Mxode's picture
upload
f6519b4
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 9480,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0010548523206751054,
"grad_norm": 1.2186888456344604,
"learning_rate": 0.00015822784810126583,
"loss": 7.4756,
"step": 10
},
{
"epoch": 0.002109704641350211,
"grad_norm": 1.1075139045715332,
"learning_rate": 0.00031645569620253165,
"loss": 6.7987,
"step": 20
},
{
"epoch": 0.0031645569620253164,
"grad_norm": 0.8604756593704224,
"learning_rate": 0.00047468354430379745,
"loss": 6.1951,
"step": 30
},
{
"epoch": 0.004219409282700422,
"grad_norm": 0.6423186659812927,
"learning_rate": 0.0006329113924050633,
"loss": 5.7039,
"step": 40
},
{
"epoch": 0.005274261603375527,
"grad_norm": 0.6575109362602234,
"learning_rate": 0.0007911392405063291,
"loss": 5.2429,
"step": 50
},
{
"epoch": 0.006329113924050633,
"grad_norm": 0.7181506156921387,
"learning_rate": 0.0009493670886075949,
"loss": 4.7705,
"step": 60
},
{
"epoch": 0.007383966244725738,
"grad_norm": 0.9496937990188599,
"learning_rate": 0.0011075949367088608,
"loss": 4.3984,
"step": 70
},
{
"epoch": 0.008438818565400843,
"grad_norm": 1.0826588869094849,
"learning_rate": 0.0012658227848101266,
"loss": 4.1411,
"step": 80
},
{
"epoch": 0.00949367088607595,
"grad_norm": 0.7389441728591919,
"learning_rate": 0.0014240506329113926,
"loss": 3.9336,
"step": 90
},
{
"epoch": 0.010548523206751054,
"grad_norm": 1.0077191591262817,
"learning_rate": 0.0015,
"loss": 3.7884,
"step": 100
},
{
"epoch": 0.011603375527426161,
"grad_norm": 0.7541709542274475,
"learning_rate": 0.0015,
"loss": 3.6578,
"step": 110
},
{
"epoch": 0.012658227848101266,
"grad_norm": 0.7902190685272217,
"learning_rate": 0.0015,
"loss": 3.5503,
"step": 120
},
{
"epoch": 0.013713080168776372,
"grad_norm": 0.7221662998199463,
"learning_rate": 0.0015,
"loss": 3.4618,
"step": 130
},
{
"epoch": 0.014767932489451477,
"grad_norm": 1.2441911697387695,
"learning_rate": 0.0015,
"loss": 3.3819,
"step": 140
},
{
"epoch": 0.015822784810126583,
"grad_norm": 0.9825108647346497,
"learning_rate": 0.0015,
"loss": 3.3147,
"step": 150
},
{
"epoch": 0.016877637130801686,
"grad_norm": 0.9007912874221802,
"learning_rate": 0.0015,
"loss": 3.2649,
"step": 160
},
{
"epoch": 0.017932489451476793,
"grad_norm": 0.86688232421875,
"learning_rate": 0.0015,
"loss": 3.1977,
"step": 170
},
{
"epoch": 0.0189873417721519,
"grad_norm": 0.8259088397026062,
"learning_rate": 0.0015,
"loss": 3.1623,
"step": 180
},
{
"epoch": 0.020042194092827006,
"grad_norm": 1.230103850364685,
"learning_rate": 0.0015,
"loss": 3.1005,
"step": 190
},
{
"epoch": 0.02109704641350211,
"grad_norm": 0.9208290576934814,
"learning_rate": 0.0015,
"loss": 3.0592,
"step": 200
},
{
"epoch": 0.022151898734177215,
"grad_norm": 0.7172234654426575,
"learning_rate": 0.0015,
"loss": 3.0197,
"step": 210
},
{
"epoch": 0.023206751054852322,
"grad_norm": 0.691364049911499,
"learning_rate": 0.0015,
"loss": 2.9838,
"step": 220
},
{
"epoch": 0.024261603375527425,
"grad_norm": 0.882768452167511,
"learning_rate": 0.0015,
"loss": 2.9276,
"step": 230
},
{
"epoch": 0.02531645569620253,
"grad_norm": 1.0899426937103271,
"learning_rate": 0.0015,
"loss": 2.898,
"step": 240
},
{
"epoch": 0.026371308016877638,
"grad_norm": 0.835606038570404,
"learning_rate": 0.0015,
"loss": 2.8611,
"step": 250
},
{
"epoch": 0.027426160337552744,
"grad_norm": 0.9906381964683533,
"learning_rate": 0.0015,
"loss": 2.8228,
"step": 260
},
{
"epoch": 0.028481012658227847,
"grad_norm": 0.837228000164032,
"learning_rate": 0.0015,
"loss": 2.7884,
"step": 270
},
{
"epoch": 0.029535864978902954,
"grad_norm": 0.9279177784919739,
"learning_rate": 0.0015,
"loss": 2.7519,
"step": 280
},
{
"epoch": 0.03059071729957806,
"grad_norm": 1.240341305732727,
"learning_rate": 0.0015,
"loss": 2.7221,
"step": 290
},
{
"epoch": 0.03164556962025317,
"grad_norm": 0.9779770374298096,
"learning_rate": 0.0015,
"loss": 2.6969,
"step": 300
},
{
"epoch": 0.03270042194092827,
"grad_norm": 1.3490874767303467,
"learning_rate": 0.0015,
"loss": 2.6669,
"step": 310
},
{
"epoch": 0.03375527426160337,
"grad_norm": 1.3380482196807861,
"learning_rate": 0.0015,
"loss": 2.6312,
"step": 320
},
{
"epoch": 0.03481012658227848,
"grad_norm": 0.8288021087646484,
"learning_rate": 0.0015,
"loss": 2.6254,
"step": 330
},
{
"epoch": 0.035864978902953586,
"grad_norm": 1.363772988319397,
"learning_rate": 0.0015,
"loss": 2.5977,
"step": 340
},
{
"epoch": 0.03691983122362869,
"grad_norm": 1.2266892194747925,
"learning_rate": 0.0015,
"loss": 2.5532,
"step": 350
},
{
"epoch": 0.0379746835443038,
"grad_norm": 0.9206980466842651,
"learning_rate": 0.0015,
"loss": 2.5316,
"step": 360
},
{
"epoch": 0.039029535864978905,
"grad_norm": 1.0549575090408325,
"learning_rate": 0.0015,
"loss": 2.5158,
"step": 370
},
{
"epoch": 0.04008438818565401,
"grad_norm": 1.217510461807251,
"learning_rate": 0.0015,
"loss": 2.4978,
"step": 380
},
{
"epoch": 0.04113924050632911,
"grad_norm": 1.0540152788162231,
"learning_rate": 0.0015,
"loss": 2.4792,
"step": 390
},
{
"epoch": 0.04219409282700422,
"grad_norm": 1.0417619943618774,
"learning_rate": 0.0015,
"loss": 2.4573,
"step": 400
},
{
"epoch": 0.043248945147679324,
"grad_norm": 0.7437645196914673,
"learning_rate": 0.0015,
"loss": 2.442,
"step": 410
},
{
"epoch": 0.04430379746835443,
"grad_norm": 1.1524747610092163,
"learning_rate": 0.0015,
"loss": 2.4248,
"step": 420
},
{
"epoch": 0.04535864978902954,
"grad_norm": 1.0705971717834473,
"learning_rate": 0.0015,
"loss": 2.3973,
"step": 430
},
{
"epoch": 0.046413502109704644,
"grad_norm": 0.9775382280349731,
"learning_rate": 0.0015,
"loss": 2.3833,
"step": 440
},
{
"epoch": 0.04746835443037975,
"grad_norm": 1.025772213935852,
"learning_rate": 0.0015,
"loss": 2.3803,
"step": 450
},
{
"epoch": 0.04852320675105485,
"grad_norm": 1.246941328048706,
"learning_rate": 0.0015,
"loss": 2.3622,
"step": 460
},
{
"epoch": 0.049578059071729956,
"grad_norm": 1.0031028985977173,
"learning_rate": 0.0015,
"loss": 2.3334,
"step": 470
},
{
"epoch": 0.05063291139240506,
"grad_norm": 1.5016298294067383,
"learning_rate": 0.0015,
"loss": 2.3327,
"step": 480
},
{
"epoch": 0.05168776371308017,
"grad_norm": 0.7244430184364319,
"learning_rate": 0.0015,
"loss": 2.3008,
"step": 490
},
{
"epoch": 0.052742616033755275,
"grad_norm": 1.213437795639038,
"learning_rate": 0.0015,
"loss": 2.2994,
"step": 500
},
{
"epoch": 0.05379746835443038,
"grad_norm": 1.2152141332626343,
"learning_rate": 0.0015,
"loss": 2.2937,
"step": 510
},
{
"epoch": 0.05485232067510549,
"grad_norm": 0.9512096047401428,
"learning_rate": 0.0015,
"loss": 2.2921,
"step": 520
},
{
"epoch": 0.05590717299578059,
"grad_norm": 0.7877956628799438,
"learning_rate": 0.0015,
"loss": 2.256,
"step": 530
},
{
"epoch": 0.056962025316455694,
"grad_norm": 1.007867693901062,
"learning_rate": 0.0015,
"loss": 2.2389,
"step": 540
},
{
"epoch": 0.0580168776371308,
"grad_norm": 1.2216438055038452,
"learning_rate": 0.0015,
"loss": 2.2341,
"step": 550
},
{
"epoch": 0.05907172995780591,
"grad_norm": 0.769692599773407,
"learning_rate": 0.0015,
"loss": 2.2007,
"step": 560
},
{
"epoch": 0.060126582278481014,
"grad_norm": 1.202372431755066,
"learning_rate": 0.0015,
"loss": 2.2073,
"step": 570
},
{
"epoch": 0.06118143459915612,
"grad_norm": 1.0290395021438599,
"learning_rate": 0.0015,
"loss": 2.2108,
"step": 580
},
{
"epoch": 0.06223628691983123,
"grad_norm": 1.367793321609497,
"learning_rate": 0.0015,
"loss": 2.1828,
"step": 590
},
{
"epoch": 0.06329113924050633,
"grad_norm": 0.8654493093490601,
"learning_rate": 0.0015,
"loss": 2.1673,
"step": 600
},
{
"epoch": 0.06434599156118144,
"grad_norm": 0.8177367448806763,
"learning_rate": 0.0015,
"loss": 2.1687,
"step": 610
},
{
"epoch": 0.06540084388185655,
"grad_norm": 1.0598950386047363,
"learning_rate": 0.0015,
"loss": 2.1564,
"step": 620
},
{
"epoch": 0.06645569620253164,
"grad_norm": 1.063185691833496,
"learning_rate": 0.0015,
"loss": 2.1392,
"step": 630
},
{
"epoch": 0.06751054852320675,
"grad_norm": 0.8926445245742798,
"learning_rate": 0.0015,
"loss": 2.1467,
"step": 640
},
{
"epoch": 0.06856540084388185,
"grad_norm": 0.7596344947814941,
"learning_rate": 0.0015,
"loss": 2.1377,
"step": 650
},
{
"epoch": 0.06962025316455696,
"grad_norm": 0.8228513598442078,
"learning_rate": 0.0015,
"loss": 2.1232,
"step": 660
},
{
"epoch": 0.07067510548523206,
"grad_norm": 0.8619157075881958,
"learning_rate": 0.0015,
"loss": 2.108,
"step": 670
},
{
"epoch": 0.07172995780590717,
"grad_norm": 2.402742862701416,
"learning_rate": 0.0015,
"loss": 2.1152,
"step": 680
},
{
"epoch": 0.07278481012658228,
"grad_norm": 0.8077136278152466,
"learning_rate": 0.0015,
"loss": 2.135,
"step": 690
},
{
"epoch": 0.07383966244725738,
"grad_norm": 0.8595909476280212,
"learning_rate": 0.0015,
"loss": 2.0909,
"step": 700
},
{
"epoch": 0.07489451476793249,
"grad_norm": 0.8992576003074646,
"learning_rate": 0.0015,
"loss": 2.0899,
"step": 710
},
{
"epoch": 0.0759493670886076,
"grad_norm": 1.1452020406723022,
"learning_rate": 0.0015,
"loss": 2.0698,
"step": 720
},
{
"epoch": 0.0770042194092827,
"grad_norm": 0.9489206075668335,
"learning_rate": 0.0015,
"loss": 2.0649,
"step": 730
},
{
"epoch": 0.07805907172995781,
"grad_norm": 0.6932376027107239,
"learning_rate": 0.0015,
"loss": 2.0725,
"step": 740
},
{
"epoch": 0.07911392405063292,
"grad_norm": 0.9433043599128723,
"learning_rate": 0.0015,
"loss": 2.0514,
"step": 750
},
{
"epoch": 0.08016877637130802,
"grad_norm": 1.7378337383270264,
"learning_rate": 0.0015,
"loss": 2.0728,
"step": 760
},
{
"epoch": 0.08122362869198312,
"grad_norm": 1.3360079526901245,
"learning_rate": 0.0015,
"loss": 2.0565,
"step": 770
},
{
"epoch": 0.08227848101265822,
"grad_norm": 0.9339075088500977,
"learning_rate": 0.0015,
"loss": 2.0525,
"step": 780
},
{
"epoch": 0.08333333333333333,
"grad_norm": 0.6675604581832886,
"learning_rate": 0.0015,
"loss": 2.0403,
"step": 790
},
{
"epoch": 0.08438818565400844,
"grad_norm": 0.6963586211204529,
"learning_rate": 0.0015,
"loss": 2.0174,
"step": 800
},
{
"epoch": 0.08544303797468354,
"grad_norm": 0.756433367729187,
"learning_rate": 0.0015,
"loss": 2.0048,
"step": 810
},
{
"epoch": 0.08649789029535865,
"grad_norm": 0.7605345845222473,
"learning_rate": 0.0015,
"loss": 2.0148,
"step": 820
},
{
"epoch": 0.08755274261603375,
"grad_norm": 0.9260172843933105,
"learning_rate": 0.0015,
"loss": 2.0224,
"step": 830
},
{
"epoch": 0.08860759493670886,
"grad_norm": 1.2491339445114136,
"learning_rate": 0.0015,
"loss": 2.0016,
"step": 840
},
{
"epoch": 0.08966244725738397,
"grad_norm": 0.9123998880386353,
"learning_rate": 0.0015,
"loss": 1.9979,
"step": 850
},
{
"epoch": 0.09071729957805907,
"grad_norm": 1.066710114479065,
"learning_rate": 0.0015,
"loss": 1.9967,
"step": 860
},
{
"epoch": 0.09177215189873418,
"grad_norm": 0.6761780381202698,
"learning_rate": 0.0015,
"loss": 1.9925,
"step": 870
},
{
"epoch": 0.09282700421940929,
"grad_norm": 0.7574045062065125,
"learning_rate": 0.0015,
"loss": 1.9863,
"step": 880
},
{
"epoch": 0.0938818565400844,
"grad_norm": 0.8483901023864746,
"learning_rate": 0.0015,
"loss": 1.9792,
"step": 890
},
{
"epoch": 0.0949367088607595,
"grad_norm": 0.9919297099113464,
"learning_rate": 0.0015,
"loss": 1.9766,
"step": 900
},
{
"epoch": 0.09599156118143459,
"grad_norm": 0.9863075613975525,
"learning_rate": 0.0015,
"loss": 1.985,
"step": 910
},
{
"epoch": 0.0970464135021097,
"grad_norm": 1.0680369138717651,
"learning_rate": 0.0015,
"loss": 1.9688,
"step": 920
},
{
"epoch": 0.0981012658227848,
"grad_norm": 0.7299118041992188,
"learning_rate": 0.0015,
"loss": 1.9583,
"step": 930
},
{
"epoch": 0.09915611814345991,
"grad_norm": 0.7109554409980774,
"learning_rate": 0.0015,
"loss": 1.9685,
"step": 940
},
{
"epoch": 0.10021097046413502,
"grad_norm": 0.6925338506698608,
"learning_rate": 0.0015,
"loss": 1.9396,
"step": 950
},
{
"epoch": 0.10126582278481013,
"grad_norm": 0.989055335521698,
"learning_rate": 0.0015,
"loss": 1.9474,
"step": 960
},
{
"epoch": 0.10232067510548523,
"grad_norm": 0.7260894179344177,
"learning_rate": 0.0015,
"loss": 1.9499,
"step": 970
},
{
"epoch": 0.10337552742616034,
"grad_norm": 1.1013668775558472,
"learning_rate": 0.0015,
"loss": 1.9332,
"step": 980
},
{
"epoch": 0.10443037974683544,
"grad_norm": 0.6925954818725586,
"learning_rate": 0.0015,
"loss": 1.9296,
"step": 990
},
{
"epoch": 0.10548523206751055,
"grad_norm": 1.757063865661621,
"learning_rate": 0.0015,
"loss": 1.9455,
"step": 1000
},
{
"epoch": 0.10654008438818566,
"grad_norm": 1.0348504781723022,
"learning_rate": 0.0015,
"loss": 1.9542,
"step": 1010
},
{
"epoch": 0.10759493670886076,
"grad_norm": 0.7209298610687256,
"learning_rate": 0.0015,
"loss": 1.919,
"step": 1020
},
{
"epoch": 0.10864978902953587,
"grad_norm": 0.7077058553695679,
"learning_rate": 0.0015,
"loss": 1.9186,
"step": 1030
},
{
"epoch": 0.10970464135021098,
"grad_norm": 0.6635832786560059,
"learning_rate": 0.0015,
"loss": 1.9155,
"step": 1040
},
{
"epoch": 0.11075949367088607,
"grad_norm": 0.8985444903373718,
"learning_rate": 0.0015,
"loss": 1.918,
"step": 1050
},
{
"epoch": 0.11181434599156118,
"grad_norm": 0.7381898164749146,
"learning_rate": 0.0015,
"loss": 1.9115,
"step": 1060
},
{
"epoch": 0.11286919831223628,
"grad_norm": 0.7176501750946045,
"learning_rate": 0.0015,
"loss": 1.9009,
"step": 1070
},
{
"epoch": 0.11392405063291139,
"grad_norm": 0.6653502583503723,
"learning_rate": 0.0015,
"loss": 1.8975,
"step": 1080
},
{
"epoch": 0.1149789029535865,
"grad_norm": 0.6939038634300232,
"learning_rate": 0.0015,
"loss": 1.9065,
"step": 1090
},
{
"epoch": 0.1160337552742616,
"grad_norm": 0.819861888885498,
"learning_rate": 0.0015,
"loss": 1.8948,
"step": 1100
},
{
"epoch": 0.11708860759493671,
"grad_norm": 0.7226906418800354,
"learning_rate": 0.0015,
"loss": 1.8988,
"step": 1110
},
{
"epoch": 0.11814345991561181,
"grad_norm": 0.9487333297729492,
"learning_rate": 0.0015,
"loss": 1.8923,
"step": 1120
},
{
"epoch": 0.11919831223628692,
"grad_norm": 0.8971943855285645,
"learning_rate": 0.0015,
"loss": 1.8775,
"step": 1130
},
{
"epoch": 0.12025316455696203,
"grad_norm": 0.7244747281074524,
"learning_rate": 0.0015,
"loss": 1.8967,
"step": 1140
},
{
"epoch": 0.12130801687763713,
"grad_norm": 0.753216028213501,
"learning_rate": 0.0015,
"loss": 1.8789,
"step": 1150
},
{
"epoch": 0.12236286919831224,
"grad_norm": 0.7540823221206665,
"learning_rate": 0.0015,
"loss": 1.8828,
"step": 1160
},
{
"epoch": 0.12341772151898735,
"grad_norm": 0.7866358757019043,
"learning_rate": 0.0015,
"loss": 1.8744,
"step": 1170
},
{
"epoch": 0.12447257383966245,
"grad_norm": 0.752935528755188,
"learning_rate": 0.0015,
"loss": 1.8599,
"step": 1180
},
{
"epoch": 0.12552742616033755,
"grad_norm": 0.6611055731773376,
"learning_rate": 0.0015,
"loss": 1.8637,
"step": 1190
},
{
"epoch": 0.12658227848101267,
"grad_norm": 1.2516565322875977,
"learning_rate": 0.0015,
"loss": 1.8793,
"step": 1200
},
{
"epoch": 0.12763713080168776,
"grad_norm": 0.6632084846496582,
"learning_rate": 0.0015,
"loss": 1.8685,
"step": 1210
},
{
"epoch": 0.12869198312236288,
"grad_norm": 0.6862037181854248,
"learning_rate": 0.0015,
"loss": 1.8506,
"step": 1220
},
{
"epoch": 0.12974683544303797,
"grad_norm": 0.8717275261878967,
"learning_rate": 0.0015,
"loss": 1.8432,
"step": 1230
},
{
"epoch": 0.1308016877637131,
"grad_norm": 1.1524758338928223,
"learning_rate": 0.0015,
"loss": 1.866,
"step": 1240
},
{
"epoch": 0.13185654008438819,
"grad_norm": 0.9125797748565674,
"learning_rate": 0.0015,
"loss": 1.8598,
"step": 1250
},
{
"epoch": 0.13291139240506328,
"grad_norm": 0.8476076722145081,
"learning_rate": 0.0015,
"loss": 1.8439,
"step": 1260
},
{
"epoch": 0.1339662447257384,
"grad_norm": 0.9498160481452942,
"learning_rate": 0.0015,
"loss": 1.8459,
"step": 1270
},
{
"epoch": 0.1350210970464135,
"grad_norm": 0.7211402654647827,
"learning_rate": 0.0015,
"loss": 1.8615,
"step": 1280
},
{
"epoch": 0.1360759493670886,
"grad_norm": 0.6844839453697205,
"learning_rate": 0.0015,
"loss": 1.8446,
"step": 1290
},
{
"epoch": 0.1371308016877637,
"grad_norm": 0.8232640027999878,
"learning_rate": 0.0015,
"loss": 1.832,
"step": 1300
},
{
"epoch": 0.13818565400843882,
"grad_norm": 0.6936826705932617,
"learning_rate": 0.0015,
"loss": 1.8344,
"step": 1310
},
{
"epoch": 0.13924050632911392,
"grad_norm": 0.9277529716491699,
"learning_rate": 0.0015,
"loss": 1.8353,
"step": 1320
},
{
"epoch": 0.14029535864978904,
"grad_norm": 0.7764731049537659,
"learning_rate": 0.0015,
"loss": 1.831,
"step": 1330
},
{
"epoch": 0.14135021097046413,
"grad_norm": 0.9051993489265442,
"learning_rate": 0.0015,
"loss": 1.8301,
"step": 1340
},
{
"epoch": 0.14240506329113925,
"grad_norm": 0.6625475883483887,
"learning_rate": 0.0015,
"loss": 1.8343,
"step": 1350
},
{
"epoch": 0.14345991561181434,
"grad_norm": 0.6393569111824036,
"learning_rate": 0.0015,
"loss": 1.8297,
"step": 1360
},
{
"epoch": 0.14451476793248946,
"grad_norm": 0.6683023571968079,
"learning_rate": 0.0015,
"loss": 1.8235,
"step": 1370
},
{
"epoch": 0.14556962025316456,
"grad_norm": 0.7225092649459839,
"learning_rate": 0.0015,
"loss": 1.8168,
"step": 1380
},
{
"epoch": 0.14662447257383968,
"grad_norm": 1.3273886442184448,
"learning_rate": 0.0015,
"loss": 1.8285,
"step": 1390
},
{
"epoch": 0.14767932489451477,
"grad_norm": 1.3077683448791504,
"learning_rate": 0.0015,
"loss": 1.8139,
"step": 1400
},
{
"epoch": 0.14873417721518986,
"grad_norm": 0.6626515984535217,
"learning_rate": 0.0015,
"loss": 1.812,
"step": 1410
},
{
"epoch": 0.14978902953586498,
"grad_norm": 0.7292690277099609,
"learning_rate": 0.0015,
"loss": 1.8165,
"step": 1420
},
{
"epoch": 0.15084388185654007,
"grad_norm": 0.7270740270614624,
"learning_rate": 0.0015,
"loss": 1.8093,
"step": 1430
},
{
"epoch": 0.1518987341772152,
"grad_norm": 0.9681549072265625,
"learning_rate": 0.0015,
"loss": 1.8034,
"step": 1440
},
{
"epoch": 0.1529535864978903,
"grad_norm": 0.6339258551597595,
"learning_rate": 0.0015,
"loss": 1.805,
"step": 1450
},
{
"epoch": 0.1540084388185654,
"grad_norm": 0.6470240950584412,
"learning_rate": 0.0015,
"loss": 1.8001,
"step": 1460
},
{
"epoch": 0.1550632911392405,
"grad_norm": 0.8807370662689209,
"learning_rate": 0.0015,
"loss": 1.8002,
"step": 1470
},
{
"epoch": 0.15611814345991562,
"grad_norm": 0.7718443274497986,
"learning_rate": 0.0015,
"loss": 1.8049,
"step": 1480
},
{
"epoch": 0.1571729957805907,
"grad_norm": 0.8173696994781494,
"learning_rate": 0.0015,
"loss": 1.7981,
"step": 1490
},
{
"epoch": 0.15822784810126583,
"grad_norm": 0.6859337687492371,
"learning_rate": 0.0015,
"loss": 1.7994,
"step": 1500
},
{
"epoch": 0.15928270042194093,
"grad_norm": 0.7179945707321167,
"learning_rate": 0.0015,
"loss": 1.7937,
"step": 1510
},
{
"epoch": 0.16033755274261605,
"grad_norm": 0.6422714591026306,
"learning_rate": 0.0015,
"loss": 1.7976,
"step": 1520
},
{
"epoch": 0.16139240506329114,
"grad_norm": 0.8165238499641418,
"learning_rate": 0.0015,
"loss": 1.7832,
"step": 1530
},
{
"epoch": 0.16244725738396623,
"grad_norm": 0.6381688714027405,
"learning_rate": 0.0015,
"loss": 1.7716,
"step": 1540
},
{
"epoch": 0.16350210970464135,
"grad_norm": 0.6977247595787048,
"learning_rate": 0.0015,
"loss": 1.7763,
"step": 1550
},
{
"epoch": 0.16455696202531644,
"grad_norm": 0.6377383470535278,
"learning_rate": 0.0015,
"loss": 1.7759,
"step": 1560
},
{
"epoch": 0.16561181434599156,
"grad_norm": 0.9726520776748657,
"learning_rate": 0.0015,
"loss": 1.7675,
"step": 1570
},
{
"epoch": 0.16666666666666666,
"grad_norm": 0.6959405541419983,
"learning_rate": 0.0015,
"loss": 1.7814,
"step": 1580
},
{
"epoch": 0.16772151898734178,
"grad_norm": 0.8633522987365723,
"learning_rate": 0.0015,
"loss": 1.7783,
"step": 1590
},
{
"epoch": 0.16877637130801687,
"grad_norm": 0.7004960179328918,
"learning_rate": 0.0015,
"loss": 1.7804,
"step": 1600
},
{
"epoch": 0.169831223628692,
"grad_norm": 0.7215518951416016,
"learning_rate": 0.0015,
"loss": 1.7773,
"step": 1610
},
{
"epoch": 0.17088607594936708,
"grad_norm": 0.6039355397224426,
"learning_rate": 0.0015,
"loss": 1.7504,
"step": 1620
},
{
"epoch": 0.1719409282700422,
"grad_norm": 0.9246911406517029,
"learning_rate": 0.0015,
"loss": 1.7648,
"step": 1630
},
{
"epoch": 0.1729957805907173,
"grad_norm": 0.8042846918106079,
"learning_rate": 0.0015,
"loss": 1.7651,
"step": 1640
},
{
"epoch": 0.17405063291139242,
"grad_norm": 0.6565026044845581,
"learning_rate": 0.0015,
"loss": 1.7564,
"step": 1650
},
{
"epoch": 0.1751054852320675,
"grad_norm": 0.7790629267692566,
"learning_rate": 0.0015,
"loss": 1.7574,
"step": 1660
},
{
"epoch": 0.17616033755274263,
"grad_norm": 0.6511490345001221,
"learning_rate": 0.0015,
"loss": 1.7633,
"step": 1670
},
{
"epoch": 0.17721518987341772,
"grad_norm": 0.6874651312828064,
"learning_rate": 0.0015,
"loss": 1.7522,
"step": 1680
},
{
"epoch": 0.17827004219409281,
"grad_norm": 0.6518383622169495,
"learning_rate": 0.0015,
"loss": 1.7446,
"step": 1690
},
{
"epoch": 0.17932489451476794,
"grad_norm": 0.9337310194969177,
"learning_rate": 0.0015,
"loss": 1.7634,
"step": 1700
},
{
"epoch": 0.18037974683544303,
"grad_norm": 1.0004547834396362,
"learning_rate": 0.0015,
"loss": 1.7414,
"step": 1710
},
{
"epoch": 0.18143459915611815,
"grad_norm": 0.9592145681381226,
"learning_rate": 0.0015,
"loss": 1.7396,
"step": 1720
},
{
"epoch": 0.18248945147679324,
"grad_norm": 0.8522003889083862,
"learning_rate": 0.0015,
"loss": 1.7394,
"step": 1730
},
{
"epoch": 0.18354430379746836,
"grad_norm": 0.6941075921058655,
"learning_rate": 0.0015,
"loss": 1.7461,
"step": 1740
},
{
"epoch": 0.18459915611814345,
"grad_norm": 0.6232513785362244,
"learning_rate": 0.0015,
"loss": 1.7665,
"step": 1750
},
{
"epoch": 0.18565400843881857,
"grad_norm": 1.227797269821167,
"learning_rate": 0.0015,
"loss": 1.7449,
"step": 1760
},
{
"epoch": 0.18670886075949367,
"grad_norm": 0.689431369304657,
"learning_rate": 0.0015,
"loss": 1.7496,
"step": 1770
},
{
"epoch": 0.1877637130801688,
"grad_norm": 0.7479903101921082,
"learning_rate": 0.0015,
"loss": 1.7448,
"step": 1780
},
{
"epoch": 0.18881856540084388,
"grad_norm": 0.6916646361351013,
"learning_rate": 0.0015,
"loss": 1.7481,
"step": 1790
},
{
"epoch": 0.189873417721519,
"grad_norm": 0.8803281188011169,
"learning_rate": 0.0015,
"loss": 1.7404,
"step": 1800
},
{
"epoch": 0.1909282700421941,
"grad_norm": 0.8555834889411926,
"learning_rate": 0.0015,
"loss": 1.7316,
"step": 1810
},
{
"epoch": 0.19198312236286919,
"grad_norm": 0.6191942691802979,
"learning_rate": 0.0015,
"loss": 1.7284,
"step": 1820
},
{
"epoch": 0.1930379746835443,
"grad_norm": 0.5880500078201294,
"learning_rate": 0.0015,
"loss": 1.7269,
"step": 1830
},
{
"epoch": 0.1940928270042194,
"grad_norm": 0.6119520664215088,
"learning_rate": 0.0015,
"loss": 1.7287,
"step": 1840
},
{
"epoch": 0.19514767932489452,
"grad_norm": 0.8929871916770935,
"learning_rate": 0.0015,
"loss": 1.731,
"step": 1850
},
{
"epoch": 0.1962025316455696,
"grad_norm": 0.8229356408119202,
"learning_rate": 0.0015,
"loss": 1.7355,
"step": 1860
},
{
"epoch": 0.19725738396624473,
"grad_norm": 0.9643657803535461,
"learning_rate": 0.0015,
"loss": 1.7365,
"step": 1870
},
{
"epoch": 0.19831223628691982,
"grad_norm": 0.6754857301712036,
"learning_rate": 0.0015,
"loss": 1.7198,
"step": 1880
},
{
"epoch": 0.19936708860759494,
"grad_norm": 0.7569988369941711,
"learning_rate": 0.0015,
"loss": 1.7248,
"step": 1890
},
{
"epoch": 0.20042194092827004,
"grad_norm": 0.7129053473472595,
"learning_rate": 0.0015,
"loss": 1.7223,
"step": 1900
},
{
"epoch": 0.20147679324894516,
"grad_norm": 1.1085187196731567,
"learning_rate": 0.0015,
"loss": 1.726,
"step": 1910
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.8156635761260986,
"learning_rate": 0.0015,
"loss": 1.7233,
"step": 1920
},
{
"epoch": 0.20358649789029537,
"grad_norm": 0.6475657224655151,
"learning_rate": 0.0015,
"loss": 1.7241,
"step": 1930
},
{
"epoch": 0.20464135021097046,
"grad_norm": 0.6763522028923035,
"learning_rate": 0.0015,
"loss": 1.7149,
"step": 1940
},
{
"epoch": 0.20569620253164558,
"grad_norm": 0.7413578629493713,
"learning_rate": 0.0015,
"loss": 1.7129,
"step": 1950
},
{
"epoch": 0.20675105485232068,
"grad_norm": 0.6201931834220886,
"learning_rate": 0.0015,
"loss": 1.7185,
"step": 1960
},
{
"epoch": 0.20780590717299577,
"grad_norm": 1.0813745260238647,
"learning_rate": 0.0015,
"loss": 1.7222,
"step": 1970
},
{
"epoch": 0.2088607594936709,
"grad_norm": 0.7272746562957764,
"learning_rate": 0.0015,
"loss": 1.7075,
"step": 1980
},
{
"epoch": 0.20991561181434598,
"grad_norm": 0.6317844986915588,
"learning_rate": 0.0015,
"loss": 1.7012,
"step": 1990
},
{
"epoch": 0.2109704641350211,
"grad_norm": 0.8212335705757141,
"learning_rate": 0.0015,
"loss": 1.7043,
"step": 2000
},
{
"epoch": 0.2120253164556962,
"grad_norm": 0.8446721434593201,
"learning_rate": 0.0015,
"loss": 1.7112,
"step": 2010
},
{
"epoch": 0.21308016877637131,
"grad_norm": 0.7635065317153931,
"learning_rate": 0.0015,
"loss": 1.7201,
"step": 2020
},
{
"epoch": 0.2141350210970464,
"grad_norm": 0.6259655952453613,
"learning_rate": 0.0015,
"loss": 1.707,
"step": 2030
},
{
"epoch": 0.21518987341772153,
"grad_norm": 0.6341089606285095,
"learning_rate": 0.0015,
"loss": 1.7077,
"step": 2040
},
{
"epoch": 0.21624472573839662,
"grad_norm": 0.6744656562805176,
"learning_rate": 0.0015,
"loss": 1.7092,
"step": 2050
},
{
"epoch": 0.21729957805907174,
"grad_norm": 0.7680793404579163,
"learning_rate": 0.0015,
"loss": 1.6997,
"step": 2060
},
{
"epoch": 0.21835443037974683,
"grad_norm": 0.855820894241333,
"learning_rate": 0.0015,
"loss": 1.6979,
"step": 2070
},
{
"epoch": 0.21940928270042195,
"grad_norm": 0.6359190940856934,
"learning_rate": 0.0015,
"loss": 1.7029,
"step": 2080
},
{
"epoch": 0.22046413502109705,
"grad_norm": 0.739136815071106,
"learning_rate": 0.0015,
"loss": 1.6946,
"step": 2090
},
{
"epoch": 0.22151898734177214,
"grad_norm": 0.6583918333053589,
"learning_rate": 0.0015,
"loss": 1.7005,
"step": 2100
},
{
"epoch": 0.22257383966244726,
"grad_norm": 0.6532536745071411,
"learning_rate": 0.0015,
"loss": 1.6911,
"step": 2110
},
{
"epoch": 0.22362869198312235,
"grad_norm": 1.0495470762252808,
"learning_rate": 0.0015,
"loss": 1.6959,
"step": 2120
},
{
"epoch": 0.22468354430379747,
"grad_norm": 0.700576901435852,
"learning_rate": 0.0015,
"loss": 1.6932,
"step": 2130
},
{
"epoch": 0.22573839662447256,
"grad_norm": 0.6586419939994812,
"learning_rate": 0.0015,
"loss": 1.684,
"step": 2140
},
{
"epoch": 0.22679324894514769,
"grad_norm": 0.6587622761726379,
"learning_rate": 0.0015,
"loss": 1.6935,
"step": 2150
},
{
"epoch": 0.22784810126582278,
"grad_norm": 0.681542694568634,
"learning_rate": 0.0015,
"loss": 1.7028,
"step": 2160
},
{
"epoch": 0.2289029535864979,
"grad_norm": 0.6412506103515625,
"learning_rate": 0.0015,
"loss": 1.6956,
"step": 2170
},
{
"epoch": 0.229957805907173,
"grad_norm": 0.7530674934387207,
"learning_rate": 0.0015,
"loss": 1.6761,
"step": 2180
},
{
"epoch": 0.2310126582278481,
"grad_norm": 0.6285898685455322,
"learning_rate": 0.0015,
"loss": 1.69,
"step": 2190
},
{
"epoch": 0.2320675105485232,
"grad_norm": 1.0234174728393555,
"learning_rate": 0.0015,
"loss": 1.6931,
"step": 2200
},
{
"epoch": 0.23312236286919832,
"grad_norm": 0.7023261189460754,
"learning_rate": 0.0015,
"loss": 1.6776,
"step": 2210
},
{
"epoch": 0.23417721518987342,
"grad_norm": 0.6849989295005798,
"learning_rate": 0.0015,
"loss": 1.6911,
"step": 2220
},
{
"epoch": 0.23523206751054854,
"grad_norm": 0.8580043911933899,
"learning_rate": 0.0015,
"loss": 1.6807,
"step": 2230
},
{
"epoch": 0.23628691983122363,
"grad_norm": 0.8025040030479431,
"learning_rate": 0.0015,
"loss": 1.682,
"step": 2240
},
{
"epoch": 0.23734177215189872,
"grad_norm": 0.6510624885559082,
"learning_rate": 0.0015,
"loss": 1.6767,
"step": 2250
},
{
"epoch": 0.23839662447257384,
"grad_norm": 0.8475005626678467,
"learning_rate": 0.0015,
"loss": 1.6703,
"step": 2260
},
{
"epoch": 0.23945147679324894,
"grad_norm": 0.5999197363853455,
"learning_rate": 0.0015,
"loss": 1.7013,
"step": 2270
},
{
"epoch": 0.24050632911392406,
"grad_norm": 0.6120131015777588,
"learning_rate": 0.0015,
"loss": 1.687,
"step": 2280
},
{
"epoch": 0.24156118143459915,
"grad_norm": 0.8520112633705139,
"learning_rate": 0.0015,
"loss": 1.67,
"step": 2290
},
{
"epoch": 0.24261603375527427,
"grad_norm": 0.6505146622657776,
"learning_rate": 0.0015,
"loss": 1.6715,
"step": 2300
},
{
"epoch": 0.24367088607594936,
"grad_norm": 0.8636083602905273,
"learning_rate": 0.0015,
"loss": 1.6736,
"step": 2310
},
{
"epoch": 0.24472573839662448,
"grad_norm": 0.6786522269248962,
"learning_rate": 0.0015,
"loss": 1.6814,
"step": 2320
},
{
"epoch": 0.24578059071729957,
"grad_norm": 0.6315372586250305,
"learning_rate": 0.0015,
"loss": 1.6611,
"step": 2330
},
{
"epoch": 0.2468354430379747,
"grad_norm": 0.6420019268989563,
"learning_rate": 0.0015,
"loss": 1.6601,
"step": 2340
},
{
"epoch": 0.2478902953586498,
"grad_norm": 0.9882775545120239,
"learning_rate": 0.0015,
"loss": 1.6627,
"step": 2350
},
{
"epoch": 0.2489451476793249,
"grad_norm": 0.7462952733039856,
"learning_rate": 0.0015,
"loss": 1.6721,
"step": 2360
},
{
"epoch": 0.25,
"grad_norm": 0.6469263434410095,
"learning_rate": 0.0015,
"loss": 1.6693,
"step": 2370
},
{
"epoch": 0.2510548523206751,
"grad_norm": 0.714177131652832,
"learning_rate": 0.0015,
"loss": 1.6692,
"step": 2380
},
{
"epoch": 0.2521097046413502,
"grad_norm": 1.074702262878418,
"learning_rate": 0.0015,
"loss": 1.6695,
"step": 2390
},
{
"epoch": 0.25316455696202533,
"grad_norm": 0.6026642918586731,
"learning_rate": 0.0015,
"loss": 1.6631,
"step": 2400
},
{
"epoch": 0.2542194092827004,
"grad_norm": 0.9359568953514099,
"learning_rate": 0.0015,
"loss": 1.6685,
"step": 2410
},
{
"epoch": 0.2552742616033755,
"grad_norm": 0.9025223851203918,
"learning_rate": 0.0015,
"loss": 1.6624,
"step": 2420
},
{
"epoch": 0.2563291139240506,
"grad_norm": 0.6915332674980164,
"learning_rate": 0.0015,
"loss": 1.665,
"step": 2430
},
{
"epoch": 0.25738396624472576,
"grad_norm": 1.0445406436920166,
"learning_rate": 0.0015,
"loss": 1.6697,
"step": 2440
},
{
"epoch": 0.25843881856540085,
"grad_norm": 0.8682367205619812,
"learning_rate": 0.0015,
"loss": 1.6582,
"step": 2450
},
{
"epoch": 0.25949367088607594,
"grad_norm": 0.6498146057128906,
"learning_rate": 0.0015,
"loss": 1.6709,
