{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 4.282427608092027, "eval_steps": 500, "global_step": 5400, "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 } ], "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": false }, "attributes": {} } }, "total_flos": 388695884562432.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }