Orbit-3-Nano / checkpoints /checkpoint-6400 /trainer_state.json
titanthree's picture
Add checkpoint-6400
7ba5613 verified
Raw History Blame Contribute Delete
119 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.5406907502977374,
"eval_steps": 500,
"global_step": 6400,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003969829297340214,
"grad_norm": 0.6961175799369812,
"learning_rate": 0.00016,
"loss": 3.2075271606445312,
"step": 10
},
{
"epoch": 0.007939658594680429,
"grad_norm": 0.9305199384689331,
"learning_rate": 0.00019978799522989268,
"loss": 3.1885648727416993,
"step": 20
},
{
"epoch": 0.011909487892020643,
"grad_norm": 0.8452768325805664,
"learning_rate": 0.00019952298926725852,
"loss": 2.9959283828735352,
"step": 30
},
{
"epoch": 0.015879317189360857,
"grad_norm": 0.9384715557098389,
"learning_rate": 0.00019925798330462438,
"loss": 3.132890510559082,
"step": 40
},
{
"epoch": 0.019849146486701073,
"grad_norm": 0.9872943162918091,
"learning_rate": 0.0001989929773419902,
"loss": 3.1314294815063475,
"step": 50
},
{
"epoch": 0.023818975784041286,
"grad_norm": 0.8298355937004089,
"learning_rate": 0.00019872797137935604,
"loss": 3.24124755859375,
"step": 60
},
{
"epoch": 0.027788805081381502,
"grad_norm": 0.8308970928192139,
"learning_rate": 0.00019846296541672188,
"loss": 3.0996198654174805,
"step": 70
},
{
"epoch": 0.031758634378721715,
"grad_norm": 0.8442180156707764,
"learning_rate": 0.0001981979594540877,
"loss": 3.0475082397460938,
"step": 80
},
{
"epoch": 0.03572846367606193,
"grad_norm": 0.9131914377212524,
"learning_rate": 0.00019793295349145357,
"loss": 3.1756013870239257,
"step": 90
},
{
"epoch": 0.03969829297340215,
"grad_norm": 0.9973925352096558,
"learning_rate": 0.0001976679475288194,
"loss": 3.159317398071289,
"step": 100
},
{
"epoch": 0.043668122270742356,
"grad_norm": 0.9906535148620605,
"learning_rate": 0.00019740294156618524,
"loss": 3.086680603027344,
"step": 110
},
{
"epoch": 0.04763795156808257,
"grad_norm": 0.8675328493118286,
"learning_rate": 0.0001971379356035511,
"loss": 3.0809303283691407,
"step": 120
},
{
"epoch": 0.05160778086542279,
"grad_norm": 1.2756576538085938,
"learning_rate": 0.00019687292964091694,
"loss": 2.965633010864258,
"step": 130
},
{
"epoch": 0.055577610162763004,
"grad_norm": 0.8540228605270386,
"learning_rate": 0.00019660792367828277,
"loss": 3.11167049407959,
"step": 140
},
{
"epoch": 0.05954743946010321,
"grad_norm": 0.8471305966377258,
"learning_rate": 0.0001963429177156486,
"loss": 2.9493082046508787,
"step": 150
},
{
"epoch": 0.06351726875744343,
"grad_norm": 1.0690556764602661,
"learning_rate": 0.00019607791175301444,
"loss": 3.1130550384521483,
"step": 160
},
{
"epoch": 0.06748709805478365,
"grad_norm": 0.9682918787002563,
"learning_rate": 0.0001958129057903803,
"loss": 3.041396713256836,
"step": 170
},
{
"epoch": 0.07145692735212386,
"grad_norm": 0.9328762292861938,
"learning_rate": 0.00019554789982774613,
"loss": 3.1088594436645507,
"step": 180
},
{
"epoch": 0.07542675664946408,
"grad_norm": 0.981307327747345,
"learning_rate": 0.000195282893865112,
"loss": 3.1146469116210938,
"step": 190
},
{
"epoch": 0.0793965859468043,
"grad_norm": 1.0752034187316895,
"learning_rate": 0.00019501788790247783,
"loss": 3.0602563858032226,
"step": 200
},
{
"epoch": 0.0833664152441445,
"grad_norm": 0.8543508052825928,
"learning_rate": 0.00019475288193984363,
"loss": 2.989186477661133,
"step": 210
},
{
"epoch": 0.08733624454148471,
"grad_norm": 0.9234327077865601,
"learning_rate": 0.0001944878759772095,
"loss": 3.0583696365356445,
"step": 220
},
{
"epoch": 0.09130607383882493,
"grad_norm": 0.8820871710777283,
"learning_rate": 0.00019422287001457533,
"loss": 2.868353271484375,
"step": 230
},
{
"epoch": 0.09527590313616514,
"grad_norm": 0.9947519898414612,
"learning_rate": 0.0001939578640519412,
"loss": 3.110881042480469,
"step": 240
},
{
"epoch": 0.09924573243350536,
"grad_norm": 0.8357694745063782,
"learning_rate": 0.00019369285808930702,
"loss": 3.01332950592041,
"step": 250
},
{
"epoch": 0.10321556173084558,
"grad_norm": 0.9128127694129944,
"learning_rate": 0.00019342785212667288,
"loss": 3.149305725097656,
"step": 260
},
{
"epoch": 0.10718539102818579,
"grad_norm": 0.9448281526565552,
"learning_rate": 0.0001931628461640387,
"loss": 3.172308921813965,
"step": 270
},
{
"epoch": 0.11115522032552601,
"grad_norm": 1.0875459909439087,
"learning_rate": 0.00019289784020140452,
"loss": 3.1815589904785155,
"step": 280
},
{
"epoch": 0.11512504962286621,
"grad_norm": 0.9254504442214966,
"learning_rate": 0.00019263283423877038,
"loss": 3.1059280395507813,
"step": 290
},
{
"epoch": 0.11909487892020643,
"grad_norm": 0.9273567795753479,
"learning_rate": 0.00019236782827613622,
"loss": 3.063690757751465,
"step": 300
},
{
"epoch": 0.12306470821754664,
"grad_norm": 0.9300341010093689,
"learning_rate": 0.00019210282231350208,
"loss": 2.989412307739258,
"step": 310
},
{
"epoch": 0.12703453751488686,
"grad_norm": 1.092647910118103,
"learning_rate": 0.0001918378163508679,
"loss": 2.9342714309692384,
"step": 320
},
{
"epoch": 0.13100436681222707,
"grad_norm": 0.7696720361709595,
"learning_rate": 0.00019157281038823375,
"loss": 2.966893768310547,
"step": 330
},
{
"epoch": 0.1349741961095673,
"grad_norm": 0.9847505688667297,
"learning_rate": 0.00019130780442559958,
"loss": 3.054021644592285,
"step": 340
},
{
"epoch": 0.1389440254069075,
"grad_norm": 1.0354719161987305,
"learning_rate": 0.00019104279846296541,
"loss": 2.9900562286376955,
"step": 350
},
{
"epoch": 0.14291385470424772,
"grad_norm": 0.8888474106788635,
"learning_rate": 0.00019077779250033128,
"loss": 3.1933818817138673,
"step": 360
},
{
"epoch": 0.14688368400158794,
"grad_norm": 1.0046141147613525,
"learning_rate": 0.0001905127865376971,
"loss": 2.920476722717285,
"step": 370
},
{
"epoch": 0.15085351329892815,
"grad_norm": 1.039939522743225,
"learning_rate": 0.00019024778057506294,
"loss": 3.1360284805297853,
"step": 380
},
{
"epoch": 0.15482334259626837,
"grad_norm": 0.9551829099655151,
"learning_rate": 0.0001899827746124288,
"loss": 3.144388198852539,
"step": 390
},
{
"epoch": 0.1587931718936086,
"grad_norm": 0.9663012027740479,
"learning_rate": 0.00018971776864979464,
"loss": 2.9875814437866213,
"step": 400
},
{
"epoch": 0.1627630011909488,
"grad_norm": 0.9962781071662903,
"learning_rate": 0.00018945276268716047,
"loss": 3.0062061309814454,
"step": 410
},
{
"epoch": 0.166732830488289,
"grad_norm": 1.001230001449585,
"learning_rate": 0.0001891877567245263,
"loss": 3.0834712982177734,
"step": 420
},
{
"epoch": 0.1707026597856292,
"grad_norm": 0.9189774990081787,
"learning_rate": 0.00018892275076189214,
"loss": 2.885240936279297,
"step": 430
},
{
"epoch": 0.17467248908296942,
"grad_norm": 1.0268958806991577,
"learning_rate": 0.000188657744799258,
"loss": 3.1581113815307615,
"step": 440
},
{
"epoch": 0.17864231838030964,
"grad_norm": 0.9994209408760071,
"learning_rate": 0.00018839273883662383,
"loss": 2.8793590545654295,
"step": 450
},
{
"epoch": 0.18261214767764986,
"grad_norm": 1.0613411664962769,
"learning_rate": 0.0001881277328739897,
"loss": 3.029997444152832,
"step": 460
},
{
"epoch": 0.18658197697499007,
"grad_norm": 1.0501508712768555,
"learning_rate": 0.0001878627269113555,
"loss": 3.150674819946289,
"step": 470
},
{
"epoch": 0.1905518062723303,
"grad_norm": 1.0706281661987305,
"learning_rate": 0.00018759772094872134,
"loss": 3.0236751556396486,
"step": 480
},
{
"epoch": 0.1945216355696705,
"grad_norm": 0.856622040271759,
"learning_rate": 0.0001873327149860872,
"loss": 2.9095502853393556,
"step": 490
},
{
"epoch": 0.19849146486701072,
"grad_norm": 0.9528838396072388,
"learning_rate": 0.00018706770902345303,
"loss": 3.001302146911621,
"step": 500
},
{
"epoch": 0.20246129416435094,
"grad_norm": 1.0344288349151611,
"learning_rate": 0.0001868027030608189,
"loss": 3.106768035888672,
"step": 510
},
{
"epoch": 0.20643112346169115,
"grad_norm": 0.9366422891616821,
"learning_rate": 0.00018653769709818473,
"loss": 2.99255428314209,
"step": 520
},
{
"epoch": 0.21040095275903137,
"grad_norm": 0.9982316493988037,
"learning_rate": 0.00018627269113555056,
"loss": 2.904929351806641,
"step": 530
},
{
"epoch": 0.21437078205637158,
"grad_norm": 1.0310651063919067,
"learning_rate": 0.0001860076851729164,
"loss": 3.076841926574707,
"step": 540
},
{
"epoch": 0.2183406113537118,
"grad_norm": 1.015830397605896,
"learning_rate": 0.00018574267921028223,
"loss": 2.995553398132324,
"step": 550
},
{
"epoch": 0.22231044065105202,
"grad_norm": 0.9116207957267761,
"learning_rate": 0.0001854776732476481,
"loss": 3.0907325744628906,
"step": 560
},
{
"epoch": 0.22628026994839223,
"grad_norm": 0.983268141746521,
"learning_rate": 0.00018521266728501392,
"loss": 2.9128732681274414,
"step": 570
},
{
"epoch": 0.23025009924573242,
"grad_norm": 1.0083118677139282,
"learning_rate": 0.00018494766132237978,
"loss": 3.06164665222168,
"step": 580
},
{
"epoch": 0.23421992854307264,
"grad_norm": 0.9371793270111084,
"learning_rate": 0.00018468265535974562,
"loss": 3.110069465637207,
"step": 590
},
{
"epoch": 0.23818975784041285,
"grad_norm": 0.8907871842384338,
"learning_rate": 0.00018441764939711142,
"loss": 2.9918779373168944,
"step": 600
},
{
"epoch": 0.24215958713775307,
"grad_norm": 1.2298108339309692,
"learning_rate": 0.00018415264343447728,
"loss": 2.995768737792969,
"step": 610
},
{
"epoch": 0.24612941643509328,
"grad_norm": 1.1198155879974365,
"learning_rate": 0.00018388763747184312,
"loss": 2.905741500854492,
"step": 620
},
{
"epoch": 0.25009924573243353,
"grad_norm": 0.9270941019058228,
"learning_rate": 0.00018362263150920898,
"loss": 3.0602294921875,
"step": 630
},
{
"epoch": 0.2540690750297737,
"grad_norm": 0.907201886177063,
"learning_rate": 0.0001833576255465748,
"loss": 3.0065471649169924,
"step": 640
},
{
"epoch": 0.25803890432711396,
"grad_norm": 1.0519545078277588,
"learning_rate": 0.00018309261958394065,
"loss": 3.029597282409668,
"step": 650
},
{
"epoch": 0.26200873362445415,
"grad_norm": 0.8725332617759705,
"learning_rate": 0.00018282761362130648,
"loss": 2.8936578750610353,
"step": 660
},
{
"epoch": 0.26597856292179434,
"grad_norm": 1.040444254875183,
"learning_rate": 0.0001825626076586723,
"loss": 3.081193542480469,
"step": 670
},
{
"epoch": 0.2699483922191346,
"grad_norm": 0.9259696006774902,
"learning_rate": 0.00018229760169603817,
"loss": 2.9740297317504885,
"step": 680
},
{
"epoch": 0.27391822151647477,
"grad_norm": 1.0812491178512573,
"learning_rate": 0.000182032595733404,
"loss": 3.0157501220703127,
"step": 690
},
{
"epoch": 0.277888050813815,
"grad_norm": 1.0642049312591553,
"learning_rate": 0.00018176758977076984,
"loss": 2.996814727783203,
"step": 700
},
{
"epoch": 0.2818578801111552,
"grad_norm": 1.0878177881240845,
"learning_rate": 0.0001815025838081357,
"loss": 3.0396562576293946,
"step": 710
},
{
"epoch": 0.28582770940849545,
"grad_norm": 0.8735718131065369,
"learning_rate": 0.00018123757784550154,
"loss": 2.8770545959472655,
"step": 720
},
{
"epoch": 0.28979753870583563,
"grad_norm": 0.9658878445625305,
"learning_rate": 0.00018097257188286737,
"loss": 3.050862121582031,
"step": 730
},
{
"epoch": 0.2937673680031759,
"grad_norm": 0.958960771560669,
"learning_rate": 0.0001807075659202332,
"loss": 3.0131216049194336,
"step": 740
},
{
"epoch": 0.29773719730051607,
"grad_norm": 1.0989668369293213,
"learning_rate": 0.00018044255995759904,
"loss": 3.0553775787353517,
"step": 750
},
{
"epoch": 0.3017070265978563,
"grad_norm": 0.9090871214866638,
"learning_rate": 0.0001801775539949649,
"loss": 2.9768255233764647,
"step": 760
},
{
"epoch": 0.3056768558951965,
"grad_norm": 0.9842050671577454,
"learning_rate": 0.00017991254803233073,
"loss": 3.028487205505371,
"step": 770
},
{
"epoch": 0.30964668519253674,
"grad_norm": 1.1661285161972046,
"learning_rate": 0.0001796475420696966,
"loss": 2.9407962799072265,
"step": 780
},
{
"epoch": 0.31361651448987693,
