{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 8.0, "eval_steps": 500, "global_step": 10088, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.007933359777865925, "grad_norm": 19.14600299430099, "learning_rate": 8.9197224975223e-07, "loss": 8.2489, "step": 10 }, { "epoch": 0.01586671955573185, "grad_norm": 18.01915458135243, "learning_rate": 1.8830525272547078e-06, "loss": 8.1177, "step": 20 }, { "epoch": 0.02380007933359778, "grad_norm": 9.33822311099433, "learning_rate": 2.8741328047571853e-06, "loss": 7.4735, "step": 30 }, { "epoch": 0.0317334391114637, "grad_norm": 4.906820841029322, "learning_rate": 3.8652130822596635e-06, "loss": 6.7427, "step": 40 }, { "epoch": 0.039666798889329634, "grad_norm": 3.0516622673060785, "learning_rate": 4.8562933597621405e-06, "loss": 6.2707, "step": 50 }, { "epoch": 0.04760015866719556, "grad_norm": 3.0497788345357493, "learning_rate": 5.847373637264618e-06, "loss": 5.9319, "step": 60 }, { "epoch": 0.055533518445061485, "grad_norm": 3.678614138586657, "learning_rate": 6.838453914767097e-06, "loss": 5.5292, "step": 70 }, { "epoch": 0.0634668782229274, "grad_norm": 4.261516409919679, "learning_rate": 7.829534192269576e-06, "loss": 4.9099, "step": 80 }, { "epoch": 0.07140023800079334, "grad_norm": 4.6638020698067875, "learning_rate": 8.820614469772052e-06, "loss": 4.2313, "step": 90 }, { "epoch": 0.07933359777865927, "grad_norm": 4.5951745742169, "learning_rate": 9.81169474727453e-06, "loss": 3.663, "step": 100 }, { "epoch": 0.0872669575565252, "grad_norm": 3.9132985800446947, "learning_rate": 1.0802775024777007e-05, "loss": 3.2806, "step": 110 }, { "epoch": 0.09520031733439112, "grad_norm": 3.864223439251188, "learning_rate": 1.1793855302279485e-05, "loss": 3.0464, "step": 120 }, { "epoch": 0.10313367711225704, "grad_norm": 3.5022825994548232, "learning_rate": 1.2784935579781963e-05, "loss": 2.8886, "step": 130 }, { "epoch": 0.11106703689012297, "grad_norm": 5.1510905388003, "learning_rate": 1.377601585728444e-05, "loss": 2.7668, "step": 140 }, { "epoch": 0.1190003966679889, "grad_norm": 7.21771904799698, "learning_rate": 1.4767096134786917e-05, "loss": 2.6692, "step": 150 }, { "epoch": 0.1269337564458548, "grad_norm": 9.497867401163038, "learning_rate": 1.5758176412289396e-05, "loss": 2.5934, "step": 160 }, { "epoch": 0.13486711622372075, "grad_norm": 8.479404348386907, "learning_rate": 1.6749256689791874e-05, "loss": 2.5471, "step": 170 }, { "epoch": 0.14280047600158668, "grad_norm": 13.33424723235721, "learning_rate": 1.7740336967294352e-05, "loss": 2.4827, "step": 180 }, { "epoch": 0.1507338357794526, "grad_norm": 9.736118445504315, "learning_rate": 1.873141724479683e-05, "loss": 2.4223, "step": 190 }, { "epoch": 0.15866719555731854, "grad_norm": 12.548385792238333, "learning_rate": 1.9722497522299307e-05, "loss": 2.3752, "step": 200 }, { "epoch": 0.16660055533518445, "grad_norm": 15.910150113440148, "learning_rate": 2.0713577799801785e-05, "loss": 2.3339, "step": 210 }, { "epoch": 0.1745339151130504, "grad_norm": 12.024121798784853, "learning_rate": 2.1704658077304263e-05, "loss": 2.2732, "step": 220 }, { "epoch": 0.1824672748909163, "grad_norm": 18.143793557621525, "learning_rate": 2.269573835480674e-05, "loss": 2.2154, "step": 230 }, { "epoch": 0.19040063466878224, "grad_norm": 11.930035885996817, "learning_rate": 2.368681863230922e-05, "loss": 2.1602, "step": 240 }, { "epoch": 0.19833399444664815, "grad_norm": 12.424586861990159, "learning_rate": 2.4677898909811696e-05, "loss": 2.0887, "step": 250 }, { "epoch": 0.2062673542245141, "grad_norm": 11.063682700121975, "learning_rate": 2.5668979187314174e-05, "loss": 1.9843, "step": 260 }, { "epoch": 0.21420071400238, "grad_norm": 9.80005982816964, "learning_rate": 2.6660059464816652e-05, "loss": 1.871, "step": 270 }, { "epoch": 0.22213407378024594, "grad_norm": 6.8486046511969985, "learning_rate": 2.7651139742319126e-05, "loss": 1.7654, "step": 280 }, { "epoch": 0.23006743355811185, "grad_norm": 4.650133260417432, "learning_rate": 2.8642220019821607e-05, "loss": 1.6631, "step": 290 }, { "epoch": 0.2380007933359778, "grad_norm": 8.484902376448611, "learning_rate": 2.9633300297324085e-05, "loss": 1.587, "step": 300 }, { "epoch": 0.2459341531138437, "grad_norm": 8.266296207037957, "learning_rate": 3.062438057482656e-05, "loss": 1.5209, "step": 310 }, { "epoch": 0.2538675128917096, "grad_norm": 6.035258103391185, "learning_rate": 3.161546085232904e-05, "loss": 1.4691, "step": 320 }, { "epoch": 0.2618008726695756, "grad_norm": 6.090335574638085, "learning_rate": 3.260654112983152e-05, "loss": 1.4213, "step": 330 }, { "epoch": 0.2697342324474415, "grad_norm": 7.504625709152101, "learning_rate": 3.359762140733399e-05, "loss": 1.3809, "step": 340 }, { "epoch": 0.2776675922253074, "grad_norm": 5.216278847793629, "learning_rate": 3.458870168483647e-05, "loss": 1.3304, "step": 350 }, { "epoch": 0.28560095200317337, "grad_norm": 5.001664767324889, "learning_rate": 3.5579781962338955e-05, "loss": 1.2886, "step": 360 }, { "epoch": 0.2935343117810393, "grad_norm": 4.035122456050597, "learning_rate": 3.657086223984143e-05, "loss": 1.2441, "step": 370 }, { "epoch": 0.3014676715589052, "grad_norm": 4.3131477681191015, "learning_rate": 3.7561942517343904e-05, "loss": 1.1973, "step": 380 }, { "epoch": 0.3094010313367711, "grad_norm": 8.188203831814278, "learning_rate": 3.855302279484638e-05, "loss": 1.175, "step": 390 }, { "epoch": 0.31733439111463707, "grad_norm": 4.755933403654607, "learning_rate": 3.9544103072348866e-05, "loss": 1.1342, "step": 400 }, { "epoch": 0.325267750892503, "grad_norm": 6.57796715979494, "learning_rate": 4.0535183349851344e-05, "loss": 1.0971, "step": 410 }, { "epoch": 0.3332011106703689, "grad_norm": 4.216167153902993, "learning_rate": 4.1526263627353815e-05, "loss": 1.0513, "step": 420 }, { "epoch": 0.3411344704482348, "grad_norm": 8.418073662598946, "learning_rate": 4.251734390485629e-05, "loss": 1.0278, "step": 430 }, { "epoch": 0.3490678302261008, "grad_norm": 4.293832971560352, "learning_rate": 4.350842418235878e-05, "loss": 0.9989, "step": 440 }, { "epoch": 0.3570011900039667, "grad_norm": 4.738542701520305, "learning_rate": 4.4499504459861255e-05, "loss": 0.9646, "step": 450 }, { "epoch": 0.3649345497818326, "grad_norm": 3.836235781620353, "learning_rate": 4.5490584737363726e-05, "loss": 0.9282, "step": 460 }, { "epoch": 0.3728679095596985, "grad_norm": 4.385628595929679, "learning_rate": 4.6481665014866204e-05, "loss": 0.9069, "step": 470 }, { "epoch": 0.3808012693375645, "grad_norm": 4.795190260431852, "learning_rate": 4.747274529236869e-05, "loss": 0.8767, "step": 480 }, { "epoch": 0.3887346291154304, "grad_norm": 5.15771361003216, "learning_rate": 4.846382556987116e-05, "loss": 0.848, "step": 490 }, { "epoch": 0.3966679888932963, "grad_norm": 5.00866147389186, "learning_rate": 4.945490584737364e-05, "loss": 0.832, "step": 500 }, { "epoch": 0.40460134867116226, "grad_norm": 3.3749478011742617, "learning_rate": 5.044598612487612e-05, "loss": 0.801, "step": 510 }, { "epoch": 0.4125347084490282, "grad_norm": 5.648100107599949, "learning_rate": 5.14370664023786e-05, "loss": 0.7907, "step": 520 }, { "epoch": 0.4204680682268941, "grad_norm": 5.180183508710987, "learning_rate": 5.242814667988107e-05, "loss": 0.7641, "step": 530 }, { "epoch": 0.42840142800476, "grad_norm": 4.795683231567134, "learning_rate": 5.341922695738355e-05, "loss": 0.7428, "step": 540 }, { "epoch": 0.43633478778262597, "grad_norm": 4.474423469358904, "learning_rate": 5.4410307234886027e-05, "loss": 0.7178, "step": 550 }, { "epoch": 0.4442681475604919, "grad_norm": 5.965818045702901, "learning_rate": 5.5401387512388504e-05, "loss": 0.7008, "step": 560 }, { "epoch": 0.4522015073383578, "grad_norm": 4.4181008640757256, "learning_rate": 5.639246778989098e-05, "loss": 0.678, "step": 570 }, { "epoch": 0.4601348671162237, "grad_norm": 4.925348313364693, "learning_rate": 5.738354806739347e-05, "loss": 0.6516, "step": 580 }, { "epoch": 0.46806822689408967, "grad_norm": 4.995771437156914, "learning_rate": 5.8374628344895944e-05, "loss": 0.6261, "step": 590 }, { "epoch": 0.4760015866719556, "grad_norm": 4.3400967454545185, "learning_rate": 5.936570862239842e-05, "loss": 0.6099, "step": 600 }, { "epoch": 0.4839349464498215, "grad_norm": 3.8437851403940013, "learning_rate": 6.035678889990089e-05, "loss": 0.5905, "step": 610 }, { "epoch": 0.4918683062276874, "grad_norm": 4.673291642824443, "learning_rate": 6.134786917740337e-05, "loss": 0.5739, "step": 620 }, { "epoch": 0.49980166600555337, "grad_norm": 4.691220121981611, "learning_rate": 6.233894945490585e-05, "loss": 0.5543, "step": 630 }, { "epoch": 0.5077350257834192, "grad_norm": 4.609768045302665, "learning_rate": 6.333002973240833e-05, "loss": 0.5265, "step": 640 }, { "epoch": 0.5156683855612852, "grad_norm": 3.2621270260578967, "learning_rate": 6.43211100099108e-05, "loss": 0.5175, "step": 650 }, { "epoch": 0.5236017453391512, "grad_norm": 3.5703471859456806, "learning_rate": 6.531219028741328e-05, "loss": 0.4972, "step": 660 }, { "epoch": 0.5315351051170171, "grad_norm": 3.530519835264167, "learning_rate": 6.630327056491576e-05, "loss": 0.4717, "step": 670 }, { "epoch": 0.539468464894883, "grad_norm": 2.392046097161058, "learning_rate": 6.729435084241824e-05, "loss": 0.4583, "step": 680 }, { "epoch": 0.5474018246727489, "grad_norm": 3.149236166223247, "learning_rate": 6.828543111992072e-05, "loss": 0.4412, "step": 690 }, { "epoch": 0.5553351844506148, "grad_norm": 2.9354688050931586, "learning_rate": 6.92765113974232e-05, "loss": 0.423, "step": 700 }, { "epoch": 0.5632685442284807, "grad_norm": 2.7127436235307125, "learning_rate": 7.026759167492567e-05, "loss": 0.4062, "step": 710 }, { "epoch": 0.5712019040063467, "grad_norm": 2.6410601628390777, "learning_rate": 7.125867195242815e-05, "loss": 0.3923, "step": 720 }, { "epoch": 0.5791352637842127, "grad_norm": 2.7996527641631386, "learning_rate": 7.224975222993063e-05, "loss": 0.3739, "step": 730 }, { "epoch": 0.5870686235620786, "grad_norm": 3.53218129873484, "learning_rate": 7.32408325074331e-05, "loss": 0.3558, "step": 740 }, { "epoch": 0.5950019833399445, "grad_norm": 3.256471177277615, "learning_rate": 7.423191278493558e-05, "loss": 0.3378, "step": 750 }, { "epoch": 0.6029353431178104, "grad_norm": 2.448855772936633, "learning_rate": 7.522299306243806e-05, "loss": 0.3318, "step": 760 }, { "epoch": 0.6108687028956763, "grad_norm": 2.8206966008765355, "learning_rate": 7.621407333994054e-05, "loss": 0.3145, "step": 770 }, { "epoch": 0.6188020626735422, "grad_norm": 2.262400615934418, "learning_rate": 7.720515361744302e-05, "loss": 0.2981, "step": 780 }, { "epoch": 0.6267354224514081, "grad_norm": 2.6691511066584894, "learning_rate": 7.81962338949455e-05, "loss": 0.2878, "step": 790 }, { "epoch": 0.6346687822292741, "grad_norm": 2.1502320771866494, "learning_rate": 7.918731417244797e-05, "loss": 0.2724, "step": 800 }, { "epoch": 0.64260214200714, "grad_norm": 2.1869290626436806, "learning_rate": 8.017839444995045e-05, "loss": 0.2589, "step": 810 }, { "epoch": 0.650535501785006, "grad_norm": 2.6277829837493476, "learning_rate": 8.116947472745293e-05, "loss": 0.2481, "step": 820 }, { "epoch": 0.6584688615628719, "grad_norm": 1.8401496589236042, "learning_rate": 8.21605550049554e-05, "loss": 0.2361, "step": 830 }, { "epoch": 0.6664022213407378, "grad_norm": 1.7862155383026863, "learning_rate": 8.315163528245788e-05, "loss": 0.222, "step": 840 }, { "epoch": 0.6743355811186037, "grad_norm": 1.9765212410305726, "learning_rate": 8.414271555996036e-05, "loss": 0.2103, "step": 850 }, { "epoch": 0.6822689408964696, "grad_norm": 2.4397348134664556, "learning_rate": 8.513379583746284e-05, "loss": 0.2021, "step": 860 }, { "epoch": 0.6902023006743356, "grad_norm": 2.296758487647809, "learning_rate": 8.612487611496532e-05, "loss": 0.1924, "step": 870 }, { "epoch": 0.6981356604522015, "grad_norm": 2.104859870669743, "learning_rate": 8.71159563924678e-05, "loss": 0.1817, "step": 880 }, { "epoch": 0.7060690202300675, "grad_norm": 1.3003724019670297, "learning_rate": 8.810703666997027e-05, "loss": 0.1732, "step": 890 }, { "epoch": 0.7140023800079334, "grad_norm": 1.3709932626535646, "learning_rate": 8.909811694747275e-05, "loss": 0.1593, "step": 900 }, { "epoch": 0.7219357397857993, "grad_norm": 1.8342489151171628, "learning_rate": 9.008919722497523e-05, "loss": 0.1551, "step": 910 }, { "epoch": 0.7298690995636652, "grad_norm": 1.762934595290678, "learning_rate": 9.10802775024777e-05, "loss": 0.1489, "step": 920 }, { "epoch": 0.7378024593415311, "grad_norm": 1.4075349928757257, "learning_rate": 9.207135777998018e-05, "loss": 0.1391, "step": 930 }, { "epoch": 0.745735819119397, "grad_norm": 1.3962752805987728, "learning_rate": 9.306243805748266e-05, "loss": 0.1332, "step": 940 }, { "epoch": 0.753669178897263, "grad_norm": 3.5595641910775173, "learning_rate": 9.405351833498514e-05, "loss": 0.1507, "step": 950 }, { "epoch": 0.761602538675129, "grad_norm": 1.2799346026465042, "learning_rate": 9.504459861248762e-05, "loss": 0.1706, "step": 960 }, { "epoch": 0.7695358984529949, "grad_norm": 0.9866434355012127, "learning_rate": 9.60356788899901e-05, "loss": 0.1241, "step": 970 }, { "epoch": 0.7774692582308608, "grad_norm": 1.2605525685265588, "learning_rate": 9.702675916749257e-05, "loss": 0.1133, "step": 980 }, { "epoch": 0.7854026180087267, "grad_norm": 1.0654588120046724, "learning_rate": 9.801783944499505e-05, "loss": 0.1066, "step": 990 }, { "epoch": 0.7933359777865926, "grad_norm": 0.761881486606698, "learning_rate": 9.900891972249753e-05, "loss": 0.1015, "step": 1000 }, { "epoch": 0.8012693375644585, "grad_norm": 1.0601392633604383, "learning_rate": 0.0001, "loss": 0.0953, "step": 1010 }, { "epoch": 0.8092026973423245, "grad_norm": 0.8224204361157935, "learning_rate": 9.999970066101032e-05, "loss": 0.0906, "step": 1020 }, { "epoch": 0.8171360571201904, "grad_norm": 0.7080480512831684, "learning_rate": 9.99988026476254e-05, "loss": 0.0866, "step": 1030 }, { "epoch": 0.8250694168980564, "grad_norm": 0.7547638843757503, "learning_rate": 9.999730597059764e-05, "loss": 0.0809, "step": 1040 }, { "epoch": 0.8330027766759223, "grad_norm": 1.1588422105514353, "learning_rate": 9.999521064784764e-05, "loss": 0.0844, "step": 1050 }, { "epoch": 0.8409361364537882, "grad_norm": 0.6385523334282596, "learning_rate": 9.999251670446386e-05, "loss": 0.0804, "step": 1060 }, { "epoch": 0.8488694962316541, "grad_norm": 0.5623095327146973, "learning_rate": 9.998922417270235e-05, "loss": 0.072, "step": 1070 }, { "epoch": 0.85680285600952, "grad_norm": 0.43678572122080983, "learning_rate": 9.99853330919865e-05, "loss": 0.0679, "step": 1080 }, { "epoch": 0.8647362157873859, "grad_norm": 0.4972727247312438, "learning_rate": 9.998084350890633e-05, "loss": 0.065, "step": 1090 }, { "epoch": 0.8726695755652519, "grad_norm": 0.5035917631799433, "learning_rate": 9.997575547721818e-05, "loss": 0.0627, "step": 1100 }, { "epoch": 0.8806029353431178, "grad_norm": 0.5182605794595788, "learning_rate": 9.997006905784388e-05, "loss": 0.0602, "step": 1110 }, { "epoch": 0.8885362951209838, "grad_norm": 0.41920699625736624, "learning_rate": 9.996378431887012e-05, "loss": 0.0579, "step": 1120 }, { "epoch": 0.8964696548988497, "grad_norm": 0.414907297365465, "learning_rate": 9.995690133554758e-05, "loss": 0.0546, "step": 1130 }, { "epoch": 0.9044030146767156, "grad_norm": 0.3702005062489374, "learning_rate": 9.99494201902901e-05, "loss": 0.0523, "step": 1140 }, { "epoch": 0.9123363744545815, "grad_norm": 0.42037834689750175, "learning_rate": 9.994134097267358e-05, "loss": 0.0516, "step": 1150 }, { "epoch": 0.9202697342324474, "grad_norm": 0.38928699328314564, "learning_rate": 9.993266377943504e-05, "loss": 0.0505, "step": 1160 }, { "epoch": 0.9282030940103134, "grad_norm": 0.3941895011795836, "learning_rate": 9.992338871447138e-05, "loss": 0.0495, "step": 1170 }, { "epoch": 0.9361364537881793, "grad_norm": 0.37452365492894446, "learning_rate": 9.991351588883811e-05, "loss": 0.0474, "step": 1180 }, { "epoch": 0.9440698135660452, "grad_norm": 0.4395025405712049, "learning_rate": 9.99030454207481e-05, "loss": 0.0467, "step": 1190 }, { "epoch": 0.9520031733439112, "grad_norm": 0.4639716702934286, "learning_rate": 9.989197743557017e-05, "loss": 0.045, "step": 1200 }, { "epoch": 0.9599365331217771, "grad_norm": 0.34908102760028864, "learning_rate": 9.988031206582746e-05, "loss": 0.0439, "step": 1210 }, { "epoch": 0.967869892899643, "grad_norm": 0.29211915294831964, "learning_rate": 9.986804945119597e-05, "loss": 0.0429, "step": 1220 }, { "epoch": 0.9758032526775089, "grad_norm": 0.2801643208729514, "learning_rate": 9.985518973850285e-05, "loss": 0.041, "step": 1230 }, { "epoch": 0.9837366124553748, "grad_norm": 0.3013251709022931, "learning_rate": 9.984173308172465e-05, "loss": 0.0403, "step": 1240 }, { "epoch": 0.9916699722332408, "grad_norm": 0.28830525385588507, "learning_rate": 9.982767964198542e-05, "loss": 0.04, "step": 1250 }, { "epoch": 0.9996033320111067, "grad_norm": 0.2756668810883745, "learning_rate": 9.98130295875549e-05, "loss": 0.0386, "step": 1260 }, { "epoch": 1.0071400238000794, "grad_norm": 0.2673295493921112, "learning_rate": 