train-armv8-O2-epoch1and2-AMD / trainer_state.json
adpretko's picture
final checkpoint after epoch 2
d8109b1 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9998454166022568,
"eval_steps": 500,
"global_step": 3235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006183335909723296,
"grad_norm": 4.7098002433776855,
"learning_rate": 5.555555555555555e-07,
"loss": 0.6494,
"step": 10
},
{
"epoch": 0.012366671819446591,
"grad_norm": 3.629424810409546,
"learning_rate": 1.1728395061728396e-06,
"loss": 0.593,
"step": 20
},
{
"epoch": 0.018550007729169888,
"grad_norm": 2.0256550312042236,
"learning_rate": 1.7901234567901235e-06,
"loss": 0.3805,
"step": 30
},
{
"epoch": 0.024733343638893183,
"grad_norm": 0.7505523562431335,
"learning_rate": 2.4074074074074075e-06,
"loss": 0.1684,
"step": 40
},
{
"epoch": 0.030916679548616478,
"grad_norm": 0.4137927293777466,
"learning_rate": 3.0246913580246917e-06,
"loss": 0.0895,
"step": 50
},
{
"epoch": 0.037100015458339776,
"grad_norm": 0.30751892924308777,
"learning_rate": 3.641975308641976e-06,
"loss": 0.0652,
"step": 60
},
{
"epoch": 0.04328335136806307,
"grad_norm": 0.1395234614610672,
"learning_rate": 4.2592592592592596e-06,
"loss": 0.0544,
"step": 70
},
{
"epoch": 0.049466687277786366,
"grad_norm": 0.13168182969093323,
"learning_rate": 4.876543209876544e-06,
"loss": 0.0468,
"step": 80
},
{
"epoch": 0.05565002318750966,
"grad_norm": 0.09889546781778336,
"learning_rate": 5.493827160493828e-06,
"loss": 0.041,
"step": 90
},
{
"epoch": 0.061833359097232955,
"grad_norm": 0.09331826120615005,
"learning_rate": 6.111111111111112e-06,
"loss": 0.039,
"step": 100
},
{
"epoch": 0.06801669500695626,
"grad_norm": 0.10898005217313766,
"learning_rate": 6.728395061728395e-06,
"loss": 0.0348,
"step": 110
},
{
"epoch": 0.07420003091667955,
"grad_norm": 0.11297845095396042,
"learning_rate": 7.34567901234568e-06,
"loss": 0.0334,
"step": 120
},
{
"epoch": 0.08038336682640285,
"grad_norm": 0.1412217915058136,
"learning_rate": 7.962962962962963e-06,
"loss": 0.0308,
"step": 130
},
{
"epoch": 0.08656670273612614,
"grad_norm": 0.5460495948791504,
"learning_rate": 8.580246913580249e-06,
"loss": 0.0305,
"step": 140
},
{
"epoch": 0.09275003864584944,
"grad_norm": 2.9898719787597656,
"learning_rate": 9.197530864197531e-06,
"loss": 0.0329,
"step": 150
},
{
"epoch": 0.09893337455557273,
"grad_norm": 0.19271120429039001,
"learning_rate": 9.814814814814815e-06,
"loss": 0.0334,
"step": 160
},
{
"epoch": 0.10511671046529603,
"grad_norm": 0.13044007122516632,
"learning_rate": 1.04320987654321e-05,
"loss": 0.0287,
"step": 170
},
{
"epoch": 0.11130004637501932,
"grad_norm": 0.10140138119459152,
"learning_rate": 1.1049382716049384e-05,
"loss": 0.0279,
"step": 180
},
{
"epoch": 0.11748338228474262,
"grad_norm": 0.10765544325113297,
"learning_rate": 1.1666666666666668e-05,
"loss": 0.0253,
"step": 190
},
{
"epoch": 0.12366671819446591,
"grad_norm": 0.12072300165891647,
"learning_rate": 1.228395061728395e-05,
"loss": 0.0257,
"step": 200
},
{
"epoch": 0.1298500541041892,
"grad_norm": 0.13082030415534973,
"learning_rate": 1.2901234567901235e-05,
"loss": 0.0239,
"step": 210
},
{
"epoch": 0.13603339001391251,
"grad_norm": 0.11062952131032944,
"learning_rate": 1.351851851851852e-05,
"loss": 0.024,
"step": 220
},
{
"epoch": 0.1422167259236358,
"grad_norm": 0.12096355855464935,
"learning_rate": 1.4135802469135805e-05,
"loss": 0.0231,
"step": 230
},
{
"epoch": 0.1484000618333591,
"grad_norm": 0.13307975232601166,
"learning_rate": 1.4753086419753087e-05,
"loss": 0.0228,
"step": 240
},
{
"epoch": 0.15458339774308238,
"grad_norm": 0.09915771335363388,
"learning_rate": 1.537037037037037e-05,
"loss": 0.022,
"step": 250
},
{
"epoch": 0.1607667336528057,
"grad_norm": 0.15398280322551727,
"learning_rate": 1.5987654320987655e-05,
"loss": 0.0213,
"step": 260
},
{
"epoch": 0.16695006956252897,
"grad_norm": 0.11246932297945023,
"learning_rate": 1.660493827160494e-05,
"loss": 0.0214,
"step": 270
},
{
"epoch": 0.17313340547225228,
"grad_norm": 0.1021815687417984,
"learning_rate": 1.7222222222222224e-05,
"loss": 0.0213,
"step": 280
},
{
"epoch": 0.17931674138197556,
"grad_norm": 0.133810892701149,
"learning_rate": 1.7839506172839506e-05,
"loss": 0.0207,
"step": 290
},
{
"epoch": 0.18550007729169887,
"grad_norm": 0.09197945147752762,
"learning_rate": 1.8456790123456792e-05,
"loss": 0.0198,
"step": 300
},
{
"epoch": 0.19168341320142215,
"grad_norm": 0.13483920693397522,
"learning_rate": 1.9074074074074075e-05,
"loss": 0.0201,
"step": 310
},
{
"epoch": 0.19786674911114546,
"grad_norm": 0.12571094930171967,
"learning_rate": 1.969135802469136e-05,
"loss": 0.0194,
"step": 320
},
{
"epoch": 0.20405008502086877,
"grad_norm": 0.11730971932411194,
"learning_rate": 1.9999854512405252e-05,
"loss": 0.0191,
"step": 330
},
{
"epoch": 0.21023342093059205,
"grad_norm": 0.10710252076387405,
"learning_rate": 1.9998690637047088e-05,
"loss": 0.0185,
"step": 340
},
{
"epoch": 0.21641675684031536,
"grad_norm": 0.13765929639339447,
"learning_rate": 1.999636302179332e-05,
"loss": 0.0187,
"step": 350
},
{
"epoch": 0.22260009275003864,
"grad_norm": 0.10426069051027298,
"learning_rate": 1.9992871937553292e-05,
"loss": 0.0181,
"step": 360
},
{
"epoch": 0.22878342865976195,
"grad_norm": 0.10840260982513428,
"learning_rate": 1.9988217790651605e-05,
"loss": 0.0174,
"step": 370
},
{
"epoch": 0.23496676456948523,
"grad_norm": 0.1033315360546112,
"learning_rate": 1.9982401122780834e-05,
"loss": 0.0183,
"step": 380
},
{
"epoch": 0.24115010047920854,
"grad_norm": 0.08180748671293259,
"learning_rate": 1.9975422610938463e-05,
"loss": 0.0173,
"step": 390
},
{
"epoch": 0.24733343638893182,
"grad_norm": 0.08223113417625427,
"learning_rate": 1.9967283067348106e-05,
"loss": 0.0173,
"step": 400
},
{
"epoch": 0.2535167722986551,
"grad_norm": 0.08103210479021072,
"learning_rate": 1.9957983439364973e-05,
"loss": 0.0165,
"step": 410
},
{
"epoch": 0.2597001082083784,
"grad_norm": 0.07913105189800262,
"learning_rate": 1.994752480936559e-05,
"loss": 0.0164,
"step": 420
},
{
"epoch": 0.2658834441181017,
"grad_norm": 0.11991153657436371,
"learning_rate": 1.9935908394621844e-05,
"loss": 0.0163,
"step": 430
},
{
"epoch": 0.27206678002782503,
"grad_norm": 0.0896911695599556,
"learning_rate": 1.992313554715929e-05,
"loss": 0.0166,
"step": 440
},
{
"epoch": 0.2782501159375483,
"grad_norm": 0.10211066901683807,
"learning_rate": 1.99092077535998e-05,
"loss": 0.0162,
"step": 450
},
{
"epoch": 0.2844334518472716,
"grad_norm": 0.09302863478660583,
"learning_rate": 1.989412663498853e-05,
"loss": 0.016,
"step": 460
},
{
"epoch": 0.2906167877569949,
"grad_norm": 0.09001875668764114,
"learning_rate": 1.9877893946605247e-05,
"loss": 0.0159,
"step": 470
},
{
"epoch": 0.2968001236667182,
"grad_norm": 0.0752326026558876,
"learning_rate": 1.9860511577760042e-05,
"loss": 0.0155,
"step": 480
},
{
"epoch": 0.3029834595764415,
"grad_norm": 0.0945795550942421,
"learning_rate": 1.9841981551573424e-05,
"loss": 0.0154,
"step": 490
},
{
"epoch": 0.30916679548616477,
"grad_norm": 0.08852316439151764,
"learning_rate": 1.9822306024740855e-05,
"loss": 0.0156,
"step": 500
},
{
"epoch": 0.3153501313958881,
"grad_norm": 0.09425287693738937,
"learning_rate": 1.980148728728174e-05,
"loss": 0.0154,
"step": 510
},
{
"epoch": 0.3215334673056114,
"grad_norm": 0.08341208100318909,
"learning_rate": 1.9779527762272877e-05,
"loss": 0.015,
"step": 520
},
{
"epoch": 0.3277168032153347,
"grad_norm": 0.06619998067617416,
"learning_rate": 1.975643000556646e-05,
"loss": 0.015,
"step": 530
},
{
"epoch": 0.33390013912505795,
"grad_norm": 0.08618766069412231,
"learning_rate": 1.973219670549259e-05,
"loss": 0.0148,
"step": 540
},
{
"epoch": 0.34008347503478126,
"grad_norm": 0.08013293147087097,
"learning_rate": 1.970683068254637e-05,
"loss": 0.0145,
"step": 550
},
{
"epoch": 0.34626681094450457,
"grad_norm": 0.0985092967748642,
"learning_rate": 1.9680334889059668e-05,
"loss": 0.0143,
"step": 560
},
{
"epoch": 0.3524501468542279,
"grad_norm": 0.057891715317964554,
"learning_rate": 1.9652712408857452e-05,
"loss": 0.0141,
"step": 570
},
{
"epoch": 0.35863348276395113,
"grad_norm": 0.09929350018501282,
"learning_rate": 1.9623966456898914e-05,
"loss": 0.0139,
"step": 580
},
{
"epoch": 0.36481681867367444,
"grad_norm": 0.09841852635145187,
"learning_rate": 1.959410037890323e-05,
"loss": 0.0146,
"step": 590
},
{
"epoch": 0.37100015458339775,
"grad_norm": 0.07560750842094421,
"learning_rate": 1.9563117650960208e-05,
"loss": 0.0146,
"step": 600
},
{
"epoch": 0.37718349049312105,
"grad_norm": 0.06976748257875443,
"learning_rate": 1.9531021879125654e-05,
"loss": 0.0141,
"step": 610
},
{
"epoch": 0.3833668264028443,
"grad_norm": 0.09913719445466995,
"learning_rate": 1.9497816799001726e-05,
"loss": 0.0139,
"step": 620
},
{
"epoch": 0.3895501623125676,
"grad_norm": 0.09037956595420837,
"learning_rate": 1.9463506275302094e-05,
"loss": 0.0135,
"step": 630
},
{
"epoch": 0.3957334982222909,
"grad_norm": 0.0735185369849205,
"learning_rate": 1.9428094301402164e-05,
"loss": 0.0136,
"step": 640
},
{
"epoch": 0.40191683413201423,
"grad_norm": 0.0814405232667923,
"learning_rate": 1.939158499887428e-05,
"loss": 0.014,
"step": 650
},
{
"epoch": 0.40810017004173754,
"grad_norm": 0.0703529343008995,
"learning_rate": 1.9353982617008024e-05,
"loss": 0.0131,
"step": 660
},
{
"epoch": 0.4142835059514608,
"grad_norm": 0.0712701678276062,
"learning_rate": 1.9315291532315626e-05,
"loss": 0.0137,
"step": 670
},
{
"epoch": 0.4204668418611841,
"grad_norm": 1.0291255712509155,
"learning_rate": 1.9275516248022613e-05,
"loss": 0.0388,
"step": 680
},
{
"epoch": 0.4266501777709074,
"grad_norm": 0.25509175658226013,
"learning_rate": 1.9234661393543668e-05,
"loss": 0.0261,
"step": 690
},
{
"epoch": 0.4328335136806307,
"grad_norm": 0.0900217741727829,
"learning_rate": 1.91927317239438e-05,
"loss": 0.0181,
"step": 700
},
{
"epoch": 0.439016849590354,
"grad_norm": 0.0761672630906105,
"learning_rate": 1.9149732119384942e-05,
"loss": 0.0163,
"step": 710
},
{
"epoch": 0.4452001855000773,
"grad_norm": 0.06312277913093567,
"learning_rate": 1.9105667584557914e-05,
"loss": 0.0145,
"step": 720
},
{
"epoch": 0.4513835214098006,
"grad_norm": 0.0726538598537445,
"learning_rate": 1.9060543248099958e-05,
"loss": 0.0144,
"step": 730
},
{
"epoch": 0.4575668573195239,
"grad_norm": 0.06511197984218597,
"learning_rate": 1.9014364361997813e-05,
"loss": 0.0143,
"step": 740
},
{
"epoch": 0.46375019322924715,
"grad_norm": 0.07237090170383453,
"learning_rate": 1.896713630097643e-05,
"loss": 0.0141,
"step": 750
},
{
"epoch": 0.46993352913897046,
"grad_norm": 0.0700288638472557,
"learning_rate": 1.8918864561873428e-05,
"loss": 0.0134,
"step": 760
},
{
"epoch": 0.47611686504869377,
"grad_norm": 0.06440385431051254,
"learning_rate": 1.8869554762999307e-05,
"loss": 0.0135,
"step": 770
},
{
"epoch": 0.4823002009584171,
"grad_norm": 0.07516197860240936,
"learning_rate": 1.881921264348355e-05,
"loss": 0.0135,
"step": 780
},
{
"epoch": 0.4884835368681404,
"grad_norm": 0.0857824832201004,
"learning_rate": 1.876784406260664e-05,
"loss": 0.0131,
"step": 790
},
{
"epoch": 0.49466687277786364,
"grad_norm": 0.056280359625816345,
"learning_rate": 1.871545499911811e-05,
"loss": 0.0131,
"step": 800
},
{
"epoch": 0.500850208687587,
"grad_norm": 0.0742514505982399,
"learning_rate": 1.866205155054067e-05,
"loss": 0.0131,
"step": 810
},
{
"epoch": 0.5070335445973102,
"grad_norm": 0.06485935300588608,
"learning_rate": 1.860763993246053e-05,
"loss": 0.0127,
"step": 820
},
{
"epoch": 0.5132168805070335,
"grad_norm": 0.06388383358716965,
"learning_rate": 1.8552226477803984e-05,
"loss": 0.0126,
"step": 830
},
{
"epoch": 0.5194002164167568,
"grad_norm": 0.08962696045637131,
"learning_rate": 1.8495817636100293e-05,
"loss": 0.0129,
"step": 840
},
{
"epoch": 0.5255835523264801,
"grad_norm": 0.05811084434390068,
"learning_rate": 1.8438419972731066e-05,
"loss": 0.0126,
"step": 850
},
{
"epoch": 0.5317668882362034,
"grad_norm": 0.055295877158641815,
"learning_rate": 1.8380040168166096e-05,
"loss": 0.0129,
"step": 860
},
{
"epoch": 0.5379502241459267,
"grad_norm": 0.07821639627218246,
"learning_rate": 1.8320685017185832e-05,
"loss": 0.0125,
"step": 870
},
{
"epoch": 0.5441335600556501,
"grad_norm": 0.052799612283706665,
"learning_rate": 1.8260361428090542e-05,
"loss": 0.0126,
"step": 880
},
{
"epoch": 0.5503168959653734,
"grad_norm": 0.053990449756383896,
"learning_rate": 1.8199076421896256e-05,
"loss": 0.0121,
"step": 890
},
{
"epoch": 0.5565002318750966,
"grad_norm": 0.05750075727701187,
"learning_rate": 1.813683713151761e-05,
"loss": 0.0124,
"step": 900
},
{
"epoch": 0.5626835677848199,
"grad_norm": 0.05828944966197014,
"learning_rate": 1.8073650800937627e-05,
"loss": 0.0123,
"step": 910
},
{
"epoch": 0.5688669036945432,
"grad_norm": 0.07432842254638672,
"learning_rate": 1.8009524784364615e-05,
"loss": 0.0122,
"step": 920
},
{
"epoch": 0.5750502396042665,
"grad_norm": 0.06380300968885422,
"learning_rate": 1.7944466545376215e-05,
"loss": 0.0126,
"step": 930
},
{
"epoch": 0.5812335755139898,
"grad_norm": 0.06373830139636993,
"learning_rate": 1.7878483656050713e-05,
"loss": 0.012,
"step": 940
},
{
"epoch": 0.5874169114237131,
"grad_norm": 0.07270880043506622,
"learning_rate": 1.7811583796085722e-05,
"loss": 0.0122,
"step": 950
},
{
"epoch": 0.5936002473334364,
"grad_norm": 0.059029556810855865,
"learning_rate": 1.7743774751904385e-05,
"loss": 0.0122,
"step": 960
},
{
"epoch": 0.5997835832431597,
"grad_norm": 0.05860303342342377,
"learning_rate": 1.7675064415749078e-05,
"loss": 0.0117,
"step": 970
},
{
"epoch": 0.605966919152883,
"grad_norm": 0.06067311763763428,
"learning_rate": 1.760546078476286e-05,
"loss": 0.0122,
"step": 980
},
{
"epoch": 0.6121502550626062,
"grad_norm": 0.06291648745536804,
"learning_rate": 1.7534971960058684e-05,
"loss": 0.012,
"step": 990
},
{
"epoch": 0.6183335909723295,
"grad_norm": 0.0688399225473404,
"learning_rate": 1.7463606145776525e-05,
"loss": 0.012,
"step": 1000
},
{
"epoch": 0.6245169268820528,
"grad_norm": 0.07030756771564484,
"learning_rate": 1.739137164812849e-05,
"loss": 0.0118,
"step": 1010
},
{
"epoch": 0.6307002627917762,
"grad_norm": 0.05231929570436478,
"learning_rate": 1.731827687443208e-05,
"loss": 0.012,
"step": 1020
},
{
"epoch": 0.6368835987014995,
"grad_norm": 0.058629412204027176,
"learning_rate": 1.7244330332131657e-05,
"loss": 0.0112,
"step": 1030
},
{
"epoch": 0.6430669346112228,
"grad_norm": 0.05891234055161476,
"learning_rate": 1.7169540627808276e-05,
"loss": 0.0115,
"step": 1040
},
{
"epoch": 0.6492502705209461,
"grad_norm": 0.05712658539414406,
"learning_rate": 1.7093916466177972e-05,
"loss": 0.0113,
"step": 1050
},
{
"epoch": 0.6554336064306694,
"grad_norm": 0.06043630093336105,
"learning_rate": 1.701746664907862e-05,
"loss": 0.0111,
"step": 1060
},
{
"epoch": 0.6616169423403926,
"grad_norm": 0.06498923152685165,
"learning_rate": 1.6940200074445503e-05,
"loss": 0.0114,
"step": 1070
},
{
"epoch": 0.6678002782501159,
"grad_norm": 0.07936729490756989,
"learning_rate": 1.6862125735275675e-05,
"loss": 0.0118,
"step": 1080
},
{
"epoch": 0.6739836141598392,
"grad_norm": 0.07405313104391098,
"learning_rate": 1.6783252718581296e-05,
"loss": 0.012,
"step": 1090
},
{
"epoch": 0.6801669500695625,
"grad_norm": 0.05247880518436432,
"learning_rate": 1.6703590204331967e-05,
"loss": 0.0114,
"step": 1100
},
{
"epoch": 0.6863502859792858,
"grad_norm": 0.05421077832579613,
"learning_rate": 1.6623147464386324e-05,
"loss": 0.0116,
"step": 1110
},
{
"epoch": 0.6925336218890091,
"grad_norm": 0.05490672215819359,
"learning_rate": 1.6541933861412848e-05,
"loss": 0.0112,
"step": 1120
},
{
"epoch": 0.6987169577987324,
"grad_norm": 0.06434568762779236,
"learning_rate": 1.645995884780019e-05,
"loss": 0.0116,
"step": 1130
},
{
"epoch": 0.7049002937084557,
"grad_norm": 0.08222652971744537,
"learning_rate": 1.637723196455699e-05,
"loss": 0.0112,
"step": 1140
},
{
"epoch": 0.7110836296181791,
"grad_norm": 0.07595206797122955,
"learning_rate": 1.629376284020143e-05,
"loss": 0.0115,
"step": 1150
},
{
"epoch": 0.7172669655279023,
"grad_norm": 0.07365500926971436,
"learning_rate": 1.620956118964053e-05,
"loss": 0.0112,
"step": 1160
},
{
"epoch": 0.7234503014376256,
"grad_norm": 0.05542070046067238,
"learning_rate": 1.6124636813039502e-05,
"loss": 0.011,
"step": 1170
},
{
"epoch": 0.7296336373473489,
"grad_norm": 0.07046163082122803,
"learning_rate": 1.6038999594681074e-05,
"loss": 0.011,
"step": 1180
},
{
"epoch": 0.7358169732570722,
"grad_norm": 0.09000487625598907,
"learning_rate": 1.5952659501815073e-05,
"loss": 0.0112,
"step": 1190
},
{
"epoch": 0.7420003091667955,
"grad_norm": 0.0723351240158081,
"learning_rate": 1.5865626583498355e-05,
"loss": 0.0112,
"step": 1200
},
{
"epoch": 0.7481836450765188,
"grad_norm": 0.05575825273990631,
"learning_rate": 1.5777910969425184e-05,
"loss": 0.0108,
"step": 1210
},
{
"epoch": 0.7543669809862421,
"grad_norm": 0.06550654768943787,
"learning_rate": 1.5689522868748276e-05,
"loss": 0.0114,
"step": 1220
},
{
"epoch": 0.7605503168959654,
"grad_norm": 0.06199701875448227,
"learning_rate": 1.5600472568890523e-05,
"loss": 0.0108,
"step": 1230
},
{
"epoch": 0.7667336528056886,
"grad_norm": 0.05217553675174713,
"learning_rate": 1.551077043434769e-05,
"loss": 0.0108,
"step": 1240
},
{
"epoch": 0.7729169887154119,
"grad_norm": 0.0468706376850605,
"learning_rate": 1.5420426905482056e-05,
"loss": 0.0108,
"step": 1250
},
{
"epoch": 0.7791003246251352,
"grad_norm": 0.050856586545705795,
"learning_rate": 1.5329452497307316e-05,
"loss": 0.0109,
"step": 1260
},
{
"epoch": 0.7852836605348585,
"grad_norm": 0.051181282848119736,
"learning_rate": 1.5237857798264701e-05,
"loss": 0.0109,
"step": 1270
},
{
"epoch": 0.7914669964445818,
"grad_norm": 0.060655757784843445,
"learning_rate": 1.5145653468990626e-05,
"loss": 0.0107,
"step": 1280
},
{
"epoch": 0.7976503323543052,
"grad_norm": 0.04917164891958237,
"learning_rate": 1.50528502410759e-05,
"loss": 0.0104,
"step": 1290
},
{
"epoch": 0.8038336682640285,
"grad_norm": 0.05182170867919922,
"learning_rate": 1.4959458915816681e-05,
"loss": 0.0107,
"step": 1300
},
{
"epoch": 0.8100170041737518,
"grad_norm": 0.08411809802055359,
"learning_rate": 1.4865490362957317e-05,
"loss": 0.0109,
"step": 1310
},
{
"epoch": 0.8162003400834751,
"grad_norm": 0.07982087135314941,
"learning_rate": 1.477095551942525e-05,
"loss": 0.011,
"step": 1320
},
{
"epoch": 0.8223836759931983,
"grad_norm": 0.05386314168572426,
"learning_rate": 1.4675865388058037e-05,
"loss": 0.0104,
"step": 1330
},
{
"epoch": 0.8285670119029216,
"grad_norm": 0.0652853399515152,
"learning_rate": 1.458023103632277e-05,
"loss": 0.0106,
"step": 1340
},
{
"epoch": 0.8347503478126449,
"grad_norm": 0.05597163736820221,
"learning_rate": 1.4484063595027924e-05,
"loss": 0.0106,
"step": 1350
},
{
"epoch": 0.8409336837223682,
"grad_norm": 0.04806815832853317,
"learning_rate": 1.438737425702786e-05,
"loss": 0.0106,
"step": 1360
},
{
"epoch": 0.8471170196320915,
"grad_norm": 0.06060324236750603,
"learning_rate": 1.4290174275920095e-05,
"loss": 0.0106,
"step": 1370
},
{
"epoch": 0.8533003555418148,
"grad_norm": 0.05702946335077286,
"learning_rate": 1.41924749647355e-05,
"loss": 0.0107,
"step": 1380
},
{
"epoch": 0.8594836914515381,
"grad_norm": 0.05112171545624733,
"learning_rate": 1.4094287694621589e-05,
"loss": 0.0104,
"step": 1390
},
{
"epoch": 0.8656670273612614,
"grad_norm": 0.05438029393553734,
"learning_rate": 1.3995623893519044e-05,
"loss": 0.0104,
"step": 1400
},
{
"epoch": 0.8718503632709848,
"grad_norm": 0.06269069015979767,
"learning_rate": 1.3896495044831622e-05,
"loss": 0.01,
"step": 1410
},
{
"epoch": 0.878033699180708,
"grad_norm": 0.06591395288705826,
"learning_rate": 1.37969126860896e-05,
"loss": 0.01,
"step": 1420
},
{
"epoch": 0.8842170350904313,
"grad_norm": 0.0547652542591095,
"learning_rate": 1.3696888407606952e-05,
"loss": 0.0103,
"step": 1430
},
{
"epoch": 0.8904003710001546,
"grad_norm": 0.05885763093829155,
"learning_rate": 1.3596433851132342e-05,
"loss": 0.0103,
"step": 1440
},
{
"epoch": 0.8965837069098779,
"grad_norm": 0.062123287469148636,
"learning_rate": 1.3495560708494167e-05,
"loss": 0.0103,
"step": 1450
},
{
"epoch": 0.9027670428196012,
"grad_norm": 0.053790923207998276,
"learning_rate": 1.3394280720239733e-05,
"loss": 0.0104,
"step": 1460
},
{
"epoch": 0.9089503787293245,
"grad_norm": 0.05707462877035141,
"learning_rate": 1.3292605674268803e-05,
"loss": 0.0103,
"step": 1470
},
{
"epoch": 0.9151337146390478,
"grad_norm": 0.05429013445973396,
"learning_rate": 1.31905474044616e-05,
"loss": 0.0104,
"step": 1480
},
{
"epoch": 0.9213170505487711,
"grad_norm": 0.04341137036681175,
"learning_rate": 1.3088117789301473e-05,
"loss": 0.0099,
"step": 1490
},
{
"epoch": 0.9275003864584943,
"grad_norm": 0.07078797370195389,
"learning_rate": 1.298532875049237e-05,
"loss": 0.0101,
"step": 1500
},
{
"epoch": 0.9336837223682176,
"grad_norm": 0.04922262579202652,
"learning_rate": 1.2882192251571282e-05,
"loss": 0.01,
"step": 1510
},
{
"epoch": 0.9398670582779409,
"grad_norm": 0.046682145446538925,
"learning_rate": 1.2778720296515816e-05,
"loss": 0.01,
"step": 1520
},
{
"epoch": 0.9460503941876642,
"grad_norm": 0.06537698954343796,
"learning_rate": 1.2674924928347055e-05,
"loss": 0.0098,
"step": 1530
},
{
"epoch": 0.9522337300973875,
"grad_norm": 0.051184337586164474,
"learning_rate": 1.2570818227727889e-05,
"loss": 0.0099,
"step": 1540
},
{
"epoch": 0.9584170660071109,
"grad_norm": 0.055025652050971985,
"learning_rate": 1.2466412311556952e-05,
"loss": 0.01,
"step": 1550
},
{
"epoch": 0.9646004019168342,
"grad_norm": 0.06580166518688202,
"learning_rate": 1.2361719331558346e-05,
"loss": 0.01,
"step": 1560
},
{
"epoch": 0.9707837378265575,
"grad_norm": 0.05151714012026787,
"learning_rate": 1.2256751472867304e-05,
"loss": 0.0099,
"step": 1570
},
{
"epoch": 0.9769670737362808,
"grad_norm": 0.05681469663977623,
"learning_rate": 1.2151520952611984e-05,
"loss": 0.0099,
"step": 1580
},
{
"epoch": 0.983150409646004,
"grad_norm": 0.07328800857067108,
"learning_rate": 1.2046040018491523e-05,
"loss": 0.0097,
"step": 1590
},
{
"epoch": 0.9893337455557273,
"grad_norm": 0.063533253967762,
"learning_rate": 1.1940320947350532e-05,
"loss": 0.0101,
"step": 1600
},
{
"epoch": 0.9955170814654506,
"grad_norm": 0.05460398644208908,
"learning_rate": 1.1834376043750201e-05,
"loss": 0.0094,
"step": 1610
},
{
"epoch": 1.001855000772917,
"grad_norm": 0.05063384771347046,
"learning_rate": 1.1728217638536196e-05,
"loss": 0.0114,
"step": 1620
},
{
"epoch": 1.0080383366826402,
"grad_norm": 0.05956198275089264,
"learning_rate": 1.1621858087403466e-05,
"loss": 0.0094,
"step": 1630
},
{
"epoch": 1.0142216725923636,
"grad_norm": 0.05441627651453018,
"learning_rate": 1.1515309769458174e-05,
"loss": 0.0094,
"step": 1640
},
{
"epoch": 1.0204050085020868,
"grad_norm": 0.045704033225774765,
"learning_rate": 1.1408585085776916e-05,
"loss": 0.0092,
"step": 1650
},
{
"epoch": 1.0265883444118102,
"grad_norm": 0.05705653131008148,
"learning_rate": 1.1301696457963352e-05,
"loss": 0.0097,
"step": 1660
},
{
"epoch": 1.0327716803215334,
"grad_norm": 0.06518584489822388,
"learning_rate": 1.1194656326702474e-05,
"loss": 0.0097,
"step": 1670
},
{
"epoch": 1.0389550162312569,
"grad_norm": 0.054483164101839066,
"learning_rate": 1.1087477150312653e-05,
"loss": 0.0094,
"step": 1680
},
{
"epoch": 1.04513835214098,
"grad_norm": 0.05142778158187866,
"learning_rate": 1.098017140329561e-05,
"loss": 0.0092,
"step": 1690
},
{
"epoch": 1.0513216880507033,
"grad_norm": 0.052420474588871,
"learning_rate": 1.0872751574884512e-05,
"loss": 0.0093,
"step": 1700
},
{
"epoch": 1.0575050239604267,
"grad_norm": 0.050652723759412766,
"learning_rate": 1.0765230167590385e-05,
"loss": 0.009,
"step": 1710
},
{
"epoch": 1.0636883598701499,
"grad_norm": 0.050776343792676926,
"learning_rate": 1.0657619695746944e-05,
"loss": 0.0093,
"step": 1720
},
{
"epoch": 1.0698716957798733,
"grad_norm": 0.052341315895318985,
"learning_rate": 1.054993268405405e-05,
"loss": 0.0093,
"step": 1730
},
{
"epoch": 1.0760550316895965,
"grad_norm": 0.04870909824967384,
"learning_rate": 1.0442181666119984e-05,
"loss": 0.0093,
"step": 1740
},
{
"epoch": 1.08223836759932,
"grad_norm": 0.04733763262629509,
"learning_rate": 1.033437918300266e-05,
"loss": 0.0093,
"step": 1750
},
{
"epoch": 1.088421703509043,
"grad_norm": 0.0712374672293663,
"learning_rate": 1.0226537781749988e-05,
"loss": 0.009,
"step": 1760
},
{
"epoch": 1.0946050394187665,
"grad_norm": 0.05846722424030304,
"learning_rate": 1.0118670013939526e-05,
"loss": 0.009,
"step": 1770
},
{
"epoch": 1.1007883753284897,
"grad_norm": 0.05150752514600754,
"learning_rate": 1.0010788434217625e-05,
"loss": 0.0091,
"step": 1780
},
{
"epoch": 1.106971711238213,
"grad_norm": 0.05533820763230324,
"learning_rate": 9.902905598838177e-06,
"loss": 0.0091,
"step": 1790
},
{
"epoch": 1.1131550471479363,
"grad_norm": 0.0494297593832016,
"learning_rate": 9.795034064201235e-06,
"loss": 0.009,
"step": 1800
},
{
"epoch": 1.1193383830576595,
"grad_norm": 0.05712695047259331,
"learning_rate": 9.68718638539156e-06,
"loss": 0.009,
"step": 1810
},
{
"epoch": 1.125521718967383,
"grad_norm": 0.058026671409606934,
"learning_rate": 9.579375114717352e-06,
"loss": 0.0092,
"step": 1820
},
{
"epoch": 1.1317050548771062,
"grad_norm": 0.06600188463926315,
"learning_rate": 9.471612800249295e-06,
"loss": 0.0095,
"step": 1830
},
{
"epoch": 1.1378883907868296,
"grad_norm": 0.05759551003575325,
"learning_rate": 9.363911984360104e-06,
"loss": 0.0088,
"step": 1840
},
{
"epoch": 1.1440717266965528,
"grad_norm": 0.048796847462654114,
"learning_rate": 9.256285202264716e-06,
"loss": 0.0093,
"step": 1850
},
{
"epoch": 1.1502550626062762,
"grad_norm": 0.04419567808508873,
"learning_rate": 9.148744980561332e-06,
"loss": 0.0089,
"step": 1860
},
{
"epoch": 1.1564383985159994,
"grad_norm": 0.044388189911842346,
"learning_rate": 9.041303835773458e-06,
"loss": 0.0093,
"step": 1870
},
{
"epoch": 1.1626217344257226,
"grad_norm": 0.05169492959976196,
"learning_rate": 8.933974272893113e-06,
"loss": 0.0091,
"step": 1880
},
{
"epoch": 1.168805070335446,
"grad_norm": 0.043622784316539764,
"learning_rate": 8.826768783925385e-06,
"loss": 0.0089,
"step": 1890
},
{
"epoch": 1.1749884062451692,
"grad_norm": 0.04748489707708359,
"learning_rate": 8.719699846434493e-06,
"loss": 0.0088,
"step": 1900
},
{
"epoch": 1.1811717421548926,
"grad_norm": 0.051559239625930786,
"learning_rate": 8.612779922091534e-06,
"loss": 0.0086,
"step": 1910
},
{
"epoch": 1.1873550780646158,
"grad_norm": 0.054223425686359406,
"learning_rate": 8.506021455224083e-06,
"loss": 0.0088,
"step": 1920
},
{
"epoch": 1.1935384139743392,
"grad_norm": 0.04965620115399361,
"learning_rate": 8.399436871367804e-06,
"loss": 0.009,
"step": 1930
},
{
"epoch": 1.1997217498840624,
"grad_norm": 0.05411219596862793,
"learning_rate": 8.293038575820249e-06,
"loss": 0.0089,
"step": 1940
},
{
"epoch": 1.2059050857937859,
"grad_norm": 0.04933865740895271,
"learning_rate": 8.186838952197019e-06,
"loss": 0.0088,
"step": 1950
},
{
"epoch": 1.212088421703509,
"grad_norm": 0.05196445807814598,
"learning_rate": 8.080850360990439e-06,
"loss": 0.0088,
"step": 1960
},
{
"epoch": 1.2182717576132323,
"grad_norm": 0.04491017013788223,
"learning_rate": 7.975085138130938e-06,
"loss": 0.0088,
"step": 1970
},
{
"epoch": 1.2244550935229557,
"grad_norm": 0.04868939146399498,
"learning_rate": 7.869555593551277e-06,
"loss": 0.009,
"step": 1980
},
{
"epoch": 1.2306384294326789,
"grad_norm": 0.05438001826405525,
"learning_rate": 7.764274009753796e-06,
"loss": 0.009,
"step": 1990
},
{
"epoch": 1.2368217653424023,
"grad_norm": 0.04552379623055458,
"learning_rate": 7.659252640380876e-06,
"loss": 0.0087,
"step": 2000
},
{
"epoch": 1.2430051012521255,
"grad_norm": 0.04229842126369476,
"learning_rate": 7.554503708788738e-06,
"loss": 0.0084,
"step": 2010
},
{
"epoch": 1.249188437161849,
"grad_norm": 0.04843869060277939,
"learning_rate": 7.450039406624792e-06,
"loss": 0.0089,
"step": 2020
},
{
"epoch": 1.255371773071572,
"grad_norm": 0.05012344941496849,
"learning_rate": 7.345871892408631e-06,
"loss": 0.0086,
"step": 2030
},
{
"epoch": 1.2615551089812955,
"grad_norm": 0.0570928156375885,
"learning_rate": 7.242013290116944e-06,
"loss": 0.0088,
"step": 2040
},
{
"epoch": 1.2677384448910187,
"grad_norm": 0.061261292546987534,
"learning_rate": 7.138475687772381e-06,
"loss": 0.0086,
"step": 2050
},
{
"epoch": 1.273921780800742,
"grad_norm": 0.046513836830854416,
"learning_rate": 7.03527113603667e-06,
"loss": 0.009,
"step": 2060
},
{
"epoch": 1.2801051167104653,
"grad_norm": 0.054141171276569366,
"learning_rate": 6.9324116468080305e-06,
"loss": 0.0087,
"step": 2070
},
{
"epoch": 1.2862884526201885,
"grad_norm": 0.04561785236001015,
"learning_rate": 6.829909191823121e-06,
"loss": 0.0085,
"step": 2080
},
{
"epoch": 1.292471788529912,
"grad_norm": 0.04863652586936951,
"learning_rate": 6.727775701263666e-06,
"loss": 0.0089,
"step": 2090
},
{
"epoch": 1.2986551244396352,
"grad_norm": 0.04651442915201187,
"learning_rate": 6.626023062367894e-06,
"loss": 0.0089,
"step": 2100
},
{
"epoch": 1.3048384603493584,
"grad_norm": 0.0455438569188118,
"learning_rate": 6.524663118047015e-06,
"loss": 0.009,
"step": 2110
},
{
"epoch": 1.3110217962590818,
"grad_norm": 0.05052075907588005,
"learning_rate": 6.423707665506812e-06,
"loss": 0.0086,
"step": 2120
},
{
"epoch": 1.3172051321688052,
"grad_norm": 0.07010175287723541,
"learning_rate": 6.323168454874582e-06,
"loss": 0.0084,
"step": 2130
},
{
"epoch": 1.3233884680785284,
"grad_norm": 0.04627922177314758,
"learning_rate": 6.223057187831548e-06,
"loss": 0.0087,
"step": 2140
},
{
"epoch": 1.3295718039882516,
"grad_norm": 0.04536746069788933,
"learning_rate": 6.123385516250911e-06,
"loss": 0.0085,
"step": 2150
},
{
"epoch": 1.335755139897975,
"grad_norm": 0.045566774904727936,
"learning_rate": 6.02416504084169e-06,
"loss": 0.0085,
"step": 2160
},
{
"epoch": 1.3419384758076982,
"grad_norm": 0.04734981432557106,
"learning_rate": 5.925407309798535e-06,
"loss": 0.0091,
"step": 2170
},
{
"epoch": 1.3481218117174216,
"grad_norm": 0.04319227859377861,
"learning_rate": 5.827123817457631e-06,
"loss": 0.0082,
"step": 2180
},
{
"epoch": 1.3543051476271448,
"grad_norm": 0.048728473484516144,
"learning_rate": 5.729326002958889e-06,
"loss": 0.0088,
"step": 2190
},
{
"epoch": 1.360488483536868,
"grad_norm": 0.059723589569330215,
"learning_rate": 5.632025248914546e-06,
"loss": 0.0083,
"step": 2200
},
{
"epoch": 1.3666718194465914,
"grad_norm": 0.059667762368917465,
"learning_rate": 5.535232880084373e-06,
"loss": 0.0088,
"step": 2210
},
{
"epoch": 1.3728551553563149,
"grad_norm": 0.046967074275016785,
"learning_rate": 5.438960162057548e-06,
"loss": 0.0084,
"step": 2220
},
{
"epoch": 1.379038491266038,
"grad_norm": 0.0653061717748642,
"learning_rate": 5.343218299941517e-06,
"loss": 0.0086,
"step": 2230
},
{
"epoch": 1.3852218271757613,
"grad_norm": 0.042991455644369125,
"learning_rate": 5.248018437057807e-06,
"loss": 0.0085,
"step": 2240
},
{
"epoch": 1.3914051630854847,
"grad_norm": 0.05747677758336067,
"learning_rate": 5.15337165364507e-06,
"loss": 0.0083,
"step": 2250
},
{
"epoch": 1.3975884989952079,
"grad_norm": 0.04957514628767967,
"learning_rate": 5.059288965569471e-06,
"loss": 0.0084,
"step": 2260
},
{
"epoch": 1.4037718349049313,
"grad_norm": 0.05094042047858238,
"learning_rate": 4.965781323042547e-06,
"loss": 0.0082,
"step": 2270
},
{
"epoch": 1.4099551708146545,
"grad_norm": 0.04698927700519562,
"learning_rate": 4.872859609346733e-06,
"loss": 0.0083,
"step": 2280
},
{
"epoch": 1.4161385067243777,
"grad_norm": 0.05430430918931961,
"learning_rate": 4.780534639568651e-06,
"loss": 0.0082,
"step": 2290
},
{
"epoch": 1.422321842634101,
"grad_norm": 0.05776724964380264,
"learning_rate": 4.68881715934037e-06,
"loss": 0.0085,
"step": 2300
},
{
"epoch": 1.4285051785438245,
"grad_norm": 0.047362685203552246,
"learning_rate": 4.597717843588708e-06,
"loss": 0.0084,
"step": 2310
},
{
"epoch": 1.4346885144535477,
"grad_norm": 0.047893136739730835,
"learning_rate": 4.5072472952928015e-06,
"loss": 0.0084,
"step": 2320
},
{
"epoch": 1.440871850363271,
"grad_norm": 0.04316046088933945,
"learning_rate": 4.417416044250026e-06,
"loss": 0.0086,
"step": 2330
},
{
"epoch": 1.4470551862729943,
"grad_norm": 0.05563107505440712,
"learning_rate": 4.328234545850441e-06,
"loss": 0.0081,
"step": 2340
},
{
"epoch": 1.4532385221827175,
"grad_norm": 0.05854015052318573,
"learning_rate": 4.239713179859901e-06,
"loss": 0.0082,
"step": 2350
},
{
"epoch": 1.459421858092441,
"grad_norm": 0.04468608647584915,
"learning_rate": 4.151862249211953e-06,
"loss": 0.0082,
"step": 2360
},
{
"epoch": 1.4656051940021642,
"grad_norm": 0.04180331900715828,
"learning_rate": 4.064691978808699e-06,
"loss": 0.0083,
"step": 2370
},
{
"epoch": 1.4717885299118874,
"grad_norm": 0.04552458971738815,
"learning_rate": 3.978212514330703e-06,
"loss": 0.0081,
"step": 2380
},
{
"epoch": 1.4779718658216108,
"grad_norm": 0.045519933104515076,
"learning_rate": 3.892433921056184e-06,
"loss": 0.0082,
"step": 2390
},
{
"epoch": 1.484155201731334,
"grad_norm": 0.04389030858874321,
"learning_rate": 3.8073661826894837e-06,
"loss": 0.0081,
"step": 2400
},
{
"epoch": 1.4903385376410574,
"grad_norm": 0.050178904086351395,
"learning_rate": 3.7230192001990963e-06,
"loss": 0.0085,
"step": 2410
},
{
"epoch": 1.4965218735507806,
"grad_norm": 0.046226803213357925,
"learning_rate": 3.6394027906652927e-06,
"loss": 0.0082,
"step": 2420
},
{
"epoch": 1.5027052094605038,
"grad_norm": 0.04952850937843323,
"learning_rate": 3.5565266861375236e-06,
"loss": 0.0081,
"step": 2430
},
{
"epoch": 1.5088885453702272,
"grad_norm": 0.046748463064432144,
"learning_rate": 3.4744005325017073e-06,
"loss": 0.0079,
"step": 2440
},
{