"step": 2460
},
{
"epoch": 0.26054852320675104,
"grad_norm": 0.773652970790863,
"learning_rate": 0.0015,
"loss": 1.6583,
"step": 2470
},
{
"epoch": 0.2616033755274262,
"grad_norm": 0.6476708054542542,
"learning_rate": 0.0015,
"loss": 1.6601,
"step": 2480
},
{
"epoch": 0.2626582278481013,
"grad_norm": 0.7904232740402222,
"learning_rate": 0.0015,
"loss": 1.6499,
"step": 2490
},
{
"epoch": 0.26371308016877637,
"grad_norm": 0.961468517780304,
"learning_rate": 0.0015,
"loss": 1.6616,
"step": 2500
},
{
"epoch": 0.26476793248945146,
"grad_norm": 0.6232831478118896,
"learning_rate": 0.0015,
"loss": 1.6544,
"step": 2510
},
{
"epoch": 0.26582278481012656,
"grad_norm": 0.6506131291389465,
"learning_rate": 0.0015,
"loss": 1.656,
"step": 2520
},
{
"epoch": 0.2668776371308017,
"grad_norm": 0.7402883768081665,
"learning_rate": 0.0015,
"loss": 1.6513,
"step": 2530
},
{
"epoch": 0.2679324894514768,
"grad_norm": 0.9401949644088745,
"learning_rate": 0.0015,
"loss": 1.6607,
"step": 2540
},
{
"epoch": 0.2689873417721519,
"grad_norm": 0.8450409770011902,
"learning_rate": 0.0015,
"loss": 1.6536,
"step": 2550
},
{
"epoch": 0.270042194092827,
"grad_norm": 0.6884193420410156,
"learning_rate": 0.0015,
"loss": 1.6565,
"step": 2560
},
{
"epoch": 0.27109704641350213,
"grad_norm": 0.6372545957565308,
"learning_rate": 0.0015,
"loss": 1.6506,
"step": 2570
},
{
"epoch": 0.2721518987341772,
"grad_norm": 0.6220951676368713,
"learning_rate": 0.0015,
"loss": 1.6522,
"step": 2580
},
{
"epoch": 0.2732067510548523,
"grad_norm": 0.8408181071281433,
"learning_rate": 0.0015,
"loss": 1.6557,
"step": 2590
},
{
"epoch": 0.2742616033755274,
"grad_norm": 0.6427624225616455,
"learning_rate": 0.0015,
"loss": 1.6569,
"step": 2600
},
{
"epoch": 0.27531645569620256,
"grad_norm": 0.8172615170478821,
"learning_rate": 0.0015,
"loss": 1.654,
"step": 2610
},
{
"epoch": 0.27637130801687765,
"grad_norm": 0.9558082222938538,
"learning_rate": 0.0015,
"loss": 1.6505,
"step": 2620
},
{
"epoch": 0.27742616033755274,
"grad_norm": 0.7412358522415161,
"learning_rate": 0.0015,
"loss": 1.6441,
"step": 2630
},
{
"epoch": 0.27848101265822783,
"grad_norm": 0.6658812165260315,
"learning_rate": 0.0015,
"loss": 1.6528,
"step": 2640
},
{
"epoch": 0.2795358649789029,
"grad_norm": 0.9503615498542786,
"learning_rate": 0.0015,
"loss": 1.6444,
"step": 2650
},
{
"epoch": 0.2805907172995781,
"grad_norm": 0.5820896625518799,
"learning_rate": 0.0015,
"loss": 1.6394,
"step": 2660
},
{
"epoch": 0.28164556962025317,
"grad_norm": 0.737178385257721,
"learning_rate": 0.0015,
"loss": 1.6411,
"step": 2670
},
{
"epoch": 0.28270042194092826,
"grad_norm": 0.6155056357383728,
"learning_rate": 0.0015,
"loss": 1.6426,
"step": 2680
},
{
"epoch": 0.28375527426160335,
"grad_norm": 0.8358173370361328,
"learning_rate": 0.0015,
"loss": 1.6418,
"step": 2690
},
{
"epoch": 0.2848101265822785,
"grad_norm": 0.6908949017524719,
"learning_rate": 0.0015,
"loss": 1.6366,
"step": 2700
},
{
"epoch": 0.2858649789029536,
"grad_norm": 0.5848388671875,
"learning_rate": 0.0015,
"loss": 1.6412,
"step": 2710
},
{
"epoch": 0.2869198312236287,
"grad_norm": 0.7473011612892151,
"learning_rate": 0.0015,
"loss": 1.6435,
"step": 2720
},
{
"epoch": 0.2879746835443038,
"grad_norm": 0.7751439809799194,
"learning_rate": 0.0015,
"loss": 1.625,
"step": 2730
},
{
"epoch": 0.2890295358649789,
"grad_norm": 0.9881758689880371,
"learning_rate": 0.0015,
"loss": 1.6345,
"step": 2740
},
{
"epoch": 0.290084388185654,
"grad_norm": 0.9220196008682251,
"learning_rate": 0.0015,
"loss": 1.641,
"step": 2750
},
{
"epoch": 0.2911392405063291,
"grad_norm": 0.6724462509155273,
"learning_rate": 0.0015,
"loss": 1.6508,
"step": 2760
},
{
"epoch": 0.2921940928270042,
"grad_norm": 0.6800593733787537,
"learning_rate": 0.0015,
"loss": 1.6453,
"step": 2770
},
{
"epoch": 0.29324894514767935,
"grad_norm": 0.6859573125839233,
"learning_rate": 0.0015,
"loss": 1.6378,
"step": 2780
},
{
"epoch": 0.29430379746835444,
"grad_norm": 0.785236120223999,
"learning_rate": 0.0015,
"loss": 1.6414,
"step": 2790
},
{
"epoch": 0.29535864978902954,
"grad_norm": 0.6508416533470154,
"learning_rate": 0.0015,
"loss": 1.6394,
"step": 2800
},
{
"epoch": 0.29641350210970463,
"grad_norm": 0.7265350222587585,
"learning_rate": 0.0015,
"loss": 1.6387,
"step": 2810
},
{
"epoch": 0.2974683544303797,
"grad_norm": 0.7346289753913879,
"learning_rate": 0.0015,
"loss": 1.642,
"step": 2820
},
{
"epoch": 0.29852320675105487,
"grad_norm": 0.5750663876533508,
"learning_rate": 0.0015,
"loss": 1.6179,
"step": 2830
},
{
"epoch": 0.29957805907172996,
"grad_norm": 0.8585373163223267,
"learning_rate": 0.0015,
"loss": 1.6211,
"step": 2840
},
{
"epoch": 0.30063291139240506,
"grad_norm": 0.6604123115539551,
"learning_rate": 0.0015,
"loss": 1.6266,
"step": 2850
},
{
"epoch": 0.30168776371308015,
"grad_norm": 0.7079337239265442,
"learning_rate": 0.0015,
"loss": 1.6293,
"step": 2860
},
{
"epoch": 0.3027426160337553,
"grad_norm": 0.9872538447380066,
"learning_rate": 0.0015,
"loss": 1.6332,
"step": 2870
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.7314916849136353,
"learning_rate": 0.0015,
"loss": 1.6207,
"step": 2880
},
{
"epoch": 0.3048523206751055,
"grad_norm": 0.6987850069999695,
"learning_rate": 0.0015,
"loss": 1.6224,
"step": 2890
},
{
"epoch": 0.3059071729957806,
"grad_norm": 0.658786416053772,
"learning_rate": 0.0015,
"loss": 1.6279,
"step": 2900
},
{
"epoch": 0.3069620253164557,
"grad_norm": 1.8567450046539307,
"learning_rate": 0.0015,
"loss": 1.6421,
"step": 2910
},
{
"epoch": 0.3080168776371308,
"grad_norm": 0.7476452589035034,
"learning_rate": 0.0015,
"loss": 1.625,
"step": 2920
},
{
"epoch": 0.3090717299578059,
"grad_norm": 0.6860174536705017,
"learning_rate": 0.0015,
"loss": 1.6179,
"step": 2930
},
{
"epoch": 0.310126582278481,
"grad_norm": 0.8042653203010559,
"learning_rate": 0.0015,
"loss": 1.6247,
"step": 2940
},
{
"epoch": 0.3111814345991561,
"grad_norm": 0.8760249018669128,
"learning_rate": 0.0015,
"loss": 1.633,
"step": 2950
},
{
"epoch": 0.31223628691983124,
"grad_norm": 1.5029122829437256,
"learning_rate": 0.0015,
"loss": 1.6288,
"step": 2960
},
{
"epoch": 0.31329113924050633,
"grad_norm": 0.8444796800613403,
"learning_rate": 0.0015,
"loss": 1.6184,
"step": 2970
},
{
"epoch": 0.3143459915611814,
"grad_norm": 0.6066569685935974,
"learning_rate": 0.0015,
"loss": 1.6102,
"step": 2980
},
{
"epoch": 0.3154008438818565,
"grad_norm": 0.6329298615455627,
"learning_rate": 0.0015,
"loss": 1.6283,
"step": 2990
},
{
"epoch": 0.31645569620253167,
"grad_norm": 0.6336819529533386,
"learning_rate": 0.0015,
"loss": 1.6086,
"step": 3000
},
{
"epoch": 0.31751054852320676,
"grad_norm": 0.6879556775093079,
"learning_rate": 0.0015,
"loss": 1.6157,
"step": 3010
},
{
"epoch": 0.31856540084388185,
"grad_norm": 0.7391799092292786,
"learning_rate": 0.0015,
"loss": 1.6197,
"step": 3020
},
{
"epoch": 0.31962025316455694,
"grad_norm": 0.5708714723587036,
"learning_rate": 0.0015,
"loss": 1.6042,
"step": 3030
},
{
"epoch": 0.3206751054852321,
"grad_norm": 0.5482274293899536,
"learning_rate": 0.0015,
"loss": 1.6152,
"step": 3040
},
{
"epoch": 0.3217299578059072,
"grad_norm": 0.5858886241912842,
"learning_rate": 0.0015,
"loss": 1.6097,
"step": 3050
},
{
"epoch": 0.3227848101265823,
"grad_norm": 0.7412643432617188,
"learning_rate": 0.0015,
"loss": 1.6219,
"step": 3060
},
{
"epoch": 0.32383966244725737,
"grad_norm": 0.5655277967453003,
"learning_rate": 0.0015,
"loss": 1.6132,
"step": 3070
},
{
"epoch": 0.32489451476793246,
"grad_norm": 0.7304298877716064,
"learning_rate": 0.0015,
"loss": 1.6011,
"step": 3080
},
{
"epoch": 0.3259493670886076,
"grad_norm": 0.8199241757392883,
"learning_rate": 0.0015,
"loss": 1.6198,
"step": 3090
},
{
"epoch": 0.3270042194092827,
"grad_norm": 0.5427016615867615,
"learning_rate": 0.0015,
"loss": 1.6049,
"step": 3100
},
{
"epoch": 0.3280590717299578,
"grad_norm": 0.5689124464988708,
"learning_rate": 0.0015,
"loss": 1.6085,
"step": 3110
},
{
"epoch": 0.3291139240506329,
"grad_norm": 0.6812858581542969,
"learning_rate": 0.0015,
"loss": 1.6118,
"step": 3120
},
{
"epoch": 0.33016877637130804,
"grad_norm": 0.5897095203399658,
"learning_rate": 0.0015,
"loss": 1.6239,
"step": 3130
},
{
"epoch": 0.33122362869198313,
"grad_norm": 0.9220945835113525,
"learning_rate": 0.0015,
"loss": 1.6001,
"step": 3140
},
{
"epoch": 0.3322784810126582,
"grad_norm": 0.9273051619529724,
"learning_rate": 0.0015,
"loss": 1.6101,
"step": 3150
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.6259636282920837,
"learning_rate": 0.0015,
"loss": 1.6116,
"step": 3160
},
{
"epoch": 0.33438818565400846,
"grad_norm": 0.6592155694961548,
"learning_rate": 0.0015,
"loss": 1.6147,
"step": 3170
},
{
"epoch": 0.33544303797468356,
"grad_norm": 0.6287351846694946,
"learning_rate": 0.0015,
"loss": 1.6056,
"step": 3180
},
{
"epoch": 0.33649789029535865,
"grad_norm": 0.982848584651947,
"learning_rate": 0.0015,
"loss": 1.6079,
"step": 3190
},
{
"epoch": 0.33755274261603374,
"grad_norm": 0.6584777235984802,
"learning_rate": 0.0015,
"loss": 1.5998,
"step": 3200
},
{
"epoch": 0.33860759493670883,
"grad_norm": 0.6095257997512817,
"learning_rate": 0.0015,
"loss": 1.6171,
"step": 3210
},
{
"epoch": 0.339662447257384,
"grad_norm": 0.6312419772148132,
"learning_rate": 0.0015,
"loss": 1.6054,
"step": 3220
},
{
"epoch": 0.3407172995780591,
"grad_norm": 0.5978062152862549,
"learning_rate": 0.0015,
"loss": 1.5997,
"step": 3230
},
{
"epoch": 0.34177215189873417,
"grad_norm": 0.5813004970550537,
"learning_rate": 0.0015,
"loss": 1.5785,
"step": 3240
},
{
"epoch": 0.34282700421940926,
"grad_norm": 0.6352910995483398,
"learning_rate": 0.0015,
"loss": 1.603,
"step": 3250
},
{
"epoch": 0.3438818565400844,
"grad_norm": 0.5998354554176331,
"learning_rate": 0.0015,
"loss": 1.5949,
"step": 3260
},
{
"epoch": 0.3449367088607595,
"grad_norm": 0.6111003756523132,
"learning_rate": 0.0015,
"loss": 1.6095,
"step": 3270
},
{
"epoch": 0.3459915611814346,
"grad_norm": 0.566878616809845,
"learning_rate": 0.0015,
"loss": 1.6106,
"step": 3280
},
{
"epoch": 0.3470464135021097,
"grad_norm": 0.6212012767791748,
"learning_rate": 0.0015,
"loss": 1.5931,
"step": 3290
},
{
"epoch": 0.34810126582278483,
"grad_norm": 0.6310331225395203,
"learning_rate": 0.0015,
"loss": 1.6032,
"step": 3300
},
{
"epoch": 0.3491561181434599,
"grad_norm": 0.6017879247665405,
"learning_rate": 0.0015,
"loss": 1.6022,
"step": 3310
},
{
"epoch": 0.350210970464135,
"grad_norm": 0.6400523781776428,
"learning_rate": 0.0015,
"loss": 1.607,
"step": 3320
},
{
"epoch": 0.3512658227848101,
"grad_norm": 0.5609440207481384,
"learning_rate": 0.0015,
"loss": 1.599,
"step": 3330
},
{
"epoch": 0.35232067510548526,
"grad_norm": 0.6740028262138367,
"learning_rate": 0.0015,
"loss": 1.6011,
"step": 3340
},
{
"epoch": 0.35337552742616035,
"grad_norm": 0.7476164102554321,
"learning_rate": 0.0015,
"loss": 1.5933,
"step": 3350
},
{
"epoch": 0.35443037974683544,
"grad_norm": 0.8268814086914062,
"learning_rate": 0.0015,
"loss": 1.5934,
"step": 3360
},
{
"epoch": 0.35548523206751054,
"grad_norm": 0.67869633436203,
"learning_rate": 0.0015,
"loss": 1.6028,
"step": 3370
},
{
"epoch": 0.35654008438818563,
"grad_norm": 0.5998952984809875,
"learning_rate": 0.0015,
"loss": 1.5976,
"step": 3380
},
{
"epoch": 0.3575949367088608,
"grad_norm": 0.8241248726844788,
"learning_rate": 0.0015,
"loss": 1.5924,
"step": 3390
},
{
"epoch": 0.35864978902953587,
"grad_norm": 0.6471379995346069,
"learning_rate": 0.0015,
"loss": 1.5999,
"step": 3400
},
{
"epoch": 0.35970464135021096,
"grad_norm": 0.6880522966384888,
"learning_rate": 0.0015,
"loss": 1.5978,
"step": 3410
},
{
"epoch": 0.36075949367088606,
"grad_norm": 0.5927819013595581,
"learning_rate": 0.0015,
"loss": 1.5752,
"step": 3420
},
{
"epoch": 0.3618143459915612,
"grad_norm": 0.623607337474823,
"learning_rate": 0.0015,
"loss": 1.5933,
"step": 3430
},
{
"epoch": 0.3628691983122363,
"grad_norm": 0.7323747277259827,
"learning_rate": 0.0015,
"loss": 1.5918,
"step": 3440
},
{
"epoch": 0.3639240506329114,
"grad_norm": 0.6532187461853027,
"learning_rate": 0.0015,
"loss": 1.588,
"step": 3450
},
{
"epoch": 0.3649789029535865,
"grad_norm": 0.5976588129997253,
"learning_rate": 0.0015,
"loss": 1.5803,
"step": 3460
},
{
"epoch": 0.36603375527426163,
"grad_norm": 0.6030303835868835,
"learning_rate": 0.0015,
"loss": 1.5976,
"step": 3470
},
{
"epoch": 0.3670886075949367,
"grad_norm": 0.6205447912216187,
"learning_rate": 0.0015,
"loss": 1.6031,
"step": 3480
},
{
"epoch": 0.3681434599156118,
"grad_norm": 0.7707657814025879,
"learning_rate": 0.0015,
"loss": 1.5968,
"step": 3490
},
{
"epoch": 0.3691983122362869,
"grad_norm": 0.651380717754364,
"learning_rate": 0.0015,
"loss": 1.5739,
"step": 3500
},
{
"epoch": 0.370253164556962,
"grad_norm": 0.6881337761878967,
"learning_rate": 0.0015,
"loss": 1.5931,
"step": 3510
},
{
"epoch": 0.37130801687763715,
"grad_norm": 0.589093804359436,
"learning_rate": 0.0015,
"loss": 1.5906,
"step": 3520
},
{
"epoch": 0.37236286919831224,
"grad_norm": 0.774224579334259,
"learning_rate": 0.0015,
"loss": 1.6007,
"step": 3530
},
{
"epoch": 0.37341772151898733,
"grad_norm": 0.579230546951294,
"learning_rate": 0.0015,
"loss": 1.5936,
"step": 3540
},
{
"epoch": 0.3744725738396624,
"grad_norm": 0.6935818195343018,
"learning_rate": 0.0015,
"loss": 1.5876,
"step": 3550
},
{
"epoch": 0.3755274261603376,
"grad_norm": 0.6207526326179504,
"learning_rate": 0.0015,
"loss": 1.5856,
"step": 3560
},
{
"epoch": 0.37658227848101267,
"grad_norm": 0.8157477974891663,
"learning_rate": 0.0015,
"loss": 1.5808,
"step": 3570
},
{
"epoch": 0.37763713080168776,
"grad_norm": 0.5693010687828064,
"learning_rate": 0.0015,
"loss": 1.5799,
"step": 3580
},
{
"epoch": 0.37869198312236285,
"grad_norm": 0.6131789088249207,
"learning_rate": 0.0015,
"loss": 1.5949,
"step": 3590
},
{
"epoch": 0.379746835443038,
"grad_norm": 0.8241103887557983,
"learning_rate": 0.0015,
"loss": 1.5764,
"step": 3600
},
{
"epoch": 0.3808016877637131,
"grad_norm": 0.7060911655426025,
"learning_rate": 0.0015,
"loss": 1.5798,
"step": 3610
},
{
"epoch": 0.3818565400843882,
"grad_norm": 0.6012842059135437,
"learning_rate": 0.0015,
"loss": 1.5833,
"step": 3620
},
{
"epoch": 0.3829113924050633,
"grad_norm": 0.7175701260566711,
"learning_rate": 0.0015,
"loss": 1.5823,
"step": 3630
},
{
"epoch": 0.38396624472573837,
"grad_norm": 0.6527909636497498,
"learning_rate": 0.0015,
"loss": 1.5838,
"step": 3640
},
{
"epoch": 0.3850210970464135,
"grad_norm": 0.5886304974555969,
"learning_rate": 0.0015,
"loss": 1.5878,
"step": 3650
},
{
"epoch": 0.3860759493670886,
"grad_norm": 0.5935819149017334,
"learning_rate": 0.0015,
"loss": 1.5827,
"step": 3660
},
{
"epoch": 0.3871308016877637,
"grad_norm": 0.5793882012367249,
"learning_rate": 0.0015,
"loss": 1.5743,
"step": 3670
},
{
"epoch": 0.3881856540084388,
"grad_norm": 0.6356774568557739,
"learning_rate": 0.0015,
"loss": 1.5783,
"step": 3680
},
{
"epoch": 0.38924050632911394,
"grad_norm": 0.6113144159317017,
"learning_rate": 0.0015,
"loss": 1.5893,
"step": 3690
},
{
"epoch": 0.39029535864978904,
"grad_norm": 0.6362749934196472,
"learning_rate": 0.0015,
"loss": 1.5917,
"step": 3700
},
{
"epoch": 0.39135021097046413,
"grad_norm": 0.8249803781509399,
"learning_rate": 0.0015,
"loss": 1.5781,
"step": 3710
},
{
"epoch": 0.3924050632911392,
"grad_norm": 0.8612954616546631,
"learning_rate": 0.0015,
"loss": 1.5608,
"step": 3720
},
{
"epoch": 0.39345991561181437,
"grad_norm": 1.7621455192565918,
"learning_rate": 0.0015,
"loss": 1.5877,
"step": 3730
},
{
"epoch": 0.39451476793248946,
"grad_norm": 1.3036092519760132,
"learning_rate": 0.0015,
"loss": 1.5916,
"step": 3740
},
{
"epoch": 0.39556962025316456,
"grad_norm": 0.9393406510353088,
"learning_rate": 0.0015,
"loss": 1.5864,
"step": 3750
},
{
"epoch": 0.39662447257383965,
"grad_norm": 0.9737085103988647,
"learning_rate": 0.0015,
"loss": 1.5715,
"step": 3760
},
{
"epoch": 0.39767932489451474,
"grad_norm": 0.8785778880119324,
"learning_rate": 0.0015,
"loss": 1.5872,
"step": 3770
},
{
"epoch": 0.3987341772151899,
"grad_norm": 0.5576906800270081,
"learning_rate": 0.0015,
"loss": 1.5743,
"step": 3780
},
{
"epoch": 0.399789029535865,
"grad_norm": 0.5176705718040466,
"learning_rate": 0.0015,
"loss": 1.5734,
"step": 3790
},
{
"epoch": 0.4008438818565401,
"grad_norm": 0.6921336650848389,
"learning_rate": 0.0015,
"loss": 1.5808,
"step": 3800
},
{
"epoch": 0.40189873417721517,
"grad_norm": 0.8691692352294922,
"learning_rate": 0.0015,
"loss": 1.5723,
"step": 3810
},
{
"epoch": 0.4029535864978903,
"grad_norm": 0.6363836526870728,
"learning_rate": 0.0015,
"loss": 1.5627,
"step": 3820
},
{
"epoch": 0.4040084388185654,
"grad_norm": 0.6024162769317627,
"learning_rate": 0.0015,
"loss": 1.5676,
"step": 3830
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.628873884677887,
"learning_rate": 0.0015,
"loss": 1.5875,
"step": 3840
},
{
"epoch": 0.4061181434599156,
"grad_norm": 0.5797845721244812,
"learning_rate": 0.0015,
"loss": 1.5663,
"step": 3850
},
{
"epoch": 0.40717299578059074,
"grad_norm": 0.7347355484962463,
"learning_rate": 0.0015,
"loss": 1.5797,
"step": 3860
},
{
"epoch": 0.40822784810126583,
"grad_norm": 0.8189390897750854,
"learning_rate": 0.0015,
"loss": 1.5715,
"step": 3870
},
{
"epoch": 0.4092827004219409,
"grad_norm": 0.5799738764762878,
"learning_rate": 0.0015,
"loss": 1.5603,
"step": 3880
},
{
"epoch": 0.410337552742616,
"grad_norm": 0.6734434366226196,
"learning_rate": 0.0015,
"loss": 1.5592,
"step": 3890
},
{
"epoch": 0.41139240506329117,
"grad_norm": 1.40089750289917,
"learning_rate": 0.0015,
"loss": 1.563,
"step": 3900
},
{
"epoch": 0.41244725738396626,
"grad_norm": 0.7925665974617004,
"learning_rate": 0.0015,
"loss": 1.5689,
"step": 3910
},
{
"epoch": 0.41350210970464135,
"grad_norm": 0.7021671533584595,
"learning_rate": 0.0015,
"loss": 1.5624,
"step": 3920
},
{
"epoch": 0.41455696202531644,
"grad_norm": 0.5486526489257812,
"learning_rate": 0.0015,
"loss": 1.5664,
"step": 3930
},
{
"epoch": 0.41561181434599154,
"grad_norm": 0.6393633484840393,
"learning_rate": 0.0015,
"loss": 1.5618,
"step": 3940
},
{
"epoch": 0.4166666666666667,
"grad_norm": 0.5666303634643555,
"learning_rate": 0.0015,
"loss": 1.5677,
"step": 3950
},
{
"epoch": 0.4177215189873418,
"grad_norm": 0.6497644782066345,
"learning_rate": 0.0015,
"loss": 1.5697,
"step": 3960
},
{
"epoch": 0.41877637130801687,
"grad_norm": 0.5655925273895264,
"learning_rate": 0.0015,
"loss": 1.5562,
"step": 3970
},
{
"epoch": 0.41983122362869196,
"grad_norm": 0.8136340975761414,
"learning_rate": 0.0015,
"loss": 1.5709,
"step": 3980
},
{
"epoch": 0.4208860759493671,
"grad_norm": 0.874515950679779,
"learning_rate": 0.0015,
"loss": 1.5607,
"step": 3990
},
{
"epoch": 0.4219409282700422,
"grad_norm": 0.5788293480873108,
"learning_rate": 0.0015,
"loss": 1.576,
"step": 4000
},
{
"epoch": 0.4229957805907173,
"grad_norm": 0.5628449320793152,
"learning_rate": 0.0015,
"loss": 1.5605,
"step": 4010
},
{
"epoch": 0.4240506329113924,
"grad_norm": 0.5451663136482239,
"learning_rate": 0.0015,
"loss": 1.5619,
"step": 4020
},
{
"epoch": 0.42510548523206754,
"grad_norm": 0.5687470436096191,
"learning_rate": 0.0015,
"loss": 1.5556,
"step": 4030
},
{
"epoch": 0.42616033755274263,
"grad_norm": 0.6235162019729614,
"learning_rate": 0.0015,
"loss": 1.5632,
"step": 4040
},
{
"epoch": 0.4272151898734177,
"grad_norm": 0.5888081192970276,
"learning_rate": 0.0015,
"loss": 1.5638,
"step": 4050
},
{
"epoch": 0.4282700421940928,
"grad_norm": 1.0608152151107788,
"learning_rate": 0.0015,
"loss": 1.5717,
"step": 4060
},
{
"epoch": 0.4293248945147679,
"grad_norm": 0.8200925588607788,
"learning_rate": 0.0015,
"loss": 1.5724,
"step": 4070
},
{
"epoch": 0.43037974683544306,
"grad_norm": 0.6000164151191711,
"learning_rate": 0.0015,
"loss": 1.5819,
"step": 4080
},
{
"epoch": 0.43143459915611815,
"grad_norm": 0.632230281829834,
"learning_rate": 0.0015,
"loss": 1.5607,
"step": 4090
},
{
"epoch": 0.43248945147679324,
"grad_norm": 0.6781566143035889,
"learning_rate": 0.0015,
"loss": 1.5579,
"step": 4100
},
{
"epoch": 0.43354430379746833,
"grad_norm": 0.6161074042320251,
"learning_rate": 0.0015,
"loss": 1.5549,
"step": 4110
},
{
"epoch": 0.4345991561181435,
"grad_norm": 0.7849648594856262,
"learning_rate": 0.0015,
"loss": 1.5565,
"step": 4120
},
{
"epoch": 0.4356540084388186,
"grad_norm": 0.7117999196052551,
"learning_rate": 0.0015,
"loss": 1.5533,
"step": 4130
},
{
"epoch": 0.43670886075949367,
"grad_norm": 0.559719443321228,
"learning_rate": 0.0015,
"loss": 1.5558,
"step": 4140
},
{
"epoch": 0.43776371308016876,
"grad_norm": 1.0182809829711914,
"learning_rate": 0.0015,
"loss": 1.5608,
"step": 4150
},
{
"epoch": 0.4388185654008439,
"grad_norm": 0.5550006628036499,
"learning_rate": 0.0015,
"loss": 1.5536,
"step": 4160
},
{
"epoch": 0.439873417721519,
"grad_norm": 0.5445747375488281,
"learning_rate": 0.0015,
"loss": 1.554,
"step": 4170
},
{
"epoch": 0.4409282700421941,
"grad_norm": 0.5958618521690369,
"learning_rate": 0.0015,
"loss": 1.5497,
"step": 4180
},
{
"epoch": 0.4419831223628692,
"grad_norm": 0.5892060995101929,
"learning_rate": 0.0015,
"loss": 1.559,
"step": 4190
},
{
"epoch": 0.4430379746835443,
"grad_norm": 0.5875847339630127,
"learning_rate": 0.0015,
"loss": 1.5576,
"step": 4200
},
{
"epoch": 0.4440928270042194,
"grad_norm": 0.7794286608695984,
"learning_rate": 0.0015,
"loss": 1.5495,
"step": 4210
},
{
"epoch": 0.4451476793248945,
"grad_norm": 0.5659272074699402,
"learning_rate": 0.0015,
"loss": 1.5555,
"step": 4220
},
{
"epoch": 0.4462025316455696,
"grad_norm": 0.5580011010169983,
"learning_rate": 0.0015,
"loss": 1.55,
"step": 4230
},
{
"epoch": 0.4472573839662447,
"grad_norm": 0.6203771829605103,
"learning_rate": 0.0015,
"loss": 1.5515,
"step": 4240
},
{
"epoch": 0.44831223628691985,
"grad_norm": 0.682816743850708,
"learning_rate": 0.0015,
"loss": 1.5575,
"step": 4250
},
{
"epoch": 0.44936708860759494,
"grad_norm": 0.6003069281578064,
"learning_rate": 0.0015,
"loss": 1.5646,
"step": 4260
},
{
"epoch": 0.45042194092827004,
"grad_norm": 0.5731751918792725,
"learning_rate": 0.0015,
"loss": 1.5498,
"step": 4270
},
{
"epoch": 0.45147679324894513,
"grad_norm": 1.3089572191238403,
"learning_rate": 0.0015,
"loss": 1.5434,
"step": 4280
},
{
"epoch": 0.4525316455696203,
"grad_norm": 0.8871122598648071,
"learning_rate": 0.0015,
"loss": 1.557,
"step": 4290
},
{
"epoch": 0.45358649789029537,
"grad_norm": 0.7614080905914307,
"learning_rate": 0.0015,
"loss": 1.5533,
"step": 4300
},
{
"epoch": 0.45464135021097046,
"grad_norm": 0.5850949883460999,
"learning_rate": 0.0015,
"loss": 1.5542,
"step": 4310
},
{
"epoch": 0.45569620253164556,
"grad_norm": 0.6122797727584839,
"learning_rate": 0.0015,
"loss": 1.5463,
"step": 4320
},
{
"epoch": 0.45675105485232065,
"grad_norm": 0.6340963840484619,
"learning_rate": 0.0015,
"loss": 1.5578,
"step": 4330
},
{
"epoch": 0.4578059071729958,
"grad_norm": 0.7485887408256531,
"learning_rate": 0.0015,
"loss": 1.5457,
"step": 4340
},
{
"epoch": 0.4588607594936709,
"grad_norm": 0.8456482887268066,
"learning_rate": 0.0015,
"loss": 1.5487,
"step": 4350
},
{
"epoch": 0.459915611814346,
"grad_norm": 0.5885475277900696,
"learning_rate": 0.0015,
"loss": 1.5621,
"step": 4360
},
{
"epoch": 0.4609704641350211,
"grad_norm": 0.7908206582069397,
"learning_rate": 0.0015,
"loss": 1.5433,
"step": 4370
},
{
"epoch": 0.4620253164556962,
"grad_norm": 0.5468949675559998,
"learning_rate": 0.0015,
"loss": 1.5446,
"step": 4380
},
{
"epoch": 0.4630801687763713,
"grad_norm": 0.5600817799568176,
"learning_rate": 0.0015,
"loss": 1.5413,
"step": 4390
},
{
"epoch": 0.4641350210970464,
"grad_norm": 0.5512800216674805,
"learning_rate": 0.0015,
"loss": 1.5467,
"step": 4400
},
{
"epoch": 0.4651898734177215,
"grad_norm": 0.7056983113288879,
"learning_rate": 0.0015,
"loss": 1.5638,
"step": 4410
},
{
"epoch": 0.46624472573839665,
"grad_norm": 0.9344276785850525,
"learning_rate": 0.0015,
"loss": 1.5457,
"step": 4420
},
{
"epoch": 0.46729957805907174,
"grad_norm": 0.8951518535614014,
"learning_rate": 0.0015,
"loss": 1.5528,
"step": 4430
},
{
"epoch": 0.46835443037974683,
"grad_norm": 0.6155388355255127,
"learning_rate": 0.0015,
"loss": 1.5467,
"step": 4440
},
{
"epoch": 0.4694092827004219,
"grad_norm": 0.7321780323982239,
"learning_rate": 0.0015,
"loss": 1.5539,
"step": 4450
},
{
"epoch": 0.4704641350210971,
"grad_norm": 0.7223412394523621,
"learning_rate": 0.0015,
"loss": 1.542,
"step": 4460
},
{
"epoch": 0.47151898734177217,
"grad_norm": 0.580058753490448,
"learning_rate": 0.0015,
"loss": 1.5512,
"step": 4470
},
{
"epoch": 0.47257383966244726,
"grad_norm": 0.626334547996521,
"learning_rate": 0.0015,
"loss": 1.5538,
"step": 4480
},
{
"epoch": 0.47362869198312235,
"grad_norm": 0.9320026636123657,
"learning_rate": 0.0015,
"loss": 1.5435,
"step": 4490
},
{
"epoch": 0.47468354430379744,
"grad_norm": 0.6046232581138611,
"learning_rate": 0.0015,
"loss": 1.5446,
"step": 4500
},
{
"epoch": 0.4757383966244726,
"grad_norm": 0.7325782775878906,
"learning_rate": 0.0015,
"loss": 1.5463,
"step": 4510
},
{
"epoch": 0.4767932489451477,
"grad_norm": 0.571357786655426,
"learning_rate": 0.0015,
"loss": 1.5328,
"step": 4520
},
{
"epoch": 0.4778481012658228,
"grad_norm": 0.5945020318031311,
"learning_rate": 0.0015,
"loss": 1.5487,
"step": 4530
},
{
"epoch": 0.47890295358649787,
"grad_norm": 0.6285760402679443,
"learning_rate": 0.0015,
"loss": 1.5507,
"step": 4540
},
{
"epoch": 0.479957805907173,
"grad_norm": 0.5691685676574707,
"learning_rate": 0.0015,
"loss": 1.5472,
"step": 4550
},
{
"epoch": 0.4810126582278481,
"grad_norm": 0.5442836880683899,
"learning_rate": 0.0015,
"loss": 1.537,
"step": 4560
},
{
"epoch": 0.4820675105485232,
"grad_norm": 0.6255428194999695,
"learning_rate": 0.0015,
"loss": 1.5193,
"step": 4570
},
{
"epoch": 0.4831223628691983,
"grad_norm": 0.554628849029541,
"learning_rate": 0.0015,
"loss": 1.5447,
"step": 4580
},
{
"epoch": 0.48417721518987344,
"grad_norm": 0.6360525488853455,
"learning_rate": 0.0015,
"loss": 1.5315,
"step": 4590
},
{
"epoch": 0.48523206751054854,
"grad_norm": 0.7742726802825928,
"learning_rate": 0.0015,
"loss": 1.5487,
"step": 4600
},
{
"epoch": 0.48628691983122363,
"grad_norm": 0.8166155219078064,
"learning_rate": 0.0015,
"loss": 1.5497,
"step": 4610
},
{
"epoch": 0.4873417721518987,
"grad_norm": 0.5890775918960571,
"learning_rate": 0.0015,
"loss": 1.527,
"step": 4620
},
{
"epoch": 0.4883966244725738,
"grad_norm": 0.6394803524017334,
"learning_rate": 0.0015,
"loss": 1.5464,
"step": 4630
},
{
"epoch": 0.48945147679324896,
"grad_norm": 0.6207756996154785,
"learning_rate": 0.0015,
"loss": 1.5459,
"step": 4640
},
{
"epoch": 0.49050632911392406,
"grad_norm": 0.5840123891830444,
"learning_rate": 0.0015,
"loss": 1.5507,
"step": 4650
},
{
"epoch": 0.49156118143459915,
"grad_norm": 0.5675470232963562,
"learning_rate": 0.0015,
"loss": 1.5353,
"step": 4660
},
{
"epoch": 0.49261603375527424,
"grad_norm": 0.5648407340049744,
"learning_rate": 0.0015,
"loss": 1.5372,
"step": 4670
},
{
"epoch": 0.4936708860759494,
"grad_norm": 0.5731132626533508,
"learning_rate": 0.0015,
"loss": 1.5304,
"step": 4680
},
{
"epoch": 0.4947257383966245,
"grad_norm": 0.8239709138870239,
"learning_rate": 0.0015,
"loss": 1.5294,
"step": 4690
},
{
"epoch": 0.4957805907172996,
"grad_norm": 0.8988879919052124,
"learning_rate": 0.0015,
"loss": 1.5411,
"step": 4700
},
{
"epoch": 0.49683544303797467,
"grad_norm": 0.5612457990646362,
"learning_rate": 0.0015,
"loss": 1.5405,
"step": 4710
},
{
"epoch": 0.4978902953586498,
"grad_norm": 0.7755550146102905,
"learning_rate": 0.0015,
"loss": 1.5394,
"step": 4720
},
{
"epoch": 0.4989451476793249,
"grad_norm": 0.5818630456924438,
"learning_rate": 0.0015,
"loss": 1.5331,
"step": 4730
},
{
"epoch": 0.5,
"grad_norm": 0.5022954344749451,
"learning_rate": 0.0015,
"loss": 1.5316,
"step": 4740
},
{
"epoch": 0.5010548523206751,
"grad_norm": 0.9314385056495667,
"learning_rate": 0.0015,
"loss": 1.5524,
"step": 4750
},
{
"epoch": 0.5021097046413502,
"grad_norm": 0.7994999885559082,
"learning_rate": 0.0015,
"loss": 1.5383,
"step": 4760
},
{
"epoch": 0.5031645569620253,
"grad_norm": 0.6273106932640076,
"learning_rate": 0.0015,
"loss": 1.5322,
"step": 4770
},
{
"epoch": 0.5042194092827004,
"grad_norm": 0.600748598575592,
"learning_rate": 0.0015,
"loss": 1.5433,
"step": 4780
},
{
"epoch": 0.5052742616033755,
"grad_norm": 0.7291206121444702,
"learning_rate": 0.0015,
"loss": 1.527,
"step": 4790
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.7947571873664856,
"learning_rate": 0.0015,
"loss": 1.5326,
"step": 4800
},
{
"epoch": 0.5073839662447257,
"grad_norm": 0.6562824845314026,
"learning_rate": 0.0015,
"loss": 1.5337,
"step": 4810
},
{
"epoch": 0.5084388185654009,
"grad_norm": 0.6121402382850647,
"learning_rate": 0.0015,
"loss": 1.554,
"step": 4820
},
{
"epoch": 0.509493670886076,
"grad_norm": 0.9882345795631409,
"learning_rate": 0.0015,
"loss": 1.5372,
"step": 4830
},
{
"epoch": 0.510548523206751,
"grad_norm": 0.6104786992073059,
"learning_rate": 0.0015,
"loss": 1.5303,
"step": 4840
},
{
"epoch": 0.5116033755274262,
"grad_norm": 0.8607736229896545,
"learning_rate": 0.0015,
"loss": 1.5444,
"step": 4850
},
{
"epoch": 0.5126582278481012,
"grad_norm": 0.6029435992240906,
"learning_rate": 0.0015,
"loss": 1.5298,
"step": 4860
},
{
"epoch": 0.5137130801687764,
"grad_norm": 0.7495979070663452,
"learning_rate": 0.0015,
"loss": 1.5225,
"step": 4870
},
{
"epoch": 0.5147679324894515,
"grad_norm": 0.6200342178344727,
"learning_rate": 0.0015,
"loss": 1.5405,
"step": 4880
},
{
"epoch": 0.5158227848101266,
"grad_norm": 0.5241333842277527,
"learning_rate": 0.0015,
"loss": 1.5306,
"step": 4890
},
{
"epoch": 0.5168776371308017,
"grad_norm": 0.5704604387283325,
"learning_rate": 0.0015,
"loss": 1.5394,
"step": 4900
},
{
"epoch": 0.5179324894514767,
"grad_norm": 0.6147574782371521,
"learning_rate": 0.0015,
"loss": 1.5335,
"step": 4910
},
{
"epoch": 0.5189873417721519,
"grad_norm": 0.6230430006980896,
"learning_rate": 0.0015,
"loss": 1.5283,
"step": 4920
},
{
"epoch": 0.520042194092827,
"grad_norm": 0.678521454334259,
"learning_rate": 0.0015,
"loss": 1.5343,
"step": 4930
},
{
"epoch": 0.5210970464135021,
"grad_norm": 0.8999406695365906,
"learning_rate": 0.0015,
"loss": 1.5324,
"step": 4940
},
{
"epoch": 0.5221518987341772,
"grad_norm": 0.7650752067565918,
"learning_rate": 0.0015,
"loss": 1.5426,
"step": 4950
},
{
"epoch": 0.5232067510548524,
"grad_norm": 0.7141919732093811,
"learning_rate": 0.0015,
"loss": 1.5218,
"step": 4960
},
{
"epoch": 0.5242616033755274,
"grad_norm": 0.5926218628883362,
"learning_rate": 0.0015,
"loss": 1.5294,
"step": 4970
},
{
"epoch": 0.5253164556962026,
"grad_norm": 0.6089852452278137,
"learning_rate": 0.0015,
"loss": 1.5233,
"step": 4980
},
{
"epoch": 0.5263713080168776,
"grad_norm": 0.7046988010406494,
"learning_rate": 0.0015,
"loss": 1.5261,
"step": 4990
},
{
"epoch": 0.5274261603375527,
"grad_norm": 0.6000553369522095,
"learning_rate": 0.0015,
"loss": 1.5373,
"step": 5000
},
{
"epoch": 0.5284810126582279,
"grad_norm": 0.5103377103805542,
"learning_rate": 0.0015,
"loss": 1.5296,
"step": 5010
},
{
"epoch": 0.5295358649789029,
"grad_norm": 0.7494931221008301,
"learning_rate": 0.0015,
"loss": 1.5253,
"step": 5020
},
{
"epoch": 0.5305907172995781,
"grad_norm": 0.5808805823326111,
"learning_rate": 0.0015,
"loss": 1.5073,
"step": 5030
},
{
"epoch": 0.5316455696202531,
"grad_norm": 0.5486202239990234,
"learning_rate": 0.0015,
"loss": 1.5374,
"step": 5040
},
{
"epoch": 0.5327004219409283,
"grad_norm": 0.5495449900627136,
"learning_rate": 0.0015,
"loss": 1.5272,
"step": 5050
},
{
"epoch": 0.5337552742616034,
"grad_norm": 0.7107338905334473,
"learning_rate": 0.0015,
"loss": 1.5377,
"step": 5060
},
{
"epoch": 0.5348101265822784,
"grad_norm": 0.6358274817466736,
"learning_rate": 0.0015,
"loss": 1.526,
"step": 5070
},
{
"epoch": 0.5358649789029536,
"grad_norm": 0.5813426375389099,
"learning_rate": 0.0015,
"loss": 1.5372,
"step": 5080
},
{
"epoch": 0.5369198312236287,
"grad_norm": 0.7069512009620667,
"learning_rate": 0.0015,
"loss": 1.529,
"step": 5090
},
{
"epoch": 0.5379746835443038,
"grad_norm": 0.5345435738563538,
"learning_rate": 0.0015,
"loss": 1.5103,
"step": 5100
},
{
"epoch": 0.5390295358649789,
"grad_norm": 0.5351413488388062,
"learning_rate": 0.0015,
"loss": 1.5245,
"step": 5110
},
{
"epoch": 0.540084388185654,
"grad_norm": 0.672135055065155,
"learning_rate": 0.0015,
"loss": 1.5229,
"step": 5120
},
{
"epoch": 0.5411392405063291,
"grad_norm": 0.5577877163887024,
"learning_rate": 0.0015,
"loss": 1.5213,
"step": 5130
},
{
"epoch": 0.5421940928270043,
"grad_norm": 0.6954469084739685,
"learning_rate": 0.0015,
"loss": 1.5422,
"step": 5140
},
{
"epoch": 0.5432489451476793,
"grad_norm": 0.721547544002533,
"learning_rate": 0.0015,
"loss": 1.5346,
"step": 5150
},
{
"epoch": 0.5443037974683544,
"grad_norm": 0.5934624075889587,
"learning_rate": 0.0015,
"loss": 1.5253,
"step": 5160
},
{
"epoch": 0.5453586497890295,
"grad_norm": 0.8245863318443298,
"learning_rate": 0.0015,
"loss": 1.5312,
"step": 5170
},
{
"epoch": 0.5464135021097046,
"grad_norm": 0.632466733455658,
"learning_rate": 0.0015,
"loss": 1.5377,
"step": 5180
},
{
"epoch": 0.5474683544303798,
"grad_norm": 0.7374550104141235,
"learning_rate": 0.0015,
"loss": 1.5187,
"step": 5190
},
{
"epoch": 0.5485232067510548,
"grad_norm": 0.8513514399528503,
"learning_rate": 0.0015,
"loss": 1.5297,
"step": 5200
},
{
"epoch": 0.54957805907173,
"grad_norm": 0.7695537209510803,
"learning_rate": 0.0015,
"loss": 1.5163,
"step": 5210
},
{
"epoch": 0.5506329113924051,
"grad_norm": 0.746254026889801,
"learning_rate": 0.0015,
"loss": 1.5258,
"step": 5220
},
{
"epoch": 0.5516877637130801,
"grad_norm": 0.5853145122528076,
"learning_rate": 0.0015,
"loss": 1.5185,
"step": 5230
},
{
"epoch": 0.5527426160337553,
"grad_norm": 0.5808828473091125,
"learning_rate": 0.0015,
"loss": 1.5253,
"step": 5240
},
{
"epoch": 0.5537974683544303,
"grad_norm": 0.7577324509620667,
"learning_rate": 0.0015,
"loss": 1.5204,
"step": 5250
},
{
"epoch": 0.5548523206751055,
"grad_norm": 0.5829281210899353,
"learning_rate": 0.0015,
"loss": 1.5279,
"step": 5260
},
{
"epoch": 0.5559071729957806,
"grad_norm": 0.8018589019775391,
"learning_rate": 0.0015,
"loss": 1.532,
"step": 5270
},
{
"epoch": 0.5569620253164557,
"grad_norm": 0.6416056752204895,
"learning_rate": 0.0015,
"loss": 1.5254,
"step": 5280
},
{
"epoch": 0.5580168776371308,
"grad_norm": 0.7368462681770325,
"learning_rate": 0.0015,
"loss": 1.5221,
"step": 5290
},
{
"epoch": 0.5590717299578059,
"grad_norm": 0.541665256023407,
"learning_rate": 0.0015,
"loss": 1.5287,
"step": 5300
},
{
"epoch": 0.560126582278481,
"grad_norm": 0.7606929540634155,
"learning_rate": 0.0015,
"loss": 1.518,
"step": 5310
},
{
"epoch": 0.5611814345991561,
"grad_norm": 0.647328794002533,
"learning_rate": 0.0015,
"loss": 1.5171,
"step": 5320
},
{
"epoch": 0.5622362869198312,
"grad_norm": 0.8251186013221741,
"learning_rate": 0.0015,
"loss": 1.5291,
"step": 5330
},
{
"epoch": 0.5632911392405063,
"grad_norm": 0.6676337122917175,
"learning_rate": 0.0015,
"loss": 1.5206,
"step": 5340
},
{
"epoch": 0.5643459915611815,
"grad_norm": 0.6196624040603638,
"learning_rate": 0.0015,
"loss": 1.5094,
"step": 5350
},
{
"epoch": 0.5654008438818565,
"grad_norm": 0.8831680417060852,
"learning_rate": 0.0015,
"loss": 1.5197,
"step": 5360
},
{
"epoch": 0.5664556962025317,
"grad_norm": 0.5513179898262024,
"learning_rate": 0.0015,
"loss": 1.5197,
"step": 5370
},
{
"epoch": 0.5675105485232067,
"grad_norm": 0.5404736399650574,
"learning_rate": 0.0015,
"loss": 1.5132,
"step": 5380
},
{
"epoch": 0.5685654008438819,
"grad_norm": 0.7697404026985168,
"learning_rate": 0.0015,
"loss": 1.516,
"step": 5390
},
{
"epoch": 0.569620253164557,
"grad_norm": 0.5695388317108154,
"learning_rate": 0.0015,
"loss": 1.5182,
"step": 5400
},
{
"epoch": 0.570675105485232,
"grad_norm": 0.636220395565033,
"learning_rate": 0.0015,
"loss": 1.5201,
"step": 5410
},
{
"epoch": 0.5717299578059072,
"grad_norm": 0.5687567591667175,
"learning_rate": 0.0015,
"loss": 1.5111,
"step": 5420
},
{
"epoch": 0.5727848101265823,
"grad_norm": 0.5303071141242981,
"learning_rate": 0.0015,
"loss": 1.5107,
"step": 5430
},
{
"epoch": 0.5738396624472574,
"grad_norm": 0.8346840143203735,
"learning_rate": 0.0015,
"loss": 1.5094,
"step": 5440
},
{
"epoch": 0.5748945147679325,
"grad_norm": 0.6687101125717163,
"learning_rate": 0.0015,
"loss": 1.5143,
"step": 5450
},
{
"epoch": 0.5759493670886076,
"grad_norm": 0.6760293245315552,
"learning_rate": 0.0015,
"loss": 1.5148,
"step": 5460
},
{
"epoch": 0.5770042194092827,
"grad_norm": 0.6732572913169861,
"learning_rate": 0.0015,
"loss": 1.5144,
"step": 5470
},
{
"epoch": 0.5780590717299579,
"grad_norm": 0.8290615677833557,
"learning_rate": 0.0015,
"loss": 1.5165,
"step": 5480
},
{
"epoch": 0.5791139240506329,
"grad_norm": 0.5455368757247925,
"learning_rate": 0.0015,
"loss": 1.5232,
"step": 5490
},
{
"epoch": 0.580168776371308,
"grad_norm": 0.5541859865188599,
"learning_rate": 0.0015,
"loss": 1.5133,
"step": 5500
},
{
"epoch": 0.5812236286919831,
"grad_norm": 0.7072927951812744,
"learning_rate": 0.0015,
"loss": 1.518,
"step": 5510
},
{
"epoch": 0.5822784810126582,
"grad_norm": 0.5367401242256165,
"learning_rate": 0.0015,
"loss": 1.5118,
"step": 5520
},
{
"epoch": 0.5833333333333334,
"grad_norm": 0.5651088356971741,
"learning_rate": 0.0015,
"loss": 1.5039,
"step": 5530
},
{
"epoch": 0.5843881856540084,
"grad_norm": 0.5084605813026428,
"learning_rate": 0.0015,
"loss": 1.5069,
"step": 5540
},
{
"epoch": 0.5854430379746836,
"grad_norm": 0.6395454406738281,
"learning_rate": 0.0015,
"loss": 1.5226,
"step": 5550
},
{
"epoch": 0.5864978902953587,
"grad_norm": 1.0206706523895264,
"learning_rate": 0.0015,
"loss": 1.5269,
"step": 5560
},
{
"epoch": 0.5875527426160337,
"grad_norm": 0.5466570854187012,
"learning_rate": 0.0015,
"loss": 1.5075,
"step": 5570
},
{
"epoch": 0.5886075949367089,
"grad_norm": 0.8130196928977966,
"learning_rate": 0.0015,
"loss": 1.5113,
"step": 5580
},
{
"epoch": 0.5896624472573839,
"grad_norm": 0.7004330158233643,
"learning_rate": 0.0015,
"loss": 1.5151,
"step": 5590
},
{
"epoch": 0.5907172995780591,
"grad_norm": 0.5617148876190186,
"learning_rate": 0.0015,
"loss": 1.5074,
"step": 5600
},
{
"epoch": 0.5917721518987342,
"grad_norm": 0.6354098916053772,
"learning_rate": 0.0015,
"loss": 1.5117,
"step": 5610
},
{
"epoch": 0.5928270042194093,
"grad_norm": 0.6242541670799255,
"learning_rate": 0.0015,
"loss": 1.5189,
"step": 5620
},
{
"epoch": 0.5938818565400844,
"grad_norm": 0.6715850234031677,
"learning_rate": 0.0015,
"loss": 1.5157,
"step": 5630
},
{
"epoch": 0.5949367088607594,
"grad_norm": 0.5890056490898132,
"learning_rate": 0.0015,
"loss": 1.5265,
"step": 5640
},
{
"epoch": 0.5959915611814346,
"grad_norm": 0.6360265612602234,
"learning_rate": 0.0015,
"loss": 1.52,
"step": 5650
},
{
"epoch": 0.5970464135021097,
"grad_norm": 0.6696105599403381,
"learning_rate": 0.0015,
"loss": 1.5175,
"step": 5660
},
{
"epoch": 0.5981012658227848,
"grad_norm": 0.5383630990982056,
"learning_rate": 0.0015,
"loss": 1.5218,
"step": 5670
},
{
"epoch": 0.5991561181434599,
"grad_norm": 0.7002443671226501,
"learning_rate": 0.0015,
"loss": 1.5228,
"step": 5680
},
{
"epoch": 0.6002109704641351,
"grad_norm": 0.5610836744308472,
"learning_rate": 0.0015,
"loss": 1.5001,
"step": 5690
},
{
"epoch": 0.6012658227848101,
"grad_norm": 0.5491235852241516,
"learning_rate": 0.0015,
"loss": 1.5048,
"step": 5700
},
{
"epoch": 0.6023206751054853,
"grad_norm": 0.5745838284492493,
"learning_rate": 0.0015,
"loss": 1.5103,
"step": 5710
},
{
"epoch": 0.6033755274261603,
"grad_norm": 0.5973967909812927,
"learning_rate": 0.0015,
"loss": 1.5142,
"step": 5720
},
{
"epoch": 0.6044303797468354,
"grad_norm": 0.7536993622779846,
"learning_rate": 0.0015,
"loss": 1.5092,
"step": 5730
},
{
"epoch": 0.6054852320675106,
"grad_norm": 0.5494557619094849,
"learning_rate": 0.0015,
"loss": 1.5117,
"step": 5740
},
{
"epoch": 0.6065400843881856,
"grad_norm": 0.8905473947525024,
"learning_rate": 0.0015,
"loss": 1.4849,
"step": 5750
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.5961906313896179,
"learning_rate": 0.0015,
"loss": 1.4959,
"step": 5760
},
{
"epoch": 0.6086497890295358,
"grad_norm": 0.7787330746650696,
"learning_rate": 0.0015,
"loss": 1.5001,
"step": 5770
},
{
"epoch": 0.609704641350211,
"grad_norm": 0.7014888525009155,
"learning_rate": 0.0015,
"loss": 1.517,
"step": 5780
},
{
"epoch": 0.6107594936708861,
"grad_norm": 0.7262461185455322,
"learning_rate": 0.0015,
"loss": 1.5071,
"step": 5790
},
{
"epoch": 0.6118143459915611,
"grad_norm": 0.550659716129303,
"learning_rate": 0.0015,
"loss": 1.5077,
"step": 5800
},
{
"epoch": 0.6128691983122363,
"grad_norm": 0.5086367726325989,
"learning_rate": 0.0015,
"loss": 1.5111,
"step": 5810
},
{
"epoch": 0.6139240506329114,
"grad_norm": 0.881079375743866,
"learning_rate": 0.0015,
"loss": 1.5241,
"step": 5820
},
{
"epoch": 0.6149789029535865,
"grad_norm": 0.6603014469146729,
"learning_rate": 0.0015,
"loss": 1.5025,
"step": 5830
},
{
"epoch": 0.6160337552742616,
"grad_norm": 0.5636839866638184,
"learning_rate": 0.0015,
"loss": 1.5192,
"step": 5840
},
{
"epoch": 0.6170886075949367,
"grad_norm": 0.5623931884765625,
"learning_rate": 0.0015,
"loss": 1.5224,
"step": 5850
},
{
"epoch": 0.6181434599156118,
"grad_norm": 0.5545900464057922,
"learning_rate": 0.0015,
"loss": 1.5094,
"step": 5860
},
{
"epoch": 0.619198312236287,
"grad_norm": 0.5038663744926453,
"learning_rate": 0.0015,
"loss": 1.5034,
"step": 5870
},
{
"epoch": 0.620253164556962,
"grad_norm": 0.634153425693512,
"learning_rate": 0.0015,
"loss": 1.5147,
"step": 5880
},
{
"epoch": 0.6213080168776371,
"grad_norm": 0.7164548635482788,
"learning_rate": 0.0015,
"loss": 1.506,
"step": 5890
},
{
"epoch": 0.6223628691983122,
"grad_norm": 0.6937264800071716,
"learning_rate": 0.0015,
"loss": 1.5062,
"step": 5900
},
{
"epoch": 0.6234177215189873,
"grad_norm": 0.5849927663803101,
"learning_rate": 0.0015,
"loss": 1.5028,
"step": 5910
},
{
"epoch": 0.6244725738396625,
"grad_norm": 0.6183120608329773,
"learning_rate": 0.0015,
"loss": 1.4944,
"step": 5920
},
{
"epoch": 0.6255274261603375,
"grad_norm": 0.6934881210327148,
"learning_rate": 0.0015,
"loss": 1.4992,
"step": 5930
},
{
"epoch": 0.6265822784810127,
"grad_norm": 0.5181916356086731,
"learning_rate": 0.0015,
"loss": 1.5054,
"step": 5940
},
{
"epoch": 0.6276371308016878,
"grad_norm": 0.5753829479217529,
"learning_rate": 0.0015,
"loss": 1.4961,
"step": 5950
},
{
"epoch": 0.6286919831223629,
"grad_norm": 0.5150362849235535,
"learning_rate": 0.0015,
"loss": 1.5038,
"step": 5960
},
{
"epoch": 0.629746835443038,
"grad_norm": 0.5351850390434265,
"learning_rate": 0.0015,
"loss": 1.5054,
"step": 5970
},
{
"epoch": 0.630801687763713,
"grad_norm": 0.6032434701919556,
"learning_rate": 0.0015,
"loss": 1.4895,
"step": 5980
},
{
"epoch": 0.6318565400843882,
"grad_norm": 0.6674609780311584,
"learning_rate": 0.0015,
"loss": 1.5079,
"step": 5990
},
{
"epoch": 0.6329113924050633,
"grad_norm": 0.6922264099121094,
"learning_rate": 0.0015,
"loss": 1.5173,
"step": 6000
},
{
"epoch": 0.6339662447257384,
"grad_norm": 0.7690093517303467,
"learning_rate": 0.0015,
"loss": 1.5061,
"step": 6010
},
{
"epoch": 0.6350210970464135,
"grad_norm": 0.859092652797699,
"learning_rate": 0.0015,
"loss": 1.5093,
"step": 6020
},
{
"epoch": 0.6360759493670886,
"grad_norm": 0.5710610151290894,
"learning_rate": 0.0015,
"loss": 1.4942,
"step": 6030
},
{
"epoch": 0.6371308016877637,
"grad_norm": 0.8480361104011536,
"learning_rate": 0.0015,
"loss": 1.511,
"step": 6040
},
{
"epoch": 0.6381856540084389,
"grad_norm": 0.5757660269737244,
"learning_rate": 0.0015,
"loss": 1.491,
"step": 6050
},
{
"epoch": 0.6392405063291139,
"grad_norm": 0.5221954584121704,
"learning_rate": 0.0015,
"loss": 1.5079,
"step": 6060
},
{
"epoch": 0.640295358649789,
"grad_norm": 0.5378153920173645,
"learning_rate": 0.0015,
"loss": 1.4992,
"step": 6070
},
{
"epoch": 0.6413502109704642,
"grad_norm": 0.5231079459190369,
"learning_rate": 0.0015,
"loss": 1.5003,
"step": 6080
},
{
"epoch": 0.6424050632911392,
"grad_norm": 0.510814368724823,
"learning_rate": 0.0015,
"loss": 1.4984,
"step": 6090
},
{
"epoch": 0.6434599156118144,
"grad_norm": 0.651321530342102,
"learning_rate": 0.0015,
"loss": 1.4928,
"step": 6100
},
{
"epoch": 0.6445147679324894,
"grad_norm": 0.621930718421936,
"learning_rate": 0.0015,
"loss": 1.5054,
"step": 6110
},
{
"epoch": 0.6455696202531646,
"grad_norm": 0.9920527935028076,
"learning_rate": 0.0015,
"loss": 1.5025,
"step": 6120
},
{
"epoch": 0.6466244725738397,
"grad_norm": 0.4889945387840271,
"learning_rate": 0.0015,
"loss": 1.4995,
"step": 6130
},
{
"epoch": 0.6476793248945147,
"grad_norm": 0.5933585166931152,
"learning_rate": 0.0015,
"loss": 1.497,
"step": 6140
},
{
"epoch": 0.6487341772151899,
"grad_norm": 0.5071729421615601,
"learning_rate": 0.0015,
"loss": 1.5068,
"step": 6150
},
{
"epoch": 0.6497890295358649,
"grad_norm": 0.5711946487426758,
"learning_rate": 0.0015,
"loss": 1.5037,
"step": 6160
},
{
"epoch": 0.6508438818565401,
"grad_norm": 0.8071870803833008,
"learning_rate": 0.0014834368975312174,
"loss": 1.4822,
"step": 6170
},
{
"epoch": 0.6518987341772152,
"grad_norm": 0.8835534453392029,
"learning_rate": 0.0014629899726345957,
"loss": 1.5064,
"step": 6180
},
{
"epoch": 0.6529535864978903,
"grad_norm": 0.5751279592514038,
"learning_rate": 0.0014428248775471316,
"loss": 1.5064,
"step": 6190
},
{
"epoch": 0.6540084388185654,
"grad_norm": 0.5673179626464844,
"learning_rate": 0.00142293772767289,
"loss": 1.4965,
"step": 6200
},
{
"epoch": 0.6550632911392406,
"grad_norm": 0.6210446357727051,
"learning_rate": 0.001403324691959192,
"loss": 1.4913,
"step": 6210
},
{
"epoch": 0.6561181434599156,
"grad_norm": 0.6187479496002197,
"learning_rate": 0.0013839819921586025,
"loss": 1.4976,
"step": 6220
},
{
"epoch": 0.6571729957805907,
"grad_norm": 0.5539543032646179,
"learning_rate": 0.0013649059021010894,
"loss": 1.486,
"step": 6230
},
{
"epoch": 0.6582278481012658,
"grad_norm": 0.5763696432113647,
"learning_rate": 0.0013460927469762154,
"loss": 1.4909,
"step": 6240
},
{
"epoch": 0.6592827004219409,
"grad_norm": 0.5042856931686401,
"learning_rate": 0.0013275389026252255,
"loss": 1.4913,
"step": 6250
},
{
"epoch": 0.6603375527426161,
"grad_norm": 0.6170564293861389,
"learning_rate": 0.0013092407948428887,
"loss": 1.4834,
"step": 6260
},
{
"epoch": 0.6613924050632911,
"grad_norm": 0.5373573303222656,
"learning_rate": 0.001291194898688966,
"loss": 1.4911,
"step": 6270
},
{
"epoch": 0.6624472573839663,
"grad_norm": 0.7562010288238525,
"learning_rate": 0.001273397737809166,
"loss": 1.4895,
"step": 6280
},
{
"epoch": 0.6635021097046413,
"grad_norm": 0.6941890716552734,
"learning_rate": 0.001255845883765463,
"loss": 1.485,
"step": 6290
},
{
"epoch": 0.6645569620253164,
"grad_norm": 0.5045270919799805,
"learning_rate": 0.001238535955375642,
"loss": 1.4783,
"step": 6300
},
{
"epoch": 0.6656118143459916,
"grad_norm": 0.4871998727321625,
"learning_rate": 0.0012214646180619506,
"loss": 1.4748,
"step": 6310
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.5172607898712158,
"learning_rate": 0.001204628583208727,
"loss": 1.4706,
"step": 6320
},
{
"epoch": 0.6677215189873418,
"grad_norm": 0.5898618102073669,
"learning_rate": 0.0011880246075288827,
"loss": 1.4782,
"step": 6330
},
{
"epoch": 0.6687763713080169,
"grad_norm": 0.5182415843009949,
"learning_rate": 0.001171649492439115,
"loss": 1.4702,
"step": 6340
},
{
"epoch": 0.669831223628692,
"grad_norm": 0.49977177381515503,
"learning_rate": 0.0011555000834437364,
"loss": 1.4745,
"step": 6350
},
{
"epoch": 0.6708860759493671,
"grad_norm": 0.8687050938606262,
"learning_rate": 0.0011395732695269908,
"loss": 1.4714,
"step": 6360
},
{
"epoch": 0.6719409282700421,
"grad_norm": 0.5547814965248108,
"learning_rate": 0.0011238659825537505,
"loss": 1.4521,
"step": 6370
},
{
"epoch": 0.6729957805907173,
"grad_norm": 0.6606083512306213,
"learning_rate": 0.0011083751966784717,
"loss": 1.4589,
"step": 6380
},
{
"epoch": 0.6740506329113924,
"grad_norm": 0.6311693787574768,
"learning_rate": 0.0010930979277622953,
"loss": 1.4735,
"step": 6390
},
{
"epoch": 0.6751054852320675,
"grad_norm": 0.5820292234420776,
"learning_rate": 0.0010780312327981854,
"loss": 1.4718,
"step": 6400
},
{
"epoch": 0.6761603375527426,
"grad_norm": 0.8019455671310425,
"learning_rate": 0.0010631722093439888,
"loss": 1.4671,
"step": 6410
},
{
"epoch": 0.6772151898734177,
"grad_norm": 0.761712908744812,
"learning_rate": 0.00104851799496331,
"loss": 1.4531,
"step": 6420
},
{
"epoch": 0.6782700421940928,
"grad_norm": 0.514397144317627,
"learning_rate": 0.0010340657666740914,
"loss": 1.4596,
"step": 6430
},
{
"epoch": 0.679324894514768,
"grad_norm": 0.49260208010673523,
"learning_rate": 0.0010198127404047975,
"loss": 1.4427,
"step": 6440
},
{
"epoch": 0.680379746835443,
"grad_norm": 0.5700627565383911,
"learning_rate": 0.0010057561704580897,
"loss": 1.4506,
"step": 6450
},
{
"epoch": 0.6814345991561181,
"grad_norm": 0.4923916459083557,
"learning_rate": 0.0009918933489818985,
"loss": 1.4626,
"step": 6460
},
{
"epoch": 0.6824894514767933,
"grad_norm": 0.6934185028076172,
"learning_rate": 0.0009782216054477827,
"loss": 1.4533,
"step": 6470
},
{
"epoch": 0.6835443037974683,
"grad_norm": 0.8038368821144104,
"learning_rate": 0.0009647383061364801,
"loss": 1.4629,
"step": 6480
},
{
"epoch": 0.6845991561181435,
"grad_norm": 0.5632407665252686,
"learning_rate": 0.0009514408536305495,
"loss": 1.4534,
"step": 6490
},
{
"epoch": 0.6856540084388185,
"grad_norm": 0.6466336250305176,
"learning_rate": 0.0009383266863140042,
"loss": 1.4668,
"step": 6500
},
{
"epoch": 0.6867088607594937,
"grad_norm": 0.6142919659614563,
"learning_rate": 0.000925393277878844,
"loss": 1.4675,
"step": 6510
},
{
"epoch": 0.6877637130801688,
"grad_norm": 0.5444507598876953,
"learning_rate": 0.0009126381368383879,
"loss": 1.4484,
"step": 6520
},
{
"epoch": 0.6888185654008439,
"grad_norm": 0.5479179620742798,
"learning_rate": 0.0009000588060473156,
"loss": 1.438,
"step": 6530
},
{
"epoch": 0.689873417721519,
"grad_norm": 0.6018847823143005,
"learning_rate": 0.0008876528622283235,
"loss": 1.4504,
"step": 6540
},
{
"epoch": 0.6909282700421941,
"grad_norm": 0.5311291217803955,
"learning_rate": 0.0008754179155053053,
"loss": 1.4477,
"step": 6550
},
{
"epoch": 0.6919831223628692,
"grad_norm": 0.5240741968154907,
"learning_rate": 0.0008633516089429683,
"loss": 1.4453,
"step": 6560
},
{
"epoch": 0.6930379746835443,
"grad_norm": 0.4696906805038452,
"learning_rate": 0.0008514516180927928,
"loss": 1.4399,
"step": 6570
},
{
"epoch": 0.6940928270042194,
"grad_norm": 0.6514448523521423,
"learning_rate": 0.0008397156505452524,
"loss": 1.4389,
"step": 6580
},
{
"epoch": 0.6951476793248945,
"grad_norm": 0.49357330799102783,
"learning_rate": 0.0008281414454882051,
"loss": 1.4449,
"step": 6590
},
{
"epoch": 0.6962025316455697,
"grad_norm": 0.5959960222244263,
"learning_rate": 0.0008167267732713704,
"loss": 1.4476,
"step": 6600
},
{
"epoch": 0.6972573839662447,
"grad_norm": 0.5163862109184265,
"learning_rate": 0.0008054694349768117,
"loss": 1.4315,
"step": 6610
},
{
"epoch": 0.6983122362869199,
"grad_norm": 0.5023782849311829,
"learning_rate": 0.0007943672619953359,
"loss": 1.4392,
"step": 6620
},
{
"epoch": 0.6993670886075949,
"grad_norm": 0.5141021609306335,
"learning_rate": 0.0007834181156087356,
"loss": 1.4311,
"step": 6630
},
{
"epoch": 0.70042194092827,
"grad_norm": 0.5499473214149475,
"learning_rate": 0.0007726198865777852,
"loss": 1.4371,
"step": 6640
},
{
"epoch": 0.7014767932489452,
"grad_norm": 0.6825447678565979,
"learning_rate": 0.0007619704947359191,
"loss": 1.4288,
"step": 6650
},
{
"epoch": 0.7025316455696202,
"grad_norm": 0.47925615310668945,
"learning_rate": 0.0007514678885885087,
"loss": 1.4312,
"step": 6660
},
{
"epoch": 0.7035864978902954,
"grad_norm": 0.48817384243011475,
"learning_rate": 0.0007411100449176633,
"loss": 1.4313,
"step": 6670
},
{
"epoch": 0.7046413502109705,
"grad_norm": 0.5127274990081787,
"learning_rate": 0.0007308949683924791,
"loss": 1.4353,
"step": 6680
},
{
"epoch": 0.7056962025316456,
"grad_norm": 0.6060632467269897,
"learning_rate": 0.000720820691184658,
"loss": 1.4261,
"step": 6690
},
{
"epoch": 0.7067510548523207,
"grad_norm": 0.553632378578186,
"learning_rate": 0.0007108852725894269,
"loss": 1.4253,
"step": 6700
},
{
"epoch": 0.7078059071729957,
"grad_norm": 0.4965393543243408,
"learning_rate": 0.000701086798651681,
"loss": 1.4279,
"step": 6710
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.539612889289856,
"learning_rate": 0.0006914233817972798,
"loss": 1.4161,
"step": 6720
},
{
"epoch": 0.709915611814346,
"grad_norm": 0.5040695071220398,
"learning_rate": 0.0006818931604694261,
"loss": 1.4232,
"step": 6730
},
{
"epoch": 0.7109704641350211,
"grad_norm": 0.4720032215118408,
"learning_rate": 0.0006724942987700563,
"loss": 1.4277,
"step": 6740
},
{
"epoch": 0.7120253164556962,
"grad_norm": 0.5587788820266724,
"learning_rate": 0.0006632249861061732,
"loss": 1.4281,
"step": 6750
},
{
"epoch": 0.7130801687763713,
"grad_norm": 0.5153480172157288,
"learning_rate": 0.0006540834368410549,
"loss": 1.4242,
"step": 6760
},
{
"epoch": 0.7141350210970464,
"grad_norm": 0.484110027551651,
"learning_rate": 0.0006450678899502701,
"loss": 1.4261,
"step": 6770
},
{
"epoch": 0.7151898734177216,
"grad_norm": 0.5298638343811035,
"learning_rate": 0.0006361766086824345,
"loss": 1.4218,
"step": 6780
},
{
"epoch": 0.7162447257383966,
"grad_norm": 0.6543071269989014,
"learning_rate": 0.000627407880224645,
"loss": 1.4302,
"step": 6790
},
{
"epoch": 0.7172995780590717,
"grad_norm": 0.5062084794044495,
"learning_rate": 0.0006187600153725225,
"loss": 1.4139,
"step": 6800
},
{
"epoch": 0.7183544303797469,
"grad_norm": 0.6418647766113281,
"learning_rate": 0.0006102313482048055,
"loss": 1.4156,
"step": 6810
},
{
"epoch": 0.7194092827004219,
"grad_norm": 0.5333866477012634,
"learning_rate": 0.0006018202357624274,
"loss": 1.4207,
"step": 6820
},
{
"epoch": 0.7204641350210971,
"grad_norm": 0.6157364249229431,
"learning_rate": 0.0005935250577320168,
"loss": 1.4154,
"step": 6830
},
{
"epoch": 0.7215189873417721,
"grad_norm": 0.49300333857536316,
"learning_rate": 0.0005853442161337618,
"loss": 1.4051,
"step": 6840
},
{
"epoch": 0.7225738396624473,
"grad_norm": 0.5593222975730896,
"learning_rate": 0.0005772761350135759,
"loss": 1.4136,
"step": 6850
},
{
"epoch": 0.7236286919831224,
"grad_norm": 0.536490797996521,
"learning_rate": 0.0005693192601395058,
"loss": 1.4077,
"step": 6860
},
{
"epoch": 0.7246835443037974,
"grad_norm": 0.526736855506897,
"learning_rate": 0.000561472058702326,
"loss": 1.4052,
"step": 6870
},
{
"epoch": 0.7257383966244726,
"grad_norm": 0.46280238032341003,
"learning_rate": 0.000553733019020258,
"loss": 1.4118,
"step": 6880
},
{
"epoch": 0.7267932489451476,
"grad_norm": 0.5084342956542969,
"learning_rate": 0.0005461006502477612,
"loss": 1.4011,
"step": 6890
},
{
"epoch": 0.7278481012658228,
"grad_norm": 0.5471343398094177,
"learning_rate": 0.0005385734820883369,
"loss": 1.4024,
"step": 6900
},
{
"epoch": 0.7289029535864979,
"grad_norm": 0.512230634689331,
"learning_rate": 0.0005311500645112907,
"loss": 1.4292,
"step": 6910
},
{
"epoch": 0.729957805907173,
"grad_norm": 0.5206472277641296,
"learning_rate": 0.0005238289674723993,
"loss": 1.4066,
"step": 6920
},
{
"epoch": 0.7310126582278481,
"grad_norm": 0.5572535991668701,
"learning_rate": 0.0005166087806384274,
"loss": 1.4115,
"step": 6930
},
{
"epoch": 0.7320675105485233,
"grad_norm": 0.5156189203262329,
"learning_rate": 0.0005094881131154418,
"loss": 1.4123,
"step": 6940
},
{
"epoch": 0.7331223628691983,
"grad_norm": 0.503831684589386,
"learning_rate": 0.0005024655931808696,
"loss": 1.41,
"step": 6950
},
{
"epoch": 0.7341772151898734,
"grad_norm": 0.47319483757019043,
"learning_rate": 0.0004955398680192508,
"loss": 1.401,
"step": 6960
},
{
"epoch": 0.7352320675105485,
"grad_norm": 0.4656444489955902,
"learning_rate": 0.000488709603461632,
"loss": 1.3946,
"step": 6970
},
{
"epoch": 0.7362869198312236,
"grad_norm": 0.4967908561229706,
"learning_rate": 0.000481973483728553,
"loss": 1.3971,
"step": 6980
},
{
"epoch": 0.7373417721518988,
"grad_norm": 0.5195211172103882,
"learning_rate": 0.0004753302111765748,
"loss": 1.3981,
"step": 6990
},
{
"epoch": 0.7383966244725738,
"grad_norm": 0.4845932424068451,
"learning_rate": 0.0004687785060483032,
"loss": 1.4101,
"step": 7000
},
{
"epoch": 0.739451476793249,
"grad_norm": 0.4656439423561096,
"learning_rate": 0.0004623171062258558,
"loss": 1.3801,
"step": 7010
},
{
"epoch": 0.740506329113924,
"grad_norm": 0.6125307083129883,
"learning_rate": 0.0004559447669877288,
"loss": 1.3971,
"step": 7020
},
{
"epoch": 0.7415611814345991,
"grad_norm": 0.5421500205993652,
"learning_rate": 0.00044966026076901413,
"loss": 1.3991,
"step": 7030
},
{
"epoch": 0.7426160337552743,
"grad_norm": 0.6151002049446106,
"learning_rate": 0.00044346237692492177,
"loss": 1.4056,
"step": 7040
},
{
"epoch": 0.7436708860759493,
"grad_norm": 0.5110399723052979,
"learning_rate": 0.0004373499214975615,
"loss": 1.392,
"step": 7050
},
{
"epoch": 0.7447257383966245,
"grad_norm": 0.4953923225402832,
"learning_rate": 0.0004313217169859396,
"loss": 1.3992,
"step": 7060
},
{
"epoch": 0.7457805907172996,
"grad_norm": 0.5136517286300659,
"learning_rate": 0.0004253766021191256,
"loss": 1.4037,
"step": 7070
},
{
"epoch": 0.7468354430379747,
"grad_norm": 0.5029652118682861,
"learning_rate": 0.00041951343163254497,
"loss": 1.399,
"step": 7080
},
{
"epoch": 0.7478902953586498,
"grad_norm": 0.4731636643409729,
"learning_rate": 0.00041373107604735626,
"loss": 1.3959,
"step": 7090
},
{
"epoch": 0.7489451476793249,
"grad_norm": 0.5364992022514343,
"learning_rate": 0.0004080284214528687,
"loss": 1.39,
"step": 7100
},
{
"epoch": 0.75,
"grad_norm": 0.5382775664329529,
"learning_rate": 0.0004024043692919589,
"loss": 1.4037,
"step": 7110
},
{
"epoch": 0.7510548523206751,
"grad_norm": 0.5207314491271973,
"learning_rate": 0.0003968578361494449,
"loss": 1.3999,
"step": 7120
},
{
"epoch": 0.7521097046413502,
"grad_norm": 0.4977900981903076,
"learning_rate": 0.000391387753543378,
"loss": 1.4073,
"step": 7130
},
{
"epoch": 0.7531645569620253,
"grad_norm": 0.5196287631988525,
"learning_rate": 0.00038599306771921023,
"loss": 1.3891,
"step": 7140
},
{
"epoch": 0.7542194092827004,
"grad_norm": 0.6989203095436096,
"learning_rate": 0.0003806727394468004,
"loss": 1.3864,
"step": 7150
},
{
"epoch": 0.7552742616033755,
"grad_norm": 0.5948571562767029,
"learning_rate": 0.0003754257438202162,
"loss": 1.3941,
"step": 7160
},
{
"epoch": 0.7563291139240507,
"grad_norm": 0.6483874320983887,
"learning_rate": 0.0003702510700602974,
"loss": 1.3999,
"step": 7170
},
{
"epoch": 0.7573839662447257,
"grad_norm": 0.5051309466362,
"learning_rate": 0.0003651477213199393,
"loss": 1.3803,
"step": 7180
},
{
"epoch": 0.7584388185654009,
"grad_norm": 0.506965696811676,
"learning_rate": 0.000360114714492061,
"loss": 1.3813,
"step": 7190
},
{
"epoch": 0.759493670886076,
"grad_norm": 0.4690167009830475,
"learning_rate": 0.0003551510800202195,
"loss": 1.3839,
"step": 7200
},
{
"epoch": 0.760548523206751,
"grad_norm": 0.5723272562026978,
"learning_rate": 0.0003502558617118353,
"loss": 1.3915,
"step": 7210
},
{
"epoch": 0.7616033755274262,
"grad_norm": 0.5610151290893555,
"learning_rate": 0.0003454281165539914,
"loss": 1.4038,
"step": 7220
},
{
"epoch": 0.7626582278481012,
"grad_norm": 0.4634474813938141,
"learning_rate": 0.00034066691453177176,
"loss": 1.4008,
"step": 7230
},
{
"epoch": 0.7637130801687764,
"grad_norm": 0.47573161125183105,
"learning_rate": 0.0003359713384491037,
"loss": 1.397,
"step": 7240
},
{
"epoch": 0.7647679324894515,
"grad_norm": 0.5212197303771973,
"learning_rate": 0.00033134048375206944,
"loss": 1.389,
"step": 7250
},
{
"epoch": 0.7658227848101266,
"grad_norm": 0.49535438418388367,
"learning_rate": 0.0003267734583546536,
"loss": 1.3859,
"step": 7260
},
{
"epoch": 0.7668776371308017,
"grad_norm": 0.5158471465110779,
"learning_rate": 0.00032226938246689157,
"loss": 1.3852,
"step": 7270
},
{
"epoch": 0.7679324894514767,
"grad_norm": 0.5728384852409363,
"learning_rate": 0.0003178273884253874,
"loss": 1.3856,
"step": 7280
},
{
"epoch": 0.7689873417721519,
"grad_norm": 0.5211251378059387,
"learning_rate": 0.0003134466205261674,
"loss": 1.3971,
"step": 7290
},
{
