{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "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": 0.1885499805212021, "learning_rate": 4.5e-07, "loss": 0.0206, "step": 10 }, { "grad_norm": 0.09837914258241653, "learning_rate": 9.5e-07, "loss": 0.0186, "step": 20 }, { "grad_norm": 0.1721290946006775, "learning_rate": 1.45e-06, "loss": 0.0201, "step": 30 }, { "grad_norm": 0.36248117685317993, "learning_rate": 1.95e-06, "loss": 0.0186, "step": 40 }, { "grad_norm": 0.10312077403068542, "learning_rate": 2.4500000000000003e-06, "loss": 0.02, "step": 50 }, { "grad_norm": 0.09769488126039505, "learning_rate": 2.95e-06, "loss": 0.0202, "step": 60 }, { "grad_norm": 0.10389938950538635, "learning_rate": 3.4500000000000004e-06, "loss": 0.0145, "step": 70 }, { "grad_norm": 0.12934468686580658, "learning_rate": 3.95e-06, "loss": 0.0193, "step": 80 }, { "grad_norm": 0.06563771516084671, "learning_rate": 4.45e-06, "loss": 0.0156, "step": 90 }, { "grad_norm": 0.07785916328430176, "learning_rate": 4.950000000000001e-06, "loss": 0.019, "step": 100 }, { "grad_norm": 0.09274876117706299, "learning_rate": 5.45e-06, "loss": 0.0188, "step": 110 }, { "grad_norm": 0.089022196829319, "learning_rate": 5.95e-06, "loss": 0.0177, "step": 120 }, { "grad_norm": 0.0954386293888092, "learning_rate": 6.45e-06, "loss": 0.0197, "step": 130 }, { "grad_norm": 0.06660106778144836, "learning_rate": 6.950000000000001e-06, "loss": 0.0193, "step": 140 }, { "grad_norm": 0.1243562176823616, "learning_rate": 7.45e-06, "loss": 0.0187, "step": 150 }, { "grad_norm": 0.0867447480559349, "learning_rate": 7.95e-06, "loss": 0.0186, "step": 160 }, { "grad_norm": 0.28182703256607056, "learning_rate": 8.45e-06, "loss": 0.0181, "step": 170 }, { "grad_norm": 0.07885300368070602, "learning_rate": 8.95e-06, "loss": 0.0175, "step": 180 }, { "grad_norm": 0.13092735409736633, "learning_rate": 9.450000000000001e-06, "loss": 0.016, "step": 190 }, { "grad_norm": 0.06927704811096191, "learning_rate": 9.950000000000001e-06, "loss": 0.0196, "step": 200 }, { "grad_norm": 0.17847253382205963, "learning_rate": 1.045e-05, "loss": 0.0191, "step": 210 }, { "grad_norm": 0.06578907370567322, "learning_rate": 1.095e-05, "loss": 0.0208, "step": 220 }, { "grad_norm": 0.08297134935855865, "learning_rate": 1.145e-05, "loss": 0.017, "step": 230 }, { "grad_norm": 0.2887559235095978, "learning_rate": 1.195e-05, "loss": 0.0184, "step": 240 }, { "grad_norm": 0.08449721336364746, "learning_rate": 1.2450000000000001e-05, "loss": 0.0176, "step": 250 }, { "grad_norm": 0.06712896376848221, "learning_rate": 1.2950000000000001e-05, "loss": 0.0175, "step": 260 }, { "grad_norm": 0.1094590499997139, "learning_rate": 1.3450000000000002e-05, "loss": 0.0235, "step": 270 }, { "grad_norm": 0.07813195139169693, "learning_rate": 1.3950000000000002e-05, "loss": 0.0185, "step": 280 }, { "grad_norm": 0.13796044886112213, "learning_rate": 1.4449999999999999e-05, "loss": 0.0229, "step": 290 }, { "grad_norm": 0.09118600189685822, "learning_rate": 1.4950000000000001e-05, "loss": 0.0185, "step": 300 }, { "grad_norm": 0.09808818250894547, "learning_rate": 1.545e-05, "loss": 0.0235, "step": 310 }, { "grad_norm": 0.15123480558395386, "learning_rate": 1.595e-05, "loss": 0.019, "step": 320 }, { "grad_norm": 0.13311117887496948, "learning_rate": 1.645e-05, "loss": 0.0194, "step": 330 }, { "grad_norm": 0.07237543165683746, "learning_rate": 1.6950000000000002e-05, "loss": 0.0179, "step": 340 }, { "grad_norm": 0.09195513278245926, "learning_rate": 1.745e-05, "loss": 0.0167, "step": 350 }, { "grad_norm": 0.11104816198348999, "learning_rate": 1.795e-05, "loss": 0.0191, "step": 360 }, { "grad_norm": 0.13291384279727936, "learning_rate": 1.845e-05, "loss": 0.0212, "step": 370 }, { "grad_norm": 0.15102817118167877, "learning_rate": 1.895e-05, "loss": 0.0185, "step": 380 }, { "grad_norm": 0.1660257875919342, "learning_rate": 1.9450000000000002e-05, "loss": 0.0194, "step": 390 }, { "grad_norm": 0.21847118437290192, "learning_rate": 1.995e-05, "loss": 0.0187, "step": 400 }, { "grad_norm": 0.09676201641559601, "learning_rate": 2.045e-05, "loss": 0.0217, "step": 410 }, { "grad_norm": 0.06816351413726807, "learning_rate": 2.095e-05, "loss": 0.0175, "step": 420 }, { "grad_norm": 0.08778628706932068, "learning_rate": 2.145e-05, "loss": 0.022, "step": 430 }, { "grad_norm": 0.11069078743457794, "learning_rate": 2.195e-05, "loss": 0.0206, "step": 440 }, { "grad_norm": 0.08844491094350815, "learning_rate": 2.245e-05, "loss": 0.0195, "step": 450 }, { "grad_norm": 0.11290254443883896, "learning_rate": 2.2950000000000002e-05, "loss": 0.0211, "step": 460 }, { "grad_norm": 0.11132366955280304, "learning_rate": 2.345e-05, "loss": 0.0233, "step": 470 }, { "grad_norm": 0.1867755800485611, "learning_rate": 2.395e-05, "loss": 0.0213, "step": 480 }, { "grad_norm": 0.133270263671875, "learning_rate": 2.445e-05, "loss": 0.0209, "step": 490 }, { "grad_norm": 0.21597158908843994, "learning_rate": 2.495e-05, "loss": 0.0193, "step": 500 }, { "grad_norm": 0.15026812255382538, "learning_rate": 2.5450000000000002e-05, "loss": 0.022, "step": 510 }, { "grad_norm": 0.12339165061712265, "learning_rate": 2.595e-05, "loss": 0.0183, "step": 520 }, { "grad_norm": 0.11109107732772827, "learning_rate": 2.6450000000000003e-05, "loss": 0.0206, "step": 530 }, { "grad_norm": 0.12221237272024155, "learning_rate": 2.6950000000000005e-05, "loss": 0.0186, "step": 540 }, { "grad_norm": 0.08705700188875198, "learning_rate": 2.7450000000000003e-05, "loss": 0.0195, "step": 550 }, { "grad_norm": 0.11547864973545074, "learning_rate": 2.7950000000000005e-05, "loss": 0.0188, "step": 560 }, { "grad_norm": 0.1908123940229416, "learning_rate": 2.845e-05, "loss": 0.0182, "step": 570 }, { "grad_norm": 0.11074212938547134, "learning_rate": 2.895e-05, "loss": 0.0192, "step": 580 }, { "grad_norm": 0.12168547511100769, "learning_rate": 2.945e-05, "loss": 0.0201, "step": 590 }, { "grad_norm": 0.11315066367387772, "learning_rate": 2.995e-05, "loss": 0.0229, "step": 600 }, { "grad_norm": 0.09272942692041397, "learning_rate": 3.045e-05, "loss": 0.0263, "step": 610 }, { "grad_norm": 0.14570583403110504, "learning_rate": 3.095e-05, "loss": 0.0202, "step": 620 }, { "grad_norm": 0.14410565793514252, "learning_rate": 3.145e-05, "loss": 0.0173, "step": 630 }, { "grad_norm": 0.11623449623584747, "learning_rate": 3.1950000000000004e-05, "loss": 0.0198, "step": 640 }, { "grad_norm": 0.10676753520965576, "learning_rate": 3.245e-05, "loss": 0.0223, "step": 650 }, { "grad_norm": 0.2214989960193634, "learning_rate": 3.295e-05, "loss": 0.0222, "step": 660 }, { "grad_norm": 0.15799355506896973, "learning_rate": 3.345000000000001e-05, "loss": 0.0251, "step": 670 }, { "grad_norm": 0.18126030266284943, "learning_rate": 3.3950000000000005e-05, "loss": 0.0242, "step": 680 }, { "grad_norm": 0.10932459682226181, "learning_rate": 3.445e-05, "loss": 0.0214, "step": 690 }, { "grad_norm": 0.20764277875423431, "learning_rate": 3.495e-05, "loss": 0.0222, "step": 700 }, { "grad_norm": 0.1682557314634323, "learning_rate": 3.545e-05, "loss": 0.0278, "step": 710 }, { "grad_norm": 0.16001200675964355, "learning_rate": 3.595e-05, "loss": 0.0219, "step": 720 }, { "grad_norm": 0.18086493015289307, "learning_rate": 3.645e-05, "loss": 0.0241, "step": 730 }, { "grad_norm": 0.11089817434549332, "learning_rate": 3.6950000000000004e-05, "loss": 0.018, "step": 740 }, { "grad_norm": 0.09976736456155777, "learning_rate": 3.745e-05, "loss": 0.0204, "step": 750 }, { "grad_norm": 0.13702629506587982, "learning_rate": 3.795e-05, "loss": 0.0187, "step": 760 }, { "grad_norm": 0.09851135313510895, "learning_rate": 3.845e-05, "loss": 0.0194, "step": 770 }, { "grad_norm": 0.13201947510242462, "learning_rate": 3.8950000000000005e-05, "loss": 0.019, "step": 780 }, { "grad_norm": 0.11131647229194641, "learning_rate": 3.9450000000000003e-05, "loss": 0.0181, "step": 790 }, { "grad_norm": 0.12948155403137207, "learning_rate": 3.995e-05, "loss": 0.0206, "step": 800 }, { "grad_norm": 0.10281801223754883, "learning_rate": 4.045000000000001e-05, "loss": 0.0218, "step": 810 }, { "grad_norm": 0.12978217005729675, "learning_rate": 4.095e-05, "loss": 0.022, "step": 820 }, { "grad_norm": 0.10837515443563461, "learning_rate": 4.145e-05, "loss": 0.0215, "step": 830 }, { "grad_norm": 0.16638995707035065, "learning_rate": 4.195e-05, "loss": 0.02, "step": 840 }, { "grad_norm": 0.09348142892122269, "learning_rate": 4.245e-05, "loss": 0.0252, "step": 850 }, { "grad_norm": 0.18574421107769012, "learning_rate": 4.295e-05, "loss": 0.0207, "step": 860 }, { "grad_norm": 0.15523070096969604, "learning_rate": 4.345e-05, "loss": 0.0222, "step": 870 }, { "grad_norm": 0.12166597694158554, "learning_rate": 4.3950000000000004e-05, "loss": 0.0211, "step": 880 }, { "grad_norm": 0.10775692760944366, "learning_rate": 4.445e-05, "loss": 0.0196, "step": 890 }, { "grad_norm": 0.16193540394306183, "learning_rate": 4.495e-05, "loss": 0.0241, "step": 900 }, { "grad_norm": 0.10164421796798706, "learning_rate": 4.545000000000001e-05, "loss": 0.0182, "step": 910 }, { "grad_norm": 0.15363559126853943, "learning_rate": 4.5950000000000006e-05, "loss": 0.0221, "step": 920 }, { "grad_norm": 0.1255577802658081, "learning_rate": 4.6450000000000004e-05, "loss": 0.0203, "step": 930 }, { "grad_norm": 0.0739310160279274, "learning_rate": 4.695e-05, "loss": 0.0188, "step": 940 }, { "grad_norm": 0.14622513949871063, "learning_rate": 4.745e-05, "loss": 0.0255, "step": 950 }, { "grad_norm": 0.1299668848514557, "learning_rate": 4.795e-05, "loss": 0.0182, "step": 960 }, { "grad_norm": 0.13341939449310303, "learning_rate": 4.845e-05, "loss": 0.0213, "step": 970 }, { "grad_norm": 0.10861328989267349, "learning_rate": 4.8950000000000004e-05, "loss": 0.0213, "step": 980 }, { "grad_norm": 0.09019789844751358, "learning_rate": 4.945e-05, "loss": 0.0236, "step": 990 }, { "grad_norm": 0.22602702677249908, "learning_rate": 4.995e-05, "loss": 0.0218, "step": 1000 }, { "grad_norm": 0.1516638547182083, "learning_rate": 5.045e-05, "loss": 0.0203, "step": 1010 }, { "grad_norm": 0.11776138097047806, "learning_rate": 5.095e-05, "loss": 0.0214, "step": 1020 }, { "grad_norm": 0.12533245980739594, "learning_rate": 5.145e-05, "loss": 0.0212, "step": 1030 }, { "grad_norm": 0.0955110490322113, "learning_rate": 5.1949999999999996e-05, "loss": 0.0239, "step": 1040 }, { "grad_norm": 0.1444927603006363, "learning_rate": 5.245e-05, "loss": 0.0205, "step": 1050 }, { "grad_norm": 0.09567553550004959, "learning_rate": 5.295e-05, "loss": 0.0264, "step": 1060 }, { "grad_norm": 0.1365964412689209, "learning_rate": 5.345e-05, "loss": 0.0244, "step": 1070 }, { "grad_norm": 0.10546471923589706, "learning_rate": 5.3950000000000004e-05, "loss": 0.0256, "step": 1080 }, { "grad_norm": 0.13762037456035614, "learning_rate": 5.445e-05, "loss": 0.0194, "step": 1090 }, { "grad_norm": 0.13343527913093567, "learning_rate": 5.495e-05, "loss": 0.0231, "step": 1100 }, { "grad_norm": 0.12880589067935944, "learning_rate": 5.545e-05, "loss": 0.0262, "step": 1110 }, { "grad_norm": 0.15767373144626617, "learning_rate": 5.5950000000000005e-05, "loss": 0.0259, "step": 1120 }, { "grad_norm": 0.1900559365749359, "learning_rate": 5.645e-05, "loss": 0.0296, "step": 1130 }, { "grad_norm": 0.15393251180648804, "learning_rate": 5.695e-05, "loss": 0.0262, "step": 1140 }, { "grad_norm": 0.12804080545902252, "learning_rate": 5.745e-05, "loss": 0.026, "step": 1150 }, { "grad_norm": 0.177230104804039, "learning_rate": 5.7950000000000006e-05, "loss": 0.0243, "step": 1160 }, { "grad_norm": 0.13721776008605957, "learning_rate": 5.8450000000000005e-05, "loss": 0.0252, "step": 1170 }, { "grad_norm": 0.13620221614837646, "learning_rate": 5.895e-05, "loss": 0.0248, "step": 1180 }, { "grad_norm": 0.14615535736083984, "learning_rate": 5.945000000000001e-05, "loss": 0.0254, "step": 1190 }, { "grad_norm": 0.09052155911922455, "learning_rate": 5.995000000000001e-05, "loss": 0.0271, "step": 1200 }, { "grad_norm": 0.1326308250427246, "learning_rate": 6.0450000000000006e-05, "loss": 0.0262, "step": 1210 }, { "grad_norm": 0.14334480464458466, "learning_rate": 6.0950000000000004e-05, "loss": 0.0222, "step": 1220 }, { "grad_norm": 0.1433686465024948, "learning_rate": 6.145e-05, "loss": 0.0272, "step": 1230 }, { "grad_norm": 0.15367963910102844, "learning_rate": 6.195e-05, "loss": 0.0249, "step": 1240 }, { "grad_norm": 0.1576729118824005, "learning_rate": 6.245000000000001e-05, "loss": 0.0257, "step": 1250 }, { "grad_norm": 0.11232073605060577, "learning_rate": 6.295e-05, "loss": 0.0206, "step": 1260 }, { "grad_norm": 0.12002817541360855, "learning_rate": 6.345e-05, "loss": 0.0244, "step": 1270 }, { "grad_norm": 0.1449766457080841, "learning_rate": 6.395e-05, "loss": 0.0224, "step": 1280 }, { "grad_norm": 0.09697163850069046, "learning_rate": 6.445e-05, "loss": 0.0231, "step": 1290 }, { "grad_norm": 0.1151171624660492, "learning_rate": 6.494999999999999e-05, "loss": 0.023, "step": 1300 }, { "grad_norm": 0.11100059747695923, "learning_rate": 6.545e-05, "loss": 0.0211, "step": 1310 }, { "grad_norm": 0.17483121156692505, "learning_rate": 6.595e-05, "loss": 0.0249, "step": 1320 }, { "grad_norm": 0.13029354810714722, "learning_rate": 6.645e-05, "loss": 0.022, "step": 1330 }, { "grad_norm": 0.11145833134651184, "learning_rate": 6.695e-05, "loss": 0.0238, "step": 1340 }, { "grad_norm": 0.11032816767692566, "learning_rate": 6.745e-05, "loss": 0.0217, "step": 1350 }, { "grad_norm": 0.16048936545848846, "learning_rate": 6.795e-05, "loss": 0.0225, "step": 1360 }, { "grad_norm": 0.1384831666946411, "learning_rate": 6.845e-05, "loss": 0.0181, "step": 1370 }, { "grad_norm": 0.1347118765115738, "learning_rate": 6.895000000000001e-05, "loss": 0.0217, "step": 1380 }, { "grad_norm": 0.12108056992292404, "learning_rate": 6.945000000000001e-05, "loss": 0.0216, "step": 1390 }, { "grad_norm": 0.14560900628566742, "learning_rate": 6.995e-05, "loss": 0.0254, "step": 1400 }, { "grad_norm": 0.09694275259971619, "learning_rate": 7.045e-05, "loss": 0.0208, "step": 1410 }, { "grad_norm": 0.13229303061962128, "learning_rate": 7.095e-05, "loss": 0.0277, "step": 1420 }, { "grad_norm": 0.08515934646129608, "learning_rate": 7.145e-05, "loss": 0.0232, "step": 1430 }, { "grad_norm": 0.13814705610275269, "learning_rate": 7.195e-05, "loss": 0.0232, "step": 1440 }, { "grad_norm": 0.12495117634534836, "learning_rate": 7.245000000000001e-05, "loss": 0.0252, "step": 1450 }, { "grad_norm": 0.12119432538747787, "learning_rate": 7.295000000000001e-05, "loss": 0.0237, "step": 1460 }, { "grad_norm": 0.11362919956445694, "learning_rate": 7.345000000000001e-05, "loss": 0.0257, "step": 1470 }, { "grad_norm": 0.19504235684871674, "learning_rate": 7.395000000000001e-05, "loss": 0.0266, "step": 1480 }, { "grad_norm": 0.11556289345026016, "learning_rate": 7.445000000000001e-05, "loss": 0.0228, "step": 1490 }, { "grad_norm": 0.12954798340797424, "learning_rate": 7.495e-05, "loss": 0.0298, "step": 1500 }, { "grad_norm": 0.12037914246320724, "learning_rate": 7.545e-05, "loss": 0.0243, "step": 1510 }, { "grad_norm": 0.13651199638843536, "learning_rate": 7.595e-05, "loss": 0.0263, "step": 1520 }, { "grad_norm": 0.17069222033023834, "learning_rate": 7.645e-05, "loss": 0.0263, "step": 1530 }, { "grad_norm": 0.19053858518600464, "learning_rate": 7.695e-05, "loss": 0.0292, "step": 1540 }, { "grad_norm": 0.2056700885295868, "learning_rate": 7.745e-05, "loss": 0.0279, "step": 1550 }, { "grad_norm": 0.4837661385536194, "learning_rate": 7.795e-05, "loss": 0.0261, "step": 1560 }, { "grad_norm": 0.18770962953567505, "learning_rate": 7.845e-05, "loss": 0.0265, "step": 1570 }, { "grad_norm": 0.1820734441280365, "learning_rate": 7.895000000000001e-05, "loss": 0.0255, "step": 1580 }, { "grad_norm": 0.13188393414020538, "learning_rate": 7.945e-05, "loss": 0.0279, "step": 1590 }, { "grad_norm": 0.17123691737651825, "learning_rate": 7.995e-05, "loss": 0.0291, "step": 1600 }, { "grad_norm": 0.13052400946617126, "learning_rate": 8.045e-05, "loss": 0.0241, "step": 1610 }, { "grad_norm": 0.1530381441116333, "learning_rate": 8.095e-05, "loss": 0.0226, "step": 1620 }, { "grad_norm": 0.14819002151489258, "learning_rate": 8.145e-05, "loss": 0.0262, "step": 1630 }, { "grad_norm": 0.1754792034626007, "learning_rate": 8.195e-05, "loss": 0.0284, "step": 1640 }, { "grad_norm": 0.16064347326755524, "learning_rate": 8.245e-05, "loss": 0.0311, "step": 1650 }, { "grad_norm": 0.12864834070205688, "learning_rate": 8.295000000000001e-05, "loss": 0.0234, "step": 1660 }, { "grad_norm": 0.21517178416252136, "learning_rate": 8.345000000000001e-05, "loss": 0.0281, "step": 1670 }, { "grad_norm": 0.1876203864812851, "learning_rate": 8.395000000000001e-05, "loss": 0.0301, "step": 1680 }, { "grad_norm": 0.16292165219783783, "learning_rate": 8.445e-05, "loss": 0.0287, "step": 1690 }, { "grad_norm": 0.13697674870491028, "learning_rate": 8.495e-05, "loss": 0.0269, "step": 1700 }, { "grad_norm": 0.2220987230539322, "learning_rate": 8.545e-05, "loss": 0.0294, "step": 1710 }, { "grad_norm": 0.11837190389633179, "learning_rate": 8.595e-05, "loss": 0.0308, "step": 1720 }, { "grad_norm": 0.22466851770877838, "learning_rate": 8.645000000000001e-05, "loss": 0.0309, "step": 1730 }, { "grad_norm": 0.16202165186405182, "learning_rate": 8.695000000000001e-05, "loss": 0.0259, "step": 1740 }, { "grad_norm": 0.14856663346290588, "learning_rate": 8.745000000000001e-05, "loss": 0.0275, "step": 1750 }, { "grad_norm": 0.1613365113735199, "learning_rate": 8.795e-05, "loss": 0.0298, "step": 1760 }, { "grad_norm": 0.165757954120636, "learning_rate": 8.845e-05, "loss": 0.0282, "step": 1770 }, { "grad_norm": 0.15920376777648926, "learning_rate": 8.895e-05, "loss": 0.0279, "step": 1780 }, { "grad_norm": 0.15578192472457886, "learning_rate": 8.945e-05, "loss": 0.0349, "step": 1790 }, { "grad_norm": 0.19890516996383667, "learning_rate": 8.995e-05, "loss": 0.0299, "step": 1800 }, { "grad_norm": 0.18577444553375244, "learning_rate": 9.045e-05, "loss": 0.0284, "step": 1810 }, { "grad_norm": 0.15409694612026215, "learning_rate": 9.095e-05, "loss": 0.0288, "step": 1820 }, { "grad_norm": 0.15441682934761047, "learning_rate": 9.145e-05, "loss": 0.0278, "step": 1830 }, { "grad_norm": 0.14898499846458435, "learning_rate": 9.195e-05, "loss": 0.0273, "step": 1840 }, { "grad_norm": 0.11693330109119415, "learning_rate": 9.245e-05, "loss": 0.0296, "step": 1850 }, { "grad_norm": 0.15856289863586426, "learning_rate": 9.295000000000001e-05, "loss": 0.0323, "step": 1860 }, { "grad_norm": 0.16372759640216827, "learning_rate": 9.345000000000001e-05, "loss": 0.0328, "step": 1870 }, { "grad_norm": 0.1808817982673645, "learning_rate": 9.395000000000001e-05, "loss": 0.0268, "step": 1880 }, { "grad_norm": 0.12193611264228821, "learning_rate": 9.445e-05, "loss": 0.0256, "step": 1890 }, { "grad_norm": 0.21117021143436432, "learning_rate": 9.495e-05, "loss": 0.0321, "step": 1900 }, { "grad_norm": 0.15645501017570496, "learning_rate": 9.545e-05, "loss": 0.0296, "step": 1910 }, { "grad_norm": 0.1574385017156601, "learning_rate": 9.595e-05, "loss": 0.0311, "step": 1920 }, { "grad_norm": 0.19142195582389832, "learning_rate": 9.645000000000001e-05, "loss": 0.0325, "step": 1930 }, { "grad_norm": 0.16684073209762573, "learning_rate": 9.695000000000001e-05, "loss": 0.0325, "step": 1940 }, { "grad_norm": 0.19213086366653442, "learning_rate": 9.745000000000001e-05, "loss": 0.0327, "step": 1950 }, { "grad_norm": 0.1260754019021988, "learning_rate": 9.795000000000001e-05, "loss": 0.0281, "step": 1960 }, { "grad_norm": 0.1572311967611313, "learning_rate": 9.845000000000001e-05, "loss": 0.0345, "step": 1970 }, { "grad_norm": 0.23428624868392944, "learning_rate": 9.895e-05, "loss": 0.0363, "step": 1980 }, { "grad_norm": 0.23022687435150146, "learning_rate": 9.945e-05, "loss": 0.0324, "step": 1990 }, { "grad_norm": 0.2089596539735794, "learning_rate": 9.995e-05, "loss": 0.0311, "step": 2000 }, { "grad_norm": 0.13797703385353088, "learning_rate": 9.99999861593158e-05, "loss": 0.0293, "step": 2010 }, { "grad_norm": 0.17777559161186218, "learning_rate": 9.999993831498517e-05, "loss": 0.029, "step": 2020 }, { "grad_norm": 0.14366154372692108, "learning_rate": 9.99998562961682e-05, "loss": 0.0294, "step": 2030 }, { "grad_norm": 0.17539174854755402, "learning_rate": 9.99997401029209e-05, "loss": 0.0323, "step": 2040 }, { "grad_norm": 0.14389024674892426, "learning_rate": 9.999958973532271e-05, "loss": 0.0288, "step": 2050 }, { "grad_norm": 0.17098680138587952, "learning_rate": 9.999940519347642e-05, "loss": 0.0272, "step": 2060 }, { "grad_norm": 0.2343844473361969, "learning_rate": 9.999918647750812e-05, "loss": 0.0283, "step": 2070 }, { "grad_norm": 0.18120041489601135, "learning_rate": 9.999893358756736e-05, "loss": 0.027, "step": 2080 }, { "grad_norm": 0.23328886926174164, "learning_rate": 9.999864652382692e-05, "loss": 0.0291, "step": 2090 }, { "grad_norm": 0.16120874881744385, "learning_rate": 9.999832528648307e-05, "loss": 0.03, "step": 2100 }, { "grad_norm": 0.16063009202480316, "learning_rate": 9.999796987575532e-05, "loss": 0.0274, "step": 2110 }, { "grad_norm": 0.1943843513727188, "learning_rate": 9.999758029188662e-05, "loss": 0.0284, "step": 2120 }, { "grad_norm": 0.14402833580970764, "learning_rate": 9.999715653514324e-05, "loss": 0.0301, "step": 2130 }, { "grad_norm": 0.1554497480392456, "learning_rate": 9.999669860581482e-05, "loss": 0.0304, "step": 2140 }, { "grad_norm": 0.15071263909339905, "learning_rate": 9.999620650421433e-05, "loss": 0.0329, "step": 2150 }, { "grad_norm": 0.17001089453697205, "learning_rate": 9.999568023067813e-05, "loss": 0.0282, "step": 2160 }, { "grad_norm": 0.11975158751010895, "learning_rate": 9.999511978556595e-05, "loss": 0.0249, "step": 2170 }, { "grad_norm": 0.21032403409481049, "learning_rate": 9.99945251692608e-05, "loss": 0.0334, "step": 2180 }, { "grad_norm": 0.16685017943382263, "learning_rate": 9.999389638216912e-05, "loss": 0.0306, "step": 2190 }, { "grad_norm": 0.18296371400356293, "learning_rate": 9.999323342472068e-05, "loss": 0.0279, "step": 2200 }, { "grad_norm": 0.1363038420677185, "learning_rate": 9.99925362973686e-05, "loss": 0.0268, "step": 2210 }, { "grad_norm": 0.13111764192581177, "learning_rate": 9.999180500058935e-05, "loss": 0.0304, "step": 2220 }, { "grad_norm": 0.19342756271362305, "learning_rate": 9.99910395348828e-05, "loss": 0.0316, "step": 2230 }, { "grad_norm": 0.18688222765922546, "learning_rate": 9.99902399007721e-05, "loss": 0.0291, "step": 2240 }, { "grad_norm": 0.136398583650589, "learning_rate": 9.99894060988038e-05, "loss": 0.0308, "step": 2250 }, { "grad_norm": 0.1587258279323578, "learning_rate": 9.998853812954783e-05, "loss": 0.0311, "step": 2260 }, { "grad_norm": 0.17053040862083435, "learning_rate": 9.998763599359739e-05, "loss": 0.0277, "step": 2270 }, { "grad_norm": 0.11653334647417068, "learning_rate": 9.998669969156912e-05, "loss": 0.0296, "step": 2280 }, { "grad_norm": 0.16500155627727509, "learning_rate": 9.998572922410294e-05, "loss": 0.0269, "step": 2290 }, { "grad_norm": 0.165052592754364, "learning_rate": 9.998472459186219e-05, "loss": 0.0248, "step": 2300 }, { "grad_norm": 0.1410936564207077, "learning_rate": 9.99836857955335e-05, "loss": 0.0314, "step": 2310 }, { "grad_norm": 0.13081537187099457, "learning_rate": 9.998261283582688e-05, "loss": 0.0244, "step": 2320 }, { "grad_norm": 0.12431687116622925, "learning_rate": 9.998150571347571e-05, "loss": 0.0311, "step": 2330 }, { "grad_norm": 0.11514468491077423, "learning_rate": 9.998036442923667e-05, "loss": 0.0289, "step": 2340 }, { "grad_norm": 0.15084555745124817, "learning_rate": 9.997918898388983e-05, "loss": 0.0308, "step": 2350 }, { "grad_norm": 0.13301090896129608, "learning_rate": 9.997797937823859e-05, "loss": 0.0268, "step": 2360 }, { "grad_norm": 0.1663820594549179, "learning_rate": 9.997673561310971e-05, "loss": 0.0266, "step": 2370 }, { "grad_norm": 0.18736472725868225, "learning_rate": 9.997545768935331e-05, "loss": 0.0255, "step": 2380 }, { "grad_norm": 0.17603975534439087, "learning_rate": 9.997414560784282e-05, "loss": 0.031, "step": 2390 }, { "grad_norm": 0.137308731675148, "learning_rate": 9.997279936947502e-05, "loss": 0.0298, "step": 2400 }, { "grad_norm": 0.14929406344890594, "learning_rate": 9.997141897517006e-05, "loss": 0.027, "step": 2410 }, { "grad_norm": 0.15157705545425415, "learning_rate": 9.997000442587145e-05, "loss": 0.0271, "step": 2420 }, { "grad_norm": 0.14492234587669373, "learning_rate": 9.9968555722546e-05, "loss": 0.028, "step": 2430 }, { "grad_norm": 0.13740433752536774, "learning_rate": 9.996707286618387e-05, "loss": 0.029, "step": 2440 }, { "grad_norm": 0.14397139847278595, "learning_rate": 9.996555585779861e-05, "loss": 0.0264, "step": 2450 }, { "grad_norm": 0.19313408434391022, "learning_rate": 9.996400469842708e-05, "loss": 0.0321, "step": 2460 }, { "grad_norm": 0.166542187333107, "learning_rate": 9.996241938912945e-05, "loss": 0.0273, "step": 2470 }, { "grad_norm": 0.10566988587379456, "learning_rate": 9.99607999309893e-05, "loss": 0.0272, "step": 2480 }, { "grad_norm": 0.12405253201723099, "learning_rate": 9.99591463251135e-05, "loss": 0.0243, "step": 2490 }, { "grad_norm": 0.14512327313423157, "learning_rate": 9.995745857263226e-05, "loss": 0.0297, "step": 2500 }, { "grad_norm": 0.0992981567978859, "learning_rate": 9.995573667469918e-05, "loss": 0.0271, "step": 2510 }, { "grad_norm": 0.09811347723007202, "learning_rate": 9.995398063249111e-05, "loss": 0.0265, "step": 2520 }, { "grad_norm": 0.09971985220909119, "learning_rate": 9.995219044720833e-05, "loss": 0.026, "step": 2530 }, { "grad_norm": 0.12056974321603775, "learning_rate": 9.99503661200744e-05, "loss": 0.0306, "step": 2540 }, { "grad_norm": 0.15542565286159515, "learning_rate": 9.994850765233624e-05, "loss": 0.034, "step": 2550 }, { "grad_norm": 0.1581292748451233, "learning_rate": 9.994661504526406e-05, "loss": 0.0299, "step": 2560 }, { "grad_norm": 0.13083837926387787, "learning_rate": 9.994468830015148e-05, "loss": 0.0282, "step": 2570 }, { "grad_norm": 0.11515465378761292, "learning_rate": 9.99427274183154e-05, "loss": 0.0317, "step": 2580 }, { "grad_norm": 0.11342388391494751, "learning_rate": 9.994073240109606e-05, "loss": 0.0309, "step": 2590 }, { "grad_norm": 0.13224202394485474, "learning_rate": 9.993870324985703e-05, "loss": 0.029, "step": 2600 }, { "grad_norm": 0.12023661285638809, "learning_rate": 9.993663996598523e-05, "loss": 0.0281, "step": 2610 }, { "grad_norm": 0.15378710627555847, "learning_rate": 9.993454255089089e-05, "loss": 0.0298, "step": 2620 }, { "grad_norm": 0.16792243719100952, "learning_rate": 9.993241100600757e-05, "loss": 0.0254, "step": 2630 }, { "grad_norm": 0.10433586686849594, "learning_rate": 9.993024533279215e-05, "loss": 0.027, "step": 2640 }, { "grad_norm": 0.13417769968509674, "learning_rate": 9.992804553272487e-05, "loss": 0.0303, "step": 2650 }, { "grad_norm": 0.14975854754447937, "learning_rate": 9.992581160730926e-05, "loss": 0.0357, "step": 2660 }, { "grad_norm": 0.17771016061306, "learning_rate": 9.992354355807218e-05, "loss": 0.0305, "step": 2670 }, { "grad_norm": 0.12858892977237701, "learning_rate": 9.992124138656385e-05, "loss": 0.0315, "step": 2680 }, { "grad_norm": 0.10708215832710266, "learning_rate": 9.991890509435774e-05, "loss": 0.0289, "step": 2690 }, { "grad_norm": 0.12570616602897644, "learning_rate": 9.991653468305071e-05, "loss": 0.0305, "step": 2700 }, { "grad_norm": 0.15006625652313232, "learning_rate": 9.991413015426292e-05, "loss": 0.0296, "step": 2710 }, { "grad_norm": 0.11508429050445557, "learning_rate": 9.991169150963781e-05, "loss": 0.0294, "step": 2720 }, { "grad_norm": 0.12216870486736298, "learning_rate": 9.990921875084221e-05, "loss": 0.0302, "step": 2730 }, { "grad_norm": 0.10050542652606964, "learning_rate": 9.990671187956622e-05, "loss": 0.0295, "step": 2740 }, { "grad_norm": 0.13682860136032104, "learning_rate": 9.990417089752324e-05, "loss": 0.0244, "step": 2750 }, { "grad_norm": 0.11236411333084106, "learning_rate": 9.990159580645001e-05, "loss": 0.0282, "step": 2760 }, { "grad_norm": 0.1313033550977707, "learning_rate": 9.98989866081066e-05, "loss": 0.0285, "step": 2770 }, { "grad_norm": 0.1536359041929245, "learning_rate": 9.989634330427636e-05, "loss": 0.0304, "step": 2780 }, { "grad_norm": 0.16180333495140076, "learning_rate": 9.989366589676596e-05, "loss": 0.0278, "step": 2790 }, { "grad_norm": 0.13546191155910492, "learning_rate": 9.98909543874054e-05, "loss": 0.032, "step": 2800 }, { "grad_norm": 0.15659016370773315, "learning_rate": 9.988820877804795e-05, "loss": 0.0323, "step": 2810 }, { "grad_norm": 0.14295190572738647, "learning_rate": 9.988542907057022e-05, "loss": 0.0292, "step": 2820 }, { "grad_norm": 0.13203337788581848, "learning_rate": 9.988261526687212e-05, "loss": 0.0299, "step": 2830 }, { "grad_norm": 0.12154544144868851, "learning_rate": 9.987976736887685e-05, "loss": 0.0309, "step": 2840 }, { "grad_norm": 0.11555686593055725, "learning_rate": 9.987688537853091e-05, "loss": 0.032, "step": 2850 }, { "grad_norm": 0.15349948406219482, "learning_rate": 9.987396929780413e-05, "loss": 0.0326, "step": 2860 }, { "grad_norm": 0.09485231339931488, "learning_rate": 9.987101912868962e-05, "loss": 0.0301, "step": 2870 }, { "grad_norm": 0.11194208264350891, "learning_rate": 9.98680348732038e-05, "loss": 0.0296, "step": 2880 }, { "grad_norm": 0.14671999216079712, "learning_rate": 9.986501653338636e-05, "loss": 0.0313, "step": 2890 }, { "grad_norm": 0.1576211303472519, "learning_rate": 9.986196411130031e-05, "loss": 0.033, "step": 2900 }, { "grad_norm": 0.15914031863212585, "learning_rate": 9.985887760903197e-05, "loss": 0.0404, "step": 2910 }, { "grad_norm": 0.14737670123577118, "learning_rate": 