"epoch": 1.5150718812799506,
"grad_norm": 0.044579874724149704,
"learning_rate": 3.3930338883575564e-06,
"loss": 0.008,
"step": 2450
},
{
"epoch": 1.5212552171896738,
"grad_norm": 0.043162211775779724,
"learning_rate": 3.3124362239060627e-06,
"loss": 0.0083,
"step": 2460
},
{
"epoch": 1.527438553099397,
"grad_norm": 0.04058588296175003,
"learning_rate": 3.2326169198472555e-06,
"loss": 0.0081,
"step": 2470
},
{
"epoch": 1.5336218890091204,
"grad_norm": 0.047970887273550034,
"learning_rate": 3.1535852662884138e-06,
"loss": 0.0081,
"step": 2480
},
{
"epoch": 1.5398052249188439,
"grad_norm": 0.05217019096016884,
"learning_rate": 3.0753504616627694e-06,
"loss": 0.0081,
"step": 2490
},
{
"epoch": 1.545988560828567,
"grad_norm": 0.04357592388987541,
"learning_rate": 2.9979216116589203e-06,
"loss": 0.0086,
"step": 2500
},
{
"epoch": 1.5521718967382903,
"grad_norm": 0.049857549369335175,
"learning_rate": 2.9213077281610336e-06,
"loss": 0.0081,
"step": 2510
},
{
"epoch": 1.5583552326480135,
"grad_norm": 0.04648921638727188,
"learning_rate": 2.8455177281999515e-06,
"loss": 0.0083,
"step": 2520
},
{
"epoch": 1.5645385685577369,
"grad_norm": 0.04828251525759697,
"learning_rate": 2.7705604329153434e-06,
"loss": 0.0085,
"step": 2530
},
{
"epoch": 1.5707219044674603,
"grad_norm": 0.05023375153541565,
"learning_rate": 2.6964445665290274e-06,
"loss": 0.0082,
"step": 2540
},
{
"epoch": 1.5769052403771835,
"grad_norm": 0.045660920441150665,
"learning_rate": 2.6231787553295583e-06,
"loss": 0.0083,
"step": 2550
},
{
"epoch": 1.5830885762869067,
"grad_norm": 0.05497366562485695,
"learning_rate": 2.550771526668221e-06,
"loss": 0.0079,
"step": 2560
},
{
"epoch": 1.58927191219663,
"grad_norm": 0.047413937747478485,
"learning_rate": 2.47923130796655e-06,
"loss": 0.008,
"step": 2570
},
{
"epoch": 1.5954552481063535,
"grad_norm": 0.04655392840504646,
"learning_rate": 2.408566425735446e-06,
"loss": 0.0082,
"step": 2580
},
{
"epoch": 1.6016385840160767,
"grad_norm": 0.047948528081178665,
"learning_rate": 2.338785104606082e-06,
"loss": 0.0082,
"step": 2590
},
{
"epoch": 1.6078219199258,
"grad_norm": 0.04257427155971527,
"learning_rate": 2.26989546637263e-06,
"loss": 0.0079,
"step": 2600
},
{
"epoch": 1.6140052558355231,
"grad_norm": 0.044703159481287,
"learning_rate": 2.2019055290469816e-06,
"loss": 0.0078,
"step": 2610
},
{
"epoch": 1.6201885917452465,
"grad_norm": 0.04206341877579689,
"learning_rate": 2.1348232059255293e-06,
"loss": 0.0081,
"step": 2620
},
{
"epoch": 1.62637192765497,
"grad_norm": 0.047870609909296036,
"learning_rate": 2.0686563046681507e-06,
"loss": 0.0078,
"step": 2630
},
{
"epoch": 1.6325552635646932,
"grad_norm": 0.05116753652691841,
"learning_rate": 2.0034125263894777e-06,
"loss": 0.0083,
"step": 2640
},
{
"epoch": 1.6387385994744164,
"grad_norm": 0.04511834681034088,
"learning_rate": 1.9390994647625693e-06,
"loss": 0.0079,
"step": 2650
},
{
"epoch": 1.6449219353841398,
"grad_norm": 0.043306078761816025,
"learning_rate": 1.8757246051350908e-06,
"loss": 0.0078,
"step": 2660
},
{
"epoch": 1.6511052712938632,
"grad_norm": 0.0442316010594368,
"learning_rate": 1.813295323658103e-06,
"loss": 0.008,
"step": 2670
},
{
"epoch": 1.6572886072035864,
"grad_norm": 0.04422750324010849,
"learning_rate": 1.7518188864275498e-06,
"loss": 0.008,
"step": 2680
},
{
"epoch": 1.6634719431133096,
"grad_norm": 0.04496246948838234,
"learning_rate": 1.6913024486385733e-06,
"loss": 0.0081,
"step": 2690
},
{
"epoch": 1.6696552790230328,
"grad_norm": 0.044760625809431076,
"learning_rate": 1.6317530537527148e-06,
"loss": 0.0078,
"step": 2700
},
{
"epoch": 1.6758386149327562,
"grad_norm": 0.04340161383152008,
"learning_rate": 1.5731776326781435e-06,
"loss": 0.0079,
"step": 2710
},
{
"epoch": 1.6820219508424796,
"grad_norm": 0.0436202809214592,
"learning_rate": 1.5155830029629614e-06,
"loss": 0.0079,
"step": 2720
},
{
"epoch": 1.6882052867522028,
"grad_norm": 0.04061950370669365,
"learning_rate": 1.4589758680017263e-06,
"loss": 0.0079,
"step": 2730
},
{
"epoch": 1.694388622661926,
"grad_norm": 0.04200872778892517,
"learning_rate": 1.403362816255236e-06,
"loss": 0.008,
"step": 2740
},
{
"epoch": 1.7005719585716494,
"grad_norm": 0.04162229970097542,
"learning_rate": 1.3487503204837182e-06,
"loss": 0.0083,
"step": 2750
},
{
"epoch": 1.7067552944813729,
"grad_norm": 0.044568002223968506,
"learning_rate": 1.2951447369934544e-06,
"loss": 0.0075,
"step": 2760
},
{
"epoch": 1.712938630391096,
"grad_norm": 0.048061616718769073,
"learning_rate": 1.2425523048969835e-06,
"loss": 0.0081,
"step": 2770
},
{
"epoch": 1.7191219663008193,
"grad_norm": 0.040904317051172256,
"learning_rate": 1.190979145386939e-06,
"loss": 0.0077,
"step": 2780
},
{
"epoch": 1.7253053022105425,
"grad_norm": 0.04392028972506523,
"learning_rate": 1.1404312610235968e-06,
"loss": 0.0078,
"step": 2790
},
{
"epoch": 1.7314886381202659,
"grad_norm": 0.04398273304104805,
"learning_rate": 1.0909145350362516e-06,
"loss": 0.0079,
"step": 2800
},
{
"epoch": 1.7376719740299893,
"grad_norm": 0.04526762291789055,
"learning_rate": 1.042434730638473e-06,
"loss": 0.0078,
"step": 2810
},
{
"epoch": 1.7438553099397125,
"grad_norm": 0.043778300285339355,
"learning_rate": 9.949974903573223e-07,
"loss": 0.0079,
"step": 2820
},
{
"epoch": 1.7500386458494357,
"grad_norm": 0.04120452329516411,
"learning_rate": 9.486083353766262e-07,
"loss": 0.0078,
"step": 2830
},
{
"epoch": 1.7562219817591591,
"grad_norm": 0.144232839345932,
"learning_rate": 9.032726648943779e-07,
"loss": 0.0078,
"step": 2840
},
{
"epoch": 1.7624053176688823,
"grad_norm": 0.0419781431555748,
"learning_rate": 8.58995755494314e-07,
"loss": 0.0083,