"epoch": 0.770042194092827,
"grad_norm": 0.5732220411300659,
"learning_rate": 0.0003091262348598378,
"loss": 1.4041,
"step": 7300
},
{
"epoch": 0.7710970464135021,
"grad_norm": 0.47373560070991516,
"learning_rate": 0.0003048653991490141,
"loss": 1.3822,
"step": 7310
},
{
"epoch": 0.7721518987341772,
"grad_norm": 0.6221377849578857,
"learning_rate": 0.00030066329258799187,
"loss": 1.3776,
"step": 7320
},
{
"epoch": 0.7732067510548524,
"grad_norm": 0.5708879828453064,
"learning_rate": 0.0002965191056846266,
"loss": 1.3808,
"step": 7330
},
{
"epoch": 0.7742616033755274,
"grad_norm": 0.603161633014679,
"learning_rate": 0.000292432040104394,
"loss": 1.3802,
"step": 7340
},
{
"epoch": 0.7753164556962026,
"grad_norm": 0.5277611613273621,
"learning_rate": 0.00028840130851659853,
"loss": 1.375,
"step": 7350
},
{
"epoch": 0.7763713080168776,
"grad_norm": 0.5325142741203308,
"learning_rate": 0.0002844261344427028,
"loss": 1.3831,
"step": 7360
},
{
"epoch": 0.7774261603375527,
"grad_norm": 0.5006171464920044,
"learning_rate": 0.0002805057521067471,
"loss": 1.3836,
"step": 7370
},
{
"epoch": 0.7784810126582279,
"grad_norm": 0.48984476923942566,
"learning_rate": 0.00027663940628783017,
"loss": 1.3707,
"step": 7380
},
{
"epoch": 0.7795358649789029,
"grad_norm": 0.4523366093635559,
"learning_rate": 0.00027282635217462393,
"loss": 1.3811,
"step": 7390
},
{
"epoch": 0.7805907172995781,
"grad_norm": 0.6498004198074341,
"learning_rate": 0.0002690658552218937,
"loss": 1.389,
"step": 7400
},
{
"epoch": 0.7816455696202531,
"grad_norm": 0.5409725308418274,
"learning_rate": 0.00026535719100899516,
"loss": 1.3711,
"step": 7410
},
{
"epoch": 0.7827004219409283,
"grad_norm": 0.5616211891174316,
"learning_rate": 0.00026169964510032245,
"loss": 1.3747,
"step": 7420
},
{
"epoch": 0.7837552742616034,
"grad_norm": 0.487123042345047,
"learning_rate": 0.00025809251290767984,
"loss": 1.3619,
"step": 7430
},
{
"epoch": 0.7848101265822784,
"grad_norm": 0.4605257511138916,
"learning_rate": 0.00025453509955454957,
"loss": 1.3643,
"step": 7440
},
{
"epoch": 0.7858649789029536,
"grad_norm": 0.5088199973106384,
"learning_rate": 0.00025102671974223175,
"loss": 1.3698,
"step": 7450
},
{
"epoch": 0.7869198312236287,
"grad_norm": 0.4879826307296753,
"learning_rate": 0.00024756669761782815,
"loss": 1.3797,
"step": 7460
},
{
"epoch": 0.7879746835443038,
"grad_norm": 0.48447489738464355,
"learning_rate": 0.0002441543666440464,
"loss": 1.368,
"step": 7470
},
{
"epoch": 0.7890295358649789,
"grad_norm": 0.5670198798179626,
"learning_rate": 0.00024078906947079878,
"loss": 1.3742,
"step": 7480
},
{
"epoch": 0.790084388185654,
"grad_norm": 0.4819045066833496,
"learning_rate": 0.00023747015780857005,
"loss": 1.383,
"step": 7490
},
{
"epoch": 0.7911392405063291,
"grad_norm": 0.49387362599372864,
"learning_rate": 0.00023419699230353144,
"loss": 1.3774,
"step": 7500
},
{
"epoch": 0.7921940928270043,
"grad_norm": 0.5370501279830933,
"learning_rate": 0.00023096894241437586,
"loss": 1.3828,
"step": 7510
},
{
"epoch": 0.7932489451476793,
"grad_norm": 0.47914108633995056,
"learning_rate": 0.00022778538629085056,
"loss": 1.3672,
"step": 7520
},
{
"epoch": 0.7943037974683544,
"grad_norm": 0.5119479298591614,
"learning_rate": 0.00022464571065396427,
"loss": 1.3694,
"step": 7530
},
{
"epoch": 0.7953586497890295,
"grad_norm": 0.5128620266914368,
"learning_rate": 0.00022154931067784521,
"loss": 1.3666,
"step": 7540
},
{
"epoch": 0.7964135021097046,
"grad_norm": 0.46946507692337036,
"learning_rate": 0.00021849558987322782,
"loss": 1.3642,
"step": 7550
},
{
"epoch": 0.7974683544303798,
"grad_norm": 0.5166003704071045,
"learning_rate": 0.0002154839599725452,
"loss": 1.365,
"step": 7560
},
{
"epoch": 0.7985232067510548,
"grad_norm": 0.5451986789703369,
"learning_rate": 0.00021251384081660544,
"loss": 1.3778,
"step": 7570
},
{
"epoch": 0.79957805907173,
"grad_norm": 0.4987885057926178,
"learning_rate": 0.0002095846602428303,
"loss": 1.3732,
"step": 7580
},
{
"epoch": 0.8006329113924051,
"grad_norm": 0.4632894694805145,
"learning_rate": 0.00020669585397503358,
"loss": 1.3643,
"step": 7590
},
{
"epoch": 0.8016877637130801,
"grad_norm": 0.4567081928253174,
"learning_rate": 0.0002038468655147195,
"loss": 1.3674,
"step": 7600
},
{
"epoch": 0.8027426160337553,
"grad_norm": 0.49346765875816345,
"learning_rate": 0.00020103714603387894,
"loss": 1.3815,
"step": 7610
},
{
"epoch": 0.8037974683544303,
"grad_norm": 0.45700404047966003,
"learning_rate": 0.00019826615426926338,
"loss": 1.3544,
"step": 7620
},
{
"epoch": 0.8048523206751055,
"grad_norm": 0.4595026671886444,
"learning_rate": 0.00019553335641811625,
"loss": 1.3763,
"step": 7630
},
{
"epoch": 0.8059071729957806,
"grad_norm": 0.5063329935073853,
"learning_rate": 0.0001928382260353415,
"loss": 1.3673,
"step": 7640
},
{
"epoch": 0.8069620253164557,
"grad_norm": 0.47150281071662903,
"learning_rate": 0.00019018024393208902,
"loss": 1.3843,
"step": 7650
},
{
"epoch": 0.8080168776371308,
"grad_norm": 0.4635440409183502,
"learning_rate": 0.00018755889807573872,
"loss": 1.3615,
"step": 7660
},
{
"epoch": 0.8090717299578059,
"grad_norm": 0.45574668049812317,
"learning_rate": 0.00018497368349126262,
"loss": 1.3672,
"step": 7670
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.5359010696411133,
"learning_rate": 0.00018242410216394648,
"loss": 1.383,
"step": 7680
},
{
"epoch": 0.8111814345991561,
"grad_norm": 0.5147023797035217,
"learning_rate": 0.0001799096629434529,
"loss": 1.3556,
"step": 7690
},
{
"epoch": 0.8122362869198312,
"grad_norm": 0.46002885699272156,
"learning_rate": 0.00017742988144920578,
"loss": 1.3666,
"step": 7700
},
{
"epoch": 0.8132911392405063,
"grad_norm": 0.4675118625164032,
"learning_rate": 0.00017498427997707976,
"loss": 1.3633,
"step": 7710
},
{
"epoch": 0.8143459915611815,
"grad_norm": 0.5563496351242065,
"learning_rate": 0.00017257238740737548,
"loss": 1.3648,
"step": 7720
},
{
"epoch": 0.8154008438818565,
"grad_norm": 0.5589823722839355,
"learning_rate": 0.00017019373911406307,
"loss": 1.3735,
"step": 7730
},
{
"epoch": 0.8164556962025317,
"grad_norm": 0.46896493434906006,
"learning_rate": 0.000167847876875277,
"loss": 1.3721,
"step": 7740
},
{
"epoch": 0.8175105485232067,
"grad_norm": 0.45898881554603577,
"learning_rate": 0.00016553434878504428,
"loss": 1.3565,
"step": 7750
},
{
"epoch": 0.8185654008438819,
"grad_norm": 0.5108206868171692,
"learning_rate": 0.00016325270916622947,
"loss": 1.3584,
"step": 7760
},
{
"epoch": 0.819620253164557,
"grad_norm": 0.4540421962738037,
"learning_rate": 0.00016100251848467966,
"loss": 1.3679,
"step": 7770
},
{
"epoch": 0.820675105485232,
"grad_norm": 0.6645315885543823,
"learning_rate": 0.0001587833432645528,
"loss": 1.3602,
"step": 7780
},
{
"epoch": 0.8217299578059072,
"grad_norm": 0.5565598011016846,
"learning_rate": 0.00015659475600481292,
"loss": 1.3754,
"step": 7790
},
{
"epoch": 0.8227848101265823,
"grad_norm": 0.4732491075992584,
"learning_rate": 0.00015443633509687688,
"loss": 1.3637,
"step": 7800
},
{
"epoch": 0.8238396624472574,
"grad_norm": 0.4488430321216583,
"learning_rate": 0.00015230766474339536,
"loss": 1.363,
"step": 7810
},
{
"epoch": 0.8248945147679325,
"grad_norm": 0.4506990313529968,
"learning_rate": 0.00015020833487815416,
"loss": 1.3719,
"step": 7820
},
{
"epoch": 0.8259493670886076,
"grad_norm": 0.46290716528892517,
"learning_rate": 0.0001481379410870792,
"loss": 1.3612,
"step": 7830
},
{
"epoch": 0.8270042194092827,
"grad_norm": 0.47624319791793823,
"learning_rate": 0.00014609608453033013,
"loss": 1.3453,
"step": 7840
},
{
"epoch": 0.8280590717299579,
"grad_norm": 0.45356878638267517,
"learning_rate": 0.00014408237186546807,
"loss": 1.3565,
"step": 7850
},
{
"epoch": 0.8291139240506329,
"grad_norm": 0.46717569231987,
"learning_rate": 0.00014209641517168273,
"loss": 1.3433,
"step": 7860
},
{
"epoch": 0.830168776371308,
"grad_norm": 0.48989081382751465,
"learning_rate": 0.00014013783187506265,
"loss": 1.3578,
"step": 7870
},
{
"epoch": 0.8312236286919831,
"grad_norm": 0.529300332069397,
"learning_rate": 0.00013820624467489697,
"loss": 1.3766,
"step": 7880
},
{
"epoch": 0.8322784810126582,
"grad_norm": 0.47537773847579956,
"learning_rate": 0.00013630128147099213,
"loss": 1.3743,
"step": 7890
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.48015323281288147,
"learning_rate": 0.00013442257529199068,
"loss": 1.3561,
"step": 7900
},
{
"epoch": 0.8343881856540084,
"grad_norm": 0.5237522125244141,
"learning_rate": 0.00013256976422467803,
"loss": 1.3634,
"step": 7910
},
{
"epoch": 0.8354430379746836,
"grad_norm": 0.4523259997367859,
"learning_rate": 0.00013074249134426366,
"loss": 1.3622,
"step": 7920
},
{
"epoch": 0.8364978902953587,
"grad_norm": 0.5076009035110474,
"learning_rate": 0.0001289404046456233,
"loss": 1.3708,
"step": 7930
},
{
"epoch": 0.8375527426160337,
"grad_norm": 0.4758736789226532,
"learning_rate": 0.0001271631569754887,
"loss": 1.3642,
"step": 7940
},
{
"epoch": 0.8386075949367089,
"grad_norm": 0.4746805429458618,
"learning_rate": 0.0001254104059655723,
"loss": 1.3666,
"step": 7950
},
{
"epoch": 0.8396624472573839,
"grad_norm": 0.45755910873413086,
"learning_rate": 0.00012368181396661337,
"loss": 1.3526,
"step": 7960
},
{
"epoch": 0.8407172995780591,
"grad_norm": 0.5004385709762573,
"learning_rate": 0.00012197704798333364,
"loss": 1.3514,
"step": 7970
},
{
"epoch": 0.8417721518987342,
"grad_norm": 0.5192686319351196,
"learning_rate": 0.00012029577961028894,
"loss": 1.3512,
"step": 7980
},
{
"epoch": 0.8428270042194093,
"grad_norm": 0.48216772079467773,
"learning_rate": 0.00011863768496860542,
"loss": 1.3634,
"step": 7990
},
{
"epoch": 0.8438818565400844,
"grad_norm": 0.46114224195480347,
"learning_rate": 0.00011700244464358777,
"loss": 1.3595,
"step": 8000
},
{
"epoch": 0.8449367088607594,
"grad_norm": 0.45775875449180603,
"learning_rate": 0.00011538974362318715,
"loss": 1.3632,
"step": 8010
},
{
"epoch": 0.8459915611814346,
"grad_norm": 0.4667496979236603,
"learning_rate": 0.00011379927123731737,
"loss": 1.3565,
"step": 8020
},
{
"epoch": 0.8470464135021097,
"grad_norm": 0.4608895778656006,
"learning_rate": 0.0001122307210980077,
"loss": 1.3568,
"step": 8030
},
{
"epoch": 0.8481012658227848,
"grad_norm": 0.47877684235572815,
"learning_rate": 0.00011068379104038026,
"loss": 1.3704,
"step": 8040
},
{
"epoch": 0.8491561181434599,
"grad_norm": 0.457285612821579,
"learning_rate": 0.00010915818306444116,
"loss": 1.3466,
"step": 8050
},
{
"epoch": 0.8502109704641351,
"grad_norm": 0.5008485913276672,
"learning_rate": 0.00010765360327767384,
"loss": 1.3548,
"step": 8060
},
{
"epoch": 0.8512658227848101,
"grad_norm": 0.47175008058547974,
"learning_rate": 0.00010616976183842376,
"loss": 1.3633,
"step": 8070
},
{
"epoch": 0.8523206751054853,
"grad_norm": 0.4591042697429657,
"learning_rate": 0.00010470637290006365,
"loss": 1.3648,
"step": 8080
},
{
"epoch": 0.8533755274261603,
"grad_norm": 0.4671987295150757,
"learning_rate": 0.00010326315455592764,
"loss": 1.352,
"step": 8090
},
{
"epoch": 0.8544303797468354,
"grad_norm": 0.4841940104961395,
"learning_rate": 0.0001018398287850053,
"loss": 1.3435,
"step": 8100
},
{
"epoch": 0.8554852320675106,
"grad_norm": 0.4897977113723755,
"learning_rate": 0.00010043612139838357,
"loss": 1.3656,
"step": 8110
},
{
"epoch": 0.8565400843881856,
"grad_norm": 0.4758096933364868,
"learning_rate": 9.905176198642719e-05,
"loss": 1.3553,
"step": 8120
},
{
"epoch": 0.8575949367088608,
"grad_norm": 0.4628910720348358,
"learning_rate": 9.76864838666871e-05,
"loss": 1.3565,
"step": 8130
},
{
"epoch": 0.8586497890295358,
"grad_norm": 0.4522228538990021,
"learning_rate": 9.634002403252676e-05,
"loss": 1.3511,
"step": 8140
},
{
"epoch": 0.859704641350211,
"grad_norm": 0.47021564841270447,
"learning_rate": 9.501212310245681e-05,
"loss": 1.3488,
"step": 8150
},
{
"epoch": 0.8607594936708861,
"grad_norm": 0.49830883741378784,
"learning_rate": 9.370252527016777e-05,
"loss": 1.3601,
"step": 8160
},
{
"epoch": 0.8618143459915611,
"grad_norm": 0.47212862968444824,
"learning_rate": 9.241097825525163e-05,
"loss": 1.3512,
"step": 8170
},
{
"epoch": 0.8628691983122363,
"grad_norm": 0.48565971851348877,
"learning_rate": 9.113723325460276e-05,
"loss": 1.358,
"step": 8180
},
{
"epoch": 0.8639240506329114,
"grad_norm": 0.5443127155303955,
"learning_rate": 8.988104489448849e-05,
"loss": 1.3521,
"step": 8190
},
{
"epoch": 0.8649789029535865,
"grad_norm": 0.4680684506893158,
"learning_rate": 8.864217118328042e-05,
"loss": 1.3639,
"step": 8200
},
{
"epoch": 0.8660337552742616,
"grad_norm": 0.46079784631729126,
"learning_rate": 8.742037346483729e-05,
"loss": 1.3559,
"step": 8210
},
{
"epoch": 0.8670886075949367,
"grad_norm": 0.45251044631004333,
"learning_rate": 8.62154163725303e-05,
"loss": 1.3637,
"step": 8220
},
{
"epoch": 0.8681434599156118,
"grad_norm": 0.5130413770675659,
"learning_rate": 8.502706778390219e-05,
"loss": 1.3585,
"step": 8230
},
{
"epoch": 0.869198312236287,
"grad_norm": 0.4337594509124756,
"learning_rate": 8.38550987759513e-05,
"loss": 1.3566,
"step": 8240
},
{
"epoch": 0.870253164556962,
"grad_norm": 0.5292634963989258,
"learning_rate": 8.269928358103191e-05,
"loss": 1.3709,
"step": 8250
},
{
"epoch": 0.8713080168776371,
"grad_norm": 0.4459943473339081,
"learning_rate": 8.155939954336243e-05,
"loss": 1.3609,
"step": 8260
},
{
"epoch": 0.8723628691983122,
"grad_norm": 0.5037257075309753,
"learning_rate": 8.043522707613312e-05,
"loss": 1.3573,
"step": 8270
},
{
"epoch": 0.8734177215189873,
"grad_norm": 0.4748310446739197,
"learning_rate": 7.932654961920486e-05,
"loss": 1.3421,
"step": 8280
},
{
"epoch": 0.8744725738396625,
"grad_norm": 0.46188482642173767,
"learning_rate": 7.823315359739135e-05,
"loss": 1.3442,
"step": 8290
},
{
"epoch": 0.8755274261603375,
"grad_norm": 0.4867098331451416,
"learning_rate": 7.715482837931577e-05,
"loss": 1.3685,
"step": 8300
},
{
"epoch": 0.8765822784810127,
"grad_norm": 0.4996362328529358,
"learning_rate": 7.6091366236835e-05,
"loss": 1.3469,
"step": 8310
},
{
"epoch": 0.8776371308016878,
"grad_norm": 0.4916406273841858,
"learning_rate": 7.504256230502289e-05,
"loss": 1.3649,
"step": 8320
},
{
"epoch": 0.8786919831223629,
"grad_norm": 0.4626372754573822,
"learning_rate": 7.400821454270524e-05,
"loss": 1.3576,
"step": 8330
},
{
"epoch": 0.879746835443038,
"grad_norm": 0.45033547282218933,
"learning_rate": 7.29881236935386e-05,
"loss": 1.3472,
"step": 8340
},
{
"epoch": 0.880801687763713,
"grad_norm": 0.44288432598114014,
"learning_rate": 7.198209324762562e-05,
"loss": 1.3494,
"step": 8350
},
{
"epoch": 0.8818565400843882,
"grad_norm": 0.49059954285621643,
"learning_rate": 7.098992940365946e-05,
"loss": 1.3443,
"step": 8360
},
{
"epoch": 0.8829113924050633,
"grad_norm": 0.45661666989326477,
"learning_rate": 7.001144103159e-05,
"loss": 1.3537,
"step": 8370
},
{
"epoch": 0.8839662447257384,
"grad_norm": 0.4595849812030792,
"learning_rate": 6.904643963580461e-05,
"loss": 1.3609,
"step": 8380
},
{
"epoch": 0.8850210970464135,
"grad_norm": 0.44966819882392883,
"learning_rate": 6.809473931881644e-05,
"loss": 1.349,
"step": 8390
},
{