"grad_norm": 1.0115314722061157,
"learning_rate": 0.00017938253610706243,
"loss": 3.02266845703125,
"step": 790
},
{
"epoch": 0.3175863437872172,
"grad_norm": 1.1729583740234375,
"learning_rate": 0.00017911753014442823,
"loss": 2.9814245223999025,
"step": 800
},
{
"epoch": 0.32155617308455736,
"grad_norm": 0.9609226584434509,
"learning_rate": 0.0001788525241817941,
"loss": 2.8289106369018553,
"step": 810
},
{
"epoch": 0.3255260023818976,
"grad_norm": 1.339521050453186,
"learning_rate": 0.00017858751821915993,
"loss": 3.0263460159301756,
"step": 820
},
{
"epoch": 0.3294958316792378,
"grad_norm": 0.8593433499336243,
"learning_rate": 0.0001783225122565258,
"loss": 3.0060649871826173,
"step": 830
},
{
"epoch": 0.333465660976578,
"grad_norm": 1.0417287349700928,
"learning_rate": 0.00017805750629389162,
"loss": 3.0265331268310547,
"step": 840
},
{
"epoch": 0.3374354902739182,
"grad_norm": 1.1134616136550903,
"learning_rate": 0.00017779250033125748,
"loss": 3.057025146484375,
"step": 850
},
{
"epoch": 0.3414053195712584,
"grad_norm": 1.0823781490325928,
"learning_rate": 0.0001775274943686233,
"loss": 3.0840169906616213,
"step": 860
},
{
"epoch": 0.34537514886859866,
"grad_norm": 1.304033637046814,
"learning_rate": 0.00017726248840598913,
"loss": 2.9510980606079102,
"step": 870
},
{
"epoch": 0.34934497816593885,
"grad_norm": 0.9548248052597046,
"learning_rate": 0.00017699748244335499,
"loss": 3.0062685012817383,
"step": 880
},
{
"epoch": 0.3533148074632791,
"grad_norm": 0.9844130277633667,
"learning_rate": 0.00017673247648072082,
"loss": 2.9815792083740233,
"step": 890
},
{
"epoch": 0.3572846367606193,
"grad_norm": 1.1588873863220215,
"learning_rate": 0.00017646747051808668,
"loss": 3.143498992919922,
"step": 900
},
{
"epoch": 0.3612544660579595,
"grad_norm": 0.9818010926246643,
"learning_rate": 0.00017620246455545251,
"loss": 2.973097229003906,
"step": 910
},
{
"epoch": 0.3652242953552997,
"grad_norm": 1.0033776760101318,
"learning_rate": 0.00017593745859281835,
"loss": 3.0639650344848635,
"step": 920
},
{
"epoch": 0.36919412465263995,
"grad_norm": 1.1256245374679565,
"learning_rate": 0.00017567245263018418,
"loss": 3.0972034454345705,
"step": 930
},
{
"epoch": 0.37316395394998014,
"grad_norm": 1.1515825986862183,
"learning_rate": 0.00017540744666755002,
"loss": 2.954210090637207,
"step": 940
},
{
"epoch": 0.3771337832473204,
"grad_norm": 1.0679875612258911,
"learning_rate": 0.00017514244070491588,
"loss": 2.9248802185058596,
"step": 950
},
{
"epoch": 0.3811036125446606,
"grad_norm": 1.3141388893127441,
"learning_rate": 0.0001748774347422817,
"loss": 2.9732879638671874,
"step": 960
},
{
"epoch": 0.3850734418420008,
"grad_norm": 0.9422619938850403,
"learning_rate": 0.00017461242877964754,
"loss": 2.8527624130249025,
"step": 970
},
{
"epoch": 0.389043271139341,
"grad_norm": 1.1796400547027588,
"learning_rate": 0.0001743474228170134,
"loss": 3.081935691833496,
"step": 980
},
{
"epoch": 0.3930131004366812,
"grad_norm": 0.9901096820831299,
"learning_rate": 0.00017408241685437924,
"loss": 2.9871950149536133,
"step": 990
},
{
"epoch": 0.39698292973402144,
"grad_norm": 0.9635742902755737,
"learning_rate": 0.00017381741089174507,
"loss": 2.909952735900879,
"step": 1000
},
{
"epoch": 0.4009527590313616,
"grad_norm": 1.094767451286316,
"learning_rate": 0.0001735524049291109,
"loss": 2.886246681213379,
"step": 1010
},
{
"epoch": 0.40492258832870187,
"grad_norm": 0.9883092045783997,
"learning_rate": 0.00017328739896647674,
"loss": 2.941468048095703,
"step": 1020
},
{
"epoch": 0.40889241762604206,
"grad_norm": 1.082886815071106,
"learning_rate": 0.0001730223930038426,
"loss": 2.8721885681152344,
"step": 1030
},
{
"epoch": 0.4128622469233823,
"grad_norm": 0.9654253125190735,
"learning_rate": 0.00017275738704120844,
"loss": 3.0437408447265626,
"step": 1040
},
{
"epoch": 0.4168320762207225,
"grad_norm": 0.9620010256767273,
"learning_rate": 0.0001724923810785743,
"loss": 2.9390357971191405,
"step": 1050
},
{
"epoch": 0.42080190551806274,
"grad_norm": 1.042639136314392,
"learning_rate": 0.0001722273751159401,
"loss": 2.928664207458496,
"step": 1060
},
{
"epoch": 0.4247717348154029,
"grad_norm": 1.1499779224395752,
"learning_rate": 0.00017196236915330594,
"loss": 2.9631744384765626,
"step": 1070
},
{
"epoch": 0.42874156411274317,
"grad_norm": 1.0400148630142212,
"learning_rate": 0.0001716973631906718,
"loss": 2.9788026809692383,
"step": 1080
},
{
"epoch": 0.43271139341008336,
"grad_norm": 1.0587307214736938,
"learning_rate": 0.00017143235722803763,
"loss": 2.9872905731201174,
"step": 1090
},
{
"epoch": 0.4366812227074236,
"grad_norm": 0.9863898158073425,
"learning_rate": 0.0001711673512654035,
"loss": 2.959150505065918,
"step": 1100
},
{
"epoch": 0.4406510520047638,
"grad_norm": 1.024854302406311,
"learning_rate": 0.00017090234530276933,
"loss": 2.906721496582031,
"step": 1110
},
{
"epoch": 0.44462088130210403,
"grad_norm": 1.2431564331054688,
"learning_rate": 0.00017063733934013516,
"loss": 2.959243583679199,
"step": 1120
},
{
"epoch": 0.4485907105994442,
"grad_norm": 1.0118112564086914,
"learning_rate": 0.000170372333377501,
"loss": 2.837575149536133,
"step": 1130
},
{
"epoch": 0.45256053989678446,
"grad_norm": 1.1883628368377686,
"learning_rate": 0.00017010732741486683,
"loss": 3.088572692871094,
"step": 1140
},
{
"epoch": 0.45653036919412465,
"grad_norm": 0.902087926864624,
"learning_rate": 0.0001698423214522327,
"loss": 2.870463180541992,
"step": 1150
},
{
"epoch": 0.46050019849146484,
"grad_norm": 1.1765600442886353,
"learning_rate": 0.00016957731548959852,
"loss": 2.907725143432617,
"step": 1160
},
{
"epoch": 0.4644700277888051,
"grad_norm": 0.9849786162376404,
"learning_rate": 0.00016931230952696438,
"loss": 2.975267791748047,
"step": 1170
},
{
"epoch": 0.4684398570861453,
"grad_norm": 1.0879027843475342,
"learning_rate": 0.00016904730356433022,
"loss": 2.9905019760131837,
"step": 1180
},
{
"epoch": 0.4724096863834855,
"grad_norm": 1.159298300743103,
"learning_rate": 0.00016878229760169605,
"loss": 2.9706029891967773,
"step": 1190
},
{
"epoch": 0.4763795156808257,
"grad_norm": 1.0128686428070068,
"learning_rate": 0.00016851729163906189,
"loss": 2.929449462890625,
"step": 1200
},
{
"epoch": 0.48034934497816595,
"grad_norm": 1.3239126205444336,
"learning_rate": 0.00016825228567642772,
"loss": 2.854958343505859,
"step": 1210
},
{
"epoch": 0.48431917427550614,
"grad_norm": 1.1641817092895508,
"learning_rate": 0.00016798727971379358,
"loss": 2.8924772262573244,
"step": 1220
},
{
"epoch": 0.4882890035728464,
"grad_norm": 1.0055742263793945,
"learning_rate": 0.00016772227375115941,
"loss": 2.9655168533325194,
"step": 1230
},
{
"epoch": 0.49225883287018657,
"grad_norm": 1.029150128364563,
"learning_rate": 0.00016745726778852525,
"loss": 2.953019905090332,
"step": 1240
},
{
"epoch": 0.4962286621675268,
"grad_norm": 1.0731725692749023,
"learning_rate": 0.00016719226182589108,
"loss": 2.8922054290771486,
"step": 1250
},
{
"epoch": 0.5001984914648671,
"grad_norm": 1.1343021392822266,
"learning_rate": 0.00016692725586325692,
"loss": 2.944770622253418,
"step": 1260
},
{
"epoch": 0.5041683207622072,
"grad_norm": 1.2642573118209839,
"learning_rate": 0.00016666224990062278,
"loss": 3.03311710357666,
"step": 1270
},
{
"epoch": 0.5081381500595474,
"grad_norm": 1.0998749732971191,
"learning_rate": 0.0001663972439379886,
"loss": 2.901491165161133,
"step": 1280
},
{
"epoch": 0.5121079793568877,
"grad_norm": 1.1051859855651855,
"learning_rate": 0.00016613223797535444,
"loss": 3.012837600708008,
"step": 1290
},
{
"epoch": 0.5160778086542279,
"grad_norm": 1.2365621328353882,
"learning_rate": 0.0001658672320127203,
"loss": 2.9026336669921875,
"step": 1300
},
{
"epoch": 0.520047637951568,
"grad_norm": 1.0339555740356445,
"learning_rate": 0.00016560222605008614,
"loss": 2.922551727294922,
"step": 1310
},
{
"epoch": 0.5240174672489083,
"grad_norm": 1.0754799842834473,
"learning_rate": 0.00016533722008745197,
"loss": 3.07440071105957,
"step": 1320
},
{
"epoch": 0.5279872965462485,
"grad_norm": 0.9746431708335876,
"learning_rate": 0.0001650722141248178,
"loss": 3.0004465103149416,
"step": 1330
},
{
"epoch": 0.5319571258435887,
"grad_norm": 1.1726387739181519,
"learning_rate": 0.00016480720816218364,
"loss": 2.94736270904541,
"step": 1340
},
{
"epoch": 0.5359269551409289,
"grad_norm": 1.149013638496399,
"learning_rate": 0.0001645422021995495,
"loss": 2.9215377807617187,
"step": 1350
},
{
"epoch": 0.5398967844382692,
"grad_norm": 1.0117363929748535,
"learning_rate": 0.00016427719623691533,
"loss": 2.914345932006836,
"step": 1360
},
{
"epoch": 0.5438666137356094,
"grad_norm": 1.356374979019165,
"learning_rate": 0.0001640121902742812,
"loss": 2.9556610107421877,
"step": 1370
},
{
"epoch": 0.5478364430329495,
"grad_norm": 1.0233981609344482,
"learning_rate": 0.00016374718431164703,
"loss": 2.7411882400512697,
"step": 1380
},
{
"epoch": 0.5518062723302898,
"grad_norm": 0.9782310128211975,
"learning_rate": 0.00016348217834901286,
"loss": 2.8227493286132814,
"step": 1390
},
{
"epoch": 0.55577610162763,
"grad_norm": 1.1206411123275757,
"learning_rate": 0.0001632171723863787,
"loss": 2.9366016387939453,
"step": 1400
},
{
"epoch": 0.5597459309249703,
"grad_norm": 1.059769868850708,
"learning_rate": 0.00016295216642374453,
"loss": 2.877486228942871,
"step": 1410
},
{
"epoch": 0.5637157602223104,
"grad_norm": 1.1128787994384766,
"learning_rate": 0.0001626871604611104,
"loss": 2.9680269241333006,
"step": 1420
},
{
"epoch": 0.5676855895196506,
"grad_norm": 1.0075215101242065,
"learning_rate": 0.00016242215449847623,
"loss": 2.9732336044311523,
"step": 1430
},
{
"epoch": 0.5716554188169909,
"grad_norm": 1.111331820487976,
"learning_rate": 0.00016215714853584209,
"loss": 2.8879764556884764,
"step": 1440
},
{
"epoch": 0.5756252481143311,
"grad_norm": 0.9936279654502869,
"learning_rate": 0.0001618921425732079,
"loss": 2.8723777770996093,
"step": 1450
},
{
"epoch": 0.5795950774116713,
"grad_norm": 1.0727992057800293,
"learning_rate": 0.00016162713661057373,
"loss": 2.7799137115478514,
"step": 1460
},
{
"epoch": 0.5835649067090115,
"grad_norm": 1.067679762840271,
"learning_rate": 0.0001613621306479396,
"loss": 2.932950019836426,
"step": 1470
},
{
"epoch": 0.5875347360063518,
"grad_norm": 1.2709579467773438,
"learning_rate": 0.00016109712468530542,
"loss": 3.027512550354004,
"step": 1480
},
{
"epoch": 0.5915045653036919,
"grad_norm": 1.0046635866165161,
"learning_rate": 0.00016083211872267128,
"loss": 2.834365463256836,
"step": 1490
},
{
"epoch": 0.5954743946010321,
"grad_norm": 1.0839996337890625,
"learning_rate": 0.00016056711276003712,
"loss": 2.955361747741699,
"step": 1500
},
{
"epoch": 0.5994442238983724,
"grad_norm": 1.0332610607147217,
"learning_rate": 0.00016030210679740295,
"loss": 2.892449378967285,
"step": 1510
},
{
"epoch": 0.6034140531957126,
"grad_norm": 1.1940085887908936,
"learning_rate": 0.00016003710083476878,
"loss": 3.0406375885009767,
"step": 1520
},
{
"epoch": 0.6073838824930528,
"grad_norm": 1.4349676370620728,
"learning_rate": 0.00015977209487213462,
"loss": 2.868655204772949,
"step": 1530
},
{
"epoch": 0.611353711790393,
"grad_norm": 0.98362797498703,
"learning_rate": 0.00015950708890950048,
"loss": 3.0159059524536134,
"step": 1540
},
{
"epoch": 0.6153235410877332,
"grad_norm": 1.1385223865509033,
"learning_rate": 0.0001592420829468663,
"loss": 2.9185848236083984,
"step": 1550
},
{
"epoch": 0.6192933703850735,
"grad_norm": 1.0622061491012573,
"learning_rate": 0.00015897707698423215,
"loss": 2.963969612121582,
"step": 1560
},
{
"epoch": 0.6232631996824136,
"grad_norm": 1.1907509565353394,
"learning_rate": 0.000158712071021598,
"loss": 3.0108991622924806,
"step": 1570
},
{
"epoch": 0.6272330289797539,
"grad_norm": 1.0997110605239868,
"learning_rate": 0.00015844706505896384,
"loss": 2.965668869018555,
"step": 1580
},
{
"epoch": 0.6312028582770941,
"grad_norm": 1.212643027305603,
"learning_rate": 0.00015818205909632967,
"loss": 2.953110694885254,
"step": 1590
},
{
"epoch": 0.6351726875744343,
"grad_norm": 1.0768868923187256,
"learning_rate": 0.0001579170531336955,
"loss": 3.17305850982666,
"step": 1600
},
{
"epoch": 0.6391425168717745,
"grad_norm": 1.4662688970565796,
"learning_rate": 0.00015765204717106134,
"loss": 2.972339057922363,
"step": 1610
},
{
"epoch": 0.6431123461691147,
"grad_norm": 1.34614896774292,
"learning_rate": 0.0001573870412084272,
"loss": 2.957012176513672,
"step": 1620
},
{
"epoch": 0.647082175466455,
"grad_norm": 1.0535542964935303,
"learning_rate": 0.00015712203524579304,
"loss": 3.014488410949707,
"step": 1630
},
{
"epoch": 0.6510520047637952,
"grad_norm": 1.0424102544784546,
"learning_rate": 0.0001568570292831589,
"loss": 2.882272720336914,
"step": 1640
},
{
"epoch": 0.6550218340611353,
"grad_norm": 1.021438717842102,
"learning_rate": 0.0001565920233205247,
"loss": 3.1003211975097655,
"step": 1650
},
{
"epoch": 0.6589916633584756,
"grad_norm": 1.237534761428833,
"learning_rate": 0.00015632701735789057,
"loss": 2.9361820220947266,
"step": 1660
},
{
"epoch": 0.6629614926558158,
"grad_norm": 1.1335283517837524,
"learning_rate": 0.0001560620113952564,
"loss": 2.932353401184082,
"step": 1670
},
{
"epoch": 0.666931321953156,
"grad_norm": 1.2225456237792969,
"learning_rate": 0.00015579700543262223,
"loss": 2.871927833557129,
"step": 1680
},
{
"epoch": 0.6709011512504962,
"grad_norm": 1.6691526174545288,
"learning_rate": 0.0001555319994699881,
"loss": 2.9645782470703126,
"step": 1690
},
{
"epoch": 0.6748709805478365,
"grad_norm": 0.9732962250709534,
"learning_rate": 0.00015526699350735393,
"loss": 2.8384857177734375,
"step": 1700
},
{
"epoch": 0.6788408098451767,
"grad_norm": 0.9814636707305908,
"learning_rate": 0.00015500198754471976,
"loss": 2.9634319305419923,
"step": 1710
},
{
"epoch": 0.6828106391425168,
"grad_norm": 1.2787647247314453,
"learning_rate": 0.0001547369815820856,
"loss": 2.973423957824707,
"step": 1720
},
{
"epoch": 0.6867804684398571,
"grad_norm": 1.2211189270019531,
"learning_rate": 0.00015447197561945143,
"loss": 3.0307428359985353,
"step": 1730
},
{
"epoch": 0.6907502977371973,
"grad_norm": 1.0328019857406616,
"learning_rate": 0.0001542069696568173,
"loss": 2.8718870162963865,
"step": 1740
},
{
"epoch": 0.6947201270345376,
"grad_norm": 1.0765776634216309,
"learning_rate": 0.00015394196369418312,
"loss": 3.050098991394043,
"step": 1750
},
{
"epoch": 0.6986899563318777,
"grad_norm": 1.0749907493591309,
"learning_rate": 0.00015367695773154899,
"loss": 3.0334381103515624,
"step": 1760
},
{
"epoch": 0.7026597856292179,
"grad_norm": 1.123230218887329,
"learning_rate": 0.00015341195176891482,
"loss": 2.948457717895508,
"step": 1770
},
{
"epoch": 0.7066296149265582,
"grad_norm": 1.2905099391937256,
"learning_rate": 0.00015314694580628065,
"loss": 2.968699264526367,
"step": 1780
},
{
"epoch": 0.7105994442238984,
"grad_norm": 1.1446195840835571,
"learning_rate": 0.0001528819398436465,
"loss": 2.848519706726074,
"step": 1790
},
{
"epoch": 0.7145692735212386,
"grad_norm": 1.073853611946106,
"learning_rate": 0.00015261693388101232,
"loss": 2.9441722869873046,
"step": 1800
},
{
"epoch": 0.7185391028185788,
"grad_norm": 1.2332338094711304,
"learning_rate": 0.00015235192791837818,
"loss": 2.9227483749389647,
"step": 1810
},
{
"epoch": 0.722508932115919,
"grad_norm": 1.0547096729278564,
"learning_rate": 0.00015208692195574402,
"loss": 3.0400369644165037,
"step": 1820
},
{
"epoch": 0.7264787614132592,
"grad_norm": 1.2107875347137451,
"learning_rate": 0.00015182191599310985,
"loss": 2.932765007019043,
"step": 1830
},
{
"epoch": 0.7304485907105994,
"grad_norm": 1.1923927068710327,
"learning_rate": 0.00015155691003047568,
"loss": 2.8786020278930664,
"step": 1840
},
{
"epoch": 0.7344184200079397,
"grad_norm": 1.3000658750534058,
"learning_rate": 0.00015129190406784152,
"loss": 3.0670289993286133,
"step": 1850
},
{
"epoch": 0.7383882493052799,
"grad_norm": 1.079545259475708,
"learning_rate": 0.00015102689810520738,
"loss": 2.9120521545410156,
"step": 1860
},
{
"epoch": 0.74235807860262,
"grad_norm": 1.164806604385376,
"learning_rate": 0.0001507618921425732,
"loss": 3.098134422302246,
"step": 1870
},
{
"epoch": 0.7463279078999603,
"grad_norm": 1.0587846040725708,
"learning_rate": 0.00015049688617993905,
"loss": 2.889917182922363,
"step": 1880
},
{
"epoch": 0.7502977371973005,
"grad_norm": 1.6976264715194702,
"learning_rate": 0.0001502318802173049,
"loss": 2.9458736419677733,
"step": 1890
},
{
"epoch": 0.7542675664946408,
"grad_norm": 1.0733898878097534,
"learning_rate": 0.00014996687425467074,
"loss": 2.872547912597656,
"step": 1900
},
{
"epoch": 0.7582373957919809,
"grad_norm": 1.560863971710205,
"learning_rate": 0.00014970186829203657,
"loss": 2.8400611877441406,
"step": 1910
},
{
"epoch": 0.7622072250893211,
"grad_norm": 1.3527973890304565,
"learning_rate": 0.0001494368623294024,
"loss": 2.9421205520629883,
"step": 1920
},
{
"epoch": 0.7661770543866614,
"grad_norm": 1.0963633060455322,
"learning_rate": 0.00014917185636676827,
"loss": 2.890008544921875,
"step": 1930
},
{
"epoch": 0.7701468836840016,
"grad_norm": 0.93892902135849,
"learning_rate": 0.0001489068504041341,
"loss": 2.8861242294311524,
"step": 1940
},
{
"epoch": 0.7741167129813418,
"grad_norm": 1.233338713645935,
"learning_rate": 0.00014864184444149994,
"loss": 2.9477554321289063,
"step": 1950
},
{
"epoch": 0.778086542278682,
"grad_norm": 1.2628257274627686,
"learning_rate": 0.0001483768384788658,
"loss": 2.8417577743530273,
"step": 1960
},
{
"epoch": 0.7820563715760223,
"grad_norm": 1.0811272859573364,
"learning_rate": 0.00014811183251623163,
"loss": 2.887801933288574,
"step": 1970
},
{
"epoch": 0.7860262008733624,
"grad_norm": 1.088047981262207,
"learning_rate": 0.00014784682655359746,
"loss": 3.039769744873047,
"step": 1980
},
{
"epoch": 0.7899960301707026,
"grad_norm": 1.2557246685028076,
"learning_rate": 0.0001475818205909633,
"loss": 2.9063520431518555,
"step": 1990
},
{
"epoch": 0.7939658594680429,
"grad_norm": 1.131279706954956,
"learning_rate": 0.00014731681462832913,
"loss": 2.995278549194336,
"step": 2000
},
{
"epoch": 0.7979356887653831,
"grad_norm": 1.127064824104309,
"learning_rate": 0.0001470783092619584,
"loss": 2.8959430694580077,
"step": 2010
},
{
"epoch": 0.8019055180627233,
"grad_norm": 1.1206923723220825,
"learning_rate": 0.00014681330329932425,
"loss": 2.988982009887695,
"step": 2020
},
{
"epoch": 0.8058753473600635,
"grad_norm": 1.2321375608444214,
"learning_rate": 0.00014654829733669008,
"loss": 2.886695671081543,
"step": 2030
},
{
"epoch": 0.8098451766574037,
"grad_norm": 1.0987670421600342,
"learning_rate": 0.00014628329137405591,
"loss": 2.9366876602172853,
"step": 2040
},
{
"epoch": 0.813815005954744,
"grad_norm": 1.2729072570800781,
"learning_rate": 0.00014601828541142177,
"loss": 2.733133316040039,
"step": 2050
},
{
"epoch": 0.8177848352520841,
"grad_norm": 1.4275444746017456,
"learning_rate": 0.0001457532794487876,
"loss": 2.9442211151123048,
"step": 2060
},
{
"epoch": 0.8217546645494244,
"grad_norm": 1.1997590065002441,
"learning_rate": 0.00014548827348615344,
"loss": 2.8335521697998045,
"step": 2070
},
{
"epoch": 0.8257244938467646,
"grad_norm": 1.4380601644515991,
"learning_rate": 0.0001452232675235193,
"loss": 3.0139677047729494,
"step": 2080
},
{
"epoch": 0.8296943231441049,
"grad_norm": 1.0848125219345093,
"learning_rate": 0.0001449582615608851,
"loss": 2.9831415176391602,
"step": 2090
},
{
"epoch": 0.833664152441445,
"grad_norm": 1.0817651748657227,
"learning_rate": 0.00014469325559825097,
"loss": 3.0775167465209963,
"step": 2100
},
{
"epoch": 0.8376339817387852,
"grad_norm": 1.1604492664337158,
"learning_rate": 0.0001444282496356168,
"loss": 2.9208209991455076,
"step": 2110
},
{
"epoch": 0.8416038110361255,
"grad_norm": 1.03305184841156,
"learning_rate": 0.00014416324367298264,
"loss": 2.9657546997070314,
"step": 2120
},
{
"epoch": 0.8455736403334657,
"grad_norm": 1.5073528289794922,
"learning_rate": 0.0001438982377103485,
"loss": 2.9726272583007813,
"step": 2130
},
{
"epoch": 0.8495434696308058,
"grad_norm": 1.332560420036316,
"learning_rate": 0.00014363323174771433,
"loss": 2.933932876586914,
"step": 2140
},
{
"epoch": 0.8535132989281461,
"grad_norm": 1.1734390258789062,
"learning_rate": 0.00014336822578508017,
"loss": 2.8562877655029295,
"step": 2150
},
{
"epoch": 0.8574831282254863,
"grad_norm": 1.1189327239990234,
"learning_rate": 0.000143103219822446,
"loss": 2.9401512145996094,
"step": 2160
},
{
"epoch": 0.8614529575228265,
"grad_norm": 1.2062342166900635,
"learning_rate": 0.00014283821385981183,
"loss": 3.0194915771484374,
"step": 2170
},
{
"epoch": 0.8654227868201667,
"grad_norm": 1.0934343338012695,
"learning_rate": 0.0001425732078971777,
"loss": 2.8667329788208007,
"step": 2180
},
{
"epoch": 0.869392616117507,
"grad_norm": 1.0804356336593628,
"learning_rate": 0.00014230820193454353,
"loss": 2.8640270233154297,
"step": 2190
},
{
"epoch": 0.8733624454148472,
"grad_norm": 1.1175063848495483,
"learning_rate": 0.0001420431959719094,
"loss": 2.998785400390625,
"step": 2200
},
{
"epoch": 0.8773322747121873,
"grad_norm": 1.038718819618225,
"learning_rate": 0.00014177819000927522,
"loss": 2.9121694564819336,
"step": 2210
},
{
"epoch": 0.8813021040095276,
"grad_norm": 1.2699339389801025,
"learning_rate": 0.00014151318404664106,
"loss": 2.925882339477539,
"step": 2220
},
{
"epoch": 0.8852719333068678,
"grad_norm": 1.0924122333526611,
"learning_rate": 0.0001412481780840069,
"loss": 2.883635902404785,
"step": 2230
},
{
"epoch": 0.8892417626042081,
"grad_norm": 1.1971683502197266,
"learning_rate": 0.00014098317212137273,
"loss": 2.8960296630859377,
"step": 2240
},
{
"epoch": 0.8932115919015482,
"grad_norm": 1.0126338005065918,
"learning_rate": 0.0001407181661587386,
"loss": 2.889109420776367,
"step": 2250
},
{
"epoch": 0.8971814211988884,
"grad_norm": 1.2543033361434937,
"learning_rate": 0.00014045316019610442,
"loss": 2.887612724304199,
"step": 2260
},
{
"epoch": 0.9011512504962287,
"grad_norm": 1.394667387008667,
"learning_rate": 0.00014018815423347028,
"loss": 2.9344444274902344,
"step": 2270
},
{
"epoch": 0.9051210797935689,
"grad_norm": 1.413507103919983,
"learning_rate": 0.00013992314827083612,
"loss": 2.91464900970459,
"step": 2280
},
{
"epoch": 0.9090909090909091,
"grad_norm": 1.258139729499817,
"learning_rate": 0.00013965814230820192,
"loss": 2.9243486404418944,
"step": 2290
},
{
"epoch": 0.9130607383882493,
"grad_norm": 0.8673571348190308,
"learning_rate": 0.00013939313634556778,
"loss": 2.7983678817749023,
"step": 2300
},
{
"epoch": 0.9170305676855895,
"grad_norm": 1.285875678062439,
"learning_rate": 0.00013912813038293362,
"loss": 3.068038558959961,
"step": 2310
},
{
"epoch": 0.9210003969829297,
"grad_norm": 1.1456619501113892,
"learning_rate": 0.00013886312442029948,
"loss": 2.982662010192871,
"step": 2320
},
{
"epoch": 0.9249702262802699,
"grad_norm": 1.028979778289795,
"learning_rate": 0.0001385981184576653,
"loss": 2.831150245666504,
"step": 2330
},
{
"epoch": 0.9289400555776102,
"grad_norm": 1.103188157081604,
"learning_rate": 0.00013833311249503115,
"loss": 2.891666603088379,
"step": 2340
},
{
"epoch": 0.9329098848749504,
"grad_norm": 1.105437159538269,
"learning_rate": 0.00013806810653239698,
"loss": 2.8475828170776367,
"step": 2350
},
{
"epoch": 0.9368797141722905,
"grad_norm": 1.1804033517837524,
"learning_rate": 0.0001378031005697628,
"loss": 2.9830224990844725,
"step": 2360
},
{
"epoch": 0.9408495434696308,
"grad_norm": 1.031882405281067,
"learning_rate": 0.00013753809460712867,
"loss": 2.820730781555176,
"step": 2370
},
{
"epoch": 0.944819372766971,
"grad_norm": 1.071161150932312,
"learning_rate": 0.0001372730886444945,
"loss": 2.9033863067626955,
"step": 2380
},
{
"epoch": 0.9487892020643113,
"grad_norm": 1.1234395503997803,
"learning_rate": 0.00013700808268186034,
"loss": 2.983052444458008,
"step": 2390
},
{
"epoch": 0.9527590313616514,
"grad_norm": 1.198830485343933,
"learning_rate": 0.0001367430767192262,
"loss": 3.0303747177124025,
"step": 2400
},
{
"epoch": 0.9567288606589917,
"grad_norm": 1.5915600061416626,
"learning_rate": 0.00013647807075659204,
"loss": 2.9372890472412108,
"step": 2410
},
{
"epoch": 0.9606986899563319,
"grad_norm": 1.2830276489257812,
"learning_rate": 0.00013621306479395787,
"loss": 2.769970703125,
"step": 2420
},
{
"epoch": 0.9646685192536721,
"grad_norm": 1.1718764305114746,
"learning_rate": 0.0001359480588313237,
"loss": 2.947464942932129,
"step": 2430
},
{
"epoch": 0.9686383485510123,
"grad_norm": 1.3351068496704102,
"learning_rate": 0.00013568305286868954,
"loss": 3.0664255142211916,
"step": 2440
},
{
"epoch": 0.9726081778483525,
"grad_norm": 1.2922847270965576,
"learning_rate": 0.0001354180469060554,
"loss": 2.882719039916992,
"step": 2450
},
{
"epoch": 0.9765780071456928,
"grad_norm": 1.3131754398345947,
"learning_rate": 0.00013515304094342123,
"loss": 2.906990814208984,
"step": 2460
},
{
"epoch": 0.9805478364430329,
"grad_norm": 1.2137771844863892,
"learning_rate": 0.0001348880349807871,
"loss": 2.9342098236083984,
"step": 2470
},
{
"epoch": 0.9845176657403731,
"grad_norm": 1.352759599685669,
"learning_rate": 0.0001346230290181529,
"loss": 2.8431310653686523,
"step": 2480
},
{
"epoch": 0.9884874950377134,
"grad_norm": 1.0442841053009033,
"learning_rate": 0.00013435802305551873,
"loss": 2.929820251464844,
"step": 2490
},
{
"epoch": 0.9924573243350536,
"grad_norm": 1.130361557006836,
"learning_rate": 0.0001340930170928846,
"loss": 2.9055503845214843,
"step": 2500
},
{
"epoch": 0.9964271536323938,
"grad_norm": 1.461983323097229,
"learning_rate": 0.00013382801113025043,
"loss": 2.9048336029052733,
"step": 2510
},
{
"epoch": 1.0003969829297341,
"grad_norm": 1.2407251596450806,
"learning_rate": 0.0001335630051676163,
"loss": 2.888891410827637,
"step": 2520
},
{
"epoch": 1.0043668122270741,
"grad_norm": 1.3580700159072876,
"learning_rate": 0.00013329799920498212,
"loss": 2.927655029296875,
"step": 2530
},
{
"epoch": 1.0083366415244144,
"grad_norm": 1.0742342472076416,
"learning_rate": 0.00013303299324234796,
"loss": 2.9614290237426757,
"step": 2540
},
{
"epoch": 1.0123064708217546,
"grad_norm": 1.2635444402694702,
"learning_rate": 0.0001327679872797138,
"loss": 2.7592201232910156,
"step": 2550
},
{
"epoch": 1.0162763001190949,
"grad_norm": 1.331705927848816,
"learning_rate": 0.00013250298131707962,
"loss": 2.8516969680786133,
"step": 2560
},
{
"epoch": 1.020246129416435,
"grad_norm": 1.1936160326004028,
"learning_rate": 0.00013223797535444549,
"loss": 3.008740234375,
"step": 2570
},
{
"epoch": 1.0242159587137754,
"grad_norm": 1.1526603698730469,
"learning_rate": 0.00013197296939181132,
"loss": 2.7761598587036134,
"step": 2580
},
{
"epoch": 1.0281857880111156,
"grad_norm": 1.0937445163726807,
"learning_rate": 0.00013170796342917718,
"loss": 2.8639547348022463,
"step": 2590
},
{
"epoch": 1.0321556173084558,
"grad_norm": 1.1443973779678345,
"learning_rate": 0.00013144295746654301,
"loss": 2.841537284851074,
"step": 2600
},
{
"epoch": 1.0361254466057959,
"grad_norm": 1.246218204498291,
"learning_rate": 0.00013117795150390885,
"loss": 2.89624137878418,
"step": 2610
},
{
"epoch": 1.040095275903136,
"grad_norm": 1.2295280694961548,
"learning_rate": 0.00013091294554127468,
"loss": 2.7112895965576174,
"step": 2620
},
{
"epoch": 1.0440651052004764,
"grad_norm": 1.1867451667785645,
"learning_rate": 0.00013064793957864052,
"loss": 2.817560577392578,
"step": 2630
},
{
"epoch": 1.0480349344978166,
"grad_norm": 1.3577598333358765,
"learning_rate": 0.00013038293361600638,
"loss": 2.899570846557617,
"step": 2640
},
{
"epoch": 1.0520047637951568,
"grad_norm": 1.1519272327423096,
"learning_rate": 0.0001301179276533722,
"loss": 2.899755096435547,
"step": 2650
},
{
"epoch": 1.055974593092497,
"grad_norm": 1.1014765501022339,
"learning_rate": 0.00012985292169073804,
"loss": 2.949821662902832,
"step": 2660
},
{
"epoch": 1.0599444223898373,
"grad_norm": 1.1649622917175293,
"learning_rate": 0.0001295879157281039,
"loss": 3.0167802810668944,
"step": 2670
},
{
"epoch": 1.0639142516871773,
"grad_norm": 1.4407117366790771,
"learning_rate": 0.0001293229097654697,
"loss": 2.934400177001953,
"step": 2680
},
{
"epoch": 1.0678840809845176,
"grad_norm": 1.1467158794403076,
"learning_rate": 0.00012905790380283557,
"loss": 2.8859203338623045,
"step": 2690
},
{
"epoch": 1.0718539102818578,
"grad_norm": 1.1657541990280151,
"learning_rate": 0.0001287928978402014,
"loss": 2.9197870254516602,
"step": 2700
},
{
"epoch": 1.075823739579198,
"grad_norm": 1.137046217918396,
"learning_rate": 0.00012852789187756724,
"loss": 2.9437820434570314,
"step": 2710
},
{
"epoch": 1.0797935688765383,
"grad_norm": 1.2877311706542969,
"learning_rate": 0.0001282628859149331,
"loss": 2.861769104003906,
"step": 2720
},
{
"epoch": 1.0837633981738786,
"grad_norm": 1.4007761478424072,
"learning_rate": 0.00012799787995229893,
"loss": 2.888690376281738,
"step": 2730
},
{
"epoch": 1.0877332274712188,
"grad_norm": 1.1097158193588257,
"learning_rate": 0.00012773287398966477,
"loss": 2.829306793212891,
"step": 2740
},
{
"epoch": 1.091703056768559,
"grad_norm": 1.3691431283950806,
"learning_rate": 0.0001274678680270306,
"loss": 2.900307846069336,
"step": 2750
},
{
"epoch": 1.095672886065899,
"grad_norm": 1.4336342811584473,
"learning_rate": 0.00012720286206439644,
"loss": 2.7967662811279297,
"step": 2760
},
{
"epoch": 1.0996427153632393,
"grad_norm": 1.2973034381866455,
"learning_rate": 0.0001269378561017623,
"loss": 2.774248504638672,
"step": 2770
},
{
"epoch": 1.1036125446605796,
"grad_norm": 1.5863878726959229,
"learning_rate": 0.00012667285013912813,
"loss": 2.7378374099731446,
"step": 2780
},
{
"epoch": 1.1075823739579198,
"grad_norm": 1.4274410009384155,
"learning_rate": 0.000126407844176494,
"loss": 3.0109136581420897,
"step": 2790
},
{
"epoch": 1.11155220325526,
"grad_norm": 1.1030832529067993,
"learning_rate": 0.00012614283821385983,
"loss": 2.7925899505615233,
"step": 2800
},
{
"epoch": 1.1155220325526003,
"grad_norm": 1.1131644248962402,
"learning_rate": 0.00012587783225122566,
"loss": 2.9009119033813477,
"step": 2810
},
{
"epoch": 1.1194918618499405,
"grad_norm": 1.3609836101531982,
"learning_rate": 0.0001256128262885915,
"loss": 2.8956506729125975,
"step": 2820
},
{
"epoch": 1.1234616911472806,
"grad_norm": 1.1545202732086182,
"learning_rate": 0.00012534782032595733,
"loss": 2.876191520690918,
"step": 2830
},
{
"epoch": 1.1274315204446208,
"grad_norm": 1.2126272916793823,
"learning_rate": 0.0001250828143633232,
"loss": 2.7961231231689454,
"step": 2840
},
{
"epoch": 1.131401349741961,
"grad_norm": 1.1984245777130127,
"learning_rate": 0.00012481780840068902,
"loss": 2.7959896087646485,
"step": 2850
},
{
"epoch": 1.1353711790393013,
"grad_norm": 1.2016088962554932,
"learning_rate": 0.00012455280243805488,
"loss": 2.867430305480957,
"step": 2860
},
{
"epoch": 1.1393410083366415,
"grad_norm": 1.1378740072250366,
"learning_rate": 0.00012428779647542072,
"loss": 2.741920280456543,
"step": 2870
},
{
"epoch": 1.1433108376339818,
"grad_norm": 1.2212181091308594,
"learning_rate": 0.00012402279051278652,
"loss": 2.7447364807128904,
"step": 2880
},
{
"epoch": 1.147280666931322,
"grad_norm": 1.0747908353805542,
"learning_rate": 0.00012375778455015238,
"loss": 2.98239688873291,
"step": 2890
},
{
"epoch": 1.1512504962286623,
"grad_norm": 1.3732624053955078,
"learning_rate": 0.00012349277858751822,
"loss": 2.932689094543457,
"step": 2900
},
{
"epoch": 1.1552203255260023,
"grad_norm": 1.2373459339141846,
"learning_rate": 0.00012322777262488408,
"loss": 2.8869907379150392,
"step": 2910
},
{
"epoch": 1.1591901548233425,
"grad_norm": 1.3174314498901367,
"learning_rate": 0.0001229627666622499,
"loss": 2.9600086212158203,
"step": 2920
},
{
"epoch": 1.1631599841206828,
"grad_norm": 1.3416168689727783,
"learning_rate": 0.00012269776069961575,
"loss": 2.840256118774414,
"step": 2930
},
{
"epoch": 1.167129813418023,
"grad_norm": 1.149091362953186,
"learning_rate": 0.00012243275473698158,
"loss": 2.807183265686035,
"step": 2940
},
{
"epoch": 1.1710996427153633,
"grad_norm": 1.1443051099777222,
"learning_rate": 0.00012216774877434741,
"loss": 2.9720365524291994,
"step": 2950
},
{
"epoch": 1.1750694720127035,
"grad_norm": 1.3129942417144775,
"learning_rate": 0.00012190274281171328,
"loss": 2.9757314682006837,
"step": 2960
},
{
"epoch": 1.1790393013100438,
"grad_norm": 1.098672866821289,
"learning_rate": 0.00012163773684907911,
"loss": 2.753942680358887,
"step": 2970
},
{
"epoch": 1.1830091306073838,
"grad_norm": 1.1337664127349854,
"learning_rate": 0.00012137273088644494,
"loss": 2.827354621887207,
"step": 2980
},
{
"epoch": 1.186978959904724,
"grad_norm": 1.1231764554977417,
"learning_rate": 0.00012110772492381079,
"loss": 2.711290740966797,
"step": 2990
},
{
"epoch": 1.1909487892020643,
"grad_norm": 1.1375696659088135,
"learning_rate": 0.00012084271896117662,
"loss": 2.904682731628418,
"step": 3000
},
{
"epoch": 1.1949186184994045,
"grad_norm": 1.1674156188964844,
"learning_rate": 0.00012057771299854248,
"loss": 2.884379768371582,
"step": 3010
},
{
"epoch": 1.1988884477967447,
"grad_norm": 1.4644557237625122,
"learning_rate": 0.00012031270703590832,
"loss": 2.801077461242676,
"step": 3020
},
{
"epoch": 1.202858277094085,
"grad_norm": 1.285089135169983,
"learning_rate": 0.00012004770107327414,
"loss": 2.8604631423950195,
"step": 3030
},
{
"epoch": 1.2068281063914252,
"grad_norm": 1.1116204261779785,
"learning_rate": 0.00011978269511064,
"loss": 2.8972236633300783,
"step": 3040
},
{
"epoch": 1.2107979356887655,
"grad_norm": 1.3030141592025757,
"learning_rate": 0.00011951768914800583,
"loss": 2.8783727645874024,
"step": 3050
},
{
"epoch": 1.2147677649861055,
"grad_norm": 1.0779145956039429,
"learning_rate": 0.00011925268318537168,
"loss": 2.85146427154541,
"step": 3060
},
{
"epoch": 1.2187375942834457,
"grad_norm": 1.7380479574203491,
"learning_rate": 0.00011898767722273751,
"loss": 2.8459522247314455,
"step": 3070
},
{
"epoch": 1.222707423580786,
"grad_norm": 1.0546619892120361,
"learning_rate": 0.00011872267126010336,
"loss": 2.805705451965332,
"step": 3080
},
{
"epoch": 1.2266772528781262,
"grad_norm": 1.533992052078247,
"learning_rate": 0.0001184576652974692,
"loss": 2.975359344482422,
"step": 3090
},
{
"epoch": 1.2306470821754665,
"grad_norm": 1.2582201957702637,
"learning_rate": 0.00011819265933483503,
"loss": 2.9268173217773437,
"step": 3100
},
{
"epoch": 1.2346169114728067,
"grad_norm": 1.2494417428970337,
"learning_rate": 0.00011792765337220089,
"loss": 2.7564210891723633,
"step": 3110
},
{
"epoch": 1.238586740770147,
"grad_norm": 1.40061354637146,
"learning_rate": 0.00011766264740956671,
"loss": 2.877336311340332,
"step": 3120
},
{
"epoch": 1.242556570067487,
"grad_norm": 1.36153244972229,
"learning_rate": 0.00011739764144693257,
"loss": 2.876156044006348,
"step": 3130
},
{
"epoch": 1.2465263993648272,
"grad_norm": 1.7986162900924683,
"learning_rate": 0.0001171326354842984,
"loss": 2.9214223861694335,
"step": 3140
},
{
"epoch": 1.2504962286621675,
"grad_norm": 1.1867148876190186,
"learning_rate": 0.00011686762952166424,
"loss": 2.8030139923095705,
"step": 3150
},
{
"epoch": 1.2544660579595077,
"grad_norm": 1.1623351573944092,
"learning_rate": 0.00011660262355903009,
"loss": 2.854610633850098,
"step": 3160
},
{
"epoch": 1.258435887256848,
"grad_norm": 1.1789515018463135,
"learning_rate": 0.00011633761759639592,
"loss": 2.840734672546387,
"step": 3170
},
{
"epoch": 1.2624057165541882,
"grad_norm": 1.3448714017868042,
"learning_rate": 0.00011607261163376177,
"loss": 2.942055511474609,
"step": 3180
},
{
"epoch": 1.2663755458515285,
"grad_norm": 1.2870416641235352,
"learning_rate": 0.0001158076056711276,
"loss": 2.928165245056152,
"step": 3190
},
