SykoLLM-V5.8-Mini / trainer_state.json
SykoSLM's picture
Upload trainer_state.json with huggingface_hub
420bae6 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.2956521739130435,
"eval_steps": 500,
"global_step": 3400,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0008695652173913044,
"grad_norm": 0.0412801168859005,
"learning_rate": 7.875e-06,
"loss": 1.986367416381836,
"step": 10
},
{
"epoch": 0.0017391304347826088,
"grad_norm": 0.04147534817457199,
"learning_rate": 1.6625e-05,
"loss": 1.8889337539672852,
"step": 20
},
{
"epoch": 0.0026086956521739132,
"grad_norm": 0.041172582656145096,
"learning_rate": 2.5374999999999997e-05,
"loss": 1.9236927032470703,
"step": 30
},
{
"epoch": 0.0034782608695652175,
"grad_norm": 0.041271984577178955,
"learning_rate": 3.4125e-05,
"loss": 1.836932373046875,
"step": 40
},
{
"epoch": 0.004347826086956522,
"grad_norm": 0.04144110530614853,
"learning_rate": 4.2875e-05,
"loss": 1.8578208923339843,
"step": 50
},
{
"epoch": 0.0052173913043478265,
"grad_norm": 0.04153507575392723,
"learning_rate": 5.1624999999999994e-05,
"loss": 1.9032476425170899,
"step": 60
},
{
"epoch": 0.00608695652173913,
"grad_norm": 0.04170163720846176,
"learning_rate": 6.0375e-05,
"loss": 1.956137466430664,
"step": 70
},
{
"epoch": 0.006956521739130435,
"grad_norm": 0.04149318486452103,
"learning_rate": 6.9125e-05,
"loss": 1.9041580200195312,
"step": 80
},
{
"epoch": 0.00782608695652174,
"grad_norm": 0.04267331585288048,
"learning_rate": 7.7875e-05,
"loss": 1.8978221893310547,
"step": 90
},
{
"epoch": 0.008695652173913044,
"grad_norm": 0.04074378311634064,
"learning_rate": 8.6625e-05,
"loss": 1.8678691864013672,
"step": 100
},
{
"epoch": 0.009565217391304347,
"grad_norm": 0.0422818660736084,
"learning_rate": 9.5375e-05,
"loss": 1.868328857421875,
"step": 110
},
{
"epoch": 0.010434782608695653,
"grad_norm": 0.041968777775764465,
"learning_rate": 0.000104125,
"loss": 1.891305923461914,
"step": 120
},
{
"epoch": 0.011304347826086957,
"grad_norm": 0.03967316821217537,
"learning_rate": 0.000112875,
"loss": 1.8953823089599608,
"step": 130
},
{
"epoch": 0.01217391304347826,
"grad_norm": 0.04263008385896683,
"learning_rate": 0.00012162499999999999,
"loss": 1.8702178955078126,
"step": 140
},
{
"epoch": 0.013043478260869565,
"grad_norm": 0.042412299662828445,
"learning_rate": 0.000130375,
"loss": 1.8251636505126954,
"step": 150
},
{
"epoch": 0.01391304347826087,
"grad_norm": 0.04295488819479942,
"learning_rate": 0.000139125,
"loss": 1.9118040084838868,
"step": 160
},
{
"epoch": 0.014782608695652174,
"grad_norm": 0.043725475668907166,
"learning_rate": 0.000147875,
"loss": 1.939938735961914,
"step": 170
},
{
"epoch": 0.01565217391304348,
"grad_norm": 0.04236534982919693,
"learning_rate": 0.000156625,
"loss": 1.9232837677001953,
"step": 180
},
{
"epoch": 0.01652173913043478,
"grad_norm": 0.04496081545948982,
"learning_rate": 0.000165375,
"loss": 1.979838752746582,
"step": 190
},
{
"epoch": 0.017391304347826087,
"grad_norm": 0.043251801282167435,
"learning_rate": 0.000174125,
"loss": 1.9420089721679688,
"step": 200
},
{
"epoch": 0.018260869565217393,
"grad_norm": 0.04351160675287247,
"learning_rate": 0.0001749997260913748,
"loss": 1.8539535522460937,
"step": 210
},
{
"epoch": 0.019130434782608695,
"grad_norm": 0.04367993399500847,
"learning_rate": 0.00017499877924894623,
"loss": 1.8686515808105468,
"step": 220
},
{
"epoch": 0.02,
"grad_norm": 0.04434121400117874,
"learning_rate": 0.000174997156098443,
"loss": 1.8759822845458984,
"step": 230
},
{
"epoch": 0.020869565217391306,
"grad_norm": 0.04514431208372116,
"learning_rate": 0.000174994856652411,
"loss": 1.9328535079956055,
"step": 240
},
{
"epoch": 0.021739130434782608,
"grad_norm": 0.04443835839629173,
"learning_rate": 0.00017499188092862342,
"loss": 1.9367387771606446,
"step": 250
},
{
"epoch": 0.022608695652173914,
"grad_norm": 0.04333459958434105,
"learning_rate": 0.00017498822895008068,
"loss": 1.9187299728393554,
"step": 260
},
{
"epoch": 0.023478260869565216,
"grad_norm": 0.04284181818366051,
"learning_rate": 0.00017498390074501016,
"loss": 1.8823593139648438,
"step": 270
},
{
"epoch": 0.02434782608695652,
"grad_norm": 0.042989395558834076,
"learning_rate": 0.00017497889634686607,
"loss": 1.9293874740600585,
"step": 280
},
{
"epoch": 0.025217391304347827,
"grad_norm": 0.04543396085500717,
"learning_rate": 0.00017497321579432908,
"loss": 1.9312299728393554,
"step": 290
},
{
"epoch": 0.02608695652173913,
"grad_norm": 0.045147597789764404,
"learning_rate": 0.00017496685913130617,
"loss": 1.905094528198242,
"step": 300
},
{
"epoch": 0.026956521739130435,
"grad_norm": 0.04503512755036354,
"learning_rate": 0.0001749598264069301,
"loss": 1.9508794784545898,
"step": 310
},
{
"epoch": 0.02782608695652174,
"grad_norm": 0.04418083652853966,
"learning_rate": 0.00017495211767555927,
"loss": 1.93798828125,
"step": 320
},
{
"epoch": 0.028695652173913042,
"grad_norm": 0.0446457639336586,
"learning_rate": 0.0001749437329967771,
"loss": 1.8692955017089843,
"step": 330
},
{
"epoch": 0.029565217391304348,
"grad_norm": 0.045488957315683365,
"learning_rate": 0.00017493467243539165,
"loss": 1.927821731567383,
"step": 340
},
{
"epoch": 0.030434782608695653,
"grad_norm": 0.04600228741765022,
"learning_rate": 0.00017492493606143506,
"loss": 1.9788978576660157,
"step": 350
},
{
"epoch": 0.03130434782608696,
"grad_norm": 0.044314391911029816,
"learning_rate": 0.00017491452395016318,
"loss": 1.8822463989257812,
"step": 360
},
{
"epoch": 0.03217391304347826,
"grad_norm": 0.04477265104651451,
"learning_rate": 0.00017490343618205474,
"loss": 1.923102569580078,
"step": 370
},
{
"epoch": 0.03304347826086956,
"grad_norm": 0.04579043760895729,
"learning_rate": 0.0001748916728428109,
"loss": 1.9119430541992188,
"step": 380
},
{
"epoch": 0.03391304347826087,
"grad_norm": 0.04229314625263214,
"learning_rate": 0.00017487923402335453,
"loss": 1.9644861221313477,
"step": 390
},
{
"epoch": 0.034782608695652174,
"grad_norm": 0.04624266177415848,
"learning_rate": 0.00017486611981982962,
"loss": 1.934438705444336,
"step": 400
},
{
"epoch": 0.035652173913043476,
"grad_norm": 0.044226158410310745,
"learning_rate": 0.00017485233033360026,
"loss": 1.9088932037353517,
"step": 410
},
{
"epoch": 0.036521739130434785,
"grad_norm": 0.0430554524064064,
"learning_rate": 0.0001748378656712501,
"loss": 1.9586982727050781,
"step": 420
},
{
"epoch": 0.03739130434782609,