"epoch": 0.8860759493670886,
"grad_norm": 0.4483475983142853,
"learning_rate": 6.71561567454532e-05,
"loss": 1.3506,
"step": 8400
},
{
"epoch": 0.8871308016877637,
"grad_norm": 0.5414402484893799,
"learning_rate": 6.623051110753948e-05,
"loss": 1.3686,
"step": 8410
},
{
"epoch": 0.8881856540084389,
"grad_norm": 0.4634023606777191,
"learning_rate": 6.531762408906607e-05,
"loss": 1.3559,
"step": 8420
},
{
"epoch": 0.8892405063291139,
"grad_norm": 0.4461284875869751,
"learning_rate": 6.441731983183912e-05,
"loss": 1.3544,
"step": 8430
},
{
"epoch": 0.890295358649789,
"grad_norm": 0.45708733797073364,
"learning_rate": 6.352942490160292e-05,
"loss": 1.3451,
"step": 8440
},
{
"epoch": 0.8913502109704642,
"grad_norm": 0.4943016469478607,
"learning_rate": 6.265376825462966e-05,
"loss": 1.3544,
"step": 8450
},
{
"epoch": 0.8924050632911392,
"grad_norm": 0.470504492521286,
"learning_rate": 6.179018120476945e-05,
"loss": 1.3537,
"step": 8460
},
{
"epoch": 0.8934599156118144,
"grad_norm": 0.4554164409637451,
"learning_rate": 6.0938497390954946e-05,
"loss": 1.3508,
"step": 8470
},
{
"epoch": 0.8945147679324894,
"grad_norm": 0.453253835439682,
"learning_rate": 6.009855274515339e-05,
"loss": 1.3449,
"step": 8480
},
{
"epoch": 0.8955696202531646,
"grad_norm": 0.48623064160346985,
"learning_rate": 5.9270185460760735e-05,
"loss": 1.354,
"step": 8490
},
{
"epoch": 0.8966244725738397,
"grad_norm": 0.45357802510261536,
"learning_rate": 5.8453235961431225e-05,
"loss": 1.3415,
"step": 8500
},
{
"epoch": 0.8976793248945147,
"grad_norm": 0.4535787105560303,
"learning_rate": 5.764754687033678e-05,
"loss": 1.3523,
"step": 8510
},
{
"epoch": 0.8987341772151899,
"grad_norm": 0.4594683349132538,
"learning_rate": 5.6852962979849836e-05,
"loss": 1.3488,
"step": 8520
},
{
"epoch": 0.8997890295358649,
"grad_norm": 0.49994322657585144,
"learning_rate": 5.6069331221644284e-05,
"loss": 1.3642,
"step": 8530
},
{
"epoch": 0.9008438818565401,
"grad_norm": 0.4636819660663605,
"learning_rate": 5.529650063720842e-05,
"loss": 1.3577,
"step": 8540
},
{
"epoch": 0.9018987341772152,
"grad_norm": 0.4792076647281647,
"learning_rate": 5.453432234876445e-05,
"loss": 1.3498,
"step": 8550
},
{
"epoch": 0.9029535864978903,
"grad_norm": 0.4741894602775574,
"learning_rate": 5.37826495305886e-05,
"loss": 1.3411,
"step": 8560
},
{
"epoch": 0.9040084388185654,
"grad_norm": 0.5298281311988831,
"learning_rate": 5.304133738072674e-05,
"loss": 1.3623,
"step": 8570
},
{
"epoch": 0.9050632911392406,
"grad_norm": 0.5063627362251282,
"learning_rate": 5.2310243093099814e-05,
"loss": 1.3535,
"step": 8580
},
{
"epoch": 0.9061181434599156,
"grad_norm": 0.5597848892211914,
"learning_rate": 5.158922582999368e-05,
"loss": 1.3568,
"step": 8590
},
{
"epoch": 0.9071729957805907,
"grad_norm": 0.48414871096611023,
"learning_rate": 5.087814669492819e-05,
"loss": 1.3481,
"step": 8600
},
{
"epoch": 0.9082278481012658,
"grad_norm": 0.5020228028297424,
"learning_rate": 5.017686870590028e-05,
"loss": 1.3479,
"step": 8610
},
{
"epoch": 0.9092827004219409,
"grad_norm": 0.4592040777206421,
"learning_rate": 4.948525676899577e-05,
"loss": 1.3463,
"step": 8620
},
{
"epoch": 0.9103375527426161,
"grad_norm": 0.47056108713150024,
"learning_rate": 4.880317765236493e-05,
"loss": 1.3561,
"step": 8630
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.47921305894851685,
"learning_rate": 4.8130499960556755e-05,
"loss": 1.3355,
"step": 8640
},
{
"epoch": 0.9124472573839663,
"grad_norm": 0.4555310010910034,
"learning_rate": 4.746709410920699e-05,
"loss": 1.3422,
"step": 8650
},
{
"epoch": 0.9135021097046413,
"grad_norm": 0.5480925440788269,
"learning_rate": 4.681283230007507e-05,
"loss": 1.3452,
"step": 8660
},
{
"epoch": 0.9145569620253164,
"grad_norm": 0.4556386470794678,
"learning_rate": 4.616758849642509e-05,
"loss": 1.3525,
"step": 8670
},
{
"epoch": 0.9156118143459916,
"grad_norm": 0.45376330614089966,
"learning_rate": 4.553123839874615e-05,
"loss": 1.3588,
"step": 8680
},
{
"epoch": 0.9166666666666666,
"grad_norm": 0.48756328225135803,
"learning_rate": 4.490365942080736e-05,
"loss": 1.3523,
"step": 8690
},
{
"epoch": 0.9177215189873418,
"grad_norm": 0.4586666524410248,
"learning_rate": 4.428473066604285e-05,
"loss": 1.3498,
"step": 8700
},
{
"epoch": 0.9187763713080169,
"grad_norm": 0.4539191722869873,
"learning_rate": 4.367433290426233e-05,
"loss": 1.3505,
"step": 8710
},
{
"epoch": 0.919831223628692,
"grad_norm": 0.4817667603492737,
"learning_rate": 4.3072348548682595e-05,
"loss": 1.3431,
"step": 8720
},
{
"epoch": 0.9208860759493671,
"grad_norm": 0.46067652106285095,
"learning_rate": 4.247866163327575e-05,
"loss": 1.3548,
"step": 8730
},
{
"epoch": 0.9219409282700421,
"grad_norm": 0.4616350829601288,
"learning_rate": 4.1893157790429404e-05,
"loss": 1.3469,
"step": 8740
},
{
"epoch": 0.9229957805907173,
"grad_norm": 0.4541439712047577,
"learning_rate": 4.1315724228915066e-05,
"loss": 1.3421,
"step": 8750
},
{
"epoch": 0.9240506329113924,
"grad_norm": 0.5221729874610901,
"learning_rate": 4.074624971216005e-05,
"loss": 1.3293,
"step": 8760
},
{
"epoch": 0.9251054852320675,
"grad_norm": 0.4680933356285095,
"learning_rate": 4.018462453681889e-05,
"loss": 1.342,
"step": 8770
},
{
"epoch": 0.9261603375527426,
"grad_norm": 0.4517577886581421,
"learning_rate": 3.963074051164014e-05,
"loss": 1.351,
"step": 8780
},
{
"epoch": 0.9272151898734177,
"grad_norm": 0.4618259072303772,
"learning_rate": 3.908449093662446e-05,
"loss": 1.3435,
"step": 8790
},
{
"epoch": 0.9282700421940928,
"grad_norm": 0.45023104548454285,
"learning_rate": 3.854577058246998e-05,
"loss": 1.3425,
"step": 8800
},
{
"epoch": 0.929324894514768,
"grad_norm": 0.47846972942352295,
"learning_rate": 3.801447567030094e-05,
"loss": 1.3586,
"step": 8810
},
{
"epoch": 0.930379746835443,
"grad_norm": 0.45514824986457825,
"learning_rate": 3.7490503851675777e-05,
"loss": 1.3471,
"step": 8820
},
{
"epoch": 0.9314345991561181,
"grad_norm": 0.44830751419067383,
"learning_rate": 3.6973754188870806e-05,
"loss": 1.3653,
"step": 8830
},
{
"epoch": 0.9324894514767933,
"grad_norm": 0.465257465839386,
"learning_rate": 3.6464127135435536e-05,
"loss": 1.3491,
"step": 8840
},
{
"epoch": 0.9335443037974683,
"grad_norm": 0.5014391541481018,
"learning_rate": 3.596152451701616e-05,
"loss": 1.3464,
"step": 8850
},
{
"epoch": 0.9345991561181435,
"grad_norm": 0.48519137501716614,
"learning_rate": 3.5465849512443226e-05,
"loss": 1.3463,
"step": 8860
},
{
"epoch": 0.9356540084388185,
"grad_norm": 0.45895054936408997,
"learning_rate": 3.4977006635080086e-05,
"loss": 1.3504,
"step": 8870
},
{
"epoch": 0.9367088607594937,
"grad_norm": 0.4757814407348633,
"learning_rate": 3.449490171442838e-05,
"loss": 1.3539,
"step": 8880
},
{
"epoch": 0.9377637130801688,
"grad_norm": 0.46523842215538025,
"learning_rate": 3.401944187798702e-05,
"loss": 1.3522,
"step": 8890
},
{
"epoch": 0.9388185654008439,
"grad_norm": 0.447443425655365,
"learning_rate": 3.355053553336137e-05,
"loss": 1.3438,
"step": 8900
},
{
"epoch": 0.939873417721519,
"grad_norm": 0.47570958733558655,
"learning_rate": 3.308809235061882e-05,
"loss": 1.3425,
"step": 8910
},
{
"epoch": 0.9409282700421941,
"grad_norm": 0.45039355754852295,
"learning_rate": 3.263202324488772e-05,
"loss": 1.3494,
"step": 8920
},
{
"epoch": 0.9419831223628692,
"grad_norm": 0.45538634061813354,
"learning_rate": 3.218224035919609e-05,
"loss": 1.3456,
"step": 8930
},
{
"epoch": 0.9430379746835443,
"grad_norm": 0.47007620334625244,
"learning_rate": 3.173865704754688e-05,
"loss": 1.3475,
"step": 8940
},
{
"epoch": 0.9440928270042194,
"grad_norm": 0.4494016170501709,
"learning_rate": 3.130118785822657e-05,
"loss": 1.3498,
"step": 8950
},
{
"epoch": 0.9451476793248945,
"grad_norm": 0.46623754501342773,
"learning_rate": 3.08697485173437e-05,
"loss": 1.3525,
"step": 8960
},
{
"epoch": 0.9462025316455697,
"grad_norm": 0.5000026226043701,
"learning_rate": 3.0444255912594442e-05,
"loss": 1.3486,
"step": 8970
},
{
"epoch": 0.9472573839662447,
"grad_norm": 0.47525277733802795,
"learning_rate": 3.002462807725185e-05,
"loss": 1.3442,
"step": 8980
},
{
"epoch": 0.9483122362869199,
"grad_norm": 0.4763094186782837,
"learning_rate": 2.9610784174375868e-05,
"loss": 1.3515,
"step": 8990
},
{
"epoch": 0.9493670886075949,
"grad_norm": 0.45728081464767456,
"learning_rate": 2.920264448124087e-05,
"loss": 1.3449,
"step": 9000
},
{
"epoch": 0.95042194092827,
"grad_norm": 0.5230275988578796,
"learning_rate": 2.8800130373977936e-05,
"loss": 1.3412,
"step": 9010
},
{
"epoch": 0.9514767932489452,
"grad_norm": 0.45506682991981506,
"learning_rate": 2.84031643124288e-05,
"loss": 1.3435,
"step": 9020
},
{
"epoch": 0.9525316455696202,
"grad_norm": 0.46359187364578247,
"learning_rate": 2.8011669825208517e-05,
"loss": 1.3625,
"step": 9030
},
{
"epoch": 0.9535864978902954,
"grad_norm": 0.4452093541622162,
"learning_rate": 2.762557149497405e-05,
"loss": 1.3399,
"step": 9040
},
{
"epoch": 0.9546413502109705,
"grad_norm": 0.45948877930641174,
"learning_rate": 2.724479494389592e-05,
"loss": 1.3497,
"step": 9050
},
{
"epoch": 0.9556962025316456,
"grad_norm": 0.44907087087631226,
"learning_rate": 2.6869266819330058e-05,
"loss": 1.3511,
"step": 9060
},
{
"epoch": 0.9567510548523207,
"grad_norm": 0.45196837186813354,
"learning_rate": 2.6498914779687228e-05,
"loss": 1.3487,
"step": 9070
},
{
"epoch": 0.9578059071729957,
"grad_norm": 0.4547522962093353,
"learning_rate": 2.6133667480497115e-05,
"loss": 1.3515,
"step": 9080
},
{
"epoch": 0.9588607594936709,
"grad_norm": 0.46276381611824036,
"learning_rate": 2.5773454560664597e-05,
"loss": 1.3532,
"step": 9090
},
{
"epoch": 0.959915611814346,
"grad_norm": 0.4469272792339325,
"learning_rate": 2.541820662891541e-05,
"loss": 1.3316,
"step": 9100
},
{
"epoch": 0.9609704641350211,
"grad_norm": 0.45500412583351135,
"learning_rate": 2.5067855250428616e-05,
"loss": 1.3432,
"step": 9110
},
{
"epoch": 0.9620253164556962,
"grad_norm": 0.4534028470516205,
"learning_rate": 2.472233293365335e-05,
"loss": 1.3511,
"step": 9120
},
{
"epoch": 0.9630801687763713,
"grad_norm": 0.47841382026672363,
"learning_rate": 2.4381573117307307e-05,
"loss": 1.337,
"step": 9130
},
{
"epoch": 0.9641350210970464,
"grad_norm": 0.463321328163147,
"learning_rate": 2.4045510157554362e-05,
"loss": 1.3586,
"step": 9140
},
{
"epoch": 0.9651898734177216,
"grad_norm": 0.45994776487350464,
"learning_rate": 2.3714079315358985e-05,
"loss": 1.3446,
"step": 9150
},
{
"epoch": 0.9662447257383966,
"grad_norm": 0.46441754698753357,
"learning_rate": 2.338721674401494e-05,
"loss": 1.3521,
"step": 9160
},
{
"epoch": 0.9672995780590717,
"grad_norm": 0.4650691747665405,
"learning_rate": 2.30648594768459e-05,
"loss": 1.3502,
"step": 9170
},
{
"epoch": 0.9683544303797469,
"grad_norm": 0.44704753160476685,
"learning_rate": 2.2746945415075523e-05,
"loss": 1.3463,
"step": 9180
},
{
"epoch": 0.9694092827004219,
"grad_norm": 0.4620136320590973,
"learning_rate": 2.2433413315864803e-05,
"loss": 1.3418,
"step": 9190
},
{
"epoch": 0.9704641350210971,
"grad_norm": 0.4611254930496216,
"learning_rate": 2.2124202780514277e-05,
"loss": 1.3463,
"step": 9200
},
{
"epoch": 0.9715189873417721,
"grad_norm": 0.4542345702648163,
"learning_rate": 2.1819254242828815e-05,
"loss": 1.3471,
"step": 9210
},
{
"epoch": 0.9725738396624473,
"grad_norm": 0.4623969495296478,
"learning_rate": 2.151850895764285e-05,
"loss": 1.3437,
"step": 9220
},
{
"epoch": 0.9736286919831224,
"grad_norm": 0.4567524790763855,
"learning_rate": 2.12219089895037e-05,
"loss": 1.3354,
"step": 9230
},
{
"epoch": 0.9746835443037974,
"grad_norm": 0.45435529947280884,
"learning_rate": 2.092939720151092e-05,
"loss": 1.329,
"step": 9240
},
{
"epoch": 0.9757383966244726,
"grad_norm": 0.46444326639175415,
"learning_rate": 2.064091724430947e-05,
"loss": 1.3422,
"step": 9250
},
{
"epoch": 0.9767932489451476,
"grad_norm": 0.46743497252464294,
"learning_rate": 2.0356413545234603e-05,
"loss": 1.3389,
"step": 9260
},
{
"epoch": 0.9778481012658228,
"grad_norm": 0.4665004014968872,
"learning_rate": 2.0075831297606357e-05,
"loss": 1.3482,
"step": 9270
},
{
"epoch": 0.9789029535864979,
"grad_norm": 0.4589768946170807,
"learning_rate": 1.9799116450171627e-05,
"loss": 1.337,
"step": 9280
},
{
"epoch": 0.979957805907173,
"grad_norm": 0.4532953202724457,
"learning_rate": 1.952621569669175e-05,
"loss": 1.3465,
"step": 9290
},
{
"epoch": 0.9810126582278481,
"grad_norm": 0.4485061764717102,
"learning_rate": 1.9257076465673605e-05,
"loss": 1.3432,
"step": 9300
},
{
"epoch": 0.9820675105485233,
"grad_norm": 0.46737146377563477,
"learning_rate": 1.899164691024229e-05,
"loss": 1.3444,
"step": 9310
},
{
"epoch": 0.9831223628691983,
"grad_norm": 0.47384166717529297,
"learning_rate": 1.872987589815331e-05,
"loss": 1.3399,
"step": 9320
},
{
"epoch": 0.9841772151898734,
"grad_norm": 0.4618329107761383,
"learning_rate": 1.8471713001942538e-05,
"loss": 1.3527,
"step": 9330
},
{
"epoch": 0.9852320675105485,
"grad_norm": 0.44944530725479126,
"learning_rate": 1.8217108489211845e-05,
"loss": 1.35,
"step": 9340
},
{
"epoch": 0.9862869198312236,
"grad_norm": 0.444137841463089,
"learning_rate": 1.7966013313048696e-05,
"loss": 1.3594,
"step": 9350
},
{
"epoch": 0.9873417721518988,
"grad_norm": 0.457152783870697,
"learning_rate": 1.7718379102577752e-05,
"loss": 1.3519,
"step": 9360
},
{
"epoch": 0.9883966244725738,
"grad_norm": 0.5073128938674927,
"learning_rate": 1.7474158153642745e-05,
"loss": 1.3498,
"step": 9370
},
{
"epoch": 0.989451476793249,
"grad_norm": 0.49374303221702576,
"learning_rate": 1.7233303419616745e-05,
"loss": 1.3331,
"step": 9380
},
{
"epoch": 0.990506329113924,
"grad_norm": 0.4625350534915924,
"learning_rate": 1.699576850233916e-05,
"loss": 1.3397,
"step": 9390
},
{
"epoch": 0.9915611814345991,
"grad_norm": 0.4780026972293854,
"learning_rate": 1.6761507643177553e-05,
"loss": 1.3526,
"step": 9400
},
{
"epoch": 0.9926160337552743,
"grad_norm": 0.47076642513275146,
"learning_rate": 1.6530475714212752e-05,
"loss": 1.3287,
"step": 9410
},
{
"epoch": 0.9936708860759493,
"grad_norm": 0.444694459438324,
"learning_rate": 1.6302628209545423e-05,
"loss": 1.349,
"step": 9420
},
{
"epoch": 0.9947257383966245,
"grad_norm": 0.45865774154663086,
"learning_rate": 1.6077921236722464e-05,
"loss": 1.3475,
"step": 9430
},
{
"epoch": 0.9957805907172996,
"grad_norm": 0.4531383812427521,
"learning_rate": 1.5856311508281594e-05,
"loss": 1.3449,
"step": 9440
},
{
"epoch": 0.9968354430379747,
"grad_norm": 0.46567508578300476,
"learning_rate": 1.5637756333412454e-05,
"loss": 1.3555,
"step": 9450
},
{
"epoch": 0.9978902953586498,
"grad_norm": 0.4613323509693146,
"learning_rate": 1.542221360973268e-05,
"loss": 1.3376,
"step": 9460
},
{
"epoch": 0.9989451476793249,
"grad_norm": 0.46998530626296997,
"learning_rate": 1.5209641815177312e-05,
"loss": 1.3494,
"step": 9470
},
{
"epoch": 1.0,
"grad_norm": 1.2927305698394775,
"learning_rate": 1.5e-05,
"loss": 1.3386,
"step": 9480
}
],
"logging_steps": 10,
"max_steps": 9480,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.832308198648013e+16,
"train_batch_size": 1024,
"trial_name": null,
"trial_params": null
}