"step": 2850
},
{
"epoch": 1.7685886535786057,
"grad_norm": 0.04286244511604309,
"learning_rate": 8.157827605317892e-07,
"loss": 0.0079,
"step": 2860
},
{
"epoch": 1.774771989488329,
"grad_norm": 0.04345642775297165,
"learning_rate": 7.736387095339748e-07,
"loss": 0.0078,
"step": 2870
},
{
"epoch": 1.7809553253980521,
"grad_norm": 0.041016895323991776,
"learning_rate": 7.325685076144795e-07,
"loss": 0.0082,
"step": 2880
},
{
"epoch": 1.7871386613077755,
"grad_norm": 0.04344632476568222,
"learning_rate": 6.925769349024459e-07,
"loss": 0.0079,
"step": 2890
},
{
"epoch": 1.793321997217499,
"grad_norm": 0.04284526780247688,
"learning_rate": 6.536686459861918e-07,
"loss": 0.0078,
"step": 2900
},
{
"epoch": 1.7995053331272222,
"grad_norm": 0.03983496129512787,
"learning_rate": 6.158481693714735e-07,
"loss": 0.0084,
"step": 2910
},
{
"epoch": 1.8056886690369454,
"grad_norm": 0.04083546996116638,
"learning_rate": 5.791199069544074e-07,
"loss": 0.0077,
"step": 2920
},
{
"epoch": 1.8118720049466688,
"grad_norm": 0.04017137736082077,
"learning_rate": 5.434881335091458e-07,
"loss": 0.0078,
"step": 2930
},
{
"epoch": 1.818055340856392,
"grad_norm": 0.041449110954999924,
"learning_rate": 5.089569961903318e-07,
"loss": 0.0079,
"step": 2940
},
{
"epoch": 1.8242386767661154,
"grad_norm": 0.040258683264255524,
"learning_rate": 4.755305140504185e-07,
"loss": 0.0082,
"step": 2950
},
{
"epoch": 1.8304220126758386,
"grad_norm": 0.03867741674184799,
"learning_rate": 4.432125775718965e-07,
"loss": 0.0076,
"step": 2960
},
{
"epoch": 1.8366053485855618,
"grad_norm": 0.04235534369945526,
"learning_rate": 4.120069482144773e-07,
"loss": 0.0078,
"step": 2970
},
{
"epoch": 1.8427886844952852,
"grad_norm": 0.041860904544591904,
"learning_rate": 3.8191725797730897e-07,
"loss": 0.008,
"step": 2980
},
{
"epoch": 1.8489720204050086,
"grad_norm": 0.04065639153122902,
"learning_rate": 3.529470089762421e-07,
"loss": 0.0077,
"step": 2990
},
{
"epoch": 1.8551553563147318,
"grad_norm": 0.03992839530110359,
"learning_rate": 3.2509957303622496e-07,
"loss": 0.0081,
"step": 3000
},
{
"epoch": 1.861338692224455,
"grad_norm": 0.04205591604113579,
"learning_rate": 2.983781912988604e-07,
"loss": 0.008,
"step": 3010
},
{
"epoch": 1.8675220281341782,
"grad_norm": 0.04142032936215401,
"learning_rate": 2.727859738451721e-07,
"loss": 0.0077,
"step": 3020
},
{
"epoch": 1.8737053640439016,
"grad_norm": 0.040039483457803726,
"learning_rate": 2.4832589933361907e-07,
"loss": 0.0076,
"step": 3030
},
{
"epoch": 1.879888699953625,
"grad_norm": 0.041868723928928375,
"learning_rate": 2.2500081465341993e-07,
"loss": 0.008,
"step": 3040
},
{
"epoch": 1.8860720358633483,
"grad_norm": 0.04264504462480545,
"learning_rate": 2.0281343459319868e-07,
"loss": 0.0081,
"step": 3050
},
{
"epoch": 1.8922553717730715,
"grad_norm": 0.04284176975488663,
"learning_rate": 1.8176634152501816e-07,
"loss": 0.0077,
"step": 3060
},
{
"epoch": 1.8984387076827949,
"grad_norm": 0.043505825102329254,
"learning_rate": 1.618619851038161e-07,
"loss": 0.0077,
"step": 3070
},
{
"epoch": 1.9046220435925183,
"grad_norm": 0.04437592998147011,
"learning_rate": 1.4310268198229115e-07,
"loss": 0.0079,
"step": 3080
},
{
"epoch": 1.9108053795022415,
"grad_norm": 0.04081235080957413,
"learning_rate": 1.2549061554127494e-07,
"loss": 0.0081,
"step": 3090
},
{
"epoch": 1.9169887154119647,
"grad_norm": 0.04169677197933197,
"learning_rate": 1.0902783563560226e-07,
"loss": 0.0078,
"step": 3100
},
{
"epoch": 1.923172051321688,
"grad_norm": 0.04049189016222954,
"learning_rate": 9.371625835553622e-08,
"loss": 0.0076,
"step": 3110
},
{
"epoch": 1.9293553872314113,
"grad_norm": 0.043342310935258865,
"learning_rate": 7.955766580375334e-08,
"loss": 0.0081,
"step": 3120
},
{
"epoch": 1.9355387231411347,
"grad_norm": 0.03951452672481537,
"learning_rate": 6.655370588792732e-08,
"loss": 0.008,
"step": 3130
},
{
"epoch": 1.941722059050858,
"grad_norm": 0.040773287415504456,
"learning_rate": 5.470589212892785e-08,
"loss": 0.0075,
"step": 3140
},
{
"epoch": 1.9479053949605811,
"grad_norm": 0.0426233746111393,
"learning_rate": 4.401560348466505e-08,
"loss": 0.0075,
"step": 3150
},
{
"epoch": 1.9540887308703045,
"grad_norm": 0.039698950946331024,
"learning_rate": 3.448408418959326e-08,
"loss": 0.0079,
"step": 3160
},
{
"epoch": 1.960272066780028,
"grad_norm": 0.04022383317351341,
"learning_rate": 2.6112443609895845e-08,
"loss": 0.0076,
"step": 3170
},
{
"epoch": 1.9664554026897512,
"grad_norm": 0.03998088836669922,
"learning_rate": 1.890165611436734e-08,
"loss": 0.0077,
"step": 3180
},
{
"epoch": 1.9726387385994744,
"grad_norm": 0.03980840742588043,
"learning_rate": 1.285256096100529e-08,
"loss": 0.0079,
"step": 3190
},
{
"epoch": 1.9788220745091976,
"grad_norm": 0.03887724131345749,
"learning_rate": 7.965862199330598e-09,
"loss": 0.0079,
"step": 3200
},
{
"epoch": 1.9843870768279488,
"grad_norm": 0.03944976627826691,
"learning_rate": 4.242128588446415e-09,
"loss": 0.007,
"step": 3210
},
{
"epoch": 1.990570412737672,
"grad_norm": 0.04082844778895378,
"learning_rate": 1.6817935308366395e-09,
"loss": 0.0074,
"step": 3220
},
{
"epoch": 1.9967537486473952,
"grad_norm": 0.04019877314567566,
"learning_rate": 2.851550219240551e-10,
"loss": 0.0069,
"step": 3230
},
{
"epoch": 1.9998454166022568,
"step": 3235,
"total_flos": 7.1289598866872074e+19,
"train_loss": 7.778825080007821e-05,
"train_runtime": 2294.6848,
"train_samples_per_second": 721.684,
"train_steps_per_second": 1.41
}
],
"logging_steps": 10,
"max_steps": 3236,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 7.1289598866872074e+19,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}