{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 2000, "global_step": 6326, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "document_regularizer_loss": 4.9235, "epoch": 0.003161555485298767, "grad_norm": 16384.04296875, "learning_rate": 2.002107481559536e-06, "loss": 848.0818, "query_regularizer_loss": 4.9461, "sparse_loss": 838.2123, "step": 20 }, { "document_regularizer_loss": 4.8004, "epoch": 0.006323110970597534, "grad_norm": 7671.6982421875, "learning_rate": 4.10958904109589e-06, "loss": 536.294, "query_regularizer_loss": 4.8017, "sparse_loss": 526.6921, "step": 40 }, { "document_regularizer_loss": 4.181, "epoch": 0.0094846664558963, "grad_norm": 5428.58251953125, "learning_rate": 6.217070600632244e-06, "loss": 243.6122, "query_regularizer_loss": 4.1751, "sparse_loss": 235.2561, "step": 60 }, { "document_regularizer_loss": 3.657, "epoch": 0.012646221941195067, "grad_norm": 3349.087890625, "learning_rate": 8.3245521601686e-06, "loss": 118.8254, "query_regularizer_loss": 3.6568, "sparse_loss": 111.5117, "step": 80 }, { "document_regularizer_loss": 3.1785, "epoch": 0.015807777426493835, "grad_norm": 1872.81982421875, "learning_rate": 1.0432033719704954e-05, "loss": 61.1316, "query_regularizer_loss": 3.1798, "sparse_loss": 54.7733, "step": 100 }, { "document_regularizer_loss": 2.7993, "epoch": 0.0189693329117926, "grad_norm": 857.3776245117188, "learning_rate": 1.2539515279241306e-05, "loss": 35.0436, "query_regularizer_loss": 2.7993, "sparse_loss": 29.445, "step": 120 }, { "document_regularizer_loss": 2.5469, "epoch": 0.022130888397091368, "grad_norm": 873.3255615234375, "learning_rate": 1.464699683877766e-05, "loss": 22.4751, "query_regularizer_loss": 2.5475, "sparse_loss": 17.3807, "step": 140 }, { "document_regularizer_loss": 2.3631, "epoch": 0.025292443882390134, "grad_norm": 1088.9566650390625, "learning_rate": 1.6754478398314015e-05, "loss": 16.6537, "query_regularizer_loss": 2.363, "sparse_loss": 11.9276, "step": 160 }, { "document_regularizer_loss": 2.177, "epoch": 0.028453999367688904, "grad_norm": 439.410400390625, "learning_rate": 1.886195995785037e-05, "loss": 15.5919, "query_regularizer_loss": 2.1773, "sparse_loss": 11.2376, "step": 180 }, { "document_regularizer_loss": 2.0062, "epoch": 0.03161555485298767, "grad_norm": 682.8355712890625, "learning_rate": 2.0969441517386723e-05, "loss": 10.4659, "query_regularizer_loss": 2.0063, "sparse_loss": 6.4535, "step": 200 }, { "document_regularizer_loss": 1.8374, "epoch": 0.034777110338286434, "grad_norm": 375.6211242675781, "learning_rate": 2.307692307692308e-05, "loss": 9.746, "query_regularizer_loss": 1.8372, "sparse_loss": 6.0714, "step": 220 }, { "document_regularizer_loss": 1.701, "epoch": 0.0379386658235852, "grad_norm": 686.1103515625, "learning_rate": 2.518440463645943e-05, "loss": 9.4233, "query_regularizer_loss": 1.7013, "sparse_loss": 6.0211, "step": 240 }, { "document_regularizer_loss": 1.5848, "epoch": 0.04110022130888397, "grad_norm": 211.80474853515625, "learning_rate": 2.7291886195995788e-05, "loss": 7.5791, "query_regularizer_loss": 1.5849, "sparse_loss": 4.4094, "step": 260 }, { "document_regularizer_loss": 1.3934, "epoch": 0.044261776794182736, "grad_norm": 78.986328125, "learning_rate": 2.9399367755532144e-05, "loss": 6.8552, "query_regularizer_loss": 1.3933, "sparse_loss": 4.0685, "step": 280 }, { "document_regularizer_loss": 1.1511, "epoch": 0.047423332279481506, "grad_norm": 68.5307846069336, "learning_rate": 3.1506849315068496e-05, "loss": 5.694, "query_regularizer_loss": 1.1512, "sparse_loss": 3.3917, "step": 300 }, { "document_regularizer_loss": 0.9269, "epoch": 0.05058488776478027, "grad_norm": 152.0952606201172, "learning_rate": 3.361433087460485e-05, "loss": 5.189, "query_regularizer_loss": 0.9268, "sparse_loss": 3.3352, "step": 320 }, { "document_regularizer_loss": 0.7798, "epoch": 0.05374644325007904, "grad_norm": 31.483890533447266, "learning_rate": 3.572181243414121e-05, "loss": 4.726, "query_regularizer_loss": 0.7797, "sparse_loss": 3.1665, "step": 340 }, { "document_regularizer_loss": 0.5923, "epoch": 0.05690799873537781, "grad_norm": 32.53569030761719, "learning_rate": 3.7829293993677554e-05, "loss": 3.9468, "query_regularizer_loss": 0.5922, "sparse_loss": 2.7623, "step": 360 }, { "document_regularizer_loss": 0.4373, "epoch": 0.06006955422067657, "grad_norm": 23.625446319580078, "learning_rate": 3.993677555321391e-05, "loss": 3.5309, "query_regularizer_loss": 0.4373, "sparse_loss": 2.6563, "step": 380 }, { "document_regularizer_loss": 0.3222, "epoch": 0.06323110970597534, "grad_norm": 37.11733627319336, "learning_rate": 4.2044257112750266e-05, "loss": 3.2516, "query_regularizer_loss": 0.3222, "sparse_loss": 2.6073, "step": 400 }, { "document_regularizer_loss": 0.2448, "epoch": 0.06639266519127411, "grad_norm": 20.662832260131836, "learning_rate": 4.415173867228662e-05, "loss": 3.1281, "query_regularizer_loss": 0.2448, "sparse_loss": 2.6385, "step": 420 }, { "document_regularizer_loss": 0.1903, "epoch": 0.06955422067657287, "grad_norm": 90.5846176147461, "learning_rate": 4.625922023182298e-05, "loss": 3.3396, "query_regularizer_loss": 0.1904, "sparse_loss": 2.9589, "step": 440 }, { "document_regularizer_loss": 0.1682, "epoch": 0.07271577616187164, "grad_norm": 145.83663940429688, "learning_rate": 4.836670179135932e-05, "loss": 2.9721, "query_regularizer_loss": 0.1682, "sparse_loss": 2.6357, "step": 460 }, { "document_regularizer_loss": 0.1365, "epoch": 0.0758773316471704, "grad_norm": 23.765230178833008, "learning_rate": 5.047418335089569e-05, "loss": 2.6706, "query_regularizer_loss": 0.1365, "sparse_loss": 2.3975, "step": 480 }, { "document_regularizer_loss": 0.1119, "epoch": 0.07903888713246918, "grad_norm": 29.82013702392578, "learning_rate": 5.2581664910432035e-05, "loss": 2.6023, "query_regularizer_loss": 0.1118, "sparse_loss": 2.3786, "step": 500 }, { "document_regularizer_loss": 0.0998, "epoch": 0.08220044261776795, "grad_norm": 26.115591049194336, "learning_rate": 5.468914646996839e-05, "loss": 2.5536, "query_regularizer_loss": 0.0998, "sparse_loss": 2.354, "step": 520 }, { "document_regularizer_loss": 0.0849, "epoch": 0.0853619981030667, "grad_norm": 11.93899154663086, "learning_rate": 5.679662802950474e-05, "loss": 2.3589, "query_regularizer_loss": 0.0849, "sparse_loss": 2.1891, "step": 540 }, { "document_regularizer_loss": 0.0726, "epoch": 0.08852355358836547, "grad_norm": 19.871112823486328, "learning_rate": 5.89041095890411e-05, "loss": 2.4359, "query_regularizer_loss": 0.0726, "sparse_loss": 2.2907, "step": 560 }, { "document_regularizer_loss": 0.0691, "epoch": 0.09168510907366424, "grad_norm": 13.610793113708496, "learning_rate": 6.101159114857745e-05, "loss": 2.4409, "query_regularizer_loss": 0.0691, "sparse_loss": 2.3028, "step": 580 }, { "document_regularizer_loss": 0.0645, "epoch": 0.09484666455896301, "grad_norm": 12.517507553100586, "learning_rate": 6.31190727081138e-05, "loss": 2.278, "query_regularizer_loss": 0.0645, "sparse_loss": 2.1489, "step": 600 }, { "document_regularizer_loss": 0.0576, "epoch": 0.09800822004426178, "grad_norm": 10.614307403564453, "learning_rate": 6.522655426765016e-05, "loss": 3.2585, "query_regularizer_loss": 0.0575, "sparse_loss": 3.1434, "step": 620 }, { "document_regularizer_loss": 0.0455, "epoch": 0.10116977552956054, "grad_norm": 12.401030540466309, "learning_rate": 6.733403582718652e-05, "loss": 2.217, "query_regularizer_loss": 0.0455, "sparse_loss": 2.1261, "step": 640 }, { "document_regularizer_loss": 0.0428, "epoch": 0.10433133101485931, "grad_norm": 7.582752227783203, "learning_rate": 6.944151738672286e-05, "loss": 2.1853, "query_regularizer_loss": 0.0428, "sparse_loss": 2.0996, "step": 660 }, { "document_regularizer_loss": 0.0381, "epoch": 0.10749288650015808, "grad_norm": 8.788124084472656, "learning_rate": 7.154899894625922e-05, "loss": 2.111, "query_regularizer_loss": 0.038, "sparse_loss": 2.0349, "step": 680 }, { "document_regularizer_loss": 0.0366, "epoch": 0.11065444198545685, "grad_norm": 8.3252592086792, "learning_rate": 7.365648050579558e-05, "loss": 2.1326, "query_regularizer_loss": 0.0366, "sparse_loss": 2.0595, "step": 700 }, { "document_regularizer_loss": 0.0357, "epoch": 0.11381599747075562, "grad_norm": 6.500911235809326, "learning_rate": 7.576396206533193e-05, "loss": 2.0903, "query_regularizer_loss": 0.0357, "sparse_loss": 2.019, "step": 720 }, { "document_regularizer_loss": 0.0345, "epoch": 0.11697755295605437, "grad_norm": 6.404702663421631, "learning_rate": 7.787144362486829e-05, "loss": 2.066, "query_regularizer_loss": 0.0345, "sparse_loss": 1.9969, "step": 740 }, { "document_regularizer_loss": 0.0343, "epoch": 0.12013910844135314, "grad_norm": 14.282183647155762, "learning_rate": 7.997892518440465e-05, "loss": 1.9859, "query_regularizer_loss": 0.0343, "sparse_loss": 1.9173, "step": 760 }, { "document_regularizer_loss": 0.0308, "epoch": 0.12330066392665191, "grad_norm": 8.478113174438477, "learning_rate": 8.208640674394099e-05, "loss": 1.9894, "query_regularizer_loss": 0.0308, "sparse_loss": 1.9278, "step": 780 }, { "document_regularizer_loss": 0.0284, "epoch": 0.12646221941195068, "grad_norm": 7.985814094543457, "learning_rate": 8.419388830347735e-05, "loss": 1.9844, "query_regularizer_loss": 0.0284, "sparse_loss": 1.9276, "step": 800 }, { "document_regularizer_loss": 0.0306, "epoch": 0.12962377489724944, "grad_norm": 7.465007781982422, "learning_rate": 8.630136986301371e-05, "loss": 1.9759, "query_regularizer_loss": 0.0307, "sparse_loss": 1.9146, "step": 820 }, { "document_regularizer_loss": 0.0309, "epoch": 