{
"epoch": 1.2703453751488687,
"grad_norm": 1.1872483491897583,
"learning_rate": 0.00011554259970849344,
"loss": 2.935594940185547,
"step": 3200
},
{
"epoch": 1.274315204446209,
"grad_norm": 1.4787572622299194,
"learning_rate": 0.0001152775937458593,
"loss": 2.8724430084228514,
"step": 3210
},
{
"epoch": 1.278285033743549,
"grad_norm": 1.4164690971374512,
"learning_rate": 0.00011501258778322512,
"loss": 2.9088714599609373,
"step": 3220
},
{
"epoch": 1.2822548630408892,
"grad_norm": 1.1791913509368896,
"learning_rate": 0.00011474758182059098,
"loss": 2.8589361190795897,
"step": 3230
},
{
"epoch": 1.2862246923382294,
"grad_norm": 1.3528516292572021,
"learning_rate": 0.00011448257585795681,
"loss": 2.8930646896362306,
"step": 3240
},
{
"epoch": 1.2901945216355697,
"grad_norm": 1.3224010467529297,
"learning_rate": 0.00011421756989532265,
"loss": 2.8452373504638673,
"step": 3250
},
{
"epoch": 1.29416435093291,
"grad_norm": 1.2920061349868774,
"learning_rate": 0.00011395256393268849,
"loss": 2.8752540588378905,
"step": 3260
},
{
"epoch": 1.2981341802302502,
"grad_norm": 1.2639166116714478,
"learning_rate": 0.00011368755797005433,
"loss": 2.88561954498291,
"step": 3270
},
{
"epoch": 1.3021040095275902,
"grad_norm": 1.2361299991607666,
"learning_rate": 0.00011342255200742017,
"loss": 2.9272403717041016,
"step": 3280
},
{
"epoch": 1.3060738388249304,
"grad_norm": 1.5476930141448975,
"learning_rate": 0.00011315754604478601,
"loss": 2.994682502746582,
"step": 3290
},
{
"epoch": 1.3100436681222707,
"grad_norm": 1.3958719968795776,
"learning_rate": 0.00011289254008215184,
"loss": 2.810057830810547,
"step": 3300
},
{
"epoch": 1.314013497419611,
"grad_norm": 1.2481234073638916,
"learning_rate": 0.0001126275341195177,
"loss": 2.8198265075683593,
"step": 3310
},
{
"epoch": 1.3179833267169512,
"grad_norm": 1.5658928155899048,
"learning_rate": 0.00011236252815688352,
"loss": 2.9725709915161134,
"step": 3320
},
{
"epoch": 1.3219531560142914,
"grad_norm": 1.1179542541503906,
"learning_rate": 0.00011209752219424938,
"loss": 2.855281448364258,
"step": 3330
},
{
"epoch": 1.3259229853116317,
"grad_norm": 1.2543385028839111,
"learning_rate": 0.00011183251623161522,
"loss": 2.870030403137207,
"step": 3340
},
{
"epoch": 1.329892814608972,
"grad_norm": 1.2362663745880127,
"learning_rate": 0.00011156751026898106,
"loss": 2.867103385925293,
"step": 3350
},
{
"epoch": 1.3338626439063122,
"grad_norm": 1.5475022792816162,
"learning_rate": 0.0001113025043063469,
"loss": 3.038361740112305,
"step": 3360
},
{
"epoch": 1.3378324732036522,
"grad_norm": 1.2643401622772217,
"learning_rate": 0.00011103749834371273,
"loss": 2.880583381652832,
"step": 3370
},
{
"epoch": 1.3418023025009924,
"grad_norm": 1.1639227867126465,
"learning_rate": 0.00011077249238107858,
"loss": 2.7907382965087892,
"step": 3380
},
{
"epoch": 1.3457721317983327,
"grad_norm": 1.1745330095291138,
"learning_rate": 0.00011050748641844441,
"loss": 2.7494991302490233,
"step": 3390
},
{
"epoch": 1.349741961095673,
"grad_norm": 1.2324411869049072,
"learning_rate": 0.00011024248045581027,
"loss": 2.938181686401367,
"step": 3400
},
{
"epoch": 1.3537117903930131,
"grad_norm": 1.1418648958206177,
"learning_rate": 0.00010997747449317611,
"loss": 2.9696788787841797,
"step": 3410
},
{
"epoch": 1.3576816196903534,
"grad_norm": 1.3676928281784058,
"learning_rate": 0.00010971246853054193,
"loss": 2.930206298828125,
"step": 3420
},
{
"epoch": 1.3616514489876934,
"grad_norm": 1.2266117334365845,
"learning_rate": 0.00010944746256790779,
"loss": 2.9436601638793944,
"step": 3430
},
{
"epoch": 1.3656212782850337,
"grad_norm": 1.296379566192627,
"learning_rate": 0.00010918245660527362,
"loss": 2.822721481323242,
"step": 3440
},
{
"epoch": 1.369591107582374,
"grad_norm": 1.5747238397598267,
"learning_rate": 0.00010891745064263947,
"loss": 2.7970211029052736,
"step": 3450
},
{
"epoch": 1.3735609368797141,
"grad_norm": 1.11448335647583,
"learning_rate": 0.0001086524446800053,
"loss": 2.850177764892578,
"step": 3460
},
{
"epoch": 1.3775307661770544,
"grad_norm": 1.585570216178894,
"learning_rate": 0.00010838743871737114,
"loss": 2.7775524139404295,
"step": 3470
},
{
"epoch": 1.3815005954743946,
"grad_norm": 1.3851438760757446,
"learning_rate": 0.00010812243275473699,
"loss": 2.841839599609375,
"step": 3480
},
{
"epoch": 1.3854704247717349,
"grad_norm": 2.0213372707366943,
"learning_rate": 0.00010785742679210282,
"loss": 2.8476282119750977,
"step": 3490
},
{
"epoch": 1.3894402540690751,
"grad_norm": 1.1296862363815308,
"learning_rate": 0.00010759242082946868,
"loss": 2.796487045288086,
"step": 3500
},
{
"epoch": 1.3934100833664154,
"grad_norm": 1.2621263265609741,
"learning_rate": 0.00010732741486683451,
"loss": 2.8717994689941406,
"step": 3510
},
{
"epoch": 1.3973799126637554,
"grad_norm": 1.1782283782958984,
"learning_rate": 0.00010706240890420033,
"loss": 2.894154167175293,
"step": 3520
},
{
"epoch": 1.4013497419610956,
"grad_norm": 1.2578155994415283,
"learning_rate": 0.0001067974029415662,
"loss": 2.699532890319824,
"step": 3530
},
{
"epoch": 1.4053195712584359,
"grad_norm": 1.1540583372116089,
"learning_rate": 0.00010653239697893203,
"loss": 2.8674644470214843,
"step": 3540
},
{
"epoch": 1.4092894005557761,
"grad_norm": 1.5898374319076538,
"learning_rate": 0.00010626739101629788,
"loss": 2.887256622314453,
"step": 3550
},
{
"epoch": 1.4132592298531164,
"grad_norm": 1.2546888589859009,
"learning_rate": 0.00010600238505366371,
"loss": 2.912336730957031,
"step": 3560
},
{
"epoch": 1.4172290591504566,
"grad_norm": 1.3821418285369873,
"learning_rate": 0.00010573737909102954,
"loss": 2.7670663833618163,
"step": 3570
},
{
"epoch": 1.4211988884477966,
"grad_norm": 1.3685969114303589,
"learning_rate": 0.00010547237312839539,
"loss": 2.8921297073364256,
"step": 3580
},
{
"epoch": 1.4251687177451369,
"grad_norm": 2.0592007637023926,
"learning_rate": 0.00010520736716576123,
"loss": 2.8367937088012694,
"step": 3590
},
{
"epoch": 1.4291385470424771,
"grad_norm": 1.2717535495758057,
"learning_rate": 0.00010494236120312709,
"loss": 2.8397924423217775,
"step": 3600
},
{
"epoch": 1.4331083763398174,
"grad_norm": 1.1194345951080322,
"learning_rate": 0.00010467735524049292,
"loss": 2.8019588470458983,
"step": 3610
},
{
"epoch": 1.4370782056371576,
"grad_norm": 1.3213400840759277,
"learning_rate": 0.00010441234927785877,
"loss": 2.8019805908203126,
"step": 3620
},
{
"epoch": 1.4410480349344978,
"grad_norm": 1.2584818601608276,
"learning_rate": 0.0001041473433152246,
"loss": 2.9595170974731446,
"step": 3630
},
{
"epoch": 1.445017864231838,
"grad_norm": 1.346086859703064,
"learning_rate": 0.00010388233735259044,
"loss": 2.8712783813476563,
"step": 3640
},
{
"epoch": 1.4489876935291783,
"grad_norm": 1.279360294342041,
"learning_rate": 0.00010361733138995628,
"loss": 2.8931833267211915,
"step": 3650
},
{
"epoch": 1.4529575228265186,
"grad_norm": 1.466051459312439,
"learning_rate": 0.00010335232542732212,
"loss": 2.963275146484375,
"step": 3660
},
{
"epoch": 1.4569273521238586,
"grad_norm": 1.3183846473693848,
"learning_rate": 0.00010308731946468796,
"loss": 2.8601329803466795,
"step": 3670
},
{
"epoch": 1.4608971814211988,
"grad_norm": 2.090052604675293,
"learning_rate": 0.0001028223135020538,
"loss": 2.9438711166381837,
"step": 3680
},
{
"epoch": 1.464867010718539,
"grad_norm": 1.3010114431381226,
"learning_rate": 0.00010255730753941963,
"loss": 2.701998710632324,
"step": 3690
},
{
"epoch": 1.4688368400158793,
"grad_norm": 1.1416301727294922,
"learning_rate": 0.00010229230157678549,
"loss": 2.846023178100586,
"step": 3700
},
{
"epoch": 1.4728066693132196,
"grad_norm": 1.3570079803466797,
"learning_rate": 0.00010202729561415131,
"loss": 3.026602554321289,
"step": 3710
},
{
"epoch": 1.4767764986105598,
"grad_norm": 1.3911274671554565,
"learning_rate": 0.00010176228965151717,
"loss": 2.9053327560424806,
"step": 3720
},
{
"epoch": 1.4807463279078998,
"grad_norm": 1.360430359840393,
"learning_rate": 0.00010149728368888301,
"loss": 2.736703872680664,
"step": 3730
},
{
"epoch": 1.48471615720524,
"grad_norm": 1.4576371908187866,
"learning_rate": 0.00010123227772624884,
"loss": 2.8896203994750977,
"step": 3740
},
{
"epoch": 1.4886859865025803,
"grad_norm": 1.4311631917953491,
"learning_rate": 0.00010096727176361469,
"loss": 2.8624494552612303,
"step": 3750
},
{
"epoch": 1.4926558157999206,
"grad_norm": 1.5123578310012817,
"learning_rate": 0.00010070226580098052,
"loss": 2.809356689453125,
"step": 3760
},
{
"epoch": 1.4966256450972608,
"grad_norm": 1.2829740047454834,
"learning_rate": 0.00010043725983834637,
"loss": 2.8745845794677733,
"step": 3770
},
{
"epoch": 1.500595474394601,
"grad_norm": 1.7028263807296753,
"learning_rate": 0.0001001722538757122,
"loss": 2.8838785171508787,
"step": 3780
},
{
"epoch": 1.5045653036919413,
"grad_norm": 1.5197185277938843,
"learning_rate": 9.990724791307805e-05,
"loss": 2.881943130493164,
"step": 3790
},
{
"epoch": 1.5085351329892815,
"grad_norm": 1.223615050315857,
"learning_rate": 9.96422419504439e-05,
"loss": 2.7202972412109374,
"step": 3800
},
{
"epoch": 1.5125049622866218,
"grad_norm": 1.8343837261199951,
"learning_rate": 9.937723598780973e-05,
"loss": 2.8640880584716797,
"step": 3810
},
{
"epoch": 1.516474791583962,
"grad_norm": 1.5116796493530273,
"learning_rate": 9.911223002517557e-05,
"loss": 2.924713134765625,
"step": 3820
},
{
"epoch": 1.520444620881302,
"grad_norm": 1.6535507440567017,
"learning_rate": 9.884722406254141e-05,
"loss": 2.7902399063110352,
"step": 3830
},
{
"epoch": 1.5244144501786423,
"grad_norm": 1.398630976676941,
"learning_rate": 9.858221809990725e-05,
"loss": 2.8293262481689454,
"step": 3840
},
{
"epoch": 1.5283842794759825,
"grad_norm": 1.4054690599441528,
"learning_rate": 9.83172121372731e-05,
"loss": 2.8115062713623047,
"step": 3850
},
{
"epoch": 1.5323541087733228,
"grad_norm": 1.4996291399002075,
"learning_rate": 9.805220617463894e-05,
"loss": 2.832067108154297,
"step": 3860
},
{
"epoch": 1.5363239380706628,
"grad_norm": 1.0884090662002563,
"learning_rate": 9.778720021200478e-05,
"loss": 2.6115888595581054,
"step": 3870
},
{
"epoch": 1.540293767368003,
"grad_norm": 1.2034204006195068,
"learning_rate": 9.752219424937061e-05,
"loss": 2.7569812774658202,
"step": 3880
},
{
"epoch": 1.5442635966653433,
"grad_norm": 1.3871127367019653,
"learning_rate": 9.725718828673646e-05,
"loss": 3.050065612792969,
"step": 3890
},
{
"epoch": 1.5482334259626835,
"grad_norm": 1.5501118898391724,
"learning_rate": 9.69921823241023e-05,
"loss": 2.8175943374633787,
"step": 3900
},
{
"epoch": 1.5522032552600238,
"grad_norm": 1.2360347509384155,
"learning_rate": 9.672717636146814e-05,
"loss": 2.7722763061523437,
"step": 3910
},
{
"epoch": 1.556173084557364,
"grad_norm": 1.2444398403167725,
"learning_rate": 9.646217039883397e-05,
"loss": 2.861615753173828,
"step": 3920
},
{
"epoch": 1.5601429138547043,
"grad_norm": 1.3740792274475098,
"learning_rate": 9.619716443619982e-05,
"loss": 2.81121826171875,
"step": 3930
},
{
"epoch": 1.5641127431520445,
"grad_norm": 1.429770588874817,
"learning_rate": 9.593215847356565e-05,
"loss": 2.922770309448242,
"step": 3940
},
{
"epoch": 1.5680825724493848,
"grad_norm": 1.2474150657653809,
"learning_rate": 9.56671525109315e-05,
"loss": 2.9116376876831054,
"step": 3950
},
{
"epoch": 1.572052401746725,
"grad_norm": 1.431827425956726,
"learning_rate": 9.540214654829735e-05,
"loss": 2.917964553833008,
"step": 3960
},
{
"epoch": 1.5760222310440652,
"grad_norm": 1.3315114974975586,
"learning_rate": 9.513714058566318e-05,
"loss": 2.8786062240600585,
"step": 3970
},
{
"epoch": 1.5799920603414053,
"grad_norm": 1.1765938997268677,
"learning_rate": 9.487213462302902e-05,
"loss": 2.8044477462768556,
"step": 3980
},
{
"epoch": 1.5839618896387455,
"grad_norm": 1.348409652709961,
"learning_rate": 9.460712866039486e-05,
"loss": 2.8915786743164062,
"step": 3990
},
{
"epoch": 1.5879317189360858,