9.985575702869093e-05, "loss": 0.0331, "step": 2920 }, { "grad_norm": 0.1211332157254219, "learning_rate": 9.985260237241008e-05, "loss": 0.0311, "step": 2930 }, { "grad_norm": 0.16923218965530396, "learning_rate": 9.984941364234557e-05, "loss": 0.033, "step": 2940 }, { "grad_norm": 0.11744227260351181, "learning_rate": 9.98461908406769e-05, "loss": 0.0312, "step": 2950 }, { "grad_norm": 0.16205552220344543, "learning_rate": 9.98429339696068e-05, "loss": 0.0374, "step": 2960 }, { "grad_norm": 0.14720788598060608, "learning_rate": 9.983964303136133e-05, "loss": 0.0337, "step": 2970 }, { "grad_norm": 0.12501254677772522, "learning_rate": 9.983631802818981e-05, "loss": 0.0292, "step": 2980 }, { "grad_norm": 0.1045399084687233, "learning_rate": 9.983295896236484e-05, "loss": 0.0282, "step": 2990 }, { "grad_norm": 0.190803661942482, "learning_rate": 9.982956583618232e-05, "loss": 0.0319, "step": 3000 }, { "grad_norm": 0.17835243046283722, "learning_rate": 9.982613865196142e-05, "loss": 0.03, "step": 3010 }, { "grad_norm": 0.12460031360387802, "learning_rate": 9.982267741204458e-05, "loss": 0.031, "step": 3020 }, { "grad_norm": 0.14160959422588348, "learning_rate": 9.981918211879753e-05, "loss": 0.026, "step": 3030 }, { "grad_norm": 0.10449188202619553, "learning_rate": 9.981565277460927e-05, "loss": 0.0245, "step": 3040 }, { "grad_norm": 0.12498742341995239, "learning_rate": 9.981208938189206e-05, "loss": 0.03, "step": 3050 }, { "grad_norm": 0.11462005972862244, "learning_rate": 9.980849194308149e-05, "loss": 0.0296, "step": 3060 }, { "grad_norm": 0.15975120663642883, "learning_rate": 9.98048604606363e-05, "loss": 0.031, "step": 3070 }, { "grad_norm": 0.12959407269954681, "learning_rate": 9.980119493703864e-05, "loss": 0.0295, "step": 3080 }, { "grad_norm": 0.14264285564422607, "learning_rate": 9.979749537479383e-05, "loss": 0.0294, "step": 3090 }, { "grad_norm": 0.15511786937713623, "learning_rate": 9.979376177643051e-05, "loss": 0.029, "step": 3100 }, { "grad_norm": 0.23999838531017303, "learning_rate": 9.978999414450052e-05, "loss": 0.0326, "step": 3110 }, { "grad_norm": 0.19824568927288055, "learning_rate": 9.978619248157904e-05, "loss": 0.0278, "step": 3120 }, { "grad_norm": 0.17711690068244934, "learning_rate": 9.978235679026445e-05, "loss": 0.0308, "step": 3130 }, { "grad_norm": 0.1748959869146347, "learning_rate": 9.97784870731784e-05, "loss": 0.0286, "step": 3140 }, { "grad_norm": 0.233074352145195, "learning_rate": 9.977458333296584e-05, "loss": 0.0289, "step": 3150 }, { "grad_norm": 0.13570450246334076, "learning_rate": 9.977064557229492e-05, "loss": 0.0267, "step": 3160 }, { "grad_norm": 0.1125078871846199, "learning_rate": 9.976667379385705e-05, "loss": 0.0275, "step": 3170 }, { "grad_norm": 0.18381014466285706, "learning_rate": 9.97626680003669e-05, "loss": 0.0264, "step": 3180 }, { "grad_norm": 0.2263237088918686, "learning_rate": 9.975862819456242e-05, "loss": 0.0278, "step": 3190 }, { "grad_norm": 0.13917167484760284, "learning_rate": 9.975455437920477e-05, "loss": 0.0316, "step": 3200 }, { "grad_norm": 0.10755997896194458, "learning_rate": 9.975044655707834e-05, "loss": 0.0268, "step": 3210 }, { "grad_norm": 0.10658248513936996, "learning_rate": 9.974630473099082e-05, "loss": 0.0271, "step": 3220 }, { "grad_norm": 0.11555879563093185, "learning_rate": 9.974212890377311e-05, "loss": 0.0272, "step": 3230 }, { "grad_norm": 0.12755411863327026, "learning_rate": 9.973791907827931e-05, "loss": 0.0245, "step": 3240 }, { "grad_norm": 0.10810429602861404, "learning_rate": 9.973367525738683e-05, "loss": 0.024, "step": 3250 }, { "grad_norm": 0.1402977705001831, "learning_rate": 9.972939744399627e-05, "loss": 0.0313, "step": 3260 }, { "grad_norm": 0.12946341931819916, "learning_rate": 9.97250856410315e-05, "loss": 0.0281, "step": 3270 }, { "grad_norm": 0.11340548098087311, "learning_rate": 9.972073985143955e-05, "loss": 0.0254, "step": 3280 }, { "grad_norm": 0.09556949138641357, "learning_rate": 9.971636007819074e-05, "loss": 0.028, "step": 3290 }, { "grad_norm": 0.13056907057762146, "learning_rate": 9.971194632427863e-05, "loss": 0.0245, "step": 3300 }, { "grad_norm": 0.13441243767738342, "learning_rate": 9.970749859271997e-05, "loss": 0.0261, "step": 3310 }, { "grad_norm": 0.13548193871974945, "learning_rate": 9.970301688655473e-05, "loss": 0.0267, "step": 3320 }, { "grad_norm": 0.15582561492919922, "learning_rate": 9.969850120884612e-05, "loss": 0.0291, "step": 3330 }, { "grad_norm": 0.1123451516032219, "learning_rate": 9.969395156268055e-05, "loss": 0.0267, "step": 3340 }, { "grad_norm": 0.09971486777067184, "learning_rate": 9.968936795116768e-05, "loss": 0.0285, "step": 3350 }, { "grad_norm": 0.11994611471891403, "learning_rate": 9.968475037744036e-05, "loss": 0.0262, "step": 3360 }, { "grad_norm": 0.14305196702480316, "learning_rate": 9.968009884465465e-05, "loss": 0.0332, "step": 3370 }, { "grad_norm": 0.13002263009548187, "learning_rate": 9.967541335598984e-05, "loss": 0.0309, "step": 3380 }, { "grad_norm": 0.13115650415420532, "learning_rate": 9.967069391464841e-05, "loss": 0.0311, "step": 3390 }, { "grad_norm": 0.13601168990135193, "learning_rate": 9.966594052385608e-05, "loss": 0.0322, "step": 3400 }, { "grad_norm": 0.1426674723625183, "learning_rate": 9.96611531868617e-05, "loss": 0.0293, "step": 3410 }, { "grad_norm": 0.1433570384979248, "learning_rate": 9.96563319069374e-05, "loss": 0.0295, "step": 3420 }, { "grad_norm": 0.16909578442573547, "learning_rate": 9.965147668737847e-05, "loss": 0.0264, "step": 3430 }, { "grad_norm": 0.1596388965845108, "learning_rate": 9.96465875315034e-05, "loss": 0.028, "step": 3440 }, { "grad_norm": 0.09162052720785141, "learning_rate": 9.96416644426539e-05, "loss": 0.0264, "step": 3450 }, { "grad_norm": 0.11600782722234726, "learning_rate": 9.963670742419485e-05, "loss": 0.0318, "step": 3460 }, { "grad_norm": 0.10963662713766098, "learning_rate": 9.96317164795143e-05, "loss": 0.0236, "step": 3470 }, { "grad_norm": 0.14214952290058136, "learning_rate": 9.962669161202356e-05, "loss": 0.0274, "step": 3480 }, { "grad_norm": 0.12937308847904205, "learning_rate": 9.962163282515705e-05, "loss": 0.029, "step": 3490 }, { "grad_norm": 0.10132495313882828, "learning_rate": 9.961654012237241e-05, "loss": 0.0354, "step": 3500 }, { "grad_norm": 0.12611524760723114, "learning_rate": 9.961141350715044e-05, "loss": 0.0311, "step": 3510 }, { "grad_norm": 0.14770229160785675, "learning_rate": 9.960625298299516e-05, "loss": 0.0289, "step": 3520 }, { "grad_norm": 0.11418575048446655, "learning_rate": 9.960105855343372e-05, "loss": 0.03, "step": 3530 }, { "grad_norm": 0.1519731730222702, "learning_rate": 9.959583022201647e-05, "loss": 0.0343, "step": 3540 }, { "grad_norm": 0.14620915055274963, "learning_rate": 9.959056799231692e-05, "loss": 0.0292, "step": 3550 }, { "grad_norm": 0.24256227910518646, "learning_rate": 9.958527186793176e-05, "loss": 0.0314, "step": 3560 }, { "grad_norm": 0.15409737825393677, "learning_rate": 9.957994185248086e-05, "loss": 0.0336, "step": 3570 }, { "grad_norm": 0.11397570371627808, "learning_rate": 9.957457794960718e-05, "loss": 0.0339, "step": 3580 }, { "grad_norm": 0.09608007967472076, "learning_rate": 9.956918016297694e-05, "loss": 0.0298, "step": 3590 }, { "grad_norm": 0.10184404253959656, "learning_rate": 9.956374849627948e-05, "loss": 0.0339, "step": 3600 }, { "grad_norm": 0.10677680373191833, "learning_rate": 9.955828295322728e-05, "loss": 0.0284, "step": 3610 }, { "grad_norm": 0.1382754147052765, "learning_rate": 9.955278353755598e-05, "loss": 0.0321, "step": 3620 }, { "grad_norm": 0.12273035198450089, "learning_rate": 9.954725025302439e-05, "loss": 0.0305, "step": 3630 }, { "grad_norm": 0.1538609117269516, "learning_rate": 9.954168310341445e-05, "loss": 0.0319, "step": 3640 }, { "grad_norm": 0.14683300256729126, "learning_rate": 9.953608209253126e-05, "loss": 0.0309, "step": 3650 }, { "grad_norm": 0.12812575697898865, "learning_rate": 9.953044722420307e-05, "loss": 0.0283, "step": 3660 }, { "grad_norm": 0.11907374113798141, "learning_rate": 9.952477850228124e-05, "loss": 0.032, "step": 3670 }, { "grad_norm": 0.12994952499866486, "learning_rate": 9.95190759306403e-05, "loss": 0.0356, "step": 3680 }, { "grad_norm": 0.14152996242046356, "learning_rate": 9.951333951317789e-05, "loss": 0.0345, "step": 3690 }, { "grad_norm": 0.12467688322067261, "learning_rate": 9.950756925381479e-05, "loss": 0.0304, "step": 3700 }, { "grad_norm": 0.14151793718338013, "learning_rate": 9.950176515649496e-05, "loss": 0.0354, "step": 3710 }, { "grad_norm": 0.15384916961193085, "learning_rate": 9.949592722518542e-05, "loss": 0.0333, "step": 3720 }, { "grad_norm": 0.17614218592643738, "learning_rate": 9.949005546387631e-05, "loss": 0.0307, "step": 3730 }, { "grad_norm": 0.12679627537727356, "learning_rate": 9.948414987658098e-05, "loss": 0.0357, "step": 3740 }, { "grad_norm": 0.14538785815238953, "learning_rate": 9.94782104673358e-05, "loss": 0.0342, "step": 3750 }, { "grad_norm": 0.08899973332881927, "learning_rate": 9.947223724020034e-05, "loss": 0.0311, "step": 3760 }, { "grad_norm": 0.1479676216840744, "learning_rate": 9.94662301992572e-05, "loss": 0.0349, "step": 3770 }, { "grad_norm": 0.12250625342130661, "learning_rate": 9.946018934861216e-05, "loss": 0.0295, "step": 3780 }, { "grad_norm": 0.17594927549362183, "learning_rate": 9.945411469239409e-05, "loss": 0.0297, "step": 3790 }, { "grad_norm": 0.16756542026996613, "learning_rate": 9.944800623475497e-05, "loss": 0.0345, "step": 3800 }, { "grad_norm": 0.15883779525756836, "learning_rate": 9.944186397986984e-05, "loss": 0.0326, "step": 3810 }, { "grad_norm": 0.14487317204475403, "learning_rate": 9.943568793193687e-05, "loss": 0.0347, "step": 3820 }, { "grad_norm": 0.13151194155216217, "learning_rate": 9.942947809517737e-05, "loss": 0.0348, "step": 3830 }, { "grad_norm": 0.14797313511371613, "learning_rate": 9.942323447383568e-05, "loss": 0.0314, "step": 3840 }, { "grad_norm": 0.17644190788269043, "learning_rate": 9.941695707217925e-05, "loss": 0.0313, "step": 3850 }, { "grad_norm": 0.10638125240802765, "learning_rate": 9.941064589449865e-05, "loss": 0.0326, "step": 3860 }, { "grad_norm": 0.22868837416172028, "learning_rate": 9.940430094510748e-05, "loss": 0.0334, "step": 3870 }, { "grad_norm": 0.12909436225891113, "learning_rate": 9.939792222834248e-05, "loss": 0.0318, "step": 3880 }, { "grad_norm": 0.10749245434999466, "learning_rate": 9.939150974856343e-05, "loss": 0.03, "step": 3890 }, { "grad_norm": 0.11890200525522232, "learning_rate": 9.93850635101532e-05, "loss": 0.0323, "step": 3900 }, { "grad_norm": 0.11279048025608063, "learning_rate": 9.937858351751775e-05, "loss": 0.0302, "step": 3910 }, { "grad_norm": 0.13113771378993988, "learning_rate": 9.937206977508604e-05, "loss": 0.0304, "step": 3920 }, { "grad_norm": 0.09151435643434525, "learning_rate": 9.936552228731022e-05, "loss": 0.0341, "step": 3930 }, { "grad_norm": 0.13711105287075043, "learning_rate": 9.93589410586654e-05, "loss": 0.0332, "step": 3940 }, { "grad_norm": 0.1559959501028061, "learning_rate": 9.93523260936498e-05, "loss": 0.033, "step": 3950 }, { "grad_norm": 0.11861798912286758, "learning_rate": 9.934567739678467e-05, "loss": 0.0293, "step": 3960 }, { "grad_norm": 0.1294163018465042, "learning_rate": 9.933899497261433e-05, "loss": 0.0336, "step": 3970 }, { "grad_norm": 0.13848619163036346, "learning_rate": 9.933227882570617e-05, "loss": 0.0306, "step": 3980 }, { "grad_norm": 0.15885598957538605, "learning_rate": 9.932552896065062e-05, "loss": 0.0291, "step": 3990 }, { "grad_norm": 0.1156715527176857, "learning_rate": 9.931874538206114e-05, "loss": 0.0309, "step": 4000 }, { "grad_norm": 0.12575118243694305, "learning_rate": 9.931192809457423e-05, "loss": 0.0271, "step": 4010 }, { "grad_norm": 0.13445301353931427, "learning_rate": 9.930507710284944e-05, "loss": 0.0287, "step": 4020 }, { "grad_norm": 0.13824178278446198, "learning_rate": 9.92981924115694e-05, "loss": 0.0305, "step": 4030 }, { "grad_norm": 0.15755516290664673, "learning_rate": 9.929127402543968e-05, "loss": 0.0308, "step": 4040 }, { "grad_norm": 0.18650053441524506, "learning_rate": 9.928432194918895e-05, "loss": 0.0298, "step": 4050 }, { "grad_norm": 0.13864253461360931, "learning_rate": 9.927733618756889e-05, "loss": 0.0352, "step": 4060 }, { "grad_norm": 0.211593359708786, "learning_rate": 9.92703167453542e-05, "loss": 0.033, "step": 4070 }, { "grad_norm": 0.1650342345237732, "learning_rate": 9.926326362734263e-05, "loss": 0.0271, "step": 4080 }, { "grad_norm": 0.1500575840473175, "learning_rate": 9.925617683835489e-05, "loss": 0.0286, "step": 4090 }, { "grad_norm": 0.11649201065301895, "learning_rate": 9.924905638323472e-05, "loss": 0.032, "step": 4100 }, { "grad_norm": 0.11799047887325287, "learning_rate": 9.92419022668489e-05, "loss": 0.0314, "step": 4110 }, { "grad_norm": 0.10159271955490112, "learning_rate": 9.923471449408723e-05, "loss": 0.0294, "step": 4120 }, { "grad_norm": 0.16360896825790405, "learning_rate": 9.922749306986244e-05, "loss": 0.0309, "step": 4130 }, { "grad_norm": 0.15599428117275238, "learning_rate": 9.922023799911033e-05, "loss": 0.0308, "step": 4140 }, { "grad_norm": 0.15360499918460846, "learning_rate": 9.921294928678968e-05, "loss": 0.028, "step": 4150 }, { "grad_norm": 0.1481134295463562, "learning_rate": 9.920562693788222e-05, "loss": 0.0238, "step": 4160 }, { "grad_norm": 0.16168344020843506, "learning_rate": 9.919827095739273e-05, "loss": 0.0307, "step": 4170 }, { "grad_norm": 0.13384684920310974, "learning_rate": 9.919088135034898e-05, "loss": 0.0311, "step": 4180 }, { "grad_norm": 0.152249276638031, "learning_rate": 9.918345812180165e-05, "loss": 0.0305, "step": 4190 }, { "grad_norm": 0.16288235783576965, "learning_rate": 9.917600127682446e-05, "loss": 0.0291, "step": 4200 }, { "grad_norm": 0.18924452364444733, "learning_rate": 9.91685108205141e-05, "loss": 0.0273, "step": 4210 }, { "grad_norm": 0.16905705630779266, "learning_rate": 9.916098675799024e-05, "loss": 0.0288, "step": 4220 }, { "grad_norm": 0.1268925964832306, "learning_rate": 9.915342909439548e-05, "loss": 0.0282, "step": 4230 }, { "grad_norm": 0.16059818863868713, "learning_rate": 9.914583783489543e-05, "loss": 0.031, "step": 4240 }, { "grad_norm": 0.14194944500923157, "learning_rate": 9.913821298467863e-05, "loss": 0.0319, "step": 4250 }, { "grad_norm": 0.11610886454582214, "learning_rate": 9.913055454895661e-05, "loss": 0.0327, "step": 4260 }, { "grad_norm": 0.1929488182067871, "learning_rate": 9.912286253296381e-05, "loss": 0.0295, "step": 4270 }, { "grad_norm": 0.1587950438261032, "learning_rate": 9.911513694195768e-05, "loss": 0.0305, "step": 4280 }, { "grad_norm": 0.1223016306757927, "learning_rate": 9.910737778121859e-05, "loss": 0.0292, "step": 4290 }, { "grad_norm": 0.13510183990001678, "learning_rate": 9.909958505604984e-05, "loss": 0.0303, "step": 4300 }, { "grad_norm": 0.15342111885547638, "learning_rate": 9.909175877177767e-05, "loss": 0.0291, "step": 4310 }, { "grad_norm": 0.15537257492542267, "learning_rate": 9.908389893375129e-05, "loss": 0.0283, "step": 4320 }, { "grad_norm": 0.09419466555118561, "learning_rate": 9.907600554734283e-05, "loss": 0.025, "step": 4330 }, { "grad_norm": 0.0995689183473587, "learning_rate": 9.906807861794734e-05, "loss": 0.0289, "step": 4340 }, { "grad_norm": 0.13394427299499512, "learning_rate": 9.906011815098279e-05, "loss": 0.0316, "step": 4350 }, { "grad_norm": 0.19666136801242828, "learning_rate": 9.90521241518901e-05, "loss": 0.0283, "step": 4360 }, { "grad_norm": 0.13816744089126587, "learning_rate": 9.904409662613308e-05, "loss": 0.0282, "step": 4370 }, { "grad_norm": 0.1351037174463272, "learning_rate": 9.903603557919849e-05, "loss": 0.0269, "step": 4380 }, { "grad_norm": 0.14647354185581207, "learning_rate": 9.902794101659594e-05, "loss": 0.033, "step": 4390 }, { "grad_norm": 0.1528237909078598, "learning_rate": 9.901981294385803e-05, "loss": 0.035, "step": 4400 }, { "grad_norm": 0.14135706424713135, "learning_rate": 9.901165136654018e-05, "loss": 0.0295, "step": 4410 }, { "grad_norm": 0.15017341077327728, "learning_rate": 9.900345629022079e-05, "loss": 0.0309, "step": 4420 }, { "grad_norm": 0.12919354438781738, "learning_rate": 9.89952277205011e-05, "loss": 0.0305, "step": 4430 }, { "grad_norm": 0.12981101870536804, "learning_rate": 9.898696566300527e-05, "loss": 0.0335, "step": 4440 }, { "grad_norm": 0.11141399294137955, "learning_rate": 9.897867012338032e-05, "loss": 0.0264, "step": 4450 }, { "grad_norm": 0.11245326697826385, "learning_rate": 9.897034110729617e-05, "loss": 0.0281, "step": 4460 }, { "grad_norm": 0.15948598086833954, "learning_rate": 9.896197862044564e-05, "loss": 0.0307, "step": 4470 }, { "grad_norm": 0.15581823885440826, "learning_rate": 9.895358266854442e-05, "loss": 0.0302, "step": 4480 }, { "grad_norm": 0.16008271276950836, "learning_rate": 9.894515325733103e-05, "loss": 0.031, "step": 4490 }, { "grad_norm": 0.13679832220077515, "learning_rate": 9.893669039256693e-05, "loss": 0.0271, "step": 4500 }, { "grad_norm": 0.13427338004112244, "learning_rate": 9.89281940800364e-05, "loss": 0.0329, "step": 4510 }, { "grad_norm": 0.08908092230558395, "learning_rate": 9.891966432554655e-05, "loss": 0.0283, "step": 4520 }, { "grad_norm": 0.11650774627923965, "learning_rate": 9.891110113492745e-05, "loss": 0.0295, "step": 4530 }, { "grad_norm": 0.14252157509326935, "learning_rate": 9.89025045140319e-05, "loss": 0.0292, "step": 4540 }, { "grad_norm": 0.11128199100494385, "learning_rate": 9.889387446873567e-05, "loss": 0.0283, "step": 4550 }, { "grad_norm": 0.12535175681114197, "learning_rate": 9.888521100493726e-05, "loss": 0.0254, "step": 4560 }, { "grad_norm": 0.12634210288524628, "learning_rate": 9.887651412855809e-05, "loss": 0.0307, "step": 4570 }, { "grad_norm": 0.15337087213993073, "learning_rate": 9.88677838455424e-05, "loss": 0.0292, "step": 4580 }, { "grad_norm": 0.12363621592521667, "learning_rate": 9.885902016185725e-05, "loss": 0.0289, "step": 4590 }, { "grad_norm": 0.11547315865755081, "learning_rate": 9.885022308349252e-05, "loss": 0.0271, "step": 4600 }, { "grad_norm": 0.0996483713388443, "learning_rate": 9.884139261646095e-05, "loss": 0.0313, "step": 4610 }, { "grad_norm": 0.12303982675075531, "learning_rate": 9.883252876679807e-05, "loss": 0.0325, "step": 4620 }, { "grad_norm": 0.12323678284883499, "learning_rate": 9.882363154056225e-05, "loss": 0.0311, "step": 4630 }, { "grad_norm": 0.13531044125556946, "learning_rate": 9.881470094383465e-05, "loss": 0.0294, "step": 4640 }, { "grad_norm": 0.13369566202163696, "learning_rate": 9.880573698271924e-05, "loss": 0.0307, "step": 4650 }, { "grad_norm": 0.13207867741584778, "learning_rate": 9.879673966334282e-05, "loss": 0.0359, "step": 4660 }, { "grad_norm": 0.1124132052063942, "learning_rate": 9.878770899185496e-05, "loss": 0.0303, "step": 4670 }, { "grad_norm": 0.1623193472623825, "learning_rate": 9.877864497442804e-05, "loss": 0.0254, "step": 4680 }, { "grad_norm": 0.1400090605020523, "learning_rate": 9.876954761725723e-05, "loss": 0.032, "step": 4690 }, { "grad_norm": 0.20034898817539215, "learning_rate": 9.876041692656052e-05, "loss": 0.0312, "step": 4700 }, { "grad_norm": 0.12346985191106796, "learning_rate": 9.87512529085786e-05, "loss": 0.0294, "step": 4710 }, { "grad_norm": 0.13754571974277496, "learning_rate": 9.874205556957503e-05, "loss": 0.0299, "step": 4720 }, { "grad_norm": 0.14072203636169434, "learning_rate": 9.873282491583608e-05, "loss": 0.0288, "step": 4730 }, { "grad_norm": 0.11869530379772186, "learning_rate": 9.872356095367084e-05, "loss": 0.0289, "step": 4740 }, { "grad_norm": 0.09897755831480026, "learning_rate": 9.87142636894111e-05, "loss": 0.0297, "step": 4750 }, { "grad_norm": 0.10574028640985489, "learning_rate": 9.870493312941148e-05, "loss": 0.0275, "step": 4760 }, { "grad_norm": 0.12462888658046722, "learning_rate": 9.869556928004933e-05, "loss": 0.0328, "step": 4770 }, { "grad_norm": 0.11893203109502792, "learning_rate": 9.868617214772476e-05, "loss": 0.0307, "step": 4780 }, { "grad_norm": 0.16910697519779205, "learning_rate": 9.867674173886059e-05, "loss": 0.0317, "step": 4790 }, { "grad_norm": 0.18454547226428986, "learning_rate": 9.866727805990245e-05, "loss": 0.0313, "step": 4800 }, { "grad_norm": 0.1425071507692337, "learning_rate": 9.865778111731865e-05, "loss": 0.0295, "step": 4810 }, { "grad_norm": 0.0847158208489418, "learning_rate": 9.864825091760028e-05, "loss": 0.0278, "step": 4820 }, { "grad_norm": 0.1591445654630661, "learning_rate": 9.863868746726112e-05, "loss": 0.0344, "step": 4830 }, { "grad_norm": 0.11879637092351913, "learning_rate": 9.862909077283773e-05, "loss": 0.0376, "step": 4840 }, { "grad_norm": 0.17099037766456604, "learning_rate": 9.861946084088933e-05, "loss": 0.0337, "step": 4850 }, { "grad_norm": 0.16577233374118805, "learning_rate": 9.860979767799792e-05, "loss": 0.0302, "step": 4860 }, { "grad_norm": 0.14668937027454376, "learning_rate": 9.860010129076813e-05, "loss": 0.0327, "step": 4870 }, { "grad_norm": 0.11291735619306564, "learning_rate": 9.859037168582741e-05, "loss": 0.0332, "step": 4880 }, { "grad_norm": 0.25719165802001953, "learning_rate": 9.85806088698258e-05, "loss": 0.0353, "step": 4890 }, { "grad_norm": 0.11225642263889313, "learning_rate": 9.857081284943612e-05, "loss": 0.0338, "step": 4900 }, { "grad_norm": 0.1288517266511917, "learning_rate": 9.856098363135385e-05, "loss": 0.032, "step": 4910 }, { "grad_norm": 0.17473100125789642, "learning_rate": 9.855112122229717e-05, "loss": 0.0318, "step": 4920 }, { "grad_norm": 0.16769173741340637, "learning_rate": 9.854122562900696e-05, "loss": 0.035, "step": 4930 }, { "grad_norm": 0.13900285959243774, "learning_rate": 9.853129685824673e-05, "loss": 0.0309, "step": 4940 }, { "grad_norm": 0.18818299472332, "learning_rate": 9.852133491680273e-05, "loss": 0.0348, "step": 4950 }, { "grad_norm": 0.17710934579372406, "learning_rate": 9.851133981148385e-05, "loss": 0.0284, "step": 4960 }, { "grad_norm": 0.16860008239746094, "learning_rate": 9.850131154912164e-05, "loss": 0.0356, "step": 4970 }, { "grad_norm": 0.12510299682617188, "learning_rate": 9.849125013657031e-05, "loss": 0.0341, "step": 4980 }, { "grad_norm": 0.168047696352005, "learning_rate": 9.848115558070676e-05, "loss": 0.0329, "step": 4990 }, { "grad_norm": 0.14191985130310059, "learning_rate": 9.847102788843055e-05, "loss": 0.0308, "step": 5000 }, { "grad_norm": 0.14457592368125916, "learning_rate": 9.84608670666638e-05, "loss": 0.0319, "step": 5010 }, { "grad_norm": 0.1412932574748993, "learning_rate": 9.845067312235138e-05, "loss": 0.0329, "step": 5020 }, { "grad_norm": 0.17743176221847534, "learning_rate": 9.844044606246074e-05, "loss": 0.0322, "step": 5030 }, { "grad_norm": 0.160648375749588, "learning_rate": 9.843018589398199e-05, "loss": 0.0302, "step": 5040 }, { "grad_norm": 0.19210588932037354, "learning_rate": 9.841989262392785e-05, "loss": 0.0321, "step": 5050 }, { "grad_norm": 0.29281091690063477, "learning_rate": 9.840956625933367e-05, "loss": 0.0315, "step": 5060 }, { "grad_norm": 0.15939855575561523, "learning_rate": 9.839920680725743e-05, "loss": 0.0274, "step": 5070 }, { "grad_norm": 0.13194029033184052, "learning_rate": 9.83888142747797e-05, "loss": 0.0314, "step": 5080 }, { "grad_norm": 0.17578157782554626, "learning_rate": 9.83783886690037e-05, "loss": 0.0321, "step": 5090 }, { "grad_norm": 0.1537005752325058, "learning_rate": 9.836792999705524e-05, "loss": 0.0335, "step": 5100 }, { "grad_norm": 0.15910406410694122, "learning_rate": 9.835743826608271e-05, "loss": 0.0294, "step": 5110 }, { "grad_norm": 0.16012096405029297, "learning_rate": 9.834691348325709e-05, "loss": 0.0349, "step": 5120 }, { "grad_norm": 0.20127712190151215, "learning_rate": 9.833635565577199e-05, "loss": 0.0276, "step": 5130 }, { "grad_norm": 0.17446377873420715, "learning_rate": 9.832576479084359e-05, "loss": 0.0303, "step": 5140 }, { "grad_norm": 0.11303416639566422, "learning_rate": 9.831514089571064e-05, "loss": 0.029, "step": 5150 }, { "grad_norm": 0.11987681686878204, "learning_rate": 9.830448397763447e-05, "loss": 0.0319, "step": 5160 }, { "grad_norm": 0.13630251586437225, "learning_rate": 9.8293794043899e-05, "loss": 0.0259, "step": 5170 }, { "grad_norm": 0.11049032211303711, "learning_rate": 9.828307110181066e-05, "loss": 0.0291, "step": 5180 }, { "grad_norm": 0.16467329859733582, "learning_rate": 9.827231515869852e-05, "loss": 0.0322, "step": 5190 }, { "grad_norm": 0.15353043377399445, "learning_rate": 9.826152622191413e-05, "loss": 0.0317, "step": 5200 }, { "grad_norm": 0.1435941606760025, "learning_rate": 9.825070429883167e-05, "loss": 0.0308, "step": 5210 }, { "grad_norm": 0.12029659003019333, "learning_rate": 9.82398493968478e-05, "loss": 0.0315, "step": 5220 }, { "grad_norm": 0.15446722507476807, "learning_rate": 9.822896152338173e-05, "loss": 0.0332, "step": 5230 }, { "grad_norm": 0.14214634895324707, "learning_rate": 9.821804068587523e-05, "loss": 0.0294, "step": 5240 }, { "grad_norm": 0.1743541657924652, "learning_rate": 9.820708689179259e-05, "loss": 0.028, "step": 5250 }, { "grad_norm": 0.11939850449562073, "learning_rate": 9.819610014862063e-05, "loss": 0.0299, "step": 5260 }, { "grad_norm": 0.13965731859207153, "learning_rate": 9.818508046386868e-05, "loss": 0.0297, "step": 5270 }, { "grad_norm": 0.14425750076770782, "learning_rate": 9.817402784506859e-05, "loss": 0.0309, "step": 5280 }, { "grad_norm": 0.13211046159267426, "learning_rate": 9.816294229977472e-05, "loss": 0.0306, "step": 5290 }, { "grad_norm": 0.1352594643831253, "learning_rate": 9.815182383556394e-05, "loss": 0.0325, "step": 5300 }, { "grad_norm": 0.13577550649642944, "learning_rate": 9.81406724600356e-05, "loss": 0.0296, "step": 5310 }, { "grad_norm": 0.14895831048488617, "learning_rate": 9.812948818081161e-05, "loss": 0.0298, "step": 5320 }, { "grad_norm": 0.144982248544693, "learning_rate": 9.811827100553623e-05, "loss": 0.0283, "step": 5330 }, { "grad_norm": 0.1312551647424698, "learning_rate": 9.810702094187638e-05, "loss": 0.0319, "step": 5340 }, { "grad_norm": 0.13265414535999298, "learning_rate": 9.809573799752135e-05, "loss": 0.0296, "step": 5350 }, { "grad_norm": 0.08128208667039871, "learning_rate": 9.808442218018288e-05, "loss": 0.0291, "step": 5360 }, { "grad_norm": 0.16145029664039612, "learning_rate": 9.807307349759527e-05, "loss": 0.0333, "step": 5370 }, { "grad_norm": 0.1585756242275238, "learning_rate": 9.806169195751525e-05, "loss": 0.0287, "step": 5380 }, { "grad_norm": 0.16165673732757568, "learning_rate": 9.805027756772194e-05, "loss": 0.0285, "step": 5390 }, { "grad_norm": 0.14438237249851227, "learning_rate": 9.803883033601702e-05, "loss": 0.0291, "step": 5400 }, { "grad_norm": 0.16120082139968872, "learning_rate": 9.802735027022453e-05, "loss": 0.032, "step": 5410 }, { "grad_norm": 0.14198072254657745, "learning_rate": 9.8015837378191e-05, "loss": 0.0279, "step": 5420 }, { "grad_norm": 0.1286853849887848, "learning_rate": 9.80042916677854e-05, "loss": 0.0302, "step": 5430 }, { "grad_norm": 0.1471838802099228, "learning_rate": 9.799271314689908e-05, "loss": 0.0298, "step": 5440 }, { "grad_norm": 0.13195747137069702, "learning_rate": 9.798110182344588e-05, "loss": 0.0338, "step": 5450 }, { "grad_norm": 0.1515452265739441, "learning_rate": 9.796945770536204e-05, "loss": 0.0271, "step": 5460 }, { "grad_norm": 0.1370011419057846, "learning_rate": 9.795778080060615e-05, "loss": 0.0311, "step": 5470 }, { "grad_norm": 0.14182420074939728, "learning_rate": 9.79460711171593e-05, "loss": 0.0274, "step": 5480 }, { "grad_norm": 0.18628908693790436, "learning_rate": 9.793432866302497e-05, "loss": 0.0306, "step": 5490 }, { "grad_norm": 0.1389557421207428, "learning_rate": 9.792255344622897e-05, "loss": 0.03, "step": 5500 }, { "grad_norm": 0.14430497586727142, "learning_rate": 9.791074547481957e-05, "loss": 0.0295, "step": 5510 }, { "grad_norm": 0.222390815615654, "learning_rate": 9.789890475686742e-05, "loss": 0.0273, "step": 5520 }, { "grad_norm": 0.1621459722518921, "learning_rate": 9.788703130046552e-05, "loss": 0.0362, "step": 5530 }, { "grad_norm": 0.10353144258260727, "learning_rate": 9.78751251137293e-05, "loss": 0.0286, "step": 5540 }, { "grad_norm": 0.20079924166202545, "learning_rate": 9.786318620479646e-05, "loss": 0.0288, "step": 5550 }, { "grad_norm": 0.20867131650447845, "learning_rate": 9.78512145818272e-05, "loss": 0.0303, "step": 5560 }, { "grad_norm": 0.16477102041244507, "learning_rate": 9.783921025300398e-05, "loss": 0.0302, "step": 5570 }, { "grad_norm": 0.11863923072814941, "learning_rate": 9.782717322653164e-05, "loss": 0.0317, "step": 5580 }, { "grad_norm": 0.0928756520152092, "learning_rate": 9.781510351063738e-05, "loss": 0.031, "step": 5590 }, { "grad_norm": 0.1299767941236496, "learning_rate": 9.780300111357075e-05, "loss": 0.031, "step": 5600 }, { "grad_norm": 0.11770078539848328, "learning_rate": 9.779086604360361e-05, "loss": 0.0338, "step": 5610 }, { "grad_norm": 0.13189104199409485, "learning_rate": 9.777869830903019e-05, "loss": 0.0312, "step": 5620 }, { "grad_norm": 0.13883882761001587, "learning_rate": 9.776649791816698e-05, "loss": 0.0285, "step": 5630 }, { "grad_norm": 0.23329204320907593, "learning_rate": 9.775426487935285e-05, "loss": 0.0312, "step": 5640 }, { "grad_norm": 0.12346314638853073, "learning_rate": 9.774199920094898e-05, "loss": 0.0251, "step": 5650 }, { "grad_norm": 0.1368986815214157, "learning_rate": 9.772970089133884e-05, "loss": 0.0313, "step": 5660 }, { "grad_norm": 0.10948159545660019, "learning_rate": 9.77173699589282e-05, "loss": 0.0285, "step": 5670 }, { "grad_norm": 0.12310419976711273, "learning_rate": 9.770500641214513e-05, "loss": 0.03, "step": 5680 }, { "grad_norm": 