9.979778309384638e-05, "loss": 0.0369, "step": 1270 }, { "epoch": 1.0150733835779453, "grad_norm": 0.28881709817045276, "learning_rate": 9.978194034341462e-05, "loss": 0.0366, "step": 1280 }, { "epoch": 1.0230067433558112, "grad_norm": 0.2443064331996704, "learning_rate": 9.976550152595379e-05, "loss": 0.035, "step": 1290 }, { "epoch": 1.0309401031336771, "grad_norm": 0.2586793532004466, "learning_rate": 9.974846683829501e-05, "loss": 0.0351, "step": 1300 }, { "epoch": 1.038873462911543, "grad_norm": 0.6237223735436187, "learning_rate": 9.973083648440416e-05, "loss": 0.0495, "step": 1310 }, { "epoch": 1.046806822689409, "grad_norm": 0.982888186735296, "learning_rate": 9.971261067537932e-05, "loss": 0.0428, "step": 1320 }, { "epoch": 1.0547401824672749, "grad_norm": 0.3721310946214082, "learning_rate": 9.969378962944829e-05, "loss": 0.0396, "step": 1330 }, { "epoch": 1.0626735422451408, "grad_norm": 0.3035794519690265, "learning_rate": 9.9674373571966e-05, "loss": 0.0352, "step": 1340 }, { "epoch": 1.0706069020230067, "grad_norm": 0.2564381074742219, "learning_rate": 9.965436273541175e-05, "loss": 0.034, "step": 1350 }, { "epoch": 1.0785402618008726, "grad_norm": 0.2627161314780947, "learning_rate": 9.963375735938651e-05, "loss": 0.0328, "step": 1360 }, { "epoch": 1.0864736215787385, "grad_norm": 0.23965693542315172, "learning_rate": 9.961255769060997e-05, "loss": 0.0318, "step": 1370 }, { "epoch": 1.0944069813566046, "grad_norm": 0.24468200257324066, "learning_rate": 9.95907639829176e-05, "loss": 0.0312, "step": 1380 }, { "epoch": 1.1023403411344705, "grad_norm": 0.2090058523545183, "learning_rate": 9.95683764972577e-05, "loss": 0.0301, "step": 1390 }, { "epoch": 1.1102737009123365, "grad_norm": 0.2583539701995491, "learning_rate": 9.954539550168816e-05, "loss": 0.0301, "step": 1400 }, { "epoch": 1.1182070606902024, "grad_norm": 0.22716588824206302, "learning_rate": 9.952182127137327e-05, "loss": 0.0293, "step": 1410 }, { "epoch": 1.1261404204680683, "grad_norm": 0.24087202301512833, "learning_rate": 9.94976540885805e-05, "loss": 0.0287, "step": 1420 }, { "epoch": 1.1340737802459342, "grad_norm": 0.23388912586015503, "learning_rate": 9.947289424267703e-05, "loss": 0.0283, "step": 1430 }, { "epoch": 1.1420071400238, "grad_norm": 0.21775302377697844, "learning_rate": 9.944754203012639e-05, "loss": 0.0273, "step": 1440 }, { "epoch": 1.149940499801666, "grad_norm": 0.2120232453240847, "learning_rate": 9.942159775448478e-05, "loss": 0.0273, "step": 1450 }, { "epoch": 1.157873859579532, "grad_norm": 0.2055361894837688, "learning_rate": 9.939506172639753e-05, "loss": 0.027, "step": 1460 }, { "epoch": 1.1658072193573978, "grad_norm": 0.22045835966574195, "learning_rate": 9.936793426359536e-05, "loss": 0.0264, "step": 1470 }, { "epoch": 1.1737405791352638, "grad_norm": 0.19936431245788608, "learning_rate": 9.934021569089058e-05, "loss": 0.0262, "step": 1480 }, { "epoch": 1.1816739389131297, "grad_norm": 0.2182877537167163, "learning_rate": 9.931190634017314e-05, "loss": 0.0256, "step": 1490 }, { "epoch": 1.1896072986909956, "grad_norm": 0.20894413268256554, "learning_rate": 9.928300655040676e-05, "loss": 0.0255, "step": 1500 }, { "epoch": 1.1975406584688615, "grad_norm": 0.217040492922683, "learning_rate": 9.925351666762479e-05, "loss": 0.0248, "step": 1510 }, { "epoch": 1.2054740182467274, "grad_norm": 0.2080798426449999, "learning_rate": 9.92234370449261e-05, "loss": 0.0247, "step": 1520 }, { "epoch": 1.2134073780245935, "grad_norm": 0.21422582180243674, "learning_rate": 9.919276804247084e-05, "loss": 0.0246, "step": 1530 }, { "epoch": 1.2213407378024592, "grad_norm": 0.22109698138341372, "learning_rate": 9.916151002747614e-05, "loss": 0.0242, "step": 1540 }, { "epoch": 1.2292740975803254, "grad_norm": 0.2191386707736148, "learning_rate": 9.912966337421169e-05, "loss": 0.0241, "step": 1550 }, { "epoch": 1.2372074573581913, "grad_norm": 0.21285746880929995, "learning_rate": 9.909722846399531e-05, "loss": 0.0235, "step": 1560 }, { "epoch": 1.2451408171360572, "grad_norm": 0.2060608693515892, "learning_rate": 9.906420568518832e-05, "loss": 0.0235, "step": 1570 }, { "epoch": 1.253074176913923, "grad_norm": 0.2165821829643641, "learning_rate": 9.903059543319096e-05, "loss": 0.0228, "step": 1580 }, { "epoch": 1.261007536691789, "grad_norm": 0.20784176983531677, "learning_rate": 9.899639811043755e-05, "loss": 0.0224, "step": 1590 }, { "epoch": 1.268940896469655, "grad_norm": 0.1941327009060369, "learning_rate": 9.896161412639178e-05, "loss": 0.0224, "step": 1600 }, { "epoch": 1.2768742562475208, "grad_norm": 0.19719426217271882, "learning_rate": 9.892624389754175e-05, "loss": 0.0218, "step": 1610 }, { "epoch": 1.2848076160253867, "grad_norm": 0.1987404294569414, "learning_rate": 9.889028784739502e-05, "loss": 0.0217, "step": 1620 }, { "epoch": 1.2927409758032526, "grad_norm": 0.36446508935174654, "learning_rate": 9.885374640647349e-05, "loss": 0.0218, "step": 1630 }, { "epoch": 1.3006743355811186, "grad_norm": 0.2153981685427444, "learning_rate": 9.881662001230827e-05, "loss": 0.0219, "step": 1640 }, { "epoch": 1.3086076953589845, "grad_norm": 0.23124643936791714, "learning_rate": 9.877890910943448e-05, "loss": 0.0215, "step": 1650 }, { "epoch": 1.3165410551368504, "grad_norm": 0.19794561936484395, "learning_rate": 9.874061414938583e-05, "loss": 0.0209, "step": 1660 }, { "epoch": 1.3244744149147163, "grad_norm": 0.1939423361167064, "learning_rate": 9.870173559068933e-05, "loss": 0.0204, "step": 1670 }, { "epoch": 1.3324077746925824, "grad_norm": 0.2157419658468051, "learning_rate": 9.86622738988597e-05, "loss": 0.02, "step": 1680 }, { "epoch": 1.3403411344704481, "grad_norm": 0.19391991216609314, "learning_rate": 9.862222954639389e-05, "loss": 0.0196, "step": 1690 }, { "epoch": 1.3482744942483142, "grad_norm": 0.19977705728669262, "learning_rate": 9.858160301276529e-05, "loss": 0.0191, "step": 1700 }, { "epoch": 1.3562078540261802, "grad_norm": 0.1658808248673058, "learning_rate": 9.854039478441817e-05, "loss": 0.0194, "step": 1710 }, { "epoch": 1.364141213804046, "grad_norm": 0.18655747400847833, "learning_rate": 9.849860535476168e-05, "loss": 0.0193, "step": 1720 }, { "epoch": 1.372074573581912, "grad_norm": 0.18591273213926962, "learning_rate": 9.845623522416407e-05, "loss": 0.0191, "step": 1730 }, { "epoch": 1.380007933359778, "grad_norm": 0.19943417251594622, "learning_rate": 9.841328489994659e-05, "loss": 0.0189, "step": 1740 }, { "epoch": 1.3879412931376438, "grad_norm": 0.20534200797212154, "learning_rate": 9.836975489637754e-05, "loss": 0.0184, "step": 1750 }, { "epoch": 1.3958746529155097, "grad_norm": 0.1911015742718259, "learning_rate": 9.832564573466598e-05, "loss": 0.0182, "step": 1760 }, { "epoch": 1.4038080126933756, "grad_norm": 0.18521458075594055, "learning_rate": 9.82809579429556e-05, "loss": 0.0178, "step": 1770 }, { "epoch": 1.4117413724712415, "grad_norm": 0.19914721587866432, "learning_rate": 9.823569205631836e-05, "loss": 0.0179, "step": 1780 }, { "epoch": 1.4196747322491075, "grad_norm": 0.19845608937674608, "learning_rate": 9.818984861674803e-05, "loss": 0.0176, "step": 1790 }, { "epoch": 1.4276080920269734, "grad_norm": 0.19674825577684443, "learning_rate": 9.814342817315374e-05, "loss": 0.0177, "step": 1800 }, { "epoch": 1.4355414518048393, "grad_norm": 0.18373733166208803, "learning_rate": 9.809643128135349e-05, "loss": 0.0171, "step": 1810 }, { "epoch": 1.4434748115827052, "grad_norm": 0.1801371927821921, "learning_rate": 9.804885850406731e-05, "loss": 0.0173, "step": 1820 }, { "epoch": 1.4514081713605713, "grad_norm": 0.17435713690708068, "learning_rate": 9.800071041091072e-05, "loss": 0.0167, "step": 1830 }, { "epoch": 1.459341531138437, "grad_norm": 0.22908445382887835, "learning_rate": 9.795198757838779e-05, "loss": 0.0168, "step": 1840 }, { "epoch": 1.4672748909163031, "grad_norm": 0.19945709262181688, "learning_rate": 9.790269058988423e-05, "loss": 0.017, "step": 1850 }, { "epoch": 1.475208250694169, "grad_norm": 0.1771428579202279, "learning_rate": 9.785282003566048e-05, "loss": 0.0169, "step": 1860 }, { "epoch": 1.483141610472035, "grad_norm": 0.16493256678410484, "learning_rate": 9.780237651284458e-05, "loss": 0.0165, "step": 1870 }, { "epoch": 1.4910749702499009, "grad_norm": 0.17934728993394863, "learning_rate": 9.775136062542509e-05, "loss": 0.0161, "step": 1880 }, { "epoch": 1.4990083300277668, "grad_norm": 0.1655943299806583, "learning_rate": 9.769977298424375e-05, "loss": 0.0157, "step": 1890 }, { "epoch": 1.5069416898056327, "grad_norm": 0.2003983514777471, "learning_rate": 9.764761420698827e-05, "loss": 0.0159, "step": 1900 }, { "epoch": 1.5148750495834986, "grad_norm": 0.17686970428136242, "learning_rate": 9.759488491818487e-05, "loss": 0.0153, "step": 1910 }, { "epoch": 1.5228084093613645, "grad_norm": 0.16545001551025132, "learning_rate": 9.754158574919082e-05, "loss": 0.0151, "step": 1920 }, { "epoch": 1.5307417691392304, "grad_norm": 0.1672222245874224, "learning_rate": 9.748771733818692e-05, "loss": 0.0149, "step": 1930 }, { "epoch": 1.5386751289170963, "grad_norm": 0.20228055159403058, "learning_rate": 9.743328033016979e-05, "loss": 0.0148, "step": 1940 }, { "epoch": 1.5466084886949623, "grad_norm": 0.1730553929947374, "learning_rate": 9.737827537694418e-05, "loss": 0.0148, "step": 1950 }, { "epoch": 1.5545418484728284, "grad_norm": 0.16306713584478874, "learning_rate": 9.732270313711518e-05, "loss": 0.0145, "step": 1960 }, { "epoch": 1.562475208250694, "grad_norm": 0.1613864429997223, "learning_rate": 9.726656427608032e-05, "loss": 0.0144, "step": 1970 }, { "epoch": 1.5704085680285602, "grad_norm": 0.15676568016378006, "learning_rate": 9.72098594660216e-05, "loss": 0.0142, "step": 1980 }, { "epoch": 1.578341927806426, "grad_norm": 0.18212353738494705, "learning_rate": 9.715258938589744e-05, "loss": 0.0138, "step": 1990 }, { "epoch": 1.586275287584292, "grad_norm": 0.1724534628055386, "learning_rate": 9.709475472143453e-05, "loss": 0.0135, "step": 2000 }, { "epoch": 1.5942086473621577, "grad_norm": 0.2030582534730454, "learning_rate": 9.703635616511972e-05, "loss": 0.0142, "step": 2010 }, { "epoch": 1.6021420071400239, "grad_norm": 0.17303461729625183, "learning_rate": 9.697739441619156e-05, "loss": 0.0136, "step": 2020 }, { "epoch": 1.6100753669178898, "grad_norm": 0.1670928864009479, "learning_rate": 9.69178701806321e-05, "loss": 0.0133, "step": 2030 }, { "epoch": 1.6180087266957557, "grad_norm": 0.17958123638082196, "learning_rate": 9.685778417115829e-05, "loss": 0.0131, "step": 2040 }, { "epoch": 1.6259420864736216, "grad_norm": 0.18217670887367746, "learning_rate": 9.679713710721356e-05, "loss": 0.0129, "step": 2050 }, { "epoch": 1.6338754462514875, "grad_norm": 0.1590392893596093, "learning_rate": 9.673592971495917e-05, "loss": 0.0125, "step": 2060 }, { "epoch": 1.6418088060293534, "grad_norm": 0.16546382830684306, "learning_rate": 9.667416272726542e-05, "loss": 0.0124, "step": 2070 }, { "epoch": 1.6497421658072193, "grad_norm": 0.16187448696505655, "learning_rate": 9.661183688370306e-05, "loss": 0.0123, "step": 2080 }, { "epoch": 1.6576755255850852, "grad_norm": 0.15132455192579947, "learning_rate": 9.654895293053427e-05, "loss": 0.0121, "step": 2090 }, { "epoch": 1.6656088853629512, "grad_norm": 0.1570850734881935, "learning_rate": 9.648551162070385e-05, "loss": 0.0117, "step": 2100 }, { "epoch": 1.6735422451408173, "grad_norm": 0.15875645484905151, "learning_rate": 9.642151371383005e-05, "loss": 0.0118, "step": 2110 }, { "epoch": 1.681475604918683, "grad_norm": 0.1438981307659869, "learning_rate": 9.635695997619565e-05, "loss": 0.0112, "step": 2120 }, { "epoch": 1.689408964696549, "grad_norm": 0.17823258729626815, "learning_rate": 9.629185118073868e-05, "loss": 0.0114, "step": 2130 }, { "epoch": 1.6973423244744148, "grad_norm": 0.1456063020084619, "learning_rate": 9.622618810704316e-05, "loss": 0.0112, "step": 2140 }, { "epoch": 1.705275684252281, "grad_norm": 0.14515257419641928, "learning_rate": 9.615997154132985e-05, "loss": 0.0111, "step": 2150 }, { "epoch": 1.7132090440301466, "grad_norm": 0.14229825582089722, "learning_rate": 9.60932022764467e-05, "loss": 0.0108, "step": 2160 }, { "epoch": 1.7211424038080128, "grad_norm": 0.1549735628184551, "learning_rate": 9.60258811118595e-05, "loss": 0.0107, "step": 2170 }, { "epoch": 1.7290757635858787, "grad_norm": 0.16276536504847472, "learning_rate": 9.595800885364226e-05, "loss": 0.0106, "step": 2180 }, { "epoch": 1.7370091233637446, "grad_norm": 0.1364292826380093, "learning_rate": 9.588958631446745e-05, "loss": 0.0106, "step": 2190 }, { "epoch": 1.7449424831416105, "grad_norm": 0.14559551979461718, "learning_rate": 9.582061431359647e-05, "loss": 0.0103, "step": 2200 }, { "epoch": 1.7528758429194764, "grad_norm": 0.1818745133582307, "learning_rate": 9.575109367686965e-05, "loss": 0.0104, "step": 2210 }, { "epoch": 1.7608092026973423, "grad_norm": 0.17914978500964376, "learning_rate": 9.568102523669649e-05, "loss": 0.0099, "step": 2220 }, { "epoch": 1.7687425624752082, "grad_norm": 0.1508452007716488, "learning_rate": 9.561040983204563e-05, "loss": 0.0103, "step": 2230 }, { "epoch": 1.7766759222530741, "grad_norm": 0.14410972407888561, "learning_rate": 9.553924830843484e-05, "loss": 0.01, "step": 2240 }, { "epoch": 1.78460928203094, "grad_norm": 0.14661639481163527, "learning_rate": 9.546754151792084e-05, "loss": 0.0096, "step": 2250 }, { "epoch": 1.7925426418088062, "grad_norm": 0.13331677970140016, "learning_rate": 9.539529031908915e-05, "loss": 0.0095, "step": 2260 }, { "epoch": 1.8004760015866719, "grad_norm": 0.1505833320965518, "learning_rate": 9.532249557704386e-05, "loss": 0.0093, "step": 2270 }, { "epoch": 1.808409361364538, "grad_norm": 0.14264267761784652, "learning_rate": 9.52491581633971e-05, "loss": 0.0094, "step": 2280 }, { "epoch": 1.8163427211424037, "grad_norm": 0.13122197713348774, "learning_rate": 9.517527895625878e-05, "loss": 0.0097, "step": 2290 }, { "epoch": 1.8242760809202698, "grad_norm": 0.1395067737860133, "learning_rate": 9.510085884022597e-05, "loss": 0.009, "step": 2300 }, { "epoch": 1.8322094406981355, "grad_norm": 0.13417802543190277, "learning_rate": 9.50258987063724e-05, "loss": 0.0091, "step": 2310 }, { "epoch": 1.8401428004760016, "grad_norm": 0.1361399896857281, "learning_rate": 9.495039945223768e-05, "loss": 0.0091, "step": 2320 }, { "epoch": 1.8480761602538676, "grad_norm": 0.12660004136574077, "learning_rate": 9.48743619818166e-05, "loss": 0.009, "step": 2330 }, { "epoch": 1.8560095200317335, "grad_norm": 0.12929036804891955, "learning_rate": 9.47977872055484e-05, "loss": 0.0087, "step": 2340 }, { "epoch": 1.8639428798095994, "grad_norm": 0.12111549685, "learning_rate": 9.472067604030567e-05, "loss": 0.0084, "step": 2350 }, { "epoch": 1.8718762395874653, "grad_norm": 0.11410224967003375, "learning_rate": 9.464302940938359e-05, "loss": 0.0083, "step": 2360 }, { "epoch": 1.8798095993653312, "grad_norm": 0.12355268332572798, "learning_rate": 9.45648482424887e-05, "loss": 0.0088, "step": 2370 }, { "epoch": 1.8877429591431971, "grad_norm": 0.15793674725938558, "learning_rate": 9.448613347572786e-05, "loss": 0.0083, "step": 2380 }, { "epoch": 1.895676318921063, "grad_norm": 0.12535858682328516, "learning_rate": 9.440688605159702e-05, "loss": 0.0084, "step": 2390 }, { "epoch": 1.903609678698929, "grad_norm": 0.13814726939668615, "learning_rate": 9.432710691896994e-05, "loss": 0.0081, "step": 2400 }, { "epoch": 1.911543038476795, "grad_norm": 0.3170609521576562, "learning_rate": 9.42467970330868e-05, "loss": 0.0083, "step": 2410 }, { "epoch": 1.9194763982546608, "grad_norm": 0.1409043652874587, "learning_rate": 9.416595735554285e-05, "loss": 0.0081, "step": 2420 }, { "epoch": 1.927409758032527, "grad_norm": 0.11628134284806954, "learning_rate": 9.408458885427673e-05, "loss": 0.008, "step": 2430 }, { "epoch": 1.9353431178103926, "grad_norm": 0.11936220506545085, "learning_rate": 9.40026925035591e-05, "loss": 0.0077, "step": 2440 }, { "epoch": 1.9432764775882587, "grad_norm": 0.13272648582859942, "learning_rate": 9.392026928398072e-05, "loss": 0.0078, "step": 2450 }, { "epoch": 1.9512098373661244, "grad_norm": 0.23619264812795585, "learning_rate": 9.383732018244097e-05, "loss": 0.0078, "step": 2460 }, { "epoch": 1.9591431971439905, "grad_norm": 0.13012160233446282, "learning_rate": 9.375384619213585e-05, "loss": 0.0078, "step": 2470 }, { "epoch": 1.9670765569218565, "grad_norm": 0.128904077298178, "learning_rate": 9.366984831254617e-05, "loss": 0.0077, "step": 2480 }, { "epoch": 1.9750099166997224, "grad_norm": 0.1269793451565523, "learning_rate": 9.358532754942552e-05, "loss": 0.0076, "step": 2490 }, { "epoch": 1.9829432764775883, "grad_norm": 0.1248949950499374, "learning_rate": 9.350028491478832e-05, "loss": 0.0074, "step": 2500 }, { "epoch": 1.9908766362554542, "grad_norm": 0.11628824723897038, "learning_rate": 9.34147214268976e-05, "loss": 0.0073, "step": 2510 }, { "epoch": 1.99880999603332, "grad_norm": 0.11767052090993746, "learning_rate": 9.33286381102529e-05, "loss": 0.007, "step": 2520 }, { "epoch": 2.0063466878222926, "grad_norm": 0.1117903784163016, "learning_rate": 9.324203599557792e-05, "loss": 0.0059, "step": 2530 }, { "epoch": 