Files
Bori-2-135M-Instruct/trainer_state.json

1295 lines
33 KiB
JSON
Raw Permalink Normal View History

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 25.0,
"eval_steps": 500,
"global_step": 1800,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.13919095258808178,
"grad_norm": 2.168393135070801,
"learning_rate": 9.000000000000001e-07,
"loss": 2.423531150817871,
"step": 10
},
{
"epoch": 0.27838190517616357,
"grad_norm": 2.0506582260131836,
"learning_rate": 1.9000000000000002e-06,
"loss": 2.4842525482177735,
"step": 20
},
{
"epoch": 0.4175728577642453,
"grad_norm": 1.8670966625213623,
"learning_rate": 2.9e-06,
"loss": 2.400492477416992,
"step": 30
},
{
"epoch": 0.5567638103523271,
"grad_norm": 2.2221310138702393,
"learning_rate": 3.900000000000001e-06,
"loss": 2.329726791381836,
"step": 40
},
{
"epoch": 0.6959547629404089,
"grad_norm": 1.5783393383026123,
"learning_rate": 4.9000000000000005e-06,
"loss": 2.3145318984985352,
"step": 50
},
{
"epoch": 0.8351457155284906,
"grad_norm": 1.7792712450027466,
"learning_rate": 5.9e-06,
"loss": 2.2815326690673827,
"step": 60
},
{
"epoch": 0.9743366681165724,
"grad_norm": 0.6695189476013184,
"learning_rate": 6.9e-06,
"loss": 2.2338001251220705,
"step": 70
},
{
"epoch": 1.1113527620704655,
"grad_norm": 0.6323113441467285,
"learning_rate": 7.9e-06,
"loss": 2.1635990142822266,
"step": 80
},
{
"epoch": 1.2505437146585472,
"grad_norm": 0.5284978747367859,
"learning_rate": 8.900000000000001e-06,
"loss": 2.146891975402832,
"step": 90
},
{
"epoch": 1.3897346672466289,
"grad_norm": 0.5480155944824219,
"learning_rate": 9.9e-06,
"loss": 2.133349609375,
"step": 100
},
{
"epoch": 1.5289256198347108,
"grad_norm": 0.5004981756210327,
"learning_rate": 1.0900000000000002e-05,
"loss": 2.0712444305419924,
"step": 110
},
{
"epoch": 1.6681165724227927,
"grad_norm": 0.5202161073684692,
"learning_rate": 1.1900000000000001e-05,
"loss": 2.062981605529785,
"step": 120
},
{
"epoch": 1.8073075250108743,
"grad_norm": 0.5805963277816772,
"learning_rate": 1.2900000000000002e-05,
"loss": 2.0578401565551756,
"step": 130
},
{
"epoch": 1.946498477598956,
"grad_norm": 0.5083374381065369,
"learning_rate": 1.39e-05,
"loss": 1.9947002410888672,
"step": 140
},
{
"epoch": 2.083514571552849,
"grad_norm": 0.5673812031745911,
"learning_rate": 1.4900000000000001e-05,
"loss": 1.9915319442749024,
"step": 150
},
{
"epoch": 2.222705524140931,
"grad_norm": 0.465129017829895,
"learning_rate": 1.5900000000000004e-05,
"loss": 2.0579460144042967,
"step": 160
},
{
"epoch": 2.3618964767290125,
"grad_norm": 0.5780671834945679,
"learning_rate": 1.69e-05,
"loss": 2.0076471328735352,
"step": 170
},
{
"epoch": 2.5010874293170944,
"grad_norm": 0.47854819893836975,
"learning_rate": 1.79e-05,
"loss": 1.9822509765625,
"step": 180
},
{
"epoch": 2.6402783819051763,
"grad_norm": 0.412657231092453,
"learning_rate": 1.8900000000000002e-05,
"loss": 1.9754974365234375,
"step": 190
},
{
"epoch": 2.7794693344932577,
"grad_norm": 0.4147932231426239,
"learning_rate": 1.9900000000000003e-05,
"loss": 1.9069629669189454,
"step": 200
},
{
"epoch": 2.9186602870813396,
"grad_norm": 0.42903852462768555,
"learning_rate": 1.9998766324816606e-05,
"loss": 1.9168815612792969,
"step": 210
},
{
"epoch": 3.0556763810352328,
"grad_norm": 0.4464038610458374,
"learning_rate": 1.9994502159417576e-05,
"loss": 1.9295040130615235,
"step": 220
