Files
Qwen3-1.7B-SFT-100k/trainer_state.json

1087 lines
28 KiB
JSON
Raw Normal View History

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 1491,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02014098690835851,
"grad_norm": 17.982622939860434,
"learning_rate": 4.000000000000001e-06,
"loss": 3.073244094848633,
"step": 10
},
{
"epoch": 0.04028197381671702,
"grad_norm": 4.416304162538307,
"learning_rate": 8.444444444444446e-06,
"loss": 2.2913312911987305,
"step": 20
},
{
"epoch": 0.06042296072507553,
"grad_norm": 2.3255969684771034,
"learning_rate": 1.288888888888889e-05,
"loss": 1.5302621841430664,
"step": 30
},
{
"epoch": 0.08056394763343404,
"grad_norm": 3.3171831937075003,
"learning_rate": 1.7333333333333336e-05,
"loss": 1.0828689575195312,
"step": 40
},
{
"epoch": 0.10070493454179255,
"grad_norm": 0.8556098890983025,
"learning_rate": 1.9999622383977725e-05,
"loss": 0.8968555450439453,
"step": 50
},
{
"epoch": 0.12084592145015106,
"grad_norm": 0.8935143428003198,
"learning_rate": 1.999537453124055e-05,
"loss": 0.8666976928710938,
"step": 60
},
{
"epoch": 0.14098690835850958,
"grad_norm": 0.9119446875586649,
"learning_rate": 1.998640881742778e-05,
"loss": 0.8250602722167969,
"step": 70
},
{
"epoch": 0.16112789526686808,
"grad_norm": 0.7723616857133194,
"learning_rate": 1.997272947439608e-05,
"loss": 0.8222087860107422,
"step": 80
},
{
"epoch": 0.18126888217522658,
"grad_norm": 0.7086293310893315,
"learning_rate": 1.995434295885598e-05,
"loss": 0.7954679489135742,
"step": 90
},
{
"epoch": 0.2014098690835851,
"grad_norm": 0.8182937768539513,
"learning_rate": 1.993125794932429e-05,
"loss": 0.8072474479675293,
"step": 100
},
{
"epoch": 0.2215508559919436,
"grad_norm": 0.7354938507346823,
"learning_rate": 1.9903485342027827e-05,
"loss": 0.8378339767456054,
"step": 110
},
{
"epoch": 0.24169184290030213,
"grad_norm": 0.7384628140878121,
"learning_rate": 1.9871038245760305e-05,
"loss": 0.7969758987426758,
"step": 120
},
{
"epoch": 0.2618328298086606,
"grad_norm": 0.7270112546895067,
"learning_rate": 1.983393197569497e-05,
"loss": 0.837348747253418,
"step": 130
},
{
"epoch": 0.28197381671701915,
"grad_norm": 1.152458861127753,
"learning_rate": 1.9792184046155733e-05,
"loss": 0.7704894065856933,
"step": 140
},
{
"epoch": 0.3021148036253776,
"grad_norm": 0.7830669702448827,
"learning_rate": 1.974581416235035e-05,
"loss": 0.7638654708862305,
"step": 150
},
{
"epoch": 0.32225579053373615,
"grad_norm": 0.662909194939404,
"learning_rate": 1.9694844211069477e-05,
"loss": 0.7630445480346679,
"step": 160
},
{
"epoch": 0.3423967774420947,
"grad_norm": 0.6068166763286809,
"learning_rate": 1.963929825035601e-05,
"loss": 0.79698166847229,
"step": 170
},
{
"epoch": 0.36253776435045315,
"grad_norm": 0.7231230225443792,
"learning_rate": 1.9579202498149562e-05,
"loss": 0.7727674961090087,
"step": 180
},
{
"epoch": 0.3826787512588117,
"grad_norm": 0.6798705825407562,
"learning_rate": 1.951458531991151e-05,
"loss": 0.7765314102172851,
"step": 190
},
{
"epoch": 0.4028197381671702,
"grad_norm": 0.601709580862295,
"learning_rate": 1.9445477215236343e-05,
"loss": 0.7855539321899414,
"step": 200
},
{
"epoch": 0.4229607250755287,
"grad_norm": 0.7559675683509207,
"learning_rate": 1.937191080345574e-05,