2.0142800476001588, "grad_norm": 0.10827884886358363, "learning_rate": 9.315491611980827e-05, "loss": 0.0057, "step": 2540 }, { "epoch": 2.0222134073780245, "grad_norm": 0.10028381904877147, "learning_rate": 9.306727952607895e-05, "loss": 0.0052, "step": 2550 }, { "epoch": 2.0301467671558906, "grad_norm": 0.09795247925220865, "learning_rate": 9.297912726371196e-05, "loss": 0.0052, "step": 2560 }, { "epoch": 2.0380801269337563, "grad_norm": 0.10143825231740725, "learning_rate": 9.289046038820362e-05, "loss": 0.005, "step": 2570 }, { "epoch": 2.0460134867116224, "grad_norm": 0.09415067512406232, "learning_rate": 9.28012799612121e-05, "loss": 0.005, "step": 2580 }, { "epoch": 2.053946846489488, "grad_norm": 0.09590593113830713, "learning_rate": 9.271158705054453e-05, "loss": 0.005, "step": 2590 }, { "epoch": 2.0618802062673542, "grad_norm": 0.11291094188649715, "learning_rate": 9.262138273014432e-05, "loss": 0.005, "step": 2600 }, { "epoch": 2.0698135660452204, "grad_norm": 0.10399465747009821, "learning_rate": 9.253066808007829e-05, "loss": 0.0051, "step": 2610 }, { "epoch": 2.077746925823086, "grad_norm": 0.09497580124640684, "learning_rate": 9.243944418652368e-05, "loss": 0.0051, "step": 2620 }, { "epoch": 2.085680285600952, "grad_norm": 0.0993516254441346, "learning_rate": 9.234771214175526e-05, "loss": 0.0051, "step": 2630 }, { "epoch": 2.093613645378818, "grad_norm": 0.10944500939451741, "learning_rate": 9.22554730441321e-05, "loss": 0.0051, "step": 2640 }, { "epoch": 2.101547005156684, "grad_norm": 0.09397689887537947, "learning_rate": 9.216272799808452e-05, "loss": 0.0049, "step": 2650 }, { "epoch": 2.1094803649345497, "grad_norm": 0.11607973538611317, "learning_rate": 9.20694781141009e-05, "loss": 0.005, "step": 2660 }, { "epoch": 2.117413724712416, "grad_norm": 0.10689694065246308, "learning_rate": 9.197572450871424e-05, "loss": 0.0049, "step": 2670 }, { "epoch": 2.1253470844902815, "grad_norm": 0.11797688389071613, "learning_rate": 9.188146830448895e-05, "loss": 0.0051, "step": 2680 }, { "epoch": 2.1332804442681477, "grad_norm": 0.10007478424257744, "learning_rate": 9.17867106300073e-05, "loss": 0.0048, "step": 2690 }, { "epoch": 2.1412138040460134, "grad_norm": 0.10670178367865263, "learning_rate": 9.169145261985593e-05, "loss": 0.0047, "step": 2700 }, { "epoch": 2.1491471638238795, "grad_norm": 0.11105464347530923, "learning_rate": 9.159569541461232e-05, "loss": 0.0047, "step": 2710 }, { "epoch": 2.157080523601745, "grad_norm": 0.10403957273950044, "learning_rate": 9.14994401608311e-05, "loss": 0.0047, "step": 2720 }, { "epoch": 2.1650138833796113, "grad_norm": 0.08862815220504634, "learning_rate": 9.140268801103022e-05, "loss": 0.0048, "step": 2730 }, { "epoch": 2.172947243157477, "grad_norm": 0.17406102479916283, "learning_rate": 9.130544012367737e-05, "loss": 0.0049, "step": 2740 }, { "epoch": 2.180880602935343, "grad_norm": 0.9984554300367096, "learning_rate": 9.12076976631759e-05, "loss": 0.0076, "step": 2750 }, { "epoch": 2.1888139627132093, "grad_norm": 0.1794409095899379, "learning_rate": 9.110946179985099e-05, "loss": 0.0086, "step": 2760 }, { "epoch": 2.196747322491075, "grad_norm": 0.17099885681701577, "learning_rate": 9.10107337099356e-05, "loss": 0.0072, "step": 2770 }, { "epoch": 2.204680682268941, "grad_norm": 0.13686932152618447, "learning_rate": 9.09115145755564e-05, "loss": 0.0063, "step": 2780 }, { "epoch": 2.212614042046807, "grad_norm": 0.13664389983421157, "learning_rate": 9.08118055847196e-05, "loss": 0.006, "step": 2790 }, { "epoch": 2.220547401824673, "grad_norm": 0.12880765577441727, "learning_rate": 9.071160793129675e-05, "loss": 0.0055, "step": 2800 }, { "epoch": 2.2284807616025386, "grad_norm": 0.10465888110547632, "learning_rate": 9.061092281501044e-05, "loss": 0.0052, "step": 2810 }, { "epoch": 2.2364141213804047, "grad_norm": 0.10776062892079107, "learning_rate": 9.050975144141988e-05, "loss": 0.0051, "step": 2820 }, { "epoch": 2.2443474811582704, "grad_norm": 0.09673629571825465, "learning_rate": 9.040809502190654e-05, "loss": 0.0049, "step": 2830 }, { "epoch": 2.2522808409361366, "grad_norm": 0.09846666092981253, "learning_rate": 9.030595477365967e-05, "loss": 0.0048, "step": 2840 }, { "epoch": 2.2602142007140023, "grad_norm": 0.1694566973365221, "learning_rate": 9.020333191966155e-05, "loss": 0.0049, "step": 2850 }, { "epoch": 2.2681475604918684, "grad_norm": 0.11789583478819116, "learning_rate": 9.010022768867306e-05, "loss": 0.0048, "step": 2860 }, { "epoch": 2.276080920269734, "grad_norm": 0.09848282297096807, "learning_rate": 8.999664331521887e-05, "loss": 0.0046, "step": 2870 }, { "epoch": 2.2840142800476, "grad_norm": 0.10259091930814387, "learning_rate": 8.989258003957263e-05, "loss": 0.0045, "step": 2880 }, { "epoch": 2.291947639825466, "grad_norm": 0.11543434434269667, "learning_rate": 8.978803910774218e-05, "loss": 0.0045, "step": 2890 }, { "epoch": 2.299880999603332, "grad_norm": 0.0936318653991697, "learning_rate": 8.968302177145461e-05, "loss": 0.0045, "step": 2900 }, { "epoch": 2.307814359381198, "grad_norm": 0.10391048932974932, "learning_rate": 8.957752928814123e-05, "loss": 0.0045, "step": 2910 }, { "epoch": 2.315747719159064, "grad_norm": 0.09676081245422562, "learning_rate": 8.947156292092257e-05, "loss": 0.0044, "step": 2920 }, { "epoch": 2.3236810789369295, "grad_norm": 0.09112725418934604, "learning_rate": 8.936512393859328e-05, "loss": 0.0045, "step": 2930 }, { "epoch": 2.3316144387147957, "grad_norm": 0.11256278250908659, "learning_rate": 8.925821361560684e-05, "loss": 0.0047, "step": 2940 }, { "epoch": 2.339547798492662, "grad_norm": 0.09790114136420414, "learning_rate": 8.915083323206035e-05, "loss": 0.0044, "step": 2950 }, { "epoch": 2.3474811582705275, "grad_norm": 0.09262306010758421, "learning_rate": 8.904298407367925e-05, "loss": 0.0043, "step": 2960 }, { "epoch": 2.3554145180483936, "grad_norm": 0.08727581119763789, "learning_rate": 8.893466743180187e-05, "loss": 0.0041, "step": 2970 }, { "epoch": 2.3633478778262593, "grad_norm": 0.08597923954803054, "learning_rate": 8.882588460336397e-05, "loss": 0.0042, "step": 2980 }, { "epoch": 2.3712812376041255, "grad_norm": 0.09609672668945411, "learning_rate": 8.871663689088321e-05, "loss": 0.0042, "step": 2990 }, { "epoch": 2.379214597381991, "grad_norm": 0.1367345211615606, "learning_rate": 8.860692560244361e-05, "loss": 0.0042, "step": 3000 }, { "epoch": 2.3871479571598573, "grad_norm": 0.13457964936513436, "learning_rate": 8.849675205167982e-05, "loss": 0.0051, "step": 3010 }, { "epoch": 2.395081316937723, "grad_norm": 0.12420549593254235, "learning_rate": 8.838611755776138e-05, "loss": 0.0046, "step": 3020 }, { "epoch": 2.403014676715589, "grad_norm": 0.10253650256405698, "learning_rate": 8.827502344537703e-05, "loss": 0.0045, "step": 3030 }, { "epoch": 2.410948036493455, "grad_norm": 0.08829366096110679, "learning_rate": 8.816347104471876e-05, "loss": 0.0044, "step": 3040 }, { "epoch": 2.418881396271321, "grad_norm": 0.08614517031706824, "learning_rate": 8.805146169146585e-05, "loss": 0.0042, "step": 3050 }, { "epoch": 2.426814756049187, "grad_norm": 0.0843132229122554, "learning_rate": 8.793899672676898e-05, "loss": 0.004, "step": 3060 }, { "epoch": 2.4347481158270528, "grad_norm": 0.0880241777856025, "learning_rate": 8.78260774972341e-05, "loss": 0.0042, "step": 3070 }, { "epoch": 2.4426814756049184, "grad_norm": 0.0831120676477387, "learning_rate": 8.771270535490635e-05, "loss": 0.0041, "step": 3080 }, { "epoch": 2.4506148353827846, "grad_norm": 0.09314275488458384, "learning_rate": 8.759888165725379e-05, "loss": 0.0039, "step": 3090 }, { "epoch": 2.4585481951606507, "grad_norm": 0.09035224703888926, "learning_rate": 8.748460776715131e-05, "loss": 0.004, "step": 3100 }, { "epoch": 2.4664815549385164, "grad_norm": 0.08183708454038133, "learning_rate": 8.73698850528641e-05, "loss": 0.004, "step": 3110 }, { "epoch": 2.4744149147163825, "grad_norm": 0.08278007701610446, "learning_rate": 8.725471488803141e-05, "loss": 0.0039, "step": 3120 }, { "epoch": 2.482348274494248, "grad_norm": 0.09013010599712637, "learning_rate": 8.713909865165009e-05, "loss": 0.0039, "step": 3130 }, { "epoch": 2.4902816342721144, "grad_norm": 0.0896002218106816, "learning_rate": 8.702303772805805e-05, "loss": 0.004, "step": 3140 }, { "epoch": 2.49821499404998, "grad_norm": 0.0892877463855606, "learning_rate": 8.690653350691762e-05, "loss": 0.004, "step": 3150 }, { "epoch": 2.506148353827846, "grad_norm": 0.10039406998279109, "learning_rate": 8.678958738319909e-05, "loss": 0.0038, "step": 3160 }, { "epoch": 2.514081713605712, "grad_norm": 0.08536316767911051, "learning_rate": 8.667220075716382e-05, "loss": 0.0038, "step": 3170 }, { "epoch": 2.522015073383578, "grad_norm": 0.07774575008129075, "learning_rate": 8.655437503434758e-05, "loss": 0.0038, "step": 3180 }, { "epoch": 2.5299484331614437, "grad_norm": 0.08337653599038533, "learning_rate": 8.643611162554366e-05, "loss": 0.0038, "step": 3190 }, { "epoch": 2.53788179293931, "grad_norm": 0.08902318662671499, "learning_rate": 8.631741194678607e-05, "loss": 0.0038, "step": 3200 }, { "epoch": 2.545815152717176, "grad_norm": 0.08479123006562005, "learning_rate": 8.619827741933244e-05, "loss": 0.0037, "step": 3210 }, { "epoch": 2.5537485124950416, "grad_norm": 0.08943099225166953, "learning_rate": 8.607870946964717e-05, "loss": 0.0038, "step": 3220 }, { "epoch": 2.5616818722729073, "grad_norm": 0.08748118262451259, "learning_rate": 8.595870952938421e-05, "loss": 0.0038, "step": 3230 }, { "epoch": 2.5696152320507735, "grad_norm": 0.08339662369863413, "learning_rate": 8.583827903537e-05, "loss": 0.0036, "step": 3240 }, { "epoch": 2.5775485918286396, "grad_norm": 0.07770861567087188, "learning_rate": 8.571741942958627e-05, "loss": 0.0036, "step": 3250 }, { "epoch": 2.5854819516065053, "grad_norm": 0.0714457769555736, "learning_rate": 8.559613215915265e-05, "loss": 0.0035, "step": 3260 }, { "epoch": 2.5934153113843714, "grad_norm": 0.06854478917601518, "learning_rate": 8.547441867630954e-05, "loss": 0.0035, "step": 3270 }, { "epoch": 2.601348671162237, "grad_norm": 0.09906389618096384, "learning_rate": 8.535228043840058e-05, "loss": 0.0036, "step": 3280 }, { "epoch": 2.6092820309401032, "grad_norm": 0.08286613493780488, "learning_rate": 8.522971890785521e-05, "loss": 0.0036, "step": 3290 }, { "epoch": 2.617215390717969, "grad_norm": 0.08091969882873308, "learning_rate": 8.510673555217125e-05, "loss": 0.0035, "step": 3300 }, { "epoch": 2.625148750495835, "grad_norm": 0.08069042220343992, "learning_rate": 8.498333184389722e-05, "loss": 0.0036, "step": 3310 }, { "epoch": 2.6330821102737008, "grad_norm": 0.07913790770498294, "learning_rate": 8.485950926061479e-05, "loss": 0.0036, "step": 3320 }, { "epoch": 2.641015470051567, "grad_norm": 0.09142937985311947, "learning_rate": 8.473526928492103e-05, "loss": 0.0034, "step": 3330 }, { "epoch": 2.6489488298294326, "grad_norm": 0.07555638243794358, "learning_rate": 8.461061340441067e-05, "loss": 0.0034, "step": 3340 }, { "epoch": 2.6568821896072987, "grad_norm": 0.09000204886375465, "learning_rate": 8.448554311165838e-05, "loss": 0.0036, "step": 3350 }, { "epoch": 2.664815549385165, "grad_norm": 0.08506056509395979, "learning_rate": 8.436005990420071e-05, "loss": 0.0035, "step": 3360 }, { "epoch": 2.6727489091630305, "grad_norm": 0.08790235841873126, "learning_rate": 8.423416528451835e-05, "loss": 0.0034, "step": 3370 }, { "epoch": 2.6806822689408962, "grad_norm": 0.09487435078751827, "learning_rate": 8.4107860760018e-05, "loss": 0.0034, "step": 3380 }, { "epoch": 2.6886156287187624, "grad_norm": 0.07811774747886152, "learning_rate": 8.398114784301445e-05, "loss": 0.0034, "step": 3390 }, { "epoch": 2.6965489884966285, "grad_norm": 0.0851786097650454, "learning_rate": 8.385402805071235e-05, "loss": 0.0035, "step": 3400 }, { "epoch": 2.704482348274494, "grad_norm": 0.08323706788726688, "learning_rate": 8.372650290518809e-05, "loss": 0.0033, "step": 3410 }, { "epoch": 2.7124157080523603, "grad_norm": 0.07906761057286403, "learning_rate": 8.359857393337161e-05, "loss": 0.0035, "step": 3420 }, { "epoch": 2.720349067830226, "grad_norm": 0.0873769646776392, "learning_rate": 8.347024266702807e-05, "loss": 0.0034, "step": 3430 }, { "epoch": 2.728282427608092, "grad_norm": 0.0866771190426881, "learning_rate": 8.334151064273955e-05, "loss": 0.0035, "step": 3440 }, { "epoch": 2.736215787385958, "grad_norm": 0.08592457072874347, "learning_rate": 8.321237940188659e-05, "loss": 0.0035, "step": 3450 }, { "epoch": 2.744149147163824, "grad_norm": 0.09029042338239049, "learning_rate": 8.308285049062983e-05, "loss": 0.0035, "step": 3460 }, { "epoch": 2.7520825069416897, "grad_norm": 0.10069787418677155, "learning_rate": 8.29529254598914e-05, "loss": 0.0035, "step": 3470 }, { "epoch": 2.760015866719556, "grad_norm": 0.08137290249084347, "learning_rate": 8.282260586533636e-05, "loss": 0.0034, "step": 3480 }, { "epoch": 2.7679492264974215, "grad_norm": 0.09100832267445944, "learning_rate": 8.269189326735418e-05, "loss": 0.0032, "step": 3490 }, { "epoch": 2.7758825862752876, "grad_norm": 0.08719461227787112, "learning_rate": 8.256078923103993e-05, "loss": 0.0031, "step": 3500 }, { "epoch": 2.7838159460531537, "grad_norm": 0.12636511512618173, "learning_rate": 8.24292953261756e-05, "loss": 0.0034, "step": 3510 }, { "epoch": 2.7917493058310194, "grad_norm": 0.08563728997361869, "learning_rate": 8.22974131272113e-05, "loss": 0.0038, "step": 3520 }, { "epoch": 2.799682665608885, "grad_norm": 0.08375124152454284, "learning_rate": 8.216514421324639e-05, "loss": 0.0035, "step": 3530 }, { "epoch": 2.8076160253867513, "grad_norm": 0.0829974589101843, "learning_rate": 8.203249016801058e-05, "loss": 0.0035, "step": 3540 }, { "epoch": 2.8155493851646174, "grad_norm": 0.08416366583567751, "learning_rate": 8.189945257984502e-05, "loss": 0.0036, "step": 3550 }, { "epoch": 2.823482744942483, "grad_norm": 0.07723514354970452, "learning_rate": 8.176603304168318e-05, "loss": 0.0033, "step": 3560 }, { "epoch": 2.831416104720349, "grad_norm": 0.07850248807988759, "learning_rate": 8.163223315103184e-05, "loss": 0.0032, "step": 3570 }, { "epoch": 2.839349464498215, "grad_norm": 0.08360156053041994, "learning_rate": 8.149805450995198e-05, "loss": 0.0033, "step": 3580 }, { "epoch": 2.847282824276081, "grad_norm": 0.07996676728546492, "learning_rate": 8.136349872503954e-05, "loss": 0.0034, "step": 3590 }, { "epoch": 2.8552161840539467, "grad_norm": 0.07805003653168172, "learning_rate": 8.122856740740626e-05, "loss": 0.0034, "step": 3600 }, { "epoch": 2.863149543831813, "grad_norm": 0.07943330761940516, "learning_rate": 8.109326217266027e-05, "loss": 0.0033, "step": 3610 }, { "epoch": 2.8710829036096785, "grad_norm": 0.07909904500133055, "learning_rate": 8.095758464088687e-05, "loss": 0.0032, "step": 3620 }, { "epoch": 2.8790162633875447, "grad_norm": 0.07561734334873721, "learning_rate": 8.08215364366291e-05, "loss": 0.003, "step": 3630 }, { "epoch": 2.8869496231654104, "grad_norm": 0.07840061628896748, "learning_rate": 8.068511918886822e-05, "loss": 0.003, "step": 3640 }, { "epoch": 2.8948829829432765, "grad_norm": 0.07498281959187345, "learning_rate": 8.054833453100427e-05, "loss": 0.0031, "step": 3650 }, { "epoch": 2.9028163427211426, "grad_norm": 0.09793216831718835, "learning_rate": 8.04111841008365e-05, "loss": 0.0031, "step": 3660 }, { "epoch": 2.9107497024990083, "grad_norm": 0.07141389192456346, "learning_rate": 8.02736695405438e-05, "loss": 0.0032, "step": 3670 }, { "epoch": 2.918683062276874, "grad_norm": 0.08088716446764796, "learning_rate": 8.013579249666488e-05, "loss": 0.0029, "step": 3680 }, { "epoch": 2.92661642205474, "grad_norm": 0.12012806096772784, "learning_rate": 7.99975546200788e-05, "loss": 0.0034, "step": 3690 }, { "epoch": 2.9345497818326063, "grad_norm": 0.10419395663860313, "learning_rate": 7.985895756598498e-05, "loss": 0.0033, "step": 3700 }, { "epoch": 2.942483141610472, "grad_norm": 0.07683470429710108, "learning_rate": 7.972000299388352e-05, "loss": 0.0031, "step": 3710 }, { "epoch": 2.950416501388338, "grad_norm": 0.07563645792357004, "learning_rate": 7.958069256755527e-05, "loss": 0.003, "step": 3720 }, { "epoch": 2.958349861166204, "grad_norm": 0.07871764079074618, "learning_rate": 7.944102795504191e-05, "loss": 0.003, "step": 3730 }, { "epoch": 2.96628322094407, "grad_norm": 0.08443507347764571, "learning_rate": 7.930101082862602e-05, "loss": 0.0031, "step": 3740 }, { "epoch": 2.9742165807219356, "grad_norm": 0.08198888831831717, "learning_rate": 7.916064286481098e-05, "loss": 0.0029, "step": 3750 }, { "epoch": 2.9821499404998018, "grad_norm": 0.07888594809011011, "learning_rate": 7.9019925744301e-05, "loss": 0.003, "step": 3760 }, { "epoch": 2.9900833002776674, "grad_norm": 0.07880889962382587, "learning_rate": 7.887886115198087e-05, "loss": 0.003, "step": 3770 }, { "epoch": 2.9980166600555336, "grad_norm": 0.07107516024122099, "learning_rate": 7.873745077689593e-05, "loss": 0.003, "step": 3780 }, { "epoch": 3.0055533518445063, "grad_norm": 0.060686105579501666, "learning_rate": 7.85956963122317e-05, "loss": 0.0022, "step": 3790 }, { "epoch": 3.013486711622372, "grad_norm": 0.05982380317449238, "learning_rate": 7.845359945529374e-05, "loss": 0.0018, "step": 3800 }, { "epoch": 3.021420071400238, "grad_norm": 0.05059954307818608, "learning_rate": 7.831116190748719e-05, "loss": 0.0016, "step": 3810 }, { "epoch": 3.029353431178104, "grad_norm": 0.060101418230209716, "learning_rate": 7.816838537429658e-05, "loss": 0.0016, "step": 3820 }, { "epoch": 3.03728679095597, "grad_norm": 0.060518008326890096, "learning_rate": 7.802527156526517e-05, "loss": 0.0016, "step": 3830 }, { "epoch": 3.0452201507338357, "grad_norm": 0.05336500181115672, "learning_rate": 7.788182219397472e-05, "loss": 0.0015, "step": 3840 }, { "epoch": 3.053153510511702, "grad_norm": 0.06782967481074778, "learning_rate": 7.773803897802483e-05, "loss": 0.0015, "step": 3850 }, { "epoch": 3.0610868702895675, "grad_norm": 0.04930132483617561, "learning_rate": 7.759392363901236e-05, "loss": 0.0015, "step": 3860 }, { "epoch": 3.0690202300674336, "grad_norm": 0.04567915962896414, "learning_rate": 7.744947790251098e-05, "loss": 0.0015, "step": 3870 }, { "epoch": 3.0769535898452993, "grad_norm": 0.0588565487407523, "learning_rate": 7.730470349805026e-05, "loss": 0.0015, "step": 3880 }, { "epoch": 3.0848869496231655, "grad_norm": 0.07877353465664082, "learning_rate": 7.71596021590952e-05, "loss": 0.0015, "step": 3890 }, { "epoch": 3.092820309401031, "grad_norm": 0.05917411272701816, "learning_rate": 7.701417562302528e-05, "loss": 0.0017, "step": 3900 }, { "epoch": 3.1007536691788973, "grad_norm": 0.09691522817610349, "learning_rate": 7.686842563111385e-05, "loss": 0.0018, "step": 3910 }, { "epoch": 3.108687028956763, "grad_norm": 0.059785872459323666, "learning_rate": 7.672235392850708e-05, "loss": 0.0016, "step": 3920 }, { "epoch": 3.116620388734629, "grad_norm": 0.06862066489790523, "learning_rate": 7.657596226420325e-05, "loss": 0.0016, "step": 3930 }, { "epoch": 3.1245537485124952, "grad_norm": 0.05634938862016537, "learning_rate": 7.642925239103163e-05, "loss": 0.0015, "step": 3940 }, { "epoch": 3.132487108290361, "grad_norm": 0.0988866674202798, "learning_rate": 7.628222606563164e-05, "loss": 0.0016, "step": 3950 }, { "epoch": 3.140420468068227, "grad_norm": 0.06323421055223716, "learning_rate": 7.613488504843176e-05, "loss": 0.0016, "step": 3960 }, { "epoch": 3.1483538278460927, "grad_norm": 0.06496828753543746, "learning_rate": 7.598723110362845e-05, "loss": 0.0016, "step": 3970 }, { "epoch": 3.156287187623959, "grad_norm": 0.05047269217363836, "learning_rate": 7.583926599916499e-05, "loss": 0.0016, "step": 3980 }, { "epoch": 3.1642205474018246, "grad_norm": 0.07981794761138315, "learning_rate": 7.56909915067104e-05, "loss": 0.0016, "step": 3990 }, { "epoch": 3.1721539071796907, "grad_norm": 0.06355526550615104, "learning_rate": 7.554240940163812e-05, "loss": 0.0016, "step": 4000 }, { "epoch": 3.1800872669575564, "grad_norm": 0.05546566207993693, "learning_rate": 7.539352146300487e-05, "loss": 0.0015, "step": 4010 }, { "epoch": 3.1880206267354225, "grad_norm": 0.05831731824989324, "learning_rate": 7.524432947352922e-05, "loss": 0.0015, "step": 4020 }, { "epoch": 3.195953986513288, "grad_norm": 0.05644067963581909, "learning_rate": 7.50948352195704e-05, "loss": 0.0015, "step": 4030 }, { "epoch": 3.2038873462911543, "grad_norm": 0.06895904309712178, "learning_rate": 7.49450404911067e-05, "loss": 0.0015, "step": 4040 }, { "epoch": 3.21182070606902, "grad_norm": 0.05534826402311216, "learning_rate": 7.479494708171427e-05, "loss": 0.0015, "step": 4050 }, { "epoch": 3.219754065846886, "grad_norm": 0.06151938192236407, "learning_rate": 7.464455678854551e-05, "loss": 0.0016, "step": 4060 }, { "epoch": 3.227687425624752, "grad_norm": 0.06417106663804124, "learning_rate": 7.44938714123075e-05, "loss": 0.0016, "step": 4070 }, { "epoch": 3.235620785402618, "grad_norm": 0.058478732718022945, "learning_rate": 7.434289275724063e-05, "loss": 0.0015, "step": 4080 }, { "epoch": 3.243554145180484, "grad_norm": 0.4226080235552495, "learning_rate": 7.419162263109675e-05, "loss": 0.0021, "step": 4090 }, { "epoch": 3.25148750495835, "grad_norm": 0.16262254877787727, "learning_rate": 7.404006284511782e-05, "loss": 0.0025, "step": 4100 }, { "epoch": 3.259420864736216, "grad_norm": 0.08967509079982082, "learning_rate": 7.38882152140139e-05, "loss": 0.002, "step": 4110 }, { "epoch": 3.2673542245140816, "grad_norm": 0.06008768673622745, "learning_rate": 7.373608155594167e-05, "loss": 0.0019, "step": 4120 }, { "epoch": 3.2752875842919478, "grad_norm": 0.14308961842178997, "learning_rate": 7.358366369248255e-05, "loss": 0.003, "step": 4130 }, { "epoch": 3.2832209440698135, "grad_norm": 4.156605079038476, "learning_rate": 7.343096344862093e-05, "loss": 0.0788, "step": 4140 }, { "epoch": 3.2911543038476796, "grad_norm": 0.7626973075773847, "learning_rate": 7.327798265272224e-05, "loss": 0.0528, "step": 4150 }, { "epoch": 3.2990876636255453, "grad_norm": 0.37088441188641186, "learning_rate": 7.31247231365112e-05, "loss": 0.013, "step": 4160 }, { "epoch": 3.3070210234034114, "grad_norm": 0.1590129844637634, "learning_rate": 7.29711867350497e-05, "loss": 0.0097, "step": 4170 }, { "epoch": 3.314954383181277, "grad_norm": 0.1357911428134929, "learning_rate": 7.281737528671507e-05, "loss": 0.005, "step": 4180 }, { "epoch": 3.3228877429591432, "grad_norm": 0.08100811758802455, "learning_rate": 7.266329063317778e-05, "loss": 0.0036, "step": 4190 }, { "epoch": 3.330821102737009, "grad_norm": 0.07955198580197219, "learning_rate": 7.250893461937964e-05, "loss": 0.0031, "step": 4200 }, { "epoch": 3.338754462514875, "grad_norm": 0.07005374446623218, "learning_rate": 7.23543090935116e-05, "loss": 0.0025, "step": 4210 }, { "epoch": 3.3466878222927408, "grad_norm": 0.08419437721011566, "learning_rate": 7.219941590699156e-05, "loss": 0.0022, "step": 4220 }, { "epoch": 3.354621182070607, "grad_norm": 0.05546188028178132, "learning_rate": 7.204425691444236e-05, "loss": 0.002, "step": 4230 }, { "epoch": 3.362554541848473, "grad_norm": 0.07348909097869522, "learning_rate": 7.188883397366943e-05, "loss": 0.002, "step": 4240 }, { "epoch": 3.3704879016263387, "grad_norm": 0.07334595012784965, "learning_rate": 7.17331489456386e-05, "loss": 0.0019, "step": 4250 }, { "epoch": 3.378421261404205, "grad_norm": 0.06827889472097157, "learning_rate": 7.157720369445386e-05, "loss": 0.0019, "step": 4260 }, { "epoch": 3.3863546211820705, "grad_norm": 0.05033182681675953, "learning_rate": 7.142100008733494e-05, "loss": 0.0018, "step": 4270 }, { "epoch": 3.3942879809599367, "grad_norm": 0.04985166432095069, "learning_rate": 7.126453999459505e-05, "loss": 0.0018, "step": 4280 }, { "epoch": 3.4022213407378024, "grad_norm": 0.05970974793998439, "learning_rate": 7.110782528961842e-05, "loss": 0.0017, "step": 4290 }, { "epoch": 3.4101547005156685, "grad_norm": 0.05451034459006807, "learning_rate": 7.095085784883792e-05, "loss": 0.0017, "step": 4300 }, { "epoch": 3.418088060293534, "grad_norm": 0.05577884840521623, "learning_rate": 7.079363955171257e-05, "loss": 0.0017, "step": 4310 }, { "epoch": 3.4260214200714003, "grad_norm": 0.06258779672375052, "learning_rate": 7.063617228070498e-05, "loss": 0.0017, "step": 4320 }, { "epoch": 3.433954779849266, "grad_norm": 0.06090809119187366, "learning_rate": 7.047845792125893e-05, "loss": 0.0017, "step": 4330 }, { "epoch": 3.441888139627132, "grad_norm": 0.056143636734633705, "learning_rate": 7.032049836177671e-05, "loss": 0.0016, "step": 4340 }, { "epoch": 3.449821499404998, "grad_norm": 0.055220090308156375, "learning_rate": 7.016229549359648e-05, "loss": 0.0016, "step": 4350 }, { "epoch": 3.457754859182864, "grad_norm": 0.059231060158582226, "learning_rate": 7.000385121096975e-05, "loss": 0.0016, "step": 4360 }, { "epoch": 3.4656882189607296, "grad_norm": 0.05605051801453069, "learning_rate": 6.984516741103857e-05, "loss": 0.0016, "step": 4370 }, { "epoch": 3.473621578738596, "grad_norm": 0.05628551484003177, "learning_rate": 6.968624599381285e-05, "loss": 0.0015, "step": 4380 }, { "epoch": 3.481554938516462, "grad_norm": 0.05896304496693092, "learning_rate": 6.952708886214768e-05, "loss": 0.0016, "step": 4390 }, { "epoch": 3.4894882982943276, "grad_norm": 0.056007923596926114, "learning_rate": 6.936769792172042e-05, "loss": 0.0015, "step": 4400 }, { "epoch": 3.4974216580721937, "grad_norm": 0.049821212778268426, "learning_rate": 6.920807508100804e-05, "loss": 0.0015, "step": 4410 }, { "epoch": 3.5053550178500594, "grad_norm": 0.04949906919634206, "learning_rate": 6.904822225126411e-05, "loss": 0.0014, "step": 4420 }, { "epoch": 3.5132883776279256, "grad_norm": 0.04892431803118073, "learning_rate": 6.8888141346496e-05, "loss": 0.0014, "step": 4430 }, { "epoch": 3.5212217374057913, "grad_norm": 0.059262410041627835, "learning_rate": 6.872783428344198e-05, "loss": 0.0015, "step": 4440 }, { "epoch": 3.5291550971836574, "grad_norm": 0.05006363328541868, "learning_rate": 6.856730298154823e-05, "loss": 0.0016, "step": 4450 }, { "epoch": 3.537088456961523, "grad_norm": 0.05450851458344542, "learning_rate": 6.840654936294585e-05, "loss": 0.0015, "step": 4460 }, { "epoch": 3.545021816739389, "grad_norm": 0.05687525869904552, "learning_rate": 6.824557535242786e-05, "loss": 0.0015, "step": 4470 }, { "epoch": 3.552955176517255, "grad_norm": 0.0678294702080013, "learning_rate": 6.808438287742618e-05, "loss": 0.0014, "step": 4480 }, { "epoch": 3.560888536295121, "grad_norm": 0.044162865371410856, "learning_rate": 6.79229738679885e-05, "loss": 0.0015, "step": 4490 }, { "epoch": 3.568821896072987, "grad_norm": 0.053724357670147424, "learning_rate": 6.776135025675523e-05, "loss": 0.0014, "step": 4500 }, { "epoch": 3.576755255850853, "grad_norm": 0.06436082409488973, "learning_rate": 6.75995139789363e-05, "loss": 0.0016, "step": 4510 }, { "epoch": 3.5846886156287185, "grad_norm": 0.06645729689807983, "learning_rate": 6.743746697228804e-05, "loss": 0.0015, "step": 4520 }, { "epoch": 3.5926219754065847, "grad_norm": 0.059281925674580835, "learning_rate": 6.727521117708993e-05, "loss": 0.0014, "step": 4530 }, { "epoch": 3.600555335184451, "grad_norm": 0.05636974000947677, "learning_rate": 6.711274853612137e-05, "loss": 0.0014, "step": 4540 }, { "epoch": 3.6084886949623165, "grad_norm": 0.05142329707461039, "learning_rate": 6.695008099463853e-05, "loss": 0.0014, "step": 4550 }, { "epoch": 3.616422054740182, "grad_norm": 0.05433458010722858, "learning_rate": 6.678721050035088e-05, "loss": 0.0014, "step": 4560 }, { "epoch": 3.6243554145180483, "grad_norm": 0.05501104677814428, "learning_rate": 6.6624139003398e-05, "loss": 0.0014, "step": 4570 }, { "epoch": 3.6322887742959145, "grad_norm": 0.04801360376445415, "learning_rate": 6.646086845632615e-05, "loss": 0.0014, "step": 4580 }, { "epoch": 3.64022213407378, "grad_norm": 0.05152512046973344, "learning_rate": 6.6297400814065e-05, "loss": 0.0014, "step": 4590 }, { "epoch": 3.6481554938516463, "grad_norm": 0.054037454993505006, "learning_rate": 6.613373803390405e-05, "loss": 0.0013, "step": 4600 }, { "epoch": 3.656088853629512, "grad_norm": 0.05137680901018049, "learning_rate": 6.59698820754694e-05, "loss": 0.0014, "step": 4610 }, { "epoch": 3.664022213407378, "grad_norm": 0.061084195875252156, "learning_rate": 6.58058349007001e-05, "loss": 0.0013, "step": 4620 }, { "epoch": 3.671955573185244, "grad_norm": 0.06458112894213228, "learning_rate": 6.564159847382478e-05, "loss": 0.0013, "step": 4630 }, { "epoch": 3.67988893296311, "grad_norm": 0.05129120947458432, "learning_rate": 6.547717476133809e-05, "loss": 0.0013, "step": 4640 }, { "epoch": 3.687822292740976, "grad_norm": 0.05469221295157648, "learning_rate": 6.531256573197715e-05, "loss": 0.0013, "step": 4650 }, { "epoch": 3.6957556525188417, "grad_norm": 0.0515566521230157, "learning_rate": 6.514777335669798e-05, "loss": 0.0013, "step": 4660 }, { "epoch": 3.7036890122967074, "grad_norm": 0.07038699029197822, "learning_rate": 6.49827996086519e-05, "loss": 0.0014, "step": 4670 }, { "epoch": 3.7116223720745736, "grad_norm": 0.06061467356501644, "learning_rate": 6.481764646316192e-05, "loss": 0.0014, "step": 4680 }, { "epoch": 3.7195557318524397, "grad_norm": 0.048898602299191234, "learning_rate": 6.465231589769905e-05, "loss": 0.0013, "step": 4690 }, { "epoch": 3.7274890916303054, "grad_norm": 0.05972630787281949, "learning_rate": 6.448680989185869e-05, "loss": 0.0013, "step": 4700 }, { "epoch": 3.735422451408171, "grad_norm": 0.05117170766880633, "learning_rate": 6.432113042733686e-05, "loss": 0.0013, "step": 4710 }, { "epoch": 3.743355811186037, "grad_norm": 0.05669351523571635, "learning_rate": 6.415527948790649e-05, "loss": 0.0013, "step": 4720 }, { "epoch": 3.7512891709639034, "grad_norm": 0.049784976209432216, "learning_rate": 6.398925905939367e-05, "loss": 0.0013, "step": 4730 }, { "epoch": 3.759222530741769, "grad_norm": 0.05834920790567014, "learning_rate": 6.382307112965394e-05, "loss": 0.0013, "step": 4740 }, { "epoch": 3.767155890519635, "grad_norm": 0.05750598148813775, "learning_rate": 6.365671768854835e-05, "loss": 0.0015, "step": 4750 }, { "epoch": 3.775089250297501, "grad_norm": 0.05557523614151382, "learning_rate": 6.349020072791976e-05, "loss": 0.0014, "step": 4760 }, { "epoch": 3.783022610075367, "grad_norm": 0.04781438681838105, "learning_rate": 6.332352224156887e-05, "loss": 0.0014, "step": 4770 }, { "epoch": 3.7909559698532327, "grad_norm": 0.0528813065250596, "learning_rate": 6.315668422523053e-05, "loss": 0.0014, "step": 4780 }, { "epoch": 3.798889329631099, "grad_norm": 0.05505484068608557, "learning_rate": 6.298968867654961e-05, "loss": 0.0014, "step": 4790 }, { "epoch": 3.806822689408965, "grad_norm": 0.04508415196364383, "learning_rate": 6.282253759505731e-05, "loss": 0.0013, "step": 4800 }, { "epoch": 3.8147560491868306, "grad_norm": 0.05196706262852486, "learning_rate": 6.265523298214707e-05, "loss": 0.0013, "step": 4810 }, { "epoch": 3.8226894089646963, "grad_norm": 0.05890686699864618, "learning_rate": 6.248777684105062e-05, "loss": 0.0013, "step": 4820 }, { "epoch": 3.8306227687425625, "grad_norm": 0.05912029515266175, "learning_rate": 6.232017117681403e-05, "loss": 0.0013, "step": 4830 }, { "epoch": 3.8385561285204286, "grad_norm": 0.05305680024094495, "learning_rate": 6.215241799627373e-05, "loss": 0.0013, "step": 4840 }, { "epoch": 3.8464894882982943, "grad_norm": 0.052113325491484945, "learning_rate": 6.198451930803242e-05, "loss": 0.0013, "step": 4850 }, { "epoch": 3.85442284807616, "grad_norm": 0.054673155435883906, "learning_rate": 6.181647712243503e-05, "loss": 0.0013, "step": 4860 }, { "epoch": 3.862356207854026, "grad_norm": 0.048627634796191435, "learning_rate": 6.164829345154471e-05, "loss": 0.0013, "step": 4870 }, { "epoch": 3.8702895676318922, "grad_norm": 0.054609957546295336, "learning_rate": 6.147997030911866e-05, "loss": 0.0013, "step": 4880 }, { "epoch": 3.878222927409758, "grad_norm": 0.056277167574111144, "learning_rate": 6.131150971058404e-05, "loss": 0.0013, "step": 4890 }, { "epoch": 3.886156287187624, "grad_norm": 0.05854408866736552, "learning_rate": 6.114291367301385e-05, "loss": 0.0013, "step": 4900 }, { "epoch": 3.8940896469654898, "grad_norm": 0.060745685291232664, "learning_rate": 6.097418421510287e-05, "loss": 0.0013, "step": 4910 }, { "epoch": 3.902023006743356, "grad_norm": 0.05706651964066593, "learning_rate": 6.080532335714322e-05, "loss": 0.0013, "step": 4920 }, { "epoch": 3.9099563665212216, "grad_norm": 0.05155534417787901, "learning_rate": 6.063633312100051e-05, "loss": 0.0013, "step": 4930 }, { "epoch": 3.9178897262990877, "grad_norm": 0.0755756222414549, "learning_rate": 6.046721553008936e-05, "loss": 0.0013, "step": 4940 }, { "epoch": 3.925823086076954, "grad_norm": 0.064098231717572, "learning_rate": 6.0297972609349364e-05, "loss": 0.0015, "step": 4950 }, { "epoch": 3.9337564458548195, "grad_norm": 0.07208222838668767, "learning_rate": 6.0128606385220686e-05, "loss": 0.0013, "step": 4960 }, { "epoch": 3.9416898056326852, "grad_norm": 0.04670376869660676, "learning_rate": 5.995911888561992e-05, "loss": 0.0013, "step": 4970 }, { "epoch": 3.9496231654105514, "grad_norm": 0.037721432923949694, "learning_rate": 5.9789512139915735e-05, "loss": 0.0013, "step": 4980 }, { "epoch": 3.9575565251884175, "grad_norm": 0.04503369158507431, "learning_rate": 5.961978817890461e-05, "loss": 0.0013, "step": 4990 }, { "epoch": 3.965489884966283, "grad_norm": 0.04814703292701314, "learning_rate": 5.9449949034786514e-05, "loss": 0.0013, "step": 5000 }, { "epoch": 3.973423244744149, "grad_norm": 0.05395632243420775, "learning_rate": 5.927999674114053e-05, "loss": 0.0013, "step": 5010 }, { "epoch": 3.981356604522015, "grad_norm": 0.04824303233268365, "learning_rate": 5.91099333329006e-05, "loss": 0.0012, "step": 5020 }, { "epoch": 3.989289964299881, "grad_norm": 0.05116862820337914, "learning_rate": 5.893976084633106e-05, "loss": 0.0013, "step": 5030 }, { "epoch": 3.997223324077747, "grad_norm": 0.06097336040692719, "learning_rate": 5.876948131900234e-05, "loss": 0.0013, "step": 5040 }, { "epoch": 4.00476001586672, "grad_norm": 0.03948711641312406, "learning_rate": 5.859909678976648e-05, "loss": 0.001, "step": 5050 }, { "epoch": 4.012693375644585, "grad_norm": 0.047669637965671155, "learning_rate": 5.8428609298732836e-05, "loss": 