0.13278533038254822, "grad_norm": 8.6587495803833, "learning_rate": 8.840885142255006e-05, "loss": 1.9294, "query_regularizer_loss": 0.0309, "sparse_loss": 1.8675, "step": 840 }, { "document_regularizer_loss": 0.0288, "epoch": 0.13594688586784698, "grad_norm": 7.235748291015625, "learning_rate": 9.051633298208641e-05, "loss": 2.0085, "query_regularizer_loss": 0.0288, "sparse_loss": 1.951, "step": 860 }, { "document_regularizer_loss": 0.0281, "epoch": 0.13910844135314573, "grad_norm": 6.58332633972168, "learning_rate": 9.262381454162276e-05, "loss": 1.9481, "query_regularizer_loss": 0.0281, "sparse_loss": 1.8919, "step": 880 }, { "document_regularizer_loss": 0.0285, "epoch": 0.14226999683844452, "grad_norm": 6.89755392074585, "learning_rate": 9.473129610115912e-05, "loss": 1.9639, "query_regularizer_loss": 0.0285, "sparse_loss": 1.9069, "step": 900 }, { "document_regularizer_loss": 0.0293, "epoch": 0.14543155232374327, "grad_norm": 6.412639617919922, "learning_rate": 9.683877766069548e-05, "loss": 1.8506, "query_regularizer_loss": 0.0293, "sparse_loss": 1.792, "step": 920 }, { "document_regularizer_loss": 0.0282, "epoch": 0.14859310780904206, "grad_norm": 8.971174240112305, "learning_rate": 9.894625922023183e-05, "loss": 1.8199, "query_regularizer_loss": 0.0283, "sparse_loss": 1.7634, "step": 940 }, { "document_regularizer_loss": 0.0283, "epoch": 0.1517546632943408, "grad_norm": 7.4108967781066895, "learning_rate": 9.999914658858875e-05, "loss": 1.7994, "query_regularizer_loss": 0.0283, "sparse_loss": 1.7428, "step": 960 }, { "document_regularizer_loss": 0.0295, "epoch": 0.15491621877963957, "grad_norm": 4.715672492980957, "learning_rate": 9.999231947209235e-05, "loss": 1.8468, "query_regularizer_loss": 0.0296, "sparse_loss": 1.7877, "step": 980 }, { "document_regularizer_loss": 0.0294, "epoch": 0.15807777426493835, "grad_norm": 6.63499641418457, "learning_rate": 9.997866617130585e-05, "loss": 1.7654, "query_regularizer_loss": 0.0294, "sparse_loss": 1.7067, "step": 1000 }, { "document_regularizer_loss": 0.0273, "epoch": 0.1612393297502371, "grad_norm": 7.484816074371338, "learning_rate": 9.995818855051458e-05, "loss": 1.7669, "query_regularizer_loss": 0.0273, "sparse_loss": 1.7123, "step": 1020 }, { "document_regularizer_loss": 0.0265, "epoch": 0.1644008852355359, "grad_norm": 16.461029052734375, "learning_rate": 9.993088940582831e-05, "loss": 1.7928, "query_regularizer_loss": 0.0265, "sparse_loss": 1.7397, "step": 1040 }, { "document_regularizer_loss": 0.0247, "epoch": 0.16756244072083465, "grad_norm": 8.150482177734375, "learning_rate": 9.989677246479949e-05, "loss": 1.73, "query_regularizer_loss": 0.0246, "sparse_loss": 1.6807, "step": 1060 }, { "document_regularizer_loss": 0.024, "epoch": 0.1707239962061334, "grad_norm": 7.778267860412598, "learning_rate": 9.985584238591425e-05, "loss": 1.6531, "query_regularizer_loss": 0.024, "sparse_loss": 1.605, "step": 1080 }, { "document_regularizer_loss": 0.0231, "epoch": 0.1738855516914322, "grad_norm": 7.546144008636475, "learning_rate": 9.980810475795631e-05, "loss": 1.622, "query_regularizer_loss": 0.0231, "sparse_loss": 1.5758, "step": 1100 }, { "document_regularizer_loss": 0.0217, "epoch": 0.17704710717673094, "grad_norm": 5.443304538726807, "learning_rate": 9.975356609924388e-05, "loss": 1.663, "query_regularizer_loss": 0.0217, "sparse_loss": 1.6196, "step": 1120 }, { "document_regularizer_loss": 0.0235, "epoch": 0.18020866266202973, "grad_norm": 6.743195056915283, "learning_rate": 9.969223385673958e-05, "loss": 1.655, "query_regularizer_loss": 0.0235, "sparse_loss": 1.608, "step": 1140 }, { "document_regularizer_loss": 0.0254, "epoch": 0.18337021814732848, "grad_norm": 7.73419713973999, "learning_rate": 9.962411640503366e-05, "loss": 1.6236, "query_regularizer_loss": 0.0254, "sparse_loss": 1.5728, "step": 1160 }, { "document_regularizer_loss": 0.026, "epoch": 0.18653177363262724, "grad_norm": 9.643766403198242, "learning_rate": 9.954922304520045e-05, "loss": 1.6167, "query_regularizer_loss": 0.026, "sparse_loss": 1.5647, "step": 1180 }, { "document_regularizer_loss": 0.0246, "epoch": 0.18969332911792602, "grad_norm": 7.896740913391113, "learning_rate": 9.94675640035283e-05, "loss": 1.595, "query_regularizer_loss": 0.0247, "sparse_loss": 1.5457, "step": 1200 }, { "document_regularizer_loss": 0.0224, "epoch": 0.19285488460322478, "grad_norm": 7.7615556716918945, "learning_rate": 9.937915043012338e-05, "loss": 1.5427, "query_regularizer_loss": 0.0223, "sparse_loss": 1.498, "step": 1220 }, { "document_regularizer_loss": 0.0216, "epoch": 0.19601644008852356, "grad_norm": 5.695051193237305, "learning_rate": 9.928399439738702e-05, "loss": 1.6241, "query_regularizer_loss": 0.0216, "sparse_loss": 1.581, "step": 1240 }, { "document_regularizer_loss": 0.0218, "epoch": 0.19917799557382232, "grad_norm": 10.770386695861816, "learning_rate": 9.918210889836742e-05, "loss": 1.4656, "query_regularizer_loss": 0.0218, "sparse_loss": 1.422, "step": 1260 }, { "document_regularizer_loss": 0.0211, "epoch": 0.20233955105912108, "grad_norm": 5.805099010467529, "learning_rate": 9.907350784498543e-05, "loss": 1.5652, "query_regularizer_loss": 0.0211, "sparse_loss": 1.5229, "step": 1280 }, { "document_regularizer_loss": 0.0215, "epoch": 0.20550110654441986, "grad_norm": 4.371960163116455, "learning_rate": 9.895820606613505e-05, "loss": 1.543, "query_regularizer_loss": 0.0215, "sparse_loss": 1.5, "step": 1300 }, { "document_regularizer_loss": 0.0221, "epoch": 0.20866266202971862, "grad_norm": 4.516541957855225, "learning_rate": 9.883621930565851e-05, "loss": 1.5816, "query_regularizer_loss": 0.022, "sparse_loss": 1.5375, "step": 1320 }, { "document_regularizer_loss": 0.0224, "epoch": 0.2118242175150174, "grad_norm": 4.348978519439697, "learning_rate": 9.870756422019661e-05, "loss": 1.6763, "query_regularizer_loss": 0.0224, "sparse_loss": 1.6315, "step": 1340 }, { "document_regularizer_loss": 0.022, "epoch": 0.21498577300031615, "grad_norm": 6.8595476150512695, "learning_rate": 9.857225837691432e-05, "loss": 1.5836, "query_regularizer_loss": 0.022, "sparse_loss": 1.5397, "step": 1360 }, { "document_regularizer_loss": 0.0217, "epoch": 0.2181473284856149, "grad_norm": 6.249878883361816, "learning_rate": 9.843032025110206e-05, "loss": 1.5415, "query_regularizer_loss": 0.0218, "sparse_loss": 1.4979, "step": 1380 }, { "document_regularizer_loss": 0.0234, "epoch": 0.2213088839709137, "grad_norm": 5.985715866088867, "learning_rate": 9.828176922365304e-05, "loss": 1.5112, "query_regularizer_loss": 0.0233, "sparse_loss": 1.4645, "step": 1400 }, { "document_regularizer_loss": 0.0217, "epoch": 0.22447043945621245, "grad_norm": 4.295719623565674, "learning_rate": 9.812662557841687e-05, "loss": 1.5745, "query_regularizer_loss": 0.0217, "sparse_loss": 1.5312, "step": 1420 }, { "document_regularizer_loss": 0.0212, "epoch": 0.22763199494151123, "grad_norm": 4.790430545806885, "learning_rate": 9.796491049942993e-05, "loss": 1.4835, "query_regularizer_loss": 0.0212, "sparse_loss": 1.4412, "step": 1440 }, { "document_regularizer_loss": 0.0215, "epoch": 0.23079355042681, "grad_norm": 5.4020490646362305, "learning_rate": 9.779664606802277e-05, "loss": 1.4984, "query_regularizer_loss": 0.0215, "sparse_loss": 1.4553, "step": 1460 }, { "document_regularizer_loss": 0.0223, "epoch": 0.23395510591210875, "grad_norm": 3.856952667236328, "learning_rate": 9.762185525980503e-05, "loss": 1.4683, "query_regularizer_loss": 0.0223, "sparse_loss": 1.4238, "step": 1480 }, { "document_regularizer_loss": 0.022, "epoch": 0.23711666139740753, "grad_norm": 5.578510284423828, "learning_rate": 9.74405619415283e-05, "loss": 1.3864, "query_regularizer_loss": 0.022, "sparse_loss": 1.3423, "step": 1500 }, { "document_regularizer_loss": 0.0213, "epoch": 0.24027821688270629, "grad_norm": 6.185074329376221, "learning_rate": 9.72527908678272e-05, "loss": 1.5377, "query_regularizer_loss": 0.0213, "sparse_loss": 1.4951, "step": 1520 }, { "document_regularizer_loss": 0.0197, "epoch": 0.24343977236800507, "grad_norm": 6.440251350402832, "learning_rate": 9.705856767783919e-05, "loss": 1.5046, "query_regularizer_loss": 0.0197, "sparse_loss": 1.4652, "step": 1540 }, { "document_regularizer_loss": 0.0195, "epoch": 0.24660132785330383, "grad_norm": 4.619477272033691, "learning_rate": 9.685791889170384e-05, "loss": 1.4879, "query_regularizer_loss": 0.0195, "sparse_loss": 1.4489, "step": 1560 }, { "document_regularizer_loss": 0.0196, "epoch": 0.24976288333860258, "grad_norm": 4.923925399780273, "learning_rate": 9.665087190694152e-05, "loss": 1.3949, "query_regularizer_loss": 0.0196, "sparse_loss": 1.3557, "step": 1580 }, { "document_regularizer_loss": 0.0198, "epoch": 0.25292443882390137, "grad_norm": 6.299711227416992, "learning_rate": 9.643745499471248e-05, "loss": 1.3844, "query_regularizer_loss": 0.0198, "sparse_loss": 1.3448, "step": 1600 }, { "document_regularizer_loss": 0.0199, "epoch": 0.25608599430920015, "grad_norm": 4.344811916351318, "learning_rate": 9.621769729595653e-05, "loss": 1.3775, "query_regularizer_loss": 0.0199, "sparse_loss": 1.3376, "step": 1620 }, { "document_regularizer_loss": 0.0198, "epoch": 0.2592475497944989, "grad_norm": 4.729667663574219, "learning_rate": 9.599162881741408e-05, "loss": 1.4741, "query_regularizer_loss": 0.0198, "sparse_loss": 1.4345, "step": 1640 }, { "document_regularizer_loss": 0.0196, "epoch": 0.26240910527979766, "grad_norm": 5.488773822784424, "learning_rate": 9.575928042752873e-05, "loss": 1.5249, "query_regularizer_loss": 0.0196, "sparse_loss": 1.4856, "step": 1660 }, { "document_regularizer_loss": 0.0181, "epoch": 0.26557066076509644, "grad_norm": 4.342692852020264, "learning_rate": 9.552068385223254e-05, "loss": 1.4609, "query_regularizer_loss": 0.0181, "sparse_loss": 1.4246, "step": 1680 }, { "document_regularizer_loss": 0.018, "epoch": 0.2687322162503952, "grad_norm": 5.443596363067627, "learning_rate": 9.527587167061386e-05, "loss": 1.384, "query_regularizer_loss": 0.018, "sparse_loss": 1.348, "step": 1700 }, { "document_regularizer_loss": 0.0187, "epoch": 0.27189377173569396, "grad_norm": 5.619431972503662, "learning_rate": 9.502487731046892e-05, "loss": 1.4199, "query_regularizer_loss": 0.0187, "sparse_loss": 1.3825, "step": 1720 }, { "document_regularizer_loss": 0.0186, "epoch": 0.27505532722099274, "grad_norm": 4.95171594619751, "learning_rate": 9.47677350437374e-05, "loss": 1.2902, "query_regularizer_loss": 0.0186, "sparse_loss": 1.253, "step": 1740 }, { "document_regularizer_loss": 0.0179, "epoch": 0.27821688270629147, "grad_norm": 5.611623287200928, "learning_rate": 9.450447998182283e-05, "loss": 1.3614, "query_regularizer_loss": 0.0179, "sparse_loss": 1.3255, "step": 1760 }, { "document_regularizer_loss": 0.0183, "epoch": 0.28137843819159025, "grad_norm": 4.829434871673584, "learning_rate": 9.423514807079828e-05, "loss": 1.3159, "query_regularizer_loss": 0.0182, "sparse_loss": 1.2794, "step": 1780 }, { "document_regularizer_loss": 0.0182, "epoch": 0.28453999367688904, "grad_norm": 4.662863254547119, "learning_rate": 9.395977608649806e-05, "loss": 1.4043, "query_regularizer_loss": 0.0183, "sparse_loss": 1.3678, "step": 1800 }, { "document_regularizer_loss": 0.0178, "epoch": 0.2877015491621878, "grad_norm": 4.188266277313232, "learning_rate": 9.367840162949632e-05, "loss": 1.3711, "query_regularizer_loss": 0.0178, "sparse_loss": 1.3355, "step": 1820 }, { "document_regularizer_loss": 0.0184, "epoch": 0.29086310464748655, "grad_norm": 5.188456058502197, "learning_rate": 9.339106311997275e-05, "loss": 1.3714, "query_regularizer_loss": 0.0184, "sparse_loss": 1.3347, "step": 1840 }, { "document_regularizer_loss": 0.0176, "epoch": 0.29402466013278533, "grad_norm": 3.629645347595215, "learning_rate": 9.309779979246656e-05, "loss": 1.4167, "query_regularizer_loss": 0.0176, "sparse_loss": 1.3815, "step": 1860 }, { "document_regularizer_loss": 0.0181, "epoch": 0.2971862156180841, "grad_norm": 4.733102321624756, "learning_rate": 9.279865169051927e-05, "loss": 1.3242, "query_regularizer_loss": 0.0181, "sparse_loss": 1.2881, "step": 1880 }, { "document_regularizer_loss": 0.0183, "epoch": 0.30034777110338284, "grad_norm": 4.081264972686768, "learning_rate": 9.249365966120691e-05, "loss": 1.3728, "query_regularizer_loss": 0.0184, "sparse_loss": 1.3361, "step": 1900 }, { "document_regularizer_loss": 0.0182, "epoch": 0.3035093265886816, "grad_norm": 4.531682014465332, "learning_rate": 9.218286534956255e-05, "loss": 1.3546, "query_regularizer_loss": 0.0183, "sparse_loss": 1.3181, "step": 1920 }, { "document_regularizer_loss": 0.0184, "epoch": 0.3066708820739804, "grad_norm": 5.468003273010254, "learning_rate": 9.186631119288997e-05, "loss": 1.3307, "query_regularizer_loss": 0.0184, "sparse_loss": 1.2939, "step": 1940 }, { "document_regularizer_loss": 0.0178, "epoch": 0.30983243755927914, "grad_norm": 4.307997226715088, "learning_rate": 9.154404041496902e-05, "loss": 1.2928, "query_regularizer_loss": 0.0179, "sparse_loss": 1.2571, "step": 1960 }, { "document_regularizer_loss": 0.0175, "epoch": 0.3129939930445779, "grad_norm": 4.240797519683838, "learning_rate": 9.121609702015367e-05, "loss": 1.3671, "query_regularizer_loss": 0.0175, "sparse_loss": 1.3322, "step": 1980 }, { "document_regularizer_loss": 0.017, "epoch": 0.3161555485298767, "grad_norm": 4.310458183288574, "learning_rate": 9.088252578736345e-05, "loss": 1.2323, "query_regularizer_loss": 0.017, "sparse_loss": 1.1983, "step": 2000 }, { "epoch": 0.3161555485298767, "eval_runtime": 124.0033, "eval_samples_per_second": 0.0, "eval_sparse-ir-eval_avg_flops": 3439.800048828125, "eval_sparse-ir-eval_corpus_active_dims": 4096.0, "eval_sparse-ir-eval_corpus_sparsity_ratio": 0.9186785260482846, "eval_sparse-ir-eval_dot_accuracy@1": 0.020195960807838434, "eval_sparse-ir-eval_dot_accuracy@100": 0.2731453709258148, "eval_sparse-ir-eval_dot_accuracy@50": 0.1893621275744851, "eval_sparse-ir-eval_dot_accuracy@8": 0.07218556288742252, "eval_sparse-ir-eval_dot_map@100": 0.04183592564683774, "eval_sparse-ir-eval_dot_mrr@10": 0.03596296613693132, "eval_sparse-ir-eval_dot_ndcg@10": 0.04659572910686561, "eval_sparse-ir-eval_dot_precision@1": 0.020195960807838434, "eval_sparse-ir-eval_dot_precision@100": 0.0027314537092581483, "eval_sparse-ir-eval_dot_precision@50": 0.0037872425514897017, "eval_sparse-ir-eval_dot_precision@8": 0.009023195360927815, "eval_sparse-ir-eval_dot_recall@1": 0.020195960807838434, "eval_sparse-ir-eval_dot_recall@100": 0.2731453709258148, "eval_sparse-ir-eval_dot_recall@50": 0.1893621275744851, "eval_sparse-ir-eval_dot_recall@8": 0.07218556288742252, "eval_sparse-ir-eval_query_active_dims": 4096.0, "eval_sparse-ir-eval_query_sparsity_ratio": 0.9186785260482846, "eval_steps_per_second": 0.0, "step": 2000 }, { "document_regularizer_loss": 0.0168, "epoch": 0.3193171040151755, "grad_norm": 4.747305393218994, "learning_rate": 9.054337226396904e-05, "loss": 1.3482, "query_regularizer_loss": 0.0168, "sparse_loss": 1.3146, "step": 2020 }, { "document_regularizer_loss": 0.0166, "epoch": 0.3224786595004742, "grad_norm": 5.303295612335205, "learning_rate": 9.019868275957321e-05, "loss": 1.3388, "query_regularizer_loss": 0.0166, "sparse_loss": 1.3056, "step": 2040 }, { "document_regularizer_loss": 0.0167, "epoch": 0.325640214985773, "grad_norm": 4.634930610656738, "learning_rate": 8.984850433968727e-05, "loss": 1.3731, "query_regularizer_loss": 0.0167, "sparse_loss": 1.3397, "step": 2060 }, { "document_regularizer_loss": 0.0164, "epoch": 0.3288017704710718, "grad_norm": 4.3774542808532715, "learning_rate": 8.94928848193047e-05, "loss": 1.2506, "query_regularizer_loss": 0.0165, "sparse_loss": 1.2177, "step": 2080 }, { "document_regularizer_loss": 0.0163, "epoch": 0.3319633259563705, "grad_norm": 3.8805744647979736, "learning_rate": 8.913187275637217e-05, "loss": 1.3351, "query_regularizer_loss": 0.0163, "sparse_loss": 1.3025, "step": 2100 }, { "document_regularizer_loss": 0.0155, "epoch": 0.3351248814416693, "grad_norm": 4.498915672302246, "learning_rate": 8.876551744515928e-05, "loss": 1.3408, "query_regularizer_loss": 0.0156, "sparse_loss": 1.3097, "step": 2120 }, { "document_regularizer_loss": 0.0153, "epoch": 0.3382864369269681, "grad_norm": 4.381481647491455, "learning_rate": 8.839386890952759e-05, "loss": 1.2679, "query_regularizer_loss": 0.0153, "sparse_loss": 1.2372, "step": 2140 }, { "document_regularizer_loss": 0.0159, "epoch": 0.3414479924122668, "grad_norm": 4.201473712921143, "learning_rate": 8.801697789610024e-05, "loss": 1.289, "query_regularizer_loss": 0.0158, "sparse_loss": 1.2573, "step": 2160 }, { "document_regularizer_loss": 0.0151, "epoch": 0.3446095478975656, "grad_norm": 4.210931777954102, "learning_rate": 8.763489586733269e-05, "loss": 1.1917, "query_regularizer_loss": 0.0151, "sparse_loss": 1.1614, "step": 2180 }, { "document_regularizer_loss": 0.0151, "epoch": 0.3477711033828644, "grad_norm": 5.02850341796875, "learning_rate": 8.724767499448586e-05, "loss": 1.2471, "query_regularizer_loss": 0.0151, "sparse_loss": 1.2168, "step": 2200 }, { "document_regularizer_loss": 0.015, "epoch": 0.35093265886816316, "grad_norm": 4.9309282302856445, "learning_rate": 8.685536815050237e-05, "loss": 1.2609, "query_regularizer_loss": 0.0151, "sparse_loss": 1.2308, "step": 2220 }, { "document_regularizer_loss": 0.0147, "epoch": 0.3540942143534619, "grad_norm": 3.9110965728759766, "learning_rate": 8.645802890278703e-05, "loss": 1.2271, "query_regularizer_loss": 0.0147, "sparse_loss": 1.1976, "step": 2240 }, { "document_regularizer_loss": 0.0157, "epoch": 0.3572557698387607, "grad_norm": 4.420169353485107, "learning_rate": 8.605571150589255e-05, "loss": 1.2712, "query_regularizer_loss": 0.0157, "sparse_loss": 1.2399, "step": 2260 }, { "document_regularizer_loss": 0.0154, "epoch": 0.36041732532405946, "grad_norm": 3.7631611824035645, "learning_rate": 8.564847089411128e-05, "loss": 1.2016, "query_regularizer_loss": 0.0155, "sparse_loss": 1.1707, "step": 2280 }, { "document_regularizer_loss": 0.0149, "epoch": 0.3635788808093582, "grad_norm": 5.38586950302124, "learning_rate": 8.52363626739743e-05, "loss": 1.1784, "query_regularizer_loss": 0.0149, "sparse_loss": 1.1486, "step": 2300 }, { "document_regularizer_loss": 0.0151, "epoch": 0.36674043629465697, "grad_norm": 4.781260967254639, "learning_rate": 8.481944311665863e-05, "loss": 1.2039, "query_regularizer_loss": 0.0151, "sparse_loss": 1.1736, "step": 2320 }, { "document_regularizer_loss": 0.0149, "epoch": 0.36990199177995575, "grad_norm": 4.140412330627441, "learning_rate": 8.439776915030368e-05, "loss": 1.2637, "query_regularizer_loss": 0.0149, "sparse_loss": 1.2339, "step": 2340 }, { "document_regularizer_loss": 0.0153, "epoch": 0.3730635472652545, "grad_norm": 4.991822719573975, "learning_rate": 8.397139835223802e-05, "loss": 1.2108, "query_regularizer_loss": 0.0152, "sparse_loss": 1.1803, "step": 2360 }, { "document_regularizer_loss": 0.0156, "epoch": 0.37622510275055326, "grad_norm": 5.662979602813721, "learning_rate": 8.354038894111754e-05, "loss": 1.2256, "query_regularizer_loss": 0.0156, "sparse_loss": 1.1943, "step": 2380 }, { "document_regularizer_loss": 