"loss": 0.7797715187072753,
"step": 210
},
{
"epoch": 0.4431017119838872,
"grad_norm": 0.6437182934735217,
"learning_rate": 1.9293920808242084e-05,
"loss": 0.7819314956665039,
"step": 220
},
{
"epoch": 0.46324269889224573,
"grad_norm": 0.6520667495447376,
"learning_rate": 1.9211544041218752e-05,
"loss": 0.7715907096862793,
"step": 230
},
{
"epoch": 0.48338368580060426,
"grad_norm": 0.6479556212268033,
"learning_rate": 1.912481938458485e-05,
"loss": 0.7307522296905518,
"step": 240
},
{
"epoch": 0.5035246727089627,
"grad_norm": 0.6205427420369014,
"learning_rate": 1.9033787772762647e-05,
"loss": 0.7505370616912842,
"step": 250
},
{
"epoch": 0.5236656596173213,
"grad_norm": 0.6301103631446919,
"learning_rate": 1.893849217307635e-05,
"loss": 0.7448963165283203,
"step": 260
},
{
"epoch": 0.5438066465256798,
"grad_norm": 0.6054604049411163,
"learning_rate": 1.8838977565471343e-05,
"loss": 0.7512392044067383,
"step": 270
},
{
"epoch": 0.5639476334340383,
"grad_norm": 0.6641842324027172,
"learning_rate": 1.873529092128343e-05,
"loss": 0.7734776496887207,
"step": 280
},
{
"epoch": 0.5840886203423967,
"grad_norm": 0.6881766972436127,
"learning_rate": 1.8627481181068185e-05,
"loss": 0.7516597747802735,
"step": 290
},
{
"epoch": 0.6042296072507553,
"grad_norm": 0.7114932246416025,
"learning_rate": 1.851559923150077e-05,
"loss": 0.7655241966247559,
"step": 300
},
{
"epoch": 0.6243705941591138,
"grad_norm": 0.6808053148104048,
"learning_rate": 1.8399697881357214e-05,
"loss": 0.7743952751159668,
"step": 310
},
{
"epoch": 0.6445115810674723,
"grad_norm": 0.633401029258308,
"learning_rate": 1.8279831836588427e-05,
"loss": 0.7196044445037841,
"step": 320
},
{
"epoch": 0.6646525679758308,
"grad_norm": 0.635189452406444,
"learning_rate": 1.8156057674498756e-05,
"loss": 0.7735816955566406,
"step": 330
},
{
"epoch": 0.6847935548841894,
"grad_norm": 0.6079022064378636,
"learning_rate": 1.8028433817041237e-05,
"loss": 0.7698081970214844,
"step": 340
},
{
"epoch": 0.7049345417925479,
"grad_norm": 0.6691748614288394,
"learning_rate": 1.7897020503242192e-05,
"loss": 0.7545667171478272,
"step": 350
},
{
"epoch": 0.7250755287009063,
"grad_norm": 0.6704403785715234,
"learning_rate": 1.7761879760768123e-05,
"loss": 0.7272197723388671,
"step": 360
},
{
"epoch": 0.7452165156092648,
"grad_norm": 0.6248181535199508,
"learning_rate": 1.7623075376648374e-05,
"loss": 0.735597801208496,
"step": 370
},
{
"epoch": 0.7653575025176234,
"grad_norm": 0.6691162481904424,
"learning_rate": 1.748067286716741e-05,
"loss": 0.7506361484527588,
"step": 380
},
{
"epoch": 0.7854984894259819,
"grad_norm": 0.5985017428584295,
"learning_rate": 1.7334739446940785e-05,
"loss": 0.7382652282714843,
"step": 390
},
{
"epoch": 0.8056394763343404,
"grad_norm": 0.674850678027072,
"learning_rate": 1.718534399718959e-05,
"loss": 0.7026462554931641,
"step": 400
},
{
"epoch": 0.8257804632426989,
"grad_norm": 0.575762458028171,
"learning_rate": 1.7032557033228184e-05,
"loss": 0.7259914398193359,
"step": 410
},
{
"epoch": 0.8459214501510574,
"grad_norm": 0.6224947992992479,
"learning_rate": 1.6876450671180667e-05,
"loss": 0.7931873798370361,
"step": 420
},
{
"epoch": 0.8660624370594159,
"grad_norm": 0.6047243356309274,
"learning_rate": 1.6717098593941753e-05,
"loss": 0.726997709274292,
"step": 430
},
{
"epoch": 0.8862034239677744,
"grad_norm": 0.618674243738717,
"learning_rate": 1.65545760163981e-05,
"loss": 0.7656588554382324,
"step": 440
},
{
"epoch": 0.9063444108761329,
"grad_norm": 0.6135808451328559,
"learning_rate": 1.6388959649926567e-05,
"loss": 0.7257630348205566,
"step": 450
},
{
"epoch": 0.9264853977844915,
"grad_norm": 0.5903369684800459,
"learning_rate": 1.62203276661861e-05,
"loss": 0.7346350669860839,
"step": 460
},
{
"epoch": 0.94662638469285,
"grad_norm": 0.6976504986685628,
"learning_rate": 1.604875966022035e-05,
"loss": 0.7640430450439453,
"step": 470
},
{
"epoch": 0.9667673716012085,
"grad_norm": 0.6035606483951823,
"learning_rate": 1.5874336612888487e-05,
"loss": 0.718746280670166,
"step": 480
},
{
"epoch": 0.9869083585095669,
"grad_norm": 0.6717571495166765,
"learning_rate": 1.5697140852641835e-05,
"loss": 0.7472597122192383,
"step": 490
},
{
"epoch": 1.0060422960725075,
"grad_norm": 0.7203444401990272,
"learning_rate": 1.5517256016664524e-05,
"loss": 0.7291147232055664,
"step": 500
},
{
"epoch": 1.0261832829808661,
"grad_norm": 0.6921694505639285,
"learning_rate": 1.533476701139633e-05,
"loss": 0.6633618354797364,
"step": 510
},
{
"epoch": 1.0463242698892246,
"grad_norm": 0.6349522071558343,
"learning_rate": 1.514975997245649e-05,
"loss": 0.6561120033264161,
"step": 520
},
{
"epoch": 1.0664652567975832,
"grad_norm": 0.6428241160936208,
"learning_rate": 1.4962322223987284e-05,
"loss": 0.651552963256836,
"step": 530
},
{
"epoch": 1.0866062437059416,
"grad_norm": 0.5614783554365259,
"learning_rate": 1.477254223743666e-05,
"loss": 0.6309733867645264,
"step": 540
},
{
"epoch": 1.1067472306143,
"grad_norm": 0.6136450539009461,
"learning_rate": 1.458050958979933e-05,
"loss": 0.6375226974487305,
"step": 550
},
{
"epoch": 1.1268882175226587,
"grad_norm": 0.5738680488992384,
"learning_rate": 1.438631492133601e-05,
"loss": 0.6501843929290771,
"step": 560
},
{
"epoch": 1.147029204431017,
"grad_norm": 0.6256698406539856,
"learning_rate": 1.4190049892790838e-05,
"loss": 0.6371057510375977,
"step": 570
},
{
"epoch": 1.1671701913393757,
"grad_norm": 0.5679189425501606,
"learning_rate": 1.3991807142127082e-05,
"loss": 0.6492814540863037,
"step": 580
},
{
"epoch": 1.1873111782477341,
"grad_norm": 0.6065846599142134,
"learning_rate": 1.3791680240801608e-05,
"loss": 0.6572648048400879,
"step": 590
},
{
"epoch": 1.2074521651560928,
"grad_norm": 0.6331239898292791,
"learning_rate": 1.358976364959876e-05,
"loss": 0.6534214496612549,
"step": 600
},
{
"epoch": 1.2275931520644512,
"grad_norm": 0.621707193585302,
"learning_rate": 1.3386152674044421e-05,
"loss": 0.6604917526245118,
"step": 610
},
{
"epoch": 1.2477341389728096,
"grad_norm": 0.6302732699277002,
"learning_rate": 1.3180943419421409e-05,
"loss": 0.6616671085357666,
"step": 620
},
{
"epoch": 1.2678751258811682,
"grad_norm": 0.6372781840535979,
"learning_rate": 1.2974232745407326e-05,
"loss": 0.6450593948364258,
"step": 630
},
{
"epoch": 1.2880161127895267,
"grad_norm": 0.5889507844368161,
"learning_rate": 1.276611822035641e-05,
"loss": 0.6392330169677735,
"step": 640
},
{
"epoch": 1.308157099697885,
"grad_norm": 0.6057569389599446,
"learning_rate": 1.2556698075246776e-05,
"loss": 0.6576297283172607,
"step": 650
},
{
"epoch": 1.3282980866062437,
"grad_norm": 0.9487405810829506,
"learning_rate": 1.2346071157315026e-05,
"loss": 0.6367029666900634,
"step": 660
},
{
"epoch": 1.3484390735146023,
"grad_norm": 0.5458544429535442,
"learning_rate": 1.2134336883399855e-05,
"loss": 0.6807767868041992,
"step": 670
},
{
"epoch": 1.3685800604229608,
"grad_norm": 0.5893035791406586,
"learning_rate": 1.1921595193016905e-05,
"loss": 0.6392487525939942,
"step": 680
},
{
"epoch": 1.3887210473313192,
"grad_norm": 0.6010604306346117,
"learning_rate": 1.1707946501186853e-05,
"loss": 0.6165999889373779,
"step": 690
},
{
"epoch": 1.4088620342396778,
"grad_norm": 0.6063068167039452,
"learning_rate": 1.1493491651039077e-05,
"loss": 0.6656947135925293,
"step": 700
},
{
"epoch": 1.4290030211480362,
"grad_norm": 0.6077129163598095,
"learning_rate": 1.1278331866213253e-05,
"loss": 0.6254322528839111,
"step": 710
},
{
"epoch": 1.4491440080563947,
"grad_norm": 0.5864287567502068,
"learning_rate": 1.1062568703081345e-05,
"loss": 0.6936765670776367,
"step": 720
},
{
"epoch": 1.4692849949647533,
"grad_norm": 0.6089793392201761,
"learning_rate": 1.0846304002812564e-05,
"loss": 0.6438858032226562,
"step": 730
},
{
"epoch": 1.4894259818731117,
"grad_norm": 0.6317374637004732,
"learning_rate": 1.0629639843303857e-05,
"loss": 0.6418643951416015,
"step": 740
},
{
"epoch": 1.5095669687814703,
"grad_norm": 0.6018478874762406,
"learning_rate": 1.0412678490998717e-05,
"loss": 0.6461330413818359,
"step": 750
},
{
"epoch": 1.5297079556898288,
"grad_norm": 0.5345160985133722,
"learning_rate": 1.0195522352616942e-05,
"loss": 0.6453229904174804,
"step": 760
},
{
"epoch": 1.5498489425981874,
"grad_norm": 0.6430235560122531,
"learning_rate": 9.978273926818233e-06,
"loss": 0.6503273010253906,
"step": 770
},
{
"epoch": 1.5699899295065458,
"grad_norm": 0.5938734205039293,
"learning_rate": 9.761035755822347e-06,
"loss": 0.6495280265808105,
"step": 780
},
{
"epoch": 1.5901309164149042,
"grad_norm": 0.5742568159918696,
"learning_rate": 9.543910377008741e-06,
"loss": 0.6502289772033691,
"step": 790
},
{
"epoch": 1.6102719033232629,
"grad_norm": 0.6183799064495971,
"learning_rate": 9.327000274518453e-06,
"loss": 0.657621192932129,
"step": 800
},
{
"epoch": 1.6304128902316215,
"grad_norm": 0.5746676315893429,
"learning_rate": 9.110407830881146e-06,
"loss": 0.6535334587097168,
"step": 810
},
{
"epoch": 1.6505538771399797,
"grad_norm": 0.6152747581447773,
"learning_rate": 8.894235278690104e-06,
"loss": 0.6586101531982422,
"step": 820
},
{
"epoch": 1.6706948640483383,
"grad_norm": 0.6175070299910487,
"learning_rate": 8.678584652347974e-06,
"loss": 0.6694258689880371,
"step": 830
},
{
"epoch": 1.690835850956697,
"grad_norm": 0.6490996006457843,
"learning_rate": 8.463557739906094e-06,
"loss": 0.6554967403411865,
"step": 840
},
{
"epoch": 1.7109768378650554,
"grad_norm": 0.6057747265761942,
"learning_rate": 8.249256035020086e-06,
"loss": 0.6524125099182129,
"step": 850
},
{
"epoch": 1.7311178247734138,
"grad_norm": 0.5781069954757457,
"learning_rate": 8.035780689044381e-06,
"loss": 0.6640297889709472,
"step": 860
},
{
"epoch": 1.7512588116817724,