0.0008, "step": 5060 }, { "epoch": 4.020626735422451, "grad_norm": 0.03982222100086071, "learning_rate": 5.825802088724348e-05, "loss": 0.0006, "step": 5070 }, { "epoch": 4.0285600952003175, "grad_norm": 0.03583925773344066, "learning_rate": 5.808733359784897e-05, "loss": 0.0006, "step": 5080 }, { "epoch": 4.036493454978183, "grad_norm": 0.0331602931996562, "learning_rate": 5.791654947428371e-05, "loss": 0.0006, "step": 5090 }, { "epoch": 4.044426814756049, "grad_norm": 0.03488588122356885, "learning_rate": 5.77456705614416e-05, "loss": 0.0006, "step": 5100 }, { "epoch": 4.0523601745339155, "grad_norm": 0.025024125641375062, "learning_rate": 5.7574698905351475e-05, "loss": 0.0005, "step": 5110 }, { "epoch": 4.060293534311781, "grad_norm": 0.03343247341697893, "learning_rate": 5.740363655315265e-05, "loss": 0.0005, "step": 5120 }, { "epoch": 4.068226894089647, "grad_norm": 0.03416031300894815, "learning_rate": 5.7232485553070383e-05, "loss": 0.0005, "step": 5130 }, { "epoch": 4.076160253867513, "grad_norm": 0.04647553297876997, "learning_rate": 5.706124795439138e-05, "loss": 0.0005, "step": 5140 }, { "epoch": 4.084093613645379, "grad_norm": 0.038665019280888835, "learning_rate": 5.688992580743924e-05, "loss": 0.0005, "step": 5150 }, { "epoch": 4.092026973423245, "grad_norm": 0.029457703913610616, "learning_rate": 5.6718521163549885e-05, "loss": 0.0005, "step": 5160 }, { "epoch": 4.0999603332011105, "grad_norm": 0.03205436854892031, "learning_rate": 5.6547036075047035e-05, "loss": 0.0005, "step": 5170 }, { "epoch": 4.107893692978976, "grad_norm": 0.03548651434079761, "learning_rate": 5.6375472595217614e-05, "loss": 0.0006, "step": 5180 }, { "epoch": 4.115827052756843, "grad_norm": 0.023340606326179264, "learning_rate": 5.620383277828718e-05, "loss": 0.0004, "step": 5190 }, { "epoch": 4.1237604125347085, "grad_norm": 0.03013140881909005, "learning_rate": 5.6032118679395306e-05, "loss": 0.0005, "step": 5200 }, { "epoch": 4.131693772312574, "grad_norm": 0.024781813458224264, "learning_rate": 5.5860332354570985e-05, "loss": 0.0005, "step": 5210 }, { "epoch": 4.139627132090441, "grad_norm": 0.029427475700634597, "learning_rate": 5.5688475860708e-05, "loss": 0.0005, "step": 5220 }, { "epoch": 4.147560491868306, "grad_norm": 0.02507137782101749, "learning_rate": 5.551655125554035e-05, "loss": 0.0004, "step": 5230 }, { "epoch": 4.155493851646172, "grad_norm": 0.01985576974963966, "learning_rate": 5.53445605976175e-05, "loss": 0.0004, "step": 5240 }, { "epoch": 4.163427211424038, "grad_norm": 0.026341526839853933, "learning_rate": 5.5172505946279894e-05, "loss": 0.0005, "step": 5250 }, { "epoch": 4.171360571201904, "grad_norm": 0.03509155676388275, "learning_rate": 5.500038936163411e-05, "loss": 0.0005, "step": 5260 }, { "epoch": 4.17929393097977, "grad_norm": 0.025101418622115294, "learning_rate": 5.482821290452834e-05, "loss": 0.0004, "step": 5270 }, { "epoch": 4.187227290757636, "grad_norm": 0.033498919182429504, "learning_rate": 5.465597863652765e-05, "loss": 0.0005, "step": 5280 }, { "epoch": 4.1951606505355015, "grad_norm": 0.029583188234299698, "learning_rate": 5.448368861988933e-05, "loss": 0.0004, "step": 5290 }, { "epoch": 4.203094010313368, "grad_norm": 0.03682857553563353, "learning_rate": 5.431134491753813e-05, "loss": 0.0005, "step": 5300 }, { "epoch": 4.211027370091234, "grad_norm": 0.03434337385458972, "learning_rate": 5.413894959304168e-05, "loss": 0.0005, "step": 5310 }, { "epoch": 4.218960729869099, "grad_norm": 0.028923209180060364, "learning_rate": 5.396650471058562e-05, "loss": 0.0005, "step": 5320 }, { "epoch": 4.226894089646965, "grad_norm": 0.026095792495590818, "learning_rate": 5.3794012334949075e-05, "loss": 0.0004, "step": 5330 }, { "epoch": 4.234827449424832, "grad_norm": 0.03550909823913911, "learning_rate": 5.362147453147975e-05, "loss": 0.0004, "step": 5340 }, { "epoch": 4.242760809202697, "grad_norm": 0.03945495186795905, "learning_rate": 5.344889336606933e-05, "loss": 0.0004, "step": 5350 }, { "epoch": 4.250694168980563, "grad_norm": 0.029173350252406232, "learning_rate": 5.3276270905128675e-05, "loss": 0.0004, "step": 5360 }, { "epoch": 4.25862752875843, "grad_norm": 0.0375999793275123, "learning_rate": 5.3103609215563125e-05, "loss": 0.0005, "step": 5370 }, { "epoch": 4.266560888536295, "grad_norm": 0.02789775378665049, "learning_rate": 5.293091036474769e-05, "loss": 0.0005, "step": 5380 }, { "epoch": 4.274494248314161, "grad_norm": 0.03589960703041237, "learning_rate": 5.275817642050234e-05, "loss": 0.0005, "step": 5390 }, { "epoch": 4.282427608092027, "grad_norm": 0.03508224658571038, "learning_rate": 5.258540945106729e-05, "loss": 0.0005, "step": 5400 }, { "epoch": 4.290360967869893, "grad_norm": 0.025447066120805977, "learning_rate": 5.2412611525078106e-05, "loss": 0.0005, "step": 5410 }, { "epoch": 4.298294327647759, "grad_norm": 0.03799788280088589, "learning_rate": 5.223978471154107e-05, "loss": 0.0005, "step": 5420 }, { "epoch": 4.306227687425625, "grad_norm": 0.029741900002682585, "learning_rate": 5.206693107980831e-05, "loss": 0.0005, "step": 5430 }, { "epoch": 4.31416104720349, "grad_norm": 0.03289905799514981, "learning_rate": 5.189405269955312e-05, "loss": 0.0005, "step": 5440 }, { "epoch": 4.322094406981357, "grad_norm": 0.03489065700819638, "learning_rate": 5.172115164074507e-05, "loss": 0.0005, "step": 5450 }, { "epoch": 4.330027766759223, "grad_norm": 0.02962757785456289, "learning_rate": 5.1548229973625296e-05, "loss": 0.0005, "step": 5460 }, { "epoch": 4.337961126537088, "grad_norm": 0.03100476460865329, "learning_rate": 5.137528976868165e-05, "loss": 0.0005, "step": 5470 }, { "epoch": 4.345894486314954, "grad_norm": 0.034344930737001345, "learning_rate": 5.1202333096624036e-05, "loss": 0.0005, "step": 5480 }, { "epoch": 4.353827846092821, "grad_norm": 0.0343313834333581, "learning_rate": 5.102936202835943e-05, "loss": 0.0004, "step": 5490 }, { "epoch": 4.361761205870686, "grad_norm": 0.030593025156389328, "learning_rate": 5.085637863496725e-05, "loss": 0.0005, "step": 5500 }, { "epoch": 4.369694565648552, "grad_norm": 0.031194634989817536, "learning_rate": 5.068338498767444e-05, "loss": 0.0005, "step": 5510 }, { "epoch": 4.3776279254264185, "grad_norm": 0.031104019786448875, "learning_rate": 5.0510383157830777e-05, "loss": 0.0005, "step": 5520 }, { "epoch": 4.385561285204284, "grad_norm": 0.0328205756669499, "learning_rate": 5.0337375216883954e-05, "loss": 0.0005, "step": 5530 }, { "epoch": 4.39349464498215, "grad_norm": 0.035024505104467765, "learning_rate": 5.016436323635486e-05, "loss": 0.0005, "step": 5540 }, { "epoch": 4.401428004760016, "grad_norm": 0.031765751908714335, "learning_rate": 4.999134928781278e-05, "loss": 0.0005, "step": 5550 }, { "epoch": 4.409361364537882, "grad_norm": 0.03903469766425352, "learning_rate": 4.981833544285048e-05, "loss": 0.0005, "step": 5560 }, { "epoch": 4.417294724315748, "grad_norm": 0.028692941300885733, "learning_rate": 4.964532377305963e-05, "loss": 0.0005, "step": 5570 }, { "epoch": 4.425228084093614, "grad_norm": 0.04007279927278608, "learning_rate": 4.947231635000569e-05, "loss": 0.0005, "step": 5580 }, { "epoch": 4.433161443871479, "grad_norm": 0.03985818429022389, "learning_rate": 4.929931524520338e-05, "loss": 0.0005, "step": 5590 }, { "epoch": 4.441094803649346, "grad_norm": 0.02822211861943674, "learning_rate": 4.912632253009173e-05, "loss": 0.0005, "step": 5600 }, { "epoch": 4.4490281634272115, "grad_norm": 0.03665255333424402, "learning_rate": 4.8953340276009325e-05, "loss": 0.0005, "step": 5610 }, { "epoch": 4.456961523205077, "grad_norm": 0.039760473358577075, "learning_rate": 4.87803705541695e-05, "loss": 0.0005, "step": 5620 }, { "epoch": 4.464894882982943, "grad_norm": 0.02758795219516536, "learning_rate": 4.8607415435635536e-05, "loss": 0.0005, "step": 5630 }, { "epoch": 4.4728282427608095, "grad_norm": 0.03129744153259456, "learning_rate": 4.843447699129579e-05, "loss": 0.0005, "step": 5640 }, { "epoch": 4.480761602538675, "grad_norm": 0.0381157808724866, "learning_rate": 4.826155729183912e-05, "loss": 0.0005, "step": 5650 }, { "epoch": 4.488694962316541, "grad_norm": 0.04115789516388352, "learning_rate": 4.8088658407729776e-05, "loss": 0.0005, "step": 5660 }, { "epoch": 4.496628322094407, "grad_norm": 0.05136266396571763, "learning_rate": 4.79157824091829e-05, "loss": 0.0005, "step": 5670 }, { "epoch": 4.504561681872273, "grad_norm": 0.04899896354854659, "learning_rate": 4.774293136613953e-05, "loss": 0.0006, "step": 5680 }, { "epoch": 4.512495041650139, "grad_norm": 0.03828807165976926, "learning_rate": 4.757010734824193e-05, "loss": 0.0006, "step": 5690 }, { "epoch": 4.5204284014280045, "grad_norm": 0.037710618834915846, "learning_rate": 4.739731242480881e-05, "loss": 0.0005, "step": 5700 }, { "epoch": 4.52836176120587, "grad_norm": 0.028998043505727787, "learning_rate": 4.7224548664810444e-05, "loss": 0.0004, "step": 5710 }, { "epoch": 4.536295120983737, "grad_norm": 0.035527619221099246, "learning_rate": 4.705181813684404e-05, "loss": 0.0006, "step": 5720 }, { "epoch": 4.5442284807616025, "grad_norm": 0.03930881671835381, "learning_rate": 4.6879122909108844e-05, "loss": 0.0005, "step": 5730 }, { "epoch": 4.552161840539468, "grad_norm": 0.02623370176397575, "learning_rate": 4.670646504938144e-05, "loss": 0.0005, "step": 5740 }, { "epoch": 4.560095200317335, "grad_norm": 0.03576531388235279, "learning_rate": 4.6533846624991043e-05, "loss": 0.0005, "step": 5750 }, { "epoch": 4.5680285600952, "grad_norm": 0.03158724188002893, "learning_rate": 4.636126970279461e-05, "loss": 0.0004, "step": 5760 }, { "epoch": 4.575961919873066, "grad_norm": 0.03921289850715458, "learning_rate": 4.6188736349152225e-05, "loss": 0.0005, "step": 5770 }, { "epoch": 4.583895279650932, "grad_norm": 0.044688185112616206, "learning_rate": 4.601624862990228e-05, "loss": 0.0005, "step": 5780 }, { "epoch": 4.591828639428798, "grad_norm": 0.039755842800635544, "learning_rate": 4.584380861033671e-05, "loss": 0.0005, "step": 5790 }, { "epoch": 4.599761999206664, "grad_norm": 0.03294594460998042, "learning_rate": 4.567141835517644e-05, "loss": 0.0005, "step": 5800 }, { "epoch": 4.60769535898453, "grad_norm": 0.05009057784151513, "learning_rate": 4.549907992854639e-05, "loss": 0.0005, "step": 5810 }, { "epoch": 4.615628718762396, "grad_norm": 0.046053559786593055, "learning_rate": 4.532679539395104e-05, "loss": 0.0005, "step": 5820 }, { "epoch": 4.623562078540262, "grad_norm": 0.039006903419640394, "learning_rate": 4.51545668142495e-05, "loss": 0.0005, "step": 5830 }, { "epoch": 4.631495438318128, "grad_norm": 0.04463671008772757, "learning_rate": 4.498239625163093e-05, "loss": 0.0006, "step": 5840 }, { "epoch": 4.639428798095993, "grad_norm": 0.031137232115157384, "learning_rate": 4.4810285767589835e-05, "loss": 0.0006, "step": 5850 }, { "epoch": 4.647362157873859, "grad_norm": 0.03745908886910379, "learning_rate": 4.463823742290134e-05, "loss": 0.0006, "step": 5860 }, { "epoch": 4.655295517651726, "grad_norm": 0.05371357767057646, "learning_rate": 4.446625327759657e-05, "loss": 0.0005, "step": 5870 }, { "epoch": 4.663228877429591, "grad_norm": 0.035529600565140186, "learning_rate": 4.429433539093792e-05, "loss": 0.0005, "step": 5880 }, { "epoch": 4.671162237207457, "grad_norm": 0.030677599118528482, "learning_rate": 4.412248582139444e-05, "loss": 0.0005, "step": 5890 }, { "epoch": 4.679095596985324, "grad_norm": 0.031952951539284166, "learning_rate": 4.395070662661722e-05, "loss": 0.0005, "step": 5900 }, { "epoch": 4.687028956763189, "grad_norm": 0.03397022273245611, "learning_rate": 4.377899986341466e-05, "loss": 0.0006, "step": 5910 }, { "epoch": 4.694962316541055, "grad_norm": 0.03900068885326374, "learning_rate": 4.360736758772794e-05, "loss": 0.0006, "step": 5920 }, { "epoch": 4.702895676318921, "grad_norm": 0.03806299844477056, "learning_rate": 4.3435811854606335e-05, "loss": 0.0005, "step": 5930 }, { "epoch": 4.710829036096787, "grad_norm": 0.03537172940295102, "learning_rate": 4.326433471818261e-05, "loss": 0.0005, "step": 5940 }, { "epoch": 4.718762395874653, "grad_norm": 0.035458508367708315, "learning_rate": 4.309293823164852e-05, "loss": 0.0005, "step": 5950 }, { "epoch": 4.726695755652519, "grad_norm": 0.0366686019533167, "learning_rate": 4.292162444723008e-05, "loss": 0.0005, "step": 5960 }, { "epoch": 4.734629115430385, "grad_norm": 0.03384684837850387, "learning_rate": 4.2750395416163106e-05, "loss": 0.0005, "step": 5970 }, { "epoch": 4.742562475208251, "grad_norm": 0.03593836394964473, "learning_rate": 4.257925318866861e-05, "loss": 0.0005, "step": 5980 }, { "epoch": 4.750495834986117, "grad_norm": 0.03622199824620536, "learning_rate": 4.240819981392823e-05, "loss": 0.0006, "step": 5990 }, { "epoch": 4.758429194763982, "grad_norm": 0.03428405143677215, "learning_rate": 4.223723734005978e-05, "loss": 0.0006, "step": 6000 }, { "epoch": 4.766362554541848, "grad_norm": 0.04137582937975696, "learning_rate": 4.206636781409258e-05, "loss": 0.0006, "step": 6010 }, { "epoch": 4.774295914319715, "grad_norm": 0.04373682676579509, "learning_rate": 4.189559328194312e-05, "loss": 0.0006, "step": 6020 }, { "epoch": 4.78222927409758, "grad_norm": 0.036311779128229894, "learning_rate": 4.172491578839043e-05, "loss": 0.0005, "step": 6030 }, { "epoch": 4.790162633875446, "grad_norm": 0.03406801123180307, "learning_rate": 4.155433737705162e-05, "loss": 0.0005, "step": 6040 }, { "epoch": 4.7980959936533125, "grad_norm": 0.0533922296433466, "learning_rate": 4.138386009035749e-05, "loss": 0.0005, "step": 6050 }, { "epoch": 4.806029353431178, "grad_norm": 0.0410802518555118, "learning_rate": 4.121348596952798e-05, "loss": 0.0005, "step": 6060 }, { "epoch": 4.813962713209044, "grad_norm": 0.033715390824791844, "learning_rate": 4.104321705454779e-05, "loss": 0.0006, "step": 6070 }, { "epoch": 4.82189607298691, "grad_norm": 0.03424950116292573, "learning_rate": 4.087305538414192e-05, "loss": 0.0005, "step": 6080 }, { "epoch": 4.829829432764776, "grad_norm": 0.04660075901825731, "learning_rate": 4.070300299575123e-05, "loss": 0.0005, "step": 6090 }, { "epoch": 4.837762792542642, "grad_norm": 0.05153809652691355, "learning_rate": 4.053306192550819e-05, "loss": 0.0005, "step": 6100 }, { "epoch": 4.8456961523205075, "grad_norm": 0.03341617795987288, "learning_rate": 4.036323420821227e-05, "loss": 0.0006, "step": 6110 }, { "epoch": 4.853629512098374, "grad_norm": 0.0361025305297232, "learning_rate": 4.0193521877305815e-05, "loss": 0.0006, "step": 6120 }, { "epoch": 4.86156287187624, "grad_norm": 0.04344739830359098, "learning_rate": 4.0023926964849503e-05, "loss": 0.0006, "step": 6130 }, { "epoch": 4.8694962316541055, "grad_norm": 0.03566737754084799, "learning_rate": 3.985445150149812e-05, "loss": 0.0005, "step": 6140 }, { "epoch": 4.877429591431971, "grad_norm": 0.034887319424396206, "learning_rate": 3.968509751647624e-05, "loss": 0.0005, "step": 6150 }, { "epoch": 4.885362951209837, "grad_norm": 0.04637642134375086, "learning_rate": 3.951586703755388e-05, "loss": 0.0005, "step": 6160 }, { "epoch": 4.8932963109877035, "grad_norm": 0.0561954722369419, "learning_rate": 3.9346762091022286e-05, "loss": 0.0006, "step": 6170 }, { "epoch": 4.901229670765569, "grad_norm": 0.04139225182798093, "learning_rate": 3.917778470166959e-05, "loss": 0.0006, "step": 6180 }, { "epoch": 4.909163030543435, "grad_norm": 0.03443358523120335, "learning_rate": 3.900893689275663e-05, "loss": 0.0006, "step": 6190 }, { "epoch": 4.917096390321301, "grad_norm": 0.03431380104504621, "learning_rate": 3.884022068599273e-05, "loss": 0.0005, "step": 6200 }, { "epoch": 4.925029750099167, "grad_norm": 0.04733066048715638, "learning_rate": 3.867163810151142e-05, "loss": 0.0006, "step": 6210 }, { "epoch": 4.932963109877033, "grad_norm": 0.042893994726431656, "learning_rate": 3.8503191157846334e-05, "loss": 0.0005, "step": 6220 }, { "epoch": 4.9408964696548985, "grad_norm": 0.03499098677282287, "learning_rate": 3.833488187190701e-05, "loss": 0.0006, "step": 6230 }, { "epoch": 4.948829829432765, "grad_norm": 0.037502624465460176, "learning_rate": 3.816671225895465e-05, "loss": 0.0005, "step": 6240 }, { "epoch": 4.956763189210631, "grad_norm": 0.04002484200540745, "learning_rate": 3.799868433257822e-05, "loss": 0.0005, "step": 6250 }, { "epoch": 4.964696548988496, "grad_norm": 0.03493086657992255, "learning_rate": 3.783080010467002e-05, "loss": 0.0005, "step": 6260 }, { "epoch": 4.972629908766363, "grad_norm": 0.04471000889901979, "learning_rate": 3.7663061585401934e-05, "loss": 0.0005, "step": 6270 }, { "epoch": 4.980563268544229, "grad_norm": 0.03624349590050735, "learning_rate": 3.749547078320108e-05, "loss": 0.0005, "step": 6280 }, { "epoch": 4.988496628322094, "grad_norm": 0.0260679574535491, "learning_rate": 3.732802970472591e-05, "loss": 0.0005, "step": 6290 }, { "epoch": 4.99642998809996, "grad_norm": 0.032708395948222266, "learning_rate": 3.716074035484217e-05, "loss": 0.0005, "step": 6300 }, { "epoch": 5.003966679888933, "grad_norm": 0.02356393622798595, "learning_rate": 3.699360473659884e-05, "loss": 0.0004, "step": 6310 }, { "epoch": 5.0119000396667985, "grad_norm": 0.030265651948823406, "learning_rate": 3.6826624851204235e-05, "loss": 