0.11432875692844391, "learning_rate": 9.769261025944003e-05, "loss": 0.0334, "step": 5690 }, { "grad_norm": 0.08589506149291992, "learning_rate": 9.768018150928552e-05, "loss": 0.0266, "step": 5700 }, { "grad_norm": 0.16488701105117798, "learning_rate": 9.766772017017654e-05, "loss": 0.0335, "step": 5710 }, { "grad_norm": 0.16456547379493713, "learning_rate": 9.765522625063032e-05, "loss": 0.033, "step": 5720 }, { "grad_norm": 0.1149597316980362, "learning_rate": 9.76426997591863e-05, "loss": 0.0342, "step": 5730 }, { "grad_norm": 0.1175214946269989, "learning_rate": 9.763014070440625e-05, "loss": 0.0285, "step": 5740 }, { "grad_norm": 0.12835101783275604, "learning_rate": 9.761754909487415e-05, "loss": 0.0293, "step": 5750 }, { "grad_norm": 0.15369701385498047, "learning_rate": 9.760492493919626e-05, "loss": 0.0292, "step": 5760 }, { "grad_norm": 0.11799885332584381, "learning_rate": 9.759226824600105e-05, "loss": 0.0291, "step": 5770 }, { "grad_norm": 0.1319350153207779, "learning_rate": 9.757957902393928e-05, "loss": 0.0325, "step": 5780 }, { "grad_norm": 0.12328489124774933, "learning_rate": 9.756685728168387e-05, "loss": 0.0281, "step": 5790 }, { "grad_norm": 0.135609969496727, "learning_rate": 9.755410302793004e-05, "loss": 0.0276, "step": 5800 }, { "grad_norm": 0.12795338034629822, "learning_rate": 9.754131627139522e-05, "loss": 0.0265, "step": 5810 }, { "grad_norm": 0.16833573579788208, "learning_rate": 9.752849702081901e-05, "loss": 0.0292, "step": 5820 }, { "grad_norm": 0.14994065463542938, "learning_rate": 9.751564528496324e-05, "loss": 0.0316, "step": 5830 }, { "grad_norm": 0.13157837092876434, "learning_rate": 9.750276107261197e-05, "loss": 0.03, "step": 5840 }, { "grad_norm": 0.15753118693828583, "learning_rate": 9.748984439257142e-05, "loss": 0.0341, "step": 5850 }, { "grad_norm": 0.17204560339450836, "learning_rate": 9.747689525367005e-05, "loss": 0.0296, "step": 5860 }, { "grad_norm": 0.15211695432662964, "learning_rate": 9.746391366475845e-05, "loss": 0.0296, "step": 5870 }, { "grad_norm": 0.1366327553987503, "learning_rate": 9.745089963470942e-05, "loss": 0.0359, "step": 5880 }, { "grad_norm": 0.1404152810573578, "learning_rate": 9.743785317241791e-05, "loss": 0.0299, "step": 5890 }, { "grad_norm": 0.1538352370262146, "learning_rate": 9.742477428680108e-05, "loss": 0.0315, "step": 5900 }, { "grad_norm": 0.1510547250509262, "learning_rate": 9.741166298679821e-05, "loss": 0.0281, "step": 5910 }, { "grad_norm": 0.1302216351032257, "learning_rate": 9.739851928137076e-05, "loss": 0.0294, "step": 5920 }, { "grad_norm": 0.11489558964967728, "learning_rate": 9.738534317950232e-05, "loss": 0.0271, "step": 5930 }, { "grad_norm": 0.0886746495962143, "learning_rate": 9.737213469019864e-05, "loss": 0.0288, "step": 5940 }, { "grad_norm": 0.09817933291196823, "learning_rate": 9.735889382248757e-05, "loss": 0.0276, "step": 5950 }, { "grad_norm": 0.14108243584632874, "learning_rate": 9.734562058541916e-05, "loss": 0.0266, "step": 5960 }, { "grad_norm": 0.1258392333984375, "learning_rate": 9.733231498806552e-05, "loss": 0.0273, "step": 5970 }, { "grad_norm": 0.13050666451454163, "learning_rate": 9.731897703952088e-05, "loss": 0.031, "step": 5980 }, { "grad_norm": 0.11958254128694534, "learning_rate": 9.730560674890165e-05, "loss": 0.0283, "step": 5990 }, { "grad_norm": 0.20901378989219666, "learning_rate": 9.729220412534628e-05, "loss": 0.0301, "step": 6000 }, { "grad_norm": 0.22273048758506775, "learning_rate": 9.72787691780153e-05, "loss": 0.0291, "step": 6010 }, { "grad_norm": 0.13543297350406647, "learning_rate": 9.726530191609142e-05, "loss": 0.0311, "step": 6020 }, { "grad_norm": 0.17966340482234955, "learning_rate": 9.725180234877937e-05, "loss": 0.0275, "step": 6030 }, { "grad_norm": 0.13217264413833618, "learning_rate": 9.723827048530595e-05, "loss": 0.0313, "step": 6040 }, { "grad_norm": 0.14630313217639923, "learning_rate": 9.722470633492011e-05, "loss": 0.0261, "step": 6050 }, { "grad_norm": 0.15577538311481476, "learning_rate": 9.721110990689278e-05, "loss": 0.0284, "step": 6060 }, { "grad_norm": 0.1326613426208496, "learning_rate": 9.719748121051699e-05, "loss": 0.0279, "step": 6070 }, { "grad_norm": 0.12177961319684982, "learning_rate": 9.718382025510785e-05, "loss": 0.0301, "step": 6080 }, { "grad_norm": 0.09021580219268799, "learning_rate": 9.717012705000248e-05, "loss": 0.0274, "step": 6090 }, { "grad_norm": 0.10481106489896774, "learning_rate": 9.715640160456006e-05, "loss": 0.0263, "step": 6100 }, { "grad_norm": 0.08577946573495865, "learning_rate": 9.714264392816181e-05, "loss": 0.0254, "step": 6110 }, { "grad_norm": 0.12203171849250793, "learning_rate": 9.712885403021095e-05, "loss": 0.0293, "step": 6120 }, { "grad_norm": 0.1585262566804886, "learning_rate": 9.711503192013276e-05, "loss": 0.0289, "step": 6130 }, { "grad_norm": 0.10581773519515991, "learning_rate": 9.710117760737454e-05, "loss": 0.0262, "step": 6140 }, { "grad_norm": 0.12545433640480042, "learning_rate": 9.708729110140554e-05, "loss": 0.0257, "step": 6150 }, { "grad_norm": 0.10892228782176971, "learning_rate": 9.707337241171709e-05, "loss": 0.0301, "step": 6160 }, { "grad_norm": 0.12374264001846313, "learning_rate": 9.705942154782245e-05, "loss": 0.028, "step": 6170 }, { "grad_norm": 0.1361217051744461, "learning_rate": 9.704543851925696e-05, "loss": 0.0305, "step": 6180 }, { "grad_norm": 0.14004649221897125, "learning_rate": 9.703142333557784e-05, "loss": 0.0314, "step": 6190 }, { "grad_norm": 0.13490745425224304, "learning_rate": 9.701737600636436e-05, "loss": 0.0298, "step": 6200 }, { "grad_norm": 0.13904361426830292, "learning_rate": 9.700329654121771e-05, "loss": 0.0289, "step": 6210 }, { "grad_norm": 0.12831024825572968, "learning_rate": 9.698918494976109e-05, "loss": 0.0292, "step": 6220 }, { "grad_norm": 0.11071411520242691, "learning_rate": 9.697504124163965e-05, "loss": 0.0269, "step": 6230 }, { "grad_norm": 0.1317506730556488, "learning_rate": 9.696086542652045e-05, "loss": 0.0267, "step": 6240 }, { "grad_norm": 0.10513166338205338, "learning_rate": 9.694665751409256e-05, "loss": 0.032, "step": 6250 }, { "grad_norm": 0.13786643743515015, "learning_rate": 9.693241751406694e-05, "loss": 0.0315, "step": 6260 }, { "grad_norm": 0.12210483849048615, "learning_rate": 9.69181454361765e-05, "loss": 0.0316, "step": 6270 }, { "grad_norm": 0.1197618767619133, "learning_rate": 9.690384129017605e-05, "loss": 0.0295, "step": 6280 }, { "grad_norm": 0.10901889950037003, "learning_rate": 9.688950508584237e-05, "loss": 0.0311, "step": 6290 }, { "grad_norm": 0.11445862054824829, "learning_rate": 9.68751368329741e-05, "loss": 0.0275, "step": 6300 }, { "grad_norm": 0.12492936104536057, "learning_rate": 9.68607365413918e-05, "loss": 0.0291, "step": 6310 }, { "grad_norm": 0.11375363916158676, "learning_rate": 9.684630422093797e-05, "loss": 0.0276, "step": 6320 }, { "grad_norm": 0.11537330597639084, "learning_rate": 9.683183988147693e-05, "loss": 0.0275, "step": 6330 }, { "grad_norm": 0.1284368634223938, "learning_rate": 9.681734353289491e-05, "loss": 0.0282, "step": 6340 }, { "grad_norm": 0.12112817168235779, "learning_rate": 9.680281518510006e-05, "loss": 0.026, "step": 6350 }, { "grad_norm": 0.11492860317230225, "learning_rate": 9.678825484802236e-05, "loss": 0.0303, "step": 6360 }, { "grad_norm": 0.12113387137651443, "learning_rate": 9.677366253161365e-05, "loss": 0.0284, "step": 6370 }, { "grad_norm": 0.1283017098903656, "learning_rate": 9.675903824584765e-05, "loss": 0.0271, "step": 6380 }, { "grad_norm": 0.12955982983112335, "learning_rate": 9.674438200071991e-05, "loss": 0.0307, "step": 6390 }, { "grad_norm": 0.10987752676010132, "learning_rate": 9.672969380624786e-05, "loss": 0.0309, "step": 6400 }, { "grad_norm": 0.12524156272411346, "learning_rate": 9.671497367247073e-05, "loss": 0.0273, "step": 6410 }, { "grad_norm": 0.1227864995598793, "learning_rate": 9.670022160944959e-05, "loss": 0.0275, "step": 6420 }, { "grad_norm": 0.12613946199417114, "learning_rate": 9.668543762726734e-05, "loss": 0.0259, "step": 6430 }, { "grad_norm": 0.1554843634366989, "learning_rate": 9.667062173602868e-05, "loss": 0.0303, "step": 6440 }, { "grad_norm": 0.14109064638614655, "learning_rate": 9.665577394586013e-05, "loss": 0.0272, "step": 6450 }, { "grad_norm": 0.1692899465560913, "learning_rate": 9.664089426691006e-05, "loss": 0.0275, "step": 6460 }, { "grad_norm": 0.12164231389760971, "learning_rate": 9.662598270934852e-05, "loss": 0.0305, "step": 6470 }, { "grad_norm": 0.09920217096805573, "learning_rate": 9.661103928336748e-05, "loss": 0.0277, "step": 6480 }, { "grad_norm": 0.09631576389074326, "learning_rate": 9.659606399918058e-05, "loss": 0.0288, "step": 6490 }, { "grad_norm": 0.07929295301437378, "learning_rate": 9.658105686702334e-05, "loss": 0.0287, "step": 6500 }, { "grad_norm": 0.12827752530574799, "learning_rate": 9.656601789715294e-05, "loss": 0.029, "step": 6510 }, { "grad_norm": 0.15072348713874817, "learning_rate": 9.655094709984842e-05, "loss": 0.0339, "step": 6520 }, { "grad_norm": 0.09238472580909729, "learning_rate": 9.653584448541048e-05, "loss": 0.0305, "step": 6530 }, { "grad_norm": 0.08577961474657059, "learning_rate": 9.652071006416166e-05, "loss": 0.0306, "step": 6540 }, { "grad_norm": 0.11022467911243439, "learning_rate": 9.650554384644617e-05, "loss": 0.0301, "step": 6550 }, { "grad_norm": 0.10836803913116455, "learning_rate": 9.649034584263e-05, "loss": 0.0322, "step": 6560 }, { "grad_norm": 0.11738304048776627, "learning_rate": 9.64751160631008e-05, "loss": 0.0326, "step": 6570 }, { "grad_norm": 0.11991164088249207, "learning_rate": 9.645985451826803e-05, "loss": 0.035, "step": 6580 }, { "grad_norm": 0.11296238005161285, "learning_rate": 9.644456121856275e-05, "loss": 0.0298, "step": 6590 }, { "grad_norm": 0.13850857317447662, "learning_rate": 9.642923617443784e-05, "loss": 0.0307, "step": 6600 }, { "grad_norm": 0.11150801181793213, "learning_rate": 9.641387939636782e-05, "loss": 0.0299, "step": 6610 }, { "grad_norm": 0.13149791955947876, "learning_rate": 9.639849089484888e-05, "loss": 0.0285, "step": 6620 }, { "grad_norm": 0.14225752651691437, "learning_rate": 9.638307068039893e-05, "loss": 0.0314, "step": 6630 }, { "grad_norm": 0.12269390374422073, "learning_rate": 9.636761876355753e-05, "loss": 0.0324, "step": 6640 }, { "grad_norm": 0.11802932620048523, "learning_rate": 9.635213515488593e-05, "loss": 0.0284, "step": 6650 }, { "grad_norm": 0.1382441222667694, "learning_rate": 9.633661986496702e-05, "loss": 0.0295, "step": 6660 }, { "grad_norm": 0.12194661051034927, "learning_rate": 9.632107290440535e-05, "loss": 0.0306, "step": 6670 }, { "grad_norm": 0.12684839963912964, "learning_rate": 9.630549428382715e-05, "loss": 0.0332, "step": 6680 }, { "grad_norm": 0.13715089857578278, "learning_rate": 9.628988401388024e-05, "loss": 0.0287, "step": 6690 }, { "grad_norm": 0.1318369358778, "learning_rate": 9.627424210523407e-05, "loss": 0.0296, "step": 6700 }, { "grad_norm": 0.1474905014038086, "learning_rate": 9.62585685685798e-05, "loss": 0.0289, "step": 6710 }, { "grad_norm": 0.15069730579853058, "learning_rate": 9.624286341463008e-05, "loss": 0.0315, "step": 6720 }, { "grad_norm": 0.13619889318943024, "learning_rate": 9.622712665411927e-05, "loss": 0.0304, "step": 6730 }, { "grad_norm": 0.10918869078159332, "learning_rate": 9.621135829780328e-05, "loss": 0.0321, "step": 6740 }, { "grad_norm": 0.11007779091596603, "learning_rate": 9.619555835645964e-05, "loss": 0.0292, "step": 6750 }, { "grad_norm": 0.10426288843154907, "learning_rate": 9.617972684088747e-05, "loss": 0.0314, "step": 6760 }, { "grad_norm": 0.10993001610040665, "learning_rate": 9.616386376190745e-05, "loss": 0.0271, "step": 6770 }, { "grad_norm": 0.12060892581939697, "learning_rate": 9.614796913036184e-05, "loss": 0.031, "step": 6780 }, { "grad_norm": 0.15469849109649658, "learning_rate": 9.61320429571145e-05, "loss": 0.0304, "step": 6790 }, { "grad_norm": 0.16319049894809723, "learning_rate": 9.611608525305078e-05, "loss": 0.0355, "step": 6800 }, { "grad_norm": 0.11188105493783951, "learning_rate": 9.610009602907764e-05, "loss": 0.0302, "step": 6810 }, { "grad_norm": 0.10680803656578064, "learning_rate": 9.608407529612357e-05, "loss": 0.0296, "step": 6820 }, { "grad_norm": 0.16108226776123047, "learning_rate": 9.606802306513857e-05, "loss": 0.0278, "step": 6830 }, { "grad_norm": 0.165028378367424, "learning_rate": 9.605193934709421e-05, "loss": 0.0284, "step": 6840 }, { "grad_norm": 0.12556092441082, "learning_rate": 9.603582415298354e-05, "loss": 0.0304, "step": 6850 }, { "grad_norm": 0.15024738013744354, "learning_rate": 9.601967749382117e-05, "loss": 0.0307, "step": 6860 }, { "grad_norm": 0.1333203911781311, "learning_rate": 9.600349938064316e-05, "loss": 0.0337, "step": 6870 }, { "grad_norm": 0.3096352815628052, "learning_rate": 9.598728982450711e-05, "loss": 0.0327, "step": 6880 }, { "grad_norm": 0.12009783089160919, "learning_rate": 9.597104883649211e-05, "loss": 0.0315, "step": 6890 }, { "grad_norm": 0.14754359424114227, "learning_rate": 9.59547764276987e-05, "loss": 0.0295, "step": 6900 }, { "grad_norm": 0.12644338607788086, "learning_rate": 9.593847260924892e-05, "loss": 0.0319, "step": 6910 }, { "grad_norm": 0.16354167461395264, "learning_rate": 9.59221373922863e-05, "loss": 0.0362, "step": 6920 }, { "grad_norm": 0.1631135791540146, "learning_rate": 9.59057707879758e-05, "loss": 0.0268, "step": 6930 }, { "grad_norm": 0.0802006945014, "learning_rate": 9.588937280750382e-05, "loss": 0.0276, "step": 6940 }, { "grad_norm": 0.1272844821214676, "learning_rate": 9.587294346207824e-05, "loss": 0.0333, "step": 6950 }, { "grad_norm": 0.15741823613643646, "learning_rate": 9.585648276292836e-05, "loss": 0.0321, "step": 6960 }, { "grad_norm": 0.13791581988334656, "learning_rate": 9.583999072130492e-05, "loss": 0.0305, "step": 6970 }, { "grad_norm": 0.11056883633136749, "learning_rate": 9.582346734848005e-05, "loss": 0.0302, "step": 6980 }, { "grad_norm": 0.1247747540473938, "learning_rate": 9.580691265574735e-05, "loss": 0.0345, "step": 6990 }, { "grad_norm": 0.13535933196544647, "learning_rate": 9.579032665442178e-05, "loss": 0.0288, "step": 7000 }, { "grad_norm": 0.1283828616142273, "learning_rate": 9.577370935583971e-05, "loss": 0.0304, "step": 7010 }, { "grad_norm": 0.11071411520242691, "learning_rate": 9.575706077135893e-05, "loss": 0.0261, "step": 7020 }, { "grad_norm": 0.290946364402771, "learning_rate": 9.574038091235857e-05, "loss": 0.0307, "step": 7030 }, { "grad_norm": 0.19254441559314728, "learning_rate": 9.572366979023916e-05, "loss": 0.0308, "step": 7040 }, { "grad_norm": 0.17995736002922058, "learning_rate": 9.57069274164226e-05, "loss": 0.0338, "step": 7050 }, { "grad_norm": 0.12612245976924896, "learning_rate": 9.569015380235213e-05, "loss": 0.0279, "step": 7060 }, { "grad_norm": 0.3300417959690094, "learning_rate": 9.567334895949238e-05, "loss": 0.0341, "step": 7070 }, { "grad_norm": 0.15364955365657806, "learning_rate": 9.565651289932928e-05, "loss": 0.0276, "step": 7080 }, { "grad_norm": 0.1227569505572319, "learning_rate": 9.563964563337015e-05, "loss": 0.0262, "step": 7090 }, { "grad_norm": 0.10855550318956375, "learning_rate": 9.562274717314357e-05, "loss": 0.0278, "step": 7100 }, { "grad_norm": 0.24683040380477905, "learning_rate": 9.56058175301995e-05, "loss": 0.0309, "step": 7110 }, { "grad_norm": 0.16517525911331177, "learning_rate": 9.558885671610918e-05, "loss": 0.0307, "step": 7120 }, { "grad_norm": 0.16336846351623535, "learning_rate": 9.557186474246517e-05, "loss": 0.0266, "step": 7130 }, { "grad_norm": 0.13334348797798157, "learning_rate": 9.555484162088133e-05, "loss": 0.0285, "step": 7140 }, { "grad_norm": 0.1894219070672989, "learning_rate": 9.553778736299279e-05, "loss": 0.0308, "step": 7150 }, { "grad_norm": 0.15781021118164062, "learning_rate": 9.552070198045599e-05, "loss": 0.0321, "step": 7160 }, { "grad_norm": 0.09600543230772018, "learning_rate": 9.550358548494863e-05, "loss": 0.0297, "step": 7170 }, { "grad_norm": 0.16912998259067535, "learning_rate": 9.548643788816965e-05, "loss": 0.0271, "step": 7180 }, { "grad_norm": 0.205084890127182, "learning_rate": 9.546925920183929e-05, "loss": 0.0349, "step": 7190 }, { "grad_norm": 0.21282480657100677, "learning_rate": 9.545204943769902e-05, "loss": 0.0296, "step": 7200 }, { "grad_norm": 0.14357496798038483, "learning_rate": 9.543480860751155e-05, "loss": 0.0234, "step": 7210 }, { "grad_norm": 0.1317637860774994, "learning_rate": 9.541753672306081e-05, "loss": 0.0313, "step": 7220 }, { "grad_norm": 0.140080064535141, "learning_rate": 9.540023379615198e-05, "loss": 0.0265, "step": 7230 }, { "grad_norm": 0.12430628389120102, "learning_rate": 9.538289983861146e-05, "loss": 0.0284, "step": 7240 }, { "grad_norm": 0.1022285744547844, "learning_rate": 9.536553486228683e-05, "loss": 0.0269, "step": 7250 }, { "grad_norm": 0.11522375792264938, "learning_rate": 9.534813887904689e-05, "loss": 0.0288, "step": 7260 }, { "grad_norm": 0.11318046599626541, "learning_rate": 9.533071190078163e-05, "loss": 0.0318, "step": 7270 }, { "grad_norm": 0.13935163617134094, "learning_rate": 9.531325393940225e-05, "loss": 0.0281, "step": 7280 }, { "grad_norm": 0.1284327656030655, "learning_rate": 9.529576500684105e-05, "loss": 0.0289, "step": 7290 }, { "grad_norm": 0.11981936544179916, "learning_rate": 9.52782451150516e-05, "loss": 0.0317, "step": 7300 }, { "grad_norm": 0.15988172590732574, "learning_rate": 9.526069427600857e-05, "loss": 0.0321, "step": 7310 }, { "grad_norm": 0.14362624287605286, "learning_rate": 9.524311250170776e-05, "loss": 0.0293, "step": 7320 }, { "grad_norm": 0.15595948696136475, "learning_rate": 9.522549980416619e-05, "loss": 0.0318, "step": 7330 }, { "grad_norm": 0.15088491141796112, "learning_rate": 9.520785619542196e-05, "loss": 0.03, "step": 7340 }, { "grad_norm": 0.09049391001462936, "learning_rate": 9.519018168753428e-05, "loss": 0.0276, "step": 7350 }, { "grad_norm": 0.12931977212429047, "learning_rate": 9.517247629258357e-05, "loss": 0.0296, "step": 7360 }, { "grad_norm": 0.11632110923528671, "learning_rate": 9.515474002267124e-05, "loss": 0.0286, "step": 7370 }, { "grad_norm": 0.11256610602140427, "learning_rate": 9.513697288991989e-05, "loss": 0.0302, "step": 7380 }, { "grad_norm": 0.08972709625959396, "learning_rate": 9.51191749064732e-05, "loss": 0.0315, "step": 7390 }, { "grad_norm": 0.13377460837364197, "learning_rate": 9.51013460844959e-05, "loss": 0.0294, "step": 7400 }, { "grad_norm": 0.12008779495954514, "learning_rate": 9.508348643617382e-05, "loss": 0.0268, "step": 7410 }, { "grad_norm": 0.11306845396757126, "learning_rate": 9.50655959737139e-05, "loss": 0.0311, "step": 7420 }, { "grad_norm": 0.1326989233493805, "learning_rate": 9.504767470934405e-05, "loss": 0.0296, "step": 7430 }, { "grad_norm": 0.1431332677602768, "learning_rate": 9.502972265531332e-05, "loss": 0.0308, "step": 7440 }, { "grad_norm": 0.1050325557589531, "learning_rate": 9.501173982389175e-05, "loss": 0.0273, "step": 7450 }, { "grad_norm": 0.11629877239465714, "learning_rate": 9.499372622737047e-05, "loss": 0.0298, "step": 7460 }, { "grad_norm": 0.1521110236644745, "learning_rate": 9.497568187806156e-05, "loss": 0.0289, "step": 7470 }, { "grad_norm": 0.13615703582763672, "learning_rate": 9.495760678829819e-05, "loss": 0.0297, "step": 7480 }, { "grad_norm": 0.13013245165348053, "learning_rate": 9.49395009704345e-05, "loss": 0.0321, "step": 7490 }, { "grad_norm": 0.16086825728416443, "learning_rate": 9.492136443684565e-05, "loss": 0.0295, "step": 7500 }, { "grad_norm": 0.14486706256866455, "learning_rate": 9.49031971999278e-05, "loss": 0.0266, "step": 7510 }, { "grad_norm": 0.1504884660243988, "learning_rate": 9.488499927209806e-05, "loss": 0.0313, "step": 7520 }, { "grad_norm": 0.17915213108062744, "learning_rate": 9.486677066579456e-05, "loss": 0.0308, "step": 7530 }, { "grad_norm": 0.14156308770179749, "learning_rate": 9.484851139347639e-05, "loss": 0.0271, "step": 7540 }, { "grad_norm": 0.1325746327638626, "learning_rate": 9.483022146762358e-05, "loss": 0.028, "step": 7550 }, { "grad_norm": 0.12551262974739075, "learning_rate": 9.48119009007371e-05, "loss": 0.0288, "step": 7560 }, { "grad_norm": 0.12816300988197327, "learning_rate": 9.479354970533892e-05, "loss": 0.0299, "step": 7570 }, { "grad_norm": 0.09647355228662491, "learning_rate": 9.477516789397189e-05, "loss": 0.0317, "step": 7580 }, { "grad_norm": 0.13393355906009674, "learning_rate": 9.475675547919983e-05, "loss": 0.0291, "step": 7590 }, { "grad_norm": 0.12224545329809189, "learning_rate": 9.47383124736074e-05, "loss": 0.0322, "step": 7600 }, { "grad_norm": 0.14440667629241943, "learning_rate": 9.471983888980028e-05, "loss": 0.028, "step": 7610 }, { "grad_norm": 0.17801237106323242, "learning_rate": 9.470133474040494e-05, "loss": 0.0294, "step": 7620 }, { "grad_norm": 0.15496407449245453, "learning_rate": 9.468280003806882e-05, "loss": 0.0296, "step": 7630 }, { "grad_norm": 0.14569731056690216, "learning_rate": 9.46642347954602e-05, "loss": 0.0297, "step": 7640 }, { "grad_norm": 0.1063489094376564, "learning_rate": 9.464563902526826e-05, "loss": 0.0303, "step": 7650 }, { "grad_norm": 0.17049100995063782, "learning_rate": 9.462701274020303e-05, "loss": 0.028, "step": 7660 }, { "grad_norm": 0.11114421486854553, "learning_rate": 9.46083559529954e-05, "loss": 0.0268, "step": 7670 }, { "grad_norm": 0.12649796903133392, "learning_rate": 9.458966867639712e-05, "loss": 0.0275, "step": 7680 }, { "grad_norm": 0.11775965988636017, "learning_rate": 9.457095092318074e-05, "loss": 0.028, "step": 7690 }, { "grad_norm": 0.14498396217823029, "learning_rate": 9.455220270613969e-05, "loss": 0.0284, "step": 7700 }, { "grad_norm": 0.14256249368190765, "learning_rate": 9.45334240380882e-05, "loss": 0.0293, "step": 7710 }, { "grad_norm": 0.14090929925441742, "learning_rate": 9.451461493186129e-05, "loss": 0.0282, "step": 7720 }, { "grad_norm": 0.1711307317018509, "learning_rate": 9.449577540031482e-05, "loss": 0.034, "step": 7730 }, { "grad_norm": 0.12716400623321533, "learning_rate": 9.447690545632544e-05, "loss": 0.0261, "step": 7740 }, { "grad_norm": 0.11395720392465591, "learning_rate": 9.445800511279058e-05, "loss": 0.027, "step": 7750 }, { "grad_norm": 0.11244361102581024, "learning_rate": 9.443907438262844e-05, "loss": 0.0274, "step": 7760 }, { "grad_norm": 0.11386767029762268, "learning_rate": 9.442011327877798e-05, "loss": 0.0235, "step": 7770 }, { "grad_norm": 0.13003621995449066, "learning_rate": 9.440112181419895e-05, "loss": 0.0296, "step": 7780 }, { "grad_norm": 0.13296537101268768, "learning_rate": 9.438210000187186e-05, "loss": 0.0274, "step": 7790 }, { "grad_norm": 0.1286478191614151, "learning_rate": 9.43630478547979e-05, "loss": 0.0282, "step": 7800 }, { "grad_norm": 0.11764876544475555, "learning_rate": 9.434396538599906e-05, "loss": 0.0307, "step": 7810 }, { "grad_norm": 0.14157810807228088, "learning_rate": 9.4324852608518e-05, "loss": 0.0317, "step": 7820 }, { "grad_norm": 0.11791733652353287, "learning_rate": 9.430570953541816e-05, "loss": 0.0341, "step": 7830 }, { "grad_norm": 0.20247730612754822, "learning_rate": 9.428653617978362e-05, "loss": 0.0297, "step": 7840 }, { "grad_norm": 0.1517757922410965, "learning_rate": 9.42673325547192e-05, "loss": 0.0286, "step": 7850 }, { "grad_norm": 0.163471058011055, "learning_rate": 9.42480986733504e-05, "loss": 0.0279, "step": 7860 }, { "grad_norm": 0.10889901220798492, "learning_rate": 9.422883454882338e-05, "loss": 0.0278, "step": 7870 }, { "grad_norm": 0.15095750987529755, "learning_rate": 9.4209540194305e-05, "loss": 0.0303, "step": 7880 }, { "grad_norm": 0.17609694600105286, "learning_rate": 9.419021562298278e-05, "loss": 0.0294, "step": 7890 }, { "grad_norm": 0.17305432260036469, "learning_rate": 9.417086084806486e-05, "loss": 0.0273, "step": 7900 }, { "grad_norm": 0.13353444635868073, "learning_rate": 9.415147588278005e-05, "loss": 0.0323, "step": 7910 }, { "grad_norm": 0.13176961243152618, "learning_rate": 9.413206074037781e-05, "loss": 0.028, "step": 7920 }, { "grad_norm": 0.15697400271892548, "learning_rate": 9.411261543412819e-05, "loss": 0.0288, "step": 7930 }, { "grad_norm": 0.10538794100284576, "learning_rate": 9.409313997732188e-05, "loss": 0.0284, "step": 7940 }, { "grad_norm": 0.12798792123794556, "learning_rate": 9.407363438327018e-05, "loss": 0.0344, "step": 7950 }, { "grad_norm": 0.1408148556947708, "learning_rate": 9.405409866530496e-05, "loss": 0.032, "step": 7960 }, { "grad_norm": 0.11852680891752243, "learning_rate": 9.40345328367787e-05, "loss": 0.0299, "step": 7970 }, { "grad_norm": 0.12036500871181488, "learning_rate": 9.401493691106446e-05, "loss": 0.0357, "step": 7980 }, { "grad_norm": 0.10441713780164719, "learning_rate": 9.39953109015559e-05, "loss": 0.0308, "step": 7990 }, { "grad_norm": 0.11159570515155792, "learning_rate": 9.397565482166718e-05, "loss": 0.0283, "step": 8000 }, { "grad_norm": 0.138644278049469, "learning_rate": 9.395596868483305e-05, "loss": 0.0314, "step": 8010 }, { "grad_norm": 0.1216169074177742, "learning_rate": 9.393625250450881e-05, "loss": 0.0311, "step": 8020 }, { "grad_norm": 0.14528360962867737, "learning_rate": 9.391650629417028e-05, "loss": 0.0293, "step": 8030 }, { "grad_norm": 0.12334354221820831, "learning_rate": 9.38967300673138e-05, "loss": 0.0289, "step": 8040 }, { "grad_norm": 0.09578478336334229, "learning_rate": 9.387692383745624e-05, "loss": 0.0269, "step": 8050 }, { "grad_norm": 0.11331698298454285, "learning_rate": 9.385708761813496e-05, "loss": 0.0288, "step": 8060 }, { "grad_norm": 0.12028075754642487, "learning_rate": 9.383722142290787e-05, "loss": 0.0272, "step": 8070 }, { "grad_norm": 0.10431206971406937, "learning_rate": 9.381732526535328e-05, "loss": 0.0312, "step": 8080 }, { "grad_norm": 0.09683418273925781, "learning_rate": 9.379739915907004e-05, "loss": 0.0261, "step": 8090 }, { "grad_norm": 0.10934101790189743, "learning_rate": 9.377744311767746e-05, "loss": 0.0268, "step": 8100 }, { "grad_norm": 0.10321927815675735, "learning_rate": 9.375745715481532e-05, "loss": 0.03, "step": 8110 }, { "grad_norm": 0.08758857101202011, "learning_rate": 9.373744128414382e-05, "loss": 0.0298, "step": 8120 }, { "grad_norm": 0.11339825391769409, "learning_rate": 9.371739551934359e-05, "loss": 0.0261, "step": 8130 }, { "grad_norm": 0.13457150757312775, "learning_rate": 9.36973198741158e-05, "loss": 0.0305, "step": 8140 }, { "grad_norm": 0.16288796067237854, "learning_rate": 9.367721436218189e-05, "loss": 0.0294, "step": 8150 }, { "grad_norm": 0.102225661277771, "learning_rate": 9.365707899728383e-05, "loss": 0.0279, "step": 8160 }, { "grad_norm": 0.09608439356088638, "learning_rate": 9.363691379318391e-05, "loss": 0.027, "step": 8170 }, { "grad_norm": 0.11881984025239944, "learning_rate": 9.361671876366491e-05, "loss": 0.0305, "step": 8180 }, { "grad_norm": 0.10668452829122543, "learning_rate": 9.35964939225299e-05, "loss": 0.0281, "step": 8190 }, { "grad_norm": 0.10614252835512161, "learning_rate": 9.35762392836024e-05, "loss": 0.0298, "step": 8200 }, { "grad_norm": 0.0864352434873581, "learning_rate": 9.355595486072624e-05, "loss": 0.0263, "step": 8210 }, { "grad_norm": 0.10655268281698227, "learning_rate": 9.353564066776563e-05, "loss": 0.0284, "step": 8220 }, { "grad_norm": 0.16699819266796112, "learning_rate": 9.351529671860516e-05, "loss": 0.0292, "step": 8230 }, { "grad_norm": 0.11768389493227005, "learning_rate": 9.349492302714969e-05, "loss": 0.0255, "step": 8240 }, { "grad_norm": 0.11024860292673111, "learning_rate": 9.347451960732447e-05, "loss": 0.0251, "step": 8250 }, { "grad_norm": 0.11358635127544403, "learning_rate": 9.345408647307506e-05, "loss": 0.025, "step": 8260 }, { "grad_norm": 0.1542224884033203, "learning_rate": 9.343362363836726e-05, "loss": 0.0333, "step": 8270 }, { "grad_norm": 0.08671465516090393, "learning_rate": 9.341313111718728e-05, "loss": 0.0251, "step": 8280 }, { "grad_norm": 0.13460762798786163, "learning_rate": 9.339260892354153e-05, "loss": 0.0269, "step": 8290 }, { "grad_norm": 0.11062833666801453, "learning_rate": 9.337205707145676e-05, "loss": 0.0251, "step": 8300 }, { "grad_norm": 0.18334470689296722, "learning_rate": 9.335147557497994e-05, "loss": 0.0316, "step": 8310 }, { "grad_norm": 0.11897672712802887, "learning_rate": 9.333086444817835e-05, "loss": 0.0275, "step": 8320 }, { "grad_norm": 0.11755608767271042, "learning_rate": 9.331022370513951e-05, "loss": 0.0282, "step": 8330 }, { "grad_norm": 0.11380372196435928, "learning_rate": 9.328955335997113e-05, "loss": 0.0258, "step": 8340 }, { "grad_norm": 0.14367012679576874, "learning_rate": 9.326885342680125e-05, "loss": 0.0298, "step": 8350 }, { "grad_norm": 0.11518554389476776, "learning_rate": 9.324812391977806e-05, "loss": 0.0278, "step": 8360 }, { "grad_norm": 0.10879780352115631, "learning_rate": 9.322736485306995e-05, "loss": 0.0227, "step": 8370 }, { "grad_norm": 0.11418196558952332, "learning_rate": 9.32065762408656e-05, "loss": 0.027, "step": 8380 }, { "grad_norm": 0.15387755632400513, "learning_rate": 9.318575809737377e-05, "loss": 0.0289, "step": 8390 }, { "grad_norm": 0.1359446495771408, "learning_rate": 9.316491043682351e-05, "loss": 0.026, "step": 8400 }, { "grad_norm": 0.15451884269714355, "learning_rate": 9.3144033273464e-05, "loss": 0.0279, "step": 8410 }, { "grad_norm": 0.10627060383558273, "learning_rate": 9.312312662156457e-05, "loss": 0.0261, "step": 8420 }, { "grad_norm": 0.1515205055475235, "learning_rate": 9.310219049541471e-05, "loss": 0.0253, "step": 8430 }, { "grad_norm": 0.11075027287006378, "learning_rate": 9.308122490932409e-05, "loss": 0.0264, "step": 8440 }, { "grad_norm": 0.09976351261138916, "learning_rate": 9.306022987762246e-05, "loss": 0.0276, "step": 8450 }, { "grad_norm": 0.1303195208311081, "learning_rate": 9.303920541465974e-05, "loss": 0.0239, "step": 8460 }, { "grad_norm": 0.10586152225732803, "learning_rate": 9.301815153480598e-05, "loss": 0.0266, "step": 8470 }, { "grad_norm": 0.13568517565727234, "learning_rate": 9.299706825245126e-05, "loss": 0.0317, "step": 8480 }, { "grad_norm": 0.133365198969841, "learning_rate": 9.297595558200583e-05, "loss": 0.0303, "step": 8490 }, { "grad_norm": 0.14718489348888397, "learning_rate": 9.295481353790001e-05, "loss": 0.0272, "step": 8500 }, { "grad_norm": 0.13845248520374298, "learning_rate": 9.293364213458417e-05, "loss": 0.0291, "step": 8510 }, { "grad_norm": 0.16029149293899536, "learning_rate": 9.291244138652877e-05, "loss": 0.0327, "step": 8520 }, { "grad_norm": 0.10453150421380997, "learning_rate": 9.289121130822431e-05, "loss": 0.0275, "step": 8530 }, { "grad_norm": 0.14703844487667084, "learning_rate": 9.286995191418137e-05, "loss": 0.0284, "step": 8540 }, { "grad_norm": 0.18353314697742462, "learning_rate": 9.284866321893056e-05, "loss": 0.0319, "step": 8550 }, { "grad_norm": 0.14856795966625214, "learning_rate": 9.282734523702244e-05, "loss": 0.0294, "step": 8560 }, { "grad_norm": 0.1483517438173294, "learning_rate": 9.280599798302771e-05, "loss": 0.0318, "step": 8570 }, { "grad_norm": 0.14491550624370575, "learning_rate": 9.278462147153699e-05, "loss": 0.0281, "step": 8580 }, { "grad_norm": 0.12586820125579834, "learning_rate": 9.276321571716094e-05, "loss": 0.0268, "step": 8590 }, { "grad_norm": 0.11629870533943176, "learning_rate": 9.274178073453017e-05, "loss": 0.0301, "step": 8600 }, { "grad_norm": 0.15247300267219543, "learning_rate": 9.27203165382953e-05, "loss": 0.0296, "step": 8610 }, { "grad_norm": 0.12051396816968918, "learning_rate": 9.26988231431269e-05, "loss": 0.0285, "step": 8620 }, { "grad_norm": 0.11124330013990402, "learning_rate": 9.267730056371551e-05, "loss": 0.0278, "step": 8630 }, { "grad_norm": 0.147922083735466, "learning_rate": 9.265574881477161e-05, "loss": 0.0309, "step": 8640 }, { "grad_norm": 0.10936543345451355, "learning_rate": 9.263416791102561e-05, "loss": 0.0315, "step": 8650 }, { "grad_norm": 0.12817972898483276, "learning_rate": 9.261255786722786e-05, "loss": 0.0293, "step": 8660 }, { "grad_norm": 0.16649557650089264, "learning_rate": 9.259091869814864e-05, "loss": 0.0285, "step": 8670 }, { "grad_norm": 0.10629771649837494, "learning_rate": 9.256925041857807e-05, "loss": 0.0324, "step": 8680 }, { "grad_norm": 0.09983040392398834, "learning_rate": 9.254755304332626e-05, "loss": 0.0301, "step": 8690 }, { "grad_norm": 0.12736079096794128, "learning_rate": 9.252582658722313e-05, "loss": 0.0268, "step": 8700 }, { "grad_norm": 0.13927336037158966, "learning_rate": 9.250407106511853e-05, "loss": 0.025, "step": 8710 }, { "grad_norm": 0.1597023606300354, "learning_rate": 9.248228649188215e-05, "loss": 0.0273, "step": 8720 }, { "grad_norm": 0.1411963254213333, "learning_rate": 9.246047288240354e-05, "loss": 0.0291, "step": 8730 }, { "grad_norm": 0.12721343338489532, "learning_rate": 9.243863025159208e-05, "loss": 0.0279, "step": 8740 }, { "grad_norm": 0.13299164175987244, "learning_rate": 9.241675861437704e-05, "loss": 0.0299, "step": 8750 }, { "grad_norm": 0.1497040092945099, "learning_rate": 9.239485798570742e-05, "loss": 0.0306, "step": 8760 }, { "grad_norm": 0.09520319104194641, "learning_rate": 9.237292838055215e-05, "loss": 0.0279, "step": 8770 }, { "grad_norm": 0.10331006348133087, "learning_rate": 9.235096981389988e-05, "loss": 0.0289, "step": 8780 }, { "grad_norm": 0.08652446419000626, "learning_rate": 9.232898230075909e-05, "loss": 0.0317, "step": 8790 }, { "grad_norm": 0.13334912061691284, "learning_rate": 9.230696585615801e-05, "loss": 0.0283, "step": 8800 }, { "grad_norm": 0.09357646852731705, "learning_rate": 9.228492049514473e-05, "loss": 0.0326, "step": 8810 }, { "grad_norm": 0.09967858344316483, "learning_rate": 9.2262846232787e-05, "loss": 0.0313, "step": 8820 }, { "grad_norm": 0.14853358268737793, "learning_rate": 9.224074308417238e-05, "loss": 0.0305, "step": 8830 }, { "grad_norm": 0.13425487279891968, "learning_rate": 9.221861106440818e-05, "loss": 0.0262, "step": 8840 }, { "grad_norm": 0.12662044167518616, "learning_rate": 9.219645018862142e-05, "loss": 0.0291, "step": 8850 }, { "grad_norm": 0.09747714549303055, "learning_rate": 9.217426047195882e-05, "loss": 0.0286, "step": 8860 }, { "grad_norm": 0.11085048317909241, "learning_rate": 9.21520419295869e-05, "loss": 0.028, "step": 8870 }, { "grad_norm": 0.17655669152736664, "learning_rate": 9.212979457669178e-05, "loss": 0.0289, "step": 8880 }, { "grad_norm": 0.11333200335502625, "learning_rate": 9.210751842847931e-05, "loss": 0.0259, "step": 8890 }, { "grad_norm": 0.1140684112906456, "learning_rate": 9.208521350017506e-05, "loss": 0.0294, "step": 8900 }, { "grad_norm": 0.13271313905715942, "learning_rate": 9.206287980702422e-05, "loss": 0.0275, "step": 8910 }, { "grad_norm": 0.109170101583004, "learning_rate": 9.204051736429166e-05, "loss": 0.0268, "step": 8920 }, { "grad_norm": 0.11512155085802078, "learning_rate": 9.20181261872619e-05, "loss": 0.0262, "step": 8930 }, { "grad_norm": 0.13126671314239502, "learning_rate": 9.19957062912391e-05, "loss": 0.0285, "step": 8940 }, { "grad_norm": 0.13014152646064758, "learning_rate": 9.197325769154704e-05, "loss": 0.0304, "step": 8950 }, { "grad_norm": 0.11316656321287155, "learning_rate": 9.195078040352914e-05, "loss": 0.0303, "step": 8960 }, { "grad_norm": 0.10034768283367157, "learning_rate": 9.192827444254839e-05, "loss": 0.0259, "step": 8970 }, { "grad_norm": 0.11259259283542633, "learning_rate": 9.190573982398744e-05, "loss": 0.0298, "step": 8980 }, { "grad_norm": 0.08610789477825165, "learning_rate": 9.188317656324845e-05, "loss": 0.0292, "step": 8990 }, { "grad_norm": 0.13704247772693634, "learning_rate": 9.186058467575322e-05, "loss": 0.0304, "step": 9000 }, { "grad_norm": 0.0978475883603096, "learning_rate": 9.183796417694309e-05, "loss": 0.0258, "step": 9010 }, { "grad_norm": 0.11762356758117676, "learning_rate": 9.181531508227897e-05, "loss": 0.0243, "step": 9020 }, { "grad_norm": 0.10183471441268921, "learning_rate": 9.179263740724125e-05, "loss": 0.0259, "step": 9030 }, { "grad_norm": 0.1156647577881813, "learning_rate": 9.176993116732996e-05, "loss": 0.0268, "step": 9040 }, { "grad_norm": 0.09210091829299927, "learning_rate": 9.17471963780646e-05, "loss": 0.0271, "step": 9050 }, { "grad_norm": 0.12988175451755524, "learning_rate": 9.172443305498414e-05, "loss": 0.0281, "step": 9060 }, { "grad_norm": 0.1413503885269165, "learning_rate": 9.170164121364714e-05, "loss": 0.0274, "step": 9070 }, { "grad_norm": 0.10808489471673965, "learning_rate": 9.167882086963158e-05, "loss": 0.0261, "step": 9080 }, { "grad_norm": 0.10120030492544174, "learning_rate": 9.165597203853499e-05, "loss": 0.0254, "step": 9090 }, { "grad_norm": 0.11393408477306366, "learning_rate": 9.163309473597428e-05, "loss": 0.0287, "step": 9100 }, { "grad_norm": 0.09172583371400833, "learning_rate": 9.161018897758591e-05, "loss": 0.0293, "step": 9110 }, { "grad_norm": 0.12477698922157288, "learning_rate": 9.158725477902572e-05, "loss": 0.0298, "step": 9120 }, { "grad_norm": 0.10048551857471466, "learning_rate": 9.156429215596905e-05, "loss": 0.0235, "step": 9130 }, { "grad_norm": 0.14396733045578003, "learning_rate": 9.154130112411061e-05, "loss": 0.0307, "step": 9140 }, { "grad_norm": 0.13169224560260773, "learning_rate": 9.151828169916455e-05, "loss": 0.0327, "step": 9150 }, { "grad_norm": 0.14610837399959564, "learning_rate": 9.149523389686445e-05, "loss": 0.0298, "step": 9160 }, { "grad_norm": 0.1812904179096222, "learning_rate": 9.147215773296327e-05, "loss": 0.0318, "step": 9170 }, { "grad_norm": 0.14370614290237427, "learning_rate": 9.144905322323334e-05, "loss": 0.0294, "step": 9180 }, { "grad_norm": 0.10460414737462997, "learning_rate": 9.142592038346635e-05, "loss": 0.0273, "step": 9190 }, { "grad_norm": 0.11263073235750198, "learning_rate": 9.140275922947343e-05, "loss": 0.0298, "step": 9200 }, { "grad_norm": 0.1396847814321518, "learning_rate": 9.137956977708496e-05, "loss": 0.0268, "step": 9210 }, { "grad_norm": 0.12378142029047012, "learning_rate": 9.135635204215075e-05, "loss": 0.0319, "step": 9220 }, { "grad_norm": 0.10785949975252151, "learning_rate": 9.133310604053986e-05, "loss": 0.0303, "step": 9230 }, { "grad_norm": 0.12469977140426636, "learning_rate": 9.130983178814077e-05, "loss": 0.028, "step": 9240 }, { "grad_norm": 0.13240855932235718, "learning_rate": 9.128652930086116e-05, "loss": 0.0288, "step": 9250 }, { "grad_norm": 0.10262436419725418, "learning_rate": 9.126319859462808e-05, "loss": 0.0286, "step": 9260 }, { "grad_norm": 0.11222714185714722, "learning_rate": 9.123983968538787e-05, "loss": 0.032, "step": 9270 }, { "grad_norm": 0.16608662903308868, "learning_rate": 9.121645258910609e-05, "loss": 0.031, "step": 9280 }, { "grad_norm": 0.17832982540130615, "learning_rate": 9.119303732176762e-05, "loss": 0.0271, "step": 9290 }, { "grad_norm": 0.12248474359512329, "learning_rate": 9.116959389937656e-05, "loss": 0.0307, "step": 9300 }, { "grad_norm": 0.1329563409090042, "learning_rate": 9.114612233795628e-05, "loss": 0.0272, "step": 9310 }, { "grad_norm": 0.1162232980132103, "learning_rate": 9.112262265354937e-05, "loss": 0.0305, "step": 9320 }, { "grad_norm": 0.16620779037475586, "learning_rate": 9.109909486221763e-05, "loss": 0.0319, "step": 9330 }, { "grad_norm": 0.1188092976808548, "learning_rate": 9.107553898004208e-05, "loss": 0.0284, "step": 9340 }, { "grad_norm": 0.12872754037380219, "learning_rate": 9.105195502312298e-05, "loss": 0.0258, "step": 9350 }, { "grad_norm": 0.14018313586711884, "learning_rate": 9.102834300757972e-05, "loss": 0.033, "step": 9360 }, { "grad_norm": 0.1571597456932068, "learning_rate": 9.100470294955087e-05, "loss": 0.0323, "step": 9370 }, { "grad_norm": 0.1739126741886139, "learning_rate": 9.098103486519423e-05, "loss": 0.0326, "step": 9380 }, { "grad_norm": 0.18614134192466736, "learning_rate": 9.095733877068666e-05, "loss": 0.0302, "step": 9390 }, { "grad_norm": 0.13503630459308624, "learning_rate": 9.093361468222426e-05, "loss": 0.0292, "step": 9400 }, { "grad_norm": 0.14536818861961365, "learning_rate": 9.090986261602221e-05, "loss": 0.0293, "step": 9410 }, { "grad_norm": 0.15094269812107086, "learning_rate": 9.088608258831482e-05, "loss": 0.0293, "step": 9420 }, { "grad_norm": 0.14773963391780853, "learning_rate": 9.08622746153555e-05, "loss": 0.0292, "step": 9430 }, { "grad_norm": 0.11252889037132263, "learning_rate": 9.083843871341679e-05, "loss": 0.034, "step": 9440 }, { "grad_norm": 0.15879718959331512, "learning_rate": 9.08145748987903e-05, "loss": 0.0284, "step": 9450 }, { "grad_norm": 0.10357075929641724, "learning_rate": 9.079068318778674e-05, "loss": 0.0291, "step": 9460 }, { "grad_norm": 0.09071110934019089, "learning_rate": 9.076676359673586e-05, "loss": 0.029, "step": 9470 }, { "grad_norm": 0.09761059284210205, "learning_rate": 9.074281614198645e-05, "loss": 0.0276, "step": 9480 }, { "grad_norm": 0.16052964329719543, "learning_rate": 9.071884083990638e-05, "loss": 0.0304, "step": 9490 }, { "grad_norm": 0.12608368694782257, "learning_rate": 9.069483770688259e-05, "loss": 0.0305, "step": 9500 }, { "grad_norm": 0.11651375889778137, "learning_rate": 9.067080675932093e-05, "loss": 0.027, "step": 9510 }, { "grad_norm": 0.12461875379085541, "learning_rate": 9.064674801364634e-05, "loss": 0.0315, "step": 9520 }, { "grad_norm": 0.11236146837472916, "learning_rate": 9.062266148630279e-05, "loss": 0.0299, "step": 9530 }, { "grad_norm": 0.15410061180591583, "learning_rate": 9.059854719375313e-05, "loss": 0.032, "step": 9540 }, { "grad_norm": 0.12623871862888336, "learning_rate": 9.057440515247932e-05, "loss": 0.0313, "step": 9550 }, { "grad_norm": 0.12670353055000305, "learning_rate": 9.055023537898215e-05, "loss": 0.0298, "step": 9560 }, { "grad_norm": 0.18570837378501892, "learning_rate": 9.052603788978149e-05, "loss": 0.0305, "step": 9570 }, { "grad_norm": 0.1415613740682602, "learning_rate": 9.050181270141606e-05, "loss": 0.0295, "step": 9580 }, { "grad_norm": 0.09376677870750427, "learning_rate": 9.047755983044355e-05, "loss": 0.0276, "step": 9590 }, { "grad_norm": 0.1493312269449234, "learning_rate": 9.045327929344058e-05, "loss": 0.0297, "step": 9600 }, { "grad_norm": 0.17793889343738556, "learning_rate": 9.042897110700268e-05, "loss": 0.0339, "step": 9610 }, { "grad_norm": 0.10527992248535156, "learning_rate": 9.040463528774423e-05, "loss": 0.0319, "step": 9620 }, { "grad_norm": 0.16174305975437164, "learning_rate": 9.038027185229856e-05, "loss": 0.032, "step": 9630 }, { "grad_norm": 0.1338992714881897, "learning_rate": 9.035588081731784e-05, "loss": 0.0284, "step": 9640 }, { "grad_norm": 0.14409905672073364, "learning_rate": 9.033146219947313e-05, "loss": 0.0334, "step": 9650 }, { "grad_norm": 0.17476682364940643, "learning_rate": 9.030701601545429e-05, "loss": 0.0308, "step": 9660 }, { "grad_norm": 0.10657326877117157, "learning_rate": 9.028254228197009e-05, "loss": 0.0319, "step": 9670 }, { "grad_norm": 0.18673844635486603, "learning_rate": 9.025804101574808e-05, "loss": 0.0302, "step": 9680 }, { "grad_norm": 0.14986000955104828, "learning_rate": 9.023351223353463e-05, "loss": 0.0287, "step": 9690 }, { "grad_norm": 0.09800340235233307, "learning_rate": 9.020895595209498e-05, "loss": 0.0274, "step": 9700 }, { "grad_norm": 0.12692870199680328, "learning_rate": 9.018437218821305e-05, "loss": 0.032, "step": 9710 }, { "grad_norm": 0.15728728473186493, "learning_rate": 9.015976095869166e-05, "loss": 0.0327, "step": 9720 }, { "grad_norm": 0.15814200043678284, "learning_rate": 9.013512228035235e-05, "loss": 0.0306, "step": 9730 }, { "grad_norm": 0.1243121325969696, "learning_rate": 9.01104561700354e-05, "loss": 0.0297, "step": 9740 }, { "grad_norm": 0.13046765327453613, "learning_rate": 9.008576264459987e-05, "loss": 0.0304, "step": 9750 }, { "grad_norm": 0.12845183908939362, "learning_rate": 9.006104172092357e-05, "loss": 0.0273, "step": 9760 }, { "grad_norm": 0.13743163645267487, "learning_rate": 9.0036293415903e-05, "loss": 0.0284, "step": 9770 }, { "grad_norm": 0.13397406041622162, "learning_rate": 9.00115177464534e-05, "loss": 0.0269, "step": 9780 }, { "grad_norm": 0.1103326752781868, "learning_rate": 8.998671472950872e-05, "loss": 0.031, "step": 9790 }, { "grad_norm": 0.1151590645313263, "learning_rate": 8.996188438202156e-05, "loss": 0.026, "step": 9800 }, { "grad_norm": 0.14763426780700684, "learning_rate": 8.993702672096324e-05, "loss": 0.0308, "step": 9810 }, { "grad_norm": 0.2143857628107071, "learning_rate": 8.991214176332377e-05, "loss": 0.0272, "step": 9820 }, { "grad_norm": 0.15543223917484283, "learning_rate": 8.988722952611171e-05, "loss": 0.033, "step": 9830 }, { "grad_norm": 0.1862386018037796, "learning_rate": 8.986229002635441e-05, "loss": 0.0315, "step": 9840 }, { "grad_norm": 0.15207436680793762, "learning_rate": 8.983732328109775e-05, "loss": 0.0314, "step": 9850 }, { "grad_norm": 0.11433095484972, "learning_rate": 8.981232930740627e-05, "loss": 0.035, "step": 9860 }, { "grad_norm": 0.16890069842338562, "learning_rate": 8.978730812236313e-05, "loss": 0.0291, "step": 9870 }, { "grad_norm": 0.15518265962600708, "learning_rate": 8.976225974307004e-05, "loss": 0.0298, "step": 9880 }, { "grad_norm": 0.1265624314546585, "learning_rate": 8.973718418664736e-05, "loss": 0.0341, "step": 9890 }, { "grad_norm": 0.16246525943279266, "learning_rate": 8.971208147023397e-05, "loss": 0.0304, "step": 9900 }, { "grad_norm": 0.16697116196155548, "learning_rate": 8.968695161098739e-05, "loss": 0.0307, "step": 9910 }, { "grad_norm": 0.1343199908733368, "learning_rate": 8.96617946260836e-05, "loss": 0.0313, "step": 9920 }, { "grad_norm": 0.11623886227607727, "learning_rate": 8.963661053271716e-05, "loss": 0.0279, "step": 9930 }, { "grad_norm": 0.12943261861801147, "learning_rate": 8.961139934810117e-05, "loss": 0.0353, "step": 9940 }, { "grad_norm": 0.16461610794067383, "learning_rate": 8.958616108946725e-05, "loss": 0.0327, "step": 9950 }, { "grad_norm": 0.0896390751004219, "learning_rate": 8.95608957740655e-05, "loss": 0.028, "step": 9960 }, { "grad_norm": 0.09333761036396027, "learning_rate": 8.953560341916452e-05, "loss": 0.0309, "step": 9970 }, { "grad_norm": 0.09899377822875977, "learning_rate": 8.95102840420514e-05, "loss": 0.028, "step": 9980 }, { "grad_norm": 0.1221500039100647, "learning_rate": 8.94849376600317e-05, "loss": 0.0289, "step": 9990 }, { "grad_norm": 0.0920102596282959, "learning_rate": 8.945956429042943e-05, "loss": 0.0304, "step": 10000 }, { "grad_norm": 0.09774758666753769, "learning_rate": 8.943416395058705e-05, "loss": 0.0303, "step": 10010 }, { "grad_norm": 0.11303262412548065, "learning_rate": 8.940873665786544e-05, "loss": 0.0299, "step": 10020 }, { "grad_norm": 0.13628514111042023, "learning_rate": 8.938328242964394e-05, "loss": 0.0293, "step": 10030 }, { "grad_norm": 0.12333880364894867, "learning_rate": 8.935780128332026e-05, "loss": 0.0338, "step": 10040 }, { "grad_norm": 0.11302664130926132, "learning_rate": 8.933229323631052e-05, "loss": 0.0271, "step": 10050 }, { "grad_norm": 0.13209666311740875, "learning_rate": 8.930675830604925e-05, "loss": 0.0326, "step": 10060 }, { "grad_norm": 0.1597147136926651, "learning_rate": 8.92811965099893e-05, "loss": 0.0297, "step": 10070 }, { "grad_norm": 0.10656121373176575, "learning_rate": 8.925560786560194e-05, "loss": 0.0295, "step": 10080 }, { "grad_norm": 0.11009500920772552, "learning_rate": 8.922999239037677e-05, "loss": 0.0274, "step": 10090 }, { "grad_norm": 0.1407836228609085, "learning_rate": 8.920435010182171e-05, "loss": 0.0311, "step": 10100 }, { "grad_norm": 0.10458751767873764, "learning_rate": 8.917868101746302e-05, "loss": 0.0298, "step": 10110 }, { "grad_norm": 0.093848817050457, "learning_rate": 8.91529851548453e-05, "loss": 0.0287, "step": 10120 }, { "grad_norm": 0.11282460391521454, "learning_rate": 8.912726253153142e-05, "loss": 0.0274, "step": 10130 }, { "grad_norm": 0.10938228666782379, "learning_rate": 8.910151316510255e-05, "loss": 0.0284, "step": 10140 }, { "grad_norm": 0.10548616945743561, "learning_rate": 8.907573707315813e-05, "loss": 0.0285, "step": 10150 }, { "grad_norm": 0.15164323151111603, "learning_rate": 8.904993427331588e-05, "loss": 0.0286, "step": 10160 }, { "grad_norm": 0.140751913189888, "learning_rate": 8.902410478321176e-05, "loss": 0.0277, "step": 10170 }, { "grad_norm": 0.15874651074409485, "learning_rate": 8.899824862050002e-05, "loss": 0.0301, "step": 10180 }, { "grad_norm": 0.12004263699054718, "learning_rate": 8.897236580285308e-05, "loss": 0.0279, "step": 10190 }, { "grad_norm": 0.13879628479480743, "learning_rate": 8.894645634796159e-05, "loss": 0.03, "step": 10200 }, { "grad_norm": 0.13652142882347107, "learning_rate": 8.892052027353444e-05, "loss": 0.031, "step": 10210 }, { "grad_norm": 0.12391958385705948, "learning_rate": 8.889455759729866e-05, "loss": 0.0271, "step": 10220 }, { "grad_norm": 0.12343933433294296, "learning_rate": 8.886856833699955e-05, "loss": 0.0297, "step": 10230 }, { "grad_norm": 0.11174112558364868, "learning_rate": 8.884255251040046e-05, "loss": 0.0273, "step": 10240 }, { "grad_norm": 0.08928477764129639, "learning_rate": 8.8816510135283e-05, "loss": 0.0273, "step": 10250 }, { "grad_norm": 0.09565801173448563, "learning_rate": 8.879044122944688e-05, "loss": 0.0298, "step": 10260 }, { "grad_norm": 0.07418955862522125, "learning_rate": 8.876434581070996e-05, "loss": 0.0283, "step": 10270 }, { "grad_norm": 0.1417301446199417, "learning_rate": 8.87382238969082e-05, "loss": 0.0355, "step": 10280 }, { "grad_norm": 0.14090022444725037, "learning_rate": 8.871207550589568e-05, "loss": 0.0316, "step": 10290 }, { "grad_norm": 0.10192500799894333, "learning_rate": 8.868590065554458e-05, "loss": 0.0276, "step": 10300 }, { "grad_norm": 0.09102647006511688, "learning_rate": 8.865969936374519e-05, "loss": 0.0285, "step": 10310 }, { "grad_norm": 0.1585114598274231, "learning_rate": 8.863347164840581e-05, "loss": 0.0313, "step": 10320 }, { "grad_norm": 0.13030043244361877, "learning_rate": 8.860721752745285e-05, "loss": 0.0263, "step": 10330 }, { "grad_norm": 0.11678861081600189, "learning_rate": 8.858093701883077e-05, "loss": 0.0271, "step": 10340 }, { "grad_norm": 0.12690754234790802, "learning_rate": 8.8554630140502e-05, "loss": 0.0268, "step": 10350 }, { "grad_norm": 0.1457594484090805, "learning_rate": 8.85282969104471e-05, "loss": 0.0251, "step": 10360 }, { "grad_norm": 0.16148334741592407, "learning_rate": 8.850193734666456e-05, "loss": 0.0306, "step": 10370 }, { "grad_norm": 0.13238495588302612, "learning_rate": 8.84755514671709e-05, "loss": 0.028, "step": 10380 }, { "grad_norm": 0.11339923739433289, "learning_rate": 8.84491392900006e-05, "loss": 0.0261, "step": 10390 }, { "grad_norm": 0.09375744313001633, "learning_rate": 8.842270083320617e-05, "loss": 0.0279, "step": 10400 }, { "grad_norm": 0.10928322374820709, "learning_rate": 8.839623611485801e-05, "loss": 0.0295, "step": 10410 }, { "grad_norm": 0.1289799064397812, "learning_rate": 8.836974515304453e-05, "loss": 0.0293, "step": 10420 }, { "grad_norm": 0.11883456259965897, "learning_rate": 8.834322796587204e-05, "loss": 0.0286, "step": 10430 }, { "grad_norm": 0.11875855922698975, "learning_rate": 8.831668457146478e-05, "loss": 0.0303, "step": 10440 }, { "grad_norm": 0.12232165783643723, "learning_rate": 8.829011498796493e-05, "loss": 0.0305, "step": 10450 }, { "grad_norm": 0.11204063892364502, "learning_rate": 8.826351923353253e-05, "loss": 0.0251, "step": 10460 }, { "grad_norm": 0.10149935632944107, "learning_rate": 8.823689732634555e-05, "loss": 0.0299, "step": 10470 }, { "grad_norm": 0.16586104035377502, "learning_rate": 8.82102492845998e-05, "loss": 0.0295, "step": 10480 }, { "grad_norm": 0.14495015144348145, "learning_rate": 8.818357512650896e-05, "loss": 0.0273, "step": 10490 }, { "grad_norm": 0.11896724998950958, "learning_rate": 8.815687487030458e-05, "loss": 0.0267, "step": 10500 }, { "grad_norm": 0.09329934418201447, "learning_rate": 8.8130148534236e-05, "loss": 0.0293, "step": 10510 }, { "grad_norm": 0.14720197021961212, "learning_rate": 8.810339613657047e-05, "loss": 0.0314, "step": 10520 }, { "grad_norm": 0.14977823197841644, "learning_rate": 8.807661769559295e-05, "loss": 0.0301, "step": 10530 }, { "grad_norm": 0.1914760321378708, "learning_rate": 8.804981322960628e-05, "loss": 0.0275, "step": 10540 }, { "grad_norm": 0.1514579951763153, "learning_rate": 8.802298275693106e-05, "loss": 0.0277, "step": 10550 }, { "grad_norm": 0.12725219130516052, "learning_rate": 8.799612629590568e-05, "loss": 0.0283, "step": 10560 }, { "grad_norm": 0.12177732586860657, "learning_rate": 8.796924386488624e-05, "loss": 0.0254, "step": 10570 }, { "grad_norm": 0.09075094014406204, "learning_rate": 8.794233548224666e-05, "loss": 0.0229, "step": 10580 }, { "grad_norm": 0.09787695109844208, "learning_rate": 8.791540116637853e-05, "loss": 0.0257, "step": 10590 }, { "grad_norm": 0.11189550161361694, "learning_rate": 8.788844093569124e-05, "loss": 0.0259, "step": 10600 }, { "grad_norm": 0.10586120188236237, "learning_rate": 8.786145480861184e-05, "loss": 0.0254, "step": 10610 }, { "grad_norm": 0.06435128301382065, "learning_rate": 8.783444280358507e-05, "loss": 0.0252, "step": 10620 }, { "grad_norm": 0.09928411990404129, "learning_rate": 8.780740493907342e-05, "loss": 0.0263, "step": 10630 }, { "grad_norm": 0.1287623792886734, "learning_rate": 8.778034123355698e-05, "loss": 0.0244, "step": 10640 }, { "grad_norm": 0.13821259140968323, "learning_rate": 8.775325170553357e-05, "loss": 0.0228, "step": 10650 }, { "grad_norm": 0.09778264909982681, "learning_rate": 8.77261363735186e-05, "loss": 0.0248, "step": 10660 }, { "grad_norm": 0.08840062469244003, "learning_rate": 8.769899525604517e-05, "loss": 0.0243, "step": 10670 }, { "grad_norm": 0.10143674910068512, "learning_rate": 8.767182837166397e-05, "loss": 0.0273, "step": 10680 }, { "grad_norm": 0.08723995834589005, "learning_rate": 8.764463573894328e-05, "loss": 0.0242, "step": 10690 }, { "grad_norm": 0.08968766033649445, "learning_rate": 8.761741737646902e-05, "loss": 0.0232, "step": 10700 }, { "grad_norm": 0.13560940325260162, "learning_rate": 8.759017330284471e-05, "loss": 0.0289, "step": 10710 }, { "grad_norm": 0.0807248130440712, "learning_rate": 8.756290353669142e-05, "loss": 0.0224, "step": 10720 }, { "grad_norm": 0.11646415293216705, "learning_rate": 8.753560809664774e-05, "loss": 0.0236, "step": 10730 }, { "grad_norm": 0.08855510503053665, "learning_rate": 8.750828700136986e-05, "loss": 0.0246, "step": 10740 }, { "grad_norm": 0.10222567617893219, "learning_rate": 8.74809402695315e-05, "loss": 0.0229, "step": 10750 }, { "grad_norm": 0.11251621693372726, "learning_rate": 8.745356791982391e-05, "loss": 0.025, "step": 10760 }, { "grad_norm": 0.08816979825496674, "learning_rate": 8.742616997095578e-05, "loss": 0.025, "step": 10770 }, { "grad_norm": 0.12043395638465881, "learning_rate": 8.739874644165341e-05, "loss": 0.0255, "step": 10780 }, { "grad_norm": 0.12305625528097153, "learning_rate": 8.737129735066048e-05, "loss": 0.0286, "step": 10790 }, { "grad_norm": 0.11573746055364609, "learning_rate": 8.734382271673821e-05, "loss": 0.0276, "step": 10800 }, { "grad_norm": 0.13614463806152344, "learning_rate": 8.731632255866525e-05, "loss": 0.0254, "step": 10810 }, { "grad_norm": 0.13069072365760803, "learning_rate": 8.728879689523767e-05, "loss": 0.0286, "step": 10820 }, { "grad_norm": 0.1296105533838272, "learning_rate": 8.726124574526905e-05, "loss": 0.0285, "step": 10830 }, { "grad_norm": 0.1450974941253662, "learning_rate": 8.72336691275903e-05, "loss": 0.0274, "step": 10840 }, { "grad_norm": 0.10583601891994476, "learning_rate": 8.720606706104979e-05, "loss": 0.0261, "step": 10850 }, { "grad_norm": 0.12349394708871841, "learning_rate": 8.71784395645133e-05, "loss": 0.0274, "step": 10860 }, { "grad_norm": 0.12830063700675964, "learning_rate": 8.715078665686392e-05, "loss": 0.0257, "step": 10870 }, { "grad_norm": 0.12938740849494934, "learning_rate": 8.712310835700218e-05, "loss": 0.0264, "step": 10880 }, { "grad_norm": 0.12112794071435928, "learning_rate": 8.709540468384591e-05, "loss": 0.0256, "step": 10890 }, { "grad_norm": 0.11775937676429749, "learning_rate": 8.706767565633033e-05, "loss": 0.0253, "step": 10900 }, { "grad_norm": 0.0879637598991394, "learning_rate": 8.7039921293408e-05, "loss": 0.0247, "step": 10910 }, { "grad_norm": 0.10105428844690323, "learning_rate": 8.70121416140487e-05, "loss": 0.0242, "step": 10920 }, { "grad_norm": 0.1275494545698166, "learning_rate": 8.698433663723962e-05, "loss": 0.0291, "step": 10930 }, { "grad_norm": 0.105502188205719, "learning_rate": 8.695650638198518e-05, "loss": 0.0253, "step": 10940 }, { "grad_norm": 0.09866601973772049, "learning_rate": 8.692865086730713e-05, "loss": 0.0282, "step": 10950 }, { "grad_norm": 0.1317633092403412, "learning_rate": 8.69007701122444e-05, "loss": 0.0261, "step": 10960 }, { "grad_norm": 0.11926168203353882, "learning_rate": 8.687286413585328e-05, "loss": 0.0262, "step": 10970 }, { "grad_norm": 0.10937410593032837, "learning_rate": 8.684493295720719e-05, "loss": 0.0286, "step": 10980 }, { "grad_norm": 0.12076296657323837, "learning_rate": 8.681697659539685e-05, "loss": 0.0285, "step": 10990 }, { "grad_norm": 0.10361894965171814, "learning_rate": 8.678899506953019e-05, "loss": 0.0343, "step": 11000 }, { "grad_norm": 0.13494662940502167, "learning_rate": 8.676098839873227e-05, "loss": 0.0263, "step": 11010 }, { "grad_norm": 0.1324995756149292, "learning_rate": 8.673295660214545e-05, "loss": 0.0284, "step": 11020 }, { "grad_norm": 0.1411599963903427, "learning_rate": 8.670489969892914e-05, "loss": 0.0316, "step": 11030 }, { "grad_norm": 0.10510364174842834, "learning_rate": 8.667681770826e-05, "loss": 0.0237, "step": 11040 }, { "grad_norm": 0.1302141696214676, "learning_rate": 8.66487106493318e-05, "loss": 0.0275, "step": 11050 }, { "grad_norm": 0.10055053234100342, "learning_rate": 8.662057854135544e-05, "loss": 0.0239, "step": 11060 }, { "grad_norm": 0.1396503746509552, "learning_rate": 8.659242140355897e-05, "loss": 0.0308, "step": 11070 }, { "grad_norm": 0.16398312151432037, "learning_rate": 8.65642392551875e-05, "loss": 0.0292, "step": 11080 }, { "grad_norm": 0.11197210103273392, "learning_rate": 8.65360321155033e-05, "loss": 0.0263, "step": 11090 }, { "grad_norm": 0.15795621275901794, "learning_rate": 8.650780000378566e-05, "loss": 0.0288, "step": 11100 }, { "grad_norm": 0.18755023181438446, "learning_rate": 8.647954293933096e-05, "loss": 0.0321, "step": 11110 }, { "grad_norm": 0.16239574551582336, "learning_rate": 8.645126094145264e-05, "loss": 0.03, "step": 11120 }, { "grad_norm": 0.15702784061431885, "learning_rate": 8.642295402948117e-05, "loss": 0.0273, "step": 11130 }, { "grad_norm": 0.15303604304790497, "learning_rate": 8.639462222276409e-05, "loss": 0.0287, "step": 11140 }, { "grad_norm": 0.1544422060251236, "learning_rate": 8.636626554066589e-05, "loss": 0.0326, "step": 11150 }, { "grad_norm": 0.11928953975439072, "learning_rate": 8.633788400256811e-05, "loss": 0.0296, "step": 11160 }, { "grad_norm": 0.10586250573396683, "learning_rate": 8.630947762786927e-05, "loss": 0.0292, "step": 11170 }, { "grad_norm": 0.13495221734046936, "learning_rate": 8.628104643598483e-05, "loss": 0.0249, "step": 11180 }, { "grad_norm": 0.11171592026948929, "learning_rate": 8.625259044634726e-05, "loss": 0.0298, "step": 11190 }, { "grad_norm": 0.10776899009943008, "learning_rate": 8.622410967840597e-05, "loss": 0.0302, "step": 11200 }, { "grad_norm": 0.101872019469738, "learning_rate": 8.619560415162731e-05, "loss": 0.0303, "step": 11210 }, { "grad_norm": 0.1412193924188614, "learning_rate": 8.616707388549447e-05, "loss": 0.027, "step": 11220 }, { "grad_norm": 0.17112579941749573, "learning_rate": 8.613851889950771e-05, "loss": 0.0313, "step": 11230 }, { "grad_norm": 0.13293351233005524, "learning_rate": 8.610993921318402e-05, "loss": 0.0279, "step": 11240 }, { "grad_norm": 0.12579594552516937, "learning_rate": 8.608133484605738e-05, "loss": 0.0323, "step": 11250 }, { "grad_norm": 0.1258835792541504, "learning_rate": 8.605270581767859e-05, "loss": 0.0255, "step": 11260 }, { "grad_norm": 0.10394752025604248, "learning_rate": 8.602405214761536e-05, "loss": 0.0264, "step": 11270 }, { "grad_norm": 0.11075276136398315, "learning_rate": 8.599537385545215e-05, "loss": 0.0279, "step": 11280 }, { "grad_norm": 0.13081242144107819, "learning_rate": 8.596667096079032e-05, "loss": 0.0291, "step": 11290 }, { "grad_norm": 0.10411479324102402, "learning_rate": 8.593794348324806e-05, "loss": 0.0301, "step": 11300 }, { "grad_norm": 0.09157048165798187, "learning_rate": 8.590919144246028e-05, "loss": 0.0262, "step": 11310 }, { "grad_norm": 0.0835331603884697, "learning_rate": 8.588041485807876e-05, "loss": 0.0253, "step": 11320 }, { "grad_norm": 0.09728085249662399, "learning_rate": 8.585161374977202e-05, "loss": 0.0298, "step": 11330 }, { "grad_norm": 0.10579380393028259, "learning_rate": 8.582278813722533e-05, "loss": 0.0315, "step": 11340 }, { "grad_norm": 0.1192263662815094, "learning_rate": 8.579393804014076e-05, "loss": 0.0261, "step": 11350 }, { "grad_norm": 0.12249188870191574, "learning_rate": 8.576506347823703e-05, "loss": 0.0257, "step": 11360 }, { "grad_norm": 0.13902221620082855, "learning_rate": 8.573616447124968e-05, "loss": 0.0317, "step": 11370 }, { "grad_norm": 0.11127708107233047, "learning_rate": 8.570724103893086e-05, "loss": 0.0267, "step": 11380 }, { "grad_norm": 0.1125873401761055, "learning_rate": 8.567829320104951e-05, "loss": 0.0264, "step": 11390 }, { "grad_norm": 0.09423176199197769, "learning_rate": 8.564932097739118e-05, "loss": 0.0258, "step": 11400 }, { "grad_norm": 0.10721530765295029, "learning_rate": 8.562032438775811e-05, "loss": 0.0295, "step": 11410 }, { "grad_norm": 0.10512181371450424, "learning_rate": 8.559130345196921e-05, "loss": 0.0286, "step": 11420 }, { "grad_norm": 0.1399877667427063, "learning_rate": 8.556225818986e-05, "loss": 0.0256, "step": 11430 }, { "grad_norm": 0.0852486863732338, "learning_rate": 8.553318862128265e-05, "loss": 0.0268, "step": 11440 }, { "grad_norm": 0.12360449880361557, "learning_rate": 8.550409476610593e-05, "loss": 0.028, "step": 11450 }, { "grad_norm": 0.13153313100337982, "learning_rate": 8.547497664421522e-05, "loss": 0.0326, "step": 11460 }, { "grad_norm": 0.10298851877450943, "learning_rate": 8.544583427551249e-05, "loss": 0.0299, "step": 11470 }, { "grad_norm": 0.09018193930387497, "learning_rate": 8.541666767991628e-05, "loss": 0.0267, "step": 11480 }, { "grad_norm": 0.0811515599489212, "learning_rate": 8.538747687736166e-05, "loss": 0.0241, "step": 11490 }, { "grad_norm": 0.12566028535366058, "learning_rate": 8.535826188780026e-05, "loss": 0.0287, "step": 11500 }, { "grad_norm": 0.10339849442243576, "learning_rate": 8.532902273120029e-05, "loss": 0.026, "step": 11510 }, { "grad_norm": 0.09859991818666458, "learning_rate": 8.52997594275464e-05, "loss": 0.0261, "step": 11520 }, { "grad_norm": 0.13668465614318848, "learning_rate": 8.52704719968398e-05, "loss": 0.0276, "step": 11530 }, { "grad_norm": 0.1273835003376007, "learning_rate": 8.524116045909818e-05, "loss": 0.0262, "step": 11540 }, { "grad_norm": 0.10758445411920547, "learning_rate": 8.521182483435569e-05, "loss": 0.0287, "step": 11550 }, { "grad_norm": 0.12030965089797974, "learning_rate": 8.518246514266295e-05, "loss": 0.0296, "step": 11560 }, { "grad_norm": 0.15149202942848206, "learning_rate": 8.515308140408703e-05, "loss": 0.0277, "step": 11570 }, { "grad_norm": 0.10496194660663605, "learning_rate": 8.512367363871145e-05, "loss": 0.0252, "step": 11580 }, { "grad_norm": 0.09403809905052185, "learning_rate": 8.509424186663614e-05, "loss": 0.0293, "step": 11590 }, { "grad_norm": 0.1329479217529297, "learning_rate": 8.506478610797743e-05, "loss": 0.0275, "step": 11600 }, { "grad_norm": 0.10430848598480225, "learning_rate": 8.503530638286805e-05, "loss": 0.0261, "step": 11610 }, { "grad_norm": 0.11412658542394638, "learning_rate": 8.500580271145712e-05, "loss": 0.0293, "step": 11620 }, { "grad_norm": 0.08500842750072479, "learning_rate": 8.497627511391014e-05, "loss": 0.0243, "step": 11630 }, { "grad_norm": 0.09556792676448822, "learning_rate": 8.494672361040891e-05, "loss": 0.0276, "step": 11640 }, { "grad_norm": 0.09949646145105362, "learning_rate": 8.491714822115162e-05, "loss": 0.0245, "step": 11650 }, { "grad_norm": 0.10121586173772812, "learning_rate": 8.488754896635277e-05, "loss": 0.0272, "step": 11660 }, { "grad_norm": 0.11802008002996445, "learning_rate": 8.485792586624317e-05, "loss": 0.028, "step": 11670 }, { "grad_norm": 0.09627264738082886, "learning_rate": 8.482827894106993e-05, "loss": 0.0244, "step": 11680 }, { "grad_norm": 0.09086113423109055, "learning_rate": 8.479860821109646e-05, "loss": 0.0273, "step": 11690 }, { "grad_norm": 0.10557037591934204, "learning_rate": 8.476891369660239e-05, "loss": 0.0283, "step": 11700 }, { "grad_norm": 0.3574118912220001, "learning_rate": 8.473919541788366e-05, "loss": 0.0303, "step": 11710 }, { "grad_norm": 0.13737650215625763, "learning_rate": 8.470945339525245e-05, "loss": 0.028, "step": 11720 }, { "grad_norm": 0.10416090488433838, "learning_rate": 8.467968764903713e-05, "loss": 0.0285, "step": 11730 }, { "grad_norm": 0.09024567902088165, "learning_rate": 8.46498981995823e-05, "loss": 0.0275, "step": 11740 }, { "grad_norm": 0.09182926267385483, "learning_rate": 8.462008506724879e-05, "loss": 0.0261, "step": 11750 }, { "grad_norm": 0.11687996238470078, "learning_rate": 8.459024827241359e-05, "loss": 0.0275, "step": 11760 }, { "grad_norm": 0.08656053990125656, "learning_rate": 8.456038783546985e-05, "loss": 0.0242, "step": 11770 }, { "grad_norm": 0.08715243637561798, "learning_rate": 8.453050377682691e-05, "loss": 0.0243, "step": 11780 }, { "grad_norm": 0.09819867461919785, "learning_rate": 8.450059611691026e-05, "loss": 0.0282, "step": 11790 }, { "grad_norm": 0.13086619973182678, "learning_rate": 8.447066487616146e-05, "loss": 0.0274, "step": 11800 }, { "grad_norm": 0.12923115491867065, "learning_rate": 8.444071007503826e-05, "loss": 0.0262, "step": 11810 }, { "grad_norm": 0.14033889770507812, "learning_rate": 8.441073173401449e-05, "loss": 0.0294, "step": 11820 }, { "grad_norm": 0.10168444365262985, "learning_rate": 8.438072987358006e-05, "loss": 0.0274, "step": 11830 }, { "grad_norm": 0.15793858468532562, "learning_rate": 8.435070451424094e-05, "loss": 0.026, "step": 11840 }, { "grad_norm": 0.09237798303365707, "learning_rate": 8.432065567651919e-05, "loss": 0.0284, "step": 11850 }, { "grad_norm": 0.1204184740781784, "learning_rate": 8.429058338095291e-05, "loss": 0.032, "step": 11860 }, { "grad_norm": 0.11205077916383743, "learning_rate": 8.426048764809624e-05, "loss": 0.0299, "step": 11870 }, { "grad_norm": 0.16490775346755981, "learning_rate": 8.423036849851932e-05, "loss": 0.0272, "step": 11880 }, { "grad_norm": 0.18416950106620789, "learning_rate": 8.42002259528083e-05, "loss": 0.0278, "step": 11890 }, { "grad_norm": 0.13162963092327118, "learning_rate": 8.417006003156532e-05, "loss": 0.0309, "step": 11900 }, { "grad_norm": 0.1065639778971672, "learning_rate": 8.413987075540852e-05, "loss": 0.027, "step": 11910 }, { "grad_norm": 0.08841574192047119, "learning_rate": 8.4109658144972e-05, "loss": 0.0278, "step": 11920 }, { "grad_norm": 0.11059822142124176, "learning_rate": 8.407942222090573e-05, "loss": 0.0261, "step": 11930 }, { "grad_norm": 0.12176115810871124, "learning_rate": 8.404916300387576e-05, "loss": 0.0275, "step": 11940 }, { "grad_norm": 0.11335583031177521, "learning_rate": 8.401888051456391e-05, "loss": 0.0269, "step": 11950 }, { "grad_norm": 0.10670951753854752, "learning_rate": 8.398857477366803e-05, "loss": 0.0277, "step": 11960 }, { "grad_norm": 0.10592199116945267, "learning_rate": 8.395824580190178e-05, "loss": 0.0289, "step": 11970 }, { "grad_norm": 0.10642940551042557, "learning_rate": 8.392789361999473e-05, "loss": 0.0243, "step": 11980 }, { "grad_norm": 0.09331389516592026, "learning_rate": 8.38975182486923e-05, "loss": 0.0248, "step": 11990 }, { "grad_norm": 0.10257644206285477, "learning_rate": 8.386711970875581e-05, "loss": 0.0252, "step": 12000 }, { "grad_norm": 0.07946138083934784, "learning_rate": 8.383669802096232e-05, "loss": 0.0279, "step": 12010 }, { "grad_norm": 0.12431667745113373, "learning_rate": 8.38062532061048e-05, "loss": 0.0281, "step": 12020 }, { "grad_norm": 0.08463780581951141, "learning_rate": 8.3775785284992e-05, "loss": 0.0262, "step": 12030 }, { "grad_norm": 0.11661824584007263, "learning_rate": 8.374529427844843e-05, "loss": 0.0243, "step": 12040 }, { "grad_norm": 0.11770286411046982, "learning_rate": 8.371478020731442e-05, "loss": 0.0293, "step": 12050 }, { "grad_norm": 0.15496249496936798, "learning_rate": 8.368424309244607e-05, "loss": 0.0285, "step": 12060 }, { "grad_norm": 0.13822950422763824, "learning_rate": 8.365368295471517e-05, "loss": 0.0258, "step": 12070 }, { "grad_norm": 0.1142486035823822, "learning_rate": 8.362309981500931e-05, "loss": 0.0259, "step": 12080 }, { "grad_norm": 0.10223314166069031, "learning_rate": 8.359249369423177e-05, "loss": 0.0243, "step": 12090 }, { "grad_norm": 0.18112923204898834, "learning_rate": 8.356186461330155e-05, "loss": 0.0283, "step": 12100 }, { "grad_norm": 0.10166671127080917, "learning_rate": 8.353121259315334e-05, "loss": 0.0234, "step": 12110 }, { "grad_norm": 0.112753726541996, "learning_rate": 8.350053765473751e-05, "loss": 0.0272, "step": 12120 }, { "grad_norm": 0.09809798747301102, "learning_rate": 8.346983981902005e-05, "loss": 0.0236, "step": 12130 }, { "grad_norm": 0.11180871725082397, "learning_rate": 8.343911910698271e-05, "loss": 0.0249, "step": 12140 }, { "grad_norm": 0.13242974877357483, "learning_rate": 8.340837553962278e-05, "loss": 0.0263, "step": 12150 }, { "grad_norm": 0.13570313155651093, "learning_rate": 8.337760913795316e-05, "loss": 0.0308, "step": 12160 }, { "grad_norm": 0.09853776544332504, "learning_rate": 8.334681992300244e-05, "loss": 0.0259, "step": 12170 }, { "grad_norm": 0.09260690212249756, "learning_rate": 8.331600791581475e-05, "loss": 0.027, "step": 12180 }, { "grad_norm": 0.12133687734603882, "learning_rate": 8.328517313744978e-05, "loss": 0.0286, "step": 12190 }, { "grad_norm": 0.1221126914024353, "learning_rate": 8.325431560898286e-05, "loss": 0.0225, "step": 12200 }, { "grad_norm": 0.13530616462230682, "learning_rate": 8.322343535150478e-05, "loss": 0.0285, "step": 12210 }, { "grad_norm": 0.13567541539669037, "learning_rate": 8.319253238612191e-05, "loss": 0.027, "step": 12220 }, { "grad_norm": 0.14347043633460999, "learning_rate": 8.316160673395614e-05, "loss": 0.0301, "step": 12230 }, { "grad_norm": 0.15663965046405792, "learning_rate": 8.313065841614487e-05, "loss": 0.0261, "step": 12240 }, { "grad_norm": 0.1384703516960144, "learning_rate": 8.309968745384096e-05, "loss": 0.0294, "step": 12250 }, { "grad_norm": 0.12099269032478333, "learning_rate": 8.306869386821282e-05, "loss": 0.0302, "step": 12260 }, { "grad_norm": 0.1348883956670761, "learning_rate": 8.30376776804442e-05, "loss": 0.0299, "step": 12270 }, { "grad_norm": 0.13037745654582977, "learning_rate": 8.300663891173443e-05, "loss": 0.03, "step": 12280 }, { "grad_norm": 0.14016762375831604, "learning_rate": 8.297557758329822e-05, "loss": 0.0266, "step": 12290 }, { "grad_norm": 0.11520743370056152, "learning_rate": 8.294449371636564e-05, "loss": 0.0266, "step": 12300 }, { "grad_norm": 0.09655636548995972, "learning_rate": 8.291338733218226e-05, "loss": 0.0266, "step": 12310 }, { "grad_norm": 0.14001591503620148, "learning_rate": 8.2882258452009e-05, "loss": 0.0262, "step": 12320 }, { "grad_norm": 0.13079126179218292, "learning_rate": 8.285110709712214e-05, "loss": 0.0256, "step": 12330 }, { "grad_norm": 0.10937929153442383, "learning_rate": 8.281993328881337e-05, "loss": 0.0263, "step": 12340 }, { "grad_norm": 0.09535573422908783, "learning_rate": 8.278873704838964e-05, "loss": 0.0282, "step": 12350 }, { "grad_norm": 0.12233728915452957, "learning_rate": 8.275751839717334e-05, "loss": 0.0251, "step": 12360 }, { "grad_norm": 0.12376698106527328, "learning_rate": 8.272627735650208e-05, "loss": 0.0246, "step": 12370 }, { "grad_norm": 0.13301728665828705, "learning_rate": 8.269501394772884e-05, "loss": 0.0326, "step": 12380 }, { "grad_norm": 0.12227606773376465, "learning_rate": 8.266372819222189e-05, "loss": 0.0269, "step": 12390 }, { "grad_norm": 0.13097888231277466, "learning_rate": 8.26324201113647e-05, "loss": 0.0292, "step": 12400 }, { "grad_norm": 0.13451452553272247, "learning_rate": 8.260108972655606e-05, "loss": 0.0299, "step": 12410 }, { "grad_norm": 0.10584114491939545, "learning_rate": 8.256973705921e-05, "loss": 0.0285, "step": 12420 }, { "grad_norm": 0.11610804498195648, "learning_rate": 8.25383621307558e-05, "loss": 0.0271, "step": 12430 }, { "grad_norm": 0.1079142764210701, "learning_rate": 8.25069649626379e-05, "loss": 0.0276, "step": 12440 }, { "grad_norm": 0.10035239160060883, "learning_rate": 8.247554557631596e-05, "loss": 0.0247, "step": 12450 }, { "grad_norm": 0.12143909186124802, "learning_rate": 8.244410399326483e-05, "loss": 0.0283, "step": 12460 }, { "grad_norm": 0.0862002894282341, "learning_rate": 8.241264023497457e-05, "loss": 0.0284, "step": 12470 }, { "grad_norm": 0.12994226813316345, "learning_rate": 8.238115432295034e-05, "loss": 0.0286, "step": 12480 }, { "grad_norm": 0.10000920295715332, "learning_rate": 8.234964627871247e-05, "loss": 0.0264, "step": 12490 }, { "grad_norm": 0.11213166266679764, "learning_rate": 8.231811612379639e-05, "loss": 0.0237, "step": 12500 }, { "grad_norm": 0.13480135798454285, "learning_rate": 8.228656387975268e-05, "loss": 0.0226, "step": 12510 }, { "grad_norm": 0.12088841944932938, "learning_rate": 8.225498956814702e-05, "loss": 0.0262, "step": 12520 }, { "grad_norm": 0.1167679950594902, "learning_rate": 8.222339321056014e-05, "loss": 0.0271, "step": 12530 }, { "grad_norm": 0.09371544420719147, "learning_rate": 8.219177482858785e-05, "loss": 0.025, "step": 12540 }, { "grad_norm": 0.12131421267986298, "learning_rate": 8.216013444384099e-05, "loss": 0.027, "step": 12550 }, { "grad_norm": 0.12567122280597687, "learning_rate": 8.21284720779455e-05, "loss": 0.0226, "step": 12560 }, { "grad_norm": 0.11687181144952774, "learning_rate": 8.209678775254231e-05, "loss": 0.0302, "step": 12570 }, { "grad_norm": 0.1398753970861435, "learning_rate": 8.206508148928733e-05, "loss": 0.0274, "step": 12580 }, { "grad_norm": 0.10974021255970001, "learning_rate": 8.203335330985151e-05, "loss": 0.0254, "step": 12590 }, { "grad_norm": 0.1445857137441635, "learning_rate": 8.200160323592076e-05, "loss": 0.0276, "step": 12600 }, { "grad_norm": 0.11861418932676315, "learning_rate": 8.196983128919598e-05, "loss": 0.0244, "step": 12610 }, { "grad_norm": 0.0987526923418045, "learning_rate": 8.193803749139295e-05, "loss": 0.0249, "step": 12620 }, { "grad_norm": 0.15828171372413635, "learning_rate": 8.190622186424244e-05, "loss": 0.0256, "step": 12630 }, { "grad_norm": 0.11511032283306122, "learning_rate": 8.187438442949016e-05, "loss": 0.0252, "step": 12640 }, { "grad_norm": 0.09208083152770996, "learning_rate": 8.184252520889668e-05, "loss": 0.0259, "step": 12650 }, { "grad_norm": 0.11003265529870987, "learning_rate": 8.181064422423748e-05, "loss": 0.0262, "step": 12660 }, { "grad_norm": 0.12354826182126999, "learning_rate": 8.177874149730289e-05, "loss": 0.0248, "step": 12670 }, { "grad_norm": 0.130209282040596, "learning_rate": 8.174681704989816e-05, "loss": 0.0233, "step": 12680 }, { "grad_norm": 0.13967429101467133, "learning_rate": 8.171487090384333e-05, "loss": 0.0288, "step": 12690 }, { "grad_norm": 0.08943973481655121, "learning_rate": 8.168290308097328e-05, "loss": 0.0287, "step": 12700 }, { "grad_norm": 0.11740205436944962, "learning_rate": 8.165091360313774e-05, "loss": 0.0249, "step": 12710 }, { "grad_norm": 0.12962156534194946, "learning_rate": 8.161890249220119e-05, "loss": 0.0278, "step": 12720 }, { "grad_norm": 0.09686656296253204, "learning_rate": 8.158686977004295e-05, "loss": 0.028, "step": 12730 }, { "grad_norm": 0.12048652768135071, "learning_rate": 8.155481545855706e-05, "loss": 0.0283, "step": 12740 }, { "grad_norm": 0.1342381089925766, "learning_rate": 8.152273957965233e-05, "loss": 0.0303, "step": 12750 }, { "grad_norm": 0.16701871156692505, "learning_rate": 8.149064215525237e-05, "loss": 0.0295, "step": 12760 }, { "grad_norm": 0.09581652283668518, "learning_rate": 8.14585232072954e-05, "loss": 0.026, "step": 12770 }, { "grad_norm": 0.12107864767313004, "learning_rate": 8.142638275773449e-05, "loss": 0.0309, "step": 12780 }, { "grad_norm": 0.13336816430091858, "learning_rate": 8.139422082853729e-05, "loss": 0.0277, "step": 12790 }, { "grad_norm": 0.09560681134462357, "learning_rate": 8.136203744168618e-05, "loss": 0.0287, "step": 12800 }, { "grad_norm": 0.10610366612672806, "learning_rate": 8.132983261917819e-05, "loss": 0.0297, "step": 12810 }, { "grad_norm": 0.10889288783073425, "learning_rate": 8.129760638302504e-05, "loss": 0.0273, "step": 12820 }, { "grad_norm": 0.10946618020534515, "learning_rate": 8.126535875525305e-05, "loss": 0.0275, "step": 12830 }, { "grad_norm": 0.09785012900829315, "learning_rate": 8.123308975790316e-05, "loss": 0.0288, "step": 12840 }, { "grad_norm": 0.09029322862625122, "learning_rate": 8.120079941303094e-05, "loss": 0.0252, "step": 12850 }, { "grad_norm": 0.10563106089830399, "learning_rate": 8.116848774270651e-05, "loss": 0.0273, "step": 12860 }, { "grad_norm": 0.09118238091468811, "learning_rate": 8.113615476901461e-05, "loss": 0.0248, "step": 12870 }, { "grad_norm": 0.10910018533468246, "learning_rate": 8.110380051405454e-05, "loss": 0.0266, "step": 12880 }, { "grad_norm": 0.12143119424581528, "learning_rate": 8.107142499994009e-05, "loss": 0.0259, "step": 12890 }, { "grad_norm": 0.1135248914361, "learning_rate": 8.103902824879966e-05, "loss": 0.0247, "step": 12900 }, { "grad_norm": 0.08806417882442474, "learning_rate": 8.10066102827761e-05, "loss": 0.0252, "step": 12910 }, { "grad_norm": 0.08443781733512878, "learning_rate": 8.097417112402676e-05, "loss": 0.028, "step": 12920 }, { "grad_norm": 0.1049533262848854, "learning_rate": 8.094171079472355e-05, "loss": 0.0271, "step": 12930 }, { "grad_norm": 0.07345869392156601, "learning_rate": 8.090922931705277e-05, "loss": 0.0226, "step": 12940 }, { "grad_norm": 0.0968589261174202, "learning_rate": 8.08767267132152e-05, "loss": 0.026, "step": 12950 }, { "grad_norm": 0.12193305045366287, "learning_rate": 8.084420300542608e-05, "loss": 0.0253, "step": 12960 }, { "grad_norm": 0.13614873588085175, "learning_rate": 8.081165821591505e-05, "loss": 0.0279, "step": 12970 }, { "grad_norm": 0.12292055040597916, "learning_rate": 8.077909236692615e-05, "loss": 0.0268, "step": 12980 }, { "grad_norm": 0.13972660899162292, "learning_rate": 8.074650548071787e-05, "loss": 0.0264, "step": 12990 }, { "grad_norm": 0.12242110073566437, "learning_rate": 8.071389757956301e-05, "loss": 0.0276, "step": 13000 }, { "grad_norm": 0.11104562878608704, "learning_rate": 8.068126868574876e-05, "loss": 0.0292, "step": 13010 }, { "grad_norm": 0.08793208748102188, "learning_rate": 8.064861882157668e-05, "loss": 0.0277, "step": 13020 }, { "grad_norm": 0.08793073892593384, "learning_rate": 8.061594800936263e-05, "loss": 0.0271, "step": 13030 }, { "grad_norm": 0.13878212869167328, "learning_rate": 8.058325627143681e-05, "loss": 0.0243, "step": 13040 }, { "grad_norm": 0.08836435526609421, "learning_rate": 8.055054363014372e-05, "loss": 0.0256, "step": 13050 }, { "grad_norm": 0.08667390793561935, "learning_rate": 8.051781010784211e-05, "loss": 0.0261, "step": 13060 }, { "grad_norm": 0.08820069581270218, "learning_rate": 8.048505572690506e-05, "loss": 0.0232, "step": 13070 }, { "grad_norm": 0.09283838421106339, "learning_rate": 8.045228050971988e-05, "loss": 0.0263, "step": 13080 }, { "grad_norm": 0.1196686178445816, "learning_rate": 8.041948447868814e-05, "loss": 0.0286, "step": 13090 }, { "grad_norm": 0.1282193660736084, "learning_rate": 8.038666765622558e-05, "loss": 0.0279, "step": 13100 }, { "grad_norm": 0.10663237422704697, "learning_rate": 8.03538300647622e-05, "loss": 0.027, "step": 13110 }, { "grad_norm": 0.10011734813451767, "learning_rate": 8.03209717267422e-05, "loss": 0.0268, "step": 13120 }, { "grad_norm": 0.10482798516750336, "learning_rate": 8.028809266462395e-05, "loss": 0.0273, "step": 13130 }, { "grad_norm": 0.11989548802375793, "learning_rate": 8.025519290087994e-05, "loss": 0.0253, "step": 13140 }, { "grad_norm": 0.10579610615968704, "learning_rate": 8.022227245799688e-05, "loss": 0.0249, "step": 13150 }, { "grad_norm": 0.09924808889627457, "learning_rate": 8.018933135847557e-05, "loss": 0.0214, "step": 13160 }, { "grad_norm": 0.09736809134483337, "learning_rate": 8.015636962483096e-05, "loss": 0.0278, "step": 13170 }, { "grad_norm": 0.09517469257116318, "learning_rate": 8.012338727959205e-05, "loss": 0.0255, "step": 13180 }, { "grad_norm": 0.09522337466478348, "learning_rate": 8.009038434530198e-05, "loss": 0.0246, "step": 13190 }, { "grad_norm": 0.12976451218128204, "learning_rate": 8.005736084451796e-05, "loss": 0.0304, "step": 13200 }, { "grad_norm": 0.10526259988546371, "learning_rate": 8.002431679981122e-05, "loss": 0.0262, "step": 13210 }, { "grad_norm": 0.15299856662750244, "learning_rate": 7.999125223376706e-05, "loss": 0.0276, "step": 13220 }, { "grad_norm": 0.11800682544708252, "learning_rate": 7.99581671689848e-05, "loss": 0.0291, "step": 13230 }, { "grad_norm": 0.08246658742427826, "learning_rate": 7.992506162807775e-05, "loss": 0.027, "step": 13240 }, { "grad_norm": 0.09556598961353302, "learning_rate": 7.989193563367328e-05, "loss": 0.0283, "step": 13250 }, { "grad_norm": 0.11927669495344162, "learning_rate": 7.985878920841266e-05, "loss": 0.0302, "step": 13260 }, { "grad_norm": 0.11198481917381287, "learning_rate": 7.982562237495117e-05, "loss": 0.0312, "step": 13270 }, { "grad_norm": 0.11648610979318619, "learning_rate": 7.979243515595802e-05, "loss": 0.0284, "step": 13280 }, { "grad_norm": 0.13559627532958984, "learning_rate": 7.975922757411636e-05, "loss": 0.0284, "step": 13290 }, { "grad_norm": 0.0941723883152008, "learning_rate": 7.972599965212329e-05, "loss": 0.028, "step": 13300 }, { "grad_norm": 0.0992496907711029, "learning_rate": 7.969275141268973e-05, "loss": 0.023, "step": 13310 }, { "grad_norm": 0.1398429572582245, "learning_rate": 7.96594828785406e-05, "loss": 0.0285, "step": 13320 }, { "grad_norm": 0.10353651642799377, "learning_rate": 7.962619407241456e-05, "loss": 0.0283, "step": 13330 }, { "grad_norm": 0.16191990673542023, "learning_rate": 7.959288501706424e-05, "loss": 0.026, "step": 13340 }, { "grad_norm": 0.14221718907356262, "learning_rate": 7.955955573525605e-05, "loss": 0.0243, "step": 13350 }, { "grad_norm": 0.13629157841205597, "learning_rate": 7.952620624977026e-05, "loss": 0.029, "step": 13360 }, { "grad_norm": 0.11034173518419266, "learning_rate": 7.949283658340089e-05, "loss": 0.0227, "step": 13370 }, { "grad_norm": 0.11739124357700348, "learning_rate": 7.945944675895585e-05, "loss": 0.0256, "step": 13380 }, { "grad_norm": 0.15318278968334198, "learning_rate": 7.942603679925671e-05, "loss": 0.0253, "step": 13390 }, { "grad_norm": 0.1875934600830078, "learning_rate": 7.93926067271389e-05, "loss": 0.0239, "step": 13400 }, { "grad_norm": 0.1035756915807724, "learning_rate": 7.935915656545155e-05, "loss": 0.0256, "step": 13410 }, { "grad_norm": 0.11682788282632828, "learning_rate": 7.932568633705752e-05, "loss": 0.0267, "step": 13420 }, { "grad_norm": 0.1173052117228508, "learning_rate": 7.929219606483341e-05, "loss": 0.0293, "step": 13430 }, { "grad_norm": 0.10087238252162933, "learning_rate": 7.925868577166948e-05, "loss": 0.0299, "step": 13440 }, { "grad_norm": 0.10725029557943344, "learning_rate": 7.922515548046974e-05, "loss": 0.0252, "step": 13450 }, { "grad_norm": 0.11569137871265411, "learning_rate": 7.919160521415179e-05, "loss": 0.0271, "step": 13460 }, { "grad_norm": 0.12000925838947296, "learning_rate": 7.915803499564694e-05, "loss": 0.0241, "step": 13470 }, { "grad_norm": 0.129752516746521, "learning_rate": 7.912444484790013e-05, "loss": 0.0227, "step": 13480 }, { "grad_norm": 0.11116205155849457, "learning_rate": 7.909083479386987e-05, "loss": 0.0294, "step": 13490 }, { "grad_norm": 0.14007337391376495, "learning_rate": 7.905720485652836e-05, "loss": 0.0266, "step": 13500 }, { "grad_norm": 0.12083592265844345, "learning_rate": 7.902355505886132e-05, "loss": 0.0283, "step": 13510 }, { "grad_norm": 0.11022412031888962, "learning_rate": 7.898988542386805e-05, "loss": 0.0221, "step": 13520 }, { "grad_norm": 0.10709687322378159, "learning_rate": 7.895619597456147e-05, "loss": 0.0249, "step": 13530 }, { "grad_norm": 0.15494662523269653, "learning_rate": 7.892248673396798e-05, "loss": 0.0283, "step": 13540 }, { "grad_norm": 0.20820946991443634, "learning_rate": 7.888875772512754e-05, "loss": 0.0276, "step": 13550 }, { "grad_norm": 0.12390106916427612, "learning_rate": 7.885500897109359e-05, "loss": 0.0275, "step": 13560 }, { "grad_norm": 0.11825954914093018, "learning_rate": 7.882124049493309e-05, "loss": 0.024, "step": 13570 }, { "grad_norm": 0.08956730365753174, "learning_rate": 7.878745231972649e-05, "loss": 0.0281, "step": 13580 }, { "grad_norm": 0.12192417681217194, "learning_rate": 7.875364446856766e-05, "loss": 0.0263, "step": 13590 }, { "grad_norm": 0.07657938450574875, "learning_rate": 7.871981696456398e-05, "loss": 0.0263, "step": 13600 }, { "grad_norm": 0.10946998745203018, "learning_rate": 7.868596983083623e-05, "loss": 0.0289, "step": 13610 }, { "grad_norm": 0.1407422423362732, "learning_rate": 7.865210309051858e-05, "loss": 0.0286, "step": 13620 }, { "grad_norm": 0.13381896913051605, "learning_rate": 7.861821676675863e-05, "loss": 0.0306, "step": 13630 }, { "grad_norm": 0.13414284586906433, "learning_rate": 7.858431088271739e-05, "loss": 0.0253, "step": 13640 }, { "grad_norm": 0.12589406967163086, "learning_rate": 7.855038546156918e-05, "loss": 0.0249, "step": 13650 }, { "grad_norm": 0.18912598490715027, "learning_rate": 7.851644052650173e-05, "loss": 0.0309, "step": 13660 }, { "grad_norm": 0.08588385581970215, "learning_rate": 7.848247610071609e-05, "loss": 0.0254, "step": 13670 }, { "grad_norm": 0.1006569191813469, "learning_rate": 7.844849220742658e-05, "loss": 0.0288, "step": 13680 }, { "grad_norm": 0.10212045907974243, "learning_rate": 7.841448886986092e-05, "loss": 0.0282, "step": 13690 }, { "grad_norm": 0.134099081158638, "learning_rate": 7.838046611126004e-05, "loss": 0.0279, "step": 13700 }, { "grad_norm": 0.11535684019327164, "learning_rate": 7.834642395487819e-05, "loss": 0.0273, "step": 13710 }, { "grad_norm": 0.1029176190495491, "learning_rate": 7.831236242398285e-05, "loss": 0.0215, "step": 13720 }, { "grad_norm": 0.10818509757518768, "learning_rate": 7.827828154185477e-05, "loss": 0.0253, "step": 13730 }, { "grad_norm": 0.1323014199733734, "learning_rate": 7.82441813317879e-05, "loss": 0.0267, "step": 13740 }, { "grad_norm": 0.10596372187137604, "learning_rate": 7.821006181708944e-05, "loss": 0.029, "step": 13750 }, { "grad_norm": 0.12147139757871628, "learning_rate": 7.81759230210797e-05, "loss": 0.0278, "step": 13760 }, { "grad_norm": 0.12238737940788269, "learning_rate": 7.814176496709227e-05, "loss": 0.0269, "step": 13770 }, { "grad_norm": 0.13143277168273926, "learning_rate": 7.810758767847385e-05, "loss": 0.0287, "step": 13780 }, { "grad_norm": 0.1328541785478592, "learning_rate": 7.807339117858427e-05, "loss": 0.0299, "step": 13790 }, { "grad_norm": 0.11097810417413712, "learning_rate": 7.803917549079655e-05, "loss": 0.0272, "step": 13800 }, { "grad_norm": 0.10806499421596527, "learning_rate": 7.800494063849679e-05, "loss": 0.0289, "step": 13810 }, { "grad_norm": 0.18074145913124084, "learning_rate": 7.797068664508416e-05, "loss": 0.0287, "step": 13820 }, { "grad_norm": 0.11759228259325027, "learning_rate": 7.793641353397096e-05, "loss": 0.029, "step": 13830 }, { "grad_norm": 0.11380765587091446, "learning_rate": 7.790212132858253e-05, "loss": 0.027, "step": 13840 }, { "grad_norm": 0.09357717633247375, "learning_rate": 7.786781005235728e-05, "loss": 0.024, "step": 13850 }, { "grad_norm": 0.13002373278141022, "learning_rate": 7.783347972874662e-05, "loss": 0.0294, "step": 13860 }, { "grad_norm": 0.11116532981395721, "learning_rate": 7.779913038121504e-05, "loss": 0.0248, "step": 13870 }, { "grad_norm": 0.15621258318424225, "learning_rate": 7.776476203323997e-05, "loss": 0.0252, "step": 13880 }, { "grad_norm": 0.0767269879579544, "learning_rate": 7.773037470831185e-05, "loss": 0.0264, "step": 13890 }, { "grad_norm": 0.09982795268297195, "learning_rate": 7.76959684299341e-05, "loss": 0.0289, "step": 13900 }, { "grad_norm": 0.1344066560268402, "learning_rate": 7.76615432216231e-05, "loss": 0.0256, "step": 13910 }, { "grad_norm": 0.09309827536344528, "learning_rate": 7.762709910690811e-05, "loss": 0.0294, "step": 13920 }, { "grad_norm": 0.13218329846858978, "learning_rate": 7.759263610933141e-05, "loss": 0.0246, "step": 13930 }, { "grad_norm": 0.1658315509557724, "learning_rate": 7.755815425244811e-05, "loss": 0.0265, "step": 13940 }, { "grad_norm": 0.10141688585281372, "learning_rate": 7.752365355982624e-05, "loss": 0.0307, "step": 13950 }, { "grad_norm": 0.11148906499147415, "learning_rate": 7.748913405504668e-05, "loss": 0.0289, "step": 13960 }, { "grad_norm": 0.15020881593227386, "learning_rate": 7.745459576170322e-05, "loss": 0.0248, "step": 13970 }, { "grad_norm": 0.1018729954957962, "learning_rate": 7.742003870340242e-05, "loss": 0.0249, "step": 13980 }, { "grad_norm": 0.09140356630086899, "learning_rate": 7.738546290376373e-05, "loss": 0.0266, "step": 13990 }, { "grad_norm": 0.09617353975772858, "learning_rate": 7.735086838641937e-05, "loss": 0.0237, "step": 14000 }, { "grad_norm": 0.12477318197488785, "learning_rate": 7.731625517501437e-05, "loss": 0.0264, "step": 14010 }, { "grad_norm": 0.10623510181903839, "learning_rate": 7.728162329320655e-05, "loss": 0.0272, "step": 14020 }, { "grad_norm": 0.138414204120636, "learning_rate": 7.724697276466645e-05, "loss": 0.0265, "step": 14030 }, { "grad_norm": 0.09865614771842957, "learning_rate": 7.721230361307738e-05, "loss": 0.0256, "step": 14040 }, { "grad_norm": 0.08903894573450089, "learning_rate": 7.71776158621354e-05, "loss": 0.0255, "step": 14050 }, { "grad_norm": 0.12328046560287476, "learning_rate": 7.714290953554925e-05, "loss": 0.028, "step": 14060 }, { "grad_norm": 0.09372229874134064, "learning_rate": 7.710818465704037e-05, "loss": 0.0287, "step": 14070 }, { "grad_norm": 0.12884511053562164, "learning_rate": 7.707344125034288e-05, "loss": 0.0269, "step": 14080 }, { "grad_norm": 0.10134650766849518, "learning_rate": 7.703867933920359e-05, "loss": 0.0272, "step": 14090 }, { "grad_norm": 0.11568118631839752, "learning_rate": 7.700389894738194e-05, "loss": 0.0271, "step": 14100 }, { "grad_norm": 0.10138683766126633, "learning_rate": 7.696910009864999e-05, "loss": 0.0263, "step": 14110 }, { "grad_norm": 0.1581132858991623, "learning_rate": 7.693428281679241e-05, "loss": 0.0277, "step": 14120 }, { "grad_norm": 0.13768436014652252, "learning_rate": 7.689944712560652e-05, "loss": 0.029, "step": 14130 }, { "grad_norm": 0.14499536156654358, "learning_rate": 7.686459304890214e-05, "loss": 0.0306, "step": 14140 }, { "grad_norm": 0.10819283127784729, "learning_rate": 7.682972061050175e-05, "loss": 0.0273, "step": 14150 }, { "grad_norm": 0.09295699745416641, "learning_rate": 7.679482983424032e-05, "loss": 0.0251, "step": 14160 }, { "grad_norm": 0.11690586805343628, "learning_rate": 7.675992074396534e-05, "loss": 0.0282, "step": 14170 }, { "grad_norm": 0.11150455474853516, "learning_rate": 7.672499336353687e-05, "loss": 0.0256, "step": 14180 }, { "grad_norm": 0.09461800009012222, "learning_rate": 7.669004771682744e-05, "loss": 0.0257, "step": 14190 }, { "grad_norm": 0.1310863047838211, "learning_rate": 7.665508382772206e-05, "loss": 0.0271, "step": 14200 }, { "grad_norm": 0.13982214033603668, "learning_rate": 7.662010172011824e-05, "loss": 0.0264, "step": 14210 }, { "grad_norm": 0.11464788764715195, "learning_rate": 7.658510141792588e-05, "loss": 0.0272, "step": 14220 }, { "grad_norm": 0.12647080421447754, "learning_rate": 7.65500829450674e-05, "loss": 0.0286, "step": 14230 }, { "grad_norm": 0.10205212980508804, "learning_rate": 7.651504632547759e-05, "loss": 0.0244, "step": 14240 }, { "grad_norm": 0.08640426397323608, "learning_rate": 7.647999158310364e-05, "loss": 0.0275, "step": 14250 }, { "grad_norm": 0.08665202558040619, "learning_rate": 7.644491874190512e-05, "loss": 0.0283, "step": 14260 }, { "grad_norm": 0.10234655439853668, "learning_rate": 7.6409827825854e-05, "loss": 0.0255, "step": 14270 }, { "grad_norm": 0.11231864988803864, "learning_rate": 7.637471885893459e-05, "loss": 0.025, "step": 14280 }, { "grad_norm": 0.09124574065208435, "learning_rate": 7.633959186514354e-05, "loss": 0.0231, "step": 14290 }, { "grad_norm": 0.11195343732833862, "learning_rate": 7.630444686848984e-05, "loss": 0.0258, "step": 14300 }, { "grad_norm": 0.11471343040466309, "learning_rate": 7.626928389299471e-05, "loss": 0.024, "step": 14310 }, { "grad_norm": 0.10579907894134521, "learning_rate": 7.623410296269175e-05, "loss": 0.0294, "step": 14320 }, { "grad_norm": 0.1102285385131836, "learning_rate": 7.61989041016268e-05, "loss": 0.0279, "step": 14330 }, { "grad_norm": 0.10693146288394928, "learning_rate": 7.616368733385793e-05, "loss": 0.0264, "step": 14340 }, { "grad_norm": 0.12358848005533218, "learning_rate": 7.612845268345547e-05, "loss": 0.0258, "step": 14350 }, { "grad_norm": 0.1063496470451355, "learning_rate": 7.609320017450199e-05, "loss": 0.0267, "step": 14360 }, { "grad_norm": 0.1256190687417984, "learning_rate": 7.605792983109222e-05, "loss": 0.0254, "step": 14370 }, { "grad_norm": 0.10332586616277695, "learning_rate": 7.602264167733313e-05, "loss": 0.0283, "step": 14380 }, { "grad_norm": 0.11555171012878418, "learning_rate": 7.598733573734384e-05, "loss": 0.0243, "step": 14390 }, { "grad_norm": 0.07906810939311981, "learning_rate": 7.595201203525561e-05, "loss": 0.0242, "step": 14400 }, { "grad_norm": 0.1100398376584053, "learning_rate": 7.591667059521187e-05, "loss": 0.0269, "step": 14410 }, { "grad_norm": 0.13536420464515686, "learning_rate": 7.588131144136815e-05, "loss": 0.0239, "step": 14420 }, { "grad_norm": 0.11071937531232834, "learning_rate": 7.584593459789212e-05, "loss": 0.0223, "step": 14430 }, { "grad_norm": 0.12823505699634552, "learning_rate": 7.58105400889635e-05, "loss": 0.0241, "step": 14440 }, { "grad_norm": 0.13027699291706085, "learning_rate": 7.57751279387741e-05, "loss": 0.0321, "step": 14450 }, { "grad_norm": 0.14149220287799835, "learning_rate": 7.573969817152782e-05, "loss": 0.0266, "step": 14460 }, { "grad_norm": 0.22982050478458405, "learning_rate": 7.570425081144052e-05, "loss": 0.0267, "step": 14470 }, { "grad_norm": 0.11322111636400223, "learning_rate": 7.56687858827402e-05, "loss": 0.0295, "step": 14480 }, { "grad_norm": 0.11191970854997635, "learning_rate": 7.563330340966675e-05, "loss": 0.0258, "step": 14490 }, { "grad_norm": 0.11112473160028458, "learning_rate": 7.559780341647212e-05, "loss": 0.0215, "step": 14500 }, { "grad_norm": 0.0947890430688858, "learning_rate": 7.556228592742026e-05, "loss": 0.0232, "step": 14510 }, { "grad_norm": 0.1117534190416336, "learning_rate": 7.552675096678696e-05, "loss": 0.0226, "step": 14520 }, { "grad_norm": 0.10915134102106094, "learning_rate": 7.549119855886012e-05, "loss": 0.0246, "step": 14530 }, { "grad_norm": 0.11092860996723175, "learning_rate": 7.545562872793941e-05, "loss": 0.026, "step": 14540 }, { "grad_norm": 0.12566520273685455, "learning_rate": 7.542004149833648e-05, "loss": 0.0253, "step": 14550 }, { "grad_norm": 0.08244369924068451, "learning_rate": 7.538443689437492e-05, "loss": 0.024, "step": 14560 }, { "grad_norm": 0.1337422877550125, "learning_rate": 7.53488149403901e-05, "loss": 0.0243, "step": 14570 }, { "grad_norm": 0.10757584124803543, "learning_rate": 7.531317566072929e-05, "loss": 0.0252, "step": 14580 }, { "grad_norm": 0.09352731704711914, "learning_rate": 7.527751907975158e-05, "loss": 0.0247, "step": 14590 }, { "grad_norm": 0.09821323305368423, "learning_rate": 7.524184522182793e-05, "loss": 0.0258, "step": 14600 }, { "grad_norm": 0.11394518613815308, "learning_rate": 7.520615411134112e-05, "loss": 0.0266, "step": 14610 }, { "grad_norm": 0.13087131083011627, "learning_rate": 7.517044577268564e-05, "loss": 0.0251, "step": 14620 }, { "grad_norm": 0.09002559632062912, "learning_rate": 7.513472023026782e-05, "loss": 0.0281, "step": 14630 }, { "grad_norm": 0.11795273423194885, "learning_rate": 7.509897750850572e-05, "loss": 0.0285, "step": 14640 }, { "grad_norm": 0.08843551576137543, "learning_rate": 7.506321763182918e-05, "loss": 0.0243, "step": 14650 }, { "grad_norm": 0.2092680186033249, "learning_rate": 7.50274406246797e-05, "loss": 0.0255, "step": 14660 }, { "grad_norm": 0.11624407023191452, "learning_rate": 7.499164651151056e-05, "loss": 0.0229, "step": 14670 }, { "grad_norm": 0.09316554665565491, "learning_rate": 7.495583531678669e-05, "loss": 0.0278, "step": 14680 }, { "grad_norm": 0.09404183179140091, "learning_rate": 7.492000706498469e-05, "loss": 0.0234, "step": 14690 }, { "grad_norm": 0.1103011891245842, "learning_rate": 7.488416178059284e-05, "loss": 0.0251, "step": 14700 }, { "grad_norm": 0.08916258811950684, "learning_rate": 7.484829948811107e-05, "loss": 0.0311, "step": 14710 }, { "grad_norm": 0.10293553024530411, "learning_rate": 7.48124202120509e-05, "loss": 0.0295, "step": 14720 }, { "grad_norm": 0.15923555195331573, "learning_rate": 7.477652397693549e-05, "loss": 0.026, "step": 14730 }, { "grad_norm": 0.1310773342847824, "learning_rate": 7.474061080729955e-05, "loss": 0.0303, "step": 14740 }, { "grad_norm": 0.1396116018295288, "learning_rate": 7.470468072768941e-05, "loss": 0.0263, "step": 14750 }, { "grad_norm": 0.1259137988090515, "learning_rate": 7.466873376266297e-05, "loss": 0.0252, "step": 14760 }, { "grad_norm": 0.1347687542438507, "learning_rate": 7.46327699367896e-05, "loss": 0.0276, "step": 14770 }, { "grad_norm": 0.13396520912647247, "learning_rate": 7.459678927465026e-05, "loss": 0.0236, "step": 14780 }, { "grad_norm": 0.15013396739959717, "learning_rate": 7.456079180083737e-05, "loss": 0.0265, "step": 14790 }, { "grad_norm": 0.08122502267360687, "learning_rate": 7.452477753995489e-05, "loss": 0.0199, "step": 14800 }, { "grad_norm": 0.1213846281170845, "learning_rate": 7.448874651661823e-05, "loss": 0.0236, "step": 14810 }, { "grad_norm": 0.10529828816652298, "learning_rate": 7.445269875545423e-05, "loss": 0.0251, "step": 14820 }, { "grad_norm": 0.11588159948587418, "learning_rate": 7.44166342811012e-05, "loss": 0.0259, "step": 14830 }, { "grad_norm": 0.09174696356058121, "learning_rate": 7.438055311820886e-05, "loss": 0.0234, "step": 14840 }, { "grad_norm": 0.10655071586370468, "learning_rate": 7.434445529143837e-05, "loss": 0.0235, "step": 14850 }, { "grad_norm": 0.18966704607009888, "learning_rate": 7.430834082546225e-05, "loss": 0.024, "step": 14860 }, { "grad_norm": 0.11604877561330795, "learning_rate": 7.427220974496438e-05, "loss": 0.0246, "step": 14870 }, { "grad_norm": 0.1650756150484085, "learning_rate": 7.423606207464005e-05, "loss": 0.0249, "step": 14880 }, { "grad_norm": 0.14413917064666748, "learning_rate": 7.419989783919578e-05, "loss": 0.0255, "step": 14890 }, { "grad_norm": 0.1283636838197708, "learning_rate": 7.416371706334956e-05, "loss": 0.0234, "step": 14900 }, { "grad_norm": 0.17666403949260712, "learning_rate": 7.412751977183056e-05, "loss": 0.0235, "step": 14910 }, { "grad_norm": 0.1381123661994934, "learning_rate": 7.409130598937932e-05, "loss": 0.0305, "step": 14920 }, { "grad_norm": 0.1567569524049759, "learning_rate": 7.40550757407476e-05, "loss": 0.0237, "step": 14930 }, { "grad_norm": 0.15358132123947144, "learning_rate": 7.401882905069843e-05, "loss": 0.0282, "step": 14940 }, { "grad_norm": 0.13603903353214264, "learning_rate": 7.39825659440061e-05, "loss": 0.0236, "step": 14950 }, { "grad_norm": 0.09044638276100159, "learning_rate": 7.394628644545609e-05, "loss": 0.0208, "step": 14960 }, { "grad_norm": 0.0869775265455246, "learning_rate": 7.390999057984507e-05, "loss": 0.0223, "step": 14970 }, { "grad_norm": 0.07122713327407837, "learning_rate": 7.387367837198097e-05, "loss": 0.023, "step": 14980 }, { "grad_norm": 0.10837635397911072, "learning_rate": 7.383734984668281e-05, "loss": 0.0271, "step": 14990 }, { "grad_norm": 0.12206853181123734, "learning_rate": 7.38010050287808e-05, "loss": 0.0235, "step": 15000 }, { "grad_norm": 0.1286841332912445, "learning_rate": 7.376464394311628e-05, "loss": 0.0239, "step": 15010 }, { "grad_norm": 0.07823383063077927, "learning_rate": 7.372826661454172e-05, "loss": 0.0226, "step": 15020 }, { "grad_norm": 0.10127204656600952, "learning_rate": 7.369187306792068e-05, "loss": 0.0258, "step": 15030 }, { "grad_norm": 0.09315725415945053, "learning_rate": 7.365546332812779e-05, "loss": 0.0242, "step": 15040 }, { "grad_norm": 0.10459643602371216, "learning_rate": 7.361903742004876e-05, "loss": 0.0236, "step": 15050 }, { "grad_norm": 0.11481944471597672, "learning_rate": 7.358259536858039e-05, "loss": 0.0238, "step": 15060 }, { "grad_norm": 0.11587508022785187, "learning_rate": 7.354613719863044e-05, "loss": 0.0246, "step": 15070 }, { "grad_norm": 0.11484523117542267, "learning_rate": 7.350966293511776e-05, "loss": 0.0217, "step": 15080 }, { "grad_norm": 0.09000763297080994, "learning_rate": 7.347317260297212e-05, "loss": 0.0238, "step": 15090 }, { "grad_norm": 0.1138652041554451, "learning_rate": 7.343666622713437e-05, "loss": 0.0222, "step": 15100 }, { "grad_norm": 0.11335600912570953, "learning_rate": 7.340014383255624e-05, "loss": 0.0255, "step": 15110 }, { "grad_norm": 0.11864104121923447, "learning_rate": 7.336360544420044e-05, "loss": 0.0273, "step": 15120 }, { "grad_norm": 0.12545648217201233, "learning_rate": 7.332705108704064e-05, "loss": 0.0229, "step": 15130 }, { "grad_norm": 0.1380547136068344, "learning_rate": 7.329048078606138e-05, "loss": 0.0211, "step": 15140 }, { "grad_norm": 0.11082867532968521, "learning_rate": 7.32538945662581e-05, "loss": 0.023, "step": 15150 }, { "grad_norm": 0.1092982217669487, "learning_rate": 7.321729245263718e-05, "loss": 0.0216, "step": 15160 }, { "grad_norm": 0.15348708629608154, "learning_rate": 7.318067447021578e-05, "loss": 0.0233, "step": 15170 }, { "grad_norm": 0.10391761362552643, "learning_rate": 7.314404064402198e-05, "loss": 0.0217, "step": 15180 }, { "grad_norm": 0.08036860078573227, "learning_rate": 7.310739099909461e-05, "loss": 0.0222, "step": 15190 }, { "grad_norm": 0.07519040256738663, "learning_rate": 7.307072556048339e-05, "loss": 0.0253, "step": 15200 }, { "grad_norm": 0.11242741346359253, "learning_rate": 7.30340443532488e-05, "loss": 0.0222, "step": 15210 }, { "grad_norm": 0.09931299835443497, "learning_rate": 7.299734740246208e-05, "loss": 0.0223, "step": 15220 }, { "grad_norm": 0.0795295462012291, "learning_rate": 7.296063473320528e-05, "loss": 0.0216, "step": 15230 }, { "grad_norm": 0.105243980884552, "learning_rate": 7.292390637057113e-05, "loss": 0.0223, "step": 15240 }, { "grad_norm": 0.4086606800556183, "learning_rate": 7.288716233966314e-05, "loss": 0.0237, "step": 15250 }, { "grad_norm": 0.11309774219989777, "learning_rate": 7.285040266559551e-05, "loss": 0.0229, "step": 15260 }, { "grad_norm": 0.0979711189866066, "learning_rate": 7.281362737349312e-05, "loss": 0.0256, "step": 15270 }, { "grad_norm": 0.11369510740041733, "learning_rate": 7.277683648849153e-05, "loss": 0.0239, "step": 15280 }, { "grad_norm": 0.16985894739627838, "learning_rate": 7.2740030035737e-05, "loss": 0.0248, "step": 15290 }, { "grad_norm": 0.10708782076835632, "learning_rate": 7.270320804038634e-05, "loss": 0.0236, "step": 15300 }, { "grad_norm": 0.10781823098659515, "learning_rate": 7.266637052760708e-05, "loss": 0.0269, "step": 15310 }, { "grad_norm": 0.08822710812091827, "learning_rate": 7.262951752257728e-05, "loss": 0.0297, "step": 15320 }, { "grad_norm": 0.12001118808984756, "learning_rate": 7.259264905048564e-05, "loss": 0.0268, "step": 15330 }, { "grad_norm": 0.10712095350027084, "learning_rate": 7.255576513653142e-05, "loss": 0.0263, "step": 15340 }, { "grad_norm": 0.08434335887432098, "learning_rate": 7.251886580592439e-05, "loss": 0.0226, "step": 15350 }, { "grad_norm": 0.07515328377485275, "learning_rate": 7.248195108388496e-05, "loss": 0.0233, "step": 15360 }, { "grad_norm": 0.11514334380626678, "learning_rate": 7.244502099564395e-05, "loss": 0.0229, "step": 15370 }, { "grad_norm": 0.09632230550050735, "learning_rate": 7.240807556644271e-05, "loss": 0.0245, "step": 15380 }, { "grad_norm": 0.12484028190374374, "learning_rate": 7.237111482153314e-05, "loss": 0.0276, "step": 15390 }, { "grad_norm": 0.11493062973022461, "learning_rate": 7.233413878617751e-05, "loss": 0.0268, "step": 15400 }, { "grad_norm": 0.12819306552410126, "learning_rate": 7.229714748564864e-05, "loss": 0.023, "step": 15410 }, { "grad_norm": 0.14203472435474396, "learning_rate": 7.22601409452297e-05, "loss": 0.0278, "step": 15420 }, { "grad_norm": 0.1003202497959137, "learning_rate": 7.222311919021433e-05, "loss": 0.0217, "step": 15430 }, { "grad_norm": 0.0989736020565033, "learning_rate": 7.218608224590655e-05, "loss": 0.0231, "step": 15440 }, { "grad_norm": 0.08382637053728104, "learning_rate": 7.214903013762074e-05, "loss": 0.0248, "step": 15450 }, { "grad_norm": 0.1012239083647728, "learning_rate": 7.21119628906817e-05, "loss": 0.0236, "step": 15460 }, { "grad_norm": 0.12966860830783844, "learning_rate": 7.207488053042454e-05, "loss": 0.0241, "step": 15470 }, { "grad_norm": 0.10548477619886398, "learning_rate": 7.203778308219467e-05, "loss": 0.0208, "step": 15480 }, { "grad_norm": 0.15607941150665283, "learning_rate": 7.200067057134787e-05, "loss": 0.0276, "step": 15490 }, { "grad_norm": 0.11690368503332138, "learning_rate": 7.196354302325019e-05, "loss": 0.0245, "step": 15500 }, { "grad_norm": 0.10100480914115906, "learning_rate": 7.192640046327795e-05, "loss": 0.025, "step": 15510 }, { "grad_norm": 0.08878485858440399, "learning_rate": 7.188924291681777e-05, "loss": 0.0238, "step": 15520 }, { "grad_norm": 0.14666205644607544, "learning_rate": 7.185207040926643e-05, "loss": 0.0305, "step": 15530 }, { "grad_norm": 0.1312023401260376, "learning_rate": 7.181488296603103e-05, "loss": 0.0256, "step": 15540 }, { "grad_norm": 0.1274571567773819, "learning_rate": 7.177768061252885e-05, "loss": 0.0237, "step": 15550 }, { "grad_norm": 0.12377388775348663, "learning_rate": 7.174046337418729e-05, "loss": 0.0218, "step": 15560 }, { "grad_norm": 0.1341186761856079, "learning_rate": 7.170323127644403e-05, "loss": 0.0212, "step": 15570 }, { "grad_norm": 0.12364830076694489, "learning_rate": 7.166598434474683e-05, "loss": 0.0249, "step": 15580 }, { "grad_norm": 0.20747633278369904, "learning_rate": 7.162872260455364e-05, "loss": 0.0243, "step": 15590 }, { "grad_norm": 0.09176817536354065, "learning_rate": 7.159144608133248e-05, "loss": 0.0237, "step": 15600 }, { "grad_norm": 0.13036973774433136, "learning_rate": 7.155415480056153e-05, "loss": 0.0236, "step": 15610 }, { "grad_norm": 0.10967782884836197, "learning_rate": 7.151684878772902e-05, "loss": 0.0238, "step": 15620 }, { "grad_norm": 0.1547185182571411, "learning_rate": 7.147952806833324e-05, "loss": 0.0204, "step": 15630 }, { "grad_norm": 0.08883103728294373, "learning_rate": 7.14421926678826e-05, "loss": 0.0242, "step": 15640 }, { "grad_norm": 0.10536426305770874, "learning_rate": 7.140484261189543e-05, "loss": 0.0271, "step": 15650 }, { "grad_norm": 0.12118902057409286, "learning_rate": 7.136747792590017e-05, "loss": 0.0242, "step": 15660 }, { "grad_norm": 0.19757097959518433, "learning_rate": 7.133009863543524e-05, "loss": 0.0228, "step": 15670 }, { "grad_norm": 0.12106046080589294, "learning_rate": 7.129270476604901e-05, "loss": 0.023, "step": 15680 }, { "grad_norm": 0.13008426129817963, "learning_rate": 7.125529634329988e-05, "loss": 0.0245, "step": 15690 }, { "grad_norm": 0.12013493478298187, "learning_rate": 7.12178733927561e-05, "loss": 0.0239, "step": 15700 }, { "grad_norm": 0.13650035858154297, "learning_rate": 7.118043593999593e-05, "loss": 0.0271, "step": 15710 }, { "grad_norm": 0.08372244238853455, "learning_rate": 7.114298401060752e-05, "loss": 0.0245, "step": 15720 }, { "grad_norm": 0.116281658411026, "learning_rate": 7.11055176301889e-05, "loss": 0.0227, "step": 15730 }, { "grad_norm": 0.11529669910669327, "learning_rate": 7.1068036824348e-05, "loss": 0.0273, "step": 15740 }, { "grad_norm": 0.12468313425779343, "learning_rate": 7.10305416187026e-05, "loss": 0.0235, "step": 15750 }, { "grad_norm": 0.1109551414847374, "learning_rate": 7.099303203888029e-05, "loss": 0.024, "step": 15760 }, { "grad_norm": 0.09102942049503326, "learning_rate": 7.095550811051855e-05, "loss": 0.0275, "step": 15770 }, { "grad_norm": 0.09194082021713257, "learning_rate": 7.09179698592646e-05, "loss": 0.0273, "step": 15780 }, { "grad_norm": 0.11326037347316742, "learning_rate": 7.088041731077551e-05, "loss": 0.026, "step": 15790 }, { "grad_norm": 0.15218192338943481, "learning_rate": 7.084285049071806e-05, "loss": 0.0268, "step": 15800 }, { "grad_norm": 0.10617762058973312, "learning_rate": 7.080526942476886e-05, "loss": 0.0258, "step": 15810 }, { "grad_norm": 0.12133777141571045, "learning_rate": 7.076767413861418e-05, "loss": 0.0295, "step": 15820 }, { "grad_norm": 0.14648836851119995, "learning_rate": 7.073006465795005e-05, "loss": 0.027, "step": 15830 }, { "grad_norm": 0.14009515941143036, "learning_rate": 7.06924410084822e-05, "loss": 0.0251, "step": 15840 }, { "grad_norm": 0.13420474529266357, "learning_rate": 7.065480321592604e-05, "loss": 0.0266, "step": 15850 }, { "grad_norm": 0.075115866959095, "learning_rate": 7.061715130600663e-05, "loss": 0.0246, "step": 15860 }, { "grad_norm": 0.0788843035697937, "learning_rate": 7.057948530445873e-05, "loss": 0.024, "step": 15870 }, { "grad_norm": 0.10432301461696625, "learning_rate": 7.054180523702668e-05, "loss": 0.0286, "step": 15880 }, { "grad_norm": 0.09690746665000916, "learning_rate": 7.050411112946442e-05, "loss": 0.0236, "step": 15890 }, { "grad_norm": 0.13146759569644928, "learning_rate": 7.046640300753557e-05, "loss": 0.0268, "step": 15900 }, { "grad_norm": 0.10066016018390656, "learning_rate": 7.042868089701325e-05, "loss": 0.027, "step": 15910 }, { "grad_norm": 0.15135496854782104, "learning_rate": 7.039094482368016e-05, "loss": 0.0254, "step": 15920 }, { "grad_norm": 0.18899472057819366, "learning_rate": 7.035319481332858e-05, "loss": 0.0311, "step": 15930 }, { "grad_norm": 0.09722673892974854, "learning_rate": 7.031543089176023e-05, "loss": 0.0237, "step": 15940 }, { "grad_norm": 0.09047382324934006, "learning_rate": 7.027765308478644e-05, "loss": 0.025, "step": 15950 }, { "grad_norm": 0.11259476840496063, "learning_rate": 7.023986141822798e-05, "loss": 0.0265, "step": 15960 }, { "grad_norm": 0.12957434356212616, "learning_rate": 7.02020559179151e-05, "loss": 0.0243, "step": 15970 }, { "grad_norm": 0.1349128633737564, "learning_rate": 7.016423660968748e-05, "loss": 0.0239, "step": 15980 }, { "grad_norm": 0.10221952199935913, "learning_rate": 7.012640351939428e-05, "loss": 0.0245, "step": 15990 }, { "grad_norm": 0.08976659178733826, "learning_rate": 7.008855667289404e-05, "loss": 0.0226, "step": 16000 }, { "grad_norm": 0.09388142824172974, "learning_rate": 7.005069609605476e-05, "loss": 0.0227, "step": 16010 }, { "grad_norm": 0.1327214390039444, "learning_rate": 7.001282181475377e-05, "loss": 0.0199, "step": 16020 }, { "grad_norm": 0.07198533415794373, "learning_rate": 6.997493385487775e-05, "loss": 0.0217, "step": 16030 }, { "grad_norm": 0.11292994022369385, "learning_rate": 6.99370322423228e-05, "loss": 0.0218, "step": 16040 }, { "grad_norm": 0.0811782255768776, "learning_rate": 6.989911700299433e-05, "loss": 0.0236, "step": 16050 }, { "grad_norm": 0.1139030009508133, "learning_rate": 6.9861188162807e-05, "loss": 0.0209, "step": 16060 }, { "grad_norm": 0.11223713308572769, "learning_rate": 6.982324574768487e-05, "loss": 0.0201, "step": 16070 }, { "grad_norm": 0.10266482830047607, "learning_rate": 6.978528978356117e-05, "loss": 0.022, "step": 16080 }, { "grad_norm": 0.07002374529838562, "learning_rate": 6.974732029637846e-05, "loss": 0.0239, "step": 16090 }, { "grad_norm": 0.06796199083328247, "learning_rate": 6.970933731208855e-05, "loss": 0.022, "step": 16100 }, { "grad_norm": 0.10215812176465988, "learning_rate": 6.967134085665244e-05, "loss": 0.0211, "step": 16110 }, { "grad_norm": 0.09600355476140976, "learning_rate": 6.963333095604034e-05, "loss": 0.0222, "step": 16120 }, { "grad_norm": 0.1066657230257988, "learning_rate": 6.959530763623166e-05, "loss": 0.0205, "step": 16130 }, { "grad_norm": 0.10581187903881073, "learning_rate": 6.955727092321497e-05, "loss": 0.0253, "step": 16140 }, { "grad_norm": 0.06090608239173889, "learning_rate": 6.951922084298803e-05, "loss": 0.0215, "step": 16150 }, { "grad_norm": 0.08135349303483963, "learning_rate": 6.948115742155769e-05, "loss": 0.0251, "step": 16160 }, { "grad_norm": 0.1222660094499588, "learning_rate": 6.944308068493996e-05, "loss": 0.0205, "step": 16170 }, { "grad_norm": 0.11638811230659485, "learning_rate": 6.940499065915992e-05, "loss": 0.0265, "step": 16180 }, { "grad_norm": 0.08529388904571533, "learning_rate": 6.936688737025173e-05, "loss": 0.023, "step": 16190 }, { "grad_norm": 0.07695186883211136, "learning_rate": 6.932877084425867e-05, "loss": 0.0232, "step": 16200 }, { "grad_norm": 0.07578609883785248, "learning_rate": 6.929064110723297e-05, "loss": 0.0199, "step": 16210 }, { "grad_norm": 0.09368785470724106, "learning_rate": 6.925249818523598e-05, "loss": 0.0225, "step": 16220 }, { "grad_norm": 0.09586215764284134, "learning_rate": 6.921434210433801e-05, "loss": 0.023, "step": 16230 }, { "grad_norm": 0.09548316895961761, "learning_rate": 6.917617289061841e-05, "loss": 0.0241, "step": 16240 }, { "grad_norm": 0.12539304792881012, "learning_rate": 6.913799057016547e-05, "loss": 0.0246, "step": 16250 }, { "grad_norm": 0.09658270329236984, "learning_rate": 6.909979516907641e-05, "loss": 0.0255, "step": 16260 }, { "grad_norm": 0.08576077967882156, "learning_rate": 6.906158671345746e-05, "loss": 0.0238, "step": 16270 }, { "grad_norm": 0.15828104317188263, "learning_rate": 6.902336522942374e-05, "loss": 0.0248, "step": 16280 }, { "grad_norm": 0.09860353171825409, "learning_rate": 6.898513074309924e-05, "loss": 0.0287, "step": 16290 }, { "grad_norm": 0.11079376935958862, "learning_rate": 6.894688328061693e-05, "loss": 0.0259, "step": 16300 }, { "grad_norm": 0.1001613438129425, "learning_rate": 6.890862286811853e-05, "loss": 0.0258, "step": 16310 }, { "grad_norm": 0.1126161590218544, "learning_rate": 6.88703495317547e-05, "loss": 0.024, "step": 16320 }, { "grad_norm": 0.09698997437953949, "learning_rate": 6.883206329768492e-05, "loss": 0.0274, "step": 16330 }, { "grad_norm": 0.08498658984899521, "learning_rate": 6.879376419207743e-05, "loss": 0.0238, "step": 16340 }, { "grad_norm": 0.10299720615148544, "learning_rate": 6.875545224110935e-05, "loss": 0.0242, "step": 16350 }, { "grad_norm": 0.11439385265111923, "learning_rate": 6.871712747096651e-05, "loss": 0.0282, "step": 16360 }, { "grad_norm": 0.12194425612688065, "learning_rate": 6.867878990784353e-05, "loss": 0.0248, "step": 16370 }, { "grad_norm": 0.06196342781186104, "learning_rate": 6.864043957794377e-05, "loss": 0.0268, "step": 16380 }, { "grad_norm": 0.0773003026843071, "learning_rate": 6.860207650747934e-05, "loss": 0.0267, "step": 16390 }, { "grad_norm": 0.10945098847150803, "learning_rate": 6.856370072267104e-05, "loss": 0.0269, "step": 16400 }, { "grad_norm": 0.08866585791110992, "learning_rate": 6.852531224974831e-05, "loss": 0.0267, "step": 16410 }, { "grad_norm": 0.08633141219615936, "learning_rate": 6.848691111494936e-05, "loss": 0.022, "step": 16420 }, { "grad_norm": 0.10028370469808578, "learning_rate": 6.844849734452097e-05, "loss": 0.024, "step": 16430 }, { "grad_norm": 0.09642697125673294, "learning_rate": 6.841007096471862e-05, "loss": 0.0272, "step": 16440 }, { "grad_norm": 0.135860875248909, "learning_rate": 6.837163200180636e-05, "loss": 0.0262, "step": 16450 }, { "grad_norm": 0.12366491556167603, "learning_rate": 6.833318048205684e-05, "loss": 0.0257, "step": 16460 }, { "grad_norm": 0.10075068473815918, "learning_rate": 6.829471643175136e-05, "loss": 0.0263, "step": 16470 }, { "grad_norm": 0.09026015549898148, "learning_rate": 6.825623987717969e-05, "loss": 0.027, "step": 16480 }, { "grad_norm": 0.09479696303606033, "learning_rate": 6.821775084464022e-05, "loss": 0.027, "step": 16490 }, { "grad_norm": 0.09237893670797348, "learning_rate": 6.817924936043982e-05, "loss": 0.0254, "step": 16500 }, { "grad_norm": 0.11549399048089981, "learning_rate": 6.81407354508939e-05, "loss": 0.0235, "step": 16510 }, { "grad_norm": 0.0950637236237526, "learning_rate": 6.810220914232636e-05, "loss": 0.0237, "step": 16520 }, { "grad_norm": 0.12480491399765015, "learning_rate": 6.806367046106959e-05, "loss": 0.0226, "step": 16530 }, { "grad_norm": 0.08088399469852448, "learning_rate": 6.802511943346435e-05, "loss": 0.0221, "step": 16540 }, { "grad_norm": 0.10249947756528854, "learning_rate": 6.798655608585997e-05, "loss": 0.0253, "step": 16550 }, { "grad_norm": 0.08941566944122314, "learning_rate": 6.79479804446141e-05, "loss": 0.0226, "step": 16560 }, { "grad_norm": 0.12886619567871094, "learning_rate": 6.790939253609284e-05, "loss": 0.023, "step": 16570 }, { "grad_norm": 0.10381213575601578, "learning_rate": 6.787079238667065e-05, "loss": 0.0243, "step": 16580 }, { "grad_norm": 0.12415983527898788, "learning_rate": 6.783218002273039e-05, "loss": 0.0237, "step": 16590 }, { "grad_norm": 0.11708206683397293, "learning_rate": 6.779355547066322e-05, "loss": 0.0231, "step": 16600 }, { "grad_norm": 0.08190847933292389, "learning_rate": 6.775491875686865e-05, "loss": 0.023, "step": 16610 }, { "grad_norm": 0.1240953728556633, "learning_rate": 6.771626990775457e-05, "loss": 0.0234, "step": 16620 }, { "grad_norm": 0.10093587636947632, "learning_rate": 6.767760894973704e-05, "loss": 0.0214, "step": 16630 }, { "grad_norm": 0.09352199733257294, "learning_rate": 6.763893590924048e-05, "loss": 0.0229, "step": 16640 }, { "grad_norm": 0.09670612215995789, "learning_rate": 6.760025081269756e-05, "loss": 0.023, "step": 16650 }, { "grad_norm": 0.09323225915431976, "learning_rate": 6.756155368654915e-05, "loss": 0.0213, "step": 16660 }, { "grad_norm": 0.1337163746356964, "learning_rate": 6.752284455724442e-05, "loss": 0.0235, "step": 16670 }, { "grad_norm": 0.10942280292510986, "learning_rate": 6.748412345124065e-05, "loss": 0.0213, "step": 16680 }, { "grad_norm": 0.16942714154720306, "learning_rate": 6.744539039500335e-05, "loss": 0.0219, "step": 16690 }, { "grad_norm": 0.10751626640558243, "learning_rate": 6.740664541500625e-05, "loss": 0.0248, "step": 16700 }, { "grad_norm": 0.11875755339860916, "learning_rate": 6.736788853773112e-05, "loss": 0.0202, "step": 16710 }, { "grad_norm": 0.11632264405488968, "learning_rate": 6.732911978966796e-05, "loss": 0.0254, "step": 16720 }, { "grad_norm": 0.08908413350582123, "learning_rate": 6.729033919731482e-05, "loss": 0.0212, "step": 16730 }, { "grad_norm": 0.09291373938322067, "learning_rate": 6.725154678717787e-05, "loss": 0.0201, "step": 16740 }, { "grad_norm": 0.10230209678411484, "learning_rate": 6.721274258577138e-05, "loss": 0.022, "step": 16750 }, { "grad_norm": 0.15454638004302979, "learning_rate": 6.717392661961763e-05, "loss": 0.0237, "step": 16760 }, { "grad_norm": 0.08721445500850677, "learning_rate": 6.713509891524697e-05, "loss": 0.0246, "step": 16770 }, { "grad_norm": 0.1041119322180748, "learning_rate": 6.709625949919777e-05, "loss": 0.0219, "step": 16780 }, { "grad_norm": 0.12266190350055695, "learning_rate": 6.705740839801642e-05, "loss": 0.024, "step": 16790 }, { "grad_norm": 0.10184130817651749, "learning_rate": 6.701854563825727e-05, "loss": 0.0235, "step": 16800 }, { "grad_norm": 0.10981734842061996, "learning_rate": 6.697967124648266e-05, "loss": 0.0258, "step": 16810 }, { "grad_norm": 0.09667760133743286, "learning_rate": 6.694078524926285e-05, "loss": 0.0256, "step": 16820 }, { "grad_norm": 0.1121491864323616, "learning_rate": 6.690188767317607e-05, "loss": 0.0228, "step": 16830 }, { "grad_norm": 0.13537812232971191, "learning_rate": 6.686297854480843e-05, "loss": 0.0248, "step": 16840 }, { "grad_norm": 0.09955497086048126, "learning_rate": 6.682405789075398e-05, "loss": 0.0221, "step": 16850 }, { "grad_norm": 0.10589995980262756, "learning_rate": 6.67851257376146e-05, "loss": 0.0224, "step": 16860 }, { "grad_norm": 0.17178408801555634, "learning_rate": 6.674618211200004e-05, "loss": 0.0224, "step": 16870 }, { "grad_norm": 0.11400720477104187, "learning_rate": 6.670722704052792e-05, "loss": 0.0261, "step": 16880 }, { "grad_norm": 0.10832507163286209, "learning_rate": 6.666826054982365e-05, "loss": 0.0268, "step": 16890 }, { "grad_norm": 0.12637580931186676, "learning_rate": 6.662928266652048e-05, "loss": 0.0301, "step": 16900 }, { "grad_norm": 0.10970087349414825, "learning_rate": 6.659029341725941e-05, "loss": 0.0235, "step": 16910 }, { "grad_norm": 0.12919950485229492, "learning_rate": 6.655129282868923e-05, "loss": 0.0222, "step": 16920 }, { "grad_norm": 0.10819017887115479, "learning_rate": 6.651228092746646e-05, "loss": 0.023, "step": 16930 }, { "grad_norm": 0.09959352016448975, "learning_rate": 6.647325774025539e-05, "loss": 0.024, "step": 16940 }, { "grad_norm": 0.10247640311717987, "learning_rate": 6.643422329372798e-05, "loss": 0.0215, "step": 16950 }, { "grad_norm": 0.09770537912845612, "learning_rate": 6.639517761456392e-05, "loss": 0.0275, "step": 16960 }, { "grad_norm": 0.09286750108003616, "learning_rate": 6.635612072945054e-05, "loss": 0.0267, "step": 16970 }, { "grad_norm": 0.09607569128274918, "learning_rate": 6.631705266508289e-05, "loss": 0.0224, "step": 16980 }, { "grad_norm": 0.1062997579574585, "learning_rate": 6.62779734481636e-05, "loss": 0.0265, "step": 16990 }, { "grad_norm": 0.1277945637702942, "learning_rate": 6.623888310540294e-05, "loss": 0.0277, "step": 17000 }, { "grad_norm": 0.09455370903015137, "learning_rate": 6.619978166351882e-05, "loss": 0.0235, "step": 17010 }, { "grad_norm": 0.11998278647661209, "learning_rate": 6.616066914923666e-05, "loss": 0.0244, "step": 17020 }, { "grad_norm": 0.10905332863330841, "learning_rate": 6.612154558928955e-05, "loss": 0.0254, "step": 17030 }, { "grad_norm": 0.12806184589862823, "learning_rate": 6.608241101041804e-05, "loss": 0.0239, "step": 17040 }, { "grad_norm": 0.08791188150644302, "learning_rate": 6.604326543937025e-05, "loss": 0.0233, "step": 17050 }, { "grad_norm": 0.11798961460590363, "learning_rate": 6.60041089029018e-05, "loss": 0.0232, "step": 17060 }, { "grad_norm": 0.07236054539680481, "learning_rate": 6.596494142777583e-05, "loss": 0.0262, "step": 17070 }, { "grad_norm": 0.12440387159585953, "learning_rate": 6.592576304076294e-05, "loss": 0.0284, "step": 17080 }, { "grad_norm": 0.10689356923103333, "learning_rate": 6.588657376864119e-05, "loss": 0.0217, "step": 17090 }, { "grad_norm": 0.12978686392307281, "learning_rate": 6.584737363819605e-05, "loss": 0.0278, "step": 17100 }, { "grad_norm": 0.1286194622516632, "learning_rate": 6.580816267622048e-05, "loss": 0.0278, "step": 17110 }, { "grad_norm": 0.12146804481744766, "learning_rate": 6.576894090951478e-05, "loss": 0.0274, "step": 17120 }, { "grad_norm": 0.1185176819562912, "learning_rate": 6.572970836488665e-05, "loss": 0.03, "step": 17130 }, { "grad_norm": 0.07865897566080093, "learning_rate": 6.569046506915119e-05, "loss": 0.0257, "step": 17140 }, { "grad_norm": 0.10252749919891357, "learning_rate": 6.56512110491308e-05, "loss": 0.0276, "step": 17150 }, { "grad_norm": 0.11371609568595886, "learning_rate": 6.561194633165523e-05, "loss": 0.0263, "step": 17160 }, { "grad_norm": 0.10276031494140625, "learning_rate": 6.557267094356155e-05, "loss": 0.0238, "step": 17170 }, { "grad_norm": 0.10430077463388443, "learning_rate": 6.553338491169414e-05, "loss": 0.026, "step": 17180 }, { "grad_norm": 0.11712552607059479, "learning_rate": 6.54940882629046e-05, "loss": 0.0241, "step": 17190 }, { "grad_norm": 0.11332487314939499, "learning_rate": 6.545478102405184e-05, "loss": 0.025, "step": 17200 }, { "grad_norm": 0.11598273366689682, "learning_rate": 6.541546322200199e-05, "loss": 0.0265, "step": 17210 }, { "grad_norm": 0.13320951163768768, "learning_rate": 6.537613488362837e-05, "loss": 0.0242, "step": 17220 }, { "grad_norm": 0.13912439346313477, "learning_rate": 6.533679603581155e-05, "loss": 0.0263, "step": 17230 }, { "grad_norm": 0.11784295737743378, "learning_rate": 6.529744670543926e-05, "loss": 0.025, "step": 17240 }, { "grad_norm": 0.08202172815799713, "learning_rate": 6.52580869194064e-05, "loss": 0.0268, "step": 17250 }, { "grad_norm": 0.1591544896364212, "learning_rate": 6.521871670461499e-05, "loss": 0.0248, "step": 17260 }, { "grad_norm": 0.09674517065286636, "learning_rate": 6.517933608797422e-05, "loss": 0.0257, "step": 17270 }, { "grad_norm": 0.1180591806769371, "learning_rate": 6.513994509640038e-05, "loss": 0.0239, "step": 17280 }, { "grad_norm": 0.12177727371454239, "learning_rate": 6.510054375681682e-05, "loss": 0.023, "step": 17290 }, { "grad_norm": 0.10447946190834045, "learning_rate": 6.506113209615398e-05, "loss": 0.026, "step": 17300 }, { "grad_norm": 0.11589759588241577, "learning_rate": 6.502171014134938e-05, "loss": 0.0246, "step": 17310 }, { "grad_norm": 0.10439684242010117, "learning_rate": 6.498227791934755e-05, "loss": 0.0216, "step": 17320 }, { "grad_norm": 0.07754261046648026, "learning_rate": 6.494283545710003e-05, "loss": 0.0252, "step": 17330 }, { "grad_norm": 0.1599469631910324, "learning_rate": 6.490338278156538e-05, "loss": 0.023, "step": 17340 }, { "grad_norm": 0.11156351119279861, "learning_rate": 6.486391991970913e-05, "loss": 0.0235, "step": 17350 }, { "grad_norm": 0.1155569851398468, "learning_rate": 6.482444689850377e-05, "loss": 0.0242, "step": 17360 }, { "grad_norm": 0.10949943959712982, "learning_rate": 6.478496374492875e-05, "loss": 0.0185, "step": 17370 }, { "grad_norm": 0.09448914974927902, "learning_rate": 6.474547048597042e-05, "loss": 0.0202, "step": 17380 }, { "grad_norm": 0.13322821259498596, "learning_rate": 6.470596714862205e-05, "loss": 0.019, "step": 17390 }, { "grad_norm": 0.10481776297092438, "learning_rate": 6.46664537598838e-05, "loss": 0.0204, "step": 17400 }, { "grad_norm": 0.1320735067129135, "learning_rate": 6.462693034676271e-05, "loss": 0.0211, "step": 17410 }, { "grad_norm": 0.12656275928020477, "learning_rate": 6.458739693627265e-05, "loss": 0.0237, "step": 17420 }, { "grad_norm": 0.10844890028238297, "learning_rate": 6.454785355543432e-05, "loss": 0.0258, "step": 17430 }, { "grad_norm": 0.09925205260515213, "learning_rate": 6.450830023127528e-05, "loss": 0.0199, "step": 17440 }, { "grad_norm": 0.0694475844502449, "learning_rate": 6.446873699082982e-05, "loss": 0.0198, "step": 17450 }, { "grad_norm": 0.10481305420398712, "learning_rate": 6.44291638611391e-05, "loss": 0.0215, "step": 17460 }, { "grad_norm": 0.11718112975358963, "learning_rate": 6.43895808692509e-05, "loss": 0.021, "step": 17470 }, { "grad_norm": 0.10071089118719101, "learning_rate": 6.434998804221986e-05, "loss": 0.0235, "step": 17480 }, { "grad_norm": 0.1058105006814003, "learning_rate": 6.431038540710732e-05, "loss": 0.0205, "step": 17490 }, { "grad_norm": 0.09795448184013367, "learning_rate": 6.427077299098129e-05, "loss": 0.021, "step": 17500 }, { "grad_norm": 0.08335208892822266, "learning_rate": 6.423115082091651e-05, "loss": 0.0218, "step": 17510 }, { "grad_norm": 0.07672671973705292, "learning_rate": 6.419151892399429e-05, "loss": 0.024, "step": 17520 }, { "grad_norm": 0.08811488002538681, "learning_rate": 6.415187732730273e-05, "loss": 0.0244, "step": 17530 }, { "grad_norm": 0.11112080514431, "learning_rate": 6.411222605793645e-05, "loss": 0.0274, "step": 17540 }, { "grad_norm": 0.11956927180290222, "learning_rate": 6.407256514299674e-05, "loss": 0.0263, "step": 17550 }, { "grad_norm": 0.10897007584571838, "learning_rate": 6.403289460959147e-05, "loss": 0.0266, "step": 17560 }, { "grad_norm": 0.11389163881540298, "learning_rate": 6.399321448483501e-05, "loss": 0.0244, "step": 17570 }, { "grad_norm": 0.09365329891443253, "learning_rate": 6.395352479584844e-05, "loss": 0.0252, "step": 17580 }, { "grad_norm": 0.07949154078960419, "learning_rate": 6.391382556975923e-05, "loss": 0.0209, "step": 17590 }, { "grad_norm": 0.10240641981363297, "learning_rate": 6.387411683370144e-05, "loss": 0.0218, "step": 17600 }, { "grad_norm": 0.09003661572933197, "learning_rate": 6.383439861481562e-05, "loss": 0.022, "step": 17610 }, { "grad_norm": 0.07642537355422974, "learning_rate": 6.379467094024879e-05, "loss": 0.0234, "step": 17620 }, { "grad_norm": 0.08223005384206772, "learning_rate": 6.375493383715445e-05, "loss": 0.0249, "step": 17630 }, { "grad_norm": 0.09750202298164368, "learning_rate": 6.371518733269254e-05, "loss": 0.0259, "step": 17640 }, { "grad_norm": 0.08342639356851578, "learning_rate": 6.367543145402942e-05, "loss": 0.0228, "step": 17650 }, { "grad_norm": 0.10924555361270905, "learning_rate": 6.363566622833785e-05, "loss": 0.0245, "step": 17660 }, { "grad_norm": 0.11869842559099197, "learning_rate": 6.359589168279698e-05, "loss": 0.0284, "step": 17670 }, { "grad_norm": 0.10604745894670486, "learning_rate": 6.355610784459235e-05, "loss": 0.0238, "step": 17680 }, { "grad_norm": 0.10276948660612106, "learning_rate": 6.351631474091585e-05, "loss": 0.0237, "step": 17690 }, { "grad_norm": 0.09585493803024292, "learning_rate": 6.347651239896566e-05, "loss": 0.0295, "step": 17700 }, { "grad_norm": 0.13081666827201843, "learning_rate": 6.343670084594633e-05, "loss": 0.0242, "step": 17710 }, { "grad_norm": 0.13620522618293762, "learning_rate": 6.339688010906866e-05, "loss": 0.025, "step": 17720 }, { "grad_norm": 0.13964970409870148, "learning_rate": 6.335705021554975e-05, "loss": 0.0274, "step": 17730 }, { "grad_norm": 0.08961072564125061, "learning_rate": 6.3317211192613e-05, "loss": 0.0257, "step": 17740 }, { "grad_norm": 0.11589616537094116, "learning_rate": 6.327736306748795e-05, "loss": 0.0212, "step": 17750 }, { "grad_norm": 0.08601880818605423, "learning_rate": 6.323750586741047e-05, "loss": 0.023, "step": 17760 }, { "grad_norm": 0.08444052934646606, "learning_rate": 6.319763961962252e-05, "loss": 0.0246, "step": 17770 }, { "grad_norm": 0.09863188862800598, "learning_rate": 6.315776435137233e-05, "loss": 0.0238, "step": 17780 }, { "grad_norm": 0.10749661177396774, "learning_rate": 6.311788008991432e-05, "loss": 0.0227, "step": 17790 }, { "grad_norm": 0.08236534148454666, "learning_rate": 6.307798686250891e-05, "loss": 0.0243, "step": 17800 }, { "grad_norm": 0.06580069661140442, "learning_rate": 6.303808469642284e-05, "loss": 0.0222, "step": 17810 }, { "grad_norm": 0.1440226286649704, "learning_rate": 6.29981736189288e-05, "loss": 0.0237, "step": 17820 }, { "grad_norm": 0.11510716378688812, "learning_rate": 6.295825365730567e-05, "loss": 0.0232, "step": 17830 }, { "grad_norm": 0.12075450271368027, "learning_rate": 6.291832483883835e-05, "loss": 0.0273, "step": 17840 }, { "grad_norm": 0.11053382605314255, "learning_rate": 6.28783871908178e-05, "loss": 0.0302, "step": 17850 }, { "grad_norm": 0.137154221534729, "learning_rate": 6.283844074054107e-05, "loss": 0.0229, "step": 17860 }, { "grad_norm": 0.12223689258098602, "learning_rate": 6.279848551531112e-05, "loss": 0.0222, "step": 17870 }, { "grad_norm": 0.09769494831562042, "learning_rate": 6.275852154243702e-05, "loss": 0.0277, "step": 17880 }, { "grad_norm": 0.0803442895412445, "learning_rate": 6.271854884923377e-05, "loss": 0.0226, "step": 17890 }, { "grad_norm": 0.10584941506385803, "learning_rate": 6.267856746302228e-05, "loss": 0.0251, "step": 17900 }, { "grad_norm": 0.1050659567117691, "learning_rate": 6.263857741112948e-05, "loss": 0.0229, "step": 17910 }, { "grad_norm": 0.07697699964046478, "learning_rate": 6.259857872088821e-05, "loss": 0.0281, "step": 17920 }, { "grad_norm": 0.0865393877029419, "learning_rate": 6.255857141963719e-05, "loss": 0.0231, "step": 17930 }, { "grad_norm": 0.1622619777917862, "learning_rate": 6.251855553472101e-05, "loss": 0.0245, "step": 17940 }, { "grad_norm": 0.09305006265640259, "learning_rate": 6.247853109349016e-05, "loss": 0.021, "step": 17950 }, { "grad_norm": 0.09097374230623245, "learning_rate": 6.243849812330098e-05, "loss": 0.0244, "step": 17960 }, { "grad_norm": 0.09013301134109497, "learning_rate": 6.239845665151563e-05, "loss": 0.0258, "step": 17970 }, { "grad_norm": 0.11781954020261765, "learning_rate": 6.235840670550204e-05, "loss": 0.0223, "step": 17980 }, { "grad_norm": 0.10218162089586258, "learning_rate": 6.231834831263403e-05, "loss": 0.0242, "step": 17990 }, { "grad_norm": 0.1066715195775032, "learning_rate": 6.22782815002911e-05, "loss": 0.0247, "step": 18000 }, { "grad_norm": 0.08213401585817337, "learning_rate": 6.223820629585852e-05, "loss": 0.0249, "step": 18010 }, { "grad_norm": 0.0877089723944664, "learning_rate": 6.219812272672737e-05, "loss": 0.0249, "step": 18020 }, { "grad_norm": 0.10619136691093445, "learning_rate": 6.215803082029434e-05, "loss": 0.025, "step": 18030 }, { "grad_norm": 0.09808290004730225, "learning_rate": 6.211793060396188e-05, "loss": 0.0244, "step": 18040 }, { "grad_norm": 0.08800256252288818, "learning_rate": 6.207782210513811e-05, "loss": 0.0256, "step": 18050 }, { "grad_norm": 0.10307000577449799, "learning_rate": 6.203770535123683e-05, "loss": 0.0256, "step": 18060 }, { "grad_norm": 0.11385779827833176, "learning_rate": 6.199758036967747e-05, "loss": 0.0225, "step": 18070 }, { "grad_norm": 0.09969981759786606, "learning_rate": 6.195744718788503e-05, "loss": 0.0253, "step": 18080 }, { "grad_norm": 0.07722598314285278, "learning_rate": 6.191730583329021e-05, "loss": 0.0215, "step": 18090 }, { "grad_norm": 0.095034658908844, "learning_rate": 6.187715633332921e-05, "loss": 0.024, "step": 18100 }, { "grad_norm": 0.11365601420402527, "learning_rate": 6.183699871544386e-05, "loss": 0.0245, "step": 18110 }, { "grad_norm": 0.09500192105770111, "learning_rate": 6.179683300708152e-05, "loss": 0.0237, "step": 18120 }, { "grad_norm": 0.19036078453063965, "learning_rate": 6.175665923569503e-05, "loss": 0.0207, "step": 18130 }, { "grad_norm": 0.12985049188137054, "learning_rate": 6.171647742874281e-05, "loss": 0.0272, "step": 18140 }, { "grad_norm": 0.10178292542695999, "learning_rate": 6.167628761368875e-05, "loss": 0.0221, "step": 18150 }, { "grad_norm": 0.1272452026605606, "learning_rate": 6.163608981800222e-05, "loss": 0.0267, "step": 18160 }, { "grad_norm": 0.09514859318733215, "learning_rate": 6.159588406915803e-05, "loss": 0.022, "step": 18170 }, { "grad_norm": 0.09883498400449753, "learning_rate": 6.155567039463639e-05, "loss": 0.0253, "step": 18180 }, { "grad_norm": 0.0685456395149231, "learning_rate": 6.151544882192302e-05, "loss": 0.0257, "step": 18190 }, { "grad_norm": 0.15536297857761383, "learning_rate": 6.147521937850895e-05, "loss": 0.0222, "step": 18200 }, { "grad_norm": 0.08635911345481873, "learning_rate": 6.143498209189066e-05, "loss": 0.0203, "step": 18210 }, { "grad_norm": 0.12486056983470917, "learning_rate": 6.139473698956993e-05, "loss": 0.0242, "step": 18220 }, { "grad_norm": 0.10807529836893082, "learning_rate": 6.13544840990539e-05, "loss": 0.0256, "step": 18230 }, { "grad_norm": 0.12176147103309631, "learning_rate": 6.131422344785507e-05, "loss": 0.0241, "step": 18240 }, { "grad_norm": 0.10545182228088379, "learning_rate": 6.127395506349119e-05, "loss": 0.0262, "step": 18250 }, { "grad_norm": 0.11150362342596054, "learning_rate": 6.123367897348533e-05, "loss": 0.0225, "step": 18260 }, { "grad_norm": 0.0764080211520195, "learning_rate": 6.119339520536584e-05, "loss": 0.0222, "step": 18270 }, { "grad_norm": 0.079001784324646, "learning_rate": 6.115310378666625e-05, "loss": 0.0207, "step": 18280 }, { "grad_norm": 0.13526524603366852, "learning_rate": 6.11128047449254e-05, "loss": 0.0268, "step": 18290 }, { "grad_norm": 0.1035805344581604, "learning_rate": 6.107249810768729e-05, "loss": 0.0256, "step": 18300 }, { "grad_norm": 0.11150595545768738, "learning_rate": 6.1032183902501125e-05, "loss": 0.0252, "step": 18310 }, { "grad_norm": 0.0841030478477478, "learning_rate": 6.099186215692131e-05, "loss": 0.0204, "step": 18320 }, { "grad_norm": 0.11937849223613739, "learning_rate": 6.095153289850734e-05, "loss": 0.0244, "step": 18330 }, { "grad_norm": 0.08741447329521179, "learning_rate": 6.0911196154823904e-05, "loss": 0.0214, "step": 18340 }, { "grad_norm": 0.09871945530176163, "learning_rate": 6.087085195344079e-05, "loss": 0.0205, "step": 18350 }, { "grad_norm": 0.10479859262704849, "learning_rate": 6.083050032193286e-05, "loss": 0.0248, "step": 18360 }, { "grad_norm": 0.114383764564991, "learning_rate": 6.0790141287880097e-05, "loss": 0.0263, "step": 18370 }, { "grad_norm": 0.10957226157188416, "learning_rate": 6.0749774878867496e-05, "loss": 0.0236, "step": 18380 }, { "grad_norm": 0.10633955150842667, "learning_rate": 6.0709401122485146e-05, "loss": 0.024, "step": 18390 }, { "grad_norm": 0.0799538642168045, "learning_rate": 6.066902004632811e-05, "loss": 0.0247, "step": 18400 }, { "grad_norm": 0.09713631123304367, "learning_rate": 6.062863167799646e-05, "loss": 0.0268, "step": 18410 }, { "grad_norm": 0.08708846569061279, "learning_rate": 6.058823604509529e-05, "loss": 0.0226, "step": 18420 }, { "grad_norm": 0.10298013687133789, "learning_rate": 6.054783317523462e-05, "loss": 0.0237, "step": 18430 }, { "grad_norm": 0.10863717645406723, "learning_rate": 6.050742309602944e-05, "loss": 0.0239, "step": 18440 }, { "grad_norm": 0.09755601733922958, "learning_rate": 6.046700583509965e-05, "loss": 0.0227, "step": 18450 }, { "grad_norm": 0.09702104330062866, "learning_rate": 6.042658142007007e-05, "loss": 0.0231, "step": 18460 }, { "grad_norm": 0.08916006982326508, "learning_rate": 6.038614987857041e-05, "loss": 0.0227, "step": 18470 }, { "grad_norm": 0.10613399744033813, "learning_rate": 6.0345711238235224e-05, "loss": 0.0204, "step": 18480 }, { "grad_norm": 0.08845461159944534, "learning_rate": 6.030526552670399e-05, "loss": 0.0221, "step": 18490 }, { "grad_norm": 0.08058935403823853, "learning_rate": 6.0264812771620925e-05, "loss": 0.0222, "step": 18500 }, { "grad_norm": 0.09265320748090744, "learning_rate": 6.022435300063512e-05, "loss": 0.0274, "step": 18510 }, { "grad_norm": 0.10735330730676651, "learning_rate": 6.0183886241400466e-05, "loss": 0.0217, "step": 18520 }, { "grad_norm": 0.10788482427597046, "learning_rate": 6.0143412521575584e-05, "loss": 0.0228, "step": 18530 }, { "grad_norm": 0.09137016534805298, "learning_rate": 6.010293186882389e-05, "loss": 0.0269, "step": 18540 }, { "grad_norm": 0.09029129892587662, "learning_rate": 6.0062444310813525e-05, "loss": 0.0216, "step": 18550 }, { "grad_norm": 0.09089964628219604, "learning_rate": 6.0021949875217355e-05, "loss": 0.0261, "step": 18560 }, { "grad_norm": 0.09535383433103561, "learning_rate": 5.998144858971295e-05, "loss": 0.0215, "step": 18570 }, { "grad_norm": 0.0931304320693016, "learning_rate": 5.994094048198257e-05, "loss": 0.0217, "step": 18580 }, { "grad_norm": 0.10483108460903168, "learning_rate": 5.990042557971307e-05, "loss": 0.0251, "step": 18590 }, { "grad_norm": 0.07814329862594604, "learning_rate": 5.985990391059607e-05, "loss": 0.0259, "step": 18600 }, { "grad_norm": 0.11408211290836334, "learning_rate": 5.981937550232771e-05, "loss": 0.0234, "step": 18610 }, { "grad_norm": 0.09944428503513336, "learning_rate": 5.9778840382608794e-05, "loss": 0.0209, "step": 18620 }, { "grad_norm": 0.11750015616416931, "learning_rate": 5.9738298579144695e-05, "loss": 0.0219, "step": 18630 }, { "grad_norm": 0.1281544268131256, "learning_rate": 5.9697750119645314e-05, "loss": 0.0219, "step": 18640 }, { "grad_norm": 0.09151139110326767, "learning_rate": 5.96571950318252e-05, "loss": 0.0227, "step": 18650 }, { "grad_norm": 0.10715504735708237, "learning_rate": 5.9616633343403316e-05, "loss": 0.0217, "step": 18660 }, { "grad_norm": 0.0965302512049675, "learning_rate": 5.957606508210324e-05, "loss": 0.0287, "step": 18670 }, { "grad_norm": 0.08906184136867523, "learning_rate": 5.953549027565297e-05, "loss": 0.0219, "step": 18680 }, { "grad_norm": 0.09744079411029816, "learning_rate": 5.949490895178501e-05, "loss": 0.0252, "step": 18690 }, { "grad_norm": 0.10199030488729477, "learning_rate": 5.945432113823632e-05, "loss": 0.0252, "step": 18700 }, { "grad_norm": 0.09390323609113693, "learning_rate": 5.9413726862748276e-05, "loss": 0.0234, "step": 18710 }, { "grad_norm": 0.08757883310317993, "learning_rate": 5.9373126153066694e-05, "loss": 0.022, "step": 18720 }, { "grad_norm": 0.06268960982561111, "learning_rate": 5.933251903694177e-05, "loss": 0.0226, "step": 18730 }, { "grad_norm": 0.10617164522409439, "learning_rate": 5.929190554212807e-05, "loss": 0.0243, "step": 18740 }, { "grad_norm": 0.1289888322353363, "learning_rate": 5.9251285696384565e-05, "loss": 0.0218, "step": 18750 }, { "grad_norm": 0.10549346357584, "learning_rate": 5.921065952747451e-05, "loss": 0.0229, "step": 18760 }, { "grad_norm": 0.11763398349285126, "learning_rate": 5.917002706316552e-05, "loss": 0.0236, "step": 18770 }, { "grad_norm": 0.12283570319414139, "learning_rate": 5.912938833122952e-05, "loss": 0.0257, "step": 18780 }, { "grad_norm": 0.11384017020463943, "learning_rate": 5.908874335944265e-05, "loss": 0.0249, "step": 18790 }, { "grad_norm": 0.07994731515645981, "learning_rate": 5.904809217558542e-05, "loss": 0.0225, "step": 18800 }, { "grad_norm": 0.18065732717514038, "learning_rate": 5.90074348074425e-05, "loss": 0.0258, "step": 18810 }, { "grad_norm": 0.11180703341960907, "learning_rate": 5.8966771282802814e-05, "loss": 0.0251, "step": 18820 }, { "grad_norm": 0.09286147356033325, "learning_rate": 5.892610162945952e-05, "loss": 0.0232, "step": 18830 }, { "grad_norm": 0.07209452241659164, "learning_rate": 5.8885425875209924e-05, "loss": 0.022, "step": 18840 }, { "grad_norm": 0.0979270413517952, "learning_rate": 5.884474404785553e-05, "loss": 0.0261, "step": 18850 }, { "grad_norm": 0.10782334953546524, "learning_rate": 5.8804056175201983e-05, "loss": 0.0224, "step": 18860 }, { "grad_norm": 0.08625907450914383, "learning_rate": 5.876336228505904e-05, "loss": 0.0246, "step": 18870 }, { "grad_norm": 0.07823051512241364, "learning_rate": 5.872266240524062e-05, "loss": 0.0246, "step": 18880 }, { "grad_norm": 0.09257593005895615, "learning_rate": 5.86819565635647e-05, "loss": 0.0224, "step": 18890 }, { "grad_norm": 0.10178530961275101, "learning_rate": 5.8641244787853334e-05, "loss": 0.025, "step": 18900 }, { "grad_norm": 0.09554295241832733, "learning_rate": 5.860052710593265e-05, "loss": 0.0231, "step": 18910 }, { "grad_norm": 0.10686822980642319, "learning_rate": 5.855980354563276e-05, "loss": 0.0272, "step": 18920 }, { "grad_norm": 0.12761549651622772, "learning_rate": 5.8519074134787874e-05, "loss": 0.0235, "step": 18930 }, { "grad_norm": 0.09829400479793549, "learning_rate": 5.847833890123614e-05, "loss": 0.024, "step": 18940 }, { "grad_norm": 0.08487319201231003, "learning_rate": 5.8437597872819737e-05, "loss": 0.0213, "step": 18950 }, { "grad_norm": 0.11980392783880234, "learning_rate": 5.839685107738473e-05, "loss": 0.0243, "step": 18960 }, { "grad_norm": 0.09534226357936859, "learning_rate": 5.835609854278118e-05, "loss": 0.0225, "step": 18970 }, { "grad_norm": 0.058375924825668335, "learning_rate": 5.831534029686308e-05, "loss": 0.0219, "step": 18980 }, { "grad_norm": 0.12928538024425507, "learning_rate": 5.82745763674883e-05, "loss": 0.0214, "step": 18990 }, { "grad_norm": 0.08795175701379776, "learning_rate": 5.823380678251861e-05, "loss": 0.0222, "step": 19000 }, { "grad_norm": 0.08208887279033661, "learning_rate": 5.81930315698196e-05, "loss": 0.0233, "step": 19010 }, { "grad_norm": 0.0738234668970108, "learning_rate": 5.815225075726076e-05, "loss": 0.0252, "step": 19020 }, { "grad_norm": 0.11246830224990845, "learning_rate": 5.811146437271543e-05, "loss": 0.0264, "step": 19030 }, { "grad_norm": 0.116236113011837, "learning_rate": 5.807067244406066e-05, "loss": 0.024, "step": 19040 }, { "grad_norm": 0.08932756632566452, "learning_rate": 5.8029874999177405e-05, "loss": 0.0234, "step": 19050 }, { "grad_norm": 0.11279036104679108, "learning_rate": 5.798907206595029e-05, "loss": 0.0224, "step": 19060 }, { "grad_norm": 0.10356727987527847, "learning_rate": 5.794826367226773e-05, "loss": 0.0246, "step": 19070 }, { "grad_norm": 0.12323072552680969, "learning_rate": 5.790744984602193e-05, "loss": 0.0235, "step": 19080 }, { "grad_norm": 0.11279084533452988, "learning_rate": 5.786663061510872e-05, "loss": 0.0262, "step": 19090 }, { "grad_norm": 0.12416516244411469, "learning_rate": 5.782580600742765e-05, "loss": 0.0244, "step": 19100 }, { "grad_norm": 0.0917622298002243, "learning_rate": 5.7784976050881965e-05, "loss": 0.0235, "step": 19110 }, { "grad_norm": 0.1011379137635231, "learning_rate": 5.774414077337855e-05, "loss": 0.0227, "step": 19120 }, { "grad_norm": 0.10309401899576187, "learning_rate": 5.770330020282796e-05, "loss": 0.0267, "step": 19130 }, { "grad_norm": 0.07254768162965775, "learning_rate": 5.7662454367144317e-05, "loss": 0.0231, "step": 19140 }, { "grad_norm": 0.09939748793840408, "learning_rate": 5.762160329424536e-05, "loss": 0.0265, "step": 19150 }, { "grad_norm": 0.10894471406936646, "learning_rate": 5.7580747012052416e-05, "loss": 0.0257, "step": 19160 }, { "grad_norm": 0.13835011422634125, "learning_rate": 5.753988554849037e-05, "loss": 0.0272, "step": 19170 }, { "grad_norm": 0.08956733345985413, "learning_rate": 5.749901893148766e-05, "loss": 0.0241, "step": 19180 }, { "grad_norm": 0.11088260263204575, "learning_rate": 5.745814718897621e-05, "loss": 0.0255, "step": 19190 }, { "grad_norm": 0.1369616836309433, "learning_rate": 5.74172703488915e-05, "loss": 0.0251, "step": 19200 }, { "grad_norm": 0.09806840121746063, "learning_rate": 5.737638843917242e-05, "loss": 0.0232, "step": 19210 }, { "grad_norm": 0.10909555107355118, "learning_rate": 5.73355014877614e-05, "loss": 0.0229, "step": 19220 }, { "grad_norm": 0.09949888288974762, "learning_rate": 5.7294609522604316e-05, "loss": 0.0253, "step": 19230 }, { "grad_norm": 0.10009405761957169, "learning_rate": 5.7253712571650376e-05, "loss": 0.0222, "step": 19240 }, { "grad_norm": 0.0837487280368805, "learning_rate": 5.721281066285229e-05, "loss": 0.0257, "step": 19250 }, { "grad_norm": 0.0703587681055069, "learning_rate": 5.717190382416615e-05, "loss": 0.0226, "step": 19260 }, { "grad_norm": 0.09217431396245956, "learning_rate": 5.713099208355135e-05, "loss": 0.0208, "step": 19270 }, { "grad_norm": 0.11323139816522598, "learning_rate": 5.709007546897074e-05, "loss": 0.0249, "step": 19280 }, { "grad_norm": 0.09748797118663788, "learning_rate": 5.704915400839037e-05, "loss": 0.0231, "step": 19290 }, { "grad_norm": 0.08301907032728195, "learning_rate": 5.700822772977971e-05, "loss": 0.0197, "step": 19300 }, { "grad_norm": 0.09721042215824127, "learning_rate": 5.696729666111148e-05, "loss": 0.0215, "step": 19310 }, { "grad_norm": 0.09500744193792343, "learning_rate": 5.692636083036168e-05, "loss": 0.0201, "step": 19320 }, { "grad_norm": 0.09421855956315994, "learning_rate": 5.688542026550958e-05, "loss": 0.023, "step": 19330 }, { "grad_norm": 0.14353683590888977, "learning_rate": 5.684447499453763e-05, "loss": 0.0237, "step": 19340 }, { "grad_norm": 0.10733184963464737, "learning_rate": 5.680352504543156e-05, "loss": 0.0229, "step": 19350 }, { "grad_norm": 0.08215567469596863, "learning_rate": 5.67625704461803e-05, "loss": 0.0201, "step": 19360 }, { "grad_norm": 0.07696142047643661, "learning_rate": 5.672161122477589e-05, "loss": 0.0198, "step": 19370 }, { "grad_norm": 0.1449175626039505, "learning_rate": 5.668064740921359e-05, "loss": 0.0234, "step": 19380 }, { "grad_norm": 0.07558810710906982, "learning_rate": 5.663967902749179e-05, "loss": 0.0216, "step": 19390 }, { "grad_norm": 0.10681329667568207, "learning_rate": 5.6598706107611965e-05, "loss": 0.0207, "step": 19400 }, { "grad_norm": 0.08569058030843735, "learning_rate": 5.655772867757876e-05, "loss": 0.0226, "step": 19410 }, { "grad_norm": 0.105869822204113, "learning_rate": 