"grad_norm": 0.5907306482202962,
"learning_rate": 7.823232463288344e-06,
"loss": 0.6420666694641113,
"step": 870
},
{
"epoch": 1.771399798590131,
"grad_norm": 0.6381794417859438,
"learning_rate": 7.611711681456493e-06,
"loss": 0.6338119506835938,
"step": 880
},
{
"epoch": 1.7915407854984893,
"grad_norm": 0.5790935289783766,
"learning_rate": 7.4013181822952484e-06,
"loss": 0.6390099048614502,
"step": 890
},
{
"epoch": 1.811681772406848,
"grad_norm": 0.5964904558967495,
"learning_rate": 7.19215127246863e-06,
"loss": 0.6316941261291504,
"step": 900
},
{
"epoch": 1.8318227593152066,
"grad_norm": 0.5909905502502826,
"learning_rate": 6.984309679685049e-06,
"loss": 0.6456757068634034,
"step": 910
},
{
"epoch": 1.851963746223565,
"grad_norm": 0.633551186863175,
"learning_rate": 6.777891506097394e-06,
"loss": 0.6603025436401367,
"step": 920
},
{
"epoch": 1.8721047331319234,
"grad_norm": 0.6405284882539302,
"learning_rate": 6.572994181998385e-06,
"loss": 0.6764796257019043,
"step": 930
},
{
"epoch": 1.892245720040282,
"grad_norm": 0.6414528571522098,
"learning_rate": 6.369714419833056e-06,
"loss": 0.632731819152832,
"step": 940
},
{
"epoch": 1.9123867069486404,
"grad_norm": 0.6234333012608522,
"learning_rate": 6.168148168550029e-06,
"loss": 0.6480500221252441,
"step": 950
},
{
"epoch": 1.9325276938569989,
"grad_norm": 0.6320726033761638,
"learning_rate": 5.968390568313194e-06,
"loss": 0.6275962352752685,
"step": 960
},
{
"epoch": 1.9526686807653575,
"grad_norm": 0.5637434358543838,
"learning_rate": 5.770535905595138e-06,
"loss": 0.6480206489562989,
"step": 970
},
{
"epoch": 1.9728096676737161,
"grad_norm": 0.5952218778997469,
"learning_rate": 5.574677568673499e-06,
"loss": 0.6083292007446289,
"step": 980
},
{
"epoch": 1.9929506545820745,
"grad_norm": 0.5734556241430626,
"learning_rate": 5.3809080035513e-06,
"loss": 0.6685420989990234,
"step": 990
},
{
"epoch": 2.012084592145015,
"grad_norm": 0.6523020868209938,
"learning_rate": 5.189318670322016e-06,
"loss": 0.6245943069458008,
"step": 1000
},
{
"epoch": 2.0322255790533736,
"grad_norm": 0.6468582760996592,
"learning_rate": 5.000000000000003e-06,
"loss": 0.5742752552032471,
"step": 1010
},
{
"epoch": 2.0523665659617323,
"grad_norm": 0.6165364026803251,
"learning_rate": 4.813041351836657e-06,
"loss": 0.6170816421508789,
"step": 1020
},
{
"epoch": 2.0725075528700905,
"grad_norm": 0.649862444578371,
"learning_rate": 4.6285309711424706e-06,
"loss": 0.5987297058105469,
"step": 1030
},
{
"epoch": 2.092648539778449,
"grad_norm": 0.6632354933054281,
"learning_rate": 4.446555947634825e-06,
"loss": 0.6134644508361816,
"step": 1040
},
{
"epoch": 2.1127895266868078,
"grad_norm": 0.6292999693586733,
"learning_rate": 4.267202174331288e-06,
"loss": 0.5931161403656006,
"step": 1050
},
{
"epoch": 2.1329305135951664,
"grad_norm": 0.6886779328643787,
"learning_rate": 4.0905543070077036e-06,
"loss": 0.5778852939605713,
"step": 1060
},
{
"epoch": 2.1530715005035246,
"grad_norm": 0.6379340378669663,
"learning_rate": 3.916695724240302e-06,
"loss": 0.5790260791778564,
"step": 1070
},
{
"epoch": 2.173212487411883,
"grad_norm": 0.587608625604733,
"learning_rate": 3.7457084880506465e-06,
"loss": 0.5704402446746826,
"step": 1080
},
{
"epoch": 2.193353474320242,
