{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.625, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.2367042452096939, "learning_rate": 4.5e-07, "loss": 0.0987, "step": 10 }, { "grad_norm": 0.14597001671791077, "learning_rate": 9.5e-07, "loss": 0.0958, "step": 20 }, { "grad_norm": 0.2929089069366455, "learning_rate": 1.45e-06, "loss": 0.0994, "step": 30 }, { "grad_norm": 0.11476625502109528, "learning_rate": 1.95e-06, "loss": 0.0932, "step": 40 }, { "grad_norm": 0.29291093349456787, "learning_rate": 2.4500000000000003e-06, "loss": 0.0962, "step": 50 }, { "grad_norm": 0.1379394233226776, "learning_rate": 2.95e-06, "loss": 0.0946, "step": 60 }, { "grad_norm": 0.10213620215654373, "learning_rate": 3.4500000000000004e-06, "loss": 0.0907, "step": 70 }, { "grad_norm": 0.23314282298088074, "learning_rate": 3.95e-06, "loss": 0.0854, "step": 80 }, { "grad_norm": 0.07606494426727295, "learning_rate": 4.45e-06, "loss": 0.0791, "step": 90 }, { "grad_norm": 0.14178502559661865, "learning_rate": 4.950000000000001e-06, "loss": 0.0774, "step": 100 }, { "grad_norm": 0.10830003023147583, "learning_rate": 5.45e-06, "loss": 0.0762, "step": 110 }, { "grad_norm": 0.14841148257255554, "learning_rate": 5.95e-06, "loss": 0.0727, "step": 120 }, { "grad_norm": 0.13756820559501648, "learning_rate": 6.45e-06, "loss": 0.0745, "step": 130 }, { "grad_norm": 0.07865367829799652, "learning_rate": 6.950000000000001e-06, "loss": 0.0787, "step": 140 }, { "grad_norm": 0.07377362251281738, "learning_rate": 7.45e-06, "loss": 0.0754, "step": 150 }, { "grad_norm": 0.07843948155641556, "learning_rate": 7.95e-06, "loss": 0.0739, "step": 160 }, { "grad_norm": 0.09336873143911362, "learning_rate": 8.45e-06, "loss": 0.0742, "step": 170 }, { "grad_norm": 0.0761382207274437, "learning_rate": 8.95e-06, "loss": 0.0715, "step": 180 }, { "grad_norm": 0.06910745799541473, "learning_rate": 9.450000000000001e-06, "loss": 0.0663, "step": 190 }, { "grad_norm": 0.060286544263362885, "learning_rate": 9.950000000000001e-06, "loss": 0.066, "step": 200 }, { "grad_norm": 0.05418948456645012, "learning_rate": 1.045e-05, "loss": 0.0664, "step": 210 }, { "grad_norm": 0.07112760841846466, "learning_rate": 1.095e-05, "loss": 0.0641, "step": 220 }, { "grad_norm": 0.05406782031059265, "learning_rate": 1.145e-05, "loss": 0.0606, "step": 230 }, { "grad_norm": 0.06886760145425797, "learning_rate": 1.195e-05, "loss": 0.0608, "step": 240 }, { "grad_norm": 0.058567389845848083, "learning_rate": 1.2450000000000001e-05, "loss": 0.0611, "step": 250 }, { "grad_norm": 0.06587164849042892, "learning_rate": 1.2950000000000001e-05, "loss": 0.0606, "step": 260 }, { "grad_norm": 0.07267028093338013, "learning_rate": 1.3450000000000002e-05, "loss": 0.0659, "step": 270 }, { "grad_norm": 0.05726030468940735, "learning_rate": 1.3950000000000002e-05, "loss": 0.0646, "step": 280 }, { "grad_norm": 0.06414954364299774, "learning_rate": 1.4449999999999999e-05, "loss": 0.0663, "step": 290 }, { "grad_norm": 0.05951888859272003, "learning_rate": 1.4950000000000001e-05, "loss": 0.0627, "step": 300 }, { "grad_norm": 0.05771953612565994, "learning_rate": 1.545e-05, "loss": 0.0649, "step": 310 }, { "grad_norm": 0.051575224846601486, "learning_rate": 1.595e-05, "loss": 0.0639, "step": 320 }, { "grad_norm": 0.048283495008945465, "learning_rate": 1.645e-05, "loss": 0.0582, "step": 330 }, { "grad_norm": 0.047337718307971954, "learning_rate": 1.6950000000000002e-05, "loss": 0.0574, "step": 340 }, { "grad_norm": 0.08193206787109375, "learning_rate": 1.745e-05, "loss": 0.0572, "step": 350 }, { "grad_norm": 0.061031121760606766, "learning_rate": 1.795e-05, "loss": 0.0573, "step": 360 }, { "grad_norm": 0.059021446853876114, "learning_rate": 1.845e-05, "loss": 0.0573, "step": 370 }, { "grad_norm": 0.04083810746669769, "learning_rate": 1.895e-05, "loss": 0.0585, "step": 380 }, { "grad_norm": 0.05405252426862717, "learning_rate": 1.9450000000000002e-05, "loss": 0.0568, "step": 390 }, { "grad_norm": 0.052873317152261734, "learning_rate": 1.995e-05, "loss": 0.0544, "step": 400 }, { "grad_norm": 0.05996895208954811, "learning_rate": 2.045e-05, "loss": 0.0531, "step": 410 }, { "grad_norm": 0.044675275683403015, "learning_rate": 2.095e-05, "loss": 0.0537, "step": 420 }, { "grad_norm": 0.073470339179039, "learning_rate": 2.145e-05, "loss": 0.0528, "step": 430 }, { "grad_norm": 0.0463869608938694, "learning_rate": 2.195e-05, "loss": 0.0524, "step": 440 }, { "grad_norm": 0.0638522282242775, "learning_rate": 2.245e-05, "loss": 0.0572, "step": 450 }, { "grad_norm": 0.0689227357506752, "learning_rate": 2.2950000000000002e-05, "loss": 0.061, "step": 460 }, { "grad_norm": 0.07094327360391617, "learning_rate": 2.345e-05, "loss": 0.062, "step": 470 }, { "grad_norm": 0.048337895423173904, "learning_rate": 2.395e-05, "loss": 0.0596, "step": 480 }, { "grad_norm": 0.04085832089185715, "learning_rate": 2.445e-05, "loss": 0.0639, "step": 490 }, { "grad_norm": 0.05852324143052101, "learning_rate": 2.495e-05, "loss": 0.061, "step": 500 }, { "grad_norm": 0.05310596898198128, "learning_rate": 2.5450000000000002e-05, "loss": 0.0591, "step": 510 }, { "grad_norm": 0.0649959146976471, "learning_rate": 2.595e-05, "loss": 0.059, "step": 520 }, { "grad_norm": 0.06832320988178253, "learning_rate": 2.6450000000000003e-05, "loss": 0.0587, "step": 530 }, { "grad_norm": 0.062165405601263046, "learning_rate": 2.6950000000000005e-05, "loss": 0.0601, "step": 540 }, { "grad_norm": 0.06534390896558762, "learning_rate": 2.7450000000000003e-05, "loss": 0.0586, "step": 550 }, { "grad_norm": 0.06276828050613403, "learning_rate": 2.7950000000000005e-05, "loss": 0.057, "step": 560 }, { "grad_norm": 0.046174000948667526, "learning_rate": 2.845e-05, "loss": 0.0573, "step": 570 }, { "grad_norm": 0.0824345052242279, "learning_rate": 2.895e-05, "loss": 0.0553, "step": 580 }, { "grad_norm": 0.06849933415651321, "learning_rate": 2.945e-05, "loss": 0.0577, "step": 590 }, { "grad_norm": 0.07585761696100235, "learning_rate": 2.995e-05, "loss": 0.0538, "step": 600 }, { "grad_norm": 0.051683682948350906, "learning_rate": 3.045e-05, "loss": 0.0563, "step": 610 }, { "grad_norm": 0.053454261273145676, "learning_rate": 3.095e-05, "loss": 0.0561, "step": 620 }, { "grad_norm": 0.07770929485559464, "learning_rate": 3.145e-05, "loss": 0.0524, "step": 630 }, { "grad_norm": 0.07696694880723953, "learning_rate": 3.1950000000000004e-05, "loss": 0.0554, "step": 640 }, { "grad_norm": 0.07260067015886307, "learning_rate": 3.245e-05, "loss": 0.0549, "step": 650 }, { "grad_norm": 0.04973814636468887, "learning_rate": 3.295e-05, "loss": 0.0529, "step": 660 }, { "grad_norm": 0.049953848123550415, "learning_rate": 3.345000000000001e-05, "loss": 0.0559, "step": 670 }, { "grad_norm": 0.04789105802774429, "learning_rate": 3.3950000000000005e-05, "loss": 0.0543, "step": 680 }, { "grad_norm": 0.06461141258478165, "learning_rate": 3.445e-05, "loss": 0.0537, "step": 690 }, { "grad_norm": 0.07403690367937088, "learning_rate": 3.495e-05, "loss": 0.0523, "step": 700 }, { "grad_norm": 0.05410892143845558, "learning_rate": 3.545e-05, "loss": 0.0536, "step": 710 }, { "grad_norm": 0.061430372297763824, "learning_rate": 3.595e-05, "loss": 0.0503, "step": 720 }, { "grad_norm": 0.059851206839084625, "learning_rate": 3.645e-05, "loss": 0.0503, "step": 730 }, { "grad_norm": 0.07542720437049866, "learning_rate": 3.6950000000000004e-05, "loss": 0.0531, "step": 740 }, { "grad_norm": 0.0805731937289238, "learning_rate": 3.745e-05, "loss": 0.0496, "step": 750 }, { "grad_norm": 0.056793082505464554, "learning_rate": 3.795e-05, "loss": 0.0515, "step": 760 }, { "grad_norm": 0.05777548626065254, "learning_rate": 3.845e-05, "loss": 0.0533, "step": 770 }, { "grad_norm": 0.04915107414126396, "learning_rate": 3.8950000000000005e-05, "loss": 0.0566, "step": 780 }, { "grad_norm": 0.05944707244634628, "learning_rate": 3.9450000000000003e-05, "loss": 0.0534, "step": 790 }, { "grad_norm": 0.060426317155361176, "learning_rate": 3.995e-05, "loss": 0.0552, "step": 800 }, { "grad_norm": 0.04674452915787697, "learning_rate": 4.045000000000001e-05, "loss": 0.0526, "step": 810 }, { "grad_norm": 0.049038417637348175, "learning_rate": 4.095e-05, "loss": 0.0533, "step": 820 }, { "grad_norm": 0.08136085420846939, "learning_rate": 4.145e-05, "loss": 0.0538, "step": 830 }, { "grad_norm": 0.1080460399389267, "learning_rate": 4.195e-05, "loss": 0.0544, "step": 840 }, { "grad_norm": 0.045583777129650116, "learning_rate": 4.245e-05, "loss": 0.0557, "step": 850 }, { "grad_norm": 0.050556447356939316, "learning_rate": 4.295e-05, "loss": 0.0518, "step": 860 }, { "grad_norm": 0.0708293467760086, "learning_rate": 4.345e-05, "loss": 0.053, "step": 870 }, { "grad_norm": 0.0604742132127285, "learning_rate": 4.3950000000000004e-05, "loss": 0.0523, "step": 880 }, { "grad_norm": 0.04935368895530701, "learning_rate": 4.445e-05, "loss": 0.053, "step": 890 }, { "grad_norm": 0.06852924078702927, "learning_rate": 4.495e-05, "loss": 0.0532, "step": 900 }, { "grad_norm": 0.07324350625276566, "learning_rate": 4.545000000000001e-05, "loss": 0.0523, "step": 910 }, { "grad_norm": 0.06833583861589432, "learning_rate": 4.5950000000000006e-05, "loss": 0.0532, "step": 920 }, { "grad_norm": 0.056056004017591476, "learning_rate": 4.6450000000000004e-05, "loss": 0.0536, "step": 930 }, { "grad_norm": 0.048354100435972214, "learning_rate": 4.695e-05, "loss": 0.0517, "step": 940 }, { "grad_norm": 0.05747619643807411, "learning_rate": 4.745e-05, "loss": 0.0528, "step": 950 }, { "grad_norm": 0.05889948084950447, "learning_rate": 4.795e-05, "loss": 0.0517, "step": 960 }, { "grad_norm": 0.051637571305036545, "learning_rate": 4.845e-05, "loss": 0.0535, "step": 970 }, { "grad_norm": 0.053299497812986374, "learning_rate": 4.8950000000000004e-05, "loss": 0.0554, "step": 980 }, { "grad_norm": 0.055623993277549744, "learning_rate": 4.945e-05, "loss": 0.0534, "step": 990 }, { "grad_norm": 0.05885191261768341, "learning_rate": 4.995e-05, "loss": 0.0563, "step": 1000 }, { "grad_norm": 0.05181998014450073, "learning_rate": 5.045e-05, "loss": 0.0534, "step": 1010 }, { "grad_norm": 0.044812433421611786, "learning_rate": 5.095e-05, "loss": 0.0537, "step": 1020 }, { "grad_norm": 0.07625119388103485, "learning_rate": 5.145e-05, "loss": 0.0572, "step": 1030 }, { "grad_norm": 0.11796906590461731, "learning_rate": 5.1949999999999996e-05, "loss": 0.0553, "step": 1040 }, { "grad_norm": 0.0881151407957077, "learning_rate": 5.245e-05, "loss": 0.0571, "step": 1050 }, { "grad_norm": 0.056303564459085464, "learning_rate": 5.295e-05, "loss": 0.0527, "step": 1060 }, { "grad_norm": 0.06179461628198624, "learning_rate": 5.345e-05, "loss": 0.0559, "step": 1070 }, { "grad_norm": 0.05705244839191437, "learning_rate": 5.3950000000000004e-05, "loss": 0.0531, "step": 1080 }, { "grad_norm": 0.06554149091243744, "learning_rate": 5.445e-05, "loss": 0.0506, "step": 1090 }, { "grad_norm": 0.0712968185544014, "learning_rate": 5.495e-05, "loss": 0.0552, "step": 1100 }, { "grad_norm": 0.08786563575267792, "learning_rate": 5.545e-05, "loss": 0.0517, "step": 1110 }, { "grad_norm": 0.06611761450767517, "learning_rate": 5.5950000000000005e-05, "loss": 0.0512, "step": 1120 }, { "grad_norm": 0.0559191033244133, "learning_rate": 5.645e-05, "loss": 0.0535, "step": 1130 }, { "grad_norm": 0.06905399262905121, "learning_rate": 5.695e-05, "loss": 0.0503, "step": 1140 }, { "grad_norm": 0.04899454116821289, "learning_rate": 5.745e-05, "loss": 0.0508, "step": 1150 }, { "grad_norm": 0.057010218501091, "learning_rate": 5.7950000000000006e-05, "loss": 0.0485, "step": 1160 }, { "grad_norm": 0.06790298223495483, "learning_rate": 5.8450000000000005e-05, "loss": 0.0469, "step": 1170 }, { "grad_norm": 0.06535037606954575, "learning_rate": 5.895e-05, "loss": 0.0467, "step": 1180 }, { "grad_norm": 0.07846511155366898, "learning_rate": 5.945000000000001e-05, "loss": 0.0455, "step": 1190 }, { "grad_norm": 0.04488803818821907, "learning_rate": 5.995000000000001e-05, "loss": 0.048, "step": 1200 }, { "grad_norm": 0.08144808560609818, "learning_rate": 6.0450000000000006e-05, "loss": 0.0471, "step": 1210 }, { "grad_norm": 0.04968644678592682, "learning_rate": 6.0950000000000004e-05, "loss": 0.0477, "step": 1220 }, { "grad_norm": 0.03707012161612511, "learning_rate": 6.145e-05, "loss": 0.0476, "step": 1230 }, { "grad_norm": 0.04717239364981651, "learning_rate": 6.195e-05, "loss": 0.0485, "step": 1240 }, { "grad_norm": 0.06056271865963936, "learning_rate": 6.245000000000001e-05, "loss": 0.0525, "step": 1250 }, { "grad_norm": 0.08031583577394485, "learning_rate": 6.295e-05, "loss": 0.049, "step": 1260 }, { "grad_norm": 0.09000271558761597, "learning_rate": 6.345e-05, "loss": 0.0464, "step": 1270 }, { "grad_norm": 0.06793270260095596, "learning_rate": 6.395e-05, "loss": 0.0513, "step": 1280 }, { "grad_norm": 0.07769488543272018, "learning_rate": 6.445e-05, "loss": 0.0449, "step": 1290 }, { "grad_norm": 0.04967029392719269, "learning_rate": 6.494999999999999e-05, "loss": 0.0482, "step": 1300 }, { "grad_norm": 0.052150458097457886, "learning_rate": 6.545e-05, "loss": 0.0479, "step": 1310 }, { "grad_norm": 0.04675132781267166, "learning_rate": 6.595e-05, "loss": 0.0486, "step": 1320 }, { "grad_norm": 0.05263039842247963, "learning_rate": 6.645e-05, "loss": 0.0488, "step": 1330 }, { "grad_norm": 0.07592596858739853, "learning_rate": 6.695e-05, "loss": 0.0478, "step": 1340 }, { "grad_norm": 0.04908451810479164, "learning_rate": 6.745e-05, "loss": 0.0494, "step": 1350 }, { "grad_norm": 0.07130961865186691, "learning_rate": 6.795e-05, "loss": 0.05, "step": 1360 }, { "grad_norm": 0.05626438930630684, "learning_rate": 6.845e-05, "loss": 0.0483, "step": 1370 }, { "grad_norm": 0.0774865597486496, "learning_rate": 6.895000000000001e-05, "loss": 0.0473, "step": 1380 }, { "grad_norm": 0.06105751171708107, "learning_rate": 6.945000000000001e-05, "loss": 0.0479, "step": 1390 }, { "grad_norm": 0.06412041187286377, "learning_rate": 6.995e-05, "loss": 0.0474, "step": 1400 }, { "grad_norm": 0.05008367449045181, "learning_rate": 7.045e-05, "loss": 0.0554, "step": 1410 }, { "grad_norm": 0.07180901616811752, "learning_rate": 7.095e-05, "loss": 0.0523, "step": 1420 }, { "grad_norm": 0.06622883677482605, "learning_rate": 7.145e-05, "loss": 0.0576, "step": 1430 }, { "grad_norm": 0.06586829572916031, "learning_rate": 7.195e-05, "loss": 0.0527, "step": 1440 }, { "grad_norm": 0.06603917479515076, "learning_rate": 7.245000000000001e-05, "loss": 0.0537, "step": 1450 }, { "grad_norm": 0.0552862174808979, "learning_rate": 7.295000000000001e-05, "loss": 0.0529, "step": 1460 }, { "grad_norm": 0.08935122191905975, "learning_rate": 7.345000000000001e-05, "loss": 0.0517, "step": 1470 }, { "grad_norm": 0.04799119383096695, "learning_rate": 7.395000000000001e-05, "loss": 0.05, "step": 1480 }, { "grad_norm": 0.07158424705266953, "learning_rate": 7.445000000000001e-05, "loss": 0.0487, "step": 1490 }, { "grad_norm": 0.059339944273233414, "learning_rate": 7.495e-05, "loss": 0.0472, "step": 1500 }, { "grad_norm": 0.05760268494486809, "learning_rate": 7.545e-05, "loss": 0.0448, "step": 1510 }, { "grad_norm": 0.060016337782144547, "learning_rate": 7.595e-05, "loss": 0.048, "step": 1520 }, { "grad_norm": 0.08550744503736496, "learning_rate": 7.645e-05, "loss": 0.0465, "step": 1530 }, { "grad_norm": 0.07722946256399155, "learning_rate": 7.695e-05, "loss": 0.0466, "step": 1540 }, { "grad_norm": 0.04797113314270973, "learning_rate": 7.745e-05, "loss": 0.0473, "step": 1550 }, { "grad_norm": 0.052482958883047104, "learning_rate": 7.795e-05, "loss": 0.0466, "step": 1560 }, { "grad_norm": 0.0645216852426529, "learning_rate": 7.845e-05, "loss": 0.0472, "step": 1570 }, { "grad_norm": 0.05997135862708092, "learning_rate": 7.895000000000001e-05, "loss": 0.0472, "step": 1580 }, { "grad_norm": 0.10403952747583389, "learning_rate": 7.945e-05, "loss": 0.0468, "step": 1590 }, { "grad_norm": 0.0654444545507431, "learning_rate": 7.995e-05, "loss": 0.0474, "step": 1600 }, { "grad_norm": 0.05725341662764549, "learning_rate": 8.045e-05, "loss": 0.0464, "step": 1610 }, { "grad_norm": 0.10814964771270752, "learning_rate": 8.095e-05, "loss": 0.0441, "step": 1620 }, { "grad_norm": 0.04941585287451744, "learning_rate": 8.145e-05, "loss": 0.0444, "step": 1630 }, { "grad_norm": 0.08012948930263519, "learning_rate": 8.195e-05, "loss": 0.044, "step": 1640 }, { "grad_norm": 0.06319572031497955, "learning_rate": 8.245e-05, "loss": 0.0467, "step": 1650 }, { "grad_norm": 0.04746998846530914, "learning_rate": 8.295000000000001e-05, "loss": 0.0437, "step": 1660 }, { "grad_norm": 0.06441157311201096, "learning_rate": 8.345000000000001e-05, "loss": 0.0479, "step": 1670 }, { "grad_norm": 0.0719904750585556, "learning_rate": 8.395000000000001e-05, "loss": 0.0487, "step": 1680 }, { "grad_norm": 0.05785597488284111, "learning_rate": 8.445e-05, "loss": 0.0482, "step": 1690 }, { "grad_norm": 0.07752306759357452, "learning_rate": 8.495e-05, "loss": 0.0479, "step": 1700 }, { "grad_norm": 0.05668678879737854, "learning_rate": 8.545e-05, "loss": 0.0454, "step": 1710 }, { "grad_norm": 0.05378156155347824, "learning_rate": 8.595e-05, "loss": 0.0465, "step": 1720 }, { "grad_norm": 0.07258997112512589, "learning_rate": 8.645000000000001e-05, "loss": 0.0492, "step": 1730 }, { "grad_norm": 0.0544121190905571, "learning_rate": 8.695000000000001e-05, "loss": 0.0455, "step": 1740 }, { "grad_norm": 0.06962165236473083, "learning_rate": 8.745000000000001e-05, "loss": 0.0453, "step": 1750 }, { "grad_norm": 0.05118221789598465, "learning_rate": 8.795e-05, "loss": 0.0461, "step": 1760 }, { "grad_norm": 0.0571901760995388, "learning_rate": 8.845e-05, "loss": 0.0464, "step": 1770 }, { "grad_norm": 0.050800010561943054, "learning_rate": 8.895e-05, "loss": 0.0448, "step": 1780 }, { "grad_norm": 0.04537017270922661, "learning_rate": 8.945e-05, "loss": 0.0457, "step": 1790 }, { "grad_norm": 0.05725893750786781, "learning_rate": 8.995e-05, "loss": 0.0479, "step": 1800 }, { "grad_norm": 0.05547211691737175, "learning_rate": 9.045e-05, "loss": 0.0433, "step": 1810 }, { "grad_norm": 0.04526804760098457, "learning_rate": 9.095e-05, "loss": 0.045, "step": 1820 }, { "grad_norm": 0.05306830629706383, "learning_rate": 9.145e-05, "loss": 0.0442, "step": 1830 }, { "grad_norm": 0.05669446289539337, "learning_rate": 9.195e-05, "loss": 0.0438, "step": 1840 }, { "grad_norm": 0.05584435164928436, "learning_rate": 9.245e-05, "loss": 0.0475, "step": 1850 }, { "grad_norm": 0.04942282661795616, "learning_rate": 9.295000000000001e-05, "loss": 0.0498, "step": 1860 }, { "grad_norm": 0.052402328699827194, "learning_rate": 9.345000000000001e-05, "loss": 0.052, "step": 1870 }, { "grad_norm": 0.0529673732817173, "learning_rate": 9.395000000000001e-05, "loss": 0.051, "step": 1880 }, { "grad_norm": 0.059187062084674835, "learning_rate": 9.445e-05, "loss": 0.0513, "step": 1890 }, { "grad_norm": 0.06676316261291504, "learning_rate": 9.495e-05, "loss": 0.0507, "step": 1900 }, { "grad_norm": 0.05016503110527992, "learning_rate": 9.545e-05, "loss": 0.0503, "step": 1910 }, { "grad_norm": 0.05649873614311218, "learning_rate": 9.595e-05, "loss": 0.0526, "step": 1920 }, { "grad_norm": 0.05375303700566292, "learning_rate": 9.645000000000001e-05, "loss": 0.0528, "step": 1930 }, { "grad_norm": 0.07709939777851105, "learning_rate": 9.695000000000001e-05, "loss": 0.0523, "step": 1940 }, { "grad_norm": 0.04950285702943802, "learning_rate": 9.745000000000001e-05, "loss": 0.0517, "step": 1950 }, { "grad_norm": 0.06602159142494202, "learning_rate": 9.795000000000001e-05, "loss": 0.0539, "step": 1960 }, { "grad_norm": 0.04525547847151756, "learning_rate": 9.845000000000001e-05, "loss": 0.0511, "step": 1970 }, { "grad_norm": 0.05090705305337906, "learning_rate": 9.895e-05, "loss": 0.0493, "step": 1980 }, { "grad_norm": 0.06536839157342911, "learning_rate": 9.945e-05, "loss": 0.0518, "step": 1990 }, { "grad_norm": 0.045681942254304886, "learning_rate": 9.995e-05, "loss": 0.0506, "step": 2000 }, { "grad_norm": 0.06368067860603333, "learning_rate": 9.99999861593158e-05, "loss": 0.0477, "step": 2010 }, { "grad_norm": 0.061842430382966995, "learning_rate": 9.999993831498517e-05, "loss": 0.0471, "step": 2020 }, { "grad_norm": 0.04350709170103073, "learning_rate": 9.99998562961682e-05, "loss": 0.0467, "step": 2030 }, { "grad_norm": 0.07890106737613678, "learning_rate": 9.99997401029209e-05, "loss": 0.046, "step": 2040 }, { "grad_norm": 0.06441490352153778, "learning_rate": 9.999958973532271e-05, "loss": 0.0497, "step": 2050 }, { "grad_norm": 0.04839397221803665, "learning_rate": 9.999940519347642e-05, "loss": 0.0465, "step": 2060 }, { "grad_norm": 0.13024568557739258, "learning_rate": 9.999918647750812e-05, "loss": 0.0479, "step": 2070 }, { "grad_norm": 0.04704037308692932, "learning_rate": 9.999893358756736e-05, "loss": 0.047, "step": 2080 }, { "grad_norm": 0.06193709373474121, "learning_rate": 9.999864652382692e-05, "loss": 0.0464, "step": 2090 }, { "grad_norm": 0.055663783103227615, "learning_rate": 9.999832528648307e-05, "loss": 0.0447, "step": 2100 }, { "grad_norm": 0.13960303366184235, "learning_rate": 9.999796987575532e-05, "loss": 0.0467, "step": 2110 }, { "grad_norm": 0.056903764605522156, "learning_rate": 9.999758029188662e-05, "loss": 0.0448, "step": 2120 }, { "grad_norm": 0.07287687808275223, "learning_rate": 9.999715653514324e-05, "loss": 0.0459, "step": 2130 }, { "grad_norm": 0.04698964208364487, "learning_rate": 9.999669860581482e-05, "loss": 0.0437, "step": 2140 }, { "grad_norm": 0.0628240704536438, "learning_rate": 9.999620650421433e-05, "loss": 0.0466, "step": 2150 }, { "grad_norm": 0.06613671034574509, "learning_rate": 9.999568023067813e-05, "loss": 0.0444, "step": 2160 }, { "grad_norm": 0.08464129269123077, "learning_rate": 9.999511978556595e-05, "loss": 0.0503, "step": 2170 }, { "grad_norm": 0.08315014094114304, "learning_rate": 9.99945251692608e-05, "loss": 0.0513, "step": 2180 }, { "grad_norm": 0.07440516352653503, "learning_rate": 9.999389638216912e-05, "loss": 0.0496, "step": 2190 }, { "grad_norm": 0.06842469424009323, "learning_rate": 9.999323342472068e-05, "loss": 0.0493, "step": 2200 }, { "grad_norm": 0.07809775322675705, "learning_rate": 9.99925362973686e-05, "loss": 0.0508, "step": 2210 }, { "grad_norm": 0.06675782799720764, "learning_rate": 9.999180500058935e-05, "loss": 0.0503, "step": 2220 }, { "grad_norm": 0.06185253709554672, "learning_rate": 9.99910395348828e-05, "loss": 0.047, "step": 2230 }, { "grad_norm": 0.09716532379388809, "learning_rate": 9.99902399007721e-05, "loss": 0.0478, "step": 2240 }, { "grad_norm": 0.051124051213264465, "learning_rate": 9.99894060988038e-05, "loss": 0.0494, "step": 2250 }, { "grad_norm": 0.047970253974199295, "learning_rate": 9.998853812954783e-05, "loss": 0.0448, "step": 2260 }, { "grad_norm": 0.05197246000170708, "learning_rate": 9.998763599359739e-05, "loss": 0.0456, "step": 2270 }, { "grad_norm": 0.05724599212408066, "learning_rate": 9.998669969156912e-05, "loss": 0.0456, "step": 2280 }, { "grad_norm": 0.05664108693599701, "learning_rate": 9.998572922410294e-05, "loss": 0.0436, "step": 2290 }, { "grad_norm": 0.04838567599654198, "learning_rate": 9.998472459186219e-05, "loss": 0.0494, "step": 2300 }, { "grad_norm": 0.08672767877578735, "learning_rate": 9.99836857955335e-05, "loss": 0.0484, "step": 2310 }, { "grad_norm": 0.2314692884683609, "learning_rate": 9.998261283582688e-05, "loss": 0.0478, "step": 2320 }, { "grad_norm": 0.08851045370101929, "learning_rate": 9.998150571347571e-05, "loss": 0.0488, "step": 2330 }, { "grad_norm": 0.064845971763134, "learning_rate": 9.998036442923667e-05, "loss": 0.0487, "step": 2340 }, { "grad_norm": 0.04243288189172745, "learning_rate": 9.997918898388983e-05, "loss": 0.0497, "step": 2350 }, { "grad_norm": 0.047868479043245316, "learning_rate": 9.997797937823859e-05, "loss": 0.0467, "step": 2360 }, { "grad_norm": 0.07111752778291702, "learning_rate": 9.997673561310971e-05, "loss": 0.0481, "step": 2370 }, { "grad_norm": 0.07211709022521973, "learning_rate": 9.997545768935331e-05, "loss": 0.0481, "step": 2380 }, { "grad_norm": 0.07223348319530487, "learning_rate": 9.997414560784282e-05, "loss": 0.0488, "step": 2390 }, { "grad_norm": 0.08666320145130157, "learning_rate": 9.997279936947502e-05, "loss": 0.0437, "step": 2400 }, { "grad_norm": 0.07779088616371155, "learning_rate": 9.997141897517006e-05, "loss": 0.0479, "step": 2410 }, { "grad_norm": 0.055702921003103256, "learning_rate": 9.997000442587145e-05, "loss": 0.0443, "step": 2420 }, { "grad_norm": 0.05179236829280853, "learning_rate": 9.9968555722546e-05, "loss": 0.0468, "step": 2430 }, { "grad_norm": 0.05055181309580803, "learning_rate": 9.996707286618387e-05, "loss": 0.0429, "step": 2440 }, { "grad_norm": 0.04722628742456436, "learning_rate": 9.996555585779861e-05, "loss": 0.0425, "step": 2450 }, { "grad_norm": 0.06612866371870041, "learning_rate": 9.996400469842708e-05, "loss": 0.0446, "step": 2460 }, { "grad_norm": 0.05022690072655678, "learning_rate": 9.996241938912945e-05, "loss": 0.0443, "step": 2470 }, { "grad_norm": 0.045986074954271317, "learning_rate": 9.99607999309893e-05, "loss": 0.0414, "step": 2480 }, { "grad_norm": 0.10119111090898514, "learning_rate": 9.99591463251135e-05, "loss": 0.0452, "step": 2490 }, { "grad_norm": 0.05102114751935005, "learning_rate": 9.995745857263226e-05, "loss": 0.0488, "step": 2500 }, { "grad_norm": 0.056538745760917664, "learning_rate": 9.995573667469918e-05, "loss": 0.0472, "step": 2510 }, { "grad_norm": 0.0566011518239975, "learning_rate": 9.995398063249111e-05, "loss": 0.0481, "step": 2520 }, { "grad_norm": 0.06752344220876694, "learning_rate": 9.995219044720833e-05, "loss": 0.0469, "step": 2530 }, { "grad_norm": 0.07237391173839569, "learning_rate": 9.99503661200744e-05, "loss": 0.046, "step": 2540 }, { "grad_norm": 0.06586555391550064, "learning_rate": 9.994850765233624e-05, "loss": 0.0502, "step": 2550 }, { "grad_norm": 0.04773681238293648, "learning_rate": 9.994661504526406e-05, "loss": 0.0483, "step": 2560 }, { "grad_norm": 0.0582672543823719, "learning_rate": 9.994468830015148e-05, "loss": 0.0481, "step": 2570 }, { "grad_norm": 0.06281842291355133, "learning_rate": 9.99427274183154e-05, "loss": 0.0465, "step": 2580 }, { "grad_norm": 0.059918712824583054, "learning_rate": 9.994073240109606e-05, "loss": 0.044, "step": 2590 }, { "grad_norm": 0.06341630965471268, "learning_rate": 9.993870324985703e-05, "loss": 0.046, "step": 2600 }, { "grad_norm": 0.09598434716463089, "learning_rate": 9.993663996598523e-05, "loss": 0.0471, "step": 2610 }, { "grad_norm": 0.0677446573972702, "learning_rate": 9.993454255089089e-05, "loss": 0.0504, "step": 2620 }, { "grad_norm": 0.06634124368429184, "learning_rate": 9.993241100600757e-05, "loss": 0.0513, "step": 2630 }, { "grad_norm": 0.06693260371685028, "learning_rate": 9.993024533279215e-05, "loss": 0.0511, "step": 2640 }, { "grad_norm": 0.0551031120121479, "learning_rate": 9.992804553272487e-05, "loss": 0.0492, "step": 2650 }, { "grad_norm": 0.05580879747867584, "learning_rate": 9.992581160730926e-05, "loss": 0.049, "step": 2660 }, { "grad_norm": 0.057869717478752136, "learning_rate": 9.992354355807218e-05, "loss": 0.0484, "step": 2670 }, { "grad_norm": 0.04472215101122856, "learning_rate": 9.992124138656385e-05, "loss": 0.0454, "step": 2680 }, { "grad_norm": 0.06821594387292862, "learning_rate": 9.991890509435774e-05, "loss": 0.0477, "step": 2690 }, { "grad_norm": 0.059958696365356445, "learning_rate": 9.991653468305071e-05, "loss": 0.0439, "step": 2700 }, { "grad_norm": 0.055043574422597885, "learning_rate": 9.991413015426292e-05, "loss": 0.0456, "step": 2710 }, { "grad_norm": 0.07571437209844589, "learning_rate": 9.991169150963781e-05, "loss": 0.0442, "step": 2720 }, { "grad_norm": 0.052696164697408676, "learning_rate": 9.990921875084221e-05, "loss": 0.0423, "step": 2730 }, { "grad_norm": 0.049476440995931625, "learning_rate": 9.990671187956622e-05, "loss": 0.0457, "step": 2740 }, { "grad_norm": 0.06535124778747559, "learning_rate": 9.990417089752324e-05, "loss": 0.0527, "step": 2750 }, { "grad_norm": 0.059893347322940826, "learning_rate": 9.990159580645001e-05, "loss": 0.052, "step": 2760 }, { "grad_norm": 0.062266603112220764, "learning_rate": 9.98989866081066e-05, "loss": 0.0497, "step": 2770 }, { "grad_norm": 0.06664110720157623, "learning_rate": 9.989634330427636e-05, "loss": 0.0501, "step": 2780 }, { "grad_norm": 0.05679744854569435, "learning_rate": 9.989366589676596e-05, "loss": 0.0507, "step": 2790 }, { "grad_norm": 0.061713334172964096, "learning_rate": 9.98909543874054e-05, "loss": 0.052, "step": 2800 }, { "grad_norm": 0.04963374137878418, "learning_rate": 9.988820877804795e-05, "loss": 0.0517, "step": 2810 }, { "grad_norm": 0.0419585183262825, "learning_rate": 9.988542907057022e-05, "loss": 0.0469, "step": 2820 }, { "grad_norm": 0.05081905797123909, "learning_rate": 9.988261526687212e-05, "loss": 0.0462, "step": 2830 }, { "grad_norm": 0.0600738488137722, "learning_rate": 9.987976736887685e-05, "loss": 0.0452, "step": 2840 }, { "grad_norm": 0.11048606038093567, "learning_rate": 9.987688537853091e-05, "loss": 0.0451, "step": 2850 }, { "grad_norm": 0.03942258656024933, "learning_rate": 9.987396929780413e-05, "loss": 0.0448, "step": 2860 }, { "grad_norm": 0.04412582889199257, "learning_rate": 9.987101912868962e-05, "loss": 0.0482, "step": 2870 }, { "grad_norm": 0.04830498993396759, "learning_rate": 9.98680348732038e-05, "loss": 0.0477, "step": 2880 }, { "grad_norm": 0.07393594086170197, "learning_rate": 9.986501653338636e-05, "loss": 0.0476, "step": 2890 }, { "grad_norm": 0.04549046978354454, "learning_rate": 9.986196411130031e-05, "loss": 0.048, "step": 2900 }, { "grad_norm": 0.07075855880975723, "learning_rate": 9.985887760903197e-05, "loss": 0.0481, "step": 2910 }, { "grad_norm": 0.09219884127378464, "learning_rate": 9.985575702869093e-05, "loss": 0.0492, "step": 2920 }, { "grad_norm": 0.05372711643576622, "learning_rate": 9.985260237241008e-05, "loss": 0.0455, "step": 2930 }, { "grad_norm": 0.07848327606916428, "learning_rate": 9.984941364234557e-05, "loss": 0.0473, "step": 2940 }, { "grad_norm": 0.040722522884607315, "learning_rate": 9.98461908406769e-05, "loss": 0.0467, "step": 2950 }, { "grad_norm": 0.052283063530921936, "learning_rate": 9.98429339696068e-05, "loss": 0.0446, "step": 2960 }, { "grad_norm": 0.06453437358140945, "learning_rate": 9.983964303136133e-05, "loss": 0.0483, "step": 2970 }, { "grad_norm": 0.07537301629781723, "learning_rate": 9.983631802818981e-05, "loss": 0.0467, "step": 2980 }, { "grad_norm": 0.09006133675575256, "learning_rate": 9.983295896236484e-05, "loss": 0.0453, "step": 2990 }, { "grad_norm": 0.07051413506269455, "learning_rate": 9.982956583618232e-05, "loss": 0.0461, "step": 3000 }, { "grad_norm": 0.07437418401241302, "learning_rate": 9.982613865196142e-05, "loss": 0.0493, "step": 3010 }, { "grad_norm": 0.052972082048654556, "learning_rate": 9.982267741204458e-05, "loss": 0.0474, "step": 3020 }, { "grad_norm": 0.06052765995264053, "learning_rate": 9.981918211879753e-05, "loss": 0.049, "step": 3030 }, { "grad_norm": 0.06627226620912552, "learning_rate": 9.981565277460927e-05, "loss": 0.0496, "step": 3040 }, { "grad_norm": 0.08694003522396088, "learning_rate": 9.981208938189206e-05, "loss": 0.0462, "step": 3050 }, { "grad_norm": 0.05517130345106125, "learning_rate": 9.980849194308149e-05, "loss": 0.0498, "step": 3060 }, { "grad_norm": 0.05107469484210014, "learning_rate": 9.98048604606363e-05, "loss": 0.0504, "step": 3070 }, { "grad_norm": 0.05309459567070007, "learning_rate": 9.980119493703864e-05, "loss": 0.0479, "step": 3080 }, { "grad_norm": 0.06065574288368225, "learning_rate": 9.979749537479383e-05, "loss": 0.0492, "step": 3090 }, { "grad_norm": 0.04890000447630882, "learning_rate": 9.979376177643051e-05, "loss": 0.0483, "step": 3100 }, { "grad_norm": 0.06031010299921036, "learning_rate": 9.978999414450052e-05, "loss": 0.0493, "step": 3110 }, { "grad_norm": 0.044559117406606674, "learning_rate": 9.978619248157904e-05, "loss": 0.048, "step": 3120 }, { "grad_norm": 0.0724765956401825, "learning_rate": 9.978235679026445e-05, "loss": 0.0483, "step": 3130 }, { "grad_norm": 0.0582214817404747, "learning_rate": 9.97784870731784e-05, "loss": 0.0467, "step": 3140 }, { "grad_norm": 0.05179806426167488, "learning_rate": 9.977458333296584e-05, "loss": 0.0494, "step": 3150 }, { "grad_norm": 0.07265907526016235, "learning_rate": 9.977064557229492e-05, "loss": 0.0466, "step": 3160 }, { "grad_norm": 0.05376670882105827, "learning_rate": 9.976667379385705e-05, "loss": 0.047, "step": 3170 }, { "grad_norm": 0.04855695366859436, "learning_rate": 9.97626680003669e-05, "loss": 0.047, "step": 3180 }, { "grad_norm": 0.06009653955698013, "learning_rate": 9.975862819456242e-05, "loss": 0.048, "step": 3190 }, { "grad_norm": 0.07684388756752014, "learning_rate": 9.975455437920477e-05, "loss": 0.0457, "step": 3200 }, { "grad_norm": 0.0738026574254036, "learning_rate": 9.975044655707834e-05, "loss": 0.0478, "step": 3210 }, { "grad_norm": 0.07000957429409027, "learning_rate": 9.974630473099082e-05, "loss": 0.0467, "step": 3220 }, { "grad_norm": 0.12336456030607224, "learning_rate": 9.974212890377311e-05, "loss": 0.0462, "step": 3230 }, { "grad_norm": 0.06025565043091774, "learning_rate": 9.973791907827931e-05, "loss": 0.0459, "step": 3240 }, { "grad_norm": 0.09145662188529968, "learning_rate": 9.973367525738683e-05, "loss": 0.0498, "step": 3250 }, { "grad_norm": 0.07158467918634415, "learning_rate": 9.972939744399627e-05, "loss": 0.0491, "step": 3260 }, { "grad_norm": 0.25944384932518005, "learning_rate": 9.97250856410315e-05, "loss": 0.0468, "step": 3270 }, { "grad_norm": 0.08413989841938019, "learning_rate": 9.972073985143955e-05, "loss": 0.0516, "step": 3280 }, { "grad_norm": 0.11324361711740494, "learning_rate": 9.971636007819074e-05, "loss": 0.05, "step": 3290 }, { "grad_norm": 0.08564183861017227, "learning_rate": 9.971194632427863e-05, "loss": 0.0468, "step": 3300 }, { "grad_norm": 0.08695483952760696, "learning_rate": 9.970749859271997e-05, "loss": 0.0496, "step": 3310 }, { "grad_norm": 0.06083517149090767, "learning_rate": 9.970301688655473e-05, "loss": 0.0491, "step": 3320 }, { "grad_norm": 0.051688484847545624, "learning_rate": 9.969850120884612e-05, "loss": 0.0482, "step": 3330 }, { "grad_norm": 0.039818331599235535, "learning_rate": 9.969395156268055e-05, "loss": 0.0466, "step": 3340 }, { "grad_norm": 0.09253015369176865, "learning_rate": 9.968936795116768e-05, "loss": 0.0463, "step": 3350 }, { "grad_norm": 0.05920790880918503, "learning_rate": 9.968475037744036e-05, "loss": 0.0484, "step": 3360 }, { "grad_norm": 0.04932876303792, "learning_rate": 9.968009884465465e-05, "loss": 0.0464, "step": 3370 }, { "grad_norm": 0.05489765852689743, "learning_rate": 9.967541335598984e-05, "loss": 0.0492, "step": 3380 }, { "grad_norm": 0.049517922103405, "learning_rate": 9.967069391464841e-05, "loss": 0.0444, "step": 3390 }, { "grad_norm": 0.06632087379693985, "learning_rate": 9.966594052385608e-05, "loss": 0.0444, "step": 3400 }, { "grad_norm": 0.0496390238404274, "learning_rate": 9.96611531868617e-05, "loss": 0.0451, "step": 3410 }, { "grad_norm": 0.08213259279727936, "learning_rate": 9.96563319069374e-05, "loss": 0.0446, "step": 3420 }, { "grad_norm": 0.04313664510846138, "learning_rate": 9.965147668737847e-05, "loss": 0.0435, "step": 3430 }, { "grad_norm": 0.052240774035453796, "learning_rate": 9.96465875315034e-05, "loss": 0.0457, "step": 3440 }, { "grad_norm": 0.04486159235239029, "learning_rate": 9.96416644426539e-05, "loss": 0.0479, "step": 3450 }, { "grad_norm": 0.05756843090057373, "learning_rate": 9.963670742419485e-05, "loss": 0.051, "step": 3460 }, { "grad_norm": 0.07407929748296738, "learning_rate": 9.96317164795143e-05, "loss": 0.0452, "step": 3470 }, { "grad_norm": 0.051698412746191025, "learning_rate": 9.962669161202356e-05, "loss": 0.0481, "step": 3480 }, { "grad_norm": 0.07663122564554214, "learning_rate": 9.962163282515705e-05, "loss": 0.0445, "step": 3490 }, { "grad_norm": 0.0690796822309494, "learning_rate": 9.961654012237241e-05, "loss": 0.0458, "step": 3500 }, { "grad_norm": 0.04611192271113396, "learning_rate": 9.961141350715044e-05, "loss": 0.0487, "step": 3510 }, { "grad_norm": 0.0689488872885704, "learning_rate": 9.960625298299516e-05, "loss": 0.046, "step": 3520 }, { "grad_norm": 0.05534031242132187, "learning_rate": 9.960105855343372e-05, "loss": 0.0459, "step": 3530 }, { "grad_norm": 0.10019628703594208, "learning_rate": 9.959583022201647e-05, "loss": 0.0451, "step": 3540 }, { "grad_norm": 0.11438955366611481, "learning_rate": 9.959056799231692e-05, "loss": 0.0466, "step": 3550 }, { "grad_norm": 0.05055440217256546, "learning_rate": 9.958527186793176e-05, "loss": 0.0473, "step": 3560 }, { "grad_norm": 0.04784643277525902, "learning_rate": 9.957994185248086e-05, "loss": 0.0481, "step": 3570 }, { "grad_norm": 0.043144673109054565, "learning_rate": 9.957457794960718e-05, "loss": 0.0431, "step": 3580 }, { "grad_norm": 0.06457866728305817, "learning_rate": 9.956918016297694e-05, "loss": 0.0463, "step": 3590 }, { "grad_norm": 0.04955054819583893, "learning_rate": 9.956374849627948e-05, "loss": 0.0426, "step": 3600 }, { "grad_norm": 0.046829331666231155, "learning_rate": 9.955828295322728e-05, "loss": 0.0414, "step": 3610 }, { "grad_norm": 0.04118868708610535, "learning_rate": 9.955278353755598e-05, "loss": 0.0469, "step": 3620 }, { "grad_norm": 0.05583234876394272, "learning_rate": 9.954725025302439e-05, "loss": 0.0394, "step": 3630 }, { "grad_norm": 0.04523902386426926, "learning_rate": 9.954168310341445e-05, "loss": 0.0439, "step": 3640 }, { "grad_norm": 0.03969503939151764, "learning_rate": 9.953608209253126e-05, "loss": 0.0404, "step": 3650 }, { "grad_norm": 0.04516305401921272, "learning_rate": 9.953044722420307e-05, "loss": 0.0441, "step": 3660 }, { "grad_norm": 0.049359146505594254, "learning_rate": 9.952477850228124e-05, "loss": 0.0396, "step": 3670 }, { "grad_norm": 0.054378692060709, "learning_rate": 9.95190759306403e-05, "loss": 0.045, "step": 3680 }, { "grad_norm": 0.048493556678295135, "learning_rate": 9.951333951317789e-05, "loss": 0.0415, "step": 3690 }, { "grad_norm": 0.05593075975775719, "learning_rate": 9.950756925381479e-05, "loss": 0.0468, "step": 3700 }, { "grad_norm": 0.04714721068739891, "learning_rate": 9.950176515649496e-05, "loss": 0.0465, "step": 3710 }, { "grad_norm": 0.044041361659765244, "learning_rate": 9.949592722518542e-05, "loss": 0.0463, "step": 3720 }, { "grad_norm": 0.07220559567213058, "learning_rate": 9.949005546387631e-05, "loss": 0.0468, "step": 3730 }, { "grad_norm": 0.057533957064151764, "learning_rate": 9.948414987658098e-05, "loss": 0.0452, "step": 3740 }, { "grad_norm": 0.04784741252660751, "learning_rate": 9.94782104673358e-05, "loss": 0.0452, "step": 3750 }, { "grad_norm": 0.05492126941680908, "learning_rate": 9.947223724020034e-05, "loss": 0.0425, "step": 3760 }, { "grad_norm": 0.04875916615128517, "learning_rate": 9.94662301992572e-05, "loss": 0.0424, "step": 3770 }, { "grad_norm": 0.06268943846225739, "learning_rate": 9.946018934861216e-05, "loss": 0.0451, "step": 3780 }, { "grad_norm": 0.058763157576322556, "learning_rate": 9.945411469239409e-05, "loss": 0.0428, "step": 3790 }, { "grad_norm": 0.05234167352318764, "learning_rate": 9.944800623475497e-05, "loss": 0.0428, "step": 3800 }, { "grad_norm": 0.07863766700029373, "learning_rate": 9.944186397986984e-05, "loss": 0.0431, "step": 3810 }, { "grad_norm": 0.07182027399539948, "learning_rate": 9.943568793193687e-05, "loss": 0.045, "step": 3820 }, { "grad_norm": 0.06332546472549438, "learning_rate": 9.942947809517737e-05, "loss": 0.0489, "step": 3830 }, { "grad_norm": 0.0637039914727211, "learning_rate": 9.942323447383568e-05, "loss": 0.0452, "step": 3840 }, { "grad_norm": 0.077239491045475, "learning_rate": 9.941695707217925e-05, "loss": 0.045, "step": 3850 }, { "grad_norm": 0.06391826272010803, "learning_rate": 9.941064589449865e-05, "loss": 0.0435, "step": 3860 }, { "grad_norm": 0.0846240371465683, "learning_rate": 9.940430094510748e-05, "loss": 0.0442, "step": 3870 }, { "grad_norm": 0.13830937445163727, "learning_rate": 9.939792222834248e-05, "loss": 0.0426, "step": 3880 }, { "grad_norm": 0.07719354331493378, "learning_rate": 9.939150974856343e-05, "loss": 0.0449, "step": 3890 }, { "grad_norm": 0.04989132285118103, "learning_rate": 9.93850635101532e-05, "loss": 0.0433, "step": 3900 }, { "grad_norm": 0.045309264212846756, "learning_rate": 9.937858351751775e-05, "loss": 0.0412, "step": 3910 }, { "grad_norm": 0.05743323266506195, "learning_rate": 9.937206977508604e-05, "loss": 0.0437, "step": 3920 }, { "grad_norm": 0.0545303039252758, "learning_rate": 9.936552228731022e-05, "loss": 0.0406, "step": 3930 }, { "grad_norm": 0.07506974786520004, "learning_rate": 9.93589410586654e-05, "loss": 0.0383, "step": 3940 }, { "grad_norm": 0.05939576402306557, "learning_rate": 9.93523260936498e-05, "loss": 0.0438, "step": 3950 }, { "grad_norm": 0.06826843321323395, "learning_rate": 9.934567739678467e-05, "loss": 0.0419, "step": 3960 }, { "grad_norm": 0.05761231854557991, "learning_rate": 9.933899497261433e-05, "loss": 0.0439, "step": 3970 }, { "grad_norm": 0.04136620834469795, "learning_rate": 9.933227882570617e-05, "loss": 0.043, "step": 3980 }, { "grad_norm": 0.0491432324051857, "learning_rate": 9.932552896065062e-05, "loss": 0.0445, "step": 3990 }, { "grad_norm": 0.07109924405813217, "learning_rate": 9.931874538206114e-05, "loss": 0.0399, "step": 4000 }, { "grad_norm": 0.04953774809837341, "learning_rate": 9.931192809457423e-05, "loss": 0.0403, "step": 4010 }, { "grad_norm": 0.043078552931547165, "learning_rate": 9.930507710284944e-05, "loss": 0.0423, "step": 4020 }, { "grad_norm": 0.07809619605541229, "learning_rate": 9.92981924115694e-05, "loss": 0.0445, "step": 4030 }, { "grad_norm": 0.0649632066488266, "learning_rate": 9.929127402543968e-05, "loss": 0.0427, "step": 4040 }, { "grad_norm": 0.0625857561826706, "learning_rate": 9.928432194918895e-05, "loss": 0.0426, "step": 4050 }, { "grad_norm": 0.05079629272222519, "learning_rate": 9.927733618756889e-05, "loss": 0.0402, "step": 4060 }, { "grad_norm": 0.0793352723121643, "learning_rate": 9.92703167453542e-05, "loss": 0.0406, "step": 4070 }, { "grad_norm": 0.0659974217414856, "learning_rate": 9.926326362734263e-05, "loss": 0.0442, "step": 4080 }, { "grad_norm": 0.056894779205322266, "learning_rate": 9.925617683835489e-05, "loss": 0.0427, "step": 4090 }, { "grad_norm": 0.0629073977470398, "learning_rate": 9.924905638323472e-05, "loss": 0.0465, "step": 4100 }, { "grad_norm": 0.06221052259206772, "learning_rate": 9.92419022668489e-05, "loss": 0.0426, "step": 4110 }, { "grad_norm": 0.05419164523482323, "learning_rate": 9.923471449408723e-05, "loss": 0.043, "step": 4120 }, { "grad_norm": 0.0566910021007061, "learning_rate": 9.922749306986244e-05, "loss": 0.0432, "step": 4130 }, { "grad_norm": 0.058813292533159256, "learning_rate": 9.922023799911033e-05, "loss": 0.0442, "step": 4140 }, { "grad_norm": 0.05137203633785248, "learning_rate": 9.921294928678968e-05, "loss": 0.0467, "step": 4150 }, { "grad_norm": 0.043598175048828125, "learning_rate": 9.920562693788222e-05, "loss": 0.0447, "step": 4160 }, { "grad_norm": 0.055789265781641006, "learning_rate": 9.919827095739273e-05, "loss": 0.0468, "step": 4170 }, { "grad_norm": 0.06915802508592606, "learning_rate": 9.919088135034898e-05, "loss": 0.0463, "step": 4180 }, { "grad_norm": 0.0431082546710968, "learning_rate": 9.918345812180165e-05, "loss": 0.0446, "step": 4190 }, { "grad_norm": 0.05634433031082153, "learning_rate": 9.917600127682446e-05, "loss": 0.0463, "step": 4200 }, { "grad_norm": 0.04613877460360527, "learning_rate": 9.91685108205141e-05, "loss": 0.0454, "step": 4210 }, { "grad_norm": 0.04718933254480362, "learning_rate": 9.916098675799024e-05, "loss": 0.044, "step": 4220 }, { "grad_norm": 0.04614855721592903, "learning_rate": 9.915342909439548e-05, "loss": 0.0443, "step": 4230 }, { "grad_norm": 0.059606652706861496, "learning_rate": 9.914583783489543e-05, "loss": 0.044, "step": 4240 }, { "grad_norm": 0.06100767105817795, "learning_rate": 9.913821298467863e-05, "loss": 0.0435, "step": 4250 }, { "grad_norm": 0.055141184478998184, "learning_rate": 9.913055454895661e-05, "loss": 0.0415, "step": 4260 }, { "grad_norm": 0.053037405014038086, "learning_rate": 9.912286253296381e-05, "loss": 0.0421, "step": 4270 }, { "grad_norm": 0.06822970509529114, "learning_rate": 9.911513694195768e-05, "loss": 0.0411, "step": 4280 }, { "grad_norm": 0.05659357085824013, "learning_rate": 9.910737778121859e-05, "loss": 0.0439, "step": 4290 }, { "grad_norm": 0.05866827443242073, "learning_rate": 9.909958505604984e-05, "loss": 0.0402, "step": 4300 }, { "grad_norm": 0.0436566025018692, "learning_rate": 9.909175877177767e-05, "loss": 0.041, "step": 4310 }, { "grad_norm": 0.07623904943466187, "learning_rate": 9.908389893375129e-05, "loss": 0.0431, "step": 4320 }, { "grad_norm": 0.03823840990662575, "learning_rate": 9.907600554734283e-05, "loss": 0.0436, "step": 4330 }, { "grad_norm": 0.060109447687864304, "learning_rate": 9.906807861794734e-05, "loss": 0.047, "step": 4340 }, { "grad_norm": 0.09651394933462143, "learning_rate": 9.906011815098279e-05, "loss": 0.0413, "step": 4350 }, { "grad_norm": 0.044976234436035156, "learning_rate": 9.90521241518901e-05, "loss": 0.0461, "step": 4360 }, { "grad_norm": 0.06315241008996964, "learning_rate": 9.904409662613308e-05, "loss": 0.0454, "step": 4370 }, { "grad_norm": 0.0405769869685173, "learning_rate": 9.903603557919849e-05, "loss": 0.0442, "step": 4380 }, { "grad_norm": 0.043944351375103, "learning_rate": 9.902794101659594e-05, "loss": 0.0463, "step": 4390 }, { "grad_norm": 0.05946527421474457, "learning_rate": 9.901981294385803e-05, "loss": 0.0436, "step": 4400 }, { "grad_norm": 0.04428481683135033, "learning_rate": 9.901165136654018e-05, "loss": 0.0422, "step": 4410 }, { "grad_norm": 0.045260775834321976, "learning_rate": 9.900345629022079e-05, "loss": 0.0427, "step": 4420 }, { "grad_norm": 0.05737404525279999, "learning_rate": 9.89952277205011e-05, "loss": 0.0418, "step": 4430 }, { "grad_norm": 0.04678422212600708, "learning_rate": 9.898696566300527e-05, "loss": 0.0412, "step": 4440 }, { "grad_norm": 0.046711377799510956, "learning_rate": 9.897867012338032e-05, "loss": 0.0403, "step": 4450 }, { "grad_norm": 0.06759215146303177, "learning_rate": 9.897034110729617e-05, "loss": 0.0415, "step": 4460 }, { "grad_norm": 0.050450798124074936, "learning_rate": 9.896197862044564e-05, "loss": 0.0432, "step": 4470 }, { "grad_norm": 0.10375609248876572, "learning_rate": 9.895358266854442e-05, "loss": 0.0423, "step": 4480 }, { "grad_norm": 0.06803005933761597, "learning_rate": 9.894515325733103e-05, "loss": 0.0406, "step": 4490 }, { "grad_norm": 0.05139078199863434, "learning_rate": 9.893669039256693e-05, "loss": 0.0455, "step": 4500 }, { "grad_norm": 0.05076520889997482, "learning_rate": 9.89281940800364e-05, "loss": 0.0421, "step": 4510 }, { "grad_norm": 0.05820799246430397, "learning_rate": 9.891966432554655e-05, "loss": 0.0417, "step": 4520 }, { "grad_norm": 0.047028765082359314, "learning_rate": 9.891110113492745e-05, "loss": 0.0417, "step": 4530 }, { "grad_norm": 0.044017594307661057, "learning_rate": 9.89025045140319e-05, "loss": 0.0435, "step": 4540 }, { "grad_norm": 0.051413774490356445, "learning_rate": 9.889387446873567e-05, "loss": 0.0438, "step": 4550 }, { "grad_norm": 0.04441475123167038, "learning_rate": 9.888521100493726e-05, "loss": 0.0422, "step": 4560 }, { "grad_norm": 0.0464717335999012, "learning_rate": 9.887651412855809e-05, "loss": 0.0429, "step": 4570 }, { "grad_norm": 0.0876283198595047, "learning_rate": 9.88677838455424e-05, "loss": 0.0443, "step": 4580 }, { "grad_norm": 0.052705805748701096, "learning_rate": 9.885902016185725e-05, "loss": 0.0455, "step": 4590 }, { "grad_norm": 0.04917502775788307, "learning_rate": 9.885022308349252e-05, "loss": 0.046, "step": 4600 }, { "grad_norm": 0.045709896832704544, "learning_rate": 9.884139261646095e-05, "loss": 0.0452, "step": 4610 }, { "grad_norm": 0.06743483990430832, "learning_rate": 9.883252876679807e-05, "loss": 0.0434, "step": 4620 }, { "grad_norm": 0.059212006628513336, "learning_rate": 9.882363154056225e-05, "loss": 0.0438, "step": 4630 }, { "grad_norm": 0.10208743065595627, "learning_rate": 9.881470094383465e-05, "loss": 0.0426, "step": 4640 }, { "grad_norm": 0.06793592870235443, "learning_rate": 9.880573698271924e-05, "loss": 0.0449, "step": 4650 }, { "grad_norm": 0.05163419619202614, "learning_rate": 9.879673966334282e-05, "loss": 0.0456, "step": 4660 }, { "grad_norm": 0.06751307100057602, "learning_rate": 9.878770899185496e-05, "loss": 0.0432, "step": 4670 }, { "grad_norm": 0.06350693851709366, "learning_rate": 9.877864497442804e-05, "loss": 0.0455, "step": 4680 }, { "grad_norm": 0.07016913592815399, "learning_rate": 9.876954761725723e-05, "loss": 0.0423, "step": 4690 }, { "grad_norm": 0.06397218257188797, "learning_rate": 9.876041692656052e-05, "loss": 0.0432, "step": 4700 }, { "grad_norm": 0.05574337765574455, "learning_rate": 9.87512529085786e-05, "loss": 0.0414, "step": 4710 }, { "grad_norm": 0.06611631065607071, "learning_rate": 9.874205556957503e-05, "loss": 0.0417, "step": 4720 }, { "grad_norm": 0.05385385453701019, "learning_rate": 9.873282491583608e-05, "loss": 0.0378, "step": 4730 }, { "grad_norm": 0.0661901980638504, "learning_rate": 9.872356095367084e-05, "loss": 0.0388, "step": 4740 }, { "grad_norm": 0.05981941521167755, "learning_rate": 9.87142636894111e-05, "loss": 0.0362, "step": 4750 }, { "grad_norm": 0.05184980481863022, "learning_rate": 9.870493312941148e-05, "loss": 0.0392, "step": 4760 }, { "grad_norm": 0.03841022774577141, "learning_rate": 9.869556928004933e-05, "loss": 0.0354, "step": 4770 }, { "grad_norm": 0.05078727379441261, "learning_rate": 9.868617214772476e-05, "loss": 0.0443, "step": 4780 }, { "grad_norm": 0.061528027057647705, "learning_rate": 9.867674173886059e-05, "loss": 0.0487, "step": 4790 }, { "grad_norm": 0.05148348957300186, "learning_rate": 9.866727805990245e-05, "loss": 0.0478, "step": 4800 }, { "grad_norm": 0.11182818561792374, "learning_rate": 9.865778111731865e-05, "loss": 0.0471, "step": 4810 }, { "grad_norm": 0.08649438619613647, "learning_rate": 9.864825091760028e-05, "loss": 0.0479, "step": 4820 }, { "grad_norm": 0.05881538242101669, "learning_rate": 9.863868746726112e-05, "loss": 0.0474, "step": 4830 }, { "grad_norm": 0.056231796741485596, "learning_rate": 9.862909077283773e-05, "loss": 0.0462, "step": 4840 }, { "grad_norm": 0.06580527871847153, "learning_rate": 9.861946084088933e-05, "loss": 0.0478, "step": 4850 }, { "grad_norm": 0.054070547223091125, "learning_rate": 9.860979767799792e-05, "loss": 0.0468, "step": 4860 }, { "grad_norm": 0.05439769849181175, "learning_rate": 9.860010129076813e-05, "loss": 0.0445, "step": 4870 }, { "grad_norm": 0.062173355370759964, "learning_rate": 9.859037168582741e-05, "loss": 0.0433, "step": 4880 }, { "grad_norm": 0.07127843052148819, "learning_rate": 9.85806088698258e-05, "loss": 0.0476, "step": 4890 }, { "grad_norm": 0.08381424099206924, "learning_rate": 9.857081284943612e-05, "loss": 0.0423, "step": 4900 }, { "grad_norm": 0.06454094499349594, "learning_rate": 9.856098363135385e-05, "loss": 0.0434, "step": 4910 }, { "grad_norm": 0.06109243631362915, "learning_rate": 9.855112122229717e-05, "loss": 0.0459, "step": 4920 }, { "grad_norm": 0.0636216253042221, "learning_rate": 9.854122562900696e-05, "loss": 0.0466, "step": 4930 }, { "grad_norm": 0.055412694811820984, "learning_rate": 9.853129685824673e-05, "loss": 0.0447, "step": 4940 }, { "grad_norm": 0.08721889555454254, "learning_rate": 9.852133491680273e-05, "loss": 0.0451, "step": 4950 }, { "grad_norm": 0.04254787415266037, "learning_rate": 9.851133981148385e-05, "loss": 0.0443, "step": 4960 }, { "grad_norm": 0.054726485162973404, "learning_rate": 9.850131154912164e-05, "loss": 0.0451, "step": 4970 }, { "grad_norm": 0.05717488378286362, "learning_rate": 9.849125013657031e-05, "loss": 0.0402, "step": 4980 }, { "grad_norm": 0.052093759179115295, "learning_rate": 9.848115558070676e-05, "loss": 0.039, "step": 4990 }, { "grad_norm": 0.04417464882135391, "learning_rate": 9.847102788843055e-05, "loss": 0.0395, "step": 5000 }, { "grad_norm": 0.07429774105548859, "learning_rate": 9.84608670666638e-05, "loss": 0.0407, "step": 5010 }, { "grad_norm": 0.04496758431196213, "learning_rate": 9.845067312235138e-05, "loss": 0.0385, "step": 5020 }, { "grad_norm": 0.09955719113349915, "learning_rate": 9.844044606246074e-05, "loss": 0.0388, "step": 5030 }, { "grad_norm": 0.07582191377878189, "learning_rate": 9.843018589398199e-05, "loss": 0.0412, "step": 5040 }, { "grad_norm": 0.046303339302539825, "learning_rate": 9.841989262392785e-05, "loss": 0.0457, "step": 5050 }, { "grad_norm": 0.059346895664930344, "learning_rate": 9.840956625933367e-05, "loss": 0.0437, "step": 5060 }, { "grad_norm": 0.061776548624038696, "learning_rate": 9.839920680725743e-05, "loss": 0.042, "step": 5070 }, { "grad_norm": 0.0537252314388752, "learning_rate": 9.83888142747797e-05, "loss": 0.04, "step": 5080 }, { "grad_norm": 0.05093883350491524, "learning_rate": 9.83783886690037e-05, "loss": 0.0447, "step": 5090 }, { "grad_norm": 0.046587023884058, "learning_rate": 9.836792999705524e-05, "loss": 0.0443, "step": 5100 }, { "grad_norm": 0.05824343115091324, "learning_rate": 9.835743826608271e-05, "loss": 0.0435, "step": 5110 }, { "grad_norm": 0.0643501803278923, "learning_rate": 9.834691348325709e-05, "loss": 0.0433, "step": 5120 }, { "grad_norm": 0.03707367554306984, "learning_rate": 9.833635565577199e-05, "loss": 0.0421, "step": 5130 }, { "grad_norm": 0.08548599481582642, "learning_rate": 9.832576479084359e-05, "loss": 0.0435, "step": 5140 }, { "grad_norm": 0.0535120852291584, "learning_rate": 9.831514089571064e-05, "loss": 0.044, "step": 5150 }, { "grad_norm": 0.062279101461172104, "learning_rate": 9.830448397763447e-05, "loss": 0.0433, "step": 5160 }, { "grad_norm": 0.06313475966453552, "learning_rate": 9.8293794043899e-05, "loss": 0.045, "step": 5170 }, { "grad_norm": 0.06006781384348869, "learning_rate": 9.828307110181066e-05, "loss": 0.0428, "step": 5180 }, { "grad_norm": 0.04552990570664406, "learning_rate": 9.827231515869852e-05, "loss": 0.0433, "step": 5190 }, { "grad_norm": 0.04895886406302452, "learning_rate": 9.826152622191413e-05, "loss": 0.0428, "step": 5200 }, { "grad_norm": 0.05123995989561081, "learning_rate": 9.825070429883167e-05, "loss": 0.0427, "step": 5210 }, { "grad_norm": 0.05791044607758522, "learning_rate": 9.82398493968478e-05, "loss": 0.0429, "step": 5220 }, { "grad_norm": 0.0713021382689476, "learning_rate": 9.822896152338173e-05, "loss": 0.0471, "step": 5230 }, { "grad_norm": 0.05635274946689606, "learning_rate": 9.821804068587523e-05, "loss": 0.0467, "step": 5240 }, { "grad_norm": 0.04692878574132919, "learning_rate": 9.820708689179259e-05, "loss": 0.0476, "step": 5250 }, { "grad_norm": 0.05848146229982376, "learning_rate": 9.819610014862063e-05, "loss": 0.0462, "step": 5260 }, { "grad_norm": 0.03826320916414261, "learning_rate": 9.818508046386868e-05, "loss": 0.0483, "step": 5270 }, { "grad_norm": 0.05528901517391205, "learning_rate": 9.817402784506859e-05, "loss": 0.0478, "step": 5280 }, { "grad_norm": 0.051738567650318146, "learning_rate": 9.816294229977472e-05, "loss": 0.0426, "step": 5290 }, { "grad_norm": 0.05163797363638878, "learning_rate": 9.815182383556394e-05, "loss": 0.0412, "step": 5300 }, { "grad_norm": 0.03770492598414421, "learning_rate": 9.81406724600356e-05, "loss": 0.0413, "step": 5310 }, { "grad_norm": 0.07742459326982498, "learning_rate": 9.812948818081161e-05, "loss": 0.0414, "step": 5320 }, { "grad_norm": 0.050320062786340714, "learning_rate": 9.811827100553623e-05, "loss": 0.0417, "step": 5330 }, { "grad_norm": 0.055196162313222885, "learning_rate": 9.810702094187638e-05, "loss": 0.0407, "step": 5340 }, { "grad_norm": 0.06390225887298584, "learning_rate": 9.809573799752135e-05, "loss": 0.0411, "step": 5350 }, { "grad_norm": 0.05394379422068596, "learning_rate": 9.808442218018288e-05, "loss": 0.0431, "step": 5360 }, { "grad_norm": 0.07818488031625748, "learning_rate": 9.807307349759527e-05, "loss": 0.0433, "step": 5370 }, { "grad_norm": 0.05923744663596153, "learning_rate": 9.806169195751525e-05, "loss": 0.043, "step": 5380 }, { "grad_norm": 0.07308276742696762, "learning_rate": 9.805027756772194e-05, "loss": 0.0417, "step": 5390 }, { "grad_norm": 0.06887039542198181, "learning_rate": 9.803883033601702e-05, "loss": 0.0416, "step": 5400 }, { "grad_norm": 0.06647087633609772, "learning_rate": 9.802735027022453e-05, "loss": 0.044, "step": 5410 }, { "grad_norm": 0.05700593441724777, "learning_rate": 9.8015837378191e-05, "loss": 0.0427, "step": 5420 }, { "grad_norm": 0.05116911232471466, "learning_rate": 9.80042916677854e-05, "loss": 0.0427, "step": 5430 }, { "grad_norm": 0.08104893565177917, "learning_rate": 9.799271314689908e-05, "loss": 0.0408, "step": 5440 }, { "grad_norm": 0.0563662126660347, "learning_rate": 9.798110182344588e-05, "loss": 0.0401, "step": 5450 }, { "grad_norm": 0.05362028256058693, "learning_rate": 9.796945770536204e-05, "loss": 0.0404, "step": 5460 }, { "grad_norm": 0.0691612958908081, "learning_rate": 9.795778080060615e-05, "loss": 0.0392, "step": 5470 }, { "grad_norm": 0.05778533220291138, "learning_rate": 9.79460711171593e-05, "loss": 0.0375, "step": 5480 }, { "grad_norm": 0.05493458732962608, "learning_rate": 9.793432866302497e-05, "loss": 0.0396, "step": 5490 }, { "grad_norm": 0.0501411110162735, "learning_rate": 9.792255344622897e-05, "loss": 0.0394, "step": 5500 }, { "grad_norm": 0.053000956773757935, "learning_rate": 9.791074547481957e-05, "loss": 0.0386, "step": 5510 }, { "grad_norm": 0.04364890977740288, "learning_rate": 9.789890475686742e-05, "loss": 0.0376, "step": 5520 }, { "grad_norm": 0.06141112744808197, "learning_rate": 9.788703130046552e-05, "loss": 0.0382, "step": 5530 }, { "grad_norm": 0.056923698633909225, "learning_rate": 9.78751251137293e-05, "loss": 0.0409, "step": 5540 }, { "grad_norm": 0.04582624509930611, "learning_rate": 9.786318620479646e-05, "loss": 0.0417, "step": 5550 }, { "grad_norm": 0.06588783115148544, "learning_rate": 9.78512145818272e-05, "loss": 0.0442, "step": 5560 }, { "grad_norm": 0.05995553731918335, "learning_rate": 9.783921025300398e-05, "loss": 0.0433, "step": 5570 }, { "grad_norm": 0.05428270995616913, "learning_rate": 9.782717322653164e-05, "loss": 0.042, "step": 5580 }, { "grad_norm": 0.06556905806064606, "learning_rate": 9.781510351063738e-05, "loss": 0.042, "step": 5590 }, { "grad_norm": 0.04956110566854477, "learning_rate": 9.780300111357075e-05, "loss": 0.0453, "step": 5600 }, { "grad_norm": 0.044103022664785385, "learning_rate": 9.779086604360361e-05, "loss": 0.0405, "step": 5610 }, { "grad_norm": 0.06309740990400314, "learning_rate": 9.777869830903019e-05, "loss": 0.0414, "step": 5620 }, { "grad_norm": 0.05292193591594696, "learning_rate": 9.776649791816698e-05, "loss": 0.0382, "step": 5630 }, { "grad_norm": 0.059573885053396225, "learning_rate": 9.775426487935285e-05, "loss": 0.0376, "step": 5640 }, { "grad_norm": 0.08558297902345657, "learning_rate": 9.774199920094898e-05, "loss": 0.0357, "step": 5650 }, { "grad_norm": 0.05572619289159775, "learning_rate": 9.772970089133884e-05, "loss": 0.035, "step": 5660 }, { "grad_norm": 0.05325151979923248, "learning_rate": 9.77173699589282e-05, "loss": 0.038, "step": 5670 }, { "grad_norm": 0.05231250077486038, "learning_rate": 9.770500641214513e-05, "loss": 0.0408, "step": 5680 }, { "grad_norm": 0.04686959832906723, "learning_rate": 9.769261025944003e-05, "loss": 0.0429, "step": 5690 }, { "grad_norm": 0.035469766706228256, "learning_rate": 9.768018150928552e-05, "loss": 0.0401, "step": 5700 }, { "grad_norm": 0.048687368631362915, "learning_rate": 9.766772017017654e-05, "loss": 0.0408, "step": 5710 }, { "grad_norm": 0.0760556310415268, "learning_rate": 9.765522625063032e-05, "loss": 0.0394, "step": 5720 }, { "grad_norm": 0.055590786039829254, "learning_rate": 9.76426997591863e-05, "loss": 0.0402, "step": 5730 }, { "grad_norm": 0.047999609261751175, "learning_rate": 9.763014070440625e-05, "loss": 0.0409, "step": 5740 }, { "grad_norm": 0.06492457538843155, "learning_rate": 9.761754909487415e-05, "loss": 0.0438, "step": 5750 }, { "grad_norm": 0.04460734501481056, "learning_rate": 9.760492493919626e-05, "loss": 0.043, "step": 5760 }, { "grad_norm": 0.09642166644334793, "learning_rate": 9.759226824600105e-05, "loss": 0.0428, "step": 5770 }, { "grad_norm": 0.05310951918363571, "learning_rate": 9.757957902393928e-05, "loss": 0.0419, "step": 5780 }, { "grad_norm": 0.053452592343091965, "learning_rate": 9.756685728168387e-05, "loss": 0.0423, "step": 5790 }, { "grad_norm": 0.05352841690182686, "learning_rate": 9.755410302793004e-05, "loss": 0.0421, "step": 5800 }, { "grad_norm": 0.06520843505859375, "learning_rate": 9.754131627139522e-05, "loss": 0.0446, "step": 5810 }, { "grad_norm": 0.06082657724618912, "learning_rate": 9.752849702081901e-05, "loss": 0.0427, "step": 5820 }, { "grad_norm": 0.059873539954423904, "learning_rate": 9.751564528496324e-05, "loss": 0.0403, "step": 5830 }, { "grad_norm": 0.05426532030105591, "learning_rate": 9.750276107261197e-05, "loss": 0.0401, "step": 5840 }, { "grad_norm": 0.058970700949430466, "learning_rate": 9.748984439257142e-05, "loss": 0.0394, "step": 5850 }, { "grad_norm": 0.052618078887462616, "learning_rate": 9.747689525367005e-05, "loss": 0.0415, "step": 5860 }, { "grad_norm": 0.06420009583234787, "learning_rate": 9.746391366475845e-05, "loss": 0.0446, "step": 5870 }, { "grad_norm": 0.09275595098733902, "learning_rate": 9.745089963470942e-05, "loss": 0.0419, "step": 5880 }, { "grad_norm": 0.0568157434463501, "learning_rate": 9.743785317241791e-05, "loss": 0.0422, "step": 5890 }, { "grad_norm": 0.057850755751132965, "learning_rate": 9.742477428680108e-05, "loss": 0.0422, "step": 5900 }, { "grad_norm": 0.0609794445335865, "learning_rate": 9.741166298679821e-05, "loss": 0.0431, "step": 5910 }, { "grad_norm": 0.05269990861415863, "learning_rate": 9.739851928137076e-05, "loss": 0.0393, "step": 5920 }, { "grad_norm": 0.06318085640668869, "learning_rate": 9.738534317950232e-05, "loss": 0.0399, "step": 5930 }, { "grad_norm": 0.039285726845264435, "learning_rate": 9.737213469019864e-05, "loss": 0.0387, "step": 5940 }, { "grad_norm": 0.0734267383813858, "learning_rate": 9.735889382248757e-05, "loss": 0.0384, "step": 5950 }, { "grad_norm": 0.05478064343333244, "learning_rate": 9.734562058541916e-05, "loss": 0.0378, "step": 5960 }, { "grad_norm": 0.04616048187017441, "learning_rate": 9.733231498806552e-05, "loss": 0.0412, "step": 5970 }, { "grad_norm": 0.05148851498961449, "learning_rate": 9.731897703952088e-05, "loss": 0.0398, "step": 5980 }, { "grad_norm": 0.07655031979084015, "learning_rate": 9.730560674890165e-05, "loss": 0.0393, "step": 5990 }, { "grad_norm": 0.06913978606462479, "learning_rate": 9.729220412534628e-05, "loss": 0.0413, "step": 6000 }, { "grad_norm": 0.057603415101766586, "learning_rate": 9.72787691780153e-05, "loss": 0.0408, "step": 6010 }, { "grad_norm": 0.04165391996502876, "learning_rate": 9.726530191609142e-05, "loss": 0.0409, "step": 6020 }, { "grad_norm": 0.09652659296989441, "learning_rate": 9.725180234877937e-05, "loss": 0.0417, "step": 6030 }, { "grad_norm": 0.05904829129576683, "learning_rate": 9.723827048530595e-05, "loss": 0.0401, "step": 6040 }, { "grad_norm": 0.035309549421072006, "learning_rate": 9.722470633492011e-05, "loss": 0.0431, "step": 6050 }, { "grad_norm": 0.049838997423648834, "learning_rate": 9.721110990689278e-05, "loss": 0.0455, "step": 6060 }, { "grad_norm": 0.06258372962474823, "learning_rate": 9.719748121051699e-05, "loss": 0.0444, "step": 6070 }, { "grad_norm": 0.060969654470682144, "learning_rate": 9.718382025510785e-05, "loss": 0.0452, "step": 6080 }, { "grad_norm": 0.05122325196862221, "learning_rate": 9.717012705000248e-05, "loss": 0.0426, "step": 6090 }, { "grad_norm": 0.04759535193443298, "learning_rate": 9.715640160456006e-05, "loss": 0.0441, "step": 6100 }, { "grad_norm": 0.041857991367578506, "learning_rate": 9.714264392816181e-05, "loss": 0.0433, "step": 6110 }, { "grad_norm": 0.06759005784988403, "learning_rate": 9.712885403021095e-05, "loss": 0.0445, "step": 6120 }, { "grad_norm": 0.07707566767930984, "learning_rate": 9.711503192013276e-05, "loss": 0.0404, "step": 6130 }, { "grad_norm": 0.06253968924283981, "learning_rate": 9.710117760737454e-05, "loss": 0.0436, "step": 6140 }, { "grad_norm": 0.07089322805404663, "learning_rate": 9.708729110140554e-05, "loss": 0.0411, "step": 6150 }, { "grad_norm": 0.07494813948869705, "learning_rate": 9.707337241171709e-05, "loss": 0.0418, "step": 6160 }, { "grad_norm": 0.05714556947350502, "learning_rate": 9.705942154782245e-05, "loss": 0.039, "step": 6170 }, { "grad_norm": 0.0634472668170929, "learning_rate": 9.704543851925696e-05, "loss": 0.0428, "step": 6180 }, { "grad_norm": 0.06839419901371002, "learning_rate": 9.703142333557784e-05, "loss": 0.0392, "step": 6190 }, { "grad_norm": 0.07753591984510422, "learning_rate": 9.701737600636436e-05, "loss": 0.0405, "step": 6200 }, { "grad_norm": 0.050061386078596115, "learning_rate": 9.700329654121771e-05, "loss": 0.0389, "step": 6210 }, { "grad_norm": 0.052719201892614365, "learning_rate": 9.698918494976109e-05, "loss": 0.039, "step": 6220 }, { "grad_norm": 0.04353303462266922, "learning_rate": 9.697504124163965e-05, "loss": 0.0377, "step": 6230 }, { "grad_norm": 0.14199867844581604, "learning_rate": 9.696086542652045e-05, "loss": 0.0427, "step": 6240 }, { "grad_norm": 0.06706629693508148, "learning_rate": 9.694665751409256e-05, "loss": 0.0431, "step": 6250 }, { "grad_norm": 0.10752968490123749, "learning_rate": 9.693241751406694e-05, "loss": 0.0425, "step": 6260 }, { "grad_norm": 0.0829964131116867, "learning_rate": 9.69181454361765e-05, "loss": 0.0401, "step": 6270 }, { "grad_norm": 0.10614679008722305, "learning_rate": 9.690384129017605e-05, "loss": 0.0427, "step": 6280 }, { "grad_norm": 0.07094824314117432, "learning_rate": 9.688950508584237e-05, "loss": 0.0417, "step": 6290 }, { "grad_norm": 0.0690661072731018, "learning_rate": 9.68751368329741e-05, "loss": 0.041, "step": 6300 }, { "grad_norm": 0.062451086938381195, "learning_rate": 9.68607365413918e-05, "loss": 0.0407, "step": 6310 }, { "grad_norm": 0.05529627576470375, "learning_rate": 9.684630422093797e-05, "loss": 0.0456, "step": 6320 }, { "grad_norm": 0.057217929512262344, "learning_rate": 9.683183988147693e-05, "loss": 0.0431, "step": 6330 }, { "grad_norm": 0.05015788972377777, "learning_rate": 9.681734353289491e-05, "loss": 0.0425, "step": 6340 }, { "grad_norm": 0.060140181332826614, "learning_rate": 9.680281518510006e-05, "loss": 0.0425, "step": 6350 }, { "grad_norm": 0.05525582656264305, "learning_rate": 9.678825484802236e-05, "loss": 0.0436, "step": 6360 }, { "grad_norm": 0.05091290920972824, "learning_rate": 9.677366253161365e-05, "loss": 0.0426, "step": 6370 }, { "grad_norm": 0.0580410398542881, "learning_rate": 9.675903824584765e-05, "loss": 0.0418, "step": 6380 }, { "grad_norm": 0.052617527544498444, "learning_rate": 9.674438200071991e-05, "loss": 0.0414, "step": 6390 }, { "grad_norm": 0.04400375485420227, "learning_rate": 9.672969380624786e-05, "loss": 0.0408, "step": 6400 }, { "grad_norm": 0.08750295639038086, "learning_rate": 9.671497367247073e-05, "loss": 0.0412, "step": 6410 }, { "grad_norm": 0.06730473786592484, "learning_rate": 9.670022160944959e-05, "loss": 0.0386, "step": 6420 }, { "grad_norm": 0.06117861345410347, "learning_rate": 9.668543762726734e-05, "loss": 0.0368, "step": 6430 }, { "grad_norm": 0.05880792811512947, "learning_rate": 9.667062173602868e-05, "loss": 0.0396, "step": 6440 }, { "grad_norm": 0.052613355219364166, "learning_rate": 9.665577394586013e-05, "loss": 0.0417, "step": 6450 }, { "grad_norm": 0.0671168640255928, "learning_rate": 9.664089426691006e-05, "loss": 0.0401, "step": 6460 }, { "grad_norm": 0.0420864075422287, "learning_rate": 9.662598270934852e-05, "loss": 0.0409, "step": 6470 }, { "grad_norm": 0.05656478554010391, "learning_rate": 9.661103928336748e-05, "loss": 0.0394, "step": 6480 }, { "grad_norm": 0.05506303161382675, "learning_rate": 9.659606399918058e-05, "loss": 0.0399, "step": 6490 }, { "grad_norm": 0.06743636727333069, "learning_rate": 9.658105686702334e-05, "loss": 0.0402, "step": 6500 }, { "grad_norm": 0.05013569816946983, "learning_rate": 9.656601789715294e-05, "loss": 0.0411, "step": 6510 }, { "grad_norm": 0.05402745306491852, "learning_rate": 9.655094709984842e-05, "loss": 0.0403, "step": 6520 }, { "grad_norm": 0.053630564361810684, "learning_rate": 9.653584448541048e-05, "loss": 0.0414, "step": 6530 }, { "grad_norm": 0.07474137097597122, "learning_rate": 9.652071006416166e-05, "loss": 0.0392, "step": 6540 }, { "grad_norm": 0.06346330791711807, "learning_rate": 9.650554384644617e-05, "loss": 0.0404, "step": 6550 }, { "grad_norm": 0.04613901674747467, "learning_rate": 9.649034584263e-05, "loss": 0.0401, "step": 6560 }, { "grad_norm": 0.08091206848621368, "learning_rate": 9.64751160631008e-05, "loss": 0.0396, "step": 6570 }, { "grad_norm": 0.03898392617702484, "learning_rate": 9.645985451826803e-05, "loss": 0.0382, "step": 6580 }, { "grad_norm": 0.07990184426307678, "learning_rate": 9.644456121856275e-05, "loss": 0.0393, "step": 6590 }, { "grad_norm": 0.06927669793367386, "learning_rate": 9.642923617443784e-05, "loss": 0.0383, "step": 6600 }, { "grad_norm": 0.06004578247666359, "learning_rate": 9.641387939636782e-05, "loss": 0.038, "step": 6610 }, { "grad_norm": 0.040226906538009644, "learning_rate": 9.639849089484888e-05, "loss": 0.039, "step": 6620 }, { "grad_norm": 0.06979980319738388, "learning_rate": 9.638307068039893e-05, "loss": 0.0398, "step": 6630 }, { "grad_norm": 0.05603966861963272, "learning_rate": 9.636761876355753e-05, "loss": 0.039, "step": 6640 }, { "grad_norm": 0.06827685236930847, "learning_rate": 9.635213515488593e-05, "loss": 0.0406, "step": 6650 }, { "grad_norm": 0.047143951058387756, "learning_rate": 9.633661986496702e-05, "loss": 0.0412, "step": 6660 }, { "grad_norm": 0.04754280298948288, "learning_rate": 9.632107290440535e-05, "loss": 0.0386, "step": 6670 }, { "grad_norm": 0.047123510390520096, "learning_rate": 9.630549428382715e-05, "loss": 0.0411, "step": 6680 }, { "grad_norm": 0.0517350398004055, "learning_rate": 9.628988401388024e-05, "loss": 0.0409, "step": 6690 }, { "grad_norm": 0.04160955548286438, "learning_rate": 9.627424210523407e-05, "loss": 0.0435, "step": 6700 }, { "grad_norm": 0.04482385143637657, "learning_rate": 9.62585685685798e-05, "loss": 0.0435, "step": 6710 }, { "grad_norm": 0.050509292632341385, "learning_rate": 9.624286341463008e-05, "loss": 0.0424, "step": 6720 }, { "grad_norm": 0.054517101496458054, "learning_rate": 9.622712665411927e-05, "loss": 0.0408, "step": 6730 }, { "grad_norm": 0.05138546600937843, "learning_rate": 9.621135829780328e-05, "loss": 0.0422, "step": 6740 }, { "grad_norm": 0.047953926026821136, "learning_rate": 9.619555835645964e-05, "loss": 0.0402, "step": 6750 }, { "grad_norm": 0.056987661868333817, "learning_rate": 9.617972684088747e-05, "loss": 0.0373, "step": 6760 }, { "grad_norm": 0.047415174543857574, "learning_rate": 9.616386376190745e-05, "loss": 0.0353, "step": 6770 }, { "grad_norm": 0.05927177146077156, "learning_rate": 9.614796913036184e-05, "loss": 0.039, "step": 6780 }, { "grad_norm": 0.05995374172925949, "learning_rate": 9.61320429571145e-05, "loss": 0.0374, "step": 6790 }, { "grad_norm": 0.055134546011686325, "learning_rate": 9.611608525305078e-05, "loss": 0.0375, "step": 6800 }, { "grad_norm": 0.04917697608470917, "learning_rate": 9.610009602907764e-05, "loss": 0.0355, "step": 6810 }, { "grad_norm": 0.055840082466602325, "learning_rate": 9.608407529612357e-05, "loss": 0.0417, "step": 6820 }, { "grad_norm": 0.05842895060777664, "learning_rate": 9.606802306513857e-05, "loss": 0.0424, "step": 6830 }, { "grad_norm": 0.07789282500743866, "learning_rate": 9.605193934709421e-05, "loss": 0.043, "step": 6840 }, { "grad_norm": 0.06607174873352051, "learning_rate": 9.603582415298354e-05, "loss": 0.042, "step": 6850 }, { "grad_norm": 0.039614371955394745, "learning_rate": 9.601967749382117e-05, "loss": 0.0401, "step": 6860 }, { "grad_norm": 0.05573767051100731, "learning_rate": 9.600349938064316e-05, "loss": 0.0404, "step": 6870 }, { "grad_norm": 0.07673023641109467, "learning_rate": 9.598728982450711e-05, "loss": 0.0416, "step": 6880 }, { "grad_norm": 0.07588963210582733, "learning_rate": 9.597104883649211e-05, "loss": 0.0394, "step": 6890 }, { "grad_norm": 0.06337156891822815, "learning_rate": 9.59547764276987e-05, "loss": 0.0371, "step": 6900 }, { "grad_norm": 0.04501822590827942, "learning_rate": 9.593847260924892e-05, "loss": 0.0368, "step": 6910 }, { "grad_norm": 0.04982190951704979, "learning_rate": 9.59221373922863e-05, "loss": 0.038, "step": 6920 }, { "grad_norm": 0.06194625422358513, "learning_rate": 9.59057707879758e-05, "loss": 0.0365, "step": 6930 }, { "grad_norm": 0.05997135862708092, "learning_rate": 9.588937280750382e-05, "loss": 0.0404, "step": 6940 }, { "grad_norm": 0.05525507032871246, "learning_rate": 9.587294346207824e-05, "loss": 0.0418, "step": 6950 }, { "grad_norm": 0.042602408677339554, "learning_rate": 9.585648276292836e-05, "loss": 0.0396, "step": 6960 }, { "grad_norm": 0.07218880951404572, "learning_rate": 9.583999072130492e-05, "loss": 0.0391, "step": 6970 }, { "grad_norm": 0.047239579260349274, "learning_rate": 9.582346734848005e-05, "loss": 0.0419, "step": 6980 }, { "grad_norm": 0.07008599489927292, "learning_rate": 9.580691265574735e-05, "loss": 0.0419, "step": 6990 }, { "grad_norm": 0.05726179853081703, "learning_rate": 9.579032665442178e-05, "loss": 0.039, "step": 7000 }, { "grad_norm": 0.06247785687446594, "learning_rate": 9.577370935583971e-05, "loss": 0.0447, "step": 7010 }, { "grad_norm": 0.08023470640182495, "learning_rate": 9.575706077135893e-05, "loss": 0.0406, "step": 7020 }, { "grad_norm": 0.06761763244867325, "learning_rate": 9.574038091235857e-05, "loss": 0.044, "step": 7030 }, { "grad_norm": 0.04810363054275513, "learning_rate": 9.572366979023916e-05, "loss": 0.0394, "step": 7040 }, { "grad_norm": 0.05483650043606758, "learning_rate": 9.57069274164226e-05, "loss": 0.0407, "step": 7050 }, { "grad_norm": 0.05316483601927757, "learning_rate": 9.569015380235213e-05, "loss": 0.0404, "step": 7060 }, { "grad_norm": 0.05690746009349823, "learning_rate": 9.567334895949238e-05, "loss": 0.0416, "step": 7070 }, { "grad_norm": 0.07938062399625778, "learning_rate": 9.565651289932928e-05, "loss": 0.0406, "step": 7080 }, { "grad_norm": 0.07565378397703171, "learning_rate": 9.563964563337015e-05, "loss": 0.0424, "step": 7090 }, { "grad_norm": 0.039081871509552, "learning_rate": 9.562274717314357e-05, "loss": 0.0417, "step": 7100 }, { "grad_norm": 0.04811544716358185, "learning_rate": 9.56058175301995e-05, "loss": 0.0416, "step": 7110 }, { "grad_norm": 0.04681483656167984, "learning_rate": 9.558885671610918e-05, "loss": 0.0406, "step": 7120 }, { "grad_norm": 0.04373253881931305, "learning_rate": 9.557186474246517e-05, "loss": 0.0417, "step": 7130 }, { "grad_norm": 0.062304552644491196, "learning_rate": 9.555484162088133e-05, "loss": 0.0458, "step": 7140 }, { "grad_norm": 0.05239774286746979, "learning_rate": 9.553778736299279e-05, "loss": 0.0454, "step": 7150 }, { "grad_norm": 0.05595233291387558, "learning_rate": 9.552070198045599e-05, "loss": 0.0462, "step": 7160 }, { "grad_norm": 0.04636126756668091, "learning_rate": 9.550358548494863e-05, "loss": 0.0457, "step": 7170 }, { "grad_norm": 0.06104600057005882, "learning_rate": 9.548643788816965e-05, "loss": 0.0448, "step": 7180 }, { "grad_norm": 0.045988477766513824, "learning_rate": 9.546925920183929e-05, "loss": 0.0441, "step": 7190 }, { "grad_norm": 0.06602706760168076, "learning_rate": 9.545204943769902e-05, "loss": 0.0441, "step": 7200 }, { "grad_norm": 0.055963002145290375, "learning_rate": 9.543480860751155e-05, "loss": 0.0375, "step": 7210 }, { "grad_norm": 0.07249027490615845, "learning_rate": 9.541753672306081e-05, "loss": 0.0403, "step": 7220 }, { "grad_norm": 0.044641152024269104, "learning_rate": 9.540023379615198e-05, "loss": 0.0389, "step": 7230 }, { "grad_norm": 0.06346680968999863, "learning_rate": 9.538289983861146e-05, "loss": 0.0391, "step": 7240 }, { "grad_norm": 0.046721912920475006, "learning_rate": 9.536553486228683e-05, "loss": 0.0379, "step": 7250 }, { "grad_norm": 0.04703133925795555, "learning_rate": 9.534813887904689e-05, "loss": 0.0381, "step": 7260 }, { "grad_norm": 0.059918809682130814, "learning_rate": 9.533071190078163e-05, "loss": 0.04, "step": 7270 }, { "grad_norm": 0.04742594063282013, "learning_rate": 9.531325393940225e-05, "loss": 0.0384, "step": 7280 }, { "grad_norm": 0.061862338334321976, "learning_rate": 9.529576500684105e-05, "loss": 0.0378, "step": 7290 }, { "grad_norm": 0.060238759964704514, "learning_rate": 9.52782451150516e-05, "loss": 0.0381, "step": 7300 }, { "grad_norm": 0.05702019855380058, "learning_rate": 9.526069427600857e-05, "loss": 0.0391, "step": 7310 }, { "grad_norm": 0.05578254163265228, "learning_rate": 9.524311250170776e-05, "loss": 0.0381, "step": 7320 }, { "grad_norm": 0.07243934273719788, "learning_rate": 9.522549980416619e-05, "loss": 0.0395, "step": 7330 }, { "grad_norm": 0.05455142259597778, "learning_rate": 9.520785619542196e-05, "loss": 0.0411, "step": 7340 }, { "grad_norm": 0.06987807154655457, "learning_rate": 9.519018168753428e-05, "loss": 0.0415, "step": 7350 }, { "grad_norm": 0.03339625522494316, "learning_rate": 9.517247629258357e-05, "loss": 0.0391, "step": 7360 }, { "grad_norm": 0.08050449192523956, "learning_rate": 9.515474002267124e-05, "loss": 0.0398, "step": 7370 }, { "grad_norm": 0.049364522099494934, "learning_rate": 9.513697288991989e-05, "loss": 0.04, "step": 7380 }, { "grad_norm": 0.055678848177194595, "learning_rate": 9.51191749064732e-05, "loss": 0.0401, "step": 7390 }, { "grad_norm": 0.04991975054144859, "learning_rate": 9.51013460844959e-05, "loss": 0.0403, "step": 7400 }, { "grad_norm": 0.07143166661262512, "learning_rate": 9.508348643617382e-05, "loss": 0.0394, "step": 7410 }, { "grad_norm": 0.045176342129707336, "learning_rate": 9.50655959737139e-05, "loss": 0.0399, "step": 7420 }, { "grad_norm": 0.047326236963272095, "learning_rate": 9.504767470934405e-05, "loss": 0.0374, "step": 7430 }, { "grad_norm": 0.046172332018613815, "learning_rate": 9.502972265531332e-05, "loss": 0.0396, "step": 7440 }, { "grad_norm": 0.05206753686070442, "learning_rate": 9.501173982389175e-05, "loss": 0.0416, "step": 7450 }, { "grad_norm": 0.06715720891952515, "learning_rate": 9.499372622737047e-05, "loss": 0.043, "step": 7460 }, { "grad_norm": 0.09169477969408035, "learning_rate": 9.497568187806156e-05, "loss": 0.0417, "step": 7470 }, { "grad_norm": 0.07226298749446869, "learning_rate": 9.495760678829819e-05, "loss": 0.0407, "step": 7480 }, { "grad_norm": 0.06754513084888458, "learning_rate": 9.49395009704345e-05, "loss": 0.0397, "step": 7490 }, { "grad_norm": 0.07376888394355774, "learning_rate": 9.492136443684565e-05, "loss": 0.0395, "step": 7500 }, { "grad_norm": 0.04711291193962097, "learning_rate": 9.49031971999278e-05, "loss": 0.0404, "step": 7510 }, { "grad_norm": 0.05586643144488335, "learning_rate": 9.488499927209806e-05, "loss": 0.0413, "step": 7520 }, { "grad_norm": 0.062248773872852325, "learning_rate": 9.486677066579456e-05, "loss": 0.0418, "step": 7530 }, { "grad_norm": 0.06531981378793716, "learning_rate": 9.484851139347639e-05, "loss": 0.0445, "step": 7540 }, { "grad_norm": 0.06840749830007553, "learning_rate": 9.483022146762358e-05, "loss": 0.0419, "step": 7550 }, { "grad_norm": 0.05689406394958496, "learning_rate": 9.48119009007371e-05, "loss": 0.0443, "step": 7560 }, { "grad_norm": 0.06378185003995895, "learning_rate": 9.479354970533892e-05, "loss": 0.0391, "step": 7570 }, { "grad_norm": 0.03934057056903839, "learning_rate": 9.477516789397189e-05, "loss": 0.0376, "step": 7580 }, { "grad_norm": 0.06111897900700569, "learning_rate": 9.475675547919983e-05, "loss": 0.0426, "step": 7590 }, { "grad_norm": 0.04265384376049042, "learning_rate": 9.47383124736074e-05, "loss": 0.0395, "step": 7600 }, { "grad_norm": 0.04598790779709816, "learning_rate": 9.471983888980028e-05, "loss": 0.0409, "step": 7610 }, { "grad_norm": 0.07745551317930222, "learning_rate": 9.470133474040494e-05, "loss": 0.04, "step": 7620 }, { "grad_norm": 0.04533305764198303, "learning_rate": 9.468280003806882e-05, "loss": 0.0407, "step": 7630 }, { "grad_norm": 0.06443537026643753, "learning_rate": 9.46642347954602e-05, "loss": 0.044, "step": 7640 }, { "grad_norm": 0.059769902378320694, "learning_rate": 9.464563902526826e-05, "loss": 0.0453, "step": 7650 }, { "grad_norm": 0.06554222851991653, "learning_rate": 9.462701274020303e-05, "loss": 0.0411, "step": 7660 }, { "grad_norm": 0.08156628161668777, "learning_rate": 9.46083559529954e-05, "loss": 0.0423, "step": 7670 }, { "grad_norm": 0.06165529042482376, "learning_rate": 9.458966867639712e-05, "loss": 0.0436, "step": 7680 }, { "grad_norm": 0.06754495948553085, "learning_rate": 9.457095092318074e-05, "loss": 0.0412, "step": 7690 }, { "grad_norm": 0.06275323033332825, "learning_rate": 9.455220270613969e-05, "loss": 0.0418, "step": 7700 }, { "grad_norm": 0.06253091990947723, "learning_rate": 9.45334240380882e-05, "loss": 0.0394, "step": 7710 }, { "grad_norm": 0.05126764252781868, "learning_rate": 9.451461493186129e-05, "loss": 0.0383, "step": 7720 }, { "grad_norm": 0.06348167359828949, "learning_rate": 9.449577540031482e-05, "loss": 0.0372, "step": 7730 }, { "grad_norm": 0.04933538660407066, "learning_rate": 9.447690545632544e-05, "loss": 0.0373, "step": 7740 }, { "grad_norm": 0.051454752683639526, "learning_rate": 9.445800511279058e-05, "loss": 0.0387, "step": 7750 }, { "grad_norm": 0.04222039133310318, "learning_rate": 9.443907438262844e-05, "loss": 0.0381, "step": 7760 }, { "grad_norm": 0.05389997735619545, "learning_rate": 9.442011327877798e-05, "loss": 0.0395, "step": 7770 }, { "grad_norm": 0.055471308529376984, "learning_rate": 9.440112181419895e-05, "loss": 0.0407, "step": 7780 }, { "grad_norm": 0.05638686195015907, "learning_rate": 9.438210000187186e-05, "loss": 0.0392, "step": 7790 }, { "grad_norm": 0.05790330842137337, "learning_rate": 9.43630478547979e-05, "loss": 0.0385, "step": 7800 }, { "grad_norm": 0.04334563389420509, "learning_rate": 9.434396538599906e-05, "loss": 0.0375, "step": 7810 }, { "grad_norm": 0.05975886434316635, "learning_rate": 9.4324852608518e-05, "loss": 0.0371, "step": 7820 }, { "grad_norm": 0.04915749654173851, "learning_rate": 9.430570953541816e-05, "loss": 0.0386, "step": 7830 }, { "grad_norm": 0.058906540274620056, "learning_rate": 9.428653617978362e-05, "loss": 0.0355, "step": 7840 }, { "grad_norm": 0.07015081495046616, "learning_rate": 9.42673325547192e-05, "loss": 0.0371, "step": 7850 }, { "grad_norm": 0.048111289739608765, "learning_rate": 9.42480986733504e-05, "loss": 0.0387, "step": 7860 }, { "grad_norm": 0.04888087138533592, "learning_rate": 9.422883454882338e-05, "loss": 0.0379, "step": 7870 }, { "grad_norm": 0.059545863419771194, "learning_rate": 9.4209540194305e-05, "loss": 0.0367, "step": 7880 }, { "grad_norm": 0.04165300726890564, "learning_rate": 9.419021562298278e-05, "loss": 0.038, "step": 7890 }, { "grad_norm": 0.044244252145290375, "learning_rate": 9.417086084806486e-05, "loss": 0.0421, "step": 7900 }, { "grad_norm": 0.0473087914288044, "learning_rate": 9.415147588278005e-05, "loss": 0.0415, "step": 7910 }, { "grad_norm": 0.07260634750127792, "learning_rate": 9.413206074037781e-05, "loss": 0.0409, "step": 7920 }, { "grad_norm": 0.06117885187268257, "learning_rate": 9.411261543412819e-05, "loss": 0.0422, "step": 7930 }, { "grad_norm": 0.044609542936086655, "learning_rate": 9.409313997732188e-05, "loss": 0.0393, "step": 7940 }, { "grad_norm": 0.06274840235710144, "learning_rate": 9.407363438327018e-05, "loss": 0.0412, "step": 7950 }, { "grad_norm": 0.08084600418806076, "learning_rate": 9.405409866530496e-05, "loss": 0.0394, "step": 7960 }, { "grad_norm": 0.06272754073143005, "learning_rate": 9.40345328367787e-05, "loss": 0.0376, "step": 7970 }, { "grad_norm": 0.06357035785913467, "learning_rate": 9.401493691106446e-05, "loss": 0.0384, "step": 7980 }, { "grad_norm": 0.060600873082876205, "learning_rate": 9.39953109015559e-05, "loss": 0.0403, "step": 7990 }, { "grad_norm": 0.04286617785692215, "learning_rate": 9.397565482166718e-05, "loss": 0.0388, "step": 8000 }, { "grad_norm": 0.03872271627187729, "learning_rate": 9.395596868483305e-05, "loss": 0.0379, "step": 8010 }, { "grad_norm": 0.06427207589149475, "learning_rate": 9.393625250450881e-05, "loss": 0.0386, "step": 8020 }, { "grad_norm": 0.04728882759809494, "learning_rate": 9.391650629417028e-05, "loss": 0.0401, "step": 8030 }, { "grad_norm": 0.04885478690266609, "learning_rate": 9.38967300673138e-05, "loss": 0.0413, "step": 8040 }, { "grad_norm": 0.06843569874763489, "learning_rate": 9.387692383745624e-05, "loss": 0.0427, "step": 8050 }, { "grad_norm": 0.06267929077148438, "learning_rate": 9.385708761813496e-05, "loss": 0.0411, "step": 8060 }, { "grad_norm": 0.04844143986701965, "learning_rate": 9.383722142290787e-05, "loss": 0.0406, "step": 8070 }, { "grad_norm": 0.045255210250616074, "learning_rate": 9.381732526535328e-05, "loss": 0.0416, "step": 8080 }, { "grad_norm": 0.061346497386693954, "learning_rate": 9.379739915907004e-05, "loss": 0.0408, "step": 8090 }, { "grad_norm": 0.05477481707930565, "learning_rate": 9.377744311767746e-05, "loss": 0.0425, "step": 8100 }, { "grad_norm": 0.0695875883102417, "learning_rate": 9.375745715481532e-05, "loss": 0.0409, "step": 8110 }, { "grad_norm": 0.05136631801724434, "learning_rate": 9.373744128414382e-05, "loss": 0.0387, "step": 8120 }, { "grad_norm": 0.05347278341650963, "learning_rate": 9.371739551934359e-05, "loss": 0.0412, "step": 8130 }, { "grad_norm": 0.044249873608350754, "learning_rate": 9.36973198741158e-05, "loss": 0.04, "step": 8140 }, { "grad_norm": 0.05221820995211601, "learning_rate": 9.367721436218189e-05, "loss": 0.0387, "step": 8150 }, { "grad_norm": 0.05047016590833664, "learning_rate": 9.365707899728383e-05, "loss": 0.0376, "step": 8160 }, { "grad_norm": 0.06902103126049042, "learning_rate": 9.363691379318391e-05, "loss": 0.0374, "step": 8170 }, { "grad_norm": 0.0509289912879467, "learning_rate": 9.361671876366491e-05, "loss": 0.0389, "step": 8180 }, { "grad_norm": 0.07417299598455429, "learning_rate": 9.35964939225299e-05, "loss": 0.036, "step": 8190 }, { "grad_norm": 0.08487609773874283, "learning_rate": 9.35762392836024e-05, "loss": 0.0374, "step": 8200 }, { "grad_norm": 0.04777513071894646, "learning_rate": 9.355595486072624e-05, "loss": 0.0366, "step": 8210 }, { "grad_norm": 0.0662907138466835, "learning_rate": 9.353564066776563e-05, "loss": 0.0398, "step": 8220 }, { "grad_norm": 0.04366435110569, "learning_rate": 9.351529671860516e-05, "loss": 0.0383, "step": 8230 }, { "grad_norm": 0.055573806166648865, "learning_rate": 9.349492302714969e-05, "loss": 0.0396, "step": 8240 }, { "grad_norm": 0.03989576920866966, "learning_rate": 9.347451960732447e-05, "loss": 0.0382, "step": 8250 }, { "grad_norm": 0.05438682436943054, "learning_rate": 9.345408647307506e-05, "loss": 0.0365, "step": 8260 }, { "grad_norm": 0.043787237256765366, "learning_rate": 9.343362363836726e-05, "loss": 0.0362, "step": 8270 }, { "grad_norm": 0.043600503355264664, "learning_rate": 9.341313111718728e-05, "loss": 0.0403, "step": 8280 }, { "grad_norm": 0.04322637617588043, "learning_rate": 9.339260892354153e-05, "loss": 0.0418, "step": 8290 }, { "grad_norm": 0.06241384148597717, "learning_rate": 9.337205707145676e-05, "loss": 0.0434, "step": 8300 }, { "grad_norm": 0.056645724922418594, "learning_rate": 9.335147557497994e-05, "loss": 0.0395, "step": 8310 }, { "grad_norm": 0.06313703209161758, "learning_rate": 9.333086444817835e-05, "loss": 0.0424, "step": 8320 }, { "grad_norm": 0.07600168883800507, "learning_rate": 9.331022370513951e-05, "loss": 0.0413, "step": 8330 }, { "grad_norm": 0.043254610151052475, "learning_rate": 9.328955335997113e-05, "loss": 0.041, "step": 8340 }, { "grad_norm": 0.05002123489975929, "learning_rate": 9.326885342680125e-05, "loss": 0.0429, "step": 8350 }, { "grad_norm": 0.07208503037691116, "learning_rate": 9.324812391977806e-05, "loss": 0.0411, "step": 8360 }, { "grad_norm": 0.06440041959285736, "learning_rate": 9.322736485306995e-05, "loss": 0.0384, "step": 8370 }, { "grad_norm": 0.05849490314722061, "learning_rate": 9.32065762408656e-05, "loss": 0.0416, "step": 8380 }, { "grad_norm": 0.046950869262218475, "learning_rate": 9.318575809737377e-05, "loss": 0.0397, "step": 8390 }, { "grad_norm": 0.0626986175775528, "learning_rate": 9.316491043682351e-05, "loss": 0.0419, "step": 8400 }, { "grad_norm": 0.054217901080846786, "learning_rate": 9.3144033273464e-05, "loss": 0.041, "step": 8410 }, { "grad_norm": 0.05510523170232773, "learning_rate": 9.312312662156457e-05, "loss": 0.042, "step": 8420 }, { "grad_norm": 0.05154014378786087, "learning_rate": 9.310219049541471e-05, "loss": 0.0419, "step": 8430 }, { "grad_norm": 0.06881421059370041, "learning_rate": 9.308122490932409e-05, "loss": 0.042, "step": 8440 }, { "grad_norm": 0.043774981051683426, "learning_rate": 9.306022987762246e-05, "loss": 0.0414, "step": 8450 }, { "grad_norm": 0.06736815720796585, "learning_rate": 9.303920541465974e-05, "loss": 0.0404, "step": 8460 }, { "grad_norm": 0.054560497403144836, "learning_rate": 9.301815153480598e-05, "loss": 0.0368, "step": 8470 }, { "grad_norm": 0.05441390722990036, "learning_rate": 9.299706825245126e-05, "loss": 0.0362, "step": 8480 }, { "grad_norm": 0.06384050101041794, "learning_rate": 9.297595558200583e-05, "loss": 0.0391, "step": 8490 }, { "grad_norm": 0.041091643273830414, "learning_rate": 9.295481353790001e-05, "loss": 0.0399, "step": 8500 }, { "grad_norm": 0.05301794782280922, "learning_rate": 9.293364213458417e-05, "loss": 0.0394, "step": 8510 }, { "grad_norm": 0.05124790593981743, "learning_rate": 9.291244138652877e-05, "loss": 0.037, "step": 8520 }, { "grad_norm": 0.04915158823132515, "learning_rate": 9.289121130822431e-05, "loss": 0.0406, "step": 8530 }, { "grad_norm": 0.05717909336090088, "learning_rate": 9.286995191418137e-05, "loss": 0.0392, "step": 8540 }, { "grad_norm": 0.07727601379156113, "learning_rate": 9.284866321893056e-05, "loss": 0.0402, "step": 8550 }, { "grad_norm": 0.06272053718566895, "learning_rate": 9.282734523702244e-05, "loss": 0.0412, "step": 8560 }, { "grad_norm": 0.043947577476501465, "learning_rate": 9.280599798302771e-05, "loss": 0.0412, "step": 8570 }, { "grad_norm": 0.08038321137428284, "learning_rate": 9.278462147153699e-05, "loss": 0.0393, "step": 8580 }, { "grad_norm": 0.04287051782011986, "learning_rate": 9.276321571716094e-05, "loss": 0.0409, "step": 8590 }, { "grad_norm": 0.04767925292253494, "learning_rate": 9.274178073453017e-05, "loss": 0.0397, "step": 8600 }, { "grad_norm": 0.054754696786403656, "learning_rate": 9.27203165382953e-05, "loss": 0.0403, "step": 8610 }, { "grad_norm": 0.06008647754788399, "learning_rate": 9.26988231431269e-05, "loss": 0.0405, "step": 8620 }, { "grad_norm": 0.05816996097564697, "learning_rate": 9.267730056371551e-05, "loss": 0.039, "step": 8630 }, { "grad_norm": 0.06311940401792526, "learning_rate": 9.265574881477161e-05, "loss": 0.0388, "step": 8640 }, { "grad_norm": 0.07520703971385956, "learning_rate": 9.263416791102561e-05, "loss": 0.0385, "step": 8650 }, { "grad_norm": 0.06654219329357147, "learning_rate": 9.261255786722786e-05, "loss": 0.0398, "step": 8660 }, { "grad_norm": 0.06703071296215057, "learning_rate": 9.259091869814864e-05, "loss": 0.0384, "step": 8670 }, { "grad_norm": 0.06920629739761353, "learning_rate": 9.256925041857807e-05, "loss": 0.0385, "step": 8680 }, { "grad_norm": 0.04542424529790878, "learning_rate": 9.254755304332626e-05, "loss": 0.0387, "step": 8690 }, { "grad_norm": 0.06183395907282829, "learning_rate": 9.252582658722313e-05, "loss": 0.0406, "step": 8700 }, { "grad_norm": 0.03966282680630684, "learning_rate": 9.250407106511853e-05, "loss": 0.0354, "step": 8710 }, { "grad_norm": 0.04913932830095291, "learning_rate": 9.248228649188215e-05, "loss": 0.0367, "step": 8720 }, { "grad_norm": 0.07163398712873459, "learning_rate": 9.246047288240354e-05, "loss": 0.0411, "step": 8730 }, { "grad_norm": 0.058525871485471725, "learning_rate": 9.243863025159208e-05, "loss": 0.0431, "step": 8740 }, { "grad_norm": 0.04942113161087036, "learning_rate": 9.241675861437704e-05, "loss": 0.0426, "step": 8750 }, { "grad_norm": 0.05471467599272728, "learning_rate": 9.239485798570742e-05, "loss": 0.0448, "step": 8760 }, { "grad_norm": 0.05450363829731941, "learning_rate": 9.237292838055215e-05, "loss": 0.0434, "step": 8770 }, { "grad_norm": 0.0416460856795311, "learning_rate": 9.235096981389988e-05, "loss": 0.0406, "step": 8780 }, { "grad_norm": 0.054874394088983536, "learning_rate": 9.232898230075909e-05, "loss": 0.0377, "step": 8790 }, { "grad_norm": 0.06133008003234863, "learning_rate": 9.230696585615801e-05, "loss": 0.0375, "step": 8800 }, { "grad_norm": 0.05541447177529335, "learning_rate": 9.228492049514473e-05, "loss": 0.035, "step": 8810 }, { "grad_norm": 0.045029833912849426, "learning_rate": 9.2262846232787e-05, "loss": 0.0371, "step": 8820 }, { "grad_norm": 0.06054483726620674, "learning_rate": 9.224074308417238e-05, "loss": 0.0366, "step": 8830 }, { "grad_norm": 0.0593223012983799, "learning_rate": 9.221861106440818e-05, "loss": 0.0358, "step": 8840 }, { "grad_norm": 0.0635245069861412, "learning_rate": 9.219645018862142e-05, "loss": 0.0398, "step": 8850 }, { "grad_norm": 0.05406925082206726, "learning_rate": 9.217426047195882e-05, "loss": 0.0376, "step": 8860 }, { "grad_norm": 0.06762140989303589, "learning_rate": 9.21520419295869e-05, "loss": 0.0383, "step": 8870 }, { "grad_norm": 0.05464353412389755, "learning_rate": 9.212979457669178e-05, "loss": 0.0388, "step": 8880 }, { "grad_norm": 0.05796137824654579, "learning_rate": 9.210751842847931e-05, "loss": 0.0374, "step": 8890 }, { "grad_norm": 0.050767842680215836, "learning_rate": 9.208521350017506e-05, "loss": 0.0374, "step": 8900 }, { "grad_norm": 0.04878038167953491, "learning_rate": 9.206287980702422e-05, "loss": 0.0378, "step": 8910 }, { "grad_norm": 0.05019514635205269, "learning_rate": 9.204051736429166e-05, "loss": 0.0398, "step": 8920 }, { "grad_norm": 0.09345724433660507, "learning_rate": 9.20181261872619e-05, "loss": 0.0415, "step": 8930 }, { "grad_norm": 0.059940923005342484, "learning_rate": 9.19957062912391e-05, "loss": 0.0402, "step": 8940 }, { "grad_norm": 0.09159446507692337, "learning_rate": 9.197325769154704e-05, "loss": 0.0405, "step": 8950 }, { "grad_norm": 0.05890351161360741, "learning_rate": 9.195078040352914e-05, "loss": 0.0413, "step": 8960 }, { "grad_norm": 0.053348518908023834, "learning_rate": 9.192827444254839e-05, "loss": 0.04, "step": 8970 }, { "grad_norm": 0.05866429954767227, "learning_rate": 9.190573982398744e-05, "loss": 0.0397, "step": 8980 }, { "grad_norm": 0.05330507457256317, "learning_rate": 9.188317656324845e-05, "loss": 0.0385, "step": 8990 }, { "grad_norm": 0.046320583671331406, "learning_rate": 9.186058467575322e-05, "loss": 0.038, "step": 9000 }, { "grad_norm": 0.07563944160938263, "learning_rate": 9.183796417694309e-05, "loss": 0.0388, "step": 9010 }, { "grad_norm": 0.05121661722660065, "learning_rate": 9.181531508227897e-05, "loss": 0.038, "step": 9020 }, { "grad_norm": 0.04112536832690239, "learning_rate": 9.179263740724125e-05, "loss": 0.036, "step": 9030 }, { "grad_norm": 0.06453230232000351, "learning_rate": 9.176993116732996e-05, "loss": 0.0389, "step": 9040 }, { "grad_norm": 0.06475291401147842, "learning_rate": 9.17471963780646e-05, "loss": 0.0368, "step": 9050 }, { "grad_norm": 0.05088533088564873, "learning_rate": 9.172443305498414e-05, "loss": 0.0373, "step": 9060 }, { "grad_norm": 0.06679406017065048, "learning_rate": 9.170164121364714e-05, "loss": 0.0399, "step": 9070 }, { "grad_norm": 0.06416536122560501, "learning_rate": 9.167882086963158e-05, "loss": 0.0379, "step": 9080 }, { "grad_norm": 0.09919245541095734, "learning_rate": 9.165597203853499e-05, "loss": 0.0392, "step": 9090 }, { "grad_norm": 0.055927738547325134, "learning_rate": 9.163309473597428e-05, "loss": 0.0367, "step": 9100 }, { "grad_norm": 0.06769727170467377, "learning_rate": 9.161018897758591e-05, "loss": 0.0386, "step": 9110 }, { "grad_norm": 0.06950589269399643, "learning_rate": 9.158725477902572e-05, "loss": 0.0378, "step": 9120 }, { "grad_norm": 0.04059762507677078, "learning_rate": 9.156429215596905e-05, "loss": 0.0375, "step": 9130 }, { "grad_norm": 0.06032635271549225, "learning_rate": 9.154130112411061e-05, "loss": 0.0375, "step": 9140 }, { "grad_norm": 0.06732000410556793, "learning_rate": 9.151828169916455e-05, "loss": 0.0387, "step": 9150 }, { "grad_norm": 0.053817570209503174, "learning_rate": 9.149523389686445e-05, "loss": 0.038, "step": 9160 }, { "grad_norm": 0.04635284096002579, "learning_rate": 9.147215773296327e-05, "loss": 0.042, "step": 9170 }, { "grad_norm": 0.047133274376392365, "learning_rate": 9.144905322323334e-05, "loss": 0.0395, "step": 9180 }, { "grad_norm": 0.06608010828495026, "learning_rate": 9.142592038346635e-05, "loss": 0.0428, "step": 9190 }, { "grad_norm": 0.057442713528871536, "learning_rate": 9.140275922947343e-05, "loss": 0.0408, "step": 9200 }, { "grad_norm": 0.056262172758579254, "learning_rate": 9.137956977708496e-05, "loss": 0.0391, "step": 9210 }, { "grad_norm": 0.06102455407381058, "learning_rate": 9.135635204215075e-05, "loss": 0.0374, "step": 9220 }, { "grad_norm": 0.04534199833869934, "learning_rate": 9.133310604053986e-05, "loss": 0.0417, "step": 9230 }, { "grad_norm": 0.044949207454919815, "learning_rate": 9.130983178814077e-05, "loss": 0.0397, "step": 9240 }, { "grad_norm": 0.05142243951559067, "learning_rate": 9.128652930086116e-05, "loss": 0.0432, "step": 9250 }, { "grad_norm": 0.03713332861661911, "learning_rate": 9.126319859462808e-05, "loss": 0.0393, "step": 9260 }, { "grad_norm": 0.06191578507423401, "learning_rate": 9.123983968538787e-05, "loss": 0.0408, "step": 9270 }, { "grad_norm": 0.08527839183807373, "learning_rate": 9.121645258910609e-05, "loss": 0.0385, "step": 9280 }, { "grad_norm": 0.04205748066306114, "learning_rate": 9.119303732176762e-05, "loss": 0.0407, "step": 9290 }, { "grad_norm": 0.06330464780330658, "learning_rate": 9.116959389937656e-05, "loss": 0.0426, "step": 9300 }, { "grad_norm": 0.045106906443834305, "learning_rate": 9.114612233795628e-05, "loss": 0.04, "step": 9310 }, { "grad_norm": 0.036323241889476776, "learning_rate": 9.112262265354937e-05, "loss": 0.0395, "step": 9320 }, { "grad_norm": 0.04824037849903107, "learning_rate": 9.109909486221763e-05, "loss": 0.0379, "step": 9330 }, { "grad_norm": 0.04820859059691429, "learning_rate": 9.107553898004208e-05, "loss": 0.0392, "step": 9340 }, { "grad_norm": 0.04748307168483734, "learning_rate": 9.105195502312298e-05, "loss": 0.0417, "step": 9350 }, { "grad_norm": 0.06050744280219078, "learning_rate": 9.102834300757972e-05, "loss": 0.0381, "step": 9360 }, { "grad_norm": 0.060097601264715195, "learning_rate": 9.100470294955087e-05, "loss": 0.0383, "step": 9370 }, { "grad_norm": 0.0486888624727726, "learning_rate": 9.098103486519423e-05, "loss": 0.0368, "step": 9380 }, { "grad_norm": 0.053843408823013306, "learning_rate": 9.095733877068666e-05, "loss": 0.036, "step": 9390 }, { "grad_norm": 0.06132175400853157, "learning_rate": 9.093361468222426e-05, "loss": 0.0376, "step": 9400 }, { "grad_norm": 0.05222520977258682, "learning_rate": 9.090986261602221e-05, "loss": 0.0376, "step": 9410 }, { "grad_norm": 0.04458325728774071, "learning_rate": 9.088608258831482e-05, "loss": 0.0388, "step": 9420 }, { "grad_norm": 0.05552945286035538, "learning_rate": 9.08622746153555e-05, "loss": 0.0414, "step": 9430 }, { "grad_norm": 0.049345944076776505, "learning_rate": 9.083843871341679e-05, "loss": 0.0419, "step": 9440 }, { "grad_norm": 0.07745874673128128, "learning_rate": 9.08145748987903e-05, "loss": 0.0429, "step": 9450 }, { "grad_norm": 0.0682804137468338, "learning_rate": 9.079068318778674e-05, "loss": 0.0419, "step": 9460 }, { "grad_norm": 0.047568414360284805, "learning_rate": 9.076676359673586e-05, "loss": 0.0393, "step": 9470 }, { "grad_norm": 0.07369004935026169, "learning_rate": 9.074281614198645e-05, "loss": 0.0417, "step": 9480 }, { "grad_norm": 0.07178745418787003, "learning_rate": 9.071884083990638e-05, "loss": 0.0424, "step": 9490 }, { "grad_norm": 0.06350478529930115, "learning_rate": 9.069483770688259e-05, "loss": 0.0401, "step": 9500 }, { "grad_norm": 0.051732711493968964, "learning_rate": 9.067080675932093e-05, "loss": 0.0394, "step": 9510 }, { "grad_norm": 0.07794926315546036, "learning_rate": 9.064674801364634e-05, "loss": 0.0421, "step": 9520 }, { "grad_norm": 0.05669085308909416, "learning_rate": 9.062266148630279e-05, "loss": 0.0425, "step": 9530 }, { "grad_norm": 0.050370872020721436, "learning_rate": 9.059854719375313e-05, "loss": 0.0387, "step": 9540 }, { "grad_norm": 0.048872530460357666, "learning_rate": 9.057440515247932e-05, "loss": 0.0404, "step": 9550 }, { "grad_norm": 0.062426675111055374, "learning_rate": 9.055023537898215e-05, "loss": 0.043, "step": 9560 }, { "grad_norm": 0.06195325404405594, "learning_rate": 9.052603788978149e-05, "loss": 0.0415, "step": 9570 }, { "grad_norm": 0.04443066939711571, "learning_rate": 9.050181270141606e-05, "loss": 0.0402, "step": 9580 }, { "grad_norm": 0.042344946414232254, "learning_rate": 9.047755983044355e-05, "loss": 0.0389, "step": 9590 }, { "grad_norm": 0.04277326539158821, "learning_rate": 9.045327929344058e-05, "loss": 0.0416, "step": 9600 }, { "grad_norm": 0.0664195567369461, "learning_rate": 9.042897110700268e-05, "loss": 0.0403, "step": 9610 }, { "grad_norm": 0.0511384978890419, "learning_rate": 9.040463528774423e-05, "loss": 0.0429, "step": 9620 }, { "grad_norm": 0.06271642446517944, "learning_rate": 9.038027185229856e-05, "loss": 0.0419, "step": 9630 }, { "grad_norm": 0.06137477979063988, "learning_rate": 9.035588081731784e-05, "loss": 0.0398, "step": 9640 }, { "grad_norm": 0.04705351963639259, "learning_rate": 9.033146219947313e-05, "loss": 0.0392, "step": 9650 }, { "grad_norm": 0.10517262667417526, "learning_rate": 9.030701601545429e-05, "loss": 0.0392, "step": 9660 }, { "grad_norm": 0.052568841725587845, "learning_rate": 9.028254228197009e-05, "loss": 0.0384, "step": 9670 }, { "grad_norm": 0.05445161834359169, "learning_rate": 9.025804101574808e-05, "loss": 0.0382, "step": 9680 }, { "grad_norm": 0.04792618006467819, "learning_rate": 9.023351223353463e-05, "loss": 0.0405, "step": 9690 }, { "grad_norm": 0.05285375937819481, "learning_rate": 9.020895595209498e-05, "loss": 0.0399, "step": 9700 }, { "grad_norm": 0.04090874269604683, "learning_rate": 9.018437218821305e-05, "loss": 0.0397, "step": 9710 }, { "grad_norm": 0.054691772907972336, "learning_rate": 9.015976095869166e-05, "loss": 0.0405, "step": 9720 }, { "grad_norm": 0.043546997010707855, "learning_rate": 9.013512228035235e-05, "loss": 0.0377, "step": 9730 }, { "grad_norm": 0.05254807323217392, "learning_rate": 9.01104561700354e-05, "loss": 0.0391, "step": 9740 }, { "grad_norm": 0.05659300461411476, "learning_rate": 9.008576264459987e-05, "loss": 0.0467, "step": 9750 }, { "grad_norm": 0.05243639647960663, "learning_rate": 9.006104172092357e-05, "loss": 0.0446, "step": 9760 }, { "grad_norm": 0.04407010227441788, "learning_rate": 9.0036293415903e-05, "loss": 0.0456, "step": 9770 }, { "grad_norm": 0.05235586315393448, "learning_rate": 9.00115177464534e-05, "loss": 0.0431, "step": 9780 }, { "grad_norm": 0.05640990659594536, "learning_rate": 8.998671472950872e-05, "loss": 0.0437, "step": 9790 }, { "grad_norm": 0.07950346171855927, "learning_rate": 8.996188438202156e-05, "loss": 0.0404, "step": 9800 }, { "grad_norm": 0.07587262243032455, "learning_rate": 8.993702672096324e-05, "loss": 0.0421, "step": 9810 }, { "grad_norm": 0.064178466796875, "learning_rate": 8.991214176332377e-05, "loss": 0.0408, "step": 9820 }, { "grad_norm": 0.049252793192863464, "learning_rate": 8.988722952611171e-05, "loss": 0.0413, "step": 9830 }, { "grad_norm": 0.08961673825979233, "learning_rate": 8.986229002635441e-05, "loss": 0.0415, "step": 9840 }, { "grad_norm": 0.0701470896601677, "learning_rate": 8.983732328109775e-05, "loss": 0.0386, "step": 9850 }, { "grad_norm": 0.04434589669108391, "learning_rate": 8.981232930740627e-05, "loss": 0.0385, "step": 9860 }, { "grad_norm": 0.04785359650850296, "learning_rate": 8.978730812236313e-05, "loss": 0.039, "step": 9870 }, { "grad_norm": 0.058615636080503464, "learning_rate": 8.976225974307004e-05, "loss": 0.0429, "step": 9880 }, { "grad_norm": 0.04750952869653702, "learning_rate": 8.973718418664736e-05, "loss": 0.0397, "step": 9890 }, { "grad_norm": 0.051584817469120026, "learning_rate": 8.971208147023397e-05, "loss": 0.0399, "step": 9900 }, { "grad_norm": 0.05359501764178276, "learning_rate": 8.968695161098739e-05, "loss": 0.039, "step": 9910 }, { "grad_norm": 0.04998241737484932, "learning_rate": 8.96617946260836e-05, "loss": 0.0372, "step": 9920 }, { "grad_norm": 0.0540221706032753, "learning_rate": 8.963661053271716e-05, "loss": 0.0389, "step": 9930 }, { "grad_norm": 0.05627027526497841, "learning_rate": 8.961139934810117e-05, "loss": 0.0367, "step": 9940 }, { "grad_norm": 0.04633694142103195, "learning_rate": 8.958616108946725e-05, "loss": 0.035, "step": 9950 }, { "grad_norm": 0.07373054325580597, "learning_rate": 8.95608957740655e-05, "loss": 0.0372, "step": 9960 }, { "grad_norm": 0.060191527009010315, "learning_rate": 8.953560341916452e-05, "loss": 0.0366, "step": 9970 }, { "grad_norm": 0.06448355317115784, "learning_rate": 8.95102840420514e-05, "loss": 0.0374, "step": 9980 }, { "grad_norm": 0.06108178198337555, "learning_rate": 8.94849376600317e-05, "loss": 0.0367, "step": 9990 }, { "grad_norm": 0.042119234800338745, "learning_rate": 8.945956429042943e-05, "loss": 0.0378, "step": 10000 }, { "grad_norm": 0.045092176645994186, "learning_rate": 8.943416395058705e-05, "loss": 0.0397, "step": 10010 }, { "grad_norm": 0.1114531084895134, "learning_rate": 8.940873665786544e-05, "loss": 0.0394, "step": 10020 }, { "grad_norm": 0.032701313495635986, "learning_rate": 8.938328242964394e-05, "loss": 0.0374, "step": 10030 }, { "grad_norm": 0.044528719037771225, "learning_rate": 8.935780128332026e-05, "loss": 0.0381, "step": 10040 }, { "grad_norm": 0.04450017958879471, "learning_rate": 8.933229323631052e-05, "loss": 0.0391, "step": 10050 }, { "grad_norm": 0.09659075736999512, "learning_rate": 8.930675830604925e-05, "loss": 0.039, "step": 10060 }, { "grad_norm": 0.0598754957318306, "learning_rate": 8.92811965099893e-05, "loss": 0.0393, "step": 10070 }, { "grad_norm": 0.06025009602308273, "learning_rate": 8.925560786560194e-05, "loss": 0.0377, "step": 10080 }, { "grad_norm": 0.05798542872071266, "learning_rate": 8.922999239037677e-05, "loss": 0.0379, "step": 10090 }, { "grad_norm": 0.05852128565311432, "learning_rate": 8.920435010182171e-05, "loss": 0.0379, "step": 10100 }, { "grad_norm": 0.046517547219991684, "learning_rate": 8.917868101746302e-05, "loss": 0.0384, "step": 10110 }, { "grad_norm": 0.08009778708219528, "learning_rate": 8.91529851548453e-05, "loss": 0.0371, "step": 10120 }, { "grad_norm": 0.04047295078635216, "learning_rate": 8.912726253153142e-05, "loss": 0.0385, "step": 10130 }, { "grad_norm": 0.048078276216983795, "learning_rate": 8.910151316510255e-05, "loss": 0.0391, "step": 10140 }, { "grad_norm": 0.06840091198682785, "learning_rate": 8.907573707315813e-05, "loss": 0.0374, "step": 10150 }, { "grad_norm": 0.06716088950634003, "learning_rate": 8.904993427331588e-05, "loss": 0.0375, "step": 10160 }, { "grad_norm": 0.04158724844455719, "learning_rate": 8.902410478321176e-05, "loss": 0.0381, "step": 10170 }, { "grad_norm": 0.04188217595219612, "learning_rate": 8.899824862050002e-05, "loss": 0.0385, "step": 10180 }, { "grad_norm": 0.057216256856918335, "learning_rate": 8.897236580285308e-05, "loss": 0.0401, "step": 10190 }, { "grad_norm": 0.05278220772743225, "learning_rate": 8.894645634796159e-05, "loss": 0.0388, "step": 10200 }, { "grad_norm": 0.047275200486183167, "learning_rate": 8.892052027353444e-05, "loss": 0.0383, "step": 10210 }, { "grad_norm": 0.04938443750143051, "learning_rate": 8.889455759729866e-05, "loss": 0.0376, "step": 10220 }, { "grad_norm": 0.04883711785078049, "learning_rate": 8.886856833699955e-05, "loss": 0.0385, "step": 10230 }, { "grad_norm": 0.04586977884173393, "learning_rate": 8.884255251040046e-05, "loss": 0.0394, "step": 10240 }, { "grad_norm": 0.038767483085393906, "learning_rate": 8.8816510135283e-05, "loss": 0.0361, "step": 10250 }, { "grad_norm": 0.038742225617170334, "learning_rate": 8.879044122944688e-05, "loss": 0.0343, "step": 10260 }, { "grad_norm": 0.05120726674795151, "learning_rate": 8.876434581070996e-05, "loss": 0.0386, "step": 10270 }, { "grad_norm": 0.051839981228113174, "learning_rate": 8.87382238969082e-05, "loss": 0.035, "step": 10280 }, { "grad_norm": 0.053634412586688995, "learning_rate": 8.871207550589568e-05, "loss": 0.0381, "step": 10290 }, { "grad_norm": 0.06382083892822266, "learning_rate": 8.868590065554458e-05, "loss": 0.0364, "step": 10300 }, { "grad_norm": 0.04906061291694641, "learning_rate": 8.865969936374519e-05, "loss": 0.0355, "step": 10310 }, { "grad_norm": 0.0476507730782032, "learning_rate": 8.863347164840581e-05, "loss": 0.0362, "step": 10320 }, { "grad_norm": 0.1304248422384262, "learning_rate": 8.860721752745285e-05, "loss": 0.0348, "step": 10330 }, { "grad_norm": 0.0522446371614933, "learning_rate": 8.858093701883077e-05, "loss": 0.0351, "step": 10340 }, { "grad_norm": 0.049987852573394775, "learning_rate": 8.8554630140502e-05, "loss": 0.0357, "step": 10350 }, { "grad_norm": 0.04844626411795616, "learning_rate": 8.85282969104471e-05, "loss": 0.0352, "step": 10360 }, { "grad_norm": 0.05947459489107132, "learning_rate": 8.850193734666456e-05, "loss": 0.0328, "step": 10370 }, { "grad_norm": 0.05051708221435547, "learning_rate": 8.84755514671709e-05, "loss": 0.0329, "step": 10380 }, { "grad_norm": 0.06332533061504364, "learning_rate": 8.84491392900006e-05, "loss": 0.0337, "step": 10390 }, { "grad_norm": 0.05211794003844261, "learning_rate": 8.842270083320617e-05, "loss": 0.0329, "step": 10400 }, { "grad_norm": 0.051630765199661255, "learning_rate": 8.839623611485801e-05, "loss": 0.0335, "step": 10410 }, { "grad_norm": 0.04328363388776779, "learning_rate": 8.836974515304453e-05, "loss": 0.0338, "step": 10420 }, { "grad_norm": 0.05560000240802765, "learning_rate": 8.834322796587204e-05, "loss": 0.0356, "step": 10430 }, { "grad_norm": 0.08819197118282318, "learning_rate": 8.831668457146478e-05, "loss": 0.0375, "step": 10440 }, { "grad_norm": 0.04797624051570892, "learning_rate": 8.829011498796493e-05, "loss": 0.0393, "step": 10450 }, { "grad_norm": 0.061436135321855545, "learning_rate": 8.826351923353253e-05, "loss": 0.0408, "step": 10460 }, { "grad_norm": 0.05312379449605942, "learning_rate": 8.823689732634555e-05, "loss": 0.0397, "step": 10470 }, { "grad_norm": 0.05271919444203377, "learning_rate": 8.82102492845998e-05, "loss": 0.0383, "step": 10480 }, { "grad_norm": 0.05630538612604141, "learning_rate": 8.818357512650896e-05, "loss": 0.0387, "step": 10490 }, { "grad_norm": 0.05551859363913536, "learning_rate": 8.815687487030458e-05, "loss": 0.0363, "step": 10500 }, { "grad_norm": 0.05256236717104912, "learning_rate": 8.8130148534236e-05, "loss": 0.0409, "step": 10510 }, { "grad_norm": 0.06078658998012543, "learning_rate": 8.810339613657047e-05, "loss": 0.0371, "step": 10520 }, { "grad_norm": 0.059426236897706985, "learning_rate": 8.807661769559295e-05, "loss": 0.0357, "step": 10530 }, { "grad_norm": 0.07844376564025879, "learning_rate": 8.804981322960628e-05, "loss": 0.0361, "step": 10540 }, { "grad_norm": 0.045583609491586685, "learning_rate": 8.802298275693106e-05, "loss": 0.0355, "step": 10550 }, { "grad_norm": 0.06570015102624893, "learning_rate": 8.799612629590568e-05, "loss": 0.035, "step": 10560 }, { "grad_norm": 0.044492729008197784, "learning_rate": 8.796924386488624e-05, "loss": 0.0383, "step": 10570 }, { "grad_norm": 0.057400789111852646, "learning_rate": 8.794233548224666e-05, "loss": 0.0375, "step": 10580 }, { "grad_norm": 0.05922670289874077, "learning_rate": 8.791540116637853e-05, "loss": 0.0363, "step": 10590 }, { "grad_norm": 0.05812140554189682, "learning_rate": 8.788844093569124e-05, "loss": 0.0362, "step": 10600 }, { "grad_norm": 0.044866520911455154, "learning_rate": 8.786145480861184e-05, "loss": 0.0351, "step": 10610 }, { "grad_norm": 0.05705462768673897, "learning_rate": 8.783444280358507e-05, "loss": 0.0374, "step": 10620 }, { "grad_norm": 0.06151163950562477, "learning_rate": 8.780740493907342e-05, "loss": 0.0364, "step": 10630 }, { "grad_norm": 0.07272005826234818, "learning_rate": 8.778034123355698e-05, "loss": 0.0372, "step": 10640 }, { "grad_norm": 0.05748465284705162, "learning_rate": 8.775325170553357e-05, "loss": 0.0382, "step": 10650 }, { "grad_norm": 0.06262219697237015, "learning_rate": 8.77261363735186e-05, "loss": 0.0395, "step": 10660 }, { "grad_norm": 0.06712506711483002, "learning_rate": 8.769899525604517e-05, "loss": 0.0382, "step": 10670 }, { "grad_norm": 0.05869012698531151, "learning_rate": 8.767182837166397e-05, "loss": 0.0379, "step": 10680 }, { "grad_norm": 0.0625663548707962, "learning_rate": 8.764463573894328e-05, "loss": 0.0327, "step": 10690 }, { "grad_norm": 0.07410836219787598, "learning_rate": 8.761741737646902e-05, "loss": 0.0335, "step": 10700 }, { "grad_norm": 0.04662543535232544, "learning_rate": 8.759017330284471e-05, "loss": 0.0358, "step": 10710 }, { "grad_norm": 0.03919145092368126, "learning_rate": 8.756290353669142e-05, "loss": 0.033, "step": 10720 }, { "grad_norm": 0.03967515751719475, "learning_rate": 8.753560809664774e-05, "loss": 0.0326, "step": 10730 }, { "grad_norm": 0.046821992844343185, "learning_rate": 8.750828700136986e-05, "loss": 0.034, "step": 10740 }, { "grad_norm": 0.043455883860588074, "learning_rate": 8.74809402695315e-05, "loss": 0.0349, "step": 10750 }, { "grad_norm": 0.11330480128526688, "learning_rate": 8.745356791982391e-05, "loss": 0.0351, "step": 10760 }, { "grad_norm": 0.13534244894981384, "learning_rate": 8.742616997095578e-05, "loss": 0.0343, "step": 10770 }, { "grad_norm": 0.06456685066223145, "learning_rate": 8.739874644165341e-05, "loss": 0.0355, "step": 10780 }, { "grad_norm": 0.06471704691648483, "learning_rate": 8.737129735066048e-05, "loss": 0.0336, "step": 10790 }, { "grad_norm": 0.0756855383515358, "learning_rate": 8.734382271673821e-05, "loss": 0.0329, "step": 10800 }, { "grad_norm": 0.13801752030849457, "learning_rate": 8.731632255866525e-05, "loss": 0.0329, "step": 10810 }, { "grad_norm": 0.048605646938085556, "learning_rate": 8.728879689523767e-05, "loss": 0.0339, "step": 10820 }, { "grad_norm": 0.05243070796132088, "learning_rate": 8.726124574526905e-05, "loss": 0.0375, "step": 10830 }, { "grad_norm": 0.06208103150129318, "learning_rate": 8.72336691275903e-05, "loss": 0.0379, "step": 10840 }, { "grad_norm": 0.043180227279663086, "learning_rate": 8.720606706104979e-05, "loss": 0.037, "step": 10850 }, { "grad_norm": 0.05216481164097786, "learning_rate": 8.71784395645133e-05, "loss": 0.0372, "step": 10860 }, { "grad_norm": 0.055575743317604065, "learning_rate": 8.715078665686392e-05, "loss": 0.038, "step": 10870 }, { "grad_norm": 0.05759371444582939, "learning_rate": 8.712310835700218e-05, "loss": 0.0392, "step": 10880 }, { "grad_norm": 0.06242604926228523, "learning_rate": 8.709540468384591e-05, "loss": 0.0371, "step": 10890 }, { "grad_norm": 0.08119571954011917, "learning_rate": 8.706767565633033e-05, "loss": 0.0377, "step": 10900 }, { "grad_norm": 0.05196016654372215, "learning_rate": 8.7039921293408e-05, "loss": 0.0395, "step": 10910 }, { "grad_norm": 0.061991047114133835, "learning_rate": 8.70121416140487e-05, "loss": 0.0389, "step": 10920 }, { "grad_norm": 0.07058275490999222, "learning_rate": 8.698433663723962e-05, "loss": 0.0401, "step": 10930 }, { "grad_norm": 0.07217667251825333, "learning_rate": 8.695650638198518e-05, "loss": 0.0399, "step": 10940 }, { "grad_norm": 0.06405582278966904, "learning_rate": 8.692865086730713e-05, "loss": 0.0386, "step": 10950 }, { "grad_norm": 0.05153418704867363, "learning_rate": 8.69007701122444e-05, "loss": 0.0372, "step": 10960 }, { "grad_norm": 0.04344450309872627, "learning_rate": 8.687286413585328e-05, "loss": 0.0377, "step": 10970 }, { "grad_norm": 0.06471835821866989, "learning_rate": 8.684493295720719e-05, "loss": 0.035, "step": 10980 }, { "grad_norm": 0.048012297600507736, "learning_rate": 8.681697659539685e-05, "loss": 0.0359, "step": 10990 }, { "grad_norm": 0.049812689423561096, "learning_rate": 8.678899506953019e-05, "loss": 0.0362, "step": 11000 }, { "grad_norm": 0.07525008916854858, "learning_rate": 8.676098839873227e-05, "loss": 0.0355, "step": 11010 }, { "grad_norm": 0.05833794176578522, "learning_rate": 8.673295660214545e-05, "loss": 0.0349, "step": 11020 }, { "grad_norm": 0.045270856469869614, "learning_rate": 8.670489969892914e-05, "loss": 0.0339, "step": 11030 }, { "grad_norm": 0.039036791771650314, "learning_rate": 8.667681770826e-05, "loss": 0.036, "step": 11040 }, { "grad_norm": 0.08344912528991699, "learning_rate": 8.66487106493318e-05, "loss": 0.0335, "step": 11050 }, { "grad_norm": 0.05608338862657547, "learning_rate": 8.662057854135544e-05, "loss": 0.0356, "step": 11060 }, { "grad_norm": 0.0466049425303936, "learning_rate": 8.659242140355897e-05, "loss": 0.0342, "step": 11070 }, { "grad_norm": 0.0484757237136364, "learning_rate": 8.65642392551875e-05, "loss": 0.0376, "step": 11080 }, { "grad_norm": 0.05199182406067848, "learning_rate": 8.65360321155033e-05, "loss": 0.0385, "step": 11090 }, { "grad_norm": 0.048331137746572495, "learning_rate": 8.650780000378566e-05, "loss": 0.0362, "step": 11100 }, { "grad_norm": 0.05684540420770645, "learning_rate": 8.647954293933096e-05, "loss": 0.0379, "step": 11110 }, { "grad_norm": 0.046220455318689346, "learning_rate": 8.645126094145264e-05, "loss": 0.0364, "step": 11120 }, { "grad_norm": 0.06671728193759918, "learning_rate": 8.642295402948117e-05, "loss": 0.0368, "step": 11130 }, { "grad_norm": 0.05766962096095085, "learning_rate": 8.639462222276409e-05, "loss": 0.0418, "step": 11140 }, { "grad_norm": 0.06258190423250198, "learning_rate": 8.636626554066589e-05, "loss": 0.0393, "step": 11150 }, { "grad_norm": 0.0445973202586174, "learning_rate": 8.633788400256811e-05, "loss": 0.0385, "step": 11160 }, { "grad_norm": 0.06350675970315933, "learning_rate": 8.630947762786927e-05, "loss": 0.0395, "step": 11170 }, { "grad_norm": 0.043886635452508926, "learning_rate": 8.628104643598483e-05, "loss": 0.0387, "step": 11180 }, { "grad_norm": 0.05372318997979164, "learning_rate": 8.625259044634726e-05, "loss": 0.0371, "step": 11190 }, { "grad_norm": 0.034441135823726654, "learning_rate": 8.622410967840597e-05, "loss": 0.0346, "step": 11200 }, { "grad_norm": 0.03918682783842087, "learning_rate": 8.619560415162731e-05, "loss": 0.0348, "step": 11210 }, { "grad_norm": 0.0404437892138958, "learning_rate": 8.616707388549447e-05, "loss": 0.0369, "step": 11220 }, { "grad_norm": 0.04336835443973541, "learning_rate": 8.613851889950771e-05, "loss": 0.0365, "step": 11230 }, { "grad_norm": 0.0537562295794487, "learning_rate": 8.610993921318402e-05, "loss": 0.0333, "step": 11240 }, { "grad_norm": 0.05779183655977249, "learning_rate": 8.608133484605738e-05, "loss": 0.0345, "step": 11250 }, { "grad_norm": 0.04849625006318092, "learning_rate": 8.605270581767859e-05, "loss": 0.0368, "step": 11260 }, { "grad_norm": 0.07375752925872803, "learning_rate": 8.602405214761536e-05, "loss": 0.0402, "step": 11270 }, { "grad_norm": 0.041092466562986374, "learning_rate": 8.599537385545215e-05, "loss": 0.0375, "step": 11280 }, { "grad_norm": 0.08284611254930496, "learning_rate": 8.596667096079032e-05, "loss": 0.0357, "step": 11290 }, { "grad_norm": 0.04600762203335762, "learning_rate": 8.593794348324806e-05, "loss": 0.0371, "step": 11300 }, { "grad_norm": 0.047460269182920456, "learning_rate": 8.590919144246028e-05, "loss": 0.0364, "step": 11310 }, { "grad_norm": 0.055617768317461014, "learning_rate": 8.588041485807876e-05, "loss": 0.0366, "step": 11320 }, { "grad_norm": 0.06011215224862099, "learning_rate": 8.585161374977202e-05, "loss": 0.0369, "step": 11330 }, { "grad_norm": 0.042358968406915665, "learning_rate": 8.582278813722533e-05, "loss": 0.0389, "step": 11340 }, { "grad_norm": 0.06373909115791321, "learning_rate": 8.579393804014076e-05, "loss": 0.0387, "step": 11350 }, { "grad_norm": 0.08846268057823181, "learning_rate": 8.576506347823703e-05, "loss": 0.0396, "step": 11360 }, { "grad_norm": 0.04236528277397156, "learning_rate": 8.573616447124968e-05, "loss": 0.0365, "step": 11370 }, { "grad_norm": 0.03539847210049629, "learning_rate": 8.570724103893086e-05, "loss": 0.0369, "step": 11380 }, { "grad_norm": 0.038362547755241394, "learning_rate": 8.567829320104951e-05, "loss": 0.036, "step": 11390 }, { "grad_norm": 0.07048764079809189, "learning_rate": 8.564932097739118e-05, "loss": 0.0354, "step": 11400 }, { "grad_norm": 0.05029568821191788, "learning_rate": 8.562032438775811e-05, "loss": 0.0356, "step": 11410 }, { "grad_norm": 0.05944446101784706, "learning_rate": 8.559130345196921e-05, "loss": 0.0344, "step": 11420 }, { "grad_norm": 0.06771279126405716, "learning_rate": 8.556225818986e-05, "loss": 0.0358, "step": 11430 }, { "grad_norm": 0.03917412459850311, "learning_rate": 8.553318862128265e-05, "loss": 0.0363, "step": 11440 }, { "grad_norm": 0.049988754093647, "learning_rate": 8.550409476610593e-05, "loss": 0.0364, "step": 11450 }, { "grad_norm": 0.05927812308073044, "learning_rate": 8.547497664421522e-05, "loss": 0.0378, "step": 11460 }, { "grad_norm": 0.05334849655628204, "learning_rate": 8.544583427551249e-05, "loss": 0.0392, "step": 11470 }, { "grad_norm": 0.04779304563999176, "learning_rate": 8.541666767991628e-05, "loss": 0.0353, "step": 11480 }, { "grad_norm": 0.044391799718141556, "learning_rate": 8.538747687736166e-05, "loss": 0.0364, "step": 11490 }, { "grad_norm": 0.05639626458287239, "learning_rate": 8.535826188780026e-05, "loss": 0.0367, "step": 11500 }, { "grad_norm": 0.055259570479393005, "learning_rate": 8.532902273120029e-05, "loss": 0.0369, "step": 11510 }, { "grad_norm": 0.05825917422771454, "learning_rate": 8.52997594275464e-05, "loss": 0.0371, "step": 11520 }, { "grad_norm": 0.04320257157087326, "learning_rate": 8.52704719968398e-05, "loss": 0.0347, "step": 11530 }, { "grad_norm": 0.05502011626958847, "learning_rate": 8.524116045909818e-05, "loss": 0.0341, "step": 11540 }, { "grad_norm": 0.05278504267334938, "learning_rate": 8.521182483435569e-05, "loss": 0.0343, "step": 11550 }, { "grad_norm": 0.048927661031484604, "learning_rate": 8.518246514266295e-05, "loss": 0.036, "step": 11560 }, { "grad_norm": 0.061313699930906296, "learning_rate": 8.515308140408703e-05, "loss": 0.0319, "step": 11570 }, { "grad_norm": 0.053450118750333786, "learning_rate": 8.512367363871145e-05, "loss": 0.0359, "step": 11580 }, { "grad_norm": 0.059659089893102646, "learning_rate": 8.509424186663614e-05, "loss": 0.0349, "step": 11590 }, { "grad_norm": 0.0914805680513382, "learning_rate": 8.506478610797743e-05, "loss": 0.0391, "step": 11600 }, { "grad_norm": 0.04321938008069992, "learning_rate": 8.503530638286805e-05, "loss": 0.0373, "step": 11610 }, { "grad_norm": 0.06375961005687714, "learning_rate": 8.500580271145712e-05, "loss": 0.0387, "step": 11620 }, { "grad_norm": 0.04758284240961075, "learning_rate": 8.497627511391014e-05, "loss": 0.0388, "step": 11630 }, { "grad_norm": 0.039600469172000885, "learning_rate": 8.494672361040891e-05, "loss": 0.0383, "step": 11640 }, { "grad_norm": 0.06179540604352951, "learning_rate": 8.491714822115162e-05, "loss": 0.0404, "step": 11650 }, { "grad_norm": 0.05329812318086624, "learning_rate": 8.488754896635277e-05, "loss": 0.038, "step": 11660 }, { "grad_norm": 0.08037851750850677, "learning_rate": 8.485792586624317e-05, "loss": 0.0402, "step": 11670 }, { "grad_norm": 0.04216975346207619, "learning_rate": 8.482827894106993e-05, "loss": 0.0376, "step": 11680 }, { "grad_norm": 0.046181194484233856, "learning_rate": 8.479860821109646e-05, "loss": 0.0424, "step": 11690 }, { "grad_norm": 0.04431464150547981, "learning_rate": 8.476891369660239e-05, "loss": 0.0378, "step": 11700 }, { "grad_norm": 0.04901775345206261, "learning_rate": 8.473919541788366e-05, "loss": 0.0388, "step": 11710 }, { "grad_norm": 0.04901784658432007, "learning_rate": 8.470945339525245e-05, "loss": 0.0357, "step": 11720 }, { "grad_norm": 0.052751392126083374, "learning_rate": 8.467968764903713e-05, "loss": 0.0372, "step": 11730 }, { "grad_norm": 0.044290583580732346, "learning_rate": 8.46498981995823e-05, "loss": 0.037, "step": 11740 }, { "grad_norm": 0.053439851850271225, "learning_rate": 8.462008506724879e-05, "loss": 0.0379, "step": 11750 }, { "grad_norm": 0.08128976076841354, "learning_rate": 8.459024827241359e-05, "loss": 0.0352, "step": 11760 }, { "grad_norm": 0.046286556869745255, "learning_rate": 8.456038783546985e-05, "loss": 0.0372, "step": 11770 }, { "grad_norm": 0.04938691481947899, "learning_rate": 8.453050377682691e-05, "loss": 0.0389, "step": 11780 }, { "grad_norm": 0.06641290336847305, "learning_rate": 8.450059611691026e-05, "loss": 0.0387, "step": 11790 }, { "grad_norm": 0.041061725467443466, "learning_rate": 8.447066487616146e-05, "loss": 0.0383, "step": 11800 }, { "grad_norm": 0.06426893919706345, "learning_rate": 8.444071007503826e-05, "loss": 0.0381, "step": 11810 }, { "grad_norm": 0.05516011267900467, "learning_rate": 8.441073173401449e-05, "loss": 0.0381, "step": 11820 }, { "grad_norm": 0.044592712074518204, "learning_rate": 8.438072987358006e-05, "loss": 0.0408, "step": 11830 }, { "grad_norm": 0.13845007121562958, "learning_rate": 8.435070451424094e-05, "loss": 0.0377, "step": 11840 }, { "grad_norm": 0.043167151510715485, "learning_rate": 8.432065567651919e-05, "loss": 0.0388, "step": 11850 }, { "grad_norm": 0.06404049694538116, "learning_rate": 8.429058338095291e-05, "loss": 0.0404, "step": 11860 }, { "grad_norm": 1.0214132070541382, "learning_rate": 8.426048764809624e-05, "loss": 0.0417, "step": 11870 }, { "grad_norm": 0.04759606346487999, "learning_rate": 8.423036849851932e-05, "loss": 0.0426, "step": 11880 }, { "grad_norm": 0.05512486398220062, "learning_rate": 8.42002259528083e-05, "loss": 0.0406, "step": 11890 }, { "grad_norm": 0.04949362576007843, "learning_rate": 8.417006003156532e-05, "loss": 0.0403, "step": 11900 }, { "grad_norm": 0.055258218199014664, "learning_rate": 8.413987075540852e-05, "loss": 0.0417, "step": 11910 }, { "grad_norm": 0.05953448265790939, "learning_rate": 8.4109658144972e-05, "loss": 0.0404, "step": 11920 }, { "grad_norm": 0.05652623996138573, "learning_rate": 8.407942222090573e-05, "loss": 0.0411, "step": 11930 }, { "grad_norm": 0.0672241598367691, "learning_rate": 8.404916300387576e-05, "loss": 0.0389, "step": 11940 }, { "grad_norm": 0.04321642220020294, "learning_rate": 8.401888051456391e-05, "loss": 0.0393, "step": 11950 }, { "grad_norm": 0.0447978638112545, "learning_rate": 8.398857477366803e-05, "loss": 0.0378, "step": 11960 }, { "grad_norm": 0.041426513344049454, "learning_rate": 8.395824580190178e-05, "loss": 0.0399, "step": 11970 }, { "grad_norm": 0.04042781889438629, "learning_rate": 8.392789361999473e-05, "loss": 0.0404, "step": 11980 }, { "grad_norm": 0.06302151083946228, "learning_rate": 8.38975182486923e-05, "loss": 0.0386, "step": 11990 }, { "grad_norm": 0.05788346752524376, "learning_rate": 8.386711970875581e-05, "loss": 0.0369, "step": 12000 }, { "grad_norm": 0.053305938839912415, "learning_rate": 8.383669802096232e-05, "loss": 0.0386, "step": 12010 }, { "grad_norm": 0.05614674836397171, "learning_rate": 8.38062532061048e-05, "loss": 0.0352, "step": 12020 }, { "grad_norm": 0.042064666748046875, "learning_rate": 8.3775785284992e-05, "loss": 0.033, "step": 12030 }, { "grad_norm": 0.048168111592531204, "learning_rate": 8.374529427844843e-05, "loss": 0.0374, "step": 12040 }, { "grad_norm": 0.05318576470017433, "learning_rate": 8.371478020731442e-05, "loss": 0.0359, "step": 12050 }, { "grad_norm": 0.051744770258665085, "learning_rate": 8.368424309244607e-05, "loss": 0.0345, "step": 12060 }, { "grad_norm": 0.05284181609749794, "learning_rate": 8.365368295471517e-05, "loss": 0.0329, "step": 12070 }, { "grad_norm": 0.0534382127225399, "learning_rate": 8.362309981500931e-05, "loss": 0.0347, "step": 12080 }, { "grad_norm": 0.050884902477264404, "learning_rate": 8.359249369423177e-05, "loss": 0.0359, "step": 12090 }, { "grad_norm": 0.04605560377240181, "learning_rate": 8.356186461330155e-05, "loss": 0.0337, "step": 12100 }, { "grad_norm": 0.046435486525297165, "learning_rate": 8.353121259315334e-05, "loss": 0.0335, "step": 12110 }, { "grad_norm": 0.04735657200217247, "learning_rate": 8.350053765473751e-05, "loss": 0.0314, "step": 12120 }, { "grad_norm": 0.05485489219427109, "learning_rate": 8.346983981902005e-05, "loss": 0.035, "step": 12130 }, { "grad_norm": 0.0344054140150547, "learning_rate": 8.343911910698271e-05, "loss": 0.0314, "step": 12140 }, { "grad_norm": 0.04384484514594078, "learning_rate": 8.340837553962278e-05, "loss": 0.0343, "step": 12150 }, { "grad_norm": 0.04781074449419975, "learning_rate": 8.337760913795316e-05, "loss": 0.0344, "step": 12160 }, { "grad_norm": 0.04516945034265518, "learning_rate": 8.334681992300244e-05, "loss": 0.0345, "step": 12170 }, { "grad_norm": 0.040340106934309006, "learning_rate": 8.331600791581475e-05, "loss": 0.0325, "step": 12180 }, { "grad_norm": 0.050508104264736176, "learning_rate": 8.328517313744978e-05, "loss": 0.0327, "step": 12190 }, { "grad_norm": 0.09817422926425934, "learning_rate": 8.325431560898286e-05, "loss": 0.0327, "step": 12200 }, { "grad_norm": 0.048671264201402664, "learning_rate": 8.322343535150478e-05, "loss": 0.033, "step": 12210 }, { "grad_norm": 0.04891505837440491, "learning_rate": 8.319253238612191e-05, "loss": 0.0348, "step": 12220 }, { "grad_norm": 0.04926146939396858, "learning_rate": 8.316160673395614e-05, "loss": 0.0359, "step": 12230 }, { "grad_norm": 0.05777312442660332, "learning_rate": 8.313065841614487e-05, "loss": 0.0372, "step": 12240 }, { "grad_norm": 0.05313461646437645, "learning_rate": 8.309968745384096e-05, "loss": 0.0361, "step": 12250 }, { "grad_norm": 0.04479417949914932, "learning_rate": 8.306869386821282e-05, "loss": 0.0349, "step": 12260 }, { "grad_norm": 0.03786585107445717, "learning_rate": 8.30376776804442e-05, "loss": 0.0342, "step": 12270 }, { "grad_norm": 0.0740462988615036, "learning_rate": 8.300663891173443e-05, "loss": 0.0341, "step": 12280 }, { "grad_norm": 0.04988993704319, "learning_rate": 8.297557758329822e-05, "loss": 0.0329, "step": 12290 }, { "grad_norm": 0.06713796406984329, "learning_rate": 8.294449371636564e-05, "loss": 0.0323, "step": 12300 }, { "grad_norm": 0.06568218022584915, "learning_rate": 8.291338733218226e-05, "loss": 0.0327, "step": 12310 }, { "grad_norm": 0.06048711761832237, "learning_rate": 8.2882258452009e-05, "loss": 0.0303, "step": 12320 }, { "grad_norm": 0.05186711251735687, "learning_rate": 8.285110709712214e-05, "loss": 0.0327, "step": 12330 }, { "grad_norm": 0.04562196880578995, "learning_rate": 8.281993328881337e-05, "loss": 0.0301, "step": 12340 }, { "grad_norm": 0.0505131334066391, "learning_rate": 8.278873704838964e-05, "loss": 0.0357, "step": 12350 }, { "grad_norm": 0.06191623583436012, "learning_rate": 8.275751839717334e-05, "loss": 0.0364, "step": 12360 }, { "grad_norm": 0.05596005544066429, "learning_rate": 8.272627735650208e-05, "loss": 0.0361, "step": 12370 }, { "grad_norm": 0.07401629537343979, "learning_rate": 8.269501394772884e-05, "loss": 0.0356, "step": 12380 }, { "grad_norm": 0.058737900108098984, "learning_rate": 8.266372819222189e-05, "loss": 0.0363, "step": 12390 }, { "grad_norm": 0.06746254116296768, "learning_rate": 8.26324201113647e-05, "loss": 0.0379, "step": 12400 }, { "grad_norm": 0.04362265393137932, "learning_rate": 8.260108972655606e-05, "loss": 0.0366, "step": 12410 }, { "grad_norm": 0.06087873503565788, "learning_rate": 8.256973705921e-05, "loss": 0.038, "step": 12420 }, { "grad_norm": 0.04629741981625557, "learning_rate": 8.25383621307558e-05, "loss": 0.0396, "step": 12430 }, { "grad_norm": 0.0813210979104042, "learning_rate": 8.25069649626379e-05, "loss": 0.0368, "step": 12440 }, { "grad_norm": 0.04421959072351456, "learning_rate": 8.247554557631596e-05, "loss": 0.0358, "step": 12450 }, { "grad_norm": 0.05333013832569122, "learning_rate": 8.244410399326483e-05, "loss": 0.0396, "step": 12460 }, { "grad_norm": 0.047729212790727615, "learning_rate": 8.241264023497457e-05, "loss": 0.0389, "step": 12470 }, { "grad_norm": 0.0481257401406765, "learning_rate": 8.238115432295034e-05, "loss": 0.0365, "step": 12480 }, { "grad_norm": 0.05106436461210251, "learning_rate": 8.234964627871247e-05, "loss": 0.0374, "step": 12490 }, { "grad_norm": 0.058268167078495026, "learning_rate": 8.231811612379639e-05, "loss": 0.0378, "step": 12500 }, { "grad_norm": 0.044447336345911026, "learning_rate": 8.228656387975268e-05, "loss": 0.035, "step": 12510 }, { "grad_norm": 0.05153888091444969, "learning_rate": 8.225498956814702e-05, "loss": 0.0363, "step": 12520 }, { "grad_norm": 0.038820601999759674, "learning_rate": 8.222339321056014e-05, "loss": 0.0353, "step": 12530 }, { "grad_norm": 0.04761352390050888, "learning_rate": 8.219177482858785e-05, "loss": 0.0371, "step": 12540 }, { "grad_norm": 0.05621815472841263, "learning_rate": 8.216013444384099e-05, "loss": 0.0359, "step": 12550 }, { "grad_norm": 0.03716948628425598, "learning_rate": 8.21284720779455e-05, "loss": 0.0375, "step": 12560 }, { "grad_norm": 0.053172990679740906, "learning_rate": 8.209678775254231e-05, "loss": 0.0372, "step": 12570 }, { "grad_norm": 0.06103495508432388, "learning_rate": 8.206508148928733e-05, "loss": 0.0368, "step": 12580 }, { "grad_norm": 0.051725130528211594, "learning_rate": 8.203335330985151e-05, "loss": 0.0343, "step": 12590 }, { "grad_norm": 0.055110104382038116, "learning_rate": 8.200160323592076e-05, "loss": 0.0354, "step": 12600 }, { "grad_norm": 0.052401356399059296, "learning_rate": 8.196983128919598e-05, "loss": 0.0343, "step": 12610 }, { "grad_norm": 0.04428107663989067, "learning_rate": 8.193803749139295e-05, "loss": 0.0341, "step": 12620 }, { "grad_norm": 0.052776653319597244, "learning_rate": 8.190622186424244e-05, "loss": 0.0362, "step": 12630 }, { "grad_norm": 0.04179810732603073, "learning_rate": 8.187438442949016e-05, "loss": 0.0339, "step": 12640 }, { "grad_norm": 0.0442713238298893, "learning_rate": 8.184252520889668e-05, "loss": 0.034, "step": 12650 }, { "grad_norm": 0.09502606093883514, "learning_rate": 8.181064422423748e-05, "loss": 0.0342, "step": 12660 }, { "grad_norm": 0.04124918580055237, "learning_rate": 8.177874149730289e-05, "loss": 0.0383, "step": 12670 }, { "grad_norm": 0.0729668140411377, "learning_rate": 8.174681704989816e-05, "loss": 0.0393, "step": 12680 }, { "grad_norm": 0.041793495416641235, "learning_rate": 8.171487090384333e-05, "loss": 0.0378, "step": 12690 }, { "grad_norm": 0.058969613164663315, "learning_rate": 8.168290308097328e-05, "loss": 0.0359, "step": 12700 }, { "grad_norm": 0.040421128273010254, "learning_rate": 8.165091360313774e-05, "loss": 0.0391, "step": 12710 }, { "grad_norm": 0.05924595519900322, "learning_rate": 8.161890249220119e-05, "loss": 0.036, "step": 12720 }, { "grad_norm": 0.03720834478735924, "learning_rate": 8.158686977004295e-05, "loss": 0.0363, "step": 12730 }, { "grad_norm": 0.04339161515235901, "learning_rate": 8.155481545855706e-05, "loss": 0.0379, "step": 12740 }, { "grad_norm": 0.041643984615802765, "learning_rate": 8.152273957965233e-05, "loss": 0.0377, "step": 12750 }, { "grad_norm": 0.04478759318590164, "learning_rate": 8.149064215525237e-05, "loss": 0.0374, "step": 12760 }, { "grad_norm": 0.046012021601200104, "learning_rate": 8.14585232072954e-05, "loss": 0.0364, "step": 12770 }, { "grad_norm": 0.062354519963264465, "learning_rate": 8.142638275773449e-05, "loss": 0.037, "step": 12780 }, { "grad_norm": 0.04869149997830391, "learning_rate": 8.139422082853729e-05, "loss": 0.0378, "step": 12790 }, { "grad_norm": 0.054789915680885315, "learning_rate": 8.136203744168618e-05, "loss": 0.0389, "step": 12800 }, { "grad_norm": 0.03844248503446579, "learning_rate": 8.132983261917819e-05, "loss": 0.037, "step": 12810 }, { "grad_norm": 0.04342883080244064, "learning_rate": 8.129760638302504e-05, "loss": 0.0359, "step": 12820 }, { "grad_norm": 0.037586916238069534, "learning_rate": 8.126535875525305e-05, "loss": 0.0383, "step": 12830 }, { "grad_norm": 0.04696829244494438, "learning_rate": 8.123308975790316e-05, "loss": 0.0376, "step": 12840 }, { "grad_norm": 0.04207571595907211, "learning_rate": 8.120079941303094e-05, "loss": 0.0372, "step": 12850 }, { "grad_norm": 0.06149030849337578, "learning_rate": 8.116848774270651e-05, "loss": 0.0376, "step": 12860 }, { "grad_norm": 0.04806893318891525, "learning_rate": 8.113615476901461e-05, "loss": 0.039, "step": 12870 }, { "grad_norm": 0.05857127159833908, "learning_rate": 8.110380051405454e-05, "loss": 0.0415, "step": 12880 }, { "grad_norm": 0.044669009745121, "learning_rate": 8.107142499994009e-05, "loss": 0.0397, "step": 12890 }, { "grad_norm": 0.07480505853891373, "learning_rate": 8.103902824879966e-05, "loss": 0.0408, "step": 12900 }, { "grad_norm": 0.04531063511967659, "learning_rate": 8.10066102827761e-05, "loss": 0.0372, "step": 12910 }, { "grad_norm": 0.08085153996944427, "learning_rate": 8.097417112402676e-05, "loss": 0.0399, "step": 12920 }, { "grad_norm": 0.04619510471820831, "learning_rate": 8.094171079472355e-05, "loss": 0.0365, "step": 12930 }, { "grad_norm": 0.05303749814629555, "learning_rate": 8.090922931705277e-05, "loss": 0.0349, "step": 12940 }, { "grad_norm": 0.053191591054201126, "learning_rate": 8.08767267132152e-05, "loss": 0.0363, "step": 12950 }, { "grad_norm": 0.0509367473423481, "learning_rate": 8.084420300542608e-05, "loss": 0.033, "step": 12960 }, { "grad_norm": 0.05290757864713669, "learning_rate": 8.081165821591505e-05, "loss": 0.037, "step": 12970 }, { "grad_norm": 0.052190691232681274, "learning_rate": 8.077909236692615e-05, "loss": 0.0355, "step": 12980 }, { "grad_norm": 0.04318593442440033, "learning_rate": 8.074650548071787e-05, "loss": 0.0366, "step": 12990 }, { "grad_norm": 0.08325114101171494, "learning_rate": 8.071389757956301e-05, "loss": 0.0351, "step": 13000 }, { "grad_norm": 0.052345626056194305, "learning_rate": 8.068126868574876e-05, "loss": 0.0358, "step": 13010 }, { "grad_norm": 0.06086638942360878, "learning_rate": 8.064861882157668e-05, "loss": 0.0354, "step": 13020 }, { "grad_norm": 0.06398896127939224, "learning_rate": 8.061594800936263e-05, "loss": 0.0351, "step": 13030 }, { "grad_norm": 0.04153592884540558, "learning_rate": 8.058325627143681e-05, "loss": 0.0327, "step": 13040 }, { "grad_norm": 0.08767485618591309, "learning_rate": 8.055054363014372e-05, "loss": 0.0334, "step": 13050 }, { "grad_norm": 0.05668815225362778, "learning_rate": 8.051781010784211e-05, "loss": 0.0368, "step": 13060 }, { "grad_norm": 0.06709006428718567, "learning_rate": 8.048505572690506e-05, "loss": 0.0344, "step": 13070 }, { "grad_norm": 0.05425124987959862, "learning_rate": 8.045228050971988e-05, "loss": 0.0347, "step": 13080 }, { "grad_norm": 0.06727461516857147, "learning_rate": 8.041948447868814e-05, "loss": 0.0343, "step": 13090 }, { "grad_norm": 0.03723764419555664, "learning_rate": 8.038666765622558e-05, "loss": 0.0359, "step": 13100 }, { "grad_norm": 0.05104079470038414, "learning_rate": 8.03538300647622e-05, "loss": 0.0339, "step": 13110 }, { "grad_norm": 0.043139562010765076, "learning_rate": 8.03209717267422e-05, "loss": 0.0342, "step": 13120 }, { "grad_norm": 0.053469132632017136, "learning_rate": 8.028809266462395e-05, "loss": 0.035, "step": 13130 }, { "grad_norm": 0.04256201162934303, "learning_rate": 8.025519290087994e-05, "loss": 0.0355, "step": 13140 }, { "grad_norm": 0.05797137692570686, "learning_rate": 8.022227245799688e-05, "loss": 0.0322, "step": 13150 }, { "grad_norm": 0.041939008980989456, "learning_rate": 8.018933135847557e-05, "loss": 0.0356, "step": 13160 }, { "grad_norm": 0.052844930440187454, "learning_rate": 8.015636962483096e-05, "loss": 0.0383, "step": 13170 }, { "grad_norm": 0.043142709881067276, "learning_rate": 8.012338727959205e-05, "loss": 0.0363, "step": 13180 }, { "grad_norm": 0.058715928345918655, "learning_rate": 8.009038434530198e-05, "loss": 0.0393, "step": 13190 }, { "grad_norm": 0.052421506494283676, "learning_rate": 8.005736084451796e-05, "loss": 0.0405, "step": 13200 }, { "grad_norm": 0.059321362525224686, "learning_rate": 8.002431679981122e-05, "loss": 0.0357, "step": 13210 }, { "grad_norm": 0.05461777001619339, "learning_rate": 7.999125223376706e-05, "loss": 0.038, "step": 13220 }, { "grad_norm": 0.05054302513599396, "learning_rate": 7.99581671689848e-05, "loss": 0.0339, "step": 13230 }, { "grad_norm": 0.04728669300675392, "learning_rate": 7.992506162807775e-05, "loss": 0.037, "step": 13240 }, { "grad_norm": 0.0482584610581398, "learning_rate": 7.989193563367328e-05, "loss": 0.0381, "step": 13250 }, { "grad_norm": 0.05647467449307442, "learning_rate": 7.985878920841266e-05, "loss": 0.0351, "step": 13260 }, { "grad_norm": 0.038493812084198, "learning_rate": 7.982562237495117e-05, "loss": 0.0359, "step": 13270 }, { "grad_norm": 0.060904841870069504, "learning_rate": 7.979243515595802e-05, "loss": 0.0349, "step": 13280 }, { "grad_norm": 0.04400007426738739, "learning_rate": 7.975922757411636e-05, "loss": 0.0381, "step": 13290 }, { "grad_norm": 0.04928687959909439, "learning_rate": 7.972599965212329e-05, "loss": 0.0357, "step": 13300 }, { "grad_norm": 0.05460944399237633, "learning_rate": 7.969275141268973e-05, "loss": 0.0342, "step": 13310 }, { "grad_norm": 0.04895928502082825, "learning_rate": 7.96594828785406e-05, "loss": 0.0375, "step": 13320 }, { "grad_norm": 0.062215108424425125, "learning_rate": 7.962619407241456e-05, "loss": 0.0386, "step": 13330 }, { "grad_norm": 0.03851885721087456, "learning_rate": 7.959288501706424e-05, "loss": 0.0371, "step": 13340 }, { "grad_norm": 0.05229327455163002, "learning_rate": 7.955955573525605e-05, "loss": 0.0395, "step": 13350 }, { "grad_norm": 0.04512520506978035, "learning_rate": 7.952620624977026e-05, "loss": 0.034, "step": 13360 }, { "grad_norm": 0.05465196445584297, "learning_rate": 7.949283658340089e-05, "loss": 0.0403, "step": 13370 }, { "grad_norm": 0.05691705271601677, "learning_rate": 7.945944675895585e-05, "loss": 0.0388, "step": 13380 }, { "grad_norm": 0.0642620325088501, "learning_rate": 7.942603679925671e-05, "loss": 0.0388, "step": 13390 }, { "grad_norm": 0.044296618551015854, "learning_rate": 7.93926067271389e-05, "loss": 0.0415, "step": 13400 }, { "grad_norm": 0.06297260522842407, "learning_rate": 7.935915656545155e-05, "loss": 0.0383, "step": 13410 }, { "grad_norm": 0.05581172928214073, "learning_rate": 7.932568633705752e-05, "loss": 0.038, "step": 13420 }, { "grad_norm": 0.07209091633558273, "learning_rate": 7.929219606483341e-05, "loss": 0.04, "step": 13430 }, { "grad_norm": 0.06993887573480606, "learning_rate": 7.925868577166948e-05, "loss": 0.0371, "step": 13440 }, { "grad_norm": 0.053997114300727844, "learning_rate": 7.922515548046974e-05, "loss": 0.0345, "step": 13450 }, { "grad_norm": 0.05089786276221275, "learning_rate": 7.919160521415179e-05, "loss": 0.0351, "step": 13460 }, { "grad_norm": 0.04238065704703331, "learning_rate": 7.915803499564694e-05, "loss": 0.0329, "step": 13470 }, { "grad_norm": 0.04865788668394089, "learning_rate": 7.912444484790013e-05, "loss": 0.0343, "step": 13480 }, { "grad_norm": 0.043780166655778885, "learning_rate": 7.909083479386987e-05, "loss": 0.0326, "step": 13490 }, { "grad_norm": 0.0532911941409111, "learning_rate": 7.905720485652836e-05, "loss": 0.0361, "step": 13500 }, { "grad_norm": 0.04784686118364334, "learning_rate": 7.902355505886132e-05, "loss": 0.0331, "step": 13510 }, { "grad_norm": 0.04627912864089012, "learning_rate": 7.898988542386805e-05, "loss": 0.036, "step": 13520 }, { "grad_norm": 0.05717354267835617, "learning_rate": 7.895619597456147e-05, "loss": 0.0341, "step": 13530 }, { "grad_norm": 0.04046240821480751, "learning_rate": 7.892248673396798e-05, "loss": 0.0344, "step": 13540 }, { "grad_norm": 0.05884886533021927, "learning_rate": 7.888875772512754e-05, "loss": 0.0331, "step": 13550 }, { "grad_norm": 0.053444743156433105, "learning_rate": 7.885500897109359e-05, "loss": 0.0321, "step": 13560 }, { "grad_norm": 0.04160422459244728, "learning_rate": 7.882124049493309e-05, "loss": 0.0348, "step": 13570 }, { "grad_norm": 0.037411369383335114, "learning_rate": 7.878745231972649e-05, "loss": 0.0337, "step": 13580 }, { "grad_norm": 0.047538094222545624, "learning_rate": 7.875364446856766e-05, "loss": 0.0369, "step": 13590 }, { "grad_norm": 0.05439785495400429, "learning_rate": 7.871981696456398e-05, "loss": 0.0348, "step": 13600 }, { "grad_norm": 0.04257991537451744, "learning_rate": 7.868596983083623e-05, "loss": 0.0374, "step": 13610 }, { "grad_norm": 0.09969007223844528, "learning_rate": 7.865210309051858e-05, "loss": 0.0378, "step": 13620 }, { "grad_norm": 0.04274920001626015, "learning_rate": 7.861821676675863e-05, "loss": 0.0419, "step": 13630 }, { "grad_norm": 0.04924515634775162, "learning_rate": 7.858431088271739e-05, "loss": 0.0408, "step": 13640 }, { "grad_norm": 0.07013842463493347, "learning_rate": 7.855038546156918e-05, "loss": 0.0404, "step": 13650 }, { "grad_norm": 0.08214777708053589, "learning_rate": 7.851644052650173e-05, "loss": 0.0396, "step": 13660 }, { "grad_norm": 0.05567934364080429, "learning_rate": 7.848247610071609e-05, "loss": 0.0377, "step": 13670 }, { "grad_norm": 0.051105063408613205, "learning_rate": 7.844849220742658e-05, "loss": 0.036, "step": 13680 }, { "grad_norm": 0.04662663862109184, "learning_rate": 7.841448886986092e-05, "loss": 0.0358, "step": 13690 }, { "grad_norm": 0.05942528322339058, "learning_rate": 7.838046611126004e-05, "loss": 0.0342, "step": 13700 }, { "grad_norm": 0.05392756685614586, "learning_rate": 7.834642395487819e-05, "loss": 0.0357, "step": 13710 }, { "grad_norm": 0.06365881860256195, "learning_rate": 7.831236242398285e-05, "loss": 0.0357, "step": 13720 }, { "grad_norm": 0.058488838374614716, "learning_rate": 7.827828154185477e-05, "loss": 0.038, "step": 13730 }, { "grad_norm": 0.06348840892314911, "learning_rate": 7.82441813317879e-05, "loss": 0.0338, "step": 13740 }, { "grad_norm": 0.052596285939216614, "learning_rate": 7.821006181708944e-05, "loss": 0.0352, "step": 13750 }, { "grad_norm": 0.04325275868177414, "learning_rate": 7.81759230210797e-05, "loss": 0.0372, "step": 13760 }, { "grad_norm": 0.05522578954696655, "learning_rate": 7.814176496709227e-05, "loss": 0.0348, "step": 13770 }, { "grad_norm": 0.06618708372116089, "learning_rate": 7.810758767847385e-05, "loss": 0.0345, "step": 13780 }, { "grad_norm": 0.04724408686161041, "learning_rate": 7.807339117858427e-05, "loss": 0.034, "step": 13790 }, { "grad_norm": 0.05623432621359825, "learning_rate": 7.803917549079655e-05, "loss": 0.0331, "step": 13800 }, { "grad_norm": 0.05172307416796684, "learning_rate": 7.800494063849679e-05, "loss": 0.0346, "step": 13810 }, { "grad_norm": 0.05508876591920853, "learning_rate": 7.797068664508416e-05, "loss": 0.0363, "step": 13820 }, { "grad_norm": 0.050485920161008835, "learning_rate": 7.793641353397096e-05, "loss": 0.0375, "step": 13830 }, { "grad_norm": 0.0791829377412796, "learning_rate": 7.790212132858253e-05, "loss": 0.0355, "step": 13840 }, { "grad_norm": 0.05350954830646515, "learning_rate": 7.786781005235728e-05, "loss": 0.0351, "step": 13850 }, { "grad_norm": 0.06372074037790298, "learning_rate": 7.783347972874662e-05, "loss": 0.0386, "step": 13860 }, { "grad_norm": 0.06104553863406181, "learning_rate": 7.779913038121504e-05, "loss": 0.0361, "step": 13870 }, { "grad_norm": 0.04869665578007698, "learning_rate": 7.776476203323997e-05, "loss": 0.035, "step": 13880 }, { "grad_norm": 0.047122687101364136, "learning_rate": 7.773037470831185e-05, "loss": 0.0368, "step": 13890 }, { "grad_norm": 0.05105523392558098, "learning_rate": 7.76959684299341e-05, "loss": 0.037, "step": 13900 }, { "grad_norm": 0.045367732644081116, "learning_rate": 7.76615432216231e-05, "loss": 0.036, "step": 13910 }, { "grad_norm": 0.06309439986944199, "learning_rate": 7.762709910690811e-05, "loss": 0.0353, "step": 13920 }, { "grad_norm": 0.03169555217027664, "learning_rate": 7.759263610933141e-05, "loss": 0.0344, "step": 13930 }, { "grad_norm": 0.07900586724281311, "learning_rate": 7.755815425244811e-05, "loss": 0.0371, "step": 13940 }, { "grad_norm": 0.046788327395915985, "learning_rate": 7.752365355982624e-05, "loss": 0.0357, "step": 13950 }, { "grad_norm": 0.054294031113386154, "learning_rate": 7.748913405504668e-05, "loss": 0.0391, "step": 13960 }, { "grad_norm": 0.049666497856378555, "learning_rate": 7.745459576170322e-05, "loss": 0.0346, "step": 13970 }, { "grad_norm": 0.051273006945848465, "learning_rate": 7.742003870340242e-05, "loss": 0.0386, "step": 13980 }, { "grad_norm": 0.04908112436532974, "learning_rate": 7.738546290376373e-05, "loss": 0.0341, "step": 13990 }, { "grad_norm": 0.03764589875936508, "learning_rate": 7.735086838641937e-05, "loss": 0.0362, "step": 14000 }, { "grad_norm": 0.05331382155418396, "learning_rate": 7.731625517501437e-05, "loss": 0.0354, "step": 14010 }, { "grad_norm": 0.052430037409067154, "learning_rate": 7.728162329320655e-05, "loss": 0.0346, "step": 14020 }, { "grad_norm": 0.04947454109787941, "learning_rate": 7.724697276466645e-05, "loss": 0.036, "step": 14030 }, { "grad_norm": 0.05224663391709328, "learning_rate": 7.721230361307738e-05, "loss": 0.0351, "step": 14040 }, { "grad_norm": 0.04277050122618675, "learning_rate": 7.71776158621354e-05, "loss": 0.0355, "step": 14050 }, { "grad_norm": 0.04690251871943474, "learning_rate": 7.714290953554925e-05, "loss": 0.0367, "step": 14060 }, { "grad_norm": 0.039461638778448105, "learning_rate": 7.710818465704037e-05, "loss": 0.0343, "step": 14070 }, { "grad_norm": 0.05983376130461693, "learning_rate": 7.707344125034288e-05, "loss": 0.0338, "step": 14080 }, { "grad_norm": 0.04302019625902176, "learning_rate": 7.703867933920359e-05, "loss": 0.033, "step": 14090 }, { "grad_norm": 0.04077642410993576, "learning_rate": 7.700389894738194e-05, "loss": 0.0309, "step": 14100 }, { "grad_norm": 0.04533116891980171, "learning_rate": 7.696910009864999e-05, "loss": 0.032, "step": 14110 }, { "grad_norm": 0.04808686673641205, "learning_rate": 7.693428281679241e-05, "loss": 0.0311, "step": 14120 }, { "grad_norm": 0.04446684569120407, "learning_rate": 7.689944712560652e-05, "loss": 0.0339, "step": 14130 }, { "grad_norm": 0.04538322612643242, "learning_rate": 7.686459304890214e-05, "loss": 0.0342, "step": 14140 }, { "grad_norm": 0.0458921417593956, "learning_rate": 7.682972061050175e-05, "loss": 0.035, "step": 14150 }, { "grad_norm": 0.0499885268509388, "learning_rate": 7.679482983424032e-05, "loss": 0.0341, "step": 14160 }, { "grad_norm": 0.054920461028814316, "learning_rate": 7.675992074396534e-05, "loss": 0.034, "step": 14170 }, { "grad_norm": 0.04797776788473129, "learning_rate": 7.672499336353687e-05, "loss": 0.0334, "step": 14180 }, { "grad_norm": 0.03674433380365372, "learning_rate": 7.669004771682744e-05, "loss": 0.0335, "step": 14190 }, { "grad_norm": 0.05536230280995369, "learning_rate": 7.665508382772206e-05, "loss": 0.0351, "step": 14200 }, { "grad_norm": 0.04165080934762955, "learning_rate": 7.662010172011824e-05, "loss": 0.0343, "step": 14210 }, { "grad_norm": 0.05277930945158005, "learning_rate": 7.658510141792588e-05, "loss": 0.0335, "step": 14220 }, { "grad_norm": 0.0426388680934906, "learning_rate": 7.65500829450674e-05, "loss": 0.035, "step": 14230 }, { "grad_norm": 0.038530830293893814, "learning_rate": 7.651504632547759e-05, "loss": 0.035, "step": 14240 }, { "grad_norm": 0.03877783194184303, "learning_rate": 7.647999158310364e-05, "loss": 0.0339, "step": 14250 }, { "grad_norm": 0.045243244618177414, "learning_rate": 7.644491874190512e-05, "loss": 0.0316, "step": 14260 }, { "grad_norm": 0.03085385262966156, "learning_rate": 7.6409827825854e-05, "loss": 0.0311, "step": 14270 }, { "grad_norm": 0.0510016493499279, "learning_rate": 7.637471885893459e-05, "loss": 0.0318, "step": 14280 }, { "grad_norm": 0.02971726283431053, "learning_rate": 7.633959186514354e-05, "loss": 0.0346, "step": 14290 }, { "grad_norm": 0.05551518127322197, "learning_rate": 7.630444686848984e-05, "loss": 0.0334, "step": 14300 }, { "grad_norm": 0.03780380263924599, "learning_rate": 7.626928389299471e-05, "loss": 0.0328, "step": 14310 }, { "grad_norm": 0.06141466647386551, "learning_rate": 7.623410296269175e-05, "loss": 0.0354, "step": 14320 }, { "grad_norm": 0.05306153744459152, "learning_rate": 7.61989041016268e-05, "loss": 0.0342, "step": 14330 }, { "grad_norm": 0.04418182745575905, "learning_rate": 7.616368733385793e-05, "loss": 0.036, "step": 14340 }, { "grad_norm": 0.05939626693725586, "learning_rate": 7.612845268345547e-05, "loss": 0.0367, "step": 14350 }, { "grad_norm": 0.05441564694046974, "learning_rate": 7.609320017450199e-05, "loss": 0.0359, "step": 14360 }, { "grad_norm": 0.06269539892673492, "learning_rate": 7.605792983109222e-05, "loss": 0.0345, "step": 14370 }, { "grad_norm": 0.05017825961112976, "learning_rate": 7.602264167733313e-05, "loss": 0.0365, "step": 14380 }, { "grad_norm": 0.04427278786897659, "learning_rate": 7.598733573734384e-05, "loss": 0.0351, "step": 14390 }, { "grad_norm": 0.04943804815411568, "learning_rate": 7.595201203525561e-05, "loss": 0.0342, "step": 14400 }, { "grad_norm": 0.05057021975517273, "learning_rate": 7.591667059521187e-05, "loss": 0.0358, "step": 14410 }, { "grad_norm": 0.052685219794511795, "learning_rate": 7.588131144136815e-05, "loss": 0.0335, "step": 14420 }, { "grad_norm": 0.042872603982686996, "learning_rate": 7.584593459789212e-05, "loss": 0.0358, "step": 14430 }, { "grad_norm": 0.05669751763343811, "learning_rate": 7.58105400889635e-05, "loss": 0.0378, "step": 14440 }, { "grad_norm": 0.05101625621318817, "learning_rate": 7.57751279387741e-05, "loss": 0.0343, "step": 14450 }, { "grad_norm": 0.047216206789016724, "learning_rate": 7.573969817152782e-05, "loss": 0.0353, "step": 14460 }, { "grad_norm": 0.05071190744638443, "learning_rate": 7.570425081144052e-05, "loss": 0.0367, "step": 14470 }, { "grad_norm": 0.0400833860039711, "learning_rate": 7.56687858827402e-05, "loss": 0.0353, "step": 14480 }, { "grad_norm": 0.053475406020879745, "learning_rate": 7.563330340966675e-05, "loss": 0.0357, "step": 14490 }, { "grad_norm": 0.04649457335472107, "learning_rate": 7.559780341647212e-05, "loss": 0.0359, "step": 14500 }, { "grad_norm": 0.05985022708773613, "learning_rate": 7.556228592742026e-05, "loss": 0.036, "step": 14510 }, { "grad_norm": 0.03349200263619423, "learning_rate": 7.552675096678696e-05, "loss": 0.0382, "step": 14520 }, { "grad_norm": 0.06086206063628197, "learning_rate": 7.549119855886012e-05, "loss": 0.0361, "step": 14530 }, { "grad_norm": 0.05320439487695694, "learning_rate": 7.545562872793941e-05, "loss": 0.0356, "step": 14540 }, { "grad_norm": 0.05127584561705589, "learning_rate": 7.542004149833648e-05, "loss": 0.0344, "step": 14550 }, { "grad_norm": 0.05351971089839935, "learning_rate": 7.538443689437492e-05, "loss": 0.036, "step": 14560 }, { "grad_norm": 0.043603189289569855, "learning_rate": 7.53488149403901e-05, "loss": 0.0355, "step": 14570 }, { "grad_norm": 0.05464618653059006, "learning_rate": 7.531317566072929e-05, "loss": 0.0371, "step": 14580 }, { "grad_norm": 0.05504905804991722, "learning_rate": 7.527751907975158e-05, "loss": 0.0388, "step": 14590 }, { "grad_norm": 0.047346167266368866, "learning_rate": 7.524184522182793e-05, "loss": 0.0376, "step": 14600 }, { "grad_norm": 0.03852618858218193, "learning_rate": 7.520615411134112e-05, "loss": 0.0387, "step": 14610 }, { "grad_norm": 0.05335785448551178, "learning_rate": 7.517044577268564e-05, "loss": 0.0388, "step": 14620 }, { "grad_norm": 0.06037973240017891, "learning_rate": 7.513472023026782e-05, "loss": 0.0352, "step": 14630 }, { "grad_norm": 0.04973027482628822, "learning_rate": 7.509897750850572e-05, "loss": 0.0355, "step": 14640 }, { "grad_norm": 0.06073792278766632, "learning_rate": 7.506321763182918e-05, "loss": 0.0339, "step": 14650 }, { "grad_norm": 0.05764419585466385, "learning_rate": 7.50274406246797e-05, "loss": 0.0334, "step": 14660 }, { "grad_norm": 0.10285790264606476, "learning_rate": 7.499164651151056e-05, "loss": 0.0334, "step": 14670 }, { "grad_norm": 0.04463162273168564, "learning_rate": 7.495583531678669e-05, "loss": 0.0353, "step": 14680 }, { "grad_norm": 0.06999067217111588, "learning_rate": 7.492000706498469e-05, "loss": 0.0331, "step": 14690 }, { "grad_norm": 0.05527466908097267, "learning_rate": 7.488416178059284e-05, "loss": 0.0329, "step": 14700 }, { "grad_norm": 0.057939931750297546, "learning_rate": 7.484829948811107e-05, "loss": 0.037, "step": 14710 }, { "grad_norm": 0.04821482300758362, "learning_rate": 7.48124202120509e-05, "loss": 0.0351, "step": 14720 }, { "grad_norm": 0.04540826380252838, "learning_rate": 7.477652397693549e-05, "loss": 0.0374, "step": 14730 }, { "grad_norm": 0.038200631737709045, "learning_rate": 7.474061080729955e-05, "loss": 0.0353, "step": 14740 }, { "grad_norm": 0.046303484588861465, "learning_rate": 7.470468072768941e-05, "loss": 0.0365, "step": 14750 }, { "grad_norm": 0.051056791096925735, "learning_rate": 7.466873376266297e-05, "loss": 0.0353, "step": 14760 }, { "grad_norm": 0.036895811557769775, "learning_rate": 7.46327699367896e-05, "loss": 0.0315, "step": 14770 }, { "grad_norm": 0.05686815083026886, "learning_rate": 7.459678927465026e-05, "loss": 0.029, "step": 14780 }, { "grad_norm": 0.11228864639997482, "learning_rate": 7.456079180083737e-05, "loss": 0.0314, "step": 14790 }, { "grad_norm": 0.05519339442253113, "learning_rate": 7.452477753995489e-05, "loss": 0.0283, "step": 14800 }, { "grad_norm": 0.04582514986395836, "learning_rate": 7.448874651661823e-05, "loss": 0.0297, "step": 14810 }, { "grad_norm": 0.04968847706913948, "learning_rate": 7.445269875545423e-05, "loss": 0.0295, "step": 14820 }, { "grad_norm": 0.08788970857858658, "learning_rate": 7.44166342811012e-05, "loss": 0.0311, "step": 14830 }, { "grad_norm": 0.04634147509932518, "learning_rate": 7.438055311820886e-05, "loss": 0.032, "step": 14840 }, { "grad_norm": 0.055957332253456116, "learning_rate": 7.434445529143837e-05, "loss": 0.0348, "step": 14850 }, { "grad_norm": 0.04317950829863548, "learning_rate": 7.430834082546225e-05, "loss": 0.0341, "step": 14860 }, { "grad_norm": 0.04536420851945877, "learning_rate": 7.427220974496438e-05, "loss": 0.0333, "step": 14870 }, { "grad_norm": 0.03184660151600838, "learning_rate": 7.423606207464005e-05, "loss": 0.0339, "step": 14880 }, { "grad_norm": 0.04269161447882652, "learning_rate": 7.419989783919578e-05, "loss": 0.0326, "step": 14890 }, { "grad_norm": 0.04712175577878952, "learning_rate": 7.416371706334956e-05, "loss": 0.0358, "step": 14900 }, { "grad_norm": 0.04841362312436104, "learning_rate": 7.412751977183056e-05, "loss": 0.0345, "step": 14910 }, { "grad_norm": 0.05212053284049034, "learning_rate": 7.409130598937932e-05, "loss": 0.0347, "step": 14920 }, { "grad_norm": 0.04941321909427643, "learning_rate": 7.40550757407476e-05, "loss": 0.034, "step": 14930 }, { "grad_norm": 0.05997121334075928, "learning_rate": 7.401882905069843e-05, "loss": 0.0344, "step": 14940 }, { "grad_norm": 0.06968791782855988, "learning_rate": 7.39825659440061e-05, "loss": 0.0358, "step": 14950 }, { "grad_norm": 0.051024049520492554, "learning_rate": 7.394628644545609e-05, "loss": 0.0359, "step": 14960 }, { "grad_norm": 0.048522595316171646, "learning_rate": 7.390999057984507e-05, "loss": 0.0327, "step": 14970 }, { "grad_norm": 0.05298072099685669, "learning_rate": 7.387367837198097e-05, "loss": 0.0335, "step": 14980 }, { "grad_norm": 0.08272150903940201, "learning_rate": 7.383734984668281e-05, "loss": 0.0318, "step": 14990 }, { "grad_norm": 0.04969868063926697, "learning_rate": 7.38010050287808e-05, "loss": 0.0322, "step": 15000 }, { "grad_norm": 0.05359876528382301, "learning_rate": 7.376464394311628e-05, "loss": 0.0302, "step": 15010 }, { "grad_norm": 0.03974495455622673, "learning_rate": 7.372826661454172e-05, "loss": 0.0291, "step": 15020 }, { "grad_norm": 0.03615577146410942, "learning_rate": 7.369187306792068e-05, "loss": 0.0297, "step": 15030 }, { "grad_norm": 0.037024375051259995, "learning_rate": 7.365546332812779e-05, "loss": 0.0299, "step": 15040 }, { "grad_norm": 0.07530766725540161, "learning_rate": 7.361903742004876e-05, "loss": 0.0306, "step": 15050 }, { "grad_norm": 0.0593290701508522, "learning_rate": 7.358259536858039e-05, "loss": 0.0281, "step": 15060 }, { "grad_norm": 0.052757423371076584, "learning_rate": 7.354613719863044e-05, "loss": 0.035, "step": 15070 }, { "grad_norm": 0.034005578607320786, "learning_rate": 7.350966293511776e-05, "loss": 0.0328, "step": 15080 }, { "grad_norm": 0.05819821357727051, "learning_rate": 7.347317260297212e-05, "loss": 0.0344, "step": 15090 }, { "grad_norm": 0.04330926015973091, "learning_rate": 7.343666622713437e-05, "loss": 0.0324, "step": 15100 }, { "grad_norm": 0.049777209758758545, "learning_rate": 7.340014383255624e-05, "loss": 0.0321, "step": 15110 }, { "grad_norm": 0.042341239750385284, "learning_rate": 7.336360544420044e-05, "loss": 0.0342, "step": 15120 }, { "grad_norm": 0.07039772719144821, "learning_rate": 7.332705108704064e-05, "loss": 0.0334, "step": 15130 }, { "grad_norm": 0.04212178289890289, "learning_rate": 7.329048078606138e-05, "loss": 0.0338, "step": 15140 }, { "grad_norm": 0.06482215970754623, "learning_rate": 7.32538945662581e-05, "loss": 0.0329, "step": 15150 }, { "grad_norm": 0.05003738775849342, "learning_rate": 7.321729245263718e-05, "loss": 0.0324, "step": 15160 }, { "grad_norm": 0.046719398349523544, "learning_rate": 7.318067447021578e-05, "loss": 0.0334, "step": 15170 }, { "grad_norm": 0.06418093293905258, "learning_rate": 7.314404064402198e-05, "loss": 0.0328, "step": 15180 }, { "grad_norm": 0.05345916002988815, "learning_rate": 7.310739099909461e-05, "loss": 0.0306, "step": 15190 }, { "grad_norm": 0.060165923088788986, "learning_rate": 7.307072556048339e-05, "loss": 0.0328, "step": 15200 }, { "grad_norm": 0.04940222576260567, "learning_rate": 7.30340443532488e-05, "loss": 0.0326, "step": 15210 }, { "grad_norm": 0.047080036252737045, "learning_rate": 7.299734740246208e-05, "loss": 0.0309, "step": 15220 }, { "grad_norm": 0.04248158261179924, "learning_rate": 7.296063473320528e-05, "loss": 0.0324, "step": 15230 }, { "grad_norm": 0.060099292546510696, "learning_rate": 7.292390637057113e-05, "loss": 0.0308, "step": 15240 }, { "grad_norm": 0.05411616712808609, "learning_rate": 7.288716233966314e-05, "loss": 0.0296, "step": 15250 }, { "grad_norm": 0.04579818621277809, "learning_rate": 7.285040266559551e-05, "loss": 0.0349, "step": 15260 }, { "grad_norm": 0.048225075006484985, "learning_rate": 7.281362737349312e-05, "loss": 0.0359, "step": 15270 }, { "grad_norm": 0.053070250898599625, "learning_rate": 7.277683648849153e-05, "loss": 0.0358, "step": 15280 }, { "grad_norm": 0.05563471093773842, "learning_rate": 7.2740030035737e-05, "loss": 0.0336, "step": 15290 }, { "grad_norm": 0.033574000000953674, "learning_rate": 7.270320804038634e-05, "loss": 0.0367, "step": 15300 }, { "grad_norm": 0.04453186318278313, "learning_rate": 7.266637052760708e-05, "loss": 0.0346, "step": 15310 }, { "grad_norm": 0.0505889467895031, "learning_rate": 7.262951752257728e-05, "loss": 0.0328, "step": 15320 }, { "grad_norm": 0.04041583463549614, "learning_rate": 7.259264905048564e-05, "loss": 0.0309, "step": 15330 }, { "grad_norm": 0.04445119947195053, "learning_rate": 7.255576513653142e-05, "loss": 0.0328, "step": 15340 }, { "grad_norm": 0.05655965209007263, "learning_rate": 7.251886580592439e-05, "loss": 0.0329, "step": 15350 }, { "grad_norm": 0.04649220407009125, "learning_rate": 7.248195108388496e-05, "loss": 0.0323, "step": 15360 }, { "grad_norm": 0.048684824258089066, "learning_rate": 7.244502099564395e-05, "loss": 0.0322, "step": 15370 }, { "grad_norm": 0.04518987610936165, "learning_rate": 7.240807556644271e-05, "loss": 0.0336, "step": 15380 }, { "grad_norm": 0.05665034055709839, "learning_rate": 7.237111482153314e-05, "loss": 0.0337, "step": 15390 }, { "grad_norm": 0.06370691955089569, "learning_rate": 7.233413878617751e-05, "loss": 0.0357, "step": 15400 }, { "grad_norm": 0.03794317692518234, "learning_rate": 7.229714748564864e-05, "loss": 0.0337, "step": 15410 }, { "grad_norm": 0.035028859972953796, "learning_rate": 7.22601409452297e-05, "loss": 0.0326, "step": 15420 }, { "grad_norm": 0.0707637295126915, "learning_rate": 7.222311919021433e-05, "loss": 0.0338, "step": 15430 }, { "grad_norm": 0.06688544899225235, "learning_rate": 7.218608224590655e-05, "loss": 0.0325, "step": 15440 }, { "grad_norm": 0.06294316798448563, "learning_rate": 7.214903013762074e-05, "loss": 0.0362, "step": 15450 }, { "grad_norm": 0.05688078701496124, "learning_rate": 7.21119628906817e-05, "loss": 0.0339, "step": 15460 }, { "grad_norm": 0.0466327965259552, "learning_rate": 7.207488053042454e-05, "loss": 0.0339, "step": 15470 }, { "grad_norm": 0.04251654073596001, "learning_rate": 7.203778308219467e-05, "loss": 0.0369, "step": 15480 }, { "grad_norm": 0.044947583228349686, "learning_rate": 7.200067057134787e-05, "loss": 0.0309, "step": 15490 }, { "grad_norm": 0.036628760397434235, "learning_rate": 7.196354302325019e-05, "loss": 0.0348, "step": 15500 }, { "grad_norm": 0.08195188641548157, "learning_rate": 7.192640046327795e-05, "loss": 0.0328, "step": 15510 }, { "grad_norm": 0.055736243724823, "learning_rate": 7.188924291681777e-05, "loss": 0.0345, "step": 15520 }, { "grad_norm": 0.08598533272743225, "learning_rate": 7.185207040926643e-05, "loss": 0.0338, "step": 15530 }, { "grad_norm": 0.05160113051533699, "learning_rate": 7.181488296603103e-05, "loss": 0.0341, "step": 15540 }, { "grad_norm": 0.09748166054487228, "learning_rate": 7.177768061252885e-05, "loss": 0.032, "step": 15550 }, { "grad_norm": 0.05128370225429535, "learning_rate": 7.174046337418729e-05, "loss": 0.0356, "step": 15560 }, { "grad_norm": 0.06322883814573288, "learning_rate": 7.170323127644403e-05, "loss": 0.0319, "step": 15570 }, { "grad_norm": 0.05370602011680603, "learning_rate": 7.166598434474683e-05, "loss": 0.0311, "step": 15580 }, { "grad_norm": 0.02950809895992279, "learning_rate": 7.162872260455364e-05, "loss": 0.0294, "step": 15590 }, { "grad_norm": 0.062088705599308014, "learning_rate": 7.159144608133248e-05, "loss": 0.0299, "step": 15600 }, { "grad_norm": 0.028365621343255043, "learning_rate": 7.155415480056153e-05, "loss": 0.0284, "step": 15610 }, { "grad_norm": 0.05437838286161423, "learning_rate": 7.151684878772902e-05, "loss": 0.028, "step": 15620 }, { "grad_norm": 0.047867145389318466, "learning_rate": 7.147952806833324e-05, "loss": 0.0272, "step": 15630 }, { "grad_norm": 0.043542828410863876, "learning_rate": 7.14421926678826e-05, "loss": 0.0301, "step": 15640 }, { "grad_norm": 0.06200586259365082, "learning_rate": 7.140484261189543e-05, "loss": 0.0314, "step": 15650 }, { "grad_norm": 0.04335838928818703, "learning_rate": 7.136747792590017e-05, "loss": 0.0321, "step": 15660 }, { "grad_norm": 0.04717600345611572, "learning_rate": 7.133009863543524e-05, "loss": 0.0344, "step": 15670 }, { "grad_norm": 0.06359857320785522, "learning_rate": 7.129270476604901e-05, "loss": 0.0302, "step": 15680 }, { "grad_norm": 0.050854649394750595, "learning_rate": 7.125529634329988e-05, "loss": 0.0321, "step": 15690 }, { "grad_norm": 0.041916679590940475, "learning_rate": 7.12178733927561e-05, "loss": 0.0318, "step": 15700 }, { "grad_norm": 0.04130154103040695, "learning_rate": 7.118043593999593e-05, "loss": 0.0293, "step": 15710 }, { "grad_norm": 0.033622223883867264, "learning_rate": 7.114298401060752e-05, "loss": 0.0301, "step": 15720 }, { "grad_norm": 0.04585988074541092, "learning_rate": 7.11055176301889e-05, "loss": 0.0277, "step": 15730 }, { "grad_norm": 0.04123583808541298, "learning_rate": 7.1068036824348e-05, "loss": 0.029, "step": 15740 }, { "grad_norm": 0.05675408989191055, "learning_rate": 7.10305416187026e-05, "loss": 0.0283, "step": 15750 }, { "grad_norm": 0.05546307936310768, "learning_rate": 7.099303203888029e-05, "loss": 0.0298, "step": 15760 }, { "grad_norm": 0.03327338024973869, "learning_rate": 7.095550811051855e-05, "loss": 0.0281, "step": 15770 }, { "grad_norm": 0.03974440321326256, "learning_rate": 7.09179698592646e-05, "loss": 0.0303, "step": 15780 }, { "grad_norm": 0.056420885026454926, "learning_rate": 7.088041731077551e-05, "loss": 0.0314, "step": 15790 }, { "grad_norm": 0.05881304666399956, "learning_rate": 7.084285049071806e-05, "loss": 0.0303, "step": 15800 }, { "grad_norm": 0.04299471154808998, "learning_rate": 7.080526942476886e-05, "loss": 0.0318, "step": 15810 }, { "grad_norm": 0.05183456465601921, "learning_rate": 7.076767413861418e-05, "loss": 0.0307, "step": 15820 }, { "grad_norm": 0.06551162153482437, "learning_rate": 7.073006465795005e-05, "loss": 0.0323, "step": 15830 }, { "grad_norm": 0.04156612604856491, "learning_rate": 7.06924410084822e-05, "loss": 0.0305, "step": 15840 }, { "grad_norm": 0.060044918209314346, "learning_rate": 7.065480321592604e-05, "loss": 0.0325, "step": 15850 }, { "grad_norm": 0.05528507009148598, "learning_rate": 7.061715130600663e-05, "loss": 0.0309, "step": 15860 }, { "grad_norm": 0.04027339443564415, "learning_rate": 7.057948530445873e-05, "loss": 0.0325, "step": 15870 }, { "grad_norm": 0.046543557196855545, "learning_rate": 7.054180523702668e-05, "loss": 0.0303, "step": 15880 }, { "grad_norm": 0.04194234684109688, "learning_rate": 7.050411112946442e-05, "loss": 0.032, "step": 15890 }, { "grad_norm": 0.05413074046373367, "learning_rate": 7.046640300753557e-05, "loss": 0.0338, "step": 15900 }, { "grad_norm": 0.06881837546825409, "learning_rate": 7.042868089701325e-05, "loss": 0.0344, "step": 15910 }, { "grad_norm": 0.03717384114861488, "learning_rate": 7.039094482368016e-05, "loss": 0.0347, "step": 15920 }, { "grad_norm": 0.04582146182656288, "learning_rate": 7.035319481332858e-05, "loss": 0.0353, "step": 15930 }, { "grad_norm": 0.051447656005620956, "learning_rate": 7.031543089176023e-05, "loss": 0.0339, "step": 15940 }, { "grad_norm": 0.04420902952551842, "learning_rate": 7.027765308478644e-05, "loss": 0.0346, "step": 15950 }, { "grad_norm": 0.050443217158317566, "learning_rate": 7.023986141822798e-05, "loss": 0.0319, "step": 15960 }, { "grad_norm": 0.04680892825126648, "learning_rate": 7.02020559179151e-05, "loss": 0.034, "step": 15970 }, { "grad_norm": 0.03812532126903534, "learning_rate": 7.016423660968748e-05, "loss": 0.0363, "step": 15980 }, { "grad_norm": 0.043833374977111816, "learning_rate": 7.012640351939428e-05, "loss": 0.0343, "step": 15990 }, { "grad_norm": 0.04392975568771362, "learning_rate": 7.008855667289404e-05, "loss": 0.0329, "step": 16000 }, { "grad_norm": 0.044667426496744156, "learning_rate": 7.005069609605476e-05, "loss": 0.0336, "step": 16010 }, { "grad_norm": 0.06065202131867409, "learning_rate": 7.001282181475377e-05, "loss": 0.0363, "step": 16020 }, { "grad_norm": 0.06378690898418427, "learning_rate": 6.997493385487775e-05, "loss": 0.0335, "step": 16030 }, { "grad_norm": 0.06438828259706497, "learning_rate": 6.99370322423228e-05, "loss": 0.0329, "step": 16040 }, { "grad_norm": 0.045307815074920654, "learning_rate": 6.989911700299433e-05, "loss": 0.0332, "step": 16050 }, { "grad_norm": 0.05173710361123085, "learning_rate": 6.9861188162807e-05, "loss": 0.0333, "step": 16060 }, { "grad_norm": 0.037699948996305466, "learning_rate": 6.982324574768487e-05, "loss": 0.0341, "step": 16070 }, { "grad_norm": 0.04524380713701248, "learning_rate": 6.978528978356117e-05, "loss": 0.0333, "step": 16080 }, { "grad_norm": 0.05117001011967659, "learning_rate": 6.974732029637846e-05, "loss": 0.033, "step": 16090 }, { "grad_norm": 0.058838099241256714, "learning_rate": 6.970933731208855e-05, "loss": 0.0331, "step": 16100 }, { "grad_norm": 0.04081297293305397, "learning_rate": 6.967134085665244e-05, "loss": 0.0348, "step": 16110 }, { "grad_norm": 0.06786549836397171, "learning_rate": 6.963333095604034e-05, "loss": 0.0343, "step": 16120 }, { "grad_norm": 0.04499155282974243, "learning_rate": 6.959530763623166e-05, "loss": 0.0335, "step": 16130 }, { "grad_norm": 0.04567566141486168, "learning_rate": 6.955727092321497e-05, "loss": 0.0332, "step": 16140 }, { "grad_norm": 0.05877925083041191, "learning_rate": 6.951922084298803e-05, "loss": 0.0355, "step": 16150 }, { "grad_norm": 0.06932735443115234, "learning_rate": 6.948115742155769e-05, "loss": 0.0361, "step": 16160 }, { "grad_norm": 0.05727159604430199, "learning_rate": 6.944308068493996e-05, "loss": 0.0337, "step": 16170 }, { "grad_norm": 0.043760865926742554, "learning_rate": 6.940499065915992e-05, "loss": 0.0332, "step": 16180 }, { "grad_norm": 0.07584185153245926, "learning_rate": 6.936688737025173e-05, "loss": 0.0347, "step": 16190 }, { "grad_norm": 0.053992532193660736, "learning_rate": 6.932877084425867e-05, "loss": 0.033, "step": 16200 }, { "grad_norm": 0.04089123010635376, "learning_rate": 6.929064110723297e-05, "loss": 0.0342, "step": 16210 }, { "grad_norm": 0.0471280999481678, "learning_rate": 6.925249818523598e-05, "loss": 0.033, "step": 16220 }, { "grad_norm": 0.04315541684627533, "learning_rate": 6.921434210433801e-05, "loss": 0.0311, "step": 16230 }, { "grad_norm": 0.04856153950095177, "learning_rate": 6.917617289061841e-05, "loss": 0.0313, "step": 16240 }, { "grad_norm": 0.05787340924143791, "learning_rate": 6.913799057016547e-05, "loss": 0.0333, "step": 16250 }, { "grad_norm": 0.036909300833940506, "learning_rate": 6.909979516907641e-05, "loss": 0.0323, "step": 16260 }, { "grad_norm": 0.04742175713181496, "learning_rate": 6.906158671345746e-05, "loss": 0.031, "step": 16270 }, { "grad_norm": 0.05087345838546753, "learning_rate": 6.902336522942374e-05, "loss": 0.0309, "step": 16280 }, { "grad_norm": 0.04741157963871956, "learning_rate": 6.898513074309924e-05, "loss": 0.0311, "step": 16290 }, { "grad_norm": 0.07127781957387924, "learning_rate": 6.894688328061693e-05, "loss": 0.0293, "step": 16300 }, { "grad_norm": 0.046605031937360764, "learning_rate": 6.890862286811853e-05, "loss": 0.0306, "step": 16310 }, { "grad_norm": 0.05408583953976631, "learning_rate": 6.88703495317547e-05, "loss": 0.0299, "step": 16320 }, { "grad_norm": 0.03945572301745415, "learning_rate": 6.883206329768492e-05, "loss": 0.0306, "step": 16330 }, { "grad_norm": 0.054980404675006866, "learning_rate": 6.879376419207743e-05, "loss": 0.0315, "step": 16340 }, { "grad_norm": 0.06479953229427338, "learning_rate": 6.875545224110935e-05, "loss": 0.0357, "step": 16350 }, { "grad_norm": 0.05414945259690285, "learning_rate": 6.871712747096651e-05, "loss": 0.0347, "step": 16360 }, { "grad_norm": 0.05302751436829567, "learning_rate": 6.867878990784353e-05, "loss": 0.0353, "step": 16370 }, { "grad_norm": 0.044091708958148956, "learning_rate": 6.864043957794377e-05, "loss": 0.0351, "step": 16380 }, { "grad_norm": 0.050862859934568405, "learning_rate": 6.860207650747934e-05, "loss": 0.0357, "step": 16390 }, { "grad_norm": 0.061585333198308945, "learning_rate": 6.856370072267104e-05, "loss": 0.0335, "step": 16400 }, { "grad_norm": 0.06703829020261765, "learning_rate": 6.852531224974831e-05, "loss": 0.0325, "step": 16410 }, { "grad_norm": 0.05450340732932091, "learning_rate": 6.848691111494936e-05, "loss": 0.0317, "step": 16420 }, { "grad_norm": 0.047875553369522095, "learning_rate": 6.844849734452097e-05, "loss": 0.032, "step": 16430 }, { "grad_norm": 0.035144463181495667, "learning_rate": 6.841007096471862e-05, "loss": 0.0314, "step": 16440 }, { "grad_norm": 0.03156546130776405, "learning_rate": 6.837163200180636e-05, "loss": 0.0305, "step": 16450 }, { "grad_norm": 0.04546022415161133, "learning_rate": 6.833318048205684e-05, "loss": 0.0301, "step": 16460 }, { "grad_norm": 0.06248312443494797, "learning_rate": 6.829471643175136e-05, "loss": 0.0322, "step": 16470 }, { "grad_norm": 0.04389537498354912, "learning_rate": 6.825623987717969e-05, "loss": 0.0323, "step": 16480 }, { "grad_norm": 0.04764392226934433, "learning_rate": 6.821775084464022e-05, "loss": 0.0315, "step": 16490 }, { "grad_norm": 0.04857713729143143, "learning_rate": 6.817924936043982e-05, "loss": 0.0332, "step": 16500 }, { "grad_norm": 0.04424086958169937, "learning_rate": 6.81407354508939e-05, "loss": 0.0324, "step": 16510 }, { "grad_norm": 0.049100663512945175, "learning_rate": 6.810220914232636e-05, "loss": 0.0325, "step": 16520 }, { "grad_norm": 0.050894517451524734, "learning_rate": 6.806367046106959e-05, "loss": 0.0322, "step": 16530 }, { "grad_norm": 0.04257355257868767, "learning_rate": 6.802511943346435e-05, "loss": 0.0337, "step": 16540 }, { "grad_norm": 0.058580975979566574, "learning_rate": 6.798655608585997e-05, "loss": 0.0311, "step": 16550 }, { "grad_norm": 0.04785739630460739, "learning_rate": 6.79479804446141e-05, "loss": 0.0304, "step": 16560 }, { "grad_norm": 0.045056503266096115, "learning_rate": 6.790939253609284e-05, "loss": 0.0307, "step": 16570 }, { "grad_norm": 0.0478798970580101, "learning_rate": 6.787079238667065e-05, "loss": 0.0316, "step": 16580 }, { "grad_norm": 0.047592416405677795, "learning_rate": 6.783218002273039e-05, "loss": 0.0308, "step": 16590 }, { "grad_norm": 0.04501166567206383, "learning_rate": 6.779355547066322e-05, "loss": 0.0309, "step": 16600 }, { "grad_norm": 0.04631853848695755, "learning_rate": 6.775491875686865e-05, "loss": 0.0354, "step": 16610 }, { "grad_norm": 0.0557868592441082, "learning_rate": 6.771626990775457e-05, "loss": 0.0341, "step": 16620 }, { "grad_norm": 0.055374789983034134, "learning_rate": 6.767760894973704e-05, "loss": 0.0333, "step": 16630 }, { "grad_norm": 0.07379171997308731, "learning_rate": 6.763893590924048e-05, "loss": 0.032, "step": 16640 }, { "grad_norm": 0.04362519457936287, "learning_rate": 6.760025081269756e-05, "loss": 0.0346, "step": 16650 }, { "grad_norm": 0.04391288757324219, "learning_rate": 6.756155368654915e-05, "loss": 0.0346, "step": 16660 }, { "grad_norm": 0.05108566954731941, "learning_rate": 6.752284455724442e-05, "loss": 0.0328, "step": 16670 }, { "grad_norm": 0.05133218318223953, "learning_rate": 6.748412345124065e-05, "loss": 0.0326, "step": 16680 }, { "grad_norm": 0.04035607725381851, "learning_rate": 6.744539039500335e-05, "loss": 0.032, "step": 16690 }, { "grad_norm": 0.03380243852734566, "learning_rate": 6.740664541500625e-05, "loss": 0.0315, "step": 16700 }, { "grad_norm": 0.05976006016135216, "learning_rate": 6.736788853773112e-05, "loss": 0.034, "step": 16710 }, { "grad_norm": 0.057180434465408325, "learning_rate": 6.732911978966796e-05, "loss": 0.0338, "step": 16720 }, { "grad_norm": 0.05488419160246849, "learning_rate": 6.729033919731482e-05, "loss": 0.0368, "step": 16730 }, { "grad_norm": 0.04778492450714111, "learning_rate": 6.725154678717787e-05, "loss": 0.0372, "step": 16740 }, { "grad_norm": 0.04983019456267357, "learning_rate": 6.721274258577138e-05, "loss": 0.0368, "step": 16750 }, { "grad_norm": 0.0431223139166832, "learning_rate": 6.717392661961763e-05, "loss": 0.0355, "step": 16760 }, { "grad_norm": 0.05342935398221016, "learning_rate": 6.713509891524697e-05, "loss": 0.0381, "step": 16770 }, { "grad_norm": 0.052466392517089844, "learning_rate": 6.709625949919777e-05, "loss": 0.0358, "step": 16780 }, { "grad_norm": 0.05967364460229874, "learning_rate": 6.705740839801642e-05, "loss": 0.0353, "step": 16790 }, { "grad_norm": 0.05741751566529274, "learning_rate": 6.701854563825727e-05, "loss": 0.0348, "step": 16800 }, { "grad_norm": 0.0450434610247612, "learning_rate": 6.697967124648266e-05, "loss": 0.0354, "step": 16810 }, { "grad_norm": 0.0513116829097271, "learning_rate": 6.694078524926285e-05, "loss": 0.0356, "step": 16820 }, { "grad_norm": 0.03918835520744324, "learning_rate": 6.690188767317607e-05, "loss": 0.0333, "step": 16830 }, { "grad_norm": 0.06463085860013962, "learning_rate": 6.686297854480843e-05, "loss": 0.0345, "step": 16840 }, { "grad_norm": 0.04628203436732292, "learning_rate": 6.682405789075398e-05, "loss": 0.0336, "step": 16850 }, { "grad_norm": 0.04935593158006668, "learning_rate": 6.67851257376146e-05, "loss": 0.0343, "step": 16860 }, { "grad_norm": 0.05007685348391533, "learning_rate": 6.674618211200004e-05, "loss": 0.0309, "step": 16870 }, { "grad_norm": 0.06849892437458038, "learning_rate": 6.670722704052792e-05, "loss": 0.0322, "step": 16880 }, { "grad_norm": 0.05875025689601898, "learning_rate": 6.666826054982365e-05, "loss": 0.0333, "step": 16890 }, { "grad_norm": 0.06279774010181427, "learning_rate": 6.662928266652048e-05, "loss": 0.0322, "step": 16900 }, { "grad_norm": 0.06664731353521347, "learning_rate": 6.659029341725941e-05, "loss": 0.0325, "step": 16910 }, { "grad_norm": 0.035484250634908676, "learning_rate": 6.655129282868923e-05, "loss": 0.0318, "step": 16920 }, { "grad_norm": 0.08640439808368683, "learning_rate": 6.651228092746646e-05, "loss": 0.0337, "step": 16930 }, { "grad_norm": 0.05875580012798309, "learning_rate": 6.647325774025539e-05, "loss": 0.0332, "step": 16940 }, { "grad_norm": 0.07748135179281235, "learning_rate": 6.643422329372798e-05, "loss": 0.0338, "step": 16950 }, { "grad_norm": 0.04192790389060974, "learning_rate": 6.639517761456392e-05, "loss": 0.0303, "step": 16960 }, { "grad_norm": 0.04612673819065094, "learning_rate": 6.635612072945054e-05, "loss": 0.0323, "step": 16970 }, { "grad_norm": 0.05385121703147888, "learning_rate": 6.631705266508289e-05, "loss": 0.0328, "step": 16980 }, { "grad_norm": 0.037992075085639954, "learning_rate": 6.62779734481636e-05, "loss": 0.0308, "step": 16990 }, { "grad_norm": 0.04413142055273056, "learning_rate": 6.623888310540294e-05, "loss": 0.0305, "step": 17000 }, { "grad_norm": 0.04642147570848465, "learning_rate": 6.619978166351882e-05, "loss": 0.0306, "step": 17010 }, { "grad_norm": 0.06569623947143555, "learning_rate": 6.616066914923666e-05, "loss": 0.0335, "step": 17020 }, { "grad_norm": 0.0472518615424633, "learning_rate": 6.612154558928955e-05, "loss": 0.0333, "step": 17030 }, { "grad_norm": 0.0490688718855381, "learning_rate": 6.608241101041804e-05, "loss": 0.0311, "step": 17040 }, { "grad_norm": 0.04840423911809921, "learning_rate": 6.604326543937025e-05, "loss": 0.0322, "step": 17050 }, { "grad_norm": 0.03886749595403671, "learning_rate": 6.60041089029018e-05, "loss": 0.0287, "step": 17060 }, { "grad_norm": 0.04239228740334511, "learning_rate": 6.596494142777583e-05, "loss": 0.0315, "step": 17070 }, { "grad_norm": 0.047152355313301086, "learning_rate": 6.592576304076294e-05, "loss": 0.0312, "step": 17080 }, { "grad_norm": 0.05413873493671417, "learning_rate": 6.588657376864119e-05, "loss": 0.0303, "step": 17090 }, { "grad_norm": 0.041122470051050186, "learning_rate": 6.584737363819605e-05, "loss": 0.0312, "step": 17100 }, { "grad_norm": 0.04846158251166344, "learning_rate": 6.580816267622048e-05, "loss": 0.0328, "step": 17110 }, { "grad_norm": 0.05130019038915634, "learning_rate": 6.576894090951478e-05, "loss": 0.0339, "step": 17120 }, { "grad_norm": 0.050091810524463654, "learning_rate": 6.572970836488665e-05, "loss": 0.0346, "step": 17130 }, { "grad_norm": 0.047592293471097946, "learning_rate": 6.569046506915119e-05, "loss": 0.035, "step": 17140 }, { "grad_norm": 0.04671717807650566, "learning_rate": 6.56512110491308e-05, "loss": 0.0328, "step": 17150 }, { "grad_norm": 0.05431406944990158, "learning_rate": 6.561194633165523e-05, "loss": 0.036, "step": 17160 }, { "grad_norm": 0.04158993810415268, "learning_rate": 6.557267094356155e-05, "loss": 0.0343, "step": 17170 }, { "grad_norm": 0.05895833298563957, "learning_rate": 6.553338491169414e-05, "loss": 0.0324, "step": 17180 }, { "grad_norm": 0.04233293607831001, "learning_rate": 6.54940882629046e-05, "loss": 0.0316, "step": 17190 }, { "grad_norm": 0.03144959732890129, "learning_rate": 6.545478102405184e-05, "loss": 0.0315, "step": 17200 }, { "grad_norm": 0.037709273397922516, "learning_rate": 6.541546322200199e-05, "loss": 0.0329, "step": 17210 }, { "grad_norm": 0.04700122028589249, "learning_rate": 6.537613488362837e-05, "loss": 0.0297, "step": 17220 }, { "grad_norm": 0.05534251406788826, "learning_rate": 6.533679603581155e-05, "loss": 0.0321, "step": 17230 }, { "grad_norm": 0.0651179626584053, "learning_rate": 6.529744670543926e-05, "loss": 0.0352, "step": 17240 }, { "grad_norm": 0.041912537068128586, "learning_rate": 6.52580869194064e-05, "loss": 0.0344, "step": 17250 }, { "grad_norm": 0.04523606598377228, "learning_rate": 6.521871670461499e-05, "loss": 0.033, "step": 17260 }, { "grad_norm": 0.05002438649535179, "learning_rate": 6.517933608797422e-05, "loss": 0.0364, "step": 17270 }, { "grad_norm": 0.04314059019088745, "learning_rate": 6.513994509640038e-05, "loss": 0.034, "step": 17280 }, { "grad_norm": 0.04527731612324715, "learning_rate": 6.510054375681682e-05, "loss": 0.0333, "step": 17290 }, { "grad_norm": 0.06955987960100174, "learning_rate": 6.506113209615398e-05, "loss": 0.0365, "step": 17300 }, { "grad_norm": 0.04268650338053703, "learning_rate": 6.502171014134938e-05, "loss": 0.0334, "step": 17310 }, { "grad_norm": 0.05059539154171944, "learning_rate": 6.498227791934755e-05, "loss": 0.0326, "step": 17320 }, { "grad_norm": 0.04008614271879196, "learning_rate": 6.494283545710003e-05, "loss": 0.0332, "step": 17330 }, { "grad_norm": 0.06082245707511902, "learning_rate": 6.490338278156538e-05, "loss": 0.0317, "step": 17340 }, { "grad_norm": 0.03095209039747715, "learning_rate": 6.486391991970913e-05, "loss": 0.0325, "step": 17350 }, { "grad_norm": 0.057393770664930344, "learning_rate": 6.482444689850377e-05, "loss": 0.0313, "step": 17360 }, { "grad_norm": 0.05742653086781502, "learning_rate": 6.478496374492875e-05, "loss": 0.0337, "step": 17370 }, { "grad_norm": 0.04826594144105911, "learning_rate": 6.474547048597042e-05, "loss": 0.0356, "step": 17380 }, { "grad_norm": 0.06840366125106812, "learning_rate": 6.470596714862205e-05, "loss": 0.0328, "step": 17390 }, { "grad_norm": 0.056852757930755615, "learning_rate": 6.46664537598838e-05, "loss": 0.0327, "step": 17400 }, { "grad_norm": 0.03675468638539314, "learning_rate": 6.462693034676271e-05, "loss": 0.0324, "step": 17410 }, { "grad_norm": 0.04593544080853462, "learning_rate": 6.458739693627265e-05, "loss": 0.0304, "step": 17420 }, { "grad_norm": 0.05093788728117943, "learning_rate": 6.454785355543432e-05, "loss": 0.0327, "step": 17430 }, { "grad_norm": 0.03585749864578247, "learning_rate": 6.450830023127528e-05, "loss": 0.0314, "step": 17440 }, { "grad_norm": 0.0605081208050251, "learning_rate": 6.446873699082982e-05, "loss": 0.0329, "step": 17450 }, { "grad_norm": 0.038347918540239334, "learning_rate": 6.44291638611391e-05, "loss": 0.0298, "step": 17460 }, { "grad_norm": 0.05895008146762848, "learning_rate": 6.43895808692509e-05, "loss": 0.0314, "step": 17470 }, { "grad_norm": 0.051310356706380844, "learning_rate": 6.434998804221986e-05, "loss": 0.0307, "step": 17480 }, { "grad_norm": 0.05368873104453087, "learning_rate": 6.431038540710732e-05, "loss": 0.0328, "step": 17490 }, { "grad_norm": 0.046339984983205795, "learning_rate": 6.427077299098129e-05, "loss": 0.032, "step": 17500 }, { "grad_norm": 0.04576772823929787, "learning_rate": 6.423115082091651e-05, "loss": 0.0323, "step": 17510 }, { "grad_norm": 0.05826675891876221, "learning_rate": 6.419151892399429e-05, "loss": 0.0321, "step": 17520 }, { "grad_norm": 0.06323760747909546, "learning_rate": 6.415187732730273e-05, "loss": 0.0325, "step": 17530 }, { "grad_norm": 0.06319580972194672, "learning_rate": 6.411222605793645e-05, "loss": 0.0335, "step": 17540 }, { "grad_norm": 0.0563216432929039, "learning_rate": 6.407256514299674e-05, "loss": 0.0315, "step": 17550 }, { "grad_norm": 0.049789298325777054, "learning_rate": 6.403289460959147e-05, "loss": 0.0311, "step": 17560 }, { "grad_norm": 0.050030335783958435, "learning_rate": 6.399321448483501e-05, "loss": 0.0293, "step": 17570 }, { "grad_norm": 0.0448915995657444, "learning_rate": 6.395352479584844e-05, "loss": 0.0307, "step": 17580 }, { "grad_norm": 0.04993382841348648, "learning_rate": 6.391382556975923e-05, "loss": 0.0291, "step": 17590 }, { "grad_norm": 0.0468224473297596, "learning_rate": 6.387411683370144e-05, "loss": 0.0291, "step": 17600 }, { "grad_norm": 0.04482126235961914, "learning_rate": 6.383439861481562e-05, "loss": 0.0307, "step": 17610 }, { "grad_norm": 0.05489707738161087, "learning_rate": 6.379467094024879e-05, "loss": 0.0327, "step": 17620 }, { "grad_norm": 0.047881681472063065, "learning_rate": 6.375493383715445e-05, "loss": 0.032, "step": 17630 }, { "grad_norm": 0.04124274477362633, "learning_rate": 6.371518733269254e-05, "loss": 0.0331, "step": 17640 }, { "grad_norm": 0.04451710730791092, "learning_rate": 6.367543145402942e-05, "loss": 0.0344, "step": 17650 }, { "grad_norm": 0.055654145777225494, "learning_rate": 6.363566622833785e-05, "loss": 0.0323, "step": 17660 }, { "grad_norm": 0.05158569663763046, "learning_rate": 6.359589168279698e-05, "loss": 0.0323, "step": 17670 }, { "grad_norm": 0.032771822065114975, "learning_rate": 6.355610784459235e-05, "loss": 0.0348, "step": 17680 }, { "grad_norm": 0.047691408544778824, "learning_rate": 6.351631474091585e-05, "loss": 0.0363, "step": 17690 }, { "grad_norm": 0.047163352370262146, "learning_rate": 6.347651239896566e-05, "loss": 0.0358, "step": 17700 }, { "grad_norm": 0.06354977190494537, "learning_rate": 6.343670084594633e-05, "loss": 0.0372, "step": 17710 }, { "grad_norm": 0.07049129903316498, "learning_rate": 6.339688010906866e-05, "loss": 0.0334, "step": 17720 }, { "grad_norm": 0.06101600080728531, "learning_rate": 6.335705021554975e-05, "loss": 0.0329, "step": 17730 }, { "grad_norm": 0.06110597029328346, "learning_rate": 6.3317211192613e-05, "loss": 0.0347, "step": 17740 }, { "grad_norm": 0.046816207468509674, "learning_rate": 6.327736306748795e-05, "loss": 0.0329, "step": 17750 }, { "grad_norm": 0.05818399414420128, "learning_rate": 6.323750586741047e-05, "loss": 0.034, "step": 17760 }, { "grad_norm": 0.06650693714618683, "learning_rate": 6.319763961962252e-05, "loss": 0.034, "step": 17770 }, { "grad_norm": 0.042645689100027084, "learning_rate": 6.315776435137233e-05, "loss": 0.0334, "step": 17780 }, { "grad_norm": 0.06609048694372177, "learning_rate": 6.311788008991432e-05, "loss": 0.0341, "step": 17790 }, { "grad_norm": 0.06532914936542511, "learning_rate": 6.307798686250891e-05, "loss": 0.0339, "step": 17800 }, { "grad_norm": 0.049560561776161194, "learning_rate": 6.303808469642284e-05, "loss": 0.0299, "step": 17810 }, { "grad_norm": 0.03690141811966896, "learning_rate": 6.29981736189288e-05, "loss": 0.0282, "step": 17820 }, { "grad_norm": 0.040507178753614426, "learning_rate": 6.295825365730567e-05, "loss": 0.0309, "step": 17830 }, { "grad_norm": 0.059820353984832764, "learning_rate": 6.291832483883835e-05, "loss": 0.0288, "step": 17840 }, { "grad_norm": 0.03962410241365433, "learning_rate": 6.28783871908178e-05, "loss": 0.0292, "step": 17850 }, { "grad_norm": 0.05174684897065163, "learning_rate": 6.283844074054107e-05, "loss": 0.0283, "step": 17860 }, { "grad_norm": 0.0762253850698471, "learning_rate": 6.279848551531112e-05, "loss": 0.031, "step": 17870 }, { "grad_norm": 0.07037408649921417, "learning_rate": 6.275852154243702e-05, "loss": 0.0307, "step": 17880 }, { "grad_norm": 0.06011907756328583, "learning_rate": 6.271854884923377e-05, "loss": 0.0302, "step": 17890 }, { "grad_norm": 0.045755621045827866, "learning_rate": 6.267856746302228e-05, "loss": 0.0282, "step": 17900 }, { "grad_norm": 0.05230601131916046, "learning_rate": 6.263857741112948e-05, "loss": 0.0306, "step": 17910 }, { "grad_norm": 0.04297307878732681, "learning_rate": 6.259857872088821e-05, "loss": 0.0288, "step": 17920 }, { "grad_norm": 0.03587720915675163, "learning_rate": 6.255857141963719e-05, "loss": 0.0305, "step": 17930 }, { "grad_norm": 0.06062362715601921, "learning_rate": 6.251855553472101e-05, "loss": 0.0329, "step": 17940 }, { "grad_norm": 0.037095800042152405, "learning_rate": 6.247853109349016e-05, "loss": 0.0328, "step": 17950 }, { "grad_norm": 0.04269901290535927, "learning_rate": 6.243849812330098e-05, "loss": 0.0333, "step": 17960 }, { "grad_norm": 0.05051492154598236, "learning_rate": 6.239845665151563e-05, "loss": 0.03, "step": 17970 }, { "grad_norm": 0.039611008018255234, "learning_rate": 6.235840670550204e-05, "loss": 0.0324, "step": 17980 }, { "grad_norm": 0.05296223238110542, "learning_rate": 6.231834831263403e-05, "loss": 0.0344, "step": 17990 }, { "grad_norm": 0.05390894412994385, "learning_rate": 6.22782815002911e-05, "loss": 0.0321, "step": 18000 }, { "grad_norm": 0.05373124033212662, "learning_rate": 6.223820629585852e-05, "loss": 0.029, "step": 18010 }, { "grad_norm": 0.047031521797180176, "learning_rate": 6.219812272672737e-05, "loss": 0.0288, "step": 18020 }, { "grad_norm": 0.047648701816797256, "learning_rate": 6.215803082029434e-05, "loss": 0.0286, "step": 18030 }, { "grad_norm": 0.042923975735902786, "learning_rate": 6.211793060396188e-05, "loss": 0.0298, "step": 18040 }, { "grad_norm": 0.040685322135686874, "learning_rate": 6.207782210513811e-05, "loss": 0.029, "step": 18050 }, { "grad_norm": 0.042587846517562866, "learning_rate": 6.203770535123683e-05, "loss": 0.0292, "step": 18060 }, { "grad_norm": 0.04886810481548309, "learning_rate": 6.199758036967747e-05, "loss": 0.0293, "step": 18070 }, { "grad_norm": 0.04558977484703064, "learning_rate": 6.195744718788503e-05, "loss": 0.03, "step": 18080 }, { "grad_norm": 0.05029740929603577, "learning_rate": 6.191730583329021e-05, "loss": 0.0298, "step": 18090 }, { "grad_norm": 0.039460696280002594, "learning_rate": 6.187715633332921e-05, "loss": 0.0293, "step": 18100 }, { "grad_norm": 0.0441880077123642, "learning_rate": 6.183699871544386e-05, "loss": 0.03, "step": 18110 }, { "grad_norm": 0.06909732520580292, "learning_rate": 6.179683300708152e-05, "loss": 0.0334, "step": 18120 }, { "grad_norm": 0.04106435924768448, "learning_rate": 6.175665923569503e-05, "loss": 0.0349, "step": 18130 }, { "grad_norm": 0.036894574761390686, "learning_rate": 6.171647742874281e-05, "loss": 0.0328, "step": 18140 }, { "grad_norm": 0.04666689783334732, "learning_rate": 6.167628761368875e-05, "loss": 0.0347, "step": 18150 }, { "grad_norm": 0.036985788494348526, "learning_rate": 6.163608981800222e-05, "loss": 0.0337, "step": 18160 }, { "grad_norm": 0.06146126613020897, "learning_rate": 6.159588406915803e-05, "loss": 0.0331, "step": 18170 }, { "grad_norm": 0.038606222718954086, "learning_rate": 6.155567039463639e-05, "loss": 0.0327, "step": 18180 }, { "grad_norm": 0.04730590060353279, "learning_rate": 6.151544882192302e-05, "loss": 0.0329, "step": 18190 }, { "grad_norm": 0.05155196413397789, "learning_rate": 6.147521937850895e-05, "loss": 0.0353, "step": 18200 }, { "grad_norm": 0.0359080545604229, "learning_rate": 6.143498209189066e-05, "loss": 0.0326, "step": 18210 }, { "grad_norm": 0.06285113096237183, "learning_rate": 6.139473698956993e-05, "loss": 0.0302, "step": 18220 }, { "grad_norm": 0.03974495083093643, "learning_rate": 6.13544840990539e-05, "loss": 0.0342, "step": 18230 }, { "grad_norm": 0.05024199187755585, "learning_rate": 6.131422344785507e-05, "loss": 0.0321, "step": 18240 }, { "grad_norm": 0.04754867032170296, "learning_rate": 6.127395506349119e-05, "loss": 0.0331, "step": 18250 }, { "grad_norm": 0.04232674092054367, "learning_rate": 6.123367897348533e-05, "loss": 0.0296, "step": 18260 }, { "grad_norm": 0.041858773678541183, "learning_rate": 6.119339520536584e-05, "loss": 0.0309, "step": 18270 }, { "grad_norm": 0.04116116464138031, "learning_rate": 6.115310378666625e-05, "loss": 0.0296, "step": 18280 }, { "grad_norm": 0.04237234964966774, "learning_rate": 6.11128047449254e-05, "loss": 0.0313, "step": 18290 }, { "grad_norm": 0.054811276495456696, "learning_rate": 6.107249810768729e-05, "loss": 0.0297, "step": 18300 }, { "grad_norm": 0.040229588747024536, "learning_rate": 6.1032183902501125e-05, "loss": 0.0276, "step": 18310 }, { "grad_norm": 0.04573875665664673, "learning_rate": 6.099186215692131e-05, "loss": 0.031, "step": 18320 }, { "grad_norm": 0.045602377504110336, "learning_rate": 6.095153289850734e-05, "loss": 0.0291, "step": 18330 }, { "grad_norm": 0.04005013406276703, "learning_rate": 6.0911196154823904e-05, "loss": 0.0278, "step": 18340 }, { "grad_norm": 0.045214999467134476, "learning_rate": 6.087085195344079e-05, "loss": 0.0274, "step": 18350 }, { "grad_norm": 0.03643769770860672, "learning_rate": 6.083050032193286e-05, "loss": 0.0281, "step": 18360 }, { "grad_norm": 0.04774174466729164, "learning_rate": 6.0790141287880097e-05, "loss": 0.0287, "step": 18370 }, { "grad_norm": 0.05746506527066231, "learning_rate": 6.0749774878867496e-05, "loss": 0.0308, "step": 18380 }, { "grad_norm": 0.0479654036462307, "learning_rate": 6.0709401122485146e-05, "loss": 0.0314, "step": 18390 }, { "grad_norm": 0.05046403035521507, "learning_rate": 6.066902004632811e-05, "loss": 0.029, "step": 18400 }, { "grad_norm": 0.04991855099797249, "learning_rate": 6.062863167799646e-05, "loss": 0.0316, "step": 18410 }, { "grad_norm": 0.05880908668041229, "learning_rate": 6.058823604509529e-05, "loss": 0.0301, "step": 18420 }, { "grad_norm": 0.048334330320358276, "learning_rate": 6.054783317523462e-05, "loss": 0.0304, "step": 18430 }, { "grad_norm": 0.046000197529792786, "learning_rate": 6.050742309602944e-05, "loss": 0.0295, "step": 18440 }, { "grad_norm": 0.0902673751115799, "learning_rate": 6.046700583509965e-05, "loss": 0.0293, "step": 18450 }, { "grad_norm": 0.045111220329999924, "learning_rate": 6.042658142007007e-05, "loss": 0.0306, "step": 18460 }, { "grad_norm": 0.05277486890554428, "learning_rate": 6.038614987857041e-05, "loss": 0.0297, "step": 18470 }, { "grad_norm": 0.056013915687799454, "learning_rate": 6.0345711238235224e-05, "loss": 0.0292, "step": 18480 }, { "grad_norm": 0.08652425557374954, "learning_rate": 6.030526552670399e-05, "loss": 0.0298, "step": 18490 }, { "grad_norm": 0.04199156537652016, "learning_rate": 6.0264812771620925e-05, "loss": 0.0282, "step": 18500 }, { "grad_norm": 0.046599131077528, "learning_rate": 6.022435300063512e-05, "loss": 0.0297, "step": 18510 }, { "grad_norm": 0.04128339886665344, "learning_rate": 6.0183886241400466e-05, "loss": 0.0321, "step": 18520 }, { "grad_norm": 0.05199650302529335, "learning_rate": 6.0143412521575584e-05, "loss": 0.0318, "step": 18530 }, { "grad_norm": 0.04690839722752571, "learning_rate": 6.010293186882389e-05, "loss": 0.0315, "step": 18540 }, { "grad_norm": 0.04383102059364319, "learning_rate": 6.0062444310813525e-05, "loss": 0.0304, "step": 18550 }, { "grad_norm": 0.05257871747016907, "learning_rate": 6.0021949875217355e-05, "loss": 0.0285, "step": 18560 }, { "grad_norm": 0.041713058948516846, "learning_rate": 5.998144858971295e-05, "loss": 0.0269, "step": 18570 }, { "grad_norm": 0.041929472237825394, "learning_rate": 5.994094048198257e-05, "loss": 0.028, "step": 18580 }, { "grad_norm": 0.05975821614265442, "learning_rate": 5.990042557971307e-05, "loss": 0.0278, "step": 18590 }, { "grad_norm": 0.043046191334724426, "learning_rate": 5.985990391059607e-05, "loss": 0.0285, "step": 18600 }, { "grad_norm": 0.043772142380476, "learning_rate": 5.981937550232771e-05, "loss": 0.0261, "step": 18610 }, { "grad_norm": 0.053474877029657364, "learning_rate": 5.9778840382608794e-05, "loss": 0.0279, "step": 18620 }, { "grad_norm": 0.03650617226958275, "learning_rate": 5.9738298579144695e-05, "loss": 0.0306, "step": 18630 }, { "grad_norm": 0.034024450927972794, "learning_rate": 5.9697750119645314e-05, "loss": 0.0327, "step": 18640 }, { "grad_norm": 0.041432470083236694, "learning_rate": 5.96571950318252e-05, "loss": 0.0329, "step": 18650 }, { "grad_norm": 0.03976539894938469, "learning_rate": 5.9616633343403316e-05, "loss": 0.0321, "step": 18660 }, { "grad_norm": 0.038044627755880356, "learning_rate": 5.957606508210324e-05, "loss": 0.0316, "step": 18670 }, { "grad_norm": 0.03460857272148132, "learning_rate": 5.953549027565297e-05, "loss": 0.0306, "step": 18680 }, { "grad_norm": 0.049688033759593964, "learning_rate": 5.949490895178501e-05, "loss": 0.03, "step": 18690 }, { "grad_norm": 0.059325870126485825, "learning_rate": 5.945432113823632e-05, "loss": 0.0289, "step": 18700 }, { "grad_norm": 0.03831303119659424, "learning_rate": 5.9413726862748276e-05, "loss": 0.0288, "step": 18710 }, { "grad_norm": 0.04999135434627533, "learning_rate": 5.9373126153066694e-05, "loss": 0.0292, "step": 18720 }, { "grad_norm": 0.04923583194613457, "learning_rate": 5.933251903694177e-05, "loss": 0.0295, "step": 18730 }, { "grad_norm": 0.07236862927675247, "learning_rate": 5.929190554212807e-05, "loss": 0.0294, "step": 18740 }, { "grad_norm": 0.04675421863794327, "learning_rate": 5.9251285696384565e-05, "loss": 0.0309, "step": 18750 }, { "grad_norm": 0.051392536610364914, "learning_rate": 5.921065952747451e-05, "loss": 0.0328, "step": 18760 }, { "grad_norm": 0.054127808660268784, "learning_rate": 5.917002706316552e-05, "loss": 0.0354, "step": 18770 }, { "grad_norm": 0.043202172964811325, "learning_rate": 5.912938833122952e-05, "loss": 0.0336, "step": 18780 }, { "grad_norm": 0.04298568516969681, "learning_rate": 5.908874335944265e-05, "loss": 0.0341, "step": 18790 }, { "grad_norm": 0.04765614494681358, "learning_rate": 5.904809217558542e-05, "loss": 0.0347, "step": 18800 }, { "grad_norm": 0.0654081478714943, "learning_rate": 5.90074348074425e-05, "loss": 0.0326, "step": 18810 }, { "grad_norm": 0.03611113876104355, "learning_rate": 5.8966771282802814e-05, "loss": 0.0315, "step": 18820 }, { "grad_norm": 0.05974261090159416, "learning_rate": 5.892610162945952e-05, "loss": 0.0296, "step": 18830 }, { "grad_norm": 0.036404095590114594, "learning_rate": 5.8885425875209924e-05, "loss": 0.0284, "step": 18840 }, { "grad_norm": 0.043861620128154755, "learning_rate": 5.884474404785553e-05, "loss": 0.0282, "step": 18850 }, { "grad_norm": 0.046998172998428345, "learning_rate": 5.8804056175201983e-05, "loss": 0.0295, "step": 18860 }, { "grad_norm": 0.045899391174316406, "learning_rate": 5.876336228505904e-05, "loss": 0.0314, "step": 18870 }, { "grad_norm": 0.048973117023706436, "learning_rate": 5.872266240524062e-05, "loss": 0.0302, "step": 18880 }, { "grad_norm": 0.06810867041349411, "learning_rate": 5.86819565635647e-05, "loss": 0.0317, "step": 18890 }, { "grad_norm": 0.03794590011239052, "learning_rate": 5.8641244787853334e-05, "loss": 0.032, "step": 18900 }, { "grad_norm": 0.04352957755327225, "learning_rate": 5.860052710593265e-05, "loss": 0.0323, "step": 18910 }, { "grad_norm": 0.05754704773426056, "learning_rate": 5.855980354563276e-05, "loss": 0.0321, "step": 18920 }, { "grad_norm": 0.04218311235308647, "learning_rate": 5.8519074134787874e-05, "loss": 0.0319, "step": 18930 }, { "grad_norm": 0.0399458184838295, "learning_rate": 5.847833890123614e-05, "loss": 0.0319, "step": 18940 }, { "grad_norm": 0.06084258109331131, "learning_rate": 5.8437597872819737e-05, "loss": 0.0283, "step": 18950 }, { "grad_norm": 0.032540395855903625, "learning_rate": 5.839685107738473e-05, "loss": 0.0307, "step": 18960 }, { "grad_norm": 0.05360624939203262, "learning_rate": 5.835609854278118e-05, "loss": 0.0319, "step": 18970 }, { "grad_norm": 0.05570428818464279, "learning_rate": 5.831534029686308e-05, "loss": 0.0323, "step": 18980 }, { "grad_norm": 0.048653654754161835, "learning_rate": 5.82745763674883e-05, "loss": 0.0315, "step": 18990 }, { "grad_norm": 0.04473493993282318, "learning_rate": 5.823380678251861e-05, "loss": 0.0291, "step": 19000 }, { "grad_norm": 0.05221635475754738, "learning_rate": 5.81930315698196e-05, "loss": 0.0317, "step": 19010 }, { "grad_norm": 0.05149189382791519, "learning_rate": 5.815225075726076e-05, "loss": 0.0312, "step": 19020 }, { "grad_norm": 0.05194469541311264, "learning_rate": 5.811146437271543e-05, "loss": 0.0329, "step": 19030 }, { "grad_norm": 0.05809265747666359, "learning_rate": 5.807067244406066e-05, "loss": 0.0331, "step": 19040 }, { "grad_norm": 0.04396812990307808, "learning_rate": 5.8029874999177405e-05, "loss": 0.0298, "step": 19050 }, { "grad_norm": 0.06698248535394669, "learning_rate": 5.798907206595029e-05, "loss": 0.0306, "step": 19060 }, { "grad_norm": 0.06568489223718643, "learning_rate": 5.794826367226773e-05, "loss": 0.0298, "step": 19070 }, { "grad_norm": 0.031866107136011124, "learning_rate": 5.790744984602193e-05, "loss": 0.0341, "step": 19080 }, { "grad_norm": 0.04410490021109581, "learning_rate": 5.786663061510872e-05, "loss": 0.0342, "step": 19090 }, { "grad_norm": 0.06078626587986946, "learning_rate": 5.782580600742765e-05, "loss": 0.0337, "step": 19100 }, { "grad_norm": 0.05284452438354492, "learning_rate": 5.7784976050881965e-05, "loss": 0.0347, "step": 19110 }, { "grad_norm": 0.03422120586037636, "learning_rate": 5.774414077337855e-05, "loss": 0.0317, "step": 19120 }, { "grad_norm": 0.05188291147351265, "learning_rate": 5.770330020282796e-05, "loss": 0.0328, "step": 19130 }, { "grad_norm": 0.040726691484451294, "learning_rate": 5.7662454367144317e-05, "loss": 0.0305, "step": 19140 }, { "grad_norm": 0.06262613087892532, "learning_rate": 5.762160329424536e-05, "loss": 0.0295, "step": 19150 }, { "grad_norm": 0.04257950931787491, "learning_rate": 5.7580747012052416e-05, "loss": 0.0287, "step": 19160 }, { "grad_norm": 0.046790190041065216, "learning_rate": 5.753988554849037e-05, "loss": 0.0287, "step": 19170 }, { "grad_norm": 0.045041099190711975, "learning_rate": 5.749901893148766e-05, "loss": 0.0279, "step": 19180 }, { "grad_norm": 0.03672128915786743, "learning_rate": 5.745814718897621e-05, "loss": 0.0259, "step": 19190 }, { "grad_norm": 0.17255553603172302, "learning_rate": 5.74172703488915e-05, "loss": 0.03, "step": 19200 }, { "grad_norm": 0.05216333270072937, "learning_rate": 5.737638843917242e-05, "loss": 0.0293, "step": 19210 }, { "grad_norm": 0.04875454306602478, "learning_rate": 5.73355014877614e-05, "loss": 0.0306, "step": 19220 }, { "grad_norm": 0.0635644868016243, "learning_rate": 5.7294609522604316e-05, "loss": 0.0321, "step": 19230 }, { "grad_norm": 0.05837872624397278, "learning_rate": 5.7253712571650376e-05, "loss": 0.0294, "step": 19240 }, { "grad_norm": 0.037657253444194794, "learning_rate": 5.721281066285229e-05, "loss": 0.0302, "step": 19250 }, { "grad_norm": 0.040150485932826996, "learning_rate": 5.717190382416615e-05, "loss": 0.03, "step": 19260 }, { "grad_norm": 0.08215399086475372, "learning_rate": 5.713099208355135e-05, "loss": 0.032, "step": 19270 }, { "grad_norm": 0.044364385306835175, "learning_rate": 5.709007546897074e-05, "loss": 0.0324, "step": 19280 }, { "grad_norm": 0.056025996804237366, "learning_rate": 5.704915400839037e-05, "loss": 0.0311, "step": 19290 }, { "grad_norm": 0.051080070436000824, "learning_rate": 5.700822772977971e-05, "loss": 0.0341, "step": 19300 }, { "grad_norm": 0.041884880512952805, "learning_rate": 5.696729666111148e-05, "loss": 0.0329, "step": 19310 }, { "grad_norm": 0.05056530609726906, "learning_rate": 5.692636083036168e-05, "loss": 0.0308, "step": 19320 }, { "grad_norm": 0.0814768448472023, "learning_rate": 5.688542026550958e-05, "loss": 0.0315, "step": 19330 }, { "grad_norm": 0.05539041385054588, "learning_rate": 5.684447499453763e-05, "loss": 0.0304, "step": 19340 }, { "grad_norm": 0.04368724301457405, "learning_rate": 5.680352504543156e-05, "loss": 0.0317, "step": 19350 }, { "grad_norm": 0.03873738273978233, "learning_rate": 5.67625704461803e-05, "loss": 0.0288, "step": 19360 }, { "grad_norm": 0.04533078148961067, "learning_rate": 5.672161122477589e-05, "loss": 0.0331, "step": 19370 }, { "grad_norm": 0.043138302862644196, "learning_rate": 5.668064740921359e-05, "loss": 0.0319, "step": 19380 }, { "grad_norm": 0.04285112023353577, "learning_rate": 5.663967902749179e-05, "loss": 0.0309, "step": 19390 }, { "grad_norm": 0.0836973786354065, "learning_rate": 5.6598706107611965e-05, "loss": 0.0307, "step": 19400 }, { "grad_norm": 0.04045240208506584, "learning_rate": 5.655772867757876e-05, "loss": 0.0339, "step": 19410 }, { "grad_norm": 0.04822771996259689, "learning_rate": 5.651674676539982e-05, "loss": 0.0317, "step": 19420 }, { "grad_norm": 0.04485323280096054, "learning_rate": 5.647576039908593e-05, "loss": 0.0318, "step": 19430 }, { "grad_norm": 0.051114264875650406, "learning_rate": 5.6434769606650864e-05, "loss": 0.0339, "step": 19440 }, { "grad_norm": 0.049366679042577744, "learning_rate": 5.639377441611143e-05, "loss": 0.0318, "step": 19450 }, { "grad_norm": 0.05105746164917946, "learning_rate": 5.635277485548751e-05, "loss": 0.0301, "step": 19460 }, { "grad_norm": 0.05901961028575897, "learning_rate": 5.631177095280186e-05, "loss": 0.0304, "step": 19470 }, { "grad_norm": 0.03931911289691925, "learning_rate": 5.627076273608027e-05, "loss": 0.0267, "step": 19480 }, { "grad_norm": 0.06202427297830582, "learning_rate": 5.622975023335148e-05, "loss": 0.0284, "step": 19490 }, { "grad_norm": 0.03698698431253433, "learning_rate": 5.618873347264716e-05, "loss": 0.0282, "step": 19500 }, { "grad_norm": 0.038017209619283676, "learning_rate": 5.614771248200188e-05, "loss": 0.0268, "step": 19510 }, { "grad_norm": 0.03065378963947296, "learning_rate": 5.6106687289453066e-05, "loss": 0.0272, "step": 19520 }, { "grad_norm": 0.04229613393545151, "learning_rate": 5.606565792304108e-05, "loss": 0.0266, "step": 19530 }, { "grad_norm": 0.04667752608656883, "learning_rate": 5.602462441080909e-05, "loss": 0.0267, "step": 19540 }, { "grad_norm": 0.058514244854450226, "learning_rate": 5.5983586780803135e-05, "loss": 0.0282, "step": 19550 }, { "grad_norm": 0.03940296918153763, "learning_rate": 5.594254506107205e-05, "loss": 0.0277, "step": 19560 }, { "grad_norm": 0.050060659646987915, "learning_rate": 5.590149927966743e-05, "loss": 0.0279, "step": 19570 }, { "grad_norm": 0.05385391414165497, "learning_rate": 5.58604494646437e-05, "loss": 0.0286, "step": 19580 }, { "grad_norm": 0.13639898598194122, "learning_rate": 5.5819395644058025e-05, "loss": 0.0311, "step": 19590 }, { "grad_norm": 0.06761667877435684, "learning_rate": 5.577833784597031e-05, "loss": 0.033, "step": 19600 }, { "grad_norm": 0.035836607217788696, "learning_rate": 5.573727609844316e-05, "loss": 0.0331, "step": 19610 }, { "grad_norm": 0.03971244394779205, "learning_rate": 5.5696210429541884e-05, "loss": 0.0328, "step": 19620 }, { "grad_norm": 0.05190500617027283, "learning_rate": 5.565514086733451e-05, "loss": 0.0336, "step": 19630 }, { "grad_norm": 0.03778998181223869, "learning_rate": 5.5614067439891657e-05, "loss": 0.0319, "step": 19640 }, { "grad_norm": 0.05142028629779816, "learning_rate": 5.557299017528666e-05, "loss": 0.0293, "step": 19650 }, { "grad_norm": 0.043339669704437256, "learning_rate": 5.5531909101595436e-05, "loss": 0.0325, "step": 19660 }, { "grad_norm": 0.04041208326816559, "learning_rate": 5.549082424689649e-05, "loss": 0.0297, "step": 19670 }, { "grad_norm": 0.054637644439935684, "learning_rate": 5.544973563927095e-05, "loss": 0.0304, "step": 19680 }, { "grad_norm": 0.05996892973780632, "learning_rate": 5.540864330680249e-05, "loss": 0.0312, "step": 19690 }, { "grad_norm": 0.04511360451579094, "learning_rate": 5.536754727757733e-05, "loss": 0.0297, "step": 19700 }, { "grad_norm": 0.04630286619067192, "learning_rate": 5.532644757968422e-05, "loss": 0.0318, "step": 19710 }, { "grad_norm": 0.07282694429159164, "learning_rate": 5.528534424121441e-05, "loss": 0.0314, "step": 19720 }, { "grad_norm": 0.042260460555553436, "learning_rate": 5.524423729026165e-05, "loss": 0.0332, "step": 19730 }, { "grad_norm": 0.04595079645514488, "learning_rate": 5.5203126754922164e-05, "loss": 0.0318, "step": 19740 }, { "grad_norm": 0.04522867873311043, "learning_rate": 5.5162012663294585e-05, "loss": 0.0301, "step": 19750 }, { "grad_norm": 0.047230254858732224, "learning_rate": 5.512089504348003e-05, "loss": 0.0312, "step": 19760 }, { "grad_norm": 0.04319560527801514, "learning_rate": 5.5079773923582e-05, "loss": 0.0318, "step": 19770 }, { "grad_norm": 0.04188758134841919, "learning_rate": 5.50386493317064e-05, "loss": 0.0299, "step": 19780 }, { "grad_norm": 0.04697016626596451, "learning_rate": 5.49975212959615e-05, "loss": 0.0313, "step": 19790 }, { "grad_norm": 0.049635857343673706, "learning_rate": 5.4956389844457904e-05, "loss": 0.0292, "step": 19800 }, { "grad_norm": 0.05143073573708534, "learning_rate": 5.491525500530859e-05, "loss": 0.0304, "step": 19810 }, { "grad_norm": 0.04676904156804085, "learning_rate": 5.487411680662882e-05, "loss": 0.029, "step": 19820 }, { "grad_norm": 0.03944479301571846, "learning_rate": 5.483297527653618e-05, "loss": 0.029, "step": 19830 }, { "grad_norm": 0.049819521605968475, "learning_rate": 5.4791830443150516e-05, "loss": 0.0296, "step": 19840 }, { "grad_norm": 0.0509910061955452, "learning_rate": 5.475068233459392e-05, "loss": 0.0337, "step": 19850 }, { "grad_norm": 0.03964246064424515, "learning_rate": 5.470953097899075e-05, "loss": 0.0283, "step": 19860 }, { "grad_norm": 0.05676116794347763, "learning_rate": 5.466837640446756e-05, "loss": 0.0286, "step": 19870 }, { "grad_norm": 0.04618774726986885, "learning_rate": 5.462721863915312e-05, "loss": 0.0291, "step": 19880 }, { "grad_norm": 0.09263963252305984, "learning_rate": 5.4586057711178374e-05, "loss": 0.0282, "step": 19890 }, { "grad_norm": 0.04335395246744156, "learning_rate": 5.454489364867642e-05, "loss": 0.0296, "step": 19900 }, { "grad_norm": 0.04800422117114067, "learning_rate": 5.4503726479782523e-05, "loss": 0.0301, "step": 19910 }, { "grad_norm": 0.04554368555545807, "learning_rate": 5.446255623263403e-05, "loss": 0.0274, "step": 19920 }, { "grad_norm": 0.03626318275928497, "learning_rate": 5.4421382935370445e-05, "loss": 0.028, "step": 19930 }, { "grad_norm": 0.04511221498250961, "learning_rate": 5.438020661613331e-05, "loss": 0.0286, "step": 19940 }, { "grad_norm": 0.05151604115962982, "learning_rate": 5.433902730306625e-05, "loss": 0.0286, "step": 19950 }, { "grad_norm": 0.04001368582248688, "learning_rate": 5.429784502431495e-05, "loss": 0.0277, "step": 19960 }, { "grad_norm": 0.04373238980770111, "learning_rate": 5.42566598080271e-05, "loss": 0.0284, "step": 19970 }, { "grad_norm": 0.09376242011785507, "learning_rate": 5.421547168235241e-05, "loss": 0.0305, "step": 19980 }, { "grad_norm": 0.047925256192684174, "learning_rate": 5.417428067544258e-05, "loss": 0.0313, "step": 19990 }, { "grad_norm": 0.048194967210292816, "learning_rate": 5.413308681545126e-05, "loss": 0.0326, "step": 20000 }, { "grad_norm": 0.06326395273208618, "learning_rate": 5.409189013053408e-05, "loss": 0.0304, "step": 20010 }, { "grad_norm": 0.06035089120268822, "learning_rate": 5.4050690648848576e-05, "loss": 0.0308, "step": 20020 }, { "grad_norm": 0.05045952647924423, "learning_rate": 5.400948839855421e-05, "loss": 0.0328, "step": 20030 }, { "grad_norm": 0.041329558938741684, "learning_rate": 5.396828340781234e-05, "loss": 0.0329, "step": 20040 }, { "grad_norm": 0.05221949890255928, "learning_rate": 5.392707570478617e-05, "loss": 0.0318, "step": 20050 }, { "grad_norm": 0.056643493473529816, "learning_rate": 5.388586531764078e-05, "loss": 0.0288, "step": 20060 }, { "grad_norm": 0.03681754320859909, "learning_rate": 5.384465227454311e-05, "loss": 0.0308, "step": 20070 }, { "grad_norm": 0.04412136971950531, "learning_rate": 5.380343660366184e-05, "loss": 0.03, "step": 20080 }, { "grad_norm": 0.0483984649181366, "learning_rate": 5.376221833316752e-05, "loss": 0.0317, "step": 20090 }, { "grad_norm": 0.04459225758910179, "learning_rate": 5.3720997491232436e-05, "loss": 0.0346, "step": 20100 }, { "grad_norm": 0.05790067836642265, "learning_rate": 5.367977410603068e-05, "loss": 0.0328, "step": 20110 }, { "grad_norm": 0.04096110165119171, "learning_rate": 5.3638548205738004e-05, "loss": 0.0337, "step": 20120 }, { "grad_norm": 0.05499361827969551, "learning_rate": 5.359731981853194e-05, "loss": 0.0342, "step": 20130 }, { "grad_norm": 0.03420053794980049, "learning_rate": 5.35560889725917e-05, "loss": 0.0334, "step": 20140 }, { "grad_norm": 0.04115643352270126, "learning_rate": 5.3514855696098176e-05, "loss": 0.0332, "step": 20150 }, { "grad_norm": 0.04052659869194031, "learning_rate": 5.347362001723394e-05, "loss": 0.0302, "step": 20160 }, { "grad_norm": 0.04120044410228729, "learning_rate": 5.3432381964183176e-05, "loss": 0.0288, "step": 20170 }, { "grad_norm": 0.04746266454458237, "learning_rate": 5.3391141565131685e-05, "loss": 0.0287, "step": 20180 }, { "grad_norm": 0.046862971037626266, "learning_rate": 5.3349898848266935e-05, "loss": 0.0291, "step": 20190 }, { "grad_norm": 0.0510731041431427, "learning_rate": 5.330865384177789e-05, "loss": 0.0271, "step": 20200 }, { "grad_norm": 0.038731206208467484, "learning_rate": 5.326740657385515e-05, "loss": 0.0284, "step": 20210 }, { "grad_norm": 0.050515316426754, "learning_rate": 5.322615707269083e-05, "loss": 0.0308, "step": 20220 }, { "grad_norm": 0.04340570792555809, "learning_rate": 5.318490536647856e-05, "loss": 0.0301, "step": 20230 }, { "grad_norm": 0.04294625669717789, "learning_rate": 5.3143651483413524e-05, "loss": 0.0302, "step": 20240 }, { "grad_norm": 0.04304599016904831, "learning_rate": 5.310239545169232e-05, "loss": 0.0319, "step": 20250 }, { "grad_norm": 0.033398766070604324, "learning_rate": 5.30611372995131e-05, "loss": 0.0315, "step": 20260 }, { "grad_norm": 0.03717504069209099, "learning_rate": 5.30198770550754e-05, "loss": 0.032, "step": 20270 }, { "grad_norm": 0.04450028017163277, "learning_rate": 5.297861474658019e-05, "loss": 0.0296, "step": 20280 }, { "grad_norm": 0.06236167997121811, "learning_rate": 5.29373504022299e-05, "loss": 0.0298, "step": 20290 }, { "grad_norm": 0.051889270544052124, "learning_rate": 5.28960840502283e-05, "loss": 0.0306, "step": 20300 }, { "grad_norm": 0.043807778507471085, "learning_rate": 5.285481571878056e-05, "loss": 0.033, "step": 20310 }, { "grad_norm": 0.04036693274974823, "learning_rate": 5.281354543609321e-05, "loss": 0.0309, "step": 20320 }, { "grad_norm": 0.0461563840508461, "learning_rate": 5.277227323037406e-05, "loss": 0.0315, "step": 20330 }, { "grad_norm": 0.052095551043748856, "learning_rate": 5.273099912983233e-05, "loss": 0.0303, "step": 20340 }, { "grad_norm": 0.060990240424871445, "learning_rate": 5.268972316267843e-05, "loss": 0.0334, "step": 20350 }, { "grad_norm": 0.054527074098587036, "learning_rate": 5.26484453571241e-05, "loss": 0.0328, "step": 20360 }, { "grad_norm": 0.03865246847271919, "learning_rate": 5.260716574138235e-05, "loss": 0.0304, "step": 20370 }, { "grad_norm": 0.04973999783396721, "learning_rate": 5.256588434366739e-05, "loss": 0.0313, "step": 20380 }, { "grad_norm": 0.04549003392457962, "learning_rate": 5.25246011921947e-05, "loss": 0.0301, "step": 20390 }, { "grad_norm": 0.04429002106189728, "learning_rate": 5.248331631518089e-05, "loss": 0.0287, "step": 20400 }, { "grad_norm": 0.04978109896183014, "learning_rate": 5.244202974084379e-05, "loss": 0.029, "step": 20410 }, { "grad_norm": 0.05241375043988228, "learning_rate": 5.240074149740239e-05, "loss": 0.0278, "step": 20420 }, { "grad_norm": 0.034102167934179306, "learning_rate": 5.2359451613076814e-05, "loss": 0.0274, "step": 20430 }, { "grad_norm": 0.037238024175167084, "learning_rate": 5.231816011608832e-05, "loss": 0.0302, "step": 20440 }, { "grad_norm": 0.03519711270928383, "learning_rate": 5.227686703465924e-05, "loss": 0.0287, "step": 20450 }, { "grad_norm": 0.05027427151799202, "learning_rate": 5.2235572397013e-05, "loss": 0.03, "step": 20460 }, { "grad_norm": 0.04080960527062416, "learning_rate": 5.2194276231374114e-05, "loss": 0.0303, "step": 20470 }, { "grad_norm": 0.04680290445685387, "learning_rate": 5.21529785659681e-05, "loss": 0.0309, "step": 20480 }, { "grad_norm": 0.05769123509526253, "learning_rate": 5.2111679429021565e-05, "loss": 0.0291, "step": 20490 }, { "grad_norm": 0.054365839809179306, "learning_rate": 5.207037884876205e-05, "loss": 0.0306, "step": 20500 }, { "grad_norm": 0.035702720284461975, "learning_rate": 5.202907685341809e-05, "loss": 0.0288, "step": 20510 }, { "grad_norm": 0.03894752264022827, "learning_rate": 5.198777347121926e-05, "loss": 0.0277, "step": 20520 }, { "grad_norm": 0.038838163018226624, "learning_rate": 5.194646873039598e-05, "loss": 0.0278, "step": 20530 }, { "grad_norm": 0.07885216921567917, "learning_rate": 5.1905162659179696e-05, "loss": 0.0292, "step": 20540 }, { "grad_norm": 0.05454135313630104, "learning_rate": 5.18638552858027e-05, "loss": 0.0316, "step": 20550 }, { "grad_norm": 0.05329703539609909, "learning_rate": 5.182254663849818e-05, "loss": 0.0339, "step": 20560 }, { "grad_norm": 0.04187990725040436, "learning_rate": 5.178123674550023e-05, "loss": 0.0317, "step": 20570 }, { "grad_norm": 0.04941069334745407, "learning_rate": 5.173992563504375e-05, "loss": 0.031, "step": 20580 }, { "grad_norm": 0.17159703373908997, "learning_rate": 5.169861333536451e-05, "loss": 0.034, "step": 20590 }, { "grad_norm": 0.0390453077852726, "learning_rate": 5.165729987469907e-05, "loss": 0.0317, "step": 20600 }, { "grad_norm": 0.03811768442392349, "learning_rate": 5.161598528128478e-05, "loss": 0.03, "step": 20610 }, { "grad_norm": 0.052078478038311005, "learning_rate": 5.157466958335981e-05, "loss": 0.0289, "step": 20620 }, { "grad_norm": 0.046267926692962646, "learning_rate": 5.1533352809163025e-05, "loss": 0.0288, "step": 20630 }, { "grad_norm": 0.04136187583208084, "learning_rate": 5.1492034986934046e-05, "loss": 0.0284, "step": 20640 }, { "grad_norm": 0.07005064189434052, "learning_rate": 5.1450716144913225e-05, "loss": 0.0316, "step": 20650 }, { "grad_norm": 0.04849500209093094, "learning_rate": 5.1409396311341595e-05, "loss": 0.0276, "step": 20660 }, { "grad_norm": 0.050885576754808426, "learning_rate": 5.136807551446089e-05, "loss": 0.0286, "step": 20670 }, { "grad_norm": 0.04156658798456192, "learning_rate": 5.132675378251346e-05, "loss": 0.0271, "step": 20680 }, { "grad_norm": 0.03702140226960182, "learning_rate": 5.1285431143742325e-05, "loss": 0.027, "step": 20690 }, { "grad_norm": 0.05691608786582947, "learning_rate": 5.1244107626391136e-05, "loss": 0.0291, "step": 20700 }, { "grad_norm": 0.045250505208969116, "learning_rate": 5.12027832587041e-05, "loss": 0.0258, "step": 20710 }, { "grad_norm": 0.048142530024051666, "learning_rate": 5.116145806892607e-05, "loss": 0.0279, "step": 20720 }, { "grad_norm": 0.04459746554493904, "learning_rate": 5.1120132085302384e-05, "loss": 0.027, "step": 20730 }, { "grad_norm": 0.04073881357908249, "learning_rate": 5.107880533607898e-05, "loss": 0.032, "step": 20740 }, { "grad_norm": 0.042392365634441376, "learning_rate": 5.103747784950231e-05, "loss": 0.0295, "step": 20750 }, { "grad_norm": 0.06399226933717728, "learning_rate": 5.09961496538193e-05, "loss": 0.0288, "step": 20760 }, { "grad_norm": 0.04945513606071472, "learning_rate": 5.095482077727742e-05, "loss": 0.0293, "step": 20770 }, { "grad_norm": 0.04487373307347298, "learning_rate": 5.091349124812452e-05, "loss": 0.0305, "step": 20780 }, { "grad_norm": 0.03840511664748192, "learning_rate": 5.087216109460897e-05, "loss": 0.0317, "step": 20790 }, { "grad_norm": 0.04555080831050873, "learning_rate": 5.083083034497954e-05, "loss": 0.0293, "step": 20800 }, { "grad_norm": 0.03385312855243683, "learning_rate": 5.07894990274854e-05, "loss": 0.0312, "step": 20810 }, { "grad_norm": 0.06125837191939354, "learning_rate": 5.074816717037614e-05, "loss": 0.0302, "step": 20820 }, { "grad_norm": 0.04960164055228233, "learning_rate": 5.070683480190165e-05, "loss": 0.032, "step": 20830 }, { "grad_norm": 0.041988588869571686, "learning_rate": 5.066550195031223e-05, "loss": 0.0291, "step": 20840 }, { "grad_norm": 0.04314441606402397, "learning_rate": 5.062416864385852e-05, "loss": 0.0313, "step": 20850 }, { "grad_norm": 0.05292556807398796, "learning_rate": 5.058283491079142e-05, "loss": 0.0287, "step": 20860 }, { "grad_norm": 0.05499349161982536, "learning_rate": 5.054150077936216e-05, "loss": 0.0295, "step": 20870 }, { "grad_norm": 0.0485992394387722, "learning_rate": 5.0500166277822214e-05, "loss": 0.0303, "step": 20880 }, { "grad_norm": 0.04439699649810791, "learning_rate": 5.0458831434423334e-05, "loss": 0.0314, "step": 20890 }, { "grad_norm": 0.04513491317629814, "learning_rate": 5.0417496277417506e-05, "loss": 0.0301, "step": 20900 }, { "grad_norm": 0.042851559817790985, "learning_rate": 5.037616083505691e-05, "loss": 0.0277, "step": 20910 }, { "grad_norm": 0.04675484821200371, "learning_rate": 5.0334825135593935e-05, "loss": 0.03, "step": 20920 }, { "grad_norm": 0.05886365473270416, "learning_rate": 5.029348920728111e-05, "loss": 0.0305, "step": 20930 }, { "grad_norm": 0.04669820889830589, "learning_rate": 5.0252153078371186e-05, "loss": 0.0293, "step": 20940 }, { "grad_norm": 0.0453062430024147, "learning_rate": 5.021081677711704e-05, "loss": 0.0322, "step": 20950 }, { "grad_norm": 0.04389895126223564, "learning_rate": 5.016948033177159e-05, "loss": 0.0277, "step": 20960 }, { "grad_norm": 0.045422930270433426, "learning_rate": 5.012814377058793e-05, "loss": 0.0313, "step": 20970 }, { "grad_norm": 0.04461435228586197, "learning_rate": 5.008680712181921e-05, "loss": 0.0303, "step": 20980 }, { "grad_norm": 0.05696148797869682, "learning_rate": 5.0045470413718645e-05, "loss": 0.0296, "step": 20990 }, { "grad_norm": 0.05820119008421898, "learning_rate": 5.00041336745395e-05, "loss": 0.0256, "step": 21000 }, { "grad_norm": 0.043037936091423035, "learning_rate": 4.996279693253499e-05, "loss": 0.0274, "step": 21010 }, { "grad_norm": 0.05910643935203552, "learning_rate": 4.992146021595847e-05, "loss": 0.0263, "step": 21020 }, { "grad_norm": 0.03967222571372986, "learning_rate": 4.988012355306313e-05, "loss": 0.0282, "step": 21030 }, { "grad_norm": 0.051408551633358, "learning_rate": 4.98387869721022e-05, "loss": 0.0306, "step": 21040 }, { "grad_norm": 0.04795463755726814, "learning_rate": 4.9797450501328866e-05, "loss": 0.028, "step": 21050 }, { "grad_norm": 0.04773436114192009, "learning_rate": 4.97561141689962e-05, "loss": 0.0326, "step": 21060 }, { "grad_norm": 0.04401761293411255, "learning_rate": 4.971477800335721e-05, "loss": 0.0299, "step": 21070 }, { "grad_norm": 0.05146123468875885, "learning_rate": 4.967344203266475e-05, "loss": 0.0311, "step": 21080 }, { "grad_norm": 0.035096630454063416, "learning_rate": 4.9632106285171584e-05, "loss": 0.0305, "step": 21090 }, { "grad_norm": 0.05734769627451897, "learning_rate": 4.959077078913031e-05, "loss": 0.0307, "step": 21100 }, { "grad_norm": 0.031226443126797676, "learning_rate": 4.954943557279333e-05, "loss": 0.0276, "step": 21110 }, { "grad_norm": 0.04077035188674927, "learning_rate": 4.9508100664412916e-05, "loss": 0.0269, "step": 21120 }, { "grad_norm": 0.0380622036755085, "learning_rate": 4.946676609224105e-05, "loss": 0.0294, "step": 21130 }, { "grad_norm": 0.04343811795115471, "learning_rate": 4.942543188452952e-05, "loss": 0.028, "step": 21140 }, { "grad_norm": 0.04217015579342842, "learning_rate": 4.938409806952988e-05, "loss": 0.0288, "step": 21150 }, { "grad_norm": 0.040499281138181686, "learning_rate": 4.93427646754934e-05, "loss": 0.0279, "step": 21160 }, { "grad_norm": 0.054571621119976044, "learning_rate": 4.930143173067108e-05, "loss": 0.027, "step": 21170 }, { "grad_norm": 0.04189007356762886, "learning_rate": 4.9260099263313565e-05, "loss": 0.0256, "step": 21180 }, { "grad_norm": 0.03988037258386612, "learning_rate": 4.921876730167123e-05, "loss": 0.0318, "step": 21190 }, { "grad_norm": 0.045528363436460495, "learning_rate": 4.917743587399409e-05, "loss": 0.0277, "step": 21200 }, { "grad_norm": 0.03936249017715454, "learning_rate": 4.913610500853178e-05, "loss": 0.0286, "step": 21210 }, { "grad_norm": 0.05129941925406456, "learning_rate": 4.909477473353354e-05, "loss": 0.0289, "step": 21220 }, { "grad_norm": 0.040182165801525116, "learning_rate": 4.9053445077248236e-05, "loss": 0.0286, "step": 21230 }, { "grad_norm": 0.03571214899420738, "learning_rate": 4.901211606792429e-05, "loss": 0.029, "step": 21240 }, { "grad_norm": 0.07589086890220642, "learning_rate": 4.89707877338097e-05, "loss": 0.028, "step": 21250 }, { "grad_norm": 0.042823147028684616, "learning_rate": 4.892946010315199e-05, "loss": 0.0265, "step": 21260 }, { "grad_norm": 0.060963813215494156, "learning_rate": 4.8888133204198204e-05, "loss": 0.0287, "step": 21270 }, { "grad_norm": 0.049677230417728424, "learning_rate": 4.8846807065194886e-05, "loss": 0.0266, "step": 21280 }, { "grad_norm": 0.0451795794069767, "learning_rate": 4.880548171438806e-05, "loss": 0.0266, "step": 21290 }, { "grad_norm": 0.04832972213625908, "learning_rate": 4.8764157180023245e-05, "loss": 0.0266, "step": 21300 }, { "grad_norm": 0.039354074746370316, "learning_rate": 4.872283349034533e-05, "loss": 0.0275, "step": 21310 }, { "grad_norm": 0.04690023511648178, "learning_rate": 4.8681510673598674e-05, "loss": 0.0284, "step": 21320 }, { "grad_norm": 0.04789527878165245, "learning_rate": 4.8640188758027046e-05, "loss": 0.0266, "step": 21330 }, { "grad_norm": 0.04200510308146477, "learning_rate": 4.859886777187357e-05, "loss": 0.0281, "step": 21340 }, { "grad_norm": 0.03513111546635628, "learning_rate": 4.855754774338077e-05, "loss": 0.0285, "step": 21350 }, { "grad_norm": 0.04505123570561409, "learning_rate": 4.851622870079048e-05, "loss": 0.0272, "step": 21360 }, { "grad_norm": 0.051604051142930984, "learning_rate": 4.847491067234389e-05, "loss": 0.0292, "step": 21370 }, { "grad_norm": 0.047741472721099854, "learning_rate": 4.843359368628146e-05, "loss": 0.0274, "step": 21380 }, { "grad_norm": 0.04954284057021141, "learning_rate": 4.8392277770842975e-05, "loss": 0.029, "step": 21390 }, { "grad_norm": 0.034857090562582016, "learning_rate": 4.83509629542675e-05, "loss": 0.0285, "step": 21400 }, { "grad_norm": 0.04712745547294617, "learning_rate": 4.830964926479329e-05, "loss": 0.0279, "step": 21410 }, { "grad_norm": 0.047309283167123795, "learning_rate": 4.826833673065785e-05, "loss": 0.0281, "step": 21420 }, { "grad_norm": 0.0442592017352581, "learning_rate": 4.822702538009794e-05, "loss": 0.0323, "step": 21430 }, { "grad_norm": 0.04897332563996315, "learning_rate": 4.818571524134945e-05, "loss": 0.0333, "step": 21440 }, { "grad_norm": 0.06492806226015091, "learning_rate": 4.8144406342647496e-05, "loss": 0.0348, "step": 21450 }, { "grad_norm": 0.03130710497498512, "learning_rate": 4.81030987122263e-05, "loss": 0.0332, "step": 21460 }, { "grad_norm": 0.06669390201568604, "learning_rate": 4.806179237831926e-05, "loss": 0.0319, "step": 21470 }, { "grad_norm": 0.0412592738866806, "learning_rate": 4.802048736915884e-05, "loss": 0.031, "step": 21480 }, { "grad_norm": 0.05423641577363014, "learning_rate": 4.797918371297666e-05, "loss": 0.0325, "step": 21490 }, { "grad_norm": 0.03736589848995209, "learning_rate": 4.793788143800334e-05, "loss": 0.0303, "step": 21500 }, { "grad_norm": 0.07421982288360596, "learning_rate": 4.789658057246862e-05, "loss": 0.0323, "step": 21510 }, { "grad_norm": 0.04316036403179169, "learning_rate": 4.7855281144601227e-05, "loss": 0.0322, "step": 21520 }, { "grad_norm": 0.05149804800748825, "learning_rate": 4.781398318262897e-05, "loss": 0.0302, "step": 21530 }, { "grad_norm": 0.05126727372407913, "learning_rate": 4.777268671477858e-05, "loss": 0.0316, "step": 21540 }, { "grad_norm": 0.03240153566002846, "learning_rate": 4.773139176927582e-05, "loss": 0.0311, "step": 21550 }, { "grad_norm": 0.061497338116168976, "learning_rate": 4.769009837434539e-05, "loss": 0.0307, "step": 21560 }, { "grad_norm": 0.034271251410245895, "learning_rate": 4.764880655821095e-05, "loss": 0.0318, "step": 21570 }, { "grad_norm": 0.053608402609825134, "learning_rate": 4.760751634909508e-05, "loss": 0.0325, "step": 21580 }, { "grad_norm": 0.040913473814725876, "learning_rate": 4.756622777521919e-05, "loss": 0.0314, "step": 21590 }, { "grad_norm": 0.05142272263765335, "learning_rate": 4.752494086480368e-05, "loss": 0.0314, "step": 21600 }, { "grad_norm": 0.03351190686225891, "learning_rate": 4.7483655646067744e-05, "loss": 0.0318, "step": 21610 }, { "grad_norm": 0.05211910232901573, "learning_rate": 4.744237214722944e-05, "loss": 0.0309, "step": 21620 }, { "grad_norm": 0.04261816293001175, "learning_rate": 4.740109039650567e-05, "loss": 0.0326, "step": 21630 }, { "grad_norm": 0.049975134432315826, "learning_rate": 4.73598104221121e-05, "loss": 0.032, "step": 21640 }, { "grad_norm": 0.046403512358665466, "learning_rate": 4.731853225226322e-05, "loss": 0.0292, "step": 21650 }, { "grad_norm": 0.04179666191339493, "learning_rate": 4.727725591517225e-05, "loss": 0.0314, "step": 21660 }, { "grad_norm": 0.046443160623311996, "learning_rate": 4.723598143905119e-05, "loss": 0.0311, "step": 21670 }, { "grad_norm": 0.0473296083509922, "learning_rate": 4.719470885211077e-05, "loss": 0.0295, "step": 21680 }, { "grad_norm": 0.03931425139307976, "learning_rate": 4.7153438182560387e-05, "loss": 0.0293, "step": 21690 }, { "grad_norm": 0.06635397672653198, "learning_rate": 4.711216945860815e-05, "loss": 0.0272, "step": 21700 }, { "grad_norm": 0.047685276716947556, "learning_rate": 4.707090270846088e-05, "loss": 0.0269, "step": 21710 }, { "grad_norm": 0.05256010219454765, "learning_rate": 4.702963796032397e-05, "loss": 0.0269, "step": 21720 }, { "grad_norm": 0.04429784417152405, "learning_rate": 4.6988375242401514e-05, "loss": 0.0266, "step": 21730 }, { "grad_norm": 0.031934741884469986, "learning_rate": 4.694711458289618e-05, "loss": 0.0267, "step": 21740 }, { "grad_norm": 0.055614545941352844, "learning_rate": 4.690585601000925e-05, "loss": 0.0282, "step": 21750 }, { "grad_norm": 0.04213646799325943, "learning_rate": 4.686459955194055e-05, "loss": 0.0297, "step": 21760 }, { "grad_norm": 0.07684100419282913, "learning_rate": 4.6823345236888504e-05, "loss": 0.0308, "step": 21770 }, { "grad_norm": 0.050804607570171356, "learning_rate": 4.678209309305002e-05, "loss": 0.0308, "step": 21780 }, { "grad_norm": 0.039800338447093964, "learning_rate": 4.674084314862057e-05, "loss": 0.028, "step": 21790 }, { "grad_norm": 0.05610353499650955, "learning_rate": 4.669959543179409e-05, "loss": 0.0325, "step": 21800 }, { "grad_norm": 0.031087247654795647, "learning_rate": 4.665834997076303e-05, "loss": 0.0305, "step": 21810 }, { "grad_norm": 0.04955196753144264, "learning_rate": 4.661710679371823e-05, "loss": 0.0315, "step": 21820 }, { "grad_norm": 0.046836793422698975, "learning_rate": 4.657586592884905e-05, "loss": 0.0305, "step": 21830 }, { "grad_norm": 0.05155163258314133, "learning_rate": 4.653462740434322e-05, "loss": 0.0303, "step": 21840 }, { "grad_norm": 0.04006950184702873, "learning_rate": 4.649339124838689e-05, "loss": 0.0312, "step": 21850 }, { "grad_norm": 0.03892180696129799, "learning_rate": 4.6452157489164574e-05, "loss": 0.03, "step": 21860 }, { "grad_norm": 0.045075155794620514, "learning_rate": 4.6410926154859155e-05, "loss": 0.029, "step": 21870 }, { "grad_norm": 0.03620024025440216, "learning_rate": 4.636969727365186e-05, "loss": 0.0287, "step": 21880 }, { "grad_norm": 0.03729855269193649, "learning_rate": 4.632847087372226e-05, "loss": 0.0283, "step": 21890 }, { "grad_norm": 0.058796461671590805, "learning_rate": 4.628724698324818e-05, "loss": 0.0294, "step": 21900 }, { "grad_norm": 0.04552392661571503, "learning_rate": 4.6246025630405795e-05, "loss": 0.0279, "step": 21910 }, { "grad_norm": 0.04604068398475647, "learning_rate": 4.6204806843369474e-05, "loss": 0.0286, "step": 21920 }, { "grad_norm": 0.05985940247774124, "learning_rate": 4.616359065031191e-05, "loss": 0.0288, "step": 21930 }, { "grad_norm": 0.037939418107271194, "learning_rate": 4.6122377079403946e-05, "loss": 0.0286, "step": 21940 }, { "grad_norm": 0.03722342476248741, "learning_rate": 4.6081166158814695e-05, "loss": 0.0303, "step": 21950 }, { "grad_norm": 0.04720357432961464, "learning_rate": 4.603995791671144e-05, "loss": 0.0286, "step": 21960 }, { "grad_norm": 0.03378818929195404, "learning_rate": 4.599875238125957e-05, "loss": 0.0275, "step": 21970 }, { "grad_norm": 0.04668780416250229, "learning_rate": 4.595754958062273e-05, "loss": 0.0272, "step": 21980 }, { "grad_norm": 0.028961339965462685, "learning_rate": 4.591634954296265e-05, "loss": 0.0292, "step": 21990 }, { "grad_norm": 0.04912031441926956, "learning_rate": 4.587515229643913e-05, "loss": 0.0278, "step": 22000 }, { "grad_norm": 0.03880075737833977, "learning_rate": 4.583395786921013e-05, "loss": 0.0259, "step": 22010 }, { "grad_norm": 0.039662666618824005, "learning_rate": 4.579276628943164e-05, "loss": 0.0278, "step": 22020 }, { "grad_norm": 0.03787447512149811, "learning_rate": 4.575157758525772e-05, "loss": 0.0275, "step": 22030 }, { "grad_norm": 0.07417988032102585, "learning_rate": 4.571039178484046e-05, "loss": 0.0288, "step": 22040 }, { "grad_norm": 0.045459624379873276, "learning_rate": 4.566920891632998e-05, "loss": 0.0278, "step": 22050 }, { "grad_norm": 0.04651174694299698, "learning_rate": 4.562802900787436e-05, "loss": 0.0289, "step": 22060 }, { "grad_norm": 0.03939501941204071, "learning_rate": 4.558685208761968e-05, "loss": 0.0269, "step": 22070 }, { "grad_norm": 0.03991120308637619, "learning_rate": 4.554567818370998e-05, "loss": 0.0291, "step": 22080 }, { "grad_norm": 0.039959266781806946, "learning_rate": 4.550450732428726e-05, "loss": 0.0282, "step": 22090 }, { "grad_norm": 0.03736316040158272, "learning_rate": 4.546333953749137e-05, "loss": 0.0257, "step": 22100 }, { "grad_norm": 0.03557121381163597, "learning_rate": 4.5422174851460154e-05, "loss": 0.0275, "step": 22110 }, { "grad_norm": 0.04820971190929413, "learning_rate": 4.538101329432924e-05, "loss": 0.0286, "step": 22120 }, { "grad_norm": 0.037222541868686676, "learning_rate": 4.5339854894232195e-05, "loss": 0.0288, "step": 22130 }, { "grad_norm": 0.0476839654147625, "learning_rate": 4.52986996793004e-05, "loss": 0.0267, "step": 22140 }, { "grad_norm": 0.05893699452280998, "learning_rate": 4.5257547677663024e-05, "loss": 0.0243, "step": 22150 }, { "grad_norm": 0.03085971437394619, "learning_rate": 4.52163989174471e-05, "loss": 0.0238, "step": 22160 }, { "grad_norm": 0.05545605346560478, "learning_rate": 4.51752534267774e-05, "loss": 0.0242, "step": 22170 }, { "grad_norm": 0.034274790436029434, "learning_rate": 4.513411123377649e-05, "loss": 0.0234, "step": 22180 }, { "grad_norm": 0.04967841878533363, "learning_rate": 4.5092972366564675e-05, "loss": 0.0239, "step": 22190 }, { "grad_norm": 0.03678249567747116, "learning_rate": 4.505183685325997e-05, "loss": 0.0239, "step": 22200 }, { "grad_norm": 0.032652679830789566, "learning_rate": 4.5010704721978125e-05, "loss": 0.028, "step": 22210 }, { "grad_norm": 0.047596175223588943, "learning_rate": 4.496957600083255e-05, "loss": 0.0282, "step": 22220 }, { "grad_norm": 0.03660891205072403, "learning_rate": 4.4928450717934343e-05, "loss": 0.0271, "step": 22230 }, { "grad_norm": 0.037910282611846924, "learning_rate": 4.488732890139227e-05, "loss": 0.0275, "step": 22240 }, { "grad_norm": 0.04081064090132713, "learning_rate": 4.4846210579312665e-05, "loss": 0.0281, "step": 22250 }, { "grad_norm": 0.041295696049928665, "learning_rate": 4.480509577979953e-05, "loss": 0.0279, "step": 22260 }, { "grad_norm": 0.05783839523792267, "learning_rate": 4.476398453095445e-05, "loss": 0.0285, "step": 22270 }, { "grad_norm": 0.04910064861178398, "learning_rate": 4.472287686087656e-05, "loss": 0.0272, "step": 22280 }, { "grad_norm": 0.03614113852381706, "learning_rate": 4.468177279766259e-05, "loss": 0.0271, "step": 22290 }, { "grad_norm": 0.04643624648451805, "learning_rate": 4.4640672369406746e-05, "loss": 0.0239, "step": 22300 }, { "grad_norm": 0.03351053223013878, "learning_rate": 4.459957560420082e-05, "loss": 0.0256, "step": 22310 }, { "grad_norm": 0.03708589822053909, "learning_rate": 4.455848253013403e-05, "loss": 0.0246, "step": 22320 }, { "grad_norm": 0.0529925562441349, "learning_rate": 4.4517393175293146e-05, "loss": 0.027, "step": 22330 }, { "grad_norm": 0.03651287779211998, "learning_rate": 4.447630756776232e-05, "loss": 0.028, "step": 22340 }, { "grad_norm": 0.056356120854616165, "learning_rate": 4.443522573562318e-05, "loss": 0.0282, "step": 22350 }, { "grad_norm": 0.042095035314559937, "learning_rate": 4.4394147706954776e-05, "loss": 0.028, "step": 22360 }, { "grad_norm": 0.04531075805425644, "learning_rate": 4.435307350983355e-05, "loss": 0.0278, "step": 22370 }, { "grad_norm": 0.039733223617076874, "learning_rate": 4.4312003172333326e-05, "loss": 0.0261, "step": 22380 }, { "grad_norm": 0.05442897230386734, "learning_rate": 4.427093672252531e-05, "loss": 0.0268, "step": 22390 }, { "grad_norm": 0.04646259546279907, "learning_rate": 4.422987418847802e-05, "loss": 0.0244, "step": 22400 }, { "grad_norm": 0.13489189743995667, "learning_rate": 4.4188815598257325e-05, "loss": 0.0257, "step": 22410 }, { "grad_norm": 0.04105103760957718, "learning_rate": 4.414776097992638e-05, "loss": 0.0247, "step": 22420 }, { "grad_norm": 0.048801638185977936, "learning_rate": 4.4106710361545595e-05, "loss": 0.0256, "step": 22430 }, { "grad_norm": 0.06587004661560059, "learning_rate": 4.406566377117272e-05, "loss": 0.0229, "step": 22440 }, { "grad_norm": 0.04906173422932625, "learning_rate": 4.40246212368627e-05, "loss": 0.0287, "step": 22450 }, { "grad_norm": 0.04925472289323807, "learning_rate": 4.3983582786667715e-05, "loss": 0.03, "step": 22460 }, { "grad_norm": 0.042444467544555664, "learning_rate": 4.394254844863716e-05, "loss": 0.0311, "step": 22470 }, { "grad_norm": 0.05737364664673805, "learning_rate": 4.390151825081762e-05, "loss": 0.0269, "step": 22480 }, { "grad_norm": 0.04508676379919052, "learning_rate": 4.386049222125286e-05, "loss": 0.0271, "step": 22490 }, { "grad_norm": 0.04784957692027092, "learning_rate": 4.3819470387983774e-05, "loss": 0.0303, "step": 22500 }, { "grad_norm": 0.049619391560554504, "learning_rate": 4.377845277904841e-05, "loss": 0.0302, "step": 22510 }, { "grad_norm": 0.04269272834062576, "learning_rate": 4.37374394224819e-05, "loss": 0.0284, "step": 22520 }, { "grad_norm": 0.03359593451023102, "learning_rate": 4.369643034631648e-05, "loss": 0.0295, "step": 22530 }, { "grad_norm": 0.03561216592788696, "learning_rate": 4.365542557858149e-05, "loss": 0.029, "step": 22540 }, { "grad_norm": 0.041708506643772125, "learning_rate": 4.361442514730329e-05, "loss": 0.0294, "step": 22550 }, { "grad_norm": 0.04185347631573677, "learning_rate": 4.357342908050528e-05, "loss": 0.0296, "step": 22560 }, { "grad_norm": 0.04484086111187935, "learning_rate": 4.3532437406207895e-05, "loss": 0.0288, "step": 22570 }, { "grad_norm": 0.09793075919151306, "learning_rate": 4.349145015242856e-05, "loss": 0.0288, "step": 22580 }, { "grad_norm": 0.038642413914203644, "learning_rate": 4.345046734718168e-05, "loss": 0.0284, "step": 22590 }, { "grad_norm": 0.057381294667720795, "learning_rate": 4.34094890184786e-05, "loss": 0.0297, "step": 22600 }, { "grad_norm": 0.038843151181936264, "learning_rate": 4.336851519432765e-05, "loss": 0.0296, "step": 22610 }, { "grad_norm": 0.0467938631772995, "learning_rate": 4.332754590273403e-05, "loss": 0.0274, "step": 22620 }, { "grad_norm": 0.044881101697683334, "learning_rate": 4.3286581171699855e-05, "loss": 0.0308, "step": 22630 }, { "grad_norm": 0.04184829071164131, "learning_rate": 4.324562102922416e-05, "loss": 0.0267, "step": 22640 }, { "grad_norm": 0.05032045394182205, "learning_rate": 4.320466550330278e-05, "loss": 0.0281, "step": 22650 }, { "grad_norm": 0.048222318291664124, "learning_rate": 4.3163714621928466e-05, "loss": 0.0264, "step": 22660 }, { "grad_norm": 0.046256743371486664, "learning_rate": 4.312276841309074e-05, "loss": 0.0259, "step": 22670 }, { "grad_norm": 0.05430758744478226, "learning_rate": 4.3081826904775945e-05, "loss": 0.0258, "step": 22680 }, { "grad_norm": 0.04299893602728844, "learning_rate": 4.3040890124967246e-05, "loss": 0.027, "step": 22690 }, { "grad_norm": 0.04906849190592766, "learning_rate": 4.2999958101644537e-05, "loss": 0.0272, "step": 22700 }, { "grad_norm": 0.046128980815410614, "learning_rate": 4.2959030862784435e-05, "loss": 0.0282, "step": 22710 }, { "grad_norm": 0.04354729503393173, "learning_rate": 4.291810843636036e-05, "loss": 0.0286, "step": 22720 }, { "grad_norm": 0.04457968845963478, "learning_rate": 4.2877190850342375e-05, "loss": 0.0319, "step": 22730 }, { "grad_norm": 0.03630239889025688, "learning_rate": 4.2836278132697294e-05, "loss": 0.0304, "step": 22740 }, { "grad_norm": 0.0344652496278286, "learning_rate": 4.279537031138855e-05, "loss": 0.0284, "step": 22750 }, { "grad_norm": 0.029645387083292007, "learning_rate": 4.275446741437625e-05, "loss": 0.0302, "step": 22760 }, { "grad_norm": 0.03503033146262169, "learning_rate": 4.2713569469617176e-05, "loss": 0.0279, "step": 22770 }, { "grad_norm": 0.040892668068408966, "learning_rate": 4.267267650506465e-05, "loss": 0.0289, "step": 22780 }, { "grad_norm": 0.03589764982461929, "learning_rate": 4.263178854866866e-05, "loss": 0.0282, "step": 22790 }, { "grad_norm": 0.04394865408539772, "learning_rate": 4.259090562837571e-05, "loss": 0.0281, "step": 22800 }, { "grad_norm": 0.036236364394426346, "learning_rate": 4.255002777212888e-05, "loss": 0.0277, "step": 22810 }, { "grad_norm": 0.04086247459053993, "learning_rate": 4.250915500786783e-05, "loss": 0.0284, "step": 22820 }, { "grad_norm": 0.05346035957336426, "learning_rate": 4.24682873635287e-05, "loss": 0.0298, "step": 22830 }, { "grad_norm": 0.028861505910754204, "learning_rate": 4.242742486704414e-05, "loss": 0.0273, "step": 22840 }, { "grad_norm": 0.04137812554836273, "learning_rate": 4.238656754634327e-05, "loss": 0.028, "step": 22850 }, { "grad_norm": 0.041316062211990356, "learning_rate": 4.234571542935168e-05, "loss": 0.0287, "step": 22860 }, { "grad_norm": 0.053932372480630875, "learning_rate": 4.230486854399144e-05, "loss": 0.0287, "step": 22870 }, { "grad_norm": 0.03693681210279465, "learning_rate": 4.226402691818098e-05, "loss": 0.0285, "step": 22880 }, { "grad_norm": 0.043953731656074524, "learning_rate": 4.2223190579835196e-05, "loss": 0.0276, "step": 22890 }, { "grad_norm": 0.052866075187921524, "learning_rate": 4.218235955686531e-05, "loss": 0.0264, "step": 22900 }, { "grad_norm": 0.03759311884641647, "learning_rate": 4.214153387717894e-05, "loss": 0.0246, "step": 22910 }, { "grad_norm": 0.044979456812143326, "learning_rate": 4.210071356868007e-05, "loss": 0.0251, "step": 22920 }, { "grad_norm": 0.0610036626458168, "learning_rate": 4.205989865926898e-05, "loss": 0.024, "step": 22930 }, { "grad_norm": 0.030375484377145767, "learning_rate": 4.2019089176842294e-05, "loss": 0.0255, "step": 22940 }, { "grad_norm": 0.04973174259066582, "learning_rate": 4.1978285149292894e-05, "loss": 0.025, "step": 22950 }, { "grad_norm": 0.07055611908435822, "learning_rate": 4.193748660450996e-05, "loss": 0.0306, "step": 22960 }, { "grad_norm": 0.05285901948809624, "learning_rate": 4.189669357037891e-05, "loss": 0.0284, "step": 22970 }, { "grad_norm": 0.07381168752908707, "learning_rate": 4.1855906074781405e-05, "loss": 0.0302, "step": 22980 }, { "grad_norm": 0.06378404796123505, "learning_rate": 4.1815124145595285e-05, "loss": 0.0321, "step": 22990 }, { "grad_norm": 0.06297926604747772, "learning_rate": 4.1774347810694644e-05, "loss": 0.0299, "step": 23000 }, { "grad_norm": 0.05649339035153389, "learning_rate": 4.17335770979497e-05, "loss": 0.0298, "step": 23010 }, { "grad_norm": 0.042780905961990356, "learning_rate": 4.169281203522687e-05, "loss": 0.0329, "step": 23020 }, { "grad_norm": 0.0480889193713665, "learning_rate": 4.1652052650388674e-05, "loss": 0.0304, "step": 23030 }, { "grad_norm": 0.04450966417789459, "learning_rate": 4.1611298971293786e-05, "loss": 0.0297, "step": 23040 }, { "grad_norm": 0.051966454833745956, "learning_rate": 4.1570551025796935e-05, "loss": 0.0286, "step": 23050 }, { "grad_norm": 0.03771984204649925, "learning_rate": 4.152980884174897e-05, "loss": 0.0305, "step": 23060 }, { "grad_norm": 0.04686366021633148, "learning_rate": 4.148907244699682e-05, "loss": 0.0295, "step": 23070 }, { "grad_norm": 0.039564166218042374, "learning_rate": 4.1448341869383395e-05, "loss": 0.03, "step": 23080 }, { "grad_norm": 0.05636744946241379, "learning_rate": 4.140761713674765e-05, "loss": 0.0298, "step": 23090 }, { "grad_norm": 0.03765176981687546, "learning_rate": 4.1366898276924574e-05, "loss": 0.0249, "step": 23100 }, { "grad_norm": 0.05539963021874428, "learning_rate": 4.132618531774512e-05, "loss": 0.0267, "step": 23110 }, { "grad_norm": 0.03659069538116455, "learning_rate": 4.128547828703622e-05, "loss": 0.0264, "step": 23120 }, { "grad_norm": 0.04860157147049904, "learning_rate": 4.1244777212620725e-05, "loss": 0.0252, "step": 23130 }, { "grad_norm": 0.04496210440993309, "learning_rate": 4.120408212231746e-05, "loss": 0.0244, "step": 23140 }, { "grad_norm": 0.048812080174684525, "learning_rate": 4.116339304394111e-05, "loss": 0.0283, "step": 23150 }, { "grad_norm": 0.042126841843128204, "learning_rate": 4.112271000530229e-05, "loss": 0.03, "step": 23160 }, { "grad_norm": 0.04395585507154465, "learning_rate": 4.10820330342075e-05, "loss": 0.0302, "step": 23170 }, { "grad_norm": 0.03824879601597786, "learning_rate": 4.1041362158459027e-05, "loss": 0.0304, "step": 23180 }, { "grad_norm": 0.037467122077941895, "learning_rate": 4.1000697405855024e-05, "loss": 0.0309, "step": 23190 }, { "grad_norm": 0.04711111634969711, "learning_rate": 4.096003880418951e-05, "loss": 0.0321, "step": 23200 }, { "grad_norm": 0.0728256106376648, "learning_rate": 4.0919386381252215e-05, "loss": 0.0305, "step": 23210 }, { "grad_norm": 0.05052027851343155, "learning_rate": 4.087874016482872e-05, "loss": 0.0305, "step": 23220 }, { "grad_norm": 0.04184744507074356, "learning_rate": 4.0838100182700295e-05, "loss": 0.0311, "step": 23230 }, { "grad_norm": 0.05561239644885063, "learning_rate": 4.079746646264402e-05, "loss": 0.0338, "step": 23240 }, { "grad_norm": 0.10176077485084534, "learning_rate": 4.075683903243262e-05, "loss": 0.0306, "step": 23250 }, { "grad_norm": 0.04455767944455147, "learning_rate": 4.071621791983462e-05, "loss": 0.029, "step": 23260 }, { "grad_norm": 0.03683802857995033, "learning_rate": 4.06756031526141e-05, "loss": 0.0308, "step": 23270 }, { "grad_norm": 0.04731821268796921, "learning_rate": 4.063499475853092e-05, "loss": 0.0278, "step": 23280 }, { "grad_norm": 0.03766752406954765, "learning_rate": 4.0594392765340506e-05, "loss": 0.0258, "step": 23290 }, { "grad_norm": 0.06373734027147293, "learning_rate": 4.0553797200793954e-05, "loss": 0.0262, "step": 23300 }, { "grad_norm": 0.04330817237496376, "learning_rate": 4.0513208092637926e-05, "loss": 0.0253, "step": 23310 }, { "grad_norm": 0.07330770790576935, "learning_rate": 4.0472625468614735e-05, "loss": 0.0234, "step": 23320 }, { "grad_norm": 0.05049319937825203, "learning_rate": 4.043204935646218e-05, "loss": 0.0254, "step": 23330 }, { "grad_norm": 0.08149690181016922, "learning_rate": 4.0391479783913675e-05, "loss": 0.0277, "step": 23340 }, { "grad_norm": 0.042728498578071594, "learning_rate": 4.0350916778698155e-05, "loss": 0.0254, "step": 23350 }, { "grad_norm": 0.05623696744441986, "learning_rate": 4.031036036854001e-05, "loss": 0.0249, "step": 23360 }, { "grad_norm": 0.043003734201192856, "learning_rate": 4.026981058115918e-05, "loss": 0.0253, "step": 23370 }, { "grad_norm": 0.050583094358444214, "learning_rate": 4.022926744427108e-05, "loss": 0.027, "step": 23380 }, { "grad_norm": 0.03952550143003464, "learning_rate": 4.018873098558654e-05, "loss": 0.0269, "step": 23390 }, { "grad_norm": 0.04505682364106178, "learning_rate": 4.014820123281186e-05, "loss": 0.0274, "step": 23400 }, { "grad_norm": 0.04132271558046341, "learning_rate": 4.0107678213648735e-05, "loss": 0.0278, "step": 23410 }, { "grad_norm": 0.04551846534013748, "learning_rate": 4.006716195579428e-05, "loss": 0.0254, "step": 23420 }, { "grad_norm": 0.046710625290870667, "learning_rate": 4.002665248694096e-05, "loss": 0.0267, "step": 23430 }, { "grad_norm": 0.048184193670749664, "learning_rate": 3.998614983477664e-05, "loss": 0.026, "step": 23440 }, { "grad_norm": 0.039953626692295074, "learning_rate": 3.994565402698448e-05, "loss": 0.0271, "step": 23450 }, { "grad_norm": 0.041824113577604294, "learning_rate": 3.9905165091242975e-05, "loss": 0.0287, "step": 23460 }, { "grad_norm": 0.04048573225736618, "learning_rate": 3.9864683055225936e-05, "loss": 0.0287, "step": 23470 }, { "grad_norm": 0.05269381031394005, "learning_rate": 3.982420794660247e-05, "loss": 0.0304, "step": 23480 }, { "grad_norm": 0.030253851786255836, "learning_rate": 3.978373979303691e-05, "loss": 0.0298, "step": 23490 }, { "grad_norm": 0.041946567595005035, "learning_rate": 3.974327862218888e-05, "loss": 0.0316, "step": 23500 }, { "grad_norm": 0.043216805905103683, "learning_rate": 3.970282446171318e-05, "loss": 0.03, "step": 23510 }, { "grad_norm": 0.03992640972137451, "learning_rate": 3.966237733925988e-05, "loss": 0.0294, "step": 23520 }, { "grad_norm": 0.04279695078730583, "learning_rate": 3.962193728247418e-05, "loss": 0.0276, "step": 23530 }, { "grad_norm": 0.039578575640916824, "learning_rate": 3.958150431899651e-05, "loss": 0.0276, "step": 23540 }, { "grad_norm": 0.03225512430071831, "learning_rate": 3.954107847646238e-05, "loss": 0.0252, "step": 23550 }, { "grad_norm": 0.03287568688392639, "learning_rate": 3.950065978250249e-05, "loss": 0.0263, "step": 23560 }, { "grad_norm": 0.039498403668403625, "learning_rate": 3.9460248264742624e-05, "loss": 0.0264, "step": 23570 }, { "grad_norm": 0.03820805996656418, "learning_rate": 3.941984395080371e-05, "loss": 0.026, "step": 23580 }, { "grad_norm": 0.05703115090727806, "learning_rate": 3.937944686830167e-05, "loss": 0.0288, "step": 23590 }, { "grad_norm": 0.04077694192528725, "learning_rate": 3.933905704484756e-05, "loss": 0.0259, "step": 23600 }, { "grad_norm": 0.037475086748600006, "learning_rate": 3.929867450804743e-05, "loss": 0.0286, "step": 23610 }, { "grad_norm": 0.03897884488105774, "learning_rate": 3.925829928550237e-05, "loss": 0.0273, "step": 23620 }, { "grad_norm": 0.04868171364068985, "learning_rate": 3.921793140480847e-05, "loss": 0.0293, "step": 23630 }, { "grad_norm": 0.04037657007575035, "learning_rate": 3.917757089355677e-05, "loss": 0.0287, "step": 23640 }, { "grad_norm": 0.048819318413734436, "learning_rate": 3.9137217779333326e-05, "loss": 0.0276, "step": 23650 }, { "grad_norm": 0.05303160101175308, "learning_rate": 3.9096872089719083e-05, "loss": 0.0254, "step": 23660 }, { "grad_norm": 0.04384654015302658, "learning_rate": 3.905653385228996e-05, "loss": 0.0284, "step": 23670 }, { "grad_norm": 0.08046158403158188, "learning_rate": 3.901620309461677e-05, "loss": 0.0302, "step": 23680 }, { "grad_norm": 0.047058697789907455, "learning_rate": 3.897587984426518e-05, "loss": 0.0306, "step": 23690 }, { "grad_norm": 0.03348928689956665, "learning_rate": 3.893556412879577e-05, "loss": 0.0287, "step": 23700 }, { "grad_norm": 0.04376688972115517, "learning_rate": 3.889525597576395e-05, "loss": 0.0287, "step": 23710 }, { "grad_norm": 0.04314300790429115, "learning_rate": 3.8854955412719965e-05, "loss": 0.029, "step": 23720 }, { "grad_norm": 0.045659635215997696, "learning_rate": 3.881466246720887e-05, "loss": 0.0251, "step": 23730 }, { "grad_norm": 0.044370025396347046, "learning_rate": 3.8774377166770484e-05, "loss": 0.0244, "step": 23740 }, { "grad_norm": 0.031824275851249695, "learning_rate": 3.8734099538939474e-05, "loss": 0.0258, "step": 23750 }, { "grad_norm": 0.03758329153060913, "learning_rate": 3.869382961124518e-05, "loss": 0.0262, "step": 23760 }, { "grad_norm": 0.04137083515524864, "learning_rate": 3.8653567411211736e-05, "loss": 0.0242, "step": 23770 }, { "grad_norm": 0.055619146674871445, "learning_rate": 3.8613312966357987e-05, "loss": 0.0258, "step": 23780 }, { "grad_norm": 0.0401938296854496, "learning_rate": 3.857306630419745e-05, "loss": 0.0273, "step": 23790 }, { "grad_norm": 0.04287959262728691, "learning_rate": 3.853282745223834e-05, "loss": 0.0274, "step": 23800 }, { "grad_norm": 0.03879475221037865, "learning_rate": 3.8492596437983546e-05, "loss": 0.0278, "step": 23810 }, { "grad_norm": 0.054630521684885025, "learning_rate": 3.8452373288930586e-05, "loss": 0.029, "step": 23820 }, { "grad_norm": 0.05127384513616562, "learning_rate": 3.841215803257159e-05, "loss": 0.0259, "step": 23830 }, { "grad_norm": 0.04269760474562645, "learning_rate": 3.83719506963933e-05, "loss": 0.0257, "step": 23840 }, { "grad_norm": 0.039085742086172104, "learning_rate": 3.8331751307877087e-05, "loss": 0.0273, "step": 23850 }, { "grad_norm": 0.04278268292546272, "learning_rate": 3.82915598944988e-05, "loss": 0.0267, "step": 23860 }, { "grad_norm": 0.04867204651236534, "learning_rate": 3.825137648372893e-05, "loss": 0.0296, "step": 23870 }, { "grad_norm": 0.03923951834440231, "learning_rate": 3.8211201103032465e-05, "loss": 0.028, "step": 23880 }, { "grad_norm": 0.057975269854068756, "learning_rate": 3.817103377986887e-05, "loss": 0.0292, "step": 23890 }, { "grad_norm": 0.03889992833137512, "learning_rate": 3.813087454169215e-05, "loss": 0.0272, "step": 23900 }, { "grad_norm": 0.05836428329348564, "learning_rate": 3.809072341595078e-05, "loss": 0.0293, "step": 23910 }, { "grad_norm": 0.05544901266694069, "learning_rate": 3.8050580430087636e-05, "loss": 0.0266, "step": 23920 }, { "grad_norm": 0.04713181406259537, "learning_rate": 3.8010445611540096e-05, "loss": 0.0255, "step": 23930 }, { "grad_norm": 0.025348540395498276, "learning_rate": 3.797031898773992e-05, "loss": 0.0261, "step": 23940 }, { "grad_norm": 0.038223810493946075, "learning_rate": 3.793020058611329e-05, "loss": 0.0266, "step": 23950 }, { "grad_norm": 0.034572165459394455, "learning_rate": 3.789009043408074e-05, "loss": 0.0266, "step": 23960 }, { "grad_norm": 0.03896702826023102, "learning_rate": 3.7849988559057194e-05, "loss": 0.0277, "step": 23970 }, { "grad_norm": 0.03970232978463173, "learning_rate": 3.78098949884519e-05, "loss": 0.027, "step": 23980 }, { "grad_norm": 0.046254586428403854, "learning_rate": 3.776980974966843e-05, "loss": 0.0265, "step": 23990 }, { "grad_norm": 0.08204708993434906, "learning_rate": 3.772973287010468e-05, "loss": 0.0268, "step": 24000 }, { "grad_norm": 0.04561780020594597, "learning_rate": 3.768966437715283e-05, "loss": 0.0265, "step": 24010 }, { "grad_norm": 0.05213429033756256, "learning_rate": 3.7649604298199274e-05, "loss": 0.026, "step": 24020 }, { "grad_norm": 0.038803815841674805, "learning_rate": 3.760955266062473e-05, "loss": 0.0274, "step": 24030 }, { "grad_norm": 0.030532682314515114, "learning_rate": 3.75695094918041e-05, "loss": 0.0255, "step": 24040 }, { "grad_norm": 0.051811233162879944, "learning_rate": 3.752947481910652e-05, "loss": 0.0274, "step": 24050 }, { "grad_norm": 0.055783070623874664, "learning_rate": 3.7489448669895324e-05, "loss": 0.0295, "step": 24060 }, { "grad_norm": 0.06532695889472961, "learning_rate": 3.744943107152798e-05, "loss": 0.0262, "step": 24070 }, { "grad_norm": 0.05944901332259178, "learning_rate": 3.7409422051356165e-05, "loss": 0.0284, "step": 24080 }, { "grad_norm": 0.043127868324518204, "learning_rate": 3.736942163672564e-05, "loss": 0.0268, "step": 24090 }, { "grad_norm": 0.04463992640376091, "learning_rate": 3.732942985497636e-05, "loss": 0.027, "step": 24100 }, { "grad_norm": 0.03995806351304054, "learning_rate": 3.728944673344228e-05, "loss": 0.0272, "step": 24110 }, { "grad_norm": 0.04469539225101471, "learning_rate": 3.72494722994515e-05, "loss": 0.0267, "step": 24120 }, { "grad_norm": 0.17956693470478058, "learning_rate": 3.720950658032617e-05, "loss": 0.0284, "step": 24130 }, { "grad_norm": 0.044288650155067444, "learning_rate": 3.716954960338249e-05, "loss": 0.0258, "step": 24140 }, { "grad_norm": 0.05254848301410675, "learning_rate": 3.712960139593066e-05, "loss": 0.0276, "step": 24150 }, { "grad_norm": 0.04009678214788437, "learning_rate": 3.708966198527493e-05, "loss": 0.0265, "step": 24160 }, { "grad_norm": 0.050845686346292496, "learning_rate": 3.704973139871349e-05, "loss": 0.0253, "step": 24170 }, { "grad_norm": 0.051513321697711945, "learning_rate": 3.700980966353853e-05, "loss": 0.0268, "step": 24180 }, { "grad_norm": 0.040438827127218246, "learning_rate": 3.696989680703619e-05, "loss": 0.0265, "step": 24190 }, { "grad_norm": 0.03772225230932236, "learning_rate": 3.69299928564865e-05, "loss": 0.0251, "step": 24200 }, { "grad_norm": 0.048683181405067444, "learning_rate": 3.689009783916345e-05, "loss": 0.0264, "step": 24210 }, { "grad_norm": 0.032188694924116135, "learning_rate": 3.6850211782334895e-05, "loss": 0.0251, "step": 24220 }, { "grad_norm": 0.03978457674384117, "learning_rate": 3.681033471326261e-05, "loss": 0.0282, "step": 24230 }, { "grad_norm": 0.04279032349586487, "learning_rate": 3.677046665920216e-05, "loss": 0.0293, "step": 24240 }, { "grad_norm": 0.03733161836862564, "learning_rate": 3.6730607647403005e-05, "loss": 0.0293, "step": 24250 }, { "grad_norm": 0.042706284672021866, "learning_rate": 3.6690757705108416e-05, "loss": 0.0302, "step": 24260 }, { "grad_norm": 0.04520031064748764, "learning_rate": 3.665091685955542e-05, "loss": 0.0268, "step": 24270 }, { "grad_norm": 0.04170075058937073, "learning_rate": 3.6611085137974896e-05, "loss": 0.0308, "step": 24280 }, { "grad_norm": 0.0480673611164093, "learning_rate": 3.657126256759143e-05, "loss": 0.0288, "step": 24290 }, { "grad_norm": 0.04259456321597099, "learning_rate": 3.653144917562335e-05, "loss": 0.0285, "step": 24300 }, { "grad_norm": 0.05231866240501404, "learning_rate": 3.649164498928277e-05, "loss": 0.0304, "step": 24310 }, { "grad_norm": 0.0455302894115448, "learning_rate": 3.645185003577546e-05, "loss": 0.0284, "step": 24320 }, { "grad_norm": 0.060951199382543564, "learning_rate": 3.6412064342300906e-05, "loss": 0.0313, "step": 24330 }, { "grad_norm": 0.0813572108745575, "learning_rate": 3.637228793605224e-05, "loss": 0.0269, "step": 24340 }, { "grad_norm": 0.0639164000749588, "learning_rate": 3.6332520844216264e-05, "loss": 0.0272, "step": 24350 }, { "grad_norm": 0.03921445086598396, "learning_rate": 3.6292763093973425e-05, "loss": 0.0267, "step": 24360 }, { "grad_norm": 0.032338958233594894, "learning_rate": 3.6253014712497754e-05, "loss": 0.0288, "step": 24370 }, { "grad_norm": 0.054412465542554855, "learning_rate": 3.621327572695692e-05, "loss": 0.0286, "step": 24380 }, { "grad_norm": 0.039128463715314865, "learning_rate": 3.617354616451211e-05, "loss": 0.0283, "step": 24390 }, { "grad_norm": 0.03859983757138252, "learning_rate": 3.6133826052318116e-05, "loss": 0.0282, "step": 24400 }, { "grad_norm": 0.035458266735076904, "learning_rate": 3.609411541752327e-05, "loss": 0.0271, "step": 24410 }, { "grad_norm": 0.05914541706442833, "learning_rate": 3.6054414287269405e-05, "loss": 0.0276, "step": 24420 }, { "grad_norm": 0.0612453930079937, "learning_rate": 3.601472268869188e-05, "loss": 0.0286, "step": 24430 }, { "grad_norm": 0.038155555725097656, "learning_rate": 3.597504064891952e-05, "loss": 0.0279, "step": 24440 }, { "grad_norm": 0.05443917214870453, "learning_rate": 3.5935368195074636e-05, "loss": 0.0284, "step": 24450 }, { "grad_norm": 0.036417070776224136, "learning_rate": 3.589570535427297e-05, "loss": 0.027, "step": 24460 }, { "grad_norm": 0.0409761406481266, "learning_rate": 3.585605215362371e-05, "loss": 0.0291, "step": 24470 }, { "grad_norm": 0.05773330107331276, "learning_rate": 3.581640862022941e-05, "loss": 0.029, "step": 24480 }, { "grad_norm": 0.04634428769350052, "learning_rate": 3.57767747811861e-05, "loss": 0.0291, "step": 24490 }, { "grad_norm": 0.044281311333179474, "learning_rate": 3.573715066358308e-05, "loss": 0.0277, "step": 24500 }, { "grad_norm": 0.04160061478614807, "learning_rate": 3.569753629450311e-05, "loss": 0.0298, "step": 24510 }, { "grad_norm": 0.047730907797813416, "learning_rate": 3.565793170102221e-05, "loss": 0.0269, "step": 24520 }, { "grad_norm": 0.05258800461888313, "learning_rate": 3.561833691020976e-05, "loss": 0.0283, "step": 24530 }, { "grad_norm": 0.04774710536003113, "learning_rate": 3.5578751949128415e-05, "loss": 0.026, "step": 24540 }, { "grad_norm": 0.05052905157208443, "learning_rate": 3.5539176844834125e-05, "loss": 0.0279, "step": 24550 }, { "grad_norm": 0.06306832283735275, "learning_rate": 3.5499611624376125e-05, "loss": 0.0282, "step": 24560 }, { "grad_norm": 0.04335474595427513, "learning_rate": 3.546005631479684e-05, "loss": 0.0275, "step": 24570 }, { "grad_norm": 0.0514860637485981, "learning_rate": 3.542051094313196e-05, "loss": 0.03, "step": 24580 }, { "grad_norm": 0.059284940361976624, "learning_rate": 3.5380975536410364e-05, "loss": 0.0282, "step": 24590 }, { "grad_norm": 0.05759191885590553, "learning_rate": 3.534145012165415e-05, "loss": 0.0294, "step": 24600 }, { "grad_norm": 0.04161602631211281, "learning_rate": 3.5301934725878546e-05, "loss": 0.0282, "step": 24610 }, { "grad_norm": 0.04158597066998482, "learning_rate": 3.526242937609197e-05, "loss": 0.0287, "step": 24620 }, { "grad_norm": 0.036129482090473175, "learning_rate": 3.522293409929595e-05, "loss": 0.025, "step": 24630 }, { "grad_norm": 0.06764905154705048, "learning_rate": 3.518344892248513e-05, "loss": 0.0273, "step": 24640 }, { "grad_norm": 0.05301393195986748, "learning_rate": 3.514397387264725e-05, "loss": 0.0269, "step": 24650 }, { "grad_norm": 0.03451194986701012, "learning_rate": 3.5104508976763176e-05, "loss": 0.0252, "step": 24660 }, { "grad_norm": 0.0433923713862896, "learning_rate": 3.506505426180674e-05, "loss": 0.0269, "step": 24670 }, { "grad_norm": 0.06292600929737091, "learning_rate": 3.502560975474488e-05, "loss": 0.0255, "step": 24680 }, { "grad_norm": 0.04971815645694733, "learning_rate": 3.4986175482537566e-05, "loss": 0.0262, "step": 24690 }, { "grad_norm": 0.056787144392728806, "learning_rate": 3.4946751472137725e-05, "loss": 0.0238, "step": 24700 }, { "grad_norm": 0.03118945099413395, "learning_rate": 3.490733775049132e-05, "loss": 0.027, "step": 24710 }, { "grad_norm": 0.051912616938352585, "learning_rate": 3.4867934344537236e-05, "loss": 0.0268, "step": 24720 }, { "grad_norm": 0.034622278064489365, "learning_rate": 3.482854128120735e-05, "loss": 0.0245, "step": 24730 }, { "grad_norm": 0.05295202508568764, "learning_rate": 3.478915858742643e-05, "loss": 0.0261, "step": 24740 }, { "grad_norm": 0.048204101622104645, "learning_rate": 3.4749786290112205e-05, "loss": 0.0279, "step": 24750 }, { "grad_norm": 0.047903627157211304, "learning_rate": 3.471042441617524e-05, "loss": 0.0272, "step": 24760 }, { "grad_norm": 0.04246718809008598, "learning_rate": 3.467107299251902e-05, "loss": 0.0276, "step": 24770 }, { "grad_norm": 0.047519806772470474, "learning_rate": 3.463173204603984e-05, "loss": 0.029, "step": 24780 }, { "grad_norm": 0.050236836075782776, "learning_rate": 3.4592401603626924e-05, "loss": 0.0282, "step": 24790 }, { "grad_norm": 0.036150973290205, "learning_rate": 3.45530816921622e-05, "loss": 0.0281, "step": 24800 }, { "grad_norm": 0.05579757317900658, "learning_rate": 3.451377233852051e-05, "loss": 0.0278, "step": 24810 }, { "grad_norm": 0.047692060470581055, "learning_rate": 3.4474473569569385e-05, "loss": 0.0294, "step": 24820 }, { "grad_norm": 0.0386558435857296, "learning_rate": 3.443518541216918e-05, "loss": 0.0264, "step": 24830 }, { "grad_norm": 0.040795307606458664, "learning_rate": 3.439590789317299e-05, "loss": 0.0264, "step": 24840 }, { "grad_norm": 0.04122069478034973, "learning_rate": 3.4356641039426607e-05, "loss": 0.0272, "step": 24850 }, { "grad_norm": 0.050189532339572906, "learning_rate": 3.431738487776857e-05, "loss": 0.0269, "step": 24860 }, { "grad_norm": 0.033886924386024475, "learning_rate": 3.4278139435030084e-05, "loss": 0.0276, "step": 24870 }, { "grad_norm": 0.04337156191468239, "learning_rate": 3.423890473803504e-05, "loss": 0.0265, "step": 24880 }, { "grad_norm": 0.03375987336039543, "learning_rate": 3.41996808136e-05, "loss": 0.0253, "step": 24890 }, { "grad_norm": 0.0404035747051239, "learning_rate": 3.416046768853413e-05, "loss": 0.0259, "step": 24900 }, { "grad_norm": 0.04160367697477341, "learning_rate": 3.412126538963925e-05, "loss": 0.0257, "step": 24910 }, { "grad_norm": 0.03954993560910225, "learning_rate": 3.4082073943709727e-05, "loss": 0.025, "step": 24920 }, { "grad_norm": 0.08371439576148987, "learning_rate": 3.404289337753258e-05, "loss": 0.0278, "step": 24930 }, { "grad_norm": 0.04766106978058815, "learning_rate": 3.400372371788736e-05, "loss": 0.0265, "step": 24940 }, { "grad_norm": 0.03993670642375946, "learning_rate": 3.3964564991546124e-05, "loss": 0.0243, "step": 24950 }, { "grad_norm": 0.039085689932107925, "learning_rate": 3.392541722527351e-05, "loss": 0.0267, "step": 24960 }, { "grad_norm": 0.04404645413160324, "learning_rate": 3.3886280445826644e-05, "loss": 0.0251, "step": 24970 }, { "grad_norm": 0.04059051349759102, "learning_rate": 3.3847154679955154e-05, "loss": 0.0244, "step": 24980 }, { "grad_norm": 0.03631792962551117, "learning_rate": 3.380803995440113e-05, "loss": 0.0235, "step": 24990 }, { "grad_norm": 0.026061393320560455, "learning_rate": 3.3768936295899115e-05, "loss": 0.0238, "step": 25000 }, { "grad_norm": 0.05207810923457146, "learning_rate": 3.3729843731176094e-05, "loss": 0.0268, "step": 25010 }, { "grad_norm": 0.042534153908491135, "learning_rate": 3.369076228695146e-05, "loss": 0.0236, "step": 25020 }, { "grad_norm": 0.047679025679826736, "learning_rate": 3.365169198993703e-05, "loss": 0.0258, "step": 25030 }, { "grad_norm": 0.04379130154848099, "learning_rate": 3.361263286683697e-05, "loss": 0.0256, "step": 25040 }, { "grad_norm": 0.03690417855978012, "learning_rate": 3.35735849443478e-05, "loss": 0.0261, "step": 25050 }, { "grad_norm": 0.036162231117486954, "learning_rate": 3.3534548249158435e-05, "loss": 0.0262, "step": 25060 }, { "grad_norm": 0.059461381286382675, "learning_rate": 3.3495522807950086e-05, "loss": 0.0297, "step": 25070 }, { "grad_norm": 0.05576873943209648, "learning_rate": 3.345650864739627e-05, "loss": 0.0286, "step": 25080 }, { "grad_norm": 0.037010449916124344, "learning_rate": 3.3417505794162794e-05, "loss": 0.0288, "step": 25090 }, { "grad_norm": 0.041895657777786255, "learning_rate": 3.3378514274907745e-05, "loss": 0.0311, "step": 25100 }, { "grad_norm": 0.043994467705488205, "learning_rate": 3.333953411628147e-05, "loss": 0.0271, "step": 25110 }, { "grad_norm": 0.03987087681889534, "learning_rate": 3.330056534492653e-05, "loss": 0.0258, "step": 25120 }, { "grad_norm": 0.07442503422498703, "learning_rate": 3.32616079874777e-05, "loss": 0.0248, "step": 25130 }, { "grad_norm": 0.03556404262781143, "learning_rate": 3.322266207056197e-05, "loss": 0.0258, "step": 25140 }, { "grad_norm": 0.045222073793411255, "learning_rate": 3.318372762079852e-05, "loss": 0.0243, "step": 25150 }, { "grad_norm": 0.044570062309503555, "learning_rate": 3.3144804664798666e-05, "loss": 0.0247, "step": 25160 }, { "grad_norm": 0.03777008503675461, "learning_rate": 3.3105893229165894e-05, "loss": 0.0266, "step": 25170 }, { "grad_norm": 0.04020845890045166, "learning_rate": 3.30669933404958e-05, "loss": 0.025, "step": 25180 }, { "grad_norm": 0.03699113056063652, "learning_rate": 3.302810502537609e-05, "loss": 0.0262, "step": 25190 }, { "grad_norm": 0.042645446956157684, "learning_rate": 3.298922831038655e-05, "loss": 0.026, "step": 25200 }, { "grad_norm": 0.04593393951654434, "learning_rate": 3.2950363222099073e-05, "loss": 0.0273, "step": 25210 }, { "grad_norm": 0.02424980141222477, "learning_rate": 3.291150978707758e-05, "loss": 0.026, "step": 25220 }, { "grad_norm": 0.03184265270829201, "learning_rate": 3.287266803187798e-05, "loss": 0.0266, "step": 25230 }, { "grad_norm": 0.033995162695646286, "learning_rate": 3.283383798304829e-05, "loss": 0.0276, "step": 25240 }, { "grad_norm": 0.03402205556631088, "learning_rate": 3.279501966712847e-05, "loss": 0.026, "step": 25250 }, { "grad_norm": 0.033541083335876465, "learning_rate": 3.275621311065047e-05, "loss": 0.0259, "step": 25260 }, { "grad_norm": 0.054479632526636124, "learning_rate": 3.271741834013822e-05, "loss": 0.0249, "step": 25270 }, { "grad_norm": 0.03649996221065521, "learning_rate": 3.267863538210756e-05, "loss": 0.0252, "step": 25280 }, { "grad_norm": 0.041210733354091644, "learning_rate": 3.2639864263066296e-05, "loss": 0.025, "step": 25290 }, { "grad_norm": 0.10408347845077515, "learning_rate": 3.26011050095141e-05, "loss": 0.0262, "step": 25300 }, { "grad_norm": 0.0421622060239315, "learning_rate": 3.256235764794259e-05, "loss": 0.0257, "step": 25310 }, { "grad_norm": 0.06181325018405914, "learning_rate": 3.2523622204835194e-05, "loss": 0.026, "step": 25320 }, { "grad_norm": 0.03606421500444412, "learning_rate": 3.2484898706667214e-05, "loss": 0.0266, "step": 25330 }, { "grad_norm": 0.05090166628360748, "learning_rate": 3.2446187179905806e-05, "loss": 0.0266, "step": 25340 }, { "grad_norm": 0.03909416124224663, "learning_rate": 3.240748765100995e-05, "loss": 0.0266, "step": 25350 }, { "grad_norm": 0.04632067680358887, "learning_rate": 3.236880014643039e-05, "loss": 0.0272, "step": 25360 }, { "grad_norm": 0.06528522074222565, "learning_rate": 3.233012469260969e-05, "loss": 0.0253, "step": 25370 }, { "grad_norm": 0.04589064419269562, "learning_rate": 3.229146131598213e-05, "loss": 0.0244, "step": 25380 }, { "grad_norm": 0.04463143274188042, "learning_rate": 3.2252810042973794e-05, "loss": 0.0264, "step": 25390 }, { "grad_norm": 0.0382172055542469, "learning_rate": 3.2214170900002456e-05, "loss": 0.0256, "step": 25400 }, { "grad_norm": 0.043241798877716064, "learning_rate": 3.217554391347758e-05, "loss": 0.0253, "step": 25410 }, { "grad_norm": 0.036056943237781525, "learning_rate": 3.213692910980037e-05, "loss": 0.0243, "step": 25420 }, { "grad_norm": 0.034732311964035034, "learning_rate": 3.2098326515363666e-05, "loss": 0.0249, "step": 25430 }, { "grad_norm": 0.04591568931937218, "learning_rate": 3.205973615655199e-05, "loss": 0.0272, "step": 25440 }, { "grad_norm": 0.047944940626621246, "learning_rate": 3.202115805974149e-05, "loss": 0.027, "step": 25450 }, { "grad_norm": 0.04528794065117836, "learning_rate": 3.1982592251299916e-05, "loss": 0.0267, "step": 25460 }, { "grad_norm": 0.046015821397304535, "learning_rate": 3.1944038757586656e-05, "loss": 0.0287, "step": 25470 }, { "grad_norm": 0.04936909303069115, "learning_rate": 3.190549760495263e-05, "loss": 0.0264, "step": 25480 }, { "grad_norm": 0.03911175951361656, "learning_rate": 3.186696881974039e-05, "loss": 0.0264, "step": 25490 }, { "grad_norm": 0.033243875950574875, "learning_rate": 3.1828452428283986e-05, "loss": 0.0262, "step": 25500 }, { "grad_norm": 0.0409788116812706, "learning_rate": 3.178994845690898e-05, "loss": 0.0233, "step": 25510 }, { "grad_norm": 0.05213499814271927, "learning_rate": 3.17514569319325e-05, "loss": 0.0274, "step": 25520 }, { "grad_norm": 0.048733387142419815, "learning_rate": 3.171297787966312e-05, "loss": 0.0271, "step": 25530 }, { "grad_norm": 0.03757680207490921, "learning_rate": 3.167451132640093e-05, "loss": 0.0265, "step": 25540 }, { "grad_norm": 0.05100459232926369, "learning_rate": 3.163605729843746e-05, "loss": 0.0255, "step": 25550 }, { "grad_norm": 0.045464202761650085, "learning_rate": 3.159761582205565e-05, "loss": 0.0261, "step": 25560 }, { "grad_norm": 0.04166946932673454, "learning_rate": 3.155918692352992e-05, "loss": 0.0257, "step": 25570 }, { "grad_norm": 0.05176934227347374, "learning_rate": 3.152077062912602e-05, "loss": 0.0263, "step": 25580 }, { "grad_norm": 0.04215496405959129, "learning_rate": 3.148236696510117e-05, "loss": 0.0264, "step": 25590 }, { "grad_norm": 0.04720129817724228, "learning_rate": 3.144397595770388e-05, "loss": 0.0251, "step": 25600 }, { "grad_norm": 0.046420373022556305, "learning_rate": 3.1405597633174036e-05, "loss": 0.0259, "step": 25610 }, { "grad_norm": 0.029064038768410683, "learning_rate": 3.136723201774289e-05, "loss": 0.0251, "step": 25620 }, { "grad_norm": 0.03799157589673996, "learning_rate": 3.132887913763295e-05, "loss": 0.0279, "step": 25630 }, { "grad_norm": 0.03398267924785614, "learning_rate": 3.129053901905806e-05, "loss": 0.0255, "step": 25640 }, { "grad_norm": 0.04669458791613579, "learning_rate": 3.125221168822335e-05, "loss": 0.028, "step": 25650 }, { "grad_norm": 0.041734762489795685, "learning_rate": 3.1213897171325154e-05, "loss": 0.025, "step": 25660 }, { "grad_norm": 0.04485572129487991, "learning_rate": 3.1175595494551116e-05, "loss": 0.0265, "step": 25670 }, { "grad_norm": 0.03851155564188957, "learning_rate": 3.1137306684080045e-05, "loss": 0.0247, "step": 25680 }, { "grad_norm": 0.04407995566725731, "learning_rate": 3.1099030766081985e-05, "loss": 0.0286, "step": 25690 }, { "grad_norm": 0.04023250937461853, "learning_rate": 3.106076776671818e-05, "loss": 0.0229, "step": 25700 }, { "grad_norm": 0.0656745508313179, "learning_rate": 3.102251771214101e-05, "loss": 0.0246, "step": 25710 }, { "grad_norm": 0.04045984521508217, "learning_rate": 3.098428062849404e-05, "loss": 0.0244, "step": 25720 }, { "grad_norm": 0.04347860440611839, "learning_rate": 3.094605654191195e-05, "loss": 0.0256, "step": 25730 }, { "grad_norm": 0.05810315161943436, "learning_rate": 3.090784547852055e-05, "loss": 0.0264, "step": 25740 }, { "grad_norm": 0.058301445096731186, "learning_rate": 3.0869647464436746e-05, "loss": 0.0243, "step": 25750 }, { "grad_norm": 0.03581609949469566, "learning_rate": 3.0831462525768496e-05, "loss": 0.0247, "step": 25760 }, { "grad_norm": 0.050760310143232346, "learning_rate": 3.079329068861488e-05, "loss": 0.0246, "step": 25770 }, { "grad_norm": 0.03548097610473633, "learning_rate": 3.075513197906597e-05, "loss": 0.0256, "step": 25780 }, { "grad_norm": 0.0646413117647171, "learning_rate": 3.071698642320286e-05, "loss": 0.0248, "step": 25790 }, { "grad_norm": 0.0675312951207161, "learning_rate": 3.067885404709772e-05, "loss": 0.0268, "step": 25800 }, { "grad_norm": 0.09470503032207489, "learning_rate": 3.0640734876813636e-05, "loss": 0.0263, "step": 25810 }, { "grad_norm": 0.049418725073337555, "learning_rate": 3.060262893840473e-05, "loss": 0.0283, "step": 25820 }, { "grad_norm": 0.03332525119185448, "learning_rate": 3.056453625791603e-05, "loss": 0.0257, "step": 25830 }, { "grad_norm": 0.03955113887786865, "learning_rate": 3.052645686138353e-05, "loss": 0.0256, "step": 25840 }, { "grad_norm": 0.04221835732460022, "learning_rate": 3.0488390774834153e-05, "loss": 0.0256, "step": 25850 }, { "grad_norm": 0.042707644402980804, "learning_rate": 3.0450338024285684e-05, "loss": 0.0266, "step": 25860 }, { "grad_norm": 0.0420917309820652, "learning_rate": 3.0412298635746855e-05, "loss": 0.0259, "step": 25870 }, { "grad_norm": 0.04580381512641907, "learning_rate": 3.03742726352172e-05, "loss": 0.0244, "step": 25880 }, { "grad_norm": 0.037255123257637024, "learning_rate": 3.0336260048687125e-05, "loss": 0.0224, "step": 25890 }, { "grad_norm": 0.0391460582613945, "learning_rate": 3.0298260902137897e-05, "loss": 0.0245, "step": 25900 }, { "grad_norm": 0.04044214263558388, "learning_rate": 3.0260275221541566e-05, "loss": 0.0231, "step": 25910 }, { "grad_norm": 0.03952762857079506, "learning_rate": 3.0222303032860987e-05, "loss": 0.0252, "step": 25920 }, { "grad_norm": 0.04155441001057625, "learning_rate": 3.018434436204979e-05, "loss": 0.024, "step": 25930 }, { "grad_norm": 0.0573115348815918, "learning_rate": 3.014639923505237e-05, "loss": 0.0263, "step": 25940 }, { "grad_norm": 0.047613151371479034, "learning_rate": 3.0108467677803863e-05, "loss": 0.0279, "step": 25950 }, { "grad_norm": 0.07399829477071762, "learning_rate": 3.0070549716230156e-05, "loss": 0.0281, "step": 25960 }, { "grad_norm": 0.046523526310920715, "learning_rate": 3.003264537624777e-05, "loss": 0.0277, "step": 25970 }, { "grad_norm": 0.034228160977363586, "learning_rate": 2.9994754683764e-05, "loss": 0.0284, "step": 25980 }, { "grad_norm": 0.044189080595970154, "learning_rate": 2.9956877664676754e-05, "loss": 0.0278, "step": 25990 }, { "grad_norm": 0.04145435243844986, "learning_rate": 2.9919014344874636e-05, "loss": 0.0272, "step": 26000 }, { "grad_norm": 0.056378673762083054, "learning_rate": 2.9881164750236857e-05, "loss": 0.0238, "step": 26010 }, { "grad_norm": 0.04534418508410454, "learning_rate": 2.984332890663326e-05, "loss": 0.0251, "step": 26020 }, { "grad_norm": 0.04474170133471489, "learning_rate": 2.9805506839924292e-05, "loss": 0.0254, "step": 26030 }, { "grad_norm": 0.06033025681972504, "learning_rate": 2.9767698575960968e-05, "loss": 0.0231, "step": 26040 }, { "grad_norm": 0.03941924124956131, "learning_rate": 2.9729904140584913e-05, "loss": 0.0244, "step": 26050 }, { "grad_norm": 0.06231696903705597, "learning_rate": 2.9692123559628234e-05, "loss": 0.0245, "step": 26060 }, { "grad_norm": 0.035608526319265366, "learning_rate": 2.9654356858913596e-05, "loss": 0.0241, "step": 26070 }, { "grad_norm": 0.049697812646627426, "learning_rate": 2.9616604064254206e-05, "loss": 0.0256, "step": 26080 }, { "grad_norm": 0.08875059336423874, "learning_rate": 2.9578865201453732e-05, "loss": 0.0243, "step": 26090 }, { "grad_norm": 0.04350545257329941, "learning_rate": 2.9541140296306335e-05, "loss": 0.0272, "step": 26100 }, { "grad_norm": 0.03787022829055786, "learning_rate": 2.9503429374596627e-05, "loss": 0.0265, "step": 26110 }, { "grad_norm": 0.04206852614879608, "learning_rate": 2.946573246209967e-05, "loss": 0.0264, "step": 26120 }, { "grad_norm": 0.04340300336480141, "learning_rate": 2.942804958458094e-05, "loss": 0.0248, "step": 26130 }, { "grad_norm": 0.04402504488825798, "learning_rate": 2.9390380767796343e-05, "loss": 0.0245, "step": 26140 }, { "grad_norm": 0.048496175557374954, "learning_rate": 2.9352726037492174e-05, "loss": 0.0252, "step": 26150 }, { "grad_norm": 0.03403433412313461, "learning_rate": 2.9315085419405052e-05, "loss": 0.0249, "step": 26160 }, { "grad_norm": 0.05920447036623955, "learning_rate": 2.927745893926199e-05, "loss": 0.025, "step": 26170 }, { "grad_norm": 0.040633562952280045, "learning_rate": 2.9239846622780358e-05, "loss": 0.0234, "step": 26180 }, { "grad_norm": 0.0365760512650013, "learning_rate": 2.9202248495667788e-05, "loss": 0.0253, "step": 26190 }, { "grad_norm": 0.03315942361950874, "learning_rate": 2.916466458362227e-05, "loss": 0.0246, "step": 26200 }, { "grad_norm": 0.05079925060272217, "learning_rate": 2.9127094912332033e-05, "loss": 0.0255, "step": 26210 }, { "grad_norm": 0.048214852809906006, "learning_rate": 2.9089539507475606e-05, "loss": 0.0269, "step": 26220 }, { "grad_norm": 0.03650747984647751, "learning_rate": 2.9051998394721748e-05, "loss": 0.0248, "step": 26230 }, { "grad_norm": 0.03874754160642624, "learning_rate": 2.901447159972948e-05, "loss": 0.0252, "step": 26240 }, { "grad_norm": 0.04196259006857872, "learning_rate": 2.8976959148148e-05, "loss": 0.0258, "step": 26250 }, { "grad_norm": 0.03698953613638878, "learning_rate": 2.8939461065616674e-05, "loss": 0.0259, "step": 26260 }, { "grad_norm": 0.037482064217329025, "learning_rate": 2.8901977377765127e-05, "loss": 0.0255, "step": 26270 }, { "grad_norm": 0.03888781741261482, "learning_rate": 2.8864508110213094e-05, "loss": 0.0254, "step": 26280 }, { "grad_norm": 0.049103591591119766, "learning_rate": 2.8827053288570503e-05, "loss": 0.0268, "step": 26290 }, { "grad_norm": 0.045192040503025055, "learning_rate": 2.8789612938437315e-05, "loss": 0.027, "step": 26300 }, { "grad_norm": 0.049693088978528976, "learning_rate": 2.8752187085403683e-05, "loss": 0.0264, "step": 26310 }, { "grad_norm": 0.03830600529909134, "learning_rate": 2.8714775755049818e-05, "loss": 0.0268, "step": 26320 }, { "grad_norm": 0.037799056619405746, "learning_rate": 2.867737897294604e-05, "loss": 0.0272, "step": 26330 }, { "grad_norm": 0.03915000334382057, "learning_rate": 2.8639996764652653e-05, "loss": 0.0271, "step": 26340 }, { "grad_norm": 0.03843235969543457, "learning_rate": 2.8602629155720084e-05, "loss": 0.0245, "step": 26350 }, { "grad_norm": 0.03770233318209648, "learning_rate": 2.8565276171688703e-05, "loss": 0.0239, "step": 26360 }, { "grad_norm": 0.037844929844141006, "learning_rate": 2.8527937838088943e-05, "loss": 0.0232, "step": 26370 }, { "grad_norm": 0.14903610944747925, "learning_rate": 2.84906141804412e-05, "loss": 0.0222, "step": 26380 }, { "grad_norm": 0.031811755150556564, "learning_rate": 2.8453305224255867e-05, "loss": 0.0228, "step": 26390 }, { "grad_norm": 0.037390295416116714, "learning_rate": 2.8416010995033216e-05, "loss": 0.023, "step": 26400 }, { "grad_norm": 0.03800824657082558, "learning_rate": 2.8378731518263524e-05, "loss": 0.0246, "step": 26410 }, { "grad_norm": 0.030258074402809143, "learning_rate": 2.834146681942696e-05, "loss": 0.0244, "step": 26420 }, { "grad_norm": 0.037430547177791595, "learning_rate": 2.8304216923993622e-05, "loss": 0.0258, "step": 26430 }, { "grad_norm": 0.0378822386264801, "learning_rate": 2.8266981857423413e-05, "loss": 0.0255, "step": 26440 }, { "grad_norm": 0.04691128805279732, "learning_rate": 2.8229761645166197e-05, "loss": 0.0267, "step": 26450 }, { "grad_norm": 0.0377044677734375, "learning_rate": 2.81925563126616e-05, "loss": 0.026, "step": 26460 }, { "grad_norm": 0.0531894750893116, "learning_rate": 2.8155365885339124e-05, "loss": 0.0291, "step": 26470 }, { "grad_norm": 0.03342267498373985, "learning_rate": 2.8118190388618093e-05, "loss": 0.0278, "step": 26480 }, { "grad_norm": 0.035287510603666306, "learning_rate": 2.8081029847907607e-05, "loss": 0.0285, "step": 26490 }, { "grad_norm": 0.05016254261136055, "learning_rate": 2.8043884288606525e-05, "loss": 0.0269, "step": 26500 }, { "grad_norm": 0.0402645580470562, "learning_rate": 2.8006753736103496e-05, "loss": 0.0276, "step": 26510 }, { "grad_norm": 0.04216066002845764, "learning_rate": 2.7969638215776918e-05, "loss": 0.0255, "step": 26520 }, { "grad_norm": 0.052211616188287735, "learning_rate": 2.793253775299487e-05, "loss": 0.0278, "step": 26530 }, { "grad_norm": 0.036245375871658325, "learning_rate": 2.7895452373115184e-05, "loss": 0.0257, "step": 26540 }, { "grad_norm": 0.04929191246628761, "learning_rate": 2.785838210148539e-05, "loss": 0.0256, "step": 26550 }, { "grad_norm": 0.03667499125003815, "learning_rate": 2.782132696344263e-05, "loss": 0.0259, "step": 26560 }, { "grad_norm": 0.0708979070186615, "learning_rate": 2.7784286984313745e-05, "loss": 0.0266, "step": 26570 }, { "grad_norm": 0.04280577600002289, "learning_rate": 2.7747262189415236e-05, "loss": 0.0246, "step": 26580 }, { "grad_norm": 0.03452114388346672, "learning_rate": 2.7710252604053205e-05, "loss": 0.0273, "step": 26590 }, { "grad_norm": 0.04912804067134857, "learning_rate": 2.767325825352332e-05, "loss": 0.0276, "step": 26600 }, { "grad_norm": 0.050549644976854324, "learning_rate": 2.7636279163110913e-05, "loss": 0.0235, "step": 26610 }, { "grad_norm": 0.04660017043352127, "learning_rate": 2.7599315358090795e-05, "loss": 0.0276, "step": 26620 }, { "grad_norm": 0.06672387570142746, "learning_rate": 2.7562366863727407e-05, "loss": 0.0253, "step": 26630 }, { "grad_norm": 0.04562574252486229, "learning_rate": 2.7525433705274695e-05, "loss": 0.0271, "step": 26640 }, { "grad_norm": 0.04677417501807213, "learning_rate": 2.748851590797614e-05, "loss": 0.0266, "step": 26650 }, { "grad_norm": 0.04176545888185501, "learning_rate": 2.7451613497064675e-05, "loss": 0.0327, "step": 26660 }, { "grad_norm": 0.052835531532764435, "learning_rate": 2.7414726497762765e-05, "loss": 0.0289, "step": 26670 }, { "grad_norm": 0.050614066421985626, "learning_rate": 2.737785493528232e-05, "loss": 0.029, "step": 26680 }, { "grad_norm": 0.04764733090996742, "learning_rate": 2.7340998834824745e-05, "loss": 0.0285, "step": 26690 }, { "grad_norm": 0.043125443160533905, "learning_rate": 2.7304158221580777e-05, "loss": 0.0283, "step": 26700 }, { "grad_norm": 0.040919121354818344, "learning_rate": 2.7267333120730675e-05, "loss": 0.0301, "step": 26710 }, { "grad_norm": 0.046208132058382034, "learning_rate": 2.7230523557444017e-05, "loss": 0.0269, "step": 26720 }, { "grad_norm": 0.04941010847687721, "learning_rate": 2.7193729556879798e-05, "loss": 0.0264, "step": 26730 }, { "grad_norm": 0.048526931554079056, "learning_rate": 2.715695114418637e-05, "loss": 0.0264, "step": 26740 }, { "grad_norm": 0.0470455177128315, "learning_rate": 2.7120188344501475e-05, "loss": 0.0255, "step": 26750 }, { "grad_norm": 0.0399119071662426, "learning_rate": 2.7083441182952067e-05, "loss": 0.0263, "step": 26760 }, { "grad_norm": 0.03410797938704491, "learning_rate": 2.7046709684654527e-05, "loss": 0.0246, "step": 26770 }, { "grad_norm": 0.04251760616898537, "learning_rate": 2.700999387471448e-05, "loss": 0.0264, "step": 26780 }, { "grad_norm": 0.04650869220495224, "learning_rate": 2.6973293778226854e-05, "loss": 0.0258, "step": 26790 }, { "grad_norm": 0.04254278168082237, "learning_rate": 2.6936609420275804e-05, "loss": 0.025, "step": 26800 }, { "grad_norm": 0.04218725115060806, "learning_rate": 2.689994082593472e-05, "loss": 0.0259, "step": 26810 }, { "grad_norm": 0.05897451192140579, "learning_rate": 2.6863288020266264e-05, "loss": 0.0254, "step": 26820 }, { "grad_norm": 0.047544851899147034, "learning_rate": 2.682665102832228e-05, "loss": 0.0255, "step": 26830 }, { "grad_norm": 0.031603965908288956, "learning_rate": 2.67900298751438e-05, "loss": 0.0255, "step": 26840 }, { "grad_norm": 0.06052102521061897, "learning_rate": 2.6753424585761067e-05, "loss": 0.0254, "step": 26850 }, { "grad_norm": 0.039691824465990067, "learning_rate": 2.671683518519341e-05, "loss": 0.0248, "step": 26860 }, { "grad_norm": 0.048046063631772995, "learning_rate": 2.668026169844936e-05, "loss": 0.0246, "step": 26870 }, { "grad_norm": 0.040471989661455154, "learning_rate": 2.6643704150526538e-05, "loss": 0.0241, "step": 26880 }, { "grad_norm": 0.03098161518573761, "learning_rate": 2.6607162566411716e-05, "loss": 0.0239, "step": 26890 }, { "grad_norm": 0.045007847249507904, "learning_rate": 2.6570636971080697e-05, "loss": 0.0249, "step": 26900 }, { "grad_norm": 0.06444563716650009, "learning_rate": 2.6534127389498364e-05, "loss": 0.025, "step": 26910 }, { "grad_norm": 0.033088333904743195, "learning_rate": 2.6497633846618696e-05, "loss": 0.0241, "step": 26920 }, { "grad_norm": 0.039013199508190155, "learning_rate": 2.6461156367384677e-05, "loss": 0.0242, "step": 26930 }, { "grad_norm": 0.03683963790535927, "learning_rate": 2.6424694976728316e-05, "loss": 0.0254, "step": 26940 }, { "grad_norm": 0.04417002201080322, "learning_rate": 2.6388249699570667e-05, "loss": 0.0241, "step": 26950 }, { "grad_norm": 0.04478374868631363, "learning_rate": 2.6351820560821672e-05, "loss": 0.0257, "step": 26960 }, { "grad_norm": 0.03633590415120125, "learning_rate": 2.631540758538034e-05, "loss": 0.0235, "step": 26970 }, { "grad_norm": 0.03264783322811127, "learning_rate": 2.6279010798134597e-05, "loss": 0.0244, "step": 26980 }, { "grad_norm": 0.06128954887390137, "learning_rate": 2.6242630223961305e-05, "loss": 0.0265, "step": 26990 }, { "grad_norm": 0.03729494288563728, "learning_rate": 2.6206265887726244e-05, "loss": 0.0252, "step": 27000 }, { "grad_norm": 0.0374571830034256, "learning_rate": 2.6169917814284066e-05, "loss": 0.0245, "step": 27010 }, { "grad_norm": 0.03409344330430031, "learning_rate": 2.6133586028478364e-05, "loss": 0.0253, "step": 27020 }, { "grad_norm": 0.04228934645652771, "learning_rate": 2.609727055514155e-05, "loss": 0.0234, "step": 27030 }, { "grad_norm": 0.03253008425235748, "learning_rate": 2.606097141909494e-05, "loss": 0.0246, "step": 27040 }, { "grad_norm": 0.040968529880046844, "learning_rate": 2.6024688645148644e-05, "loss": 0.0245, "step": 27050 }, { "grad_norm": 0.03996624797582626, "learning_rate": 2.5988422258101564e-05, "loss": 0.0235, "step": 27060 }, { "grad_norm": 0.04977913573384285, "learning_rate": 2.5952172282741453e-05, "loss": 0.0226, "step": 27070 }, { "grad_norm": 0.054068442434072495, "learning_rate": 2.5915938743844853e-05, "loss": 0.023, "step": 27080 }, { "grad_norm": 0.04283708333969116, "learning_rate": 2.5879721666177003e-05, "loss": 0.0231, "step": 27090 }, { "grad_norm": 0.0499827079474926, "learning_rate": 2.5843521074491972e-05, "loss": 0.0241, "step": 27100 }, { "grad_norm": 0.06404633820056915, "learning_rate": 2.5807336993532487e-05, "loss": 0.0251, "step": 27110 }, { "grad_norm": 0.041569434106349945, "learning_rate": 2.577116944803004e-05, "loss": 0.0241, "step": 27120 }, { "grad_norm": 0.04582272097468376, "learning_rate": 2.5735018462704818e-05, "loss": 0.0253, "step": 27130 }, { "grad_norm": 0.04854002967476845, "learning_rate": 2.5698884062265665e-05, "loss": 0.023, "step": 27140 }, { "grad_norm": 0.049256257712841034, "learning_rate": 2.5662766271410134e-05, "loss": 0.0265, "step": 27150 }, { "grad_norm": 0.04030991718173027, "learning_rate": 2.5626665114824343e-05, "loss": 0.0258, "step": 27160 }, { "grad_norm": 0.0533609613776207, "learning_rate": 2.5590580617183148e-05, "loss": 0.0271, "step": 27170 }, { "grad_norm": 0.059522245079278946, "learning_rate": 2.5554512803149912e-05, "loss": 0.0288, "step": 27180 }, { "grad_norm": 0.045237671583890915, "learning_rate": 2.5518461697376662e-05, "loss": 0.0245, "step": 27190 }, { "grad_norm": 0.04363376647233963, "learning_rate": 2.548242732450402e-05, "loss": 0.0268, "step": 27200 }, { "grad_norm": 0.03591596707701683, "learning_rate": 2.5446409709161095e-05, "loss": 0.0266, "step": 27210 }, { "grad_norm": 0.0422450415790081, "learning_rate": 2.541040887596561e-05, "loss": 0.0271, "step": 27220 }, { "grad_norm": 0.052765510976314545, "learning_rate": 2.537442484952378e-05, "loss": 0.0276, "step": 27230 }, { "grad_norm": 0.08847755193710327, "learning_rate": 2.533845765443037e-05, "loss": 0.0268, "step": 27240 }, { "grad_norm": 0.05953560397028923, "learning_rate": 2.530250731526863e-05, "loss": 0.025, "step": 27250 }, { "grad_norm": 0.04069187864661217, "learning_rate": 2.5266573856610253e-05, "loss": 0.0247, "step": 27260 }, { "grad_norm": 0.06236827000975609, "learning_rate": 2.5230657303015403e-05, "loss": 0.0237, "step": 27270 }, { "grad_norm": 0.043950166553258896, "learning_rate": 2.5194757679032728e-05, "loss": 0.0255, "step": 27280 }, { "grad_norm": 0.03225109353661537, "learning_rate": 2.5158875009199278e-05, "loss": 0.0245, "step": 27290 }, { "grad_norm": 0.03694595769047737, "learning_rate": 2.5123009318040537e-05, "loss": 0.0254, "step": 27300 }, { "grad_norm": 0.06842074543237686, "learning_rate": 2.508716063007034e-05, "loss": 0.0255, "step": 27310 }, { "grad_norm": 0.0498795360326767, "learning_rate": 2.5051328969790934e-05, "loss": 0.0245, "step": 27320 }, { "grad_norm": 0.036108408123254776, "learning_rate": 2.501551436169292e-05, "loss": 0.0265, "step": 27330 }, { "grad_norm": 0.06332111358642578, "learning_rate": 2.4979716830255255e-05, "loss": 0.0257, "step": 27340 }, { "grad_norm": 0.03274569660425186, "learning_rate": 2.4943936399945233e-05, "loss": 0.0246, "step": 27350 }, { "grad_norm": 0.03744745999574661, "learning_rate": 2.4908173095218412e-05, "loss": 0.0233, "step": 27360 }, { "grad_norm": 0.03593941405415535, "learning_rate": 2.4872426940518663e-05, "loss": 0.0258, "step": 27370 }, { "grad_norm": 0.0413743294775486, "learning_rate": 2.4836697960278156e-05, "loss": 0.0261, "step": 27380 }, { "grad_norm": 0.03884699195623398, "learning_rate": 2.480098617891732e-05, "loss": 0.0247, "step": 27390 }, { "grad_norm": 0.03047751635313034, "learning_rate": 2.4765291620844837e-05, "loss": 0.0233, "step": 27400 }, { "grad_norm": 0.030411820858716965, "learning_rate": 2.472961431045756e-05, "loss": 0.0249, "step": 27410 }, { "grad_norm": 0.04267521947622299, "learning_rate": 2.4693954272140622e-05, "loss": 0.0231, "step": 27420 }, { "grad_norm": 0.03189150616526604, "learning_rate": 2.4658311530267315e-05, "loss": 0.0225, "step": 27430 }, { "grad_norm": 0.04103327542543411, "learning_rate": 2.4622686109199124e-05, "loss": 0.0225, "step": 27440 }, { "grad_norm": 0.04399267956614494, "learning_rate": 2.4587078033285695e-05, "loss": 0.0249, "step": 27450 }, { "grad_norm": 0.054286662489175797, "learning_rate": 2.45514873268648e-05, "loss": 0.0227, "step": 27460 }, { "grad_norm": 0.03901870176196098, "learning_rate": 2.4515914014262336e-05, "loss": 0.0229, "step": 27470 }, { "grad_norm": 0.04394494369626045, "learning_rate": 2.4480358119792345e-05, "loss": 0.0261, "step": 27480 }, { "grad_norm": 0.046928729861974716, "learning_rate": 2.4444819667756942e-05, "loss": 0.0229, "step": 27490 }, { "grad_norm": 0.040602993220090866, "learning_rate": 2.4409298682446346e-05, "loss": 0.0259, "step": 27500 }, { "grad_norm": 0.04485655203461647, "learning_rate": 2.437379518813877e-05, "loss": 0.0243, "step": 27510 }, { "grad_norm": 0.06098052114248276, "learning_rate": 2.4338309209100547e-05, "loss": 0.0263, "step": 27520 }, { "grad_norm": 0.05391869321465492, "learning_rate": 2.4302840769586004e-05, "loss": 0.0238, "step": 27530 }, { "grad_norm": 0.037250783294439316, "learning_rate": 2.42673898938375e-05, "loss": 0.0269, "step": 27540 }, { "grad_norm": 0.05694030970335007, "learning_rate": 2.4231956606085343e-05, "loss": 0.0262, "step": 27550 }, { "grad_norm": 0.04089662805199623, "learning_rate": 2.419654093054789e-05, "loss": 0.0273, "step": 27560 }, { "grad_norm": 0.057572778314352036, "learning_rate": 2.4161142891431375e-05, "loss": 0.0255, "step": 27570 }, { "grad_norm": 0.04602838307619095, "learning_rate": 2.412576251293005e-05, "loss": 0.025, "step": 27580 }, { "grad_norm": 0.05531005933880806, "learning_rate": 2.4090399819226068e-05, "loss": 0.0265, "step": 27590 }, { "grad_norm": 0.04312954843044281, "learning_rate": 2.4055054834489514e-05, "loss": 0.0264, "step": 27600 }, { "grad_norm": 0.046638231724500656, "learning_rate": 2.401972758287832e-05, "loss": 0.0235, "step": 27610 }, { "grad_norm": 0.04141467437148094, "learning_rate": 2.398441808853834e-05, "loss": 0.024, "step": 27620 }, { "grad_norm": 0.05191562697291374, "learning_rate": 2.3949126375603288e-05, "loss": 0.0248, "step": 27630 }, { "grad_norm": 0.0390949621796608, "learning_rate": 2.3913852468194724e-05, "loss": 0.0254, "step": 27640 }, { "grad_norm": 0.15220214426517487, "learning_rate": 2.387859639042201e-05, "loss": 0.024, "step": 27650 }, { "grad_norm": 0.026360781863331795, "learning_rate": 2.3843358166382368e-05, "loss": 0.023, "step": 27660 }, { "grad_norm": 0.03805338218808174, "learning_rate": 2.3808137820160757e-05, "loss": 0.0237, "step": 27670 }, { "grad_norm": 0.04295172914862633, "learning_rate": 2.3772935375829975e-05, "loss": 0.022, "step": 27680 }, { "grad_norm": 0.03112499974668026, "learning_rate": 2.3737750857450553e-05, "loss": 0.0232, "step": 27690 }, { "grad_norm": 0.033912114799022675, "learning_rate": 2.3702584289070805e-05, "loss": 0.0246, "step": 27700 }, { "grad_norm": 0.046336688101291656, "learning_rate": 2.36674356947267e-05, "loss": 0.0236, "step": 27710 }, { "grad_norm": 0.0441116988658905, "learning_rate": 2.3632305098442004e-05, "loss": 0.0227, "step": 27720 }, { "grad_norm": 0.030455101281404495, "learning_rate": 2.3597192524228156e-05, "loss": 0.0249, "step": 27730 }, { "grad_norm": 0.03506918251514435, "learning_rate": 2.356209799608424e-05, "loss": 0.0256, "step": 27740 }, { "grad_norm": 0.06637784838676453, "learning_rate": 2.352702153799704e-05, "loss": 0.0265, "step": 27750 }, { "grad_norm": 0.05508854240179062, "learning_rate": 2.3491963173941018e-05, "loss": 0.0248, "step": 27760 }, { "grad_norm": 0.043285973370075226, "learning_rate": 2.3456922927878196e-05, "loss": 0.0259, "step": 27770 }, { "grad_norm": 0.0681888684630394, "learning_rate": 2.3421900823758257e-05, "loss": 0.025, "step": 27780 }, { "grad_norm": 0.03379374369978905, "learning_rate": 2.3386896885518496e-05, "loss": 0.0264, "step": 27790 }, { "grad_norm": 0.03584817424416542, "learning_rate": 2.335191113708378e-05, "loss": 0.0259, "step": 27800 }, { "grad_norm": 0.049346406012773514, "learning_rate": 2.331694360236651e-05, "loss": 0.0282, "step": 27810 }, { "grad_norm": 0.03674435243010521, "learning_rate": 2.3281994305266702e-05, "loss": 0.0234, "step": 27820 }, { "grad_norm": 0.03985724598169327, "learning_rate": 2.3247063269671826e-05, "loss": 0.0246, "step": 27830 }, { "grad_norm": 0.02850833348929882, "learning_rate": 2.321215051945695e-05, "loss": 0.0259, "step": 27840 }, { "grad_norm": 0.030431663617491722, "learning_rate": 2.3177256078484588e-05, "loss": 0.0255, "step": 27850 }, { "grad_norm": 0.046498093754053116, "learning_rate": 2.3142379970604798e-05, "loss": 0.023, "step": 27860 }, { "grad_norm": 0.048123590648174286, "learning_rate": 2.3107522219655025e-05, "loss": 0.0251, "step": 27870 }, { "grad_norm": 0.041559331119060516, "learning_rate": 2.3072682849460236e-05, "loss": 0.0244, "step": 27880 }, { "grad_norm": 0.04774794355034828, "learning_rate": 2.303786188383281e-05, "loss": 0.0235, "step": 27890 }, { "grad_norm": 0.04321659728884697, "learning_rate": 2.300305934657257e-05, "loss": 0.0241, "step": 27900 }, { "grad_norm": 0.041511572897434235, "learning_rate": 2.2968275261466677e-05, "loss": 0.0245, "step": 27910 }, { "grad_norm": 0.0402936153113842, "learning_rate": 2.293350965228977e-05, "loss": 0.0264, "step": 27920 }, { "grad_norm": 0.047865159809589386, "learning_rate": 2.2898762542803776e-05, "loss": 0.0264, "step": 27930 }, { "grad_norm": 0.056008849292993546, "learning_rate": 2.286403395675803e-05, "loss": 0.0279, "step": 27940 }, { "grad_norm": 0.042001087218523026, "learning_rate": 2.28293239178892e-05, "loss": 0.026, "step": 27950 }, { "grad_norm": 0.03823995962738991, "learning_rate": 2.2794632449921287e-05, "loss": 0.0273, "step": 27960 }, { "grad_norm": 0.052499182522296906, "learning_rate": 2.275995957656555e-05, "loss": 0.0271, "step": 27970 }, { "grad_norm": 0.040908489376306534, "learning_rate": 2.272530532152058e-05, "loss": 0.0261, "step": 27980 }, { "grad_norm": 0.04203177988529205, "learning_rate": 2.2690669708472233e-05, "loss": 0.0236, "step": 27990 }, { "grad_norm": 0.045121025294065475, "learning_rate": 2.2656052761093655e-05, "loss": 0.0246, "step": 28000 }, { "grad_norm": 0.044839952141046524, "learning_rate": 2.262145450304517e-05, "loss": 0.0251, "step": 28010 }, { "grad_norm": 0.04089561104774475, "learning_rate": 2.2586874957974352e-05, "loss": 0.0262, "step": 28020 }, { "grad_norm": 0.04258809983730316, "learning_rate": 2.2552314149516012e-05, "loss": 0.0256, "step": 28030 }, { "grad_norm": 0.04129762947559357, "learning_rate": 2.2517772101292133e-05, "loss": 0.0244, "step": 28040 }, { "grad_norm": 0.04108230024576187, "learning_rate": 2.248324883691188e-05, "loss": 0.0254, "step": 28050 }, { "grad_norm": 0.0314999595284462, "learning_rate": 2.24487443799716e-05, "loss": 0.0238, "step": 28060 }, { "grad_norm": 0.03693634644150734, "learning_rate": 2.241425875405472e-05, "loss": 0.0232, "step": 28070 }, { "grad_norm": 0.03138679265975952, "learning_rate": 2.2379791982731868e-05, "loss": 0.0226, "step": 28080 }, { "grad_norm": 0.03435776010155678, "learning_rate": 2.2345344089560756e-05, "loss": 0.023, "step": 28090 }, { "grad_norm": 0.044256892055273056, "learning_rate": 2.2310915098086206e-05, "loss": 0.022, "step": 28100 }, { "grad_norm": 0.050092317163944244, "learning_rate": 2.227650503184009e-05, "loss": 0.026, "step": 28110 }, { "grad_norm": 0.03423536941409111, "learning_rate": 2.2242113914341357e-05, "loss": 0.024, "step": 28120 }, { "grad_norm": 0.0415583997964859, "learning_rate": 2.220774176909602e-05, "loss": 0.0268, "step": 28130 }, { "grad_norm": 0.07284404337406158, "learning_rate": 2.2173388619597114e-05, "loss": 0.0268, "step": 28140 }, { "grad_norm": 0.05466019734740257, "learning_rate": 2.21390544893247e-05, "loss": 0.0264, "step": 28150 }, { "grad_norm": 0.03765086829662323, "learning_rate": 2.210473940174585e-05, "loss": 0.0285, "step": 28160 }, { "grad_norm": 0.04390807822346687, "learning_rate": 2.207044338031456e-05, "loss": 0.0245, "step": 28170 }, { "grad_norm": 0.034864746034145355, "learning_rate": 2.203616644847186e-05, "loss": 0.0252, "step": 28180 }, { "grad_norm": 0.03710898384451866, "learning_rate": 2.200190862964571e-05, "loss": 0.0235, "step": 28190 }, { "grad_norm": 0.040098682045936584, "learning_rate": 2.1967669947251024e-05, "loss": 0.0232, "step": 28200 }, { "grad_norm": 0.054116640239953995, "learning_rate": 2.1933450424689583e-05, "loss": 0.0235, "step": 28210 }, { "grad_norm": 0.03819897025823593, "learning_rate": 2.1899250085350142e-05, "loss": 0.0242, "step": 28220 }, { "grad_norm": 0.043332427740097046, "learning_rate": 2.1865068952608277e-05, "loss": 0.0227, "step": 28230 }, { "grad_norm": 0.030520791187882423, "learning_rate": 2.1830907049826487e-05, "loss": 0.0231, "step": 28240 }, { "grad_norm": 0.03601795807480812, "learning_rate": 2.179676440035411e-05, "loss": 0.0236, "step": 28250 }, { "grad_norm": 0.03576650843024254, "learning_rate": 2.1762641027527337e-05, "loss": 0.0255, "step": 28260 }, { "grad_norm": 0.034912314265966415, "learning_rate": 2.1728536954669143e-05, "loss": 0.0246, "step": 28270 }, { "grad_norm": 0.03493485972285271, "learning_rate": 2.169445220508936e-05, "loss": 0.0231, "step": 28280 }, { "grad_norm": 0.05169730260968208, "learning_rate": 2.166038680208461e-05, "loss": 0.0242, "step": 28290 }, { "grad_norm": 0.031728796660900116, "learning_rate": 2.162634076893823e-05, "loss": 0.0251, "step": 28300 }, { "grad_norm": 0.04143313691020012, "learning_rate": 2.1592314128920388e-05, "loss": 0.0252, "step": 28310 }, { "grad_norm": 0.056868020445108414, "learning_rate": 2.155830690528799e-05, "loss": 0.0257, "step": 28320 }, { "grad_norm": 0.044911306351423264, "learning_rate": 2.1524319121284613e-05, "loss": 0.0263, "step": 28330 }, { "grad_norm": 0.041519343852996826, "learning_rate": 2.1490350800140607e-05, "loss": 0.0269, "step": 28340 }, { "grad_norm": 0.04984421655535698, "learning_rate": 2.1456401965073002e-05, "loss": 0.0236, "step": 28350 }, { "grad_norm": 0.03556450456380844, "learning_rate": 2.1422472639285524e-05, "loss": 0.0248, "step": 28360 }, { "grad_norm": 0.052456334233284, "learning_rate": 2.13885628459685e-05, "loss": 0.025, "step": 28370 }, { "grad_norm": 0.033078063279390335, "learning_rate": 2.135467260829901e-05, "loss": 0.0218, "step": 28380 }, { "grad_norm": 0.03790844604372978, "learning_rate": 2.1320801949440654e-05, "loss": 0.0218, "step": 28390 }, { "grad_norm": 0.03180667757987976, "learning_rate": 2.1286950892543744e-05, "loss": 0.0213, "step": 28400 }, { "grad_norm": 0.040669262409210205, "learning_rate": 2.125311946074515e-05, "loss": 0.0213, "step": 28410 }, { "grad_norm": 0.04328538477420807, "learning_rate": 2.1219307677168355e-05, "loss": 0.0225, "step": 28420 }, { "grad_norm": 0.031209571287035942, "learning_rate": 2.118551556492336e-05, "loss": 0.0211, "step": 28430 }, { "grad_norm": 0.03014606051146984, "learning_rate": 2.1151743147106774e-05, "loss": 0.0221, "step": 28440 }, { "grad_norm": 0.06238967552781105, "learning_rate": 2.111799044680172e-05, "loss": 0.0217, "step": 28450 }, { "grad_norm": 0.035210754722356796, "learning_rate": 2.1084257487077873e-05, "loss": 0.022, "step": 28460 }, { "grad_norm": 0.037764888256788254, "learning_rate": 2.1050544290991357e-05, "loss": 0.0224, "step": 28470 }, { "grad_norm": 0.03930547460913658, "learning_rate": 2.101685088158486e-05, "loss": 0.0237, "step": 28480 }, { "grad_norm": 0.026478027924895287, "learning_rate": 2.0983177281887472e-05, "loss": 0.0229, "step": 28490 }, { "grad_norm": 0.03229299932718277, "learning_rate": 2.0949523514914798e-05, "loss": 0.0215, "step": 28500 }, { "grad_norm": 0.05216970294713974, "learning_rate": 2.0915889603668876e-05, "loss": 0.0233, "step": 28510 }, { "grad_norm": 0.041040826588869095, "learning_rate": 2.0882275571138175e-05, "loss": 0.0232, "step": 28520 }, { "grad_norm": 0.04093573987483978, "learning_rate": 2.0848681440297545e-05, "loss": 0.022, "step": 28530 }, { "grad_norm": 0.03460615873336792, "learning_rate": 2.081510723410827e-05, "loss": 0.0224, "step": 28540 }, { "grad_norm": 0.08274950087070465, "learning_rate": 2.0781552975518003e-05, "loss": 0.0241, "step": 28550 }, { "grad_norm": 0.03473050147294998, "learning_rate": 2.074801868746078e-05, "loss": 0.023, "step": 28560 }, { "grad_norm": 0.038741689175367355, "learning_rate": 2.0714504392856955e-05, "loss": 0.0224, "step": 28570 }, { "grad_norm": 0.03422350063920021, "learning_rate": 2.0681010114613215e-05, "loss": 0.0235, "step": 28580 }, { "grad_norm": 0.04289880394935608, "learning_rate": 2.0647535875622597e-05, "loss": 0.024, "step": 28590 }, { "grad_norm": 0.037730153650045395, "learning_rate": 2.0614081698764432e-05, "loss": 0.0224, "step": 28600 }, { "grad_norm": 0.05413195118308067, "learning_rate": 2.0580647606904334e-05, "loss": 0.0243, "step": 28610 }, { "grad_norm": 0.029056526720523834, "learning_rate": 2.0547233622894208e-05, "loss": 0.0255, "step": 28620 }, { "grad_norm": 0.04007687792181969, "learning_rate": 2.0513839769572157e-05, "loss": 0.0252, "step": 28630 }, { "grad_norm": 0.04933862388134003, "learning_rate": 2.0480466069762584e-05, "loss": 0.0253, "step": 28640 }, { "grad_norm": 0.03576682507991791, "learning_rate": 2.0447112546276104e-05, "loss": 0.0231, "step": 28650 }, { "grad_norm": 0.05756193399429321, "learning_rate": 2.0413779221909547e-05, "loss": 0.0259, "step": 28660 }, { "grad_norm": 0.03426724672317505, "learning_rate": 2.0380466119445912e-05, "loss": 0.0243, "step": 28670 }, { "grad_norm": 0.07894433289766312, "learning_rate": 2.0347173261654373e-05, "loss": 0.025, "step": 28680 }, { "grad_norm": 0.03768209367990494, "learning_rate": 2.03139006712903e-05, "loss": 0.023, "step": 28690 }, { "grad_norm": 0.05792359635233879, "learning_rate": 2.028064837109519e-05, "loss": 0.0231, "step": 28700 }, { "grad_norm": 0.038888879120349884, "learning_rate": 2.0247416383796685e-05, "loss": 0.023, "step": 28710 }, { "grad_norm": 0.04775875061750412, "learning_rate": 2.0214204732108548e-05, "loss": 0.0223, "step": 28720 }, { "grad_norm": 0.035211872309446335, "learning_rate": 2.0181013438730596e-05, "loss": 0.0212, "step": 28730 }, { "grad_norm": 0.04904521629214287, "learning_rate": 2.0147842526348783e-05, "loss": 0.0235, "step": 28740 }, { "grad_norm": 0.05016692727804184, "learning_rate": 2.011469201763511e-05, "loss": 0.023, "step": 28750 }, { "grad_norm": 0.04193210229277611, "learning_rate": 2.0081561935247665e-05, "loss": 0.0255, "step": 28760 }, { "grad_norm": 0.03405604511499405, "learning_rate": 2.0048452301830523e-05, "loss": 0.0257, "step": 28770 }, { "grad_norm": 0.045563917607069016, "learning_rate": 2.0015363140013788e-05, "loss": 0.0253, "step": 28780 }, { "grad_norm": 0.03510799631476402, "learning_rate": 1.9982294472413606e-05, "loss": 0.0239, "step": 28790 }, { "grad_norm": 0.04913138225674629, "learning_rate": 1.9949246321632103e-05, "loss": 0.0235, "step": 28800 }, { "grad_norm": 0.0571175292134285, "learning_rate": 1.9916218710257377e-05, "loss": 0.0234, "step": 28810 }, { "grad_norm": 0.051827047020196915, "learning_rate": 1.988321166086351e-05, "loss": 0.0242, "step": 28820 }, { "grad_norm": 0.02648179791867733, "learning_rate": 1.9850225196010468e-05, "loss": 0.0232, "step": 28830 }, { "grad_norm": 0.0414431169629097, "learning_rate": 1.981725933824421e-05, "loss": 0.0239, "step": 28840 }, { "grad_norm": 0.03926323354244232, "learning_rate": 1.978431411009661e-05, "loss": 0.0242, "step": 28850 }, { "grad_norm": 0.033052898943424225, "learning_rate": 1.9751389534085375e-05, "loss": 0.0224, "step": 28860 }, { "grad_norm": 0.04243791848421097, "learning_rate": 1.9718485632714184e-05, "loss": 0.0237, "step": 28870 }, { "grad_norm": 0.05042881518602371, "learning_rate": 1.968560242847251e-05, "loss": 0.0228, "step": 28880 }, { "grad_norm": 0.06805835664272308, "learning_rate": 1.965273994383573e-05, "loss": 0.023, "step": 28890 }, { "grad_norm": 0.037190161645412445, "learning_rate": 1.961989820126504e-05, "loss": 0.0256, "step": 28900 }, { "grad_norm": 0.04751140996813774, "learning_rate": 1.958707722320746e-05, "loss": 0.022, "step": 28910 }, { "grad_norm": 0.030458224937319756, "learning_rate": 1.955427703209584e-05, "loss": 0.0248, "step": 28920 }, { "grad_norm": 0.035640228539705276, "learning_rate": 1.9521497650348764e-05, "loss": 0.0248, "step": 28930 }, { "grad_norm": 0.03986964374780655, "learning_rate": 1.948873910037067e-05, "loss": 0.0238, "step": 28940 }, { "grad_norm": 0.045549873262643814, "learning_rate": 1.9456001404551678e-05, "loss": 0.027, "step": 28950 }, { "grad_norm": 0.04229596257209778, "learning_rate": 1.942328458526771e-05, "loss": 0.0256, "step": 28960 }, { "grad_norm": 0.04050430655479431, "learning_rate": 1.9390588664880427e-05, "loss": 0.0248, "step": 28970 }, { "grad_norm": 0.04315851628780365, "learning_rate": 1.9357913665737145e-05, "loss": 0.0251, "step": 28980 }, { "grad_norm": 0.03940408304333687, "learning_rate": 1.932525961017093e-05, "loss": 0.0245, "step": 28990 }, { "grad_norm": 0.06669913977384567, "learning_rate": 1.9292626520500533e-05, "loss": 0.0255, "step": 29000 }, { "grad_norm": 0.03159944713115692, "learning_rate": 1.9260014419030354e-05, "loss": 0.0246, "step": 29010 }, { "grad_norm": 0.1186399906873703, "learning_rate": 1.9227423328050475e-05, "loss": 0.0244, "step": 29020 }, { "grad_norm": 0.047463759779930115, "learning_rate": 1.9194853269836582e-05, "loss": 0.0245, "step": 29030 }, { "grad_norm": 0.028051650151610374, "learning_rate": 1.916230426664999e-05, "loss": 0.0241, "step": 29040 }, { "grad_norm": 0.04315502941608429, "learning_rate": 1.912977634073765e-05, "loss": 0.0247, "step": 29050 }, { "grad_norm": 0.03302328288555145, "learning_rate": 1.9097269514332083e-05, "loss": 0.0221, "step": 29060 }, { "grad_norm": 0.047692492604255676, "learning_rate": 1.9064783809651433e-05, "loss": 0.0217, "step": 29070 }, { "grad_norm": 0.03863929584622383, "learning_rate": 1.9032319248899333e-05, "loss": 0.0249, "step": 29080 }, { "grad_norm": 0.03528451547026634, "learning_rate": 1.8999875854265015e-05, "loss": 0.0249, "step": 29090 }, { "grad_norm": 0.03595513477921486, "learning_rate": 1.8967453647923232e-05, "loss": 0.0283, "step": 29100 }, { "grad_norm": 0.039214327931404114, "learning_rate": 1.893505265203427e-05, "loss": 0.0281, "step": 29110 }, { "grad_norm": 0.037795279175043106, "learning_rate": 1.8902672888743907e-05, "loss": 0.0255, "step": 29120 }, { "grad_norm": 0.029169045388698578, "learning_rate": 1.8870314380183396e-05, "loss": 0.0262, "step": 29130 }, { "grad_norm": 0.042625267058610916, "learning_rate": 1.8837977148469448e-05, "loss": 0.0226, "step": 29140 }, { "grad_norm": 0.061229195445775986, "learning_rate": 1.880566121570429e-05, "loss": 0.0246, "step": 29150 }, { "grad_norm": 0.038261931389570236, "learning_rate": 1.877336660397554e-05, "loss": 0.0231, "step": 29160 }, { "grad_norm": 0.027746355161070824, "learning_rate": 1.874109333535628e-05, "loss": 0.0226, "step": 29170 }, { "grad_norm": 0.02227967604994774, "learning_rate": 1.870884143190496e-05, "loss": 0.0217, "step": 29180 }, { "grad_norm": 0.03499946743249893, "learning_rate": 1.867661091566546e-05, "loss": 0.0207, "step": 29190 }, { "grad_norm": 0.046982791274785995, "learning_rate": 1.864440180866704e-05, "loss": 0.022, "step": 29200 }, { "grad_norm": 0.042514700442552567, "learning_rate": 1.8612214132924317e-05, "loss": 0.0216, "step": 29210 }, { "grad_norm": 0.03375662863254547, "learning_rate": 1.858004791043728e-05, "loss": 0.0224, "step": 29220 }, { "grad_norm": 0.04742666706442833, "learning_rate": 1.854790316319123e-05, "loss": 0.0231, "step": 29230 }, { "grad_norm": 0.04050751030445099, "learning_rate": 1.8515779913156766e-05, "loss": 0.0221, "step": 29240 }, { "grad_norm": 0.04903649538755417, "learning_rate": 1.848367818228986e-05, "loss": 0.022, "step": 29250 }, { "grad_norm": 0.04851599782705307, "learning_rate": 1.8451597992531733e-05, "loss": 0.0227, "step": 29260 }, { "grad_norm": 0.03919871523976326, "learning_rate": 1.8419539365808914e-05, "loss": 0.0243, "step": 29270 }, { "grad_norm": 0.04366392269730568, "learning_rate": 1.838750232403313e-05, "loss": 0.0221, "step": 29280 }, { "grad_norm": 0.041741736233234406, "learning_rate": 1.835548688910142e-05, "loss": 0.023, "step": 29290 }, { "grad_norm": 0.048385411500930786, "learning_rate": 1.8323493082896037e-05, "loss": 0.0209, "step": 29300 }, { "grad_norm": 0.03629417344927788, "learning_rate": 1.8291520927284454e-05, "loss": 0.0214, "step": 29310 }, { "grad_norm": 0.05200496315956116, "learning_rate": 1.8259570444119305e-05, "loss": 0.0217, "step": 29320 }, { "grad_norm": 0.03767458721995354, "learning_rate": 1.8227641655238488e-05, "loss": 0.0206, "step": 29330 }, { "grad_norm": 0.05372998118400574, "learning_rate": 1.819573458246498e-05, "loss": 0.0233, "step": 29340 }, { "grad_norm": 0.035634152591228485, "learning_rate": 1.816384924760699e-05, "loss": 0.0218, "step": 29350 }, { "grad_norm": 0.03440430760383606, "learning_rate": 1.813198567245784e-05, "loss": 0.0226, "step": 29360 }, { "grad_norm": 0.03775963559746742, "learning_rate": 1.8100143878796006e-05, "loss": 0.0216, "step": 29370 }, { "grad_norm": 0.044724855571985245, "learning_rate": 1.8068323888385015e-05, "loss": 0.0244, "step": 29380 }, { "grad_norm": 0.06092628836631775, "learning_rate": 1.803652572297355e-05, "loss": 0.0249, "step": 29390 }, { "grad_norm": 0.05645186826586723, "learning_rate": 1.8004749404295353e-05, "loss": 0.0264, "step": 29400 }, { "grad_norm": 0.05125168338418007, "learning_rate": 1.797299495406926e-05, "loss": 0.0271, "step": 29410 }, { "grad_norm": 0.03605259582400322, "learning_rate": 1.7941262393999103e-05, "loss": 0.0258, "step": 29420 }, { "grad_norm": 0.0398700013756752, "learning_rate": 1.7909551745773816e-05, "loss": 0.0244, "step": 29430 }, { "grad_norm": 0.033219873905181885, "learning_rate": 1.7877863031067304e-05, "loss": 0.0263, "step": 29440 }, { "grad_norm": 0.04357511177659035, "learning_rate": 1.7846196271538516e-05, "loss": 0.0281, "step": 29450 }, { "grad_norm": 0.03738533705472946, "learning_rate": 1.7814551488831384e-05, "loss": 0.0285, "step": 29460 }, { "grad_norm": 0.05857086554169655, "learning_rate": 1.7782928704574835e-05, "loss": 0.0273, "step": 29470 }, { "grad_norm": 0.04240554943680763, "learning_rate": 1.775132794038271e-05, "loss": 0.0259, "step": 29480 }, { "grad_norm": 0.05849405750632286, "learning_rate": 1.7719749217853855e-05, "loss": 0.0263, "step": 29490 }, { "grad_norm": 0.03246699273586273, "learning_rate": 1.7688192558572038e-05, "loss": 0.0248, "step": 29500 }, { "grad_norm": 0.06867790222167969, "learning_rate": 1.7656657984105906e-05, "loss": 0.0244, "step": 29510 }, { "grad_norm": 0.043995074927806854, "learning_rate": 1.7625145516009068e-05, "loss": 0.0255, "step": 29520 }, { "grad_norm": 0.04639024659991264, "learning_rate": 1.7593655175820005e-05, "loss": 0.0237, "step": 29530 }, { "grad_norm": 0.03963720053434372, "learning_rate": 1.7562186985062046e-05, "loss": 0.0249, "step": 29540 }, { "grad_norm": 0.04184795171022415, "learning_rate": 1.7530740965243403e-05, "loss": 0.0222, "step": 29550 }, { "grad_norm": 0.0778517872095108, "learning_rate": 1.7499317137857153e-05, "loss": 0.0213, "step": 29560 }, { "grad_norm": 0.06613922119140625, "learning_rate": 1.7467915524381184e-05, "loss": 0.0226, "step": 29570 }, { "grad_norm": 0.03754349797964096, "learning_rate": 1.7436536146278182e-05, "loss": 0.0243, "step": 29580 }, { "grad_norm": 0.04870352894067764, "learning_rate": 1.7405179024995688e-05, "loss": 0.0278, "step": 29590 }, { "grad_norm": 0.0509861558675766, "learning_rate": 1.737384418196596e-05, "loss": 0.0274, "step": 29600 }, { "grad_norm": 0.04885049909353256, "learning_rate": 1.734253163860609e-05, "loss": 0.0269, "step": 29610 }, { "grad_norm": 0.032920319586992264, "learning_rate": 1.7311241416317896e-05, "loss": 0.0239, "step": 29620 }, { "grad_norm": 0.09906832873821259, "learning_rate": 1.7279973536487982e-05, "loss": 0.0261, "step": 29630 }, { "grad_norm": 0.052919935435056686, "learning_rate": 1.724872802048761e-05, "loss": 0.0266, "step": 29640 }, { "grad_norm": 0.10147158801555634, "learning_rate": 1.7217504889672803e-05, "loss": 0.0243, "step": 29650 }, { "grad_norm": 0.04450851306319237, "learning_rate": 1.7186304165384287e-05, "loss": 0.0228, "step": 29660 }, { "grad_norm": 0.04547782242298126, "learning_rate": 1.7155125868947475e-05, "loss": 0.024, "step": 29670 }, { "grad_norm": 0.05128708854317665, "learning_rate": 1.7123970021672404e-05, "loss": 0.0245, "step": 29680 }, { "grad_norm": 0.04557230323553085, "learning_rate": 1.709283664485384e-05, "loss": 0.022, "step": 29690 }, { "grad_norm": 0.035388581454753876, "learning_rate": 1.7061725759771113e-05, "loss": 0.0223, "step": 29700 }, { "grad_norm": 0.05114679038524628, "learning_rate": 1.7030637387688248e-05, "loss": 0.0244, "step": 29710 }, { "grad_norm": 0.04725366830825806, "learning_rate": 1.6999571549853836e-05, "loss": 0.0211, "step": 29720 }, { "grad_norm": 0.045777179300785065, "learning_rate": 1.696852826750112e-05, "loss": 0.0222, "step": 29730 }, { "grad_norm": 0.047655172646045685, "learning_rate": 1.6937507561847844e-05, "loss": 0.0227, "step": 29740 }, { "grad_norm": 0.0500541515648365, "learning_rate": 1.6906509454096385e-05, "loss": 0.0232, "step": 29750 }, { "grad_norm": 0.13640174269676208, "learning_rate": 1.687553396543367e-05, "loss": 0.0222, "step": 29760 }, { "grad_norm": 0.044246915727853775, "learning_rate": 1.6844581117031154e-05, "loss": 0.0219, "step": 29770 }, { "grad_norm": 0.040537308901548386, "learning_rate": 1.681365093004481e-05, "loss": 0.0222, "step": 29780 }, { "grad_norm": 0.08399252593517303, "learning_rate": 1.678274342561511e-05, "loss": 0.0233, "step": 29790 }, { "grad_norm": 0.03915110602974892, "learning_rate": 1.675185862486706e-05, "loss": 0.022, "step": 29800 }, { "grad_norm": 0.04285934567451477, "learning_rate": 1.6720996548910127e-05, "loss": 0.0224, "step": 29810 }, { "grad_norm": 0.04563598334789276, "learning_rate": 1.6690157218838247e-05, "loss": 0.0223, "step": 29820 }, { "grad_norm": 0.03976026177406311, "learning_rate": 1.665934065572984e-05, "loss": 0.0241, "step": 29830 }, { "grad_norm": 0.060388341546058655, "learning_rate": 1.6628546880647688e-05, "loss": 0.0259, "step": 29840 }, { "grad_norm": 0.03643405809998512, "learning_rate": 1.6597775914639076e-05, "loss": 0.0246, "step": 29850 }, { "grad_norm": 0.03948723524808884, "learning_rate": 1.6567027778735654e-05, "loss": 0.0239, "step": 29860 }, { "grad_norm": 0.039187561720609665, "learning_rate": 1.653630249395351e-05, "loss": 0.024, "step": 29870 }, { "grad_norm": 0.04541468620300293, "learning_rate": 1.6505600081293072e-05, "loss": 0.024, "step": 29880 }, { "grad_norm": 0.04098577797412872, "learning_rate": 1.647492056173912e-05, "loss": 0.0258, "step": 29890 }, { "grad_norm": 0.0377429835498333, "learning_rate": 1.644426395626085e-05, "loss": 0.0232, "step": 29900 }, { "grad_norm": 0.037036627531051636, "learning_rate": 1.641363028581175e-05, "loss": 0.0257, "step": 29910 }, { "grad_norm": 0.04790761321783066, "learning_rate": 1.638301957132965e-05, "loss": 0.0279, "step": 29920 }, { "grad_norm": 0.05382158234715462, "learning_rate": 1.6352431833736703e-05, "loss": 0.0264, "step": 29930 }, { "grad_norm": 0.0362967811524868, "learning_rate": 1.6321867093939298e-05, "loss": 0.0264, "step": 29940 }, { "grad_norm": 0.27284878492355347, "learning_rate": 1.629132537282817e-05, "loss": 0.0265, "step": 29950 }, { "grad_norm": 0.05626853182911873, "learning_rate": 1.62608066912783e-05, "loss": 0.0252, "step": 29960 }, { "grad_norm": 0.05464104935526848, "learning_rate": 1.623031107014893e-05, "loss": 0.0234, "step": 29970 }, { "grad_norm": 0.04691712558269501, "learning_rate": 1.619983853028351e-05, "loss": 0.0237, "step": 29980 }, { "grad_norm": 0.03376277908682823, "learning_rate": 1.6169389092509724e-05, "loss": 0.0229, "step": 29990 }, { "grad_norm": 0.0456075444817543, "learning_rate": 1.6138962777639494e-05, "loss": 0.0255, "step": 30000 }, { "grad_norm": 0.04175880178809166, "learning_rate": 1.610855960646891e-05, "loss": 0.0241, "step": 30010 }, { "grad_norm": 0.05184519663453102, "learning_rate": 1.607817959977826e-05, "loss": 0.0266, "step": 30020 }, { "grad_norm": 0.07915632426738739, "learning_rate": 1.6047822778332005e-05, "loss": 0.0248, "step": 30030 }, { "grad_norm": 0.039955127984285355, "learning_rate": 1.6017489162878713e-05, "loss": 0.029, "step": 30040 }, { "grad_norm": 0.04298907145857811, "learning_rate": 1.5987178774151147e-05, "loss": 0.0256, "step": 30050 }, { "grad_norm": 0.03676952049136162, "learning_rate": 1.5956891632866184e-05, "loss": 0.0261, "step": 30060 }, { "grad_norm": 0.045373547822237015, "learning_rate": 1.592662775972476e-05, "loss": 0.0273, "step": 30070 }, { "grad_norm": 0.07277703285217285, "learning_rate": 1.5896387175411986e-05, "loss": 0.0249, "step": 30080 }, { "grad_norm": 0.029851051047444344, "learning_rate": 1.586616990059699e-05, "loss": 0.0238, "step": 30090 }, { "grad_norm": 0.03332846611738205, "learning_rate": 1.5835975955933002e-05, "loss": 0.022, "step": 30100 }, { "grad_norm": 0.06608742475509644, "learning_rate": 1.5805805362057295e-05, "loss": 0.0238, "step": 30110 }, { "grad_norm": 0.039395421743392944, "learning_rate": 1.57756581395912e-05, "loss": 0.0228, "step": 30120 }, { "grad_norm": 0.13230709731578827, "learning_rate": 1.574553430914006e-05, "loss": 0.0232, "step": 30130 }, { "grad_norm": 0.03844660148024559, "learning_rate": 1.5715433891293206e-05, "loss": 0.0224, "step": 30140 }, { "grad_norm": 0.09296836704015732, "learning_rate": 1.5685356906624016e-05, "loss": 0.0227, "step": 30150 }, { "grad_norm": 0.05768269672989845, "learning_rate": 1.56553033756898e-05, "loss": 0.0244, "step": 30160 }, { "grad_norm": 0.058506693691015244, "learning_rate": 1.5625273319031873e-05, "loss": 0.024, "step": 30170 }, { "grad_norm": 0.06759831309318542, "learning_rate": 1.5595266757175518e-05, "loss": 0.0252, "step": 30180 }, { "grad_norm": 0.037631623446941376, "learning_rate": 1.55652837106299e-05, "loss": 0.0255, "step": 30190 }, { "grad_norm": 0.045434657484292984, "learning_rate": 1.5535324199888164e-05, "loss": 0.0236, "step": 30200 }, { "grad_norm": 0.0766875371336937, "learning_rate": 1.5505388245427355e-05, "loss": 0.0234, "step": 30210 }, { "grad_norm": 0.05254736542701721, "learning_rate": 1.5475475867708418e-05, "loss": 0.0235, "step": 30220 }, { "grad_norm": 0.03970908746123314, "learning_rate": 1.5445587087176195e-05, "loss": 0.0225, "step": 30230 }, { "grad_norm": 0.08768696337938309, "learning_rate": 1.5415721924259346e-05, "loss": 0.0222, "step": 30240 }, { "grad_norm": 0.040029097348451614, "learning_rate": 1.538588039937047e-05, "loss": 0.0214, "step": 30250 }, { "grad_norm": 0.04507242143154144, "learning_rate": 1.535606253290594e-05, "loss": 0.0237, "step": 30260 }, { "grad_norm": 0.05643310025334358, "learning_rate": 1.5326268345245993e-05, "loss": 0.0227, "step": 30270 }, { "grad_norm": 0.03222034499049187, "learning_rate": 1.5296497856754693e-05, "loss": 0.0235, "step": 30280 }, { "grad_norm": 0.0456925667822361, "learning_rate": 1.5266751087779858e-05, "loss": 0.0257, "step": 30290 }, { "grad_norm": 0.045035120099782944, "learning_rate": 1.5237028058653142e-05, "loss": 0.0227, "step": 30300 }, { "grad_norm": 0.034548208117485046, "learning_rate": 1.5207328789689946e-05, "loss": 0.0234, "step": 30310 }, { "grad_norm": 0.04152727127075195, "learning_rate": 1.5177653301189454e-05, "loss": 0.0236, "step": 30320 }, { "grad_norm": 0.058033835142850876, "learning_rate": 1.5148001613434587e-05, "loss": 0.0246, "step": 30330 }, { "grad_norm": 0.03674709051847458, "learning_rate": 1.5118373746691983e-05, "loss": 0.0215, "step": 30340 }, { "grad_norm": 0.05320112779736519, "learning_rate": 1.5088769721211992e-05, "loss": 0.0219, "step": 30350 }, { "grad_norm": 0.031050359830260277, "learning_rate": 1.505918955722871e-05, "loss": 0.0237, "step": 30360 }, { "grad_norm": 0.03447182476520538, "learning_rate": 1.5029633274959886e-05, "loss": 0.0221, "step": 30370 }, { "grad_norm": 0.03257173299789429, "learning_rate": 1.5000100894606989e-05, "loss": 0.024, "step": 30380 }, { "grad_norm": 0.05575406923890114, "learning_rate": 1.4970592436355079e-05, "loss": 0.0226, "step": 30390 }, { "grad_norm": 0.03911479562520981, "learning_rate": 1.4941107920372938e-05, "loss": 0.0256, "step": 30400 }, { "grad_norm": 0.043278370052576065, "learning_rate": 1.4911647366812947e-05, "loss": 0.0238, "step": 30410 }, { "grad_norm": 0.04424995556473732, "learning_rate": 1.4882210795811114e-05, "loss": 0.0235, "step": 30420 }, { "grad_norm": 0.03277837857604027, "learning_rate": 1.485279822748708e-05, "loss": 0.0228, "step": 30430 }, { "grad_norm": 0.038805294781923294, "learning_rate": 1.4823409681944034e-05, "loss": 0.0227, "step": 30440 }, { "grad_norm": 0.033775538206100464, "learning_rate": 1.4794045179268763e-05, "loss": 0.0233, "step": 30450 }, { "grad_norm": 0.04156673327088356, "learning_rate": 1.4764704739531643e-05, "loss": 0.022, "step": 30460 }, { "grad_norm": 0.06737097352743149, "learning_rate": 1.473538838278658e-05, "loss": 0.0225, "step": 30470 }, { "grad_norm": 0.031252481043338776, "learning_rate": 1.4706096129071045e-05, "loss": 0.0222, "step": 30480 }, { "grad_norm": 0.03839314356446266, "learning_rate": 1.4676827998405996e-05, "loss": 0.0224, "step": 30490 }, { "grad_norm": 0.03229150176048279, "learning_rate": 1.4647584010795928e-05, "loss": 0.0209, "step": 30500 }, { "grad_norm": 0.029047101736068726, "learning_rate": 1.4618364186228834e-05, "loss": 0.0229, "step": 30510 }, { "grad_norm": 0.030410580337047577, "learning_rate": 1.4589168544676207e-05, "loss": 0.0226, "step": 30520 }, { "grad_norm": 0.04155963286757469, "learning_rate": 1.455999710609296e-05, "loss": 0.0219, "step": 30530 }, { "grad_norm": 0.05277044698596001, "learning_rate": 1.453084989041753e-05, "loss": 0.0216, "step": 30540 }, { "grad_norm": 0.0480753518640995, "learning_rate": 1.4501726917571728e-05, "loss": 0.0215, "step": 30550 }, { "grad_norm": 0.16072197258472443, "learning_rate": 1.4472628207460848e-05, "loss": 0.021, "step": 30560 }, { "grad_norm": 0.036718644201755524, "learning_rate": 1.4443553779973584e-05, "loss": 0.0223, "step": 30570 }, { "grad_norm": 0.04860648885369301, "learning_rate": 1.4414503654982037e-05, "loss": 0.0228, "step": 30580 }, { "grad_norm": 0.044646069407463074, "learning_rate": 1.4385477852341706e-05, "loss": 0.0236, "step": 30590 }, { "grad_norm": 0.03561843931674957, "learning_rate": 1.4356476391891421e-05, "loss": 0.0223, "step": 30600 }, { "grad_norm": 0.04003528505563736, "learning_rate": 1.4327499293453423e-05, "loss": 0.0243, "step": 30610 }, { "grad_norm": 0.03714337944984436, "learning_rate": 1.4298546576833304e-05, "loss": 0.0213, "step": 30620 }, { "grad_norm": 0.03141923248767853, "learning_rate": 1.426961826181994e-05, "loss": 0.0218, "step": 30630 }, { "grad_norm": 0.05357448011636734, "learning_rate": 1.4240714368185592e-05, "loss": 0.0215, "step": 30640 }, { "grad_norm": 0.03035716526210308, "learning_rate": 1.4211834915685767e-05, "loss": 0.0223, "step": 30650 }, { "grad_norm": 0.04429818317294121, "learning_rate": 1.4182979924059313e-05, "loss": 0.0204, "step": 30660 }, { "grad_norm": 0.06705047190189362, "learning_rate": 1.415414941302835e-05, "loss": 0.0229, "step": 30670 }, { "grad_norm": 0.04697441682219505, "learning_rate": 1.4125343402298257e-05, "loss": 0.0227, "step": 30680 }, { "grad_norm": 0.03545653447508812, "learning_rate": 1.4096561911557688e-05, "loss": 0.0212, "step": 30690 }, { "grad_norm": 0.04257125407457352, "learning_rate": 1.4067804960478481e-05, "loss": 0.0217, "step": 30700 }, { "grad_norm": 0.04810172691941261, "learning_rate": 1.4039072568715784e-05, "loss": 0.0211, "step": 30710 }, { "grad_norm": 0.03738876059651375, "learning_rate": 1.401036475590788e-05, "loss": 0.0237, "step": 30720 }, { "grad_norm": 0.035576630383729935, "learning_rate": 1.3981681541676301e-05, "loss": 0.0236, "step": 30730 }, { "grad_norm": 0.0456770658493042, "learning_rate": 1.3953022945625772e-05, "loss": 0.0253, "step": 30740 }, { "grad_norm": 0.05638749524950981, "learning_rate": 1.3924388987344138e-05, "loss": 0.0229, "step": 30750 }, { "grad_norm": 0.030574480071663857, "learning_rate": 1.3895779686402465e-05, "loss": 0.0227, "step": 30760 }, { "grad_norm": 0.04061077907681465, "learning_rate": 1.3867195062354931e-05, "loss": 0.0247, "step": 30770 }, { "grad_norm": 0.04390500485897064, "learning_rate": 1.3838635134738858e-05, "loss": 0.0234, "step": 30780 }, { "grad_norm": 0.044929180294275284, "learning_rate": 1.3810099923074704e-05, "loss": 0.0232, "step": 30790 }, { "grad_norm": 0.059412021189928055, "learning_rate": 1.3781589446866006e-05, "loss": 0.023, "step": 30800 }, { "grad_norm": 0.041990865021944046, "learning_rate": 1.3753103725599387e-05, "loss": 0.0206, "step": 30810 }, { "grad_norm": 0.04544004797935486, "learning_rate": 1.3724642778744579e-05, "loss": 0.0226, "step": 30820 }, { "grad_norm": 0.05534037947654724, "learning_rate": 1.3696206625754376e-05, "loss": 0.0214, "step": 30830 }, { "grad_norm": 0.06869778782129288, "learning_rate": 1.3667795286064639e-05, "loss": 0.025, "step": 30840 }, { "grad_norm": 0.03551354631781578, "learning_rate": 1.3639408779094215e-05, "loss": 0.0228, "step": 30850 }, { "grad_norm": 0.04990348592400551, "learning_rate": 1.361104712424503e-05, "loss": 0.0237, "step": 30860 }, { "grad_norm": 0.04079180583357811, "learning_rate": 1.358271034090201e-05, "loss": 0.0249, "step": 30870 }, { "grad_norm": 0.037688933312892914, "learning_rate": 1.3554398448433076e-05, "loss": 0.0241, "step": 30880 }, { "grad_norm": 0.02981441468000412, "learning_rate": 1.3526111466189156e-05, "loss": 0.0227, "step": 30890 }, { "grad_norm": 0.05300643667578697, "learning_rate": 1.3497849413504121e-05, "loss": 0.024, "step": 30900 }, { "grad_norm": 0.03890770301222801, "learning_rate": 1.3469612309694802e-05, "loss": 0.0232, "step": 30910 }, { "grad_norm": 0.03133013844490051, "learning_rate": 1.3441400174061008e-05, "loss": 0.022, "step": 30920 }, { "grad_norm": 0.043500207364559174, "learning_rate": 1.3413213025885469e-05, "loss": 0.02, "step": 30930 }, { "grad_norm": 0.0369785837829113, "learning_rate": 1.3385050884433853e-05, "loss": 0.0222, "step": 30940 }, { "grad_norm": 0.07237369567155838, "learning_rate": 1.3356913768954676e-05, "loss": 0.0204, "step": 30950 }, { "grad_norm": 0.04782838001847267, "learning_rate": 1.3328801698679415e-05, "loss": 0.0215, "step": 30960 }, { "grad_norm": 0.056151457130908966, "learning_rate": 1.3300714692822413e-05, "loss": 0.0225, "step": 30970 }, { "grad_norm": 0.03910408169031143, "learning_rate": 1.3272652770580857e-05, "loss": 0.0243, "step": 30980 }, { "grad_norm": 0.041583821177482605, "learning_rate": 1.3244615951134831e-05, "loss": 0.023, "step": 30990 }, { "grad_norm": 0.036988068372011185, "learning_rate": 1.3216604253647219e-05, "loss": 0.0251, "step": 31000 }, { "grad_norm": 0.114605613052845, "learning_rate": 1.3188617697263738e-05, "loss": 0.0243, "step": 31010 }, { "grad_norm": 0.041806530207395554, "learning_rate": 1.3160656301112956e-05, "loss": 0.0246, "step": 31020 }, { "grad_norm": 0.03574593737721443, "learning_rate": 1.3132720084306211e-05, "loss": 0.0244, "step": 31030 }, { "grad_norm": 0.18542297184467316, "learning_rate": 1.3104809065937668e-05, "loss": 0.0255, "step": 31040 }, { "grad_norm": 0.05446215718984604, "learning_rate": 1.3076923265084212e-05, "loss": 0.0239, "step": 31050 }, { "grad_norm": 0.04294700548052788, "learning_rate": 1.3049062700805543e-05, "loss": 0.0239, "step": 31060 }, { "grad_norm": 0.07313230633735657, "learning_rate": 1.3021227392144092e-05, "loss": 0.0236, "step": 31070 }, { "grad_norm": 0.04504365846514702, "learning_rate": 1.2993417358125037e-05, "loss": 0.0236, "step": 31080 }, { "grad_norm": 0.044858433306217194, "learning_rate": 1.2965632617756257e-05, "loss": 0.0231, "step": 31090 }, { "grad_norm": 0.030425144359469414, "learning_rate": 1.2937873190028387e-05, "loss": 0.0234, "step": 31100 }, { "grad_norm": 0.04064952954649925, "learning_rate": 1.2910139093914697e-05, "loss": 0.023, "step": 31110 }, { "grad_norm": 0.03991818055510521, "learning_rate": 1.2882430348371205e-05, "loss": 0.0216, "step": 31120 }, { "grad_norm": 0.03345935046672821, "learning_rate": 1.2854746972336567e-05, "loss": 0.0219, "step": 31130 }, { "grad_norm": 0.04822108522057533, "learning_rate": 1.2827088984732132e-05, "loss": 0.0222, "step": 31140 }, { "grad_norm": 0.03068983554840088, "learning_rate": 1.279945640446184e-05, "loss": 0.0222, "step": 31150 }, { "grad_norm": 0.08023936301469803, "learning_rate": 1.2771849250412315e-05, "loss": 0.0206, "step": 31160 }, { "grad_norm": 0.032598044723272324, "learning_rate": 1.2744267541452792e-05, "loss": 0.0235, "step": 31170 }, { "grad_norm": 0.044549595564603806, "learning_rate": 1.2716711296435112e-05, "loss": 0.0244, "step": 31180 }, { "grad_norm": 0.06926105171442032, "learning_rate": 1.268918053419369e-05, "loss": 0.0227, "step": 31190 }, { "grad_norm": 0.029186313971877098, "learning_rate": 1.2661675273545564e-05, "loss": 0.0233, "step": 31200 }, { "grad_norm": 0.03395861014723778, "learning_rate": 1.26341955332903e-05, "loss": 0.0224, "step": 31210 }, { "grad_norm": 0.03704312816262245, "learning_rate": 1.2606741332210048e-05, "loss": 0.0236, "step": 31220 }, { "grad_norm": 0.03796081617474556, "learning_rate": 1.2579312689069495e-05, "loss": 0.024, "step": 31230 }, { "grad_norm": 0.047054775059223175, "learning_rate": 1.2551909622615881e-05, "loss": 0.0205, "step": 31240 }, { "grad_norm": 0.04290385544300079, "learning_rate": 1.252453215157891e-05, "loss": 0.0236, "step": 31250 }, { "grad_norm": 0.0346832200884819, "learning_rate": 1.2497180294670841e-05, "loss": 0.0216, "step": 31260 }, { "grad_norm": 0.03385329991579056, "learning_rate": 1.2469854070586423e-05, "loss": 0.0236, "step": 31270 }, { "grad_norm": 0.03243541717529297, "learning_rate": 1.2442553498002845e-05, "loss": 0.0231, "step": 31280 }, { "grad_norm": 0.04064340516924858, "learning_rate": 1.241527859557981e-05, "loss": 0.0222, "step": 31290 }, { "grad_norm": 0.027730390429496765, "learning_rate": 1.238802938195946e-05, "loss": 0.022, "step": 31300 }, { "grad_norm": 0.04568805918097496, "learning_rate": 1.2360805875766357e-05, "loss": 0.0223, "step": 31310 }, { "grad_norm": 0.031761348247528076, "learning_rate": 1.2333608095607524e-05, "loss": 0.0208, "step": 31320 }, { "grad_norm": 0.03928312659263611, "learning_rate": 1.230643606007238e-05, "loss": 0.0206, "step": 31330 }, { "grad_norm": 0.036089442670345306, "learning_rate": 1.2279289787732778e-05, "loss": 0.0203, "step": 31340 }, { "grad_norm": 0.036811042577028275, "learning_rate": 1.2252169297142906e-05, "loss": 0.0209, "step": 31350 }, { "grad_norm": 0.03288887441158295, "learning_rate": 1.2225074606839392e-05, "loss": 0.0234, "step": 31360 }, { "grad_norm": 0.02575700730085373, "learning_rate": 1.2198005735341173e-05, "loss": 0.0215, "step": 31370 }, { "grad_norm": 0.03309909254312515, "learning_rate": 1.2170962701149584e-05, "loss": 0.0218, "step": 31380 }, { "grad_norm": 0.03473755717277527, "learning_rate": 1.2143945522748285e-05, "loss": 0.0209, "step": 31390 }, { "grad_norm": 0.04524755850434303, "learning_rate": 1.2116954218603266e-05, "loss": 0.0219, "step": 31400 }, { "grad_norm": 0.050709839910268784, "learning_rate": 1.2089988807162817e-05, "loss": 0.0214, "step": 31410 }, { "grad_norm": 0.032424453645944595, "learning_rate": 1.2063049306857543e-05, "loss": 0.0212, "step": 31420 }, { "grad_norm": 0.061628084629774094, "learning_rate": 1.2036135736100346e-05, "loss": 0.0208, "step": 31430 }, { "grad_norm": 0.05560489743947983, "learning_rate": 1.200924811328641e-05, "loss": 0.0226, "step": 31440 }, { "grad_norm": 0.0435616597533226, "learning_rate": 1.1982386456793143e-05, "loss": 0.0233, "step": 31450 }, { "grad_norm": 0.037396933883428574, "learning_rate": 1.1955550784980257e-05, "loss": 0.0253, "step": 31460 }, { "grad_norm": 0.06159604713320732, "learning_rate": 1.1928741116189663e-05, "loss": 0.0226, "step": 31470 }, { "grad_norm": 0.03379058092832565, "learning_rate": 1.1901957468745529e-05, "loss": 0.0244, "step": 31480 }, { "grad_norm": 0.03276272863149643, "learning_rate": 1.1875199860954217e-05, "loss": 0.0234, "step": 31490 }, { "grad_norm": 0.04127391055226326, "learning_rate": 1.1848468311104316e-05, "loss": 0.024, "step": 31500 }, { "grad_norm": 0.05616943910717964, "learning_rate": 1.1821762837466566e-05, "loss": 0.0246, "step": 31510 }, { "grad_norm": 0.026899613440036774, "learning_rate": 1.179508345829392e-05, "loss": 0.0229, "step": 31520 }, { "grad_norm": 0.030971132218837738, "learning_rate": 1.1768430191821478e-05, "loss": 0.0216, "step": 31530 }, { "grad_norm": 0.04127822816371918, "learning_rate": 1.174180305626651e-05, "loss": 0.0225, "step": 31540 }, { "grad_norm": 0.04953974112868309, "learning_rate": 1.1715202069828401e-05, "loss": 0.0234, "step": 31550 }, { "grad_norm": 0.051243674010038376, "learning_rate": 1.1688627250688661e-05, "loss": 0.024, "step": 31560 }, { "grad_norm": 0.11939767748117447, "learning_rate": 1.1662078617010935e-05, "loss": 0.0247, "step": 31570 }, { "grad_norm": 0.05761849507689476, "learning_rate": 1.1635556186940972e-05, "loss": 0.025, "step": 31580 }, { "grad_norm": 0.03918596729636192, "learning_rate": 1.1609059978606602e-05, "loss": 0.0234, "step": 31590 }, { "grad_norm": 0.09747567772865295, "learning_rate": 1.1582590010117745e-05, "loss": 0.0225, "step": 31600 }, { "grad_norm": 0.05939994007349014, "learning_rate": 1.1556146299566345e-05, "loss": 0.0205, "step": 31610 }, { "grad_norm": 0.06805165857076645, "learning_rate": 1.1529728865026452e-05, "loss": 0.0222, "step": 31620 }, { "grad_norm": 0.04267434775829315, "learning_rate": 1.150333772455412e-05, "loss": 0.0207, "step": 31630 }, { "grad_norm": 0.02675367146730423, "learning_rate": 1.1476972896187466e-05, "loss": 0.0202, "step": 31640 }, { "grad_norm": 0.03695881739258766, "learning_rate": 1.1450634397946592e-05, "loss": 0.0209, "step": 31650 }, { "grad_norm": 0.047648508101701736, "learning_rate": 1.142432224783359e-05, "loss": 0.0217, "step": 31660 }, { "grad_norm": 0.11493810266256332, "learning_rate": 1.1398036463832584e-05, "loss": 0.0193, "step": 31670 }, { "grad_norm": 0.034341808408498764, "learning_rate": 1.1371777063909661e-05, "loss": 0.024, "step": 31680 }, { "grad_norm": 0.05002209171652794, "learning_rate": 1.1345544066012874e-05, "loss": 0.0254, "step": 31690 }, { "grad_norm": 0.02866225875914097, "learning_rate": 1.131933748807224e-05, "loss": 0.026, "step": 31700 }, { "grad_norm": 0.03365984186530113, "learning_rate": 1.1293157347999683e-05, "loss": 0.0247, "step": 31710 }, { "grad_norm": 0.0496782623231411, "learning_rate": 1.1267003663689097e-05, "loss": 0.0244, "step": 31720 }, { "grad_norm": 0.04268568754196167, "learning_rate": 1.1240876453016275e-05, "loss": 0.025, "step": 31730 }, { "grad_norm": 0.1125185415148735, "learning_rate": 1.1214775733838933e-05, "loss": 0.0243, "step": 31740 }, { "grad_norm": 0.17431725561618805, "learning_rate": 1.118870152399666e-05, "loss": 0.0227, "step": 31750 }, { "grad_norm": 0.048765629529953, "learning_rate": 1.1162653841310905e-05, "loss": 0.0204, "step": 31760 }, { "grad_norm": 0.03384198248386383, "learning_rate": 1.1136632703585043e-05, "loss": 0.0205, "step": 31770 }, { "grad_norm": 0.04206444323062897, "learning_rate": 1.111063812860426e-05, "loss": 0.0211, "step": 31780 }, { "grad_norm": 0.031392112374305725, "learning_rate": 1.1084670134135616e-05, "loss": 0.0213, "step": 31790 }, { "grad_norm": 0.04869675263762474, "learning_rate": 1.1058728737927998e-05, "loss": 0.0222, "step": 31800 }, { "grad_norm": 0.03337191417813301, "learning_rate": 1.1032813957712086e-05, "loss": 0.0215, "step": 31810 }, { "grad_norm": 0.04047969728708267, "learning_rate": 1.1006925811200385e-05, "loss": 0.0223, "step": 31820 }, { "grad_norm": 0.0388643741607666, "learning_rate": 1.0981064316087231e-05, "loss": 0.0232, "step": 31830 }, { "grad_norm": 0.025297196581959724, "learning_rate": 1.0955229490048675e-05, "loss": 0.022, "step": 31840 }, { "grad_norm": 0.05466605722904205, "learning_rate": 1.0929421350742603e-05, "loss": 0.0225, "step": 31850 }, { "grad_norm": 0.0403306782245636, "learning_rate": 1.090363991580861e-05, "loss": 0.0212, "step": 31860 }, { "grad_norm": 0.03888816386461258, "learning_rate": 1.0877885202868083e-05, "loss": 0.0227, "step": 31870 }, { "grad_norm": 0.04405313357710838, "learning_rate": 1.0852157229524118e-05, "loss": 0.0231, "step": 31880 }, { "grad_norm": 0.04881252348423004, "learning_rate": 1.0826456013361542e-05, "loss": 0.0208, "step": 31890 }, { "grad_norm": 0.05914316326379776, "learning_rate": 1.080078157194691e-05, "loss": 0.021, "step": 31900 }, { "grad_norm": 0.03672154247760773, "learning_rate": 1.077513392282844e-05, "loss": 0.0225, "step": 31910 }, { "grad_norm": 0.04704362526535988, "learning_rate": 1.0749513083536073e-05, "loss": 0.0218, "step": 31920 }, { "grad_norm": 0.055848512798547745, "learning_rate": 1.072391907158139e-05, "loss": 0.0219, "step": 31930 }, { "grad_norm": 0.05291841924190521, "learning_rate": 1.0698351904457676e-05, "loss": 0.0211, "step": 31940 }, { "grad_norm": 0.03971601277589798, "learning_rate": 1.0672811599639849e-05, "loss": 0.0223, "step": 31950 }, { "grad_norm": 0.027719207108020782, "learning_rate": 1.0647298174584453e-05, "loss": 0.0228, "step": 31960 }, { "grad_norm": 0.04068226367235184, "learning_rate": 1.0621811646729673e-05, "loss": 0.0207, "step": 31970 }, { "grad_norm": 0.06288432329893112, "learning_rate": 1.0596352033495328e-05, "loss": 0.0203, "step": 31980 }, { "grad_norm": 0.05197663977742195, "learning_rate": 1.057091935228281e-05, "loss": 0.0218, "step": 31990 }, { "grad_norm": 0.0409017875790596, "learning_rate": 1.0545513620475134e-05, "loss": 0.0223, "step": 32000 }, { "grad_norm": 0.03586411103606224, "learning_rate": 1.0520134855436863e-05, "loss": 0.0215, "step": 32010 }, { "grad_norm": 0.052838537842035294, "learning_rate": 1.049478307451413e-05, "loss": 0.0216, "step": 32020 }, { "grad_norm": 0.026109538972377777, "learning_rate": 1.0469458295034656e-05, "loss": 0.0224, "step": 32030 }, { "grad_norm": 0.04347730427980423, "learning_rate": 1.044416053430769e-05, "loss": 0.0222, "step": 32040 }, { "grad_norm": 0.03506696969270706, "learning_rate": 1.0418889809624021e-05, "loss": 0.0217, "step": 32050 }, { "grad_norm": 0.060303132981061935, "learning_rate": 1.0393646138255925e-05, "loss": 0.0222, "step": 32060 }, { "grad_norm": 0.04016118124127388, "learning_rate": 1.036842953745723e-05, "loss": 0.0198, "step": 32070 }, { "grad_norm": 0.04157622531056404, "learning_rate": 1.034324002446324e-05, "loss": 0.0225, "step": 32080 }, { "grad_norm": 0.03619833290576935, "learning_rate": 1.0318077616490751e-05, "loss": 0.0222, "step": 32090 }, { "grad_norm": 0.06038830429315567, "learning_rate": 1.0292942330738054e-05, "loss": 0.0205, "step": 32100 }, { "grad_norm": 0.0387483686208725, "learning_rate": 1.026783418438485e-05, "loss": 0.0198, "step": 32110 }, { "grad_norm": 0.044958822429180145, "learning_rate": 1.0242753194592319e-05, "loss": 0.0228, "step": 32120 }, { "grad_norm": 0.07012530416250229, "learning_rate": 1.0217699378503092e-05, "loss": 0.0214, "step": 32130 }, { "grad_norm": 0.05052102357149124, "learning_rate": 1.0192672753241218e-05, "loss": 0.022, "step": 32140 }, { "grad_norm": 0.039519499987363815, "learning_rate": 1.0167673335912164e-05, "loss": 0.0224, "step": 32150 }, { "grad_norm": 0.039319202303886414, "learning_rate": 1.0142701143602779e-05, "loss": 0.0227, "step": 32160 }, { "grad_norm": 0.07576177269220352, "learning_rate": 1.0117756193381328e-05, "loss": 0.0228, "step": 32170 }, { "grad_norm": 0.03859199583530426, "learning_rate": 1.0092838502297453e-05, "loss": 0.0229, "step": 32180 }, { "grad_norm": 0.06368175148963928, "learning_rate": 1.0067948087382162e-05, "loss": 0.0235, "step": 32190 }, { "grad_norm": 0.035080958157777786, "learning_rate": 1.0043084965647826e-05, "loss": 0.0237, "step": 32200 }, { "grad_norm": 0.044293101876974106, "learning_rate": 1.001824915408815e-05, "loss": 0.0252, "step": 32210 }, { "grad_norm": 0.09508024901151657, "learning_rate": 9.99344066967815e-06, "loss": 0.022, "step": 32220 }, { "grad_norm": 0.04608726128935814, "learning_rate": 9.968659529374219e-06, "loss": 0.0233, "step": 32230 }, { "grad_norm": 0.04010193794965744, "learning_rate": 9.943905750114025e-06, "loss": 0.0234, "step": 32240 }, { "grad_norm": 0.04444396123290062, "learning_rate": 9.919179348816559e-06, "loss": 0.0243, "step": 32250 }, { "grad_norm": 0.04228738322854042, "learning_rate": 9.894480342382056e-06, "loss": 0.0228, "step": 32260 }, { "grad_norm": 0.08438821136951447, "learning_rate": 9.869808747692061e-06, "loss": 0.0224, "step": 32270 }, { "grad_norm": 0.0791582241654396, "learning_rate": 9.84516458160939e-06, "loss": 0.0226, "step": 32280 }, { "grad_norm": 0.03237992152571678, "learning_rate": 9.820547860978097e-06, "loss": 0.0214, "step": 32290 }, { "grad_norm": 0.057591453194618225, "learning_rate": 9.795958602623461e-06, "loss": 0.0222, "step": 32300 }, { "grad_norm": 0.037733253091573715, "learning_rate": 9.77139682335203e-06, "loss": 0.0201, "step": 32310 }, { "grad_norm": 0.029751762747764587, "learning_rate": 9.74686253995153e-06, "loss": 0.0207, "step": 32320 }, { "grad_norm": 0.048600081354379654, "learning_rate": 9.722355769190911e-06, "loss": 0.0223, "step": 32330 }, { "grad_norm": 0.04840546473860741, "learning_rate": 9.697876527820333e-06, "loss": 0.0237, "step": 32340 }, { "grad_norm": 0.056910160928964615, "learning_rate": 9.673424832571132e-06, "loss": 0.0221, "step": 32350 }, { "grad_norm": 0.03471437841653824, "learning_rate": 9.649000700155791e-06, "loss": 0.0217, "step": 32360 }, { "grad_norm": 0.09012334048748016, "learning_rate": 9.624604147267984e-06, "loss": 0.022, "step": 32370 }, { "grad_norm": 0.03678888827562332, "learning_rate": 9.600235190582524e-06, "loss": 0.024, "step": 32380 }, { "grad_norm": 0.04358452931046486, "learning_rate": 9.57589384675538e-06, "loss": 0.0254, "step": 32390 }, { "grad_norm": 0.04202694818377495, "learning_rate": 9.551580132423598e-06, "loss": 0.0256, "step": 32400 }, { "grad_norm": 0.035179633647203445, "learning_rate": 9.527294064205405e-06, "loss": 0.0263, "step": 32410 }, { "grad_norm": 0.03521718084812164, "learning_rate": 9.503035658700072e-06, "loss": 0.0259, "step": 32420 }, { "grad_norm": 0.05389253795146942, "learning_rate": 9.478804932488e-06, "loss": 0.0265, "step": 32430 }, { "grad_norm": 0.04006361588835716, "learning_rate": 9.454601902130667e-06, "loss": 0.0226, "step": 32440 }, { "grad_norm": 0.03682072088122368, "learning_rate": 9.430426584170626e-06, "loss": 0.0217, "step": 32450 }, { "grad_norm": 0.04399817809462547, "learning_rate": 9.40627899513145e-06, "loss": 0.0201, "step": 32460 }, { "grad_norm": 0.03404362499713898, "learning_rate": 9.38215915151781e-06, "loss": 0.0208, "step": 32470 }, { "grad_norm": 0.028883930295705795, "learning_rate": 9.358067069815396e-06, "loss": 0.0208, "step": 32480 }, { "grad_norm": 0.04010457918047905, "learning_rate": 9.334002766490901e-06, "loss": 0.0229, "step": 32490 }, { "grad_norm": 0.04488876834511757, "learning_rate": 9.309966257992054e-06, "loss": 0.0227, "step": 32500 }, { "grad_norm": 0.03203745186328888, "learning_rate": 9.2859575607476e-06, "loss": 0.0221, "step": 32510 }, { "grad_norm": 0.04209626466035843, "learning_rate": 9.26197669116723e-06, "loss": 0.0226, "step": 32520 }, { "grad_norm": 0.039200786501169205, "learning_rate": 9.238023665641655e-06, "loss": 0.0215, "step": 32530 }, { "grad_norm": 0.03960935026407242, "learning_rate": 9.21409850054254e-06, "loss": 0.0235, "step": 32540 }, { "grad_norm": 0.03587385267019272, "learning_rate": 9.190201212222526e-06, "loss": 0.0218, "step": 32550 }, { "grad_norm": 0.04269804432988167, "learning_rate": 9.166331817015155e-06, "loss": 0.0212, "step": 32560 }, { "grad_norm": 0.04672133922576904, "learning_rate": 9.142490331234965e-06, "loss": 0.0227, "step": 32570 }, { "grad_norm": 0.03473415598273277, "learning_rate": 9.11867677117736e-06, "loss": 0.0207, "step": 32580 }, { "grad_norm": 0.09414838254451752, "learning_rate": 9.09489115311869e-06, "loss": 0.0221, "step": 32590 }, { "grad_norm": 0.04502332583069801, "learning_rate": 9.071133493316202e-06, "loss": 0.024, "step": 32600 }, { "grad_norm": 0.037127599120140076, "learning_rate": 9.047403808008053e-06, "loss": 0.0229, "step": 32610 }, { "grad_norm": 0.03987783566117287, "learning_rate": 9.023702113413224e-06, "loss": 0.0243, "step": 32620 }, { "grad_norm": 0.028606602922081947, "learning_rate": 9.00002842573162e-06, "loss": 0.0231, "step": 32630 }, { "grad_norm": 0.03944665938615799, "learning_rate": 8.976382761143981e-06, "loss": 0.0246, "step": 32640 }, { "grad_norm": 0.047829557210206985, "learning_rate": 8.952765135811909e-06, "loss": 0.0241, "step": 32650 }, { "grad_norm": 0.028183622285723686, "learning_rate": 8.929175565877795e-06, "loss": 0.0226, "step": 32660 }, { "grad_norm": 0.05381942540407181, "learning_rate": 8.905614067464924e-06, "loss": 0.0221, "step": 32670 }, { "grad_norm": 0.030367638915777206, "learning_rate": 8.882080656677333e-06, "loss": 0.0215, "step": 32680 }, { "grad_norm": 0.047099411487579346, "learning_rate": 8.85857534959989e-06, "loss": 0.0221, "step": 32690 }, { "grad_norm": 0.03239577263593674, "learning_rate": 8.835098162298256e-06, "loss": 0.0222, "step": 32700 }, { "grad_norm": 0.051170941442251205, "learning_rate": 8.811649110818875e-06, "loss": 0.0232, "step": 32710 }, { "grad_norm": 0.04657788202166557, "learning_rate": 8.78822821118893e-06, "loss": 0.0229, "step": 32720 }, { "grad_norm": 0.046669892966747284, "learning_rate": 8.764835479416405e-06, "loss": 0.025, "step": 32730 }, { "grad_norm": 0.039647676050662994, "learning_rate": 8.741470931490003e-06, "loss": 0.0236, "step": 32740 }, { "grad_norm": 0.06874950975179672, "learning_rate": 8.718134583379189e-06, "loss": 0.0237, "step": 32750 }, { "grad_norm": 0.035700295120477676, "learning_rate": 8.69482645103412e-06, "loss": 0.0226, "step": 32760 }, { "grad_norm": 0.035969171673059464, "learning_rate": 8.671546550385679e-06, "loss": 0.0232, "step": 32770 }, { "grad_norm": 0.03885240480303764, "learning_rate": 8.648294897345466e-06, "loss": 0.0205, "step": 32780 }, { "grad_norm": 0.039105284959077835, "learning_rate": 8.62507150780576e-06, "loss": 0.0214, "step": 32790 }, { "grad_norm": 0.043878182768821716, "learning_rate": 8.601876397639542e-06, "loss": 0.0223, "step": 32800 }, { "grad_norm": 0.05127972736954689, "learning_rate": 8.578709582700456e-06, "loss": 0.0216, "step": 32810 }, { "grad_norm": 0.028421318158507347, "learning_rate": 8.555571078822783e-06, "loss": 0.0218, "step": 32820 }, { "grad_norm": 0.03519587591290474, "learning_rate": 8.532460901821477e-06, "loss": 0.0221, "step": 32830 }, { "grad_norm": 0.033751022070646286, "learning_rate": 8.509379067492128e-06, "loss": 0.0205, "step": 32840 }, { "grad_norm": 0.05677512288093567, "learning_rate": 8.486325591610967e-06, "loss": 0.0218, "step": 32850 }, { "grad_norm": 0.031102919951081276, "learning_rate": 8.463300489934811e-06, "loss": 0.0218, "step": 32860 }, { "grad_norm": 0.03432675451040268, "learning_rate": 8.440303778201093e-06, "loss": 0.0214, "step": 32870 }, { "grad_norm": 0.03334285318851471, "learning_rate": 8.417335472127858e-06, "loss": 0.0222, "step": 32880 }, { "grad_norm": 0.05269911140203476, "learning_rate": 8.394395587413722e-06, "loss": 0.025, "step": 32890 }, { "grad_norm": 0.04230576008558273, "learning_rate": 8.371484139737889e-06, "loss": 0.0229, "step": 32900 }, { "grad_norm": 0.034006040543317795, "learning_rate": 8.348601144760127e-06, "loss": 0.023, "step": 32910 }, { "grad_norm": 0.0428704097867012, "learning_rate": 8.325746618120716e-06, "loss": 0.022, "step": 32920 }, { "grad_norm": 0.05151943117380142, "learning_rate": 8.302920575440531e-06, "loss": 0.0217, "step": 32930 }, { "grad_norm": 0.03232322633266449, "learning_rate": 8.280123032320953e-06, "loss": 0.0214, "step": 32940 }, { "grad_norm": 0.040603313595056534, "learning_rate": 8.257354004343892e-06, "loss": 0.0215, "step": 32950 }, { "grad_norm": 0.0484699048101902, "learning_rate": 8.234613507071754e-06, "loss": 0.0222, "step": 32960 }, { "grad_norm": 0.037082113325595856, "learning_rate": 8.211901556047475e-06, "loss": 0.0223, "step": 32970 }, { "grad_norm": 0.04435017704963684, "learning_rate": 8.189218166794427e-06, "loss": 0.0208, "step": 32980 }, { "grad_norm": 0.029081938788294792, "learning_rate": 8.166563354816509e-06, "loss": 0.0233, "step": 32990 }, { "grad_norm": 0.06527487933635712, "learning_rate": 8.143937135598074e-06, "loss": 0.0232, "step": 33000 }, { "grad_norm": 0.07351461797952652, "learning_rate": 8.121339524603943e-06, "loss": 0.0227, "step": 33010 }, { "grad_norm": 0.03363712131977081, "learning_rate": 8.098770537279338e-06, "loss": 0.0236, "step": 33020 }, { "grad_norm": 0.03637590631842613, "learning_rate": 8.076230189049966e-06, "loss": 0.0225, "step": 33030 }, { "grad_norm": 0.1483805924654007, "learning_rate": 8.05371849532196e-06, "loss": 0.0213, "step": 33040 }, { "grad_norm": 0.0353095680475235, "learning_rate": 8.031235471481813e-06, "loss": 0.0216, "step": 33050 }, { "grad_norm": 0.08698521554470062, "learning_rate": 8.008781132896486e-06, "loss": 0.022, "step": 33060 }, { "grad_norm": 0.04622401297092438, "learning_rate": 7.986355494913312e-06, "loss": 0.0216, "step": 33070 }, { "grad_norm": 0.04043344780802727, "learning_rate": 7.963958572859975e-06, "loss": 0.0213, "step": 33080 }, { "grad_norm": 0.038275234401226044, "learning_rate": 7.941590382044583e-06, "loss": 0.0225, "step": 33090 }, { "grad_norm": 0.035281162708997726, "learning_rate": 7.919250937755574e-06, "loss": 0.0227, "step": 33100 }, { "grad_norm": 0.13467547297477722, "learning_rate": 7.896940255261764e-06, "loss": 0.0239, "step": 33110 }, { "grad_norm": 0.03864181414246559, "learning_rate": 7.874658349812269e-06, "loss": 0.0237, "step": 33120 }, { "grad_norm": 0.03433383256196976, "learning_rate": 7.85240523663659e-06, "loss": 0.0236, "step": 33130 }, { "grad_norm": 0.03728759288787842, "learning_rate": 7.830180930944487e-06, "loss": 0.0239, "step": 33140 }, { "grad_norm": 0.03226301446557045, "learning_rate": 7.80798544792608e-06, "loss": 0.0222, "step": 33150 }, { "grad_norm": 0.034762296825647354, "learning_rate": 7.785818802751776e-06, "loss": 0.0219, "step": 33160 }, { "grad_norm": 0.05363379418849945, "learning_rate": 7.763681010572277e-06, "loss": 0.023, "step": 33170 }, { "grad_norm": 0.03978962078690529, "learning_rate": 7.741572086518528e-06, "loss": 0.0223, "step": 33180 }, { "grad_norm": 0.04233279824256897, "learning_rate": 7.719492045701782e-06, "loss": 0.0199, "step": 33190 }, { "grad_norm": 0.058275043964385986, "learning_rate": 7.697440903213549e-06, "loss": 0.0239, "step": 33200 }, { "grad_norm": 0.03549265116453171, "learning_rate": 7.675418674125573e-06, "loss": 0.0225, "step": 33210 }, { "grad_norm": 0.037979040294885635, "learning_rate": 7.65342537348983e-06, "loss": 0.0227, "step": 33220 }, { "grad_norm": 0.03983355686068535, "learning_rate": 7.631461016338549e-06, "loss": 0.0218, "step": 33230 }, { "grad_norm": 0.054613009095191956, "learning_rate": 7.609525617684138e-06, "loss": 0.0231, "step": 33240 }, { "grad_norm": 0.02995535172522068, "learning_rate": 7.587619192519252e-06, "loss": 0.0218, "step": 33250 }, { "grad_norm": 0.05474540591239929, "learning_rate": 7.565741755816725e-06, "loss": 0.023, "step": 33260 }, { "grad_norm": 0.04351922497153282, "learning_rate": 7.543893322529588e-06, "loss": 0.0222, "step": 33270 }, { "grad_norm": 0.0411502979695797, "learning_rate": 7.522073907591015e-06, "loss": 0.0233, "step": 33280 }, { "grad_norm": 0.033715929836034775, "learning_rate": 7.500283525914381e-06, "loss": 0.0212, "step": 33290 }, { "grad_norm": 0.042218778282403946, "learning_rate": 7.478522192393211e-06, "loss": 0.0208, "step": 33300 }, { "grad_norm": 0.03910578414797783, "learning_rate": 7.456789921901181e-06, "loss": 0.0212, "step": 33310 }, { "grad_norm": 0.041201118379831314, "learning_rate": 7.435086729292079e-06, "loss": 0.0224, "step": 33320 }, { "grad_norm": 0.0491308756172657, "learning_rate": 7.413412629399824e-06, "loss": 0.0233, "step": 33330 }, { "grad_norm": 0.021599754691123962, "learning_rate": 7.391767637038466e-06, "loss": 0.0187, "step": 33340 }, { "grad_norm": 0.07400321960449219, "learning_rate": 7.370151767002159e-06, "loss": 0.0229, "step": 33350 }, { "grad_norm": 0.028744371607899666, "learning_rate": 7.348565034065147e-06, "loss": 0.0225, "step": 33360 }, { "grad_norm": 0.04068112373352051, "learning_rate": 7.327007452981765e-06, "loss": 0.0234, "step": 33370 }, { "grad_norm": 0.048889610916376114, "learning_rate": 7.305479038486396e-06, "loss": 0.0229, "step": 33380 }, { "grad_norm": 0.05255451425909996, "learning_rate": 7.283979805293517e-06, "loss": 0.0232, "step": 33390 }, { "grad_norm": 0.02753354236483574, "learning_rate": 7.262509768097653e-06, "loss": 0.0229, "step": 33400 }, { "grad_norm": 0.04876343905925751, "learning_rate": 7.241068941573381e-06, "loss": 0.0231, "step": 33410 }, { "grad_norm": 0.04240617901086807, "learning_rate": 7.219657340375296e-06, "loss": 0.0241, "step": 33420 }, { "grad_norm": 0.049132831394672394, "learning_rate": 7.198274979138014e-06, "loss": 0.024, "step": 33430 }, { "grad_norm": 0.039917320013046265, "learning_rate": 7.176921872476189e-06, "loss": 0.0225, "step": 33440 }, { "grad_norm": 0.03917012736201286, "learning_rate": 7.1555980349844674e-06, "loss": 0.0227, "step": 33450 }, { "grad_norm": 0.0378575474023819, "learning_rate": 7.134303481237492e-06, "loss": 0.023, "step": 33460 }, { "grad_norm": 0.046207573264837265, "learning_rate": 7.113038225789892e-06, "loss": 0.0238, "step": 33470 }, { "grad_norm": 0.03336196020245552, "learning_rate": 7.091802283176258e-06, "loss": 0.0209, "step": 33480 }, { "grad_norm": 0.06210804358124733, "learning_rate": 7.070595667911162e-06, "loss": 0.0216, "step": 33490 }, { "grad_norm": 0.13015955686569214, "learning_rate": 7.049418394489127e-06, "loss": 0.0211, "step": 33500 }, { "grad_norm": 0.042652685195207596, "learning_rate": 7.028270477384636e-06, "loss": 0.0206, "step": 33510 }, { "grad_norm": 0.03383871540427208, "learning_rate": 7.0071519310520695e-06, "loss": 0.0207, "step": 33520 }, { "grad_norm": 0.06154133006930351, "learning_rate": 6.98606276992575e-06, "loss": 0.0222, "step": 33530 }, { "grad_norm": 0.03877688944339752, "learning_rate": 6.965003008419929e-06, "loss": 0.0224, "step": 33540 }, { "grad_norm": 0.09899210929870605, "learning_rate": 6.943972660928755e-06, "loss": 0.0227, "step": 33550 }, { "grad_norm": 0.0479511022567749, "learning_rate": 6.922971741826273e-06, "loss": 0.0227, "step": 33560 }, { "grad_norm": 0.05907992273569107, "learning_rate": 6.902000265466424e-06, "loss": 0.0233, "step": 33570 }, { "grad_norm": 0.03820709511637688, "learning_rate": 6.881058246182998e-06, "loss": 0.0233, "step": 33580 }, { "grad_norm": 0.05108008533716202, "learning_rate": 6.860145698289666e-06, "loss": 0.0233, "step": 33590 }, { "grad_norm": 0.03689424321055412, "learning_rate": 6.839262636079985e-06, "loss": 0.0233, "step": 33600 }, { "grad_norm": 0.03970819339156151, "learning_rate": 6.818409073827292e-06, "loss": 0.0234, "step": 33610 }, { "grad_norm": 0.032696232199668884, "learning_rate": 6.79758502578483e-06, "loss": 0.0213, "step": 33620 }, { "grad_norm": 0.054018691182136536, "learning_rate": 6.776790506185621e-06, "loss": 0.022, "step": 33630 }, { "grad_norm": 0.040597401559352875, "learning_rate": 6.756025529242533e-06, "loss": 0.0217, "step": 33640 }, { "grad_norm": 0.0794173926115036, "learning_rate": 6.735290109148223e-06, "loss": 0.0231, "step": 33650 }, { "grad_norm": 0.03967982158064842, "learning_rate": 6.714584260075163e-06, "loss": 0.0247, "step": 33660 }, { "grad_norm": 0.048810869455337524, "learning_rate": 6.693907996175619e-06, "loss": 0.0224, "step": 33670 }, { "grad_norm": 0.03817664086818695, "learning_rate": 6.673261331581593e-06, "loss": 0.0246, "step": 33680 }, { "grad_norm": 0.03469765931367874, "learning_rate": 6.652644280404912e-06, "loss": 0.0228, "step": 33690 }, { "grad_norm": 0.03086056001484394, "learning_rate": 6.6320568567371175e-06, "loss": 0.0221, "step": 33700 }, { "grad_norm": 0.05507516488432884, "learning_rate": 6.611499074649518e-06, "loss": 0.0226, "step": 33710 }, { "grad_norm": 0.04548953101038933, "learning_rate": 6.590970948193187e-06, "loss": 0.0236, "step": 33720 }, { "grad_norm": 0.04417731612920761, "learning_rate": 6.5704724913988825e-06, "loss": 0.0243, "step": 33730 }, { "grad_norm": 0.03617874160408974, "learning_rate": 6.550003718277109e-06, "loss": 0.0239, "step": 33740 }, { "grad_norm": 0.041469741612672806, "learning_rate": 6.529564642818087e-06, "loss": 0.0238, "step": 33750 }, { "grad_norm": 0.03030477836728096, "learning_rate": 6.50915527899173e-06, "loss": 0.0245, "step": 33760 }, { "grad_norm": 0.05469844117760658, "learning_rate": 6.4887756407476575e-06, "loss": 0.0252, "step": 33770 }, { "grad_norm": 0.06373290717601776, "learning_rate": 6.468425742015149e-06, "loss": 0.0219, "step": 33780 }, { "grad_norm": 0.03859850391745567, "learning_rate": 6.448105596703152e-06, "loss": 0.0211, "step": 33790 }, { "grad_norm": 0.08265098929405212, "learning_rate": 6.427815218700317e-06, "loss": 0.0199, "step": 33800 }, { "grad_norm": 0.04254092648625374, "learning_rate": 6.407554621874911e-06, "loss": 0.0202, "step": 33810 }, { "grad_norm": 0.10619206726551056, "learning_rate": 6.3873238200748864e-06, "loss": 0.019, "step": 33820 }, { "grad_norm": 0.03232989087700844, "learning_rate": 6.367122827127775e-06, "loss": 0.0194, "step": 33830 }, { "grad_norm": 0.031040210276842117, "learning_rate": 6.346951656840777e-06, "loss": 0.0224, "step": 33840 }, { "grad_norm": 0.03306521102786064, "learning_rate": 6.326810323000698e-06, "loss": 0.0227, "step": 33850 }, { "grad_norm": 0.03577335178852081, "learning_rate": 6.306698839373948e-06, "loss": 0.0238, "step": 33860 }, { "grad_norm": 0.03706294298171997, "learning_rate": 6.286617219706547e-06, "loss": 0.0252, "step": 33870 }, { "grad_norm": 0.036167774349451065, "learning_rate": 6.26656547772409e-06, "loss": 0.0235, "step": 33880 }, { "grad_norm": 0.057401061058044434, "learning_rate": 6.246543627131734e-06, "loss": 0.0234, "step": 33890 }, { "grad_norm": 0.05533439666032791, "learning_rate": 6.22655168161424e-06, "loss": 0.0212, "step": 33900 }, { "grad_norm": 0.034643109887838364, "learning_rate": 6.20658965483591e-06, "loss": 0.0216, "step": 33910 }, { "grad_norm": 0.05223296210169792, "learning_rate": 6.186657560440612e-06, "loss": 0.0206, "step": 33920 }, { "grad_norm": 0.04602174088358879, "learning_rate": 6.166755412051734e-06, "loss": 0.0208, "step": 33930 }, { "grad_norm": 0.04323755204677582, "learning_rate": 6.1468832232721985e-06, "loss": 0.0207, "step": 33940 }, { "grad_norm": 0.03570808842778206, "learning_rate": 6.127041007684476e-06, "loss": 0.0221, "step": 33950 }, { "grad_norm": 0.04287606477737427, "learning_rate": 6.107228778850527e-06, "loss": 0.0218, "step": 33960 }, { "grad_norm": 0.041730158030986786, "learning_rate": 6.087446550311837e-06, "loss": 0.0256, "step": 33970 }, { "grad_norm": 0.047418586909770966, "learning_rate": 6.067694335589353e-06, "loss": 0.0227, "step": 33980 }, { "grad_norm": 0.029164893552660942, "learning_rate": 6.047972148183528e-06, "loss": 0.0208, "step": 33990 }, { "grad_norm": 0.03632812574505806, "learning_rate": 6.0282800015742936e-06, "loss": 0.0222, "step": 34000 }, { "grad_norm": 0.03031337261199951, "learning_rate": 6.008617909221054e-06, "loss": 0.0216, "step": 34010 }, { "grad_norm": 0.030121993273496628, "learning_rate": 5.988985884562664e-06, "loss": 0.0212, "step": 34020 }, { "grad_norm": 0.05334597826004028, "learning_rate": 5.969383941017414e-06, "loss": 0.0227, "step": 34030 }, { "grad_norm": 0.0397113598883152, "learning_rate": 5.949812091983059e-06, "loss": 0.0228, "step": 34040 }, { "grad_norm": 0.0407915897667408, "learning_rate": 5.9302703508367695e-06, "loss": 0.0242, "step": 34050 }, { "grad_norm": 0.06231004744768143, "learning_rate": 5.910758730935157e-06, "loss": 0.0233, "step": 34060 }, { "grad_norm": 0.03139529377222061, "learning_rate": 5.8912772456142016e-06, "loss": 0.0223, "step": 34070 }, { "grad_norm": 0.04600043222308159, "learning_rate": 5.871825908189338e-06, "loss": 0.0241, "step": 34080 }, { "grad_norm": 0.03414798900485039, "learning_rate": 5.852404731955352e-06, "loss": 0.0235, "step": 34090 }, { "grad_norm": 0.04231952503323555, "learning_rate": 5.833013730186443e-06, "loss": 0.0223, "step": 34100 }, { "grad_norm": 0.042365144938230515, "learning_rate": 5.813652916136181e-06, "loss": 0.0189, "step": 34110 }, { "grad_norm": 0.032940447330474854, "learning_rate": 5.794322303037508e-06, "loss": 0.0195, "step": 34120 }, { "grad_norm": 0.0409935787320137, "learning_rate": 5.775021904102701e-06, "loss": 0.0182, "step": 34130 }, { "grad_norm": 0.05133889988064766, "learning_rate": 5.755751732523401e-06, "loss": 0.0199, "step": 34140 }, { "grad_norm": 0.03567693009972572, "learning_rate": 5.736511801470595e-06, "loss": 0.0187, "step": 34150 }, { "grad_norm": 0.02623561955988407, "learning_rate": 5.717302124094603e-06, "loss": 0.0198, "step": 34160 }, { "grad_norm": 0.04213135689496994, "learning_rate": 5.6981227135250396e-06, "loss": 0.0203, "step": 34170 }, { "grad_norm": 0.043247658759355545, "learning_rate": 5.678973582870867e-06, "loss": 0.0201, "step": 34180 }, { "grad_norm": 0.045056141912937164, "learning_rate": 5.659854745220317e-06, "loss": 0.0212, "step": 34190 }, { "grad_norm": 0.0249613169580698, "learning_rate": 5.640766213640947e-06, "loss": 0.02, "step": 34200 }, { "grad_norm": 0.05051187798380852, "learning_rate": 5.6217080011795934e-06, "loss": 0.0228, "step": 34210 }, { "grad_norm": 0.05761801823973656, "learning_rate": 5.602680120862364e-06, "loss": 0.022, "step": 34220 }, { "grad_norm": 0.03257124871015549, "learning_rate": 5.583682585694616e-06, "loss": 0.0241, "step": 34230 }, { "grad_norm": 0.03449214622378349, "learning_rate": 5.564715408661009e-06, "loss": 0.0231, "step": 34240 }, { "grad_norm": 0.035067569464445114, "learning_rate": 5.5457786027254265e-06, "loss": 0.0237, "step": 34250 }, { "grad_norm": 0.03350548818707466, "learning_rate": 5.526872180830977e-06, "loss": 0.0254, "step": 34260 }, { "grad_norm": 0.04277621954679489, "learning_rate": 5.507996155900041e-06, "loss": 0.0238, "step": 34270 }, { "grad_norm": 0.040710147470235825, "learning_rate": 5.489150540834203e-06, "loss": 0.0238, "step": 34280 }, { "grad_norm": 0.04774301499128342, "learning_rate": 5.47033534851425e-06, "loss": 0.0227, "step": 34290 }, { "grad_norm": 0.05457582324743271, "learning_rate": 5.451550591800198e-06, "loss": 0.0255, "step": 34300 }, { "grad_norm": 0.049560435116291046, "learning_rate": 5.43279628353125e-06, "loss": 0.025, "step": 34310 }, { "grad_norm": 0.05376192554831505, "learning_rate": 5.4140724365258075e-06, "loss": 0.0235, "step": 34320 }, { "grad_norm": 0.037002936005592346, "learning_rate": 5.395379063581424e-06, "loss": 0.0242, "step": 34330 }, { "grad_norm": 0.06648995727300644, "learning_rate": 5.376716177474866e-06, "loss": 0.0238, "step": 34340 }, { "grad_norm": 0.07566165924072266, "learning_rate": 5.358083790962015e-06, "loss": 0.0247, "step": 34350 }, { "grad_norm": 0.03809267655014992, "learning_rate": 5.33948191677795e-06, "loss": 0.0239, "step": 34360 }, { "grad_norm": 0.04277642443776131, "learning_rate": 5.320910567636866e-06, "loss": 0.0239, "step": 34370 }, { "grad_norm": 0.04159902408719063, "learning_rate": 5.302369756232123e-06, "loss": 0.0245, "step": 34380 }, { "grad_norm": 0.0488785095512867, "learning_rate": 5.283859495236165e-06, "loss": 0.0235, "step": 34390 }, { "grad_norm": 0.0476045124232769, "learning_rate": 5.265379797300602e-06, "loss": 0.0243, "step": 34400 }, { "grad_norm": 0.0376841276884079, "learning_rate": 5.246930675056117e-06, "loss": 0.024, "step": 34410 }, { "grad_norm": 0.03478951379656792, "learning_rate": 5.228512141112535e-06, "loss": 0.0241, "step": 34420 }, { "grad_norm": 0.03871544077992439, "learning_rate": 5.21012420805872e-06, "loss": 0.0244, "step": 34430 }, { "grad_norm": 0.043241143226623535, "learning_rate": 5.191766888462679e-06, "loss": 0.0253, "step": 34440 }, { "grad_norm": 0.038458604365587234, "learning_rate": 5.173440194871437e-06, "loss": 0.0246, "step": 34450 }, { "grad_norm": 0.029019679874181747, "learning_rate": 5.155144139811141e-06, "loss": 0.0251, "step": 34460 }, { "grad_norm": 0.042357973754405975, "learning_rate": 5.13687873578696e-06, "loss": 0.0237, "step": 34470 }, { "grad_norm": 0.040684256702661514, "learning_rate": 5.11864399528314e-06, "loss": 0.0245, "step": 34480 }, { "grad_norm": 0.03470860794186592, "learning_rate": 5.1004399307629346e-06, "loss": 0.0244, "step": 34490 }, { "grad_norm": 0.040557075291872025, "learning_rate": 5.082266554668652e-06, "loss": 0.027, "step": 34500 }, { "grad_norm": 0.043569087982177734, "learning_rate": 5.06412387942164e-06, "loss": 0.0241, "step": 34510 }, { "grad_norm": 0.05192790552973747, "learning_rate": 5.046011917422239e-06, "loss": 0.024, "step": 34520 }, { "grad_norm": 0.045733820647001266, "learning_rate": 5.0279306810498026e-06, "loss": 0.0258, "step": 34530 }, { "grad_norm": 0.03631332144141197, "learning_rate": 5.009880182662679e-06, "loss": 0.0227, "step": 34540 }, { "grad_norm": 0.1022462472319603, "learning_rate": 4.991860434598217e-06, "loss": 0.0212, "step": 34550 }, { "grad_norm": 0.04181404039263725, "learning_rate": 4.973871449172746e-06, "loss": 0.0206, "step": 34560 }, { "grad_norm": 0.03776983544230461, "learning_rate": 4.9559132386815695e-06, "loss": 0.0219, "step": 34570 }, { "grad_norm": 0.04153115302324295, "learning_rate": 4.93798581539896e-06, "loss": 0.0194, "step": 34580 }, { "grad_norm": 0.031416814774274826, "learning_rate": 4.920089191578131e-06, "loss": 0.0216, "step": 34590 }, { "grad_norm": 0.05648430436849594, "learning_rate": 4.90222337945126e-06, "loss": 0.0209, "step": 34600 }, { "grad_norm": 0.03228159248828888, "learning_rate": 4.8843883912294605e-06, "loss": 0.0204, "step": 34610 }, { "grad_norm": 0.09059508144855499, "learning_rate": 4.866584239102795e-06, "loss": 0.0222, "step": 34620 }, { "grad_norm": 0.03327091783285141, "learning_rate": 4.8488109352402164e-06, "loss": 0.0232, "step": 34630 }, { "grad_norm": 0.06627437472343445, "learning_rate": 4.831068491789603e-06, "loss": 0.0213, "step": 34640 }, { "grad_norm": 0.037464674562215805, "learning_rate": 4.81335692087776e-06, "loss": 0.0222, "step": 34650 }, { "grad_norm": 0.05233924463391304, "learning_rate": 4.795676234610374e-06, "loss": 0.0213, "step": 34660 }, { "grad_norm": 0.04604183882474899, "learning_rate": 4.778026445072037e-06, "loss": 0.0235, "step": 34670 }, { "grad_norm": 0.043832436203956604, "learning_rate": 4.760407564326208e-06, "loss": 0.0207, "step": 34680 }, { "grad_norm": 0.036828674376010895, "learning_rate": 4.742819604415222e-06, "loss": 0.0227, "step": 34690 }, { "grad_norm": 0.037221185863018036, "learning_rate": 4.725262577360284e-06, "loss": 0.0223, "step": 34700 }, { "grad_norm": 0.04950142279267311, "learning_rate": 4.707736495161458e-06, "loss": 0.022, "step": 34710 }, { "grad_norm": 0.042575959116220474, "learning_rate": 4.690241369797671e-06, "loss": 0.025, "step": 34720 }, { "grad_norm": 0.04243168234825134, "learning_rate": 4.672777213226659e-06, "loss": 0.0228, "step": 34730 }, { "grad_norm": 0.02911657840013504, "learning_rate": 4.655344037385001e-06, "loss": 0.0217, "step": 34740 }, { "grad_norm": 0.03514567017555237, "learning_rate": 4.6379418541881215e-06, "loss": 0.0202, "step": 34750 }, { "grad_norm": 0.024988003075122833, "learning_rate": 4.620570675530245e-06, "loss": 0.0205, "step": 34760 }, { "grad_norm": 0.028730088844895363, "learning_rate": 4.603230513284407e-06, "loss": 0.0203, "step": 34770 }, { "grad_norm": 0.0393696129322052, "learning_rate": 4.58592137930246e-06, "loss": 0.0204, "step": 34780 }, { "grad_norm": 0.024039335548877716, "learning_rate": 4.568643285415009e-06, "loss": 0.0198, "step": 34790 }, { "grad_norm": 0.035010918974876404, "learning_rate": 4.551396243431483e-06, "loss": 0.0211, "step": 34800 }, { "grad_norm": 0.04228962957859039, "learning_rate": 4.534180265140075e-06, "loss": 0.0242, "step": 34810 }, { "grad_norm": 0.035306647419929504, "learning_rate": 4.5169953623077356e-06, "loss": 0.0218, "step": 34820 }, { "grad_norm": 0.035059984773397446, "learning_rate": 4.499841546680189e-06, "loss": 0.0229, "step": 34830 }, { "grad_norm": 0.039788711816072464, "learning_rate": 4.4827188299819025e-06, "loss": 0.0202, "step": 34840 }, { "grad_norm": 0.053852252662181854, "learning_rate": 4.465627223916086e-06, "loss": 0.0214, "step": 34850 }, { "grad_norm": 0.033023279160261154, "learning_rate": 4.448566740164706e-06, "loss": 0.0233, "step": 34860 }, { "grad_norm": 0.030494581907987595, "learning_rate": 4.431537390388429e-06, "loss": 0.0237, "step": 34870 }, { "grad_norm": 0.04084949567914009, "learning_rate": 4.414539186226674e-06, "loss": 0.023, "step": 34880 }, { "grad_norm": 0.04089001566171646, "learning_rate": 4.397572139297529e-06, "loss": 0.0214, "step": 34890 }, { "grad_norm": 0.04937664046883583, "learning_rate": 4.380636261197829e-06, "loss": 0.0219, "step": 34900 }, { "grad_norm": 0.061189740896224976, "learning_rate": 4.36373156350307e-06, "loss": 0.0216, "step": 34910 }, { "grad_norm": 0.049920715391635895, "learning_rate": 4.3468580577674665e-06, "loss": 0.0218, "step": 34920 }, { "grad_norm": 0.0821775421500206, "learning_rate": 4.330015755523908e-06, "loss": 0.0241, "step": 34930 }, { "grad_norm": 0.0489848256111145, "learning_rate": 4.313204668283926e-06, "loss": 0.0243, "step": 34940 }, { "grad_norm": 0.05765592306852341, "learning_rate": 4.296424807537752e-06, "loss": 0.0242, "step": 34950 }, { "grad_norm": 0.04654596373438835, "learning_rate": 4.279676184754267e-06, "loss": 0.0228, "step": 34960 }, { "grad_norm": 0.048527587205171585, "learning_rate": 4.262958811380985e-06, "loss": 0.0242, "step": 34970 }, { "grad_norm": 0.062224820256233215, "learning_rate": 4.246272698844095e-06, "loss": 0.0237, "step": 34980 }, { "grad_norm": 0.042014043778181076, "learning_rate": 4.229617858548374e-06, "loss": 0.0227, "step": 34990 }, { "grad_norm": 0.03691699355840683, "learning_rate": 4.212994301877271e-06, "loss": 0.0219, "step": 35000 }, { "grad_norm": 0.03565715253353119, "learning_rate": 4.196402040192804e-06, "loss": 0.021, "step": 35010 }, { "grad_norm": 0.03538196161389351, "learning_rate": 4.179841084835639e-06, "loss": 0.0223, "step": 35020 }, { "grad_norm": 0.051452796906232834, "learning_rate": 4.163311447125046e-06, "loss": 0.0206, "step": 35030 }, { "grad_norm": 0.03611215576529503, "learning_rate": 4.146813138358851e-06, "loss": 0.0216, "step": 35040 }, { "grad_norm": 0.042668018490076065, "learning_rate": 4.130346169813509e-06, "loss": 0.0216, "step": 35050 }, { "grad_norm": 0.03946681693196297, "learning_rate": 4.113910552744038e-06, "loss": 0.0214, "step": 35060 }, { "grad_norm": 0.03540675342082977, "learning_rate": 4.097506298384024e-06, "loss": 0.0204, "step": 35070 }, { "grad_norm": 0.045506801456213, "learning_rate": 4.081133417945626e-06, "loss": 0.0229, "step": 35080 }, { "grad_norm": 0.03338050842285156, "learning_rate": 4.0647919226195484e-06, "loss": 0.0205, "step": 35090 }, { "grad_norm": 0.05194700136780739, "learning_rate": 4.04848182357504e-06, "loss": 0.0206, "step": 35100 }, { "grad_norm": 0.02886483445763588, "learning_rate": 4.032203131959911e-06, "loss": 0.0228, "step": 35110 }, { "grad_norm": 0.03711486980319023, "learning_rate": 4.015955858900483e-06, "loss": 0.0241, "step": 35120 }, { "grad_norm": 0.04649556055665016, "learning_rate": 3.999740015501641e-06, "loss": 0.0234, "step": 35130 }, { "grad_norm": 0.04392290487885475, "learning_rate": 3.9835556128467236e-06, "loss": 0.0271, "step": 35140 }, { "grad_norm": 0.03929334878921509, "learning_rate": 3.967402661997644e-06, "loss": 0.0252, "step": 35150 }, { "grad_norm": 0.031558334827423096, "learning_rate": 3.951281173994781e-06, "loss": 0.0258, "step": 35160 }, { "grad_norm": 0.045632489025592804, "learning_rate": 3.935191159857016e-06, "loss": 0.0246, "step": 35170 }, { "grad_norm": 0.04327121376991272, "learning_rate": 3.919132630581746e-06, "loss": 0.0251, "step": 35180 }, { "grad_norm": 0.039546407759189606, "learning_rate": 3.903105597144802e-06, "loss": 0.0237, "step": 35190 }, { "grad_norm": 0.04573062062263489, "learning_rate": 3.887110070500505e-06, "loss": 0.0224, "step": 35200 }, { "grad_norm": 0.07868589460849762, "learning_rate": 3.871146061581654e-06, "loss": 0.021, "step": 35210 }, { "grad_norm": 0.025884374976158142, "learning_rate": 3.855213581299505e-06, "loss": 0.0232, "step": 35220 }, { "grad_norm": 0.03272281587123871, "learning_rate": 3.839312640543763e-06, "loss": 0.0224, "step": 35230 }, { "grad_norm": 0.033219534903764725, "learning_rate": 3.823443250182551e-06, "loss": 0.0227, "step": 35240 }, { "grad_norm": 0.157049760222435, "learning_rate": 3.80760542106246e-06, "loss": 0.0195, "step": 35250 }, { "grad_norm": 0.03930722922086716, "learning_rate": 3.7917991640084972e-06, "loss": 0.02, "step": 35260 }, { "grad_norm": 0.06516987085342407, "learning_rate": 3.7760244898240925e-06, "loss": 0.0225, "step": 35270 }, { "grad_norm": 0.03398682549595833, "learning_rate": 3.760281409291078e-06, "loss": 0.0223, "step": 35280 }, { "grad_norm": 0.0418245866894722, "learning_rate": 3.7445699331697113e-06, "loss": 0.0206, "step": 35290 }, { "grad_norm": 0.039542555809020996, "learning_rate": 3.7288900721986262e-06, "loss": 0.0221, "step": 35300 }, { "grad_norm": 0.04571922495961189, "learning_rate": 3.7132418370948663e-06, "loss": 0.0216, "step": 35310 }, { "grad_norm": 0.06266781687736511, "learning_rate": 3.6976252385538535e-06, "loss": 0.0209, "step": 35320 }, { "grad_norm": 0.057195473462343216, "learning_rate": 3.6820402872493832e-06, "loss": 0.0214, "step": 35330 }, { "grad_norm": 0.04591064527630806, "learning_rate": 3.6664869938336343e-06, "loss": 0.0214, "step": 35340 }, { "grad_norm": 0.029403846710920334, "learning_rate": 3.650965368937115e-06, "loss": 0.0191, "step": 35350 }, { "grad_norm": 0.036047182977199554, "learning_rate": 3.6354754231687214e-06, "loss": 0.0221, "step": 35360 }, { "grad_norm": 0.03764341399073601, "learning_rate": 3.620017167115697e-06, "loss": 0.0216, "step": 35370 }, { "grad_norm": 0.03714681789278984, "learning_rate": 3.604590611343589e-06, "loss": 0.0225, "step": 35380 }, { "grad_norm": 0.036074746400117874, "learning_rate": 3.589195766396325e-06, "loss": 0.0215, "step": 35390 }, { "grad_norm": 0.0415482223033905, "learning_rate": 3.573832642796121e-06, "loss": 0.0224, "step": 35400 }, { "grad_norm": 0.032050393521785736, "learning_rate": 3.5585012510435336e-06, "loss": 0.0216, "step": 35410 }, { "grad_norm": 0.03658490628004074, "learning_rate": 3.5432016016174184e-06, "loss": 0.0216, "step": 35420 }, { "grad_norm": 0.039615780115127563, "learning_rate": 3.5279337049749506e-06, "loss": 0.0203, "step": 35430 }, { "grad_norm": 0.1365239918231964, "learning_rate": 3.5126975715515985e-06, "loss": 0.0229, "step": 35440 }, { "grad_norm": 0.10460600256919861, "learning_rate": 3.4974932117610994e-06, "loss": 0.0209, "step": 35450 }, { "grad_norm": 0.044283170253038406, "learning_rate": 3.4823206359955065e-06, "loss": 0.0231, "step": 35460 }, { "grad_norm": 0.05698508769273758, "learning_rate": 3.4671798546251133e-06, "loss": 0.0205, "step": 35470 }, { "grad_norm": 0.03785049542784691, "learning_rate": 3.4520708779985088e-06, "loss": 0.021, "step": 35480 }, { "grad_norm": 0.10535518079996109, "learning_rate": 3.4369937164425493e-06, "loss": 0.0216, "step": 35490 }, { "grad_norm": 0.0525081492960453, "learning_rate": 3.421948380262313e-06, "loss": 0.0194, "step": 35500 }, { "grad_norm": 0.030300822108983994, "learning_rate": 3.406934879741158e-06, "loss": 0.0177, "step": 35510 }, { "grad_norm": 0.03372390940785408, "learning_rate": 3.391953225140665e-06, "loss": 0.0207, "step": 35520 }, { "grad_norm": 0.02861492708325386, "learning_rate": 3.3770034267006533e-06, "loss": 0.0196, "step": 35530 }, { "grad_norm": 0.024369409307837486, "learning_rate": 3.362085494639178e-06, "loss": 0.0201, "step": 35540 }, { "grad_norm": 0.03249485790729523, "learning_rate": 3.3471994391524986e-06, "loss": 0.021, "step": 35550 }, { "grad_norm": 0.040912143886089325, "learning_rate": 3.3323452704150825e-06, "loss": 0.0199, "step": 35560 }, { "grad_norm": 0.09592052549123764, "learning_rate": 3.3175229985796253e-06, "loss": 0.0207, "step": 35570 }, { "grad_norm": 0.047311536967754364, "learning_rate": 3.3027326337770115e-06, "loss": 0.0192, "step": 35580 }, { "grad_norm": 0.04659786447882652, "learning_rate": 3.2879741861163216e-06, "loss": 0.0198, "step": 35590 }, { "grad_norm": 0.04196186363697052, "learning_rate": 3.2732476656847976e-06, "loss": 0.0189, "step": 35600 }, { "grad_norm": 0.05120830237865448, "learning_rate": 3.2585530825478883e-06, "loss": 0.0189, "step": 35610 }, { "grad_norm": 0.03599473461508751, "learning_rate": 3.243890446749204e-06, "loss": 0.0199, "step": 35620 }, { "grad_norm": 0.08062446117401123, "learning_rate": 3.2292597683105107e-06, "loss": 0.0224, "step": 35630 }, { "grad_norm": 0.09694268554449081, "learning_rate": 3.214661057231755e-06, "loss": 0.022, "step": 35640 }, { "grad_norm": 0.03123961202800274, "learning_rate": 3.2000943234910097e-06, "loss": 0.0215, "step": 35650 }, { "grad_norm": 0.0696980208158493, "learning_rate": 3.1855595770444835e-06, "loss": 0.0233, "step": 35660 }, { "grad_norm": 0.03387654572725296, "learning_rate": 3.1710568278265574e-06, "loss": 0.0222, "step": 35670 }, { "grad_norm": 0.04491005837917328, "learning_rate": 3.156586085749719e-06, "loss": 0.024, "step": 35680 }, { "grad_norm": 0.05252356082201004, "learning_rate": 3.1421473607045893e-06, "loss": 0.0261, "step": 35690 }, { "grad_norm": 0.02932259626686573, "learning_rate": 3.127740662559886e-06, "loss": 0.0236, "step": 35700 }, { "grad_norm": 0.06280732899904251, "learning_rate": 3.1133660011624655e-06, "loss": 0.0243, "step": 35710 }, { "grad_norm": 0.02925565280020237, "learning_rate": 3.099023386337263e-06, "loss": 0.0215, "step": 35720 }, { "grad_norm": 0.04471457004547119, "learning_rate": 3.0847128278873317e-06, "loss": 0.0239, "step": 35730 }, { "grad_norm": 0.033495813608169556, "learning_rate": 3.0704343355938027e-06, "loss": 0.0216, "step": 35740 }, { "grad_norm": 0.037688761949539185, "learning_rate": 3.056187919215886e-06, "loss": 0.0222, "step": 35750 }, { "grad_norm": 0.06169980764389038, "learning_rate": 3.041973588490865e-06, "loss": 0.0215, "step": 35760 }, { "grad_norm": 0.030803006142377853, "learning_rate": 3.0277913531341128e-06, "loss": 0.0223, "step": 35770 }, { "grad_norm": 0.036401644349098206, "learning_rate": 3.0136412228390476e-06, "loss": 0.0219, "step": 35780 }, { "grad_norm": 0.032573152333498, "learning_rate": 2.999523207277155e-06, "loss": 0.0222, "step": 35790 }, { "grad_norm": 0.03805851936340332, "learning_rate": 2.9854373160979552e-06, "loss": 0.0203, "step": 35800 }, { "grad_norm": 0.05617113783955574, "learning_rate": 2.9713835589290362e-06, "loss": 0.0217, "step": 35810 }, { "grad_norm": 0.04665475711226463, "learning_rate": 2.9573619453759913e-06, "loss": 0.0226, "step": 35820 }, { "grad_norm": 0.034773487597703934, "learning_rate": 2.943372485022489e-06, "loss": 0.0221, "step": 35830 }, { "grad_norm": 0.05160670727491379, "learning_rate": 2.9294151874301623e-06, "loss": 0.0238, "step": 35840 }, { "grad_norm": 0.03182518854737282, "learning_rate": 2.9154900621387093e-06, "loss": 0.0213, "step": 35850 }, { "grad_norm": 0.04061645641922951, "learning_rate": 2.9015971186658164e-06, "loss": 0.0235, "step": 35860 }, { "grad_norm": 0.03864875063300133, "learning_rate": 2.887736366507182e-06, "loss": 0.0229, "step": 35870 }, { "grad_norm": 0.047357577830553055, "learning_rate": 2.873907815136501e-06, "loss": 0.0207, "step": 35880 }, { "grad_norm": 0.05651531368494034, "learning_rate": 2.8601114740054556e-06, "loss": 0.0214, "step": 35890 }, { "grad_norm": 0.04226518049836159, "learning_rate": 2.8463473525437155e-06, "loss": 0.0196, "step": 35900 }, { "grad_norm": 0.04464986175298691, "learning_rate": 2.8326154601589216e-06, "loss": 0.0202, "step": 35910 }, { "grad_norm": 0.03425787761807442, "learning_rate": 2.818915806236705e-06, "loss": 0.0216, "step": 35920 }, { "grad_norm": 0.033257149159908295, "learning_rate": 2.8052484001406478e-06, "loss": 0.021, "step": 35930 }, { "grad_norm": 0.05677444487810135, "learning_rate": 2.791613251212277e-06, "loss": 0.0218, "step": 35940 }, { "grad_norm": 0.022627050057053566, "learning_rate": 2.7780103687711166e-06, "loss": 0.0204, "step": 35950 }, { "grad_norm": 0.06086407229304314, "learning_rate": 2.764439762114579e-06, "loss": 0.0207, "step": 35960 }, { "grad_norm": 0.030806628987193108, "learning_rate": 2.750901440518061e-06, "loss": 0.0218, "step": 35970 }, { "grad_norm": 0.11129768192768097, "learning_rate": 2.7373954132348776e-06, "loss": 0.023, "step": 35980 }, { "grad_norm": 0.027777787297964096, "learning_rate": 2.7239216894962784e-06, "loss": 0.0222, "step": 35990 }, { "grad_norm": 0.11605488508939743, "learning_rate": 2.710480278511407e-06, "loss": 0.0221, "step": 36000 }, { "grad_norm": 0.03506011515855789, "learning_rate": 2.6970711894673483e-06, "loss": 0.0234, "step": 36010 }, { "grad_norm": 0.060330796986818314, "learning_rate": 2.6836944315291047e-06, "loss": 0.0194, "step": 36020 }, { "grad_norm": 0.03741972893476486, "learning_rate": 2.67035001383954e-06, "loss": 0.0214, "step": 36030 }, { "grad_norm": 0.03946124389767647, "learning_rate": 2.6570379455194527e-06, "loss": 0.02, "step": 36040 }, { "grad_norm": 0.03402768075466156, "learning_rate": 2.6437582356675206e-06, "loss": 0.0212, "step": 36050 }, { "grad_norm": 0.16418032348155975, "learning_rate": 2.6305108933602775e-06, "loss": 0.021, "step": 36060 }, { "grad_norm": 0.033079713582992554, "learning_rate": 2.6172959276521803e-06, "loss": 0.0211, "step": 36070 }, { "grad_norm": 0.03410976752638817, "learning_rate": 2.6041133475755153e-06, "loss": 0.0226, "step": 36080 }, { "grad_norm": 0.07053786516189575, "learning_rate": 2.590963162140475e-06, "loss": 0.0217, "step": 36090 }, { "grad_norm": 0.038097307085990906, "learning_rate": 2.5778453803350587e-06, "loss": 0.023, "step": 36100 }, { "grad_norm": 0.031606338918209076, "learning_rate": 2.5647600111251723e-06, "loss": 0.0229, "step": 36110 }, { "grad_norm": 0.05843028798699379, "learning_rate": 2.551707063454528e-06, "loss": 0.0242, "step": 36120 }, { "grad_norm": 0.05425525829195976, "learning_rate": 2.5386865462446897e-06, "loss": 0.0247, "step": 36130 }, { "grad_norm": 0.03790887072682381, "learning_rate": 2.52569846839506e-06, "loss": 0.024, "step": 36140 }, { "grad_norm": 0.03196831792593002, "learning_rate": 2.512742838782883e-06, "loss": 0.0226, "step": 36150 }, { "grad_norm": 0.04058605805039406, "learning_rate": 2.4998196662631924e-06, "loss": 0.025, "step": 36160 }, { "grad_norm": 0.034596577286720276, "learning_rate": 2.4869289596688604e-06, "loss": 0.0216, "step": 36170 }, { "grad_norm": 0.07164624333381653, "learning_rate": 2.4740707278105567e-06, "loss": 0.0214, "step": 36180 }, { "grad_norm": 0.02833396941423416, "learning_rate": 2.461244979476779e-06, "loss": 0.0238, "step": 36190 }, { "grad_norm": 0.044237170368433, "learning_rate": 2.4484517234337866e-06, "loss": 0.0219, "step": 36200 }, { "grad_norm": 0.04216456785798073, "learning_rate": 2.4356909684256634e-06, "loss": 0.0213, "step": 36210 }, { "grad_norm": 0.03273652493953705, "learning_rate": 2.4229627231742545e-06, "loss": 0.0216, "step": 36220 }, { "grad_norm": 0.04354054480791092, "learning_rate": 2.4102669963791956e-06, "loss": 0.0212, "step": 36230 }, { "grad_norm": 0.044517967849969864, "learning_rate": 2.3976037967179065e-06, "loss": 0.0202, "step": 36240 }, { "grad_norm": 0.04880671203136444, "learning_rate": 2.3849731328455582e-06, "loss": 0.0194, "step": 36250 }, { "grad_norm": 0.04218175634741783, "learning_rate": 2.3723750133950894e-06, "loss": 0.0204, "step": 36260 }, { "grad_norm": 0.04227535054087639, "learning_rate": 2.3598094469771957e-06, "loss": 0.0213, "step": 36270 }, { "grad_norm": 0.041676346212625504, "learning_rate": 2.347276442180324e-06, "loss": 0.0219, "step": 36280 }, { "grad_norm": 0.029147641733288765, "learning_rate": 2.3347760075706715e-06, "loss": 0.0198, "step": 36290 }, { "grad_norm": 0.039502013474702835, "learning_rate": 2.3223081516921595e-06, "loss": 0.0223, "step": 36300 }, { "grad_norm": 0.03473476693034172, "learning_rate": 2.3098728830664494e-06, "loss": 0.0207, "step": 36310 }, { "grad_norm": 0.03730541840195656, "learning_rate": 2.2974702101929314e-06, "loss": 0.0226, "step": 36320 }, { "grad_norm": 0.0414801724255085, "learning_rate": 2.2851001415487185e-06, "loss": 0.0207, "step": 36330 }, { "grad_norm": 0.047849737107753754, "learning_rate": 2.272762685588631e-06, "loss": 0.0203, "step": 36340 }, { "grad_norm": 0.07202187925577164, "learning_rate": 2.260457850745218e-06, "loss": 0.0188, "step": 36350 }, { "grad_norm": 0.06453447788953781, "learning_rate": 2.2481856454286963e-06, "loss": 0.0208, "step": 36360 }, { "grad_norm": 0.03651243448257446, "learning_rate": 2.235946078027018e-06, "loss": 0.0194, "step": 36370 }, { "grad_norm": 0.02729095332324505, "learning_rate": 2.223739156905802e-06, "loss": 0.0211, "step": 36380 }, { "grad_norm": 0.08406355232000351, "learning_rate": 2.2115648904083796e-06, "loss": 0.0238, "step": 36390 }, { "grad_norm": 0.04897258058190346, "learning_rate": 2.199423286855734e-06, "loss": 0.0253, "step": 36400 }, { "grad_norm": 0.05420144274830818, "learning_rate": 2.1873143545465323e-06, "loss": 0.0234, "step": 36410 }, { "grad_norm": 0.07458694279193878, "learning_rate": 2.1752381017571267e-06, "loss": 0.0218, "step": 36420 }, { "grad_norm": 0.0660233423113823, "learning_rate": 2.1631945367415142e-06, "loss": 0.0237, "step": 36430 }, { "grad_norm": 0.05209420621395111, "learning_rate": 2.15118366773136e-06, "loss": 0.0224, "step": 36440 }, { "grad_norm": 0.030076853930950165, "learning_rate": 2.1392055029359926e-06, "loss": 0.0238, "step": 36450 }, { "grad_norm": 0.03981539607048035, "learning_rate": 2.127260050542357e-06, "loss": 0.0234, "step": 36460 }, { "grad_norm": 0.15211321413516998, "learning_rate": 2.1153473187150608e-06, "loss": 0.0217, "step": 36470 }, { "grad_norm": 0.03034614399075508, "learning_rate": 2.1034673155963467e-06, "loss": 0.0228, "step": 36480 }, { "grad_norm": 0.047791093587875366, "learning_rate": 2.0916200493060976e-06, "loss": 0.0214, "step": 36490 }, { "grad_norm": 0.03900253400206566, "learning_rate": 2.0798055279417916e-06, "loss": 0.0223, "step": 36500 }, { "grad_norm": 0.10923036932945251, "learning_rate": 2.0680237595785356e-06, "loss": 0.0215, "step": 36510 }, { "grad_norm": 0.034594032913446426, "learning_rate": 2.0562747522690774e-06, "loss": 0.0213, "step": 36520 }, { "grad_norm": 0.044664572924375534, "learning_rate": 2.0445585140437373e-06, "loss": 0.0236, "step": 36530 }, { "grad_norm": 0.06361228227615356, "learning_rate": 2.0328750529104602e-06, "loss": 0.0228, "step": 36540 }, { "grad_norm": 0.037367403507232666, "learning_rate": 2.021224376854791e-06, "loss": 0.0222, "step": 36550 }, { "grad_norm": 0.03991701826453209, "learning_rate": 2.0096064938398386e-06, "loss": 0.025, "step": 36560 }, { "grad_norm": 0.041693635284900665, "learning_rate": 1.998021411806328e-06, "loss": 0.0213, "step": 36570 }, { "grad_norm": 0.03142857551574707, "learning_rate": 1.9864691386725586e-06, "loss": 0.0224, "step": 36580 }, { "grad_norm": 0.026617590337991714, "learning_rate": 1.9749496823343814e-06, "loss": 0.0244, "step": 36590 }, { "grad_norm": 0.03802197054028511, "learning_rate": 1.963463050665254e-06, "loss": 0.0236, "step": 36600 }, { "grad_norm": 0.032829660922288895, "learning_rate": 1.9520092515161682e-06, "loss": 0.0262, "step": 36610 }, { "grad_norm": 0.047630030661821365, "learning_rate": 1.940588292715695e-06, "loss": 0.0246, "step": 36620 }, { "grad_norm": 0.03546590730547905, "learning_rate": 1.929200182069951e-06, "loss": 0.0245, "step": 36630 }, { "grad_norm": 0.04400722682476044, "learning_rate": 1.9178449273625997e-06, "loss": 0.0264, "step": 36640 }, { "grad_norm": 0.03555662930011749, "learning_rate": 1.90652253635486e-06, "loss": 0.0216, "step": 36650 }, { "grad_norm": 0.030306892469525337, "learning_rate": 1.8952330167854693e-06, "loss": 0.0232, "step": 36660 }, { "grad_norm": 0.05710633099079132, "learning_rate": 1.8839763763707107e-06, "loss": 0.023, "step": 36670 }, { "grad_norm": 0.03179145231842995, "learning_rate": 1.8727526228043912e-06, "loss": 0.0225, "step": 36680 }, { "grad_norm": 0.037634652107954025, "learning_rate": 1.8615617637578408e-06, "loss": 0.0232, "step": 36690 }, { "grad_norm": 0.022633131593465805, "learning_rate": 1.850403806879908e-06, "loss": 0.0216, "step": 36700 }, { "grad_norm": 0.030085787177085876, "learning_rate": 1.8392787597969475e-06, "loss": 0.0197, "step": 36710 }, { "grad_norm": 0.029858199879527092, "learning_rate": 1.828186630112827e-06, "loss": 0.0211, "step": 36720 }, { "grad_norm": 0.03415573015809059, "learning_rate": 1.8171274254089098e-06, "loss": 0.0197, "step": 36730 }, { "grad_norm": 0.058369606733322144, "learning_rate": 1.806101153244061e-06, "loss": 0.018, "step": 36740 }, { "grad_norm": 0.03278065845370293, "learning_rate": 1.7951078211546401e-06, "loss": 0.0191, "step": 36750 }, { "grad_norm": 0.027413642033934593, "learning_rate": 1.7841474366544764e-06, "loss": 0.0192, "step": 36760 }, { "grad_norm": 0.03174826502799988, "learning_rate": 1.7732200072348825e-06, "loss": 0.0213, "step": 36770 }, { "grad_norm": 0.037425536662340164, "learning_rate": 1.7623255403646677e-06, "loss": 0.02, "step": 36780 }, { "grad_norm": 0.03296070545911789, "learning_rate": 1.7514640434900865e-06, "loss": 0.0211, "step": 36790 }, { "grad_norm": 0.03472207486629486, "learning_rate": 1.7406355240348838e-06, "loss": 0.0226, "step": 36800 }, { "grad_norm": 0.04453844949603081, "learning_rate": 1.7298399894002392e-06, "loss": 0.0217, "step": 36810 }, { "grad_norm": 0.04382908716797829, "learning_rate": 1.7190774469647952e-06, "loss": 0.0236, "step": 36820 }, { "grad_norm": 0.03529861941933632, "learning_rate": 1.7083479040846617e-06, "loss": 0.0204, "step": 36830 }, { "grad_norm": 0.035790037363767624, "learning_rate": 1.697651368093367e-06, "loss": 0.0196, "step": 36840 }, { "grad_norm": 0.032145075500011444, "learning_rate": 1.6869878463019017e-06, "loss": 0.0193, "step": 36850 }, { "grad_norm": 0.03943804278969765, "learning_rate": 1.6763573459986858e-06, "loss": 0.0205, "step": 36860 }, { "grad_norm": 0.045988794416189194, "learning_rate": 1.6657598744495516e-06, "loss": 0.0205, "step": 36870 }, { "grad_norm": 0.033256128430366516, "learning_rate": 1.6551954388977774e-06, "loss": 0.019, "step": 36880 }, { "grad_norm": 0.0465000718832016, "learning_rate": 1.644664046564054e-06, "loss": 0.0214, "step": 36890 }, { "grad_norm": 0.03446193039417267, "learning_rate": 1.634165704646501e-06, "loss": 0.0222, "step": 36900 }, { "grad_norm": 0.03438636660575867, "learning_rate": 1.6237004203206175e-06, "loss": 0.0214, "step": 36910 }, { "grad_norm": 0.04074688255786896, "learning_rate": 1.6132682007393373e-06, "loss": 0.0212, "step": 36920 }, { "grad_norm": 0.03942296281456947, "learning_rate": 1.6028690530329847e-06, "loss": 0.0203, "step": 36930 }, { "grad_norm": 0.048089951276779175, "learning_rate": 1.5925029843092742e-06, "loss": 0.0218, "step": 36940 }, { "grad_norm": 0.04014762490987778, "learning_rate": 1.5821700016533214e-06, "loss": 0.0228, "step": 36950 }, { "grad_norm": 0.041572462767362595, "learning_rate": 1.5718701121276268e-06, "loss": 0.0207, "step": 36960 }, { "grad_norm": 0.045679181814193726, "learning_rate": 1.5616033227720483e-06, "loss": 0.0215, "step": 36970 }, { "grad_norm": 0.04610154777765274, "learning_rate": 1.5513696406038558e-06, "loss": 0.0224, "step": 36980 }, { "grad_norm": 0.04570995271205902, "learning_rate": 1.5411690726176653e-06, "loss": 0.023, "step": 36990 }, { "grad_norm": 0.046875786036252975, "learning_rate": 1.5310016257854832e-06, "loss": 0.0218, "step": 37000 }, { "grad_norm": 0.02756178006529808, "learning_rate": 1.5208673070566504e-06, "loss": 0.0218, "step": 37010 }, { "grad_norm": 0.03126571699976921, "learning_rate": 1.5107661233578818e-06, "loss": 0.0213, "step": 37020 }, { "grad_norm": 0.041378214955329895, "learning_rate": 1.5006980815932437e-06, "loss": 0.0232, "step": 37030 }, { "grad_norm": 0.0348002053797245, "learning_rate": 1.490663188644148e-06, "loss": 0.0219, "step": 37040 }, { "grad_norm": 0.06328056752681732, "learning_rate": 1.4806614513693418e-06, "loss": 0.0217, "step": 37050 }, { "grad_norm": 0.046016380190849304, "learning_rate": 1.4706928766049289e-06, "loss": 0.0204, "step": 37060 }, { "grad_norm": 0.03985212743282318, "learning_rate": 1.4607574711643257e-06, "loss": 0.0211, "step": 37070 }, { "grad_norm": 0.07342346012592316, "learning_rate": 1.4508552418382949e-06, "loss": 0.0221, "step": 37080 }, { "grad_norm": 0.038806039839982986, "learning_rate": 1.4409861953949112e-06, "loss": 0.0237, "step": 37090 }, { "grad_norm": 0.04205073416233063, "learning_rate": 1.43115033857959e-06, "loss": 0.0236, "step": 37100 }, { "grad_norm": 0.04342159256339073, "learning_rate": 1.4213476781150258e-06, "loss": 0.0225, "step": 37110 }, { "grad_norm": 0.055819351226091385, "learning_rate": 1.411578220701254e-06, "loss": 0.025, "step": 37120 }, { "grad_norm": 0.03286270424723625, "learning_rate": 1.4018419730156108e-06, "loss": 0.0213, "step": 37130 }, { "grad_norm": 0.048330072313547134, "learning_rate": 1.3921389417127283e-06, "loss": 0.024, "step": 37140 }, { "grad_norm": 0.042975712567567825, "learning_rate": 1.3824691334245298e-06, "loss": 0.0234, "step": 37150 }, { "grad_norm": 0.06340604275465012, "learning_rate": 1.3728325547602505e-06, "loss": 0.0219, "step": 37160 }, { "grad_norm": 0.0662156343460083, "learning_rate": 1.3632292123063828e-06, "loss": 0.0212, "step": 37170 }, { "grad_norm": 0.04524853825569153, "learning_rate": 1.3536591126267317e-06, "loss": 0.0221, "step": 37180 }, { "grad_norm": 0.04562412202358246, "learning_rate": 1.3441222622623761e-06, "loss": 0.0219, "step": 37190 }, { "grad_norm": 0.04947201907634735, "learning_rate": 1.3346186677316575e-06, "loss": 0.0223, "step": 37200 }, { "grad_norm": 0.02866511233150959, "learning_rate": 1.3251483355301852e-06, "loss": 0.0199, "step": 37210 }, { "grad_norm": 0.04488096013665199, "learning_rate": 1.315711272130854e-06, "loss": 0.0206, "step": 37220 }, { "grad_norm": 0.0426475889980793, "learning_rate": 1.306307483983804e-06, "loss": 0.0214, "step": 37230 }, { "grad_norm": 0.03571192920207977, "learning_rate": 1.2969369775164275e-06, "loss": 0.0198, "step": 37240 }, { "grad_norm": 0.041604653000831604, "learning_rate": 1.28759975913339e-06, "loss": 0.0208, "step": 37250 }, { "grad_norm": 0.04308483004570007, "learning_rate": 1.2782958352165863e-06, "loss": 0.0223, "step": 37260 }, { "grad_norm": 0.05647021532058716, "learning_rate": 1.2690252121251633e-06, "loss": 0.0217, "step": 37270 }, { "grad_norm": 0.04827849194407463, "learning_rate": 1.2597878961954968e-06, "loss": 0.0227, "step": 37280 }, { "grad_norm": 0.034507378935813904, "learning_rate": 1.250583893741214e-06, "loss": 0.0224, "step": 37290 }, { "grad_norm": 0.05313101410865784, "learning_rate": 1.2414132110531607e-06, "loss": 0.0231, "step": 37300 }, { "grad_norm": 0.044000137597322464, "learning_rate": 1.2322758543994062e-06, "loss": 0.0218, "step": 37310 }, { "grad_norm": 0.03225364908576012, "learning_rate": 1.2231718300252603e-06, "loss": 0.02, "step": 37320 }, { "grad_norm": 0.042806822806596756, "learning_rate": 1.214101144153229e-06, "loss": 0.0234, "step": 37330 }, { "grad_norm": 0.02208271250128746, "learning_rate": 1.205063802983042e-06, "loss": 0.0191, "step": 37340 }, { "grad_norm": 0.0553194023668766, "learning_rate": 1.1960598126916301e-06, "loss": 0.0184, "step": 37350 }, { "grad_norm": 0.049245454370975494, "learning_rate": 1.187089179433154e-06, "loss": 0.0214, "step": 37360 }, { "grad_norm": 0.03625768795609474, "learning_rate": 1.178151909338937e-06, "loss": 0.0202, "step": 37370 }, { "grad_norm": 0.025169728323817253, "learning_rate": 1.1692480085175317e-06, "loss": 0.0201, "step": 37380 }, { "grad_norm": 0.031138354912400246, "learning_rate": 1.1603774830546587e-06, "loss": 0.0199, "step": 37390 }, { "grad_norm": 0.03412410616874695, "learning_rate": 1.1515403390132573e-06, "loss": 0.0201, "step": 37400 }, { "grad_norm": 0.04695436730980873, "learning_rate": 1.1427365824334124e-06, "loss": 0.0224, "step": 37410 }, { "grad_norm": 0.03570137545466423, "learning_rate": 1.133966219332422e-06, "loss": 0.0239, "step": 37420 }, { "grad_norm": 0.0405246838927269, "learning_rate": 1.1252292557047405e-06, "loss": 0.0227, "step": 37430 }, { "grad_norm": 0.03959868475794792, "learning_rate": 1.1165256975219974e-06, "loss": 0.0221, "step": 37440 }, { "grad_norm": 0.08851056545972824, "learning_rate": 1.1078555507330002e-06, "loss": 0.0226, "step": 37450 }, { "grad_norm": 0.0590183287858963, "learning_rate": 1.0992188212637089e-06, "loss": 0.0228, "step": 37460 }, { "grad_norm": 0.04999823123216629, "learning_rate": 1.0906155150172458e-06, "loss": 0.0221, "step": 37470 }, { "grad_norm": 0.027637088671326637, "learning_rate": 1.0820456378738852e-06, "loss": 0.0205, "step": 37480 }, { "grad_norm": 0.09732401371002197, "learning_rate": 1.0735091956910692e-06, "loss": 0.0195, "step": 37490 }, { "grad_norm": 0.03616158664226532, "learning_rate": 1.0650061943033695e-06, "loss": 0.0205, "step": 37500 }, { "grad_norm": 0.03311160206794739, "learning_rate": 1.0565366395225041e-06, "loss": 0.0193, "step": 37510 }, { "grad_norm": 0.0450395829975605, "learning_rate": 1.0481005371373366e-06, "loss": 0.0199, "step": 37520 }, { "grad_norm": 0.03810988366603851, "learning_rate": 1.0396978929138602e-06, "loss": 0.02, "step": 37530 }, { "grad_norm": 0.033456362783908844, "learning_rate": 1.0313287125952032e-06, "loss": 0.0218, "step": 37540 }, { "grad_norm": 0.09506454318761826, "learning_rate": 1.0229930019016288e-06, "loss": 0.0251, "step": 37550 }, { "grad_norm": 0.022935567423701286, "learning_rate": 1.0146907665305183e-06, "loss": 0.0233, "step": 37560 }, { "grad_norm": 0.05627082288265228, "learning_rate": 1.0064220121563551e-06, "loss": 0.0223, "step": 37570 }, { "grad_norm": 0.033009808510541916, "learning_rate": 9.98186744430768e-07, "loss": 0.0218, "step": 37580 }, { "grad_norm": 0.05388855189085007, "learning_rate": 9.899849689824826e-07, "loss": 0.0224, "step": 37590 }, { "grad_norm": 0.13200756907463074, "learning_rate": 9.818166914173365e-07, "loss": 0.0212, "step": 37600 }, { "grad_norm": 0.06655251234769821, "learning_rate": 9.736819173182642e-07, "loss": 0.0198, "step": 37610 }, { "grad_norm": 0.039223555475473404, "learning_rate": 9.655806522453126e-07, "loss": 0.0201, "step": 37620 }, { "grad_norm": 0.045760735869407654, "learning_rate": 9.575129017356187e-07, "loss": 0.0202, "step": 37630 }, { "grad_norm": 0.04405907541513443, "learning_rate": 9.494786713034109e-07, "loss": 0.0183, "step": 37640 }, { "grad_norm": 0.041872087866067886, "learning_rate": 9.414779664400075e-07, "loss": 0.0213, "step": 37650 }, { "grad_norm": 0.03694995865225792, "learning_rate": 9.335107926138342e-07, "loss": 0.021, "step": 37660 }, { "grad_norm": 0.04030304029583931, "learning_rate": 9.255771552703519e-07, "loss": 0.0198, "step": 37670 }, { "grad_norm": 0.026166990399360657, "learning_rate": 9.176770598321449e-07, "loss": 0.0207, "step": 37680 }, { "grad_norm": 0.025762047618627548, "learning_rate": 9.098105116988498e-07, "loss": 0.0215, "step": 37690 }, { "grad_norm": 0.03230335935950279, "learning_rate": 9.019775162471767e-07, "loss": 0.021, "step": 37700 }, { "grad_norm": 0.03224270045757294, "learning_rate": 8.94178078830904e-07, "loss": 0.0222, "step": 37710 }, { "grad_norm": 0.03919527679681778, "learning_rate": 8.864122047808731e-07, "loss": 0.0212, "step": 37720 }, { "grad_norm": 0.0487036295235157, "learning_rate": 8.786798994049882e-07, "loss": 0.0207, "step": 37730 }, { "grad_norm": 0.06809131056070328, "learning_rate": 8.709811679882052e-07, "loss": 0.0196, "step": 37740 }, { "grad_norm": 0.0651586502790451, "learning_rate": 8.633160157925313e-07, "loss": 0.0226, "step": 37750 }, { "grad_norm": 0.029980242252349854, "learning_rate": 8.556844480570369e-07, "loss": 0.0211, "step": 37760 }, { "grad_norm": 0.0383610762655735, "learning_rate": 8.480864699978164e-07, "loss": 0.0211, "step": 37770 }, { "grad_norm": 0.039366018027067184, "learning_rate": 8.405220868080154e-07, "loss": 0.0203, "step": 37780 }, { "grad_norm": 0.044155023992061615, "learning_rate": 8.329913036578263e-07, "loss": 0.0213, "step": 37790 }, { "grad_norm": 0.04157377406954765, "learning_rate": 8.254941256944593e-07, "loss": 0.0219, "step": 37800 }, { "grad_norm": 0.0429031066596508, "learning_rate": 8.18030558042171e-07, "loss": 0.0225, "step": 37810 }, { "grad_norm": 0.039135441184043884, "learning_rate": 8.106006058022419e-07, "loss": 0.0219, "step": 37820 }, { "grad_norm": 0.05315380170941353, "learning_rate": 8.03204274052971e-07, "loss": 0.0233, "step": 37830 }, { "grad_norm": 0.052048757672309875, "learning_rate": 7.958415678496756e-07, "loss": 0.0209, "step": 37840 }, { "grad_norm": 0.05806933343410492, "learning_rate": 7.885124922247022e-07, "loss": 0.0203, "step": 37850 }, { "grad_norm": 0.04938412830233574, "learning_rate": 7.812170521874051e-07, "loss": 0.0193, "step": 37860 }, { "grad_norm": 0.03530612960457802, "learning_rate": 7.739552527241456e-07, "loss": 0.0196, "step": 37870 }, { "grad_norm": 0.03132877126336098, "learning_rate": 7.667270987983033e-07, "loss": 0.02, "step": 37880 }, { "grad_norm": 0.03749268129467964, "learning_rate": 7.595325953502375e-07, "loss": 0.0203, "step": 37890 }, { "grad_norm": 0.029544223099946976, "learning_rate": 7.523717472973368e-07, "loss": 0.021, "step": 37900 }, { "grad_norm": 0.03914153575897217, "learning_rate": 7.452445595339696e-07, "loss": 0.024, "step": 37910 }, { "grad_norm": 0.038749706000089645, "learning_rate": 7.381510369315059e-07, "loss": 0.0242, "step": 37920 }, { "grad_norm": 0.05691090226173401, "learning_rate": 7.310911843382895e-07, "loss": 0.0246, "step": 37930 }, { "grad_norm": 0.0370161272585392, "learning_rate": 7.240650065796716e-07, "loss": 0.0219, "step": 37940 }, { "grad_norm": 0.05006939917802811, "learning_rate": 7.170725084579777e-07, "loss": 0.0255, "step": 37950 }, { "grad_norm": 0.03325002267956734, "learning_rate": 7.101136947525233e-07, "loss": 0.0237, "step": 37960 }, { "grad_norm": 0.055385150015354156, "learning_rate": 7.031885702195707e-07, "loss": 0.0231, "step": 37970 }, { "grad_norm": 0.05899950861930847, "learning_rate": 6.962971395924e-07, "loss": 0.0244, "step": 37980 }, { "grad_norm": 0.03853357583284378, "learning_rate": 6.894394075812217e-07, "loss": 0.0252, "step": 37990 }, { "grad_norm": 0.04067906364798546, "learning_rate": 6.826153788732359e-07, "loss": 0.0245, "step": 38000 }, { "grad_norm": 0.03409135341644287, "learning_rate": 6.758250581326009e-07, "loss": 0.0221, "step": 38010 }, { "grad_norm": 0.03570278361439705, "learning_rate": 6.690684500004485e-07, "loss": 0.023, "step": 38020 }, { "grad_norm": 0.041380636394023895, "learning_rate": 6.623455590948458e-07, "loss": 0.0257, "step": 38030 }, { "grad_norm": 0.03053576685488224, "learning_rate": 6.556563900108226e-07, "loss": 0.0227, "step": 38040 }, { "grad_norm": 0.04075581580400467, "learning_rate": 6.490009473203662e-07, "loss": 0.024, "step": 38050 }, { "grad_norm": 0.02858196571469307, "learning_rate": 6.423792355724156e-07, "loss": 0.0216, "step": 38060 }, { "grad_norm": 0.04874531924724579, "learning_rate": 6.35791259292845e-07, "loss": 0.0233, "step": 38070 }, { "grad_norm": 0.04100551828742027, "learning_rate": 6.292370229844691e-07, "loss": 0.0222, "step": 38080 }, { "grad_norm": 0.03316871449351311, "learning_rate": 6.227165311270488e-07, "loss": 0.0217, "step": 38090 }, { "grad_norm": 0.0361994244158268, "learning_rate": 6.16229788177275e-07, "loss": 0.0222, "step": 38100 }, { "grad_norm": 0.03813052177429199, "learning_rate": 6.097767985687841e-07, "loss": 0.0205, "step": 38110 }, { "grad_norm": 0.03546948358416557, "learning_rate": 6.033575667121317e-07, "loss": 0.0227, "step": 38120 }, { "grad_norm": 0.03222384303808212, "learning_rate": 5.969720969947967e-07, "loss": 0.022, "step": 38130 }, { "grad_norm": 0.03551296517252922, "learning_rate": 5.906203937811883e-07, "loss": 0.0223, "step": 38140 }, { "grad_norm": 0.03737106919288635, "learning_rate": 5.843024614126447e-07, "loss": 0.0202, "step": 38150 }, { "grad_norm": 0.03354562446475029, "learning_rate": 5.780183042074061e-07, "loss": 0.0195, "step": 38160 }, { "grad_norm": 0.05413394793868065, "learning_rate": 5.717679264606312e-07, "loss": 0.019, "step": 38170 }, { "grad_norm": 0.07333846390247345, "learning_rate": 5.655513324444028e-07, "loss": 0.0221, "step": 38180 }, { "grad_norm": 0.052268706262111664, "learning_rate": 5.593685264076942e-07, "loss": 0.0185, "step": 38190 }, { "grad_norm": 0.03774308040738106, "learning_rate": 5.532195125764029e-07, "loss": 0.0203, "step": 38200 }, { "grad_norm": 0.030783237889409065, "learning_rate": 5.471042951533167e-07, "loss": 0.018, "step": 38210 }, { "grad_norm": 0.04451846331357956, "learning_rate": 5.410228783181371e-07, "loss": 0.0199, "step": 38220 }, { "grad_norm": 0.04017477482557297, "learning_rate": 5.349752662274444e-07, "loss": 0.018, "step": 38230 }, { "grad_norm": 0.04498263821005821, "learning_rate": 5.289614630147321e-07, "loss": 0.0182, "step": 38240 }, { "grad_norm": 0.03912936896085739, "learning_rate": 5.229814727903737e-07, "loss": 0.0209, "step": 38250 }, { "grad_norm": 0.06946945935487747, "learning_rate": 5.170352996416439e-07, "loss": 0.0198, "step": 38260 }, { "grad_norm": 0.03251708298921585, "learning_rate": 5.111229476326918e-07, "loss": 0.0194, "step": 38270 }, { "grad_norm": 0.0410856232047081, "learning_rate": 5.052444208045459e-07, "loss": 0.0193, "step": 38280 }, { "grad_norm": 0.028276201337575912, "learning_rate": 4.993997231751312e-07, "loss": 0.0186, "step": 38290 }, { "grad_norm": 0.043496742844581604, "learning_rate": 4.935888587392468e-07, "loss": 0.0219, "step": 38300 }, { "grad_norm": 0.044543929398059845, "learning_rate": 4.878118314685598e-07, "loss": 0.0207, "step": 38310 }, { "grad_norm": 0.058940257877111435, "learning_rate": 4.820686453116119e-07, "loss": 0.0227, "step": 38320 }, { "grad_norm": 0.039958782494068146, "learning_rate": 4.763593041938186e-07, "loss": 0.022, "step": 38330 }, { "grad_norm": 0.05400353670120239, "learning_rate": 4.706838120174584e-07, "loss": 0.023, "step": 38340 }, { "grad_norm": 0.03967221453785896, "learning_rate": 4.650421726616838e-07, "loss": 0.0216, "step": 38350 }, { "grad_norm": 0.03241384029388428, "learning_rate": 4.594343899824993e-07, "loss": 0.0223, "step": 38360 }, { "grad_norm": 0.032942526042461395, "learning_rate": 4.538604678127667e-07, "loss": 0.0216, "step": 38370 }, { "grad_norm": 0.04007360339164734, "learning_rate": 4.483204099622107e-07, "loss": 0.021, "step": 38380 }, { "grad_norm": 0.038342200219631195, "learning_rate": 4.4281422021740793e-07, "loss": 0.0199, "step": 38390 }, { "grad_norm": 0.05788779631257057, "learning_rate": 4.373419023417924e-07, "loss": 0.0218, "step": 38400 }, { "grad_norm": 0.05799879878759384, "learning_rate": 4.319034600756389e-07, "loss": 0.0196, "step": 38410 }, { "grad_norm": 0.04026064649224281, "learning_rate": 4.264988971360739e-07, "loss": 0.0234, "step": 38420 }, { "grad_norm": 0.038907404989004135, "learning_rate": 4.211282172170594e-07, "loss": 0.0218, "step": 38430 }, { "grad_norm": 0.03122934140264988, "learning_rate": 4.1579142398940894e-07, "loss": 0.0217, "step": 38440 }, { "grad_norm": 0.038650449365377426, "learning_rate": 4.1048852110076585e-07, "loss": 0.0204, "step": 38450 }, { "grad_norm": 0.04615982249379158, "learning_rate": 4.052195121756197e-07, "loss": 0.0233, "step": 38460 }, { "grad_norm": 0.038503557443618774, "learning_rate": 3.9998440081528974e-07, "loss": 0.0196, "step": 38470 }, { "grad_norm": 0.037525735795497894, "learning_rate": 3.9478319059791936e-07, "loss": 0.0193, "step": 38480 }, { "grad_norm": 0.038621045649051666, "learning_rate": 3.8961588507848143e-07, "loss": 0.0214, "step": 38490 }, { "grad_norm": 0.033917203545570374, "learning_rate": 3.844824877887954e-07, "loss": 0.0216, "step": 38500 }, { "grad_norm": 0.03379565849900246, "learning_rate": 3.7938300223748224e-07, "loss": 0.0219, "step": 38510 }, { "grad_norm": 0.041691746562719345, "learning_rate": 3.743174319099929e-07, "loss": 0.0226, "step": 38520 }, { "grad_norm": 0.03441058099269867, "learning_rate": 3.6928578026859673e-07, "loss": 0.0212, "step": 38530 }, { "grad_norm": 0.04268146678805351, "learning_rate": 3.642880507523816e-07, "loss": 0.0254, "step": 38540 }, { "grad_norm": 0.036921512335538864, "learning_rate": 3.5932424677724285e-07, "loss": 0.0236, "step": 38550 }, { "grad_norm": 0.04158921539783478, "learning_rate": 3.5439437173589996e-07, "loss": 0.0233, "step": 38560 }, { "grad_norm": 0.041686393320560455, "learning_rate": 3.4949842899787423e-07, "loss": 0.0236, "step": 38570 }, { "grad_norm": 0.12755542993545532, "learning_rate": 3.446364219094944e-07, "loss": 0.0239, "step": 38580 }, { "grad_norm": 0.03696272149682045, "learning_rate": 3.3980835379389675e-07, "loss": 0.0234, "step": 38590 }, { "grad_norm": 0.0889318510890007, "learning_rate": 3.350142279510249e-07, "loss": 0.0232, "step": 38600 }, { "grad_norm": 0.03545480966567993, "learning_rate": 3.3025404765761324e-07, "loss": 0.024, "step": 38610 }, { "grad_norm": 0.040283750742673874, "learning_rate": 3.255278161672037e-07, "loss": 0.0246, "step": 38620 }, { "grad_norm": 0.054110731929540634, "learning_rate": 3.20835536710129e-07, "loss": 0.021, "step": 38630 }, { "grad_norm": 0.040209610015153885, "learning_rate": 3.1617721249351804e-07, "loss": 0.023, "step": 38640 }, { "grad_norm": 0.05089676007628441, "learning_rate": 3.115528467012963e-07, "loss": 0.0222, "step": 38650 }, { "grad_norm": 0.042497631162405014, "learning_rate": 3.0696244249416305e-07, "loss": 0.023, "step": 38660 }, { "grad_norm": 0.047186363488435745, "learning_rate": 3.024060030096254e-07, "loss": 0.0216, "step": 38670 }, { "grad_norm": 0.03744322061538696, "learning_rate": 2.9788353136196435e-07, "loss": 0.0227, "step": 38680 }, { "grad_norm": 0.08156673610210419, "learning_rate": 2.9339503064224615e-07, "loss": 0.0229, "step": 38690 }, { "grad_norm": 0.04187962785363197, "learning_rate": 2.889405039183224e-07, "loss": 0.0215, "step": 38700 }, { "grad_norm": 0.07749072462320328, "learning_rate": 2.8451995423481315e-07, "loss": 0.0227, "step": 38710 }, { "grad_norm": 0.04577113687992096, "learning_rate": 2.801333846131293e-07, "loss": 0.0222, "step": 38720 }, { "grad_norm": 0.0340002179145813, "learning_rate": 2.757807980514504e-07, "loss": 0.0235, "step": 38730 }, { "grad_norm": 0.030911777168512344, "learning_rate": 2.714621975247189e-07, "loss": 0.021, "step": 38740 }, { "grad_norm": 0.04006553068757057, "learning_rate": 2.671775859846626e-07, "loss": 0.0207, "step": 38750 }, { "grad_norm": 0.04211931303143501, "learning_rate": 2.629269663597722e-07, "loss": 0.0206, "step": 38760 }, { "grad_norm": 0.042233314365148544, "learning_rate": 2.587103415553127e-07, "loss": 0.0218, "step": 38770 }, { "grad_norm": 0.055743228644132614, "learning_rate": 2.545277144533009e-07, "loss": 0.021, "step": 38780 }, { "grad_norm": 0.026278166100382805, "learning_rate": 2.5037908791252227e-07, "loss": 0.021, "step": 38790 }, { "grad_norm": 0.04241716489195824, "learning_rate": 2.462644647685308e-07, "loss": 0.0216, "step": 38800 }, { "grad_norm": 0.034295983612537384, "learning_rate": 2.421838478336269e-07, "loss": 0.0197, "step": 38810 }, { "grad_norm": 0.03148418292403221, "learning_rate": 2.3813723989687398e-07, "loss": 0.0217, "step": 38820 }, { "grad_norm": 0.037615444511175156, "learning_rate": 2.3412464372409293e-07, "loss": 0.0234, "step": 38830 }, { "grad_norm": 0.025750543922185898, "learning_rate": 2.301460620578455e-07, "loss": 0.0216, "step": 38840 }, { "grad_norm": 0.06085406243801117, "learning_rate": 2.262014976174731e-07, "loss": 0.0214, "step": 38850 }, { "grad_norm": 0.03321724757552147, "learning_rate": 2.2229095309903025e-07, "loss": 0.0218, "step": 38860 }, { "grad_norm": 0.030064184218645096, "learning_rate": 2.184144311753511e-07, "loss": 0.0181, "step": 38870 }, { "grad_norm": 0.03707718104124069, "learning_rate": 2.14571934495994e-07, "loss": 0.0175, "step": 38880 }, { "grad_norm": 0.030529366806149483, "learning_rate": 2.1076346568726923e-07, "loss": 0.0179, "step": 38890 }, { "grad_norm": 0.025019697844982147, "learning_rate": 2.0698902735223346e-07, "loss": 0.0197, "step": 38900 }, { "grad_norm": 0.03511465713381767, "learning_rate": 2.0324862207067863e-07, "loss": 0.019, "step": 38910 }, { "grad_norm": 0.027567904442548752, "learning_rate": 1.9954225239913195e-07, "loss": 0.0182, "step": 38920 }, { "grad_norm": 0.03986330330371857, "learning_rate": 1.9586992087087252e-07, "loss": 0.0211, "step": 38930 }, { "grad_norm": 0.06027914211153984, "learning_rate": 1.9223162999588705e-07, "loss": 0.0195, "step": 38940 }, { "grad_norm": 0.025156032294034958, "learning_rate": 1.886273822609308e-07, "loss": 0.0203, "step": 38950 }, { "grad_norm": 0.03706533461809158, "learning_rate": 1.8505718012946094e-07, "loss": 0.0216, "step": 38960 }, { "grad_norm": 0.02275983802974224, "learning_rate": 1.8152102604168664e-07, "loss": 0.0219, "step": 38970 }, { "grad_norm": 0.035244304686784744, "learning_rate": 1.7801892241452457e-07, "loss": 0.0221, "step": 38980 }, { "grad_norm": 0.022063206881284714, "learning_rate": 1.745508716416433e-07, "loss": 0.0218, "step": 38990 }, { "grad_norm": 0.04322776943445206, "learning_rate": 1.7111687609341342e-07, "loss": 0.0217, "step": 39000 }, { "grad_norm": 0.03581369295716286, "learning_rate": 1.6771693811693523e-07, "loss": 0.0237, "step": 39010 }, { "grad_norm": 0.03681626915931702, "learning_rate": 1.643510600360443e-07, "loss": 0.02, "step": 39020 }, { "grad_norm": 0.04761947691440582, "learning_rate": 1.610192441512781e-07, "loss": 0.0212, "step": 39030 }, { "grad_norm": 0.03287721052765846, "learning_rate": 1.5772149273990956e-07, "loss": 0.0216, "step": 39040 }, { "grad_norm": 0.02924688719213009, "learning_rate": 1.5445780805591337e-07, "loss": 0.0181, "step": 39050 }, { "grad_norm": 0.039822857826948166, "learning_rate": 1.5122819232998852e-07, "loss": 0.021, "step": 39060 }, { "grad_norm": 0.03876229375600815, "learning_rate": 1.4803264776955262e-07, "loss": 0.0227, "step": 39070 }, { "grad_norm": 0.03554261475801468, "learning_rate": 1.448711765587196e-07, "loss": 0.0224, "step": 39080 }, { "grad_norm": 0.041210826486349106, "learning_rate": 1.4174378085833885e-07, "loss": 0.0223, "step": 39090 }, { "grad_norm": 0.061877451837062836, "learning_rate": 1.3865046280594486e-07, "loss": 0.0228, "step": 39100 }, { "grad_norm": 0.04689539596438408, "learning_rate": 1.3559122451579643e-07, "loss": 0.0225, "step": 39110 }, { "grad_norm": 0.03842697665095329, "learning_rate": 1.3256606807884874e-07, "loss": 0.0237, "step": 39120 }, { "grad_norm": 0.050530433654785156, "learning_rate": 1.2957499556277563e-07, "loss": 0.023, "step": 39130 }, { "grad_norm": 0.03071979247033596, "learning_rate": 1.266180090119473e-07, "loss": 0.0218, "step": 39140 }, { "grad_norm": 0.037183646112680435, "learning_rate": 1.2369511044742488e-07, "loss": 0.0215, "step": 39150 }, { "grad_norm": 0.028612889349460602, "learning_rate": 1.2080630186698805e-07, "loss": 0.0215, "step": 39160 }, { "grad_norm": 0.03216112032532692, "learning_rate": 1.1795158524511851e-07, "loss": 0.0224, "step": 39170 }, { "grad_norm": 0.0374920554459095, "learning_rate": 1.1513096253297773e-07, "loss": 0.0203, "step": 39180 }, { "grad_norm": 0.03479057177901268, "learning_rate": 1.1234443565843466e-07, "loss": 0.0201, "step": 39190 }, { "grad_norm": 0.04489215463399887, "learning_rate": 1.0959200652606028e-07, "loss": 0.0198, "step": 39200 }, { "grad_norm": 0.050454720854759216, "learning_rate": 1.0687367701711082e-07, "loss": 0.0195, "step": 39210 }, { "grad_norm": 0.06850565969944, "learning_rate": 1.0418944898953898e-07, "loss": 0.019, "step": 39220 }, { "grad_norm": 0.040050070732831955, "learning_rate": 1.0153932427798829e-07, "loss": 0.0205, "step": 39230 }, { "grad_norm": 0.03855996951460838, "learning_rate": 9.892330469379874e-08, "loss": 0.0203, "step": 39240 }, { "grad_norm": 0.0479377843439579, "learning_rate": 9.634139202498449e-08, "loss": 0.0216, "step": 39250 }, { "grad_norm": 0.034256476908922195, "learning_rate": 9.379358803627281e-08, "loss": 0.02, "step": 39260 }, { "grad_norm": 0.04863450303673744, "learning_rate": 9.12798944690485e-08, "loss": 0.021, "step": 39270 }, { "grad_norm": 0.02274571731686592, "learning_rate": 8.880031304141501e-08, "loss": 0.0205, "step": 39280 }, { "grad_norm": 0.03696684166789055, "learning_rate": 8.635484544812777e-08, "loss": 0.0227, "step": 39290 }, { "grad_norm": 0.030391458421945572, "learning_rate": 8.394349336064422e-08, "loss": 0.0235, "step": 39300 }, { "grad_norm": 0.06930005550384521, "learning_rate": 8.156625842710153e-08, "loss": 0.0218, "step": 39310 }, { "grad_norm": 0.03625176101922989, "learning_rate": 7.92231422723222e-08, "loss": 0.0205, "step": 39320 }, { "grad_norm": 0.0888049378991127, "learning_rate": 7.691414649780293e-08, "loss": 0.0201, "step": 39330 }, { "grad_norm": 0.037473179399967194, "learning_rate": 7.463927268171467e-08, "loss": 0.0202, "step": 39340 }, { "grad_norm": 0.03329338878393173, "learning_rate": 7.239852237891365e-08, "loss": 0.0211, "step": 39350 }, { "grad_norm": 0.03996864706277847, "learning_rate": 7.019189712093588e-08, "loss": 0.0211, "step": 39360 }, { "grad_norm": 0.049301646649837494, "learning_rate": 6.801939841598603e-08, "loss": 0.0212, "step": 39370 }, { "grad_norm": 0.05603655427694321, "learning_rate": 6.588102774894856e-08, "loss": 0.0224, "step": 39380 }, { "grad_norm": 0.05174688249826431, "learning_rate": 6.377678658137653e-08, "loss": 0.0217, "step": 39390 }, { "grad_norm": 0.05443846434354782, "learning_rate": 6.17066763514973e-08, "loss": 0.0218, "step": 39400 }, { "grad_norm": 0.05411950871348381, "learning_rate": 5.967069847421792e-08, "loss": 0.0235, "step": 39410 }, { "grad_norm": 0.040968768298625946, "learning_rate": 5.7668854341108583e-08, "loss": 0.0232, "step": 39420 }, { "grad_norm": 0.04629532992839813, "learning_rate": 5.570114532040815e-08, "loss": 0.0233, "step": 39430 }, { "grad_norm": 0.04874425008893013, "learning_rate": 5.3767572757029706e-08, "loss": 0.0225, "step": 39440 }, { "grad_norm": 0.05863459035754204, "learning_rate": 5.186813797254386e-08, "loss": 0.0222, "step": 39450 }, { "grad_norm": 0.0386526994407177, "learning_rate": 5.000284226521213e-08, "loss": 0.0216, "step": 39460 }, { "grad_norm": 0.034621238708496094, "learning_rate": 4.8171686909931346e-08, "loss": 0.0216, "step": 39470 }, { "grad_norm": 0.04949817433953285, "learning_rate": 4.637467315828925e-08, "loss": 0.0226, "step": 39480 }, { "grad_norm": 0.04308539256453514, "learning_rate": 4.461180223852002e-08, "loss": 0.0222, "step": 39490 }, { "grad_norm": 0.04854131117463112, "learning_rate": 4.288307535553204e-08, "loss": 0.0257, "step": 39500 }, { "grad_norm": 0.03499859943985939, "learning_rate": 4.1188493690896836e-08, "loss": 0.0236, "step": 39510 }, { "grad_norm": 0.03653721883893013, "learning_rate": 3.952805840284346e-08, "loss": 0.0245, "step": 39520 }, { "grad_norm": 0.058325886726379395, "learning_rate": 3.790177062625855e-08, "loss": 0.024, "step": 39530 }, { "grad_norm": 0.03893456235527992, "learning_rate": 3.6309631472702946e-08, "loss": 0.0238, "step": 39540 }, { "grad_norm": 0.03651663288474083, "learning_rate": 3.475164203038395e-08, "loss": 0.0238, "step": 39550 }, { "grad_norm": 0.04294687882065773, "learning_rate": 3.3227803364171974e-08, "loss": 0.0204, "step": 39560 }, { "grad_norm": 0.039896782487630844, "learning_rate": 3.173811651560055e-08, "loss": 0.0196, "step": 39570 }, { "grad_norm": 0.04002702608704567, "learning_rate": 3.0282582502855204e-08, "loss": 0.0203, "step": 39580 }, { "grad_norm": 0.028753502294421196, "learning_rate": 2.8861202320773496e-08, "loss": 0.0185, "step": 39590 }, { "grad_norm": 0.039750706404447556, "learning_rate": 2.7473976940867175e-08, "loss": 0.0195, "step": 39600 }, { "grad_norm": 0.03558996319770813, "learning_rate": 2.6120907311277808e-08, "loss": 0.0179, "step": 39610 }, { "grad_norm": 0.03520963340997696, "learning_rate": 2.480199435682673e-08, "loss": 0.0193, "step": 39620 }, { "grad_norm": 0.04793660342693329, "learning_rate": 2.3517238978976174e-08, "loss": 0.0214, "step": 39630 }, { "grad_norm": 0.06966735422611237, "learning_rate": 2.2266642055834842e-08, "loss": 0.0222, "step": 39640 }, { "grad_norm": 0.056878864765167236, "learning_rate": 2.105020444219119e-08, "loss": 0.0216, "step": 39650 }, { "grad_norm": 0.041569121181964874, "learning_rate": 1.9867926969452388e-08, "loss": 0.0229, "step": 39660 }, { "grad_norm": 0.038974739611148834, "learning_rate": 1.8719810445705366e-08, "loss": 0.0226, "step": 39670 }, { "grad_norm": 0.03264053910970688, "learning_rate": 1.760585565567241e-08, "loss": 0.0218, "step": 39680 }, { "grad_norm": 0.04608098790049553, "learning_rate": 1.6526063360727816e-08, "loss": 0.0229, "step": 39690 }, { "grad_norm": 0.05378098785877228, "learning_rate": 1.5480434298908997e-08, "loss": 0.0205, "step": 39700 }, { "grad_norm": 0.03352312743663788, "learning_rate": 1.4468969184883164e-08, "loss": 0.0182, "step": 39710 }, { "grad_norm": 0.04764281585812569, "learning_rate": 1.3491668709980643e-08, "loss": 0.02, "step": 39720 }, { "grad_norm": 0.0523667074739933, "learning_rate": 1.254853354217822e-08, "loss": 0.0189, "step": 39730 }, { "grad_norm": 0.03060653805732727, "learning_rate": 1.1639564326099139e-08, "loss": 0.0182, "step": 39740 }, { "grad_norm": 0.05435839667916298, "learning_rate": 1.0764761683018653e-08, "loss": 0.0213, "step": 39750 }, { "grad_norm": 0.06521041691303253, "learning_rate": 9.924126210852925e-09, "loss": 0.0205, "step": 39760 }, { "grad_norm": 0.08527195453643799, "learning_rate": 9.117658484170122e-09, "loss": 0.0229, "step": 39770 }, { "grad_norm": 0.03584129735827446, "learning_rate": 8.345359054179324e-09, "loss": 0.0239, "step": 39780 }, { "grad_norm": 0.04345797374844551, "learning_rate": 7.60722844874162e-09, "loss": 0.0246, "step": 39790 }, { "grad_norm": 0.04982290044426918, "learning_rate": 6.903267172364558e-09, "loss": 0.0248, "step": 39800 }, { "grad_norm": 0.05088656395673752, "learning_rate": 6.23347570619659e-09, "loss": 0.0238, "step": 39810 }, { "grad_norm": 0.036908525973558426, "learning_rate": 5.59785450803818e-09, "loss": 0.0225, "step": 39820 }, { "grad_norm": 0.04228844493627548, "learning_rate": 4.996404012325151e-09, "loss": 0.0212, "step": 39830 }, { "grad_norm": 0.04956022650003433, "learning_rate": 4.429124630145331e-09, "loss": 0.02, "step": 39840 }, { "grad_norm": 0.03603565692901611, "learning_rate": 3.89601674922746e-09, "loss": 0.02, "step": 39850 }, { "grad_norm": 0.028023896738886833, "learning_rate": 3.3970807339467336e-09, "loss": 0.0189, "step": 39860 }, { "grad_norm": 0.0394427552819252, "learning_rate": 2.9323169253248073e-09, "loss": 0.0203, "step": 39870 }, { "grad_norm": 0.048035118728876114, "learning_rate": 2.501725641018693e-09, "loss": 0.0218, "step": 39880 }, { "grad_norm": 0.06611286848783493, "learning_rate": 2.105307175331861e-09, "loss": 0.0225, "step": 39890 }, { "grad_norm": 0.05107645317912102, "learning_rate": 1.7430617992197918e-09, "loss": 0.0236, "step": 39900 }, { "grad_norm": 0.0615951232612133, "learning_rate": 1.4149897602677708e-09, "loss": 0.0239, "step": 39910 }, { "grad_norm": 0.040727682411670685, "learning_rate": 1.1210912827130937e-09, "loss": 0.0234, "step": 39920 }, { "grad_norm": 0.03947996348142624, "learning_rate": 8.613665674284121e-10, "loss": 0.0264, "step": 39930 }, { "grad_norm": 0.04128251597285271, "learning_rate": 6.358157919383878e-10, "loss": 0.0238, "step": 39940 }, { "grad_norm": 0.044754333794116974, "learning_rate": 4.4443911040303877e-10, "loss": 0.0266, "step": 39950 }, { "grad_norm": 0.040919229388237, "learning_rate": 2.8723665362329066e-10, "loss": 0.023, "step": 39960 }, { "grad_norm": 0.03778012841939926, "learning_rate": 1.6420852905207895e-10, "loss": 0.024, "step": 39970 }, { "grad_norm": 0.02815219946205616, "learning_rate": 7.535482076659328e-11, "loss": 0.0214, "step": 39980 }, { "grad_norm": 0.04162062332034111, "learning_rate": 2.0675589512686443e-11, "loss": 0.0237, "step": 39990 }, { "grad_norm": 0.0247944388538599, "learning_rate": 1.7087265491433358e-13, "loss": 0.0232, "step": 40000 } ], "logging_steps": 10, "max_steps": 40000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 5000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 64, "trial_name": null, "trial_params": null }