},
{
"epoch": 3.1948673336233147,
"grad_norm": 0.4937732517719269,
"learning_rate": 1.9987193571841865e-05,
"loss": 1.9452518463134765,
"step": 230
},
{
"epoch": 3.334058286211396,
"grad_norm": 0.5101185441017151,
"learning_rate": 1.9976842788356054e-05,
"loss": 1.8960039138793945,
"step": 240
},
{
"epoch": 3.473249238799478,
"grad_norm": 0.44784998893737793,
"learning_rate": 1.9963452961909065e-05,
"loss": 1.9037799835205078,
"step": 250
},
{
"epoch": 3.61244019138756,
"grad_norm": 0.43229031562805176,
"learning_rate": 1.9947028171171742e-05,
"loss": 1.9300691604614257,
"step": 260
},
{
"epoch": 3.751631143975642,
"grad_norm": 0.4695909321308136,
"learning_rate": 1.9927573419294456e-05,
"loss": 1.8757774353027343,
"step": 270
},
{
"epoch": 3.8908220965637232,
"grad_norm": 0.32979893684387207,
"learning_rate": 1.990509463238309e-05,
"loss": 1.8736637115478516,
"step": 280
},
{
"epoch": 4.027838190517616,
"grad_norm": 0.44490286707878113,
"learning_rate": 1.9879598657693894e-05,
"loss": 1.9053909301757812,
"step": 290
},
{
"epoch": 4.167029143105698,
"grad_norm": 0.38493791222572327,
"learning_rate": 1.985109326154774e-05,
"loss": 1.925973892211914,
"step": 300
},
{
"epoch": 4.30622009569378,
"grad_norm": 0.346566766500473,
"learning_rate": 1.981958712696444e-05,
"loss": 1.834067153930664,
"step": 310
},
{
"epoch": 4.445411048281862,
"grad_norm": 0.41095346212387085,
"learning_rate": 1.9785089851017788e-05,
"loss": 1.831857681274414,
"step": 320
},
{
"epoch": 4.5846020008699435,
"grad_norm": 0.3395974040031433,
"learning_rate": 1.974761194191222e-05,
"loss": 1.8029518127441406,
"step": 330
},
{
"epoch": 4.723792953458025,
"grad_norm": 0.3473803699016571,
"learning_rate": 1.970716481578191e-05,
"loss": 1.9018579483032227,
"step": 340
},
{
"epoch": 4.862983906046107,
"grad_norm": 0.361186683177948,
"learning_rate": 1.9663760793213297e-05,
"loss": 1.814296340942383,
"step": 350
},
{
"epoch": 5.0,
"grad_norm": 0.4093976318836212,
"learning_rate": 1.9617413095492114e-05,
"loss": 1.8802352905273438,
"step": 360
},
{
"epoch": 5.1391909525880815,
"grad_norm": 0.32077041268348694,
"learning_rate": 1.956813584057608e-05,
"loss": 1.8466520309448242,
"step": 370
},
{
"epoch": 5.278381905176164,
"grad_norm": 0.35901033878326416,
"learning_rate": 1.9515944038794384e-05,
"loss": 1.802210807800293,
"step": 380
},
{
"epoch": 5.417572857764245,
"grad_norm": 0.4042360186576843,
"learning_rate": 1.9460853588275454e-05,
"loss": 1.7823505401611328,
"step": 390
},
{
"epoch": 5.556763810352328,
"grad_norm": 0.3464180827140808,
"learning_rate": 1.940288127010419e-05,
"loss": 1.7850902557373047,
"step": 400
},
{
"epoch": 5.695954762940409,
"grad_norm": 0.41673702001571655,
"learning_rate": 1.9342044743210295e-05,
"loss": 1.8721931457519532,
"step": 410
},
{
"epoch": 5.8351457155284905,
"grad_norm": 0.4328930675983429,
"learning_rate": 1.92783625389892e-05,
"loss": 1.8219058990478516,
"step": 420
},
{
"epoch": 5.974336668116573,
"grad_norm": 0.36878204345703125,
"learning_rate": 1.9211854055657216e-05,
"loss": 1.8349163055419921,
"step": 430
},
{
"epoch": 6.1113527620704655,
"grad_norm": 0.33796119689941406,
"learning_rate": 1.9142539552342638e-05,
"loss": 1.7687013626098633,
"step": 440
},
{
"epoch": 6.250543714658547,
"grad_norm": 0.31589892506599426,
"learning_rate": 1.907044014291465e-05,
"loss": 1.7794736862182616,