"grad_norm": 0.04383225739002228,
"learning_rate": 0.00017482272594458156,
"loss": 1.9529533386230469,
"step": 430
},
{
"epoch": 0.03826086956521739,
"grad_norm": 0.04541724920272827,
"learning_rate": 0.00017480691127061475,
"loss": 1.9128194808959962,
"step": 440
},
{
"epoch": 0.0391304347826087,
"grad_norm": 0.045937202870845795,
"learning_rate": 0.00017479042177158665,
"loss": 1.9061151504516602,
"step": 450
},
{
"epoch": 0.04,
"grad_norm": 0.04901459440588951,
"learning_rate": 0.00017477325757495036,
"loss": 1.9450111389160156,
"step": 460
},
{
"epoch": 0.0408695652173913,
"grad_norm": 0.04544888064265251,
"learning_rate": 0.00017475541881337373,
"loss": 1.8662300109863281,
"step": 470
},
{
"epoch": 0.04173913043478261,
"grad_norm": 0.044371675699949265,
"learning_rate": 0.00017473690562473874,
"loss": 1.9551380157470704,
"step": 480
},
{
"epoch": 0.042608695652173914,
"grad_norm": 0.04671357944607735,
"learning_rate": 0.00017471771815214017,
"loss": 1.941476821899414,
"step": 490
},
{
"epoch": 0.043478260869565216,
"grad_norm": 0.04723666235804558,
"learning_rate": 0.00017469785654388454,
"loss": 1.9640892028808594,
"step": 500
},
{
"epoch": 0.04434782608695652,
"grad_norm": 0.04654547944664955,
"learning_rate": 0.00017467732095348905,
"loss": 1.944451141357422,
"step": 510
},
{
"epoch": 0.04521739130434783,
"grad_norm": 0.0464133694767952,
"learning_rate": 0.00017465611153968035,
"loss": 1.9126354217529298,
"step": 520
},
{
"epoch": 0.04608695652173913,
"grad_norm": 0.04279818385839462,
"learning_rate": 0.00017463422846639325,
"loss": 1.9004449844360352,
"step": 530
},
{
"epoch": 0.04695652173913043,
"grad_norm": 0.045143257826566696,
"learning_rate": 0.0001746116719027695,
"loss": 1.8843421936035156,
"step": 540
},
{
"epoch": 0.04782608695652174,
"grad_norm": 0.04402666911482811,
"learning_rate": 0.00017458844202315656,
"loss": 1.9196475982666015,
"step": 550
},
{
"epoch": 0.04869565217391304,
"grad_norm": 0.04636216163635254,
"learning_rate": 0.0001745645390071061,
"loss": 1.9364383697509766,
"step": 560
},
{
"epoch": 0.049565217391304345,
"grad_norm": 0.044472336769104004,
"learning_rate": 0.00017453996303937278,
"loss": 1.9093303680419922,
"step": 570
},
{
"epoch": 0.050434782608695654,
"grad_norm": 0.04402562230825424,
"learning_rate": 0.00017451471430991261,
"loss": 1.9382781982421875,
"step": 580
},
{
"epoch": 0.051304347826086956,
"grad_norm": 0.045037273317575455,
"learning_rate": 0.0001744887930138818,
"loss": 1.9814619064331054,
"step": 590
},
{
"epoch": 0.05217391304347826,
"grad_norm": 0.04517680034041405,
"learning_rate": 0.0001744621993516348,
"loss": 1.8577213287353516,
"step": 600
},
{
"epoch": 0.05304347826086957,
"grad_norm": 0.044585905969142914,
"learning_rate": 0.0001744349335287232,
"loss": 1.906428909301758,
"step": 610
},
{
"epoch": 0.05391304347826087,
"grad_norm": 0.047429412603378296,
"learning_rate": 0.00017440699575589392,
"loss": 1.956072998046875,
"step": 620
},
{
"epoch": 0.05478260869565217,
"grad_norm": 0.04587710648775101,
"learning_rate": 0.0001743783862490876,
"loss": 1.9825408935546875,
"step": 630
},
{
"epoch": 0.05565217391304348,
"grad_norm": 0.044858191162347794,
"learning_rate": 0.00017434910522943687,
"loss": 1.9249252319335937,
"step": 640
},
{
"epoch": 0.05652173913043478,
"grad_norm": 0.046517107635736465,
"learning_rate": 0.0001743191529232648,
"loss": 1.9157310485839845,
"step": 650
},
{
"epoch": 0.057391304347826085,
"grad_norm": 0.04853269085288048,
"learning_rate": 0.00017428852956208308,
"loss": 1.9744232177734375,
"step": 660
},
{
"epoch": 0.058260869565217394,
"grad_norm": 0.04673819616436958,
"learning_rate": 0.0001742572353825901,
"loss": 1.984130859375,
"step": 670
},
{
"epoch": 0.059130434782608696,
"grad_norm": 0.0440078005194664,
"learning_rate": 0.00017422527062666934,
"loss": 1.841623306274414,
"step": 680
},
{
"epoch": 0.06,
"grad_norm": 0.04755179211497307,
"learning_rate": 0.00017419263554138738,
"loss": 1.8390884399414062,
"step": 690
},
{
"epoch": 0.06086956521739131,
"grad_norm": 0.04695327207446098,
"learning_rate": 0.00017415933037899197,
"loss": 1.9590534210205077,
"step": 700
},
{
"epoch": 0.06173913043478261,
"grad_norm": 0.04554755613207817,
"learning_rate": 0.00017412535539691013,
"loss": 1.9262825012207032,
"step": 710
},
{
"epoch": 0.06260869565217392,
"grad_norm": 0.04517935588955879,
"learning_rate": 0.00017409071085774613,
"loss": 1.9321512222290038,
"step": 720
},
{
"epoch": 0.06347826086956522,
"grad_norm": 0.045931559056043625,
"learning_rate": 0.00017405539702927954,
"loss": 1.8974014282226563,
"step": 730
},
{
"epoch": 0.06434782608695652,
"grad_norm": 0.04443606361746788,
"learning_rate": 0.00017401941418446301,
"loss": 1.9469221115112305,
"step": 740
},
{
"epoch": 0.06521739130434782,
"grad_norm": 0.04661998152732849,
"learning_rate": 0.00017398276260142034,
"loss": 1.9025968551635741,
"step": 750
},
{
"epoch": 0.06608695652173913,
"grad_norm": 0.044536832720041275,
"learning_rate": 0.00017394544256344412,
"loss": 1.9783390045166016,
"step": 760
},
{
"epoch": 0.06695652173913043,
"grad_norm": 0.04502754285931587,
"learning_rate": 0.0001739074543589938,
"loss": 1.9079425811767579,
"step": 770
},
{
"epoch": 0.06782608695652174,
"grad_norm": 0.045174986124038696,
"learning_rate": 0.00017386879828169315,
"loss": 1.9875041961669921,
"step": 780
},
{
"epoch": 0.06869565217391305,
"grad_norm": 0.04711156338453293,
"learning_rate": 0.00017382947463032833,
"loss": 1.9256580352783204,
"step": 790
},
{
"epoch": 0.06956521739130435,
"grad_norm": 0.046600691974163055,
"learning_rate": 0.00017378948370884526,
"loss": 1.9062789916992187,
"step": 800
},
{
"epoch": 0.07043478260869565,
"grad_norm": 0.04497547820210457,
"learning_rate": 0.00017374882582634756,
"loss": 1.9336435317993164,
"step": 810
},
{
"epoch": 0.07130434782608695,
"grad_norm": 0.045298535376787186,
"learning_rate": 0.00017370750129709392,
"loss": 1.9081130981445313,
"step": 820
},
{
"epoch": 0.07217391304347825,
"grad_norm": 0.04361579939723015,
"learning_rate": 0.00017366551044049576,
"loss": 1.9321014404296875,
"step": 830
},
{
"epoch": 0.07304347826086957,
"grad_norm": 0.04433725029230118,
"learning_rate": 0.00017362285358111485,
"loss": 1.8640398025512694,
"step": 840
},
{
"epoch": 0.07391304347826087,
"grad_norm": 0.04579870030283928,
"learning_rate": 0.0001735795310486607,
"loss": 1.9060062408447265,
"step": 850
},
{
"epoch": 0.07478260869565218,
"grad_norm": 0.04460195451974869,
"learning_rate": 0.000173535543177988,
"loss": 1.9636484146118165,
"step": 860
},
{
"epoch": 0.07565217391304348,
"grad_norm": 0.046220432966947556,
"learning_rate": 0.00017349089030909407,
"loss": 1.928147315979004,
"step": 870
},
{
"epoch": 0.07652173913043478,
"grad_norm": 0.04606110602617264,
"learning_rate": 0.00017344557278711628,
"loss": 1.9536296844482421,
"step": 880
},
{
"epoch": 0.07739130434782608,
"grad_norm": 0.045817017555236816,
"learning_rate": 0.00017339959096232928,
"loss": 1.9305814743041991,
"step": 890
},
{
"epoch": 0.0782608695652174,
"grad_norm": 0.04592573270201683,
"learning_rate": 0.00017335294519014243,
"loss": 1.8813438415527344,
"step": 900
},
{
"epoch": 0.0791304347826087,
"grad_norm": 0.04669544845819473,
"learning_rate": 0.0001733056358310968,
"loss": 1.9484172821044923,
"step": 910
},
{
"epoch": 0.08,
"grad_norm": 0.04449376091361046,
"learning_rate": 0.0001732576632508627,
"loss": 1.8734165191650392,
"step": 920
},
{
"epoch": 0.0808695652173913,
"grad_norm": 0.04569607973098755,
"learning_rate": 0.00017320902782023666,
"loss": 1.904270553588867,
"step": 930
},
{
"epoch": 0.0817391304347826,
"grad_norm": 0.047245677560567856,
"learning_rate": 0.00017315972991513857,
"loss": 1.944266128540039,
"step": 940
},
{
"epoch": 0.08260869565217391,
"grad_norm": 0.04679643735289574,
"learning_rate": 0.00017310976991660891,
"loss": 1.9376203536987304,
"step": 950
},
{
"epoch": 0.08347826086956522,
"grad_norm": 0.04880222678184509,
"learning_rate": 0.00017305914821080562,
"loss": 2.0033763885498046,
"step": 960
},
{
"epoch": 0.08434782608695653,
"grad_norm": 0.04653260484337807,
"learning_rate": 0.0001730078651890012,
"loss": 1.877172088623047,
"step": 970
},
{
"epoch": 0.08521739130434783,
"grad_norm": 0.04537530615925789,
"learning_rate": 0.00017295592124757974,
"loss": 1.8832229614257812,
"step": 980
},
{
"epoch": 0.08608695652173913,
"grad_norm": 0.0438741110265255,
"learning_rate": 0.00017290331678803373,
"loss": 1.8948379516601563,
"step": 990
},
{
"epoch": 0.08695652173913043,
"grad_norm": 0.04543806612491608,
"learning_rate": 0.00017285005221696114,
"loss": 1.8686225891113282,
"step": 1000
},
{
"epoch": 0.08782608695652173,
"grad_norm": 0.044838301837444305,
"learning_rate": 0.0001727961279460621,
"loss": 1.9170583724975585,
"step": 1010
},
{
"epoch": 0.08869565217391304,
"grad_norm": 0.04491033777594566,
"learning_rate": 0.00017274154439213577,
"loss": 1.9348651885986328,
"step": 1020
},
{
"epoch": 0.08956521739130435,
"grad_norm": 0.045102160423994064,
"learning_rate": 0.00017268630197707715,
"loss": 1.9238882064819336,
"step": 1030
},
{
"epoch": 0.09043478260869565,
"grad_norm": 0.044371772557497025,
"learning_rate": 0.00017263040112787387,
"loss": 1.8916519165039063,
"step": 1040
},
{
"epoch": 0.09130434782608696,
"grad_norm": 0.04538256302475929,
"learning_rate": 0.00017257384227660271,
"loss": 1.8470991134643555,
"step": 1050
},
{
"epoch": 0.09217391304347826,
"grad_norm": 0.046616047620773315,
"learning_rate": 0.00017251662586042644,
"loss": 1.9685466766357422,
"step": 1060
},
{
"epoch": 0.09304347826086956,
"grad_norm": 0.04723796248435974,
"learning_rate": 0.0001724587523215904,
"loss": 1.8623455047607422,
"step": 1070
},
{
"epoch": 0.09391304347826086,
"grad_norm": 0.04548574984073639,
"learning_rate": 0.00017240022210741893,
"loss": 1.8760478973388672,
"step": 1080
},
{
"epoch": 0.09478260869565218,
"grad_norm": 0.04325335472822189,
"learning_rate": 0.00017234103567031225,
"loss": 1.839696502685547,
"step": 1090
},
{
"epoch": 0.09565217391304348,
"grad_norm": 0.04418843984603882,
"learning_rate": 0.00017228119346774257,
"loss": 1.9864925384521483,
"step": 1100
},
{
"epoch": 0.09652173913043478,
"grad_norm": 0.04545409604907036,
"learning_rate": 0.00017222069596225078,
"loss": 1.9569078445434571,
"step": 1110
},
{
"epoch": 0.09739130434782609,
"grad_norm": 0.047114428132772446,
"learning_rate": 0.00017215954362144285,
"loss": 1.94586181640625,
"step": 1120
},
{
"epoch": 0.09826086956521739,
"grad_norm": 0.046218354254961014,
"learning_rate": 0.00017209773691798628,
"loss": 1.959282684326172,
"step": 1130
},
{
"epoch": 0.09913043478260869,
"grad_norm": 0.04634036496281624,
"learning_rate": 0.00017203527632960619,
"loss": 1.8650602340698241,
"step": 1140
},
{
"epoch": 0.1,
"grad_norm": 0.048119544982910156,
"learning_rate": 0.0001719721623390819,
"loss": 1.9122610092163086,
"step": 1150
},
{
"epoch": 0.10086956521739131,
"grad_norm": 0.045855700969696045,
"learning_rate": 0.00017190839543424312,
"loss": 1.9363866806030274,
"step": 1160
},
{
"epoch": 0.10173913043478261,
"grad_norm": 0.04780857637524605,
"learning_rate": 0.0001718439761079661,
"loss": 1.9255781173706055,
"step": 1170
},
{
"epoch": 0.10260869565217391,
"grad_norm": 0.044750381261110306,
"learning_rate": 0.00017177890485816993,
"loss": 1.8942161560058595,
"step": 1180
},
{
"epoch": 0.10347826086956521,
"grad_norm": 0.0476827546954155,
"learning_rate": 0.0001717131821878126,
"loss": 1.9411766052246093,
"step": 1190
},
{
"epoch": 0.10434782608695652,
"grad_norm": 0.046074964106082916,
"learning_rate": 0.0001716468086048872,
"loss": 1.8827497482299804,
"step": 1200
},
{
"epoch": 0.10521739130434783,
"grad_norm": 0.04519912600517273,
"learning_rate": 0.0001715797846224179,
"loss": 1.9211563110351562,
"step": 1210
},
{
"epoch": 0.10608695652173913,
"grad_norm": 0.04722365736961365,
"learning_rate": 0.00017151211075845603,
"loss": 1.9217552185058593,
"step": 1220
},
{
"epoch": 0.10695652173913044,
"grad_norm": 0.046000994741916656,
"learning_rate": 0.00017144378753607608,
"loss": 1.968286895751953,
"step": 1230
},
{
"epoch": 0.10782608695652174,
"grad_norm": 0.04773630201816559,
"learning_rate": 0.00017137481548337174,
"loss": 1.9427530288696289,
"step": 1240
},
{
"epoch": 0.10869565217391304,
"grad_norm": 0.04625891149044037,
"learning_rate": 0.0001713051951334516,
"loss": 1.8612529754638671,
"step": 1250
},
{
"epoch": 0.10956521739130434,
"grad_norm": 0.04403012990951538,
"learning_rate": 0.00017123492702443522,
"loss": 1.91204776763916,
"step": 1260
},
{
"epoch": 0.11043478260869566,
"grad_norm": 0.04643299803137779,
"learning_rate": 0.00017116401169944896,
"loss": 1.9106037139892578,
"step": 1270
},
{
"epoch": 0.11130434782608696,
"grad_norm": 0.04438823461532593,
"learning_rate": 0.00017109244970662165,
"loss": 1.9446937561035156,
"step": 1280
},
{
"epoch": 0.11217391304347826,
"grad_norm": 0.04284852743148804,
"learning_rate": 0.0001710202415990805,
"loss": 1.8916801452636718,
"step": 1290
},
{
"epoch": 0.11304347826086956,
"grad_norm": 0.04538673162460327,
"learning_rate": 0.00017094738793494674,
"loss": 1.9826034545898437,
"step": 1300
},
{
"epoch": 0.11391304347826087,
"grad_norm": 0.045958247035741806,
"learning_rate": 0.0001708738892773313,
"loss": 1.918802261352539,
"step": 1310
},
{
"epoch": 0.11478260869565217,
"grad_norm": 0.043768592178821564,
"learning_rate": 0.00017079974619433054,
"loss": 1.9888408660888672,
"step": 1320
},
{
"epoch": 0.11565217391304349,
"grad_norm": 0.04695571959018707,
"learning_rate": 0.00017072495925902174,
"loss": 1.9092866897583007,
"step": 1330
},
{
"epoch": 0.11652173913043479,
"grad_norm": 0.04605986550450325,
"learning_rate": 0.00017064952904945885,
"loss": 1.9453262329101562,
"step": 1340
},
{
"epoch": 0.11739130434782609,
"grad_norm": 0.047600872814655304,
"learning_rate": 0.00017057345614866776,
"loss": 1.932131576538086,
"step": 1350
},
{
"epoch": 0.11826086956521739,
"grad_norm": 0.04641648009419441,
"learning_rate": 0.00017049674114464204,
"loss": 1.9330526351928712,
"step": 1360
},
{
"epoch": 0.1191304347826087,
"grad_norm": 0.04527211934328079,
"learning_rate": 0.0001704193846303382,
"loss": 1.9231037139892577,
"step": 1370
},
{
"epoch": 0.12,
"grad_norm": 0.049683600664138794,
"learning_rate": 0.0001703413872036714,
"loss": 1.988457489013672,
"step": 1380
},
{
"epoch": 0.1208695652173913,
"grad_norm": 0.045173171907663345,
"learning_rate": 0.0001702627494675104,
"loss": 1.952442741394043,
"step": 1390
},
{
"epoch": 0.12173913043478261,
"grad_norm": 0.04802786186337471,
"learning_rate": 0.00017018347202967326,
"loss": 1.915532112121582,
"step": 1400
},
{
"epoch": 0.12260869565217392,
"grad_norm": 0.04501411318778992,
"learning_rate": 0.0001701035555029225,
"loss": 1.9840251922607421,
"step": 1410
},
{
"epoch": 0.12347826086956522,
"grad_norm": 0.043427255004644394,
"learning_rate": 0.00017002300050496038,
"loss": 1.8686183929443358,
"step": 1420
},
{
"epoch": 0.12434782608695652,
"grad_norm": 0.04784533753991127,
"learning_rate": 0.00016994180765842406,
"loss": 1.8943784713745118,
"step": 1430
},
{
"epoch": 0.12521739130434784,
"grad_norm": 0.044591404497623444,
"learning_rate": 0.0001698599775908809,
"loss": 1.8551948547363282,
"step": 1440
},
{
"epoch": 0.12608695652173912,
"grad_norm": 0.04831645265221596,
"learning_rate": 0.0001697775109348236,
"loss": 1.890420913696289,
"step": 1450
},
{
"epoch": 0.12695652173913044,
"grad_norm": 0.04690225422382355,
"learning_rate": 0.0001696944083276652,
"loss": 1.9546632766723633,
"step": 1460
},
{
"epoch": 0.12782608695652173,
"grad_norm": 0.04488781467080116,
"learning_rate": 0.00016961067041173432,
"loss": 1.8585077285766602,
"step": 1470
},
{
"epoch": 0.12869565217391304,
"grad_norm": 0.04684244468808174,
"learning_rate": 0.00016952629783426992,
"loss": 1.9355541229248048,
"step": 1480
},
{
"epoch": 0.12956521739130436,
"grad_norm": 0.04604167491197586,
"learning_rate": 0.00016944129124741664,
"loss": 1.8615631103515624,
"step": 1490
},
{
"epoch": 0.13043478260869565,
"grad_norm": 0.048263128846883774,
"learning_rate": 0.00016935565130821956,
"loss": 1.9315719604492188,
"step": 1500
},
{
"epoch": 0.13130434782608696,
"grad_norm": 0.04678445681929588,
"learning_rate": 0.0001692693786786191,
"loss": 1.9562299728393555,
"step": 1510
},
{
"epoch": 0.13217391304347825,
"grad_norm": 0.04416315257549286,
"learning_rate": 0.0001691824740254461,
"loss": 1.8920721054077148,
"step": 1520
},
{
"epoch": 0.13304347826086957,
"grad_norm": 0.044212035834789276,
"learning_rate": 0.00016909493802041628,
"loss": 1.8729904174804688,
"step": 1530
},
{
"epoch": 0.13391304347826086,
"grad_norm": 0.046913452446460724,
"learning_rate": 0.0001690067713401256,
"loss": 1.9575483322143554,
"step": 1540
},
{
"epoch": 0.13478260869565217,
"grad_norm": 0.04588250815868378,
"learning_rate": 0.00016891797466604444,
"loss": 1.9156288146972655,
"step": 1550
},
{
"epoch": 0.1356521739130435,
"grad_norm": 0.04520321637392044,
"learning_rate": 0.00016882854868451285,
"loss": 1.9655925750732421,
"step": 1560
},
{
"epoch": 0.13652173913043478,
"grad_norm": 0.04560220241546631,
"learning_rate": 0.00016873849408673486,
"loss": 1.892664337158203,
"step": 1570
},
{
"epoch": 0.1373913043478261,
"grad_norm": 0.04517337679862976,
"learning_rate": 0.00016864781156877335,
"loss": 1.9007608413696289,
"step": 1580
},
{
"epoch": 0.13826086956521738,
"grad_norm": 0.046717483550310135,
"learning_rate": 0.00016855650183154463,
"loss": 1.9165136337280273,
"step": 1590
},
{
"epoch": 0.1391304347826087,
"grad_norm": 0.04691331461071968,
"learning_rate": 0.0001684645655808129,
"loss": 1.8728553771972656,
"step": 1600
},
{
"epoch": 0.14,
"grad_norm": 0.04566628113389015,
"learning_rate": 0.00016837200352718507,
"loss": 1.883155059814453,
"step": 1610
},
{
"epoch": 0.1408695652173913,
"grad_norm": 0.04991758242249489,
"learning_rate": 0.00016827881638610496,
"loss": 1.9257890701293945,
"step": 1620
},
{
"epoch": 0.14173913043478262,
"grad_norm": 0.04652441665530205,
"learning_rate": 0.0001681850048778479,
"loss": 1.9192398071289063,
"step": 1630
},
{
"epoch": 0.1426086956521739,
"grad_norm": 0.04636216163635254,
"learning_rate": 0.00016809056972751527,
"loss": 1.9711288452148437,
"step": 1640
},
{
"epoch": 0.14347826086956522,
"grad_norm": 0.04579229652881622,
"learning_rate": 0.00016799551166502873,
"loss": 1.9829723358154296,
"step": 1650
},
{
"epoch": 0.1443478260869565,
"grad_norm": 0.044148631393909454,
"learning_rate": 0.00016789983142512463,
"loss": 1.912728500366211,
"step": 1660
},
{
"epoch": 0.14521739130434783,
"grad_norm": 0.046788472682237625,
"learning_rate": 0.00016780352974734834,
"loss": 1.938332748413086,
"step": 1670
},
{
"epoch": 0.14608695652173914,
"grad_norm": 0.04814344644546509,
"learning_rate": 0.00016770660737604862,
"loss": 1.9140899658203125,
"step": 1680
},
{
"epoch": 0.14695652173913043,
"grad_norm": 0.046944644302129745,
"learning_rate": 0.00016760906506037166,
"loss": 1.9023950576782227,
"step": 1690
},
{
"epoch": 0.14782608695652175,
"grad_norm": 0.04715333133935928,
"learning_rate": 0.00016751090355425548,
"loss": 1.9201131820678712,
"step": 1700
},
{
"epoch": 0.14869565217391303,
"grad_norm": 0.045025281608104706,
"learning_rate": 0.00016741212361642407,
