{ "best_metric": null, "best_model_checkpoint": null, "epoch": 5.0, "eval_steps": 500, "global_step": 21660, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0023084025854108957, "grad_norm": 4.518136024475098, "learning_rate": 9.99538319482918e-06, "loss": 0.0901, "step": 10 }, { "epoch": 0.0046168051708217915, "grad_norm": 0.7349718809127808, "learning_rate": 9.990766389658357e-06, "loss": 0.3089, "step": 20 }, { "epoch": 0.006925207756232687, "grad_norm": 9.184749603271484, "learning_rate": 9.986149584487536e-06, "loss": 0.2896, "step": 30 }, { "epoch": 0.009233610341643583, "grad_norm": 10.729694366455078, "learning_rate": 9.981532779316714e-06, "loss": 0.2194, "step": 40 }, { "epoch": 0.011542012927054479, "grad_norm": 6.5318989753723145, "learning_rate": 9.976915974145893e-06, "loss": 0.2622, "step": 50 }, { "epoch": 0.013850415512465374, "grad_norm": 1.6825273036956787, "learning_rate": 9.97229916897507e-06, "loss": 0.0971, "step": 60 }, { "epoch": 0.016158818097876268, "grad_norm": 13.866430282592773, "learning_rate": 9.96768236380425e-06, "loss": 0.1541, "step": 70 }, { "epoch": 0.018467220683287166, "grad_norm": 0.09123529493808746, "learning_rate": 9.963065558633427e-06, "loss": 0.2593, "step": 80 }, { "epoch": 0.02077562326869806, "grad_norm": 8.347816467285156, "learning_rate": 9.958448753462606e-06, "loss": 0.1499, "step": 90 }, { "epoch": 0.023084025854108958, "grad_norm": 7.898433685302734, "learning_rate": 9.953831948291783e-06, "loss": 0.137, "step": 100 }, { "epoch": 0.025392428439519853, "grad_norm": 12.087220191955566, "learning_rate": 9.949215143120962e-06, "loss": 0.1205, "step": 110 }, { "epoch": 0.027700831024930747, "grad_norm": 7.061990737915039, "learning_rate": 9.94459833795014e-06, "loss": 0.2906, "step": 120 }, { "epoch": 0.030009233610341645, "grad_norm": 6.957597732543945, "learning_rate": 9.939981532779317e-06, "loss": 0.1156, "step": 130 }, { "epoch": 0.032317636195752536, "grad_norm": 4.130475997924805, "learning_rate": 9.935364727608496e-06, "loss": 0.2686, "step": 140 }, { "epoch": 0.03462603878116344, "grad_norm": 4.95039701461792, "learning_rate": 9.930747922437673e-06, "loss": 0.049, "step": 150 }, { "epoch": 0.03693444136657433, "grad_norm": 0.016249021515250206, "learning_rate": 9.926131117266852e-06, "loss": 0.2209, "step": 160 }, { "epoch": 0.039242843951985226, "grad_norm": 5.67888069152832, "learning_rate": 9.92151431209603e-06, "loss": 0.0999, "step": 170 }, { "epoch": 0.04155124653739612, "grad_norm": 4.460853099822998, "learning_rate": 9.916897506925209e-06, "loss": 0.1367, "step": 180 }, { "epoch": 0.043859649122807015, "grad_norm": 2.987170696258545, "learning_rate": 9.912280701754386e-06, "loss": 0.1781, "step": 190 }, { "epoch": 0.046168051708217916, "grad_norm": 1.5171048641204834, "learning_rate": 9.907663896583565e-06, "loss": 0.1776, "step": 200 }, { "epoch": 0.04847645429362881, "grad_norm": 15.294007301330566, "learning_rate": 9.903047091412743e-06, "loss": 0.2982, "step": 210 }, { "epoch": 0.050784856879039705, "grad_norm": 0.3504500985145569, "learning_rate": 9.898430286241922e-06, "loss": 0.0849, "step": 220 }, { "epoch": 0.0530932594644506, "grad_norm": 2.0193207263946533, "learning_rate": 9.8938134810711e-06, "loss": 0.2644, "step": 230 }, { "epoch": 0.055401662049861494, "grad_norm": 13.084386825561523, "learning_rate": 9.889196675900278e-06, "loss": 0.1378, "step": 240 }, { "epoch": 0.05771006463527239, "grad_norm": 6.731300354003906, "learning_rate": 9.884579870729456e-06, "loss": 0.3476, "step": 250 }, { "epoch": 0.06001846722068329, "grad_norm": 15.974516868591309, "learning_rate": 9.879963065558635e-06, "loss": 0.1693, "step": 260 }, { "epoch": 0.062326869806094184, "grad_norm": 5.969636917114258, "learning_rate": 9.875346260387812e-06, "loss": 0.1751, "step": 270 }, { "epoch": 0.06463527239150507, "grad_norm": 3.4812111854553223, "learning_rate": 9.870729455216991e-06, "loss": 0.139, "step": 280 }, { "epoch": 0.06694367497691597, "grad_norm": 2.747326612472534, "learning_rate": 9.866112650046169e-06, "loss": 0.2624, "step": 290 }, { "epoch": 0.06925207756232687, "grad_norm": 4.280517578125, "learning_rate": 9.861495844875348e-06, "loss": 0.1476, "step": 300 }, { "epoch": 0.07156048014773776, "grad_norm": 1.3420567512512207, "learning_rate": 9.856879039704525e-06, "loss": 0.2018, "step": 310 }, { "epoch": 0.07386888273314866, "grad_norm": 9.673868179321289, "learning_rate": 9.852262234533704e-06, "loss": 0.1798, "step": 320 }, { "epoch": 0.07617728531855955, "grad_norm": 3.046891927719116, "learning_rate": 9.847645429362882e-06, "loss": 0.2243, "step": 330 }, { "epoch": 0.07848568790397045, "grad_norm": 0.40016505122184753, "learning_rate": 9.84302862419206e-06, "loss": 0.1231, "step": 340 }, { "epoch": 0.08079409048938135, "grad_norm": 4.051836967468262, "learning_rate": 9.838411819021238e-06, "loss": 0.2221, "step": 350 }, { "epoch": 0.08310249307479224, "grad_norm": 1.4235812425613403, "learning_rate": 9.833795013850417e-06, "loss": 0.169, "step": 360 }, { "epoch": 0.08541089566020314, "grad_norm": 1.3389965295791626, "learning_rate": 9.829178208679594e-06, "loss": 0.202, "step": 370 }, { "epoch": 0.08771929824561403, "grad_norm": 6.450026035308838, "learning_rate": 9.824561403508772e-06, "loss": 0.0796, "step": 380 }, { "epoch": 0.09002770083102493, "grad_norm": 3.4097864627838135, "learning_rate": 9.819944598337951e-06, "loss": 0.1239, "step": 390 }, { "epoch": 0.09233610341643583, "grad_norm": 3.103640079498291, "learning_rate": 9.815327793167128e-06, "loss": 0.0573, "step": 400 }, { "epoch": 0.09464450600184672, "grad_norm": 0.8760634064674377, "learning_rate": 9.810710987996307e-06, "loss": 0.2037, "step": 410 }, { "epoch": 0.09695290858725762, "grad_norm": 1.9501415491104126, "learning_rate": 9.806094182825485e-06, "loss": 0.2325, "step": 420 }, { "epoch": 0.09926131117266851, "grad_norm": 0.14746245741844177, "learning_rate": 9.801477377654664e-06, "loss": 0.0433, "step": 430 }, { "epoch": 0.10156971375807941, "grad_norm": 10.241828918457031, "learning_rate": 9.796860572483841e-06, "loss": 0.2351, "step": 440 }, { "epoch": 0.1038781163434903, "grad_norm": 1.1289986371994019, "learning_rate": 9.79224376731302e-06, "loss": 0.062, "step": 450 }, { "epoch": 0.1061865189289012, "grad_norm": 10.396888732910156, "learning_rate": 9.787626962142198e-06, "loss": 0.2071, "step": 460 }, { "epoch": 0.1084949215143121, "grad_norm": 1.2174731492996216, "learning_rate": 9.783010156971377e-06, "loss": 0.0761, "step": 470 }, { "epoch": 0.11080332409972299, "grad_norm": 6.478285312652588, "learning_rate": 9.778393351800554e-06, "loss": 0.0876, "step": 480 }, { "epoch": 0.11311172668513389, "grad_norm": 6.522837162017822, "learning_rate": 9.773776546629733e-06, "loss": 0.3174, "step": 490 }, { "epoch": 0.11542012927054478, "grad_norm": 4.25970983505249, "learning_rate": 9.76915974145891e-06, "loss": 0.1484, "step": 500 }, { "epoch": 0.11772853185595568, "grad_norm": 6.3719024658203125, "learning_rate": 9.76454293628809e-06, "loss": 0.0913, "step": 510 }, { "epoch": 0.12003693444136658, "grad_norm": 6.4946441650390625, "learning_rate": 9.759926131117267e-06, "loss": 0.2221, "step": 520 }, { "epoch": 0.12234533702677747, "grad_norm": 1.0337347984313965, "learning_rate": 9.755309325946446e-06, "loss": 0.1504, "step": 530 }, { "epoch": 0.12465373961218837, "grad_norm": 6.716022491455078, "learning_rate": 9.750692520775623e-06, "loss": 0.3298, "step": 540 }, { "epoch": 0.12696214219759927, "grad_norm": 0.982958972454071, "learning_rate": 9.746075715604803e-06, "loss": 0.1637, "step": 550 }, { "epoch": 0.12927054478301014, "grad_norm": 2.2563958168029785, "learning_rate": 9.74145891043398e-06, "loss": 0.1235, "step": 560 }, { "epoch": 0.13157894736842105, "grad_norm": 2.639207363128662, "learning_rate": 9.736842105263159e-06, "loss": 0.1916, "step": 570 }, { "epoch": 0.13388734995383195, "grad_norm": 9.504934310913086, "learning_rate": 9.732225300092338e-06, "loss": 0.1779, "step": 580 }, { "epoch": 0.13619575253924285, "grad_norm": 8.144251823425293, "learning_rate": 9.727608494921515e-06, "loss": 0.1563, "step": 590 }, { "epoch": 0.13850415512465375, "grad_norm": 0.12685340642929077, "learning_rate": 9.722991689750695e-06, "loss": 0.2688, "step": 600 }, { "epoch": 0.14081255771006462, "grad_norm": 2.690690279006958, "learning_rate": 9.718374884579872e-06, "loss": 0.1462, "step": 610 }, { "epoch": 0.14312096029547552, "grad_norm": 4.394099235534668, "learning_rate": 9.713758079409051e-06, "loss": 0.3113, "step": 620 }, { "epoch": 0.14542936288088643, "grad_norm": 0.9470362067222595, "learning_rate": 9.709141274238227e-06, "loss": 0.2729, "step": 630 }, { "epoch": 0.14773776546629733, "grad_norm": 1.6863257884979248, "learning_rate": 9.704524469067406e-06, "loss": 0.2146, "step": 640 }, { "epoch": 0.15004616805170823, "grad_norm": 9.275301933288574, "learning_rate": 9.699907663896583e-06, "loss": 0.2194, "step": 650 }, { "epoch": 0.1523545706371191, "grad_norm": 10.41015338897705, "learning_rate": 9.695290858725762e-06, "loss": 0.2853, "step": 660 }, { "epoch": 0.15466297322253, "grad_norm": 3.2390172481536865, "learning_rate": 9.69067405355494e-06, "loss": 0.0957, "step": 670 }, { "epoch": 0.1569713758079409, "grad_norm": 6.6949849128723145, "learning_rate": 9.686057248384119e-06, "loss": 0.1774, "step": 680 }, { "epoch": 0.1592797783933518, "grad_norm": 5.19643497467041, "learning_rate": 9.681440443213298e-06, "loss": 0.1679, "step": 690 }, { "epoch": 0.1615881809787627, "grad_norm": 0.33260342478752136, "learning_rate": 9.676823638042475e-06, "loss": 0.0561, "step": 700 }, { "epoch": 0.16389658356417358, "grad_norm": 11.740571022033691, "learning_rate": 9.672206832871654e-06, "loss": 0.1117, "step": 710 }, { "epoch": 0.16620498614958448, "grad_norm": 3.4466536045074463, "learning_rate": 9.667590027700832e-06, "loss": 0.2143, "step": 720 }, { "epoch": 0.16851338873499538, "grad_norm": 6.085760593414307, "learning_rate": 9.66297322253001e-06, "loss": 0.094, "step": 730 }, { "epoch": 0.17082179132040629, "grad_norm": 4.806663513183594, "learning_rate": 9.658356417359188e-06, "loss": 0.1469, "step": 740 }, { "epoch": 0.1731301939058172, "grad_norm": 9.830680847167969, "learning_rate": 9.653739612188367e-06, "loss": 0.3836, "step": 750 }, { "epoch": 0.17543859649122806, "grad_norm": 6.321448802947998, "learning_rate": 9.649122807017545e-06, "loss": 0.3127, "step": 760 }, { "epoch": 0.17774699907663896, "grad_norm": 0.8883340954780579, "learning_rate": 9.644506001846724e-06, "loss": 0.2311, "step": 770 }, { "epoch": 0.18005540166204986, "grad_norm": 0.36167535185813904, "learning_rate": 9.639889196675901e-06, "loss": 0.1058, "step": 780 }, { "epoch": 0.18236380424746076, "grad_norm": 7.073151588439941, "learning_rate": 9.63527239150508e-06, "loss": 0.1334, "step": 790 }, { "epoch": 0.18467220683287167, "grad_norm": 2.5660276412963867, "learning_rate": 9.630655586334257e-06, "loss": 0.1754, "step": 800 }, { "epoch": 0.18698060941828254, "grad_norm": 0.5900049209594727, "learning_rate": 9.626038781163436e-06, "loss": 0.0877, "step": 810 }, { "epoch": 0.18928901200369344, "grad_norm": 2.075150728225708, "learning_rate": 9.621421975992614e-06, "loss": 0.1663, "step": 820 }, { "epoch": 0.19159741458910434, "grad_norm": 6.630813121795654, "learning_rate": 9.616805170821793e-06, "loss": 0.1554, "step": 830 }, { "epoch": 0.19390581717451524, "grad_norm": 4.084090709686279, "learning_rate": 9.61218836565097e-06, "loss": 0.1748, "step": 840 }, { "epoch": 0.19621421975992612, "grad_norm": 1.9874173402786255, "learning_rate": 9.60757156048015e-06, "loss": 0.2409, "step": 850 }, { "epoch": 0.19852262234533702, "grad_norm": 5.994325637817383, "learning_rate": 9.602954755309327e-06, "loss": 0.1676, "step": 860 }, { "epoch": 0.20083102493074792, "grad_norm": 7.272033214569092, "learning_rate": 9.598337950138506e-06, "loss": 0.1236, "step": 870 }, { "epoch": 0.20313942751615882, "grad_norm": 0.7912845611572266, "learning_rate": 9.593721144967683e-06, "loss": 0.0946, "step": 880 }, { "epoch": 0.20544783010156972, "grad_norm": 1.7485767602920532, "learning_rate": 9.58910433979686e-06, "loss": 0.1679, "step": 890 }, { "epoch": 0.2077562326869806, "grad_norm": 2.479883909225464, "learning_rate": 9.58448753462604e-06, "loss": 0.1502, "step": 900 }, { "epoch": 0.2100646352723915, "grad_norm": 5.910335063934326, "learning_rate": 9.579870729455217e-06, "loss": 0.082, "step": 910 }, { "epoch": 0.2123730378578024, "grad_norm": 1.2047022581100464, "learning_rate": 9.575253924284396e-06, "loss": 0.1838, "step": 920 }, { "epoch": 0.2146814404432133, "grad_norm": 3.5549404621124268, "learning_rate": 9.570637119113574e-06, "loss": 0.1784, "step": 930 }, { "epoch": 0.2169898430286242, "grad_norm": 4.833847522735596, "learning_rate": 9.566020313942753e-06, "loss": 0.1015, "step": 940 }, { "epoch": 0.21929824561403508, "grad_norm": 5.299631118774414, "learning_rate": 9.56140350877193e-06, "loss": 0.0985, "step": 950 }, { "epoch": 0.22160664819944598, "grad_norm": 10.254817962646484, "learning_rate": 9.556786703601109e-06, "loss": 0.494, "step": 960 }, { "epoch": 0.22391505078485688, "grad_norm": 5.614777088165283, "learning_rate": 9.552169898430286e-06, "loss": 0.1996, "step": 970 }, { "epoch": 0.22622345337026778, "grad_norm": 4.300576210021973, "learning_rate": 9.547553093259466e-06, "loss": 0.1987, "step": 980 }, { "epoch": 0.22853185595567868, "grad_norm": 10.567139625549316, "learning_rate": 9.542936288088643e-06, "loss": 0.2799, "step": 990 }, { "epoch": 0.23084025854108955, "grad_norm": 9.390183448791504, "learning_rate": 9.538319482917822e-06, "loss": 0.1817, "step": 1000 }, { "epoch": 0.23314866112650046, "grad_norm": 7.043652057647705, "learning_rate": 9.533702677747e-06, "loss": 0.1833, "step": 1010 }, { "epoch": 0.23545706371191136, "grad_norm": 5.536339282989502, "learning_rate": 9.529085872576178e-06, "loss": 0.1897, "step": 1020 }, { "epoch": 0.23776546629732226, "grad_norm": 8.518433570861816, "learning_rate": 9.524469067405356e-06, "loss": 0.3156, "step": 1030 }, { "epoch": 0.24007386888273316, "grad_norm": 11.972070693969727, "learning_rate": 9.519852262234535e-06, "loss": 0.134, "step": 1040 }, { "epoch": 0.24238227146814403, "grad_norm": 7.265350818634033, "learning_rate": 9.515235457063712e-06, "loss": 0.3054, "step": 1050 }, { "epoch": 0.24469067405355494, "grad_norm": 10.719847679138184, "learning_rate": 9.510618651892891e-06, "loss": 0.2455, "step": 1060 }, { "epoch": 0.24699907663896584, "grad_norm": 9.438804626464844, "learning_rate": 9.506001846722069e-06, "loss": 0.2328, "step": 1070 }, { "epoch": 0.24930747922437674, "grad_norm": 4.239076614379883, "learning_rate": 9.501385041551248e-06, "loss": 0.1635, "step": 1080 }, { "epoch": 0.2516158818097876, "grad_norm": 5.953492641448975, "learning_rate": 9.496768236380425e-06, "loss": 0.1399, "step": 1090 }, { "epoch": 0.25392428439519854, "grad_norm": 1.620262861251831, "learning_rate": 9.492151431209604e-06, "loss": 0.1419, "step": 1100 }, { "epoch": 0.2562326869806094, "grad_norm": 1.141279935836792, "learning_rate": 9.487534626038782e-06, "loss": 0.1209, "step": 1110 }, { "epoch": 0.2585410895660203, "grad_norm": 0.8355013132095337, "learning_rate": 9.48291782086796e-06, "loss": 0.0818, "step": 1120 }, { "epoch": 0.2608494921514312, "grad_norm": 0.5204324722290039, "learning_rate": 9.478301015697138e-06, "loss": 0.2279, "step": 1130 }, { "epoch": 0.2631578947368421, "grad_norm": 9.81224250793457, "learning_rate": 9.473684210526315e-06, "loss": 0.1757, "step": 1140 }, { "epoch": 0.265466297322253, "grad_norm": 9.512256622314453, "learning_rate": 9.469067405355495e-06, "loss": 0.2873, "step": 1150 }, { "epoch": 0.2677746999076639, "grad_norm": 5.788575649261475, "learning_rate": 9.464450600184672e-06, "loss": 0.1019, "step": 1160 }, { "epoch": 0.27008310249307477, "grad_norm": 0.005946075078099966, "learning_rate": 9.459833795013851e-06, "loss": 0.1408, "step": 1170 }, { "epoch": 0.2723915050784857, "grad_norm": 5.019355773925781, "learning_rate": 9.455216989843028e-06, "loss": 0.2246, "step": 1180 }, { "epoch": 0.27469990766389657, "grad_norm": 7.404417037963867, "learning_rate": 9.450600184672207e-06, "loss": 0.0867, "step": 1190 }, { "epoch": 0.2770083102493075, "grad_norm": 0.025951171293854713, "learning_rate": 9.445983379501385e-06, "loss": 0.2332, "step": 1200 }, { "epoch": 0.2793167128347184, "grad_norm": 0.6903806924819946, "learning_rate": 9.441366574330564e-06, "loss": 0.146, "step": 1210 }, { "epoch": 0.28162511542012925, "grad_norm": 0.4754585921764374, "learning_rate": 9.436749769159741e-06, "loss": 0.1483, "step": 1220 }, { "epoch": 0.2839335180055402, "grad_norm": 4.7122483253479, "learning_rate": 9.43213296398892e-06, "loss": 0.2111, "step": 1230 }, { "epoch": 0.28624192059095105, "grad_norm": 4.9761271476745605, "learning_rate": 9.427516158818098e-06, "loss": 0.2371, "step": 1240 }, { "epoch": 0.288550323176362, "grad_norm": 9.827065467834473, "learning_rate": 9.422899353647277e-06, "loss": 0.3087, "step": 1250 }, { "epoch": 0.29085872576177285, "grad_norm": 0.41259774565696716, "learning_rate": 9.418282548476454e-06, "loss": 0.1081, "step": 1260 }, { "epoch": 0.2931671283471837, "grad_norm": 1.8536396026611328, "learning_rate": 9.413665743305633e-06, "loss": 0.2183, "step": 1270 }, { "epoch": 0.29547553093259465, "grad_norm": 3.651834726333618, "learning_rate": 9.409048938134812e-06, "loss": 0.1272, "step": 1280 }, { "epoch": 0.29778393351800553, "grad_norm": 2.255277156829834, "learning_rate": 9.40443213296399e-06, "loss": 0.3248, "step": 1290 }, { "epoch": 0.30009233610341646, "grad_norm": 0.9870880842208862, "learning_rate": 9.399815327793169e-06, "loss": 0.1309, "step": 1300 }, { "epoch": 0.30240073868882733, "grad_norm": 8.100092887878418, "learning_rate": 9.395198522622346e-06, "loss": 0.1945, "step": 1310 }, { "epoch": 0.3047091412742382, "grad_norm": 6.449978828430176, "learning_rate": 9.390581717451525e-06, "loss": 0.2893, "step": 1320 }, { "epoch": 0.30701754385964913, "grad_norm": 8.208856582641602, "learning_rate": 9.385964912280703e-06, "loss": 0.3231, "step": 1330 }, { "epoch": 0.30932594644506, "grad_norm": 7.377975940704346, "learning_rate": 9.381348107109882e-06, "loss": 0.2315, "step": 1340 }, { "epoch": 0.31163434903047094, "grad_norm": 8.44047737121582, "learning_rate": 9.376731301939059e-06, "loss": 0.2417, "step": 1350 }, { "epoch": 0.3139427516158818, "grad_norm": 1.1026626825332642, "learning_rate": 9.372114496768238e-06, "loss": 0.146, "step": 1360 }, { "epoch": 0.3162511542012927, "grad_norm": 6.599093914031982, "learning_rate": 9.367497691597416e-06, "loss": 0.2079, "step": 1370 }, { "epoch": 0.3185595567867036, "grad_norm": 6.624666213989258, "learning_rate": 9.362880886426595e-06, "loss": 0.2617, "step": 1380 }, { "epoch": 0.3208679593721145, "grad_norm": 0.857266902923584, "learning_rate": 9.358264081255772e-06, "loss": 0.0697, "step": 1390 }, { "epoch": 0.3231763619575254, "grad_norm": 2.5935277938842773, "learning_rate": 9.35364727608495e-06, "loss": 0.2199, "step": 1400 }, { "epoch": 0.3254847645429363, "grad_norm": 7.3394455909729, "learning_rate": 9.349030470914129e-06, "loss": 0.2001, "step": 1410 }, { "epoch": 0.32779316712834716, "grad_norm": 3.1454150676727295, "learning_rate": 9.344413665743306e-06, "loss": 0.1214, "step": 1420 }, { "epoch": 0.3301015697137581, "grad_norm": 7.05164909362793, "learning_rate": 9.339796860572485e-06, "loss": 0.1382, "step": 1430 }, { "epoch": 0.33240997229916897, "grad_norm": 8.544759750366211, "learning_rate": 9.335180055401662e-06, "loss": 0.202, "step": 1440 }, { "epoch": 0.3347183748845799, "grad_norm": 1.9893354177474976, "learning_rate": 9.330563250230841e-06, "loss": 0.2795, "step": 1450 }, { "epoch": 0.33702677746999077, "grad_norm": 5.580140113830566, "learning_rate": 9.325946445060019e-06, "loss": 0.2303, "step": 1460 }, { "epoch": 0.33933518005540164, "grad_norm": 2.9136245250701904, "learning_rate": 9.321329639889198e-06, "loss": 0.1115, "step": 1470 }, { "epoch": 0.34164358264081257, "grad_norm": 9.274630546569824, "learning_rate": 9.316712834718375e-06, "loss": 0.1377, "step": 1480 }, { "epoch": 0.34395198522622344, "grad_norm": 0.033000919967889786, "learning_rate": 9.312096029547554e-06, "loss": 0.2577, "step": 1490 }, { "epoch": 0.3462603878116344, "grad_norm": 8.987251281738281, "learning_rate": 9.307479224376732e-06, "loss": 0.2595, "step": 1500 }, { "epoch": 0.34856879039704525, "grad_norm": 7.6237053871154785, "learning_rate": 9.30286241920591e-06, "loss": 0.3364, "step": 1510 }, { "epoch": 0.3508771929824561, "grad_norm": 1.9691364765167236, "learning_rate": 9.298245614035088e-06, "loss": 0.1996, "step": 1520 }, { "epoch": 0.35318559556786705, "grad_norm": 4.991685390472412, "learning_rate": 9.293628808864267e-06, "loss": 0.1637, "step": 1530 }, { "epoch": 0.3554939981532779, "grad_norm": 2.543471097946167, "learning_rate": 9.289012003693445e-06, "loss": 0.1888, "step": 1540 }, { "epoch": 0.35780240073868885, "grad_norm": 0.18523567914962769, "learning_rate": 9.284395198522624e-06, "loss": 0.2239, "step": 1550 }, { "epoch": 0.3601108033240997, "grad_norm": 5.461203098297119, "learning_rate": 9.279778393351801e-06, "loss": 0.1584, "step": 1560 }, { "epoch": 0.3624192059095106, "grad_norm": 26.16046142578125, "learning_rate": 9.27516158818098e-06, "loss": 0.1336, "step": 1570 }, { "epoch": 0.36472760849492153, "grad_norm": 3.9863812923431396, "learning_rate": 9.270544783010158e-06, "loss": 0.1406, "step": 1580 }, { "epoch": 0.3670360110803324, "grad_norm": 6.076376914978027, "learning_rate": 9.265927977839337e-06, "loss": 0.3124, "step": 1590 }, { "epoch": 0.36934441366574333, "grad_norm": 2.2239086627960205, "learning_rate": 9.261311172668514e-06, "loss": 0.144, "step": 1600 }, { "epoch": 0.3716528162511542, "grad_norm": 7.535514831542969, "learning_rate": 9.256694367497693e-06, "loss": 0.1381, "step": 1610 }, { "epoch": 0.3739612188365651, "grad_norm": 1.4443440437316895, "learning_rate": 9.25207756232687e-06, "loss": 0.1727, "step": 1620 }, { "epoch": 0.376269621421976, "grad_norm": 0.6596614122390747, "learning_rate": 9.24746075715605e-06, "loss": 0.2527, "step": 1630 }, { "epoch": 0.3785780240073869, "grad_norm": 7.379482746124268, "learning_rate": 9.242843951985227e-06, "loss": 0.1727, "step": 1640 }, { "epoch": 0.3808864265927978, "grad_norm": 1.6362881660461426, "learning_rate": 9.238227146814404e-06, "loss": 0.0302, "step": 1650 }, { "epoch": 0.3831948291782087, "grad_norm": 3.241750717163086, "learning_rate": 9.233610341643583e-06, "loss": 0.1562, "step": 1660 }, { "epoch": 0.38550323176361956, "grad_norm": 3.7530391216278076, "learning_rate": 9.22899353647276e-06, "loss": 0.2132, "step": 1670 }, { "epoch": 0.3878116343490305, "grad_norm": 5.971234321594238, "learning_rate": 9.22437673130194e-06, "loss": 0.2187, "step": 1680 }, { "epoch": 0.39012003693444136, "grad_norm": 6.949127674102783, "learning_rate": 9.219759926131117e-06, "loss": 0.2094, "step": 1690 }, { "epoch": 0.39242843951985223, "grad_norm": 1.806967854499817, "learning_rate": 9.215143120960296e-06, "loss": 0.1645, "step": 1700 }, { "epoch": 0.39473684210526316, "grad_norm": 9.919262886047363, "learning_rate": 9.210526315789474e-06, "loss": 0.3011, "step": 1710 }, { "epoch": 0.39704524469067404, "grad_norm": 13.103031158447266, "learning_rate": 9.205909510618653e-06, "loss": 0.333, "step": 1720 }, { "epoch": 0.39935364727608497, "grad_norm": 5.377729415893555, "learning_rate": 9.20129270544783e-06, "loss": 0.2529, "step": 1730 }, { "epoch": 0.40166204986149584, "grad_norm": 1.0202577114105225, "learning_rate": 9.19667590027701e-06, "loss": 0.1588, "step": 1740 }, { "epoch": 0.4039704524469067, "grad_norm": 2.622170925140381, "learning_rate": 9.192059095106187e-06, "loss": 0.1431, "step": 1750 }, { "epoch": 0.40627885503231764, "grad_norm": 0.314001202583313, "learning_rate": 9.187442289935366e-06, "loss": 0.1185, "step": 1760 }, { "epoch": 0.4085872576177285, "grad_norm": 4.181126594543457, "learning_rate": 9.182825484764543e-06, "loss": 0.1942, "step": 1770 }, { "epoch": 0.41089566020313945, "grad_norm": 5.7715959548950195, "learning_rate": 9.178208679593722e-06, "loss": 0.1199, "step": 1780 }, { "epoch": 0.4132040627885503, "grad_norm": 1.7448700666427612, "learning_rate": 9.1735918744229e-06, "loss": 0.2491, "step": 1790 }, { "epoch": 0.4155124653739612, "grad_norm": 1.511138677597046, "learning_rate": 9.168975069252079e-06, "loss": 0.3079, "step": 1800 }, { "epoch": 0.4178208679593721, "grad_norm": 0.836219310760498, "learning_rate": 9.164358264081256e-06, "loss": 0.1399, "step": 1810 }, { "epoch": 0.420129270544783, "grad_norm": 3.6991126537323, "learning_rate": 9.159741458910435e-06, "loss": 0.1483, "step": 1820 }, { "epoch": 0.4224376731301939, "grad_norm": 2.4861156940460205, "learning_rate": 9.155124653739612e-06, "loss": 0.1648, "step": 1830 }, { "epoch": 0.4247460757156048, "grad_norm": 7.801022529602051, "learning_rate": 9.150507848568791e-06, "loss": 0.1629, "step": 1840 }, { "epoch": 0.42705447830101567, "grad_norm": 0.09382744878530502, "learning_rate": 9.14589104339797e-06, "loss": 0.2144, "step": 1850 }, { "epoch": 0.4293628808864266, "grad_norm": 18.142040252685547, "learning_rate": 9.141274238227148e-06, "loss": 0.297, "step": 1860 }, { "epoch": 0.4316712834718375, "grad_norm": 6.909928321838379, "learning_rate": 9.136657433056327e-06, "loss": 0.0843, "step": 1870 }, { "epoch": 0.4339796860572484, "grad_norm": 1.391675353050232, "learning_rate": 9.132040627885504e-06, "loss": 0.1386, "step": 1880 }, { "epoch": 0.4362880886426593, "grad_norm": 8.40285587310791, "learning_rate": 9.127423822714682e-06, "loss": 0.115, "step": 1890 }, { "epoch": 0.43859649122807015, "grad_norm": 8.314501762390137, "learning_rate": 9.12280701754386e-06, "loss": 0.234, "step": 1900 }, { "epoch": 0.4409048938134811, "grad_norm": 1.1695400476455688, "learning_rate": 9.118190212373038e-06, "loss": 0.2568, "step": 1910 }, { "epoch": 0.44321329639889195, "grad_norm": 10.582197189331055, "learning_rate": 9.113573407202216e-06, "loss": 0.1879, "step": 1920 }, { "epoch": 0.4455216989843029, "grad_norm": 9.855074882507324, "learning_rate": 9.108956602031395e-06, "loss": 0.2116, "step": 1930 }, { "epoch": 0.44783010156971376, "grad_norm": 2.4227991104125977, "learning_rate": 9.104339796860572e-06, "loss": 0.2096, "step": 1940 }, { "epoch": 0.45013850415512463, "grad_norm": 6.169090747833252, "learning_rate": 9.099722991689751e-06, "loss": 0.0909, "step": 1950 }, { "epoch": 0.45244690674053556, "grad_norm": 4.098794460296631, "learning_rate": 9.09510618651893e-06, "loss": 0.1646, "step": 1960 }, { "epoch": 0.45475530932594643, "grad_norm": 10.102387428283691, "learning_rate": 9.090489381348108e-06, "loss": 0.1538, "step": 1970 }, { "epoch": 0.45706371191135736, "grad_norm": 8.377202987670898, "learning_rate": 9.085872576177287e-06, "loss": 0.102, "step": 1980 }, { "epoch": 0.45937211449676824, "grad_norm": 7.5883941650390625, "learning_rate": 9.081255771006464e-06, "loss": 0.24, "step": 1990 }, { "epoch": 0.4616805170821791, "grad_norm": 1.7626484632492065, "learning_rate": 9.076638965835643e-06, "loss": 0.1121, "step": 2000 }, { "epoch": 0.46398891966759004, "grad_norm": 1.3545483350753784, "learning_rate": 9.07202216066482e-06, "loss": 0.1017, "step": 2010 }, { "epoch": 0.4662973222530009, "grad_norm": 1.3195403814315796, "learning_rate": 9.067405355494e-06, "loss": 0.075, "step": 2020 }, { "epoch": 0.46860572483841184, "grad_norm": 5.284424781799316, "learning_rate": 9.062788550323177e-06, "loss": 0.209, "step": 2030 }, { "epoch": 0.4709141274238227, "grad_norm": 5.874682903289795, "learning_rate": 9.058171745152356e-06, "loss": 0.1442, "step": 2040 }, { "epoch": 0.4732225300092336, "grad_norm": 0.37103739380836487, "learning_rate": 9.053554939981533e-06, "loss": 0.176, "step": 2050 }, { "epoch": 0.4755309325946445, "grad_norm": 10.780657768249512, "learning_rate": 9.048938134810713e-06, "loss": 0.1491, "step": 2060 }, { "epoch": 0.4778393351800554, "grad_norm": 7.184582710266113, "learning_rate": 9.04432132963989e-06, "loss": 0.2337, "step": 2070 }, { "epoch": 0.4801477377654663, "grad_norm": 2.4554800987243652, "learning_rate": 9.039704524469069e-06, "loss": 0.1151, "step": 2080 }, { "epoch": 0.4824561403508772, "grad_norm": 12.154218673706055, "learning_rate": 9.035087719298246e-06, "loss": 0.1973, "step": 2090 }, { "epoch": 0.48476454293628807, "grad_norm": 0.6119669079780579, "learning_rate": 9.030470914127425e-06, "loss": 0.1062, "step": 2100 }, { "epoch": 0.487072945521699, "grad_norm": 0.5228668451309204, "learning_rate": 9.025854108956603e-06, "loss": 0.2023, "step": 2110 }, { "epoch": 0.48938134810710987, "grad_norm": 9.339476585388184, "learning_rate": 9.021237303785782e-06, "loss": 0.2159, "step": 2120 }, { "epoch": 0.4916897506925208, "grad_norm": 0.13014836609363556, "learning_rate": 9.01662049861496e-06, "loss": 0.1892, "step": 2130 }, { "epoch": 0.4939981532779317, "grad_norm": 7.348382472991943, "learning_rate": 9.012003693444138e-06, "loss": 0.1781, "step": 2140 }, { "epoch": 0.49630655586334255, "grad_norm": 0.30433371663093567, "learning_rate": 9.007386888273316e-06, "loss": 0.0728, "step": 2150 }, { "epoch": 0.4986149584487535, "grad_norm": 3.4963645935058594, "learning_rate": 9.002770083102493e-06, "loss": 0.1225, "step": 2160 }, { "epoch": 0.5009233610341643, "grad_norm": 9.128937721252441, "learning_rate": 8.998153277931672e-06, "loss": 0.2525, "step": 2170 }, { "epoch": 0.5032317636195752, "grad_norm": 6.375169277191162, "learning_rate": 8.99353647276085e-06, "loss": 0.1635, "step": 2180 }, { "epoch": 0.5055401662049861, "grad_norm": 0.610079288482666, "learning_rate": 8.988919667590029e-06, "loss": 0.1339, "step": 2190 }, { "epoch": 0.5078485687903971, "grad_norm": 0.23317770659923553, "learning_rate": 8.984302862419206e-06, "loss": 0.1603, "step": 2200 }, { "epoch": 0.510156971375808, "grad_norm": 2.9289543628692627, "learning_rate": 8.979686057248385e-06, "loss": 0.2054, "step": 2210 }, { "epoch": 0.5124653739612188, "grad_norm": 10.74867057800293, "learning_rate": 8.975069252077562e-06, "loss": 0.2978, "step": 2220 }, { "epoch": 0.5147737765466297, "grad_norm": 13.992448806762695, "learning_rate": 8.970452446906742e-06, "loss": 0.0812, "step": 2230 }, { "epoch": 0.5170821791320406, "grad_norm": 6.574431896209717, "learning_rate": 8.965835641735919e-06, "loss": 0.226, "step": 2240 }, { "epoch": 0.5193905817174516, "grad_norm": 8.116408348083496, "learning_rate": 8.961218836565098e-06, "loss": 0.32, "step": 2250 }, { "epoch": 0.5216989843028624, "grad_norm": 0.78421550989151, "learning_rate": 8.956602031394275e-06, "loss": 0.1636, "step": 2260 }, { "epoch": 0.5240073868882733, "grad_norm": 0.781618595123291, "learning_rate": 8.951985226223454e-06, "loss": 0.2135, "step": 2270 }, { "epoch": 0.5263157894736842, "grad_norm": 8.12641429901123, "learning_rate": 8.947368421052632e-06, "loss": 0.1113, "step": 2280 }, { "epoch": 0.528624192059095, "grad_norm": 4.973891735076904, "learning_rate": 8.942751615881811e-06, "loss": 0.0812, "step": 2290 }, { "epoch": 0.530932594644506, "grad_norm": 10.630931854248047, "learning_rate": 8.938134810710988e-06, "loss": 0.2721, "step": 2300 }, { "epoch": 0.5332409972299169, "grad_norm": 4.901821613311768, "learning_rate": 8.933518005540167e-06, "loss": 0.0514, "step": 2310 }, { "epoch": 0.5355493998153278, "grad_norm": 0.31838470697402954, "learning_rate": 8.928901200369345e-06, "loss": 0.1498, "step": 2320 }, { "epoch": 0.5378578024007387, "grad_norm": 3.5705583095550537, "learning_rate": 8.924284395198524e-06, "loss": 0.2902, "step": 2330 }, { "epoch": 0.5401662049861495, "grad_norm": 3.606900215148926, "learning_rate": 8.919667590027701e-06, "loss": 0.1574, "step": 2340 }, { "epoch": 0.5424746075715605, "grad_norm": 1.8169214725494385, "learning_rate": 8.91505078485688e-06, "loss": 0.231, "step": 2350 }, { "epoch": 0.5447830101569714, "grad_norm": 0.022802285850048065, "learning_rate": 8.910433979686058e-06, "loss": 0.1842, "step": 2360 }, { "epoch": 0.5470914127423823, "grad_norm": 7.0141496658325195, "learning_rate": 8.905817174515237e-06, "loss": 0.1983, "step": 2370 }, { "epoch": 0.5493998153277931, "grad_norm": 0.25370287895202637, "learning_rate": 8.901200369344414e-06, "loss": 0.1429, "step": 2380 }, { "epoch": 0.551708217913204, "grad_norm": 0.483079195022583, "learning_rate": 8.896583564173593e-06, "loss": 0.1743, "step": 2390 }, { "epoch": 0.554016620498615, "grad_norm": 5.299391269683838, "learning_rate": 8.89196675900277e-06, "loss": 0.2134, "step": 2400 }, { "epoch": 0.5563250230840259, "grad_norm": 1.2383846044540405, "learning_rate": 8.887349953831948e-06, "loss": 0.0912, "step": 2410 }, { "epoch": 0.5586334256694367, "grad_norm": 0.8941829204559326, "learning_rate": 8.882733148661127e-06, "loss": 0.1249, "step": 2420 }, { "epoch": 0.5609418282548476, "grad_norm": 6.666571140289307, "learning_rate": 8.878116343490304e-06, "loss": 0.2344, "step": 2430 }, { "epoch": 0.5632502308402585, "grad_norm": 1.338377833366394, "learning_rate": 8.873499538319484e-06, "loss": 0.1461, "step": 2440 }, { "epoch": 0.5655586334256695, "grad_norm": 3.791501760482788, "learning_rate": 8.868882733148661e-06, "loss": 0.1497, "step": 2450 }, { "epoch": 0.5678670360110804, "grad_norm": 2.7001969814300537, "learning_rate": 8.86426592797784e-06, "loss": 0.2116, "step": 2460 }, { "epoch": 0.5701754385964912, "grad_norm": 4.568991661071777, "learning_rate": 8.859649122807017e-06, "loss": 0.1449, "step": 2470 }, { "epoch": 0.5724838411819021, "grad_norm": 1.7414610385894775, "learning_rate": 8.855032317636196e-06, "loss": 0.0863, "step": 2480 }, { "epoch": 0.574792243767313, "grad_norm": 0.5982003808021545, "learning_rate": 8.850415512465374e-06, "loss": 0.1286, "step": 2490 }, { "epoch": 0.577100646352724, "grad_norm": 1.3507001399993896, "learning_rate": 8.845798707294553e-06, "loss": 0.2061, "step": 2500 }, { "epoch": 0.5794090489381348, "grad_norm": 3.2266528606414795, "learning_rate": 8.84118190212373e-06, "loss": 0.1486, "step": 2510 }, { "epoch": 0.5817174515235457, "grad_norm": 0.191385418176651, "learning_rate": 8.83656509695291e-06, "loss": 0.1415, "step": 2520 }, { "epoch": 0.5840258541089566, "grad_norm": 0.8593127727508545, "learning_rate": 8.831948291782087e-06, "loss": 0.1024, "step": 2530 }, { "epoch": 0.5863342566943675, "grad_norm": 2.335806369781494, "learning_rate": 8.827331486611266e-06, "loss": 0.2457, "step": 2540 }, { "epoch": 0.5886426592797784, "grad_norm": 4.868105411529541, "learning_rate": 8.822714681440445e-06, "loss": 0.1876, "step": 2550 }, { "epoch": 0.5909510618651893, "grad_norm": 9.985259056091309, "learning_rate": 8.818097876269622e-06, "loss": 0.285, "step": 2560 }, { "epoch": 0.5932594644506002, "grad_norm": 5.933226585388184, "learning_rate": 8.813481071098801e-06, "loss": 0.1476, "step": 2570 }, { "epoch": 0.5955678670360111, "grad_norm": 8.517694473266602, "learning_rate": 8.808864265927979e-06, "loss": 0.1955, "step": 2580 }, { "epoch": 0.5978762696214219, "grad_norm": 0.5734942555427551, "learning_rate": 8.804247460757158e-06, "loss": 0.2412, "step": 2590 }, { "epoch": 0.6001846722068329, "grad_norm": 5.922646522521973, "learning_rate": 8.799630655586335e-06, "loss": 0.2437, "step": 2600 }, { "epoch": 0.6024930747922438, "grad_norm": 0.948183536529541, "learning_rate": 8.795013850415514e-06, "loss": 0.2197, "step": 2610 }, { "epoch": 0.6048014773776547, "grad_norm": 2.1584763526916504, "learning_rate": 8.790397045244692e-06, "loss": 0.1741, "step": 2620 }, { "epoch": 0.6071098799630655, "grad_norm": 7.596801280975342, "learning_rate": 8.78578024007387e-06, "loss": 0.1451, "step": 2630 }, { "epoch": 0.6094182825484764, "grad_norm": 0.7149803042411804, "learning_rate": 8.781163434903048e-06, "loss": 0.0927, "step": 2640 }, { "epoch": 0.6117266851338874, "grad_norm": 0.6864413619041443, "learning_rate": 8.776546629732225e-06, "loss": 0.1811, "step": 2650 }, { "epoch": 0.6140350877192983, "grad_norm": 1.6932225227355957, "learning_rate": 8.771929824561405e-06, "loss": 0.1401, "step": 2660 }, { "epoch": 0.6163434903047091, "grad_norm": 2.8230690956115723, "learning_rate": 8.767313019390582e-06, "loss": 0.187, "step": 2670 }, { "epoch": 0.61865189289012, "grad_norm": 0.6588670015335083, "learning_rate": 8.762696214219761e-06, "loss": 0.0805, "step": 2680 }, { "epoch": 0.6209602954755309, "grad_norm": 5.022006988525391, "learning_rate": 8.758079409048938e-06, "loss": 0.2703, "step": 2690 }, { "epoch": 0.6232686980609419, "grad_norm": 0.051676761358976364, "learning_rate": 8.753462603878117e-06, "loss": 0.0952, "step": 2700 }, { "epoch": 0.6255771006463527, "grad_norm": 3.2634942531585693, "learning_rate": 8.748845798707295e-06, "loss": 0.1788, "step": 2710 }, { "epoch": 0.6278855032317636, "grad_norm": 0.012699434533715248, "learning_rate": 8.744228993536474e-06, "loss": 0.1272, "step": 2720 }, { "epoch": 0.6301939058171745, "grad_norm": 1.1832557916641235, "learning_rate": 8.739612188365651e-06, "loss": 0.2542, "step": 2730 }, { "epoch": 0.6325023084025854, "grad_norm": 4.28739070892334, "learning_rate": 8.73499538319483e-06, "loss": 0.1201, "step": 2740 }, { "epoch": 0.6348107109879964, "grad_norm": 1.4598315954208374, "learning_rate": 8.730378578024008e-06, "loss": 0.1131, "step": 2750 }, { "epoch": 0.6371191135734072, "grad_norm": 2.8734846115112305, "learning_rate": 8.725761772853187e-06, "loss": 0.0912, "step": 2760 }, { "epoch": 0.6394275161588181, "grad_norm": 1.6884006261825562, "learning_rate": 8.721144967682364e-06, "loss": 0.2134, "step": 2770 }, { "epoch": 0.641735918744229, "grad_norm": 0.029618360102176666, "learning_rate": 8.716528162511543e-06, "loss": 0.2105, "step": 2780 }, { "epoch": 0.6440443213296398, "grad_norm": 6.83729362487793, "learning_rate": 8.71191135734072e-06, "loss": 0.2517, "step": 2790 }, { "epoch": 0.6463527239150508, "grad_norm": 4.557281970977783, "learning_rate": 8.7072945521699e-06, "loss": 0.2907, "step": 2800 }, { "epoch": 0.6486611265004617, "grad_norm": 4.419384002685547, "learning_rate": 8.702677746999077e-06, "loss": 0.14, "step": 2810 }, { "epoch": 0.6509695290858726, "grad_norm": 7.11807918548584, "learning_rate": 8.698060941828256e-06, "loss": 0.2452, "step": 2820 }, { "epoch": 0.6532779316712835, "grad_norm": 2.684034585952759, "learning_rate": 8.693444136657434e-06, "loss": 0.1657, "step": 2830 }, { "epoch": 0.6555863342566943, "grad_norm": 2.968884229660034, "learning_rate": 8.688827331486613e-06, "loss": 0.1818, "step": 2840 }, { "epoch": 0.6578947368421053, "grad_norm": 5.446342468261719, "learning_rate": 8.68421052631579e-06, "loss": 0.1598, "step": 2850 }, { "epoch": 0.6602031394275162, "grad_norm": 4.752799034118652, "learning_rate": 8.679593721144969e-06, "loss": 0.0894, "step": 2860 }, { "epoch": 0.6625115420129271, "grad_norm": 1.5078086853027344, "learning_rate": 8.674976915974147e-06, "loss": 0.1544, "step": 2870 }, { "epoch": 0.6648199445983379, "grad_norm": 1.139893889427185, "learning_rate": 8.670360110803326e-06, "loss": 0.1674, "step": 2880 }, { "epoch": 0.6671283471837488, "grad_norm": 0.03082580305635929, "learning_rate": 8.665743305632503e-06, "loss": 0.2105, "step": 2890 }, { "epoch": 0.6694367497691598, "grad_norm": 0.029552439227700233, "learning_rate": 8.66112650046168e-06, "loss": 0.2003, "step": 2900 }, { "epoch": 0.6717451523545707, "grad_norm": 9.0057373046875, "learning_rate": 8.65650969529086e-06, "loss": 0.2301, "step": 2910 }, { "epoch": 0.6740535549399815, "grad_norm": 1.9247156381607056, "learning_rate": 8.651892890120037e-06, "loss": 0.1002, "step": 2920 }, { "epoch": 0.6763619575253924, "grad_norm": 7.208845615386963, "learning_rate": 8.647276084949216e-06, "loss": 0.1913, "step": 2930 }, { "epoch": 0.6786703601108033, "grad_norm": 6.815885066986084, "learning_rate": 8.642659279778393e-06, "loss": 0.1772, "step": 2940 }, { "epoch": 0.6809787626962143, "grad_norm": 0.07034219056367874, "learning_rate": 8.638042474607572e-06, "loss": 0.1395, "step": 2950 }, { "epoch": 0.6832871652816251, "grad_norm": 2.6379406452178955, "learning_rate": 8.63342566943675e-06, "loss": 0.213, "step": 2960 }, { "epoch": 0.685595567867036, "grad_norm": 8.080329895019531, "learning_rate": 8.628808864265929e-06, "loss": 0.213, "step": 2970 }, { "epoch": 0.6879039704524469, "grad_norm": 0.46268343925476074, "learning_rate": 8.624192059095106e-06, "loss": 0.1013, "step": 2980 }, { "epoch": 0.6902123730378578, "grad_norm": 6.735634803771973, "learning_rate": 8.619575253924285e-06, "loss": 0.2183, "step": 2990 }, { "epoch": 0.6925207756232687, "grad_norm": 8.794092178344727, "learning_rate": 8.614958448753463e-06, "loss": 0.0705, "step": 3000 }, { "epoch": 0.6948291782086796, "grad_norm": 0.08418063074350357, "learning_rate": 8.610341643582642e-06, "loss": 0.0917, "step": 3010 }, { "epoch": 0.6971375807940905, "grad_norm": 4.231406211853027, "learning_rate": 8.605724838411819e-06, "loss": 0.1316, "step": 3020 }, { "epoch": 0.6994459833795014, "grad_norm": 0.9246407747268677, "learning_rate": 8.601108033240998e-06, "loss": 0.0585, "step": 3030 }, { "epoch": 0.7017543859649122, "grad_norm": 11.197175979614258, "learning_rate": 8.596491228070176e-06, "loss": 0.2107, "step": 3040 }, { "epoch": 0.7040627885503232, "grad_norm": 5.508973598480225, "learning_rate": 8.591874422899355e-06, "loss": 0.2448, "step": 3050 }, { "epoch": 0.7063711911357341, "grad_norm": 5.923152446746826, "learning_rate": 8.587257617728532e-06, "loss": 0.0883, "step": 3060 }, { "epoch": 0.708679593721145, "grad_norm": 11.441405296325684, "learning_rate": 8.582640812557711e-06, "loss": 0.3493, "step": 3070 }, { "epoch": 0.7109879963065558, "grad_norm": 1.7312946319580078, "learning_rate": 8.578024007386888e-06, "loss": 0.1599, "step": 3080 }, { "epoch": 0.7132963988919667, "grad_norm": 7.770139694213867, "learning_rate": 8.573407202216068e-06, "loss": 0.2575, "step": 3090 }, { "epoch": 0.7156048014773777, "grad_norm": 4.013419151306152, "learning_rate": 8.568790397045245e-06, "loss": 0.1331, "step": 3100 }, { "epoch": 0.7179132040627886, "grad_norm": 4.1502790451049805, "learning_rate": 8.564173591874424e-06, "loss": 0.0954, "step": 3110 }, { "epoch": 0.7202216066481995, "grad_norm": 8.959720611572266, "learning_rate": 8.559556786703603e-06, "loss": 0.3433, "step": 3120 }, { "epoch": 0.7225300092336103, "grad_norm": 2.727738380432129, "learning_rate": 8.55493998153278e-06, "loss": 0.0849, "step": 3130 }, { "epoch": 0.7248384118190212, "grad_norm": 6.403547763824463, "learning_rate": 8.55032317636196e-06, "loss": 0.2253, "step": 3140 }, { "epoch": 0.7271468144044322, "grad_norm": 16.912302017211914, "learning_rate": 8.545706371191135e-06, "loss": 0.2936, "step": 3150 }, { "epoch": 0.7294552169898431, "grad_norm": 1.6519527435302734, "learning_rate": 8.541089566020314e-06, "loss": 0.2582, "step": 3160 }, { "epoch": 0.7317636195752539, "grad_norm": 1.9267312288284302, "learning_rate": 8.536472760849492e-06, "loss": 0.101, "step": 3170 }, { "epoch": 0.7340720221606648, "grad_norm": 1.5446139574050903, "learning_rate": 8.53185595567867e-06, "loss": 0.2388, "step": 3180 }, { "epoch": 0.7363804247460757, "grad_norm": 0.07017502188682556, "learning_rate": 8.527239150507848e-06, "loss": 0.2017, "step": 3190 }, { "epoch": 0.7386888273314867, "grad_norm": 0.07057207077741623, "learning_rate": 8.522622345337027e-06, "loss": 0.22, "step": 3200 }, { "epoch": 0.7409972299168975, "grad_norm": 6.549715042114258, "learning_rate": 8.518005540166205e-06, "loss": 0.2471, "step": 3210 }, { "epoch": 0.7433056325023084, "grad_norm": 10.777205467224121, "learning_rate": 8.513388734995384e-06, "loss": 0.0954, "step": 3220 }, { "epoch": 0.7456140350877193, "grad_norm": 3.3119924068450928, "learning_rate": 8.508771929824563e-06, "loss": 0.0707, "step": 3230 }, { "epoch": 0.7479224376731302, "grad_norm": 9.866997718811035, "learning_rate": 8.50415512465374e-06, "loss": 0.1698, "step": 3240 }, { "epoch": 0.7502308402585411, "grad_norm": 4.566750526428223, "learning_rate": 8.49953831948292e-06, "loss": 0.2585, "step": 3250 }, { "epoch": 0.752539242843952, "grad_norm": 9.86897087097168, "learning_rate": 8.494921514312097e-06, "loss": 0.1458, "step": 3260 }, { "epoch": 0.7548476454293629, "grad_norm": 5.148314476013184, "learning_rate": 8.490304709141276e-06, "loss": 0.1554, "step": 3270 }, { "epoch": 0.7571560480147738, "grad_norm": 9.085517883300781, "learning_rate": 8.485687903970453e-06, "loss": 0.24, "step": 3280 }, { "epoch": 0.7594644506001846, "grad_norm": 0.2084989994764328, "learning_rate": 8.481071098799632e-06, "loss": 0.2083, "step": 3290 }, { "epoch": 0.7617728531855956, "grad_norm": 11.02973461151123, "learning_rate": 8.47645429362881e-06, "loss": 0.2082, "step": 3300 }, { "epoch": 0.7640812557710065, "grad_norm": 3.157027244567871, "learning_rate": 8.471837488457989e-06, "loss": 0.1888, "step": 3310 }, { "epoch": 0.7663896583564174, "grad_norm": 5.361010551452637, "learning_rate": 8.467220683287166e-06, "loss": 0.0868, "step": 3320 }, { "epoch": 0.7686980609418282, "grad_norm": 6.1732096672058105, "learning_rate": 8.462603878116345e-06, "loss": 0.1713, "step": 3330 }, { "epoch": 0.7710064635272391, "grad_norm": 5.8650946617126465, "learning_rate": 8.457987072945522e-06, "loss": 0.127, "step": 3340 }, { "epoch": 0.7733148661126501, "grad_norm": 4.028465747833252, "learning_rate": 8.453370267774701e-06, "loss": 0.097, "step": 3350 }, { "epoch": 0.775623268698061, "grad_norm": 1.5145978927612305, "learning_rate": 8.448753462603879e-06, "loss": 0.1583, "step": 3360 }, { "epoch": 0.7779316712834718, "grad_norm": 3.635915756225586, "learning_rate": 8.444136657433058e-06, "loss": 0.2075, "step": 3370 }, { "epoch": 0.7802400738688827, "grad_norm": 0.03490106016397476, "learning_rate": 8.439519852262235e-06, "loss": 0.1947, "step": 3380 }, { "epoch": 0.7825484764542936, "grad_norm": 10.461822509765625, "learning_rate": 8.434903047091414e-06, "loss": 0.0826, "step": 3390 }, { "epoch": 0.7848568790397045, "grad_norm": 7.9899420738220215, "learning_rate": 8.430286241920592e-06, "loss": 0.2917, "step": 3400 }, { "epoch": 0.7871652816251155, "grad_norm": 7.81199312210083, "learning_rate": 8.42566943674977e-06, "loss": 0.1562, "step": 3410 }, { "epoch": 0.7894736842105263, "grad_norm": 1.9472380876541138, "learning_rate": 8.421052631578948e-06, "loss": 0.1568, "step": 3420 }, { "epoch": 0.7917820867959372, "grad_norm": 4.842315196990967, "learning_rate": 8.416435826408126e-06, "loss": 0.1681, "step": 3430 }, { "epoch": 0.7940904893813481, "grad_norm": 27.294620513916016, "learning_rate": 8.411819021237305e-06, "loss": 0.1717, "step": 3440 }, { "epoch": 0.796398891966759, "grad_norm": 8.655313491821289, "learning_rate": 8.407202216066482e-06, "loss": 0.3171, "step": 3450 }, { "epoch": 0.7987072945521699, "grad_norm": 5.681049346923828, "learning_rate": 8.402585410895661e-06, "loss": 0.1738, "step": 3460 }, { "epoch": 0.8010156971375808, "grad_norm": 0.6661124229431152, "learning_rate": 8.397968605724839e-06, "loss": 0.1324, "step": 3470 }, { "epoch": 0.8033240997229917, "grad_norm": 0.6557456851005554, "learning_rate": 8.393351800554018e-06, "loss": 0.1253, "step": 3480 }, { "epoch": 0.8056325023084026, "grad_norm": 6.639368057250977, "learning_rate": 8.388734995383195e-06, "loss": 0.1955, "step": 3490 }, { "epoch": 0.8079409048938134, "grad_norm": 2.957080841064453, "learning_rate": 8.384118190212374e-06, "loss": 0.1188, "step": 3500 }, { "epoch": 0.8102493074792244, "grad_norm": 6.867522239685059, "learning_rate": 8.379501385041551e-06, "loss": 0.1443, "step": 3510 }, { "epoch": 0.8125577100646353, "grad_norm": 2.938314199447632, "learning_rate": 8.37488457987073e-06, "loss": 0.176, "step": 3520 }, { "epoch": 0.8148661126500462, "grad_norm": 0.040740109980106354, "learning_rate": 8.370267774699908e-06, "loss": 0.0661, "step": 3530 }, { "epoch": 0.817174515235457, "grad_norm": 2.256371021270752, "learning_rate": 8.365650969529087e-06, "loss": 0.1342, "step": 3540 }, { "epoch": 0.8194829178208679, "grad_norm": 11.778620719909668, "learning_rate": 8.361034164358264e-06, "loss": 0.1291, "step": 3550 }, { "epoch": 0.8217913204062789, "grad_norm": 7.650779724121094, "learning_rate": 8.356417359187443e-06, "loss": 0.2504, "step": 3560 }, { "epoch": 0.8240997229916898, "grad_norm": 0.4972974359989166, "learning_rate": 8.35180055401662e-06, "loss": 0.2238, "step": 3570 }, { "epoch": 0.8264081255771006, "grad_norm": 0.31301605701446533, "learning_rate": 8.3471837488458e-06, "loss": 0.2013, "step": 3580 }, { "epoch": 0.8287165281625115, "grad_norm": 5.242720127105713, "learning_rate": 8.342566943674977e-06, "loss": 0.192, "step": 3590 }, { "epoch": 0.8310249307479224, "grad_norm": 1.1165562868118286, "learning_rate": 8.337950138504156e-06, "loss": 0.2283, "step": 3600 }, { "epoch": 0.8333333333333334, "grad_norm": 5.6433329582214355, "learning_rate": 8.333333333333334e-06, "loss": 0.1737, "step": 3610 }, { "epoch": 0.8356417359187442, "grad_norm": 2.8478212356567383, "learning_rate": 8.328716528162513e-06, "loss": 0.2074, "step": 3620 }, { "epoch": 0.8379501385041551, "grad_norm": 1.121833086013794, "learning_rate": 8.32409972299169e-06, "loss": 0.2048, "step": 3630 }, { "epoch": 0.840258541089566, "grad_norm": 4.590597152709961, "learning_rate": 8.31948291782087e-06, "loss": 0.1114, "step": 3640 }, { "epoch": 0.8425669436749769, "grad_norm": 3.8996849060058594, "learning_rate": 8.314866112650047e-06, "loss": 0.0985, "step": 3650 }, { "epoch": 0.8448753462603878, "grad_norm": 9.282421112060547, "learning_rate": 8.310249307479224e-06, "loss": 0.0823, "step": 3660 }, { "epoch": 0.8471837488457987, "grad_norm": 0.4110245406627655, "learning_rate": 8.305632502308403e-06, "loss": 0.1489, "step": 3670 }, { "epoch": 0.8494921514312096, "grad_norm": 0.560570478439331, "learning_rate": 8.30101569713758e-06, "loss": 0.1714, "step": 3680 }, { "epoch": 0.8518005540166205, "grad_norm": 4.820432662963867, "learning_rate": 8.29639889196676e-06, "loss": 0.2155, "step": 3690 }, { "epoch": 0.8541089566020313, "grad_norm": 0.06603138893842697, "learning_rate": 8.291782086795937e-06, "loss": 0.0704, "step": 3700 }, { "epoch": 0.8564173591874423, "grad_norm": 8.31919002532959, "learning_rate": 8.287165281625116e-06, "loss": 0.2275, "step": 3710 }, { "epoch": 0.8587257617728532, "grad_norm": 1.474788784980774, "learning_rate": 8.282548476454293e-06, "loss": 0.2416, "step": 3720 }, { "epoch": 0.8610341643582641, "grad_norm": 5.689862251281738, "learning_rate": 8.277931671283472e-06, "loss": 0.084, "step": 3730 }, { "epoch": 0.863342566943675, "grad_norm": 1.4493321180343628, "learning_rate": 8.27331486611265e-06, "loss": 0.1913, "step": 3740 }, { "epoch": 0.8656509695290858, "grad_norm": 6.178581714630127, "learning_rate": 8.268698060941829e-06, "loss": 0.1903, "step": 3750 }, { "epoch": 0.8679593721144968, "grad_norm": 7.550521373748779, "learning_rate": 8.264081255771006e-06, "loss": 0.1644, "step": 3760 }, { "epoch": 0.8702677746999077, "grad_norm": 0.4819086790084839, "learning_rate": 8.259464450600185e-06, "loss": 0.2667, "step": 3770 }, { "epoch": 0.8725761772853186, "grad_norm": 2.676957130432129, "learning_rate": 8.254847645429363e-06, "loss": 0.0307, "step": 3780 }, { "epoch": 0.8748845798707294, "grad_norm": 3.926511526107788, "learning_rate": 8.250230840258542e-06, "loss": 0.2962, "step": 3790 }, { "epoch": 0.8771929824561403, "grad_norm": 0.15894842147827148, "learning_rate": 8.24561403508772e-06, "loss": 0.198, "step": 3800 }, { "epoch": 0.8795013850415513, "grad_norm": 3.0533387660980225, "learning_rate": 8.240997229916898e-06, "loss": 0.2216, "step": 3810 }, { "epoch": 0.8818097876269622, "grad_norm": 3.607330322265625, "learning_rate": 8.236380424746077e-06, "loss": 0.2361, "step": 3820 }, { "epoch": 0.884118190212373, "grad_norm": 6.245927810668945, "learning_rate": 8.231763619575255e-06, "loss": 0.2664, "step": 3830 }, { "epoch": 0.8864265927977839, "grad_norm": 1.1220074892044067, "learning_rate": 8.227146814404434e-06, "loss": 0.2193, "step": 3840 }, { "epoch": 0.8887349953831948, "grad_norm": 7.891664981842041, "learning_rate": 8.222530009233611e-06, "loss": 0.1886, "step": 3850 }, { "epoch": 0.8910433979686058, "grad_norm": 1.1142659187316895, "learning_rate": 8.21791320406279e-06, "loss": 0.1643, "step": 3860 }, { "epoch": 0.8933518005540166, "grad_norm": 1.0188210010528564, "learning_rate": 8.213296398891968e-06, "loss": 0.2573, "step": 3870 }, { "epoch": 0.8956602031394275, "grad_norm": 1.5279109477996826, "learning_rate": 8.208679593721147e-06, "loss": 0.1179, "step": 3880 }, { "epoch": 0.8979686057248384, "grad_norm": 0.9006258845329285, "learning_rate": 8.204062788550324e-06, "loss": 0.086, "step": 3890 }, { "epoch": 0.9002770083102493, "grad_norm": 0.4039612412452698, "learning_rate": 8.199445983379503e-06, "loss": 0.1671, "step": 3900 }, { "epoch": 0.9025854108956602, "grad_norm": 4.117120265960693, "learning_rate": 8.194829178208679e-06, "loss": 0.1978, "step": 3910 }, { "epoch": 0.9048938134810711, "grad_norm": 8.001017570495605, "learning_rate": 8.190212373037858e-06, "loss": 0.1339, "step": 3920 }, { "epoch": 0.907202216066482, "grad_norm": 4.532901287078857, "learning_rate": 8.185595567867037e-06, "loss": 0.417, "step": 3930 }, { "epoch": 0.9095106186518929, "grad_norm": 0.5681565403938293, "learning_rate": 8.180978762696214e-06, "loss": 0.0796, "step": 3940 }, { "epoch": 0.9118190212373037, "grad_norm": 2.5583901405334473, "learning_rate": 8.176361957525394e-06, "loss": 0.3093, "step": 3950 }, { "epoch": 0.9141274238227147, "grad_norm": 4.248782634735107, "learning_rate": 8.171745152354571e-06, "loss": 0.1073, "step": 3960 }, { "epoch": 0.9164358264081256, "grad_norm": 0.4341060221195221, "learning_rate": 8.16712834718375e-06, "loss": 0.1543, "step": 3970 }, { "epoch": 0.9187442289935365, "grad_norm": 1.084920883178711, "learning_rate": 8.162511542012927e-06, "loss": 0.1382, "step": 3980 }, { "epoch": 0.9210526315789473, "grad_norm": 8.02807331085205, "learning_rate": 8.157894736842106e-06, "loss": 0.1705, "step": 3990 }, { "epoch": 0.9233610341643582, "grad_norm": 6.406728267669678, "learning_rate": 8.153277931671284e-06, "loss": 0.1939, "step": 4000 }, { "epoch": 0.9256694367497692, "grad_norm": 0.7076385021209717, "learning_rate": 8.148661126500463e-06, "loss": 0.1844, "step": 4010 }, { "epoch": 0.9279778393351801, "grad_norm": 7.5065598487854, "learning_rate": 8.14404432132964e-06, "loss": 0.1937, "step": 4020 }, { "epoch": 0.930286241920591, "grad_norm": 2.018402338027954, "learning_rate": 8.13942751615882e-06, "loss": 0.174, "step": 4030 }, { "epoch": 0.9325946445060018, "grad_norm": 4.676671504974365, "learning_rate": 8.134810710987997e-06, "loss": 0.1498, "step": 4040 }, { "epoch": 0.9349030470914127, "grad_norm": 3.8964033126831055, "learning_rate": 8.130193905817176e-06, "loss": 0.1475, "step": 4050 }, { "epoch": 0.9372114496768237, "grad_norm": 0.6416448354721069, "learning_rate": 8.125577100646353e-06, "loss": 0.1098, "step": 4060 }, { "epoch": 0.9395198522622346, "grad_norm": 0.4627153277397156, "learning_rate": 8.120960295475532e-06, "loss": 0.1971, "step": 4070 }, { "epoch": 0.9418282548476454, "grad_norm": 3.1341536045074463, "learning_rate": 8.11634349030471e-06, "loss": 0.1556, "step": 4080 }, { "epoch": 0.9441366574330563, "grad_norm": 8.236727714538574, "learning_rate": 8.111726685133889e-06, "loss": 0.1973, "step": 4090 }, { "epoch": 0.9464450600184672, "grad_norm": 9.807982444763184, "learning_rate": 8.107109879963066e-06, "loss": 0.1444, "step": 4100 }, { "epoch": 0.9487534626038782, "grad_norm": 0.607775092124939, "learning_rate": 8.102493074792245e-06, "loss": 0.2249, "step": 4110 }, { "epoch": 0.951061865189289, "grad_norm": 0.0064311218447983265, "learning_rate": 8.097876269621423e-06, "loss": 0.1164, "step": 4120 }, { "epoch": 0.9533702677746999, "grad_norm": 1.9682823419570923, "learning_rate": 8.093259464450602e-06, "loss": 0.1664, "step": 4130 }, { "epoch": 0.9556786703601108, "grad_norm": 0.1796337068080902, "learning_rate": 8.088642659279779e-06, "loss": 0.0976, "step": 4140 }, { "epoch": 0.9579870729455217, "grad_norm": 2.5847933292388916, "learning_rate": 8.084025854108958e-06, "loss": 0.2195, "step": 4150 }, { "epoch": 0.9602954755309326, "grad_norm": 1.2782025337219238, "learning_rate": 8.079409048938135e-06, "loss": 0.1616, "step": 4160 }, { "epoch": 0.9626038781163435, "grad_norm": 6.094910621643066, "learning_rate": 8.074792243767313e-06, "loss": 0.281, "step": 4170 }, { "epoch": 0.9649122807017544, "grad_norm": 0.46985533833503723, "learning_rate": 8.070175438596492e-06, "loss": 0.1701, "step": 4180 }, { "epoch": 0.9672206832871653, "grad_norm": 0.41220489144325256, "learning_rate": 8.06555863342567e-06, "loss": 0.1823, "step": 4190 }, { "epoch": 0.9695290858725761, "grad_norm": 0.05810916796326637, "learning_rate": 8.060941828254848e-06, "loss": 0.1169, "step": 4200 }, { "epoch": 0.9718374884579871, "grad_norm": 2.851436138153076, "learning_rate": 8.056325023084026e-06, "loss": 0.1922, "step": 4210 }, { "epoch": 0.974145891043398, "grad_norm": 1.4276131391525269, "learning_rate": 8.051708217913205e-06, "loss": 0.1672, "step": 4220 }, { "epoch": 0.9764542936288089, "grad_norm": 0.03251894935965538, "learning_rate": 8.047091412742382e-06, "loss": 0.1026, "step": 4230 }, { "epoch": 0.9787626962142197, "grad_norm": 6.590252876281738, "learning_rate": 8.042474607571561e-06, "loss": 0.1478, "step": 4240 }, { "epoch": 0.9810710987996306, "grad_norm": 1.2194827795028687, "learning_rate": 8.037857802400739e-06, "loss": 0.1922, "step": 4250 }, { "epoch": 0.9833795013850416, "grad_norm": 0.10244120657444, "learning_rate": 8.033240997229918e-06, "loss": 0.1003, "step": 4260 }, { "epoch": 0.9856879039704525, "grad_norm": 2.5781946182250977, "learning_rate": 8.028624192059095e-06, "loss": 0.1253, "step": 4270 }, { "epoch": 0.9879963065558633, "grad_norm": 6.615061283111572, "learning_rate": 8.024007386888274e-06, "loss": 0.1525, "step": 4280 }, { "epoch": 0.9903047091412742, "grad_norm": 4.588455677032471, "learning_rate": 8.019390581717452e-06, "loss": 0.281, "step": 4290 }, { "epoch": 0.9926131117266851, "grad_norm": 4.515980243682861, "learning_rate": 8.01477377654663e-06, "loss": 0.0448, "step": 4300 }, { "epoch": 0.9949215143120961, "grad_norm": 5.931595802307129, "learning_rate": 8.010156971375808e-06, "loss": 0.2194, "step": 4310 }, { "epoch": 0.997229916897507, "grad_norm": 0.3821454346179962, "learning_rate": 8.005540166204987e-06, "loss": 0.2029, "step": 4320 }, { "epoch": 0.9995383194829178, "grad_norm": 0.06136353313922882, "learning_rate": 8.000923361034165e-06, "loss": 0.1044, "step": 4330 }, { "epoch": 1.0018467220683287, "grad_norm": 4.806890487670898, "learning_rate": 7.996306555863344e-06, "loss": 0.1538, "step": 4340 }, { "epoch": 1.0041551246537397, "grad_norm": 0.5629621148109436, "learning_rate": 7.991689750692521e-06, "loss": 0.2269, "step": 4350 }, { "epoch": 1.0064635272391504, "grad_norm": 5.226959228515625, "learning_rate": 7.9870729455217e-06, "loss": 0.2343, "step": 4360 }, { "epoch": 1.0087719298245614, "grad_norm": 0.7424554228782654, "learning_rate": 7.982456140350877e-06, "loss": 0.1572, "step": 4370 }, { "epoch": 1.0110803324099722, "grad_norm": 4.007261753082275, "learning_rate": 7.977839335180056e-06, "loss": 0.1134, "step": 4380 }, { "epoch": 1.0133887349953832, "grad_norm": 8.55867862701416, "learning_rate": 7.973222530009236e-06, "loss": 0.1848, "step": 4390 }, { "epoch": 1.0156971375807942, "grad_norm": 0.7979418039321899, "learning_rate": 7.968605724838413e-06, "loss": 0.1253, "step": 4400 }, { "epoch": 1.018005540166205, "grad_norm": 11.240249633789062, "learning_rate": 7.96398891966759e-06, "loss": 0.1621, "step": 4410 }, { "epoch": 1.020313942751616, "grad_norm": 3.8235790729522705, "learning_rate": 7.959372114496768e-06, "loss": 0.1828, "step": 4420 }, { "epoch": 1.0226223453370267, "grad_norm": 0.04937285929918289, "learning_rate": 7.954755309325947e-06, "loss": 0.053, "step": 4430 }, { "epoch": 1.0249307479224377, "grad_norm": 5.2850213050842285, "learning_rate": 7.950138504155124e-06, "loss": 0.2751, "step": 4440 }, { "epoch": 1.0272391505078486, "grad_norm": 0.0649990439414978, "learning_rate": 7.945521698984303e-06, "loss": 0.1219, "step": 4450 }, { "epoch": 1.0295475530932594, "grad_norm": 3.24285626411438, "learning_rate": 7.94090489381348e-06, "loss": 0.1301, "step": 4460 }, { "epoch": 1.0318559556786704, "grad_norm": 1.6931140422821045, "learning_rate": 7.93628808864266e-06, "loss": 0.0896, "step": 4470 }, { "epoch": 1.0341643582640812, "grad_norm": 0.7782284021377563, "learning_rate": 7.931671283471837e-06, "loss": 0.1471, "step": 4480 }, { "epoch": 1.0364727608494921, "grad_norm": 4.752656936645508, "learning_rate": 7.927054478301016e-06, "loss": 0.1937, "step": 4490 }, { "epoch": 1.0387811634349031, "grad_norm": 0.6479209661483765, "learning_rate": 7.922437673130195e-06, "loss": 0.1246, "step": 4500 }, { "epoch": 1.0410895660203139, "grad_norm": 9.935562133789062, "learning_rate": 7.917820867959373e-06, "loss": 0.1142, "step": 4510 }, { "epoch": 1.0433979686057249, "grad_norm": 3.0006847381591797, "learning_rate": 7.913204062788552e-06, "loss": 0.097, "step": 4520 }, { "epoch": 1.0457063711911356, "grad_norm": 4.007802963256836, "learning_rate": 7.908587257617729e-06, "loss": 0.1299, "step": 4530 }, { "epoch": 1.0480147737765466, "grad_norm": 2.044865608215332, "learning_rate": 7.903970452446908e-06, "loss": 0.1708, "step": 4540 }, { "epoch": 1.0503231763619576, "grad_norm": 3.5933120250701904, "learning_rate": 7.899353647276086e-06, "loss": 0.0899, "step": 4550 }, { "epoch": 1.0526315789473684, "grad_norm": 0.012179333716630936, "learning_rate": 7.894736842105265e-06, "loss": 0.158, "step": 4560 }, { "epoch": 1.0549399815327793, "grad_norm": 5.514457702636719, "learning_rate": 7.890120036934442e-06, "loss": 0.1148, "step": 4570 }, { "epoch": 1.05724838411819, "grad_norm": 3.768728017807007, "learning_rate": 7.885503231763621e-06, "loss": 0.3045, "step": 4580 }, { "epoch": 1.059556786703601, "grad_norm": 2.062894344329834, "learning_rate": 7.880886426592798e-06, "loss": 0.0932, "step": 4590 }, { "epoch": 1.061865189289012, "grad_norm": 7.800497531890869, "learning_rate": 7.876269621421978e-06, "loss": 0.0924, "step": 4600 }, { "epoch": 1.0641735918744228, "grad_norm": 4.241180419921875, "learning_rate": 7.871652816251155e-06, "loss": 0.1322, "step": 4610 }, { "epoch": 1.0664819944598338, "grad_norm": 1.3674288988113403, "learning_rate": 7.867036011080334e-06, "loss": 0.1182, "step": 4620 }, { "epoch": 1.0687903970452446, "grad_norm": 0.2856109142303467, "learning_rate": 7.862419205909511e-06, "loss": 0.2477, "step": 4630 }, { "epoch": 1.0710987996306556, "grad_norm": 5.879015922546387, "learning_rate": 7.85780240073869e-06, "loss": 0.045, "step": 4640 }, { "epoch": 1.0734072022160666, "grad_norm": 10.425411224365234, "learning_rate": 7.853185595567868e-06, "loss": 0.1411, "step": 4650 }, { "epoch": 1.0757156048014773, "grad_norm": 5.851372718811035, "learning_rate": 7.848568790397045e-06, "loss": 0.1715, "step": 4660 }, { "epoch": 1.0780240073868883, "grad_norm": 3.7951362133026123, "learning_rate": 7.843951985226224e-06, "loss": 0.1504, "step": 4670 }, { "epoch": 1.080332409972299, "grad_norm": 10.139032363891602, "learning_rate": 7.839335180055402e-06, "loss": 0.1855, "step": 4680 }, { "epoch": 1.08264081255771, "grad_norm": 0.4991055727005005, "learning_rate": 7.83471837488458e-06, "loss": 0.0819, "step": 4690 }, { "epoch": 1.084949215143121, "grad_norm": 0.4465656578540802, "learning_rate": 7.830101569713758e-06, "loss": 0.145, "step": 4700 }, { "epoch": 1.0872576177285318, "grad_norm": 0.4848790466785431, "learning_rate": 7.825484764542937e-06, "loss": 0.315, "step": 4710 }, { "epoch": 1.0895660203139428, "grad_norm": 4.088257312774658, "learning_rate": 7.820867959372115e-06, "loss": 0.0998, "step": 4720 }, { "epoch": 1.0918744228993535, "grad_norm": 8.607985496520996, "learning_rate": 7.816251154201294e-06, "loss": 0.1132, "step": 4730 }, { "epoch": 1.0941828254847645, "grad_norm": 2.8763158321380615, "learning_rate": 7.811634349030471e-06, "loss": 0.1356, "step": 4740 }, { "epoch": 1.0964912280701755, "grad_norm": 7.248260974884033, "learning_rate": 7.80701754385965e-06, "loss": 0.1666, "step": 4750 }, { "epoch": 1.0987996306555863, "grad_norm": 1.1353635787963867, "learning_rate": 7.802400738688827e-06, "loss": 0.1783, "step": 4760 }, { "epoch": 1.1011080332409973, "grad_norm": 2.150517463684082, "learning_rate": 7.797783933518007e-06, "loss": 0.1907, "step": 4770 }, { "epoch": 1.103416435826408, "grad_norm": 4.966104030609131, "learning_rate": 7.793167128347184e-06, "loss": 0.1418, "step": 4780 }, { "epoch": 1.105724838411819, "grad_norm": 0.1536702960729599, "learning_rate": 7.788550323176363e-06, "loss": 0.0984, "step": 4790 }, { "epoch": 1.10803324099723, "grad_norm": 0.11235389858484268, "learning_rate": 7.78393351800554e-06, "loss": 0.1195, "step": 4800 }, { "epoch": 1.1103416435826408, "grad_norm": 4.2295942306518555, "learning_rate": 7.77931671283472e-06, "loss": 0.1564, "step": 4810 }, { "epoch": 1.1126500461680517, "grad_norm": 9.70842170715332, "learning_rate": 7.774699907663897e-06, "loss": 0.1065, "step": 4820 }, { "epoch": 1.1149584487534625, "grad_norm": 3.682037830352783, "learning_rate": 7.770083102493076e-06, "loss": 0.2343, "step": 4830 }, { "epoch": 1.1172668513388735, "grad_norm": 2.330042839050293, "learning_rate": 7.765466297322253e-06, "loss": 0.0533, "step": 4840 }, { "epoch": 1.1195752539242845, "grad_norm": 1.348418951034546, "learning_rate": 7.760849492151432e-06, "loss": 0.2092, "step": 4850 }, { "epoch": 1.1218836565096952, "grad_norm": 0.12500044703483582, "learning_rate": 7.75623268698061e-06, "loss": 0.1195, "step": 4860 }, { "epoch": 1.1241920590951062, "grad_norm": 7.640311241149902, "learning_rate": 7.751615881809789e-06, "loss": 0.1231, "step": 4870 }, { "epoch": 1.126500461680517, "grad_norm": 0.8686632513999939, "learning_rate": 7.746999076638966e-06, "loss": 0.1721, "step": 4880 }, { "epoch": 1.128808864265928, "grad_norm": 0.28602880239486694, "learning_rate": 7.742382271468145e-06, "loss": 0.0913, "step": 4890 }, { "epoch": 1.131117266851339, "grad_norm": 0.28318125009536743, "learning_rate": 7.737765466297323e-06, "loss": 0.1647, "step": 4900 }, { "epoch": 1.1334256694367497, "grad_norm": 1.848526954650879, "learning_rate": 7.733148661126502e-06, "loss": 0.2331, "step": 4910 }, { "epoch": 1.1357340720221607, "grad_norm": 0.11867644637823105, "learning_rate": 7.728531855955679e-06, "loss": 0.0607, "step": 4920 }, { "epoch": 1.1380424746075715, "grad_norm": 6.992359161376953, "learning_rate": 7.723915050784857e-06, "loss": 0.0408, "step": 4930 }, { "epoch": 1.1403508771929824, "grad_norm": 28.76036834716797, "learning_rate": 7.719298245614036e-06, "loss": 0.2178, "step": 4940 }, { "epoch": 1.1426592797783934, "grad_norm": 2.440525531768799, "learning_rate": 7.714681440443213e-06, "loss": 0.093, "step": 4950 }, { "epoch": 1.1449676823638042, "grad_norm": 0.001492099603638053, "learning_rate": 7.710064635272392e-06, "loss": 0.2124, "step": 4960 }, { "epoch": 1.1472760849492152, "grad_norm": 0.0038797783199697733, "learning_rate": 7.70544783010157e-06, "loss": 0.0723, "step": 4970 }, { "epoch": 1.149584487534626, "grad_norm": 5.044519901275635, "learning_rate": 7.700831024930749e-06, "loss": 0.1172, "step": 4980 }, { "epoch": 1.151892890120037, "grad_norm": 2.8185760974884033, "learning_rate": 7.696214219759926e-06, "loss": 0.1057, "step": 4990 }, { "epoch": 1.154201292705448, "grad_norm": 4.505733013153076, "learning_rate": 7.691597414589105e-06, "loss": 0.0768, "step": 5000 }, { "epoch": 1.1565096952908587, "grad_norm": 1.1265312433242798, "learning_rate": 7.686980609418282e-06, "loss": 0.1504, "step": 5010 }, { "epoch": 1.1588180978762697, "grad_norm": 0.6987248063087463, "learning_rate": 7.682363804247461e-06, "loss": 0.0697, "step": 5020 }, { "epoch": 1.1611265004616804, "grad_norm": 9.511496543884277, "learning_rate": 7.677746999076639e-06, "loss": 0.139, "step": 5030 }, { "epoch": 1.1634349030470914, "grad_norm": 7.314056873321533, "learning_rate": 7.673130193905818e-06, "loss": 0.1918, "step": 5040 }, { "epoch": 1.1657433056325024, "grad_norm": 0.1578429639339447, "learning_rate": 7.668513388734995e-06, "loss": 0.1648, "step": 5050 }, { "epoch": 1.1680517082179132, "grad_norm": 0.12806738913059235, "learning_rate": 7.663896583564174e-06, "loss": 0.1409, "step": 5060 }, { "epoch": 1.1703601108033241, "grad_norm": 0.6459066271781921, "learning_rate": 7.659279778393352e-06, "loss": 0.0807, "step": 5070 }, { "epoch": 1.172668513388735, "grad_norm": 2.657371997833252, "learning_rate": 7.65466297322253e-06, "loss": 0.1329, "step": 5080 }, { "epoch": 1.1749769159741459, "grad_norm": 0.42850738763809204, "learning_rate": 7.65004616805171e-06, "loss": 0.1352, "step": 5090 }, { "epoch": 1.1772853185595569, "grad_norm": 4.224897861480713, "learning_rate": 7.645429362880887e-06, "loss": 0.0963, "step": 5100 }, { "epoch": 1.1795937211449676, "grad_norm": 0.03889668732881546, "learning_rate": 7.640812557710066e-06, "loss": 0.2824, "step": 5110 }, { "epoch": 1.1819021237303786, "grad_norm": 1.9942296743392944, "learning_rate": 7.636195752539244e-06, "loss": 0.143, "step": 5120 }, { "epoch": 1.1842105263157894, "grad_norm": 0.30719640851020813, "learning_rate": 7.631578947368423e-06, "loss": 0.0256, "step": 5130 }, { "epoch": 1.1865189289012004, "grad_norm": 7.15642786026001, "learning_rate": 7.6269621421976e-06, "loss": 0.1109, "step": 5140 }, { "epoch": 1.1888273314866113, "grad_norm": 0.5269545912742615, "learning_rate": 7.622345337026778e-06, "loss": 0.0589, "step": 5150 }, { "epoch": 1.1911357340720221, "grad_norm": 0.32469239830970764, "learning_rate": 7.617728531855957e-06, "loss": 0.1345, "step": 5160 }, { "epoch": 1.193444136657433, "grad_norm": 10.661999702453613, "learning_rate": 7.613111726685134e-06, "loss": 0.0929, "step": 5170 }, { "epoch": 1.1957525392428439, "grad_norm": 0.7502566576004028, "learning_rate": 7.608494921514312e-06, "loss": 0.0939, "step": 5180 }, { "epoch": 1.1980609418282548, "grad_norm": 17.891868591308594, "learning_rate": 7.6038781163434905e-06, "loss": 0.1671, "step": 5190 }, { "epoch": 1.2003693444136658, "grad_norm": 0.01886000484228134, "learning_rate": 7.599261311172669e-06, "loss": 0.1278, "step": 5200 }, { "epoch": 1.2026777469990766, "grad_norm": 10.954153060913086, "learning_rate": 7.594644506001847e-06, "loss": 0.1743, "step": 5210 }, { "epoch": 1.2049861495844876, "grad_norm": 0.1088632345199585, "learning_rate": 7.590027700831025e-06, "loss": 0.1036, "step": 5220 }, { "epoch": 1.2072945521698983, "grad_norm": 0.2393483966588974, "learning_rate": 7.585410895660203e-06, "loss": 0.1614, "step": 5230 }, { "epoch": 1.2096029547553093, "grad_norm": 0.4398743808269501, "learning_rate": 7.580794090489382e-06, "loss": 0.1082, "step": 5240 }, { "epoch": 1.2119113573407203, "grad_norm": 10.58081340789795, "learning_rate": 7.57617728531856e-06, "loss": 0.1787, "step": 5250 }, { "epoch": 1.214219759926131, "grad_norm": 0.6137261986732483, "learning_rate": 7.571560480147738e-06, "loss": 0.0998, "step": 5260 }, { "epoch": 1.216528162511542, "grad_norm": 1.791608214378357, "learning_rate": 7.566943674976916e-06, "loss": 0.2145, "step": 5270 }, { "epoch": 1.2188365650969528, "grad_norm": 0.24770161509513855, "learning_rate": 7.5623268698060945e-06, "loss": 0.1488, "step": 5280 }, { "epoch": 1.2211449676823638, "grad_norm": 0.0609627440571785, "learning_rate": 7.557710064635273e-06, "loss": 0.1372, "step": 5290 }, { "epoch": 1.2234533702677748, "grad_norm": 8.085432052612305, "learning_rate": 7.553093259464451e-06, "loss": 0.1522, "step": 5300 }, { "epoch": 1.2257617728531855, "grad_norm": 2.8689088821411133, "learning_rate": 7.548476454293629e-06, "loss": 0.225, "step": 5310 }, { "epoch": 1.2280701754385965, "grad_norm": 5.723049640655518, "learning_rate": 7.5438596491228074e-06, "loss": 0.1176, "step": 5320 }, { "epoch": 1.2303785780240073, "grad_norm": 3.165386915206909, "learning_rate": 7.539242843951986e-06, "loss": 0.1326, "step": 5330 }, { "epoch": 1.2326869806094183, "grad_norm": 3.740962266921997, "learning_rate": 7.534626038781164e-06, "loss": 0.1327, "step": 5340 }, { "epoch": 1.2349953831948293, "grad_norm": 2.3535711765289307, "learning_rate": 7.530009233610342e-06, "loss": 0.1848, "step": 5350 }, { "epoch": 1.23730378578024, "grad_norm": 7.17512321472168, "learning_rate": 7.525392428439521e-06, "loss": 0.2253, "step": 5360 }, { "epoch": 1.239612188365651, "grad_norm": 4.354602813720703, "learning_rate": 7.5207756232686994e-06, "loss": 0.1347, "step": 5370 }, { "epoch": 1.2419205909510618, "grad_norm": 4.1893310546875, "learning_rate": 7.516158818097878e-06, "loss": 0.1292, "step": 5380 }, { "epoch": 1.2442289935364728, "grad_norm": 3.6551003456115723, "learning_rate": 7.511542012927056e-06, "loss": 0.2606, "step": 5390 }, { "epoch": 1.2465373961218837, "grad_norm": 7.738318920135498, "learning_rate": 7.506925207756234e-06, "loss": 0.1563, "step": 5400 }, { "epoch": 1.2488457987072945, "grad_norm": 6.813594818115234, "learning_rate": 7.502308402585412e-06, "loss": 0.2221, "step": 5410 }, { "epoch": 1.2511542012927055, "grad_norm": 7.761618614196777, "learning_rate": 7.497691597414589e-06, "loss": 0.1801, "step": 5420 }, { "epoch": 1.2534626038781163, "grad_norm": 0.7811102867126465, "learning_rate": 7.493074792243767e-06, "loss": 0.1256, "step": 5430 }, { "epoch": 1.2557710064635272, "grad_norm": 0.5396248698234558, "learning_rate": 7.488457987072945e-06, "loss": 0.0679, "step": 5440 }, { "epoch": 1.2580794090489382, "grad_norm": 0.18834218382835388, "learning_rate": 7.4838411819021236e-06, "loss": 0.1547, "step": 5450 }, { "epoch": 1.260387811634349, "grad_norm": 11.427350044250488, "learning_rate": 7.479224376731302e-06, "loss": 0.16, "step": 5460 }, { "epoch": 1.26269621421976, "grad_norm": 7.803902626037598, "learning_rate": 7.47460757156048e-06, "loss": 0.135, "step": 5470 }, { "epoch": 1.2650046168051707, "grad_norm": 8.828819274902344, "learning_rate": 7.469990766389659e-06, "loss": 0.1224, "step": 5480 }, { "epoch": 1.2673130193905817, "grad_norm": 7.974379062652588, "learning_rate": 7.465373961218837e-06, "loss": 0.2412, "step": 5490 }, { "epoch": 1.2696214219759927, "grad_norm": 15.236988067626953, "learning_rate": 7.4607571560480156e-06, "loss": 0.1278, "step": 5500 }, { "epoch": 1.2719298245614035, "grad_norm": 2.727228879928589, "learning_rate": 7.456140350877194e-06, "loss": 0.1386, "step": 5510 }, { "epoch": 1.2742382271468145, "grad_norm": 0.8179554343223572, "learning_rate": 7.451523545706372e-06, "loss": 0.1788, "step": 5520 }, { "epoch": 1.2765466297322252, "grad_norm": 4.186853408813477, "learning_rate": 7.44690674053555e-06, "loss": 0.2543, "step": 5530 }, { "epoch": 1.2788550323176362, "grad_norm": 1.9172505140304565, "learning_rate": 7.4422899353647285e-06, "loss": 0.1484, "step": 5540 }, { "epoch": 1.2811634349030472, "grad_norm": 4.140894412994385, "learning_rate": 7.437673130193907e-06, "loss": 0.2277, "step": 5550 }, { "epoch": 1.283471837488458, "grad_norm": 5.878515720367432, "learning_rate": 7.433056325023085e-06, "loss": 0.1114, "step": 5560 }, { "epoch": 1.285780240073869, "grad_norm": 9.318262100219727, "learning_rate": 7.428439519852263e-06, "loss": 0.1451, "step": 5570 }, { "epoch": 1.2880886426592797, "grad_norm": 5.865315914154053, "learning_rate": 7.423822714681441e-06, "loss": 0.141, "step": 5580 }, { "epoch": 1.2903970452446907, "grad_norm": 6.645530700683594, "learning_rate": 7.41920590951062e-06, "loss": 0.1839, "step": 5590 }, { "epoch": 1.2927054478301017, "grad_norm": 3.454763889312744, "learning_rate": 7.414589104339798e-06, "loss": 0.1257, "step": 5600 }, { "epoch": 1.2950138504155124, "grad_norm": 4.323970794677734, "learning_rate": 7.409972299168976e-06, "loss": 0.1573, "step": 5610 }, { "epoch": 1.2973222530009234, "grad_norm": 3.3374648094177246, "learning_rate": 7.405355493998154e-06, "loss": 0.1147, "step": 5620 }, { "epoch": 1.2996306555863342, "grad_norm": 5.09335470199585, "learning_rate": 7.4007386888273325e-06, "loss": 0.085, "step": 5630 }, { "epoch": 1.3019390581717452, "grad_norm": 1.1174424886703491, "learning_rate": 7.396121883656511e-06, "loss": 0.1635, "step": 5640 }, { "epoch": 1.3042474607571561, "grad_norm": 0.8146535754203796, "learning_rate": 7.391505078485689e-06, "loss": 0.2071, "step": 5650 }, { "epoch": 1.306555863342567, "grad_norm": 2.6559276580810547, "learning_rate": 7.386888273314867e-06, "loss": 0.1236, "step": 5660 }, { "epoch": 1.3088642659279779, "grad_norm": 0.05767223238945007, "learning_rate": 7.382271468144045e-06, "loss": 0.1795, "step": 5670 }, { "epoch": 1.3111726685133887, "grad_norm": 0.780188798904419, "learning_rate": 7.377654662973223e-06, "loss": 0.1938, "step": 5680 }, { "epoch": 1.3134810710987996, "grad_norm": 7.373915672302246, "learning_rate": 7.373037857802401e-06, "loss": 0.2673, "step": 5690 }, { "epoch": 1.3157894736842106, "grad_norm": 2.371856451034546, "learning_rate": 7.368421052631579e-06, "loss": 0.1697, "step": 5700 }, { "epoch": 1.3180978762696214, "grad_norm": 5.01724910736084, "learning_rate": 7.3638042474607575e-06, "loss": 0.1427, "step": 5710 }, { "epoch": 1.3204062788550324, "grad_norm": 0.5977950096130371, "learning_rate": 7.359187442289936e-06, "loss": 0.1391, "step": 5720 }, { "epoch": 1.3227146814404431, "grad_norm": 9.537864685058594, "learning_rate": 7.354570637119114e-06, "loss": 0.201, "step": 5730 }, { "epoch": 1.3250230840258541, "grad_norm": 0.461699903011322, "learning_rate": 7.349953831948292e-06, "loss": 0.0619, "step": 5740 }, { "epoch": 1.327331486611265, "grad_norm": 2.652895927429199, "learning_rate": 7.3453370267774704e-06, "loss": 0.0556, "step": 5750 }, { "epoch": 1.3296398891966759, "grad_norm": 1.1949410438537598, "learning_rate": 7.340720221606649e-06, "loss": 0.0996, "step": 5760 }, { "epoch": 1.3319482917820868, "grad_norm": 2.2707009315490723, "learning_rate": 7.336103416435827e-06, "loss": 0.0924, "step": 5770 }, { "epoch": 1.3342566943674976, "grad_norm": 0.1404481828212738, "learning_rate": 7.331486611265005e-06, "loss": 0.0643, "step": 5780 }, { "epoch": 1.3365650969529086, "grad_norm": 0.007576497737318277, "learning_rate": 7.326869806094183e-06, "loss": 0.0911, "step": 5790 }, { "epoch": 1.3388734995383196, "grad_norm": 3.610656976699829, "learning_rate": 7.3222530009233616e-06, "loss": 0.2311, "step": 5800 }, { "epoch": 1.3411819021237303, "grad_norm": 5.810819149017334, "learning_rate": 7.31763619575254e-06, "loss": 0.0358, "step": 5810 }, { "epoch": 1.3434903047091413, "grad_norm": 1.4628088474273682, "learning_rate": 7.313019390581718e-06, "loss": 0.0531, "step": 5820 }, { "epoch": 1.345798707294552, "grad_norm": 0.9402265548706055, "learning_rate": 7.308402585410896e-06, "loss": 0.1453, "step": 5830 }, { "epoch": 1.348107109879963, "grad_norm": 4.542391777038574, "learning_rate": 7.3037857802400745e-06, "loss": 0.2416, "step": 5840 }, { "epoch": 1.350415512465374, "grad_norm": 2.8139114379882812, "learning_rate": 7.299168975069253e-06, "loss": 0.0568, "step": 5850 }, { "epoch": 1.3527239150507848, "grad_norm": 0.10611671209335327, "learning_rate": 7.294552169898431e-06, "loss": 0.1641, "step": 5860 }, { "epoch": 1.3550323176361958, "grad_norm": 5.248083114624023, "learning_rate": 7.289935364727609e-06, "loss": 0.0362, "step": 5870 }, { "epoch": 1.3573407202216066, "grad_norm": 2.2302982807159424, "learning_rate": 7.285318559556787e-06, "loss": 0.1584, "step": 5880 }, { "epoch": 1.3596491228070176, "grad_norm": 0.8745562434196472, "learning_rate": 7.280701754385966e-06, "loss": 0.2016, "step": 5890 }, { "epoch": 1.3619575253924285, "grad_norm": 13.594534873962402, "learning_rate": 7.276084949215144e-06, "loss": 0.2086, "step": 5900 }, { "epoch": 1.3642659279778393, "grad_norm": 0.24438448250293732, "learning_rate": 7.271468144044322e-06, "loss": 0.1955, "step": 5910 }, { "epoch": 1.3665743305632503, "grad_norm": 1.7107964754104614, "learning_rate": 7.2668513388734995e-06, "loss": 0.1891, "step": 5920 }, { "epoch": 1.368882733148661, "grad_norm": 2.745070219039917, "learning_rate": 7.262234533702678e-06, "loss": 0.0686, "step": 5930 }, { "epoch": 1.371191135734072, "grad_norm": 15.58464527130127, "learning_rate": 7.257617728531856e-06, "loss": 0.1608, "step": 5940 }, { "epoch": 1.373499538319483, "grad_norm": 2.7004234790802, "learning_rate": 7.253000923361034e-06, "loss": 0.1288, "step": 5950 }, { "epoch": 1.3758079409048938, "grad_norm": 4.858201026916504, "learning_rate": 7.248384118190212e-06, "loss": 0.2206, "step": 5960 }, { "epoch": 1.3781163434903048, "grad_norm": 9.247483253479004, "learning_rate": 7.243767313019391e-06, "loss": 0.1422, "step": 5970 }, { "epoch": 1.3804247460757155, "grad_norm": 2.872438430786133, "learning_rate": 7.239150507848569e-06, "loss": 0.0947, "step": 5980 }, { "epoch": 1.3827331486611265, "grad_norm": 0.6808979511260986, "learning_rate": 7.234533702677747e-06, "loss": 0.1601, "step": 5990 }, { "epoch": 1.3850415512465375, "grad_norm": 6.9575605392456055, "learning_rate": 7.229916897506925e-06, "loss": 0.2808, "step": 6000 }, { "epoch": 1.3873499538319483, "grad_norm": 10.154241561889648, "learning_rate": 7.2253000923361035e-06, "loss": 0.2266, "step": 6010 }, { "epoch": 1.3896583564173592, "grad_norm": 0.5315317511558533, "learning_rate": 7.220683287165282e-06, "loss": 0.2006, "step": 6020 }, { "epoch": 1.39196675900277, "grad_norm": 7.316532135009766, "learning_rate": 7.21606648199446e-06, "loss": 0.135, "step": 6030 }, { "epoch": 1.394275161588181, "grad_norm": 0.09342466294765472, "learning_rate": 7.211449676823638e-06, "loss": 0.1131, "step": 6040 }, { "epoch": 1.396583564173592, "grad_norm": 7.437924861907959, "learning_rate": 7.206832871652817e-06, "loss": 0.0911, "step": 6050 }, { "epoch": 1.3988919667590027, "grad_norm": 3.494609832763672, "learning_rate": 7.2022160664819955e-06, "loss": 0.1099, "step": 6060 }, { "epoch": 1.4012003693444137, "grad_norm": 12.25583267211914, "learning_rate": 7.197599261311174e-06, "loss": 0.2117, "step": 6070 }, { "epoch": 1.4035087719298245, "grad_norm": 0.3987880349159241, "learning_rate": 7.192982456140352e-06, "loss": 0.092, "step": 6080 }, { "epoch": 1.4058171745152355, "grad_norm": 7.968601226806641, "learning_rate": 7.18836565096953e-06, "loss": 0.1579, "step": 6090 }, { "epoch": 1.4081255771006465, "grad_norm": 4.037213325500488, "learning_rate": 7.1837488457987084e-06, "loss": 0.1286, "step": 6100 }, { "epoch": 1.4104339796860572, "grad_norm": 10.96886920928955, "learning_rate": 7.179132040627887e-06, "loss": 0.2313, "step": 6110 }, { "epoch": 1.4127423822714682, "grad_norm": 7.142239093780518, "learning_rate": 7.174515235457065e-06, "loss": 0.1446, "step": 6120 }, { "epoch": 1.415050784856879, "grad_norm": 1.1031750440597534, "learning_rate": 7.169898430286243e-06, "loss": 0.1272, "step": 6130 }, { "epoch": 1.41735918744229, "grad_norm": 2.233212471008301, "learning_rate": 7.165281625115421e-06, "loss": 0.0607, "step": 6140 }, { "epoch": 1.419667590027701, "grad_norm": 0.06565926969051361, "learning_rate": 7.1606648199445996e-06, "loss": 0.1306, "step": 6150 }, { "epoch": 1.4219759926131117, "grad_norm": 0.7157612442970276, "learning_rate": 7.156048014773778e-06, "loss": 0.036, "step": 6160 }, { "epoch": 1.4242843951985227, "grad_norm": 1.5251933336257935, "learning_rate": 7.151431209602956e-06, "loss": 0.0743, "step": 6170 }, { "epoch": 1.4265927977839334, "grad_norm": 4.516425132751465, "learning_rate": 7.146814404432133e-06, "loss": 0.2198, "step": 6180 }, { "epoch": 1.4289012003693444, "grad_norm": 9.113761901855469, "learning_rate": 7.142197599261312e-06, "loss": 0.1955, "step": 6190 }, { "epoch": 1.4312096029547554, "grad_norm": 6.974668502807617, "learning_rate": 7.13758079409049e-06, "loss": 0.1184, "step": 6200 }, { "epoch": 1.4335180055401662, "grad_norm": 3.2737364768981934, "learning_rate": 7.132963988919668e-06, "loss": 0.0753, "step": 6210 }, { "epoch": 1.4358264081255772, "grad_norm": 4.3788533210754395, "learning_rate": 7.128347183748846e-06, "loss": 0.0932, "step": 6220 }, { "epoch": 1.438134810710988, "grad_norm": 0.5225698947906494, "learning_rate": 7.1237303785780246e-06, "loss": 0.0939, "step": 6230 }, { "epoch": 1.440443213296399, "grad_norm": 13.48166275024414, "learning_rate": 7.119113573407203e-06, "loss": 0.3673, "step": 6240 }, { "epoch": 1.44275161588181, "grad_norm": 9.31886100769043, "learning_rate": 7.114496768236381e-06, "loss": 0.0867, "step": 6250 }, { "epoch": 1.4450600184672207, "grad_norm": 3.050671339035034, "learning_rate": 7.109879963065559e-06, "loss": 0.1645, "step": 6260 }, { "epoch": 1.4473684210526316, "grad_norm": 5.100831508636475, "learning_rate": 7.1052631578947375e-06, "loss": 0.2498, "step": 6270 }, { "epoch": 1.4496768236380424, "grad_norm": 8.201932907104492, "learning_rate": 7.100646352723916e-06, "loss": 0.073, "step": 6280 }, { "epoch": 1.4519852262234534, "grad_norm": 8.057510375976562, "learning_rate": 7.096029547553094e-06, "loss": 0.0483, "step": 6290 }, { "epoch": 1.4542936288088644, "grad_norm": 3.353121042251587, "learning_rate": 7.091412742382272e-06, "loss": 0.219, "step": 6300 }, { "epoch": 1.4566020313942751, "grad_norm": 2.3961400985717773, "learning_rate": 7.08679593721145e-06, "loss": 0.2361, "step": 6310 }, { "epoch": 1.4589104339796861, "grad_norm": 6.363582134246826, "learning_rate": 7.082179132040629e-06, "loss": 0.0731, "step": 6320 }, { "epoch": 1.4612188365650969, "grad_norm": 0.40576833486557007, "learning_rate": 7.077562326869807e-06, "loss": 0.1059, "step": 6330 }, { "epoch": 1.4635272391505079, "grad_norm": 5.753564357757568, "learning_rate": 7.072945521698985e-06, "loss": 0.105, "step": 6340 }, { "epoch": 1.4658356417359188, "grad_norm": 1.83617103099823, "learning_rate": 7.068328716528163e-06, "loss": 0.3554, "step": 6350 }, { "epoch": 1.4681440443213296, "grad_norm": 6.0678486824035645, "learning_rate": 7.0637119113573415e-06, "loss": 0.2172, "step": 6360 }, { "epoch": 1.4704524469067406, "grad_norm": 5.073874473571777, "learning_rate": 7.05909510618652e-06, "loss": 0.1529, "step": 6370 }, { "epoch": 1.4727608494921514, "grad_norm": 6.467820644378662, "learning_rate": 7.054478301015698e-06, "loss": 0.0919, "step": 6380 }, { "epoch": 1.4750692520775623, "grad_norm": 10.857748031616211, "learning_rate": 7.049861495844876e-06, "loss": 0.1794, "step": 6390 }, { "epoch": 1.4773776546629733, "grad_norm": 0.9127181172370911, "learning_rate": 7.0452446906740544e-06, "loss": 0.0725, "step": 6400 }, { "epoch": 1.479686057248384, "grad_norm": 11.431780815124512, "learning_rate": 7.040627885503233e-06, "loss": 0.2027, "step": 6410 }, { "epoch": 1.481994459833795, "grad_norm": 0.48204249143600464, "learning_rate": 7.036011080332411e-06, "loss": 0.106, "step": 6420 }, { "epoch": 1.4843028624192058, "grad_norm": 1.6743664741516113, "learning_rate": 7.031394275161588e-06, "loss": 0.1177, "step": 6430 }, { "epoch": 1.4866112650046168, "grad_norm": 1.391472578048706, "learning_rate": 7.0267774699907665e-06, "loss": 0.2619, "step": 6440 }, { "epoch": 1.4889196675900278, "grad_norm": 0.07073758542537689, "learning_rate": 7.022160664819945e-06, "loss": 0.1124, "step": 6450 }, { "epoch": 1.4912280701754386, "grad_norm": 0.7767134308815002, "learning_rate": 7.017543859649123e-06, "loss": 0.0935, "step": 6460 }, { "epoch": 1.4935364727608496, "grad_norm": 0.04671125113964081, "learning_rate": 7.012927054478301e-06, "loss": 0.0869, "step": 6470 }, { "epoch": 1.4958448753462603, "grad_norm": 4.7529730796813965, "learning_rate": 7.008310249307479e-06, "loss": 0.1813, "step": 6480 }, { "epoch": 1.4981532779316713, "grad_norm": 5.929255962371826, "learning_rate": 7.003693444136658e-06, "loss": 0.1045, "step": 6490 }, { "epoch": 1.5004616805170823, "grad_norm": 0.16102801263332367, "learning_rate": 6.999076638965836e-06, "loss": 0.1911, "step": 6500 }, { "epoch": 1.502770083102493, "grad_norm": 2.3684637546539307, "learning_rate": 6.994459833795014e-06, "loss": 0.0597, "step": 6510 }, { "epoch": 1.5050784856879038, "grad_norm": 2.377720832824707, "learning_rate": 6.989843028624192e-06, "loss": 0.2462, "step": 6520 }, { "epoch": 1.5073868882733148, "grad_norm": 9.177262306213379, "learning_rate": 6.9852262234533706e-06, "loss": 0.2552, "step": 6530 }, { "epoch": 1.5096952908587258, "grad_norm": 6.268819808959961, "learning_rate": 6.980609418282549e-06, "loss": 0.1443, "step": 6540 }, { "epoch": 1.5120036934441368, "grad_norm": 6.041132926940918, "learning_rate": 6.975992613111727e-06, "loss": 0.1145, "step": 6550 }, { "epoch": 1.5143120960295475, "grad_norm": 0.26514285802841187, "learning_rate": 6.971375807940905e-06, "loss": 0.0139, "step": 6560 }, { "epoch": 1.5166204986149583, "grad_norm": 6.405760288238525, "learning_rate": 6.9667590027700835e-06, "loss": 0.0787, "step": 6570 }, { "epoch": 1.5189289012003693, "grad_norm": 1.9073222875595093, "learning_rate": 6.962142197599262e-06, "loss": 0.1712, "step": 6580 }, { "epoch": 1.5212373037857803, "grad_norm": 3.0030696392059326, "learning_rate": 6.95752539242844e-06, "loss": 0.129, "step": 6590 }, { "epoch": 1.5235457063711912, "grad_norm": 2.5851778984069824, "learning_rate": 6.952908587257618e-06, "loss": 0.2646, "step": 6600 }, { "epoch": 1.525854108956602, "grad_norm": 4.750157356262207, "learning_rate": 6.948291782086796e-06, "loss": 0.0622, "step": 6610 }, { "epoch": 1.5281625115420128, "grad_norm": 0.045969005674123764, "learning_rate": 6.943674976915975e-06, "loss": 0.1446, "step": 6620 }, { "epoch": 1.5304709141274238, "grad_norm": 4.824373245239258, "learning_rate": 6.939058171745154e-06, "loss": 0.2411, "step": 6630 }, { "epoch": 1.5327793167128347, "grad_norm": 1.8297144174575806, "learning_rate": 6.934441366574332e-06, "loss": 0.1691, "step": 6640 }, { "epoch": 1.5350877192982457, "grad_norm": 1.1578670740127563, "learning_rate": 6.92982456140351e-06, "loss": 0.1804, "step": 6650 }, { "epoch": 1.5373961218836565, "grad_norm": 2.3993418216705322, "learning_rate": 6.925207756232688e-06, "loss": 0.0822, "step": 6660 }, { "epoch": 1.5397045244690672, "grad_norm": 9.510886192321777, "learning_rate": 6.920590951061867e-06, "loss": 0.2206, "step": 6670 }, { "epoch": 1.5420129270544782, "grad_norm": 8.793068885803223, "learning_rate": 6.915974145891043e-06, "loss": 0.2275, "step": 6680 }, { "epoch": 1.5443213296398892, "grad_norm": 6.262901306152344, "learning_rate": 6.911357340720221e-06, "loss": 0.1047, "step": 6690 }, { "epoch": 1.5466297322253002, "grad_norm": 1.0575183629989624, "learning_rate": 6.9067405355494e-06, "loss": 0.167, "step": 6700 }, { "epoch": 1.548938134810711, "grad_norm": 0.05064845085144043, "learning_rate": 6.902123730378578e-06, "loss": 0.1211, "step": 6710 }, { "epoch": 1.5512465373961217, "grad_norm": 0.10975093394517899, "learning_rate": 6.897506925207756e-06, "loss": 0.1064, "step": 6720 }, { "epoch": 1.5535549399815327, "grad_norm": 2.0717873573303223, "learning_rate": 6.892890120036934e-06, "loss": 0.0193, "step": 6730 }, { "epoch": 1.5558633425669437, "grad_norm": 2.670248508453369, "learning_rate": 6.888273314866113e-06, "loss": 0.1196, "step": 6740 }, { "epoch": 1.5581717451523547, "grad_norm": 5.354759216308594, "learning_rate": 6.883656509695292e-06, "loss": 0.0519, "step": 6750 }, { "epoch": 1.5604801477377654, "grad_norm": 0.5386488437652588, "learning_rate": 6.87903970452447e-06, "loss": 0.0826, "step": 6760 }, { "epoch": 1.5627885503231762, "grad_norm": 0.0072530172765254974, "learning_rate": 6.874422899353648e-06, "loss": 0.14, "step": 6770 }, { "epoch": 1.5650969529085872, "grad_norm": 2.1837263107299805, "learning_rate": 6.869806094182826e-06, "loss": 0.1318, "step": 6780 }, { "epoch": 1.5674053554939982, "grad_norm": 0.02525436133146286, "learning_rate": 6.8651892890120045e-06, "loss": 0.2601, "step": 6790 }, { "epoch": 1.5697137580794092, "grad_norm": 2.096130609512329, "learning_rate": 6.860572483841183e-06, "loss": 0.2065, "step": 6800 }, { "epoch": 1.57202216066482, "grad_norm": 0.0018772504990920424, "learning_rate": 6.855955678670361e-06, "loss": 0.117, "step": 6810 }, { "epoch": 1.5743305632502307, "grad_norm": 1.6277235746383667, "learning_rate": 6.851338873499539e-06, "loss": 0.1531, "step": 6820 }, { "epoch": 1.5766389658356417, "grad_norm": 6.536577224731445, "learning_rate": 6.8467220683287174e-06, "loss": 0.194, "step": 6830 }, { "epoch": 1.5789473684210527, "grad_norm": 2.6445486545562744, "learning_rate": 6.842105263157896e-06, "loss": 0.2106, "step": 6840 }, { "epoch": 1.5812557710064636, "grad_norm": 5.757327079772949, "learning_rate": 6.837488457987074e-06, "loss": 0.0925, "step": 6850 }, { "epoch": 1.5835641735918744, "grad_norm": 8.323981285095215, "learning_rate": 6.832871652816252e-06, "loss": 0.1539, "step": 6860 }, { "epoch": 1.5858725761772852, "grad_norm": 5.712255954742432, "learning_rate": 6.82825484764543e-06, "loss": 0.1121, "step": 6870 }, { "epoch": 1.5881809787626961, "grad_norm": 0.8227988481521606, "learning_rate": 6.8236380424746086e-06, "loss": 0.0887, "step": 6880 }, { "epoch": 1.5904893813481071, "grad_norm": 1.1026835441589355, "learning_rate": 6.819021237303787e-06, "loss": 0.2628, "step": 6890 }, { "epoch": 1.5927977839335181, "grad_norm": 0.9900234341621399, "learning_rate": 6.814404432132965e-06, "loss": 0.1345, "step": 6900 }, { "epoch": 1.5951061865189289, "grad_norm": 3.250154733657837, "learning_rate": 6.809787626962143e-06, "loss": 0.2296, "step": 6910 }, { "epoch": 1.5974145891043396, "grad_norm": 6.97826623916626, "learning_rate": 6.8051708217913215e-06, "loss": 0.0896, "step": 6920 }, { "epoch": 1.5997229916897506, "grad_norm": 12.831768035888672, "learning_rate": 6.800554016620499e-06, "loss": 0.1729, "step": 6930 }, { "epoch": 1.6020313942751616, "grad_norm": 6.787016868591309, "learning_rate": 6.795937211449677e-06, "loss": 0.3281, "step": 6940 }, { "epoch": 1.6043397968605726, "grad_norm": 4.8044891357421875, "learning_rate": 6.791320406278855e-06, "loss": 0.1209, "step": 6950 }, { "epoch": 1.6066481994459834, "grad_norm": 0.15000203251838684, "learning_rate": 6.7867036011080335e-06, "loss": 0.0964, "step": 6960 }, { "epoch": 1.6089566020313941, "grad_norm": 6.843861103057861, "learning_rate": 6.782086795937212e-06, "loss": 0.1262, "step": 6970 }, { "epoch": 1.611265004616805, "grad_norm": 5.8809380531311035, "learning_rate": 6.77746999076639e-06, "loss": 0.1646, "step": 6980 }, { "epoch": 1.613573407202216, "grad_norm": 8.259763717651367, "learning_rate": 6.772853185595568e-06, "loss": 0.1607, "step": 6990 }, { "epoch": 1.615881809787627, "grad_norm": 8.264200210571289, "learning_rate": 6.7682363804247465e-06, "loss": 0.1167, "step": 7000 }, { "epoch": 1.6181902123730378, "grad_norm": 4.43766975402832, "learning_rate": 6.763619575253925e-06, "loss": 0.1631, "step": 7010 }, { "epoch": 1.6204986149584486, "grad_norm": 10.75998306274414, "learning_rate": 6.759002770083103e-06, "loss": 0.1543, "step": 7020 }, { "epoch": 1.6228070175438596, "grad_norm": 0.18856525421142578, "learning_rate": 6.754385964912281e-06, "loss": 0.1572, "step": 7030 }, { "epoch": 1.6251154201292706, "grad_norm": 0.04454103112220764, "learning_rate": 6.749769159741459e-06, "loss": 0.137, "step": 7040 }, { "epoch": 1.6274238227146816, "grad_norm": 3.4214189052581787, "learning_rate": 6.745152354570638e-06, "loss": 0.0829, "step": 7050 }, { "epoch": 1.6297322253000923, "grad_norm": 2.196315050125122, "learning_rate": 6.740535549399816e-06, "loss": 0.1462, "step": 7060 }, { "epoch": 1.632040627885503, "grad_norm": 2.3843271732330322, "learning_rate": 6.735918744228994e-06, "loss": 0.206, "step": 7070 }, { "epoch": 1.634349030470914, "grad_norm": 5.313675403594971, "learning_rate": 6.731301939058172e-06, "loss": 0.1532, "step": 7080 }, { "epoch": 1.636657433056325, "grad_norm": 0.5028116703033447, "learning_rate": 6.7266851338873505e-06, "loss": 0.101, "step": 7090 }, { "epoch": 1.638965835641736, "grad_norm": 0.4982970952987671, "learning_rate": 6.722068328716529e-06, "loss": 0.2086, "step": 7100 }, { "epoch": 1.6412742382271468, "grad_norm": 1.3126336336135864, "learning_rate": 6.717451523545707e-06, "loss": 0.1619, "step": 7110 }, { "epoch": 1.6435826408125576, "grad_norm": 7.027770042419434, "learning_rate": 6.712834718374885e-06, "loss": 0.2548, "step": 7120 }, { "epoch": 1.6458910433979685, "grad_norm": 0.8990707993507385, "learning_rate": 6.7082179132040634e-06, "loss": 0.0769, "step": 7130 }, { "epoch": 1.6481994459833795, "grad_norm": 0.11296223104000092, "learning_rate": 6.703601108033242e-06, "loss": 0.198, "step": 7140 }, { "epoch": 1.6505078485687905, "grad_norm": 0.10602495074272156, "learning_rate": 6.69898430286242e-06, "loss": 0.1268, "step": 7150 }, { "epoch": 1.6528162511542013, "grad_norm": 2.000558614730835, "learning_rate": 6.694367497691598e-06, "loss": 0.1331, "step": 7160 }, { "epoch": 1.655124653739612, "grad_norm": 4.645630359649658, "learning_rate": 6.689750692520776e-06, "loss": 0.1702, "step": 7170 }, { "epoch": 1.657433056325023, "grad_norm": 1.5484449863433838, "learning_rate": 6.685133887349954e-06, "loss": 0.2008, "step": 7180 }, { "epoch": 1.659741458910434, "grad_norm": 0.8385854959487915, "learning_rate": 6.680517082179132e-06, "loss": 0.1722, "step": 7190 }, { "epoch": 1.662049861495845, "grad_norm": 0.3394019305706024, "learning_rate": 6.67590027700831e-06, "loss": 0.0954, "step": 7200 }, { "epoch": 1.6643582640812558, "grad_norm": 7.03579044342041, "learning_rate": 6.671283471837488e-06, "loss": 0.1966, "step": 7210 }, { "epoch": 1.6666666666666665, "grad_norm": 7.899814605712891, "learning_rate": 6.666666666666667e-06, "loss": 0.2173, "step": 7220 }, { "epoch": 1.6689750692520775, "grad_norm": 6.237720966339111, "learning_rate": 6.662049861495845e-06, "loss": 0.1248, "step": 7230 }, { "epoch": 1.6712834718374885, "grad_norm": 2.157731771469116, "learning_rate": 6.657433056325023e-06, "loss": 0.1114, "step": 7240 }, { "epoch": 1.6735918744228995, "grad_norm": 2.4611334800720215, "learning_rate": 6.652816251154201e-06, "loss": 0.1001, "step": 7250 }, { "epoch": 1.6759002770083102, "grad_norm": 3.5052037239074707, "learning_rate": 6.6481994459833796e-06, "loss": 0.0506, "step": 7260 }, { "epoch": 1.678208679593721, "grad_norm": 10.980010986328125, "learning_rate": 6.643582640812558e-06, "loss": 0.0978, "step": 7270 }, { "epoch": 1.680517082179132, "grad_norm": 2.5022213459014893, "learning_rate": 6.638965835641736e-06, "loss": 0.1426, "step": 7280 }, { "epoch": 1.682825484764543, "grad_norm": 0.012976584024727345, "learning_rate": 6.634349030470914e-06, "loss": 0.1531, "step": 7290 }, { "epoch": 1.685133887349954, "grad_norm": 1.8009392023086548, "learning_rate": 6.6297322253000925e-06, "loss": 0.2101, "step": 7300 }, { "epoch": 1.6874422899353647, "grad_norm": 1.5941520929336548, "learning_rate": 6.625115420129271e-06, "loss": 0.1352, "step": 7310 }, { "epoch": 1.6897506925207755, "grad_norm": 5.420823097229004, "learning_rate": 6.62049861495845e-06, "loss": 0.0714, "step": 7320 }, { "epoch": 1.6920590951061865, "grad_norm": 10.23454761505127, "learning_rate": 6.615881809787628e-06, "loss": 0.1828, "step": 7330 }, { "epoch": 1.6943674976915974, "grad_norm": 7.362997531890869, "learning_rate": 6.611265004616806e-06, "loss": 0.1637, "step": 7340 }, { "epoch": 1.6966759002770084, "grad_norm": 7.657659530639648, "learning_rate": 6.6066481994459845e-06, "loss": 0.1133, "step": 7350 }, { "epoch": 1.6989843028624192, "grad_norm": 0.1771342009305954, "learning_rate": 6.602031394275163e-06, "loss": 0.1376, "step": 7360 }, { "epoch": 1.70129270544783, "grad_norm": 13.190537452697754, "learning_rate": 6.597414589104341e-06, "loss": 0.2935, "step": 7370 }, { "epoch": 1.703601108033241, "grad_norm": 1.702020287513733, "learning_rate": 6.592797783933519e-06, "loss": 0.017, "step": 7380 }, { "epoch": 1.705909510618652, "grad_norm": 1.9911080598831177, "learning_rate": 6.588180978762697e-06, "loss": 0.1629, "step": 7390 }, { "epoch": 1.708217913204063, "grad_norm": 7.369743347167969, "learning_rate": 6.583564173591876e-06, "loss": 0.0678, "step": 7400 }, { "epoch": 1.7105263157894737, "grad_norm": 6.85450553894043, "learning_rate": 6.578947368421054e-06, "loss": 0.1496, "step": 7410 }, { "epoch": 1.7128347183748844, "grad_norm": 8.551612854003906, "learning_rate": 6.574330563250232e-06, "loss": 0.1042, "step": 7420 }, { "epoch": 1.7151431209602954, "grad_norm": 2.9744086265563965, "learning_rate": 6.5697137580794094e-06, "loss": 0.233, "step": 7430 }, { "epoch": 1.7174515235457064, "grad_norm": 0.07635639607906342, "learning_rate": 6.565096952908588e-06, "loss": 0.1139, "step": 7440 }, { "epoch": 1.7197599261311174, "grad_norm": 1.504429578781128, "learning_rate": 6.560480147737766e-06, "loss": 0.1059, "step": 7450 }, { "epoch": 1.7220683287165282, "grad_norm": 2.9266772270202637, "learning_rate": 6.555863342566944e-06, "loss": 0.0567, "step": 7460 }, { "epoch": 1.724376731301939, "grad_norm": 3.605612277984619, "learning_rate": 6.551246537396122e-06, "loss": 0.1037, "step": 7470 }, { "epoch": 1.72668513388735, "grad_norm": 0.9540549516677856, "learning_rate": 6.546629732225301e-06, "loss": 0.1119, "step": 7480 }, { "epoch": 1.7289935364727609, "grad_norm": 6.184191703796387, "learning_rate": 6.542012927054479e-06, "loss": 0.1474, "step": 7490 }, { "epoch": 1.7313019390581719, "grad_norm": 8.588491439819336, "learning_rate": 6.537396121883657e-06, "loss": 0.141, "step": 7500 }, { "epoch": 1.7336103416435826, "grad_norm": 1.8794986009597778, "learning_rate": 6.532779316712835e-06, "loss": 0.0944, "step": 7510 }, { "epoch": 1.7359187442289934, "grad_norm": 9.552380561828613, "learning_rate": 6.5281625115420135e-06, "loss": 0.1744, "step": 7520 }, { "epoch": 1.7382271468144044, "grad_norm": 0.387534499168396, "learning_rate": 6.523545706371192e-06, "loss": 0.2037, "step": 7530 }, { "epoch": 1.7405355493998154, "grad_norm": 5.804172515869141, "learning_rate": 6.51892890120037e-06, "loss": 0.1304, "step": 7540 }, { "epoch": 1.7428439519852263, "grad_norm": 0.7988699078559875, "learning_rate": 6.514312096029548e-06, "loss": 0.1473, "step": 7550 }, { "epoch": 1.745152354570637, "grad_norm": 0.7697465419769287, "learning_rate": 6.509695290858726e-06, "loss": 0.1124, "step": 7560 }, { "epoch": 1.7474607571560479, "grad_norm": 7.298343181610107, "learning_rate": 6.505078485687905e-06, "loss": 0.2562, "step": 7570 }, { "epoch": 1.7497691597414589, "grad_norm": 4.50553035736084, "learning_rate": 6.500461680517083e-06, "loss": 0.126, "step": 7580 }, { "epoch": 1.7520775623268698, "grad_norm": 12.981157302856445, "learning_rate": 6.495844875346261e-06, "loss": 0.1195, "step": 7590 }, { "epoch": 1.7543859649122808, "grad_norm": 4.404174327850342, "learning_rate": 6.491228070175439e-06, "loss": 0.1919, "step": 7600 }, { "epoch": 1.7566943674976916, "grad_norm": 0.5157092809677124, "learning_rate": 6.4866112650046176e-06, "loss": 0.0951, "step": 7610 }, { "epoch": 1.7590027700831024, "grad_norm": 0.16650424897670746, "learning_rate": 6.481994459833796e-06, "loss": 0.1248, "step": 7620 }, { "epoch": 1.7613111726685133, "grad_norm": 8.10939884185791, "learning_rate": 6.477377654662974e-06, "loss": 0.0544, "step": 7630 }, { "epoch": 1.7636195752539243, "grad_norm": 2.1119225025177, "learning_rate": 6.472760849492152e-06, "loss": 0.1149, "step": 7640 }, { "epoch": 1.7659279778393353, "grad_norm": 2.523036479949951, "learning_rate": 6.4681440443213305e-06, "loss": 0.1567, "step": 7650 }, { "epoch": 1.768236380424746, "grad_norm": 10.765961647033691, "learning_rate": 6.463527239150509e-06, "loss": 0.1761, "step": 7660 }, { "epoch": 1.7705447830101568, "grad_norm": 0.755062460899353, "learning_rate": 6.458910433979687e-06, "loss": 0.0507, "step": 7670 }, { "epoch": 1.7728531855955678, "grad_norm": 3.490917921066284, "learning_rate": 6.454293628808865e-06, "loss": 0.0596, "step": 7680 }, { "epoch": 1.7751615881809788, "grad_norm": 0.43090829253196716, "learning_rate": 6.4496768236380425e-06, "loss": 0.1474, "step": 7690 }, { "epoch": 1.7774699907663898, "grad_norm": 4.917365550994873, "learning_rate": 6.445060018467221e-06, "loss": 0.0974, "step": 7700 }, { "epoch": 1.7797783933518005, "grad_norm": 10.694178581237793, "learning_rate": 6.440443213296399e-06, "loss": 0.3028, "step": 7710 }, { "epoch": 1.7820867959372113, "grad_norm": 5.929460525512695, "learning_rate": 6.435826408125577e-06, "loss": 0.0749, "step": 7720 }, { "epoch": 1.7843951985226223, "grad_norm": 4.399403095245361, "learning_rate": 6.4312096029547555e-06, "loss": 0.176, "step": 7730 }, { "epoch": 1.7867036011080333, "grad_norm": 0.027071785181760788, "learning_rate": 6.426592797783934e-06, "loss": 0.131, "step": 7740 }, { "epoch": 1.7890120036934443, "grad_norm": 8.557962417602539, "learning_rate": 6.421975992613112e-06, "loss": 0.1057, "step": 7750 }, { "epoch": 1.791320406278855, "grad_norm": 0.7661490440368652, "learning_rate": 6.41735918744229e-06, "loss": 0.2144, "step": 7760 }, { "epoch": 1.7936288088642658, "grad_norm": 9.75623607635498, "learning_rate": 6.412742382271468e-06, "loss": 0.1276, "step": 7770 }, { "epoch": 1.7959372114496768, "grad_norm": 6.364738464355469, "learning_rate": 6.408125577100647e-06, "loss": 0.1435, "step": 7780 }, { "epoch": 1.7982456140350878, "grad_norm": 1.4771772623062134, "learning_rate": 6.403508771929825e-06, "loss": 0.0973, "step": 7790 }, { "epoch": 1.8005540166204987, "grad_norm": 5.833428382873535, "learning_rate": 6.398891966759003e-06, "loss": 0.1377, "step": 7800 }, { "epoch": 1.8028624192059095, "grad_norm": 5.6276140213012695, "learning_rate": 6.394275161588181e-06, "loss": 0.0747, "step": 7810 }, { "epoch": 1.8051708217913203, "grad_norm": 6.839344501495361, "learning_rate": 6.3896583564173595e-06, "loss": 0.2604, "step": 7820 }, { "epoch": 1.8074792243767313, "grad_norm": 0.39363521337509155, "learning_rate": 6.385041551246538e-06, "loss": 0.1624, "step": 7830 }, { "epoch": 1.8097876269621422, "grad_norm": 5.885039329528809, "learning_rate": 6.380424746075716e-06, "loss": 0.1182, "step": 7840 }, { "epoch": 1.8120960295475532, "grad_norm": 3.342357873916626, "learning_rate": 6.375807940904894e-06, "loss": 0.1371, "step": 7850 }, { "epoch": 1.814404432132964, "grad_norm": 0.826147735118866, "learning_rate": 6.3711911357340724e-06, "loss": 0.0978, "step": 7860 }, { "epoch": 1.8167128347183747, "grad_norm": 0.12890218198299408, "learning_rate": 6.366574330563251e-06, "loss": 0.0874, "step": 7870 }, { "epoch": 1.8190212373037857, "grad_norm": 3.2980127334594727, "learning_rate": 6.361957525392429e-06, "loss": 0.1183, "step": 7880 }, { "epoch": 1.8213296398891967, "grad_norm": 0.6489532589912415, "learning_rate": 6.357340720221607e-06, "loss": 0.0617, "step": 7890 }, { "epoch": 1.8236380424746077, "grad_norm": 0.6646090149879456, "learning_rate": 6.352723915050786e-06, "loss": 0.2681, "step": 7900 }, { "epoch": 1.8259464450600185, "grad_norm": 4.692173957824707, "learning_rate": 6.348107109879964e-06, "loss": 0.1338, "step": 7910 }, { "epoch": 1.8282548476454292, "grad_norm": 3.6965296268463135, "learning_rate": 6.343490304709143e-06, "loss": 0.1116, "step": 7920 }, { "epoch": 1.8305632502308402, "grad_norm": 6.4236907958984375, "learning_rate": 6.338873499538321e-06, "loss": 0.1017, "step": 7930 }, { "epoch": 1.8328716528162512, "grad_norm": 7.079874515533447, "learning_rate": 6.334256694367497e-06, "loss": 0.1418, "step": 7940 }, { "epoch": 1.8351800554016622, "grad_norm": 8.222419738769531, "learning_rate": 6.329639889196676e-06, "loss": 0.1577, "step": 7950 }, { "epoch": 1.837488457987073, "grad_norm": 8.236457824707031, "learning_rate": 6.325023084025854e-06, "loss": 0.2812, "step": 7960 }, { "epoch": 1.8397968605724837, "grad_norm": 7.066323280334473, "learning_rate": 6.320406278855032e-06, "loss": 0.1444, "step": 7970 }, { "epoch": 1.8421052631578947, "grad_norm": 4.454392433166504, "learning_rate": 6.31578947368421e-06, "loss": 0.0752, "step": 7980 }, { "epoch": 1.8444136657433057, "grad_norm": 8.278393745422363, "learning_rate": 6.3111726685133886e-06, "loss": 0.1849, "step": 7990 }, { "epoch": 1.8467220683287167, "grad_norm": 0.8757348656654358, "learning_rate": 6.306555863342567e-06, "loss": 0.0798, "step": 8000 }, { "epoch": 1.8490304709141274, "grad_norm": 5.9545183181762695, "learning_rate": 6.301939058171746e-06, "loss": 0.1675, "step": 8010 }, { "epoch": 1.8513388734995382, "grad_norm": 19.9125919342041, "learning_rate": 6.297322253000924e-06, "loss": 0.1884, "step": 8020 }, { "epoch": 1.8536472760849492, "grad_norm": 9.5575532913208, "learning_rate": 6.292705447830102e-06, "loss": 0.1534, "step": 8030 }, { "epoch": 1.8559556786703602, "grad_norm": 1.6205804347991943, "learning_rate": 6.2880886426592805e-06, "loss": 0.0776, "step": 8040 }, { "epoch": 1.8582640812557711, "grad_norm": 15.89978313446045, "learning_rate": 6.283471837488459e-06, "loss": 0.0875, "step": 8050 }, { "epoch": 1.860572483841182, "grad_norm": 10.861461639404297, "learning_rate": 6.278855032317637e-06, "loss": 0.1827, "step": 8060 }, { "epoch": 1.8628808864265927, "grad_norm": 1.8603498935699463, "learning_rate": 6.274238227146815e-06, "loss": 0.0956, "step": 8070 }, { "epoch": 1.8651892890120036, "grad_norm": 0.018299419432878494, "learning_rate": 6.2696214219759935e-06, "loss": 0.0195, "step": 8080 }, { "epoch": 1.8674976915974146, "grad_norm": 1.9744303226470947, "learning_rate": 6.265004616805172e-06, "loss": 0.2294, "step": 8090 }, { "epoch": 1.8698060941828256, "grad_norm": 0.05369226261973381, "learning_rate": 6.26038781163435e-06, "loss": 0.1401, "step": 8100 }, { "epoch": 1.8721144967682364, "grad_norm": 0.937160849571228, "learning_rate": 6.255771006463528e-06, "loss": 0.052, "step": 8110 }, { "epoch": 1.8744228993536471, "grad_norm": 8.414374351501465, "learning_rate": 6.251154201292706e-06, "loss": 0.1244, "step": 8120 }, { "epoch": 1.8767313019390581, "grad_norm": 0.0035069389268755913, "learning_rate": 6.246537396121885e-06, "loss": 0.1082, "step": 8130 }, { "epoch": 1.8790397045244691, "grad_norm": 8.114951133728027, "learning_rate": 6.241920590951063e-06, "loss": 0.1264, "step": 8140 }, { "epoch": 1.88134810710988, "grad_norm": 0.08601807057857513, "learning_rate": 6.237303785780241e-06, "loss": 0.135, "step": 8150 }, { "epoch": 1.8836565096952909, "grad_norm": 0.6841493844985962, "learning_rate": 6.232686980609419e-06, "loss": 