0.0003, "step": 6320 }, { "epoch": 5.019833399444665, "grad_norm": 0.034346781358936866, "learning_rate": 3.665980269800195e-05, "loss": 0.0003, "step": 6330 }, { "epoch": 5.027766759222531, "grad_norm": 0.022042568415201184, "learning_rate": 3.6493140274446956e-05, "loss": 0.0002, "step": 6340 }, { "epoch": 5.0357001190003965, "grad_norm": 0.015964635458602636, "learning_rate": 3.6326639576081736e-05, "loss": 0.0002, "step": 6350 }, { "epoch": 5.043633478778262, "grad_norm": 0.02141195229548223, "learning_rate": 3.6160302596512306e-05, "loss": 0.0002, "step": 6360 }, { "epoch": 5.051566838556129, "grad_norm": 0.020960358356966326, "learning_rate": 3.599413132738442e-05, "loss": 0.0002, "step": 6370 }, { "epoch": 5.0595001983339944, "grad_norm": 0.016085199303949093, "learning_rate": 3.582812775835967e-05, "loss": 0.0002, "step": 6380 }, { "epoch": 5.06743355811186, "grad_norm": 0.024489125240812853, "learning_rate": 3.566229387709166e-05, "loss": 0.0002, "step": 6390 }, { "epoch": 5.075366917889727, "grad_norm": 0.01899430001222409, "learning_rate": 3.549663166920228e-05, "loss": 0.0002, "step": 6400 }, { "epoch": 5.083300277667592, "grad_norm": 0.017618632910005552, "learning_rate": 3.53311431182578e-05, "loss": 0.0002, "step": 6410 }, { "epoch": 5.091233637445458, "grad_norm": 0.019192313779158823, "learning_rate": 3.516583020574532e-05, "loss": 0.0002, "step": 6420 }, { "epoch": 5.099166997223324, "grad_norm": 0.015755209037333868, "learning_rate": 3.5000694911048786e-05, "loss": 0.0002, "step": 6430 }, { "epoch": 5.10710035700119, "grad_norm": 0.024627836886461137, "learning_rate": 3.48357392114255e-05, "loss": 0.0002, "step": 6440 }, { "epoch": 5.115033716779056, "grad_norm": 0.02081657793996447, "learning_rate": 3.467096508198238e-05, "loss": 0.0002, "step": 6450 }, { "epoch": 5.122967076556922, "grad_norm": 0.03144908977146878, "learning_rate": 3.450637449565226e-05, "loss": 0.0002, "step": 6460 }, { "epoch": 5.130900436334787, "grad_norm": 0.01966486711658438, "learning_rate": 3.434196942317036e-05, "loss": 0.0002, "step": 6470 }, { "epoch": 5.138833796112654, "grad_norm": 0.018404707165616697, "learning_rate": 3.417775183305061e-05, "loss": 0.0002, "step": 6480 }, { "epoch": 5.14676715589052, "grad_norm": 0.01971609636403434, "learning_rate": 3.4013723691562075e-05, "loss": 0.0002, "step": 6490 }, { "epoch": 5.154700515668385, "grad_norm": 0.02204756262770674, "learning_rate": 3.384988696270551e-05, "loss": 0.0002, "step": 6500 }, { "epoch": 5.162633875446251, "grad_norm": 0.025646812443634467, "learning_rate": 3.368624360818974e-05, "loss": 0.0002, "step": 6510 }, { "epoch": 5.170567235224118, "grad_norm": 0.014589233724654323, "learning_rate": 3.352279558740823e-05, "loss": 0.0002, "step": 6520 }, { "epoch": 5.178500595001983, "grad_norm": 0.026356250203196954, "learning_rate": 3.3359544857415604e-05, "loss": 0.0002, "step": 6530 }, { "epoch": 5.186433954779849, "grad_norm": 0.022556246851185607, "learning_rate": 3.3196493372904175e-05, "loss": 0.0002, "step": 6540 }, { "epoch": 5.194367314557716, "grad_norm": 0.018649837004425372, "learning_rate": 3.3033643086180655e-05, "loss": 0.0002, "step": 6550 }, { "epoch": 5.202300674335581, "grad_norm": 0.012399955523642373, "learning_rate": 3.287099594714259e-05, "loss": 0.0002, "step": 6560 }, { "epoch": 5.210234034113447, "grad_norm": 0.02242060681788757, "learning_rate": 3.2708553903255266e-05, "loss": 0.0002, "step": 6570 }, { "epoch": 5.218167393891313, "grad_norm": 0.02494821461029386, "learning_rate": 3.2546318899528126e-05, "loss": 0.0002, "step": 6580 }, { "epoch": 5.226100753669179, "grad_norm": 0.014392880099203894, "learning_rate": 3.238429287849166e-05, "loss": 0.0002, "step": 6590 }, { "epoch": 5.234034113447045, "grad_norm": 0.020621007936465537, "learning_rate": 3.22224777801741e-05, "loss": 0.0002, "step": 6600 }, { "epoch": 5.241967473224911, "grad_norm": 0.015239896276040802, "learning_rate": 3.206087554207815e-05, "loss": 0.0002, "step": 6610 }, { "epoch": 5.249900833002776, "grad_norm": 0.02554475940570255, "learning_rate": 3.189948809915786e-05, "loss": 0.0002, "step": 6620 }, { "epoch": 5.257834192780643, "grad_norm": 0.01766584278677565, "learning_rate": 3.173831738379539e-05, "loss": 0.0002, "step": 6630 }, { "epoch": 5.265767552558509, "grad_norm": 0.010670892014779415, "learning_rate": 3.157736532577789e-05, "loss": 0.0002, "step": 6640 }, { "epoch": 5.273700912336374, "grad_norm": 0.017620381766375547, "learning_rate": 3.141663385227441e-05, "loss": 0.0002, "step": 6650 }, { "epoch": 5.28163427211424, "grad_norm": 0.02386083956530999, "learning_rate": 3.1256124887812846e-05, "loss": 0.0002, "step": 6660 }, { "epoch": 5.2895676318921065, "grad_norm": 0.02176652303568066, "learning_rate": 3.109584035425686e-05, "loss": 0.0002, "step": 6670 }, { "epoch": 5.297500991669972, "grad_norm": 0.02546996515513367, "learning_rate": 3.093578217078283e-05, "loss": 0.0002, "step": 6680 }, { "epoch": 5.305434351447838, "grad_norm": 0.021978207402568426, "learning_rate": 3.077595225385697e-05, "loss": 0.0002, "step": 6690 }, { "epoch": 5.313367711225704, "grad_norm": 0.01746622319467791, "learning_rate": 3.0616352517212333e-05, "loss": 0.0002, "step": 6700 }, { "epoch": 5.32130107100357, "grad_norm": 0.01813118204104857, "learning_rate": 3.045698487182582e-05, "loss": 0.0002, "step": 6710 }, { "epoch": 5.329234430781436, "grad_norm": 0.021293381295506352, "learning_rate": 3.0297851225895514e-05, "loss": 0.0002, "step": 6720 }, { "epoch": 5.337167790559302, "grad_norm": 0.019197024726085598, "learning_rate": 3.0138953484817538e-05, "loss": 0.0002, "step": 6730 }, { "epoch": 5.345101150337168, "grad_norm": 0.01717763088589667, "learning_rate": 2.998029355116348e-05, "loss": 0.0002, "step": 6740 }, { "epoch": 5.353034510115034, "grad_norm": 0.03317515928898506, "learning_rate": 2.9821873324657523e-05, "loss": 0.0002, "step": 6750 }, { "epoch": 5.3609678698928995, "grad_norm": 0.0173587328768651, "learning_rate": 2.9663694702153676e-05, "loss": 0.0002, "step": 6760 }, { "epoch": 5.368901229670765, "grad_norm": 0.01513031712905152, "learning_rate": 2.950575957761311e-05, "loss": 0.0002, "step": 6770 }, { "epoch": 5.376834589448632, "grad_norm": 0.025054349014416034, "learning_rate": 2.934806984208146e-05, "loss": 0.0002, "step": 6780 }, { "epoch": 5.3847679492264975, "grad_norm": 0.018499956482419755, "learning_rate": 2.9190627383666123e-05, "loss": 0.0002, "step": 6790 }, { "epoch": 5.392701309004363, "grad_norm": 0.023171987381965235, "learning_rate": 2.9033434087513823e-05, "loss": 0.0002, "step": 6800 }, { "epoch": 5.400634668782229, "grad_norm": 0.026437498108805427, "learning_rate": 2.887649183578781e-05, "loss": 0.0002, "step": 6810 }, { "epoch": 5.408568028560095, "grad_norm": 0.01719043179559044, "learning_rate": 2.8719802507645512e-05, "loss": 0.0002, "step": 6820 }, { "epoch": 5.416501388337961, "grad_norm": 0.016476084825567765, "learning_rate": 2.8563367979215923e-05, "loss": 0.0002, "step": 6830 }, { "epoch": 5.424434748115827, "grad_norm": 0.018370014045199303, "learning_rate": 2.840719012357719e-05, "loss": 0.0002, "step": 6840 }, { "epoch": 5.4323681078936925, "grad_norm": 0.016896202259222674, "learning_rate": 2.8251270810734187e-05, "loss": 0.0002, "step": 6850 }, { "epoch": 5.440301467671559, "grad_norm": 0.026715331808124586, "learning_rate": 2.8095611907596063e-05, "loss": 0.0002, "step": 6860 }, { "epoch": 5.448234827449425, "grad_norm": 0.021974323944017538, "learning_rate": 2.794021527795402e-05, "loss": 0.0002, "step": 6870 }, { "epoch": 5.4561681872272905, "grad_norm": 0.011872479543677119, "learning_rate": 2.778508278245883e-05, "loss": 0.0002, "step": 6880 }, { "epoch": 5.464101547005157, "grad_norm": 0.025449305193458935, "learning_rate": 2.7630216278598686e-05, "loss": 0.0002, "step": 6890 }, { "epoch": 5.472034906783023, "grad_norm": 0.022266506174472914, "learning_rate": 2.7475617620676895e-05, "loss": 0.0002, "step": 6900 }, { "epoch": 5.479968266560888, "grad_norm": 0.012463980949771454, "learning_rate": 2.732128865978968e-05, "loss": 0.0001, "step": 6910 }, { "epoch": 5.487901626338754, "grad_norm": 0.01970893118506279, "learning_rate": 2.7167231243804103e-05, "loss": 0.0002, "step": 6920 }, { "epoch": 5.495834986116621, "grad_norm": 0.036361307147187286, "learning_rate": 2.7013447217335762e-05, "loss": 0.0002, "step": 6930 }, { "epoch": 5.503768345894486, "grad_norm": 0.01610261562945896, "learning_rate": 2.685993842172686e-05, "loss": 0.0002, "step": 6940 }, { "epoch": 5.511701705672352, "grad_norm": 0.06650240227903932, "learning_rate": 2.6706706695024153e-05, "loss": 0.0002, "step": 6950 }, { "epoch": 5.519635065450219, "grad_norm": 0.016264669439375323, "learning_rate": 2.6553753871956822e-05, "loss": 0.0002, "step": 6960 }, { "epoch": 5.527568425228084, "grad_norm": 0.015497113999516445, "learning_rate": 2.6401081783914615e-05, "loss": 0.0002, "step": 6970 }, { "epoch": 5.53550178500595, "grad_norm": 0.021740610152103258, "learning_rate": 2.6248692258925873e-05, "loss": 0.0002, "step": 6980 }, { "epoch": 5.543435144783816, "grad_norm": 0.018457275613358366, "learning_rate": 2.609658712163565e-05, "loss": 0.0002, "step": 6990 }, { "epoch": 5.551368504561681, "grad_norm": 0.010591722671401276, "learning_rate": 2.5944768193283888e-05, "loss": 0.0002, "step": 7000 }, { "epoch": 5.559301864339548, "grad_norm": 0.02051791714180893, "learning_rate": 2.5793237291683548e-05, "loss": 0.0002, "step": 7010 }, { "epoch": 5.567235224117414, "grad_norm": 0.023490830377365157, "learning_rate": 2.5641996231198905e-05, "loss": 0.0002, "step": 7020 }, { "epoch": 5.575168583895279, "grad_norm": 0.02930447130987043, "learning_rate": 2.549104682272384e-05, "loss": 0.0002, "step": 7030 }, { "epoch": 5.583101943673146, "grad_norm": 0.028475396159914408, "learning_rate": 2.534039087366008e-05, "loss": 0.0002, "step": 7040 }, { "epoch": 5.591035303451012, "grad_norm": 0.021782700354104772, "learning_rate": 2.5190030187895607e-05, "loss": 0.0002, "step": 7050 }, { "epoch": 5.598968663228877, "grad_norm": 0.016194571464477902, "learning_rate": 2.5039966565783048e-05, "loss": 0.0002, "step": 7060 }, { "epoch": 5.606902023006743, "grad_norm": 0.03666417202420427, "learning_rate": 2.4890201804118118e-05, "loss": 0.0002, "step": 7070 }, { "epoch": 5.61483538278461, "grad_norm": 0.021039221248939265, "learning_rate": 2.474073769611812e-05, "loss": 0.0002, "step": 7080 }, { "epoch": 5.622768742562475, "grad_norm": 0.014287064889706932, "learning_rate": 2.4591576031400448e-05, "loss": 0.0002, "step": 7090 }, { "epoch": 5.630702102340341, "grad_norm": 0.02244623715762893, "learning_rate": 2.4442718595961206e-05, "loss": 0.0002, "step": 7100 }, { "epoch": 5.6386354621182075, "grad_norm": 0.01872352069361142, "learning_rate": 2.429416717215378e-05, "loss": 0.0002, "step": 7110 }, { "epoch": 5.646568821896073, "grad_norm": 0.019330511252306934, "learning_rate": 2.4145923538667426e-05, "loss": 0.0002, "step": 7120 }, { "epoch": 5.654502181673939, "grad_norm": 0.014456111144621683, "learning_rate": 2.399798947050619e-05, "loss": 0.0002, "step": 7130 }, { "epoch": 5.662435541451805, "grad_norm": 0.019941430664394487, "learning_rate": 2.3850366738967428e-05, "loss": 0.0002, "step": 7140 }, { "epoch": 5.67036890122967, "grad_norm": 0.01404718320094563, "learning_rate": 2.3703057111620707e-05, "loss": 0.0001, "step": 7150 }, { "epoch": 5.678302261007537, "grad_norm": 0.01986729273911193, "learning_rate": 2.355606235228664e-05, "loss": 0.0002, "step": 7160 }, { "epoch": 5.686235620785403, "grad_norm": 0.019702701423304155, "learning_rate": 2.3409384221015713e-05, "loss": 0.0002, "step": 7170 }, { "epoch": 5.694168980563268, "grad_norm": 0.0193427837891847, "learning_rate": 2.3263024474067308e-05, "loss": 0.0002, "step": 7180 }, { "epoch": 5.702102340341135, "grad_norm": 0.018901586404883854, "learning_rate": 2.311698486388856e-05, "loss": 0.0002, "step": 7190 }, { "epoch": 5.7100357001190005, "grad_norm": 0.01644867285535476, "learning_rate": 2.2971267139093433e-05, "loss": 0.0002, "step": 7200 }, { "epoch": 5.717969059896866, "grad_norm": 0.01610048218584611, "learning_rate": 2.2825873044441797e-05, "loss": 0.0002, "step": 7210 }, { "epoch": 5.725902419674732, "grad_norm": 0.014905068968530293, "learning_rate": 2.2680804320818515e-05, "loss": 0.0002, "step": 7220 }, { "epoch": 5.7338357794525985, "grad_norm": 0.02238053090195203, "learning_rate": 2.253606270521258e-05, "loss": 0.0002, "step": 7230 }, { "epoch": 5.741769139230464, "grad_norm": 0.01553956648466813, "learning_rate": 2.239164993069634e-05, "loss": 0.0002, "step": 7240 }, { "epoch": 5.74970249900833, "grad_norm": 0.01819646073812141, "learning_rate": 2.224756772640479e-05, "loss": 0.0002, "step": 7250 }, { "epoch": 5.757635858786196, "grad_norm": 0.021274892059503548, "learning_rate": 2.2103817817514794e-05, "loss": 0.0002, "step": 7260 }, { "epoch": 5.765569218564062, "grad_norm": 0.021862688281632903, "learning_rate": 2.1960401925224388e-05, "loss": 0.0002, "step": 7270 }, { "epoch": 5.773502578341928, "grad_norm": 0.012083274738059514, "learning_rate": 2.1817321766732363e-05, "loss": 0.0002, "step": 7280 }, { "epoch": 5.7814359381197935, "grad_norm": 0.018553574215593092, "learning_rate": 2.1674579055217485e-05, "loss": 0.0002, "step": 7290 }, { "epoch": 5.789369297897659, "grad_norm": 0.023292330357945906, "learning_rate": 2.153217549981818e-05, "loss": 0.0002, "step": 7300 }, { "epoch": 5.797302657675526, "grad_norm": 0.009024956091772902, "learning_rate": 2.139011280561184e-05, "loss": 0.0002, "step": 7310 }, { "epoch": 5.8052360174533915, "grad_norm": 0.012123543126878187, "learning_rate": 2.1248392673594596e-05, "loss": 0.0001, "step": 7320 }, { "epoch": 5.813169377231257, "grad_norm": 0.021396980248990182, "learning_rate": 2.110701680066094e-05, "loss": 0.0002, "step": 7330 }, { "epoch": 5.821102737009124, "grad_norm": 0.01960372255603896, "learning_rate": 2.0965986879583282e-05, "loss": 0.0002, "step": 7340 }, { "epoch": 5.829036096786989, "grad_norm": 0.023333924413006867, "learning_rate": 2.0825304598991796e-05, "loss": 0.0002, "step": 7350 }, { "epoch": 5.836969456564855, "grad_norm": 0.013901522359384202, "learning_rate": 2.0684971643354144e-05, "loss": 0.0001, "step": 7360 }, { "epoch": 5.844902816342721, "grad_norm": 0.017930151782612508, "learning_rate": 2.054498969295534e-05, "loss": 0.0001, "step": 7370 }, { "epoch": 5.852836176120587, "grad_norm": 0.022554525794180696, "learning_rate": 2.04053604238776e-05, "loss": 0.0001, "step": 7380 }, { "epoch": 5.860769535898453, "grad_norm": 0.014900887338642126, "learning_rate": 2.0266085507980283e-05, "loss": 0.0002, "step": 7390 }, { "epoch": 5.868702895676319, "grad_norm": 0.023172511561526655, "learning_rate": 2.0127166612879928e-05, "loss": 0.0002, "step": 7400 }, { "epoch": 5.876636255454185, "grad_norm": 0.018548117621674808, "learning_rate": 1.998860540193021e-05, "loss": 0.0002, "step": 7410 }, { "epoch": 5.884569615232051, "grad_norm": 0.013304242847968033, "learning_rate": 1.9850403534201983e-05, "loss": 0.0002, "step": 7420 }, { "epoch": 5.892502975009917, "grad_norm": 0.01967598052135776, "learning_rate": 1.9712562664463602e-05, "loss": 0.0002, "step": 7430 }, { "epoch": 5.900436334787782, "grad_norm": 0.021017506242715006, "learning_rate": 1.95750844431609e-05, "loss": 0.0002, "step": 7440 }, { "epoch": 5.908369694565648, "grad_norm": 0.026497676925691795, "learning_rate": 1.94379705163976e-05, "loss": 0.0002, "step": 7450 }, { "epoch": 5.916303054343515, "grad_norm": 0.027292274193786416, "learning_rate": 1.930122252591543e-05, "loss": 0.0002, "step": 7460 }, { "epoch": 5.92423641412138, "grad_norm": 0.024287017983999428, "learning_rate": 1.91648421090746e-05, "loss": 0.0002, "step": 7470 }, { "epoch": 5.932169773899246, "grad_norm": 0.02442653600388173, "learning_rate": 1.902883089883418e-05, "loss": 0.0001, "step": 7480 }, { "epoch": 5.940103133677113, "grad_norm": 0.01045605936644682, "learning_rate": 1.8893190523732502e-05, "loss": 0.0001, "step": 7490 }, { "epoch": 5.948036493454978, "grad_norm": 0.012709825255949471, "learning_rate": 1.8757922607867674e-05, "loss": 0.0001, "step": 7500 }, { "epoch": 5.955969853232844, "grad_norm": 0.014410929816608277, "learning_rate": 1.8623028770878137e-05, "loss": 0.0001, "step": 7510 }, { "epoch": 5.96390321301071, "grad_norm": 0.02029576549571299, "learning_rate": 1.8488510627923293e-05, "loss": 0.0002, "step": 7520 }, { "epoch": 5.971836572788576, "grad_norm": 0.028269671850583093, "learning_rate": 1.835436978966415e-05, "loss": 0.0001, "step": 7530 }, { "epoch": 5.979769932566442, "grad_norm": 0.012014144262357295, "learning_rate": 1.8220607862243998e-05, "loss": 0.0002, "step": 7540 }, { "epoch": 5.987703292344308, "grad_norm": 0.021492090609164464, "learning_rate": 1.8087226447269283e-05, "loss": 0.0001, "step": 7550 }, { "epoch": 5.995636652122174, "grad_norm": 0.012629069588511219, "learning_rate": 1.795422714179032e-05, "loss": 0.0002, "step": 7560 }, { "epoch": 6.003173343911146, "grad_norm": 0.011533624404427682, "learning_rate": 1.78216115382822e-05, "loss": 0.0001, "step": 7570 }, { "epoch": 6.011106703689013, "grad_norm": 