"loss": 1.885734748840332,
"step": 1710
},
{
"epoch": 0.14956521739130435,
"grad_norm": 0.04626510292291641,
"learning_rate": 0.00016731272601038137,
"loss": 1.939837646484375,
"step": 1720
},
{
"epoch": 0.15043478260869564,
"grad_norm": 0.04652830585837364,
"learning_rate": 0.00016721271150440568,
"loss": 1.880080223083496,
"step": 1730
},
{
"epoch": 0.15130434782608695,
"grad_norm": 0.044095490127801895,
"learning_rate": 0.00016711208087154334,
"loss": 1.9228801727294922,
"step": 1740
},
{
"epoch": 0.15217391304347827,
"grad_norm": 0.04645839333534241,
"learning_rate": 0.00016701083488960302,
"loss": 1.939291763305664,
"step": 1750
},
{
"epoch": 0.15304347826086956,
"grad_norm": 0.04704873263835907,
"learning_rate": 0.00016690897434114965,
"loss": 1.9802566528320313,
"step": 1760
},
{
"epoch": 0.15391304347826087,
"grad_norm": 0.044596049934625626,
"learning_rate": 0.00016680650001349838,
"loss": 1.9283245086669922,
"step": 1770
},
{
"epoch": 0.15478260869565216,
"grad_norm": 0.04547381401062012,
"learning_rate": 0.0001667034126987084,
"loss": 1.882451629638672,
"step": 1780
},
{
"epoch": 0.15565217391304348,
"grad_norm": 0.04559491574764252,
"learning_rate": 0.00016659971319357694,
"loss": 1.9023427963256836,
"step": 1790
},
{
"epoch": 0.1565217391304348,
"grad_norm": 0.04723436385393143,
"learning_rate": 0.00016649540229963304,
"loss": 1.9267019271850585,
"step": 1800
},
{
"epoch": 0.15739130434782608,
"grad_norm": 0.04553121328353882,
"learning_rate": 0.00016639048082313138,
"loss": 1.8939033508300782,
"step": 1810
},
{
"epoch": 0.1582608695652174,
"grad_norm": 0.046808917075395584,
"learning_rate": 0.000166284949575046,
"loss": 1.8823970794677733,
"step": 1820
},
{
"epoch": 0.1591304347826087,
"grad_norm": 0.04732776805758476,
"learning_rate": 0.00016617880937106413,
"loss": 1.8905269622802734,
"step": 1830
},
{
"epoch": 0.16,
"grad_norm": 0.04512569308280945,
"learning_rate": 0.00016607206103157982,
"loss": 1.891006851196289,
"step": 1840
},
{
"epoch": 0.1608695652173913,
"grad_norm": 0.0452556274831295,
"learning_rate": 0.0001659647053816876,
"loss": 1.8488157272338868,
"step": 1850
},
{
"epoch": 0.1617391304347826,
"grad_norm": 0.04695947468280792,
"learning_rate": 0.00016585674325117613,
"loss": 1.8792579650878907,
"step": 1860
},
{
"epoch": 0.16260869565217392,
"grad_norm": 0.046590715646743774,
"learning_rate": 0.0001657481754745217,
"loss": 1.905181884765625,
"step": 1870
},
{
"epoch": 0.1634782608695652,
"grad_norm": 0.04671342298388481,
"learning_rate": 0.00016563900289088187,
"loss": 1.879676055908203,
"step": 1880
},
{
"epoch": 0.16434782608695653,
"grad_norm": 0.04555339738726616,
"learning_rate": 0.0001655292263440891,
"loss": 1.9276203155517577,
"step": 1890
},
{
"epoch": 0.16521739130434782,
"grad_norm": 0.04350179061293602,
"learning_rate": 0.00016541884668264383,
"loss": 1.8957508087158204,
"step": 1900
},
{
"epoch": 0.16608695652173913,
"grad_norm": 0.04489278420805931,
"learning_rate": 0.0001653078647597084,
"loss": 1.9363662719726562,
"step": 1910
},
{
"epoch": 0.16695652173913045,
"grad_norm": 0.04279284551739693,
"learning_rate": 0.00016519628143310015,
"loss": 1.953699493408203,
"step": 1920
},
{
"epoch": 0.16782608695652174,
"grad_norm": 0.04596225544810295,
"learning_rate": 0.00016508409756528486,
"loss": 1.9585641860961913,
"step": 1930
},
{
"epoch": 0.16869565217391305,
"grad_norm": 0.04413312301039696,
"learning_rate": 0.00016497131402337016,
"loss": 1.8821155548095703,
"step": 1940
},
{
"epoch": 0.16956521739130434,
"grad_norm": 0.04498974233865738,
"learning_rate": 0.00016485793167909864,
"loss": 1.9019199371337892,
"step": 1950
},
{
"epoch": 0.17043478260869566,
"grad_norm": 0.04706019163131714,
"learning_rate": 0.00016474395140884143,
"loss": 1.8370037078857422,
"step": 1960
},
{
"epoch": 0.17130434782608694,
"grad_norm": 0.04894762858748436,
"learning_rate": 0.0001646293740935911,
"loss": 1.9752735137939452,
"step": 1970
},
{
"epoch": 0.17217391304347826,
"grad_norm": 0.04635745286941528,
"learning_rate": 0.00016451420061895498,
"loss": 1.9637317657470703,
"step": 1980
},
{
"epoch": 0.17304347826086958,
"grad_norm": 0.04573964327573776,
"learning_rate": 0.00016439843187514836,
"loss": 1.9148262023925782,
"step": 1990
},
{
"epoch": 0.17391304347826086,
"grad_norm": 0.045651230961084366,
"learning_rate": 0.00016428206875698762,
"loss": 1.8789260864257813,
"step": 2000
},
{
"epoch": 0.17478260869565218,
"grad_norm": 0.047087036073207855,
"learning_rate": 0.00016416511216388316,
"loss": 1.913290786743164,
"step": 2010
},
{
"epoch": 0.17565217391304347,
"grad_norm": 0.045222099870443344,
"learning_rate": 0.0001640475629998326,
"loss": 1.9029350280761719,
"step": 2020
},
{
"epoch": 0.17652173913043478,
"grad_norm": 0.04745696485042572,
"learning_rate": 0.00016392942217341382,
"loss": 1.9319425582885743,
"step": 2030
},
{
"epoch": 0.17739130434782607,
"grad_norm": 0.047783516347408295,
"learning_rate": 0.00016381069059777769,
"loss": 1.8861225128173829,
"step": 2040
},
{
"epoch": 0.1782608695652174,
"grad_norm": 0.04685746133327484,
"learning_rate": 0.00016369136919064132,
"loss": 1.990982437133789,
"step": 2050
},
{
"epoch": 0.1791304347826087,
"grad_norm": 0.0454823300242424,
"learning_rate": 0.0001635714588742808,
"loss": 1.859075164794922,
"step": 2060
},
{
"epoch": 0.18,
"grad_norm": 0.044694554060697556,
"learning_rate": 0.00016345096057552409,
"loss": 1.926215934753418,
"step": 2070
},
{
"epoch": 0.1808695652173913,
"grad_norm": 0.04416250064969063,
"learning_rate": 0.00016332987522574383,
"loss": 1.8868453979492188,
"step": 2080
},
{
"epoch": 0.1817391304347826,
"grad_norm": 0.04497243091464043,
"learning_rate": 0.00016320820376085026,
"loss": 1.902182388305664,
"step": 2090
},
{
"epoch": 0.1826086956521739,
"grad_norm": 0.0451941043138504,
"learning_rate": 0.00016308594712128384,
"loss": 1.960106658935547,
"step": 2100
},
{
"epoch": 0.18347826086956523,
"grad_norm": 0.046650130301713943,
"learning_rate": 0.0001629631062520081,
"loss": 1.9553258895874024,
"step": 2110
},
{
"epoch": 0.18434782608695652,
"grad_norm": 0.04722883552312851,
"learning_rate": 0.0001628396821025022,
"loss": 1.9351499557495118,
"step": 2120
},
{
"epoch": 0.18521739130434783,
"grad_norm": 0.04475412890315056,
"learning_rate": 0.00016271567562675378,
"loss": 1.8991554260253907,
"step": 2130
},
{
"epoch": 0.18608695652173912,
"grad_norm": 0.0482119657099247,
