{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 5000, "global_step": 3174, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00945179584120983, "grad_norm": 13739.59375, "learning_rate": 6.000000000000001e-08, "loss": 14.4504, "step": 10 }, { "epoch": 0.01890359168241966, "grad_norm": 2555.72998046875, "learning_rate": 2.6e-07, "loss": 14.1577, "step": 20 }, { "epoch": 0.02835538752362949, "grad_norm": 9887.001953125, "learning_rate": 4.0000000000000003e-07, "loss": 13.151, "step": 30 }, { "epoch": 0.03780718336483932, "grad_norm": 9856.8603515625, "learning_rate": 5.6e-07, "loss": 13.2907, "step": 40 }, { "epoch": 0.04725897920604915, "grad_norm": 5284.92724609375, "learning_rate": 7.6e-07, "loss": 12.378, "step": 50 }, { "epoch": 0.05671077504725898, "grad_norm": 3841.773681640625, "learning_rate": 9.600000000000001e-07, "loss": 9.7428, "step": 60 }, { "epoch": 0.0661625708884688, "grad_norm": 5774.35205078125, "learning_rate": 1.1600000000000001e-06, "loss": 7.6229, "step": 70 }, { "epoch": 0.07561436672967864, "grad_norm": 2795.67626953125, "learning_rate": 1.3600000000000001e-06, "loss": 6.3023, "step": 80 }, { "epoch": 0.08506616257088846, "grad_norm": 2087.883056640625, "learning_rate": 1.56e-06, "loss": 5.7153, "step": 90 }, { "epoch": 0.0945179584120983, "grad_norm": 3819.43310546875, "learning_rate": 1.76e-06, "loss": 5.3327, "step": 100 }, { "epoch": 0.10396975425330812, "grad_norm": 4049.7900390625, "learning_rate": 1.9600000000000003e-06, "loss": 4.9699, "step": 110 }, { "epoch": 0.11342155009451796, "grad_norm": 659.6953125, "learning_rate": 2.16e-06, "loss": 4.6378, "step": 120 }, { "epoch": 0.12287334593572778, "grad_norm": 951.3776245117188, "learning_rate": 2.3600000000000003e-06, "loss": 4.3035, "step": 130 }, { "epoch": 0.1323251417769376, "grad_norm": 820.498046875, "learning_rate": 2.56e-06, "loss": 4.0183, "step": 140 }, { "epoch": 0.14177693761814744, "grad_norm": 1631.3897705078125, "learning_rate": 2.7600000000000003e-06, "loss": 3.7, "step": 150 }, { "epoch": 0.15122873345935728, "grad_norm": 1278.76904296875, "learning_rate": 2.96e-06, "loss": 3.3597, "step": 160 }, { "epoch": 0.16068052930056712, "grad_norm": 831.0169067382812, "learning_rate": 3.1600000000000002e-06, "loss": 3.2386, "step": 170 }, { "epoch": 0.17013232514177692, "grad_norm": 1020.6453857421875, "learning_rate": 3.3600000000000004e-06, "loss": 2.9075, "step": 180 }, { "epoch": 0.17958412098298676, "grad_norm": 142.8883819580078, "learning_rate": 3.5600000000000002e-06, "loss": 2.6557, "step": 190 }, { "epoch": 0.1890359168241966, "grad_norm": 1896.7252197265625, "learning_rate": 3.7600000000000004e-06, "loss": 2.3728, "step": 200 }, { "epoch": 0.19848771266540643, "grad_norm": 356.39935302734375, "learning_rate": 3.96e-06, "loss": 2.1427, "step": 210 }, { "epoch": 0.20793950850661624, "grad_norm": 70.32272338867188, "learning_rate": 4.16e-06, "loss": 1.8594, "step": 220 }, { "epoch": 0.21739130434782608, "grad_norm": 184.2218475341797, "learning_rate": 4.360000000000001e-06, "loss": 1.5881, "step": 230 }, { "epoch": 0.22684310018903592, "grad_norm": 260.4848937988281, "learning_rate": 4.56e-06, "loss": 1.4066, "step": 240 }, { "epoch": 0.23629489603024575, "grad_norm": 455.8525085449219, "learning_rate": 4.76e-06, "loss": 1.0801, "step": 250 }, { "epoch": 0.24574669187145556, "grad_norm": 105.61797332763672, "learning_rate": 4.960000000000001e-06, "loss": 0.927, "step": 260 }, { "epoch": 0.2551984877126654, "grad_norm": 68.5740737915039, "learning_rate": 5.1600000000000006e-06, "loss": 0.7127, "step": 270 }, { "epoch": 0.2646502835538752, "grad_norm": 208.01817321777344, "learning_rate": 5.36e-06, "loss": 0.5245, "step": 280 }, { "epoch": 0.2741020793950851, "grad_norm": 69.70160675048828, "learning_rate": 5.560000000000001e-06, "loss": 0.4756, "step": 290 }, { "epoch": 0.2835538752362949, "grad_norm": 42.45973587036133, "learning_rate": 5.76e-06, "loss": 0.3668, "step": 300 }, { "epoch": 0.29300567107750475, "grad_norm": 14.809382438659668, "learning_rate": 5.9600000000000005e-06, "loss": 0.3122, "step": 310 }, { "epoch": 0.30245746691871456, "grad_norm": 106.4734115600586, "learning_rate": 6.16e-06, "loss": 0.2707, "step": 320 }, { "epoch": 0.31190926275992437, "grad_norm": 248.66326904296875, "learning_rate": 6.360000000000001e-06, "loss": 0.2752, "step": 330 }, { "epoch": 0.32136105860113423, "grad_norm": 1199.41064453125, "learning_rate": 6.560000000000001e-06, "loss": 0.287, "step": 340 }, { "epoch": 0.33081285444234404, "grad_norm": 92.04813385009766, "learning_rate": 6.760000000000001e-06, "loss": 0.2436, "step": 350 }, { "epoch": 0.34026465028355385, "grad_norm": 42.634403228759766, "learning_rate": 6.96e-06, "loss": 0.2851, "step": 360 }, { "epoch": 0.3497164461247637, "grad_norm": 6404.83837890625, "learning_rate": 7.16e-06, "loss": 2.0508, "step": 370 }, { "epoch": 0.3591682419659735, "grad_norm": 109.61998748779297, "learning_rate": 7.360000000000001e-06, "loss": 1.6968, "step": 380 }, { "epoch": 0.3686200378071834, "grad_norm": 92.08924102783203, "learning_rate": 7.5600000000000005e-06, "loss": 0.2509, "step": 390 }, { "epoch": 0.3780718336483932, "grad_norm": 31.336467742919922, "learning_rate": 7.76e-06, "loss": 0.2022, "step": 400 }, { "epoch": 0.387523629489603, "grad_norm": 21.53646469116211, "learning_rate": 7.960000000000002e-06, "loss": 0.2043, "step": 410 }, { "epoch": 0.39697542533081287, "grad_norm": 103.8397216796875, "learning_rate": 8.16e-06, "loss": 0.1877, "step": 420 }, { "epoch": 0.4064272211720227, "grad_norm": 13.6918306350708, "learning_rate": 8.36e-06, "loss": 0.1542, "step": 430 }, { "epoch": 0.4158790170132325, "grad_norm": 564.9049072265625, "learning_rate": 8.560000000000001e-06, "loss": 0.1714, "step": 440 }, { "epoch": 0.42533081285444235, "grad_norm": 3.5834062099456787, "learning_rate": 8.76e-06, "loss": 0.1856, "step": 450 }, { "epoch": 0.43478260869565216, "grad_norm": 6.6105637550354, "learning_rate": 8.96e-06, "loss": 0.1631, "step": 460 }, { "epoch": 0.444234404536862, "grad_norm": 18.874755859375, "learning_rate": 9.16e-06, "loss": 0.1556, "step": 470 }, { "epoch": 0.45368620037807184, "grad_norm": 4.490022659301758, "learning_rate": 9.360000000000002e-06, "loss": 0.1196, "step": 480 }, { "epoch": 0.46313799621928164, "grad_norm": 647.4920654296875, "learning_rate": 9.56e-06, "loss": 0.1807, "step": 490 }, { "epoch": 0.4725897920604915, "grad_norm": 5.233212471008301, "learning_rate": 9.760000000000001e-06, "loss": 0.2179, "step": 500 }, { "epoch": 0.4820415879017013, "grad_norm": 10.719651222229004, "learning_rate": 9.960000000000001e-06, "loss": 0.1338, "step": 510 }, { "epoch": 0.4914933837429111, "grad_norm": 16.664508819580078, "learning_rate": 9.970082273747197e-06, "loss": 1.2648, "step": 520 }, { "epoch": 0.500945179584121, "grad_norm": 11.62501335144043, "learning_rate": 9.93268511593119e-06, "loss": 0.1396, "step": 530 }, { "epoch": 0.5103969754253308, "grad_norm": 139.4752960205078, "learning_rate": 9.895287958115183e-06, "loss": 0.1359, "step": 540 }, { "epoch": 0.5198487712665406, "grad_norm": 7.7168145179748535, "learning_rate": 9.857890800299179e-06, "loss": 0.1376, "step": 550 }, { "epoch": 0.5293005671077504, "grad_norm": 32.059261322021484, "learning_rate": 9.820493642483172e-06, "loss": 0.1389, "step": 560 }, { "epoch": 0.5387523629489603, "grad_norm": 7.257635116577148, "learning_rate": 9.783096484667165e-06, "loss": 0.1427, "step": 570 }, { "epoch": 0.5482041587901701, "grad_norm": 7.264402389526367, "learning_rate": 9.74569932685116e-06, "loss": 0.1254, "step": 580 }, { "epoch": 0.55765595463138, "grad_norm": 15.918863296508789, "learning_rate": 9.708302169035154e-06, "loss": 0.1375, "step": 590 }, { "epoch": 0.5671077504725898, "grad_norm": 3.28000807762146, "learning_rate": 9.670905011219147e-06, "loss": 0.1018, "step": 600 }, { "epoch": 0.5765595463137996, "grad_norm": 3.7656383514404297, "learning_rate": 9.633507853403143e-06, "loss": 0.1003, "step": 610 }, { "epoch": 0.5860113421550095, "grad_norm": 2.568873882293701, "learning_rate": 9.596110695587136e-06, "loss": 0.1004, "step": 620 }, { "epoch": 0.5954631379962193, "grad_norm": 95.7247314453125, "learning_rate": 9.558713537771131e-06, "loss": 0.1197, "step": 630 }, { "epoch": 0.6049149338374291, "grad_norm": 1.9285805225372314, "learning_rate": 9.521316379955125e-06, "loss": 0.1489, "step": 640 }, { "epoch": 0.6143667296786389, "grad_norm": 2368.34423828125, "learning_rate": 9.483919222139118e-06, "loss": 0.1099, "step": 650 }, { "epoch": 0.6238185255198487, "grad_norm": 127.53451538085938, "learning_rate": 9.446522064323113e-06, "loss": 0.2462, "step": 660 }, { "epoch": 0.6332703213610587, "grad_norm": 3.0768001079559326, "learning_rate": 9.409124906507107e-06, "loss": 0.8238, "step": 670 }, { "epoch": 0.6427221172022685, "grad_norm": 10.293018341064453, "learning_rate": 9.3717277486911e-06, "loss": 0.1209, "step": 680 }, { "epoch": 0.6521739130434783, "grad_norm": 2.587555408477783, "learning_rate": 9.334330590875095e-06, "loss": 0.1323, "step": 690 }, { "epoch": 0.6616257088846881, "grad_norm": 3.1364760398864746, "learning_rate": 9.296933433059089e-06, "loss": 0.1179, "step": 700 }, { "epoch": 0.6710775047258979, "grad_norm": 7.582152366638184, "learning_rate": 9.259536275243082e-06, "loss": 0.1037, "step": 710 }, { "epoch": 0.6805293005671077, "grad_norm": 1070.914794921875, "learning_rate": 9.222139117427075e-06, "loss": 0.106, "step": 720 }, { "epoch": 0.6899810964083176, "grad_norm": 24.98986053466797, "learning_rate": 9.18474195961107e-06, "loss": 0.1497, "step": 730 }, { "epoch": 0.6994328922495274, "grad_norm": 2.334745407104492, "learning_rate": 9.147344801795064e-06, "loss": 0.0928, "step": 740 }, { "epoch": 0.7088846880907372, "grad_norm": 48.43152618408203, "learning_rate": 9.109947643979057e-06, "loss": 0.0976, "step": 750 }, { "epoch": 0.718336483931947, "grad_norm": 12.193922996520996, "learning_rate": 9.072550486163052e-06, "loss": 0.1007, "step": 760 }, { "epoch": 0.7277882797731569, "grad_norm": 65.48289489746094, "learning_rate": 9.035153328347046e-06, "loss": 0.0961, "step": 770 }, { "epoch": 0.7372400756143668, "grad_norm": 9.52373218536377, "learning_rate": 8.99775617053104e-06, "loss": 0.094, "step": 780 }, { "epoch": 0.7466918714555766, "grad_norm": 31.735408782958984, "learning_rate": 8.960359012715034e-06, "loss": 0.0781, "step": 790 }, { "epoch": 0.7561436672967864, "grad_norm": 3.400327682495117, "learning_rate": 8.922961854899028e-06, "loss": 0.0797, "step": 800 }, { "epoch": 0.7655954631379962, "grad_norm": 523.4867553710938, "learning_rate": 8.896783844427824e-06, "loss": 0.7303, "step": 810 }, { "epoch": 0.775047258979206, "grad_norm": 13.026022911071777, "learning_rate": 8.859386686611819e-06, "loss": 0.1286, "step": 820 }, { "epoch": 0.7844990548204159, "grad_norm": 292.6473083496094, "learning_rate": 8.821989528795813e-06, "loss": 0.1002, "step": 830 }, { "epoch": 0.7939508506616257, "grad_norm": 10.678335189819336, "learning_rate": 8.784592370979806e-06, "loss": 0.5203, "step": 840 }, { "epoch": 0.8034026465028355, "grad_norm": 1.7254323959350586, "learning_rate": 8.747195213163801e-06, "loss": 0.0882, "step": 850 }, { "epoch": 0.8128544423440454, "grad_norm": 61.3662223815918, "learning_rate": 8.709798055347795e-06, "loss": 0.0789, "step": 860 }, { "epoch": 0.8223062381852552, "grad_norm": 9330.720703125, "learning_rate": 8.672400897531788e-06, "loss": 1.3957, "step": 870 }, { "epoch": 0.831758034026465, "grad_norm": 8887.767578125, "learning_rate": 8.635003739715783e-06, "loss": 2.697, "step": 880 }, { "epoch": 0.8412098298676749, "grad_norm": 14477.5322265625, "learning_rate": 8.597606581899777e-06, "loss": 1.646, "step": 890 }, { "epoch": 0.8506616257088847, "grad_norm": 2064.92724609375, "learning_rate": 8.56020942408377e-06, "loss": 1.0418, "step": 900 }, { "epoch": 0.8601134215500945, "grad_norm": 53.577369689941406, "learning_rate": 8.522812266267765e-06, "loss": 0.3661, "step": 910 }, { "epoch": 0.8695652173913043, "grad_norm": 11.63904094696045, "learning_rate": 8.485415108451759e-06, "loss": 0.1646, "step": 920 }, { "epoch": 0.8790170132325141, "grad_norm": 313.92529296875, "learning_rate": 8.448017950635754e-06, "loss": 0.2555, "step": 930 }, { "epoch": 0.888468809073724, "grad_norm": 27.883251190185547, "learning_rate": 8.410620792819745e-06, "loss": 0.0899, "step": 940 }, { "epoch": 0.8979206049149339, "grad_norm": 1.2697465419769287, "learning_rate": 8.37322363500374e-06, "loss": 0.093, "step": 950 }, { "epoch": 0.9073724007561437, "grad_norm": 658.786376953125, "learning_rate": 8.335826477187734e-06, "loss": 0.0879, "step": 960 }, { "epoch": 0.9168241965973535, "grad_norm": 12.427763938903809, "learning_rate": 8.298429319371727e-06, "loss": 0.068, "step": 970 }, { "epoch": 0.9262759924385633, "grad_norm": 2670.397216796875, "learning_rate": 8.261032161555723e-06, "loss": 0.7053, "step": 980 }, { "epoch": 0.9357277882797732, "grad_norm": 147.85740661621094, "learning_rate": 8.223635003739716e-06, "loss": 1.1022, "step": 990 }, { "epoch": 0.945179584120983, "grad_norm": 73.45378112792969, "learning_rate": 8.18623784592371e-06, "loss": 0.3059, "step": 1000 }, { "epoch": 0.9546313799621928, "grad_norm": 7.717315196990967, "learning_rate": 8.148840688107704e-06, "loss": 0.1488, "step": 1010 }, { "epoch": 0.9640831758034026, "grad_norm": 9.288606643676758, "learning_rate": 8.111443530291698e-06, "loss": 0.1161, "step": 1020 }, { "epoch": 0.9735349716446124, "grad_norm": 12.23811149597168, "learning_rate": 8.074046372475693e-06, "loss": 0.0874, "step": 1030 }, { "epoch": 0.9829867674858223, "grad_norm": 3.111706256866455, "learning_rate": 8.036649214659686e-06, "loss": 0.1212, "step": 1040 }, { "epoch": 0.9924385633270322, "grad_norm": 7.152887344360352, "learning_rate": 7.99925205684368e-06, "loss": 0.1249, "step": 1050 }, { "epoch": 1.001890359168242, "grad_norm": 1.6718932390213013, "learning_rate": 7.961854899027675e-06, "loss": 0.1261, "step": 1060 }, { "epoch": 1.011342155009452, "grad_norm": 1.6129039525985718, "learning_rate": 7.924457741211668e-06, "loss": 0.0767, "step": 1070 }, { "epoch": 1.0207939508506616, "grad_norm": 10.476449966430664, "learning_rate": 7.887060583395662e-06, "loss": 0.0995, "step": 1080 }, { "epoch": 1.0302457466918715, "grad_norm": 5.244194030761719, "learning_rate": 7.849663425579657e-06, "loss": 0.1115, "step": 1090 }, { "epoch": 1.0396975425330812, "grad_norm": 3.1006393432617188, "learning_rate": 7.81226626776365e-06, "loss": 0.0878, "step": 1100 }, { "epoch": 1.0491493383742911, "grad_norm": 8.808783531188965, "learning_rate": 7.774869109947646e-06, "loss": 0.105, "step": 1110 }, { "epoch": 1.0586011342155008, "grad_norm": 5.994989395141602, "learning_rate": 7.737471952131639e-06, "loss": 0.1088, "step": 1120 }, { "epoch": 1.0680529300567108, "grad_norm": 2.1118083000183105, "learning_rate": 7.700074794315632e-06, "loss": 0.1051, "step": 1130 }, { "epoch": 1.0775047258979207, "grad_norm": 5.780063629150391, "learning_rate": 7.662677636499628e-06, "loss": 0.0792, "step": 1140 }, { "epoch": 1.0869565217391304, "grad_norm": 0.76413893699646, "learning_rate": 7.625280478683621e-06, "loss": 0.1034, "step": 1150 }, { "epoch": 1.0964083175803403, "grad_norm": 1.0356695652008057, "learning_rate": 7.587883320867615e-06, "loss": 0.0698, "step": 1160 }, { "epoch": 1.10586011342155, "grad_norm": 1.114787220954895, "learning_rate": 7.550486163051609e-06, "loss": 0.0847, "step": 1170 }, { "epoch": 1.11531190926276, "grad_norm": 45.184532165527344, "learning_rate": 7.513089005235603e-06, "loss": 0.1104, "step": 1180 }, { "epoch": 1.1247637051039698, "grad_norm": 1.0391864776611328, "learning_rate": 7.475691847419597e-06, "loss": 0.0829, "step": 1190 }, { "epoch": 1.1342155009451795, "grad_norm": 5.533662796020508, "learning_rate": 7.4382946896035916e-06, "loss": 0.0853, "step": 1200 }, { "epoch": 1.1436672967863895, "grad_norm": 4.5780229568481445, "learning_rate": 7.400897531787585e-06, "loss": 0.0656, "step": 1210 }, { "epoch": 1.1531190926275992, "grad_norm": 1.4483156204223633, "learning_rate": 7.363500373971578e-06, "loss": 0.0748, "step": 1220 }, { "epoch": 1.162570888468809, "grad_norm": 1.0575523376464844, "learning_rate": 7.326103216155573e-06, "loss": 0.0761, "step": 1230 }, { "epoch": 1.172022684310019, "grad_norm": 1.2220410108566284, "learning_rate": 7.288706058339566e-06, "loss": 0.0885, "step": 1240 }, { "epoch": 1.1814744801512287, "grad_norm": 9.184976577758789, "learning_rate": 7.25130890052356e-06, "loss": 0.0747, "step": 1250 }, { "epoch": 1.1909262759924386, "grad_norm": 2.5088539123535156, "learning_rate": 7.213911742707555e-06, "loss": 0.0878, "step": 1260 }, { "epoch": 1.2003780718336483, "grad_norm": 0.895652711391449, "learning_rate": 7.176514584891548e-06, "loss": 0.0648, "step": 1270 }, { "epoch": 1.2098298676748582, "grad_norm": 3.6389918327331543, "learning_rate": 7.139117427075542e-06, "loss": 0.069, "step": 1280 }, { "epoch": 1.2192816635160681, "grad_norm": 1.7553330659866333, "learning_rate": 7.101720269259537e-06, "loss": 0.085, "step": 1290 }, { "epoch": 1.2287334593572778, "grad_norm": 1.7131526470184326, "learning_rate": 7.064323111443531e-06, "loss": 0.0821, "step": 1300 }, { "epoch": 1.2381852551984878, "grad_norm": 1.947082281112671, "learning_rate": 7.026925953627524e-06, "loss": 0.079, "step": 1310 }, { "epoch": 1.2476370510396975, "grad_norm": 0.9147713780403137, "learning_rate": 6.989528795811519e-06, "loss": 0.0854, "step": 1320 }, { "epoch": 1.2570888468809074, "grad_norm": 1.334012746810913, "learning_rate": 6.952131637995513e-06, "loss": 0.1571, "step": 1330 }, { "epoch": 1.2665406427221173, "grad_norm": 1.3452980518341064, "learning_rate": 6.914734480179507e-06, "loss": 0.0861, "step": 1340 }, { "epoch": 1.275992438563327, "grad_norm": 0.8652889132499695, "learning_rate": 6.877337322363501e-06, "loss": 0.0755, "step": 1350 }, { "epoch": 1.285444234404537, "grad_norm": 0.6752623319625854, "learning_rate": 6.839940164547495e-06, "loss": 0.0671, "step": 1360 }, { "epoch": 1.2948960302457466, "grad_norm": 2.3604907989501953, "learning_rate": 6.802543006731489e-06, "loss": 0.0675, "step": 1370 }, { "epoch": 1.3043478260869565, "grad_norm": 84.10884094238281, "learning_rate": 6.7651458489154835e-06, "loss": 0.1184, "step": 1380 }, { "epoch": 1.3137996219281662, "grad_norm": 584.1233520507812, "learning_rate": 6.727748691099477e-06, "loss": 0.1766, "step": 1390 }, { "epoch": 1.3232514177693762, "grad_norm": 5.4759202003479, "learning_rate": 6.690351533283471e-06, "loss": 0.2643, "step": 1400 }, { "epoch": 1.332703213610586, "grad_norm": 8.900699615478516, "learning_rate": 6.6529543754674655e-06, "loss": 0.27, "step": 1410 }, { "epoch": 1.3421550094517958, "grad_norm": 31.19775390625, "learning_rate": 6.615557217651459e-06, "loss": 0.1707, "step": 1420 }, { "epoch": 1.3516068052930057, "grad_norm": 2.57696270942688, "learning_rate": 6.578160059835453e-06, "loss": 0.1375, "step": 1430 }, { "epoch": 1.3610586011342156, "grad_norm": 2.712918996810913, "learning_rate": 6.5407629020194475e-06, "loss": 0.1319, "step": 1440 }, { "epoch": 1.3705103969754253, "grad_norm": 2.8474884033203125, "learning_rate": 6.503365744203442e-06, "loss": 0.0871, "step": 1450 }, { "epoch": 1.3799621928166352, "grad_norm": 3.187256336212158, "learning_rate": 6.465968586387435e-06, "loss": 0.0876, "step": 1460 }, { "epoch": 1.389413988657845, "grad_norm": 1.1946818828582764, "learning_rate": 6.4285714285714295e-06, "loss": 0.0732, "step": 1470 }, { "epoch": 1.3988657844990549, "grad_norm": 2.49619722366333, "learning_rate": 6.391174270755424e-06, "loss": 0.0858, "step": 1480 }, { "epoch": 1.4083175803402646, "grad_norm": 1.9434871673583984, "learning_rate": 6.353777112939416e-06, "loss": 0.1049, "step": 1490 }, { "epoch": 1.4177693761814745, "grad_norm": 1.6459031105041504, "learning_rate": 6.316379955123411e-06, "loss": 0.0706, "step": 1500 }, { "epoch": 1.4272211720226844, "grad_norm": 2.103151798248291, "learning_rate": 6.278982797307405e-06, "loss": 0.0871, "step": 1510 }, { "epoch": 1.436672967863894, "grad_norm": 1.688470721244812, "learning_rate": 6.241585639491398e-06, "loss": 0.0662, "step": 1520 }, { "epoch": 1.446124763705104, "grad_norm": 1.6121007204055786, "learning_rate": 6.204188481675393e-06, "loss": 0.0702, "step": 1530 }, { "epoch": 1.455576559546314, "grad_norm": 1.3256118297576904, "learning_rate": 6.166791323859387e-06, "loss": 0.0867, "step": 1540 }, { "epoch": 1.4650283553875236, "grad_norm": 2.683123826980591, "learning_rate": 6.129394166043381e-06, "loss": 0.0658, "step": 1550 }, { "epoch": 1.4744801512287333, "grad_norm": 11.660983085632324, "learning_rate": 6.091997008227375e-06, "loss": 0.0785, "step": 1560 }, { "epoch": 1.4839319470699432, "grad_norm": 1.9295110702514648, "learning_rate": 6.054599850411369e-06, "loss": 0.085, "step": 1570 }, { "epoch": 1.4933837429111532, "grad_norm": 0.9210951924324036, "learning_rate": 6.017202692595363e-06, "loss": 0.0974, "step": 1580 }, { "epoch": 1.5028355387523629, "grad_norm": 2.4897303581237793, "learning_rate": 5.979805534779357e-06, "loss": 0.0708, "step": 1590 }, { "epoch": 1.5122873345935728, "grad_norm": 2.9634382724761963, "learning_rate": 5.942408376963351e-06, "loss": 0.075, "step": 1600 }, { "epoch": 1.5217391304347827, "grad_norm": 3.2284462451934814, "learning_rate": 5.905011219147345e-06, "loss": 0.1634, "step": 1610 }, { "epoch": 1.5311909262759924, "grad_norm": 2.838763475418091, "learning_rate": 5.867614061331339e-06, "loss": 0.1089, "step": 1620 }, { "epoch": 1.5406427221172023, "grad_norm": 1.9930697679519653, "learning_rate": 5.830216903515334e-06, "loss": 0.0653, "step": 1630 }, { "epoch": 1.5500945179584122, "grad_norm": 1.8953261375427246, "learning_rate": 5.792819745699327e-06, "loss": 0.1015, "step": 1640 }, { "epoch": 1.559546313799622, "grad_norm": 0.7137608528137207, "learning_rate": 5.755422587883321e-06, "loss": 0.0608, "step": 1650 }, { "epoch": 1.5689981096408316, "grad_norm": 1.0345617532730103, "learning_rate": 5.718025430067316e-06, "loss": 0.0767, "step": 1660 }, { "epoch": 1.5784499054820416, "grad_norm": 1.0763232707977295, "learning_rate": 5.680628272251309e-06, "loss": 0.089, "step": 1670 }, { "epoch": 1.5879017013232515, "grad_norm": 13.364884376525879, "learning_rate": 5.643231114435303e-06, "loss": 0.0671, "step": 1680 }, { "epoch": 1.5973534971644612, "grad_norm": 3.817563772201538, "learning_rate": 5.605833956619298e-06, "loss": 0.1104, "step": 1690 }, { "epoch": 1.606805293005671, "grad_norm": 1.828808307647705, "learning_rate": 5.568436798803292e-06, "loss": 0.0953, "step": 1700 }, { "epoch": 1.616257088846881, "grad_norm": 3.8507823944091797, "learning_rate": 5.531039640987285e-06, "loss": 0.0838, "step": 1710 }, { "epoch": 1.6257088846880907, "grad_norm": 3.9542222023010254, "learning_rate": 5.49364248317128e-06, "loss": 0.0792, "step": 1720 }, { "epoch": 1.6351606805293004, "grad_norm": 2.827173948287964, "learning_rate": 5.456245325355274e-06, "loss": 0.0869, "step": 1730 }, { "epoch": 1.6446124763705106, "grad_norm": 8.706243515014648, "learning_rate": 5.418848167539268e-06, "loss": 0.0782, "step": 1740 }, { "epoch": 1.6540642722117203, "grad_norm": 2.825563430786133, "learning_rate": 5.381451009723262e-06, "loss": 0.1127, "step": 1750 }, { "epoch": 1.66351606805293, "grad_norm": 152.21490478515625, "learning_rate": 5.344053851907256e-06, "loss": 0.1218, "step": 1760 }, { "epoch": 1.6729678638941399, "grad_norm": 6.626716613769531, "learning_rate": 5.306656694091249e-06, "loss": 0.0866, "step": 1770 }, { "epoch": 1.6824196597353498, "grad_norm": 1.9379291534423828, "learning_rate": 5.269259536275243e-06, "loss": 0.0829, "step": 1780 }, { "epoch": 1.6918714555765595, "grad_norm": 18.214658737182617, "learning_rate": 5.231862378459237e-06, "loss": 0.0873, "step": 1790 }, { "epoch": 1.7013232514177694, "grad_norm": 17.1606502532959, "learning_rate": 5.194465220643231e-06, "loss": 0.1232, "step": 1800 }, { "epoch": 1.7107750472589793, "grad_norm": 142.15260314941406, "learning_rate": 5.157068062827225e-06, "loss": 0.1019, "step": 1810 }, { "epoch": 1.720226843100189, "grad_norm": 124.0704345703125, "learning_rate": 5.119670905011219e-06, "loss": 0.4851, "step": 1820 }, { "epoch": 1.7296786389413987, "grad_norm": 51.26754379272461, "learning_rate": 5.082273747195213e-06, "loss": 0.394, "step": 1830 }, { "epoch": 1.7391304347826086, "grad_norm": 56.5087890625, "learning_rate": 5.044876589379208e-06, "loss": 0.1282, "step": 1840 }, { "epoch": 1.7485822306238186, "grad_norm": 25.078203201293945, "learning_rate": 5.007479431563201e-06, "loss": 0.0931, "step": 1850 }, { "epoch": 1.7580340264650283, "grad_norm": 1.4061124324798584, "learning_rate": 4.970082273747195e-06, "loss": 0.0745, "step": 1860 }, { "epoch": 1.7674858223062382, "grad_norm": 16.42283058166504, "learning_rate": 4.93268511593119e-06, "loss": 0.0939, "step": 1870 }, { "epoch": 1.776937618147448, "grad_norm": 2.5964503288269043, "learning_rate": 4.895287958115184e-06, "loss": 0.0865, "step": 1880 }, { "epoch": 1.7863894139886578, "grad_norm": 4.37824821472168, "learning_rate": 4.857890800299177e-06, "loss": 0.0872, "step": 1890 }, { "epoch": 1.7958412098298677, "grad_norm": 3.0010366439819336, "learning_rate": 4.820493642483172e-06, "loss": 0.1027, "step": 1900 }, { "epoch": 1.8052930056710776, "grad_norm": 9.359098434448242, "learning_rate": 4.783096484667166e-06, "loss": 0.1273, "step": 1910 }, { "epoch": 1.8147448015122873, "grad_norm": 1.1968307495117188, "learning_rate": 4.74569932685116e-06, "loss": 0.1062, "step": 1920 }, { "epoch": 1.824196597353497, "grad_norm": 8.978023529052734, "learning_rate": 4.708302169035154e-06, "loss": 0.073, "step": 1930 }, { "epoch": 1.833648393194707, "grad_norm": 4.380856513977051, "learning_rate": 4.670905011219148e-06, "loss": 0.0807, "step": 1940 }, { "epoch": 1.8431001890359169, "grad_norm": 1.6301466226577759, "learning_rate": 4.633507853403142e-06, "loss": 0.069, "step": 1950 }, { "epoch": 1.8525519848771266, "grad_norm": 1.8033560514450073, "learning_rate": 4.5961106955871356e-06, "loss": 0.1026, "step": 1960 }, { "epoch": 1.8620037807183365, "grad_norm": 1.4674009084701538, "learning_rate": 4.55871353777113e-06, "loss": 0.0716, "step": 1970 }, { "epoch": 1.8714555765595464, "grad_norm": 1.8336230516433716, "learning_rate": 4.521316379955123e-06, "loss": 0.0804, "step": 1980 }, { "epoch": 1.8809073724007561, "grad_norm": 1.2159379720687866, "learning_rate": 4.4839192221391176e-06, "loss": 0.097, "step": 1990 }, { "epoch": 1.8903591682419658, "grad_norm": 4.437017440795898, "learning_rate": 4.446522064323112e-06, "loss": 0.0833, "step": 2000 }, { "epoch": 1.899810964083176, "grad_norm": 1.8714925050735474, "learning_rate": 4.409124906507105e-06, "loss": 0.0764, "step": 2010 }, { "epoch": 1.9092627599243857, "grad_norm": 1.631150722503662, "learning_rate": 4.3717277486910996e-06, "loss": 0.0827, "step": 2020 }, { "epoch": 1.9187145557655954, "grad_norm": 1.145902395248413, "learning_rate": 4.334330590875094e-06, "loss": 0.0575, "step": 2030 }, { "epoch": 1.9281663516068053, "grad_norm": 1.2010490894317627, "learning_rate": 4.296933433059088e-06, "loss": 0.086, "step": 2040 }, { "epoch": 1.9376181474480152, "grad_norm": 1.181751012802124, "learning_rate": 4.2595362752430815e-06, "loss": 0.085, "step": 2050 }, { "epoch": 1.947069943289225, "grad_norm": 1.8226746320724487, "learning_rate": 4.222139117427076e-06, "loss": 0.081, "step": 2060 }, { "epoch": 1.9565217391304348, "grad_norm": 2.128484010696411, "learning_rate": 4.18474195961107e-06, "loss": 0.0924, "step": 2070 }, { "epoch": 1.9659735349716447, "grad_norm": 3.3163959980010986, "learning_rate": 4.147344801795064e-06, "loss": 0.0878, "step": 2080 }, { "epoch": 1.9754253308128544, "grad_norm": 3.319842576980591, "learning_rate": 4.109947643979058e-06, "loss": 0.0771, "step": 2090 }, { "epoch": 1.9848771266540641, "grad_norm": 1.4262229204177856, "learning_rate": 4.072550486163052e-06, "loss": 0.0701, "step": 2100 }, { "epoch": 1.994328922495274, "grad_norm": 0.9300116896629333, "learning_rate": 4.035153328347046e-06, "loss": 0.071, "step": 2110 }, { "epoch": 2.003780718336484, "grad_norm": 3.044982671737671, "learning_rate": 3.99775617053104e-06, "loss": 0.093, "step": 2120 }, { "epoch": 2.0132325141776937, "grad_norm": 1.7930673360824585, "learning_rate": 3.960359012715034e-06, "loss": 0.0543, "step": 2130 }, { "epoch": 2.022684310018904, "grad_norm": 1.6980185508728027, "learning_rate": 3.9229618548990275e-06, "loss": 0.0611, "step": 2140 }, { "epoch": 2.0321361058601135, "grad_norm": 1.0368515253067017, "learning_rate": 3.885564697083022e-06, "loss": 0.047, "step": 2150 }, { "epoch": 2.041587901701323, "grad_norm": 3.497382640838623, "learning_rate": 3.848167539267016e-06, "loss": 0.0695, "step": 2160 }, { "epoch": 2.051039697542533, "grad_norm": 3.369415521621704, "learning_rate": 3.81077038145101e-06, "loss": 0.1011, "step": 2170 }, { "epoch": 2.060491493383743, "grad_norm": 1.3048640489578247, "learning_rate": 3.773373223635004e-06, "loss": 0.1194, "step": 2180 }, { "epoch": 2.0699432892249527, "grad_norm": 1.1383615732192993, "learning_rate": 3.735976065818998e-06, "loss": 0.0582, "step": 2190 }, { "epoch": 2.0793950850661624, "grad_norm": 1.8952641487121582, "learning_rate": 3.6985789080029923e-06, "loss": 0.0564, "step": 2200 }, { "epoch": 2.0888468809073726, "grad_norm": 0.8446295261383057, "learning_rate": 3.661181750186986e-06, "loss": 0.0501, "step": 2210 }, { "epoch": 2.0982986767485823, "grad_norm": 1.8624228239059448, "learning_rate": 3.62378459237098e-06, "loss": 0.0668, "step": 2220 }, { "epoch": 2.107750472589792, "grad_norm": 0.6371021866798401, "learning_rate": 3.5863874345549743e-06, "loss": 0.1011, "step": 2230 }, { "epoch": 2.1172022684310017, "grad_norm": 0.8320020437240601, "learning_rate": 3.548990276738968e-06, "loss": 0.0577, "step": 2240 }, { "epoch": 2.126654064272212, "grad_norm": 1.212850570678711, "learning_rate": 3.5115931189229625e-06, "loss": 0.0515, "step": 2250 }, { "epoch": 2.1361058601134215, "grad_norm": 1.5608317852020264, "learning_rate": 3.474195961106956e-06, "loss": 0.0549, "step": 2260 }, { "epoch": 2.145557655954631, "grad_norm": 1.6277782917022705, "learning_rate": 3.4367988032909497e-06, "loss": 0.0775, "step": 2270 }, { "epoch": 2.1550094517958414, "grad_norm": 0.8742659091949463, "learning_rate": 3.399401645474944e-06, "loss": 0.0702, "step": 2280 }, { "epoch": 2.164461247637051, "grad_norm": 2.2621870040893555, "learning_rate": 3.362004487658938e-06, "loss": 0.3185, "step": 2290 }, { "epoch": 2.1739130434782608, "grad_norm": 3.3026936054229736, "learning_rate": 3.324607329842932e-06, "loss": 0.0706, "step": 2300 }, { "epoch": 2.183364839319471, "grad_norm": 1.3657283782958984, "learning_rate": 3.287210172026926e-06, "loss": 0.074, "step": 2310 }, { "epoch": 2.1928166351606806, "grad_norm": 2.3479464054107666, "learning_rate": 3.2498130142109203e-06, "loss": 0.0859, "step": 2320 }, { "epoch": 2.2022684310018903, "grad_norm": 1.2391554117202759, "learning_rate": 3.212415856394914e-06, "loss": 0.0505, "step": 2330 }, { "epoch": 2.2117202268431, "grad_norm": 2.2902774810791016, "learning_rate": 3.1750186985789084e-06, "loss": 0.0659, "step": 2340 }, { "epoch": 2.22117202268431, "grad_norm": 1.3776533603668213, "learning_rate": 3.1376215407629023e-06, "loss": 0.0591, "step": 2350 }, { "epoch": 2.23062381852552, "grad_norm": 2.1550002098083496, "learning_rate": 3.1002243829468966e-06, "loss": 0.0797, "step": 2360 }, { "epoch": 2.2400756143667295, "grad_norm": 1.9426394701004028, "learning_rate": 3.0628272251308904e-06, "loss": 0.0986, "step": 2370 }, { "epoch": 2.2495274102079397, "grad_norm": 1.2983957529067993, "learning_rate": 3.0254300673148847e-06, "loss": 0.0454, "step": 2380 }, { "epoch": 2.2589792060491494, "grad_norm": 1.7152113914489746, "learning_rate": 2.9880329094988786e-06, "loss": 0.0508, "step": 2390 }, { "epoch": 2.268431001890359, "grad_norm": 0.9719337224960327, "learning_rate": 2.950635751682872e-06, "loss": 0.0581, "step": 2400 }, { "epoch": 2.2778827977315688, "grad_norm": 0.8925687670707703, "learning_rate": 2.9132385938668663e-06, "loss": 0.082, "step": 2410 }, { "epoch": 2.287334593572779, "grad_norm": 4.842050075531006, "learning_rate": 2.87584143605086e-06, "loss": 0.0731, "step": 2420 }, { "epoch": 2.2967863894139886, "grad_norm": 1.5382630825042725, "learning_rate": 2.8384442782348544e-06, "loss": 0.0595, "step": 2430 }, { "epoch": 2.3062381852551983, "grad_norm": 0.8771998286247253, "learning_rate": 2.8010471204188483e-06, "loss": 0.0584, "step": 2440 }, { "epoch": 2.3156899810964084, "grad_norm": 1.2943544387817383, "learning_rate": 2.7636499626028425e-06, "loss": 0.0674, "step": 2450 }, { "epoch": 2.325141776937618, "grad_norm": 2.6974918842315674, "learning_rate": 2.7262528047868364e-06, "loss": 0.0606, "step": 2460 }, { "epoch": 2.334593572778828, "grad_norm": 1.4208108186721802, "learning_rate": 2.6888556469708307e-06, "loss": 0.0731, "step": 2470 }, { "epoch": 2.344045368620038, "grad_norm": 1.310441493988037, "learning_rate": 2.6514584891548245e-06, "loss": 0.1353, "step": 2480 }, { "epoch": 2.3534971644612477, "grad_norm": 1.5753083229064941, "learning_rate": 2.6140613313388184e-06, "loss": 0.0648, "step": 2490 }, { "epoch": 2.3629489603024574, "grad_norm": 0.9591426253318787, "learning_rate": 2.5766641735228127e-06, "loss": 0.042, "step": 2500 }, { "epoch": 2.3724007561436675, "grad_norm": 1.4900124073028564, "learning_rate": 2.5392670157068065e-06, "loss": 0.495, "step": 2510 }, { "epoch": 2.381852551984877, "grad_norm": 1.3494096994400024, "learning_rate": 2.5018698578908008e-06, "loss": 0.0621, "step": 2520 }, { "epoch": 2.391304347826087, "grad_norm": 3.060147762298584, "learning_rate": 2.4644727000747946e-06, "loss": 0.0506, "step": 2530 }, { "epoch": 2.4007561436672966, "grad_norm": 4.2206292152404785, "learning_rate": 2.4270755422587885e-06, "loss": 0.0627, "step": 2540 }, { "epoch": 2.4102079395085068, "grad_norm": 0.9725443720817566, "learning_rate": 2.3896783844427828e-06, "loss": 0.0461, "step": 2550 }, { "epoch": 2.4196597353497165, "grad_norm": 3.230731964111328, "learning_rate": 2.3522812266267766e-06, "loss": 0.0399, "step": 2560 }, { "epoch": 2.429111531190926, "grad_norm": 2.5825319290161133, "learning_rate": 2.3148840688107705e-06, "loss": 0.0595, "step": 2570 }, { "epoch": 2.4385633270321363, "grad_norm": 1.2526172399520874, "learning_rate": 2.2774869109947643e-06, "loss": 0.0526, "step": 2580 }, { "epoch": 2.448015122873346, "grad_norm": 2.662144184112549, "learning_rate": 2.2400897531787586e-06, "loss": 0.0611, "step": 2590 }, { "epoch": 2.4574669187145557, "grad_norm": 1.0309128761291504, "learning_rate": 2.2026925953627525e-06, "loss": 0.0768, "step": 2600 }, { "epoch": 2.4669187145557654, "grad_norm": 1.1861653327941895, "learning_rate": 2.1652954375467468e-06, "loss": 0.0611, "step": 2610 }, { "epoch": 2.4763705103969755, "grad_norm": 2.0343453884124756, "learning_rate": 2.1278982797307406e-06, "loss": 0.0637, "step": 2620 }, { "epoch": 2.4858223062381852, "grad_norm": 1.182011365890503, "learning_rate": 2.090501121914735e-06, "loss": 0.0451, "step": 2630 }, { "epoch": 2.495274102079395, "grad_norm": 0.9003073573112488, "learning_rate": 2.0531039640987283e-06, "loss": 0.0526, "step": 2640 }, { "epoch": 2.504725897920605, "grad_norm": 2.235811471939087, "learning_rate": 2.0157068062827226e-06, "loss": 0.0753, "step": 2650 }, { "epoch": 2.5141776937618148, "grad_norm": 5.556035995483398, "learning_rate": 1.9783096484667165e-06, "loss": 0.1743, "step": 2660 }, { "epoch": 2.5236294896030245, "grad_norm": 1.437063455581665, "learning_rate": 1.9409124906507107e-06, "loss": 0.0511, "step": 2670 }, { "epoch": 2.5330812854442346, "grad_norm": 0.8690747022628784, "learning_rate": 1.9035153328347048e-06, "loss": 0.0634, "step": 2680 }, { "epoch": 2.5425330812854443, "grad_norm": 1.155195713043213, "learning_rate": 1.8661181750186989e-06, "loss": 0.0545, "step": 2690 }, { "epoch": 2.551984877126654, "grad_norm": 0.990226686000824, "learning_rate": 1.828721017202693e-06, "loss": 0.0704, "step": 2700 }, { "epoch": 2.561436672967864, "grad_norm": 1.3275784254074097, "learning_rate": 1.7913238593866866e-06, "loss": 0.048, "step": 2710 }, { "epoch": 2.570888468809074, "grad_norm": 0.962841808795929, "learning_rate": 1.7539267015706806e-06, "loss": 0.0469, "step": 2720 }, { "epoch": 2.5803402646502835, "grad_norm": 3.382091522216797, "learning_rate": 1.7165295437546747e-06, "loss": 0.0715, "step": 2730 }, { "epoch": 2.5897920604914932, "grad_norm": 1.1871252059936523, "learning_rate": 1.6791323859386688e-06, "loss": 0.0667, "step": 2740 }, { "epoch": 2.599243856332703, "grad_norm": 2.9719226360321045, "learning_rate": 1.6417352281226628e-06, "loss": 0.0602, "step": 2750 }, { "epoch": 2.608695652173913, "grad_norm": 1.1596277952194214, "learning_rate": 1.604338070306657e-06, "loss": 0.061, "step": 2760 }, { "epoch": 2.618147448015123, "grad_norm": 1.5264619588851929, "learning_rate": 1.566940912490651e-06, "loss": 0.055, "step": 2770 }, { "epoch": 2.6275992438563325, "grad_norm": 1.292921781539917, "learning_rate": 1.5295437546746446e-06, "loss": 0.0637, "step": 2780 }, { "epoch": 2.6370510396975426, "grad_norm": 4.510688781738281, "learning_rate": 1.4921465968586387e-06, "loss": 0.0734, "step": 2790 }, { "epoch": 2.6465028355387523, "grad_norm": 0.9487069249153137, "learning_rate": 1.4547494390426328e-06, "loss": 0.0467, "step": 2800 }, { "epoch": 2.655954631379962, "grad_norm": 1.8110833168029785, "learning_rate": 1.4173522812266268e-06, "loss": 0.0479, "step": 2810 }, { "epoch": 2.665406427221172, "grad_norm": 1.7220536470413208, "learning_rate": 1.3799551234106209e-06, "loss": 0.0665, "step": 2820 }, { "epoch": 2.674858223062382, "grad_norm": 2.3197669982910156, "learning_rate": 1.342557965594615e-06, "loss": 0.0559, "step": 2830 }, { "epoch": 2.6843100189035916, "grad_norm": 1.2168952226638794, "learning_rate": 1.305160807778609e-06, "loss": 0.0597, "step": 2840 }, { "epoch": 2.6937618147448017, "grad_norm": 2.3563144207000732, "learning_rate": 1.2677636499626029e-06, "loss": 0.0698, "step": 2850 }, { "epoch": 2.7032136105860114, "grad_norm": 1.7858084440231323, "learning_rate": 1.230366492146597e-06, "loss": 0.0493, "step": 2860 }, { "epoch": 2.712665406427221, "grad_norm": 0.999500036239624, "learning_rate": 1.192969334330591e-06, "loss": 0.0555, "step": 2870 }, { "epoch": 2.7221172022684312, "grad_norm": 2.3788414001464844, "learning_rate": 1.1555721765145849e-06, "loss": 0.0466, "step": 2880 }, { "epoch": 2.731568998109641, "grad_norm": 1.2717732191085815, "learning_rate": 1.118175018698579e-06, "loss": 0.0573, "step": 2890 }, { "epoch": 2.7410207939508506, "grad_norm": 1.2164254188537598, "learning_rate": 1.080777860882573e-06, "loss": 0.0613, "step": 2900 }, { "epoch": 2.7504725897920603, "grad_norm": 1.483476996421814, "learning_rate": 1.043380703066567e-06, "loss": 0.0621, "step": 2910 }, { "epoch": 2.7599243856332705, "grad_norm": 0.919227659702301, "learning_rate": 1.0059835452505611e-06, "loss": 0.0551, "step": 2920 }, { "epoch": 2.76937618147448, "grad_norm": 1.3688384294509888, "learning_rate": 9.685863874345552e-07, "loss": 0.0457, "step": 2930 }, { "epoch": 2.77882797731569, "grad_norm": 2.6967413425445557, "learning_rate": 9.311892296185491e-07, "loss": 0.0522, "step": 2940 }, { "epoch": 2.7882797731568996, "grad_norm": 1.768357276916504, "learning_rate": 8.93792071802543e-07, "loss": 0.0507, "step": 2950 }, { "epoch": 2.7977315689981097, "grad_norm": 1.3043689727783203, "learning_rate": 8.563949139865371e-07, "loss": 0.0496, "step": 2960 }, { "epoch": 2.8071833648393194, "grad_norm": 1.3980363607406616, "learning_rate": 8.189977561705311e-07, "loss": 0.06, "step": 2970 }, { "epoch": 2.816635160680529, "grad_norm": 2.036473512649536, "learning_rate": 7.816005983545251e-07, "loss": 0.0519, "step": 2980 }, { "epoch": 2.8260869565217392, "grad_norm": 1.1731027364730835, "learning_rate": 7.442034405385192e-07, "loss": 0.0537, "step": 2990 }, { "epoch": 2.835538752362949, "grad_norm": 1.2910258769989014, "learning_rate": 7.068062827225131e-07, "loss": 0.0439, "step": 3000 }, { "epoch": 2.8449905482041586, "grad_norm": 0.6368816494941711, "learning_rate": 6.694091249065071e-07, "loss": 0.0541, "step": 3010 }, { "epoch": 2.854442344045369, "grad_norm": 1.0319303274154663, "learning_rate": 6.320119670905012e-07, "loss": 0.0337, "step": 3020 }, { "epoch": 2.8638941398865785, "grad_norm": 1.931760311126709, "learning_rate": 5.946148092744952e-07, "loss": 0.0555, "step": 3030 }, { "epoch": 2.873345935727788, "grad_norm": 1.5144755840301514, "learning_rate": 5.572176514584892e-07, "loss": 0.0555, "step": 3040 }, { "epoch": 2.8827977315689983, "grad_norm": 9.888813972473145, "learning_rate": 5.198204936424832e-07, "loss": 0.0606, "step": 3050 }, { "epoch": 2.892249527410208, "grad_norm": 10.548643112182617, "learning_rate": 4.824233358264772e-07, "loss": 0.064, "step": 3060 }, { "epoch": 2.9017013232514177, "grad_norm": 1.567958950996399, "learning_rate": 4.4502617801047125e-07, "loss": 0.0644, "step": 3070 }, { "epoch": 2.911153119092628, "grad_norm": 2.924102783203125, "learning_rate": 4.0762902019446526e-07, "loss": 0.0584, "step": 3080 }, { "epoch": 2.9206049149338376, "grad_norm": 2.3769917488098145, "learning_rate": 3.702318623784593e-07, "loss": 0.054, "step": 3090 }, { "epoch": 2.9300567107750473, "grad_norm": 2.570713996887207, "learning_rate": 3.328347045624533e-07, "loss": 0.0709, "step": 3100 }, { "epoch": 2.939508506616257, "grad_norm": 1.1489148139953613, "learning_rate": 2.954375467464473e-07, "loss": 0.0605, "step": 3110 }, { "epoch": 2.9489603024574667, "grad_norm": 1.264321208000183, "learning_rate": 2.580403889304413e-07, "loss": 0.0684, "step": 3120 }, { "epoch": 2.958412098298677, "grad_norm": 1.6857093572616577, "learning_rate": 2.2064323111443533e-07, "loss": 0.0579, "step": 3130 }, { "epoch": 2.9678638941398865, "grad_norm": 2.269068717956543, "learning_rate": 1.8324607329842932e-07, "loss": 0.0606, "step": 3140 }, { "epoch": 2.977315689981096, "grad_norm": 1.0288152694702148, "learning_rate": 1.4584891548242336e-07, "loss": 0.0577, "step": 3150 }, { "epoch": 2.9867674858223063, "grad_norm": 0.940092146396637, "learning_rate": 1.0845175766641736e-07, "loss": 0.0412, "step": 3160 }, { "epoch": 2.996219281663516, "grad_norm": 1.5147079229354858, "learning_rate": 7.105459985041138e-08, "loss": 0.0586, "step": 3170 } ], "logging_steps": 10, "max_steps": 3174, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 5000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 7.245637388284723e+16, "train_batch_size": 1, "trial_name": null, "trial_params": null }