0.0148, "epoch": 0.37938665823585205, "grad_norm": 5.262570858001709, "learning_rate": 8.310479976897594e-05, "loss": 1.228, "query_regularizer_loss": 0.0148, "sparse_loss": 1.1984, "step": 2400 }, { "document_regularizer_loss": 0.0145, "epoch": 0.38254821372115083, "grad_norm": 4.193863868713379, "learning_rate": 8.266469031318885e-05, "loss": 1.2128, "query_regularizer_loss": 0.0145, "sparse_loss": 1.1838, "step": 2420 }, { "document_regularizer_loss": 0.0147, "epoch": 0.38570976920644956, "grad_norm": 6.101235389709473, "learning_rate": 8.222012066835249e-05, "loss": 1.2234, "query_regularizer_loss": 0.0147, "sparse_loss": 1.1941, "step": 2440 }, { "document_regularizer_loss": 0.015, "epoch": 0.38887132469174834, "grad_norm": 4.011259078979492, "learning_rate": 8.177115153807808e-05, "loss": 1.2604, "query_regularizer_loss": 0.015, "sparse_loss": 1.2304, "step": 2460 }, { "document_regularizer_loss": 0.0147, "epoch": 0.3920328801770471, "grad_norm": 4.433645725250244, "learning_rate": 8.131784422670312e-05, "loss": 1.2839, "query_regularizer_loss": 0.0147, "sparse_loss": 1.2545, "step": 2480 }, { "document_regularizer_loss": 0.0143, "epoch": 0.39519443566234586, "grad_norm": 4.499387741088867, "learning_rate": 8.086026063092051e-05, "loss": 1.247, "query_regularizer_loss": 0.0143, "sparse_loss": 1.2184, "step": 2500 }, { "document_regularizer_loss": 0.0141, "epoch": 0.39835599114764464, "grad_norm": 3.8418445587158203, "learning_rate": 8.039846323132702e-05, "loss": 1.1972, "query_regularizer_loss": 0.0141, "sparse_loss": 1.169, "step": 2520 }, { "document_regularizer_loss": 0.0144, "epoch": 0.4015175466329434, "grad_norm": 4.753872871398926, "learning_rate": 7.993251508389176e-05, "loss": 1.1735, "query_regularizer_loss": 0.0144, "sparse_loss": 1.1446, "step": 2540 }, { "document_regularizer_loss": 0.0147, "epoch": 0.40467910211824215, "grad_norm": 5.184446811676025, "learning_rate": 7.94624798113464e-05, "loss": 1.1969, "query_regularizer_loss": 0.0147, "sparse_loss": 1.1675, "step": 2560 }, { "document_regularizer_loss": 0.0139, "epoch": 0.40784065760354093, "grad_norm": 3.993455648422241, "learning_rate": 7.898842159449758e-05, "loss": 1.1012, "query_regularizer_loss": 0.0139, "sparse_loss": 1.0734, "step": 2580 }, { "document_regularizer_loss": 0.0139, "epoch": 0.4110022130888397, "grad_norm": 3.5968716144561768, "learning_rate": 7.85104051634636e-05, "loss": 1.186, "query_regularizer_loss": 0.0139, "sparse_loss": 1.1582, "step": 2600 }, { "document_regularizer_loss": 0.0138, "epoch": 0.4141637685741385, "grad_norm": 3.9869585037231445, "learning_rate": 7.802849578883577e-05, "loss": 1.2217, "query_regularizer_loss": 0.0139, "sparse_loss": 1.194, "step": 2620 }, { "document_regularizer_loss": 0.0139, "epoch": 0.41732532405943723, "grad_norm": 4.377500534057617, "learning_rate": 7.754275927276604e-05, "loss": 1.1483, "query_regularizer_loss": 0.0139, "sparse_loss": 1.1206, "step": 2640 }, { "document_regularizer_loss": 0.0142, "epoch": 0.420486879544736, "grad_norm": 4.8011980056762695, "learning_rate": 7.705326193998207e-05, "loss": 1.1347, "query_regularizer_loss": 0.0142, "sparse_loss": 1.1062, "step": 2660 }, { "document_regularizer_loss": 0.0142, "epoch": 0.4236484350300348, "grad_norm": 5.086633682250977, "learning_rate": 7.656007062873101e-05, "loss": 1.1717, "query_regularizer_loss": 0.0142, "sparse_loss": 1.1433, "step": 2680 }, { "document_regularizer_loss": 0.0141, "epoch": 0.4268099905153335, "grad_norm": 6.009626388549805, "learning_rate": 7.606325268165298e-05, "loss": 1.1639, "query_regularizer_loss": 0.0141, "sparse_loss": 1.1357, "step": 2700 }, { "document_regularizer_loss": 0.014, "epoch": 0.4299715460006323, "grad_norm": 3.995734214782715, "learning_rate": 7.556287593658593e-05, "loss": 1.1368, "query_regularizer_loss": 0.014, "sparse_loss": 1.1088, "step": 2720 }, { "document_regularizer_loss": 0.014, "epoch": 0.4331331014859311, "grad_norm": 4.510756015777588, "learning_rate": 7.505900871730262e-05, "loss": 1.0291, "query_regularizer_loss": 0.0139, "sparse_loss": 1.0012, "step": 2740 }, { "document_regularizer_loss": 0.0141, "epoch": 0.4362946569712298, "grad_norm": 3.8161864280700684, "learning_rate": 7.455171982418147e-05, "loss": 1.1819, "query_regularizer_loss": 0.014, "sparse_loss": 1.1538, "step": 2760 }, { "document_regularizer_loss": 0.0133, "epoch": 0.4394562124565286, "grad_norm": 4.2942118644714355, "learning_rate": 7.404107852481215e-05, "loss": 1.1746, "query_regularizer_loss": 0.0133, "sparse_loss": 1.148, "step": 2780 }, { "document_regularizer_loss": 0.0135, "epoch": 0.4426177679418274, "grad_norm": 4.7691473960876465, "learning_rate": 7.352715454453752e-05, "loss": 1.0858, "query_regularizer_loss": 0.0135, "sparse_loss": 1.0588, "step": 2800 }, { "document_regularizer_loss": 0.014, "epoch": 0.4457793234271262, "grad_norm": 4.439463138580322, "learning_rate": 7.301001805693299e-05, "loss": 1.1154, "query_regularizer_loss": 0.014, "sparse_loss": 1.0873, "step": 2820 }, { "document_regularizer_loss": 0.0136, "epoch": 0.4489408789124249, "grad_norm": 4.452930927276611, "learning_rate": 7.24897396742247e-05, "loss": 1.1061, "query_regularizer_loss": 0.0136, "sparse_loss": 1.0788, "step": 2840 }, { "document_regularizer_loss": 0.0136, "epoch": 0.4521024343977237, "grad_norm": 4.004344940185547, "learning_rate": 7.196639043764778e-05, "loss": 1.1019, "query_regularizer_loss": 0.0136, "sparse_loss": 1.0747, "step": 2860 }, { "document_regularizer_loss": 0.0132, "epoch": 0.45526398988302247, "grad_norm": 4.1848320960998535, "learning_rate": 7.144004180774607e-05, "loss": 1.1331, "query_regularizer_loss": 0.0132, "sparse_loss": 1.1067, "step": 2880 }, { "document_regularizer_loss": 0.0131, "epoch": 0.4584255453683212, "grad_norm": 5.027848720550537, "learning_rate": 7.091076565461454e-05, "loss": 1.1157, "query_regularizer_loss": 0.0131, "sparse_loss": 1.0895, "step": 2900 }, { "document_regularizer_loss": 0.0129, "epoch": 0.46158710085362, "grad_norm": 4.188879013061523, "learning_rate": 7.037863424808584e-05, "loss": 0.998, "query_regularizer_loss": 0.0129, "sparse_loss": 0.9723, "step": 2920 }, { "document_regularizer_loss": 0.0129, "epoch": 0.46474865633891876, "grad_norm": 4.0563154220581055, "learning_rate": 6.98437202478622e-05, "loss": 1.1248, "query_regularizer_loss": 0.0129, "sparse_loss": 1.099, "step": 2940 }, { "document_regularizer_loss": 0.0125, "epoch": 0.4679102118242175, "grad_norm": 4.203678131103516, "learning_rate": 6.93060966935942e-05, "loss": 1.1063, "query_regularizer_loss": 0.0125, "sparse_loss": 1.0812, "step": 2960 }, { "document_regularizer_loss": 0.0119, "epoch": 0.4710717673095163, "grad_norm": 3.845771074295044, "learning_rate": 6.876583699490753e-05, "loss": 1.1335, "query_regularizer_loss": 0.0119, "sparse_loss": 1.1097, "step": 2980 }, { "document_regularizer_loss": 0.0122, "epoch": 0.47423332279481506, "grad_norm": 4.067045211791992, "learning_rate": 6.82230149213793e-05, "loss": 1.0592, "query_regularizer_loss": 0.0122, "sparse_loss": 1.0349, "step": 3000 }, { "document_regularizer_loss": 0.0118, "epoch": 0.47739487828011384, "grad_norm": 5.053692817687988, "learning_rate": 6.76777045924653e-05, "loss": 1.0198, "query_regularizer_loss": 0.0118, "sparse_loss": 0.9962, "step": 3020 }, { "document_regularizer_loss": 0.0118, "epoch": 0.48055643376541257, "grad_norm": 4.124359607696533, "learning_rate": 6.712998046737926e-05, "loss": 1.0803, "query_regularizer_loss": 0.0118, "sparse_loss": 1.0567, "step": 3040 }, { "document_regularizer_loss": 0.0114, "epoch": 0.48371798925071136, "grad_norm": 4.783193111419678, "learning_rate": 6.657991733492586e-05, "loss": 1.0682, "query_regularizer_loss": 0.0114, "sparse_loss": 1.0454, "step": 3060 }, { "document_regularizer_loss": 0.0117, "epoch": 0.48687954473601014, "grad_norm": 4.906167507171631, "learning_rate": 6.602759030328878e-05, "loss": 1.0574, "query_regularizer_loss": 0.0118, "sparse_loss": 1.0338, "step": 3080 }, { "document_regularizer_loss": 0.0115, "epoch": 0.49004110022130887, "grad_norm": 4.0294036865234375, "learning_rate": 6.547307478977504e-05, "loss": 1.0531, "query_regularizer_loss": 0.0115, "sparse_loss": 1.0301, "step": 3100 }, { "document_regularizer_loss": 0.0116, "epoch": 0.49320265570660765, "grad_norm": 4.094285488128662, "learning_rate": 6.49164465105172e-05, "loss": 1.1137, "query_regularizer_loss": 0.0116, "sparse_loss": 1.0906, "step": 3120 }, { "document_regularizer_loss": 0.0117, "epoch": 0.49636421119190643, "grad_norm": 4.972708702087402, "learning_rate": 6.435778147013468e-05, "loss": 1.082, "query_regularizer_loss": 0.0118, "sparse_loss": 1.0585, "step": 3140 }, { "document_regularizer_loss": 0.0117, "epoch": 0.49952576667720516, "grad_norm": 3.6531054973602295, "learning_rate": 6.379715595135571e-05, "loss": 1.0133, "query_regularizer_loss": 0.0118, "sparse_loss": 0.9898, "step": 3160 }, { "document_regularizer_loss": 0.0116, "epoch": 0.502687322162504, "grad_norm": 5.076261520385742, "learning_rate": 6.323464650460145e-05, "loss": 1.1048, "query_regularizer_loss": 0.0116, "sparse_loss": 1.0816, "step": 3180 }, { "document_regularizer_loss": 0.0114, "epoch": 0.5058488776478027, "grad_norm": 4.060733318328857, "learning_rate": 6.267032993753337e-05, "loss": 1.1139, "query_regularizer_loss": 0.0114, "sparse_loss": 1.091, "step": 3200 }, { "document_regularizer_loss": 0.0114, "epoch": 0.5090104331331015, "grad_norm": 5.026739120483398, "learning_rate": 6.210428330456546e-05, "loss": 1.0924, "query_regularizer_loss": 0.0113, "sparse_loss": 