0.1232, "step": 8160 }, { "epoch": 1.8859649122807016, "grad_norm": 0.34537649154663086, "learning_rate": 6.2280701754385975e-06, "loss": 0.0829, "step": 8170 }, { "epoch": 1.8882733148661126, "grad_norm": 8.30687141418457, "learning_rate": 6.223453370267776e-06, "loss": 0.0796, "step": 8180 }, { "epoch": 1.8905817174515236, "grad_norm": 9.79694938659668, "learning_rate": 6.218836565096953e-06, "loss": 0.1882, "step": 8190 }, { "epoch": 1.8928901200369346, "grad_norm": 6.208591461181641, "learning_rate": 6.214219759926131e-06, "loss": 0.2588, "step": 8200 }, { "epoch": 1.8951985226223453, "grad_norm": 2.372145891189575, "learning_rate": 6.20960295475531e-06, "loss": 0.1919, "step": 8210 }, { "epoch": 1.897506925207756, "grad_norm": 8.687252044677734, "learning_rate": 6.204986149584488e-06, "loss": 0.2623, "step": 8220 }, { "epoch": 1.899815327793167, "grad_norm": 6.273497581481934, "learning_rate": 6.200369344413666e-06, "loss": 0.1165, "step": 8230 }, { "epoch": 1.902123730378578, "grad_norm": 2.975783109664917, "learning_rate": 6.195752539242844e-06, "loss": 0.0985, "step": 8240 }, { "epoch": 1.904432132963989, "grad_norm": 3.606961965560913, "learning_rate": 6.1911357340720225e-06, "loss": 0.0989, "step": 8250 }, { "epoch": 1.9067405355493998, "grad_norm": 0.389692097902298, "learning_rate": 6.186518928901201e-06, "loss": 0.1624, "step": 8260 }, { "epoch": 1.9090489381348106, "grad_norm": 8.54810905456543, "learning_rate": 6.181902123730379e-06, "loss": 0.1967, "step": 8270 }, { "epoch": 1.9113573407202216, "grad_norm": 3.235846996307373, "learning_rate": 6.177285318559557e-06, "loss": 0.0824, "step": 8280 }, { "epoch": 1.9136657433056325, "grad_norm": 0.16667799651622772, "learning_rate": 6.172668513388735e-06, "loss": 0.0534, "step": 8290 }, { "epoch": 1.9159741458910435, "grad_norm": 0.10937054455280304, "learning_rate": 6.168051708217914e-06, "loss": 0.089, "step": 8300 }, { "epoch": 1.9182825484764543, "grad_norm": 1.9354497194290161, "learning_rate": 6.163434903047092e-06, "loss": 0.1156, "step": 8310 }, { "epoch": 1.920590951061865, "grad_norm": 0.8759872317314148, "learning_rate": 6.15881809787627e-06, "loss": 0.3605, "step": 8320 }, { "epoch": 1.922899353647276, "grad_norm": 6.817061901092529, "learning_rate": 6.154201292705448e-06, "loss": 0.1679, "step": 8330 }, { "epoch": 1.925207756232687, "grad_norm": 0.8531666398048401, "learning_rate": 6.1495844875346266e-06, "loss": 0.2253, "step": 8340 }, { "epoch": 1.927516158818098, "grad_norm": 0.42465031147003174, "learning_rate": 6.144967682363805e-06, "loss": 0.0866, "step": 8350 }, { "epoch": 1.9298245614035088, "grad_norm": 1.4570600986480713, "learning_rate": 6.140350877192983e-06, "loss": 0.0814, "step": 8360 }, { "epoch": 1.9321329639889195, "grad_norm": 0.46014854311943054, "learning_rate": 6.135734072022161e-06, "loss": 0.1907, "step": 8370 }, { "epoch": 1.9344413665743305, "grad_norm": 1.4643794298171997, "learning_rate": 6.1311172668513395e-06, "loss": 0.2251, "step": 8380 }, { "epoch": 1.9367497691597415, "grad_norm": 8.7191743850708, "learning_rate": 6.126500461680518e-06, "loss": 0.1714, "step": 8390 }, { "epoch": 1.9390581717451525, "grad_norm": 0.6021118760108948, "learning_rate": 6.121883656509696e-06, "loss": 0.1401, "step": 8400 }, { "epoch": 1.9413665743305633, "grad_norm": 0.47443175315856934, "learning_rate": 6.117266851338874e-06, "loss": 0.2086, "step": 8410 }, { "epoch": 1.943674976915974, "grad_norm": 4.75976037979126, "learning_rate": 6.112650046168052e-06, "loss": 0.1009, "step": 8420 }, { "epoch": 1.945983379501385, "grad_norm": 5.859301567077637, "learning_rate": 6.108033240997231e-06, "loss": 0.0853, "step": 8430 }, { "epoch": 1.948291782086796, "grad_norm": 6.5929083824157715, "learning_rate": 6.103416435826408e-06, "loss": 0.0887, "step": 8440 }, { "epoch": 1.950600184672207, "grad_norm": 10.656343460083008, "learning_rate": 6.098799630655586e-06, "loss": 0.2523, "step": 8450 }, { "epoch": 1.9529085872576177, "grad_norm": 3.823875665664673, "learning_rate": 6.0941828254847645e-06, "loss": 0.2889, "step": 8460 }, { "epoch": 1.9552169898430285, "grad_norm": 1.346490740776062, "learning_rate": 6.089566020313943e-06, "loss": 0.1102, "step": 8470 }, { "epoch": 1.9575253924284395, "grad_norm": 1.3046754598617554, "learning_rate": 6.084949215143121e-06, "loss": 0.2301, "step": 8480 }, { "epoch": 1.9598337950138505, "grad_norm": 0.03322243317961693, "learning_rate": 6.080332409972299e-06, "loss": 0.0365, "step": 8490 }, { "epoch": 1.9621421975992615, "grad_norm": 1.0472577810287476, "learning_rate": 6.075715604801477e-06, "loss": 0.153, "step": 8500 }, { "epoch": 1.9644506001846722, "grad_norm": 5.539668083190918, "learning_rate": 6.071098799630656e-06, "loss": 0.057, "step": 8510 }, { "epoch": 1.966759002770083, "grad_norm": 1.9032143354415894, "learning_rate": 6.066481994459834e-06, "loss": 0.2137, "step": 8520 }, { "epoch": 1.969067405355494, "grad_norm": 6.5532307624816895, "learning_rate": 6.061865189289012e-06, "loss": 0.2677, "step": 8530 }, { "epoch": 1.971375807940905, "grad_norm": 5.149965286254883, "learning_rate": 6.05724838411819e-06, "loss": 0.1398, "step": 8540 }, { "epoch": 1.973684210526316, "grad_norm": 7.169659614562988, "learning_rate": 6.0526315789473685e-06, "loss": 0.1636, "step": 8550 }, { "epoch": 1.9759926131117267, "grad_norm": 6.840461730957031, "learning_rate": 6.048014773776547e-06, "loss": 0.2315, "step": 8560 }, { "epoch": 1.9783010156971375, "grad_norm": 1.6913082599639893, "learning_rate": 6.043397968605725e-06, "loss": 0.0966, "step": 8570 }, { "epoch": 1.9806094182825484, "grad_norm": 2.4741079807281494, "learning_rate": 6.038781163434903e-06, "loss": 0.1515, "step": 8580 }, { "epoch": 1.9829178208679594, "grad_norm": 4.566931247711182, "learning_rate": 6.034164358264082e-06, "loss": 0.1615, "step": 8590 }, { "epoch": 1.9852262234533704, "grad_norm": 0.8300578594207764, "learning_rate": 6.0295475530932605e-06, "loss": 0.1022, "step": 8600 }, { "epoch": 1.9875346260387812, "grad_norm": 0.03683464601635933, "learning_rate": 6.024930747922439e-06, "loss": 0.1052, "step": 8610 }, { "epoch": 1.989843028624192, "grad_norm": 4.96811580657959, "learning_rate": 6.020313942751617e-06, "loss": 0.2101, "step": 8620 }, { "epoch": 1.992151431209603, "grad_norm": 9.449173927307129, "learning_rate": 6.015697137580795e-06, "loss": 0.3838, "step": 8630 }, { "epoch": 1.994459833795014, "grad_norm": 5.175396919250488, "learning_rate": 6.011080332409973e-06, "loss": 0.2618, "step": 8640 }, { "epoch": 1.9967682363804249, "grad_norm": 9.836182594299316, "learning_rate": 6.006463527239152e-06, "loss": 0.196, "step": 8650 }, { "epoch": 1.9990766389658357, "grad_norm": 4.681766510009766, "learning_rate": 6.00184672206833e-06, "loss": 0.1649, "step": 8660 }, { "epoch": 2.0013850415512464, "grad_norm": 2.927532196044922, "learning_rate": 5.997229916897508e-06, "loss": 0.1959, "step": 8670 }, { "epoch": 2.0036934441366574, "grad_norm": 0.05056557059288025, "learning_rate": 5.992613111726686e-06, "loss": 0.1221, "step": 8680 }, { "epoch": 2.0060018467220684, "grad_norm": 0.10122841596603394, "learning_rate": 5.987996306555863e-06, "loss": 0.0872, "step": 8690 }, { "epoch": 2.0083102493074794, "grad_norm": 5.571578025817871, "learning_rate": 5.983379501385042e-06, "loss": 0.0557, "step": 8700 }, { "epoch": 2.01061865189289, "grad_norm": 1.3636723756790161, "learning_rate": 5.97876269621422e-06, "loss": 0.0712, "step": 8710 }, { "epoch": 2.012927054478301, "grad_norm": 3.220374822616577, "learning_rate": 5.974145891043398e-06, "loss": 0.0405, "step": 8720 }, { "epoch": 2.015235457063712, "grad_norm": 8.186657905578613, "learning_rate": 5.969529085872577e-06, "loss": 0.1472, "step": 8730 }, { "epoch": 2.017543859649123, "grad_norm": 1.016586184501648, "learning_rate": 5.964912280701755e-06, "loss": 0.0584, "step": 8740 }, { "epoch": 2.019852262234534, "grad_norm": 0.9086069464683533, "learning_rate": 5.960295475530933e-06, "loss": 0.1755, "step": 8750 }, { "epoch": 2.0221606648199444, "grad_norm": 0.016934020444750786, "learning_rate": 5.955678670360111e-06, "loss": 0.2406, "step": 8760 }, { "epoch": 2.0244690674053554, "grad_norm": 0.6315302848815918, "learning_rate": 5.9510618651892895e-06, "loss": 0.1454, "step": 8770 }, { "epoch": 2.0267774699907664, "grad_norm": 3.5935792922973633, "learning_rate": 5.946445060018468e-06, "loss": 0.1484, "step": 8780 }, { "epoch": 2.0290858725761773, "grad_norm": 6.510880470275879, "learning_rate": 5.941828254847646e-06, "loss": 0.1997, "step": 8790 }, { "epoch": 2.0313942751615883, "grad_norm": 2.0462422370910645, "learning_rate": 5.937211449676824e-06, "loss": 0.0329, "step": 8800 }, { "epoch": 2.033702677746999, "grad_norm": 0.016499172896146774, "learning_rate": 5.9325946445060025e-06, "loss": 0.0978, "step": 8810 }, { "epoch": 2.03601108033241, "grad_norm": 0.08158121258020401, "learning_rate": 5.927977839335181e-06, "loss": 0.2987, "step": 8820 }, { "epoch": 2.038319482917821, "grad_norm": 1.0147842168807983, "learning_rate": 5.923361034164359e-06, "loss": 0.0758, "step": 8830 }, { "epoch": 2.040627885503232, "grad_norm": 1.7109171152114868, "learning_rate": 5.918744228993537e-06, "loss": 0.14, "step": 8840 }, { "epoch": 2.042936288088643, "grad_norm": 0.7527329921722412, "learning_rate": 5.914127423822715e-06, "loss": 0.0826, "step": 8850 }, { "epoch": 2.0452446906740533, "grad_norm": 0.3407728374004364, "learning_rate": 5.909510618651894e-06, "loss": 0.0815, "step": 8860 }, { "epoch": 2.0475530932594643, "grad_norm": 3.8781118392944336, "learning_rate": 5.904893813481072e-06, "loss": 0.1201, "step": 8870 }, { "epoch": 2.0498614958448753, "grad_norm": 0.057001981884241104, "learning_rate": 5.90027700831025e-06, "loss": 0.0481, "step": 8880 }, { "epoch": 2.0521698984302863, "grad_norm": 4.004337787628174, "learning_rate": 5.895660203139428e-06, "loss": 0.1359, "step": 8890 }, { "epoch": 2.0544783010156973, "grad_norm": 0.3880602717399597, "learning_rate": 5.8910433979686065e-06, "loss": 0.128, "step": 8900 }, { "epoch": 2.056786703601108, "grad_norm": 0.04380026459693909, "learning_rate": 5.886426592797785e-06, "loss": 0.0731, "step": 8910 }, { "epoch": 2.059095106186519, "grad_norm": 0.022117508575320244, "learning_rate": 5.881809787626963e-06, "loss": 0.1556, "step": 8920 }, { "epoch": 2.06140350877193, "grad_norm": 4.597250938415527, "learning_rate": 5.877192982456141e-06, "loss": 0.114, "step": 8930 }, { "epoch": 2.063711911357341, "grad_norm": 0.033978935331106186, "learning_rate": 5.8725761772853194e-06, "loss": 0.0575, "step": 8940 }, { "epoch": 2.0660203139427518, "grad_norm": 6.101284980773926, "learning_rate": 5.867959372114497e-06, "loss": 0.0505, "step": 8950 }, { "epoch": 2.0683287165281623, "grad_norm": 0.05929476395249367, "learning_rate": 5.863342566943675e-06, "loss": 0.0402, "step": 8960 }, { "epoch": 2.0706371191135733, "grad_norm": 5.941567420959473, "learning_rate": 5.858725761772853e-06, "loss": 0.1594, "step": 8970 }, { "epoch": 2.0729455216989843, "grad_norm": 8.183211326599121, "learning_rate": 5.8541089566020315e-06, "loss": 0.0838, "step": 8980 }, { "epoch": 2.0752539242843953, "grad_norm": 0.09150294959545135, "learning_rate": 5.84949215143121e-06, "loss": 0.1315, "step": 8990 }, { "epoch": 2.0775623268698062, "grad_norm": 9.495864868164062, "learning_rate": 5.844875346260388e-06, "loss": 0.1332, "step": 9000 }, { "epoch": 2.079870729455217, "grad_norm": 1.3700872659683228, "learning_rate": 5.840258541089566e-06, "loss": 0.1399, "step": 9010 }, { "epoch": 2.0821791320406278, "grad_norm": 6.932931423187256, "learning_rate": 5.835641735918744e-06, "loss": 0.1696, "step": 9020 }, { "epoch": 2.0844875346260388, "grad_norm": 4.011542320251465, "learning_rate": 5.831024930747923e-06, "loss": 0.1356, "step": 9030 }, { "epoch": 2.0867959372114497, "grad_norm": 10.626606941223145, "learning_rate": 5.826408125577101e-06, "loss": 0.1291, "step": 9040 }, { "epoch": 2.0891043397968607, "grad_norm": 7.218416213989258, "learning_rate": 5.821791320406279e-06, "loss": 0.0941, "step": 9050 }, { "epoch": 2.0914127423822713, "grad_norm": 0.42429065704345703, "learning_rate": 5.817174515235457e-06, "loss": 0.2245, "step": 9060 }, { "epoch": 2.0937211449676822, "grad_norm": 0.7302764058113098, "learning_rate": 5.8125577100646356e-06, "loss": 0.1242, "step": 9070 }, { "epoch": 2.0960295475530932, "grad_norm": 7.304187774658203, "learning_rate": 5.807940904893814e-06, "loss": 0.0707, "step": 9080 }, { "epoch": 2.098337950138504, "grad_norm": 0.15638557076454163, "learning_rate": 5.803324099722992e-06, "loss": 0.0206, "step": 9090 }, { "epoch": 2.100646352723915, "grad_norm": 0.28185614943504333, "learning_rate": 5.79870729455217e-06, "loss": 0.1083, "step": 9100 }, { "epoch": 2.1029547553093257, "grad_norm": 14.079293251037598, "learning_rate": 5.7940904893813485e-06, "loss": 0.3238, "step": 9110 }, { "epoch": 2.1052631578947367, "grad_norm": 2.9056591987609863, "learning_rate": 5.789473684210527e-06, "loss": 0.1397, "step": 9120 }, { "epoch": 2.1075715604801477, "grad_norm": 6.2419843673706055, "learning_rate": 5.784856879039705e-06, "loss": 0.0863, "step": 9130 }, { "epoch": 2.1098799630655587, "grad_norm": 2.238790273666382, "learning_rate": 5.780240073868883e-06, "loss": 0.1536, "step": 9140 }, { "epoch": 2.1121883656509697, "grad_norm": 0.2927044630050659, "learning_rate": 5.775623268698061e-06, "loss": 0.0902, "step": 9150 }, { "epoch": 2.11449676823638, "grad_norm": 4.019207954406738, "learning_rate": 5.77100646352724e-06, "loss": 0.1232, "step": 9160 }, { "epoch": 2.116805170821791, "grad_norm": 7.978782653808594, "learning_rate": 5.766389658356419e-06, "loss": 0.1065, "step": 9170 }, { "epoch": 2.119113573407202, "grad_norm": 2.3987929821014404, "learning_rate": 5.761772853185597e-06, "loss": 0.1332, "step": 9180 }, { "epoch": 2.121421975992613, "grad_norm": 5.822368144989014, "learning_rate": 5.757156048014775e-06, "loss": 0.1453, "step": 9190 }, { "epoch": 2.123730378578024, "grad_norm": 2.1533074378967285, "learning_rate": 5.752539242843952e-06, "loss": 0.1478, "step": 9200 }, { "epoch": 2.1260387811634347, "grad_norm": 0.6697908043861389, "learning_rate": 5.74792243767313e-06, "loss": 0.0964, "step": 9210 }, { "epoch": 2.1283471837488457, "grad_norm": 3.2779664993286133, "learning_rate": 5.743305632502308e-06, "loss": 0.0917, "step": 9220 }, { "epoch": 2.1306555863342567, "grad_norm": 0.6165647506713867, "learning_rate": 5.738688827331486e-06, "loss": 0.1052, "step": 9230 }, { "epoch": 2.1329639889196677, "grad_norm": 5.503302574157715, "learning_rate": 5.734072022160665e-06, "loss": 0.0448, "step": 9240 }, { "epoch": 2.1352723915050786, "grad_norm": 7.877943992614746, "learning_rate": 5.729455216989843e-06, "loss": 0.339, "step": 9250 }, { "epoch": 2.137580794090489, "grad_norm": 8.723716735839844, "learning_rate": 5.724838411819021e-06, "loss": 0.1391, "step": 9260 }, { "epoch": 2.1398891966759, "grad_norm": 0.8852521777153015, "learning_rate": 5.720221606648199e-06, "loss": 0.1099, "step": 9270 }, { "epoch": 2.142197599261311, "grad_norm": 1.1055375337600708, "learning_rate": 5.715604801477378e-06, "loss": 0.0919, "step": 9280 }, { "epoch": 2.144506001846722, "grad_norm": 2.017138719558716, "learning_rate": 5.710987996306557e-06, "loss": 0.096, "step": 9290 }, { "epoch": 2.146814404432133, "grad_norm": 8.649641990661621, "learning_rate": 5.706371191135735e-06, "loss": 0.093, "step": 9300 }, { "epoch": 2.1491228070175437, "grad_norm": 7.006896495819092, "learning_rate": 5.701754385964913e-06, "loss": 0.1401, "step": 9310 }, { "epoch": 2.1514312096029546, "grad_norm": 7.177443981170654, "learning_rate": 5.697137580794091e-06, "loss": 0.1195, "step": 9320 }, { "epoch": 2.1537396121883656, "grad_norm": 1.8384499549865723, "learning_rate": 5.6925207756232695e-06, "loss": 0.1036, "step": 9330 }, { "epoch": 2.1560480147737766, "grad_norm": 6.480041027069092, "learning_rate": 5.687903970452448e-06, "loss": 0.1999, "step": 9340 }, { "epoch": 2.1583564173591876, "grad_norm": 11.267412185668945, "learning_rate": 5.683287165281626e-06, "loss": 0.1152, "step": 9350 }, { "epoch": 2.160664819944598, "grad_norm": 0.8537270426750183, "learning_rate": 5.678670360110804e-06, "loss": 0.1377, "step": 9360 }, { "epoch": 2.162973222530009, "grad_norm": 0.18044432997703552, "learning_rate": 5.674053554939982e-06, "loss": 0.11, "step": 9370 }, { "epoch": 2.16528162511542, "grad_norm": 9.148309707641602, "learning_rate": 5.669436749769161e-06, "loss": 0.0663, "step": 9380 }, { "epoch": 2.167590027700831, "grad_norm": 1.588027000427246, "learning_rate": 5.664819944598339e-06, "loss": 0.1553, "step": 9390 }, { "epoch": 2.169898430286242, "grad_norm": 4.9665656089782715, "learning_rate": 5.660203139427517e-06, "loss": 0.1243, "step": 9400 }, { "epoch": 2.1722068328716526, "grad_norm": 9.686516761779785, "learning_rate": 5.655586334256695e-06, "loss": 0.0922, "step": 9410 }, { "epoch": 2.1745152354570636, "grad_norm": 1.307399868965149, "learning_rate": 5.6509695290858736e-06, "loss": 0.1356, "step": 9420 }, { "epoch": 2.1768236380424746, "grad_norm": 0.22611260414123535, "learning_rate": 5.646352723915052e-06, "loss": 0.1403, "step": 9430 }, { "epoch": 2.1791320406278856, "grad_norm": 0.4205482304096222, "learning_rate": 5.64173591874423e-06, "loss": 0.0477, "step": 9440 }, { "epoch": 2.1814404432132966, "grad_norm": 0.18651635944843292, "learning_rate": 5.637119113573407e-06, "loss": 0.184, "step": 9450 }, { "epoch": 2.183748845798707, "grad_norm": 1.2028234004974365, "learning_rate": 5.632502308402586e-06, "loss": 0.2016, "step": 9460 }, { "epoch": 2.186057248384118, "grad_norm": 0.3975241184234619, "learning_rate": 5.627885503231764e-06, "loss": 0.0962, "step": 9470 }, { "epoch": 2.188365650969529, "grad_norm": 1.0354328155517578, "learning_rate": 5.623268698060942e-06, "loss": 0.1062, "step": 9480 }, { "epoch": 2.19067405355494, "grad_norm": 1.9186370372772217, "learning_rate": 5.61865189289012e-06, "loss": 0.177, "step": 9490 }, { "epoch": 2.192982456140351, "grad_norm": 0.0028579889331012964, "learning_rate": 5.6140350877192985e-06, "loss": 0.151, "step": 9500 }, { "epoch": 2.1952908587257616, "grad_norm": 0.3903660178184509, "learning_rate": 5.609418282548477e-06, "loss": 0.1244, "step": 9510 }, { "epoch": 2.1975992613111726, "grad_norm": 7.859496593475342, "learning_rate": 5.604801477377655e-06, "loss": 0.1241, "step": 9520 }, { "epoch": 2.1999076638965835, "grad_norm": 6.646348476409912, "learning_rate": 5.600184672206833e-06, "loss": 0.0771, "step": 9530 }, { "epoch": 2.2022160664819945, "grad_norm": 1.8306361436843872, "learning_rate": 5.5955678670360115e-06, "loss": 0.1297, "step": 9540 }, { "epoch": 2.2045244690674055, "grad_norm": 3.601691484451294, "learning_rate": 5.59095106186519e-06, "loss": 0.0734, "step": 9550 }, { "epoch": 2.206832871652816, "grad_norm": 4.080760955810547, "learning_rate": 5.586334256694368e-06, "loss": 0.1103, "step": 9560 }, { "epoch": 2.209141274238227, "grad_norm": 0.003511419054120779, "learning_rate": 5.581717451523546e-06, "loss": 0.1649, "step": 9570 }, { "epoch": 2.211449676823638, "grad_norm": 5.900886535644531, "learning_rate": 5.577100646352724e-06, "loss": 0.0621, "step": 9580 }, { "epoch": 2.213758079409049, "grad_norm": 10.087888717651367, "learning_rate": 5.572483841181903e-06, "loss": 0.1529, "step": 9590 }, { "epoch": 2.21606648199446, "grad_norm": 4.802870750427246, "learning_rate": 5.567867036011081e-06, "loss": 0.2889, "step": 9600 }, { "epoch": 2.2183748845798705, "grad_norm": 0.5660249590873718, "learning_rate": 5.563250230840259e-06, "loss": 0.0746, "step": 9610 }, { "epoch": 2.2206832871652815, "grad_norm": 0.02569517120718956, "learning_rate": 5.558633425669437e-06, "loss": 0.2151, "step": 9620 }, { "epoch": 2.2229916897506925, "grad_norm": 0.36746013164520264, "learning_rate": 5.5540166204986155e-06, "loss": 0.0417, "step": 9630 }, { "epoch": 2.2253000923361035, "grad_norm": 0.27406227588653564, "learning_rate": 5.549399815327794e-06, "loss": 0.0808, "step": 9640 }, { "epoch": 2.2276084949215145, "grad_norm": 4.953248023986816, "learning_rate": 5.544783010156972e-06, "loss": 0.0811, "step": 9650 }, { "epoch": 2.229916897506925, "grad_norm": 1.3286144733428955, "learning_rate": 5.54016620498615e-06, "loss": 0.1202, "step": 9660 }, { "epoch": 2.232225300092336, "grad_norm": 8.907193183898926, "learning_rate": 5.5355493998153284e-06, "loss": 0.1912, "step": 9670 }, { "epoch": 2.234533702677747, "grad_norm": 3.4373843669891357, "learning_rate": 5.530932594644507e-06, "loss": 0.1474, "step": 9680 }, { "epoch": 2.236842105263158, "grad_norm": 3.297705888748169, "learning_rate": 5.526315789473685e-06, "loss": 0.2656, "step": 9690 }, { "epoch": 2.239150507848569, "grad_norm": 9.207077980041504, "learning_rate": 5.521698984302862e-06, "loss": 0.0875, "step": 9700 }, { "epoch": 2.2414589104339795, "grad_norm": 3.8544671535491943, "learning_rate": 5.5170821791320405e-06, "loss": 0.1239, "step": 9710 }, { "epoch": 2.2437673130193905, "grad_norm": 3.8043951988220215, "learning_rate": 5.512465373961219e-06, "loss": 0.1111, "step": 9720 }, { "epoch": 2.2460757156048015, "grad_norm": 0.03035220317542553, "learning_rate": 5.507848568790397e-06, "loss": 0.1158, "step": 9730 }, { "epoch": 2.2483841181902124, "grad_norm": 0.02507336065173149, "learning_rate": 5.503231763619575e-06, "loss": 0.0818, "step": 9740 }, { "epoch": 2.2506925207756234, "grad_norm": 4.862039089202881, "learning_rate": 5.498614958448753e-06, "loss": 0.0325, "step": 9750 }, { "epoch": 2.253000923361034, "grad_norm": 9.605551719665527, "learning_rate": 5.493998153277932e-06, "loss": 0.1031, "step": 9760 }, { "epoch": 2.255309325946445, "grad_norm": 0.0015900577418506145, "learning_rate": 5.48938134810711e-06, "loss": 0.1237, "step": 9770 }, { "epoch": 2.257617728531856, "grad_norm": 5.533511161804199, "learning_rate": 5.484764542936288e-06, "loss": 0.053, "step": 9780 }, { "epoch": 2.259926131117267, "grad_norm": 0.8163972496986389, "learning_rate": 5.480147737765466e-06, "loss": 0.0309, "step": 9790 }, { "epoch": 2.262234533702678, "grad_norm": 1.4856349229812622, "learning_rate": 5.4755309325946446e-06, "loss": 0.091, "step": 9800 }, { "epoch": 2.2645429362880884, "grad_norm": 0.011693392880260944, "learning_rate": 5.470914127423823e-06, "loss": 0.1015, "step": 9810 }, { "epoch": 2.2668513388734994, "grad_norm": 0.07746884971857071, "learning_rate": 5.466297322253001e-06, "loss": 0.0912, "step": 9820 }, { "epoch": 2.2691597414589104, "grad_norm": 5.117270469665527, "learning_rate": 5.461680517082179e-06, "loss": 0.1078, "step": 9830 }, { "epoch": 2.2714681440443214, "grad_norm": 0.1702166646718979, "learning_rate": 5.4570637119113575e-06, "loss": 0.1336, "step": 9840 }, { "epoch": 2.2737765466297324, "grad_norm": 3.8074378967285156, "learning_rate": 5.452446906740536e-06, "loss": 0.0773, "step": 9850 }, { "epoch": 2.276084949215143, "grad_norm": 0.06646876782178879, "learning_rate": 5.447830101569715e-06, "loss": 0.1249, "step": 9860 }, { "epoch": 2.278393351800554, "grad_norm": 1.825014352798462, "learning_rate": 5.443213296398893e-06, "loss": 0.0956, "step": 9870 }, { "epoch": 2.280701754385965, "grad_norm": 5.45120096206665, "learning_rate": 5.438596491228071e-06, "loss": 0.109, "step": 9880 }, { "epoch": 2.283010156971376, "grad_norm": 0.7723276019096375, "learning_rate": 5.4339796860572495e-06, "loss": 0.0286, "step": 9890 }, { "epoch": 2.285318559556787, "grad_norm": 2.1377758979797363, "learning_rate": 5.429362880886428e-06, "loss": 0.1554, "step": 9900 }, { "epoch": 2.2876269621421974, "grad_norm": 14.516889572143555, "learning_rate": 5.424746075715606e-06, "loss": 0.1882, "step": 9910 }, { "epoch": 2.2899353647276084, "grad_norm": 0.36585190892219543, "learning_rate": 5.420129270544784e-06, "loss": 0.1274, "step": 9920 }, { "epoch": 2.2922437673130194, "grad_norm": 9.257481575012207, "learning_rate": 5.415512465373962e-06, "loss": 0.2736, "step": 9930 }, { "epoch": 2.2945521698984304, "grad_norm": 0.12996172904968262, "learning_rate": 5.410895660203141e-06, "loss": 0.0807, "step": 9940 }, { "epoch": 2.2968605724838413, "grad_norm": 9.617722511291504, "learning_rate": 5.406278855032317e-06, "loss": 0.0812, "step": 9950 }, { "epoch": 2.299168975069252, "grad_norm": 15.020936012268066, "learning_rate": 5.401662049861495e-06, "loss": 0.181, "step": 9960 }, { "epoch": 2.301477377654663, "grad_norm": 1.0838297605514526, "learning_rate": 5.3970452446906744e-06, "loss": 0.1372, "step": 9970 }, { "epoch": 2.303785780240074, "grad_norm": 7.529239654541016, "learning_rate": 5.392428439519853e-06, "loss": 0.1077, "step": 9980 }, { "epoch": 2.306094182825485, "grad_norm": 8.316734313964844, "learning_rate": 5.387811634349031e-06, "loss": 0.1192, "step": 9990 }, { "epoch": 2.308402585410896, "grad_norm": 19.527170181274414, "learning_rate": 5.383194829178209e-06, "loss": 0.1068, "step": 10000 }, { "epoch": 2.3107109879963064, "grad_norm": 0.010472903028130531, "learning_rate": 5.378578024007387e-06, "loss": 0.2407, "step": 10010 }, { "epoch": 2.3130193905817173, "grad_norm": 0.011368873529136181, "learning_rate": 5.373961218836566e-06, "loss": 0.2354, "step": 10020 }, { "epoch": 2.3153277931671283, "grad_norm": 2.553821086883545, "learning_rate": 5.369344413665744e-06, "loss": 0.1708, "step": 10030 }, { "epoch": 2.3176361957525393, "grad_norm": 0.07721114903688431, "learning_rate": 5.364727608494922e-06, "loss": 0.1417, "step": 10040 }, { "epoch": 2.3199445983379503, "grad_norm": 2.3833868503570557, "learning_rate": 5.3601108033241e-06, "loss": 0.1596, "step": 10050 }, { "epoch": 2.322253000923361, "grad_norm": 1.267509937286377, "learning_rate": 5.3554939981532785e-06, "loss": 0.1137, "step": 10060 }, { "epoch": 2.324561403508772, "grad_norm": 0.006082138977944851, "learning_rate": 5.350877192982457e-06, "loss": 0.0816, "step": 10070 }, { "epoch": 2.326869806094183, "grad_norm": 1.6724847555160522, "learning_rate": 5.346260387811635e-06, "loss": 0.1174, "step": 10080 }, { "epoch": 2.329178208679594, "grad_norm": 3.9419944286346436, "learning_rate": 5.341643582640813e-06, "loss": 0.0286, "step": 10090 }, { "epoch": 2.331486611265005, "grad_norm": 3.089473009109497, "learning_rate": 5.337026777469991e-06, "loss": 0.0545, "step": 10100 }, { "epoch": 2.3337950138504153, "grad_norm": 7.198444366455078, "learning_rate": 5.33240997229917e-06, "loss": 0.075, "step": 10110 }, { "epoch": 2.3361034164358263, "grad_norm": 1.4369611740112305, "learning_rate": 5.327793167128348e-06, "loss": 0.0721, "step": 10120 }, { "epoch": 2.3384118190212373, "grad_norm": 9.33039379119873, "learning_rate": 5.323176361957526e-06, "loss": 0.1526, "step": 10130 }, { "epoch": 2.3407202216066483, "grad_norm": 0.6000414490699768, "learning_rate": 5.318559556786704e-06, "loss": 0.1733, "step": 10140 }, { "epoch": 2.3430286241920593, "grad_norm": 0.27287518978118896, "learning_rate": 5.3139427516158826e-06, "loss": 0.0913, "step": 10150 }, { "epoch": 2.34533702677747, "grad_norm": 0.010472988709807396, "learning_rate": 5.309325946445061e-06, "loss": 0.1022, "step": 10160 }, { "epoch": 2.347645429362881, "grad_norm": 11.53478717803955, "learning_rate": 5.304709141274239e-06, "loss": 0.2519, "step": 10170 }, { "epoch": 2.3499538319482918, "grad_norm": 3.843574285507202, "learning_rate": 5.300092336103417e-06, "loss": 0.1698, "step": 10180 }, { "epoch": 2.3522622345337028, "grad_norm": 10.747403144836426, "learning_rate": 5.2954755309325955e-06, "loss": 0.1246, "step": 10190 }, { "epoch": 2.3545706371191137, "grad_norm": 2.907240629196167, "learning_rate": 5.290858725761774e-06, "loss": 0.0391, "step": 10200 }, { "epoch": 2.3568790397045243, "grad_norm": 0.05704546347260475, "learning_rate": 5.286241920590951e-06, "loss": 0.06, "step": 10210 }, { "epoch": 2.3591874422899353, "grad_norm": 8.052515983581543, "learning_rate": 5.281625115420129e-06, "loss": 0.1085, "step": 10220 }, { "epoch": 2.3614958448753463, "grad_norm": 6.959723472595215, "learning_rate": 5.2770083102493075e-06, "loss": 0.1959, "step": 10230 }, { "epoch": 2.3638042474607572, "grad_norm": 0.8250983953475952, "learning_rate": 5.272391505078486e-06, "loss": 0.1978, "step": 10240 }, { "epoch": 2.366112650046168, "grad_norm": 0.8531680107116699, "learning_rate": 5.267774699907664e-06, "loss": 0.1688, "step": 10250 }, { "epoch": 2.3684210526315788, "grad_norm": 3.994530439376831, "learning_rate": 5.263157894736842e-06, "loss": 0.1565, "step": 10260 }, { "epoch": 2.3707294552169897, "grad_norm": 5.152223587036133, "learning_rate": 5.2585410895660205e-06, "loss": 0.0608, "step": 10270 }, { "epoch": 2.3730378578024007, "grad_norm": 0.958200991153717, "learning_rate": 5.253924284395199e-06, "loss": 0.0772, "step": 10280 }, { "epoch": 2.3753462603878117, "grad_norm": 12.468168258666992, "learning_rate": 5.249307479224377e-06, "loss": 0.1942, "step": 10290 }, { "epoch": 2.3776546629732227, "grad_norm": 0.5707756876945496, "learning_rate": 5.244690674053555e-06, "loss": 0.0483, "step": 10300 }, { "epoch": 2.3799630655586332, "grad_norm": 7.6824493408203125, "learning_rate": 5.240073868882733e-06, "loss": 0.2506, "step": 10310 }, { "epoch": 2.3822714681440442, "grad_norm": 3.3208401203155518, "learning_rate": 5.235457063711912e-06, "loss": 0.0919, "step": 10320 }, { "epoch": 2.384579870729455, "grad_norm": 18.769147872924805, "learning_rate": 5.23084025854109e-06, "loss": 0.2225, "step": 10330 }, { "epoch": 2.386888273314866, "grad_norm": 0.3650968074798584, "learning_rate": 5.226223453370268e-06, "loss": 0.2222, "step": 10340 }, { "epoch": 2.389196675900277, "grad_norm": 7.498579502105713, "learning_rate": 5.221606648199446e-06, "loss": 0.0518, "step": 10350 }, { "epoch": 2.3915050784856877, "grad_norm": 0.03944002091884613, "learning_rate": 5.2169898430286245e-06, "loss": 0.1112, "step": 10360 }, { "epoch": 2.3938134810710987, "grad_norm": 0.035453423857688904, "learning_rate": 5.212373037857803e-06, "loss": 0.0622, "step": 10370 }, { "epoch": 2.3961218836565097, "grad_norm": 0.359327495098114, "learning_rate": 5.207756232686981e-06, "loss": 0.0757, "step": 10380 }, { "epoch": 2.3984302862419207, "grad_norm": 0.8615443110466003, "learning_rate": 5.203139427516159e-06, "loss": 0.1608, "step": 10390 }, { "epoch": 2.4007386888273317, "grad_norm": 13.809239387512207, "learning_rate": 5.1985226223453374e-06, "loss": 0.1539, "step": 10400 }, { "epoch": 2.403047091412742, "grad_norm": 2.998382329940796, "learning_rate": 5.193905817174516e-06, "loss": 0.0656, "step": 10410 }, { "epoch": 2.405355493998153, "grad_norm": 1.6741342544555664, "learning_rate": 5.189289012003694e-06, "loss": 0.0862, "step": 10420 }, { "epoch": 2.407663896583564, "grad_norm": 1.6313860416412354, "learning_rate": 5.184672206832873e-06, "loss": 0.0716, "step": 10430 }, { "epoch": 2.409972299168975, "grad_norm": 4.321824550628662, "learning_rate": 5.180055401662051e-06, "loss": 0.0553, "step": 10440 }, { "epoch": 2.412280701754386, "grad_norm": 1.8620705604553223, "learning_rate": 5.175438596491229e-06, "loss": 0.0516, "step": 10450 }, { "epoch": 2.4145891043397967, "grad_norm": 1.8643032312393188, "learning_rate": 5.170821791320406e-06, "loss": 0.1259, "step": 10460 }, { "epoch": 2.4168975069252077, "grad_norm": 1.228969693183899, "learning_rate": 5.166204986149584e-06, "loss": 0.0642, "step": 10470 }, { "epoch": 2.4192059095106186, "grad_norm": 0.0082811638712883, "learning_rate": 5.161588180978762e-06, "loss": 0.2539, "step": 10480 }, { "epoch": 2.4215143120960296, "grad_norm": 0.0017055664211511612, "learning_rate": 5.156971375807941e-06, "loss": 0.1083, "step": 10490 }, { "epoch": 2.4238227146814406, "grad_norm": 0.024952754378318787, "learning_rate": 5.152354570637119e-06, "loss": 0.162, "step": 10500 }, { "epoch": 2.426131117266851, "grad_norm": 0.05639614537358284, "learning_rate": 5.147737765466297e-06, "loss": 0.1719, "step": 10510 }, { "epoch": 2.428439519852262, "grad_norm": 0.2340765744447708, "learning_rate": 5.143120960295475e-06, "loss": 0.0371, "step": 10520 }, { "epoch": 2.430747922437673, "grad_norm": 0.228525772690773, "learning_rate": 5.1385041551246535e-06, "loss": 0.174, "step": 10530 }, { "epoch": 2.433056325023084, "grad_norm": 0.45449307560920715, "learning_rate": 5.133887349953832e-06, "loss": 0.1605, "step": 10540 }, { "epoch": 2.435364727608495, "grad_norm": 5.817375659942627, "learning_rate": 5.129270544783011e-06, "loss": 0.1704, "step": 10550 }, { "epoch": 2.4376731301939056, "grad_norm": 0.13261659443378448, "learning_rate": 5.124653739612189e-06, "loss": 0.1107, "step": 10560 }, { "epoch": 2.4399815327793166, "grad_norm": 0.17769721150398254, "learning_rate": 5.120036934441367e-06, "loss": 0.2034, "step": 10570 }, { "epoch": 2.4422899353647276, "grad_norm": 10.103926658630371, "learning_rate": 5.1154201292705455e-06, "loss": 0.1, "step": 10580 }, { "epoch": 2.4445983379501386, "grad_norm": 4.391120433807373, "learning_rate": 5.110803324099724e-06, "loss": 0.1804, "step": 10590 }, { "epoch": 2.4469067405355496, "grad_norm": 4.3361616134643555, "learning_rate": 5.106186518928902e-06, "loss": 0.0433, "step": 10600 }, { "epoch": 2.44921514312096, "grad_norm": 7.4146294593811035, "learning_rate": 5.10156971375808e-06, "loss": 0.0698, "step": 10610 }, { "epoch": 2.451523545706371, "grad_norm": 7.45351505279541, "learning_rate": 5.0969529085872585e-06, "loss": 0.1882, "step": 10620 }, { "epoch": 2.453831948291782, "grad_norm": 7.367565631866455, "learning_rate": 5.092336103416437e-06, "loss": 0.1808, "step": 10630 }, { "epoch": 2.456140350877193, "grad_norm": 8.811199188232422, "learning_rate": 5.087719298245615e-06, "loss": 0.1969, "step": 10640 }, { "epoch": 2.458448753462604, "grad_norm": 0.30227410793304443, "learning_rate": 5.083102493074793e-06, "loss": 0.068, "step": 10650 }, { "epoch": 2.4607571560480146, "grad_norm": 0.021973516792058945, "learning_rate": 5.078485687903971e-06, "loss": 0.133, "step": 10660 }, { "epoch": 2.4630655586334256, "grad_norm": 0.20362035930156708, "learning_rate": 5.07386888273315e-06, "loss": 0.1689, "step": 10670 }, { "epoch": 2.4653739612188366, "grad_norm": 8.237716674804688, "learning_rate": 5.069252077562328e-06, "loss": 0.1539, "step": 10680 }, { "epoch": 2.4676823638042475, "grad_norm": 1.113956332206726, "learning_rate": 5.064635272391506e-06, "loss": 0.0879, "step": 10690 }, { "epoch": 2.4699907663896585, "grad_norm": 5.183690071105957, "learning_rate": 5.060018467220684e-06, "loss": 0.1298, "step": 10700 }, { "epoch": 2.472299168975069, "grad_norm": 5.826737403869629, "learning_rate": 5.055401662049862e-06, "loss": 0.1101, "step": 10710 }, { "epoch": 2.47460757156048, "grad_norm": 0.6018840670585632, "learning_rate": 5.05078485687904e-06, "loss": 0.0893, "step": 10720 }, { "epoch": 2.476915974145891, "grad_norm": 1.2729594707489014, "learning_rate": 5.046168051708218e-06, "loss": 0.0419, "step": 10730 }, { "epoch": 2.479224376731302, "grad_norm": 0.5178221464157104, "learning_rate": 5.041551246537396e-06, "loss": 0.1132, "step": 10740 }, { "epoch": 2.481532779316713, "grad_norm": 0.6894585490226746, "learning_rate": 5.036934441366575e-06, "loss": 0.106, "step": 10750 }, { "epoch": 2.4838411819021236, "grad_norm": 10.935760498046875, "learning_rate": 5.032317636195753e-06, "loss": 0.0915, "step": 10760 }, { "epoch": 2.4861495844875345, "grad_norm": 0.35753488540649414, "learning_rate": 5.027700831024931e-06, "loss": 0.1277, "step": 10770 }, { "epoch": 2.4884579870729455, "grad_norm": 5.115323543548584, "learning_rate": 5.023084025854109e-06, "loss": 0.14, "step": 10780 }, { "epoch": 2.4907663896583565, "grad_norm": 3.7961530685424805, "learning_rate": 5.0184672206832875e-06, "loss": 0.1547, "step": 10790 }, { "epoch": 2.4930747922437675, "grad_norm": 0.9842548370361328, "learning_rate": 5.013850415512466e-06, "loss": 0.0896, "step": 10800 }, { "epoch": 2.495383194829178, "grad_norm": 0.001225674874149263, "learning_rate": 5.009233610341644e-06, "loss": 0.088, "step": 10810 }, { "epoch": 2.497691597414589, "grad_norm": 0.20593582093715668, "learning_rate": 5.004616805170822e-06, "loss": 0.1703, "step": 10820 }, { "epoch": 2.5, "grad_norm": 8.60975170135498, "learning_rate": 5e-06, "loss": 0.1198, "step": 10830 }, { "epoch": 2.502308402585411, "grad_norm": 12.95661735534668, "learning_rate": 4.995383194829179e-06, "loss": 0.1492, "step": 10840 }, { "epoch": 2.504616805170822, "grad_norm": 0.016272690147161484, "learning_rate": 4.990766389658357e-06, "loss": 0.0816, "step": 10850 }, { "epoch": 2.5069252077562325, "grad_norm": 4.308964729309082, "learning_rate": 4.986149584487535e-06, "loss": 0.0893, "step": 10860 }, { "epoch": 2.5092336103416435, "grad_norm": 0.2542669177055359, "learning_rate": 4.981532779316713e-06, "loss": 0.054, "step": 10870 }, { "epoch": 2.5115420129270545, "grad_norm": 0.8899867534637451, "learning_rate": 4.9769159741458916e-06, "loss": 0.0916, "step": 10880 }, { "epoch": 2.5138504155124655, "grad_norm": 4.631340980529785, "learning_rate": 4.97229916897507e-06, "loss": 0.1524, "step": 10890 }, { "epoch": 2.5161588180978764, "grad_norm": 2.2988791465759277, "learning_rate": 4.967682363804248e-06, "loss": 0.0474, "step": 10900 }, { "epoch": 2.518467220683287, "grad_norm": 11.104069709777832, "learning_rate": 4.963065558633426e-06, "loss": 0.1028, "step": 10910 }, { "epoch": 2.520775623268698, "grad_norm": 0.3736654818058014, "learning_rate": 4.9584487534626045e-06, "loss": 0.0933, "step": 10920 }, { "epoch": 2.523084025854109, "grad_norm": 7.188614845275879, "learning_rate": 4.953831948291783e-06, "loss": 0.1667, "step": 10930 }, { "epoch": 2.52539242843952, "grad_norm": 0.136319100856781, "learning_rate": 4.949215143120961e-06, "loss": 0.1315, "step": 10940 }, { "epoch": 2.527700831024931, "grad_norm": 9.168624877929688, "learning_rate": 4.944598337950139e-06, "loss": 0.1315, "step": 10950 }, { "epoch": 2.5300092336103415, "grad_norm": 10.19312572479248, "learning_rate": 4.939981532779317e-06, "loss": 0.2534, "step": 10960 }, { "epoch": 2.5323176361957525, "grad_norm": 9.690261840820312, "learning_rate": 4.935364727608496e-06, "loss": 0.1123, "step": 10970 }, { "epoch": 2.5346260387811634, "grad_norm": 0.002922193845734, "learning_rate": 4.930747922437674e-06, "loss": 0.089, "step": 10980 }, { "epoch": 2.5369344413665744, "grad_norm": 7.273950099945068, "learning_rate": 4.926131117266852e-06, "loss": 0.1011, "step": 10990 }, { "epoch": 2.5392428439519854, "grad_norm": 14.686577796936035, "learning_rate": 4.92151431209603e-06, "loss": 0.2818, "step": 11000 }, { "epoch": 2.541551246537396, "grad_norm": 6.286073684692383, "learning_rate": 4.9168975069252085e-06, "loss": 0.1788, "step": 11010 }, { "epoch": 2.543859649122807, "grad_norm": 5.268906593322754, "learning_rate": 4.912280701754386e-06, "loss": 0.2233, "step": 11020 }, { "epoch": 2.546168051708218, "grad_norm": 2.672182559967041, "learning_rate": 4.907663896583564e-06, "loss": 0.171, "step": 11030 }, { "epoch": 2.548476454293629, "grad_norm": 0.9314679503440857, "learning_rate": 4.903047091412742e-06, "loss": 0.06, "step": 11040 }, { "epoch": 2.55078485687904, "grad_norm": 8.238770484924316, "learning_rate": 4.898430286241921e-06, "loss": 0.1555, "step": 11050 }, { "epoch": 2.5530932594644504, "grad_norm": 9.039568901062012, "learning_rate": 4.893813481071099e-06, "loss": 0.2041, "step": 11060 }, { "epoch": 2.5554016620498614, "grad_norm": 0.08301632851362228, "learning_rate": 4.889196675900277e-06, "loss": 0.2525, "step": 11070 }, { "epoch": 2.5577100646352724, "grad_norm": 0.5256550908088684, "learning_rate": 4.884579870729455e-06, "loss": 0.0738, "step": 11080 }, { "epoch": 2.5600184672206834, "grad_norm": 0.0030590256210416555, "learning_rate": 4.8799630655586335e-06, "loss": 0.0867, "step": 11090 }, { "epoch": 2.5623268698060944, "grad_norm": 0.07534092664718628, "learning_rate": 4.875346260387812e-06, "loss": 0.0211, "step": 11100 }, { "epoch": 2.564635272391505, "grad_norm": 0.2707156240940094, "learning_rate": 4.87072945521699e-06, "loss": 0.1175, "step": 11110 }, { "epoch": 2.566943674976916, "grad_norm": 6.710625648498535, "learning_rate": 4.866112650046169e-06, "loss": 0.1548, "step": 11120 }, { "epoch": 2.569252077562327, "grad_norm": 8.996347427368164, "learning_rate": 4.861495844875347e-06, "loss": 0.1408, "step": 11130 }, { "epoch": 2.571560480147738, "grad_norm": 0.4904603660106659, "learning_rate": 4.8568790397045255e-06, "loss": 0.0109, "step": 11140 }, { "epoch": 2.573868882733149, "grad_norm": 6.742771148681641, "learning_rate": 4.852262234533703e-06, "loss": 0.1513, "step": 11150 }, { "epoch": 2.5761772853185594, "grad_norm": 0.23859672248363495, "learning_rate": 4.847645429362881e-06, "loss": 0.0954, "step": 11160 }, { "epoch": 2.5784856879039704, "grad_norm": 4.727524757385254, "learning_rate": 4.843028624192059e-06, "loss": 0.2321, "step": 11170 }, { "epoch": 2.5807940904893814, "grad_norm": 0.841078519821167, "learning_rate": 4.8384118190212376e-06, "loss": 0.0832, "step": 11180 }, { "epoch": 2.5831024930747923, "grad_norm": 0.2316310852766037, "learning_rate": 4.833795013850416e-06, "loss": 0.1029, "step": 11190 }, { "epoch": 2.5854108956602033, "grad_norm": 0.025395721197128296, "learning_rate": 4.829178208679594e-06, "loss": 0.1506, "step": 11200 }, { "epoch": 2.587719298245614, "grad_norm": 8.19233512878418, "learning_rate": 4.824561403508772e-06, "loss": 0.0953, "step": 11210 }, { "epoch": 2.590027700831025, "grad_norm": 0.6950219869613647, "learning_rate": 4.8199445983379505e-06, "loss": 0.0725, "step": 11220 }, { "epoch": 2.592336103416436, "grad_norm": 9.826360702514648, "learning_rate": 4.815327793167129e-06, "loss": 0.1368, "step": 11230 }, { "epoch": 2.594644506001847, "grad_norm": 5.526108264923096, "learning_rate": 4.810710987996307e-06, "loss": 0.0871, "step": 11240 }, { "epoch": 2.596952908587258, "grad_norm": 3.102935314178467, "learning_rate": 4.806094182825485e-06, "loss": 0.1603, "step": 11250 }, { "epoch": 2.5992613111726683, "grad_norm": 2.393739938735962, "learning_rate": 4.801477377654663e-06, "loss": 0.3449, "step": 11260 }, { "epoch": 2.6015697137580793, "grad_norm": 5.931601524353027, "learning_rate": 4.796860572483842e-06, "loss": 0.1194, "step": 11270 }, { "epoch": 2.6038781163434903, "grad_norm": 4.0177717208862305, "learning_rate": 4.79224376731302e-06, "loss": 0.1451, "step": 11280 }, { "epoch": 2.6061865189289013, "grad_norm": 6.002121925354004, "learning_rate": 4.787626962142198e-06, "loss": 0.0524, "step": 11290 }, { "epoch": 2.6084949215143123, "grad_norm": 2.448683977127075, "learning_rate": 4.783010156971376e-06, "loss": 0.1586, "step": 11300 }, { "epoch": 2.610803324099723, "grad_norm": 0.02205214649438858, "learning_rate": 4.7783933518005545e-06, "loss": 0.051, "step": 11310 }, { "epoch": 2.613111726685134, "grad_norm": 1.4181525707244873, "learning_rate": 4.773776546629733e-06, "loss": 0.17, "step": 11320 }, { "epoch": 2.615420129270545, "grad_norm": 4.464173316955566, "learning_rate": 4.769159741458911e-06, "loss": 0.1973, "step": 11330 }, { "epoch": 2.6177285318559558, "grad_norm": 9.001364707946777, "learning_rate": 4.764542936288089e-06, "loss": 0.1334, "step": 11340 }, { "epoch": 2.6200369344413668, "grad_norm": 0.4343610405921936, "learning_rate": 4.7599261311172675e-06, "loss": 0.0886, "step": 11350 }, { "epoch": 2.6223453370267773, "grad_norm": 2.682781934738159, "learning_rate": 4.755309325946446e-06, "loss": 0.1994, "step": 11360 }, { "epoch": 2.6246537396121883, "grad_norm": 0.1579998880624771, "learning_rate": 4.750692520775624e-06, "loss": 0.0586, "step": 11370 }, { "epoch": 2.6269621421975993, "grad_norm": 2.7223992347717285, "learning_rate": 4.746075715604802e-06, "loss": 0.1036, "step": 11380 }, { "epoch": 2.6292705447830103, "grad_norm": 0.3042351007461548, "learning_rate": 4.74145891043398e-06, "loss": 0.1053, "step": 11390 }, { "epoch": 2.6315789473684212, "grad_norm": 0.18436291813850403, "learning_rate": 4.736842105263158e-06, "loss": 0.0866, "step": 11400 }, { "epoch": 2.633887349953832, "grad_norm": 6.090860843658447, "learning_rate": 4.732225300092336e-06, "loss": 0.0962, "step": 11410 }, { "epoch": 2.6361957525392428, "grad_norm": 0.6144652366638184, "learning_rate": 4.727608494921514e-06, "loss": 0.0716, "step": 11420 }, { "epoch": 2.6385041551246537, "grad_norm": 0.385552316904068, "learning_rate": 4.7229916897506924e-06, "loss": 0.0344, "step": 11430 }, { "epoch": 2.6408125577100647, "grad_norm": 7.219420909881592, "learning_rate": 4.718374884579871e-06, "loss": 0.2256, "step": 11440 }, { "epoch": 2.6431209602954757, "grad_norm": 2.0240821838378906, "learning_rate": 4.713758079409049e-06, "loss": 0.1099, "step": 11450 }, { "epoch": 2.6454293628808863, "grad_norm": 3.666261672973633, "learning_rate": 4.709141274238227e-06, "loss": 0.0943, "step": 11460 }, { "epoch": 2.6477377654662972, "grad_norm": 0.025211500003933907, "learning_rate": 4.704524469067406e-06, "loss": 0.0438, "step": 11470 }, { "epoch": 2.6500461680517082, "grad_norm": 8.360804557800293, "learning_rate": 4.699907663896584e-06, "loss": 0.0867, "step": 11480 }, { "epoch": 2.652354570637119, "grad_norm": 0.030948152765631676, "learning_rate": 4.695290858725763e-06, "loss": 0.1312, "step": 11490 }, { "epoch": 2.65466297322253, "grad_norm": 2.7387523651123047, "learning_rate": 4.690674053554941e-06, "loss": 0.1682, "step": 11500 }, { "epoch": 2.6569713758079407, "grad_norm": 12.855392456054688, "learning_rate": 4.686057248384119e-06, "loss": 0.1544, "step": 11510 }, { "epoch": 2.6592797783933517, "grad_norm": 7.6028547286987305, "learning_rate": 4.681440443213297e-06, "loss": 0.1219, "step": 11520 }, { "epoch": 2.6615881809787627, "grad_norm": 0.24558210372924805, "learning_rate": 4.676823638042475e-06, "loss": 0.1672, "step": 11530 }, { "epoch": 2.6638965835641737, "grad_norm": 2.8545150756835938, "learning_rate": 4.672206832871653e-06, "loss": 0.1525, "step": 11540 }, { "epoch": 2.6662049861495847, "grad_norm": 6.165798187255859, "learning_rate": 4.667590027700831e-06, "loss": 0.1455, "step": 11550 }, { "epoch": 2.668513388734995, "grad_norm": 6.052087306976318, "learning_rate": 4.662973222530009e-06, "loss": 0.0924, "step": 11560 }, { "epoch": 2.670821791320406, "grad_norm": 0.026141835376620293, "learning_rate": 4.658356417359188e-06, "loss": 0.1824, "step": 11570 }, { "epoch": 2.673130193905817, "grad_norm": 5.248561859130859, "learning_rate": 4.653739612188366e-06, "loss": 0.0746, "step": 11580 }, { "epoch": 2.675438596491228, "grad_norm": 0.4188123345375061, "learning_rate": 4.649122807017544e-06, "loss": 0.0493, "step": 11590 }, { "epoch": 2.677746999076639, "grad_norm": 11.177963256835938, "learning_rate": 4.644506001846722e-06, "loss": 0.0745, "step": 11600 }, { "epoch": 2.6800554016620497, "grad_norm": 7.531991958618164, "learning_rate": 4.6398891966759005e-06, "loss": 0.1461, "step": 11610 }, { "epoch": 2.6823638042474607, "grad_norm": 0.7496140599250793, "learning_rate": 4.635272391505079e-06, "loss": 0.0574, "step": 11620 }, { "epoch": 2.6846722068328717, "grad_norm": 0.823010265827179, "learning_rate": 4.630655586334257e-06, "loss": 0.1198, "step": 11630 }, { "epoch": 2.6869806094182827, "grad_norm": 7.933892250061035, "learning_rate": 4.626038781163435e-06, "loss": 0.09, "step": 11640 }, { "epoch": 2.6892890120036936, "grad_norm": 4.41282844543457, "learning_rate": 4.6214219759926135e-06, "loss": 0.112, "step": 11650 }, { "epoch": 2.691597414589104, "grad_norm": 0.33152124285697937, "learning_rate": 4.616805170821792e-06, "loss": 0.0565, "step": 11660 }, { "epoch": 2.693905817174515, "grad_norm": 0.028384560719132423, "learning_rate": 4.61218836565097e-06, "loss": 0.0685, "step": 11670 }, { "epoch": 2.696214219759926, "grad_norm": 6.035626411437988, "learning_rate": 4.607571560480148e-06, "loss": 0.1751, "step": 11680 }, { "epoch": 2.698522622345337, "grad_norm": 0.001319588627666235, "learning_rate": 4.602954755309326e-06, "loss": 0.0724, "step": 11690 }, { "epoch": 2.700831024930748, "grad_norm": 1.2319642305374146, "learning_rate": 4.598337950138505e-06, "loss": 0.2252, "step": 11700 }, { "epoch": 2.7031394275161587, "grad_norm": 12.681593894958496, "learning_rate": 4.593721144967683e-06, "loss": 0.1636, "step": 11710 }, { "epoch": 2.7054478301015696, "grad_norm": 10.092299461364746, "learning_rate": 4.589104339796861e-06, "loss": 0.1498, "step": 11720 }, { "epoch": 2.7077562326869806, "grad_norm": 0.002909166505560279, "learning_rate": 4.584487534626039e-06, "loss": 0.1948, "step": 11730 }, { "epoch": 2.7100646352723916, "grad_norm": 8.839336395263672, "learning_rate": 4.5798707294552175e-06, "loss": 0.2027, "step": 11740 }, { "epoch": 2.7123730378578026, "grad_norm": 0.16871292889118195, "learning_rate": 4.575253924284396e-06, "loss": 0.1081, "step": 11750 }, { "epoch": 2.714681440443213, "grad_norm": 0.009426860138773918, "learning_rate": 4.570637119113574e-06, "loss": 0.0936, "step": 11760 }, { "epoch": 2.716989843028624, "grad_norm": 8.538941383361816, "learning_rate": 4.566020313942752e-06, "loss": 0.2144, "step": 11770 }, { "epoch": 2.719298245614035, "grad_norm": 5.874807357788086, "learning_rate": 4.56140350877193e-06, "loss": 0.1103, "step": 11780 }, { "epoch": 2.721606648199446, "grad_norm": 1.0961629152297974, "learning_rate": 4.556786703601108e-06, "loss": 0.1151, "step": 11790 }, { "epoch": 2.723915050784857, "grad_norm": 2.0657474994659424, "learning_rate": 4.552169898430286e-06, "loss": 0.1354, "step": 11800 }, { "epoch": 2.7262234533702676, "grad_norm": 0.7834126353263855, "learning_rate": 4.547553093259465e-06, "loss": 0.0505, "step": 11810 }, { "epoch": 2.7285318559556786, "grad_norm": 4.23038911819458, "learning_rate": 4.542936288088643e-06, "loss": 0.2024, "step": 11820 }, { "epoch": 2.7308402585410896, "grad_norm": 8.258891105651855, "learning_rate": 4.538319482917822e-06, "loss": 0.0796, "step": 11830 }, { "epoch": 2.7331486611265006, "grad_norm": 0.3063031733036041, "learning_rate": 4.533702677747e-06, "loss": 0.0795, "step": 11840 }, { "epoch": 2.7354570637119116, "grad_norm": 2.1137850284576416, "learning_rate": 4.529085872576178e-06, "loss": 0.0801, "step": 11850 }, { "epoch": 2.737765466297322, "grad_norm": 4.891541481018066, "learning_rate": 4.524469067405356e-06, "loss": 0.1934, "step": 11860 }, { "epoch": 2.740073868882733, "grad_norm": 10.987977981567383, "learning_rate": 4.5198522622345345e-06, "loss": 0.2616, "step": 11870 }, { "epoch": 2.742382271468144, "grad_norm": 5.472017765045166, "learning_rate": 4.515235457063713e-06, "loss": 0.1969, "step": 11880 }, { "epoch": 2.744690674053555, "grad_norm": 1.2341314554214478, "learning_rate": 4.510618651892891e-06, "loss": 0.268, "step": 11890 }, { "epoch": 2.746999076638966, "grad_norm": 3.740185022354126, "learning_rate": 4.506001846722069e-06, "loss": 0.1929, "step": 11900 }, { "epoch": 2.7493074792243766, "grad_norm": 5.091491222381592, "learning_rate": 4.5013850415512466e-06, "loss": 0.0784, "step": 11910 }, { "epoch": 2.7516158818097876, "grad_norm": 3.84535551071167, "learning_rate": 4.496768236380425e-06, "loss": 0.0996, "step": 11920 }, { "epoch": 2.7539242843951985, "grad_norm": 6.8564276695251465, "learning_rate": 4.492151431209603e-06, "loss": 0.2739, "step": 11930 }, { "epoch": 2.7562326869806095, "grad_norm": 0.9228309392929077, "learning_rate": 4.487534626038781e-06, "loss": 0.1234, "step": 11940 }, { "epoch": 2.7585410895660205, "grad_norm": 0.5705520510673523, "learning_rate": 4.4829178208679595e-06, "loss": 0.1878, "step": 11950 }, { "epoch": 2.760849492151431, "grad_norm": 2.723750352859497, "learning_rate": 4.478301015697138e-06, "loss": 0.0808, "step": 11960 }, { "epoch": 2.763157894736842, "grad_norm": 0.039775166660547256, "learning_rate": 4.473684210526316e-06, "loss": 0.1177, "step": 11970 }, { "epoch": 2.765466297322253, "grad_norm": 3.499601125717163, "learning_rate": 4.469067405355494e-06, "loss": 0.199, "step": 11980 }, { "epoch": 2.767774699907664, "grad_norm": 0.47708064317703247, "learning_rate": 4.464450600184672e-06, "loss": 0.0252, "step": 11990 }, { "epoch": 2.770083102493075, "grad_norm": 0.0033902195282280445, "learning_rate": 4.459833795013851e-06, "loss": 0.1332, "step": 12000 }, { "epoch": 2.7723915050784855, "grad_norm": 3.745899200439453, "learning_rate": 4.455216989843029e-06, "loss": 0.0595, "step": 12010 }, { "epoch": 2.7746999076638965, "grad_norm": 4.700807094573975, "learning_rate": 4.450600184672207e-06, "loss": 0.1559, "step": 12020 }, { "epoch": 2.7770083102493075, "grad_norm": 1.475287914276123, "learning_rate": 4.445983379501385e-06, "loss": 0.3529, "step": 12030 }, { "epoch": 2.7793167128347185, "grad_norm": 6.322391510009766, "learning_rate": 4.4413665743305635e-06, "loss": 0.1018, "step": 12040 }, { "epoch": 2.7816251154201295, "grad_norm": 0.14255914092063904, "learning_rate": 4.436749769159742e-06, "loss": 0.1201, "step": 12050 }, { "epoch": 2.78393351800554, "grad_norm": 2.42327618598938, "learning_rate": 4.43213296398892e-06, "loss": 0.0888, "step": 12060 }, { "epoch": 2.786241920590951, "grad_norm": 0.27212220430374146, "learning_rate": 4.427516158818098e-06, "loss": 0.1177, "step": 12070 }, { "epoch": 2.788550323176362, "grad_norm": 4.134885787963867, "learning_rate": 4.4228993536472764e-06, "loss": 0.2691, "step": 12080 }, { "epoch": 2.790858725761773, "grad_norm": 0.6160984039306641, "learning_rate": 4.418282548476455e-06, "loss": 0.107, "step": 12090 }, { "epoch": 2.793167128347184, "grad_norm": 7.349871635437012, "learning_rate": 4.413665743305633e-06, "loss": 0.2196, "step": 12100 }, { "epoch": 2.7954755309325945, "grad_norm": 3.8904452323913574, "learning_rate": 4.409048938134811e-06, "loss": 0.1124, "step": 12110 }, { "epoch": 2.7977839335180055, "grad_norm": 0.42588678002357483, "learning_rate": 4.404432132963989e-06, "loss": 0.1423, "step": 12120 }, { "epoch": 2.8000923361034165, "grad_norm": 8.574092864990234, "learning_rate": 4.399815327793168e-06, "loss": 0.1425, "step": 12130 }, { "epoch": 2.8024007386888274, "grad_norm": 0.49055933952331543, "learning_rate": 4.395198522622346e-06, "loss": 0.1008, "step": 12140 }, { "epoch": 2.8047091412742384, "grad_norm": 8.238758087158203, "learning_rate": 4.390581717451524e-06, "loss": 0.142, "step": 12150 }, { "epoch": 2.807017543859649, "grad_norm": 2.543043613433838, "learning_rate": 4.385964912280702e-06, "loss": 0.163, "step": 12160 }, { "epoch": 2.80932594644506, "grad_norm": 10.594868659973145, "learning_rate": 4.3813481071098805e-06, "loss": 0.2544, "step": 12170 }, { "epoch": 2.811634349030471, "grad_norm": 4.137728691101074, "learning_rate": 4.376731301939059e-06, "loss": 0.0655, "step": 12180 }, { "epoch": 2.813942751615882, "grad_norm": 1.3153051137924194, "learning_rate": 4.372114496768237e-06, "loss": 0.147, "step": 12190 }, { "epoch": 2.816251154201293, "grad_norm": 9.323812484741211, "learning_rate": 4.367497691597415e-06, "loss": 0.0993, "step": 12200 }, { "epoch": 2.8185595567867034, "grad_norm": 0.6330371499061584, "learning_rate": 4.362880886426593e-06, "loss": 0.1303, "step": 12210 }, { "epoch": 2.8208679593721144, "grad_norm": 7.022908687591553, "learning_rate": 4.358264081255772e-06, "loss": 0.1969, "step": 12220 }, { "epoch": 2.8231763619575254, "grad_norm": 12.385854721069336, "learning_rate": 4.35364727608495e-06, "loss": 0.1082, "step": 12230 }, { "epoch": 2.8254847645429364, "grad_norm": 3.009413003921509, "learning_rate": 4.349030470914128e-06, "loss": 0.1227, "step": 12240 }, { "epoch": 2.8277931671283474, "grad_norm": 0.6962842345237732, "learning_rate": 4.344413665743306e-06, "loss": 0.1166, "step": 12250 }, { "epoch": 2.830101569713758, "grad_norm": 1.5894197225570679, "learning_rate": 4.3397968605724846e-06, "loss": 0.1863, "step": 12260 }, { "epoch": 2.832409972299169, "grad_norm": 8.911307334899902, "learning_rate": 4.335180055401663e-06, "loss": 0.0857, "step": 12270 }, { "epoch": 2.83471837488458, "grad_norm": 7.859243392944336, "learning_rate": 4.33056325023084e-06, "loss": 0.2345, "step": 12280 }, { "epoch": 2.837026777469991, "grad_norm": 0.01118907704949379, "learning_rate": 4.325946445060018e-06, "loss": 0.077, "step": 12290 }, { "epoch": 2.839335180055402, "grad_norm": 7.9256205558776855, "learning_rate": 4.321329639889197e-06, "loss": 0.1413, "step": 12300 }, { "epoch": 2.8416435826408124, "grad_norm": 0.17542405426502228, "learning_rate": 4.316712834718375e-06, "loss": 0.1531, "step": 12310 }, { "epoch": 2.8439519852262234, "grad_norm": 3.5374441146850586, "learning_rate": 4.312096029547553e-06, "loss": 0.0499, "step": 12320 }, { "epoch": 2.8462603878116344, "grad_norm": 5.826684474945068, "learning_rate": 4.307479224376731e-06, "loss": 0.2864, "step": 12330 }, { "epoch": 2.8485687903970454, "grad_norm": 3.758387565612793, "learning_rate": 4.3028624192059095e-06, "loss": 0.0447, "step": 12340 }, { "epoch": 2.8508771929824563, "grad_norm": 5.596960067749023, "learning_rate": 4.298245614035088e-06, "loss": 0.0791, "step": 12350 }, { "epoch": 2.853185595567867, "grad_norm": 2.401756525039673, "learning_rate": 4.293628808864266e-06, "loss": 0.1292, "step": 12360 }, { "epoch": 2.855493998153278, "grad_norm": 11.880175590515137, "learning_rate": 4.289012003693444e-06, "loss": 0.1456, "step": 12370 }, { "epoch": 2.857802400738689, "grad_norm": 0.018428804352879524, "learning_rate": 4.2843951985226225e-06, "loss": 0.1234, "step": 12380 }, { "epoch": 2.8601108033241, "grad_norm": 6.749403953552246, "learning_rate": 4.2797783933518015e-06, "loss": 0.1732, "step": 12390 }, { "epoch": 2.862419205909511, "grad_norm": 0.01735766790807247, "learning_rate": 4.27516158818098e-06, "loss": 0.0069, "step": 12400 }, { "epoch": 2.8647276084949214, "grad_norm": 1.4365239143371582, "learning_rate": 4.270544783010157e-06, "loss": 0.1788, "step": 12410 }, { "epoch": 2.8670360110803323, "grad_norm": 1.1888684034347534, "learning_rate": 4.265927977839335e-06, "loss": 0.2305, "step": 12420 }, { "epoch": 2.8693444136657433, "grad_norm": 0.08861195296049118, "learning_rate": 4.261311172668514e-06, "loss": 0.1234, "step": 12430 }, { "epoch": 2.8716528162511543, "grad_norm": 3.8148953914642334, "learning_rate": 4.256694367497692e-06, "loss": 0.0796, "step": 12440 }, { "epoch": 2.8739612188365653, "grad_norm": 0.24537815153598785, "learning_rate": 4.25207756232687e-06, "loss": 0.0898, "step": 12450 }, { "epoch": 2.876269621421976, "grad_norm": 1.9411917924880981, "learning_rate": 4.247460757156048e-06, "loss": 0.1337, "step": 12460 }, { "epoch": 2.878578024007387, "grad_norm": 1.2744687795639038, "learning_rate": 4.2428439519852265e-06, "loss": 0.0618, "step": 12470 }, { "epoch": 2.880886426592798, "grad_norm": 1.1889690160751343, "learning_rate": 4.238227146814405e-06, "loss": 0.1463, "step": 12480 }, { "epoch": 2.883194829178209, "grad_norm": 5.925126552581787, "learning_rate": 4.233610341643583e-06, "loss": 0.0362, "step": 12490 }, { "epoch": 2.88550323176362, "grad_norm": 2.906893730163574, "learning_rate": 4.228993536472761e-06, "loss": 0.1441, "step": 12500 }, { "epoch": 2.8878116343490303, "grad_norm": 2.8232715129852295, "learning_rate": 4.2243767313019394e-06, "loss": 0.0549, "step": 12510 }, { "epoch": 2.8901200369344413, "grad_norm": 0.09884170442819595, "learning_rate": 4.219759926131118e-06, "loss": 0.1591, "step": 12520 }, { "epoch": 2.8924284395198523, "grad_norm": 0.19307319819927216, "learning_rate": 4.215143120960296e-06, "loss": 0.1517, "step": 12530 }, { "epoch": 2.8947368421052633, "grad_norm": 0.003301033051684499, "learning_rate": 4.210526315789474e-06, "loss": 0.0571, "step": 12540 }, { "epoch": 2.8970452446906743, "grad_norm": 6.7552289962768555, "learning_rate": 4.205909510618652e-06, "loss": 0.0659, "step": 12550 }, { "epoch": 2.899353647276085, "grad_norm": 4.592986583709717, "learning_rate": 4.2012927054478306e-06, "loss": 0.1829, "step": 12560 }, { "epoch": 2.901662049861496, "grad_norm": 0.06313756108283997, "learning_rate": 4.196675900277009e-06, "loss": 0.0409, "step": 12570 }, { "epoch": 2.9039704524469068, "grad_norm": 6.815639019012451, "learning_rate": 4.192059095106187e-06, "loss": 0.1495, "step": 12580 }, { "epoch": 2.9062788550323178, "grad_norm": 0.12222865968942642, "learning_rate": 4.187442289935365e-06, "loss": 0.1085, "step": 12590 }, { "epoch": 2.9085872576177287, "grad_norm": 1.250489354133606, "learning_rate": 4.1828254847645435e-06, "loss": 0.0609, "step": 12600 }, { "epoch": 2.9108956602031393, "grad_norm": 0.01571718417108059, "learning_rate": 4.178208679593722e-06, "loss": 0.1259, "step": 12610 }, { "epoch": 2.9132040627885503, "grad_norm": 4.453795909881592, "learning_rate": 4.1735918744229e-06, "loss": 0.0872, "step": 12620 }, { "epoch": 2.9155124653739612, "grad_norm": 3.9681210182607174e-05, "learning_rate": 4.168975069252078e-06, "loss": 0.2371, "step": 12630 }, { "epoch": 2.9178208679593722, "grad_norm": 1.2263693809509277, "learning_rate": 4.164358264081256e-06, "loss": 0.0928, "step": 12640 }, { "epoch": 2.920129270544783, "grad_norm": 3.0415117740631104, "learning_rate": 4.159741458910435e-06, "loss": 0.1723, "step": 12650 }, { "epoch": 2.9224376731301938, "grad_norm": 1.183361530303955, "learning_rate": 4.155124653739612e-06, "loss": 0.1371, "step": 12660 }, { "epoch": 2.9247460757156047, "grad_norm": 12.683422088623047, "learning_rate": 4.15050784856879e-06, "loss": 0.2489, "step": 12670 }, { "epoch": 2.9270544783010157, "grad_norm": 9.827859878540039, "learning_rate": 4.1458910433979685e-06, "loss": 0.278, "step": 12680 }, { "epoch": 2.9293628808864267, "grad_norm": 0.7906888127326965, "learning_rate": 4.141274238227147e-06, "loss": 0.1311, "step": 12690 }, { "epoch": 2.9316712834718377, "grad_norm": 2.8516433238983154, "learning_rate": 4.136657433056325e-06, "loss": 0.1022, "step": 12700 }, { "epoch": 2.9339796860572482, "grad_norm": 0.5616424679756165, "learning_rate": 4.132040627885503e-06, "loss": 0.107, "step": 12710 }, { "epoch": 2.936288088642659, "grad_norm": 1.1421198844909668, "learning_rate": 4.127423822714681e-06, "loss": 0.1424, "step": 12720 }, { "epoch": 2.93859649122807, "grad_norm": 2.0235655307769775, "learning_rate": 4.12280701754386e-06, "loss": 0.0645, "step": 12730 }, { "epoch": 2.940904893813481, "grad_norm": 0.027226995676755905, "learning_rate": 4.118190212373039e-06, "loss": 0.0533, "step": 12740 }, { "epoch": 2.943213296398892, "grad_norm": 1.8449541330337524, "learning_rate": 4.113573407202217e-06, "loss": 0.2813, "step": 12750 }, { "epoch": 2.9455216989843027, "grad_norm": 0.10684164613485336, "learning_rate": 4.108956602031395e-06, "loss": 0.0962, "step": 12760 }, { "epoch": 2.9478301015697137, "grad_norm": 3.032719612121582, "learning_rate": 4.104339796860573e-06, "loss": 0.1026, "step": 12770 }, { "epoch": 2.9501385041551247, "grad_norm": 4.871971607208252, "learning_rate": 4.099722991689752e-06, "loss": 0.1319, "step": 12780 }, { "epoch": 2.9524469067405357, "grad_norm": 0.6279833912849426, "learning_rate": 4.095106186518929e-06, "loss": 0.1044, "step": 12790 }, { "epoch": 2.9547553093259467, "grad_norm": 7.019402027130127, "learning_rate": 4.090489381348107e-06, "loss": 0.0706, "step": 12800 }, { "epoch": 2.957063711911357, "grad_norm": 0.11368291079998016, "learning_rate": 4.0858725761772854e-06, "loss": 0.1078, "step": 12810 }, { "epoch": 2.959372114496768, "grad_norm": 0.004620388615876436, "learning_rate": 4.081255771006464e-06, "loss": 0.272, "step": 12820 }, { "epoch": 2.961680517082179, "grad_norm": 11.176382064819336, "learning_rate": 4.076638965835642e-06, "loss": 0.0886, "step": 12830 }, { "epoch": 2.96398891966759, "grad_norm": 9.425949096679688, "learning_rate": 4.07202216066482e-06, "loss": 0.1563, "step": 12840 }, { "epoch": 2.966297322253001, "grad_norm": 1.3736194372177124, "learning_rate": 4.067405355493998e-06, "loss": 0.114, "step": 12850 }, { "epoch": 2.9686057248384117, "grad_norm": 7.303421974182129, "learning_rate": 4.062788550323177e-06, "loss": 0.2275, "step": 12860 }, { "epoch": 2.9709141274238227, "grad_norm": 2.4428906440734863, "learning_rate": 4.058171745152355e-06, "loss": 0.1747, "step": 12870 }, { "epoch": 2.9732225300092336, "grad_norm": 1.0928312540054321, "learning_rate": 4.053554939981533e-06, "loss": 0.0763, "step": 12880 }, { "epoch": 2.9755309325946446, "grad_norm": 1.023366928100586, "learning_rate": 4.048938134810711e-06, "loss": 0.0645, "step": 12890 }, { "epoch": 2.9778393351800556, "grad_norm": 0.009234298951923847, "learning_rate": 4.0443213296398895e-06, "loss": 0.0417, "step": 12900 }, { "epoch": 2.980147737765466, "grad_norm": 11.426445960998535, "learning_rate": 4.039704524469068e-06, "loss": 0.2068, "step": 12910 }, { "epoch": 2.982456140350877, "grad_norm": 0.015169547870755196, "learning_rate": 4.035087719298246e-06, "loss": 0.0614, "step": 12920 }, { "epoch": 2.984764542936288, "grad_norm": 2.518937349319458, "learning_rate": 4.030470914127424e-06, "loss": 0.1195, "step": 12930 }, { "epoch": 2.987072945521699, "grad_norm": 9.604926109313965, "learning_rate": 4.025854108956602e-06, "loss": 0.1733, "step": 12940 }, { "epoch": 2.98938134810711, "grad_norm": 1.9895973205566406, "learning_rate": 4.021237303785781e-06, "loss": 0.0484, "step": 12950 }, { "epoch": 2.9916897506925206, "grad_norm": 0.7261149883270264, "learning_rate": 4.016620498614959e-06, "loss": 0.1033, "step": 12960 }, { "epoch": 2.9939981532779316, "grad_norm": 7.013209342956543, "learning_rate": 4.012003693444137e-06, "loss": 0.1557, "step": 12970 }, { "epoch": 2.9963065558633426, "grad_norm": 0.10759413242340088, "learning_rate": 4.007386888273315e-06, "loss": 0.2148, "step": 12980 }, { "epoch": 2.9986149584487536, "grad_norm": 1.3662234544754028, "learning_rate": 4.0027700831024936e-06, "loss": 0.0721, "step": 12990 }, { "epoch": 3.0009233610341646, "grad_norm": 7.676759243011475, "learning_rate": 3.998153277931672e-06, "loss": 0.1231, "step": 13000 }, { "epoch": 3.003231763619575, "grad_norm": 0.27375108003616333, "learning_rate": 3.99353647276085e-06, "loss": 0.1197, "step": 13010 }, { "epoch": 3.005540166204986, "grad_norm": 6.033203125, "learning_rate": 3.988919667590028e-06, "loss": 0.1631, "step": 13020 }, { "epoch": 3.007848568790397, "grad_norm": 0.006016091909259558, "learning_rate": 3.9843028624192065e-06, "loss": 0.0933, "step": 13030 }, { "epoch": 3.010156971375808, "grad_norm": 8.702679634094238, "learning_rate": 3.979686057248384e-06, "loss": 0.24, "step": 13040 }, { "epoch": 3.012465373961219, "grad_norm": 8.987473487854004, "learning_rate": 3.975069252077562e-06, "loss": 0.2273, "step": 13050 }, { "epoch": 3.0147737765466296, "grad_norm": 0.5691729784011841, "learning_rate": 3.97045244690674e-06, "loss": 0.0607, "step": 13060 }, { "epoch": 3.0170821791320406, "grad_norm": 0.48847314715385437, "learning_rate": 3.9658356417359185e-06, "loss": 0.0853, "step": 13070 }, { "epoch": 3.0193905817174516, "grad_norm": 2.7002453804016113, "learning_rate": 3.961218836565098e-06, "loss": 0.0211, "step": 13080 }, { "epoch": 3.0216989843028625, "grad_norm": 3.363935947418213, "learning_rate": 3.956602031394276e-06, "loss": 0.0823, "step": 13090 }, { "epoch": 3.0240073868882735, "grad_norm": 0.459314227104187, "learning_rate": 3.951985226223454e-06, "loss": 0.0877, "step": 13100 }, { "epoch": 3.026315789473684, "grad_norm": 0.00035472147283144295, "learning_rate": 3.947368421052632e-06, "loss": 0.0786, "step": 13110 }, { "epoch": 3.028624192059095, "grad_norm": 0.009792185388505459, "learning_rate": 3.9427516158818105e-06, "loss": 0.1428, "step": 13120 }, { "epoch": 3.030932594644506, "grad_norm": 3.769273519515991, "learning_rate": 3.938134810710989e-06, "loss": 0.0992, "step": 13130 }, { "epoch": 3.033240997229917, "grad_norm": 2.1976845264434814, "learning_rate": 3.933518005540167e-06, "loss": 0.151, "step": 13140 }, { "epoch": 3.035549399815328, "grad_norm": 5.4209089279174805, "learning_rate": 3.928901200369345e-06, "loss": 0.104, "step": 13150 }, { "epoch": 3.0378578024007385, "grad_norm": 0.21197374165058136, "learning_rate": 3.924284395198523e-06, "loss": 0.164, "step": 13160 }, { "epoch": 3.0401662049861495, "grad_norm": 11.942008972167969, "learning_rate": 3.919667590027701e-06, "loss": 0.1581, "step": 13170 }, { "epoch": 3.0424746075715605, "grad_norm": 4.588008880615234, "learning_rate": 3.915050784856879e-06, "loss": 0.0492, "step": 13180 }, { "epoch": 3.0447830101569715, "grad_norm": 10.034119606018066, "learning_rate": 3.910433979686057e-06, "loss": 0.1067, "step": 13190 }, { "epoch": 3.0470914127423825, "grad_norm": 0.38339802622795105, "learning_rate": 3.9058171745152355e-06, "loss": 0.0395, "step": 13200 }, { "epoch": 3.049399815327793, "grad_norm": 0.34745603799819946, "learning_rate": 3.901200369344414e-06, "loss": 0.0566, "step": 13210 }, { "epoch": 3.051708217913204, "grad_norm": 0.4207463562488556, "learning_rate": 3.896583564173592e-06, "loss": 0.2866, "step": 13220 }, { "epoch": 3.054016620498615, "grad_norm": 3.144205093383789, "learning_rate": 3.89196675900277e-06, "loss": 0.0626, "step": 13230 }, { "epoch": 3.056325023084026, "grad_norm": 0.787864625453949, "learning_rate": 3.8873499538319484e-06, "loss": 0.1187, "step": 13240 }, { "epoch": 3.058633425669437, "grad_norm": 9.620296478271484, "learning_rate": 3.882733148661127e-06, "loss": 0.0937, "step": 13250 }, { "epoch": 3.0609418282548475, "grad_norm": 0.13369058072566986, "learning_rate": 3.878116343490305e-06, "loss": 0.1931, "step": 13260 }, { "epoch": 3.0632502308402585, "grad_norm": 0.0021691815927624702, "learning_rate": 3.873499538319483e-06, "loss": 0.0301, "step": 13270 }, { "epoch": 3.0655586334256695, "grad_norm": 5.102521896362305, "learning_rate": 3.868882733148661e-06, "loss": 0.0707, "step": 13280 }, { "epoch": 3.0678670360110805, "grad_norm": 3.165531635284424, "learning_rate": 3.8642659279778396e-06, "loss": 0.2115, "step": 13290 }, { "epoch": 3.0701754385964914, "grad_norm": 3.643566608428955, "learning_rate": 3.859649122807018e-06, "loss": 0.1491, "step": 13300 }, { "epoch": 3.072483841181902, "grad_norm": 7.703869342803955, "learning_rate": 3.855032317636196e-06, "loss": 0.2112, "step": 13310 }, { "epoch": 3.074792243767313, "grad_norm": 0.04833643510937691, "learning_rate": 3.850415512465374e-06, "loss": 0.1045, "step": 13320 }, { "epoch": 3.077100646352724, "grad_norm": 5.805057525634766, "learning_rate": 3.8457987072945525e-06, "loss": 0.0514, "step": 13330 }, { "epoch": 3.079409048938135, "grad_norm": 0.630721390247345, "learning_rate": 3.841181902123731e-06, "loss": 0.0333, "step": 13340 }, { "epoch": 3.081717451523546, "grad_norm": 8.413884162902832, "learning_rate": 3.836565096952909e-06, "loss": 0.2165, "step": 13350 }, { "epoch": 3.0840258541089565, "grad_norm": 3.9678401947021484, "learning_rate": 3.831948291782087e-06, "loss": 0.1899, "step": 13360 }, { "epoch": 3.0863342566943675, "grad_norm": 5.244140148162842, "learning_rate": 3.827331486611265e-06, "loss": 0.0802, "step": 13370 }, { "epoch": 3.0886426592797784, "grad_norm": 0.21963419020175934, "learning_rate": 3.822714681440444e-06, "loss": 0.2449, "step": 13380 }, { "epoch": 3.0909510618651894, "grad_norm": 0.7885622978210449, "learning_rate": 3.818097876269622e-06, "loss": 0.0942, "step": 13390 }, { "epoch": 3.0932594644506004, "grad_norm": 0.05466680973768234, "learning_rate": 3.8134810710988e-06, "loss": 0.0459, "step": 13400 }, { "epoch": 3.095567867036011, "grad_norm": 0.007173283491283655, "learning_rate": 3.8088642659279783e-06, "loss": 0.121, "step": 13410 }, { "epoch": 3.097876269621422, "grad_norm": 11.440203666687012, "learning_rate": 3.804247460757156e-06, "loss": 0.161, "step": 13420 }, { "epoch": 3.100184672206833, "grad_norm": 6.782459735870361, "learning_rate": 3.7996306555863343e-06, "loss": 0.0893, "step": 13430 }, { "epoch": 3.102493074792244, "grad_norm": 0.37497928738594055, "learning_rate": 3.7950138504155126e-06, "loss": 0.0184, "step": 13440 }, { "epoch": 3.104801477377655, "grad_norm": 4.443853855133057, "learning_rate": 3.790397045244691e-06, "loss": 0.222, "step": 13450 }, { "epoch": 3.1071098799630654, "grad_norm": 0.7745585441589355, "learning_rate": 3.785780240073869e-06, "loss": 0.1432, "step": 13460 }, { "epoch": 3.1094182825484764, "grad_norm": 0.8787555694580078, "learning_rate": 3.7811634349030473e-06, "loss": 0.0507, "step": 13470 }, { "epoch": 3.1117266851338874, "grad_norm": 7.299482345581055, "learning_rate": 3.7765466297322255e-06, "loss": 0.1505, "step": 13480 }, { "epoch": 3.1140350877192984, "grad_norm": 0.07280556857585907, "learning_rate": 3.7719298245614037e-06, "loss": 0.0756, "step": 13490 }, { "epoch": 3.1163434903047094, "grad_norm": 1.853515386581421, "learning_rate": 3.767313019390582e-06, "loss": 0.0962, "step": 13500 }, { "epoch": 3.11865189289012, "grad_norm": 6.079598426818848, "learning_rate": 3.7626962142197606e-06, "loss": 0.1614, "step": 13510 }, { "epoch": 3.120960295475531, "grad_norm": 2.489668369293213, "learning_rate": 3.758079409048939e-06, "loss": 0.1418, "step": 13520 }, { "epoch": 3.123268698060942, "grad_norm": 1.5554605722427368, "learning_rate": 3.753462603878117e-06, "loss": 0.2152, "step": 13530 }, { "epoch": 3.125577100646353, "grad_norm": 1.0239243507385254, "learning_rate": 3.7488457987072944e-06, "loss": 0.0959, "step": 13540 }, { "epoch": 3.127885503231764, "grad_norm": 2.946845769882202, "learning_rate": 3.7442289935364727e-06, "loss": 0.1254, "step": 13550 }, { "epoch": 3.1301939058171744, "grad_norm": 5.748038291931152, "learning_rate": 3.739612188365651e-06, "loss": 0.2081, "step": 13560 }, { "epoch": 3.1325023084025854, "grad_norm": 4.811450481414795, "learning_rate": 3.7349953831948295e-06, "loss": 0.1132, "step": 13570 }, { "epoch": 3.1348107109879964, "grad_norm": 0.20056384801864624, "learning_rate": 3.7303785780240078e-06, "loss": 