"step": 450
},
{
"epoch": 6.389734667246629,
"grad_norm": 0.3584829866886139,
"learning_rate": 1.8995577789551806e-05,
"loss": 1.7861099243164062,
"step": 460
},
{
"epoch": 6.528925619834711,
"grad_norm": 0.39660242199897766,
"learning_rate": 1.8917975296052143e-05,
"loss": 1.8276309967041016,
"step": 470
},
{
"epoch": 6.668116572422792,
"grad_norm": 0.3751063048839569,
"learning_rate": 1.8837656300886937e-05,
"loss": 1.7932762145996093,
"step": 480
},
{
"epoch": 6.8073075250108745,
"grad_norm": 0.33827248215675354,
"learning_rate": 1.875464527000018e-05,
"loss": 1.8122461318969727,
"step": 490
},
{
"epoch": 6.946498477598956,
"grad_norm": 0.34659793972969055,
"learning_rate": 1.866896748935603e-05,
"loss": 1.7623111724853515,
"step": 500
},
{
"epoch": 7.083514571552849,
"grad_norm": 0.3210960328578949,
"learning_rate": 1.858064905723645e-05,
"loss": 1.749332809448242,
"step": 510
},
{
"epoch": 7.222705524140931,
"grad_norm": 0.3873847723007202,
"learning_rate": 1.8489716876291417e-05,
"loss": 1.7265895843505858,
"step": 520
},
{
"epoch": 7.3618964767290125,
"grad_norm": 0.370370477437973,
"learning_rate": 1.8396198645344133e-05,
"loss": 1.7626024246215821,
"step": 530
},
{
"epoch": 7.501087429317094,
"grad_norm": 0.3540331721305847,
"learning_rate": 1.8300122850953678e-05,
"loss": 1.761727523803711,
"step": 540
},
{
"epoch": 7.640278381905176,
"grad_norm": 0.38898929953575134,
"learning_rate": 1.8201518758737726e-05,
"loss": 1.7989364624023438,
"step": 550
},
{
"epoch": 7.779469334493258,
"grad_norm": 0.4159281253814697,
"learning_rate": 1.8100416404457962e-05,
"loss": 1.73597412109375,
"step": 560
},
{
"epoch": 7.91866028708134,
"grad_norm": 0.372927188873291,
"learning_rate": 1.799684658487091e-05,
"loss": 1.7885486602783203,
"step": 570
},
{
"epoch": 8.055676381035232,
"grad_norm": 0.3615380823612213,
"learning_rate": 1.789084084834691e-05,
"loss": 1.7422874450683594,
"step": 580
},
{
"epoch": 8.194867333623314,
"grad_norm": 0.4041246175765991,
"learning_rate": 1.778243148526021e-05,
"loss": 1.738204002380371,
"step": 590
},
{
"epoch": 8.334058286211397,
"grad_norm": 0.38440802693367004,
"learning_rate": 1.7671651518153e-05,
"loss": 1.722575569152832,
"step": 600
},
{
"epoch": 8.473249238799479,
"grad_norm": 0.38214221596717834,
"learning_rate": 1.7558534691676396e-05,
"loss": 1.7379753112792968,
"step": 610
},
{
"epoch": 8.61244019138756,
"grad_norm": 0.33185893297195435,
"learning_rate": 1.744311546231154e-05,
"loss": 1.720932388305664,
"step": 620
},
{
"epoch": 8.751631143975642,
"grad_norm": 0.3739125728607178,
"learning_rate": 1.732542898787379e-05,
"loss": 1.7539859771728517,
"step": 630
},
{
"epoch": 8.890822096563724,
"grad_norm": 0.3843406140804291,
"learning_rate": 1.7205511116803306e-05,
"loss": 1.7045206069946288,
"step": 640
},
{
"epoch": 9.027838190517617,
"grad_norm": 0.3675142228603363,
"learning_rate": 1.708339837724529e-05,
"loss": 1.7155038833618164,
"step": 650
},
{
"epoch": 9.167029143105697,
"grad_norm": 0.3326404094696045,
"learning_rate": 1.6959127965923144e-05,
"loss": 1.7345653533935548,
"step": 660
},
{
"epoch": 9.30622009569378,
"grad_norm": 0.4057214856147766,
"learning_rate": 1.6832737736807994e-05,
"loss": 1.7187301635742187,
"step": 670
},
{
"epoch": 9.445411048281862,
"grad_norm": 0.4393128454685211,
"learning_rate": 1.6704266189587992e-05,
"loss": 1.733585739135742,
"step": 680
},
{
"epoch": 9.584602000869943,
"grad_norm": 0.35559985041618347,
"learning_rate": 1.657375245794096e-05,
"loss": 1.709829330444336,
"step": 690
},
{
"epoch": 9.723792953458025,
"grad_norm": 0.3496800363063812,
"learning_rate": 1.644123629761387e-05,
"loss": 1.6870285034179688,
"step": 700
},
{
"epoch": 9.862983906046107,
"grad_norm": 0.3656679093837738,
"learning_rate": 1.6306758074312866e-05,
"loss": 1.7305381774902344,
"step": 710
},
{
"epoch": 10.0,
"grad_norm": 0.3679341673851013,
"learning_rate": 1.617035875140749e-05,
"loss": 1.6515853881835938,
"step": 720
},
{
"epoch": 10.139190952588082,
"grad_norm": 0.36722251772880554,
"learning_rate": 1.6032079877452825e-05,
"loss": 1.6483562469482422,
"step": 730
},
{
"epoch": 10.278381905176163,
"grad_norm": 0.3640598952770233,
"learning_rate": 1.5891963573533424e-05,
"loss": 1.686029052734375,
"step": 740
},
{
"epoch": 10.417572857764245,
"grad_norm": 0.3718581795692444,
"learning_rate": 1.575005252043279e-05,
"loss": 1.7063976287841798,
"step": 750
},
{
"epoch": 10.556763810352328,
"grad_norm": 0.33007678389549255,
"learning_rate": 1.560638994563242e-05,
"loss": 1.6735488891601562,
"step": 760
},
{
"epoch": 10.695954762940408,
"grad_norm": 0.3436570465564728,
"learning_rate": 1.5461019610144292e-05,
"loss": 1.6896970748901368,
"step": 770
},
{
"epoch": 10.83514571552849,
"grad_norm": 0.3507939577102661,
"learning_rate": 1.531398579518083e-05,
"loss": 1.6960151672363282,
"step": 780
},
{
"epoch": 10.974336668116573,
"grad_norm": 0.407681941986084,
"learning_rate": 1.516533328866642e-05,
"loss": 1.693899154663086,
"step": 790
},
{
"epoch": 11.111352762070466,
"grad_norm": 0.39356550574302673,
"learning_rate": 1.5015107371594576e-05,
"loss": 1.6978179931640625,
"step": 800
},
{
"epoch": 11.250543714658548,
"grad_norm": 0.3161419928073883,
"learning_rate": 1.4863353804234906e-05,
"loss": 1.6530536651611327,
"step": 810
},
{
"epoch": 11.389734667246628,
"grad_norm": 0.34999293088912964,
"learning_rate": 1.47101188121941e-05,
"loss": 1.663174819946289,
"step": 820
},
{
"epoch": 11.52892561983471,
"grad_norm": 0.37010619044303894,
"learning_rate": 1.4555449072335157e-05,
"loss": 1.641597366333008,
"step": 830
},
{
"epoch": 11.668116572422793,
"grad_norm": 0.35215842723846436,
"learning_rate": 1.4399391698559153e-05,
"loss": 1.618931198120117,
"step": 840
},
{
"epoch": 11.807307525010874,
"grad_norm": 0.3630053400993347,
"learning_rate": 1.4241994227453902e-05,
"loss": 1.7444683074951173,
"step": 850
},
{
"epoch": 11.946498477598956,
"grad_norm": 0.35291457176208496,
"learning_rate": 1.408330460381385e-05,
"loss": 1.6455070495605468,
"step": 860
},
{
"epoch": 12.083514571552849,
"grad_norm": 0.3664669394493103,
"learning_rate": 1.3923371166035615e-05,
"loss": 1.6963356018066407,
"step": 870
},
{
"epoch": 12.222705524140931,
"grad_norm": 0.35901519656181335,
"learning_rate": 1.3762242631393656e-05,
"loss": 1.6133247375488282,
"step": 880
},
{
"epoch": 12.361896476729013,
"grad_norm": 0.33618736267089844,
"learning_rate": 1.3599968081200515e-05,
"loss": 1.657137680053711,
"step": 890
},
{
"epoch": 12.501087429317094,
"grad_norm": 0.35423800349235535,
"learning_rate": 1.3436596945856164e-05,
"loss": 1.6295785903930664,
"step": 900
},
{
"epoch": 12.640278381905176,
"grad_norm": 0.37102779746055603,
"learning_rate": 1.327217898979104e-05,
"loss": 1.663412857055664,
"step": 910
},
{
"epoch": 12.779469334493259,
"grad_norm": 0.3388361632823944,
"learning_rate": 1.310676429630732e-05,
"loss": 1.6444122314453125,
"step": 920
},
{
"epoch": 12.91866028708134,
"grad_norm": 0.4413791000843048,
"learning_rate": 1.294040325232304e-05,
"loss": 1.6420936584472656,
"step": 930
},
{
"epoch": 13.055676381035232,
"grad_norm": 0.3519588112831116,
"learning_rate": 1.2773146533023782e-05,
"loss": 1.6441299438476562,
"step": 940
},
{
"epoch": 13.194867333623314,
"grad_norm": 0.37189796566963196,
"learning_rate": 1.2605045086426487e-05,
"loss": 1.6411401748657226,
"step": 950
},
{
"epoch": 13.334058286211397,
"grad_norm": 0.36288321018218994,
"learning_rate": 1.2436150117860226e-05,
"loss": 1.6302228927612306,
"step": 960
},
{
"epoch": 13.473249238799479,
"grad_norm": 0.3815094828605652,
"learning_rate": 1.2266513074368552e-05,
"loss": 1.6279544830322266,
"step": 970
},
{
"epoch": 13.61244019138756,
"grad_norm": 0.3317127525806427,
"learning_rate": 1.2096185629038219e-05,
"loss": 1.6380729675292969,
"step": 980
},
{
"epoch": 13.751631143975642,
"grad_norm": 0.33991727232933044,
"learning_rate": 1.1925219665259076e-05,
"loss": 1.6174076080322266,
"step": 990
},
{
"epoch": 13.890822096563724,
"grad_norm": 0.3842763602733612,
"learning_rate": 1.1753667260919872e-05,
"loss": 1.6469120025634765,
"step": 1000
},
{
"epoch": 14.027838190517617,
"grad_norm": 0.3835509121417999,
"learning_rate": 1.1581580672544839e-05,
"loss": 1.6305667877197265,
"step": 1010
},
{
"epoch": 14.167029143105697,
"grad_norm": 0.3624890446662903,
"learning_rate": 1.1409012319375828e-05,
"loss": 1.6266782760620118,
"step": 1020
},
{
"epoch": 14.30622009569378,
"grad_norm": 0.4243997633457184,
"learning_rate": 1.1236014767404929e-05,
"loss": 1.5902006149291992,
"step": 1030
},
{
"epoch": 14.445411048281862,
"grad_norm": 0.3468796908855438,
"learning_rate": 1.1062640713362333e-05,
"loss": 1.6216354370117188,
"step": 1040
},
{
"epoch": 14.584602000869943,
"grad_norm": 0.401908814907074,
"learning_rate": 1.0888942968664417e-05,
"loss": 1.6114536285400392,
"step": 1050
},
{
"epoch": 14.723792953458025,
"grad_norm": 0.4970390796661377,
"learning_rate": 1.071497444332686e-05,
"loss": 1.6442710876464843,
"step": 1060
},
{
"epoch": 14.862983906046107,
"grad_norm": 0.3688914477825165,
"learning_rate": 1.0540788129847757e-05,
"loss": 1.6298063278198243,
"step": 1070
},
{
"epoch": 15.0,
"grad_norm": 0.434993714094162,
"learning_rate": 1.0366437087065564e-05,
"loss": 1.595756435394287,
"step": 1080
},
{
"epoch": 15.139190952588082,
"grad_norm": 0.35529911518096924,
"learning_rate": 1.01919744239969e-05,
"loss": 1.6257198333740235,
"step": 1090
},
{
"epoch": 15.278381905176163,
"grad_norm": 0.34072160720825195,
"learning_rate": 1.0017453283658984e-05,
"loss": 1.5960454940795898,
"step": 1100
},
{
"epoch": 15.417572857764245,
"grad_norm": 0.3407002091407776,
"learning_rate": 9.842926826881796e-06,
"loss": 1.6124040603637695,
"step": 1110
},
{
"epoch": 15.556763810352328,
"grad_norm": 0.40393882989883423,
"learning_rate": 9.668448216114739e-06,
"loss": 1.5963356018066406,
"step": 1120
},
{
"epoch": 15.695954762940408,
"grad_norm": 0.3404715955257416,
"learning_rate": 9.494070599232868e-06,
"loss": 1.5878984451293945,
"step": 1130
},
{
"epoch": 15.83514571552849,
"grad_norm": 0.36653557419776917,
"learning_rate": 9.319847093347522e-06,
"loss": 1.6171627044677734,
"step": 1140
},
{
"epoch": 15.974336668116573,
"grad_norm": 0.3662278354167938,
"learning_rate": 9.145830768626326e-06,
"loss": 1.5852025985717773,
"step": 1150
},
{
"epoch": 16.111352762070464,
"grad_norm": 0.38278138637542725,
"learning_rate": 8.972074632127533e-06,
"loss": 1.6375057220458984,
"step": 1160
},
{
"epoch": 16.250543714658548,
"grad_norm": 0.37945541739463806,
"learning_rate": 8.79863161165353e-06,
"loss": 1.6030035018920898,
"step": 1170
},
{
"epoch": 16.38973466724663,
"grad_norm": 0.4779614806175232,
"learning_rate": 8.625554539628536e-06,
"loss": 1.5862947463989259,
"step": 1180
},
{
"epoch": 16.52892561983471,
"grad_norm": 0.4305327236652374,
"learning_rate": 8.452896137005322e-06,
"loss": 1.6087259292602538,
"step": 1190
},
{
"epoch": 16.668116572422793,
"grad_norm": 0.3351025879383087,
"learning_rate": 8.280708997205904e-06,
"loss": 1.540428924560547,
"step": 1200
},
{
"epoch": 16.807307525010874,
"grad_norm": 0.4297463595867157,
"learning_rate": 8.109045570101086e-06,
"loss": 1.5917155265808105,
"step": 1210
},
{
"epoch": 16.946498477598958,
"grad_norm": 0.34814122319221497,
"learning_rate": 7.937958146033706e-06,
"loss": 1.552242374420166,
"step": 1220
},
{
"epoch": 17.08351457155285,
"grad_norm": 0.3431946933269501,
"learning_rate": 7.767498839890489e-06,
"loss": 1.5931344985961915,
"step": 1230
},
{
"epoch": 17.22270552414093,
"grad_norm": 0.4247678816318512,
"learning_rate": 7.597719575227364e-06,
"loss": 1.5802928924560546,
"step": 1240
},
{
"epoch": 17.36189647672901,
"grad_norm": 0.34527385234832764,
"learning_rate": 7.428672068453041e-06,
"loss": 1.5672491073608399,
"step": 1250
},
{
"epoch": 17.501087429317096,
"grad_norm": 0.387896329164505,
"learning_rate": 7.260407813075676e-06,
"loss": 1.5785249710083007,
"step": 1260
},
{
"epoch": 17.640278381905176,
"grad_norm": 0.4005770981311798,
"learning_rate": 7.092978064017475e-06,
"loss": 1.607071876525879,
"step": 1270
},
{
"epoch": 17.779469334493257,
"grad_norm": 0.34315192699432373,
"learning_rate": 6.92643382200193e-06,
"loss": 1.6127098083496094,
"step": 1280
},
{
"epoch": 17.91866028708134,
"grad_norm": 0.38598817586898804,
"learning_rate": 6.7608258180185085e-06,
"loss": 1.5511341094970703,
"step": 1290
},
{
"epoch": 18.055676381035234,
"grad_norm": 0.37196919322013855,
"learning_rate": 6.596204497869501e-06,
"loss": 1.5802863121032715,
"step": 1300
},
{
"epoch": 18.194867333623314,
"grad_norm": 0.3532085120677948,
"learning_rate": 6.432620006803747e-06,
"loss": 1.6016632080078126,
"step": 1310
},
{
"epoch": 18.334058286211395,
"grad_norm": 0.3635902404785156,
"learning_rate": 6.2701221742419106e-06,
"loss": 1.587782096862793,
"step": 1320
},
{
"epoch": 18.47324923879948,
"grad_norm": 0.3914374113082886,
"learning_rate": 6.108760498597939e-06,
"loss": 1.564741325378418,
"step": 1330
},
{
"epoch": 18.61244019138756,
"grad_norm": 0.36254847049713135,
"learning_rate": 5.948584132201376e-06,
"loss": 1.5382339477539062,
"step": 1340
},
{
"epoch": 18.75163114397564,
"grad_norm": 0.40806812047958374,
"learning_rate": 5.789641866325091e-06,
"loss": 1.5827655792236328,
"step": 1350
},
{
"epoch": 18.890822096563724,
"grad_norm": 0.40871936082839966,
"learning_rate": 5.631982116322981e-06,
"loss": 1.5698270797729492,
"step": 1360
},
{
"epoch": 19.027838190517617,
"grad_norm": 0.34864017367362976,
"learning_rate": 5.475652906882173e-06,
"loss": 1.547338104248047,
"step": 1370
},
{
"epoch": 19.167029143105697,
"grad_norm": 0.344682902097702,
"learning_rate": 5.3207018573942684e-06,
"loss": 1.5822577476501465,
"step": 1380
},
{
"epoch": 19.30622009569378,
"grad_norm": 0.34946802258491516,
"learning_rate": 5.167176167449977e-06,
"loss": 1.5616905212402343,
"step": 1390
},
{
"epoch": 19.445411048281862,
"grad_norm": 0.3829444944858551,
"learning_rate": 5.015122602461698e-06,
"loss": 1.583076286315918,
"step": 1400
},
{
"epoch": 19.584602000869943,
"grad_norm": 0.3944872319698334,
"learning_rate": 4.864587479418302e-06,
"loss": 1.5577040672302247,
"step": 1410
},
{
"epoch": 19.723792953458027,
"grad_norm": 0.37802907824516296,
"learning_rate": 4.71561665277653e-06,
"loss": 1.559574508666992,
"step": 1420
},
{
"epoch": 19.862983906046107,
"grad_norm": 0.3113175928592682,
"learning_rate": 4.568255500493292e-06,
"loss": 1.5564599990844727,
"step": 1430
},
{
"epoch": 20.0,
"grad_norm": 0.3849709630012512,
"learning_rate": 4.422548910203099e-06,
"loss": 1.5471788406372071,
"step": 1440
},
{
"epoch": 20.13919095258808,
"grad_norm": 0.3226722478866577,
"learning_rate": 4.27854126554484e-06,
"loss": 1.5596961975097656,
"step": 1450
},
{
"epoch": 20.278381905176165,
"grad_norm": 0.33885228633880615,
"learning_rate": 4.136276432642107e-06,
"loss": 1.5046983718872071,
"step": 1460
},
{
"epoch": 20.417572857764245,
"grad_norm": 0.3818175196647644,
"learning_rate": 3.9957977467411615e-06,
"loss": 1.5638971328735352,
"step": 1470
},
{
"epoch": 20.556763810352326,
"grad_norm": 0.36039480566978455,
"learning_rate": 3.857147999010568e-06,
"loss": 1.5903932571411132,
"step": 1480
},
{
"epoch": 20.69595476294041,
"grad_norm": 0.3746209144592285,
"learning_rate": 3.7203694235066224e-06,
"loss": 1.5735241889953613,
"step": 1490
},
{
"epoch": 20.83514571552849,
"grad_norm": 0.38420161604881287,
"learning_rate": 3.5855036843084213e-06,
"loss": 1.5623096466064452,
"step": 1500
},
{
"epoch": 20.97433666811657,
"grad_norm": 0.4188708961009979,
"learning_rate": 3.452591862826603e-06,
"loss": 1.5697534561157227,
"step": 1510
},
{
"epoch": 21.111352762070464,
"grad_norm": 0.3920036554336548,
"learning_rate": 3.3216744452895356e-06,
"loss": 1.5524249076843262,
"step": 1520
},
{
"epoch": 21.250543714658548,
"grad_norm": 0.38039422035217285,
"learning_rate": 3.192791310410822e-06,
"loss": 1.5278741836547851,
"step": 1530
},
{
"epoch": 21.38973466724663,
"grad_norm": 0.3081962466239929,
"learning_rate": 3.0659817172418694e-06,
"loss": 1.5423470497131349,
"step": 1540
},
{
"epoch": 21.52892561983471,
"grad_norm": 0.3509298264980316,
"learning_rate": 2.9412842932131904e-06,
"loss": 1.5035577774047852,
"step": 1550
},
{
"epoch": 21.668116572422793,
"grad_norm": 0.33368051052093506,
"learning_rate": 2.8187370223681134e-06,
"loss": 1.555666732788086,
"step": 1560
},
{
"epoch": 21.807307525010874,
"grad_norm": 0.33778148889541626,
"learning_rate": 2.698377233792476e-06,
"loss": 1.5625399589538573,
"step": 1570
},
{
"epoch": 21.946498477598958,
"grad_norm": 0.42140209674835205,
"learning_rate": 2.5802415902438373e-06,
"loss": 1.5932843208312988,
"step": 1580
},
{
"epoch": 22.08351457155285,
"grad_norm": 0.3272395730018616,
"learning_rate": 2.464366076983623e-06,
"loss": 1.59045467376709,
"step": 1590
},
{
"epoch": 22.22270552414093,
"grad_norm": 0.33958563208580017,
"learning_rate": 2.3507859908156828e-06,
"loss": 1.5175203323364257,
"step": 1600
},
{
"epoch": 22.36189647672901,
"grad_norm": 0.40000298619270325,
"learning_rate": 2.2395359293345396e-06,
"loss": 1.5577522277832032,
"step": 1610
},
{
"epoch": 22.501087429317096,
"grad_norm": 0.39034193754196167,
"learning_rate": 2.130649780386628e-06,
"loss": 1.5607427597045898,
"step": 1620
},
{
"epoch": 22.640278381905176,
"grad_norm": 0.3313250243663788,
"learning_rate": 2.024160711747717e-06,
"loss": 1.5696653366088866,
"step": 1630
},
{
"epoch": 22.779469334493257,
"grad_norm": 0.38273027539253235,
"learning_rate": 1.9201011610196972e-06,
"loss": 1.5454700469970704,
"step": 1640
},
{
"epoch": 22.91866028708134,
"grad_norm": 0.36865147948265076,
"learning_rate": 1.818502825749764e-06,
"loss": 1.5079867362976074,
"step": 1650
},
{
"epoch": 23.055676381035234,
"grad_norm": 0.42405128479003906,
"learning_rate": 1.7193966537750561e-06,
"loss": 1.5665024757385253,
"step": 1660
},
{
"epoch": 23.194867333623314,
"grad_norm": 0.4113521873950958,
"learning_rate": 1.6228128337956128e-06,
"loss": 1.5511959075927735,
"step": 1670
},
{
"epoch": 23.334058286211395,
"grad_norm": 0.39184561371803284,
"learning_rate": 1.5287807861786308e-06,
"loss": 1.5294864654541016,
"step": 1680
},
{
"epoch": 23.47324923879948,
"grad_norm": 0.368387907743454,
"learning_rate": 1.4373291539967182e-06,
"loss": 1.5911618232727052,
"step": 1690
},
{
"epoch": 23.61244019138756,
"grad_norm": 0.37242400646209717,
"learning_rate": 1.3484857943029572e-06,
"loss": 1.5394433975219726,
"step": 1700
},
{
"epoch": 23.75163114397564,
"grad_norm": 0.42866021394729614,
"learning_rate": 1.2622777696453482e-06,
"loss": 1.5390392303466798,
"step": 1710
},
{
"epoch": 23.890822096563724,
"grad_norm": 0.38126400113105774,
"learning_rate": 1.1787313398233235e-06,
"loss": 1.5386754989624023,
"step": 1720
},
{
"epoch": 24.027838190517617,
"grad_norm": 0.358623743057251,
"learning_rate": 1.097871953888735e-06,
"loss": 1.5404573440551759,
"step": 1730
},
{
"epoch": 24.167029143105697,
"grad_norm": 0.35931700468063354,
"learning_rate": 1.0197242423938447e-06,
"loss": 1.5361785888671875,
"step": 1740
},
{
"epoch": 24.30622009569378,
"grad_norm": 0.36846256256103516,
"learning_rate": 9.44312009888606e-07,
"loss": 1.560469150543213,
"step": 1750
},
{
"epoch": 24.445411048281862,
"grad_norm": 0.40911176800727844,
"learning_rate": 8.716582276695729e-07,
"loss": 1.5527557373046874,
"step": 1760
},
{
"epoch": 24.584602000869943,
"grad_norm": 0.36536332964897156,
"learning_rate": 8.017850267826233e-07,
"loss": 1.5348384857177735,
"step": 1770
},
{
"epoch": 24.723792953458027,
"grad_norm": 0.36238783597946167,
"learning_rate": 7.347136912816277e-07,
"loss": 1.5461329460144042,
"step": 1780
},
{
"epoch": 24.862983906046107,
"grad_norm": 0.37549588084220886,
"learning_rate": 6.704646517451108e-07,
"loss": 1.5619462966918944,
"step": 1790
},
{
"epoch": 25.0,
"grad_norm": 0.41407886147499084,
"learning_rate": 6.090574790529091e-07,
"loss": 1.5449966430664062,
"step": 1800
}
],
"logging_steps": 10,
"max_steps": 2000,
"num_input_tokens_seen": 0,
"num_train_epochs": 28,
"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": 4.724013614392934e+16,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}