0.02113691818544228, "learning_rate": 1.768938122462576e-05, "loss": 0.0001, "step": 7580 }, { "epoch": 6.019040063466878, "grad_norm": 0.01577669752749457, "learning_rate": 1.7557537784088533e-05, "loss": 0.0001, "step": 7590 }, { "epoch": 6.026973423244744, "grad_norm": 0.005599698698857246, "learning_rate": 1.7426082795305855e-05, "loss": 0.0001, "step": 7600 }, { "epoch": 6.03490678302261, "grad_norm": 0.003717254102951315, "learning_rate": 1.7295017832261807e-05, "loss": 0.0001, "step": 7610 }, { "epoch": 6.042840142800476, "grad_norm": 0.00694598158357044, "learning_rate": 1.7164344464270547e-05, "loss": 0.0001, "step": 7620 }, { "epoch": 6.050773502578342, "grad_norm": 0.02184595432737024, "learning_rate": 1.703406425595746e-05, "loss": 0.0001, "step": 7630 }, { "epoch": 6.058706862356208, "grad_norm": 0.015095600319938752, "learning_rate": 1.690417876724037e-05, "loss": 0.0001, "step": 7640 }, { "epoch": 6.066640222134073, "grad_norm": 0.004702523735727343, "learning_rate": 1.6774689553310914e-05, "loss": 0.0001, "step": 7650 }, { "epoch": 6.07457358191194, "grad_norm": 0.007872665713307917, "learning_rate": 1.664559816461591e-05, "loss": 0.0001, "step": 7660 }, { "epoch": 6.082506941689806, "grad_norm": 0.003136208147533061, "learning_rate": 1.6516906146838797e-05, "loss": 0.0001, "step": 7670 }, { "epoch": 6.090440301467671, "grad_norm": 0.0033339738207317005, "learning_rate": 1.6388615040881112e-05, "loss": 0.0001, "step": 7680 }, { "epoch": 6.098373661245537, "grad_norm": 0.017713177509744313, "learning_rate": 1.6260726382844045e-05, "loss": 0.0001, "step": 7690 }, { "epoch": 6.106307021023404, "grad_norm": 0.0035563956736452667, "learning_rate": 1.6133241704010095e-05, "loss": 0.0001, "step": 7700 }, { "epoch": 6.114240380801269, "grad_norm": 0.009330011664238864, "learning_rate": 1.6006162530824644e-05, "loss": 0.0001, "step": 7710 }, { "epoch": 6.122173740579135, "grad_norm": 0.011123998601190626, "learning_rate": 1.5879490384877743e-05, "loss": 0.0001, "step": 7720 }, { "epoch": 6.130107100357002, "grad_norm": 0.002705928202103928, "learning_rate": 1.5753226782885883e-05, "loss": 0.0001, "step": 7730 }, { "epoch": 6.138040460134867, "grad_norm": 0.0023955822626147523, "learning_rate": 1.5627373236673807e-05, "loss": 0.0001, "step": 7740 }, { "epoch": 6.145973819912733, "grad_norm": 0.005670929396809823, "learning_rate": 1.5501931253156504e-05, "loss": 0.0001, "step": 7750 }, { "epoch": 6.153907179690599, "grad_norm": 0.0022117086569406145, "learning_rate": 1.537690233432098e-05, "loss": 0.0001, "step": 7760 }, { "epoch": 6.161840539468465, "grad_norm": 0.00482733635648037, "learning_rate": 1.5252287977208451e-05, "loss": 0.0001, "step": 7770 }, { "epoch": 6.169773899246331, "grad_norm": 0.0031761653300209657, "learning_rate": 1.5128089673896384e-05, "loss": 0.0001, "step": 7780 }, { "epoch": 6.177707259024197, "grad_norm": 0.0037913782491479854, "learning_rate": 1.5004308911480542e-05, "loss": 0.0001, "step": 7790 }, { "epoch": 6.185640618802062, "grad_norm": 0.011710443947767653, "learning_rate": 1.4880947172057257e-05, "loss": 0.0001, "step": 7800 }, { "epoch": 6.193573978579929, "grad_norm": 0.0027321625829977497, "learning_rate": 1.4758005932705677e-05, "loss": 0.0001, "step": 7810 }, { "epoch": 6.2015073383577946, "grad_norm": 0.0035432977297696266, "learning_rate": 1.4635486665470043e-05, "loss": 0.0001, "step": 7820 }, { "epoch": 6.20944069813566, "grad_norm": 0.00361898277525427, "learning_rate": 1.4513390837342116e-05, "loss": 0.0001, "step": 7830 }, { "epoch": 6.217374057913526, "grad_norm": 0.005677162576617751, "learning_rate": 1.4391719910243545e-05, "loss": 0.0001, "step": 7840 }, { "epoch": 6.2253074176913925, "grad_norm": 0.0025307706345765704, "learning_rate": 1.427047534100846e-05, "loss": 0.0001, "step": 7850 }, { "epoch": 6.233240777469258, "grad_norm": 0.007475128015275152, "learning_rate": 1.4149658581365921e-05, "loss": 0.0001, "step": 7860 }, { "epoch": 6.241174137247124, "grad_norm": 0.007675457302854138, "learning_rate": 1.4029271077922595e-05, "loss": 0.0001, "step": 7870 }, { "epoch": 6.2491074970249905, "grad_norm": 0.0059657871459808595, "learning_rate": 1.3909314272145436e-05, "loss": 0.0001, "step": 7880 }, { "epoch": 6.257040856802856, "grad_norm": 0.0035916644832025553, "learning_rate": 1.3789789600344377e-05, "loss": 0.0001, "step": 7890 }, { "epoch": 6.264974216580722, "grad_norm": 0.0018890639303880236, "learning_rate": 1.367069849365526e-05, "loss": 0.0001, "step": 7900 }, { "epoch": 6.2729075763585875, "grad_norm": 0.00872666855647669, "learning_rate": 1.355204237802249e-05, "loss": 0.0001, "step": 7910 }, { "epoch": 6.280840936136454, "grad_norm": 0.009402394590679793, "learning_rate": 1.3433822674182139e-05, "loss": 0.0001, "step": 7920 }, { "epoch": 6.28877429591432, "grad_norm": 0.0029095442713952687, "learning_rate": 1.3316040797644908e-05, "loss": 0.0001, "step": 7930 }, { "epoch": 6.2967076556921855, "grad_norm": 0.0026675520104683347, "learning_rate": 1.3198698158679101e-05, "loss": 0.0001, "step": 7940 }, { "epoch": 6.304641015470051, "grad_norm": 0.002585967410511431, "learning_rate": 1.3081796162293796e-05, "loss": 0.0001, "step": 7950 }, { "epoch": 6.312574375247918, "grad_norm": 0.003696586206022054, "learning_rate": 1.2965336208222013e-05, "loss": 0.0001, "step": 7960 }, { "epoch": 6.3205077350257834, "grad_norm": 0.008547175502304006, "learning_rate": 1.2849319690903932e-05, "loss": 0.0001, "step": 7970 }, { "epoch": 6.328441094803649, "grad_norm": 0.012604909611857767, "learning_rate": 1.2733747999470292e-05, "loss": 0.0001, "step": 7980 }, { "epoch": 6.336374454581515, "grad_norm": 0.0021832477759726973, "learning_rate": 1.2618622517725558e-05, "loss": 0.0001, "step": 7990 }, { "epoch": 6.344307814359381, "grad_norm": 0.002599077457749917, "learning_rate": 1.2503944624131592e-05, "loss": 0.0001, "step": 8000 }, { "epoch": 6.352241174137247, "grad_norm": 0.0036187129840702876, "learning_rate": 1.238971569179097e-05, "loss": 0.0001, "step": 8010 }, { "epoch": 6.360174533915113, "grad_norm": 0.015994396924192456, "learning_rate": 1.2275937088430628e-05, "loss": 0.0001, "step": 8020 }, { "epoch": 6.368107893692979, "grad_norm": 0.0037236256932899753, "learning_rate": 1.2162610176385441e-05, "loss": 0.0001, "step": 8030 }, { "epoch": 6.376041253470845, "grad_norm": 0.0034664584032565345, "learning_rate": 1.2049736312581943e-05, "loss": 0.0001, "step": 8040 }, { "epoch": 6.383974613248711, "grad_norm": 0.006253640846222608, "learning_rate": 1.1937316848522101e-05, "loss": 0.0001, "step": 8050 }, { "epoch": 6.391907973026576, "grad_norm": 0.0018825865449311637, "learning_rate": 1.1825353130267025e-05, "loss": 0.0001, "step": 8060 }, { "epoch": 6.399841332804443, "grad_norm": 0.011783938034103112, "learning_rate": 1.1713846498420971e-05, "loss": 0.0001, "step": 8070 }, { "epoch": 6.407774692582309, "grad_norm": 0.003962082751478537, "learning_rate": 1.1602798288115263e-05, "loss": 0.0001, "step": 8080 }, { "epoch": 6.415708052360174, "grad_norm": 0.013224288963615423, "learning_rate": 1.1492209828992256e-05, "loss": 0.0001, "step": 8090 }, { "epoch": 6.42364141213804, "grad_norm": 0.0013442861680880809, "learning_rate": 1.1382082445189451e-05, "loss": 0.0001, "step": 8100 }, { "epoch": 6.431574771915907, "grad_norm": 0.0017602975972726455, "learning_rate": 1.1272417455323642e-05, "loss": 0.0001, "step": 8110 }, { "epoch": 6.439508131693772, "grad_norm": 0.003216462083754715, "learning_rate": 1.1163216172475106e-05, "loss": 0.0001, "step": 8120 }, { "epoch": 6.447441491471638, "grad_norm": 0.0021929530822368828, "learning_rate": 1.105447990417195e-05, "loss": 0.0001, "step": 8130 }, { "epoch": 6.455374851249504, "grad_norm": 0.0038238905504317604, "learning_rate": 1.0946209952374303e-05, "loss": 0.0001, "step": 8140 }, { "epoch": 6.46330821102737, "grad_norm": 0.002317942288983082, "learning_rate": 1.083840761345894e-05, "loss": 0.0001, "step": 8150 }, { "epoch": 6.471241570805236, "grad_norm": 0.010913741502800998, "learning_rate": 1.0731074178203559e-05, "loss": 0.0001, "step": 8160 }, { "epoch": 6.479174930583102, "grad_norm": 0.004040908756499227, "learning_rate": 1.062421093177145e-05, "loss": 0.0001, "step": 8170 }, { "epoch": 6.487108290360968, "grad_norm": 0.011990959198424234, "learning_rate": 1.0517819153696068e-05, "loss": 0.0001, "step": 8180 }, { "epoch": 6.495041650138834, "grad_norm": 0.0028543893686056618, "learning_rate": 1.0411900117865681e-05, "loss": 0.0001, "step": 8190 }, { "epoch": 6.5029750099167, "grad_norm": 0.005237899035322297, "learning_rate": 1.0306455092508228e-05, "loss": 0.0001, "step": 8200 }, { "epoch": 6.510908369694565, "grad_norm": 0.0077348176079224855, "learning_rate": 1.0201485340175954e-05, "loss": 0.0001, "step": 8210 }, { "epoch": 6.518841729472432, "grad_norm": 0.0026071526308679032, "learning_rate": 1.0096992117730453e-05, "loss": 0.0001, "step": 8220 }, { "epoch": 6.526775089250298, "grad_norm": 0.0029496858570794895, "learning_rate": 9.99297667632757e-06, "loss": 0.0001, "step": 8230 }, { "epoch": 6.534708449028163, "grad_norm": 0.003080748242877757, "learning_rate": 9.889440261402377e-06, "loss": 0.0001, "step": 8240 }, { "epoch": 6.542641808806029, "grad_norm": 0.004183472745177338, "learning_rate": 9.786384112654313e-06, "loss": 0.0001, "step": 8250 }, { "epoch": 6.5505751685838955, "grad_norm": 0.0036521893671293978, "learning_rate": 9.683809464032313e-06, "loss": 0.0001, "step": 8260 }, { "epoch": 6.558508528361761, "grad_norm": 0.013912062736315433, "learning_rate": 9.581717543720032e-06, "loss": 0.0001, "step": 8270 }, { "epoch": 6.566441888139627, "grad_norm": 0.003945324330868108, "learning_rate": 9.480109574121199e-06, "loss": 0.0001, "step": 8280 }, { "epoch": 6.574375247917493, "grad_norm": 0.0034799483358547946, "learning_rate": 9.378986771844839e-06, "loss": 0.0001, "step": 8290 }, { "epoch": 6.582308607695359, "grad_norm": 0.002033250061809301, "learning_rate": 9.27835034769089e-06, "loss": 0.0001, "step": 8300 }, { "epoch": 6.590241967473225, "grad_norm": 0.0032980898107513486, "learning_rate": 9.178201506635565e-06, "loss": 0.0001, "step": 8310 }, { "epoch": 6.598175327251091, "grad_norm": 0.004474514381064928, "learning_rate": 9.07854144781698e-06, "loss": 0.0001, "step": 8320 }, { "epoch": 6.606108687028957, "grad_norm": 0.013349018197290438, "learning_rate": 8.979371364520783e-06, "loss": 0.0001, "step": 8330 }, { "epoch": 6.614042046806823, "grad_norm": 0.0025534328027148195, "learning_rate": 8.88069244416586e-06, "loss": 0.0, "step": 8340 }, { "epoch": 6.6219754065846885, "grad_norm": 0.0023116634026737483, "learning_rate": 8.782505868290198e-06, "loss": 0.0001, "step": 8350 }, { "epoch": 6.629908766362554, "grad_norm": 0.004273641661059992, "learning_rate": 8.684812812536558e-06, "loss": 0.0001, "step": 8360 }, { "epoch": 6.637842126140421, "grad_norm": 0.0045929145321804195, "learning_rate": 8.58761444663857e-06, "loss": 0.0001, "step": 8370 }, { "epoch": 6.6457754859182865, "grad_norm": 0.0019315519828308053, "learning_rate": 8.490911934406687e-06, "loss": 0.0, "step": 8380 }, { "epoch": 6.653708845696152, "grad_norm": 0.002705981852950197, "learning_rate": 8.394706433714189e-06, "loss": 0.0, "step": 8390 }, { "epoch": 6.661642205474018, "grad_norm": 0.0024053921961185213, "learning_rate": 8.298999096483374e-06, "loss": 0.0, "step": 8400 }, { "epoch": 6.669575565251884, "grad_norm": 0.0044210431297837475, "learning_rate": 8.20379106867174e-06, "loss": 0.0, "step": 8410 }, { "epoch": 6.67750892502975, "grad_norm": 0.007131121700793568, "learning_rate": 8.109083490258274e-06, "loss": 0.0, "step": 8420 }, { "epoch": 6.685442284807616, "grad_norm": 0.007455569169196668, "learning_rate": 8.014877495229844e-06, "loss": 0.0001, "step": 8430 }, { "epoch": 6.6933756445854815, "grad_norm": 0.004185721238089578, "learning_rate": 7.9211742115675e-06, "loss": 0.0001, "step": 8440 }, { "epoch": 6.701309004363348, "grad_norm": 0.0037407304085371185, "learning_rate": 7.82797476123312e-06, "loss": 0.0001, "step": 8450 }, { "epoch": 6.709242364141214, "grad_norm": 0.003440094019653483, "learning_rate": 7.735280260155875e-06, "loss": 0.0001, "step": 8460 }, { "epoch": 6.7171757239190795, "grad_norm": 0.0019999798409724855, "learning_rate": 7.643091818218896e-06, "loss": 0.0001, "step": 8470 }, { "epoch": 6.725109083696946, "grad_norm": 0.005179335784672986, "learning_rate": 7.551410539245973e-06, "loss": 0.0001, "step": 8480 }, { "epoch": 6.733042443474812, "grad_norm": 0.009777941349573382, "learning_rate": 7.460237520988367e-06, "loss": 0.0001, "step": 8490 }, { "epoch": 6.740975803252677, "grad_norm": 0.0013914184952641948, "learning_rate": 7.369573855111661e-06, "loss": 0.0, "step": 8500 }, { "epoch": 6.748909163030543, "grad_norm": 0.0024862936065105294, "learning_rate": 7.279420627182659e-06, "loss": 0.0, "step": 8510 }, { "epoch": 6.75684252280841, "grad_norm": 0.0025615020669894806, "learning_rate": 7.1897789166563755e-06, "loss": 0.0001, "step": 8520 }, { "epoch": 6.764775882586275, "grad_norm": 0.0032990615432745038, "learning_rate": 7.100649796863212e-06, "loss": 0.0, "step": 8530 }, { "epoch": 6.772709242364141, "grad_norm": 0.008889185369011465, "learning_rate": 7.012034334995976e-06, "loss": 0.0001, "step": 8540 }, { "epoch": 6.780642602142007, "grad_norm": 0.004538058763919351, "learning_rate": 6.9239335920971816e-06, "loss": 0.0, "step": 8550 }, { "epoch": 6.788575961919873, "grad_norm": 0.003570696189995056, "learning_rate": 6.836348623046329e-06, "loss": 0.0001, "step": 8560 }, { "epoch": 6.796509321697739, "grad_norm": 0.003657558277366337, "learning_rate": 6.749280476547248e-06, "loss": 0.0001, "step": 8570 }, { "epoch": 6.804442681475605, "grad_norm": 0.002177325137931005, "learning_rate": 6.6627301951156125e-06, "loss": 0.0, "step": 8580 }, { "epoch": 6.81237604125347, "grad_norm": 0.0028257680551589427, "learning_rate": 6.576698815066334e-06, "loss": 0.0, "step": 8590 }, { "epoch": 6.820309401031337, "grad_norm": 0.0013095105321840328, "learning_rate": 6.4911873665013e-06, "loss": 0.0001, "step": 8600 }, { "epoch": 6.828242760809203, "grad_norm": 0.0013752126406687754, "learning_rate": 6.406196873296921e-06, "loss": 0.0, "step": 8610 }, { "epoch": 6.836176120587068, "grad_norm": 0.0030297043679555144, "learning_rate": 6.321728353091938e-06, "loss": 0.0001, "step": 8620 }, { "epoch": 6.844109480364935, "grad_norm": 0.001269434570177823, "learning_rate": 6.237782817275201e-06, "loss": 0.0001, "step": 8630 }, { "epoch": 6.852042840142801, "grad_norm": 0.0020931791112185905, "learning_rate": 6.154361270973585e-06, "loss": 0.0, "step": 8640 }, { "epoch": 6.859976199920666, "grad_norm": 0.008900765050326668, "learning_rate": 6.071464713039959e-06, "loss": 0.0, "step": 8650 }, { "epoch": 6.867909559698532, "grad_norm": 0.0069214232332908045, "learning_rate": 5.989094136041201e-06, "loss": 0.0001, "step": 8660 }, { "epoch": 6.875842919476399, "grad_norm": 0.0031155457645847887, "learning_rate": 5.907250526246294e-06, "loss": 0.0, "step": 8670 }, { "epoch": 6.883776279254264, "grad_norm": 0.0014196379493027088, "learning_rate": 5.8259348636146125e-06, "loss": 0.0001, "step": 8680 }, { "epoch": 6.89170963903213, "grad_norm": 0.008578660727259371, "learning_rate": 5.745148121784066e-06, "loss": 0.0, "step": 8690 }, { "epoch": 6.899642998809996, "grad_norm": 0.0019484792114888411, "learning_rate": 5.66489126805953e-06, "loss": 0.0, "step": 8700 }, { "epoch": 6.907576358587862, "grad_norm": 0.0022578897659263405, "learning_rate": 5.585165263401221e-06, "loss": 0.0, "step": 8710 }, { "epoch": 6.915509718365728, "grad_norm": 0.0021183689091518114, "learning_rate": 5.505971062413201e-06, "loss": 0.0, "step": 8720 }, { "epoch": 6.923443078143594, "grad_norm": 0.002467972104563849, "learning_rate": 5.4273096133319824e-06, "loss": 0.0, "step": 8730 }, { "epoch": 6.931376437921459, "grad_norm": 0.008255847673330399, "learning_rate": 5.3491818580150675e-06, "loss": 0.0, "step": 8740 }, { "epoch": 6.939309797699326, "grad_norm": 0.002659483732448167, "learning_rate": 5.271588731929833e-06, "loss": 0.0, "step": 8750 }, { "epoch": 6.947243157477192, "grad_norm": 0.003447826409870969, "learning_rate": 5.1945311641421744e-06, "loss": 0.0, "step": 8760 }, { "epoch": 6.955176517255057, "grad_norm": 0.0087848804812965, "learning_rate": 5.118010077305474e-06, "loss": 0.0001, "step": 8770 }, { "epoch": 6.963109877032924, "grad_norm": 0.0018339128265004274, "learning_rate": 5.042026387649535e-06, "loss": 0.0, "step": 8780 }, { "epoch": 6.9710432368107895, "grad_norm": 0.0037509703948807624, "learning_rate": 4.966581004969567e-06, "loss": 0.0001, "step": 8790 }, { "epoch": 6.978976596588655, "grad_norm": 0.00926945018014254, "learning_rate": 4.891674832615389e-06, "loss": 0.0, "step": 8800 }, { "epoch": 6.986909956366521, "grad_norm": 0.01816254726333236, "learning_rate": 4.8173087674805165e-06, "loss": 0.0, "step": 8810 }, { "epoch": 6.9948433161443875, "grad_norm": 0.001631686151376323, "learning_rate": 4.743483699991425e-06, "loss": 0.0, "step": 8820 }, { "epoch": 7.002380007933359, "grad_norm": 0.00112925170126183, "learning_rate": 