"grad_norm": 0.578613328498253,
"learning_rate": 3.5776733051719935e-06,
"loss": 0.5825461864471435,
"step": 1090
},
{
"epoch": 2.2134944612286,
"grad_norm": 0.5893218858500716,
"learning_rate": 3.412669488955346e-06,
"loss": 0.5475648880004883,
"step": 1100
},
{
"epoch": 2.2336354481369587,
"grad_norm": 0.5728337624083586,
"learning_rate": 3.2507749219332065e-06,
"loss": 0.5859562873840332,
"step": 1110
},
{
"epoch": 2.2537764350453173,
"grad_norm": 0.5865892490266207,
"learning_rate": 3.0920660190586893e-06,
"loss": 0.5725872039794921,
"step": 1120
},
{
"epoch": 2.2739174219536755,
"grad_norm": 0.6813688024576977,
"learning_rate": 2.9366176916373024e-06,
"loss": 0.5810467720031738,
"step": 1130
},
{
"epoch": 2.294058408862034,
"grad_norm": 0.5564488853837758,
"learning_rate": 2.7845033119684996e-06,
"loss": 0.5589536666870117,
"step": 1140
},
{
"epoch": 2.314199395770393,
"grad_norm": 0.6221441575834261,
"learning_rate": 2.635794678713611e-06,
"loss": 0.5693657875061036,
"step": 1150
},
{
"epoch": 2.3343403826787514,
"grad_norm": 0.6142938109628727,
"learning_rate": 2.4905619830065685e-06,
"loss": 0.5591839790344239,
"step": 1160
},
{
"epoch": 2.3544813695871096,
"grad_norm": 0.6040241814867302,
"learning_rate": 2.3488737753233827e-06,
"loss": 0.5525025367736817,
"step": 1170
},
{
"epoch": 2.3746223564954683,
"grad_norm": 0.5485294568778827,
"learning_rate": 2.210796933126005e-06,
"loss": 0.5566308498382568,
"step": 1180
},
{
"epoch": 2.394763343403827,
"grad_norm": 0.5991094942819453,
"learning_rate": 2.0763966292958704e-06,
"loss": 0.5564406394958497,
"step": 1190
},
{
"epoch": 2.4149043303121855,
"grad_norm": 0.5790075897305983,
"learning_rate": 1.94573630137201e-06,
"loss": 0.5888455390930176,
"step": 1200
},
{
"epoch": 2.4350453172205437,
"grad_norm": 0.5831452304720957,
"learning_rate": 1.8188776216082604e-06,
"loss": 0.5436674594879151,
"step": 1210
},
{
"epoch": 2.4551863041289024,
"grad_norm": 0.5538350696710048,
"learning_rate": 1.6958804678636743e-06,
"loss": 0.5791024208068848,
"step": 1220
},
{
"epoch": 2.475327291037261,
"grad_norm": 0.6004824349649384,
"learning_rate": 1.5768028953399083e-06,
"loss": 0.5947460174560547,
"step": 1230
},
{
"epoch": 2.495468277945619,
"grad_norm": 0.6013072332149644,
"learning_rate": 1.4617011091789135e-06,
"loss": 0.6078821659088135,
"step": 1240
},
{
"epoch": 2.515609264853978,
"grad_norm": 0.5770517074629651,
"learning_rate": 1.3506294379338557e-06,
"loss": 0.576723861694336,
"step": 1250
},
{
"epoch": 2.5357502517623365,
"grad_norm": 0.6118862530109742,
"learning_rate": 1.2436403079258064e-06,
"loss": 0.5891146659851074,
"step": 1260
},
{
"epoch": 2.5558912386706947,
"grad_norm": 0.6128217947924604,
"learning_rate": 1.1407842184982786e-06,
"loss": 0.5659162044525147,
"step": 1270
},
{
"epoch": 2.5760322255790533,
"grad_norm": 0.6192024814414648,
"learning_rate": 1.04210971818131e-06,
"loss": 0.5638632774353027,
"step": 1280
},
{
"epoch": 2.596173212487412,
"grad_norm": 0.5898775110317098,
"learning_rate": 9.476633817763481e-07,
"loss": 0.556513786315918,
"step": 1290
},
{
"epoch": 2.61631419939577,
"grad_norm": 0.6039004982453571,
"learning_rate": 8.574897883727384e-07,
"loss": 0.5709822654724122,
"step": 1300
},
{
"epoch": 2.636455186304129,
"grad_norm": 0.6240148606283491,
"learning_rate": 7.716315003061948e-07,
"loss": 0.602019214630127,
"step": 1310
},
{
"epoch": 2.6565961732124874,
"grad_norm": 0.6274955798619531,
"learning_rate": 6.901290430691842e-07,
"loss": 0.5823301315307617,
"step": 1320
},
{
"epoch": 2.676737160120846,
"grad_norm": 0.6252907977807435,
"learning_rate": 6.130208861827203e-07,
"loss": 0.571865463256836,
"step": 1330
},
{
"epoch": 2.6968781470292047,
"grad_norm": 0.6144848901151767,
"learning_rate": 5.40343425038573e-07,
"loss": 0.5906769752502441,
"step": 1340
},
{
"epoch": 2.717019133937563,
"grad_norm": 0.7200971287683633,
"learning_rate": 4.7213096372049404e-07,
"loss": 0.5911761283874511,
"step": 1350
},
{
"epoch": 2.7371601208459215,
"grad_norm": 0.6104854405112541,
"learning_rate": 4.084156988125232e-07,
"loss": 0.5763310432434082,
"step": 1360
},
{
"epoch": 2.75730110775428,
"grad_norm": 0.5464330677309381,
"learning_rate": 3.4922770420206754e-07,
"loss": 0.5731249332427979,
"step": 1370
},
{
"epoch": 2.7774420946626384,
"grad_norm": 0.6216489629404571,
"learning_rate": 2.9459491688488604e-07,
"loss": 0.5795963764190674,
"step": 1380
},
{
"epoch": 2.797583081570997,
"grad_norm": 0.62677128317226,
"learning_rate": 2.4454312377871105e-07,
"loss": 0.580109691619873,
"step": 1390
},
{
"epoch": 2.8177240684793556,
"grad_norm": 0.5976820703242487,
"learning_rate": 1.9909594955170752e-07,
"loss": 0.6225358009338379,
"step": 1400
},
{
"epoch": 2.837865055387714,
"grad_norm": 0.633710024857856,
"learning_rate": 1.5827484547151772e-07,
"loss": 0.5751284599304199,
"step": 1410
},
{
"epoch": 2.8580060422960725,
"grad_norm": 0.606609361073119,
"learning_rate": 1.2209907928017794e-07,
"loss": 0.6069238662719727,
"step": 1420
},
{
"epoch": 2.878147029204431,
"grad_norm": 0.6047426574267594,
"learning_rate": 9.058572609964788e-08,
"loss": 0.5654289245605468,
"step": 1430
},
{
"epoch": 2.8982880161127893,
"grad_norm": 0.6131232265425541,
"learning_rate": 6.374966037229202e-08,
"loss": 0.5597522735595704,
"step": 1440
},
{
"epoch": 2.918429003021148,
"grad_norm": 0.5824519340175841,
"learning_rate": 4.1603548840062344e-08,
"loss": 0.5661287784576416,
"step": 1450
},
{
"epoch": 2.9385699899295066,
"grad_norm": 0.5758313584108831,
"learning_rate": 2.4157844565747546e-08,
"loss": 0.5586350440979004,
"step": 1460
},
{
"epoch": 2.958710976837865,
"grad_norm": 0.6676715022794658,
"learning_rate": 1.14207819990797e-08,
"loss": 0.5903618335723877,
"step": 1470
},
{
"epoch": 2.9788519637462234,
"grad_norm": 0.6109078286891169,
"learning_rate": 3.3983730900377654e-09,
"loss": 0.5927345275878906,
"step": 1480
},
{
"epoch": 2.998992950654582,
"grad_norm": 0.569947241319677,
"learning_rate": 9.44044511796971e-11,
"loss": 0.571845006942749,
"step": 1490
},
{
"epoch": 3.0,
"step": 1491,
"total_flos": 145115603533824.0,
"train_loss": 0.698847218818268,
"train_runtime": 3222.3497,
"train_samples_per_second": 29.579,
"train_steps_per_second": 0.463
}
],
"logging_steps": 10,
"max_steps": 1491,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 300,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 145115603533824.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}