1.0697, "step": 3220 }, { "document_regularizer_loss": 0.0114, "epoch": 0.5121719886184003, "grad_norm": 4.5130815505981445, "learning_rate": 6.153658389634312e-05, "loss": 1.0344, "query_regularizer_loss": 0.0114, "sparse_loss": 1.0117, "step": 3240 }, { "document_regularizer_loss": 0.0117, "epoch": 0.515333544103699, "grad_norm": 4.630151748657227, "learning_rate": 6.096730922918927e-05, "loss": 1.0427, "query_regularizer_loss": 0.0117, "sparse_loss": 1.0193, "step": 3260 }, { "document_regularizer_loss": 0.0112, "epoch": 0.5184950995889978, "grad_norm": 5.487637042999268, "learning_rate": 6.0396537034520094e-05, "loss": 1.0867, "query_regularizer_loss": 0.0112, "sparse_loss": 1.0644, "step": 3280 }, { "document_regularizer_loss": 0.0111, "epoch": 0.5216566550742966, "grad_norm": 4.014909744262695, "learning_rate": 5.982434524823113e-05, "loss": 1.0218, "query_regularizer_loss": 0.0111, "sparse_loss": 0.9996, "step": 3300 }, { "document_regularizer_loss": 0.0118, "epoch": 0.5248182105595953, "grad_norm": 4.691087245941162, "learning_rate": 5.9250812000055563e-05, "loss": 1.084, "query_regularizer_loss": 0.0117, "sparse_loss": 1.0605, "step": 3320 }, { "document_regularizer_loss": 0.0114, "epoch": 0.527979766044894, "grad_norm": 4.6859331130981445, "learning_rate": 5.867601560289608e-05, "loss": 1.0198, "query_regularizer_loss": 0.0114, "sparse_loss": 0.997, "step": 3340 }, { "document_regularizer_loss": 0.0112, "epoch": 0.5311413215301929, "grad_norm": 4.256602764129639, "learning_rate": 5.8100034542131575e-05, "loss": 1.0805, "query_regularizer_loss": 0.0112, "sparse_loss": 1.0581, "step": 3360 }, { "document_regularizer_loss": 0.0109, "epoch": 0.5343028770154916, "grad_norm": 4.678290367126465, "learning_rate": 5.7522947464900436e-05, "loss": 0.959, "query_regularizer_loss": 0.0109, "sparse_loss": 0.9373, "step": 3380 }, { "document_regularizer_loss": 0.0112, "epoch": 0.5374644325007903, "grad_norm": 3.6500658988952637, "learning_rate": 5.694483316936173e-05, "loss": 1.0043, "query_regularizer_loss": 0.0112, "sparse_loss": 0.9819, "step": 3400 }, { "document_regularizer_loss": 0.0112, "epoch": 0.5406259879860892, "grad_norm": 3.8542468547821045, "learning_rate": 5.636577059393566e-05, "loss": 1.0425, "query_regularizer_loss": 0.0112, "sparse_loss": 1.0201, "step": 3420 }, { "document_regularizer_loss": 0.0116, "epoch": 0.5437875434713879, "grad_norm": 4.720110893249512, "learning_rate": 5.5785838806524994e-05, "loss": 1.0322, "query_regularizer_loss": 0.0116, "sparse_loss": 1.009, "step": 3440 }, { "document_regularizer_loss": 0.0114, "epoch": 0.5469490989566866, "grad_norm": 6.155559539794922, "learning_rate": 5.5205116993718774e-05, "loss": 1.0394, "query_regularizer_loss": 0.0114, "sparse_loss": 1.0166, "step": 3460 }, { "document_regularizer_loss": 0.011, "epoch": 0.5501106544419855, "grad_norm": 4.21190881729126, "learning_rate": 5.462368444997975e-05, "loss": 1.0332, "query_regularizer_loss": 0.0111, "sparse_loss": 1.0111, "step": 3480 }, { "document_regularizer_loss": 0.0109, "epoch": 0.5532722099272842, "grad_norm": 4.882951736450195, "learning_rate": 5.404162056681724e-05, "loss": 1.0251, "query_regularizer_loss": 0.0109, "sparse_loss": 1.0033, "step": 3500 }, { "document_regularizer_loss": 0.011, "epoch": 0.5564337654125829, "grad_norm": 5.064315319061279, "learning_rate": 5.3459004821946544e-05, "loss": 1.0058, "query_regularizer_loss": 0.0111, "sparse_loss": 0.9837, "step": 3520 }, { "document_regularizer_loss": 0.011, "epoch": 0.5595953208978818, "grad_norm": 4.2110514640808105, "learning_rate": 5.287591676843675e-05, "loss": 1.0643, "query_regularizer_loss": 0.011, "sparse_loss": 1.0422, "step": 3540 }, { "document_regularizer_loss": 0.0107, "epoch": 0.5627568763831805, "grad_norm": 4.3318681716918945, "learning_rate": 5.229243602384819e-05, "loss": 1.0501, "query_regularizer_loss": 0.0107, "sparse_loss": 1.0287, "step": 3560 }, { "document_regularizer_loss": 0.0106, "epoch": 0.5659184318684793, "grad_norm": 3.779938220977783, "learning_rate": 5.170864225936103e-05, "loss": 0.9583, "query_regularizer_loss": 0.0106, "sparse_loss": 0.9372, "step": 3580 }, { "document_regularizer_loss": 0.011, "epoch": 0.5690799873537781, "grad_norm": 5.420925617218018, "learning_rate": 5.112461518889663e-05, "loss": 1.0118, "query_regularizer_loss": 0.011, "sparse_loss": 0.9897, "step": 3600 }, { "document_regularizer_loss": 0.0107, "epoch": 0.5722415428390768, "grad_norm": 5.2877888679504395, "learning_rate": 5.054043455823306e-05, "loss": 0.9915, "query_regularizer_loss": 0.0107, "sparse_loss": 0.9701, "step": 3620 }, { "document_regularizer_loss": 0.0104, "epoch": 0.5754030983243756, "grad_norm": 5.362128734588623, "learning_rate": 4.995618013411621e-05, "loss": 0.9232, "query_regularizer_loss": 0.0103, "sparse_loss": 0.9025, "step": 3640 }, { "document_regularizer_loss": 0.0103, "epoch": 0.5785646538096744, "grad_norm": 4.5731377601623535, "learning_rate": 4.93719316933681e-05, "loss": 0.9832, "query_regularizer_loss": 0.0103, "sparse_loss": 0.9626, "step": 3660 }, { "document_regularizer_loss": 0.0106, "epoch": 0.5817262092949731, "grad_norm": 4.423415660858154, "learning_rate": 4.878776901199367e-05, "loss": 0.9539, "query_regularizer_loss": 0.0107, "sparse_loss": 0.9327, "step": 3680 }, { "document_regularizer_loss": 0.0108, "epoch": 0.5848877647802719, "grad_norm": 4.719078540802002, "learning_rate": 4.820377185428794e-05, "loss": 0.9561, "query_regularizer_loss": 0.0108, "sparse_loss": 0.9346, "step": 3700 }, { "document_regularizer_loss": 0.0107, "epoch": 0.5880493202655707, "grad_norm": 5.694866180419922, "learning_rate": 4.762001996194457e-05, "loss": 0.9345, "query_regularizer_loss": 0.0107, "sparse_loss": 0.9132, "step": 3720 }, { "document_regularizer_loss": 0.0106, "epoch": 0.5912108757508694, "grad_norm": 4.105371475219727, "learning_rate": 4.7036593043167486e-05, "loss": 0.9471, "query_regularizer_loss": 0.0106, "sparse_loss": 0.9259, "step": 3740 }, { "document_regularizer_loss": 0.0107, "epoch": 0.5943724312361682, "grad_norm": 4.429325580596924, "learning_rate": 4.645357076178729e-05, "loss": 0.9831, "query_regularizer_loss": 0.0107, "sparse_loss": 0.9617, "step": 3760 }, { "document_regularizer_loss": 0.0101, "epoch": 0.597533986721467, "grad_norm": 4.601746559143066, "learning_rate": 4.5871032726383386e-05, "loss": 1.0021, "query_regularizer_loss": 0.0101, "sparse_loss": 0.9818, "step": 3780 }, { "document_regularizer_loss": 0.0099, "epoch": 0.6006955422067657, "grad_norm": 4.143659591674805, "learning_rate": 4.52890584794141e-05, "loss": 1.0178, "query_regularizer_loss": 0.0099, "sparse_loss": 0.9981, "step": 3800 }, { "document_regularizer_loss": 0.0098, "epoch": 0.6038570976920645, "grad_norm": 4.553821086883545, "learning_rate": 4.47077274863554e-05, "loss": 0.9706, "query_regularizer_loss": 0.0098, "sparse_loss": 0.951, "step": 3820 }, { "document_regularizer_loss": 0.0101, "epoch": 0.6070186531773633, "grad_norm": 4.846632957458496, "learning_rate": 4.4127119124850364e-05, "loss": 0.9825, "query_regularizer_loss": 0.0101, "sparse_loss": 0.9623, "step": 3840 }, { "document_regularizer_loss": 0.0099, "epoch": 0.610180208662662, "grad_norm": 4.342281818389893, "learning_rate": 4.35473126738706e-05, "loss": 1.016, "query_regularizer_loss": 0.01, "sparse_loss": 0.9961, "step": 3860 }, { "document_regularizer_loss": 0.0101, "epoch": 0.6133417641479608, "grad_norm": 4.643177509307861, "learning_rate": 4.2968387302891126e-05, "loss": 0.8974, "query_regularizer_loss": 0.0101, "sparse_loss": 0.8772, "step": 3880 }, { "document_regularizer_loss": 0.0103, "epoch": 0.6165033196332595, "grad_norm": 4.818831920623779, "learning_rate": 4.239042206108016e-05, "loss": 0.8818, "query_regularizer_loss": 0.0103, "sparse_loss": 0.8613, "step": 3900 }, { "document_regularizer_loss": 0.01, "epoch": 0.6196648751185583, "grad_norm": 4.226445198059082, "learning_rate": 4.181349586650545e-05, "loss": 0.913, "query_regularizer_loss": 0.01, "sparse_loss": 0.893, "step": 3920 }, { "document_regularizer_loss": 0.01, "epoch": 0.6228264306038571, "grad_norm": 5.203937530517578, "learning_rate": 4.123768749535834e-05, "loss": 0.9275, "query_regularizer_loss": 0.01, "sparse_loss": 0.9076, "step": 3940 }, { "document_regularizer_loss": 0.01, "epoch": 0.6259879860891558, "grad_norm": 5.980523109436035, "learning_rate": 4.066307557119741e-05, "loss": 0.9632, "query_regularizer_loss": 0.01, "sparse_loss": 0.9432, "step": 3960 }, { "document_regularizer_loss": 0.01, "epoch": 0.6291495415744547, "grad_norm": 4.493658065795898, "learning_rate": 4.008973855421277e-05, "loss": 0.9866, "query_regularizer_loss": 0.01, "sparse_loss": 0.9666, "step": 3980 }, { "document_regularizer_loss": 0.0099, "epoch": 0.6323110970597534, "grad_norm": 5.311489105224609, "learning_rate": 3.9517754730512734e-05, "loss": 0.9693, "query_regularizer_loss": 0.0099, "sparse_loss": 0.9495, "step": 4000 }, { "epoch": 0.6323110970597534, "eval_runtime": 117.9967, "eval_samples_per_second": 0.0, "eval_sparse-ir-eval_avg_flops": 3206.58056640625, "eval_sparse-ir-eval_corpus_active_dims": 4096.0, "eval_sparse-ir-eval_corpus_sparsity_ratio": 0.9186785260482846, "eval_sparse-ir-eval_dot_accuracy@1": 0.046790641871625675, "eval_sparse-ir-eval_dot_accuracy@100": 0.39992001599680066, "eval_sparse-ir-eval_dot_accuracy@50": 0.30813837232553487, "eval_sparse-ir-eval_dot_accuracy@8": 0.13937212557488501, "eval_sparse-ir-eval_dot_map@100": 0.08368707063784206, "eval_sparse-ir-eval_dot_mrr@10": 0.07560733884969031, "eval_sparse-ir-eval_dot_ndcg@10": 0.09403480206033758, "eval_sparse-ir-eval_dot_precision@1": 0.046790641871625675, "eval_sparse-ir-eval_dot_precision@100": 0.003999200159968006, "eval_sparse-ir-eval_dot_precision@50": 0.006162767446510699, "eval_sparse-ir-eval_dot_precision@8": 0.017421515696860627, "eval_sparse-ir-eval_dot_recall@1": 0.046790641871625675, "eval_sparse-ir-eval_dot_recall@100": 0.39992001599680066, "eval_sparse-ir-eval_dot_recall@50": 0.30813837232553487, "eval_sparse-ir-eval_dot_recall@8": 0.13937212557488501, "eval_sparse-ir-eval_query_active_dims": 4096.0, "eval_sparse-ir-eval_query_sparsity_ratio": 0.9186785260482846, "eval_steps_per_second": 0.0, "step": 4000 }, { "document_regularizer_loss": 0.0095, "epoch": 0.6354726525450521, "grad_norm": 4.420332431793213, "learning_rate": 3.894720220143433e-05, "loss": 0.9294, "query_regularizer_loss": 0.0096, "sparse_loss": 0.9103, "step": 4020 }, { "document_regularizer_loss": 0.0098, "epoch": 0.638634208030351, "grad_norm": 4.270013809204102, "learning_rate": 3.837815887287894e-05, "loss": 0.9444, "query_regularizer_loss": 0.0098, "sparse_loss": 0.9248, "step": 4040 }, { "document_regularizer_loss": 0.0095, "epoch": 0.6417957635156497, "grad_norm": 4.951900482177734, "learning_rate": 3.781070244467469e-05, "loss": 0.9298, "query_regularizer_loss": 0.0095, "sparse_loss": 0.9108, "step": 4060 }, { "document_regularizer_loss": 0.0094, "epoch": 0.6449573190009484, "grad_norm": 4.797527313232422, "learning_rate": 3.724491039996691e-05, "loss": 0.8788, "query_regularizer_loss": 0.0095, "sparse_loss": 0.8599, "step": 4080 }, { "document_regularizer_loss": 0.0093, "epoch": 0.6481188744862473, "grad_norm": 4.516369819641113, "learning_rate": 3.668085999463825e-05, "loss": 0.9033, "query_regularizer_loss": 0.0093, "sparse_loss": 0.8847, "step": 4100 }, { "document_regularizer_loss": 0.0094, "epoch": 0.651280429971546, "grad_norm": 5.438692092895508, "learning_rate": 3.611862824675982e-05, "loss": 0.9328, "query_regularizer_loss": 0.0094, "sparse_loss": 0.914, "step": 4120 }, { "document_regularizer_loss": 0.0092, "epoch": 0.6544419854568447, "grad_norm": 4.26346492767334, "learning_rate": 3.555829192607475e-05, "loss": 0.9284, "query_regularizer_loss": 0.0093, "sparse_loss": 0.9099, "step": 4140 }, { "document_regularizer_loss": 0.0092, "epoch": 0.6576035409421436, "grad_norm": 5.638837814331055, "learning_rate": 3.499992754351572e-05, "loss": 0.9507, "query_regularizer_loss": 0.0092, "sparse_loss": 0.9323, "step": 4160 }, { "document_regularizer_loss": 0.0092, "epoch": 0.6607650964274423, "grad_norm": 3.8766870498657227, "learning_rate": 3.44436113407578e-05, "loss": 0.9688, "query_regularizer_loss": 0.0092, "sparse_loss": 0.9503, "step": 4180 }, { "document_regularizer_loss": 0.0093, "epoch": 0.663926651912741, "grad_norm": 4.3223161697387695, "learning_rate": 3.388941927980803e-05, "loss": 0.9124, "query_regularizer_loss": 0.0093, "sparse_loss": 0.8937, "step": 4200 }, { "document_regularizer_loss": 0.0096, "epoch": 0.6670882073980399, "grad_norm": 4.250365257263184, "learning_rate": 3.3337427032633195e-05, "loss": 0.8412, "query_regularizer_loss": 0.0096, "sparse_loss": 0.8219, "step": 4220 }, { "document_regularizer_loss": 0.0093, "epoch": 0.6702497628833386, "grad_norm": 4.510296821594238, "learning_rate": 3.278770997082729e-05, "loss": 0.9065, "query_regularizer_loss": 0.0093, "sparse_loss": 0.8878, "step": 4240 }, { "document_regularizer_loss": 0.0088, "epoch": 0.6734113183686373, "grad_norm": 4.981862545013428, "learning_rate": 3.224034315531992e-05, "loss": 0.8408, "query_regularizer_loss": 0.0088, "sparse_loss": 0.8231, "step": 4260 }, { "document_regularizer_loss": 0.009, "epoch": 0.6765728738539362, "grad_norm": 5.086022853851318, "learning_rate": 3.169540132612706e-05, "loss": 0.9456, "query_regularizer_loss": 0.009, "sparse_loss": 0.9276, "step": 4280 }, { "document_regularizer_loss": 0.009, "epoch": 0.6797344293392349, "grad_norm": 4.191604137420654, "learning_rate": 3.11529588921458e-05, "loss": 0.8641, "query_regularizer_loss": 0.009, "sparse_loss": 0.8462, "step": 4300 }, { "document_regularizer_loss": 0.0091, "epoch": 0.6828959848245336, "grad_norm": 4.860748767852783, "learning_rate": 3.061308992099407e-05, "loss": 0.894, "query_regularizer_loss": 0.009, "sparse_loss": 0.8759, "step": 4320 }, { "document_regularizer_loss": 0.0092, "epoch": 0.6860575403098325, "grad_norm": 4.463791847229004, "learning_rate": 3.00758681288973e-05, "loss": 0.8876, "query_regularizer_loss": 0.0092, "sparse_loss": 0.8693, "step": 4340 }, { "document_regularizer_loss": 0.0091, "epoch": 0.6892190957951312, "grad_norm": 4.914661884307861, "learning_rate": 2.95413668706227e-05, "loss": 0.935, "query_regularizer_loss": 0.0091, "sparse_loss": 0.9167, "step": 4360 }, { "document_regularizer_loss": 0.0088, "epoch": 0.69238065128043, "grad_norm": 4.835674285888672, "learning_rate": 2.9009659129463096e-05, "loss": 0.8805, "query_regularizer_loss": 0.0088, "sparse_loss": 0.863, "step": 4380 }, { "document_regularizer_loss": 0.0088, "epoch": 0.6955422067657288, "grad_norm": 4.686373710632324, "learning_rate": 2.848081750727154e-05, "loss": 0.8278, "query_regularizer_loss": 0.0088, "sparse_loss": 0.8103, "step": 4400 }, { "document_regularizer_loss": 0.0089, "epoch": 0.6987037622510275, "grad_norm": 4.177417278289795, "learning_rate": 2.7954914214547757e-05, "loss": 0.8243, "query_regularizer_loss": 0.0089, "sparse_loss": 0.8065, "step": 4420 }, { "document_regularizer_loss": 0.009, "epoch": 0.7018653177363263, "grad_norm": 4.196948051452637, "learning_rate": 2.743202106057832e-05, "loss": 0.8886, "query_regularizer_loss": 0.009, "sparse_loss": 0.8705, "step": 4440 }, { "document_regularizer_loss": 0.009, "epoch": 0.705026873221625, "grad_norm": 4.922492027282715, "learning_rate": 2.6912209443631396e-05, "loss": 0.8486, "query_regularizer_loss": 0.009, "sparse_loss": 0.8305, "step": 4460 }, { "document_regularizer_loss": 0.0089, "epoch": 0.7081884287069238, "grad_norm": 4.493139743804932, "learning_rate": 2.639555034120777e-05, "loss": 0.8496, "query_regularizer_loss": 0.0088, "sparse_loss": 0.8319, "step": 4480 }, { "document_regularizer_loss": 0.0089, "epoch": 0.7113499841922226, "grad_norm": 5.216787338256836, "learning_rate": 2.588211430034919e-05, "loss": 0.8937, "query_regularizer_loss": 0.0089, "sparse_loss": 0.8758, "step": 4500 }, { "document_regularizer_loss": 0.0085, "epoch": 0.7145115396775213, "grad_norm": 5.056200981140137, "learning_rate": 2.5371971428005526e-05, "loss": 0.931, "query_regularizer_loss": 0.0085, "sparse_loss": 0.9141, "step": 4520 }, { "document_regularizer_loss": 0.0083, "epoch": 0.7176730951628201, "grad_norm": 4.5209527015686035, "learning_rate": 2.4865191381462076e-05, "loss": 0.8276, "query_regularizer_loss": 0.0083, "sparse_loss": 0.811, "step": 4540 }, { "document_regularizer_loss": 0.0085, "epoch": 0.7208346506481189, "grad_norm": 5.0197014808654785, "learning_rate": 2.4361843358828285e-05, "loss": 0.8825, "query_regularizer_loss": 0.0085, "sparse_loss": 0.8656, "step": 4560 }, { "document_regularizer_loss": 0.0086, "epoch": 0.7239962061334176, "grad_norm": 4.634271144866943, "learning_rate": 2.386199608958898e-05, "loss": 0.8827, "query_regularizer_loss": 0.0086, "sparse_loss": 0.8655, "step": 4580 }, { "document_regularizer_loss": 0.0086, "epoch": 0.7271577616187164, "grad_norm": 4.368088245391846, "learning_rate": 2.3365717825219872e-05, "loss": 0.8466, "query_regularizer_loss": 0.0086, "sparse_loss": 0.8294, "step": 4600 }, { "document_regularizer_loss": 0.0087, "epoch": 0.7303193171040152, "grad_norm": 4.631587028503418, "learning_rate": 2.2873076329868075e-05, "loss": 0.9349, "query_regularizer_loss": 0.0087, "sparse_loss": 0.9174, "step": 4620 }, { "document_regularizer_loss": 0.0085, "epoch": 0.7334808725893139, "grad_norm": 5.4434027671813965, "learning_rate": 2.2384138871099297e-05, "loss": 0.8376, "query_regularizer_loss": 0.0085, "sparse_loss": 0.8205, "step": 4640 }, { "document_regularizer_loss": 0.0084, "epoch": 0.7366424280746127, "grad_norm": 4.6973700523376465, "learning_rate": 2.189897221071289e-05, "loss": 0.8292, "query_regularizer_loss": 0.0084, "sparse_loss": 0.8124, "step": 4660 }, { "document_regularizer_loss": 0.0083, "epoch": 0.7398039835599115, "grad_norm": 3.986987590789795, "learning_rate": 2.141764259562576e-05, "loss": 0.8591, "query_regularizer_loss": 0.0083, "sparse_loss": 0.8425, "step": 4680 }, { "document_regularizer_loss": 0.0084, "epoch": 0.7429655390452102, "grad_norm": 3.8863649368286133, "learning_rate": 2.0940215748826813e-05, "loss": 0.7949, "query_regularizer_loss": 0.0084, "sparse_loss": 0.7782, "step": 4700 }, { "document_regularizer_loss": 0.0084, "epoch": 0.746127094530509, "grad_norm": 5.146430492401123, "learning_rate": 2.0466756860402724e-05, "loss": 0.9626, "query_regularizer_loss": 0.0084, "sparse_loss": 0.9457, "step": 4720 }, { "document_regularizer_loss": 0.0083, "epoch": 0.7492886500158078, "grad_norm": 4.899904727935791, "learning_rate": 1.9997330578636715e-05, "loss": 0.8082, "query_regularizer_loss": 0.0083, "sparse_loss": 0.7916, "step": 4740 }, { "document_regularizer_loss": 0.0083, "epoch": 0.7524502055011065, "grad_norm": 4.620065689086914, "learning_rate": 1.9532001001181043e-05, "loss": 0.7986, "query_regularizer_loss": 0.0083, "sparse_loss": 0.782, "step": 4760 }, { "document_regularizer_loss": 0.0085, "epoch": 0.7556117609864053, "grad_norm": 5.927558898925781, "learning_rate": 1.9070831666304867e-05, "loss": 0.8056, "query_regularizer_loss": 0.0085, "sparse_loss": 0.7887, "step": 4780 }, { "document_regularizer_loss": 0.0084, "epoch": 0.7587733164717041, "grad_norm": 4.596479892730713, "learning_rate": 1.8613885544218486e-05, "loss": 0.857, "query_regularizer_loss": 0.0085, "sparse_loss": 0.8402, "step": 4800 }, { "document_regularizer_loss": 0.0083, "epoch": 0.7619348719570028, "grad_norm": 3.580037832260132, "learning_rate": 1.8161225028474978e-05, "loss": 0.8355, "query_regularizer_loss": 0.0083, "sparse_loss": 0.8188, "step": 4820 }, { "document_regularizer_loss": 0.0082, "epoch": 0.7650964274423017, "grad_norm": 5.386363506317139, "learning_rate": 1.771291192745081e-05, "loss": 0.8903, "query_regularizer_loss": 0.0082, "sparse_loss": 0.8739, "step": 4840 }, { "document_regularizer_loss": 0.008, "epoch": 0.7682579829276004, "grad_norm": 4.491359233856201, "learning_rate": 1.7269007455906162e-05, "loss": 0.8429, "query_regularizer_loss": 0.008, "sparse_loss": 0.8268, "step": 4860 }, { "document_regularizer_loss": 0.0081, "epoch": 0.7714195384128991, "grad_norm": 5.180741786956787, "learning_rate": 1.682957222662636e-05, "loss": 0.796, "query_regularizer_loss": 0.0081, "sparse_loss": 0.7799, "step": 4880 }, { "document_regularizer_loss": 0.0082, "epoch": 0.774581093898198, "grad_norm": 5.093390941619873, "learning_rate": 1.6394666242145633e-05, "loss": 0.7887, "query_regularizer_loss": 0.0082, "sparse_loss": 0.7722, "step": 4900 }, { "document_regularizer_loss": 0.0082, "epoch": 0.7777426493834967, "grad_norm": 4.695287227630615, "learning_rate": 1.5964348886553997e-05, "loss": 0.887, "query_regularizer_loss": 0.0082, "sparse_loss": 0.8706, "step": 4920 }, { "document_regularizer_loss": 0.008, "epoch": 0.7809042048687954, "grad_norm": 3.9537525177001953, "learning_rate": 1.553867891738864e-05, "loss": 0.8574, "query_regularizer_loss": 0.008, "sparse_loss": 0.8414, "step": 4940 }, { "document_regularizer_loss": 0.0079, "epoch": 0.7840657603540943, "grad_norm": 4.744511127471924, "learning_rate": 1.5117714457611038e-05, "loss": 0.879, "query_regularizer_loss": 0.008, "sparse_loss": 0.8631, "step": 4960 }, { "document_regularizer_loss": 0.0079, "epoch": 0.787227315839393, "grad_norm": 5.013537406921387, "learning_rate": 1.4701512987670374e-05, "loss": 0.8083, "query_regularizer_loss": 0.008, "sparse_loss": 0.7925, "step": 4980 }, { "document_regularizer_loss": 0.0081, "epoch": 0.7903888713246917, "grad_norm": 4.365302085876465, "learning_rate": 1.4290131337655094e-05, "loss": 0.797, "query_regularizer_loss": 0.0081, "sparse_loss": 0.7808, "step": 5000 }, { "document_regularizer_loss": 0.0083, "epoch": 0.7935504268099906, "grad_norm": 4.093557357788086, "learning_rate": 1.3883625679532951e-05, "loss": 0.7833, "query_regularizer_loss": 0.0083, "sparse_loss": 0.7668, "step": 5020 }, { "document_regularizer_loss": 0.0081, "epoch": 0.7967119822952893, "grad_norm": 4.502114772796631, "learning_rate": 1.3482051519480992e-05, "loss": 0.8312, "query_regularizer_loss": 0.0081, "sparse_loss": 0.8149, "step": 5040 }, { "document_regularizer_loss": 0.0081, "epoch": 0.799873537780588, "grad_norm": 5.223890781402588, "learning_rate": 1.308546369030661e-05, "loss": 0.8098, "query_regularizer_loss": 0.0081, "sparse_loss": 0.7937, "step": 5060 }, { "document_regularizer_loss": 0.0082, "epoch": 0.8030350932658868, "grad_norm": 5.283390998840332, "learning_rate": 1.2693916343960272e-05, "loss": 0.7976, "query_regularizer_loss": 0.0082, "sparse_loss": 0.7812, "step": 5080 }, { "document_regularizer_loss": 0.0083, "epoch": 0.8061966487511856, "grad_norm": 3.989577054977417, "learning_rate": 1.2307462944141446e-05, "loss": 0.7444, "query_regularizer_loss": 0.0083, "sparse_loss": 0.7278, "step": 5100 }, { "document_regularizer_loss": 0.0084, "epoch": 0.8093582042364843, "grad_norm": 4.379826545715332, "learning_rate": 1.1926156258998439e-05, "loss": 0.8059, "query_regularizer_loss": 0.0084, "sparse_loss": 0.7891, "step": 5120 }, { "document_regularizer_loss": 0.0083, "epoch": 0.8125197597217831, "grad_norm": 4.5320658683776855, "learning_rate": 1.1550048353923094e-05, "loss": 0.8265, "query_regularizer_loss": 0.0083, "sparse_loss": 0.8099, "step": 5140 }, { "document_regularizer_loss": 0.0083, "epoch": 0.8156813152070819, "grad_norm": 4.99989128112793, "learning_rate": 1.1179190584441652e-05, "loss": 0.8062, "query_regularizer_loss": 0.0083, "sparse_loss": 0.7896, "step": 5160 }, { "document_regularizer_loss": 0.0082, "epoch": 0.8188428706923806, "grad_norm": 4.583944797515869, "learning_rate": 1.0813633589202321e-05, "loss": 0.7823, "query_regularizer_loss": 0.0082, "sparse_loss": 0.7659, "step": 5180 }, { "document_regularizer_loss": 0.0082, "epoch": 0.8220044261776794, "grad_norm": 4.400042533874512, "learning_rate": 1.0453427283060907e-05, "loss": 0.8586, "query_regularizer_loss": 0.0082, "sparse_loss": 0.8423, "step": 5200 }, { "document_regularizer_loss": 0.0082, "epoch": 0.8251659816629782, "grad_norm": 4.8341450691223145, "learning_rate": 1.0098620850265211e-05, "loss": 0.8439, "query_regularizer_loss": 0.0081, "sparse_loss": 0.8276, "step": 5220 }, { "document_regularizer_loss": 0.0081, "epoch": 0.828327537148277, "grad_norm": 4.303659439086914, "learning_rate": 9.749262737739123e-06, "loss": 0.9102, "query_regularizer_loss": 0.0081, "sparse_loss": 0.894, "step": 5240 }, { "document_regularizer_loss": 0.0079, "epoch": 0.8314890926335757, "grad_norm": 4.474036693572998, "learning_rate": 9.405400648467533e-06, "loss": 0.8299, "query_regularizer_loss": 0.0079, "sparse_loss": 0.8142, "step": 5260 }, { "document_regularizer_loss": 0.0079, "epoch": 0.8346506481188745, "grad_norm": 4.432526588439941, "learning_rate": 9.067081534982674e-06, "loss": 0.848, "query_regularizer_loss": 0.0079, "sparse_loss": 0.8322, "step": 5280 }, { "document_regularizer_loss": 0.0079, "epoch": 0.8378122036041733, "grad_norm": 5.212082862854004, "learning_rate": 8.734351592953023e-06, "loss": 0.7742, "query_regularizer_loss": 0.0079, "sparse_loss": 0.7584, "step": 5300 }, { "document_regularizer_loss": 0.008, "epoch": 0.840973759089472, "grad_norm": 5.37485408782959, "learning_rate": 8.407256254875573e-06, "loss": 0.8484, "query_regularizer_loss": 0.008, "sparse_loss": 0.8324, "step": 5320 }, { "document_regularizer_loss": 0.0079, "epoch": 0.8441353145747708, "grad_norm": 5.1344475746154785, "learning_rate": 8.0858401838722e-06, "loss": 0.769, "query_regularizer_loss": 0.0079, "sparse_loss": 0.7531, "step": 5340 }, { "document_regularizer_loss": 0.008, "epoch": 0.8472968700600696, "grad_norm": 5.680397987365723, "learning_rate": 7.770147267591205e-06, "loss": 0.8132, "query_regularizer_loss": 0.0079, "sparse_loss": 0.7973, "step": 5360 }, { "document_regularizer_loss": 0.0079, "epoch": 0.8504584255453683, "grad_norm": 5.465423583984375, "learning_rate": 7.4602206122146915e-06, "loss": 0.821, "query_regularizer_loss": 0.0079, "sparse_loss": 0.8052, "step": 5380 }, { "document_regularizer_loss": 0.0079, "epoch": 0.853619981030667, "grad_norm": 4.828605651855469, "learning_rate": 7.156102536572551e-06, "loss": 0.8055, "query_regularizer_loss": 0.0079, "sparse_loss": 0.7898, "step": 5400 }, { "document_regularizer_loss": 0.0079, "epoch": 0.8567815365159659, "grad_norm": 5.933167457580566, "learning_rate": 6.857834566364207e-06, "loss": 0.8342, "query_regularizer_loss": 0.0079, "sparse_loss": 0.8183, "step": 5420 }, { "document_regularizer_loss": 0.0078, "epoch": 0.8599430920012646, "grad_norm": 4.389745235443115, "learning_rate": 6.56545742848837e-06, "loss": 0.8397, "query_regularizer_loss": 0.0078, "sparse_loss": 0.8241, "step": 5440 }, { "document_regularizer_loss": 0.0078, "epoch": 0.8631046474865633, "grad_norm": 5.277611255645752, "learning_rate": 6.279011045482086e-06, "loss": 0.8156, "query_regularizer_loss": 0.0078, "sparse_loss": 0.8, "step": 5460 }, { "document_regularizer_loss": 0.0078, "epoch": 0.8662662029718622, "grad_norm": 4.179137229919434, "learning_rate": 5.998534530069477e-06, "loss": 0.729, "query_regularizer_loss": 0.0079, "sparse_loss": 0.7133, "step": 5480 }, { "document_regularizer_loss": 0.0079, "epoch": 0.8694277584571609, "grad_norm": 5.640354156494141, "learning_rate": 5.724066179821114e-06, "loss": 0.8196, "query_regularizer_loss": 0.0078, "sparse_loss": 0.8039, "step": 5500 }, { "document_regularizer_loss": 0.0078, "epoch": 0.8725893139424596, "grad_norm": 4.522198677062988, "learning_rate": 5.455643471924693e-06, "loss": 0.828, "query_regularizer_loss": 0.0078, "sparse_loss": 0.8123, "step": 5520 }, { "document_regularizer_loss": 0.0077, "epoch": 0.8757508694277585, "grad_norm": 5.020621299743652, "learning_rate": 5.193303058067745e-06, "loss": 0.7892, "query_regularizer_loss": 0.0077, "sparse_loss": 0.7738, "step": 5540 }, { "document_regularizer_loss": 0.0077, "epoch": 0.8789124249130572, "grad_norm": 4.668663024902344, "learning_rate": 4.937080759433011e-06, "loss": 0.7577, "query_regularizer_loss": 0.0077, "sparse_loss": 0.7423, "step": 5560 }, { "document_regularizer_loss": 0.0078, "epoch": 0.8820739803983559, "grad_norm": 5.669675350189209, "learning_rate": 4.687011561807275e-06, "loss": 0.8399, "query_regularizer_loss": 0.0078, "sparse_loss": 0.8243, "step": 5580 }, { "document_regularizer_loss": 0.0077, "epoch": 0.8852355358836548, "grad_norm": 10.074179649353027, "learning_rate": 4.443129610804231e-06, "loss": 0.7002, "query_regularizer_loss": 0.0077, "sparse_loss": 0.6847, "step": 5600 }, { "document_regularizer_loss": 0.0078, "epoch": 0.8883970913689535, "grad_norm": 4.835682392120361, "learning_rate": 4.2054682072020735e-06, "loss": 0.8063, "query_regularizer_loss": 0.0077, "sparse_loss": 0.7909, "step": 5620 }, { "document_regularizer_loss": 0.0077, "epoch": 0.8915586468542523, "grad_norm": 4.431399345397949, "learning_rate": 3.9740598023964795e-06, "loss": 0.8414, "query_regularizer_loss": 0.0077, "sparse_loss": 0.826, "step": 5640 }, { "document_regularizer_loss": 0.0077, "epoch": 0.8947202023395511, "grad_norm": 4.4177656173706055, "learning_rate": 3.748935993969516e-06, "loss": 0.812, "query_regularizer_loss": 0.0077, "sparse_loss": 0.7966, "step": 5660 }, { "document_regularizer_loss": 0.0077, "epoch": 0.8978817578248498, "grad_norm": 4.823155403137207, "learning_rate": 3.5301275213751815e-06, "loss": 0.8235, "query_regularizer_loss": 0.0077, "sparse_loss": 0.8082, "step": 5680 }, { "document_regularizer_loss": 0.0077, "epoch": 0.9010433133101486, "grad_norm": 4.244443893432617, "learning_rate": 3.3176642617420816e-06, "loss": 0.7945, "query_regularizer_loss": 0.0077, "sparse_loss": 0.7792, "step": 5700 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9042048687954474, "grad_norm": 4.172726631164551, "learning_rate": 3.1115752257938556e-06, "loss": 0.8551, "query_regularizer_loss": 0.0076, "sparse_loss": 0.8399, "step": 5720 }, { "document_regularizer_loss": 0.0077, "epoch": 0.9073664242807461, "grad_norm": 5.08225679397583, "learning_rate": 2.9118885538879814e-06, "loss": 0.861, "query_regularizer_loss": 0.0077, "sparse_loss": 0.8456, "step": 5740 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9105279797660449, "grad_norm": 3.611447811126709, "learning_rate": 2.7186315121732475e-06, "loss": 0.7446, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7293, "step": 5760 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9136895352513437, "grad_norm": 4.82051944732666, "learning_rate": 2.531830488866832e-06, "loss": 0.8001, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7848, "step": 5780 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9168510907366424, "grad_norm": 4.792423725128174, "learning_rate": 2.3515109906510292e-06, "loss": 0.7581, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7428, "step": 5800 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9200126462219412, "grad_norm": 4.480869293212891, "learning_rate": 2.1776976391905125e-06, "loss": 0.8134, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7982, "step": 5820 }, { "document_regularizer_loss": 0.0077, "epoch": 0.92317420170724, "grad_norm": 4.211410045623779, "learning_rate": 2.010414167770347e-06, "loss": 0.8014, "query_regularizer_loss": 0.0077, "sparse_loss": 0.786, "step": 5840 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9263357571925387, "grad_norm": 5.381926536560059, "learning_rate": 1.849683418055337e-06, "loss": 0.8245, "query_regularizer_loss": 0.0076, "sparse_loss": 0.8092, "step": 5860 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9294973126778375, "grad_norm": 5.152965068817139, "learning_rate": 1.6955273369711366e-06, "loss": 0.78, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7648, "step": 5880 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9326588681631363, "grad_norm": 5.174445629119873, "learning_rate": 1.5479669737075154e-06, "loss": 0.7105, "query_regularizer_loss": 0.0076, "sparse_loss": 0.6953, "step": 5900 }, { "document_regularizer_loss": 0.0076, "epoch": 0.935820423648435, "grad_norm": 5.0146989822387695, "learning_rate": 1.407022476844172e-06, "loss": 0.8057, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7904, "step": 5920 }, { "document_regularizer_loss": 0.0077, "epoch": 0.9389819791337338, "grad_norm": 4.791074275970459, "learning_rate": 1.2727130915995755e-06, "loss": 0.8372, "query_regularizer_loss": 0.0077, "sparse_loss": 0.8218, "step": 5940 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9421435346190326, "grad_norm": 3.7246296405792236, "learning_rate": 1.1450571572031388e-06, "loss": 0.7693, "query_regularizer_loss": 0.0077, "sparse_loss": 0.754, "step": 5960 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9453050901043313, "grad_norm": 5.010093688964844, "learning_rate": 1.0240721043910828e-06, "loss": 0.7894, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7742, "step": 5980 }, { "document_regularizer_loss": 0.0077, "epoch": 0.9484666455896301, "grad_norm": 3.6110615730285645, "learning_rate": 9.097744530263796e-07, "loss": 0.847, "query_regularizer_loss": 0.0077, "sparse_loss": 0.8317, "step": 6000 }, { "epoch": 0.9484666455896301, "eval_runtime": 112.1405, "eval_samples_per_second": 0.0, "eval_sparse-ir-eval_avg_flops": 2975.3544921875, "eval_sparse-ir-eval_corpus_active_dims": 4096.0, "eval_sparse-ir-eval_corpus_sparsity_ratio": 0.9186785260482846, "eval_sparse-ir-eval_dot_accuracy@1": 0.04939012197560488, "eval_sparse-ir-eval_dot_accuracy@100": 0.41091781643671266, "eval_sparse-ir-eval_dot_accuracy@50": 0.3137372525494901, "eval_sparse-ir-eval_dot_accuracy@8": 0.14417116576684663, "eval_sparse-ir-eval_dot_map@100": 0.08678875016406272, "eval_sparse-ir-eval_dot_mrr@10": 0.07828410508374503, "eval_sparse-ir-eval_dot_ndcg@10": 0.0971973401543676, "eval_sparse-ir-eval_dot_precision@1": 0.04939012197560488, "eval_sparse-ir-eval_dot_precision@100": 0.0041091781643671265, "eval_sparse-ir-eval_dot_precision@50": 0.006274745050989803, "eval_sparse-ir-eval_dot_precision@8": 0.01802139572085583, "eval_sparse-ir-eval_dot_recall@1": 0.04939012197560488, "eval_sparse-ir-eval_dot_recall@100": 0.41091781643671266, "eval_sparse-ir-eval_dot_recall@50": 0.3137372525494901, "eval_sparse-ir-eval_dot_recall@8": 0.14417116576684663, "eval_sparse-ir-eval_query_active_dims": 4096.0, "eval_sparse-ir-eval_query_sparsity_ratio": 0.9186785260482846, "eval_steps_per_second": 0.0, "step": 6000 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9516282010749288, "grad_norm": 4.8956098556518555, "learning_rate": 8.021798098430234e-07, "loss": 0.8242, "query_regularizer_loss": 0.0076, "sparse_loss": 0.809, "step": 6020 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9547897565602277, "grad_norm": 5.695012092590332, "learning_rate": 7.0130286631504e-07, "loss": 0.7428, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7276, "step": 6040 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9579513120455264, "grad_norm": 5.0694475173950195, "learning_rate": 6.071573966504374e-07, "loss": 0.8008, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7855, "step": 6060 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9611128675308251, "grad_norm": 4.476046085357666, "learning_rate": 5.19756255910403e-07, "loss": 0.7523, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7371, "step": 6080 }, { "document_regularizer_loss": 0.0075, "epoch": 0.964274423016124, "grad_norm": 4.1265692710876465, "learning_rate": 4.391113782540368e-07, "loss": 0.7182, "query_regularizer_loss": 0.0075, "sparse_loss": 0.7031, "step": 6100 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9674359785014227, "grad_norm": 4.413667678833008, "learning_rate": 3.6523377530877647e-07, "loss": 0.8748, "query_regularizer_loss": 0.0076, "sparse_loss": 0.8596, "step": 6120 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9705975339867214, "grad_norm": 4.976710319519043, "learning_rate": 2.9813353466683346e-07, "loss": 0.8329, "query_regularizer_loss": 0.0076, "sparse_loss": 0.8176, "step": 6140 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9737590894720203, "grad_norm": 4.861936569213867, "learning_rate": 2.3781981850778402e-07, "loss": 0.8176, "query_regularizer_loss": 0.0076, "sparse_loss": 0.8025, "step": 6160 }, { "document_regularizer_loss": 0.0075, "epoch": 0.976920644957319, "grad_norm": 6.24556303024292, "learning_rate": 1.8430086234752508e-07, "loss": 0.7237, "query_regularizer_loss": 0.0075, "sparse_loss": 0.7086, "step": 6180 }, { "document_regularizer_loss": 0.0075, "epoch": 0.9800822004426177, "grad_norm": 4.751788139343262, "learning_rate": 1.3758397391376297e-07, "loss": 0.7097, "query_regularizer_loss": 0.0075, "sparse_loss": 0.6947, "step": 6200 }, { "document_regularizer_loss": 0.0075, "epoch": 0.9832437559279166, "grad_norm": 4.708479881286621, "learning_rate": 9.767553214815595e-08, "loss": 0.7186, "query_regularizer_loss": 0.0075, "sparse_loss": 0.7035, "step": 6220 }, { "document_regularizer_loss": 0.0077, "epoch": 0.9864053114132153, "grad_norm": 5.60076904296875, "learning_rate": 6.45809863353497e-08, "loss": 0.8387, "query_regularizer_loss": 0.0077, "sparse_loss": 0.8234, "step": 6240 }, { "document_regularizer_loss": 0.0076, "epoch": 0.989566866898514, "grad_norm": 5.529794216156006, "learning_rate": 3.8304855358861636e-08, "loss": 0.7929, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7777, "step": 6260 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9927284223838129, "grad_norm": 5.613116264343262, "learning_rate": 1.8850727084068765e-08, "loss": 0.7785, "query_regularizer_loss": 0.0077, "sparse_loss": 0.7632, "step": 6280 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9958899778691116, "grad_norm": 5.6589484214782715, "learning_rate": 6.221257868321839e-09, "loss": 0.8018, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7865, "step": 6300 }, { "document_regularizer_loss": 0.0076, "epoch": 0.9990515333544103, "grad_norm": 5.379122257232666, "learning_rate": 4.1817219820217137e-10, "loss": 0.7283, "query_regularizer_loss": 0.0076, "sparse_loss": 0.7131, "step": 6320 } ], "logging_steps": 20, "max_steps": 6326, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 2000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 64, "trial_name": null, "trial_params": null }