"grad_norm": 1.4071879386901855,
"learning_rate": 9.434212269776071e-05,
"loss": 2.8507789611816405,
"step": 4000
},
{
"epoch": 1.591901548233426,
"grad_norm": 1.3064533472061157,
"learning_rate": 9.410361733138996e-05,
"loss": 2.7582677841186523,
"step": 4010
},
{
"epoch": 1.595871377530766,
"grad_norm": 1.2912938594818115,
"learning_rate": 9.38386113687558e-05,
"loss": 2.882757568359375,
"step": 4020
},
{
"epoch": 1.5998412068281063,
"grad_norm": 1.1329375505447388,
"learning_rate": 9.357360540612164e-05,
"loss": 2.815750312805176,
"step": 4030
},
{
"epoch": 1.6038110361254465,
"grad_norm": 1.337375521659851,
"learning_rate": 9.330859944348749e-05,
"loss": 2.8128747940063477,
"step": 4040
},
{
"epoch": 1.6077808654227868,
"grad_norm": 1.2815532684326172,
"learning_rate": 9.304359348085333e-05,
"loss": 2.7639719009399415,
"step": 4050
},
{
"epoch": 1.611750694720127,
"grad_norm": 1.1914876699447632,
"learning_rate": 9.277858751821916e-05,
"loss": 2.894421195983887,
"step": 4060
},
{
"epoch": 1.6157205240174672,
"grad_norm": 1.2822132110595703,
"learning_rate": 9.2513581555585e-05,
"loss": 2.754531669616699,
"step": 4070
},
{
"epoch": 1.6196903533148075,
"grad_norm": 1.2767010927200317,
"learning_rate": 9.224857559295085e-05,
"loss": 2.884870719909668,
"step": 4080
},
{
"epoch": 1.6236601826121477,
"grad_norm": 1.5516023635864258,
"learning_rate": 9.198356963031669e-05,
"loss": 2.928493309020996,
"step": 4090
},
{
"epoch": 1.627630011909488,
"grad_norm": 1.2238390445709229,
"learning_rate": 9.171856366768252e-05,
"loss": 2.753667449951172,
"step": 4100
},
{
"epoch": 1.6315998412068282,
"grad_norm": 1.3094969987869263,
"learning_rate": 9.145355770504837e-05,
"loss": 2.8846521377563477,
"step": 4110
},
{
"epoch": 1.6355696705041685,
"grad_norm": 1.1217821836471558,
"learning_rate": 9.11885517424142e-05,
"loss": 2.8405046463012695,
"step": 4120
},
{
"epoch": 1.6395394998015085,
"grad_norm": 1.1809136867523193,
"learning_rate": 9.092354577978005e-05,
"loss": 2.8990488052368164,
"step": 4130
},
{
"epoch": 1.6435093290988487,
"grad_norm": 1.4486050605773926,
"learning_rate": 9.06585398171459e-05,
"loss": 2.801529884338379,
"step": 4140
},
{
"epoch": 1.647479158396189,
"grad_norm": 1.3603150844573975,
"learning_rate": 9.039353385451173e-05,
"loss": 2.7627063751220704,
"step": 4150
},
{
"epoch": 1.6514489876935292,
"grad_norm": 1.3169219493865967,
"learning_rate": 9.012852789187757e-05,
"loss": 2.844338798522949,
"step": 4160
},
{
"epoch": 1.6554188169908692,
"grad_norm": 1.4041424989700317,
"learning_rate": 8.986352192924341e-05,
"loss": 2.820200729370117,
"step": 4170
},
{
"epoch": 1.6593886462882095,
"grad_norm": 1.900181531906128,
"learning_rate": 8.959851596660926e-05,
"loss": 2.798480415344238,
"step": 4180
},
{
"epoch": 1.6633584755855497,
"grad_norm": 1.3303025960922241,
"learning_rate": 8.93335100039751e-05,
"loss": 2.7450151443481445,
"step": 4190
},
{
"epoch": 1.66732830488289,
"grad_norm": 1.173895239830017,
"learning_rate": 8.906850404134094e-05,
"loss": 2.807242584228516,
"step": 4200
},
{
"epoch": 1.6712981341802302,
"grad_norm": 1.3657634258270264,
"learning_rate": 8.880349807870677e-05,
"loss": 2.8270692825317383,
"step": 4210
},
{
"epoch": 1.6752679634775705,
"grad_norm": 1.1091082096099854,
"learning_rate": 8.853849211607261e-05,
"loss": 2.79335994720459,
"step": 4220
},
{
"epoch": 1.6792377927749107,
"grad_norm": 1.2687846422195435,
"learning_rate": 8.827348615343846e-05,
"loss": 2.942344093322754,
"step": 4230
},
{
"epoch": 1.683207622072251,
"grad_norm": 1.5438424348831177,
"learning_rate": 8.80084801908043e-05,
"loss": 2.7900604248046874,
"step": 4240
},
{
"epoch": 1.6871774513695912,
"grad_norm": 1.2709591388702393,
"learning_rate": 8.774347422817014e-05,
"loss": 2.9074878692626953,
"step": 4250
},
{
"epoch": 1.6911472806669314,
"grad_norm": 1.2722247838974,
"learning_rate": 8.747846826553597e-05,
"loss": 2.8957965850830076,
"step": 4260
},
{
"epoch": 1.6951171099642717,
"grad_norm": 1.359768271446228,
"learning_rate": 8.721346230290182e-05,
"loss": 2.944797706604004,
"step": 4270
},
{
"epoch": 1.699086939261612,
"grad_norm": 1.3503456115722656,
"learning_rate": 8.694845634026765e-05,
"loss": 2.9122066497802734,
"step": 4280
},
{
"epoch": 1.703056768558952,
"grad_norm": 1.1977394819259644,
"learning_rate": 8.66834503776335e-05,
"loss": 2.7401096343994142,
"step": 4290
},
{
"epoch": 1.7070265978562922,
"grad_norm": 1.3613430261611938,
"learning_rate": 8.641844441499935e-05,
"loss": 2.8979299545288084,
"step": 4300
},
{
"epoch": 1.7109964271536324,
"grad_norm": 1.4135546684265137,
"learning_rate": 8.615343845236518e-05,
"loss": 2.8803421020507813,
"step": 4310
},
{
"epoch": 1.7149662564509724,
"grad_norm": 1.291506052017212,
"learning_rate": 8.588843248973101e-05,
"loss": 2.8269506454467774,
"step": 4320
},
{
"epoch": 1.7189360857483127,
"grad_norm": 1.3398776054382324,
"learning_rate": 8.562342652709686e-05,
"loss": 2.8653587341308593,
"step": 4330
},
{
"epoch": 1.722905915045653,
"grad_norm": 1.2929062843322754,
"learning_rate": 8.535842056446271e-05,
"loss": 2.8511186599731446,
"step": 4340
},
{
"epoch": 1.7268757443429932,
"grad_norm": 1.2606959342956543,
"learning_rate": 8.509341460182854e-05,
"loss": 2.6914939880371094,
"step": 4350
},
{
"epoch": 1.7308455736403334,
"grad_norm": 1.3255246877670288,
"learning_rate": 8.482840863919439e-05,
"loss": 2.7858762741088867,
"step": 4360
},
{
"epoch": 1.7348154029376737,
"grad_norm": 1.5206516981124878,
"learning_rate": 8.456340267656022e-05,
"loss": 2.8856891632080077,
"step": 4370
},
{
"epoch": 1.738785232235014,
"grad_norm": 1.3062914609909058,
"learning_rate": 8.429839671392606e-05,
"loss": 2.7817665100097657,
"step": 4380
},
{
"epoch": 1.7427550615323542,
"grad_norm": 1.5466359853744507,
"learning_rate": 8.40333907512919e-05,
"loss": 2.8702486038208006,
"step": 4390
},
{
"epoch": 1.7467248908296944,
"grad_norm": 1.2387479543685913,
"learning_rate": 8.376838478865775e-05,
"loss": 2.744974899291992,
"step": 4400
},
{
"epoch": 1.7506947201270346,
"grad_norm": 1.3228589296340942,
"learning_rate": 8.350337882602359e-05,
"loss": 2.8770301818847654,
"step": 4410
},
{
"epoch": 1.7546645494243749,
"grad_norm": 1.5744258165359497,
"learning_rate": 8.323837286338943e-05,
"loss": 2.8095497131347655,
"step": 4420
},
{
"epoch": 1.7586343787217151,
"grad_norm": 1.2943992614746094,
"learning_rate": 8.297336690075527e-05,
"loss": 2.832097816467285,
"step": 4430
},
{
"epoch": 1.7626042080190552,
"grad_norm": 1.3430354595184326,
"learning_rate": 8.270836093812112e-05,
"loss": 2.7784374237060545,
"step": 4440
},
{
"epoch": 1.7665740373163954,
"grad_norm": 1.2830846309661865,
"learning_rate": 8.244335497548695e-05,
"loss": 2.8552547454833985,
"step": 4450
},
{
"epoch": 1.7705438666137356,
"grad_norm": 1.3120782375335693,
"learning_rate": 8.21783490128528e-05,
"loss": 2.7444028854370117,
"step": 4460
},
{
"epoch": 1.7745136959110757,
"grad_norm": 1.4138376712799072,
"learning_rate": 8.191334305021864e-05,
"loss": 2.858925628662109,
"step": 4470
},
{
"epoch": 1.778483525208416,
"grad_norm": 1.3830326795578003,
"learning_rate": 8.164833708758446e-05,
"loss": 2.922714614868164,
"step": 4480
},
{
"epoch": 1.7824533545057561,
"grad_norm": 1.632527470588684,
"learning_rate": 8.138333112495031e-05,
"loss": 2.792901039123535,
"step": 4490
},
{
"epoch": 1.7864231838030964,
"grad_norm": 1.2483190298080444,
"learning_rate": 8.111832516231616e-05,
"loss": 2.812052536010742,
"step": 4500
},
{
"epoch": 1.7903930131004366,
"grad_norm": 1.3889878988265991,
"learning_rate": 8.085331919968199e-05,
"loss": 2.801294708251953,
"step": 4510
},
{
"epoch": 1.7943628423977769,
"grad_norm": 1.2944639921188354,
"learning_rate": 8.058831323704784e-05,
"loss": 2.910211944580078,
"step": 4520
},
{
"epoch": 1.7983326716951171,
"grad_norm": 1.5790624618530273,
"learning_rate": 8.032330727441367e-05,
"loss": 2.836952781677246,
"step": 4530
},
{
"epoch": 1.8023025009924574,
"grad_norm": 1.2622289657592773,
"learning_rate": 8.005830131177952e-05,
"loss": 2.611729621887207,
"step": 4540
},
{
"epoch": 1.8062723302897976,
"grad_norm": 1.5101556777954102,
"learning_rate": 7.979329534914535e-05,
"loss": 2.816328430175781,
"step": 4550
},
{
"epoch": 1.8102421595871379,
"grad_norm": 1.2184019088745117,
"learning_rate": 7.95282893865112e-05,
"loss": 2.9140066146850585,
"step": 4560
},
{
"epoch": 1.814211988884478,
"grad_norm": 1.2687621116638184,
"learning_rate": 7.926328342387705e-05,
"loss": 2.780377197265625,
"step": 4570
},
{
"epoch": 1.8181818181818183,
"grad_norm": 1.6863418817520142,
"learning_rate": 7.899827746124288e-05,
"loss": 2.8988069534301757,
"step": 4580
},
{
"epoch": 1.8221516474791584,
"grad_norm": 1.3914945125579834,
"learning_rate": 7.873327149860872e-05,
"loss": 2.8935794830322266,
"step": 4590
},
{
"epoch": 1.8261214767764986,
"grad_norm": 1.3118557929992676,
"learning_rate": 7.846826553597456e-05,
"loss": 2.715951919555664,
"step": 4600
},
{
"epoch": 1.8300913060738389,
"grad_norm": 1.3855705261230469,
"learning_rate": 7.82032595733404e-05,
"loss": 2.86264705657959,
"step": 4610
},
{
"epoch": 1.8340611353711789,
"grad_norm": 1.5338412523269653,
"learning_rate": 7.793825361070625e-05,
"loss": 2.8531450271606444,
"step": 4620
},
{
"epoch": 1.8380309646685191,
"grad_norm": 1.235976219177246,
"learning_rate": 7.767324764807209e-05,
"loss": 2.9074989318847657,
"step": 4630
},
{
"epoch": 1.8420007939658594,
"grad_norm": 1.3477051258087158,
"learning_rate": 7.740824168543793e-05,
"loss": 2.757794189453125,
"step": 4640
},
{
"epoch": 1.8459706232631996,
"grad_norm": 1.3555721044540405,
"learning_rate": 7.714323572280376e-05,
"loss": 2.885786247253418,
"step": 4650
},
{
"epoch": 1.8499404525605398,
"grad_norm": 1.4486838579177856,
"learning_rate": 7.687822976016961e-05,
"loss": 2.816962242126465,
"step": 4660
},
{
"epoch": 1.85391028185788,
"grad_norm": 1.4096218347549438,
"learning_rate": 7.661322379753546e-05,
"loss": 2.817583465576172,
"step": 4670
},
{
"epoch": 1.8578801111552203,
"grad_norm": 1.3793836832046509,
"learning_rate": 7.634821783490129e-05,
"loss": 2.8635271072387694,
"step": 4680
},
{
"epoch": 1.8618499404525606,
"grad_norm": 1.192571997642517,
"learning_rate": 7.608321187226714e-05,
"loss": 2.86818790435791,
"step": 4690
},
{
"epoch": 1.8658197697499008,
"grad_norm": 1.4668769836425781,
"learning_rate": 7.581820590963297e-05,
"loss": 2.849831962585449,
"step": 4700
},
{
"epoch": 1.869789599047241,
"grad_norm": 1.4557112455368042,
"learning_rate": 7.55531999469988e-05,
"loss": 3.024089050292969,
"step": 4710
},
{
"epoch": 1.8737594283445813,
"grad_norm": 1.3126049041748047,
"learning_rate": 7.528819398436465e-05,
"loss": 2.8115358352661133,
"step": 4720
},
{
"epoch": 1.8777292576419216,
"grad_norm": 1.4214657545089722,
"learning_rate": 7.50231880217305e-05,
"loss": 2.9371572494506837,
"step": 4730
},
{
"epoch": 1.8816990869392616,
"grad_norm": 1.3505725860595703,
"learning_rate": 7.475818205909633e-05,
"loss": 2.790787124633789,
"step": 4740
},
{
"epoch": 1.8856689162366018,
"grad_norm": 1.5866847038269043,
"learning_rate": 7.449317609646217e-05,
"loss": 2.853039360046387,
"step": 4750
},
{
"epoch": 1.889638745533942,
"grad_norm": 1.3959614038467407,
"learning_rate": 7.422817013382801e-05,
"loss": 2.9222713470458985,
"step": 4760
},
{
"epoch": 1.8936085748312823,
"grad_norm": 1.214270830154419,
"learning_rate": 7.396316417119386e-05,
"loss": 2.780070495605469,
"step": 4770
},
{
"epoch": 1.8975784041286223,
"grad_norm": 1.179607629776001,
"learning_rate": 7.36981582085597e-05,
"loss": 2.9518333435058595,
"step": 4780
},
{
"epoch": 1.9015482334259626,
"grad_norm": 1.3093786239624023,
"learning_rate": 7.343315224592554e-05,
"loss": 2.809184455871582,
"step": 4790
},
{
"epoch": 1.9055180627233028,
"grad_norm": 1.2070693969726562,
"learning_rate": 7.316814628329138e-05,
"loss": 2.7718027114868162,
"step": 4800
},
{
"epoch": 1.909487892020643,
"grad_norm": 1.3962334394454956,
"learning_rate": 7.290314032065721e-05,
"loss": 2.8573562622070314,
"step": 4810
},
{
"epoch": 1.9134577213179833,
"grad_norm": 1.2028801441192627,
"learning_rate": 7.263813435802306e-05,
"loss": 2.845962715148926,
"step": 4820
},
{
"epoch": 1.9174275506153236,
"grad_norm": 1.3764971494674683,
"learning_rate": 7.23731283953889e-05,
"loss": 2.801960563659668,
"step": 4830
},
{
"epoch": 1.9213973799126638,
"grad_norm": 1.4100943803787231,
"learning_rate": 7.210812243275474e-05,
"loss": 3.0971054077148437,
"step": 4840
},
{
"epoch": 1.925367209210004,
"grad_norm": 1.2439861297607422,
"learning_rate": 7.184311647012059e-05,
"loss": 2.8608491897583006,
"step": 4850
},
{
"epoch": 1.9293370385073443,
"grad_norm": 1.5379648208618164,
"learning_rate": 7.157811050748642e-05,
"loss": 2.846979331970215,
"step": 4860
},
{
"epoch": 1.9333068678046845,
"grad_norm": 1.3032444715499878,
"learning_rate": 7.131310454485227e-05,
"loss": 2.9017236709594725,
"step": 4870
},
{
"epoch": 1.9372766971020248,
"grad_norm": 1.1875441074371338,
"learning_rate": 7.10480985822181e-05,
"loss": 2.7739852905273437,
"step": 4880
},
{
"epoch": 1.9412465263993648,
"grad_norm": 1.426823616027832,
"learning_rate": 7.078309261958395e-05,
"loss": 2.8642370223999025,
"step": 4890
},
{
"epoch": 1.945216355696705,
"grad_norm": 1.2701075077056885,
"learning_rate": 7.051808665694978e-05,
"loss": 2.7887300491333007,
"step": 4900
},
{
"epoch": 1.9491861849940453,
"grad_norm": 1.4069523811340332,
"learning_rate": 7.025308069431562e-05,
"loss": 2.9955364227294923,
"step": 4910
},
{
"epoch": 1.9531560142913855,
"grad_norm": 1.48947012424469,
"learning_rate": 6.998807473168146e-05,
"loss": 2.897355842590332,
"step": 4920
},
{
"epoch": 1.9571258435887255,
"grad_norm": 1.2420414686203003,
"learning_rate": 6.972306876904731e-05,
"loss": 2.952886772155762,
"step": 4930
},
{
"epoch": 1.9610956728860658,
"grad_norm": 1.5975414514541626,
"learning_rate": 6.945806280641314e-05,
"loss": 2.7640342712402344,
"step": 4940
},
{
"epoch": 1.965065502183406,
"grad_norm": 1.2006224393844604,
"learning_rate": 6.919305684377899e-05,
"loss": 2.708919334411621,
"step": 4950
},
{
"epoch": 1.9690353314807463,
"grad_norm": 1.7260173559188843,
"learning_rate": 6.892805088114484e-05,
"loss": 2.7740182876586914,
"step": 4960
},
{
"epoch": 1.9730051607780865,
"grad_norm": 1.3943583965301514,
"learning_rate": 6.866304491851066e-05,
"loss": 2.823019790649414,
"step": 4970
},
{
"epoch": 1.9769749900754268,
"grad_norm": 1.3960636854171753,
"learning_rate": 6.839803895587651e-05,
"loss": 2.8009057998657227,
"step": 4980
},
{
"epoch": 1.980944819372767,
"grad_norm": 1.3609869480133057,
"learning_rate": 6.813303299324235e-05,
"loss": 2.8227596282958984,
"step": 4990
},
{
"epoch": 1.9849146486701073,
"grad_norm": 1.3699599504470825,
"learning_rate": 6.786802703060819e-05,
"loss": 2.91951961517334,
"step": 5000
},
{
"epoch": 1.9888844779674475,
"grad_norm": 1.3474290370941162,
"learning_rate": 6.760302106797404e-05,
"loss": 2.9053911209106444,
"step": 5010
},
{
"epoch": 1.9928543072647877,
"grad_norm": 1.3683226108551025,
"learning_rate": 6.733801510533987e-05,
"loss": 2.8106914520263673,
"step": 5020
},
{
"epoch": 1.996824136562128,
"grad_norm": 1.2184762954711914,
"learning_rate": 6.707300914270572e-05,
"loss": 2.732072448730469,
"step": 5030
},
{
"epoch": 2.0007939658594682,
"grad_norm": 1.539412021636963,
"learning_rate": 6.680800318007155e-05,
"loss": 2.8226951599121093,
"step": 5040
},
{
"epoch": 2.0047637951568085,
"grad_norm": 1.2428901195526123,
"learning_rate": 6.65429972174374e-05,
"loss": 2.836050033569336,
"step": 5050
},
{
"epoch": 2.0087336244541483,
"grad_norm": 1.3540750741958618,
"learning_rate": 6.627799125480325e-05,
"loss": 2.7182661056518556,
"step": 5060
},
{
"epoch": 2.0127034537514885,
"grad_norm": 1.2656959295272827,
"learning_rate": 6.601298529216907e-05,
"loss": 2.694625663757324,
"step": 5070
},
{
"epoch": 2.0166732830488288,
"grad_norm": 1.2895488739013672,
"learning_rate": 6.574797932953491e-05,
"loss": 2.7980525970458983,
"step": 5080
},
{
"epoch": 2.020643112346169,
"grad_norm": 1.299148678779602,
"learning_rate": 6.548297336690076e-05,
"loss": 2.7926309585571287,
"step": 5090
},
{
"epoch": 2.0246129416435092,
"grad_norm": 1.5304101705551147,
"learning_rate": 6.52179674042666e-05,
"loss": 2.7923309326171877,
"step": 5100
},
{
"epoch": 2.0285827709408495,
"grad_norm": 1.2151813507080078,
"learning_rate": 6.495296144163244e-05,
"loss": 2.867671585083008,
"step": 5110
},
{
"epoch": 2.0325526002381897,
"grad_norm": 1.2785301208496094,
"learning_rate": 6.468795547899829e-05,
"loss": 2.8732872009277344,
"step": 5120
},
{
"epoch": 2.03652242953553,
"grad_norm": 1.3549985885620117,
"learning_rate": 6.442294951636412e-05,
"loss": 2.8315383911132814,
"step": 5130
},
{
"epoch": 2.04049225883287,
"grad_norm": 1.2033029794692993,
"learning_rate": 6.415794355372996e-05,
"loss": 2.8878578186035155,
"step": 5140
},
{
"epoch": 2.0444620881302105,
"grad_norm": 1.4432929754257202,
"learning_rate": 6.38929375910958e-05,
"loss": 2.7286848068237304,
"step": 5150
},
{
"epoch": 2.0484319174275507,
"grad_norm": 1.0933128595352173,
"learning_rate": 6.362793162846165e-05,
"loss": 2.6944265365600586,
"step": 5160
},
{
"epoch": 2.052401746724891,
"grad_norm": 1.4498447179794312,
"learning_rate": 6.336292566582748e-05,
"loss": 2.8956008911132813,
"step": 5170
},
{
"epoch": 2.056371576022231,
"grad_norm": 1.3581401109695435,
"learning_rate": 6.309791970319332e-05,
"loss": 2.8658390045166016,
"step": 5180
},
{
"epoch": 2.0603414053195714,
"grad_norm": 1.6925913095474243,
"learning_rate": 6.283291374055917e-05,
"loss": 2.71678466796875,
"step": 5190
},
{
"epoch": 2.0643112346169117,
"grad_norm": 1.4804787635803223,
"learning_rate": 6.2567907777925e-05,
"loss": 2.772994041442871,
"step": 5200
},
{
"epoch": 2.0682810639142515,
"grad_norm": 1.6423425674438477,
"learning_rate": 6.230290181529085e-05,
"loss": 2.7334659576416014,
"step": 5210
},
{
"epoch": 2.0722508932115917,
"grad_norm": 1.4786124229431152,
"learning_rate": 6.20378958526567e-05,
"loss": 2.7266508102416993,
"step": 5220
},
{
"epoch": 2.076220722508932,
"grad_norm": 1.6273307800292969,
"learning_rate": 6.177288989002253e-05,
"loss": 2.896736717224121,
"step": 5230
},
{
"epoch": 2.080190551806272,
"grad_norm": 1.372231125831604,
"learning_rate": 6.150788392738836e-05,
"loss": 2.8753612518310545,
"step": 5240
},
{
"epoch": 2.0841603811036125,
"grad_norm": 1.0888550281524658,
"learning_rate": 6.124287796475421e-05,
"loss": 2.6621337890625,
"step": 5250
},
{
"epoch": 2.0881302104009527,
"grad_norm": 1.3228888511657715,
"learning_rate": 6.097787200212005e-05,
"loss": 2.8179800033569338,
"step": 5260
},
{
"epoch": 2.092100039698293,
"grad_norm": 1.3481526374816895,
"learning_rate": 6.071286603948589e-05,
"loss": 2.6514272689819336,
"step": 5270
},
{
"epoch": 2.096069868995633,
"grad_norm": 2.2035071849823,
"learning_rate": 6.044786007685174e-05,
"loss": 2.8128807067871096,
"step": 5280
},
{
"epoch": 2.1000396982929734,
"grad_norm": 1.4449114799499512,
"learning_rate": 6.0182854114217565e-05,
"loss": 2.7257884979248046,
"step": 5290
},
{
"epoch": 2.1040095275903137,
"grad_norm": 1.4004982709884644,
"learning_rate": 5.991784815158341e-05,
"loss": 2.8827852249145507,
"step": 5300
},
{
"epoch": 2.107979356887654,
"grad_norm": 1.5591332912445068,
"learning_rate": 5.965284218894925e-05,
"loss": 2.759036636352539,
"step": 5310
},
{
"epoch": 2.111949186184994,
"grad_norm": 1.5215182304382324,
"learning_rate": 5.9387836226315094e-05,
"loss": 2.7620918273925783,
"step": 5320
},
{
"epoch": 2.1159190154823344,
"grad_norm": 1.4247337579727173,
"learning_rate": 5.912283026368094e-05,
"loss": 2.6711416244506836,
"step": 5330
},
{
"epoch": 2.1198888447796747,
"grad_norm": 1.348399043083191,
"learning_rate": 5.885782430104677e-05,
"loss": 2.7836679458618163,
"step": 5340
},
{
"epoch": 2.123858674077015,
"grad_norm": 1.6695053577423096,
"learning_rate": 5.8592818338412615e-05,
"loss": 2.748013496398926,
"step": 5350
},
{
"epoch": 2.1278285033743547,
"grad_norm": 1.5424185991287231,
"learning_rate": 5.8327812375778456e-05,
"loss": 2.8054782867431642,
"step": 5360
},
{
"epoch": 2.131798332671695,
"grad_norm": 1.4576644897460938,
"learning_rate": 5.8062806413144297e-05,
"loss": 2.8069080352783202,
"step": 5370
},
{
"epoch": 2.135768161969035,
"grad_norm": 1.3622318506240845,
"learning_rate": 5.7797800450510144e-05,
"loss": 2.732857894897461,
"step": 5380
},
{
"epoch": 2.1397379912663754,
"grad_norm": 1.2806458473205566,
"learning_rate": 5.7532794487875985e-05,
"loss": 2.7497119903564453,
"step": 5390
},
{
"epoch": 2.1437078205637157,
"grad_norm": 1.2458010911941528,
"learning_rate": 5.726778852524182e-05,
"loss": 2.8465991973876954,
"step": 5400
},
{
"epoch": 2.147677649861056,
"grad_norm": 1.266097903251648,
"learning_rate": 5.700278256260766e-05,
"loss": 2.7699043273925783,
"step": 5410
},
{
"epoch": 2.151647479158396,
"grad_norm": 1.7706705331802368,
"learning_rate": 5.67377765999735e-05,
"loss": 2.890908622741699,
"step": 5420
},
{
"epoch": 2.1556173084557364,
"grad_norm": 1.672590732574463,
"learning_rate": 5.647277063733935e-05,
"loss": 2.7777748107910156,
"step": 5430
},
{
"epoch": 2.1595871377530766,
"grad_norm": 1.4306832551956177,
"learning_rate": 5.620776467470519e-05,
"loss": 2.8492876052856446,
"step": 5440
},
{
"epoch": 2.163556967050417,
"grad_norm": 1.3680981397628784,
"learning_rate": 5.594275871207102e-05,
"loss": 2.8353946685791014,
"step": 5450
},
{
"epoch": 2.167526796347757,
"grad_norm": 1.3657315969467163,
"learning_rate": 5.567775274943686e-05,
"loss": 2.784601593017578,
"step": 5460
},
{
"epoch": 2.1714966256450974,
"grad_norm": 1.4296693801879883,
"learning_rate": 5.54127467868027e-05,
"loss": 2.7476505279541015,
"step": 5470
},
{
"epoch": 2.1754664549424376,
"grad_norm": 1.3640064001083374,
"learning_rate": 5.514774082416855e-05,
"loss": 2.7358591079711916,
"step": 5480
},
{
"epoch": 2.179436284239778,
"grad_norm": 1.3160929679870605,
"learning_rate": 5.488273486153439e-05,
"loss": 2.7519712448120117,
"step": 5490
},
{
"epoch": 2.183406113537118,
"grad_norm": 1.2436435222625732,
"learning_rate": 5.461772889890023e-05,
"loss": 2.8443069458007812,
"step": 5500
},
{
"epoch": 2.187375942834458,
"grad_norm": 1.849849820137024,
"learning_rate": 5.4352722936266065e-05,
"loss": 2.8738945007324217,
"step": 5510
},
{
"epoch": 2.191345772131798,
"grad_norm": 1.4771231412887573,
"learning_rate": 5.4087716973631905e-05,
"loss": 2.8680547714233398,
"step": 5520
},
{
"epoch": 2.1953156014291384,
"grad_norm": 1.2482953071594238,
"learning_rate": 5.382271101099775e-05,
"loss": 2.8199131011962892,
"step": 5530
},
{
"epoch": 2.1992854307264786,
"grad_norm": 1.4146586656570435,
"learning_rate": 5.355770504836359e-05,
"loss": 2.853030204772949,
"step": 5540
},
{
"epoch": 2.203255260023819,
"grad_norm": 1.3868707418441772,
"learning_rate": 5.3292699085729434e-05,
"loss": 2.830170440673828,
"step": 5550
},
{
"epoch": 2.207225089321159,
"grad_norm": 1.6373202800750732,
"learning_rate": 5.302769312309527e-05,
"loss": 2.839306640625,
"step": 5560
},
{
"epoch": 2.2111949186184994,
"grad_norm": 1.8044177293777466,
"learning_rate": 5.276268716046111e-05,
"loss": 2.6578227996826174,
"step": 5570
},
{
"epoch": 2.2151647479158396,
"grad_norm": 1.2935256958007812,
"learning_rate": 5.2497681197826956e-05,
"loss": 2.7891679763793946,
"step": 5580
},
{
"epoch": 2.21913457721318,
"grad_norm": 1.8723143339157104,
"learning_rate": 5.2232675235192796e-05,
"loss": 2.8123620986938476,
"step": 5590
},
{
"epoch": 2.22310440651052,
"grad_norm": 1.4469753503799438,
"learning_rate": 5.196766927255864e-05,
"loss": 2.7508686065673826,
"step": 5600
},
{
"epoch": 2.2270742358078603,
"grad_norm": 1.2512681484222412,
"learning_rate": 5.170266330992447e-05,
"loss": 2.814630126953125,
"step": 5610
},
{
"epoch": 2.2310440651052006,
"grad_norm": 1.4430116415023804,
"learning_rate": 5.143765734729031e-05,
"loss": 2.683193016052246,
"step": 5620
},
{
"epoch": 2.235013894402541,
"grad_norm": 1.8242406845092773,
"learning_rate": 5.117265138465616e-05,
"loss": 2.685003662109375,
"step": 5630
},
{
"epoch": 2.238983723699881,
"grad_norm": 1.5419832468032837,
"learning_rate": 5.0907645422022e-05,
"loss": 2.7486963272094727,
"step": 5640
},
{
"epoch": 2.2429535529972213,
"grad_norm": 1.4663166999816895,
"learning_rate": 5.064263945938784e-05,
"loss": 2.775518608093262,
"step": 5650
},
{
"epoch": 2.246923382294561,
"grad_norm": 1.4978145360946655,
"learning_rate": 5.037763349675369e-05,
"loss": 2.7634124755859375,
"step": 5660
},
{
"epoch": 2.2508932115919014,
"grad_norm": 1.3765963315963745,
"learning_rate": 5.0112627534119514e-05,
"loss": 2.7787755966186523,
"step": 5670
},
{
"epoch": 2.2548630408892416,
"grad_norm": 1.3982337713241577,
"learning_rate": 4.984762157148536e-05,
"loss": 2.7080947875976564,
"step": 5680
},
{
"epoch": 2.258832870186582,
"grad_norm": 1.8514288663864136,
"learning_rate": 4.95826156088512e-05,
"loss": 2.806821823120117,
"step": 5690
},
{
"epoch": 2.262802699483922,
"grad_norm": 1.647444486618042,
"learning_rate": 4.931760964621704e-05,
"loss": 2.866771697998047,
"step": 5700
},
{
"epoch": 2.2667725287812623,
"grad_norm": 1.4562606811523438,
"learning_rate": 4.905260368358288e-05,
"loss": 2.8650636672973633,
"step": 5710
},
{
"epoch": 2.2707423580786026,
"grad_norm": 1.514732003211975,
"learning_rate": 4.8787597720948724e-05,
"loss": 2.725859260559082,
"step": 5720
},
{
"epoch": 2.274712187375943,
"grad_norm": 1.3854975700378418,
"learning_rate": 4.8522591758314565e-05,
"loss": 2.778037452697754,
"step": 5730
},
{
"epoch": 2.278682016673283,
"grad_norm": 1.5461442470550537,
"learning_rate": 4.8257585795680405e-05,
"loss": 2.8365272521972655,
"step": 5740
},
{
"epoch": 2.2826518459706233,
"grad_norm": 1.4813385009765625,
"learning_rate": 4.7992579833046246e-05,
"loss": 2.637844467163086,
"step": 5750
},
{
"epoch": 2.2866216752679636,
"grad_norm": 1.2352749109268188,
"learning_rate": 4.7727573870412086e-05,
"loss": 2.7269769668579102,
"step": 5760
},
{
"epoch": 2.290591504565304,
"grad_norm": 1.2570537328720093,
"learning_rate": 4.746256790777793e-05,
"loss": 2.6165693283081053,
"step": 5770
},
{
"epoch": 2.294561333862644,
"grad_norm": 1.2438743114471436,
"learning_rate": 4.719756194514377e-05,
"loss": 2.817479705810547,
"step": 5780
},
{
"epoch": 2.2985311631599843,
"grad_norm": 1.5795632600784302,
"learning_rate": 4.693255598250961e-05,
"loss": 2.808450698852539,
"step": 5790
},
{
"epoch": 2.3025009924573245,
"grad_norm": 1.2858535051345825,
"learning_rate": 4.666755001987545e-05,
"loss": 2.8606801986694337,
"step": 5800
},
{
"epoch": 2.3064708217546643,
"grad_norm": 1.8110570907592773,
"learning_rate": 4.6402544057241296e-05,
"loss": 2.8314798355102537,
"step": 5810
},
{
"epoch": 2.3104406510520046,
"grad_norm": 1.4152132272720337,
"learning_rate": 4.613753809460713e-05,
"loss": 2.8347152709960937,
"step": 5820
},
{
"epoch": 2.314410480349345,
"grad_norm": 1.5837576389312744,
"learning_rate": 4.587253213197297e-05,
"loss": 2.7796628952026365,
"step": 5830
},
{
"epoch": 2.318380309646685,
"grad_norm": 1.35812246799469,
"learning_rate": 4.560752616933881e-05,
"loss": 2.6421552658081056,
"step": 5840
},
{
"epoch": 2.3223501389440253,
"grad_norm": 1.700376033782959,
"learning_rate": 4.534252020670465e-05,
"loss": 2.794668197631836,
"step": 5850
},
{
"epoch": 2.3263199682413656,
"grad_norm": 1.3668570518493652,
"learning_rate": 4.50775142440705e-05,
"loss": 2.721013069152832,
"step": 5860
},
{
"epoch": 2.330289797538706,
"grad_norm": 1.268725872039795,
"learning_rate": 4.481250828143633e-05,
"loss": 2.8375295639038085,
"step": 5870
},
{
"epoch": 2.334259626836046,
"grad_norm": 1.4931402206420898,
"learning_rate": 4.454750231880217e-05,
"loss": 2.852950668334961,
"step": 5880
},
{
"epoch": 2.3382294561333863,
"grad_norm": 1.426378607749939,
"learning_rate": 4.428249635616802e-05,
"loss": 2.81459903717041,
"step": 5890
},
{
"epoch": 2.3421992854307265,
"grad_norm": 1.5145498514175415,
"learning_rate": 4.4017490393533855e-05,
"loss": 2.8200088500976563,
"step": 5900
},
{
"epoch": 2.3461691147280668,
"grad_norm": 1.280332326889038,
"learning_rate": 4.3752484430899695e-05,
"loss": 2.7844158172607423,
"step": 5910
},
{
"epoch": 2.350138944025407,
"grad_norm": 1.3290445804595947,
"learning_rate": 4.3487478468265536e-05,
"loss": 2.713772773742676,
"step": 5920
},
{
"epoch": 2.3541087733227473,
"grad_norm": 1.8138554096221924,
"learning_rate": 4.3222472505631376e-05,
"loss": 2.605319023132324,
"step": 5930
},
{
"epoch": 2.3580786026200875,
"grad_norm": 1.9153021574020386,
"learning_rate": 4.2957466542997224e-05,
"loss": 2.8222110748291014,
"step": 5940
},
{
"epoch": 2.3620484319174277,
"grad_norm": 1.6975692510604858,
"learning_rate": 4.269246058036306e-05,
"loss": 2.794956588745117,
"step": 5950
},
{
"epoch": 2.3660182612147675,
"grad_norm": 1.559992790222168,
"learning_rate": 4.24274546177289e-05,
"loss": 2.7832340240478515,
"step": 5960
},
{
"epoch": 2.369988090512108,
"grad_norm": 1.5059814453125,
"learning_rate": 4.2162448655094745e-05,
"loss": 2.760076141357422,
"step": 5970
},
{
"epoch": 2.373957919809448,
"grad_norm": 1.3382420539855957,
"learning_rate": 4.189744269246058e-05,
"loss": 2.719432067871094,
"step": 5980
},
{
"epoch": 2.3779277491067883,
"grad_norm": 1.3866031169891357,
"learning_rate": 4.1632436729826427e-05,
"loss": 2.7717851638793944,
"step": 5990
},
{
"epoch": 2.3818975784041285,
"grad_norm": 1.3926609754562378,
"learning_rate": 4.136743076719226e-05,
"loss": 2.861334037780762,
"step": 6000
},
{
"epoch": 2.3858674077014688,
"grad_norm": 1.8084567785263062,
"learning_rate": 4.112892540082152e-05,
"loss": 2.9400596618652344,
"step": 6010
},
{
"epoch": 2.389837236998809,
"grad_norm": 1.3029987812042236,
"learning_rate": 4.086391943818736e-05,
"loss": 2.702388381958008,
"step": 6020
},
{
"epoch": 2.3938070662961493,
"grad_norm": 1.6704697608947754,
"learning_rate": 4.05989134755532e-05,
"loss": 2.894243049621582,
"step": 6030
},
{
"epoch": 2.3977768955934895,
"grad_norm": 1.625575065612793,
"learning_rate": 4.033390751291904e-05,
"loss": 2.8416337966918945,
"step": 6040
},
{
"epoch": 2.4017467248908297,
"grad_norm": 1.3440985679626465,
"learning_rate": 4.006890155028488e-05,
"loss": 2.732625389099121,
"step": 6050
},
{
"epoch": 2.40571655418817,
"grad_norm": 1.6972922086715698,
"learning_rate": 3.980389558765072e-05,
"loss": 2.9266534805297852,
"step": 6060
},
{
"epoch": 2.4096863834855102,
"grad_norm": 1.3665430545806885,
"learning_rate": 3.953888962501657e-05,
"loss": 2.6411718368530273,
"step": 6070
},
{
"epoch": 2.4136562127828505,
"grad_norm": 1.355370283126831,
"learning_rate": 3.9273883662382404e-05,
"loss": 2.7438932418823243,
"step": 6080
},
{
"epoch": 2.4176260420801907,
"grad_norm": 1.483780860900879,
"learning_rate": 3.9008877699748245e-05,
"loss": 2.8010843276977537,
"step": 6090
},
{
"epoch": 2.421595871377531,
"grad_norm": 1.7616933584213257,
"learning_rate": 3.8743871737114086e-05,
"loss": 2.8067392349243163,
"step": 6100
},
{
"epoch": 2.4255657006748708,
"grad_norm": 1.3735806941986084,
"learning_rate": 3.8478865774479926e-05,
"loss": 2.817402648925781,
"step": 6110
},
{
"epoch": 2.429535529972211,
"grad_norm": 1.2895877361297607,
"learning_rate": 3.8213859811845774e-05,
"loss": 2.8315813064575197,
"step": 6120
},
{
"epoch": 2.4335053592695512,
"grad_norm": 1.4236490726470947,
"learning_rate": 3.794885384921161e-05,
"loss": 2.7428449630737304,
"step": 6130
},
{
"epoch": 2.4374751885668915,
"grad_norm": 1.5362484455108643,
"learning_rate": 3.768384788657745e-05,
"loss": 2.6830842971801756,
"step": 6140
},
{
"epoch": 2.4414450178642317,
"grad_norm": 1.3560556173324585,
"learning_rate": 3.7418841923943295e-05,
"loss": 2.847624397277832,
"step": 6150
},
{
"epoch": 2.445414847161572,
"grad_norm": 1.5367388725280762,
"learning_rate": 3.715383596130913e-05,
"loss": 2.837147331237793,
"step": 6160
},
{
"epoch": 2.4493846764589122,
"grad_norm": 1.4494478702545166,
"learning_rate": 3.688882999867497e-05,
"loss": 2.6767187118530273,
"step": 6170
},
{
"epoch": 2.4533545057562525,
"grad_norm": 1.8024524450302124,
"learning_rate": 3.662382403604081e-05,
"loss": 2.785817527770996,
"step": 6180
},
{
"epoch": 2.4573243350535927,
"grad_norm": 1.6272693872451782,
"learning_rate": 3.635881807340665e-05,
"loss": 2.9185848236083984,
"step": 6190
},
{
"epoch": 2.461294164350933,
"grad_norm": 1.4263159036636353,
"learning_rate": 3.60938121107725e-05,
"loss": 2.6653270721435547,
"step": 6200
},
{
"epoch": 2.465263993648273,
"grad_norm": 1.450008511543274,
"learning_rate": 3.582880614813833e-05,
"loss": 2.7907215118408204,
"step": 6210
},
{
"epoch": 2.4692338229456134,
"grad_norm": 1.6371126174926758,
"learning_rate": 3.556380018550417e-05,
"loss": 2.788674736022949,
"step": 6220
},
{
"epoch": 2.4732036522429537,
"grad_norm": 1.4910118579864502,
"learning_rate": 3.529879422287002e-05,
"loss": 2.8709230422973633,
"step": 6230
},
{
"epoch": 2.477173481540294,
"grad_norm": 1.7096924781799316,
"learning_rate": 3.5033788260235854e-05,
"loss": 2.8957002639770506,
"step": 6240
},
{
"epoch": 2.481143310837634,
"grad_norm": 1.4181935787200928,
"learning_rate": 3.47687822976017e-05,
"loss": 2.8151878356933593,
"step": 6250
},
{
"epoch": 2.485113140134974,
"grad_norm": 1.5911320447921753,
"learning_rate": 3.4503776334967535e-05,
"loss": 2.771291732788086,
"step": 6260
},
{
"epoch": 2.489082969432314,
"grad_norm": 1.488732933998108,
"learning_rate": 3.4238770372333376e-05,
"loss": 2.85223503112793,
"step": 6270
},
{
"epoch": 2.4930527987296545,
"grad_norm": 2.1942832469940186,
"learning_rate": 3.397376440969922e-05,
"loss": 2.972767639160156,
"step": 6280
},
{
"epoch": 2.4970226280269947,
"grad_norm": 1.427437424659729,
"learning_rate": 3.370875844706506e-05,
"loss": 2.80262451171875,
"step": 6290
},
{
"epoch": 2.500992457324335,
"grad_norm": 1.4462391138076782,
"learning_rate": 3.3443752484430904e-05,
"loss": 2.9569679260253907,
"step": 6300
},
{
"epoch": 2.504962286621675,
"grad_norm": 1.2452316284179688,
"learning_rate": 3.3178746521796745e-05,
"loss": 2.791957664489746,
"step": 6310
},
{
"epoch": 2.5089321159190154,
"grad_norm": 1.22199285030365,
"learning_rate": 3.291374055916258e-05,
"loss": 2.9822917938232423,
"step": 6320
},
{
"epoch": 2.5129019452163557,
"grad_norm": 1.3499411344528198,
"learning_rate": 3.2648734596528426e-05,
"loss": 2.8133171081542967,
"step": 6330
},
{
"epoch": 2.516871774513696,
"grad_norm": 1.2252686023712158,
"learning_rate": 3.2383728633894267e-05,
"loss": 2.8636957168579102,
"step": 6340
},
{
"epoch": 2.520841603811036,
"grad_norm": 1.4534387588500977,
"learning_rate": 3.211872267126011e-05,
"loss": 2.7793609619140627,
"step": 6350
},
{
"epoch": 2.5248114331083764,
"grad_norm": 1.6919007301330566,
"learning_rate": 3.185371670862595e-05,
"loss": 2.747213363647461,
"step": 6360
},
{
"epoch": 2.5287812624057167,
"grad_norm": 1.3403940200805664,
"learning_rate": 3.158871074599178e-05,
"loss": 2.6384368896484376,
"step": 6370
},
{
"epoch": 2.532751091703057,
"grad_norm": 1.5563888549804688,
"learning_rate": 3.132370478335763e-05,
"loss": 2.7293819427490233,
"step": 6380
},
{
"epoch": 2.536720921000397,
"grad_norm": 1.7954516410827637,
"learning_rate": 3.105869882072347e-05,
"loss": 2.8856603622436525,
"step": 6390
},
{
"epoch": 2.5406907502977374,
"grad_norm": 1.6983191967010498,
"learning_rate": 3.079369285808931e-05,
"loss": 2.8485260009765625,
"step": 6400
}
],
"logging_steps": 10,
"max_steps": 7557,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.8126554281312256e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}