"learning_rate": 0.0001625910877832514,
"loss": 1.9038169860839844,
"step": 2140
},
{
"epoch": 0.18695652173913044,
"grad_norm": 0.04647846147418022,
"learning_rate": 0.0001624659195349772,
"loss": 1.8781488418579102,
"step": 2150
},
{
"epoch": 0.18782608695652173,
"grad_norm": 0.04708510637283325,
"learning_rate": 0.00016234017184939953,
"loss": 1.951561164855957,
"step": 2160
},
{
"epoch": 0.18869565217391304,
"grad_norm": 0.04458949714899063,
"learning_rate": 0.00016221384569846539,
"loss": 1.9635744094848633,
"step": 2170
},
{
"epoch": 0.18956521739130436,
"grad_norm": 0.04679902642965317,
"learning_rate": 0.00016208694205859287,
"loss": 1.941354751586914,
"step": 2180
},
{
"epoch": 0.19043478260869565,
"grad_norm": 0.04542798921465874,
"learning_rate": 0.0001619594619106638,
"loss": 1.9266124725341798,
"step": 2190
},
{
"epoch": 0.19130434782608696,
"grad_norm": 0.04602201655507088,
"learning_rate": 0.00016183140624001594,
"loss": 1.968562126159668,
"step": 2200
},
{
"epoch": 0.19217391304347825,
"grad_norm": 0.045000337064266205,
"learning_rate": 0.0001617027760364355,
"loss": 1.9628772735595703,
"step": 2210
},
{
"epoch": 0.19304347826086957,
"grad_norm": 0.043177004903554916,
"learning_rate": 0.00016157357229414941,
"loss": 1.918726348876953,
"step": 2220
},
{
"epoch": 0.19391304347826088,
"grad_norm": 0.045546501874923706,
"learning_rate": 0.00016144379601181773,
"loss": 1.8972591400146483,
"step": 2230
},
{
"epoch": 0.19478260869565217,
"grad_norm": 0.04712609574198723,
"learning_rate": 0.00016131344819252585,
"loss": 1.8813795089721679,
"step": 2240
},
{
"epoch": 0.1956521739130435,
"grad_norm": 0.04688692092895508,
"learning_rate": 0.00016118252984377674,
"loss": 1.9304527282714843,
"step": 2250
},
{
"epoch": 0.19652173913043477,
"grad_norm": 0.04640372470021248,
"learning_rate": 0.00016105104197748328,
"loss": 1.9001270294189454,
"step": 2260
},
{
"epoch": 0.1973913043478261,
"grad_norm": 0.04748000577092171,
"learning_rate": 0.00016091898560996026,
"loss": 1.8875625610351563,
"step": 2270
},
{
"epoch": 0.19826086956521738,
"grad_norm": 0.04630134627223015,
"learning_rate": 0.00016078636176191662,
"loss": 1.880694580078125,
"step": 2280
},
{
"epoch": 0.1991304347826087,
"grad_norm": 0.04616926237940788,
"learning_rate": 0.00016065317145844756,
"loss": 1.919076156616211,
"step": 2290
},
{
"epoch": 0.2,
"grad_norm": 0.04642150551080704,
"learning_rate": 0.00016051941572902664,
"loss": 1.9145648956298829,
"step": 2300
},
{
"epoch": 0.2008695652173913,
"grad_norm": 0.04683443531394005,
"learning_rate": 0.00016038509560749778,
"loss": 1.998061180114746,
"step": 2310
},
{
"epoch": 0.20173913043478262,
"grad_norm": 0.04580046609044075,
"learning_rate": 0.00016025021213206722,
"loss": 1.8663105010986327,
"step": 2320
},
{
"epoch": 0.2026086956521739,
"grad_norm": 0.046270985156297684,
"learning_rate": 0.00016011476634529567,
"loss": 1.9407358169555664,
"step": 2330
},
{
"epoch": 0.20347826086956522,
"grad_norm": 0.047260086983442307,
"learning_rate": 0.00015997875929409005,
"loss": 1.8749610900878906,
"step": 2340
},
{
"epoch": 0.20434782608695654,
"grad_norm": 0.04815071448683739,
"learning_rate": 0.00015984219202969544,
"loss": 1.9252590179443358,
"step": 2350
},
{
"epoch": 0.20521739130434782,
"grad_norm": 0.04547717794775963,
"learning_rate": 0.00015970506560768715,
"loss": 1.8544036865234375,
"step": 2360
},
{
"epoch": 0.20608695652173914,
"grad_norm": 0.04534965381026268,
"learning_rate": 0.00015956738108796227,
"loss": 1.9171245574951172,
"step": 2370
},
{
"epoch": 0.20695652173913043,
"grad_norm": 0.0445626862347126,
"learning_rate": 0.00015942913953473172,
"loss": 1.8566919326782227,
"step": 2380
},
{
"epoch": 0.20782608695652174,
"grad_norm": 0.04675230383872986,
"learning_rate": 0.00015929034201651185,
"loss": 1.9092636108398438,
"step": 2390
},
{
"epoch": 0.20869565217391303,
"grad_norm": 0.04733717441558838,
"learning_rate": 0.00015915098960611628,
"loss": 1.9153020858764649,
"step": 2400
},
{
"epoch": 0.20956521739130435,
"grad_norm": 0.04689701274037361,
"learning_rate": 0.0001590110833806476,
"loss": 1.8826875686645508,
"step": 2410
},
{
"epoch": 0.21043478260869566,
"grad_norm": 0.04545078054070473,
"learning_rate": 0.000158870624421489,
"loss": 1.8986186981201172,
"step": 2420
},
{
"epoch": 0.21130434782608695,
"grad_norm": 0.04620114341378212,
"learning_rate": 0.00015872961381429592,
"loss": 1.9260257720947265,
"step": 2430
},
{
"epoch": 0.21217391304347827,
"grad_norm": 0.047381263226270676,
"learning_rate": 0.00015858805264898777,
"loss": 1.928044319152832,
"step": 2440
},
{
"epoch": 0.21304347826086956,
"grad_norm": 0.04613082483410835,
"learning_rate": 0.00015844594201973935,
"loss": 1.9063533782958983,
"step": 2450
},
{
"epoch": 0.21391304347826087,
"grad_norm": 0.04579924792051315,
"learning_rate": 0.00015830328302497235,
"loss": 1.8655834197998047,
"step": 2460
},
{
"epoch": 0.21478260869565216,
"grad_norm": 0.04783694073557854,
"learning_rate": 0.0001581600767673472,
"loss": 1.8869522094726563,
"step": 2470
},
{
"epoch": 0.21565217391304348,
"grad_norm": 0.04724857956171036,
"learning_rate": 0.0001580163243537541,
"loss": 1.911419677734375,
"step": 2480
},
{
"epoch": 0.2165217391304348,
"grad_norm": 0.046219322830438614,
"learning_rate": 0.00015787202689530476,
"loss": 1.878607940673828,
"step": 2490
},
{
"epoch": 0.21739130434782608,
"grad_norm": 0.04494387283921242,
"learning_rate": 0.00015772718550732375,
"loss": 1.9234241485595702,
"step": 2500
},
{
"epoch": 0.2182608695652174,
"grad_norm": 0.04502217471599579,
"learning_rate": 0.00015758180130933985,
"loss": 1.915613555908203,
"step": 2510
},
{
"epoch": 0.21913043478260869,
"grad_norm": 0.04806201159954071,
"learning_rate": 0.00015743587542507736,
"loss": 1.8788349151611328,
"step": 2520
},
{
"epoch": 0.22,
"grad_norm": 0.04505891352891922,
"learning_rate": 0.00015728940898244748,
"loss": 1.892514419555664,
"step": 2530
},
{
"epoch": 0.22086956521739132,
"grad_norm": 0.047691989690065384,
"learning_rate": 0.00015714240311353963,
"loss": 1.919693374633789,
"step": 2540
},
{
"epoch": 0.2217391304347826,
"grad_norm": 0.04654225334525108,
"learning_rate": 0.00015699485895461252,
"loss": 1.8307640075683593,
"step": 2550
},
{
"epoch": 0.22260869565217392,
"grad_norm": 0.04682806134223938,
"learning_rate": 0.00015684677764608564,
"loss": 1.8828701019287108,
"step": 2560
},
{
"epoch": 0.2234782608695652,
"grad_norm": 0.04656338319182396,
"learning_rate": 0.00015669816033253014,
"loss": 1.877579689025879,
"step": 2570
},
{
"epoch": 0.22434782608695653,
"grad_norm": 0.0479404553771019,
"learning_rate": 0.00015654900816266024,
"loss": 1.894699478149414,
"step": 2580
},
{
"epoch": 0.2252173913043478,
"grad_norm": 0.047481246292591095,
"learning_rate": 0.00015639932228932418,
"loss": 1.8986604690551758,
"step": 2590
},
{
"epoch": 0.22608695652173913,
"grad_norm": 0.04713468998670578,
"learning_rate": 0.0001562491038694955,
"loss": 1.9140901565551758,
"step": 2600
},
{
"epoch": 0.22695652173913045,
"grad_norm": 0.046244435012340546,
"learning_rate": 0.0001560983540642638,
"loss": 1.8460613250732423,
"step": 2610
},
{
"epoch": 0.22782608695652173,
"grad_norm": 0.0466030091047287,
"learning_rate": 0.00015594707403882607,
"loss": 1.9079011917114257,
"step": 2620
},
{
"epoch": 0.22869565217391305,
"grad_norm": 0.04684355854988098,
"learning_rate": 0.00015579526496247752,
"loss": 1.923720932006836,
"step": 2630
},
{
"epoch": 0.22956521739130434,
"grad_norm": 0.04607565701007843,
"learning_rate": 0.00015564292800860252,
"loss": 1.8996570587158204,
"step": 2640
},
{
"epoch": 0.23043478260869565,
"grad_norm": 0.0482359416782856,
"learning_rate": 0.00015549006435466566,
"loss": 1.8546255111694336,
"step": 2650
},
{
"epoch": 0.23130434782608697,
"grad_norm": 0.04718746244907379,
"learning_rate": 0.0001553366751822026,
"loss": 1.9036752700805664,
"step": 2660
},
{
"epoch": 0.23217391304347826,
"grad_norm": 0.04525565356016159,
"learning_rate": 0.00015518276167681078,
"loss": 1.878532600402832,
"step": 2670
},
{
"epoch": 0.23304347826086957,
"grad_norm": 0.045305028557777405,
"learning_rate": 0.0001550283250281405,
"loss": 1.8548662185668945,
"step": 2680
},
{
"epoch": 0.23391304347826086,
"grad_norm": 0.04721986874938011,
"learning_rate": 0.00015487336642988563,
"loss": 1.898862075805664,
"step": 2690
},
{
"epoch": 0.23478260869565218,
"grad_norm": 0.0476800911128521,
"learning_rate": 0.00015471788707977424,
"loss": 1.8859397888183593,
"step": 2700
},
{
"epoch": 0.23565217391304347,
"grad_norm": 0.04801041632890701,
"learning_rate": 0.00015456188817955963,
"loss": 1.884882926940918,
"step": 2710
},
{
"epoch": 0.23652173913043478,
"grad_norm": 0.04432157799601555,
"learning_rate": 0.00015440537093501077,
"loss": 1.9631668090820313,
"step": 2720
},
{
"epoch": 0.2373913043478261,
"grad_norm": 0.045344945043325424,
"learning_rate": 0.00015424833655590313,
"loss": 1.9439357757568358,
"step": 2730
},
{
"epoch": 0.2382608695652174,
"grad_norm": 0.046976640820503235,
"learning_rate": 0.00015409078625600926,
"loss": 1.8975433349609374,
"step": 2740
},
{
"epoch": 0.2391304347826087,
"grad_norm": 0.04438307136297226,
"learning_rate": 0.0001539327212530895,
"loss": 1.870760154724121,
"step": 2750
},
{
"epoch": 0.24,
"grad_norm": 0.04586169123649597,
"learning_rate": 0.00015377414276888246,
"loss": 1.8693674087524415,
"step": 2760
},
{
"epoch": 0.2408695652173913,
"grad_norm": 0.04416853189468384,
"learning_rate": 0.00015361505202909564,
"loss": 1.8614776611328125,
"step": 2770
},
{
"epoch": 0.2417391304347826,
"grad_norm": 0.04704264551401138,
"learning_rate": 0.0001534554502633959,
"loss": 1.9111223220825195,
"step": 2780
},
{
"epoch": 0.2426086956521739,
"grad_norm": 0.04835261031985283,
"learning_rate": 0.00015329533870540003,
"loss": 1.9522029876708984,
"step": 2790
},
{
"epoch": 0.24347826086956523,
"grad_norm": 0.048007503151893616,
"learning_rate": 0.00015313471859266522,
"loss": 1.9032674789428712,
"step": 2800
},
{
"epoch": 0.24434782608695652,
"grad_norm": 0.04623431712388992,
"learning_rate": 0.00015297359116667934,
"loss": 1.911952018737793,
"step": 2810
},
{
"epoch": 0.24521739130434783,
"grad_norm": 0.045389365404844284,
"learning_rate": 0.00015281195767285162,
"loss": 1.945793914794922,
"step": 2820
},
{
"epoch": 0.24608695652173912,
"grad_norm": 0.046615757048130035,
"learning_rate": 0.00015264981936050269,
"loss": 1.8489097595214843,
"step": 2830
},
{
"epoch": 0.24695652173913044,
"grad_norm": 0.04565439373254776,
"learning_rate": 0.00015248717748285521,
"loss": 1.7807727813720704,
"step": 2840
},
{
"epoch": 0.24782608695652175,
"grad_norm": 0.04634851589798927,
"learning_rate": 0.00015232403329702406,
"loss": 1.8626737594604492,
"step": 2850
},
{
"epoch": 0.24869565217391304,
"grad_norm": 0.043341245502233505,
"learning_rate": 0.00015216038806400655,
"loss": 1.885053825378418,
"step": 2860
},
{
"epoch": 0.24956521739130436,
"grad_norm": 0.044841159135103226,
"learning_rate": 0.00015199624304867285,
"loss": 1.877364158630371,
"step": 2870
},
{
"epoch": 0.25043478260869567,
"grad_norm": 0.045520294457674026,
"learning_rate": 0.00015183159951975605,
"loss": 1.896337127685547,
"step": 2880
},
{
"epoch": 0.25130434782608696,
"grad_norm": 0.04885012283921242,
"learning_rate": 0.00015166645874984248,
"loss": 1.9359861373901368,
"step": 2890
},
{
"epoch": 0.25217391304347825,
"grad_norm": 0.04632116109132767,
"learning_rate": 0.0001515008220153618,
"loss": 1.8857734680175782,
"step": 2900
},
{
"epoch": 0.2530434782608696,
"grad_norm": 0.047859784215688705,
"learning_rate": 0.0001513346905965771,
"loss": 1.8554449081420898,
"step": 2910
},
{
"epoch": 0.2539130434782609,
"grad_norm": 0.046239182353019714,
"learning_rate": 0.00015116806577757512,
"loss": 1.7961738586425782,
"step": 2920
},
{
"epoch": 0.25478260869565217,
"grad_norm": 0.04520752280950546,
"learning_rate": 0.00015100094884625624,
"loss": 1.8724737167358398,
"step": 2930
},
{
"epoch": 0.25565217391304346,
"grad_norm": 0.046254873275756836,
"learning_rate": 0.00015083334109432454,
"loss": 1.8898199081420899,
"step": 2940
},
{
"epoch": 0.2565217391304348,
"grad_norm": 0.04747215658426285,
"learning_rate": 0.00015066524381727778,
"loss": 1.8878746032714844,
"step": 2950
},
{
"epoch": 0.2573913043478261,
"grad_norm": 0.048975683748722076,
"learning_rate": 0.00015049665831439753,
"loss": 1.8283885955810546,
"step": 2960
},
{
"epoch": 0.2582608695652174,
"grad_norm": 0.04566105827689171,
"learning_rate": 0.0001503275858887389,
"loss": 1.840024185180664,
"step": 2970
},
{
"epoch": 0.2591304347826087,
"grad_norm": 0.046260230243206024,
"learning_rate": 0.0001501580278471207,
"loss": 1.867376708984375,
"step": 2980
},
{
"epoch": 0.26,
"grad_norm": 0.046284548938274384,
"learning_rate": 0.00014998798550011516,
"loss": 1.886456298828125,
"step": 2990
},
{
"epoch": 0.2608695652173913,
"grad_norm": 0.046349357813596725,
"learning_rate": 0.00014981746016203793,
"loss": 1.8599205017089844,
"step": 3000
},
{
"epoch": 0.2617391304347826,
"grad_norm": 0.046956196427345276,
"learning_rate": 0.0001496464531509378,
"loss": 1.8856639862060547,
"step": 3010
},
{
"epoch": 0.26260869565217393,
"grad_norm": 0.045312799513339996,
"learning_rate": 0.00014947496578858666,
"loss": 1.8629747390747071,
"step": 3020
},
{
"epoch": 0.2634782608695652,
"grad_norm": 0.04610862210392952,
"learning_rate": 0.00014930299940046907,
"loss": 1.8872554779052735,
"step": 3030
},
{
"epoch": 0.2643478260869565,
"grad_norm": 0.04584391042590141,
"learning_rate": 0.0001491305553157723,
"loss": 1.883194351196289,
"step": 3040
},
{
"epoch": 0.26521739130434785,
"grad_norm": 0.04532180353999138,
"learning_rate": 0.00014895763486737578,
"loss": 1.8846633911132813,
"step": 3050
},
{
"epoch": 0.26608695652173914,
"grad_norm": 0.04704463481903076,
"learning_rate": 0.000148784239391841,
"loss": 1.8317132949829102,
"step": 3060
},
{
"epoch": 0.2669565217391304,
"grad_norm": 0.04611065611243248,
"learning_rate": 0.000148610370229401,
"loss": 1.9712543487548828,
"step": 3070
},
{
"epoch": 0.2678260869565217,
"grad_norm": 0.04814252257347107,
"learning_rate": 0.00014843602872395025,
"loss": 1.94564208984375,
"step": 3080
},
{
"epoch": 0.26869565217391306,
"grad_norm": 0.046260274946689606,
"learning_rate": 0.00014826121622303398,
"loss": 1.8456733703613282,
"step": 3090
},
{
"epoch": 0.26956521739130435,
"grad_norm": 0.04497453570365906,
"learning_rate": 0.00014808593407783795,
"loss": 1.8342208862304688,
"step": 3100
},
{
"epoch": 0.27043478260869563,
"grad_norm": 0.045834191143512726,
"learning_rate": 0.00014791018364317807,
"loss": 1.8588268280029296,
"step": 3110
},
{
"epoch": 0.271304347826087,
"grad_norm": 0.04720929265022278,
"learning_rate": 0.00014773396627748963,
"loss": 1.8708805084228515,
"step": 3120
},
{
"epoch": 0.27217391304347827,
"grad_norm": 0.044134072959423065,
"learning_rate": 0.00014755728334281713,
"loss": 1.8183509826660156,
"step": 3130
},
{
"epoch": 0.27304347826086955,
"grad_norm": 0.04427867382764816,
"learning_rate": 0.00014738013620480355,
"loss": 1.9084312438964843,
"step": 3140
},
{
"epoch": 0.27391304347826084,
"grad_norm": 0.04593170806765556,
"learning_rate": 0.0001472025262326799,
"loss": 1.9480342864990234,
"step": 3150
},
{
"epoch": 0.2747826086956522,
"grad_norm": 0.04596893861889839,
"learning_rate": 0.0001470244547992546,
"loss": 1.8195436477661133,
"step": 3160
},
{
"epoch": 0.2756521739130435,
"grad_norm": 0.04655667021870613,
"learning_rate": 0.00014684592328090276,
"loss": 1.9056392669677735,
"step": 3170
},
{
"epoch": 0.27652173913043476,
"grad_norm": 0.045723024755716324,
"learning_rate": 0.00014666693305755575,
"loss": 1.8669166564941406,
"step": 3180
},
{
"epoch": 0.2773913043478261,
"grad_norm": 0.04697049781680107,
"learning_rate": 0.00014648748551269043,
"loss": 1.8303703308105468,
"step": 3190
},
{
"epoch": 0.2782608695652174,
"grad_norm": 0.04601813852787018,
"learning_rate": 0.00014630758203331836,
"loss": 1.8695808410644532,
"step": 3200
},
{
"epoch": 0.2791304347826087,
"grad_norm": 0.04566989839076996,
"learning_rate": 0.00014612722400997525,
"loss": 1.8917108535766602,
"step": 3210
},
{
"epoch": 0.28,
"grad_norm": 0.04773734137415886,
"learning_rate": 0.00014594641283671017,
"loss": 1.895745086669922,
"step": 3220
},
{
"epoch": 0.2808695652173913,
"grad_norm": 0.04546461999416351,
"learning_rate": 0.0001457651499110746,
"loss": 1.8925960540771485,
"step": 3230
},
{
"epoch": 0.2817391304347826,
"grad_norm": 0.04565638676285744,
"learning_rate": 0.00014558343663411197,
"loss": 1.8597404479980468,
"step": 3240
},
{
"epoch": 0.2826086956521739,
"grad_norm": 0.045438554137945175,
"learning_rate": 0.00014540127441034644,
"loss": 1.8530876159667968,
"step": 3250
},
{
"epoch": 0.28347826086956524,
"grad_norm": 0.0473264716565609,
"learning_rate": 0.00014521866464777233,
"loss": 1.8691961288452148,
"step": 3260
},
{
"epoch": 0.2843478260869565,
"grad_norm": 0.04689843952655792,
"learning_rate": 0.00014503560875784316,
"loss": 1.8995010375976562,
"step": 3270
},
{
"epoch": 0.2852173913043478,
"grad_norm": 0.04655284807085991,
"learning_rate": 0.00014485210815546066,
"loss": 1.8623058319091796,
"step": 3280
},
{
"epoch": 0.28608695652173916,
"grad_norm": 0.04491620883345604,
"learning_rate": 0.0001446681642589639,
"loss": 1.912955093383789,
"step": 3290
},
{
"epoch": 0.28695652173913044,
"grad_norm": 0.04738801345229149,
"learning_rate": 0.0001444837784901184,
"loss": 1.8688699722290039,
"step": 3300
},
{
"epoch": 0.28782608695652173,
"grad_norm": 0.04594497010111809,
"learning_rate": 0.00014429895227410493,
"loss": 1.9313240051269531,
"step": 3310
},
{
"epoch": 0.288695652173913,
"grad_norm": 0.047535549849271774,
"learning_rate": 0.0001441136870395088,
"loss": 1.9088796615600585,
"step": 3320
},
{
"epoch": 0.28956521739130436,
"grad_norm": 0.04802543297410011,
"learning_rate": 0.00014392798421830846,
"loss": 1.8703914642333985,
"step": 3330
},
{
"epoch": 0.29043478260869565,
"grad_norm": 0.046663422137498856,
"learning_rate": 0.00014374184524586477,
"loss": 1.8450454711914062,
"step": 3340
},
{
"epoch": 0.29130434782608694,
"grad_norm": 0.046551477164030075,
"learning_rate": 0.00014355527156090965,
"loss": 1.9085540771484375,
"step": 3350
},
{
"epoch": 0.2921739130434783,
"grad_norm": 0.045060548931360245,
"learning_rate": 0.00014336826460553513,
"loss": 1.8219585418701172,
"step": 3360
},
{
"epoch": 0.29304347826086957,
"grad_norm": 0.04710334166884422,
"learning_rate": 0.00014318082582518208,
"loss": 1.9055330276489257,
"step": 3370
},
{
"epoch": 0.29391304347826086,
"grad_norm": 0.045783352106809616,
"learning_rate": 0.00014299295666862912,
"loss": 1.8600608825683593,
"step": 3380
},
{
"epoch": 0.29478260869565215,
"grad_norm": 0.04774748533964157,
"learning_rate": 0.00014280465858798144,
"loss": 1.8205097198486329,
"step": 3390
},
{
"epoch": 0.2956521739130435,
"grad_norm": 0.048090871423482895,
"learning_rate": 0.00014261593303865943,
"loss": 1.8969783782958984,
"step": 3400
}
],
"logging_steps": 10,
"max_steps": 11500,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"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": 3.019142697320448e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}