4.670200514096995e-06, "loss": 0.0, "step": 8830 }, { "epoch": 7.010313367711226, "grad_norm": 0.0017963464310088158, "learning_rate": 4.597460087257805e-06, "loss": 0.0, "step": 8840 }, { "epoch": 7.018246727489092, "grad_norm": 0.002542430703599005, "learning_rate": 4.5252632904356936e-06, "loss": 0.0, "step": 8850 }, { "epoch": 7.026180087266957, "grad_norm": 0.011786494282585387, "learning_rate": 4.453610988083312e-06, "loss": 0.0, "step": 8860 }, { "epoch": 7.034113447044824, "grad_norm": 0.001293139663894738, "learning_rate": 4.382504038133755e-06, "loss": 0.0, "step": 8870 }, { "epoch": 7.04204680682269, "grad_norm": 0.000945962995392387, "learning_rate": 4.311943291990362e-06, "loss": 0.0, "step": 8880 }, { "epoch": 7.049980166600555, "grad_norm": 0.0009142844034458246, "learning_rate": 4.241929594516386e-06, "loss": 0.0, "step": 8890 }, { "epoch": 7.057913526378421, "grad_norm": 0.0009695239575419341, "learning_rate": 4.172463784025044e-06, "loss": 0.0, "step": 8900 }, { "epoch": 7.0658468861562875, "grad_norm": 0.0008416673603162244, "learning_rate": 4.103546692269333e-06, "loss": 0.0, "step": 8910 }, { "epoch": 7.073780245934153, "grad_norm": 0.0013893935779380015, "learning_rate": 4.035179144432172e-06, "loss": 0.0, "step": 8920 }, { "epoch": 7.081713605712019, "grad_norm": 0.0028070819339044476, "learning_rate": 3.967361959116461e-06, "loss": 0.0, "step": 8930 }, { "epoch": 7.089646965489885, "grad_norm": 0.00080669700090187, "learning_rate": 3.9000959483353025e-06, "loss": 0.0, "step": 8940 }, { "epoch": 7.097580325267751, "grad_norm": 0.0009230021276189036, "learning_rate": 3.833381917502299e-06, "loss": 0.0, "step": 8950 }, { "epoch": 7.105513685045617, "grad_norm": 0.0011525098157638392, "learning_rate": 3.7672206654218834e-06, "loss": 0.0, "step": 8960 }, { "epoch": 7.113447044823483, "grad_norm": 0.0018122209259069186, "learning_rate": 3.7016129842797122e-06, "loss": 0.0, "step": 8970 }, { "epoch": 7.121380404601348, "grad_norm": 0.0014412370096594106, "learning_rate": 3.636559659633304e-06, "loss": 0.0, "step": 8980 }, { "epoch": 7.129313764379215, "grad_norm": 0.0009042269572004987, "learning_rate": 3.5720614704024934e-06, "loss": 0.0, "step": 8990 }, { "epoch": 7.1372471241570805, "grad_norm": 0.0010217282566214166, "learning_rate": 3.5081191888602095e-06, "loss": 0.0, "step": 9000 }, { "epoch": 7.145180483934946, "grad_norm": 0.0008787151762567979, "learning_rate": 3.444733580623155e-06, "loss": 0.0, "step": 9010 }, { "epoch": 7.153113843712813, "grad_norm": 0.0012276451057078826, "learning_rate": 3.381905404642688e-06, "loss": 0.0, "step": 9020 }, { "epoch": 7.1610472034906785, "grad_norm": 0.001155820801865784, "learning_rate": 3.3196354131957374e-06, "loss": 0.0, "step": 9030 }, { "epoch": 7.168980563268544, "grad_norm": 0.0010146420895872708, "learning_rate": 3.2579243518757297e-06, "loss": 0.0, "step": 9040 }, { "epoch": 7.17691392304641, "grad_norm": 0.001457070930367323, "learning_rate": 3.1967729595837547e-06, "loss": 0.0, "step": 9050 }, { "epoch": 7.184847282824276, "grad_norm": 0.0010529583460812777, "learning_rate": 3.136181968519647e-06, "loss": 0.0, "step": 9060 }, { "epoch": 7.192780642602142, "grad_norm": 0.000797723835794973, "learning_rate": 3.076152104173241e-06, "loss": 0.0, "step": 9070 }, { "epoch": 7.200714002380008, "grad_norm": 0.0008241674828238871, "learning_rate": 3.016684085315702e-06, "loss": 0.0, "step": 9080 }, { "epoch": 7.2086473621578735, "grad_norm": 0.001385587657312191, "learning_rate": 2.9577786239908866e-06, "loss": 0.0, "step": 9090 }, { "epoch": 7.21658072193574, "grad_norm": 0.0009342225830852496, "learning_rate": 2.899436425506863e-06, "loss": 0.0, "step": 9100 }, { "epoch": 7.224514081713606, "grad_norm": 0.0008715338545560816, "learning_rate": 2.8416581884274206e-06, "loss": 0.0, "step": 9110 }, { "epoch": 7.2324474414914715, "grad_norm": 0.0011034795571218663, "learning_rate": 2.7844446045637017e-06, "loss": 0.0, "step": 9120 }, { "epoch": 7.240380801269337, "grad_norm": 0.0014800698144384976, "learning_rate": 2.727796358965984e-06, "loss": 0.0, "step": 9130 }, { "epoch": 7.248314161047204, "grad_norm": 0.001431277616914537, "learning_rate": 2.6717141299154034e-06, "loss": 0.0, "step": 9140 }, { "epoch": 7.256247520825069, "grad_norm": 0.001014598122662335, "learning_rate": 2.616198588915869e-06, "loss": 0.0, "step": 9150 }, { "epoch": 7.264180880602935, "grad_norm": 0.0009021204295892132, "learning_rate": 2.5612504006860272e-06, "loss": 0.0, "step": 9160 }, { "epoch": 7.272114240380802, "grad_norm": 0.0012099250486595546, "learning_rate": 2.506870223151264e-06, "loss": 0.0, "step": 9170 }, { "epoch": 7.280047600158667, "grad_norm": 0.0012695764507401467, "learning_rate": 2.4530587074359015e-06, "loss": 0.0, "step": 9180 }, { "epoch": 7.287980959936533, "grad_norm": 0.0009330081322142683, "learning_rate": 2.3998164978553228e-06, "loss": 0.0, "step": 9190 }, { "epoch": 7.295914319714399, "grad_norm": 0.0018665814548439327, "learning_rate": 2.3471442319082827e-06, "loss": 0.0, "step": 9200 }, { "epoch": 7.303847679492265, "grad_norm": 0.0006962323020714295, "learning_rate": 2.2950425402693156e-06, "loss": 0.0, "step": 9210 }, { "epoch": 7.311781039270131, "grad_norm": 0.000900428596898322, "learning_rate": 2.2435120467811156e-06, "loss": 0.0, "step": 9220 }, { "epoch": 7.319714399047997, "grad_norm": 0.0007492661658477411, "learning_rate": 2.19255336844712e-06, "loss": 0.0, "step": 9230 }, { "epoch": 7.327647758825862, "grad_norm": 0.0011648793001289015, "learning_rate": 2.1421671154240986e-06, "loss": 0.0, "step": 9240 }, { "epoch": 7.335581118603729, "grad_norm": 0.0008750847637548515, "learning_rate": 2.09235389101487e-06, "loss": 0.0, "step": 9250 }, { "epoch": 7.343514478381595, "grad_norm": 0.0009029795833748402, "learning_rate": 2.0431142916610358e-06, "loss": 0.0, "step": 9260 }, { "epoch": 7.35144783815946, "grad_norm": 0.0009786343781213582, "learning_rate": 1.9944489069358584e-06, "loss": 0.0, "step": 9270 }, { "epoch": 7.359381197937326, "grad_norm": 0.0007604871893811181, "learning_rate": 1.9463583195372393e-06, "loss": 0.0, "step": 9280 }, { "epoch": 7.367314557715193, "grad_norm": 0.0015553199120644122, "learning_rate": 1.89884310528069e-06, "loss": 0.0, "step": 9290 }, { "epoch": 7.375247917493058, "grad_norm": 0.0010210805007361942, "learning_rate": 1.8519038330924554e-06, "loss": 0.0, "step": 9300 }, { "epoch": 7.383181277270924, "grad_norm": 0.0010629138512852903, "learning_rate": 1.8055410650027083e-06, "loss": 0.0, "step": 9310 }, { "epoch": 7.391114637048791, "grad_norm": 0.0009634063161324694, "learning_rate": 1.7597553561388091e-06, "loss": 0.0, "step": 9320 }, { "epoch": 7.399047996826656, "grad_norm": 0.001085598797070503, "learning_rate": 1.7145472547186903e-06, "loss": 0.0, "step": 9330 }, { "epoch": 7.406981356604522, "grad_norm": 0.0011344005715411016, "learning_rate": 1.6699173020442272e-06, "loss": 0.0, "step": 9340 }, { "epoch": 7.414914716382388, "grad_norm": 0.0009306965265734756, "learning_rate": 1.6258660324948326e-06, "loss": 0.0, "step": 9350 }, { "epoch": 7.422848076160254, "grad_norm": 0.00103931747933224, "learning_rate": 1.5823939735210014e-06, "loss": 0.0, "step": 9360 }, { "epoch": 7.43078143593812, "grad_norm": 0.0008196453807465721, "learning_rate": 1.5395016456380195e-06, "loss": 0.0, "step": 9370 }, { "epoch": 7.438714795715986, "grad_norm": 0.0012094086453035785, "learning_rate": 1.4971895624197318e-06, "loss": 0.0, "step": 9380 }, { "epoch": 7.446648155493851, "grad_norm": 0.0010737061404356033, "learning_rate": 1.4554582304923902e-06, "loss": 0.0, "step": 9390 }, { "epoch": 7.454581515271718, "grad_norm": 0.0009482462125706409, "learning_rate": 1.4143081495285816e-06, "loss": 0.0, "step": 9400 }, { "epoch": 7.4625148750495836, "grad_norm": 0.0008898491731135208, "learning_rate": 1.3737398122412592e-06, "loss": 0.0, "step": 9410 }, { "epoch": 7.470448234827449, "grad_norm": 0.0009432951387028076, "learning_rate": 1.333753704377816e-06, "loss": 0.0, "step": 9420 }, { "epoch": 7.478381594605315, "grad_norm": 0.0008457643258641981, "learning_rate": 1.2943503047143035e-06, "loss": 0.0, "step": 9430 }, { "epoch": 7.4863149543831815, "grad_norm": 0.002191690933238259, "learning_rate": 1.2555300850496777e-06, "loss": 0.0, "step": 9440 }, { "epoch": 7.494248314161047, "grad_norm": 0.0010511987572010295, "learning_rate": 1.2172935102001515e-06, "loss": 0.0, "step": 9450 }, { "epoch": 7.502181673938913, "grad_norm": 0.0010953609518889065, "learning_rate": 1.1796410379936285e-06, "loss": 0.0, "step": 9460 }, { "epoch": 7.5101150337167795, "grad_norm": 0.0013770546303157102, "learning_rate": 1.1425731192642231e-06, "loss": 0.0, "step": 9470 }, { "epoch": 7.518048393494645, "grad_norm": 0.000843062391334003, "learning_rate": 1.1060901978468818e-06, "loss": 0.0, "step": 9480 }, { "epoch": 7.525981753272511, "grad_norm": 0.0010465575099789092, "learning_rate": 1.0701927105720377e-06, "loss": 0.0, "step": 9490 }, { "epoch": 7.5339151130503765, "grad_norm": 0.000938772661979012, "learning_rate": 1.0348810872603753e-06, "loss": 0.0, "step": 9500 }, { "epoch": 7.541848472828243, "grad_norm": 0.000821438323544369, "learning_rate": 1.0001557507177405e-06, "loss": 0.0, "step": 9510 }, { "epoch": 7.549781832606109, "grad_norm": 0.0008108851578648631, "learning_rate": 9.660171167300113e-07, "loss": 0.0, "step": 9520 }, { "epoch": 7.5577151923839745, "grad_norm": 0.0051432877907656915, "learning_rate": 9.324655940581573e-07, "loss": 0.0, "step": 9530 }, { "epoch": 7.56564855216184, "grad_norm": 0.0009052482637096377, "learning_rate": 8.995015844333377e-07, "loss": 0.0, "step": 9540 }, { "epoch": 7.573581911939707, "grad_norm": 0.0013490729692116723, "learning_rate": 8.671254825520725e-07, "loss": 0.0, "step": 9550 }, { "epoch": 7.5815152717175724, "grad_norm": 0.0015353729831592007, "learning_rate": 8.353376760715681e-07, "loss": 0.0, "step": 9560 }, { "epoch": 7.589448631495438, "grad_norm": 0.0009038699288686859, "learning_rate": 8.041385456049988e-07, "loss": 0.0, "step": 9570 }, { "epoch": 7.597381991273304, "grad_norm": 0.0013444217107359327, "learning_rate": 7.735284647170272e-07, "loss": 0.0, "step": 9580 }, { "epoch": 7.60531535105117, "grad_norm": 0.0011263142224788069, "learning_rate": 7.435077999192741e-07, "loss": 0.0, "step": 9590 }, { "epoch": 7.613248710829036, "grad_norm": 0.0011640315711710341, "learning_rate": 7.140769106659617e-07, "loss": 0.0, "step": 9600 }, { "epoch": 7.621182070606902, "grad_norm": 0.0009110139587373453, "learning_rate": 6.85236149349594e-07, "loss": 0.0, "step": 9610 }, { "epoch": 7.629115430384768, "grad_norm": 0.0013811748561037669, "learning_rate": 6.569858612967384e-07, "loss": 0.0, "step": 9620 }, { "epoch": 7.637048790162634, "grad_norm": 0.0010860793917171084, "learning_rate": 6.293263847639175e-07, "loss": 0.0, "step": 9630 }, { "epoch": 7.6449821499405, "grad_norm": 0.0010907587473306963, "learning_rate": 6.022580509335129e-07, "loss": 0.0, "step": 9640 }, { "epoch": 7.652915509718365, "grad_norm": 0.0012472234199027285, "learning_rate": 5.757811839098293e-07, "loss": 0.0, "step": 9650 }, { "epoch": 7.660848869496232, "grad_norm": 0.0011338365844874106, "learning_rate": 5.498961007152137e-07, "loss": 0.0, "step": 9660 }, { "epoch": 7.668782229274098, "grad_norm": 0.0007636416132129641, "learning_rate": 5.246031112862593e-07, "loss": 0.0, "step": 9670 }, { "epoch": 7.676715589051963, "grad_norm": 0.0009830243268813448, "learning_rate": 4.999025184700745e-07, "loss": 0.0, "step": 9680 }, { "epoch": 7.684648948829829, "grad_norm": 0.0009931939313208113, "learning_rate": 4.757946180206807e-07, "loss": 0.0, "step": 9690 }, { "epoch": 7.692582308607696, "grad_norm": 0.0011286596129988268, "learning_rate": 4.5227969859545917e-07, "loss": 0.0, "step": 9700 }, { "epoch": 7.700515668385561, "grad_norm": 0.0008222334275844989, "learning_rate": 4.29358041751704e-07, "loss": 0.0, "step": 9710 }, { "epoch": 7.708449028163427, "grad_norm": 0.0008045269845985444, "learning_rate": 4.0702992194324164e-07, "loss": 0.0, "step": 9720 }, { "epoch": 7.716382387941293, "grad_norm": 0.0009940292615833253, "learning_rate": 3.8529560651713313e-07, "loss": 0.0, "step": 9730 }, { "epoch": 7.724315747719159, "grad_norm": 0.0007268274444565032, "learning_rate": 3.641553557105049e-07, "loss": 0.0, "step": 9740 }, { "epoch": 7.732249107497025, "grad_norm": 0.0011614380847196136, "learning_rate": 3.4360942264741736e-07, "loss": 0.0, "step": 9750 }, { "epoch": 7.740182467274891, "grad_norm": 0.0010157424677401034, "learning_rate": 3.2365805333581246e-07, "loss": 0.0, "step": 9760 }, { "epoch": 7.748115827052757, "grad_norm": 0.0010790111062041613, "learning_rate": 3.043014866646043e-07, "loss": 0.0, "step": 9770 }, { "epoch": 7.756049186830623, "grad_norm": 0.0010183144853066492, "learning_rate": 2.855399544007986e-07, "loss": 0.0, "step": 9780 }, { "epoch": 7.763982546608489, "grad_norm": 0.0009632923377566276, "learning_rate": 2.673736811867222e-07, "loss": 0.0, "step": 9790 }, { "epoch": 7.771915906386354, "grad_norm": 0.0010471167839244788, "learning_rate": 2.498028845373257e-07, "loss": 0.0, "step": 9800 }, { "epoch": 7.779849266164221, "grad_norm": 0.001067974246811032, "learning_rate": 2.3282777483758511e-07, "loss": 0.0, "step": 9810 }, { "epoch": 7.787782625942087, "grad_norm": 0.000730452731778674, "learning_rate": 2.1644855533999864e-07, "loss": 0.0, "step": 9820 }, { "epoch": 7.795715985719952, "grad_norm": 0.0014108752184380858, "learning_rate": 2.0066542216212181e-07, "loss": 0.0, "step": 9830 }, { "epoch": 7.803649345497818, "grad_norm": 0.0010708351031329653, "learning_rate": 1.854785642842416e-07, "loss": 0.0, "step": 9840 }, { "epoch": 7.8115827052756845, "grad_norm": 0.0008621277760036482, "learning_rate": 1.7088816354710046e-07, "loss": 0.0, "step": 9850 }, { "epoch": 7.81951606505355, "grad_norm": 0.001000136887339331, "learning_rate": 1.5689439464973698e-07, "loss": 0.0, "step": 9860 }, { "epoch": 7.827449424831416, "grad_norm": 0.0006705146208669689, "learning_rate": 1.434974251473764e-07, "loss": 0.0, "step": 9870 }, { "epoch": 7.835382784609282, "grad_norm": 0.0010349790503916465, "learning_rate": 1.3069741544943226e-07, "loss": 0.0, "step": 9880 }, { "epoch": 7.843316144387148, "grad_norm": 0.0013587198986446398, "learning_rate": 1.1849451881758012e-07, "loss": 0.0, "step": 9890 }, { "epoch": 7.851249504165014, "grad_norm": 0.00098495123836121, "learning_rate": 1.0688888136392572e-07, "loss": 0.0, "step": 9900 }, { "epoch": 7.85918286394288, "grad_norm": 0.0011365769614274793, "learning_rate": 9.588064204927305e-08, "loss": 0.0, "step": 9910 }, { "epoch": 7.867116223720746, "grad_norm": 0.0013003767909848553, "learning_rate": 8.546993268141457e-08, "loss": 0.0, "step": 9920 }, { "epoch": 7.875049583498612, "grad_norm": 0.0007765853244875164, "learning_rate": 7.565687791361575e-08, "loss": 0.0, "step": 9930 }, { "epoch": 7.8829829432764775, "grad_norm": 0.0010487334235602414, "learning_rate": 6.644159524306637e-08, "loss": 0.0, "step": 9940 }, { "epoch": 7.890916303054343, "grad_norm": 0.0010407897434088931, "learning_rate": 5.782419500950376e-08, "loss": 0.0, "step": 9950 }, { "epoch": 7.89884966283221, "grad_norm": 0.0007338937205371856, "learning_rate": 4.9804780393875036e-08, "loss": 0.0, "step": 9960 }, { "epoch": 7.9067830226100755, "grad_norm": 0.0008196846409708543, "learning_rate": 4.238344741712696e-08, "loss": 0.0, "step": 9970 }, { "epoch": 7.914716382387941, "grad_norm": 0.0010986838351581703, "learning_rate": 3.556028493903463e-08, "loss": 0.0, "step": 9980 }, { "epoch": 7.922649742165807, "grad_norm": 0.0011616254376772364, "learning_rate": 2.933537465713565e-08, "loss": 0.0, "step": 9990 }, { "epoch": 7.930583101943673, "grad_norm": 0.0010248022607911517, "learning_rate": 2.3708791105764293e-08, "loss": 0.0, "step": 10000 }, { "epoch": 7.938516461721539, "grad_norm": 0.000893312452536662, "learning_rate": 1.8680601655152174e-08, "loss": 0.0, "step": 10010 }, { "epoch": 7.946449821499405, "grad_norm": 0.0011836379486032619, "learning_rate": 1.4250866510628902e-08, "loss": 0.0, "step": 10020 }, { "epoch": 7.9543831812772705, "grad_norm": 0.0008919613735066287, "learning_rate": 1.0419638711894886e-08, "loss": 0.0, "step": 10030 }, { "epoch": 7.962316541055137, "grad_norm": 0.0013697779609313282, "learning_rate": 7.186964132377405e-09, "loss": 0.0, "step": 10040 }, { "epoch": 7.970249900833003, "grad_norm": 0.0016797518546899826, "learning_rate": 4.552881478697702e-09, "loss": 0.0, "step": 10050 }, { "epoch": 7.9781832606108685, "grad_norm": 0.0015311575958115535, "learning_rate": 2.517422290210236e-09, "loss": 0.0, "step": 10060 }, { "epoch": 7.986116620388735, "grad_norm": 0.0008314207396903481, "learning_rate": 1.0806109386030105e-09, "loss": 0.0, "step": 10070 }, { "epoch": 7.994049980166601, "grad_norm": 0.0007295176287307578, "learning_rate": 2.424646276200093e-10, "loss": 0.0, "step": 10080 } ], "logging_steps": 10, "max_steps": 10088, "num_input_tokens_seen": 0, "num_train_epochs": 8, "save_steps": 1800, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 726150384451584.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }