476,
"grad_norm": 3.1319570541381836,
"learning_rate": 1.0265927055501476e-05,
"loss": 0.14830927848815917,
"step": 5000
},
{
"epoch": 0.5079493824891801,
"grad_norm": 1.86725914478302,
"learning_rate": 1.023310292145275e-05,
"loss": 0.13299996852874757,
"step": 5010
},
{
"epoch": 0.5089632535121126,
"grad_norm": 5.76659631729126,
"learning_rate": 1.020027627433314e-05,
"loss": 0.14041578769683838,
"step": 5020
},
{
"epoch": 0.5099771245350451,
"grad_norm": 3.184755802154541,
"learning_rate": 1.0167447468045083e-05,
"loss": 0.13832666873931884,
"step": 5030
},
{
"epoch": 0.5109909955579776,
"grad_norm": 3.908865451812744,
"learning_rate": 1.0134616856514282e-05,
"loss": 0.15212610960006714,
"step": 5040
},
{
"epoch": 0.51200486658091,
"grad_norm": 3.225447654724121,
"learning_rate": 1.0101784793685914e-05,
"loss": 0.15245879888534547,
"step": 5050
},
{
"epoch": 0.5130187376038425,
"grad_norm": 4.439329147338867,
"learning_rate": 1.0068951633520802e-05,
"loss": 0.13295621871948243,
"step": 5060
},
{
"epoch": 0.514032608626775,
"grad_norm": 3.2054994106292725,
"learning_rate": 1.0036117729991595e-05,
"loss": 0.1339601159095764,
"step": 5070
},
{
"epoch": 0.5150464796497075,
"grad_norm": 4.5582780838012695,
"learning_rate": 1.0003283437078951e-05,
"loss": 0.14823698997497559,
"step": 5080
},
{
"epoch": 0.5160603506726401,
"grad_norm": 3.0242674350738525,
"learning_rate": 9.970449108767742e-06,
"loss": 0.1389592170715332,
"step": 5090
},
{
"epoch": 0.5170742216955726,
"grad_norm": 3.3237133026123047,
"learning_rate": 9.937615099043206e-06,
"loss": 0.14187226295471192,
"step": 5100
},
{
"epoch": 0.5180880927185051,
"grad_norm": 3.5458712577819824,
"learning_rate": 9.904781761887147e-06,
"loss": 0.147489333152771,
"step": 5110
},
{
"epoch": 0.5191019637414376,
"grad_norm": 2.6628382205963135,
"learning_rate": 9.871949451274137e-06,
"loss": 0.14312341213226318,
"step": 5120
},
{
"epoch": 0.5201158347643701,
"grad_norm": 1.9601985216140747,
"learning_rate": 9.839118521167656e-06,
"loss": 0.13303396701812745,
"step": 5130
},
{
"epoch": 0.5211297057873026,
"grad_norm": 3.96596622467041,
"learning_rate": 9.80628932551631e-06,
"loss": 0.14449589252471923,
"step": 5140
},
{
"epoch": 0.522143576810235,
"grad_norm": 3.226538896560669,
"learning_rate": 9.773462218250014e-06,
"loss": 0.13667025566101074,
"step": 5150
},
{
"epoch": 0.5231574478331675,
"grad_norm": 4.236742973327637,
"learning_rate": 9.740637553276162e-06,
"loss": 0.1524426221847534,
"step": 5160
},
{
"epoch": 0.5241713188561,
"grad_norm": 2.9533586502075195,
"learning_rate": 9.707815684475817e-06,
"loss": 0.1370355486869812,
"step": 5170
},
{
"epoch": 0.5251851898790325,
"grad_norm": 3.8765130043029785,
"learning_rate": 9.674996965699895e-06,
"loss": 0.13194715976715088,
"step": 5180
},
{
"epoch": 0.526199060901965,
"grad_norm": 3.651693344116211,
"learning_rate": 9.642181750765354e-06,
"loss": 0.14788544178009033,
"step": 5190
},
{
"epoch": 0.5272129319248975,
"grad_norm": 2.1815619468688965,
"learning_rate": 9.60937039345138e-06,
"loss": 0.15143814086914062,
"step": 5200
},
{
"epoch": 0.52822680294783,
"grad_norm": 3.5286946296691895,
"learning_rate": 9.576563247495566e-06,
"loss": 0.1322399854660034,
"step": 5210
},
{
"epoch": 0.5292406739707625,
"grad_norm": 2.597372531890869,
"learning_rate": 9.543760666590105e-06,
"loss": 0.14120746850967408,
"step": 5220
},
{
"epoch": 0.530254544993695,
"grad_norm": 2.749856948852539,
"learning_rate": 9.510963004377974e-06,
"loss": 0.13882296085357665,
"step": 5230
},
{
"epoch": 0.5312684160166274,
"grad_norm": 2.7055041790008545,
"learning_rate": 9.478170614449114e-06,
"loss": 0.1560375213623047,
"step": 5240
},
{
"epoch": 0.5322822870395599,
"grad_norm": 2.444040298461914,
"learning_rate": 9.445383850336648e-06,
"loss": 0.14090545177459718,
"step": 5250
},
{
"epoch": 0.5332961580624924,
"grad_norm": 4.34225606918335,
"learning_rate": 9.412603065513021e-06,
"loss": 0.1516825795173645,
"step": 5260
},
{
"epoch": 0.534310029085425,
"grad_norm": 3.3723366260528564,
"learning_rate": 9.379828613386225e-06,
"loss": 0.15602221488952636,
"step": 5270
},
{
"epoch": 0.5353239001083575,
"grad_norm": 2.109825372695923,
"learning_rate": 9.347060847295991e-06,
"loss": 0.1413689970970154,
"step": 5280
},
{
"epoch": 0.53633777113129,
"grad_norm": 2.397494316101074,
"learning_rate": 9.314300120509952e-06,
"loss": 0.1450955867767334,
"step": 5290
},
{
"epoch": 0.5373516421542225,
"grad_norm": 2.7441093921661377,
"learning_rate": 9.281546786219861e-06,
"loss": 0.13447961807250977,
"step": 5300
},
{
"epoch": 0.538365513177155,
"grad_norm": 2.711240291595459,
"learning_rate": 9.24880119753777e-06,
"loss": 0.13954005241394044,
"step": 5310
},
{
"epoch": 0.5393793842000875,
"grad_norm": 4.243502140045166,
"learning_rate": 9.216063707492218e-06,
"loss": 0.13719184398651124,
"step": 5320
},
{
"epoch": 0.54039325522302,
"grad_norm": 2.613917589187622,
"learning_rate": 9.183334669024449e-06,
"loss": 0.14381712675094604,
"step": 5330
},
{
"epoch": 0.5414071262459524,
"grad_norm": 3.7209067344665527,
"learning_rate": 9.150614434984578e-06,
"loss": 0.14506936073303223,
"step": 5340
},
{
"epoch": 0.5424209972688849,
"grad_norm": 4.315561294555664,
"learning_rate": 9.117903358127801e-06,
"loss": 0.1292236328125,
"step": 5350
},
{
"epoch": 0.5434348682918174,
"grad_norm": 2.5302693843841553,
"learning_rate": 9.085201791110602e-06,
"loss": 0.1260303497314453,
"step": 5360
},
{
"epoch": 0.5444487393147499,
"grad_norm": 3.4451229572296143,
"learning_rate": 9.052510086486922e-06,
"loss": 0.1305840015411377,
"step": 5370
},
{
"epoch": 0.5454626103376824,
"grad_norm": 1.997789740562439,
"learning_rate": 9.019828596704394e-06,
"loss": 0.14048842191696168,
"step": 5380
},
{
"epoch": 0.5464764813606149,
"grad_norm": 2.9708194732666016,
"learning_rate": 8.987157674100519e-06,
"loss": 0.13699736595153808,
"step": 5390
},
{
"epoch": 0.5474903523835474,
"grad_norm": 2.5471973419189453,
"learning_rate": 8.954497670898858e-06,
"loss": 0.13051559925079345,
"step": 5400
},
{
"epoch": 0.5485042234064799,
"grad_norm": 4.02651834487915,
"learning_rate": 8.92184893920528e-06,
"loss": 0.130690336227417,
"step": 5410
},
{
"epoch": 0.5495180944294124,
"grad_norm": 2.6350808143615723,
"learning_rate": 8.889211831004112e-06,
"loss": 0.1422417402267456,
"step": 5420
},
{
"epoch": 0.5505319654523448,
"grad_norm": 4.701023578643799,
"learning_rate": 8.856586698154382e-06,
"loss": 0.14483355283737182,
"step": 5430
},
{
"epoch": 0.5515458364752774,
"grad_norm": 3.377810478210449,
"learning_rate": 8.823973892386007e-06,
"loss": 0.15473124980926514,
"step": 5440
},
{
"epoch": 0.5525597074982099,
"grad_norm": 3.6960642337799072,
"learning_rate": 8.791373765296006e-06,
"loss": 0.13917756080627441,
"step": 5450
},
{
"epoch": 0.5535735785211424,
"grad_norm": 2.605804204940796,
"learning_rate": 8.758786668344711e-06,
"loss": 0.13410532474517822,
"step": 5460
},
{
"epoch": 0.5545874495440749,
"grad_norm": 2.731872081756592,
"learning_rate": 8.726212952851978e-06,
"loss": 0.1451941967010498,
"step": 5470
},
{
"epoch": 0.5556013205670074,
"grad_norm": 1.9896572828292847,
"learning_rate": 8.693652969993396e-06,
"loss": 0.13926795721054078,
"step": 5480
},
{
"epoch": 0.5566151915899399,
"grad_norm": 2.1392672061920166,
"learning_rate": 8.661107070796504e-06,
"loss": 0.13833966255187988,
"step": 5490
},
{
"epoch": 0.5576290626128724,
"grad_norm": 2.968776226043701,
"learning_rate": 8.628575606137005e-06,
"loss": 0.13946595191955566,
"step": 5500
},
{
"epoch": 0.5586429336358049,
"grad_norm": 4.038211345672607,
"learning_rate": 8.59605892673499e-06,
"loss": 0.15292141437530518,
"step": 5510
},
{
"epoch": 0.5596568046587374,
"grad_norm": 3.32049298286438,
"learning_rate": 8.563557383151143e-06,
"loss": 0.13712563514709472,
"step": 5520
},
{
"epoch": 0.5606706756816698,
"grad_norm": 3.1748876571655273,
"learning_rate": 8.531071325782965e-06,
"loss": 0.14831210374832154,
"step": 5530
},
{
"epoch": 0.5616845467046023,
"grad_norm": 2.158181667327881,
"learning_rate": 8.49860110486102e-06,
"loss": 0.12468826770782471,
"step": 5540
},
{
"epoch": 0.5626984177275348,
"grad_norm": 2.7808425426483154,
"learning_rate": 8.466147070445123e-06,
"loss": 0.13523318767547607,
"step": 5550
},
{
"epoch": 0.5637122887504673,
"grad_norm": 3.9292819499969482,
"learning_rate": 8.433709572420586e-06,
"loss": 0.1281266689300537,
"step": 5560
},
{
"epoch": 0.5647261597733998,
"grad_norm": 2.866576671600342,
"learning_rate": 8.40128896049445e-06,
"loss": 0.13159582614898682,
"step": 5570
},
{
"epoch": 0.5657400307963323,
"grad_norm": 2.305171489715576,
"learning_rate": 8.368885584191698e-06,
"loss": 0.12322475910186767,
"step": 5580
},
{
"epoch": 0.5667539018192648,
"grad_norm": 3.5697221755981445,
"learning_rate": 8.336499792851505e-06,
"loss": 0.13272398710250854,
"step": 5590
},
{
"epoch": 0.5677677728421973,
"grad_norm": 3.3337748050689697,
"learning_rate": 8.304131935623461e-06,
"loss": 0.13962501287460327,
"step": 5600
},
{
"epoch": 0.5687816438651299,
"grad_norm": 2.4586572647094727,
"learning_rate": 8.271782361463806e-06,
"loss": 0.13514838218688965,
"step": 5610
},
{
"epoch": 0.5697955148880623,
"grad_norm": 2.3035106658935547,
"learning_rate": 8.239451419131676e-06,
"loss": 0.1251867651939392,
"step": 5620
},
{
"epoch": 0.5708093859109948,
"grad_norm": 3.967775821685791,
"learning_rate": 8.207139457185328e-06,
"loss": 0.13274972438812255,
"step": 5630
},
{
"epoch": 0.5718232569339273,
"grad_norm": 5.767806053161621,
"learning_rate": 8.174846823978412e-06,
"loss": 0.13744549751281737,
"step": 5640
},
{
"epoch": 0.5728371279568598,
"grad_norm": 3.1191351413726807,
"learning_rate": 8.142573867656182e-06,
"loss": 0.12564929723739623,
"step": 5650
},
{
"epoch": 0.5738509989797923,
"grad_norm": 3.916274070739746,
"learning_rate": 8.110320936151746e-06,
"loss": 0.11837705373764038,
"step": 5660
},
{
"epoch": 0.5748648700027248,
"grad_norm": 6.145861625671387,
"learning_rate": 8.078088377182352e-06,
"loss": 0.10878785848617553,
"step": 5670
},
{
"epoch": 0.5758787410256573,
"grad_norm": 2.69242525100708,
"learning_rate": 8.045876538245595e-06,
"loss": 0.12540196180343627,
"step": 5680
},
{
"epoch": 0.5768926120485898,
"grad_norm": 2.389148473739624,
"learning_rate": 8.013685766615687e-06,
"loss": 0.14135003089904785,
"step": 5690
},
{
"epoch": 0.5779064830715223,
"grad_norm": 2.7626302242279053,
"learning_rate": 7.981516409339726e-06,
"loss": 0.143448805809021,
"step": 5700
},
{
"epoch": 0.5789203540944547,
"grad_norm": 2.7355239391326904,
"learning_rate": 7.94936881323393e-06,
"loss": 0.1356850266456604,
"step": 5710
},
{
"epoch": 0.5799342251173872,
"grad_norm": 4.117995738983154,
"learning_rate": 7.917243324879924e-06,
"loss": 0.14854304790496825,
"step": 5720
},
{
"epoch": 0.5809480961403197,
"grad_norm": 2.2346887588500977,
"learning_rate": 7.88514029062098e-06,
"loss": 0.12667829990386964,
"step": 5730
},
{
"epoch": 0.5819619671632522,
"grad_norm": 2.621720790863037,
"learning_rate": 7.853060056558294e-06,
"loss": 0.12541188001632692,
"step": 5740
},
{
"epoch": 0.5829758381861847,
"grad_norm": 2.870051145553589,
"learning_rate": 7.821002968547266e-06,
"loss": 0.14867613315582276,
"step": 5750
},
{
"epoch": 0.5839897092091172,
"grad_norm": 2.1742770671844482,
"learning_rate": 7.788969372193744e-06,
"loss": 0.13156120777130126,
"step": 5760
},
{
"epoch": 0.5850035802320497,
"grad_norm": 3.277129650115967,
"learning_rate": 7.75695961285032e-06,
"loss": 0.12340745925903321,
"step": 5770
},
{
"epoch": 0.5860174512549823,
"grad_norm": 2.4195051193237305,
"learning_rate": 7.724974035612611e-06,
"loss": 0.12762395143508912,
"step": 5780
},
{
"epoch": 0.5870313222779148,
"grad_norm": 2.737187147140503,
"learning_rate": 7.693012985315503e-06,
"loss": 0.14598875045776366,
"step": 5790
},
{
"epoch": 0.5880451933008473,
"grad_norm": 4.48429012298584,
"learning_rate": 7.661076806529486e-06,
"loss": 0.12410306930541992,
"step": 5800
},
{
"epoch": 0.5890590643237797,
"grad_norm": 2.7455196380615234,
"learning_rate": 7.629165843556896e-06,
"loss": 0.13966749906539916,
"step": 5810
},
{
"epoch": 0.5900729353467122,
"grad_norm": 2.5052459239959717,
"learning_rate": 7.597280440428216e-06,
"loss": 0.13433483839035035,
"step": 5820
},
{
"epoch": 0.5910868063696447,
"grad_norm": 2.603473424911499,
"learning_rate": 7.565420940898381e-06,
"loss": 0.12661619186401368,
"step": 5830
},
{
"epoch": 0.5921006773925772,
"grad_norm": 3.3487226963043213,
"learning_rate": 7.5335876884430495e-06,
"loss": 0.14274101257324218,
"step": 5840
},
{
"epoch": 0.5931145484155097,
"grad_norm": 2.813497304916382,
"learning_rate": 7.501781026254922e-06,
"loss": 0.13043177127838135,
"step": 5850
},
{
"epoch": 0.5941284194384422,
"grad_norm": 3.0368547439575195,
"learning_rate": 7.4700012972400224e-06,
"loss": 0.13536282777786254,
"step": 5860
},
{
"epoch": 0.5951422904613747,
"grad_norm": 2.3660271167755127,
"learning_rate": 7.438248844014011e-06,
"loss": 0.13313745260238646,
"step": 5870
},
{
"epoch": 0.5961561614843072,
"grad_norm": 3.4228696823120117,
"learning_rate": 7.4065240088984944e-06,
"loss": 0.12414616346359253,
"step": 5880
},
{
"epoch": 0.5971700325072397,
"grad_norm": 3.453580379486084,
"learning_rate": 7.3748271339173224e-06,
"loss": 0.1362109899520874,
"step": 5890
},
{
"epoch": 0.5981839035301721,
"grad_norm": 3.5125832557678223,
"learning_rate": 7.34315856079291e-06,
"loss": 0.14772225618362428,
"step": 5900
},
{
"epoch": 0.5991977745531046,
"grad_norm": 3.229559898376465,
"learning_rate": 7.311518630942562e-06,
"loss": 0.13679978847503663,
"step": 5910
},
{
"epoch": 0.6002116455760371,
"grad_norm": 2.663947820663452,
"learning_rate": 7.279907685474759e-06,
"loss": 0.13754661083221437,
"step": 5920
},
{
"epoch": 0.6012255165989696,
"grad_norm": 3.3940374851226807,
"learning_rate": 7.248326065185527e-06,
"loss": 0.13147175312042236,
"step": 5930
},
{
"epoch": 0.6022393876219021,
"grad_norm": 2.655005931854248,
"learning_rate": 7.216774110554724e-06,
"loss": 0.12423207759857177,
"step": 5940
},
{
"epoch": 0.6032532586448346,
"grad_norm": 2.6113064289093018,
"learning_rate": 7.185252161742385e-06,
"loss": 0.13161710500717164,
"step": 5950
},
{
"epoch": 0.6042671296677672,
"grad_norm": 2.2770113945007324,
"learning_rate": 7.153760558585059e-06,
"loss": 0.11782714128494262,
"step": 5960
},
{
"epoch": 0.6052810006906997,
"grad_norm": 3.8537843227386475,
"learning_rate": 7.122299640592135e-06,
"loss": 0.1270732045173645,
"step": 5970
},
{
"epoch": 0.6062948717136322,
"grad_norm": 2.2156670093536377,
"learning_rate": 7.090869746942191e-06,
"loss": 0.12284700870513916,
"step": 5980
},
{
"epoch": 0.6073087427365647,
"grad_norm": 3.2629640102386475,
"learning_rate": 7.05947121647933e-06,
"loss": 0.13652844429016114,
"step": 5990
},
{
"epoch": 0.6083226137594971,
"grad_norm": 3.59788179397583,
"learning_rate": 7.028104387709531e-06,
"loss": 0.1218494176864624,
"step": 6000
},
{
"epoch": 0.6093364847824296,
"grad_norm": 2.8342368602752686,
"learning_rate": 6.996769598797001e-06,
"loss": 0.11924595832824707,
"step": 6010
},
{
"epoch": 0.6103503558053621,
"grad_norm": 3.567408323287964,
"learning_rate": 6.965467187560524e-06,
"loss": 0.11976990699768067,
"step": 6020
},
{
"epoch": 0.6113642268282946,
"grad_norm": 3.005990505218506,
"learning_rate": 6.934197491469818e-06,
"loss": 0.1092523455619812,
"step": 6030
},
{
"epoch": 0.6123780978512271,
"grad_norm": 3.618170976638794,
"learning_rate": 6.902960847641915e-06,
"loss": 0.13374135494232178,
"step": 6040
},
{
"epoch": 0.6133919688741596,
"grad_norm": 3.890486001968384,
"learning_rate": 6.8717575928374915e-06,
"loss": 0.13332891464233398,
"step": 6050
},
{
"epoch": 0.6144058398970921,
"grad_norm": 2.6272149085998535,
"learning_rate": 6.840588063457278e-06,
"loss": 0.12563180923461914,
"step": 6060
},
{
"epoch": 0.6154197109200246,
"grad_norm": 2.831005334854126,
"learning_rate": 6.809452595538403e-06,
"loss": 0.13246108293533326,
"step": 6070
},
{
"epoch": 0.616433581942957,
"grad_norm": 2.5452754497528076,
"learning_rate": 6.778351524750777e-06,
"loss": 0.11854002475738526,
"step": 6080
},
{
"epoch": 0.6174474529658895,
"grad_norm": 2.7081029415130615,
"learning_rate": 6.747285186393488e-06,
"loss": 0.127579927444458,
"step": 6090
},
{
"epoch": 0.618461323988822,
"grad_norm": 4.759993553161621,
"learning_rate": 6.716253915391167e-06,
"loss": 0.13936355113983154,
"step": 6100
},
{
"epoch": 0.6194751950117545,
"grad_norm": 1.9744373559951782,
"learning_rate": 6.685258046290381e-06,
"loss": 0.12731306552886962,
"step": 6110
},
{
"epoch": 0.620489066034687,
"grad_norm": 3.5431594848632812,
"learning_rate": 6.654297913256047e-06,
"loss": 0.14302924871444703,
"step": 6120
},
{
"epoch": 0.6215029370576196,
"grad_norm": 2.790893077850342,
"learning_rate": 6.623373850067794e-06,
"loss": 0.13724067211151122,
"step": 6130
},
{
"epoch": 0.6225168080805521,
"grad_norm": 4.765311241149902,
"learning_rate": 6.592486190116405e-06,
"loss": 0.13404587507247925,
"step": 6140
},
{
"epoch": 0.6235306791034846,
"grad_norm": 4.108818054199219,
"learning_rate": 6.561635266400181e-06,
"loss": 0.11459832191467285,
"step": 6150
},
{
"epoch": 0.6245445501264171,
"grad_norm": 5.287535667419434,
"learning_rate": 6.5308214115213785e-06,
"loss": 0.13368606567382812,
"step": 6160
},
{
"epoch": 0.6255584211493496,
"grad_norm": 2.4635119438171387,
"learning_rate": 6.500044957682627e-06,
"loss": 0.12908401489257812,
"step": 6170
},
{
"epoch": 0.626572292172282,
"grad_norm": 2.659508466720581,
"learning_rate": 6.469306236683324e-06,
"loss": 0.11673092842102051,
"step": 6180
},
{
"epoch": 0.6275861631952145,
"grad_norm": 2.8278465270996094,
"learning_rate": 6.438605579916078e-06,
"loss": 0.13038581609725952,
"step": 6190
},
{
"epoch": 0.628600034218147,
"grad_norm": 4.357350826263428,
"learning_rate": 6.4079433183631255e-06,
"loss": 0.12441782951354981,
"step": 6200
},
{
"epoch": 0.6296139052410795,
"grad_norm": 3.119800090789795,
"learning_rate": 6.377319782592767e-06,
"loss": 0.11600160598754883,
"step": 6210
},
{
"epoch": 0.630627776264012,
"grad_norm": 4.034447193145752,
"learning_rate": 6.346735302755805e-06,
"loss": 0.13251206874847413,
"step": 6220
},
{
"epoch": 0.6316416472869445,
"grad_norm": 3.305497407913208,
"learning_rate": 6.316190208581979e-06,
"loss": 0.1198967456817627,
"step": 6230
},
{
"epoch": 0.632655518309877,
"grad_norm": 3.0117650032043457,
"learning_rate": 6.285684829376413e-06,
"loss": 0.11986765861511231,
"step": 6240
},
{
"epoch": 0.6336693893328095,
"grad_norm": 4.478419780731201,
"learning_rate": 6.25521949401607e-06,
"loss": 0.11757739782333373,
"step": 6250
},
{
"epoch": 0.634683260355742,
"grad_norm": 4.012950897216797,
"learning_rate": 6.2247945309461955e-06,
"loss": 0.12513408660888672,
"step": 6260
},
{
"epoch": 0.6356971313786745,
"grad_norm": 3.94972562789917,
"learning_rate": 6.194410268176796e-06,
"loss": 0.1264721155166626,
"step": 6270
},
{
"epoch": 0.6367110024016069,
"grad_norm": 2.730064868927002,
"learning_rate": 6.164067033279073e-06,
"loss": 0.11956477165222168,
"step": 6280
},
{
"epoch": 0.6377248734245394,
"grad_norm": 3.129274368286133,
"learning_rate": 6.133765153381918e-06,
"loss": 0.13525882959365845,
"step": 6290
},
{
"epoch": 0.638738744447472,
"grad_norm": 3.2582685947418213,
"learning_rate": 6.103504955168382e-06,
"loss": 0.12026245594024658,
"step": 6300
},
{
"epoch": 0.6397526154704045,
"grad_norm": 3.5812673568725586,
"learning_rate": 6.07328676487214e-06,
"loss": 0.1319317936897278,
"step": 6310
},
{
"epoch": 0.640766486493337,
"grad_norm": 3.111353635787964,
"learning_rate": 6.043110908273973e-06,
"loss": 0.12165262699127197,
"step": 6320
},
{
"epoch": 0.6417803575162695,
"grad_norm": 4.7924981117248535,
"learning_rate": 6.012977710698288e-06,
"loss": 0.10895328521728516,
"step": 6330
},
{
"epoch": 0.642794228539202,
"grad_norm": 4.90350341796875,
"learning_rate": 5.982887497009566e-06,
"loss": 0.1305592656135559,
"step": 6340
},
{
"epoch": 0.6438080995621345,
"grad_norm": 3.278956413269043,
"learning_rate": 5.9528405916088926e-06,
"loss": 0.11411793231964111,
"step": 6350
},
{
"epoch": 0.644821970585067,
"grad_norm": 5.162795543670654,
"learning_rate": 5.9228373184304455e-06,
"loss": 0.12445547580718994,
"step": 6360
},
{
"epoch": 0.6458358416079995,
"grad_norm": 3.5281789302825928,
"learning_rate": 5.892878000938001e-06,
"loss": 0.11668292284011841,
"step": 6370
},
{
"epoch": 0.6468497126309319,
"grad_norm": 3.2116007804870605,
"learning_rate": 5.86296296212146e-06,
"loss": 0.11492093801498413,
"step": 6380
},
{
"epoch": 0.6478635836538644,
"grad_norm": 2.5329136848449707,
"learning_rate": 5.833092524493349e-06,
"loss": 0.11985704898834229,
"step": 6390
},
{
"epoch": 0.6488774546767969,
"grad_norm": 3.386889696121216,
"learning_rate": 5.80326701008536e-06,
"loss": 0.11552813053131103,
"step": 6400
},
{
"epoch": 0.6498913256997294,
"grad_norm": 3.757209539413452,
"learning_rate": 5.773486740444862e-06,
"loss": 0.1108394742012024,
"step": 6410
},
{
"epoch": 0.6509051967226619,
"grad_norm": 2.823276996612549,
"learning_rate": 5.743752036631443e-06,
"loss": 0.1239701747894287,
"step": 6420
},
{
"epoch": 0.6519190677455944,
"grad_norm": 4.323822021484375,
"learning_rate": 5.714063219213456e-06,
"loss": 0.11108143329620361,
"step": 6430
},
{
"epoch": 0.6529329387685269,
"grad_norm": 3.168882369995117,
"learning_rate": 5.684420608264549e-06,
"loss": 0.11870837211608887,
"step": 6440
},
{
"epoch": 0.6539468097914594,
"grad_norm": 4.06667947769165,
"learning_rate": 5.654824523360219e-06,
"loss": 0.10694868564605713,
"step": 6450
},
{
"epoch": 0.6549606808143918,
"grad_norm": 3.502532482147217,
"learning_rate": 5.625275283574372e-06,
"loss": 0.11015803813934326,
"step": 6460
},
{
"epoch": 0.6559745518373244,
"grad_norm": 4.827873706817627,
"learning_rate": 5.595773207475872e-06,
"loss": 0.11638245582580567,
"step": 6470
},
{
"epoch": 0.6569884228602569,
"grad_norm": 3.4224705696105957,
"learning_rate": 5.566318613125128e-06,
"loss": 0.12504308223724364,
"step": 6480
},
{
"epoch": 0.6580022938831894,
"grad_norm": 2.7117559909820557,
"learning_rate": 5.536911818070636e-06,
"loss": 0.11904535293579102,
"step": 6490
},
{
"epoch": 0.6590161649061219,
"grad_norm": 2.634927988052368,
"learning_rate": 5.507553139345574e-06,
"loss": 0.11809990406036378,
"step": 6500
},
{
"epoch": 0.6600300359290544,
"grad_norm": 2.7234811782836914,
"learning_rate": 5.478242893464389e-06,
"loss": 0.10842573642730713,
"step": 6510
},
{
"epoch": 0.6610439069519869,
"grad_norm": 4.091216087341309,
"learning_rate": 5.448981396419366e-06,
"loss": 0.11989302635192871,
"step": 6520
},
{
"epoch": 0.6620577779749194,
"grad_norm": 2.9693658351898193,
"learning_rate": 5.419768963677233e-06,
"loss": 0.11543912887573242,
"step": 6530
},
{
"epoch": 0.6630716489978519,
"grad_norm": 3.6583468914031982,
"learning_rate": 5.39060591017576e-06,
"loss": 0.12606061697006227,
"step": 6540
},
{
"epoch": 0.6640855200207844,
"grad_norm": 2.6029796600341797,
"learning_rate": 5.3614925503203586e-06,
"loss": 0.12360832691192628,
"step": 6550
},
{
"epoch": 0.6650993910437168,
"grad_norm": 5.494139194488525,
"learning_rate": 5.332429197980706e-06,
"loss": 0.11253817081451416,
"step": 6560
},
{
"epoch": 0.6661132620666493,
"grad_norm": 3.9320521354675293,
"learning_rate": 5.303416166487337e-06,
"loss": 0.12184336185455322,
"step": 6570
},
{
"epoch": 0.6671271330895818,
"grad_norm": 5.509915351867676,
"learning_rate": 5.274453768628289e-06,
"loss": 0.12528070211410522,
"step": 6580
},
{
"epoch": 0.6681410041125143,
"grad_norm": 2.9067180156707764,
"learning_rate": 5.245542316645718e-06,
"loss": 0.11270208358764648,
"step": 6590
},
{
"epoch": 0.6691548751354468,
"grad_norm": 3.995389461517334,
"learning_rate": 5.2166821222325306e-06,
"loss": 0.10754237174987794,
"step": 6600
},
{
"epoch": 0.6701687461583793,
"grad_norm": 2.7081081867218018,
"learning_rate": 5.187873496529041e-06,
"loss": 0.10937079191207885,
"step": 6610
},
{
"epoch": 0.6711826171813118,
"grad_norm": 4.117220878601074,
"learning_rate": 5.15911675011959e-06,
"loss": 0.12457695007324218,
"step": 6620
},
{
"epoch": 0.6721964882042443,
"grad_norm": 2.8138976097106934,
"learning_rate": 5.130412193029213e-06,
"loss": 0.117437744140625,
"step": 6630
},
{
"epoch": 0.6732103592271768,
"grad_norm": 3.8193001747131348,
"learning_rate": 5.101760134720306e-06,
"loss": 0.11478121280670166,
"step": 6640
},
{
"epoch": 0.6742242302501094,
"grad_norm": 3.5127668380737305,
"learning_rate": 5.073160884089263e-06,
"loss": 0.1244268536567688,
"step": 6650
},
{
"epoch": 0.6752381012730418,
"grad_norm": 3.0123836994171143,
"learning_rate": 5.044614749463173e-06,
"loss": 0.10863211154937744,
"step": 6660
},
{
"epoch": 0.6762519722959743,
"grad_norm": 3.290250301361084,
"learning_rate": 5.0161220385964745e-06,
"loss": 0.10864517688751221,
"step": 6670
},
{
"epoch": 0.6772658433189068,
"grad_norm": 2.559648275375366,
"learning_rate": 4.987683058667651e-06,
"loss": 0.11527155637741089,
"step": 6680
},
{
"epoch": 0.6782797143418393,
"grad_norm": 4.497267723083496,
"learning_rate": 4.9592981162759215e-06,
"loss": 0.12280102968215942,
"step": 6690
},
{
"epoch": 0.6792935853647718,
"grad_norm": 2.4327831268310547,
"learning_rate": 4.930967517437922e-06,
"loss": 0.10697582960128785,
"step": 6700
},
{
"epoch": 0.6803074563877043,
"grad_norm": 5.957965850830078,
"learning_rate": 4.902691567584412e-06,
"loss": 0.11532588005065918,
"step": 6710
},
{
"epoch": 0.6813213274106368,
"grad_norm": 3.598313808441162,
"learning_rate": 4.87447057155699e-06,
"loss": 0.1099928379058838,
"step": 6720
},
{
"epoch": 0.6823351984335693,
"grad_norm": 4.173735618591309,
"learning_rate": 4.846304833604792e-06,
"loss": 0.11035137176513672,
"step": 6730
},
{
"epoch": 0.6833490694565018,
"grad_norm": 6.358034610748291,
"learning_rate": 4.8181946573812236e-06,
"loss": 0.12221287488937378,
"step": 6740
},
{
"epoch": 0.6843629404794342,
"grad_norm": 1.996628761291504,
"learning_rate": 4.790140345940688e-06,
"loss": 0.10727550983428955,
"step": 6750
},
{
"epoch": 0.6853768115023667,
"grad_norm": 4.452577114105225,
"learning_rate": 4.762142201735299e-06,
"loss": 0.1096308708190918,
"step": 6760
},
{
"epoch": 0.6863906825252992,
"grad_norm": 4.759271621704102,
"learning_rate": 4.734200526611652e-06,
"loss": 0.10596942901611328,
"step": 6770
},
{
"epoch": 0.6874045535482317,
"grad_norm": 3.545624017715454,
"learning_rate": 4.706315621807537e-06,
"loss": 0.11316285133361817,
"step": 6780
},
{
"epoch": 0.6884184245711642,
"grad_norm": 3.554701328277588,
"learning_rate": 4.678487787948711e-06,
"loss": 0.11098406314849854,
"step": 6790
},
{
"epoch": 0.6894322955940967,
"grad_norm": 4.247222900390625,
"learning_rate": 4.650717325045655e-06,
"loss": 0.10941197872161865,
"step": 6800
},
{
"epoch": 0.6904461666170292,
"grad_norm": 4.068756103515625,
"learning_rate": 4.623004532490328e-06,
"loss": 0.11889941692352295,
"step": 6810
},
{
"epoch": 0.6914600376399618,
"grad_norm": 3.7360899448394775,
"learning_rate": 4.595349709052964e-06,
"loss": 0.10977576971054077,
"step": 6820
},
{
"epoch": 0.6924739086628943,
"grad_norm": 3.2178421020507812,
"learning_rate": 4.5677531528788196e-06,
"loss": 0.10204930305480957,
"step": 6830
},
{
"epoch": 0.6934877796858268,
"grad_norm": 3.7818236351013184,
"learning_rate": 4.54021516148498e-06,
"loss": 0.1089769721031189,
"step": 6840
},
{
"epoch": 0.6945016507087592,
"grad_norm": 3.872225761413574,
"learning_rate": 4.5127360317571475e-06,
"loss": 0.11058665513992309,
"step": 6850
},
{
"epoch": 0.6955155217316917,
"grad_norm": 3.279524087905884,
"learning_rate": 4.485316059946438e-06,
"loss": 0.10155479907989502,
"step": 6860
},
{
"epoch": 0.6965293927546242,
"grad_norm": 5.33518123626709,
"learning_rate": 4.457955541666188e-06,
"loss": 0.11589955091476441,
"step": 6870
},
{
"epoch": 0.6975432637775567,
"grad_norm": 4.027236461639404,
"learning_rate": 4.430654771888776e-06,
"loss": 0.09957776069641114,
"step": 6880
},
{
"epoch": 0.6985571348004892,
"grad_norm": 5.866827487945557,
"learning_rate": 4.403414044942422e-06,
"loss": 0.10335280895233154,
"step": 6890
},
{
"epoch": 0.6995710058234217,
"grad_norm": 4.475950717926025,
"learning_rate": 4.376233654508042e-06,
"loss": 0.10277276039123535,
"step": 6900
},
{
"epoch": 0.7005848768463542,
"grad_norm": 4.117318153381348,
"learning_rate": 4.349113893616056e-06,
"loss": 0.09870555400848388,
"step": 6910
},
{
"epoch": 0.7015987478692867,
"grad_norm": 4.379005432128906,
"learning_rate": 4.322055054643244e-06,
"loss": 0.10079617500305176,
"step": 6920
},
{
"epoch": 0.7026126188922192,
"grad_norm": 4.7671709060668945,
"learning_rate": 4.295057429309589e-06,
"loss": 0.11881935596466064,
"step": 6930
},
{
"epoch": 0.7036264899151516,
"grad_norm": 4.0872673988342285,
"learning_rate": 4.268121308675132e-06,
"loss": 0.09912853240966797,
"step": 6940
},
{
"epoch": 0.7046403609380841,
"grad_norm": 3.417915105819702,
"learning_rate": 4.241246983136834e-06,
"loss": 0.1078688621520996,
"step": 6950
},
{
"epoch": 0.7056542319610166,
"grad_norm": 3.042978525161743,
"learning_rate": 4.214434742425455e-06,
"loss": 0.09791347980499268,
"step": 6960
},
{
"epoch": 0.7066681029839491,
"grad_norm": 4.524270057678223,
"learning_rate": 4.1876848756024116e-06,
"loss": 0.1063512921333313,
"step": 6970
},
{
"epoch": 0.7076819740068816,
"grad_norm": 2.8843939304351807,
"learning_rate": 4.1609976710566735e-06,
"loss": 0.12488405704498291,
"step": 6980
},
{
"epoch": 0.7086958450298142,
"grad_norm": 2.5707852840423584,
"learning_rate": 4.134373416501652e-06,
"loss": 0.10664181709289551,
"step": 6990
},
{
"epoch": 0.7097097160527467,
"grad_norm": 3.3396098613739014,
"learning_rate": 4.107812398972093e-06,
"loss": 0.11485625505447387,
"step": 7000
},
{
"epoch": 0.7107235870756792,
"grad_norm": 3.7509665489196777,
"learning_rate": 4.081314904821001e-06,
"loss": 0.10062665939331054,
"step": 7010
},
{
"epoch": 0.7117374580986117,
"grad_norm": 4.215871334075928,
"learning_rate": 4.054881219716521e-06,
"loss": 0.11119086742401123,
"step": 7020
},
{
"epoch": 0.7127513291215442,
"grad_norm": 4.858083248138428,
"learning_rate": 4.0285116286388914e-06,
"loss": 0.09426209926605225,
"step": 7030
},
{
"epoch": 0.7137652001444766,
"grad_norm": 6.0166473388671875,
"learning_rate": 4.0022064158773445e-06,
"loss": 0.11742116212844848,
"step": 7040
},
{
"epoch": 0.7147790711674091,
"grad_norm": 3.2609286308288574,
"learning_rate": 3.9759658650270575e-06,
"loss": 0.11036663055419922,
"step": 7050
},
{
"epoch": 0.7157929421903416,
"grad_norm": 4.492204189300537,
"learning_rate": 3.94979025898609e-06,
"loss": 0.10044634342193604,
"step": 7060
},
{
"epoch": 0.7168068132132741,
"grad_norm": 3.360687017440796,
"learning_rate": 3.9236798799523375e-06,
"loss": 0.1072852373123169,
"step": 7070
},
{
"epoch": 0.7178206842362066,
"grad_norm": 4.381544589996338,
"learning_rate": 3.89763500942048e-06,
"loss": 0.11510798931121827,
"step": 7080
},
{
"epoch": 0.7188345552591391,
"grad_norm": 2.5754785537719727,
"learning_rate": 3.871655928178962e-06,
"loss": 0.10218755006790162,
"step": 7090
},
{
"epoch": 0.7198484262820716,
"grad_norm": 4.045770168304443,
"learning_rate": 3.8457429163069534e-06,
"loss": 0.11071910858154296,
"step": 7100
},
{
"epoch": 0.7208622973050041,
"grad_norm": 3.4769110679626465,
"learning_rate": 3.819896253171332e-06,
"loss": 0.11043792963027954,
"step": 7110
},
{
"epoch": 0.7218761683279366,
"grad_norm": 4.183675289154053,
"learning_rate": 3.7941162174236745e-06,
"loss": 0.09668819904327393,
"step": 7120
},
{
"epoch": 0.722890039350869,
"grad_norm": 5.783495903015137,
"learning_rate": 3.7684030869972464e-06,
"loss": 0.08785664439201354,
"step": 7130
},
{
"epoch": 0.7239039103738015,
"grad_norm": 5.215122222900391,
"learning_rate": 3.742757139104023e-06,
"loss": 0.1129606008529663,
"step": 7140
},
{
"epoch": 0.724917781396734,
"grad_norm": 3.7518675327301025,
"learning_rate": 3.7171786502316767e-06,
"loss": 0.11115808486938476,
"step": 7150
},
{
"epoch": 0.7259316524196666,
"grad_norm": 1.9049906730651855,
"learning_rate": 3.691667896140605e-06,
"loss": 0.09753057360649109,
"step": 7160
},
{
"epoch": 0.7269455234425991,
"grad_norm": 3.287567615509033,
"learning_rate": 3.6662251518609747e-06,
"loss": 0.10345501899719238,
"step": 7170
},
{
"epoch": 0.7279593944655316,
"grad_norm": 4.340944290161133,
"learning_rate": 3.64085069168973e-06,
"loss": 0.09775218367576599,
"step": 7180
},
{
"epoch": 0.7289732654884641,
"grad_norm": 3.730257511138916,
"learning_rate": 3.6155447891876518e-06,
"loss": 0.10205883979797363,
"step": 7190
},
{
"epoch": 0.7299871365113966,
"grad_norm": 1.9651705026626587,
"learning_rate": 3.590307717176401e-06,
"loss": 0.09002344608306885,
"step": 7200
},
{
"epoch": 0.7310010075343291,
"grad_norm": 4.629663944244385,
"learning_rate": 3.565139747735581e-06,
"loss": 0.09352742433547974,
"step": 7210
},
{
"epoch": 0.7320148785572616,
"grad_norm": 3.4427878856658936,
"learning_rate": 3.5400411521998123e-06,
"loss": 0.08346476554870605,
"step": 7220
},
{
"epoch": 0.733028749580194,
"grad_norm": 4.608767032623291,
"learning_rate": 3.5150122011557888e-06,
"loss": 0.1056178092956543,
"step": 7230
},
{
"epoch": 0.7340426206031265,
"grad_norm": 3.6405434608459473,
"learning_rate": 3.4900531644393744e-06,
"loss": 0.09827905893325806,
"step": 7240
},
{
"epoch": 0.735056491626059,
"grad_norm": 4.988274097442627,
"learning_rate": 3.4651643111326895e-06,
"loss": 0.09997121095657349,
"step": 7250
},
{
"epoch": 0.7360703626489915,
"grad_norm": 4.005931854248047,
"learning_rate": 3.440345909561208e-06,
"loss": 0.08496606945991517,
"step": 7260
},
{
"epoch": 0.737084233671924,
"grad_norm": 2.8140532970428467,
"learning_rate": 3.415598227290877e-06,
"loss": 0.09334189891815185,
"step": 7270
},
{
"epoch": 0.7380981046948565,
"grad_norm": 4.083347797393799,
"learning_rate": 3.390921531125211e-06,
"loss": 0.08948659896850586,
"step": 7280
},
{
"epoch": 0.739111975717789,
"grad_norm": 4.835827350616455,
"learning_rate": 3.366316087102429e-06,
"loss": 0.0869488000869751,
"step": 7290
},
{
"epoch": 0.7401258467407215,
"grad_norm": 4.2866435050964355,
"learning_rate": 3.3417821604925937e-06,
"loss": 0.09308248162269592,
"step": 7300
},
{
"epoch": 0.741139717763654,
"grad_norm": 4.676980495452881,
"learning_rate": 3.3173200157947315e-06,
"loss": 0.10867881774902344,
"step": 7310
},
{
"epoch": 0.7421535887865864,
"grad_norm": 5.795026779174805,
"learning_rate": 3.2929299167339955e-06,
"loss": 0.09959579706192016,
"step": 7320
},
{
"epoch": 0.7431674598095189,
"grad_norm": 3.178743600845337,
"learning_rate": 3.2686121262588165e-06,
"loss": 0.09881712198257446,
"step": 7330
},
{
"epoch": 0.7441813308324515,
"grad_norm": 4.3362717628479,
"learning_rate": 3.244366906538071e-06,
"loss": 0.10258498191833496,
"step": 7340
},
{
"epoch": 0.745195201855384,
"grad_norm": 2.979397773742676,
"learning_rate": 3.2201945189582597e-06,
"loss": 0.09613606929779053,
"step": 7350
},
{
"epoch": 0.7462090728783165,
"grad_norm": 3.6704070568084717,
"learning_rate": 3.1960952241206745e-06,
"loss": 0.09370797276496887,
"step": 7360
},
{
"epoch": 0.747222943901249,
"grad_norm": 5.913214206695557,
"learning_rate": 3.172069281838603e-06,
"loss": 0.09167356491088867,
"step": 7370
},
{
"epoch": 0.7482368149241815,
"grad_norm": 5.8804521560668945,
"learning_rate": 3.148116951134522e-06,
"loss": 0.09761692881584168,
"step": 7380
},
{
"epoch": 0.749250685947114,
"grad_norm": 2.887545585632324,
"learning_rate": 3.1242384902373024e-06,
"loss": 0.09275962710380554,
"step": 7390
},
{
"epoch": 0.7502645569700465,
"grad_norm": 3.007314920425415,
"learning_rate": 3.1004341565794373e-06,
"loss": 0.10005519390106202,
"step": 7400
},
{
"epoch": 0.751278427992979,
"grad_norm": 5.379554271697998,
"learning_rate": 3.0767042067942475e-06,
"loss": 0.09788405299186706,
"step": 7410
},
{
"epoch": 0.7522922990159114,
"grad_norm": 4.7895612716674805,
"learning_rate": 3.0530488967131265e-06,
"loss": 0.08829056620597839,
"step": 7420
},
{
"epoch": 0.7533061700388439,
"grad_norm": 4.062852382659912,
"learning_rate": 3.0294684813627917e-06,
"loss": 0.08906958103179932,
"step": 7430
},
{
"epoch": 0.7543200410617764,
"grad_norm": 5.67551326751709,
"learning_rate": 3.0059632149625118e-06,
"loss": 0.08716234564781189,
"step": 7440
},
{
"epoch": 0.7553339120847089,
"grad_norm": 4.380494117736816,
"learning_rate": 2.982533350921383e-06,
"loss": 0.09460397362709046,
"step": 7450
},
{
"epoch": 0.7563477831076414,
"grad_norm": 2.275362253189087,
"learning_rate": 2.959179141835591e-06,
"loss": 0.08684351444244384,
"step": 7460
},
{
"epoch": 0.7573616541305739,
"grad_norm": 3.053177833557129,
"learning_rate": 2.935900839485688e-06,
"loss": 0.09221785068511963,
"step": 7470
},
{
"epoch": 0.7583755251535064,
"grad_norm": 5.197820663452148,
"learning_rate": 2.9126986948338898e-06,
"loss": 0.0799995243549347,
"step": 7480
},
{
"epoch": 0.7593893961764389,
"grad_norm": 2.7164454460144043,
"learning_rate": 2.889572958021347e-06,
"loss": 0.08904615044593811,
"step": 7490
},
{
"epoch": 0.7604032671993713,
"grad_norm": 3.8327910900115967,
"learning_rate": 2.866523878365468e-06,
"loss": 0.07715175151824952,
"step": 7500
},
{
"epoch": 0.761417138222304,
"grad_norm": 3.9158542156219482,
"learning_rate": 2.8435517043572204e-06,
"loss": 0.08948068618774414,
"step": 7510
},
{
"epoch": 0.7624310092452364,
"grad_norm": 4.099941730499268,
"learning_rate": 2.820656683658456e-06,
"loss": 0.09695887565612793,
"step": 7520
},
{
"epoch": 0.7634448802681689,
"grad_norm": 4.444246292114258,
"learning_rate": 2.797839063099249e-06,
"loss": 0.08869975805282593,
"step": 7530
},
{
"epoch": 0.7644587512911014,
"grad_norm": 3.452589511871338,
"learning_rate": 2.775099088675216e-06,
"loss": 0.10242993831634521,
"step": 7540
},
{
"epoch": 0.7654726223140339,
"grad_norm": 4.679555892944336,
"learning_rate": 2.7524370055448755e-06,
"loss": 0.09647311568260193,
"step": 7550
},
{
"epoch": 0.7664864933369664,
"grad_norm": 3.5535950660705566,
"learning_rate": 2.7298530580270154e-06,
"loss": 0.08267072439193726,
"step": 7560
},
{
"epoch": 0.7675003643598989,
"grad_norm": 4.4267473220825195,
"learning_rate": 2.7073474895980333e-06,
"loss": 0.09025751948356628,
"step": 7570
},
{
"epoch": 0.7685142353828314,
"grad_norm": 4.029013633728027,
"learning_rate": 2.6849205428893334e-06,
"loss": 0.08459845185279846,
"step": 7580
},
{
"epoch": 0.7695281064057639,
"grad_norm": 2.4955527782440186,
"learning_rate": 2.662572459684699e-06,
"loss": 0.08028301000595092,
"step": 7590
},
{
"epoch": 0.7705419774286963,
"grad_norm": 3.302767753601074,
"learning_rate": 2.640303480917693e-06,
"loss": 0.0870816469192505,
"step": 7600
},
{
"epoch": 0.7715558484516288,
"grad_norm": 3.1298558712005615,
"learning_rate": 2.618113846669059e-06,
"loss": 0.08608932495117187,
"step": 7610
},
{
"epoch": 0.7725697194745613,
"grad_norm": 5.863857269287109,
"learning_rate": 2.596003796164127e-06,
"loss": 0.08926304578781127,
"step": 7620
},
{
"epoch": 0.7735835904974938,
"grad_norm": 3.27186918258667,
"learning_rate": 2.5739735677702416e-06,
"loss": 0.08512946367263793,
"step": 7630
},
{
"epoch": 0.7745974615204263,
"grad_norm": 4.9715752601623535,
"learning_rate": 2.552023398994188e-06,
"loss": 0.08647626638412476,
"step": 7640
},
{
"epoch": 0.7756113325433588,
"grad_norm": 5.376694679260254,
"learning_rate": 2.5301535264796274e-06,
"loss": 0.0960479199886322,
"step": 7650
},
{
"epoch": 0.7766252035662913,
"grad_norm": 6.782083034515381,
"learning_rate": 2.508364186004566e-06,
"loss": 0.09775331616401672,
"step": 7660
},
{
"epoch": 0.7776390745892238,
"grad_norm": 5.304043292999268,
"learning_rate": 2.486655612478782e-06,
"loss": 0.08757140636444091,
"step": 7670
},
{
"epoch": 0.7786529456121564,
"grad_norm": 4.488348007202148,
"learning_rate": 2.465028039941316e-06,
"loss": 0.08265872001647949,
"step": 7680
},
{
"epoch": 0.7796668166350889,
"grad_norm": 3.9967384338378906,
"learning_rate": 2.443481701557946e-06,
"loss": 0.08648114800453185,
"step": 7690
},
{
"epoch": 0.7806806876580213,
"grad_norm": 3.7726054191589355,
"learning_rate": 2.4220168296186587e-06,
"loss": 0.08333401679992676,
"step": 7700
},
{
"epoch": 0.7816945586809538,
"grad_norm": 4.3952131271362305,
"learning_rate": 2.4006336555351615e-06,
"loss": 0.08247607946395874,
"step": 7710
},
{
"epoch": 0.7827084297038863,
"grad_norm": 4.9461350440979,
"learning_rate": 2.3793324098383796e-06,
"loss": 0.08987885117530822,
"step": 7720
},
{
"epoch": 0.7837223007268188,
"grad_norm": 6.3852386474609375,
"learning_rate": 2.35811332217597e-06,
"loss": 0.07547081112861634,
"step": 7730
},
{
"epoch": 0.7847361717497513,
"grad_norm": 3.7512640953063965,
"learning_rate": 2.3369766213098553e-06,
"loss": 0.08143643140792847,
"step": 7740
},
{
"epoch": 0.7857500427726838,
"grad_norm": 3.9396536350250244,
"learning_rate": 2.315922535113743e-06,
"loss": 0.08094037771224975,
"step": 7750
},
{
"epoch": 0.7867639137956163,
"grad_norm": 4.576452255249023,
"learning_rate": 2.2949512905706793e-06,
"loss": 0.08206928968429565,
"step": 7760
},
{
"epoch": 0.7877777848185488,
"grad_norm": 5.834323883056641,
"learning_rate": 2.274063113770597e-06,
"loss": 0.07476301789283753,
"step": 7770
},
{
"epoch": 0.7887916558414813,
"grad_norm": 4.068087100982666,
"learning_rate": 2.2532582299078775e-06,
"loss": 0.09134551286697387,
"step": 7780
},
{
"epoch": 0.7898055268644137,
"grad_norm": 3.458346366882324,
"learning_rate": 2.2325368632789345e-06,
"loss": 0.08150627613067626,
"step": 7790
},
{
"epoch": 0.7908193978873462,
"grad_norm": 4.828081130981445,
"learning_rate": 2.211899237279779e-06,
"loss": 0.08178857564926148,
"step": 7800
},
{
"epoch": 0.7918332689102787,
"grad_norm": 3.63834285736084,
"learning_rate": 2.1913455744036138e-06,
"loss": 0.10948653221130371,
"step": 7810
},
{
"epoch": 0.7928471399332112,
"grad_norm": 3.7400805950164795,
"learning_rate": 2.1708760962384545e-06,
"loss": 0.09584753513336182,
"step": 7820
},
{
"epoch": 0.7938610109561437,
"grad_norm": 5.6095662117004395,
"learning_rate": 2.1504910234647112e-06,
"loss": 0.08254612684249878,
"step": 7830
},
{
"epoch": 0.7948748819790762,
"grad_norm": 4.4293413162231445,
"learning_rate": 2.130190575852832e-06,
"loss": 0.0781608521938324,
"step": 7840
},
{
"epoch": 0.7958887530020088,
"grad_norm": 4.166698932647705,
"learning_rate": 2.109974972260921e-06,
"loss": 0.07337145209312439,
"step": 7850
},
{
"epoch": 0.7969026240249413,
"grad_norm": 4.298923969268799,
"learning_rate": 2.0898444306323807e-06,
"loss": 0.07692206501960755,
"step": 7860
},
{
"epoch": 0.7979164950478738,
"grad_norm": 3.5728261470794678,
"learning_rate": 2.069799167993577e-06,
"loss": 0.0906355321407318,
"step": 7870
},
{
"epoch": 0.7989303660708063,
"grad_norm": 5.480486869812012,
"learning_rate": 2.0498394004514745e-06,
"loss": 0.08317444324493409,
"step": 7880
},
{
"epoch": 0.7999442370937387,
"grad_norm": 4.030394077301025,
"learning_rate": 2.029965343191326e-06,
"loss": 0.09432138204574585,
"step": 7890
},
{
"epoch": 0.8009581081166712,
"grad_norm": 4.4059367179870605,
"learning_rate": 2.0101772104743422e-06,
"loss": 0.08797411322593689,
"step": 7900
},
{
"epoch": 0.8019719791396037,
"grad_norm": 3.2624707221984863,
"learning_rate": 1.990475215635388e-06,
"loss": 0.08104816675186158,
"step": 7910
},
{
"epoch": 0.8029858501625362,
"grad_norm": 4.162240982055664,
"learning_rate": 1.9708595710806787e-06,
"loss": 0.07154384851455689,
"step": 7920
},
{
"epoch": 0.8039997211854687,
"grad_norm": 4.724563121795654,
"learning_rate": 1.9513304882854966e-06,
"loss": 0.07427242398262024,
"step": 7930
},
{
"epoch": 0.8050135922084012,
"grad_norm": 6.564634799957275,
"learning_rate": 1.9318881777918973e-06,
"loss": 0.07599087357521057,
"step": 7940
},
{
"epoch": 0.8060274632313337,
"grad_norm": 5.843585014343262,
"learning_rate": 1.9125328492064575e-06,
"loss": 0.07820788621902466,
"step": 7950
},
{
"epoch": 0.8070413342542662,
"grad_norm": 4.897693157196045,
"learning_rate": 1.8932647111980007e-06,
"loss": 0.08159160614013672,
"step": 7960
},
{
"epoch": 0.8080552052771987,
"grad_norm": 3.4772603511810303,
"learning_rate": 1.8740839714953552e-06,
"loss": 0.07544232606887817,
"step": 7970
},
{
"epoch": 0.8090690763001311,
"grad_norm": 4.972339630126953,
"learning_rate": 1.8549908368851099e-06,
"loss": 0.09030557274818421,
"step": 7980
},
{
"epoch": 0.8100829473230636,
"grad_norm": 4.231588363647461,
"learning_rate": 1.8359855132093884e-06,
"loss": 0.06983974575996399,
"step": 7990
},
{
"epoch": 0.8110968183459961,
"grad_norm": 4.495914459228516,
"learning_rate": 1.8170682053636358e-06,
"loss": 0.08559158444404602,
"step": 8000
},
{
"epoch": 0.8121106893689286,
"grad_norm": 5.225792407989502,
"learning_rate": 1.7982391172943936e-06,
"loss": 0.07759113311767578,
"step": 8010
},
{
"epoch": 0.8131245603918611,
"grad_norm": 5.047466278076172,
"learning_rate": 1.7794984519971126e-06,
"loss": 0.09562692642211915,
"step": 8020
},
{
"epoch": 0.8141384314147937,
"grad_norm": 3.8912978172302246,
"learning_rate": 1.7608464115139712e-06,
"loss": 0.07346256971359252,
"step": 8030
},
{
"epoch": 0.8151523024377262,
"grad_norm": 3.0911762714385986,
"learning_rate": 1.7422831969316724e-06,
"loss": 0.08886501789093018,
"step": 8040
},
{
"epoch": 0.8161661734606587,
"grad_norm": 4.10170841217041,
"learning_rate": 1.7238090083793003e-06,
"loss": 0.07181520462036133,
"step": 8050
},
{
"epoch": 0.8171800444835912,
"grad_norm": 4.242997646331787,
"learning_rate": 1.7054240450261594e-06,
"loss": 0.06896800994873047,
"step": 8060
},
{
"epoch": 0.8181939155065237,
"grad_norm": 4.576391696929932,
"learning_rate": 1.6871285050796093e-06,
"loss": 0.0828772246837616,
"step": 8070
},
{
"epoch": 0.8192077865294561,
"grad_norm": 3.8591134548187256,
"learning_rate": 1.6689225857829527e-06,
"loss": 0.07515162229537964,
"step": 8080
},
{
"epoch": 0.8202216575523886,
"grad_norm": 4.734580993652344,
"learning_rate": 1.6508064834132875e-06,
"loss": 0.08277686834335327,
"step": 8090
},
{
"epoch": 0.8212355285753211,
"grad_norm": 6.300182342529297,
"learning_rate": 1.632780393279404e-06,
"loss": 0.07737209796905517,
"step": 8100
},
{
"epoch": 0.8222493995982536,
"grad_norm": 4.55206298828125,
"learning_rate": 1.614844509719674e-06,
"loss": 0.07655172348022461,
"step": 8110
},
{
"epoch": 0.8232632706211861,
"grad_norm": 4.130857944488525,
"learning_rate": 1.5969990260999569e-06,
"loss": 0.08142294883728027,
"step": 8120
},
{
"epoch": 0.8242771416441186,
"grad_norm": 3.4378278255462646,
"learning_rate": 1.5792441348115116e-06,
"loss": 0.07142369747161866,
"step": 8130
},
{
"epoch": 0.8252910126670511,
"grad_norm": 5.728525161743164,
"learning_rate": 1.5615800272689352e-06,
"loss": 0.06666783690452575,
"step": 8140
},
{
"epoch": 0.8263048836899836,
"grad_norm": 3.637394905090332,
"learning_rate": 1.5440068939080776e-06,
"loss": 0.06981600522994995,
"step": 8150
},
{
"epoch": 0.827318754712916,
"grad_norm": 4.84428596496582,
"learning_rate": 1.5265249241840153e-06,
"loss": 0.07258843183517456,
"step": 8160
},
{
"epoch": 0.8283326257358485,
"grad_norm": 4.921848297119141,
"learning_rate": 1.5091343065689756e-06,
"loss": 0.07620466947555542,
"step": 8170
},
{
"epoch": 0.829346496758781,
"grad_norm": 4.347558975219727,
"learning_rate": 1.4918352285503357e-06,
"loss": 0.06912872195243835,
"step": 8180
},
{
"epoch": 0.8303603677817135,
"grad_norm": 6.265475273132324,
"learning_rate": 1.4746278766285905e-06,
"loss": 0.07449768185615539,
"step": 8190
},
{
"epoch": 0.8313742388046461,
"grad_norm": 5.178585052490234,
"learning_rate": 1.4575124363153315e-06,
"loss": 0.08348404169082642,
"step": 8200
},
{
"epoch": 0.8323881098275786,
"grad_norm": 5.991110801696777,
"learning_rate": 1.4404890921312654e-06,
"loss": 0.07296459674835205,
"step": 8210
},
{
"epoch": 0.8334019808505111,
"grad_norm": 4.3999738693237305,
"learning_rate": 1.4235580276042071e-06,
"loss": 0.07096297740936279,
"step": 8220
},
{
"epoch": 0.8344158518734436,
"grad_norm": 3.254610776901245,
"learning_rate": 1.4067194252671157e-06,
"loss": 0.06543638706207275,
"step": 8230
},
{
"epoch": 0.8354297228963761,
"grad_norm": 4.318779468536377,
"learning_rate": 1.3899734666561138e-06,
"loss": 0.08265292644500732,
"step": 8240
},
{
"epoch": 0.8364435939193086,
"grad_norm": 4.862679958343506,
"learning_rate": 1.373320332308541e-06,
"loss": 0.08003490567207336,
"step": 8250
},
{
"epoch": 0.837457464942241,
"grad_norm": 4.257755756378174,
"learning_rate": 1.3567602017609981e-06,
"loss": 0.07166021466255187,
"step": 8260
},
{
"epoch": 0.8384713359651735,
"grad_norm": 4.13678503036499,
"learning_rate": 1.3402932535474277e-06,
"loss": 0.06801553964614868,
"step": 8270
},
{
"epoch": 0.839485206988106,
"grad_norm": 4.516883373260498,
"learning_rate": 1.3239196651971654e-06,
"loss": 0.08092795610427857,
"step": 8280
},
{
"epoch": 0.8404990780110385,
"grad_norm": 3.651750326156616,
"learning_rate": 1.3076396132330516e-06,
"loss": 0.062251228094100955,
"step": 8290
},
{
"epoch": 0.841512949033971,
"grad_norm": 3.7092175483703613,
"learning_rate": 1.2914532731694996e-06,
"loss": 0.060022014379501346,
"step": 8300
},
{
"epoch": 0.8425268200569035,
"grad_norm": 4.45656681060791,
"learning_rate": 1.2753608195106314e-06,
"loss": 0.06412867307662964,
"step": 8310
},
{
"epoch": 0.843540691079836,
"grad_norm": 5.378104209899902,
"learning_rate": 1.2593624257483838e-06,
"loss": 0.07710906267166137,
"step": 8320
},
{
"epoch": 0.8445545621027685,
"grad_norm": 4.533139228820801,
"learning_rate": 1.243458264360634e-06,
"loss": 0.07199961543083191,
"step": 8330
},
{
"epoch": 0.845568433125701,
"grad_norm": 4.633220195770264,
"learning_rate": 1.2276485068093447e-06,
"loss": 0.06943961381912231,
"step": 8340
},
{
"epoch": 0.8465823041486334,
"grad_norm": 4.588976860046387,
"learning_rate": 1.2119333235387232e-06,
"loss": 0.06936721801757813,
"step": 8350
},
{
"epoch": 0.8475961751715659,
"grad_norm": 3.4223034381866455,
"learning_rate": 1.1963128839733674e-06,
"loss": 0.06148537993431091,
"step": 8360
},
{
"epoch": 0.8486100461944985,
"grad_norm": 4.408962726593018,
"learning_rate": 1.1807873565164507e-06,
"loss": 0.07375574111938477,
"step": 8370
},
{
"epoch": 0.849623917217431,
"grad_norm": 4.830119609832764,
"learning_rate": 1.1653569085479067e-06,
"loss": 0.08017903566360474,
"step": 8380
},
{
"epoch": 0.8506377882403635,
"grad_norm": 4.281261444091797,
"learning_rate": 1.150021706422615e-06,
"loss": 0.08258112668991088,
"step": 8390
},
{
"epoch": 0.851651659263296,
"grad_norm": 5.443359851837158,
"learning_rate": 1.1347819154686246e-06,
"loss": 0.07334780693054199,
"step": 8400
},
{
"epoch": 0.8526655302862285,
"grad_norm": 3.4962246417999268,
"learning_rate": 1.1196376999853531e-06,
"loss": 0.06386159062385559,
"step": 8410
},
{
"epoch": 0.853679401309161,
"grad_norm": 3.4677231311798096,
"learning_rate": 1.104589223241831e-06,
"loss": 0.07561634778976441,
"step": 8420
},
{
"epoch": 0.8546932723320935,
"grad_norm": 3.3005125522613525,
"learning_rate": 1.089636647474923e-06,
"loss": 0.07525822520256042,
"step": 8430
},
{
"epoch": 0.855707143355026,
"grad_norm": 4.639155864715576,
"learning_rate": 1.0747801338876006e-06,
"loss": 0.06314131021499633,
"step": 8440
},
{
"epoch": 0.8567210143779584,
"grad_norm": 4.6455278396606445,
"learning_rate": 1.0600198426471908e-06,
"loss": 0.061797142028808594,
"step": 8450
},
{
"epoch": 0.8577348854008909,
"grad_norm": 4.246340751647949,
"learning_rate": 1.0453559328836517e-06,
"loss": 0.07186439633369446,
"step": 8460
},
{
"epoch": 0.8587487564238234,
"grad_norm": 4.872790336608887,
"learning_rate": 1.0307885626878555e-06,
"loss": 0.06340059638023376,
"step": 8470
},
{
"epoch": 0.8597626274467559,
"grad_norm": 4.028145790100098,
"learning_rate": 1.0163178891098912e-06,
"loss": 0.06009650230407715,
"step": 8480
},
{
"epoch": 0.8607764984696884,
"grad_norm": 4.883366584777832,
"learning_rate": 1.0019440681573633e-06,
"loss": 0.06926445960998535,
"step": 8490
},
{
"epoch": 0.8617903694926209,
"grad_norm": 3.5113470554351807,
"learning_rate": 9.876672547937117e-07,
"loss": 0.06527281999588012,
"step": 8500
},
{
"epoch": 0.8628042405155534,
"grad_norm": 4.0587158203125,
"learning_rate": 9.734876029365448e-07,
"loss": 0.060712647438049314,
"step": 8510
},
{
"epoch": 0.8638181115384859,
"grad_norm": 5.724323272705078,
"learning_rate": 9.594052654559737e-07,
"loss": 0.062471520900726316,
"step": 8520
},
{
"epoch": 0.8648319825614184,
"grad_norm": 4.131713390350342,
"learning_rate": 9.454203941729767e-07,
"loss": 0.05695462226867676,
"step": 8530
},
{
"epoch": 0.865845853584351,
"grad_norm": 3.5871546268463135,
"learning_rate": 9.315331398577454e-07,
"loss": 0.0620286226272583,
"step": 8540
},
{
"epoch": 0.8668597246072834,
"grad_norm": 4.538206577301025,
"learning_rate": 9.177436522280714e-07,
"loss": 0.06718026995658874,
"step": 8550
},
{
"epoch": 0.8678735956302159,
"grad_norm": 5.070430278778076,
"learning_rate": 9.04052079947727e-07,
"loss": 0.07798833847045898,
"step": 8560
},
{
"epoch": 0.8688874666531484,
"grad_norm": 4.813652038574219,
"learning_rate": 8.904585706248637e-07,
"loss": 0.06214483976364136,
"step": 8570
},
{
"epoch": 0.8699013376760809,
"grad_norm": 4.205121994018555,
"learning_rate": 8.76963270810427e-07,
"loss": 0.06835060715675353,
"step": 8580
},
{
"epoch": 0.8709152086990134,
"grad_norm": 3.8200302124023438,
"learning_rate": 8.635663259965632e-07,
"loss": 0.07072184681892395,
"step": 8590
},
{
"epoch": 0.8719290797219459,
"grad_norm": 4.4455461502075195,
"learning_rate": 8.502678806150588e-07,
"loss": 0.055426597595214844,
"step": 8600
},
{
"epoch": 0.8729429507448784,
"grad_norm": 3.10363507270813,
"learning_rate": 8.370680780357909e-07,
"loss": 0.06317970752716065,
"step": 8610
},
{
"epoch": 0.8739568217678109,
"grad_norm": 3.0124571323394775,
"learning_rate": 8.23967060565164e-07,
"loss": 0.055988001823425296,
"step": 8620
},
{
"epoch": 0.8749706927907434,
"grad_norm": 4.844436168670654,
"learning_rate": 8.109649694445898e-07,
"loss": 0.05884669423103332,
"step": 8630
},
{
"epoch": 0.8759845638136758,
"grad_norm": 4.722231388092041,
"learning_rate": 7.980619448489601e-07,
"loss": 0.06499350666999817,
"step": 8640
},
{
"epoch": 0.8769984348366083,
"grad_norm": 3.4042773246765137,
"learning_rate": 7.852581258851333e-07,
"loss": 0.05211971998214722,
"step": 8650
},
{
"epoch": 0.8780123058595408,
"grad_norm": 3.5941426753997803,
"learning_rate": 7.725536505904396e-07,
"loss": 0.045614618062973025,
"step": 8660
},
{
"epoch": 0.8790261768824733,
"grad_norm": 4.255887508392334,
"learning_rate": 7.599486559311887e-07,
"loss": 0.058591830730438235,
"step": 8670
},
{
"epoch": 0.8800400479054058,
"grad_norm": 5.157286643981934,
"learning_rate": 7.474432778011953e-07,
"loss": 0.05858051180839539,
"step": 8680
},
{
"epoch": 0.8810539189283383,
"grad_norm": 4.712939262390137,
"learning_rate": 7.350376510203105e-07,
"loss": 0.0570667564868927,
"step": 8690
},
{
"epoch": 0.8820677899512708,
"grad_norm": 5.699842929840088,
"learning_rate": 7.227319093329732e-07,
"loss": 0.07239587306976318,
"step": 8700
},
{
"epoch": 0.8830816609742033,
"grad_norm": 3.7360153198242188,
"learning_rate": 7.105261854067691e-07,
"loss": 0.05815955400466919,
"step": 8710
},
{
"epoch": 0.8840955319971359,
"grad_norm": 3.5813300609588623,
"learning_rate": 6.984206108309932e-07,
"loss": 0.06460291147232056,
"step": 8720
},
{
"epoch": 0.8851094030200684,
"grad_norm": 2.8663229942321777,
"learning_rate": 6.864153161152354e-07,
"loss": 0.04623095989227295,
"step": 8730
},
{
"epoch": 0.8861232740430008,
"grad_norm": 4.130201816558838,
"learning_rate": 6.745104306879769e-07,
"loss": 0.07243728041648864,
"step": 8740
},
{
"epoch": 0.8871371450659333,
"grad_norm": 4.734703063964844,
"learning_rate": 6.627060828951881e-07,
"loss": 0.06279332637786865,
"step": 8750
},
{
"epoch": 0.8881510160888658,
"grad_norm": 3.365182876586914,
"learning_rate": 6.510023999989501e-07,
"loss": 0.0670632541179657,
"step": 8760
},
{
"epoch": 0.8891648871117983,
"grad_norm": 4.467462062835693,
"learning_rate": 6.393995081760795e-07,
"loss": 0.05444055199623108,
"step": 8770
},
{
"epoch": 0.8901787581347308,
"grad_norm": 4.5228424072265625,
"learning_rate": 6.278975325167691e-07,
"loss": 0.05567265748977661,
"step": 8780
},
{
"epoch": 0.8911926291576633,
"grad_norm": 8.498067855834961,
"learning_rate": 6.164965970232417e-07,
"loss": 0.08640431165695191,
"step": 8790
},
{
"epoch": 0.8922065001805958,
"grad_norm": 4.87403678894043,
"learning_rate": 6.051968246084117e-07,
"loss": 0.06385751962661743,
"step": 8800
},
{
"epoch": 0.8932203712035283,
"grad_norm": 4.9440131187438965,
"learning_rate": 5.939983370945535e-07,
"loss": 0.06448200941085816,
"step": 8810
},
{
"epoch": 0.8942342422264608,
"grad_norm": 3.3038432598114014,
"learning_rate": 5.829012552120039e-07,
"loss": 0.06067747473716736,
"step": 8820
},
{
"epoch": 0.8952481132493932,
"grad_norm": 2.8009018898010254,
"learning_rate": 5.719056985978388e-07,
"loss": 0.05692199468612671,
"step": 8830
},
{
"epoch": 0.8962619842723257,
"grad_norm": 4.704399108886719,
"learning_rate": 5.610117857946062e-07,
"loss": 0.05390816926956177,
"step": 8840
},
{
"epoch": 0.8972758552952582,
"grad_norm": 3.707298517227173,
"learning_rate": 5.502196342490318e-07,
"loss": 0.05538478493690491,
"step": 8850
},
{
"epoch": 0.8982897263181907,
"grad_norm": 3.8044562339782715,
"learning_rate": 5.395293603107566e-07,
"loss": 0.0492157518863678,
"step": 8860
},
{
"epoch": 0.8993035973411232,
"grad_norm": 2.7145748138427734,
"learning_rate": 5.2894107923109e-07,
"loss": 0.058959460258483885,
"step": 8870
},
{
"epoch": 0.9003174683640557,
"grad_norm": 5.880849838256836,
"learning_rate": 5.184549051617571e-07,
"loss": 0.06578001976013184,
"step": 8880
},
{
"epoch": 0.9013313393869883,
"grad_norm": 3.389134407043457,
"learning_rate": 5.08070951153673e-07,
"loss": 0.05168793201446533,
"step": 8890
},
{
"epoch": 0.9023452104099208,
"grad_norm": 2.6303889751434326,
"learning_rate": 4.977893291557256e-07,
"loss": 0.051916593313217164,
"step": 8900
},
{
"epoch": 0.9033590814328533,
"grad_norm": 4.501885414123535,
"learning_rate": 4.876101500135621e-07,
"loss": 0.06377041935920716,
"step": 8910
},
{
"epoch": 0.9043729524557858,
"grad_norm": 3.382725477218628,
"learning_rate": 4.775335234684042e-07,
"loss": 0.05729907751083374,
"step": 8920
},
{
"epoch": 0.9053868234787182,
"grad_norm": 5.059771537780762,
"learning_rate": 4.675595581558556e-07,
"loss": 0.05842411518096924,
"step": 8930
},
{
"epoch": 0.9064006945016507,
"grad_norm": 5.198675155639648,
"learning_rate": 4.576883616047345e-07,
"loss": 0.07761399745941162,
"step": 8940
},
{
"epoch": 0.9074145655245832,
"grad_norm": 3.379863739013672,
"learning_rate": 4.479200402359174e-07,
"loss": 0.054069888591766355,
"step": 8950
},
{
"epoch": 0.9084284365475157,
"grad_norm": 4.736568927764893,
"learning_rate": 4.382546993611836e-07,
"loss": 0.057700687646865846,
"step": 8960
},
{
"epoch": 0.9094423075704482,
"grad_norm": 4.177674293518066,
"learning_rate": 4.28692443182086e-07,
"loss": 0.0665976881980896,
"step": 8970
},
{
"epoch": 0.9104561785933807,
"grad_norm": 2.713918924331665,
"learning_rate": 4.192333747888311e-07,
"loss": 0.048197883367538455,
"step": 8980
},
{
"epoch": 0.9114700496163132,
"grad_norm": 4.6662750244140625,
"learning_rate": 4.0987759615915837e-07,
"loss": 0.06197211742401123,
"step": 8990
},
{
"epoch": 0.9124839206392457,
"grad_norm": 5.572573184967041,
"learning_rate": 4.0062520815724816e-07,
"loss": 0.05858563780784607,
"step": 9000
},
{
"epoch": 0.9134977916621781,
"grad_norm": 4.838371753692627,
"learning_rate": 3.9147631053263027e-07,
"loss": 0.052366888523101805,
"step": 9010
},
{
"epoch": 0.9145116626851106,
"grad_norm": 4.843835353851318,
"learning_rate": 3.824310019191102e-07,
"loss": 0.06615509986877441,
"step": 9020
},
{
"epoch": 0.9155255337080431,
"grad_norm": 3.2296056747436523,
"learning_rate": 3.734893798337058e-07,
"loss": 0.05357252359390259,
"step": 9030
},
{
"epoch": 0.9165394047309756,
"grad_norm": 2.8668088912963867,
"learning_rate": 3.646515406755946e-07,
"loss": 0.05796706676483154,
"step": 9040
},
{
"epoch": 0.9175532757539081,
"grad_norm": 4.347914695739746,
"learning_rate": 3.559175797250769e-07,
"loss": 0.053299999237060545,
"step": 9050
},
{
"epoch": 0.9185671467768407,
"grad_norm": 4.839900016784668,
"learning_rate": 3.4728759114254774e-07,
"loss": 0.057792121171951295,
"step": 9060
},
{
"epoch": 0.9195810177997732,
"grad_norm": 4.779026508331299,
"learning_rate": 3.3876166796747877e-07,
"loss": 0.05886713862419128,
"step": 9070
},
{
"epoch": 0.9205948888227057,
"grad_norm": 3.022145986557007,
"learning_rate": 3.303399021174214e-07,
"loss": 0.05389755964279175,
"step": 9080
},
{
"epoch": 0.9216087598456382,
"grad_norm": 4.586666584014893,
"learning_rate": 3.220223843870096e-07,
"loss": 0.045734137296676636,
"step": 9090
},
{
"epoch": 0.9226226308685707,
"grad_norm": 6.984045505523682,
"learning_rate": 3.1380920444698093e-07,
"loss": 0.060455596446990965,
"step": 9100
},
{
"epoch": 0.9236365018915031,
"grad_norm": 4.81547737121582,
"learning_rate": 3.057004508432193e-07,
"loss": 0.047420310974121097,
"step": 9110
},
{
"epoch": 0.9246503729144356,
"grad_norm": 2.596825122833252,
"learning_rate": 2.976962109957882e-07,
"loss": 0.05280296206474304,
"step": 9120
},
{
"epoch": 0.9256642439373681,
"grad_norm": 3.1084349155426025,
"learning_rate": 2.8979657119799667e-07,
"loss": 0.05177220106124878,
"step": 9130
},
{
"epoch": 0.9266781149603006,
"grad_norm": 4.437162399291992,
"learning_rate": 2.8200161661546377e-07,
"loss": 0.05403250455856323,
"step": 9140
},
{
"epoch": 0.9276919859832331,
"grad_norm": 3.970893383026123,
"learning_rate": 2.7431143128520243e-07,
"loss": 0.05441449880599976,
"step": 9150
},
{
"epoch": 0.9287058570061656,
"grad_norm": 5.47275972366333,
"learning_rate": 2.667260981147157e-07,
"loss": 0.04801787137985229,
"step": 9160
},
{
"epoch": 0.9297197280290981,
"grad_norm": 6.234987258911133,
"learning_rate": 2.5924569888109654e-07,
"loss": 0.06317727565765381,
"step": 9170
},
{
"epoch": 0.9307335990520306,
"grad_norm": 3.816516876220703,
"learning_rate": 2.5187031423015597e-07,
"loss": 0.046700885891914366,
"step": 9180
},
{
"epoch": 0.9317474700749631,
"grad_norm": 4.063258647918701,
"learning_rate": 2.4460002367554523e-07,
"loss": 0.04776942729949951,
"step": 9190
},
{
"epoch": 0.9327613410978955,
"grad_norm": 6.303117275238037,
"learning_rate": 2.3743490559789973e-07,
"loss": 0.0622767448425293,
"step": 9200
},
{
"epoch": 0.933775212120828,
"grad_norm": 2.4068617820739746,
"learning_rate": 2.3037503724399946e-07,
"loss": 0.0416769802570343,
"step": 9210
},
{
"epoch": 0.9347890831437605,
"grad_norm": 4.629761219024658,
"learning_rate": 2.2342049472592996e-07,
"loss": 0.05060495138168335,
"step": 9220
},
{
"epoch": 0.9358029541666931,
"grad_norm": 5.992824554443359,
"learning_rate": 2.1657135302026177e-07,
"loss": 0.06019677519798279,
"step": 9230
},
{
"epoch": 0.9368168251896256,
"grad_norm": 4.353598594665527,
"learning_rate": 2.09827685967251e-07,
"loss": 0.04423293471336365,
"step": 9240
},
{
"epoch": 0.9378306962125581,
"grad_norm": 2.8788037300109863,
"learning_rate": 2.0318956627003227e-07,
"loss": 0.03794194459915161,
"step": 9250
},
{
"epoch": 0.9388445672354906,
"grad_norm": 5.680798053741455,
"learning_rate": 1.9665706549384044e-07,
"loss": 0.05324877500534057,
"step": 9260
},
{
"epoch": 0.9398584382584231,
"grad_norm": 5.529104709625244,
"learning_rate": 1.9023025406523787e-07,
"loss": 0.06142445802688599,
"step": 9270
},
{
"epoch": 0.9408723092813556,
"grad_norm": 4.749696731567383,
"learning_rate": 1.8390920127135613e-07,
"loss": 0.05096818208694458,
"step": 9280
},
{
"epoch": 0.9418861803042881,
"grad_norm": 4.975162982940674,
"learning_rate": 1.7769397525914668e-07,
"loss": 0.05544185042381287,
"step": 9290
},
{
"epoch": 0.9429000513272205,
"grad_norm": 3.0535106658935547,
"learning_rate": 1.7158464303465017e-07,
"loss": 0.05222514271736145,
"step": 9300
},
{
"epoch": 0.943913922350153,
"grad_norm": 6.289915084838867,
"learning_rate": 1.6558127046226945e-07,
"loss": 0.05867892503738403,
"step": 9310
},
{
"epoch": 0.9449277933730855,
"grad_norm": 5.3161115646362305,
"learning_rate": 1.5968392226406337e-07,
"loss": 0.0561317503452301,
"step": 9320
},
{
"epoch": 0.945941664396018,
"grad_norm": 7.5520524978637695,
"learning_rate": 1.5389266201904506e-07,
"loss": 0.04963110089302063,
"step": 9330
},
{
"epoch": 0.9469555354189505,
"grad_norm": 2.940347909927368,
"learning_rate": 1.482075521625026e-07,
"loss": 0.048455798625946046,
"step": 9340
},
{
"epoch": 0.947969406441883,
"grad_norm": 4.4893269538879395,
"learning_rate": 1.4262865398531834e-07,
"loss": 0.04709072411060333,
"step": 9350
},
{
"epoch": 0.9489832774648155,
"grad_norm": 5.17371940612793,
"learning_rate": 1.3715602763331172e-07,
"loss": 0.05423116087913513,
"step": 9360
},
{
"epoch": 0.949997148487748,
"grad_norm": 4.860888481140137,
"learning_rate": 1.3178973210659528e-07,
"loss": 0.054334777593612674,
"step": 9370
},
{
"epoch": 0.9510110195106805,
"grad_norm": 4.819238185882568,
"learning_rate": 1.2652982525892975e-07,
"loss": 0.05044026374816894,
"step": 9380
},
{
"epoch": 0.9520248905336129,
"grad_norm": 6.42965030670166,
"learning_rate": 1.213763637971066e-07,
"loss": 0.060467517375946044,
"step": 9390
},
{
"epoch": 0.9530387615565454,
"grad_norm": 5.912594318389893,
"learning_rate": 1.1632940328033526e-07,
"loss": 0.05043342113494873,
"step": 9400
},
{
"epoch": 0.954052632579478,
"grad_norm": 6.430222511291504,
"learning_rate": 1.1138899811964477e-07,
"loss": 0.05034192204475403,
"step": 9410
},
{
"epoch": 0.9550665036024105,
"grad_norm": 5.617626667022705,
"learning_rate": 1.0655520157729194e-07,
"loss": 0.043424731492996214,
"step": 9420
},
{
"epoch": 0.956080374625343,
"grad_norm": 2.405336856842041,
"learning_rate": 1.0182806576619742e-07,
"loss": 0.05420509576797485,
"step": 9430
},
{
"epoch": 0.9570942456482755,
"grad_norm": 6.126251697540283,
"learning_rate": 9.72076416493728e-08,
"loss": 0.0422305703163147,
"step": 9440
},
{
"epoch": 0.958108116671208,
"grad_norm": 5.654440879821777,
"learning_rate": 9.269397903937883e-08,
"loss": 0.057630491256713864,
"step": 9450
},
{
"epoch": 0.9591219876941405,
"grad_norm": 4.640891075134277,
"learning_rate": 8.828712659778471e-08,
"loss": 0.048209396004676816,
"step": 9460
},
{
"epoch": 0.960135858717073,
"grad_norm": 5.749047756195068,
"learning_rate": 8.398713183464857e-08,
"loss": 0.03655834794044495,
"step": 9470
},
{
"epoch": 0.9611497297400055,
"grad_norm": 6.037595748901367,
"learning_rate": 7.979404110799338e-08,
"loss": 0.059028875827789304,
"step": 9480
},
{
"epoch": 0.9621636007629379,
"grad_norm": 2.625040292739868,
"learning_rate": 7.570789962332181e-08,
"loss": 0.042799767851829526,
"step": 9490
},
{
"epoch": 0.9631774717858704,
"grad_norm": 4.518804550170898,
"learning_rate": 7.172875143311996e-08,
"loss": 0.052408069372177124,
"step": 9500
},
{
"epoch": 0.9641913428088029,
"grad_norm": 6.310920715332031,
"learning_rate": 6.785663943638333e-08,
"loss": 0.05600939989089966,
"step": 9510
},
{
"epoch": 0.9652052138317354,
"grad_norm": 3.3058269023895264,
"learning_rate": 6.409160537815818e-08,
"loss": 0.04984386563301087,
"step": 9520
},
{
"epoch": 0.9662190848546679,
"grad_norm": 5.386301517486572,
"learning_rate": 6.043368984908759e-08,
"loss": 0.059482693672180176,
"step": 9530
},
{
"epoch": 0.9672329558776004,
"grad_norm": 4.383225440979004,
"learning_rate": 5.688293228497399e-08,
"loss": 0.05753294825553894,
"step": 9540
},
{
"epoch": 0.9682468269005329,
"grad_norm": 5.520195007324219,
"learning_rate": 5.34393709663561e-08,
"loss": 0.04877977967262268,
"step": 9550
},
{
"epoch": 0.9692606979234654,
"grad_norm": 3.361797571182251,
"learning_rate": 5.0103043018097096e-08,
"loss": 0.05434563755989075,
"step": 9560
},
{
"epoch": 0.9702745689463979,
"grad_norm": 3.9871559143066406,
"learning_rate": 4.687398440897828e-08,
"loss": 0.05255262851715088,
"step": 9570
},
{
"epoch": 0.9712884399693305,
"grad_norm": 3.269758462905884,
"learning_rate": 4.3752229951321553e-08,
"loss": 0.052621197700500486,
"step": 9580
},
{
"epoch": 0.9723023109922629,
"grad_norm": 4.416233539581299,
"learning_rate": 4.0737813300602e-08,
"loss": 0.05528765320777893,
"step": 9590
},
{
"epoch": 0.9733161820151954,
"grad_norm": 3.4947350025177,
"learning_rate": 3.783076695509702e-08,
"loss": 0.048863542079925534,
"step": 9600
},
{
"epoch": 0.9743300530381279,
"grad_norm": 3.3298087120056152,
"learning_rate": 3.5031122255524406e-08,
"loss": 0.04724987149238587,
"step": 9610
},
{
"epoch": 0.9753439240610604,
"grad_norm": 3.323794364929199,
"learning_rate": 3.233890938471373e-08,
"loss": 0.060399866104125975,
"step": 9620
},
{
"epoch": 0.9763577950839929,
"grad_norm": 2.9769816398620605,
"learning_rate": 2.9754157367277715e-08,
"loss": 0.056619775295257566,
"step": 9630
},
{
"epoch": 0.9773716661069254,
"grad_norm": 5.619937419891357,
"learning_rate": 2.72768940692969e-08,
"loss": 0.046914905309677124,
"step": 9640
},
{
"epoch": 0.9783855371298579,
"grad_norm": 4.764586925506592,
"learning_rate": 2.4907146198023256e-08,
"loss": 0.04930989146232605,
"step": 9650
},
{
"epoch": 0.9793994081527904,
"grad_norm": 7.0540666580200195,
"learning_rate": 2.2644939301588174e-08,
"loss": 0.05243293046951294,
"step": 9660
},
{
"epoch": 0.9804132791757229,
"grad_norm": 4.6112895011901855,
"learning_rate": 2.049029776873268e-08,
"loss": 0.03863253891468048,
"step": 9670
},
{
"epoch": 0.9814271501986553,
"grad_norm": 2.8966825008392334,
"learning_rate": 1.8443244828536544e-08,
"loss": 0.055679994821548465,
"step": 9680
},
{
"epoch": 0.9824410212215878,
"grad_norm": 4.302661418914795,
"learning_rate": 1.650380255017625e-08,
"loss": 0.05167467594146728,
"step": 9690
},
{
"epoch": 0.9834548922445203,
"grad_norm": 2.8718514442443848,
"learning_rate": 1.4671991842680756e-08,
"loss": 0.04157778024673462,
"step": 9700
},
{
"epoch": 0.9844687632674528,
"grad_norm": 4.957472324371338,
"learning_rate": 1.294783245470943e-08,
"loss": 0.05514119863510132,
"step": 9710
},
{
"epoch": 0.9854826342903853,
"grad_norm": 3.3154122829437256,
"learning_rate": 1.13313429743378e-08,
"loss": 0.04565766751766205,
"step": 9720
},
{
"epoch": 0.9864965053133178,
"grad_norm": 4.853787899017334,
"learning_rate": 9.822540828858807e-09,
"loss": 0.059190398454666136,
"step": 9730
},
{
"epoch": 0.9875103763362503,
"grad_norm": 3.376783847808838,
"learning_rate": 8.421442284591852e-09,
"loss": 0.055718207359313966,
"step": 9740
},
{
"epoch": 0.9885242473591829,
"grad_norm": 3.923891305923462,
"learning_rate": 7.128062446709605e-09,
"loss": 0.04465946555137634,
"step": 9750
},
{
"epoch": 0.9895381183821154,
"grad_norm": 5.462875843048096,
"learning_rate": 5.942415259074796e-09,
"loss": 0.0454116016626358,
"step": 9760
},
{
"epoch": 0.9905519894050479,
"grad_norm": 4.120839595794678,
"learning_rate": 4.86451350409145e-09,
"loss": 0.05331336259841919,
"step": 9770
},
{
"epoch": 0.9915658604279803,
"grad_norm": 4.6509270668029785,
"learning_rate": 3.894368802563886e-09,
"loss": 0.04725489020347595,
"step": 9780
},
{
"epoch": 0.9925797314509128,
"grad_norm": 2.7159740924835205,
"learning_rate": 3.0319916135723713e-09,
"loss": 0.05010777711868286,
"step": 9790
},
{
"epoch": 0.9935936024738453,
"grad_norm": 5.265696048736572,
"learning_rate": 2.277391234363213e-09,
"loss": 0.045841658115386964,
"step": 9800
},
{
"epoch": 0.9946074734967778,
"grad_norm": 6.023398399353027,
"learning_rate": 1.6305758002443939e-09,
"loss": 0.04846552610397339,
"step": 9810
},
{
"epoch": 0.9956213445197103,
"grad_norm": 4.8127946853637695,
"learning_rate": 1.091552284500086e-09,
"loss": 0.057257080078125,
"step": 9820
},
{
"epoch": 0.9966352155426428,
"grad_norm": 5.342715263366699,
"learning_rate": 6.603264983173762e-10,
"loss": 0.057232248783111575,
"step": 9830
},
{
"epoch": 0.9976490865655753,
"grad_norm": 3.8298466205596924,
"learning_rate": 3.3690309071965355e-10,
"loss": 0.0546303391456604,
"step": 9840
},
{
"epoch": 0.9986629575885078,
"grad_norm": 3.6711020469665527,
"learning_rate": 1.2128554851664754e-10,
"loss": 0.044484928250312805,
"step": 9850
},
{
"epoch": 0.9996768286114402,
"grad_norm": 3.259277820587158,
"learning_rate": 1.3476196272232955e-11,
"loss": 0.04408583641052246,
"step": 9860
},
{
"epoch": 1.0,
"step": 9864,
"total_flos": 5.070703459485876e+19,
"train_loss": 0.13390739868785537,
"train_runtime": 226718.4641,
"train_samples_per_second": 2.784,
"train_steps_per_second": 0.044
}
],
"logging_steps": 10,
"max_steps": 9864,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5.070703459485876e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}