0.0726, "step": 13580 }, { "epoch": 3.1371191135734073, "grad_norm": 7.290846824645996, "learning_rate": 3.725761772853186e-06, "loss": 0.1021, "step": 13590 }, { "epoch": 3.1394275161588183, "grad_norm": 0.22795188426971436, "learning_rate": 3.7211449676823642e-06, "loss": 0.0689, "step": 13600 }, { "epoch": 3.141735918744229, "grad_norm": 2.670126438140869, "learning_rate": 3.7165281625115425e-06, "loss": 0.1375, "step": 13610 }, { "epoch": 3.14404432132964, "grad_norm": 3.748411178588867, "learning_rate": 3.7119113573407207e-06, "loss": 0.0861, "step": 13620 }, { "epoch": 3.146352723915051, "grad_norm": 0.16348066926002502, "learning_rate": 3.707294552169899e-06, "loss": 0.1028, "step": 13630 }, { "epoch": 3.148661126500462, "grad_norm": 0.003457755548879504, "learning_rate": 3.702677746999077e-06, "loss": 0.0795, "step": 13640 }, { "epoch": 3.150969529085873, "grad_norm": 0.05064443126320839, "learning_rate": 3.6980609418282554e-06, "loss": 0.0907, "step": 13650 }, { "epoch": 3.1532779316712833, "grad_norm": 6.89999532699585, "learning_rate": 3.6934441366574336e-06, "loss": 0.1559, "step": 13660 }, { "epoch": 3.1555863342566943, "grad_norm": 1.538375973701477, "learning_rate": 3.6888273314866114e-06, "loss": 0.0668, "step": 13670 }, { "epoch": 3.1578947368421053, "grad_norm": 0.6473358273506165, "learning_rate": 3.6842105263157896e-06, "loss": 0.0582, "step": 13680 }, { "epoch": 3.1602031394275163, "grad_norm": 10.123824119567871, "learning_rate": 3.679593721144968e-06, "loss": 0.102, "step": 13690 }, { "epoch": 3.1625115420129273, "grad_norm": 2.378610610961914, "learning_rate": 3.674976915974146e-06, "loss": 0.2194, "step": 13700 }, { "epoch": 3.164819944598338, "grad_norm": 7.982519149780273, "learning_rate": 3.6703601108033243e-06, "loss": 0.1049, "step": 13710 }, { "epoch": 3.167128347183749, "grad_norm": 3.909001588821411, "learning_rate": 3.6657433056325026e-06, "loss": 0.0495, "step": 13720 }, { "epoch": 3.16943674976916, "grad_norm": 0.03413324058055878, "learning_rate": 3.6611265004616808e-06, "loss": 0.0917, "step": 13730 }, { "epoch": 3.1717451523545708, "grad_norm": 3.6728029251098633, "learning_rate": 3.656509695290859e-06, "loss": 0.1013, "step": 13740 }, { "epoch": 3.1740535549399818, "grad_norm": 0.2892889678478241, "learning_rate": 3.6518928901200372e-06, "loss": 0.0924, "step": 13750 }, { "epoch": 3.1763619575253923, "grad_norm": 1.0608230829238892, "learning_rate": 3.6472760849492155e-06, "loss": 0.0661, "step": 13760 }, { "epoch": 3.1786703601108033, "grad_norm": 0.0012446482433006167, "learning_rate": 3.6426592797783937e-06, "loss": 0.1019, "step": 13770 }, { "epoch": 3.1809787626962143, "grad_norm": 14.463258743286133, "learning_rate": 3.638042474607572e-06, "loss": 0.1039, "step": 13780 }, { "epoch": 3.1832871652816253, "grad_norm": 0.01538106333464384, "learning_rate": 3.6334256694367497e-06, "loss": 0.0733, "step": 13790 }, { "epoch": 3.1855955678670362, "grad_norm": 0.10405433923006058, "learning_rate": 3.628808864265928e-06, "loss": 0.0215, "step": 13800 }, { "epoch": 3.187903970452447, "grad_norm": 2.4062962532043457, "learning_rate": 3.624192059095106e-06, "loss": 0.0148, "step": 13810 }, { "epoch": 3.1902123730378578, "grad_norm": 1.6283870935440063, "learning_rate": 3.6195752539242844e-06, "loss": 0.1014, "step": 13820 }, { "epoch": 3.1925207756232687, "grad_norm": 0.7465366721153259, "learning_rate": 3.6149584487534626e-06, "loss": 0.0191, "step": 13830 }, { "epoch": 3.1948291782086797, "grad_norm": 7.680554389953613, "learning_rate": 3.610341643582641e-06, "loss": 0.0413, "step": 13840 }, { "epoch": 3.1971375807940907, "grad_norm": 0.02895492874085903, "learning_rate": 3.605724838411819e-06, "loss": 0.0455, "step": 13850 }, { "epoch": 3.1994459833795013, "grad_norm": 0.35189956426620483, "learning_rate": 3.6011080332409978e-06, "loss": 0.1721, "step": 13860 }, { "epoch": 3.2017543859649122, "grad_norm": 0.003074290696531534, "learning_rate": 3.596491228070176e-06, "loss": 0.0925, "step": 13870 }, { "epoch": 3.2040627885503232, "grad_norm": 1.3636423349380493, "learning_rate": 3.5918744228993542e-06, "loss": 0.1384, "step": 13880 }, { "epoch": 3.206371191135734, "grad_norm": 0.6792395710945129, "learning_rate": 3.5872576177285324e-06, "loss": 0.0597, "step": 13890 }, { "epoch": 3.208679593721145, "grad_norm": 0.10410724580287933, "learning_rate": 3.5826408125577107e-06, "loss": 0.0498, "step": 13900 }, { "epoch": 3.2109879963065557, "grad_norm": 4.637862205505371, "learning_rate": 3.578024007386889e-06, "loss": 0.0483, "step": 13910 }, { "epoch": 3.2132963988919667, "grad_norm": 0.39197665452957153, "learning_rate": 3.5734072022160667e-06, "loss": 0.1221, "step": 13920 }, { "epoch": 3.2156048014773777, "grad_norm": 1.9769436120986938, "learning_rate": 3.568790397045245e-06, "loss": 0.0108, "step": 13930 }, { "epoch": 3.2179132040627887, "grad_norm": 8.299297332763672, "learning_rate": 3.564173591874423e-06, "loss": 0.1059, "step": 13940 }, { "epoch": 3.2202216066481997, "grad_norm": 5.5726470947265625, "learning_rate": 3.5595567867036014e-06, "loss": 0.1487, "step": 13950 }, { "epoch": 3.22253000923361, "grad_norm": 0.32682740688323975, "learning_rate": 3.5549399815327796e-06, "loss": 0.1232, "step": 13960 }, { "epoch": 3.224838411819021, "grad_norm": 6.204406261444092, "learning_rate": 3.550323176361958e-06, "loss": 0.0716, "step": 13970 }, { "epoch": 3.227146814404432, "grad_norm": 6.579569339752197, "learning_rate": 3.545706371191136e-06, "loss": 0.1497, "step": 13980 }, { "epoch": 3.229455216989843, "grad_norm": 12.647995948791504, "learning_rate": 3.5410895660203143e-06, "loss": 0.1388, "step": 13990 }, { "epoch": 3.231763619575254, "grad_norm": 4.913413047790527, "learning_rate": 3.5364727608494925e-06, "loss": 0.0348, "step": 14000 }, { "epoch": 3.2340720221606647, "grad_norm": 13.637798309326172, "learning_rate": 3.5318559556786708e-06, "loss": 0.1231, "step": 14010 }, { "epoch": 3.2363804247460757, "grad_norm": 6.473240375518799, "learning_rate": 3.527239150507849e-06, "loss": 0.1193, "step": 14020 }, { "epoch": 3.2386888273314867, "grad_norm": 3.0292248725891113, "learning_rate": 3.5226223453370272e-06, "loss": 0.1484, "step": 14030 }, { "epoch": 3.2409972299168976, "grad_norm": 3.2937562465667725, "learning_rate": 3.5180055401662054e-06, "loss": 0.0691, "step": 14040 }, { "epoch": 3.2433056325023086, "grad_norm": 0.06690353900194168, "learning_rate": 3.5133887349953833e-06, "loss": 0.0621, "step": 14050 }, { "epoch": 3.245614035087719, "grad_norm": 0.006138617638498545, "learning_rate": 3.5087719298245615e-06, "loss": 0.0151, "step": 14060 }, { "epoch": 3.24792243767313, "grad_norm": 3.079624891281128, "learning_rate": 3.5041551246537397e-06, "loss": 0.0477, "step": 14070 }, { "epoch": 3.250230840258541, "grad_norm": 10.834182739257812, "learning_rate": 3.499538319482918e-06, "loss": 0.2138, "step": 14080 }, { "epoch": 3.252539242843952, "grad_norm": 9.931973457336426, "learning_rate": 3.494921514312096e-06, "loss": 0.0919, "step": 14090 }, { "epoch": 3.254847645429363, "grad_norm": 1.2647910118103027, "learning_rate": 3.4903047091412744e-06, "loss": 0.0569, "step": 14100 }, { "epoch": 3.2571560480147737, "grad_norm": 0.011304380372166634, "learning_rate": 3.4856879039704526e-06, "loss": 0.0735, "step": 14110 }, { "epoch": 3.2594644506001846, "grad_norm": 0.013899344950914383, "learning_rate": 3.481071098799631e-06, "loss": 0.0872, "step": 14120 }, { "epoch": 3.2617728531855956, "grad_norm": 9.755290031433105, "learning_rate": 3.476454293628809e-06, "loss": 0.2599, "step": 14130 }, { "epoch": 3.2640812557710066, "grad_norm": 1.1771888732910156, "learning_rate": 3.4718374884579873e-06, "loss": 0.3479, "step": 14140 }, { "epoch": 3.2663896583564176, "grad_norm": 0.5214719176292419, "learning_rate": 3.467220683287166e-06, "loss": 0.1274, "step": 14150 }, { "epoch": 3.268698060941828, "grad_norm": 0.09512855857610703, "learning_rate": 3.462603878116344e-06, "loss": 0.0699, "step": 14160 }, { "epoch": 3.271006463527239, "grad_norm": 0.07729105651378632, "learning_rate": 3.4579870729455216e-06, "loss": 0.1731, "step": 14170 }, { "epoch": 3.27331486611265, "grad_norm": 2.452829599380493, "learning_rate": 3.4533702677747e-06, "loss": 0.1376, "step": 14180 }, { "epoch": 3.275623268698061, "grad_norm": 0.05896938592195511, "learning_rate": 3.448753462603878e-06, "loss": 0.1356, "step": 14190 }, { "epoch": 3.277931671283472, "grad_norm": 3.13010835647583, "learning_rate": 3.4441366574330567e-06, "loss": 0.0334, "step": 14200 }, { "epoch": 3.2802400738688826, "grad_norm": 0.08988390117883682, "learning_rate": 3.439519852262235e-06, "loss": 0.1009, "step": 14210 }, { "epoch": 3.2825484764542936, "grad_norm": 0.23293976485729218, "learning_rate": 3.434903047091413e-06, "loss": 0.0587, "step": 14220 }, { "epoch": 3.2848568790397046, "grad_norm": 4.495115280151367, "learning_rate": 3.4302862419205914e-06, "loss": 0.0314, "step": 14230 }, { "epoch": 3.2871652816251156, "grad_norm": 5.483779430389404, "learning_rate": 3.4256694367497696e-06, "loss": 0.0845, "step": 14240 }, { "epoch": 3.2894736842105265, "grad_norm": 4.726720333099365, "learning_rate": 3.421052631578948e-06, "loss": 0.0792, "step": 14250 }, { "epoch": 3.291782086795937, "grad_norm": 0.006577345076948404, "learning_rate": 3.416435826408126e-06, "loss": 0.1486, "step": 14260 }, { "epoch": 3.294090489381348, "grad_norm": 1.1988341808319092, "learning_rate": 3.4118190212373043e-06, "loss": 0.125, "step": 14270 }, { "epoch": 3.296398891966759, "grad_norm": 6.639676094055176, "learning_rate": 3.4072022160664825e-06, "loss": 0.0501, "step": 14280 }, { "epoch": 3.29870729455217, "grad_norm": 0.13016977906227112, "learning_rate": 3.4025854108956607e-06, "loss": 0.13, "step": 14290 }, { "epoch": 3.301015697137581, "grad_norm": 1.7371689081192017, "learning_rate": 3.3979686057248385e-06, "loss": 0.0634, "step": 14300 }, { "epoch": 3.3033240997229916, "grad_norm": 8.90865707397461, "learning_rate": 3.3933518005540168e-06, "loss": 0.1409, "step": 14310 }, { "epoch": 3.3056325023084026, "grad_norm": 0.790031909942627, "learning_rate": 3.388734995383195e-06, "loss": 0.0697, "step": 14320 }, { "epoch": 3.3079409048938135, "grad_norm": 5.816183090209961, "learning_rate": 3.3841181902123732e-06, "loss": 0.0828, "step": 14330 }, { "epoch": 3.3102493074792245, "grad_norm": 2.7938485145568848, "learning_rate": 3.3795013850415515e-06, "loss": 0.1699, "step": 14340 }, { "epoch": 3.3125577100646355, "grad_norm": 2.894359588623047, "learning_rate": 3.3748845798707297e-06, "loss": 0.1017, "step": 14350 }, { "epoch": 3.314866112650046, "grad_norm": 1.35383141040802, "learning_rate": 3.370267774699908e-06, "loss": 0.1228, "step": 14360 }, { "epoch": 3.317174515235457, "grad_norm": 5.159501075744629, "learning_rate": 3.365650969529086e-06, "loss": 0.0863, "step": 14370 }, { "epoch": 3.319482917820868, "grad_norm": 0.2287338525056839, "learning_rate": 3.3610341643582644e-06, "loss": 0.1491, "step": 14380 }, { "epoch": 3.321791320406279, "grad_norm": 0.05021412670612335, "learning_rate": 3.3564173591874426e-06, "loss": 0.0741, "step": 14390 }, { "epoch": 3.32409972299169, "grad_norm": 7.520995616912842, "learning_rate": 3.351800554016621e-06, "loss": 0.1164, "step": 14400 }, { "epoch": 3.3264081255771005, "grad_norm": 3.08172869682312, "learning_rate": 3.347183748845799e-06, "loss": 0.0745, "step": 14410 }, { "epoch": 3.3287165281625115, "grad_norm": 10.243539810180664, "learning_rate": 3.342566943674977e-06, "loss": 0.0762, "step": 14420 }, { "epoch": 3.3310249307479225, "grad_norm": 0.01597224362194538, "learning_rate": 3.337950138504155e-06, "loss": 0.1802, "step": 14430 }, { "epoch": 3.3333333333333335, "grad_norm": 1.1900466680526733, "learning_rate": 3.3333333333333333e-06, "loss": 0.0729, "step": 14440 }, { "epoch": 3.3356417359187445, "grad_norm": 0.02962113916873932, "learning_rate": 3.3287165281625116e-06, "loss": 0.1395, "step": 14450 }, { "epoch": 3.337950138504155, "grad_norm": 8.536209106445312, "learning_rate": 3.3240997229916898e-06, "loss": 0.0942, "step": 14460 }, { "epoch": 3.340258541089566, "grad_norm": 10.208232879638672, "learning_rate": 3.319482917820868e-06, "loss": 0.1554, "step": 14470 }, { "epoch": 3.342566943674977, "grad_norm": 6.863511085510254, "learning_rate": 3.3148661126500462e-06, "loss": 0.0684, "step": 14480 }, { "epoch": 3.344875346260388, "grad_norm": 10.011826515197754, "learning_rate": 3.310249307479225e-06, "loss": 0.1523, "step": 14490 }, { "epoch": 3.347183748845799, "grad_norm": 13.80631160736084, "learning_rate": 3.305632502308403e-06, "loss": 0.1518, "step": 14500 }, { "epoch": 3.3494921514312095, "grad_norm": 3.74145770072937, "learning_rate": 3.3010156971375813e-06, "loss": 0.0569, "step": 14510 }, { "epoch": 3.3518005540166205, "grad_norm": 0.1434367150068283, "learning_rate": 3.2963988919667596e-06, "loss": 0.101, "step": 14520 }, { "epoch": 3.3541089566020315, "grad_norm": 1.3357571363449097, "learning_rate": 3.291782086795938e-06, "loss": 0.0386, "step": 14530 }, { "epoch": 3.3564173591874424, "grad_norm": 0.8745942711830139, "learning_rate": 3.287165281625116e-06, "loss": 0.0503, "step": 14540 }, { "epoch": 3.3587257617728534, "grad_norm": 11.09388542175293, "learning_rate": 3.282548476454294e-06, "loss": 0.2291, "step": 14550 }, { "epoch": 3.361034164358264, "grad_norm": 0.00524839386343956, "learning_rate": 3.277931671283472e-06, "loss": 0.1217, "step": 14560 }, { "epoch": 3.363342566943675, "grad_norm": 12.373751640319824, "learning_rate": 3.2733148661126503e-06, "loss": 0.1478, "step": 14570 }, { "epoch": 3.365650969529086, "grad_norm": 7.651139736175537, "learning_rate": 3.2686980609418285e-06, "loss": 0.1553, "step": 14580 }, { "epoch": 3.367959372114497, "grad_norm": 7.118612766265869, "learning_rate": 3.2640812557710068e-06, "loss": 0.0554, "step": 14590 }, { "epoch": 3.370267774699908, "grad_norm": 10.486122131347656, "learning_rate": 3.259464450600185e-06, "loss": 0.0374, "step": 14600 }, { "epoch": 3.3725761772853184, "grad_norm": 2.519848346710205, "learning_rate": 3.254847645429363e-06, "loss": 0.0404, "step": 14610 }, { "epoch": 3.3748845798707294, "grad_norm": 0.060396481305360794, "learning_rate": 3.2502308402585414e-06, "loss": 0.1099, "step": 14620 }, { "epoch": 3.3771929824561404, "grad_norm": 1.0960288047790527, "learning_rate": 3.2456140350877197e-06, "loss": 0.1348, "step": 14630 }, { "epoch": 3.3795013850415514, "grad_norm": 10.786575317382812, "learning_rate": 3.240997229916898e-06, "loss": 0.0955, "step": 14640 }, { "epoch": 3.3818097876269624, "grad_norm": 2.139821767807007, "learning_rate": 3.236380424746076e-06, "loss": 0.169, "step": 14650 }, { "epoch": 3.384118190212373, "grad_norm": 4.978335380554199, "learning_rate": 3.2317636195752544e-06, "loss": 0.0457, "step": 14660 }, { "epoch": 3.386426592797784, "grad_norm": 11.686244010925293, "learning_rate": 3.2271468144044326e-06, "loss": 0.1212, "step": 14670 }, { "epoch": 3.388734995383195, "grad_norm": 3.8788068294525146, "learning_rate": 3.2225300092336104e-06, "loss": 0.013, "step": 14680 }, { "epoch": 3.391043397968606, "grad_norm": 15.2543363571167, "learning_rate": 3.2179132040627886e-06, "loss": 0.1388, "step": 14690 }, { "epoch": 3.393351800554017, "grad_norm": 7.03340482711792, "learning_rate": 3.213296398891967e-06, "loss": 0.1229, "step": 14700 }, { "epoch": 3.3956602031394274, "grad_norm": 0.6374154090881348, "learning_rate": 3.208679593721145e-06, "loss": 0.1032, "step": 14710 }, { "epoch": 3.3979686057248384, "grad_norm": 0.780220091342926, "learning_rate": 3.2040627885503233e-06, "loss": 0.1615, "step": 14720 }, { "epoch": 3.4002770083102494, "grad_norm": 0.27147895097732544, "learning_rate": 3.1994459833795015e-06, "loss": 0.0647, "step": 14730 }, { "epoch": 3.4025854108956604, "grad_norm": 0.012577136047184467, "learning_rate": 3.1948291782086798e-06, "loss": 0.0664, "step": 14740 }, { "epoch": 3.4048938134810713, "grad_norm": 3.57993483543396, "learning_rate": 3.190212373037858e-06, "loss": 0.0817, "step": 14750 }, { "epoch": 3.407202216066482, "grad_norm": 2.7401812076568604, "learning_rate": 3.1855955678670362e-06, "loss": 0.1435, "step": 14760 }, { "epoch": 3.409510618651893, "grad_norm": 8.122635841369629, "learning_rate": 3.1809787626962144e-06, "loss": 0.0654, "step": 14770 }, { "epoch": 3.411819021237304, "grad_norm": 0.17859013378620148, "learning_rate": 3.176361957525393e-06, "loss": 0.2021, "step": 14780 }, { "epoch": 3.414127423822715, "grad_norm": 0.4929366111755371, "learning_rate": 3.1717451523545713e-06, "loss": 0.0935, "step": 14790 }, { "epoch": 3.416435826408126, "grad_norm": 13.312911987304688, "learning_rate": 3.1671283471837487e-06, "loss": 0.1066, "step": 14800 }, { "epoch": 3.4187442289935364, "grad_norm": 7.135969638824463, "learning_rate": 3.162511542012927e-06, "loss": 0.0741, "step": 14810 }, { "epoch": 3.4210526315789473, "grad_norm": 0.977172315120697, "learning_rate": 3.157894736842105e-06, "loss": 0.03, "step": 14820 }, { "epoch": 3.4233610341643583, "grad_norm": 2.9585630893707275, "learning_rate": 3.1532779316712834e-06, "loss": 0.1705, "step": 14830 }, { "epoch": 3.4256694367497693, "grad_norm": 4.252414703369141, "learning_rate": 3.148661126500462e-06, "loss": 0.041, "step": 14840 }, { "epoch": 3.4279778393351803, "grad_norm": 1.7192718982696533, "learning_rate": 3.1440443213296403e-06, "loss": 0.141, "step": 14850 }, { "epoch": 3.430286241920591, "grad_norm": 12.114724159240723, "learning_rate": 3.1394275161588185e-06, "loss": 0.3049, "step": 14860 }, { "epoch": 3.432594644506002, "grad_norm": 9.545573234558105, "learning_rate": 3.1348107109879967e-06, "loss": 0.1434, "step": 14870 }, { "epoch": 3.434903047091413, "grad_norm": 4.291155815124512, "learning_rate": 3.130193905817175e-06, "loss": 0.0823, "step": 14880 }, { "epoch": 3.437211449676824, "grad_norm": 0.006219394039362669, "learning_rate": 3.125577100646353e-06, "loss": 0.0599, "step": 14890 }, { "epoch": 3.439519852262235, "grad_norm": 0.022235281765460968, "learning_rate": 3.1209602954755314e-06, "loss": 0.0483, "step": 14900 }, { "epoch": 3.4418282548476453, "grad_norm": 1.5321310758590698, "learning_rate": 3.1163434903047096e-06, "loss": 0.09, "step": 14910 }, { "epoch": 3.4441366574330563, "grad_norm": 0.08648065477609634, "learning_rate": 3.111726685133888e-06, "loss": 0.1383, "step": 14920 }, { "epoch": 3.4464450600184673, "grad_norm": 0.003573131049051881, "learning_rate": 3.1071098799630657e-06, "loss": 0.1015, "step": 14930 }, { "epoch": 3.4487534626038783, "grad_norm": 1.8613256216049194, "learning_rate": 3.102493074792244e-06, "loss": 0.09, "step": 14940 }, { "epoch": 3.4510618651892893, "grad_norm": 0.4596211314201355, "learning_rate": 3.097876269621422e-06, "loss": 0.1186, "step": 14950 }, { "epoch": 3.4533702677747, "grad_norm": 0.012255740351974964, "learning_rate": 3.0932594644506004e-06, "loss": 0.0577, "step": 14960 }, { "epoch": 3.455678670360111, "grad_norm": 5.410879135131836, "learning_rate": 3.0886426592797786e-06, "loss": 0.2096, "step": 14970 }, { "epoch": 3.4579870729455218, "grad_norm": 2.9461112022399902, "learning_rate": 3.084025854108957e-06, "loss": 0.0756, "step": 14980 }, { "epoch": 3.4602954755309328, "grad_norm": 0.9145554900169373, "learning_rate": 3.079409048938135e-06, "loss": 0.061, "step": 14990 }, { "epoch": 3.4626038781163437, "grad_norm": 1.8118585348129272, "learning_rate": 3.0747922437673133e-06, "loss": 0.1222, "step": 15000 }, { "epoch": 3.4649122807017543, "grad_norm": 3.5605287551879883, "learning_rate": 3.0701754385964915e-06, "loss": 0.1706, "step": 15010 }, { "epoch": 3.4672206832871653, "grad_norm": 9.877135276794434, "learning_rate": 3.0655586334256697e-06, "loss": 0.261, "step": 15020 }, { "epoch": 3.4695290858725762, "grad_norm": 0.003916747402399778, "learning_rate": 3.060941828254848e-06, "loss": 0.1345, "step": 15030 }, { "epoch": 3.4718374884579872, "grad_norm": 1.626537561416626, "learning_rate": 3.056325023084026e-06, "loss": 0.0439, "step": 15040 }, { "epoch": 3.474145891043398, "grad_norm": 0.010578769259154797, "learning_rate": 3.051708217913204e-06, "loss": 0.0355, "step": 15050 }, { "epoch": 3.4764542936288088, "grad_norm": 11.334758758544922, "learning_rate": 3.0470914127423822e-06, "loss": 0.1787, "step": 15060 }, { "epoch": 3.4787626962142197, "grad_norm": 0.321943998336792, "learning_rate": 3.0424746075715605e-06, "loss": 0.085, "step": 15070 }, { "epoch": 3.4810710987996307, "grad_norm": 0.9875873923301697, "learning_rate": 3.0378578024007387e-06, "loss": 0.0748, "step": 15080 }, { "epoch": 3.4833795013850417, "grad_norm": 9.079259872436523, "learning_rate": 3.033240997229917e-06, "loss": 0.1636, "step": 15090 }, { "epoch": 3.4856879039704527, "grad_norm": 4.437586784362793, "learning_rate": 3.028624192059095e-06, "loss": 0.0873, "step": 15100 }, { "epoch": 3.4879963065558632, "grad_norm": 0.03537740930914879, "learning_rate": 3.0240073868882734e-06, "loss": 0.0435, "step": 15110 }, { "epoch": 3.490304709141274, "grad_norm": 0.39910420775413513, "learning_rate": 3.0193905817174516e-06, "loss": 0.1657, "step": 15120 }, { "epoch": 3.492613111726685, "grad_norm": 0.023489344865083694, "learning_rate": 3.0147737765466303e-06, "loss": 0.0444, "step": 15130 }, { "epoch": 3.494921514312096, "grad_norm": 6.7024993896484375, "learning_rate": 3.0101569713758085e-06, "loss": 0.0674, "step": 15140 }, { "epoch": 3.497229916897507, "grad_norm": 7.002059459686279, "learning_rate": 3.0055401662049867e-06, "loss": 0.1806, "step": 15150 }, { "epoch": 3.4995383194829177, "grad_norm": 8.779173851013184, "learning_rate": 3.000923361034165e-06, "loss": 0.0627, "step": 15160 }, { "epoch": 3.5018467220683287, "grad_norm": 8.67514419555664, "learning_rate": 2.996306555863343e-06, "loss": 0.0635, "step": 15170 }, { "epoch": 3.5041551246537397, "grad_norm": 9.346287727355957, "learning_rate": 2.991689750692521e-06, "loss": 0.115, "step": 15180 }, { "epoch": 3.5064635272391507, "grad_norm": 6.76889181137085, "learning_rate": 2.987072945521699e-06, "loss": 0.0683, "step": 15190 }, { "epoch": 3.5087719298245617, "grad_norm": 10.451617240905762, "learning_rate": 2.9824561403508774e-06, "loss": 0.2157, "step": 15200 }, { "epoch": 3.511080332409972, "grad_norm": 0.34595000743865967, "learning_rate": 2.9778393351800557e-06, "loss": 0.115, "step": 15210 }, { "epoch": 3.513388734995383, "grad_norm": 8.750661849975586, "learning_rate": 2.973222530009234e-06, "loss": 0.1266, "step": 15220 }, { "epoch": 3.515697137580794, "grad_norm": 5.47856330871582, "learning_rate": 2.968605724838412e-06, "loss": 0.0726, "step": 15230 }, { "epoch": 3.518005540166205, "grad_norm": 25.498395919799805, "learning_rate": 2.9639889196675903e-06, "loss": 0.1717, "step": 15240 }, { "epoch": 3.520313942751616, "grad_norm": 0.7776657938957214, "learning_rate": 2.9593721144967686e-06, "loss": 0.1422, "step": 15250 }, { "epoch": 3.5226223453370267, "grad_norm": 0.018797045573592186, "learning_rate": 2.954755309325947e-06, "loss": 0.1058, "step": 15260 }, { "epoch": 3.5249307479224377, "grad_norm": 0.13654901087284088, "learning_rate": 2.950138504155125e-06, "loss": 0.0159, "step": 15270 }, { "epoch": 3.5272391505078486, "grad_norm": 0.0027757291682064533, "learning_rate": 2.9455216989843033e-06, "loss": 0.0787, "step": 15280 }, { "epoch": 3.5295475530932596, "grad_norm": 4.7459821701049805, "learning_rate": 2.9409048938134815e-06, "loss": 0.1269, "step": 15290 }, { "epoch": 3.5318559556786706, "grad_norm": 0.011034720577299595, "learning_rate": 2.9362880886426597e-06, "loss": 0.0432, "step": 15300 }, { "epoch": 3.534164358264081, "grad_norm": 0.1139032393693924, "learning_rate": 2.9316712834718375e-06, "loss": 0.1202, "step": 15310 }, { "epoch": 3.536472760849492, "grad_norm": 0.29917529225349426, "learning_rate": 2.9270544783010157e-06, "loss": 0.0243, "step": 15320 }, { "epoch": 3.538781163434903, "grad_norm": 0.6147027611732483, "learning_rate": 2.922437673130194e-06, "loss": 0.1621, "step": 15330 }, { "epoch": 3.541089566020314, "grad_norm": 0.062258604913949966, "learning_rate": 2.917820867959372e-06, "loss": 0.1966, "step": 15340 }, { "epoch": 3.543397968605725, "grad_norm": 10.856746673583984, "learning_rate": 2.9132040627885504e-06, "loss": 0.2091, "step": 15350 }, { "epoch": 3.5457063711911356, "grad_norm": 0.11241677403450012, "learning_rate": 2.9085872576177287e-06, "loss": 0.1378, "step": 15360 }, { "epoch": 3.5480147737765466, "grad_norm": 0.06720955669879913, "learning_rate": 2.903970452446907e-06, "loss": 0.1179, "step": 15370 }, { "epoch": 3.5503231763619576, "grad_norm": 0.011267607100307941, "learning_rate": 2.899353647276085e-06, "loss": 0.0977, "step": 15380 }, { "epoch": 3.5526315789473686, "grad_norm": 9.50528335571289, "learning_rate": 2.8947368421052634e-06, "loss": 0.2466, "step": 15390 }, { "epoch": 3.5549399815327796, "grad_norm": 0.09559822827577591, "learning_rate": 2.8901200369344416e-06, "loss": 0.1043, "step": 15400 }, { "epoch": 3.55724838411819, "grad_norm": 0.939671516418457, "learning_rate": 2.88550323176362e-06, "loss": 0.0927, "step": 15410 }, { "epoch": 3.559556786703601, "grad_norm": 4.562279224395752, "learning_rate": 2.8808864265927985e-06, "loss": 0.0781, "step": 15420 }, { "epoch": 3.561865189289012, "grad_norm": 1.7657698392868042, "learning_rate": 2.876269621421976e-06, "loss": 0.1011, "step": 15430 }, { "epoch": 3.564173591874423, "grad_norm": 2.1192398071289062, "learning_rate": 2.871652816251154e-06, "loss": 0.1572, "step": 15440 }, { "epoch": 3.566481994459834, "grad_norm": 0.6786194443702698, "learning_rate": 2.8670360110803323e-06, "loss": 0.1445, "step": 15450 }, { "epoch": 3.5687903970452446, "grad_norm": 3.0834736824035645, "learning_rate": 2.8624192059095105e-06, "loss": 0.069, "step": 15460 }, { "epoch": 3.5710987996306556, "grad_norm": 0.0033159321174025536, "learning_rate": 2.857802400738689e-06, "loss": 0.1198, "step": 15470 }, { "epoch": 3.5734072022160666, "grad_norm": 0.0765325129032135, "learning_rate": 2.8531855955678674e-06, "loss": 0.1148, "step": 15480 }, { "epoch": 3.5757156048014775, "grad_norm": 0.3174864649772644, "learning_rate": 2.8485687903970456e-06, "loss": 0.0863, "step": 15490 }, { "epoch": 3.5780240073868885, "grad_norm": 6.153985500335693, "learning_rate": 2.843951985226224e-06, "loss": 0.1139, "step": 15500 }, { "epoch": 3.580332409972299, "grad_norm": 10.814231872558594, "learning_rate": 2.839335180055402e-06, "loss": 0.0774, "step": 15510 }, { "epoch": 3.58264081255771, "grad_norm": 6.4611101150512695, "learning_rate": 2.8347183748845803e-06, "loss": 0.115, "step": 15520 }, { "epoch": 3.584949215143121, "grad_norm": 0.04866408929228783, "learning_rate": 2.8301015697137586e-06, "loss": 0.1828, "step": 15530 }, { "epoch": 3.587257617728532, "grad_norm": 0.230972558259964, "learning_rate": 2.8254847645429368e-06, "loss": 0.0699, "step": 15540 }, { "epoch": 3.589566020313943, "grad_norm": 6.951765537261963, "learning_rate": 2.820867959372115e-06, "loss": 0.2572, "step": 15550 }, { "epoch": 3.5918744228993535, "grad_norm": 0.015099984593689442, "learning_rate": 2.816251154201293e-06, "loss": 0.0563, "step": 15560 }, { "epoch": 3.5941828254847645, "grad_norm": 8.978866577148438, "learning_rate": 2.811634349030471e-06, "loss": 0.075, "step": 15570 }, { "epoch": 3.5964912280701755, "grad_norm": 0.08920139819383621, "learning_rate": 2.8070175438596493e-06, "loss": 0.1095, "step": 15580 }, { "epoch": 3.598799630655586, "grad_norm": 3.822021484375, "learning_rate": 2.8024007386888275e-06, "loss": 0.0344, "step": 15590 }, { "epoch": 3.6011080332409975, "grad_norm": 1.2789679765701294, "learning_rate": 2.7977839335180057e-06, "loss": 0.0533, "step": 15600 }, { "epoch": 3.603416435826408, "grad_norm": 0.004921296611428261, "learning_rate": 2.793167128347184e-06, "loss": 0.0545, "step": 15610 }, { "epoch": 3.605724838411819, "grad_norm": 1.6983273029327393, "learning_rate": 2.788550323176362e-06, "loss": 0.042, "step": 15620 }, { "epoch": 3.60803324099723, "grad_norm": 1.825598955154419, "learning_rate": 2.7839335180055404e-06, "loss": 0.0311, "step": 15630 }, { "epoch": 3.6103416435826405, "grad_norm": 1.3642513751983643, "learning_rate": 2.7793167128347186e-06, "loss": 0.0742, "step": 15640 }, { "epoch": 3.612650046168052, "grad_norm": 2.4744393825531006, "learning_rate": 2.774699907663897e-06, "loss": 0.1913, "step": 15650 }, { "epoch": 3.6149584487534625, "grad_norm": 3.559018135070801, "learning_rate": 2.770083102493075e-06, "loss": 0.055, "step": 15660 }, { "epoch": 3.6172668513388735, "grad_norm": 6.687820911407471, "learning_rate": 2.7654662973222533e-06, "loss": 0.0569, "step": 15670 }, { "epoch": 3.6195752539242845, "grad_norm": 1.0267345905303955, "learning_rate": 2.760849492151431e-06, "loss": 0.1447, "step": 15680 }, { "epoch": 3.621883656509695, "grad_norm": 0.11089525371789932, "learning_rate": 2.7562326869806094e-06, "loss": 0.0286, "step": 15690 }, { "epoch": 3.6241920590951064, "grad_norm": 9.259074211120605, "learning_rate": 2.7516158818097876e-06, "loss": 0.0605, "step": 15700 }, { "epoch": 3.626500461680517, "grad_norm": 0.1097044125199318, "learning_rate": 2.746999076638966e-06, "loss": 0.0927, "step": 15710 }, { "epoch": 3.628808864265928, "grad_norm": 5.6350836753845215, "learning_rate": 2.742382271468144e-06, "loss": 0.1076, "step": 15720 }, { "epoch": 3.631117266851339, "grad_norm": 10.61784553527832, "learning_rate": 2.7377654662973223e-06, "loss": 0.119, "step": 15730 }, { "epoch": 3.6334256694367495, "grad_norm": 6.789185523986816, "learning_rate": 2.7331486611265005e-06, "loss": 0.071, "step": 15740 }, { "epoch": 3.635734072022161, "grad_norm": 6.860452651977539, "learning_rate": 2.7285318559556787e-06, "loss": 0.0738, "step": 15750 }, { "epoch": 3.6380424746075715, "grad_norm": 0.008029925636947155, "learning_rate": 2.7239150507848574e-06, "loss": 0.0833, "step": 15760 }, { "epoch": 3.6403508771929824, "grad_norm": 0.1483672857284546, "learning_rate": 2.7192982456140356e-06, "loss": 0.0807, "step": 15770 }, { "epoch": 3.6426592797783934, "grad_norm": 7.238580226898193, "learning_rate": 2.714681440443214e-06, "loss": 0.217, "step": 15780 }, { "epoch": 3.644967682363804, "grad_norm": 7.89875602722168, "learning_rate": 2.710064635272392e-06, "loss": 0.0482, "step": 15790 }, { "epoch": 3.6472760849492154, "grad_norm": 5.84787130355835, "learning_rate": 2.7054478301015703e-06, "loss": 0.1189, "step": 15800 }, { "epoch": 3.649584487534626, "grad_norm": 3.236947536468506, "learning_rate": 2.7008310249307477e-06, "loss": 0.1108, "step": 15810 }, { "epoch": 3.651892890120037, "grad_norm": 0.09606441855430603, "learning_rate": 2.6962142197599263e-06, "loss": 0.0466, "step": 15820 }, { "epoch": 3.654201292705448, "grad_norm": 1.1421186923980713, "learning_rate": 2.6915974145891046e-06, "loss": 0.1142, "step": 15830 }, { "epoch": 3.6565096952908585, "grad_norm": 6.388090133666992, "learning_rate": 2.686980609418283e-06, "loss": 0.0359, "step": 15840 }, { "epoch": 3.65881809787627, "grad_norm": 1.261206865310669, "learning_rate": 2.682363804247461e-06, "loss": 0.0535, "step": 15850 }, { "epoch": 3.6611265004616804, "grad_norm": 0.26815763115882874, "learning_rate": 2.6777469990766392e-06, "loss": 0.0937, "step": 15860 }, { "epoch": 3.6634349030470914, "grad_norm": 0.2905905246734619, "learning_rate": 2.6731301939058175e-06, "loss": 0.1274, "step": 15870 }, { "epoch": 3.6657433056325024, "grad_norm": 0.0038062131498008966, "learning_rate": 2.6685133887349957e-06, "loss": 0.0899, "step": 15880 }, { "epoch": 3.668051708217913, "grad_norm": 12.367846488952637, "learning_rate": 2.663896583564174e-06, "loss": 0.0675, "step": 15890 }, { "epoch": 3.6703601108033244, "grad_norm": 6.238649845123291, "learning_rate": 2.659279778393352e-06, "loss": 0.2909, "step": 15900 }, { "epoch": 3.672668513388735, "grad_norm": 0.0018944531911984086, "learning_rate": 2.6546629732225304e-06, "loss": 0.0848, "step": 15910 }, { "epoch": 3.674976915974146, "grad_norm": 0.8859285116195679, "learning_rate": 2.6500461680517086e-06, "loss": 0.0981, "step": 15920 }, { "epoch": 3.677285318559557, "grad_norm": 4.115336894989014, "learning_rate": 2.645429362880887e-06, "loss": 0.1095, "step": 15930 }, { "epoch": 3.6795937211449674, "grad_norm": 11.718485832214355, "learning_rate": 2.6408125577100647e-06, "loss": 0.0914, "step": 15940 }, { "epoch": 3.681902123730379, "grad_norm": 0.04772588610649109, "learning_rate": 2.636195752539243e-06, "loss": 0.2492, "step": 15950 }, { "epoch": 3.6842105263157894, "grad_norm": 0.22151795029640198, "learning_rate": 2.631578947368421e-06, "loss": 0.0826, "step": 15960 }, { "epoch": 3.6865189289012004, "grad_norm": 0.0894412249326706, "learning_rate": 2.6269621421975993e-06, "loss": 0.0768, "step": 15970 }, { "epoch": 3.6888273314866113, "grad_norm": 4.416106700897217, "learning_rate": 2.6223453370267776e-06, "loss": 0.1348, "step": 15980 }, { "epoch": 3.691135734072022, "grad_norm": 0.0027838642708957195, "learning_rate": 2.617728531855956e-06, "loss": 0.0872, "step": 15990 }, { "epoch": 3.6934441366574333, "grad_norm": 0.11230922490358353, "learning_rate": 2.613111726685134e-06, "loss": 0.0553, "step": 16000 }, { "epoch": 3.695752539242844, "grad_norm": 0.4442640244960785, "learning_rate": 2.6084949215143123e-06, "loss": 0.1334, "step": 16010 }, { "epoch": 3.698060941828255, "grad_norm": 0.23532544076442719, "learning_rate": 2.6038781163434905e-06, "loss": 0.0488, "step": 16020 }, { "epoch": 3.700369344413666, "grad_norm": 1.0589299201965332, "learning_rate": 2.5992613111726687e-06, "loss": 0.1483, "step": 16030 }, { "epoch": 3.7026777469990764, "grad_norm": 4.646289348602295, "learning_rate": 2.594644506001847e-06, "loss": 0.0782, "step": 16040 }, { "epoch": 3.704986149584488, "grad_norm": 9.458477020263672, "learning_rate": 2.5900277008310256e-06, "loss": 0.0587, "step": 16050 }, { "epoch": 3.7072945521698983, "grad_norm": 1.2729982137680054, "learning_rate": 2.585410895660203e-06, "loss": 0.2213, "step": 16060 }, { "epoch": 3.7096029547553093, "grad_norm": 3.390852689743042, "learning_rate": 2.580794090489381e-06, "loss": 0.0527, "step": 16070 }, { "epoch": 3.7119113573407203, "grad_norm": 8.382437705993652, "learning_rate": 2.5761772853185594e-06, "loss": 0.0582, "step": 16080 }, { "epoch": 3.714219759926131, "grad_norm": 0.11722227185964584, "learning_rate": 2.5715604801477377e-06, "loss": 0.1354, "step": 16090 }, { "epoch": 3.7165281625115423, "grad_norm": 1.3579716682434082, "learning_rate": 2.566943674976916e-06, "loss": 0.1118, "step": 16100 }, { "epoch": 3.718836565096953, "grad_norm": 0.34972548484802246, "learning_rate": 2.5623268698060945e-06, "loss": 0.0574, "step": 16110 }, { "epoch": 3.721144967682364, "grad_norm": 3.4655861854553223, "learning_rate": 2.5577100646352728e-06, "loss": 0.0585, "step": 16120 }, { "epoch": 3.723453370267775, "grad_norm": 2.356037139892578, "learning_rate": 2.553093259464451e-06, "loss": 0.1005, "step": 16130 }, { "epoch": 3.7257617728531853, "grad_norm": 15.801155090332031, "learning_rate": 2.5484764542936292e-06, "loss": 0.1139, "step": 16140 }, { "epoch": 3.7280701754385968, "grad_norm": 5.944174766540527, "learning_rate": 2.5438596491228075e-06, "loss": 0.0458, "step": 16150 }, { "epoch": 3.7303785780240073, "grad_norm": 0.20397046208381653, "learning_rate": 2.5392428439519857e-06, "loss": 0.0447, "step": 16160 }, { "epoch": 3.7326869806094183, "grad_norm": 9.588594436645508, "learning_rate": 2.534626038781164e-06, "loss": 0.1521, "step": 16170 }, { "epoch": 3.7349953831948293, "grad_norm": 6.161980152130127, "learning_rate": 2.530009233610342e-06, "loss": 0.166, "step": 16180 }, { "epoch": 3.73730378578024, "grad_norm": 5.900730133056641, "learning_rate": 2.52539242843952e-06, "loss": 0.0988, "step": 16190 }, { "epoch": 3.7396121883656512, "grad_norm": 0.04483022168278694, "learning_rate": 2.520775623268698e-06, "loss": 0.0716, "step": 16200 }, { "epoch": 3.7419205909510618, "grad_norm": 10.756357192993164, "learning_rate": 2.5161588180978764e-06, "loss": 0.0906, "step": 16210 }, { "epoch": 3.7442289935364728, "grad_norm": 0.21383854746818542, "learning_rate": 2.5115420129270546e-06, "loss": 0.0779, "step": 16220 }, { "epoch": 3.7465373961218837, "grad_norm": 4.471083164215088, "learning_rate": 2.506925207756233e-06, "loss": 0.058, "step": 16230 }, { "epoch": 3.7488457987072943, "grad_norm": 0.0658981129527092, "learning_rate": 2.502308402585411e-06, "loss": 0.0706, "step": 16240 }, { "epoch": 3.7511542012927057, "grad_norm": 10.466422080993652, "learning_rate": 2.4976915974145893e-06, "loss": 0.1705, "step": 16250 }, { "epoch": 3.7534626038781163, "grad_norm": 10.537128448486328, "learning_rate": 2.4930747922437675e-06, "loss": 0.061, "step": 16260 }, { "epoch": 3.7557710064635272, "grad_norm": 8.104669570922852, "learning_rate": 2.4884579870729458e-06, "loss": 0.1439, "step": 16270 }, { "epoch": 3.7580794090489382, "grad_norm": 0.3424845039844513, "learning_rate": 2.483841181902124e-06, "loss": 0.0458, "step": 16280 }, { "epoch": 3.7603878116343488, "grad_norm": 1.9809235334396362, "learning_rate": 2.4792243767313022e-06, "loss": 0.0201, "step": 16290 }, { "epoch": 3.76269621421976, "grad_norm": 10.497163772583008, "learning_rate": 2.4746075715604805e-06, "loss": 0.0654, "step": 16300 }, { "epoch": 3.7650046168051707, "grad_norm": 0.5199874639511108, "learning_rate": 2.4699907663896587e-06, "loss": 0.2, "step": 16310 }, { "epoch": 3.7673130193905817, "grad_norm": 6.773366928100586, "learning_rate": 2.465373961218837e-06, "loss": 0.1401, "step": 16320 }, { "epoch": 3.7696214219759927, "grad_norm": 0.007049506064504385, "learning_rate": 2.460757156048015e-06, "loss": 0.1863, "step": 16330 }, { "epoch": 3.7719298245614032, "grad_norm": 6.620320796966553, "learning_rate": 2.456140350877193e-06, "loss": 0.0551, "step": 16340 }, { "epoch": 3.7742382271468147, "grad_norm": 2.904639720916748, "learning_rate": 2.451523545706371e-06, "loss": 0.0757, "step": 16350 }, { "epoch": 3.776546629732225, "grad_norm": 8.675143241882324, "learning_rate": 2.4469067405355494e-06, "loss": 0.2653, "step": 16360 }, { "epoch": 3.778855032317636, "grad_norm": 5.641013145446777, "learning_rate": 2.4422899353647276e-06, "loss": 0.0425, "step": 16370 }, { "epoch": 3.781163434903047, "grad_norm": 0.05047835409641266, "learning_rate": 2.437673130193906e-06, "loss": 0.0885, "step": 16380 }, { "epoch": 3.7834718374884577, "grad_norm": 0.05300053209066391, "learning_rate": 2.4330563250230845e-06, "loss": 0.0364, "step": 16390 }, { "epoch": 3.785780240073869, "grad_norm": 0.08056043833494186, "learning_rate": 2.4284395198522627e-06, "loss": 0.0726, "step": 16400 }, { "epoch": 3.7880886426592797, "grad_norm": 0.7794252634048462, "learning_rate": 2.4238227146814406e-06, "loss": 0.0587, "step": 16410 }, { "epoch": 3.7903970452446907, "grad_norm": 2.1419363021850586, "learning_rate": 2.4192059095106188e-06, "loss": 0.0813, "step": 16420 }, { "epoch": 3.7927054478301017, "grad_norm": 2.0678517818450928, "learning_rate": 2.414589104339797e-06, "loss": 0.1013, "step": 16430 }, { "epoch": 3.795013850415512, "grad_norm": 10.26939582824707, "learning_rate": 2.4099722991689752e-06, "loss": 0.1484, "step": 16440 }, { "epoch": 3.7973222530009236, "grad_norm": 10.651616096496582, "learning_rate": 2.4053554939981535e-06, "loss": 0.0801, "step": 16450 }, { "epoch": 3.799630655586334, "grad_norm": 1.389552354812622, "learning_rate": 2.4007386888273317e-06, "loss": 0.0926, "step": 16460 }, { "epoch": 3.801939058171745, "grad_norm": 5.801025390625, "learning_rate": 2.39612188365651e-06, "loss": 0.1072, "step": 16470 }, { "epoch": 3.804247460757156, "grad_norm": 12.096196174621582, "learning_rate": 2.391505078485688e-06, "loss": 0.1237, "step": 16480 }, { "epoch": 3.8065558633425667, "grad_norm": 0.1424197554588318, "learning_rate": 2.3868882733148664e-06, "loss": 0.0906, "step": 16490 }, { "epoch": 3.808864265927978, "grad_norm": 9.554216384887695, "learning_rate": 2.3822714681440446e-06, "loss": 0.0528, "step": 16500 }, { "epoch": 3.8111726685133887, "grad_norm": 2.1867902278900146, "learning_rate": 2.377654662973223e-06, "loss": 0.1224, "step": 16510 }, { "epoch": 3.8134810710987996, "grad_norm": 1.3387869596481323, "learning_rate": 2.373037857802401e-06, "loss": 0.089, "step": 16520 }, { "epoch": 3.8157894736842106, "grad_norm": 0.10991424322128296, "learning_rate": 2.368421052631579e-06, "loss": 0.0696, "step": 16530 }, { "epoch": 3.818097876269621, "grad_norm": 6.807132244110107, "learning_rate": 2.363804247460757e-06, "loss": 0.1015, "step": 16540 }, { "epoch": 3.8204062788550326, "grad_norm": 6.293307781219482, "learning_rate": 2.3591874422899353e-06, "loss": 0.1057, "step": 16550 }, { "epoch": 3.822714681440443, "grad_norm": 1.5220431089401245, "learning_rate": 2.3545706371191136e-06, "loss": 0.0487, "step": 16560 }, { "epoch": 3.825023084025854, "grad_norm": 0.011065010912716389, "learning_rate": 2.349953831948292e-06, "loss": 0.0589, "step": 16570 }, { "epoch": 3.827331486611265, "grad_norm": 0.915833592414856, "learning_rate": 2.3453370267774704e-06, "loss": 0.1128, "step": 16580 }, { "epoch": 3.8296398891966756, "grad_norm": 9.216570854187012, "learning_rate": 2.3407202216066487e-06, "loss": 0.1074, "step": 16590 }, { "epoch": 3.831948291782087, "grad_norm": 4.286947250366211, "learning_rate": 2.3361034164358265e-06, "loss": 0.1162, "step": 16600 }, { "epoch": 3.8342566943674976, "grad_norm": 8.311838150024414, "learning_rate": 2.3314866112650047e-06, "loss": 0.1913, "step": 16610 }, { "epoch": 3.8365650969529086, "grad_norm": 1.3367292881011963, "learning_rate": 2.326869806094183e-06, "loss": 0.0653, "step": 16620 }, { "epoch": 3.8388734995383196, "grad_norm": 8.780695915222168, "learning_rate": 2.322253000923361e-06, "loss": 0.1615, "step": 16630 }, { "epoch": 3.84118190212373, "grad_norm": 2.856950283050537, "learning_rate": 2.3176361957525394e-06, "loss": 0.063, "step": 16640 }, { "epoch": 3.8434903047091415, "grad_norm": 7.390763282775879, "learning_rate": 2.3130193905817176e-06, "loss": 0.118, "step": 16650 }, { "epoch": 3.845798707294552, "grad_norm": 13.196995735168457, "learning_rate": 2.308402585410896e-06, "loss": 0.1395, "step": 16660 }, { "epoch": 3.848107109879963, "grad_norm": 0.0663808286190033, "learning_rate": 2.303785780240074e-06, "loss": 0.1051, "step": 16670 }, { "epoch": 3.850415512465374, "grad_norm": 3.9963021278381348, "learning_rate": 2.2991689750692523e-06, "loss": 0.1081, "step": 16680 }, { "epoch": 3.8527239150507846, "grad_norm": 2.4646544456481934, "learning_rate": 2.2945521698984305e-06, "loss": 0.0733, "step": 16690 }, { "epoch": 3.855032317636196, "grad_norm": 2.620331287384033, "learning_rate": 2.2899353647276088e-06, "loss": 0.0318, "step": 16700 }, { "epoch": 3.8573407202216066, "grad_norm": 4.9823760986328125, "learning_rate": 2.285318559556787e-06, "loss": 0.1085, "step": 16710 }, { "epoch": 3.8596491228070176, "grad_norm": 0.9780207872390747, "learning_rate": 2.280701754385965e-06, "loss": 0.2531, "step": 16720 }, { "epoch": 3.8619575253924285, "grad_norm": 0.053154800087213516, "learning_rate": 2.276084949215143e-06, "loss": 0.0604, "step": 16730 }, { "epoch": 3.864265927977839, "grad_norm": 10.986392974853516, "learning_rate": 2.2714681440443217e-06, "loss": 0.1026, "step": 16740 }, { "epoch": 3.8665743305632505, "grad_norm": 2.464510202407837, "learning_rate": 2.2668513388735e-06, "loss": 0.1142, "step": 16750 }, { "epoch": 3.868882733148661, "grad_norm": 0.14334821701049805, "learning_rate": 2.262234533702678e-06, "loss": 0.0575, "step": 16760 }, { "epoch": 3.871191135734072, "grad_norm": 5.617680072784424, "learning_rate": 2.2576177285318564e-06, "loss": 0.1545, "step": 16770 }, { "epoch": 3.873499538319483, "grad_norm": 0.5689931511878967, "learning_rate": 2.2530009233610346e-06, "loss": 0.0292, "step": 16780 }, { "epoch": 3.8758079409048936, "grad_norm": 0.1201648861169815, "learning_rate": 2.2483841181902124e-06, "loss": 0.0869, "step": 16790 }, { "epoch": 3.878116343490305, "grad_norm": 0.11815714836120605, "learning_rate": 2.2437673130193906e-06, "loss": 0.0072, "step": 16800 }, { "epoch": 3.8804247460757155, "grad_norm": 1.5870321989059448, "learning_rate": 2.239150507848569e-06, "loss": 0.0996, "step": 16810 }, { "epoch": 3.8827331486611265, "grad_norm": 1.9803186655044556, "learning_rate": 2.234533702677747e-06, "loss": 0.2171, "step": 16820 }, { "epoch": 3.8850415512465375, "grad_norm": 4.469898223876953, "learning_rate": 2.2299168975069253e-06, "loss": 0.1586, "step": 16830 }, { "epoch": 3.887349953831948, "grad_norm": 6.31825590133667, "learning_rate": 2.2253000923361035e-06, "loss": 0.0161, "step": 16840 }, { "epoch": 3.8896583564173595, "grad_norm": 10.417545318603516, "learning_rate": 2.2206832871652818e-06, "loss": 0.1996, "step": 16850 }, { "epoch": 3.89196675900277, "grad_norm": 0.6706535220146179, "learning_rate": 2.21606648199446e-06, "loss": 0.056, "step": 16860 }, { "epoch": 3.894275161588181, "grad_norm": 4.172085285186768, "learning_rate": 2.2114496768236382e-06, "loss": 0.1292, "step": 16870 }, { "epoch": 3.896583564173592, "grad_norm": 1.5664077997207642, "learning_rate": 2.2068328716528165e-06, "loss": 0.1767, "step": 16880 }, { "epoch": 3.8988919667590025, "grad_norm": 0.2762158513069153, "learning_rate": 2.2022160664819947e-06, "loss": 0.022, "step": 16890 }, { "epoch": 3.901200369344414, "grad_norm": 1.0671693086624146, "learning_rate": 2.197599261311173e-06, "loss": 0.0315, "step": 16900 }, { "epoch": 3.9035087719298245, "grad_norm": 0.06970557570457458, "learning_rate": 2.192982456140351e-06, "loss": 0.0347, "step": 16910 }, { "epoch": 3.9058171745152355, "grad_norm": 1.616258978843689, "learning_rate": 2.1883656509695294e-06, "loss": 0.0688, "step": 16920 }, { "epoch": 3.9081255771006465, "grad_norm": 5.086021900177002, "learning_rate": 2.1837488457987076e-06, "loss": 0.1778, "step": 16930 }, { "epoch": 3.910433979686057, "grad_norm": 10.797608375549316, "learning_rate": 2.179132040627886e-06, "loss": 0.114, "step": 16940 }, { "epoch": 3.912742382271468, "grad_norm": 3.6901097297668457, "learning_rate": 2.174515235457064e-06, "loss": 0.1104, "step": 16950 }, { "epoch": 3.915050784856879, "grad_norm": 13.063434600830078, "learning_rate": 2.1698984302862423e-06, "loss": 0.0914, "step": 16960 }, { "epoch": 3.91735918744229, "grad_norm": 4.555662631988525, "learning_rate": 2.16528162511542e-06, "loss": 0.1045, "step": 16970 }, { "epoch": 3.919667590027701, "grad_norm": 3.498223066329956, "learning_rate": 2.1606648199445983e-06, "loss": 0.1101, "step": 16980 }, { "epoch": 3.9219759926131115, "grad_norm": 14.24549674987793, "learning_rate": 2.1560480147737765e-06, "loss": 0.1992, "step": 16990 }, { "epoch": 3.9242843951985225, "grad_norm": 0.0043205199763178825, "learning_rate": 2.1514312096029548e-06, "loss": 0.0268, "step": 17000 }, { "epoch": 3.9265927977839334, "grad_norm": 1.3474671840667725, "learning_rate": 2.146814404432133e-06, "loss": 0.0457, "step": 17010 }, { "epoch": 3.9289012003693444, "grad_norm": 7.950743675231934, "learning_rate": 2.1421975992613112e-06, "loss": 0.0864, "step": 17020 }, { "epoch": 3.9312096029547554, "grad_norm": 4.570095062255859, "learning_rate": 2.13758079409049e-06, "loss": 0.1531, "step": 17030 }, { "epoch": 3.933518005540166, "grad_norm": 11.769319534301758, "learning_rate": 2.1329639889196677e-06, "loss": 0.1503, "step": 17040 }, { "epoch": 3.935826408125577, "grad_norm": 0.05433037132024765, "learning_rate": 2.128347183748846e-06, "loss": 0.0771, "step": 17050 }, { "epoch": 3.938134810710988, "grad_norm": 13.737153053283691, "learning_rate": 2.123730378578024e-06, "loss": 0.1123, "step": 17060 }, { "epoch": 3.940443213296399, "grad_norm": 2.8180816173553467, "learning_rate": 2.1191135734072024e-06, "loss": 0.0692, "step": 17070 }, { "epoch": 3.94275161588181, "grad_norm": 7.970756530761719, "learning_rate": 2.1144967682363806e-06, "loss": 0.2341, "step": 17080 }, { "epoch": 3.9450600184672204, "grad_norm": 5.409698486328125, "learning_rate": 2.109879963065559e-06, "loss": 0.1217, "step": 17090 }, { "epoch": 3.9473684210526314, "grad_norm": 0.005585066974163055, "learning_rate": 2.105263157894737e-06, "loss": 0.1562, "step": 17100 }, { "epoch": 3.9496768236380424, "grad_norm": 2.0422773361206055, "learning_rate": 2.1006463527239153e-06, "loss": 0.0777, "step": 17110 }, { "epoch": 3.9519852262234534, "grad_norm": 1.7330297231674194, "learning_rate": 2.0960295475530935e-06, "loss": 0.116, "step": 17120 }, { "epoch": 3.9542936288088644, "grad_norm": 3.7982141971588135, "learning_rate": 2.0914127423822717e-06, "loss": 0.0576, "step": 17130 }, { "epoch": 3.956602031394275, "grad_norm": 7.212771415710449, "learning_rate": 2.08679593721145e-06, "loss": 0.0893, "step": 17140 }, { "epoch": 3.958910433979686, "grad_norm": 15.783292770385742, "learning_rate": 2.082179132040628e-06, "loss": 0.2041, "step": 17150 }, { "epoch": 3.961218836565097, "grad_norm": 13.14898681640625, "learning_rate": 2.077562326869806e-06, "loss": 0.1455, "step": 17160 }, { "epoch": 3.963527239150508, "grad_norm": 0.04724092409014702, "learning_rate": 2.0729455216989842e-06, "loss": 0.0495, "step": 17170 }, { "epoch": 3.965835641735919, "grad_norm": 7.008227825164795, "learning_rate": 2.0683287165281625e-06, "loss": 0.2045, "step": 17180 }, { "epoch": 3.9681440443213294, "grad_norm": 10.595121383666992, "learning_rate": 2.0637119113573407e-06, "loss": 0.1677, "step": 17190 }, { "epoch": 3.9704524469067404, "grad_norm": 4.675482749938965, "learning_rate": 2.0590951061865193e-06, "loss": 0.0636, "step": 17200 }, { "epoch": 3.9727608494921514, "grad_norm": 0.12224403023719788, "learning_rate": 2.0544783010156976e-06, "loss": 0.1179, "step": 17210 }, { "epoch": 3.9750692520775623, "grad_norm": 12.846823692321777, "learning_rate": 2.049861495844876e-06, "loss": 0.1488, "step": 17220 }, { "epoch": 3.9773776546629733, "grad_norm": 2.5073230266571045, "learning_rate": 2.0452446906740536e-06, "loss": 0.1049, "step": 17230 }, { "epoch": 3.979686057248384, "grad_norm": 0.03585357591509819, "learning_rate": 2.040627885503232e-06, "loss": 0.1493, "step": 17240 }, { "epoch": 3.981994459833795, "grad_norm": 4.717536926269531, "learning_rate": 2.03601108033241e-06, "loss": 0.0618, "step": 17250 }, { "epoch": 3.984302862419206, "grad_norm": 4.684089183807373, "learning_rate": 2.0313942751615883e-06, "loss": 0.0866, "step": 17260 }, { "epoch": 3.986611265004617, "grad_norm": 0.0415482297539711, "learning_rate": 2.0267774699907665e-06, "loss": 0.113, "step": 17270 }, { "epoch": 3.988919667590028, "grad_norm": 14.08647346496582, "learning_rate": 2.0221606648199448e-06, "loss": 0.3745, "step": 17280 }, { "epoch": 3.9912280701754383, "grad_norm": 17.470605850219727, "learning_rate": 2.017543859649123e-06, "loss": 0.0706, "step": 17290 }, { "epoch": 3.9935364727608493, "grad_norm": 10.449420928955078, "learning_rate": 2.012927054478301e-06, "loss": 0.1789, "step": 17300 }, { "epoch": 3.9958448753462603, "grad_norm": 15.79150104522705, "learning_rate": 2.0083102493074794e-06, "loss": 0.1931, "step": 17310 }, { "epoch": 3.9981532779316713, "grad_norm": 3.792760133743286, "learning_rate": 2.0036934441366577e-06, "loss": 0.0486, "step": 17320 }, { "epoch": 4.000461680517082, "grad_norm": 0.5567103028297424, "learning_rate": 1.999076638965836e-06, "loss": 0.1129, "step": 17330 }, { "epoch": 4.002770083102493, "grad_norm": 0.05515914782881737, "learning_rate": 1.994459833795014e-06, "loss": 0.0371, "step": 17340 }, { "epoch": 4.005078485687904, "grad_norm": 7.6160054206848145, "learning_rate": 1.989843028624192e-06, "loss": 0.204, "step": 17350 }, { "epoch": 4.007386888273315, "grad_norm": 7.692667484283447, "learning_rate": 1.98522622345337e-06, "loss": 0.0946, "step": 17360 }, { "epoch": 4.009695290858725, "grad_norm": 0.4839926064014435, "learning_rate": 1.980609418282549e-06, "loss": 0.0735, "step": 17370 }, { "epoch": 4.012003693444137, "grad_norm": 2.701791763305664, "learning_rate": 1.975992613111727e-06, "loss": 0.0446, "step": 17380 }, { "epoch": 4.014312096029547, "grad_norm": 8.76486587524414, "learning_rate": 1.9713758079409053e-06, "loss": 0.1321, "step": 17390 }, { "epoch": 4.016620498614959, "grad_norm": 5.284134387969971, "learning_rate": 1.9667590027700835e-06, "loss": 0.1055, "step": 17400 }, { "epoch": 4.018928901200369, "grad_norm": 0.1691371500492096, "learning_rate": 1.9621421975992613e-06, "loss": 0.0756, "step": 17410 }, { "epoch": 4.02123730378578, "grad_norm": 11.886601448059082, "learning_rate": 1.9575253924284395e-06, "loss": 0.3296, "step": 17420 }, { "epoch": 4.023545706371191, "grad_norm": 10.572009086608887, "learning_rate": 1.9529085872576178e-06, "loss": 0.1479, "step": 17430 }, { "epoch": 4.025854108956602, "grad_norm": 0.012676590122282505, "learning_rate": 1.948291782086796e-06, "loss": 0.0801, "step": 17440 }, { "epoch": 4.028162511542013, "grad_norm": 2.1128697395324707, "learning_rate": 1.9436749769159742e-06, "loss": 0.0371, "step": 17450 }, { "epoch": 4.030470914127424, "grad_norm": 0.5411245226860046, "learning_rate": 1.9390581717451524e-06, "loss": 0.0109, "step": 17460 }, { "epoch": 4.032779316712834, "grad_norm": 8.834590911865234, "learning_rate": 1.9344413665743307e-06, "loss": 0.1718, "step": 17470 }, { "epoch": 4.035087719298246, "grad_norm": 4.457376480102539, "learning_rate": 1.929824561403509e-06, "loss": 0.1301, "step": 17480 }, { "epoch": 4.037396121883656, "grad_norm": 0.0689697191119194, "learning_rate": 1.925207756232687e-06, "loss": 0.1183, "step": 17490 }, { "epoch": 4.039704524469068, "grad_norm": 0.150483176112175, "learning_rate": 1.9205909510618654e-06, "loss": 0.0924, "step": 17500 }, { "epoch": 4.042012927054478, "grad_norm": 3.862537384033203, "learning_rate": 1.9159741458910436e-06, "loss": 0.0923, "step": 17510 }, { "epoch": 4.044321329639889, "grad_norm": 6.420804500579834, "learning_rate": 1.911357340720222e-06, "loss": 0.0522, "step": 17520 }, { "epoch": 4.0466297322253, "grad_norm": 7.919189453125, "learning_rate": 1.9067405355494e-06, "loss": 0.0548, "step": 17530 }, { "epoch": 4.048938134810711, "grad_norm": 1.2139182090759277, "learning_rate": 1.902123730378578e-06, "loss": 0.0761, "step": 17540 }, { "epoch": 4.051246537396122, "grad_norm": 0.019797885790467262, "learning_rate": 1.8975069252077563e-06, "loss": 0.2109, "step": 17550 }, { "epoch": 4.053554939981533, "grad_norm": 12.0720796585083, "learning_rate": 1.8928901200369345e-06, "loss": 0.0886, "step": 17560 }, { "epoch": 4.055863342566943, "grad_norm": 6.289129734039307, "learning_rate": 1.8882733148661127e-06, "loss": 0.1237, "step": 17570 }, { "epoch": 4.058171745152355, "grad_norm": 0.06416813284158707, "learning_rate": 1.883656509695291e-06, "loss": 0.1428, "step": 17580 }, { "epoch": 4.060480147737765, "grad_norm": 2.9999382495880127, "learning_rate": 1.8790397045244694e-06, "loss": 0.0816, "step": 17590 }, { "epoch": 4.062788550323177, "grad_norm": 4.974567413330078, "learning_rate": 1.8744228993536472e-06, "loss": 0.1655, "step": 17600 }, { "epoch": 4.065096952908587, "grad_norm": 5.961575031280518, "learning_rate": 1.8698060941828254e-06, "loss": 0.1642, "step": 17610 }, { "epoch": 4.067405355493998, "grad_norm": 7.574702262878418, "learning_rate": 1.8651892890120039e-06, "loss": 0.1049, "step": 17620 }, { "epoch": 4.069713758079409, "grad_norm": 0.34732726216316223, "learning_rate": 1.8605724838411821e-06, "loss": 0.0489, "step": 17630 }, { "epoch": 4.07202216066482, "grad_norm": 0.03999030590057373, "learning_rate": 1.8559556786703603e-06, "loss": 0.0798, "step": 17640 }, { "epoch": 4.074330563250231, "grad_norm": 7.563187122344971, "learning_rate": 1.8513388734995386e-06, "loss": 0.0564, "step": 17650 }, { "epoch": 4.076638965835642, "grad_norm": 0.022275689989328384, "learning_rate": 1.8467220683287168e-06, "loss": 0.0841, "step": 17660 }, { "epoch": 4.078947368421052, "grad_norm": 0.14487551152706146, "learning_rate": 1.8421052631578948e-06, "loss": 0.0609, "step": 17670 }, { "epoch": 4.081255771006464, "grad_norm": 11.210583686828613, "learning_rate": 1.837488457987073e-06, "loss": 0.0955, "step": 17680 }, { "epoch": 4.083564173591874, "grad_norm": 13.005208015441895, "learning_rate": 1.8328716528162513e-06, "loss": 0.0856, "step": 17690 }, { "epoch": 4.085872576177286, "grad_norm": 0.009737399406731129, "learning_rate": 1.8282548476454295e-06, "loss": 0.061, "step": 17700 }, { "epoch": 4.088180978762696, "grad_norm": 0.18837927281856537, "learning_rate": 1.8236380424746077e-06, "loss": 0.0244, "step": 17710 }, { "epoch": 4.090489381348107, "grad_norm": 0.06044956296682358, "learning_rate": 1.819021237303786e-06, "loss": 0.1074, "step": 17720 }, { "epoch": 4.092797783933518, "grad_norm": 0.04484780132770538, "learning_rate": 1.814404432132964e-06, "loss": 0.0527, "step": 17730 }, { "epoch": 4.095106186518929, "grad_norm": 2.089848041534424, "learning_rate": 1.8097876269621422e-06, "loss": 0.0477, "step": 17740 }, { "epoch": 4.09741458910434, "grad_norm": 0.8328580260276794, "learning_rate": 1.8051708217913204e-06, "loss": 0.0627, "step": 17750 }, { "epoch": 4.099722991689751, "grad_norm": 4.158858776092529, "learning_rate": 1.8005540166204989e-06, "loss": 0.1899, "step": 17760 }, { "epoch": 4.102031394275161, "grad_norm": 4.327850341796875, "learning_rate": 1.7959372114496771e-06, "loss": 0.1697, "step": 17770 }, { "epoch": 4.104339796860573, "grad_norm": 7.390630722045898, "learning_rate": 1.7913204062788553e-06, "loss": 0.0682, "step": 17780 }, { "epoch": 4.106648199445983, "grad_norm": 0.24370385706424713, "learning_rate": 1.7867036011080334e-06, "loss": 0.0586, "step": 17790 }, { "epoch": 4.108956602031395, "grad_norm": 0.3343433439731598, "learning_rate": 1.7820867959372116e-06, "loss": 0.0269, "step": 17800 }, { "epoch": 4.111265004616805, "grad_norm": 1.1882919073104858, "learning_rate": 1.7774699907663898e-06, "loss": 0.0825, "step": 17810 }, { "epoch": 4.113573407202216, "grad_norm": 5.757431983947754, "learning_rate": 1.772853185595568e-06, "loss": 0.1235, "step": 17820 }, { "epoch": 4.115881809787627, "grad_norm": 1.8218317031860352, "learning_rate": 1.7682363804247463e-06, "loss": 0.0971, "step": 17830 }, { "epoch": 4.118190212373038, "grad_norm": 2.606081247329712, "learning_rate": 1.7636195752539245e-06, "loss": 0.0706, "step": 17840 }, { "epoch": 4.120498614958449, "grad_norm": 1.4898006916046143, "learning_rate": 1.7590027700831027e-06, "loss": 0.0727, "step": 17850 }, { "epoch": 4.12280701754386, "grad_norm": 0.18175561726093292, "learning_rate": 1.7543859649122807e-06, "loss": 0.1295, "step": 17860 }, { "epoch": 4.12511542012927, "grad_norm": 8.361217498779297, "learning_rate": 1.749769159741459e-06, "loss": 0.1003, "step": 17870 }, { "epoch": 4.127423822714682, "grad_norm": 13.528752326965332, "learning_rate": 1.7451523545706372e-06, "loss": 0.0388, "step": 17880 }, { "epoch": 4.129732225300092, "grad_norm": 0.17072995007038116, "learning_rate": 1.7405355493998154e-06, "loss": 0.1673, "step": 17890 }, { "epoch": 4.1320406278855035, "grad_norm": 9.364889144897461, "learning_rate": 1.7359187442289937e-06, "loss": 0.157, "step": 17900 }, { "epoch": 4.134349030470914, "grad_norm": 9.533273696899414, "learning_rate": 1.731301939058172e-06, "loss": 0.1244, "step": 17910 }, { "epoch": 4.136657433056325, "grad_norm": 0.2143106609582901, "learning_rate": 1.72668513388735e-06, "loss": 0.1403, "step": 17920 }, { "epoch": 4.138965835641736, "grad_norm": 0.4234224855899811, "learning_rate": 1.7220683287165283e-06, "loss": 0.0936, "step": 17930 }, { "epoch": 4.141274238227147, "grad_norm": 0.0015247706323862076, "learning_rate": 1.7174515235457066e-06, "loss": 0.0346, "step": 17940 }, { "epoch": 4.143582640812558, "grad_norm": 8.444857597351074, "learning_rate": 1.7128347183748848e-06, "loss": 0.0839, "step": 17950 }, { "epoch": 4.1458910433979685, "grad_norm": 2.9204680919647217, "learning_rate": 1.708217913204063e-06, "loss": 0.0416, "step": 17960 }, { "epoch": 4.148199445983379, "grad_norm": 0.23721843957901, "learning_rate": 1.7036011080332413e-06, "loss": 0.0273, "step": 17970 }, { "epoch": 4.1505078485687905, "grad_norm": 0.3443264961242676, "learning_rate": 1.6989843028624193e-06, "loss": 0.0662, "step": 17980 }, { "epoch": 4.152816251154201, "grad_norm": 1.521956205368042, "learning_rate": 1.6943674976915975e-06, "loss": 0.1224, "step": 17990 }, { "epoch": 4.1551246537396125, "grad_norm": 6.6898579597473145, "learning_rate": 1.6897506925207757e-06, "loss": 0.0727, "step": 18000 }, { "epoch": 4.157433056325023, "grad_norm": 0.749911367893219, "learning_rate": 1.685133887349954e-06, "loss": 0.0231, "step": 18010 }, { "epoch": 4.159741458910434, "grad_norm": 2.4709060192108154, "learning_rate": 1.6805170821791322e-06, "loss": 0.0401, "step": 18020 }, { "epoch": 4.162049861495845, "grad_norm": 0.5970932841300964, "learning_rate": 1.6759002770083104e-06, "loss": 0.0615, "step": 18030 }, { "epoch": 4.1643582640812555, "grad_norm": 1.4250962734222412, "learning_rate": 1.6712834718374884e-06, "loss": 0.1061, "step": 18040 }, { "epoch": 4.166666666666667, "grad_norm": 7.599595069885254, "learning_rate": 1.6666666666666667e-06, "loss": 0.1943, "step": 18050 }, { "epoch": 4.1689750692520775, "grad_norm": 0.1758354753255844, "learning_rate": 1.6620498614958449e-06, "loss": 0.1029, "step": 18060 }, { "epoch": 4.171283471837488, "grad_norm": 0.019367141649127007, "learning_rate": 1.6574330563250231e-06, "loss": 0.0501, "step": 18070 }, { "epoch": 4.1735918744228995, "grad_norm": 8.911870827432722e-05, "learning_rate": 1.6528162511542016e-06, "loss": 0.1196, "step": 18080 }, { "epoch": 4.17590027700831, "grad_norm": 9.273355484008789, "learning_rate": 1.6481994459833798e-06, "loss": 0.1282, "step": 18090 }, { "epoch": 4.178208679593721, "grad_norm": 2.6502904891967773, "learning_rate": 1.643582640812558e-06, "loss": 0.1614, "step": 18100 }, { "epoch": 4.180517082179132, "grad_norm": 9.313275337219238, "learning_rate": 1.638965835641736e-06, "loss": 0.1026, "step": 18110 }, { "epoch": 4.1828254847645425, "grad_norm": 6.981542587280273, "learning_rate": 1.6343490304709143e-06, "loss": 0.0976, "step": 18120 }, { "epoch": 4.185133887349954, "grad_norm": 7.823934555053711, "learning_rate": 1.6297322253000925e-06, "loss": 0.1702, "step": 18130 }, { "epoch": 4.1874422899353645, "grad_norm": 0.015078183263540268, "learning_rate": 1.6251154201292707e-06, "loss": 0.079, "step": 18140 }, { "epoch": 4.189750692520776, "grad_norm": 0.5002915263175964, "learning_rate": 1.620498614958449e-06, "loss": 0.015, "step": 18150 }, { "epoch": 4.1920590951061865, "grad_norm": 3.6707632541656494, "learning_rate": 1.6158818097876272e-06, "loss": 0.06, "step": 18160 }, { "epoch": 4.194367497691597, "grad_norm": 7.1134185791015625, "learning_rate": 1.6112650046168052e-06, "loss": 0.1528, "step": 18170 }, { "epoch": 4.196675900277008, "grad_norm": 0.5840227007865906, "learning_rate": 1.6066481994459834e-06, "loss": 0.0746, "step": 18180 }, { "epoch": 4.198984302862419, "grad_norm": 7.29752779006958, "learning_rate": 1.6020313942751617e-06, "loss": 0.1147, "step": 18190 }, { "epoch": 4.20129270544783, "grad_norm": 2.0548648834228516, "learning_rate": 1.5974145891043399e-06, "loss": 0.0285, "step": 18200 }, { "epoch": 4.203601108033241, "grad_norm": 12.844947814941406, "learning_rate": 1.5927977839335181e-06, "loss": 0.1169, "step": 18210 }, { "epoch": 4.2059095106186515, "grad_norm": 0.6720438003540039, "learning_rate": 1.5881809787626965e-06, "loss": 0.1875, "step": 18220 }, { "epoch": 4.208217913204063, "grad_norm": 0.4104534983634949, "learning_rate": 1.5835641735918744e-06, "loss": 0.0932, "step": 18230 }, { "epoch": 4.2105263157894735, "grad_norm": 9.412288665771484, "learning_rate": 1.5789473684210526e-06, "loss": 0.0587, "step": 18240 }, { "epoch": 4.212834718374885, "grad_norm": 11.267598152160645, "learning_rate": 1.574330563250231e-06, "loss": 0.1059, "step": 18250 }, { "epoch": 4.215143120960295, "grad_norm": 0.21640585362911224, "learning_rate": 1.5697137580794093e-06, "loss": 0.1436, "step": 18260 }, { "epoch": 4.217451523545706, "grad_norm": 0.047597870230674744, "learning_rate": 1.5650969529085875e-06, "loss": 0.0213, "step": 18270 }, { "epoch": 4.219759926131117, "grad_norm": 1.3516931533813477, "learning_rate": 1.5604801477377657e-06, "loss": 0.1453, "step": 18280 }, { "epoch": 4.222068328716528, "grad_norm": 2.1602530479431152, "learning_rate": 1.555863342566944e-06, "loss": 0.0831, "step": 18290 }, { "epoch": 4.224376731301939, "grad_norm": 4.4824748039245605, "learning_rate": 1.551246537396122e-06, "loss": 0.0992, "step": 18300 }, { "epoch": 4.22668513388735, "grad_norm": 0.6479166150093079, "learning_rate": 1.5466297322253002e-06, "loss": 0.0479, "step": 18310 }, { "epoch": 4.22899353647276, "grad_norm": 0.003623060416430235, "learning_rate": 1.5420129270544784e-06, "loss": 0.0726, "step": 18320 }, { "epoch": 4.231301939058172, "grad_norm": 1.3259658813476562, "learning_rate": 1.5373961218836566e-06, "loss": 0.1497, "step": 18330 }, { "epoch": 4.233610341643582, "grad_norm": 6.869663715362549, "learning_rate": 1.5327793167128349e-06, "loss": 0.0566, "step": 18340 }, { "epoch": 4.235918744228994, "grad_norm": 0.3856455981731415, "learning_rate": 1.528162511542013e-06, "loss": 0.0742, "step": 18350 }, { "epoch": 4.238227146814404, "grad_norm": 1.9238159656524658, "learning_rate": 1.5235457063711911e-06, "loss": 0.1013, "step": 18360 }, { "epoch": 4.240535549399815, "grad_norm": 0.0026226695626974106, "learning_rate": 1.5189289012003693e-06, "loss": 0.055, "step": 18370 }, { "epoch": 4.242843951985226, "grad_norm": 4.205273628234863, "learning_rate": 1.5143120960295476e-06, "loss": 0.0365, "step": 18380 }, { "epoch": 4.245152354570637, "grad_norm": 0.03291506692767143, "learning_rate": 1.5096952908587258e-06, "loss": 0.0219, "step": 18390 }, { "epoch": 4.247460757156048, "grad_norm": 3.9213287830352783, "learning_rate": 1.5050784856879042e-06, "loss": 0.0936, "step": 18400 }, { "epoch": 4.249769159741459, "grad_norm": 0.001965222880244255, "learning_rate": 1.5004616805170825e-06, "loss": 0.2006, "step": 18410 }, { "epoch": 4.252077562326869, "grad_norm": 8.61193561553955, "learning_rate": 1.4958448753462605e-06, "loss": 0.1056, "step": 18420 }, { "epoch": 4.254385964912281, "grad_norm": 0.0003543222264852375, "learning_rate": 1.4912280701754387e-06, "loss": 0.1152, "step": 18430 }, { "epoch": 4.256694367497691, "grad_norm": 0.03553229942917824, "learning_rate": 1.486611265004617e-06, "loss": 0.1278, "step": 18440 }, { "epoch": 4.259002770083103, "grad_norm": 0.3012474775314331, "learning_rate": 1.4819944598337952e-06, "loss": 0.0447, "step": 18450 }, { "epoch": 4.261311172668513, "grad_norm": 0.0004998059594072402, "learning_rate": 1.4773776546629734e-06, "loss": 0.0848, "step": 18460 }, { "epoch": 4.263619575253924, "grad_norm": 0.02571015991270542, "learning_rate": 1.4727608494921516e-06, "loss": 0.054, "step": 18470 }, { "epoch": 4.265927977839335, "grad_norm": 1.6443620920181274, "learning_rate": 1.4681440443213299e-06, "loss": 0.1124, "step": 18480 }, { "epoch": 4.268236380424746, "grad_norm": 2.0169291496276855, "learning_rate": 1.4635272391505079e-06, "loss": 0.0635, "step": 18490 }, { "epoch": 4.270544783010157, "grad_norm": 0.0017683185869827867, "learning_rate": 1.458910433979686e-06, "loss": 0.0545, "step": 18500 }, { "epoch": 4.272853185595568, "grad_norm": 7.548868656158447, "learning_rate": 1.4542936288088643e-06, "loss": 0.1501, "step": 18510 }, { "epoch": 4.275161588180978, "grad_norm": 7.761920928955078, "learning_rate": 1.4496768236380426e-06, "loss": 0.1276, "step": 18520 }, { "epoch": 4.27746999076639, "grad_norm": 8.311779022216797, "learning_rate": 1.4450600184672208e-06, "loss": 0.0804, "step": 18530 }, { "epoch": 4.2797783933518, "grad_norm": 0.03365530073642731, "learning_rate": 1.4404432132963992e-06, "loss": 0.0866, "step": 18540 }, { "epoch": 4.282086795937212, "grad_norm": 0.11477167904376984, "learning_rate": 1.435826408125577e-06, "loss": 0.1802, "step": 18550 }, { "epoch": 4.284395198522622, "grad_norm": 11.957294464111328, "learning_rate": 1.4312096029547553e-06, "loss": 0.1311, "step": 18560 }, { "epoch": 4.286703601108033, "grad_norm": 0.3100244104862213, "learning_rate": 1.4265927977839337e-06, "loss": 0.1249, "step": 18570 }, { "epoch": 4.289012003693444, "grad_norm": 3.5263280868530273, "learning_rate": 1.421975992613112e-06, "loss": 0.026, "step": 18580 }, { "epoch": 4.291320406278855, "grad_norm": 5.12006139755249, "learning_rate": 1.4173591874422902e-06, "loss": 0.2176, "step": 18590 }, { "epoch": 4.293628808864266, "grad_norm": 2.9324889183044434, "learning_rate": 1.4127423822714684e-06, "loss": 0.0984, "step": 18600 }, { "epoch": 4.295937211449677, "grad_norm": 6.247832298278809, "learning_rate": 1.4081255771006464e-06, "loss": 0.0625, "step": 18610 }, { "epoch": 4.298245614035087, "grad_norm": 0.5543584227561951, "learning_rate": 1.4035087719298246e-06, "loss": 0.0226, "step": 18620 }, { "epoch": 4.300554016620499, "grad_norm": 0.045898132026195526, "learning_rate": 1.3988919667590029e-06, "loss": 0.0512, "step": 18630 }, { "epoch": 4.302862419205909, "grad_norm": 0.8284197449684143, "learning_rate": 1.394275161588181e-06, "loss": 0.0806, "step": 18640 }, { "epoch": 4.305170821791321, "grad_norm": 1.3980050086975098, "learning_rate": 1.3896583564173593e-06, "loss": 0.0398, "step": 18650 }, { "epoch": 4.307479224376731, "grad_norm": 0.285629004240036, "learning_rate": 1.3850415512465375e-06, "loss": 0.0898, "step": 18660 }, { "epoch": 4.309787626962142, "grad_norm": 10.167963027954102, "learning_rate": 1.3804247460757156e-06, "loss": 0.0465, "step": 18670 }, { "epoch": 4.312096029547553, "grad_norm": 0.05650853365659714, "learning_rate": 1.3758079409048938e-06, "loss": 0.1037, "step": 18680 }, { "epoch": 4.314404432132964, "grad_norm": 0.3463301360607147, "learning_rate": 1.371191135734072e-06, "loss": 0.1745, "step": 18690 }, { "epoch": 4.316712834718375, "grad_norm": 0.7532675862312317, "learning_rate": 1.3665743305632503e-06, "loss": 0.0531, "step": 18700 }, { "epoch": 4.319021237303786, "grad_norm": 0.707317590713501, "learning_rate": 1.3619575253924287e-06, "loss": 0.0314, "step": 18710 }, { "epoch": 4.321329639889196, "grad_norm": 0.03190390765666962, "learning_rate": 1.357340720221607e-06, "loss": 0.0543, "step": 18720 }, { "epoch": 4.323638042474608, "grad_norm": 1.1432627439498901, "learning_rate": 1.3527239150507852e-06, "loss": 0.08, "step": 18730 }, { "epoch": 4.325946445060018, "grad_norm": 0.03065003640949726, "learning_rate": 1.3481071098799632e-06, "loss": 0.1757, "step": 18740 }, { "epoch": 4.32825484764543, "grad_norm": 1.1875349283218384, "learning_rate": 1.3434903047091414e-06, "loss": 0.1004, "step": 18750 }, { "epoch": 4.33056325023084, "grad_norm": 0.18091168999671936, "learning_rate": 1.3388734995383196e-06, "loss": 0.0426, "step": 18760 }, { "epoch": 4.332871652816251, "grad_norm": 11.40159797668457, "learning_rate": 1.3342566943674979e-06, "loss": 0.0911, "step": 18770 }, { "epoch": 4.335180055401662, "grad_norm": 0.5495089292526245, "learning_rate": 1.329639889196676e-06, "loss": 0.0757, "step": 18780 }, { "epoch": 4.337488457987073, "grad_norm": 0.38059499859809875, "learning_rate": 1.3250230840258543e-06, "loss": 0.0755, "step": 18790 }, { "epoch": 4.339796860572484, "grad_norm": 11.468230247497559, "learning_rate": 1.3204062788550323e-06, "loss": 0.1349, "step": 18800 }, { "epoch": 4.342105263157895, "grad_norm": 2.189826488494873, "learning_rate": 1.3157894736842106e-06, "loss": 0.1038, "step": 18810 }, { "epoch": 4.344413665743305, "grad_norm": 5.178741931915283, "learning_rate": 1.3111726685133888e-06, "loss": 0.1205, "step": 18820 }, { "epoch": 4.346722068328717, "grad_norm": 0.009501563385128975, "learning_rate": 1.306555863342567e-06, "loss": 0.0772, "step": 18830 }, { "epoch": 4.349030470914127, "grad_norm": 9.182662010192871, "learning_rate": 1.3019390581717452e-06, "loss": 0.1399, "step": 18840 }, { "epoch": 4.351338873499539, "grad_norm": 2.9312639236450195, "learning_rate": 1.2973222530009235e-06, "loss": 0.0711, "step": 18850 }, { "epoch": 4.353647276084949, "grad_norm": 13.856345176696777, "learning_rate": 1.2927054478301015e-06, "loss": 0.1443, "step": 18860 }, { "epoch": 4.35595567867036, "grad_norm": 0.026102133095264435, "learning_rate": 1.2880886426592797e-06, "loss": 0.1102, "step": 18870 }, { "epoch": 4.358264081255771, "grad_norm": 10.875211715698242, "learning_rate": 1.283471837488458e-06, "loss": 0.1348, "step": 18880 }, { "epoch": 4.360572483841182, "grad_norm": 0.7638143301010132, "learning_rate": 1.2788550323176364e-06, "loss": 0.1005, "step": 18890 }, { "epoch": 4.362880886426593, "grad_norm": 15.387251853942871, "learning_rate": 1.2742382271468146e-06, "loss": 0.1619, "step": 18900 }, { "epoch": 4.365189289012004, "grad_norm": 0.012922042049467564, "learning_rate": 1.2696214219759928e-06, "loss": 0.0777, "step": 18910 }, { "epoch": 4.367497691597414, "grad_norm": 9.09872055053711, "learning_rate": 1.265004616805171e-06, "loss": 0.1269, "step": 18920 }, { "epoch": 4.369806094182826, "grad_norm": 15.013482093811035, "learning_rate": 1.260387811634349e-06, "loss": 0.1903, "step": 18930 }, { "epoch": 4.372114496768236, "grad_norm": 1.723223328590393, "learning_rate": 1.2557710064635273e-06, "loss": 0.1051, "step": 18940 }, { "epoch": 4.374422899353648, "grad_norm": 0.036664124578237534, "learning_rate": 1.2511542012927055e-06, "loss": 0.0776, "step": 18950 }, { "epoch": 4.376731301939058, "grad_norm": 0.4533037543296814, "learning_rate": 1.2465373961218838e-06, "loss": 0.0254, "step": 18960 }, { "epoch": 4.379039704524469, "grad_norm": 11.614164352416992, "learning_rate": 1.241920590951062e-06, "loss": 0.1115, "step": 18970 }, { "epoch": 4.38134810710988, "grad_norm": 2.6653497219085693, "learning_rate": 1.2373037857802402e-06, "loss": 0.1085, "step": 18980 }, { "epoch": 4.383656509695291, "grad_norm": 1.6848649978637695, "learning_rate": 1.2326869806094185e-06, "loss": 0.0464, "step": 18990 }, { "epoch": 4.385964912280702, "grad_norm": 0.36225515604019165, "learning_rate": 1.2280701754385965e-06, "loss": 0.0312, "step": 19000 }, { "epoch": 4.388273314866113, "grad_norm": 8.291478157043457, "learning_rate": 1.2234533702677747e-06, "loss": 0.1256, "step": 19010 }, { "epoch": 4.390581717451523, "grad_norm": 0.14928275346755981, "learning_rate": 1.218836565096953e-06, "loss": 0.0202, "step": 19020 }, { "epoch": 4.392890120036935, "grad_norm": 0.11870255321264267, "learning_rate": 1.2142197599261314e-06, "loss": 0.1611, "step": 19030 }, { "epoch": 4.395198522622345, "grad_norm": 0.08544490486383438, "learning_rate": 1.2096029547553094e-06, "loss": 0.0334, "step": 19040 }, { "epoch": 4.3975069252077565, "grad_norm": 4.483710289001465, "learning_rate": 1.2049861495844876e-06, "loss": 0.0538, "step": 19050 }, { "epoch": 4.399815327793167, "grad_norm": 0.07368674874305725, "learning_rate": 1.2003693444136658e-06, "loss": 0.0837, "step": 19060 }, { "epoch": 4.402123730378578, "grad_norm": 6.6922926902771, "learning_rate": 1.195752539242844e-06, "loss": 0.1468, "step": 19070 }, { "epoch": 4.404432132963989, "grad_norm": 0.22682537138462067, "learning_rate": 1.1911357340720223e-06, "loss": 0.1224, "step": 19080 }, { "epoch": 4.4067405355494, "grad_norm": 0.2628570795059204, "learning_rate": 1.1865189289012005e-06, "loss": 0.0377, "step": 19090 }, { "epoch": 4.409048938134811, "grad_norm": 0.6884846687316895, "learning_rate": 1.1819021237303786e-06, "loss": 0.0802, "step": 19100 }, { "epoch": 4.411357340720222, "grad_norm": 0.24220415949821472, "learning_rate": 1.1772853185595568e-06, "loss": 0.0431, "step": 19110 }, { "epoch": 4.413665743305632, "grad_norm": 0.0012206712272018194, "learning_rate": 1.1726685133887352e-06, "loss": 0.113, "step": 19120 }, { "epoch": 4.4159741458910435, "grad_norm": 0.010857267305254936, "learning_rate": 1.1680517082179132e-06, "loss": 0.1441, "step": 19130 }, { "epoch": 4.418282548476454, "grad_norm": 8.617463111877441, "learning_rate": 1.1634349030470915e-06, "loss": 0.0969, "step": 19140 }, { "epoch": 4.4205909510618655, "grad_norm": 5.146607875823975, "learning_rate": 1.1588180978762697e-06, "loss": 0.02, "step": 19150 }, { "epoch": 4.422899353647276, "grad_norm": 0.26189592480659485, "learning_rate": 1.154201292705448e-06, "loss": 0.0311, "step": 19160 }, { "epoch": 4.425207756232687, "grad_norm": 7.718921184539795, "learning_rate": 1.1495844875346262e-06, "loss": 0.1022, "step": 19170 }, { "epoch": 4.427516158818098, "grad_norm": 2.4889941215515137, "learning_rate": 1.1449676823638044e-06, "loss": 0.0521, "step": 19180 }, { "epoch": 4.4298245614035086, "grad_norm": 2.502051591873169, "learning_rate": 1.1403508771929824e-06, "loss": 0.0608, "step": 19190 }, { "epoch": 4.43213296398892, "grad_norm": 0.08653046935796738, "learning_rate": 1.1357340720221608e-06, "loss": 0.0878, "step": 19200 }, { "epoch": 4.4344413665743305, "grad_norm": 1.4791200160980225, "learning_rate": 1.131117266851339e-06, "loss": 0.1725, "step": 19210 }, { "epoch": 4.436749769159741, "grad_norm": 7.290171146392822, "learning_rate": 1.1265004616805173e-06, "loss": 0.0668, "step": 19220 }, { "epoch": 4.4390581717451525, "grad_norm": 1.1946375370025635, "learning_rate": 1.1218836565096953e-06, "loss": 0.1583, "step": 19230 }, { "epoch": 4.441366574330563, "grad_norm": 2.806880235671997, "learning_rate": 1.1172668513388735e-06, "loss": 0.185, "step": 19240 }, { "epoch": 4.4436749769159745, "grad_norm": 0.06182056665420532, "learning_rate": 1.1126500461680518e-06, "loss": 0.0462, "step": 19250 }, { "epoch": 4.445983379501385, "grad_norm": 0.002978579606860876, "learning_rate": 1.10803324099723e-06, "loss": 0.07, "step": 19260 }, { "epoch": 4.4482917820867955, "grad_norm": 0.003829601686447859, "learning_rate": 1.1034164358264082e-06, "loss": 0.0548, "step": 19270 }, { "epoch": 4.450600184672207, "grad_norm": 16.720489501953125, "learning_rate": 1.0987996306555865e-06, "loss": 0.2019, "step": 19280 }, { "epoch": 4.4529085872576175, "grad_norm": 0.0027888657059520483, "learning_rate": 1.0941828254847647e-06, "loss": 0.0074, "step": 19290 }, { "epoch": 4.455216989843029, "grad_norm": 10.662086486816406, "learning_rate": 1.089566020313943e-06, "loss": 0.1479, "step": 19300 }, { "epoch": 4.4575253924284395, "grad_norm": 0.003766958834603429, "learning_rate": 1.0849492151431211e-06, "loss": 0.1004, "step": 19310 }, { "epoch": 4.45983379501385, "grad_norm": 1.0769058465957642, "learning_rate": 1.0803324099722992e-06, "loss": 0.0615, "step": 19320 }, { "epoch": 4.4621421975992615, "grad_norm": 0.16774635016918182, "learning_rate": 1.0757156048014774e-06, "loss": 0.0265, "step": 19330 }, { "epoch": 4.464450600184672, "grad_norm": 0.08873479068279266, "learning_rate": 1.0710987996306556e-06, "loss": 0.1162, "step": 19340 }, { "epoch": 4.466759002770083, "grad_norm": 0.5571467876434326, "learning_rate": 1.0664819944598338e-06, "loss": 0.0407, "step": 19350 }, { "epoch": 4.469067405355494, "grad_norm": 0.019094835966825485, "learning_rate": 1.061865189289012e-06, "loss": 0.1246, "step": 19360 }, { "epoch": 4.4713758079409045, "grad_norm": 8.37247085571289, "learning_rate": 1.0572483841181903e-06, "loss": 0.0413, "step": 19370 }, { "epoch": 4.473684210526316, "grad_norm": 18.04896354675293, "learning_rate": 1.0526315789473685e-06, "loss": 0.0736, "step": 19380 }, { "epoch": 4.4759926131117265, "grad_norm": 2.294590950012207, "learning_rate": 1.0480147737765468e-06, "loss": 0.0591, "step": 19390 }, { "epoch": 4.478301015697138, "grad_norm": 2.386199712753296, "learning_rate": 1.043397968605725e-06, "loss": 0.1517, "step": 19400 }, { "epoch": 4.480609418282548, "grad_norm": 0.054773665964603424, "learning_rate": 1.038781163434903e-06, "loss": 0.0803, "step": 19410 }, { "epoch": 4.482917820867959, "grad_norm": 6.038883686065674, "learning_rate": 1.0341643582640812e-06, "loss": 0.0406, "step": 19420 }, { "epoch": 4.48522622345337, "grad_norm": 6.4689412117004395, "learning_rate": 1.0295475530932597e-06, "loss": 0.0617, "step": 19430 }, { "epoch": 4.487534626038781, "grad_norm": 0.3081076741218567, "learning_rate": 1.024930747922438e-06, "loss": 0.0852, "step": 19440 }, { "epoch": 4.489843028624192, "grad_norm": 11.17514419555664, "learning_rate": 1.020313942751616e-06, "loss": 0.1791, "step": 19450 }, { "epoch": 4.492151431209603, "grad_norm": 0.5829828381538391, "learning_rate": 1.0156971375807941e-06, "loss": 0.1899, "step": 19460 }, { "epoch": 4.4944598337950135, "grad_norm": 5.834979057312012, "learning_rate": 1.0110803324099724e-06, "loss": 0.0804, "step": 19470 }, { "epoch": 4.496768236380425, "grad_norm": 7.586325168609619, "learning_rate": 1.0064635272391506e-06, "loss": 0.0406, "step": 19480 }, { "epoch": 4.499076638965835, "grad_norm": 9.50064468383789, "learning_rate": 1.0018467220683288e-06, "loss": 0.074, "step": 19490 }, { "epoch": 4.501385041551247, "grad_norm": 0.0067270915023982525, "learning_rate": 9.97229916897507e-07, "loss": 0.0677, "step": 19500 }, { "epoch": 4.503693444136657, "grad_norm": 0.010083766654133797, "learning_rate": 9.92613111726685e-07, "loss": 0.1195, "step": 19510 }, { "epoch": 4.506001846722068, "grad_norm": 9.135502815246582, "learning_rate": 9.879963065558635e-07, "loss": 0.1195, "step": 19520 }, { "epoch": 4.508310249307479, "grad_norm": 6.888404846191406, "learning_rate": 9.833795013850417e-07, "loss": 0.044, "step": 19530 }, { "epoch": 4.51061865189289, "grad_norm": 0.008340762928128242, "learning_rate": 9.787626962142198e-07, "loss": 0.0918, "step": 19540 }, { "epoch": 4.512927054478301, "grad_norm": 0.08295182138681412, "learning_rate": 9.74145891043398e-07, "loss": 0.14, "step": 19550 }, { "epoch": 4.515235457063712, "grad_norm": 0.00909690372645855, "learning_rate": 9.695290858725762e-07, "loss": 0.0964, "step": 19560 }, { "epoch": 4.517543859649123, "grad_norm": 0.36127379536628723, "learning_rate": 9.649122807017545e-07, "loss": 0.0595, "step": 19570 }, { "epoch": 4.519852262234534, "grad_norm": 0.10293540358543396, "learning_rate": 9.602954755309327e-07, "loss": 0.1007, "step": 19580 }, { "epoch": 4.522160664819944, "grad_norm": 8.337013244628906, "learning_rate": 9.55678670360111e-07, "loss": 0.0435, "step": 19590 }, { "epoch": 4.524469067405356, "grad_norm": 7.4408955574035645, "learning_rate": 9.51061865189289e-07, "loss": 0.0312, "step": 19600 }, { "epoch": 4.526777469990766, "grad_norm": 0.755298912525177, "learning_rate": 9.464450600184673e-07, "loss": 0.0997, "step": 19610 }, { "epoch": 4.529085872576177, "grad_norm": 5.594273567199707, "learning_rate": 9.418282548476455e-07, "loss": 0.0936, "step": 19620 }, { "epoch": 4.531394275161588, "grad_norm": 8.472539901733398, "learning_rate": 9.372114496768236e-07, "loss": 0.1007, "step": 19630 }, { "epoch": 4.533702677746999, "grad_norm": 0.7538583874702454, "learning_rate": 9.325946445060019e-07, "loss": 0.0897, "step": 19640 }, { "epoch": 4.53601108033241, "grad_norm": 0.34898680448532104, "learning_rate": 9.279778393351802e-07, "loss": 0.1016, "step": 19650 }, { "epoch": 4.538319482917821, "grad_norm": 0.07763586938381195, "learning_rate": 9.233610341643584e-07, "loss": 0.1774, "step": 19660 }, { "epoch": 4.540627885503231, "grad_norm": 2.740216016769409, "learning_rate": 9.187442289935365e-07, "loss": 0.0373, "step": 19670 }, { "epoch": 4.542936288088643, "grad_norm": 0.03562672436237335, "learning_rate": 9.141274238227148e-07, "loss": 0.1049, "step": 19680 }, { "epoch": 4.545244690674053, "grad_norm": 0.684116542339325, "learning_rate": 9.09510618651893e-07, "loss": 0.1576, "step": 19690 }, { "epoch": 4.547553093259465, "grad_norm": 11.174668312072754, "learning_rate": 9.048938134810711e-07, "loss": 0.1779, "step": 19700 }, { "epoch": 4.549861495844875, "grad_norm": 12.515663146972656, "learning_rate": 9.002770083102494e-07, "loss": 0.1352, "step": 19710 }, { "epoch": 4.552169898430286, "grad_norm": 0.4817483127117157, "learning_rate": 8.956602031394277e-07, "loss": 0.0364, "step": 19720 }, { "epoch": 4.554478301015697, "grad_norm": 0.08345703780651093, "learning_rate": 8.910433979686058e-07, "loss": 0.0753, "step": 19730 }, { "epoch": 4.556786703601108, "grad_norm": 5.428301811218262, "learning_rate": 8.86426592797784e-07, "loss": 0.0569, "step": 19740 }, { "epoch": 4.559095106186519, "grad_norm": 0.022812476381659508, "learning_rate": 8.818097876269622e-07, "loss": 0.0874, "step": 19750 }, { "epoch": 4.56140350877193, "grad_norm": 9.988215446472168, "learning_rate": 8.771929824561404e-07, "loss": 0.2024, "step": 19760 }, { "epoch": 4.56371191135734, "grad_norm": 1.9239765405654907, "learning_rate": 8.725761772853186e-07, "loss": 0.1324, "step": 19770 }, { "epoch": 4.566020313942752, "grad_norm": 0.22725607454776764, "learning_rate": 8.679593721144968e-07, "loss": 0.0592, "step": 19780 }, { "epoch": 4.568328716528162, "grad_norm": 4.433720111846924, "learning_rate": 8.63342566943675e-07, "loss": 0.1111, "step": 19790 }, { "epoch": 4.570637119113574, "grad_norm": 0.4482247829437256, "learning_rate": 8.587257617728533e-07, "loss": 0.0489, "step": 19800 }, { "epoch": 4.572945521698984, "grad_norm": 0.5948168039321899, "learning_rate": 8.541089566020315e-07, "loss": 0.0361, "step": 19810 }, { "epoch": 4.575253924284395, "grad_norm": 0.11281322687864304, "learning_rate": 8.494921514312096e-07, "loss": 0.1049, "step": 19820 }, { "epoch": 4.577562326869806, "grad_norm": 15.677630424499512, "learning_rate": 8.448753462603879e-07, "loss": 0.1681, "step": 19830 }, { "epoch": 4.579870729455217, "grad_norm": 5.464344501495361, "learning_rate": 8.402585410895661e-07, "loss": 0.0883, "step": 19840 }, { "epoch": 4.582179132040628, "grad_norm": 8.245606422424316, "learning_rate": 8.356417359187442e-07, "loss": 0.0481, "step": 19850 }, { "epoch": 4.584487534626039, "grad_norm": 0.34970471262931824, "learning_rate": 8.310249307479224e-07, "loss": 0.1155, "step": 19860 }, { "epoch": 4.586795937211449, "grad_norm": 1.144962191581726, "learning_rate": 8.264081255771008e-07, "loss": 0.0813, "step": 19870 }, { "epoch": 4.589104339796861, "grad_norm": 0.7480217218399048, "learning_rate": 8.21791320406279e-07, "loss": 0.0698, "step": 19880 }, { "epoch": 4.591412742382271, "grad_norm": 1.2967151403427124, "learning_rate": 8.171745152354571e-07, "loss": 0.1008, "step": 19890 }, { "epoch": 4.593721144967683, "grad_norm": 0.4681199789047241, "learning_rate": 8.125577100646354e-07, "loss": 0.2091, "step": 19900 }, { "epoch": 4.596029547553093, "grad_norm": 0.5502877235412598, "learning_rate": 8.079409048938136e-07, "loss": 0.1269, "step": 19910 }, { "epoch": 4.598337950138504, "grad_norm": 3.06742525100708, "learning_rate": 8.033240997229917e-07, "loss": 0.1394, "step": 19920 }, { "epoch": 4.600646352723915, "grad_norm": 2.2365834712982178, "learning_rate": 7.987072945521699e-07, "loss": 0.1324, "step": 19930 }, { "epoch": 4.602954755309326, "grad_norm": 6.717630863189697, "learning_rate": 7.940904893813483e-07, "loss": 0.1003, "step": 19940 }, { "epoch": 4.605263157894737, "grad_norm": 0.40891891717910767, "learning_rate": 7.894736842105263e-07, "loss": 0.0148, "step": 19950 }, { "epoch": 4.607571560480148, "grad_norm": 8.703094482421875, "learning_rate": 7.848568790397046e-07, "loss": 0.1084, "step": 19960 }, { "epoch": 4.609879963065558, "grad_norm": 0.43557193875312805, "learning_rate": 7.802400738688829e-07, "loss": 0.0919, "step": 19970 }, { "epoch": 4.61218836565097, "grad_norm": 0.0077831982634961605, "learning_rate": 7.75623268698061e-07, "loss": 0.163, "step": 19980 }, { "epoch": 4.61449676823638, "grad_norm": 0.0003239473153371364, "learning_rate": 7.710064635272392e-07, "loss": 0.0903, "step": 19990 }, { "epoch": 4.616805170821792, "grad_norm": 0.139947772026062, "learning_rate": 7.663896583564174e-07, "loss": 0.1546, "step": 20000 }, { "epoch": 4.619113573407202, "grad_norm": 1.185225486755371, "learning_rate": 7.617728531855956e-07, "loss": 0.1097, "step": 20010 }, { "epoch": 4.621421975992613, "grad_norm": 2.5803513526916504, "learning_rate": 7.571560480147738e-07, "loss": 0.0869, "step": 20020 }, { "epoch": 4.623730378578024, "grad_norm": 0.020152220502495766, "learning_rate": 7.525392428439521e-07, "loss": 0.0836, "step": 20030 }, { "epoch": 4.626038781163435, "grad_norm": 0.0038780674804002047, "learning_rate": 7.479224376731302e-07, "loss": 0.0142, "step": 20040 }, { "epoch": 4.628347183748846, "grad_norm": 12.621441841125488, "learning_rate": 7.433056325023085e-07, "loss": 0.0584, "step": 20050 }, { "epoch": 4.630655586334257, "grad_norm": 15.029406547546387, "learning_rate": 7.386888273314867e-07, "loss": 0.0944, "step": 20060 }, { "epoch": 4.632963988919667, "grad_norm": 12.991839408874512, "learning_rate": 7.340720221606649e-07, "loss": 0.084, "step": 20070 }, { "epoch": 4.635272391505079, "grad_norm": 0.4588876962661743, "learning_rate": 7.29455216989843e-07, "loss": 0.0783, "step": 20080 }, { "epoch": 4.637580794090489, "grad_norm": 0.018094852566719055, "learning_rate": 7.248384118190213e-07, "loss": 0.0833, "step": 20090 }, { "epoch": 4.639889196675901, "grad_norm": 10.912367820739746, "learning_rate": 7.202216066481996e-07, "loss": 0.2626, "step": 20100 }, { "epoch": 4.642197599261311, "grad_norm": 18.362335205078125, "learning_rate": 7.156048014773776e-07, "loss": 0.1586, "step": 20110 }, { "epoch": 4.644506001846722, "grad_norm": 21.22774887084961, "learning_rate": 7.10987996306556e-07, "loss": 0.1198, "step": 20120 }, { "epoch": 4.646814404432133, "grad_norm": 0.0047484454698860645, "learning_rate": 7.063711911357342e-07, "loss": 0.222, "step": 20130 }, { "epoch": 4.649122807017544, "grad_norm": 2.948723793029785, "learning_rate": 7.017543859649123e-07, "loss": 0.1411, "step": 20140 }, { "epoch": 4.651431209602955, "grad_norm": 0.5264045596122742, "learning_rate": 6.971375807940905e-07, "loss": 0.0097, "step": 20150 }, { "epoch": 4.653739612188366, "grad_norm": 2.253751516342163, "learning_rate": 6.925207756232688e-07, "loss": 0.1362, "step": 20160 }, { "epoch": 4.656048014773776, "grad_norm": 6.633666038513184, "learning_rate": 6.879039704524469e-07, "loss": 0.0996, "step": 20170 }, { "epoch": 4.658356417359188, "grad_norm": 0.6632509827613831, "learning_rate": 6.832871652816251e-07, "loss": 0.0798, "step": 20180 }, { "epoch": 4.660664819944598, "grad_norm": 4.981107711791992, "learning_rate": 6.786703601108035e-07, "loss": 0.0758, "step": 20190 }, { "epoch": 4.66297322253001, "grad_norm": 9.494646072387695, "learning_rate": 6.740535549399816e-07, "loss": 0.0525, "step": 20200 }, { "epoch": 4.66528162511542, "grad_norm": 7.384730815887451, "learning_rate": 6.694367497691598e-07, "loss": 0.1978, "step": 20210 }, { "epoch": 4.667590027700831, "grad_norm": 6.360969066619873, "learning_rate": 6.64819944598338e-07, "loss": 0.09, "step": 20220 }, { "epoch": 4.669898430286242, "grad_norm": 12.208895683288574, "learning_rate": 6.602031394275162e-07, "loss": 0.1932, "step": 20230 }, { "epoch": 4.672206832871653, "grad_norm": 10.788138389587402, "learning_rate": 6.555863342566944e-07, "loss": 0.1152, "step": 20240 }, { "epoch": 4.674515235457064, "grad_norm": 6.225203990936279, "learning_rate": 6.509695290858726e-07, "loss": 0.0651, "step": 20250 }, { "epoch": 4.676823638042475, "grad_norm": 0.04782375693321228, "learning_rate": 6.463527239150507e-07, "loss": 0.0555, "step": 20260 }, { "epoch": 4.679132040627885, "grad_norm": 9.396204948425293, "learning_rate": 6.41735918744229e-07, "loss": 0.106, "step": 20270 }, { "epoch": 4.6814404432132966, "grad_norm": 0.1343054473400116, "learning_rate": 6.371191135734073e-07, "loss": 0.2473, "step": 20280 }, { "epoch": 4.683748845798707, "grad_norm": 0.08104892075061798, "learning_rate": 6.325023084025855e-07, "loss": 0.0642, "step": 20290 }, { "epoch": 4.6860572483841185, "grad_norm": 8.993147850036621, "learning_rate": 6.278855032317637e-07, "loss": 0.1304, "step": 20300 }, { "epoch": 4.688365650969529, "grad_norm": 0.38738447427749634, "learning_rate": 6.232686980609419e-07, "loss": 0.0752, "step": 20310 }, { "epoch": 4.69067405355494, "grad_norm": 0.5128989219665527, "learning_rate": 6.186518928901201e-07, "loss": 0.0396, "step": 20320 }, { "epoch": 4.692982456140351, "grad_norm": 6.087538719177246, "learning_rate": 6.140350877192982e-07, "loss": 0.0464, "step": 20330 }, { "epoch": 4.695290858725762, "grad_norm": 9.987509727478027, "learning_rate": 6.094182825484765e-07, "loss": 0.049, "step": 20340 }, { "epoch": 4.697599261311173, "grad_norm": 0.502367377281189, "learning_rate": 6.048014773776547e-07, "loss": 0.0843, "step": 20350 }, { "epoch": 4.6999076638965835, "grad_norm": 10.970908164978027, "learning_rate": 6.001846722068329e-07, "loss": 0.0589, "step": 20360 }, { "epoch": 4.702216066481994, "grad_norm": 3.8757376670837402, "learning_rate": 5.955678670360112e-07, "loss": 0.0488, "step": 20370 }, { "epoch": 4.7045244690674055, "grad_norm": 3.9657037258148193, "learning_rate": 5.909510618651893e-07, "loss": 0.0417, "step": 20380 }, { "epoch": 4.706832871652816, "grad_norm": 0.14002129435539246, "learning_rate": 5.863342566943676e-07, "loss": 0.0518, "step": 20390 }, { "epoch": 4.7091412742382275, "grad_norm": 0.011854655109345913, "learning_rate": 5.817174515235457e-07, "loss": 0.094, "step": 20400 }, { "epoch": 4.711449676823638, "grad_norm": 7.836629867553711, "learning_rate": 5.77100646352724e-07, "loss": 0.0698, "step": 20410 }, { "epoch": 4.713758079409049, "grad_norm": 0.5280254483222961, "learning_rate": 5.724838411819022e-07, "loss": 0.1219, "step": 20420 }, { "epoch": 4.71606648199446, "grad_norm": 4.025824546813965, "learning_rate": 5.678670360110804e-07, "loss": 0.0875, "step": 20430 }, { "epoch": 4.7183748845798705, "grad_norm": 0.5322102308273315, "learning_rate": 5.632502308402586e-07, "loss": 0.1294, "step": 20440 }, { "epoch": 4.720683287165282, "grad_norm": 0.7324974536895752, "learning_rate": 5.586334256694368e-07, "loss": 0.0306, "step": 20450 }, { "epoch": 4.7229916897506925, "grad_norm": 0.259613573551178, "learning_rate": 5.54016620498615e-07, "loss": 0.0572, "step": 20460 }, { "epoch": 4.725300092336103, "grad_norm": 0.17638278007507324, "learning_rate": 5.493998153277932e-07, "loss": 0.0413, "step": 20470 }, { "epoch": 4.7276084949215145, "grad_norm": 8.890875816345215, "learning_rate": 5.447830101569715e-07, "loss": 0.1008, "step": 20480 }, { "epoch": 4.729916897506925, "grad_norm": 0.48535436391830444, "learning_rate": 5.401662049861496e-07, "loss": 0.0584, "step": 20490 }, { "epoch": 4.732225300092336, "grad_norm": 4.537534236907959, "learning_rate": 5.355493998153278e-07, "loss": 0.0453, "step": 20500 }, { "epoch": 4.734533702677747, "grad_norm": 0.912034273147583, "learning_rate": 5.30932594644506e-07, "loss": 0.202, "step": 20510 }, { "epoch": 4.7368421052631575, "grad_norm": 10.345976829528809, "learning_rate": 5.263157894736843e-07, "loss": 0.0534, "step": 20520 }, { "epoch": 4.739150507848569, "grad_norm": 1.0590749979019165, "learning_rate": 5.216989843028625e-07, "loss": 0.0454, "step": 20530 }, { "epoch": 4.7414589104339795, "grad_norm": 1.1391432285308838, "learning_rate": 5.170821791320406e-07, "loss": 0.1757, "step": 20540 }, { "epoch": 4.743767313019391, "grad_norm": 0.007721558213233948, "learning_rate": 5.12465373961219e-07, "loss": 0.0672, "step": 20550 }, { "epoch": 4.7460757156048015, "grad_norm": 2.642233371734619, "learning_rate": 5.078485687903971e-07, "loss": 0.0376, "step": 20560 }, { "epoch": 4.748384118190212, "grad_norm": 5.2271575927734375, "learning_rate": 5.032317636195753e-07, "loss": 0.1353, "step": 20570 }, { "epoch": 4.750692520775623, "grad_norm": 10.931775093078613, "learning_rate": 4.986149584487535e-07, "loss": 0.1866, "step": 20580 }, { "epoch": 4.753000923361034, "grad_norm": 0.07494260370731354, "learning_rate": 4.939981532779318e-07, "loss": 0.068, "step": 20590 }, { "epoch": 4.755309325946445, "grad_norm": 4.855413913726807, "learning_rate": 4.893813481071099e-07, "loss": 0.0854, "step": 20600 }, { "epoch": 4.757617728531856, "grad_norm": 0.2879292964935303, "learning_rate": 4.847645429362881e-07, "loss": 0.0491, "step": 20610 }, { "epoch": 4.7599261311172665, "grad_norm": 9.912734985351562, "learning_rate": 4.801477377654663e-07, "loss": 0.2316, "step": 20620 }, { "epoch": 4.762234533702678, "grad_norm": 0.07501928508281708, "learning_rate": 4.755309325946445e-07, "loss": 0.0644, "step": 20630 }, { "epoch": 4.7645429362880884, "grad_norm": 3.7826623916625977, "learning_rate": 4.7091412742382274e-07, "loss": 0.0966, "step": 20640 }, { "epoch": 4.7668513388735, "grad_norm": 7.7684006690979, "learning_rate": 4.6629732225300097e-07, "loss": 0.1305, "step": 20650 }, { "epoch": 4.76915974145891, "grad_norm": 8.805274963378906, "learning_rate": 4.616805170821792e-07, "loss": 0.1821, "step": 20660 }, { "epoch": 4.771468144044321, "grad_norm": 0.16345404088497162, "learning_rate": 4.570637119113574e-07, "loss": 0.0689, "step": 20670 }, { "epoch": 4.773776546629732, "grad_norm": 11.634016990661621, "learning_rate": 4.5244690674053555e-07, "loss": 0.0416, "step": 20680 }, { "epoch": 4.776084949215143, "grad_norm": 1.3202722072601318, "learning_rate": 4.4783010156971383e-07, "loss": 0.0983, "step": 20690 }, { "epoch": 4.778393351800554, "grad_norm": 7.432985305786133, "learning_rate": 4.43213296398892e-07, "loss": 0.0958, "step": 20700 }, { "epoch": 4.780701754385965, "grad_norm": 0.4654413163661957, "learning_rate": 4.385964912280702e-07, "loss": 0.0863, "step": 20710 }, { "epoch": 4.783010156971375, "grad_norm": 5.309205055236816, "learning_rate": 4.339796860572484e-07, "loss": 0.1204, "step": 20720 }, { "epoch": 4.785318559556787, "grad_norm": 0.16075949370861053, "learning_rate": 4.2936288088642664e-07, "loss": 0.0769, "step": 20730 }, { "epoch": 4.787626962142197, "grad_norm": 0.794579267501831, "learning_rate": 4.247460757156048e-07, "loss": 0.0557, "step": 20740 }, { "epoch": 4.789935364727609, "grad_norm": 2.8942463397979736, "learning_rate": 4.2012927054478305e-07, "loss": 0.1536, "step": 20750 }, { "epoch": 4.792243767313019, "grad_norm": 7.191786766052246, "learning_rate": 4.155124653739612e-07, "loss": 0.06, "step": 20760 }, { "epoch": 4.79455216989843, "grad_norm": 1.1968438625335693, "learning_rate": 4.108956602031395e-07, "loss": 0.048, "step": 20770 }, { "epoch": 4.796860572483841, "grad_norm": 0.27007827162742615, "learning_rate": 4.062788550323177e-07, "loss": 0.0487, "step": 20780 }, { "epoch": 4.799168975069252, "grad_norm": 0.5440189838409424, "learning_rate": 4.0166204986149586e-07, "loss": 0.1343, "step": 20790 }, { "epoch": 4.801477377654663, "grad_norm": 1.352490782737732, "learning_rate": 3.9704524469067414e-07, "loss": 0.0318, "step": 20800 }, { "epoch": 4.803785780240074, "grad_norm": 2.4697468280792236, "learning_rate": 3.924284395198523e-07, "loss": 0.0778, "step": 20810 }, { "epoch": 4.806094182825484, "grad_norm": 1.888457179069519, "learning_rate": 3.878116343490305e-07, "loss": 0.0543, "step": 20820 }, { "epoch": 4.808402585410896, "grad_norm": 1.7298115491867065, "learning_rate": 3.831948291782087e-07, "loss": 0.0248, "step": 20830 }, { "epoch": 4.810710987996306, "grad_norm": 7.328147888183594, "learning_rate": 3.785780240073869e-07, "loss": 0.0514, "step": 20840 }, { "epoch": 4.813019390581718, "grad_norm": 0.3497525453567505, "learning_rate": 3.739612188365651e-07, "loss": 0.0104, "step": 20850 }, { "epoch": 4.815327793167128, "grad_norm": 0.24612508714199066, "learning_rate": 3.6934441366574335e-07, "loss": 0.0414, "step": 20860 }, { "epoch": 4.817636195752539, "grad_norm": 7.450209617614746, "learning_rate": 3.647276084949215e-07, "loss": 0.0574, "step": 20870 }, { "epoch": 4.81994459833795, "grad_norm": 0.0030814206693321466, "learning_rate": 3.601108033240998e-07, "loss": 0.2245, "step": 20880 }, { "epoch": 4.822253000923361, "grad_norm": 13.84593677520752, "learning_rate": 3.55493998153278e-07, "loss": 0.1905, "step": 20890 }, { "epoch": 4.824561403508772, "grad_norm": 9.884056091308594, "learning_rate": 3.5087719298245616e-07, "loss": 0.0928, "step": 20900 }, { "epoch": 4.826869806094183, "grad_norm": 12.368430137634277, "learning_rate": 3.462603878116344e-07, "loss": 0.1702, "step": 20910 }, { "epoch": 4.829178208679593, "grad_norm": 0.004777924157679081, "learning_rate": 3.4164358264081256e-07, "loss": 0.0305, "step": 20920 }, { "epoch": 4.831486611265005, "grad_norm": 1.671309232711792, "learning_rate": 3.370267774699908e-07, "loss": 0.0577, "step": 20930 }, { "epoch": 4.833795013850415, "grad_norm": 2.441744089126587, "learning_rate": 3.32409972299169e-07, "loss": 0.0481, "step": 20940 }, { "epoch": 4.836103416435827, "grad_norm": 6.544705867767334, "learning_rate": 3.277931671283472e-07, "loss": 0.1265, "step": 20950 }, { "epoch": 4.838411819021237, "grad_norm": 1.5451968908309937, "learning_rate": 3.2317636195752537e-07, "loss": 0.0498, "step": 20960 }, { "epoch": 4.840720221606648, "grad_norm": 0.747820258140564, "learning_rate": 3.1855955678670365e-07, "loss": 0.0316, "step": 20970 }, { "epoch": 4.843028624192059, "grad_norm": 11.4855318069458, "learning_rate": 3.1394275161588183e-07, "loss": 0.2469, "step": 20980 }, { "epoch": 4.84533702677747, "grad_norm": 11.642210960388184, "learning_rate": 3.0932594644506006e-07, "loss": 0.0909, "step": 20990 }, { "epoch": 4.847645429362881, "grad_norm": 0.7561269402503967, "learning_rate": 3.0470914127423823e-07, "loss": 0.0744, "step": 21000 }, { "epoch": 4.849953831948292, "grad_norm": 1.744206428527832, "learning_rate": 3.0009233610341646e-07, "loss": 0.0484, "step": 21010 }, { "epoch": 4.852262234533702, "grad_norm": 9.418831825256348, "learning_rate": 2.9547553093259464e-07, "loss": 0.0779, "step": 21020 }, { "epoch": 4.854570637119114, "grad_norm": 11.402624130249023, "learning_rate": 2.9085872576177287e-07, "loss": 0.0689, "step": 21030 }, { "epoch": 4.856879039704524, "grad_norm": 3.7089037895202637, "learning_rate": 2.862419205909511e-07, "loss": 0.0937, "step": 21040 }, { "epoch": 4.859187442289936, "grad_norm": 1.1794441938400269, "learning_rate": 2.816251154201293e-07, "loss": 0.0197, "step": 21050 }, { "epoch": 4.861495844875346, "grad_norm": 11.01574993133545, "learning_rate": 2.770083102493075e-07, "loss": 0.0489, "step": 21060 }, { "epoch": 4.863804247460757, "grad_norm": 2.568364381790161, "learning_rate": 2.7239150507848573e-07, "loss": 0.034, "step": 21070 }, { "epoch": 4.866112650046168, "grad_norm": 7.690331935882568, "learning_rate": 2.677746999076639e-07, "loss": 0.0587, "step": 21080 }, { "epoch": 4.868421052631579, "grad_norm": 0.007998675107955933, "learning_rate": 2.6315789473684213e-07, "loss": 0.0348, "step": 21090 }, { "epoch": 4.87072945521699, "grad_norm": 8.070531845092773, "learning_rate": 2.585410895660203e-07, "loss": 0.1067, "step": 21100 }, { "epoch": 4.873037857802401, "grad_norm": 0.05041287839412689, "learning_rate": 2.5392428439519854e-07, "loss": 0.0613, "step": 21110 }, { "epoch": 4.875346260387811, "grad_norm": 10.597179412841797, "learning_rate": 2.4930747922437677e-07, "loss": 0.1482, "step": 21120 }, { "epoch": 4.877654662973223, "grad_norm": 1.2775416374206543, "learning_rate": 2.4469067405355494e-07, "loss": 0.0622, "step": 21130 }, { "epoch": 4.879963065558633, "grad_norm": 0.0014759311452507973, "learning_rate": 2.4007386888273317e-07, "loss": 0.0273, "step": 21140 }, { "epoch": 4.882271468144045, "grad_norm": 0.2902587056159973, "learning_rate": 2.3545706371191137e-07, "loss": 0.0176, "step": 21150 }, { "epoch": 4.884579870729455, "grad_norm": 6.944189548492432, "learning_rate": 2.308402585410896e-07, "loss": 0.0724, "step": 21160 }, { "epoch": 4.886888273314866, "grad_norm": 11.473503112792969, "learning_rate": 2.2622345337026778e-07, "loss": 0.1155, "step": 21170 }, { "epoch": 4.889196675900277, "grad_norm": 0.3364136815071106, "learning_rate": 2.21606648199446e-07, "loss": 0.0767, "step": 21180 }, { "epoch": 4.891505078485688, "grad_norm": 0.0041140224784612656, "learning_rate": 2.169898430286242e-07, "loss": 0.0508, "step": 21190 }, { "epoch": 4.893813481071099, "grad_norm": 0.10550861060619354, "learning_rate": 2.123730378578024e-07, "loss": 0.0777, "step": 21200 }, { "epoch": 4.89612188365651, "grad_norm": 11.309544563293457, "learning_rate": 2.077562326869806e-07, "loss": 0.0837, "step": 21210 }, { "epoch": 4.89843028624192, "grad_norm": 0.0018829823238775134, "learning_rate": 2.0313942751615884e-07, "loss": 0.1165, "step": 21220 }, { "epoch": 4.900738688827332, "grad_norm": 0.0050180647522211075, "learning_rate": 1.9852262234533707e-07, "loss": 0.0188, "step": 21230 }, { "epoch": 4.903047091412742, "grad_norm": 2.4538888931274414, "learning_rate": 1.9390581717451524e-07, "loss": 0.0303, "step": 21240 }, { "epoch": 4.905355493998154, "grad_norm": 0.2505981922149658, "learning_rate": 1.8928901200369345e-07, "loss": 0.1006, "step": 21250 }, { "epoch": 4.907663896583564, "grad_norm": 2.9865217208862305, "learning_rate": 1.8467220683287168e-07, "loss": 0.0787, "step": 21260 }, { "epoch": 4.909972299168975, "grad_norm": 0.11575239896774292, "learning_rate": 1.800554016620499e-07, "loss": 0.0297, "step": 21270 }, { "epoch": 4.912280701754386, "grad_norm": 5.265392780303955, "learning_rate": 1.7543859649122808e-07, "loss": 0.1201, "step": 21280 }, { "epoch": 4.914589104339797, "grad_norm": 4.96639347076416, "learning_rate": 1.7082179132040628e-07, "loss": 0.0309, "step": 21290 }, { "epoch": 4.916897506925208, "grad_norm": 13.720710754394531, "learning_rate": 1.662049861495845e-07, "loss": 0.0833, "step": 21300 }, { "epoch": 4.919205909510619, "grad_norm": 0.30121421813964844, "learning_rate": 1.6158818097876269e-07, "loss": 0.0186, "step": 21310 }, { "epoch": 4.921514312096029, "grad_norm": 4.877224445343018, "learning_rate": 1.5697137580794091e-07, "loss": 0.0775, "step": 21320 }, { "epoch": 4.923822714681441, "grad_norm": 1.3914650678634644, "learning_rate": 1.5235457063711912e-07, "loss": 0.0693, "step": 21330 }, { "epoch": 4.926131117266851, "grad_norm": 10.681194305419922, "learning_rate": 1.4773776546629732e-07, "loss": 0.1837, "step": 21340 }, { "epoch": 4.928439519852263, "grad_norm": 7.103105068206787, "learning_rate": 1.4312096029547555e-07, "loss": 0.1013, "step": 21350 }, { "epoch": 4.930747922437673, "grad_norm": 6.090207576751709, "learning_rate": 1.3850415512465375e-07, "loss": 0.2113, "step": 21360 }, { "epoch": 4.933056325023084, "grad_norm": 11.182722091674805, "learning_rate": 1.3388734995383195e-07, "loss": 0.1107, "step": 21370 }, { "epoch": 4.935364727608495, "grad_norm": 10.59288501739502, "learning_rate": 1.2927054478301015e-07, "loss": 0.1848, "step": 21380 }, { "epoch": 4.937673130193906, "grad_norm": 3.021940231323242, "learning_rate": 1.2465373961218838e-07, "loss": 0.1125, "step": 21390 }, { "epoch": 4.939981532779317, "grad_norm": 5.304302215576172, "learning_rate": 1.2003693444136658e-07, "loss": 0.1683, "step": 21400 }, { "epoch": 4.942289935364728, "grad_norm": 0.5288513898849487, "learning_rate": 1.154201292705448e-07, "loss": 0.0865, "step": 21410 }, { "epoch": 4.944598337950138, "grad_norm": 4.0574212074279785, "learning_rate": 1.10803324099723e-07, "loss": 0.0446, "step": 21420 }, { "epoch": 4.94690674053555, "grad_norm": 0.8631877303123474, "learning_rate": 1.061865189289012e-07, "loss": 0.0583, "step": 21430 }, { "epoch": 4.94921514312096, "grad_norm": 0.6526022553443909, "learning_rate": 1.0156971375807942e-07, "loss": 0.0826, "step": 21440 }, { "epoch": 4.9515235457063715, "grad_norm": 1.8015707731246948, "learning_rate": 9.695290858725762e-08, "loss": 0.0249, "step": 21450 }, { "epoch": 4.953831948291782, "grad_norm": 2.896164894104004, "learning_rate": 9.233610341643584e-08, "loss": 0.0091, "step": 21460 }, { "epoch": 4.956140350877193, "grad_norm": 0.05567614734172821, "learning_rate": 8.771929824561404e-08, "loss": 0.0227, "step": 21470 }, { "epoch": 4.958448753462604, "grad_norm": 0.47444531321525574, "learning_rate": 8.310249307479226e-08, "loss": 0.167, "step": 21480 }, { "epoch": 4.960757156048015, "grad_norm": 1.8904088735580444, "learning_rate": 7.848568790397046e-08, "loss": 0.0398, "step": 21490 }, { "epoch": 4.963065558633426, "grad_norm": 0.0800403282046318, "learning_rate": 7.386888273314866e-08, "loss": 0.1243, "step": 21500 }, { "epoch": 4.965373961218837, "grad_norm": 14.902820587158203, "learning_rate": 6.925207756232687e-08, "loss": 0.1058, "step": 21510 }, { "epoch": 4.967682363804247, "grad_norm": 9.65396499633789, "learning_rate": 6.463527239150508e-08, "loss": 0.1016, "step": 21520 }, { "epoch": 4.9699907663896585, "grad_norm": 0.09172435849905014, "learning_rate": 6.001846722068329e-08, "loss": 0.0642, "step": 21530 }, { "epoch": 4.972299168975069, "grad_norm": 1.702339768409729, "learning_rate": 5.54016620498615e-08, "loss": 0.1295, "step": 21540 }, { "epoch": 4.9746075715604805, "grad_norm": 0.2060338407754898, "learning_rate": 5.078485687903971e-08, "loss": 0.1959, "step": 21550 }, { "epoch": 4.976915974145891, "grad_norm": 0.27414509654045105, "learning_rate": 4.616805170821792e-08, "loss": 0.0491, "step": 21560 }, { "epoch": 4.979224376731302, "grad_norm": 0.013999445363879204, "learning_rate": 4.155124653739613e-08, "loss": 0.0401, "step": 21570 }, { "epoch": 4.981532779316713, "grad_norm": 0.3103227913379669, "learning_rate": 3.693444136657433e-08, "loss": 0.0881, "step": 21580 }, { "epoch": 4.9838411819021236, "grad_norm": 0.5369472503662109, "learning_rate": 3.231763619575254e-08, "loss": 0.2265, "step": 21590 }, { "epoch": 4.986149584487535, "grad_norm": 0.40786945819854736, "learning_rate": 2.770083102493075e-08, "loss": 0.019, "step": 21600 }, { "epoch": 4.9884579870729455, "grad_norm": 11.795389175415039, "learning_rate": 2.308402585410896e-08, "loss": 0.0944, "step": 21610 }, { "epoch": 4.990766389658356, "grad_norm": 5.759951591491699, "learning_rate": 1.8467220683287165e-08, "loss": 0.0789, "step": 21620 }, { "epoch": 4.9930747922437675, "grad_norm": 0.3841859996318817, "learning_rate": 1.3850415512465375e-08, "loss": 0.1099, "step": 21630 }, { "epoch": 4.995383194829178, "grad_norm": 0.061368267983198166, "learning_rate": 9.233610341643582e-09, "loss": 0.0573, "step": 21640 }, { "epoch": 4.9976915974145895, "grad_norm": 9.162101745605469, "learning_rate": 4.616805170821791e-09, "loss": 0.179, "step": 21650 }, { "epoch": 5.0, "grad_norm": 0.008566767908632755, "learning_rate": 0.0, "loss": 0.0364, "step": 21660 } ], "logging_steps": 10, "max_steps": 21660, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 3, "trial_name": null, "trial_params": null }