5.651674676539982e-05, "loss": 0.0213, "step": 19420 }, { "grad_norm": 0.09118977934122086, "learning_rate": 5.647576039908593e-05, "loss": 0.0203, "step": 19430 }, { "grad_norm": 0.10558758676052094, "learning_rate": 5.6434769606650864e-05, "loss": 0.0239, "step": 19440 }, { "grad_norm": 0.0985925942659378, "learning_rate": 5.639377441611143e-05, "loss": 0.0206, "step": 19450 }, { "grad_norm": 0.10598591715097427, "learning_rate": 5.635277485548751e-05, "loss": 0.0197, "step": 19460 }, { "grad_norm": 0.1156068742275238, "learning_rate": 5.631177095280186e-05, "loss": 0.0232, "step": 19470 }, { "grad_norm": 0.11462223529815674, "learning_rate": 5.627076273608027e-05, "loss": 0.0269, "step": 19480 }, { "grad_norm": 0.11431332677602768, "learning_rate": 5.622975023335148e-05, "loss": 0.0257, "step": 19490 }, { "grad_norm": 0.09453441947698593, "learning_rate": 5.618873347264716e-05, "loss": 0.0209, "step": 19500 }, { "grad_norm": 0.09783489257097244, "learning_rate": 5.614771248200188e-05, "loss": 0.0227, "step": 19510 }, { "grad_norm": 0.11431585997343063, "learning_rate": 5.6106687289453066e-05, "loss": 0.0199, "step": 19520 }, { "grad_norm": 0.1021452248096466, "learning_rate": 5.606565792304108e-05, "loss": 0.022, "step": 19530 }, { "grad_norm": 0.11729742586612701, "learning_rate": 5.602462441080909e-05, "loss": 0.0224, "step": 19540 }, { "grad_norm": 0.10811246186494827, "learning_rate": 5.5983586780803135e-05, "loss": 0.0267, "step": 19550 }, { "grad_norm": 0.09754738956689835, "learning_rate": 5.594254506107205e-05, "loss": 0.0219, "step": 19560 }, { "grad_norm": 0.09738756716251373, "learning_rate": 5.590149927966743e-05, "loss": 0.0262, "step": 19570 }, { "grad_norm": 0.1365385204553604, "learning_rate": 5.58604494646437e-05, "loss": 0.0235, "step": 19580 }, { "grad_norm": 0.09698004275560379, "learning_rate": 5.5819395644058025e-05, "loss": 0.0205, "step": 19590 }, { "grad_norm": 0.08613412827253342, "learning_rate": 5.577833784597031e-05, "loss": 0.0212, "step": 19600 }, { "grad_norm": 0.1127835065126419, "learning_rate": 5.573727609844316e-05, "loss": 0.019, "step": 19610 }, { "grad_norm": 0.10981825739145279, "learning_rate": 5.5696210429541884e-05, "loss": 0.0233, "step": 19620 }, { "grad_norm": 0.10792407393455505, "learning_rate": 5.565514086733451e-05, "loss": 0.0248, "step": 19630 }, { "grad_norm": 0.07715045660734177, "learning_rate": 5.5614067439891657e-05, "loss": 0.0199, "step": 19640 }, { "grad_norm": 0.08429106324911118, "learning_rate": 5.557299017528666e-05, "loss": 0.0202, "step": 19650 }, { "grad_norm": 0.08680865168571472, "learning_rate": 5.5531909101595436e-05, "loss": 0.0256, "step": 19660 }, { "grad_norm": 0.10301195085048676, "learning_rate": 5.549082424689649e-05, "loss": 0.0245, "step": 19670 }, { "grad_norm": 0.09871276468038559, "learning_rate": 5.544973563927095e-05, "loss": 0.0199, "step": 19680 }, { "grad_norm": 0.10401936620473862, "learning_rate": 5.540864330680249e-05, "loss": 0.0213, "step": 19690 }, { "grad_norm": 0.11151929199695587, "learning_rate": 5.536754727757733e-05, "loss": 0.029, "step": 19700 }, { "grad_norm": 0.11848022043704987, "learning_rate": 5.532644757968422e-05, "loss": 0.0234, "step": 19710 }, { "grad_norm": 0.10717544704675674, "learning_rate": 5.528534424121441e-05, "loss": 0.024, "step": 19720 }, { "grad_norm": 0.10368771106004715, "learning_rate": 5.524423729026165e-05, "loss": 0.0235, "step": 19730 }, { "grad_norm": 0.1126420646905899, "learning_rate": 5.5203126754922164e-05, "loss": 0.0252, "step": 19740 }, { "grad_norm": 0.08629179745912552, "learning_rate": 5.5162012663294585e-05, "loss": 0.0243, "step": 19750 }, { "grad_norm": 0.11611433327198029, "learning_rate": 5.512089504348003e-05, "loss": 0.0232, "step": 19760 }, { "grad_norm": 0.11485461890697479, "learning_rate": 5.5079773923582e-05, "loss": 0.025, "step": 19770 }, { "grad_norm": 0.08752971142530441, "learning_rate": 5.50386493317064e-05, "loss": 0.0239, "step": 19780 }, { "grad_norm": 0.1061713770031929, "learning_rate": 5.49975212959615e-05, "loss": 0.023, "step": 19790 }, { "grad_norm": 0.0901169702410698, "learning_rate": 5.4956389844457904e-05, "loss": 0.0209, "step": 19800 }, { "grad_norm": 0.10209798067808151, "learning_rate": 5.491525500530859e-05, "loss": 0.0225, "step": 19810 }, { "grad_norm": 0.10239914804697037, "learning_rate": 5.487411680662882e-05, "loss": 0.0222, "step": 19820 }, { "grad_norm": 0.0605178102850914, "learning_rate": 5.483297527653618e-05, "loss": 0.021, "step": 19830 }, { "grad_norm": 0.10753088444471359, "learning_rate": 5.4791830443150516e-05, "loss": 0.0252, "step": 19840 }, { "grad_norm": 0.08590072393417358, "learning_rate": 5.475068233459392e-05, "loss": 0.025, "step": 19850 }, { "grad_norm": 0.13675731420516968, "learning_rate": 5.470953097899075e-05, "loss": 0.0241, "step": 19860 }, { "grad_norm": 0.10397081077098846, "learning_rate": 5.466837640446756e-05, "loss": 0.0242, "step": 19870 }, { "grad_norm": 0.09642099589109421, "learning_rate": 5.462721863915312e-05, "loss": 0.0202, "step": 19880 }, { "grad_norm": 0.065709687769413, "learning_rate": 5.4586057711178374e-05, "loss": 0.0224, "step": 19890 }, { "grad_norm": 0.13568030297756195, "learning_rate": 5.454489364867642e-05, "loss": 0.0225, "step": 19900 }, { "grad_norm": 0.08185358345508575, "learning_rate": 5.4503726479782523e-05, "loss": 0.0204, "step": 19910 }, { "grad_norm": 0.08872582018375397, "learning_rate": 5.446255623263403e-05, "loss": 0.0203, "step": 19920 }, { "grad_norm": 0.12019002437591553, "learning_rate": 5.4421382935370445e-05, "loss": 0.0203, "step": 19930 }, { "grad_norm": 0.08029188215732574, "learning_rate": 5.438020661613331e-05, "loss": 0.0207, "step": 19940 }, { "grad_norm": 0.0980476513504982, "learning_rate": 5.433902730306625e-05, "loss": 0.0224, "step": 19950 }, { "grad_norm": 0.11460363119840622, "learning_rate": 5.429784502431495e-05, "loss": 0.0217, "step": 19960 }, { "grad_norm": 0.1039150059223175, "learning_rate": 5.42566598080271e-05, "loss": 0.0235, "step": 19970 }, { "grad_norm": 0.08290011435747147, "learning_rate": 5.421547168235241e-05, "loss": 0.0206, "step": 19980 }, { "grad_norm": 0.09412630647420883, "learning_rate": 5.417428067544258e-05, "loss": 0.0242, "step": 19990 }, { "grad_norm": 0.08623050153255463, "learning_rate": 5.413308681545126e-05, "loss": 0.0217, "step": 20000 }, { "grad_norm": 0.11029179394245148, "learning_rate": 5.409189013053408e-05, "loss": 0.0225, "step": 20010 }, { "grad_norm": 0.1139431819319725, "learning_rate": 5.4050690648848576e-05, "loss": 0.0208, "step": 20020 }, { "grad_norm": 0.06876649707555771, "learning_rate": 5.400948839855421e-05, "loss": 0.0241, "step": 20030 }, { "grad_norm": 0.18629515171051025, "learning_rate": 5.396828340781234e-05, "loss": 0.0273, "step": 20040 }, { "grad_norm": 0.17006203532218933, "learning_rate": 5.392707570478617e-05, "loss": 0.0261, "step": 20050 }, { "grad_norm": 0.13303059339523315, "learning_rate": 5.388586531764078e-05, "loss": 0.0254, "step": 20060 }, { "grad_norm": 0.14061567187309265, "learning_rate": 5.384465227454311e-05, "loss": 0.0223, "step": 20070 }, { "grad_norm": 0.12148621678352356, "learning_rate": 5.380343660366184e-05, "loss": 0.0231, "step": 20080 }, { "grad_norm": 0.12470987439155579, "learning_rate": 5.376221833316752e-05, "loss": 0.0233, "step": 20090 }, { "grad_norm": 0.1378064900636673, "learning_rate": 5.3720997491232436e-05, "loss": 0.0231, "step": 20100 }, { "grad_norm": 0.1116957738995552, "learning_rate": 5.367977410603068e-05, "loss": 0.0233, "step": 20110 }, { "grad_norm": 0.14751620590686798, "learning_rate": 5.3638548205738004e-05, "loss": 0.0223, "step": 20120 }, { "grad_norm": 0.15373462438583374, "learning_rate": 5.359731981853194e-05, "loss": 0.0213, "step": 20130 }, { "grad_norm": 0.09329284727573395, "learning_rate": 5.35560889725917e-05, "loss": 0.026, "step": 20140 }, { "grad_norm": 0.10393887758255005, "learning_rate": 5.3514855696098176e-05, "loss": 0.0262, "step": 20150 }, { "grad_norm": 0.10952384769916534, "learning_rate": 5.347362001723394e-05, "loss": 0.0262, "step": 20160 }, { "grad_norm": 0.10652220994234085, "learning_rate": 5.3432381964183176e-05, "loss": 0.021, "step": 20170 }, { "grad_norm": 0.1469632238149643, "learning_rate": 5.3391141565131685e-05, "loss": 0.0244, "step": 20180 }, { "grad_norm": 0.09232247620820999, "learning_rate": 5.3349898848266935e-05, "loss": 0.0221, "step": 20190 }, { "grad_norm": 0.10425262898206711, "learning_rate": 5.330865384177789e-05, "loss": 0.0284, "step": 20200 }, { "grad_norm": 0.10844796895980835, "learning_rate": 5.326740657385515e-05, "loss": 0.0226, "step": 20210 }, { "grad_norm": 0.12075062096118927, "learning_rate": 5.322615707269083e-05, "loss": 0.0246, "step": 20220 }, { "grad_norm": 0.09591244906187057, "learning_rate": 5.318490536647856e-05, "loss": 0.0212, "step": 20230 }, { "grad_norm": 0.1300162672996521, "learning_rate": 5.3143651483413524e-05, "loss": 0.0252, "step": 20240 }, { "grad_norm": 0.11177203059196472, "learning_rate": 5.310239545169232e-05, "loss": 0.0186, "step": 20250 }, { "grad_norm": 0.07703730463981628, "learning_rate": 5.30611372995131e-05, "loss": 0.0218, "step": 20260 }, { "grad_norm": 0.08072656393051147, "learning_rate": 5.30198770550754e-05, "loss": 0.0222, "step": 20270 }, { "grad_norm": 0.09615100920200348, "learning_rate": 5.297861474658019e-05, "loss": 0.0192, "step": 20280 }, { "grad_norm": 0.06872208416461945, "learning_rate": 5.29373504022299e-05, "loss": 0.0193, "step": 20290 }, { "grad_norm": 0.10126229375600815, "learning_rate": 5.28960840502283e-05, "loss": 0.0206, "step": 20300 }, { "grad_norm": 0.0794999897480011, "learning_rate": 5.285481571878056e-05, "loss": 0.019, "step": 20310 }, { "grad_norm": 0.07188725471496582, "learning_rate": 5.281354543609321e-05, "loss": 0.0203, "step": 20320 }, { "grad_norm": 0.06111828610301018, "learning_rate": 5.277227323037406e-05, "loss": 0.0213, "step": 20330 }, { "grad_norm": 0.12517230212688446, "learning_rate": 5.273099912983233e-05, "loss": 0.0267, "step": 20340 }, { "grad_norm": 0.08403833210468292, "learning_rate": 5.268972316267843e-05, "loss": 0.0194, "step": 20350 }, { "grad_norm": 0.11597929149866104, "learning_rate": 5.26484453571241e-05, "loss": 0.0228, "step": 20360 }, { "grad_norm": 0.07966574281454086, "learning_rate": 5.260716574138235e-05, "loss": 0.0205, "step": 20370 }, { "grad_norm": 0.07653860747814178, "learning_rate": 5.256588434366739e-05, "loss": 0.0192, "step": 20380 }, { "grad_norm": 0.0718764141201973, "learning_rate": 5.25246011921947e-05, "loss": 0.0215, "step": 20390 }, { "grad_norm": 0.09509217739105225, "learning_rate": 5.248331631518089e-05, "loss": 0.0191, "step": 20400 }, { "grad_norm": 0.08997789770364761, "learning_rate": 5.244202974084379e-05, "loss": 0.0217, "step": 20410 }, { "grad_norm": 0.09017548710107803, "learning_rate": 5.240074149740239e-05, "loss": 0.0195, "step": 20420 }, { "grad_norm": 0.09962678700685501, "learning_rate": 5.2359451613076814e-05, "loss": 0.0214, "step": 20430 }, { "grad_norm": 0.09386661648750305, "learning_rate": 5.231816011608832e-05, "loss": 0.0224, "step": 20440 }, { "grad_norm": 0.09860027581453323, "learning_rate": 5.227686703465924e-05, "loss": 0.0191, "step": 20450 }, { "grad_norm": 0.10182858258485794, "learning_rate": 5.2235572397013e-05, "loss": 0.0216, "step": 20460 }, { "grad_norm": 0.0967497006058693, "learning_rate": 5.2194276231374114e-05, "loss": 0.0207, "step": 20470 }, { "grad_norm": 0.1174670159816742, "learning_rate": 5.21529785659681e-05, "loss": 0.0193, "step": 20480 }, { "grad_norm": 0.07318185269832611, "learning_rate": 5.2111679429021565e-05, "loss": 0.0193, "step": 20490 }, { "grad_norm": 0.12158562988042831, "learning_rate": 5.207037884876205e-05, "loss": 0.0192, "step": 20500 }, { "grad_norm": 0.11162712424993515, "learning_rate": 5.202907685341809e-05, "loss": 0.0236, "step": 20510 }, { "grad_norm": 0.09560039639472961, "learning_rate": 5.198777347121926e-05, "loss": 0.0214, "step": 20520 }, { "grad_norm": 0.09790568798780441, "learning_rate": 5.194646873039598e-05, "loss": 0.0197, "step": 20530 }, { "grad_norm": 0.082989402115345, "learning_rate": 5.1905162659179696e-05, "loss": 0.023, "step": 20540 }, { "grad_norm": 0.3078877925872803, "learning_rate": 5.18638552858027e-05, "loss": 0.0212, "step": 20550 }, { "grad_norm": 0.0908958911895752, "learning_rate": 5.182254663849818e-05, "loss": 0.0236, "step": 20560 }, { "grad_norm": 0.07455337792634964, "learning_rate": 5.178123674550023e-05, "loss": 0.0218, "step": 20570 }, { "grad_norm": 0.0978764072060585, "learning_rate": 5.173992563504375e-05, "loss": 0.0231, "step": 20580 }, { "grad_norm": 0.08269522339105606, "learning_rate": 5.169861333536451e-05, "loss": 0.0198, "step": 20590 }, { "grad_norm": 0.099516861140728, "learning_rate": 5.165729987469907e-05, "loss": 0.0245, "step": 20600 }, { "grad_norm": 0.1286890059709549, "learning_rate": 5.161598528128478e-05, "loss": 0.0242, "step": 20610 }, { "grad_norm": 0.08021556586027145, "learning_rate": 5.157466958335981e-05, "loss": 0.0193, "step": 20620 }, { "grad_norm": 0.07921124249696732, "learning_rate": 5.1533352809163025e-05, "loss": 0.0188, "step": 20630 }, { "grad_norm": 0.06401781737804413, "learning_rate": 5.1492034986934046e-05, "loss": 0.0204, "step": 20640 }, { "grad_norm": 0.10123753547668457, "learning_rate": 5.1450716144913225e-05, "loss": 0.0222, "step": 20650 }, { "grad_norm": 0.15261198580265045, "learning_rate": 5.1409396311341595e-05, "loss": 0.0248, "step": 20660 }, { "grad_norm": 0.10691164433956146, "learning_rate": 5.136807551446089e-05, "loss": 0.0221, "step": 20670 }, { "grad_norm": 0.09075574576854706, "learning_rate": 5.132675378251346e-05, "loss": 0.0191, "step": 20680 }, { "grad_norm": 0.0808049663901329, "learning_rate": 5.1285431143742325e-05, "loss": 0.0218, "step": 20690 }, { "grad_norm": 0.09792470932006836, "learning_rate": 5.1244107626391136e-05, "loss": 0.0215, "step": 20700 }, { "grad_norm": 0.10082340240478516, "learning_rate": 5.12027832587041e-05, "loss": 0.0204, "step": 20710 }, { "grad_norm": 0.0751948282122612, "learning_rate": 5.116145806892607e-05, "loss": 0.022, "step": 20720 }, { "grad_norm": 0.0771419033408165, "learning_rate": 5.1120132085302384e-05, "loss": 0.022, "step": 20730 }, { "grad_norm": 0.09767752140760422, "learning_rate": 5.107880533607898e-05, "loss": 0.0204, "step": 20740 }, { "grad_norm": 0.12676164507865906, "learning_rate": 5.103747784950231e-05, "loss": 0.0203, "step": 20750 }, { "grad_norm": 0.09440334886312485, "learning_rate": 5.09961496538193e-05, "loss": 0.021, "step": 20760 }, { "grad_norm": 0.0927821546792984, "learning_rate": 5.095482077727742e-05, "loss": 0.0219, "step": 20770 }, { "grad_norm": 0.10984862595796585, "learning_rate": 5.091349124812452e-05, "loss": 0.0256, "step": 20780 }, { "grad_norm": 0.16898944973945618, "learning_rate": 5.087216109460897e-05, "loss": 0.0237, "step": 20790 }, { "grad_norm": 0.08814775943756104, "learning_rate": 5.083083034497954e-05, "loss": 0.0239, "step": 20800 }, { "grad_norm": 0.08687160909175873, "learning_rate": 5.07894990274854e-05, "loss": 0.0203, "step": 20810 }, { "grad_norm": 0.08443199098110199, "learning_rate": 5.074816717037614e-05, "loss": 0.0236, "step": 20820 }, { "grad_norm": 0.11787860840559006, "learning_rate": 5.070683480190165e-05, "loss": 0.0215, "step": 20830 }, { "grad_norm": 0.09607773274183273, "learning_rate": 5.066550195031223e-05, "loss": 0.0232, "step": 20840 }, { "grad_norm": 0.07003391534090042, "learning_rate": 5.062416864385852e-05, "loss": 0.0205, "step": 20850 }, { "grad_norm": 0.08027158677577972, "learning_rate": 5.058283491079142e-05, "loss": 0.0192, "step": 20860 }, { "grad_norm": 0.1596497744321823, "learning_rate": 5.054150077936216e-05, "loss": 0.0216, "step": 20870 }, { "grad_norm": 0.09560534358024597, "learning_rate": 5.0500166277822214e-05, "loss": 0.022, "step": 20880 }, { "grad_norm": 0.0857379361987114, "learning_rate": 5.0458831434423334e-05, "loss": 0.0187, "step": 20890 }, { "grad_norm": 0.10968194901943207, "learning_rate": 5.0417496277417506e-05, "loss": 0.0225, "step": 20900 }, { "grad_norm": 0.10435151308774948, "learning_rate": 5.037616083505691e-05, "loss": 0.0217, "step": 20910 }, { "grad_norm": 0.09486225992441177, "learning_rate": 5.0334825135593935e-05, "loss": 0.0231, "step": 20920 }, { "grad_norm": 0.10033102333545685, "learning_rate": 5.029348920728111e-05, "loss": 0.0206, "step": 20930 }, { "grad_norm": 0.11347873508930206, "learning_rate": 5.0252153078371186e-05, "loss": 0.0224, "step": 20940 }, { "grad_norm": 0.08258987218141556, "learning_rate": 5.021081677711704e-05, "loss": 0.0212, "step": 20950 }, { "grad_norm": 0.08439753949642181, "learning_rate": 5.016948033177159e-05, "loss": 0.0245, "step": 20960 }, { "grad_norm": 0.10196356475353241, "learning_rate": 5.012814377058793e-05, "loss": 0.0232, "step": 20970 }, { "grad_norm": 0.08307132869958878, "learning_rate": 5.008680712181921e-05, "loss": 0.022, "step": 20980 }, { "grad_norm": 0.15539349615573883, "learning_rate": 5.0045470413718645e-05, "loss": 0.0256, "step": 20990 }, { "grad_norm": 0.12809066474437714, "learning_rate": 5.00041336745395e-05, "loss": 0.0285, "step": 21000 }, { "grad_norm": 0.07646511495113373, "learning_rate": 4.996279693253499e-05, "loss": 0.0203, "step": 21010 }, { "grad_norm": 0.08647359907627106, "learning_rate": 4.992146021595847e-05, "loss": 0.0211, "step": 21020 }, { "grad_norm": 0.08904685080051422, "learning_rate": 4.988012355306313e-05, "loss": 0.0218, "step": 21030 }, { "grad_norm": 0.08488861471414566, "learning_rate": 4.98387869721022e-05, "loss": 0.0196, "step": 21040 }, { "grad_norm": 0.1195412427186966, "learning_rate": 4.9797450501328866e-05, "loss": 0.0237, "step": 21050 }, { "grad_norm": 0.09538999944925308, "learning_rate": 4.97561141689962e-05, "loss": 0.0245, "step": 21060 }, { "grad_norm": 0.10885710269212723, "learning_rate": 4.971477800335721e-05, "loss": 0.0205, "step": 21070 }, { "grad_norm": 0.10165664553642273, "learning_rate": 4.967344203266475e-05, "loss": 0.0263, "step": 21080 }, { "grad_norm": 0.10404407978057861, "learning_rate": 4.9632106285171584e-05, "loss": 0.0207, "step": 21090 }, { "grad_norm": 0.09765078872442245, "learning_rate": 4.959077078913031e-05, "loss": 0.0237, "step": 21100 }, { "grad_norm": 0.12518349289894104, "learning_rate": 4.954943557279333e-05, "loss": 0.021, "step": 21110 }, { "grad_norm": 0.10025406628847122, "learning_rate": 4.9508100664412916e-05, "loss": 0.0224, "step": 21120 }, { "grad_norm": 0.09345949441194534, "learning_rate": 4.946676609224105e-05, "loss": 0.0205, "step": 21130 }, { "grad_norm": 0.08670105785131454, "learning_rate": 4.942543188452952e-05, "loss": 0.0219, "step": 21140 }, { "grad_norm": 0.10684351623058319, "learning_rate": 4.938409806952988e-05, "loss": 0.022, "step": 21150 }, { "grad_norm": 0.13698922097682953, "learning_rate": 4.93427646754934e-05, "loss": 0.0208, "step": 21160 }, { "grad_norm": 0.08596085011959076, "learning_rate": 4.930143173067108e-05, "loss": 0.024, "step": 21170 }, { "grad_norm": 0.14422181248664856, "learning_rate": 4.9260099263313565e-05, "loss": 0.0209, "step": 21180 }, { "grad_norm": 0.08708267658948898, "learning_rate": 4.921876730167123e-05, "loss": 0.0199, "step": 21190 }, { "grad_norm": 0.0838143453001976, "learning_rate": 4.917743587399409e-05, "loss": 0.0216, "step": 21200 }, { "grad_norm": 0.10118640214204788, "learning_rate": 4.913610500853178e-05, "loss": 0.0208, "step": 21210 }, { "grad_norm": 0.09435057640075684, "learning_rate": 4.909477473353354e-05, "loss": 0.021, "step": 21220 }, { "grad_norm": 0.08675588667392731, "learning_rate": 4.9053445077248236e-05, "loss": 0.0229, "step": 21230 }, { "grad_norm": 0.11872036010026932, "learning_rate": 4.901211606792429e-05, "loss": 0.0234, "step": 21240 }, { "grad_norm": 0.09607300162315369, "learning_rate": 4.89707877338097e-05, "loss": 0.0207, "step": 21250 }, { "grad_norm": 0.08719946444034576, "learning_rate": 4.892946010315199e-05, "loss": 0.0235, "step": 21260 }, { "grad_norm": 0.0960119292140007, "learning_rate": 4.8888133204198204e-05, "loss": 0.025, "step": 21270 }, { "grad_norm": 0.08671178668737411, "learning_rate": 4.8846807065194886e-05, "loss": 0.0233, "step": 21280 }, { "grad_norm": 0.10044905543327332, "learning_rate": 4.880548171438806e-05, "loss": 0.0213, "step": 21290 }, { "grad_norm": 0.0781872346997261, "learning_rate": 4.8764157180023245e-05, "loss": 0.0234, "step": 21300 }, { "grad_norm": 0.08940071612596512, "learning_rate": 4.872283349034533e-05, "loss": 0.0182, "step": 21310 }, { "grad_norm": 0.0828508660197258, "learning_rate": 4.8681510673598674e-05, "loss": 0.0198, "step": 21320 }, { "grad_norm": 0.0983411893248558, "learning_rate": 4.8640188758027046e-05, "loss": 0.0193, "step": 21330 }, { "grad_norm": 0.10469300299882889, "learning_rate": 4.859886777187357e-05, "loss": 0.0211, "step": 21340 }, { "grad_norm": 0.1029466986656189, "learning_rate": 4.855754774338077e-05, "loss": 0.0192, "step": 21350 }, { "grad_norm": 0.0968843400478363, "learning_rate": 4.851622870079048e-05, "loss": 0.0239, "step": 21360 }, { "grad_norm": 0.19529154896736145, "learning_rate": 4.847491067234389e-05, "loss": 0.0227, "step": 21370 }, { "grad_norm": 0.10435697436332703, "learning_rate": 4.843359368628146e-05, "loss": 0.0208, "step": 21380 }, { "grad_norm": 0.0849507749080658, "learning_rate": 4.8392277770842975e-05, "loss": 0.0206, "step": 21390 }, { "grad_norm": 0.07761618494987488, "learning_rate": 4.83509629542675e-05, "loss": 0.0246, "step": 21400 }, { "grad_norm": 0.08613349497318268, "learning_rate": 4.830964926479329e-05, "loss": 0.0203, "step": 21410 }, { "grad_norm": 0.08155950158834457, "learning_rate": 4.826833673065785e-05, "loss": 0.0213, "step": 21420 }, { "grad_norm": 0.06687166541814804, "learning_rate": 4.822702538009794e-05, "loss": 0.0214, "step": 21430 }, { "grad_norm": 0.11575765162706375, "learning_rate": 4.818571524134945e-05, "loss": 0.0217, "step": 21440 }, { "grad_norm": 0.14859235286712646, "learning_rate": 4.8144406342647496e-05, "loss": 0.0233, "step": 21450 }, { "grad_norm": 0.07817673683166504, "learning_rate": 4.81030987122263e-05, "loss": 0.0191, "step": 21460 }, { "grad_norm": 0.10507060587406158, "learning_rate": 4.806179237831926e-05, "loss": 0.0215, "step": 21470 }, { "grad_norm": 0.10623795539140701, "learning_rate": 4.802048736915884e-05, "loss": 0.0211, "step": 21480 }, { "grad_norm": 0.06767714023590088, "learning_rate": 4.797918371297666e-05, "loss": 0.0207, "step": 21490 }, { "grad_norm": 0.08123829960823059, "learning_rate": 4.793788143800334e-05, "loss": 0.0211, "step": 21500 }, { "grad_norm": 0.07835143059492111, "learning_rate": 4.789658057246862e-05, "loss": 0.0191, "step": 21510 }, { "grad_norm": 0.09023585170507431, "learning_rate": 4.7855281144601227e-05, "loss": 0.0205, "step": 21520 }, { "grad_norm": 0.06799225509166718, "learning_rate": 4.781398318262897e-05, "loss": 0.0202, "step": 21530 }, { "grad_norm": 0.1364372968673706, "learning_rate": 4.777268671477858e-05, "loss": 0.0197, "step": 21540 }, { "grad_norm": 0.12988990545272827, "learning_rate": 4.773139176927582e-05, "loss": 0.0174, "step": 21550 }, { "grad_norm": 0.08441752940416336, "learning_rate": 4.769009837434539e-05, "loss": 0.0173, "step": 21560 }, { "grad_norm": 0.09880684316158295, "learning_rate": 4.764880655821095e-05, "loss": 0.0194, "step": 21570 }, { "grad_norm": 0.07720812410116196, "learning_rate": 4.760751634909508e-05, "loss": 0.0197, "step": 21580 }, { "grad_norm": 0.0696038082242012, "learning_rate": 4.756622777521919e-05, "loss": 0.023, "step": 21590 }, { "grad_norm": 0.10342488437891006, "learning_rate": 4.752494086480368e-05, "loss": 0.0183, "step": 21600 }, { "grad_norm": 0.09064651280641556, "learning_rate": 4.7483655646067744e-05, "loss": 0.0223, "step": 21610 }, { "grad_norm": 0.09831590205430984, "learning_rate": 4.744237214722944e-05, "loss": 0.0202, "step": 21620 }, { "grad_norm": 0.10788817703723907, "learning_rate": 4.740109039650567e-05, "loss": 0.0201, "step": 21630 }, { "grad_norm": 0.11089778691530228, "learning_rate": 4.73598104221121e-05, "loss": 0.0206, "step": 21640 }, { "grad_norm": 0.08799317479133606, "learning_rate": 4.731853225226322e-05, "loss": 0.0187, "step": 21650 }, { "grad_norm": 0.06753381341695786, "learning_rate": 4.727725591517225e-05, "loss": 0.0185, "step": 21660 }, { "grad_norm": 0.07734974473714828, "learning_rate": 4.723598143905119e-05, "loss": 0.0199, "step": 21670 }, { "grad_norm": 0.10752318799495697, "learning_rate": 4.719470885211077e-05, "loss": 0.0218, "step": 21680 }, { "grad_norm": 0.09976155310869217, "learning_rate": 4.7153438182560387e-05, "loss": 0.0208, "step": 21690 }, { "grad_norm": 0.08038180321455002, "learning_rate": 4.711216945860815e-05, "loss": 0.0219, "step": 21700 }, { "grad_norm": 0.08184488862752914, "learning_rate": 4.707090270846088e-05, "loss": 0.0183, "step": 21710 }, { "grad_norm": 0.09182453900575638, "learning_rate": 4.702963796032397e-05, "loss": 0.0203, "step": 21720 }, { "grad_norm": 0.10214152932167053, "learning_rate": 4.6988375242401514e-05, "loss": 0.0212, "step": 21730 }, { "grad_norm": 0.18896549940109253, "learning_rate": 4.694711458289618e-05, "loss": 0.0171, "step": 21740 }, { "grad_norm": 0.09457434713840485, "learning_rate": 4.690585601000925e-05, "loss": 0.019, "step": 21750 }, { "grad_norm": 0.10914354771375656, "learning_rate": 4.686459955194055e-05, "loss": 0.0212, "step": 21760 }, { "grad_norm": 0.08721782267093658, "learning_rate": 4.6823345236888504e-05, "loss": 0.0198, "step": 21770 }, { "grad_norm": 0.07320486754179001, "learning_rate": 4.678209309305002e-05, "loss": 0.02, "step": 21780 }, { "grad_norm": 0.09828253835439682, "learning_rate": 4.674084314862057e-05, "loss": 0.0213, "step": 21790 }, { "grad_norm": 0.13961242139339447, "learning_rate": 4.669959543179409e-05, "loss": 0.0191, "step": 21800 }, { "grad_norm": 0.08816280961036682, "learning_rate": 4.665834997076303e-05, "loss": 0.022, "step": 21810 }, { "grad_norm": 0.07064077258110046, "learning_rate": 4.661710679371823e-05, "loss": 0.0169, "step": 21820 }, { "grad_norm": 0.11053887009620667, "learning_rate": 4.657586592884905e-05, "loss": 0.0214, "step": 21830 }, { "grad_norm": 0.11212071776390076, "learning_rate": 4.653462740434322e-05, "loss": 0.0187, "step": 21840 }, { "grad_norm": 0.08129701763391495, "learning_rate": 4.649339124838689e-05, "loss": 0.0217, "step": 21850 }, { "grad_norm": 0.09281933307647705, "learning_rate": 4.6452157489164574e-05, "loss": 0.0188, "step": 21860 }, { "grad_norm": 0.10990185290575027, "learning_rate": 4.6410926154859155e-05, "loss": 0.0212, "step": 21870 }, { "grad_norm": 0.14211222529411316, "learning_rate": 4.636969727365186e-05, "loss": 0.0194, "step": 21880 }, { "grad_norm": 0.07190936803817749, "learning_rate": 4.632847087372226e-05, "loss": 0.0195, "step": 21890 }, { "grad_norm": 0.062119439244270325, "learning_rate": 4.628724698324818e-05, "loss": 0.0181, "step": 21900 }, { "grad_norm": 0.08105101436376572, "learning_rate": 4.6246025630405795e-05, "loss": 0.0205, "step": 21910 }, { "grad_norm": 0.08208007365465164, "learning_rate": 4.6204806843369474e-05, "loss": 0.0162, "step": 21920 }, { "grad_norm": 0.09581466019153595, "learning_rate": 4.616359065031191e-05, "loss": 0.0233, "step": 21930 }, { "grad_norm": 0.12534725666046143, "learning_rate": 4.6122377079403946e-05, "loss": 0.0209, "step": 21940 }, { "grad_norm": 0.10396719723939896, "learning_rate": 4.6081166158814695e-05, "loss": 0.0203, "step": 21950 }, { "grad_norm": 0.083697110414505, "learning_rate": 4.603995791671144e-05, "loss": 0.0189, "step": 21960 }, { "grad_norm": 0.09695810079574585, "learning_rate": 4.599875238125957e-05, "loss": 0.0185, "step": 21970 }, { "grad_norm": 0.09865886718034744, "learning_rate": 4.595754958062273e-05, "loss": 0.0249, "step": 21980 }, { "grad_norm": 0.10379046946763992, "learning_rate": 4.591634954296265e-05, "loss": 0.0216, "step": 21990 }, { "grad_norm": 0.15789960324764252, "learning_rate": 4.587515229643913e-05, "loss": 0.0192, "step": 22000 }, { "grad_norm": 0.07572514563798904, "learning_rate": 4.583395786921013e-05, "loss": 0.0197, "step": 22010 }, { "grad_norm": 0.07175593078136444, "learning_rate": 4.579276628943164e-05, "loss": 0.0202, "step": 22020 }, { "grad_norm": 0.11757007241249084, "learning_rate": 4.575157758525772e-05, "loss": 0.0214, "step": 22030 }, { "grad_norm": 0.09621905535459518, "learning_rate": 4.571039178484046e-05, "loss": 0.0191, "step": 22040 }, { "grad_norm": 0.09203602373600006, "learning_rate": 4.566920891632998e-05, "loss": 0.0195, "step": 22050 }, { "grad_norm": 0.1207689717411995, "learning_rate": 4.562802900787436e-05, "loss": 0.0193, "step": 22060 }, { "grad_norm": 0.1022612527012825, "learning_rate": 4.558685208761968e-05, "loss": 0.0182, "step": 22070 }, { "grad_norm": 0.0718463584780693, "learning_rate": 4.554567818370998e-05, "loss": 0.0222, "step": 22080 }, { "grad_norm": 0.09394227713346481, "learning_rate": 4.550450732428726e-05, "loss": 0.0214, "step": 22090 }, { "grad_norm": 0.12058904021978378, "learning_rate": 4.546333953749137e-05, "loss": 0.0198, "step": 22100 }, { "grad_norm": 0.099493607878685, "learning_rate": 4.5422174851460154e-05, "loss": 0.0224, "step": 22110 }, { "grad_norm": 0.14024761319160461, "learning_rate": 4.538101329432924e-05, "loss": 0.0197, "step": 22120 }, { "grad_norm": 0.07876201719045639, "learning_rate": 4.5339854894232195e-05, "loss": 0.0182, "step": 22130 }, { "grad_norm": 0.14605239033699036, "learning_rate": 4.52986996793004e-05, "loss": 0.0192, "step": 22140 }, { "grad_norm": 0.0998225212097168, "learning_rate": 4.5257547677663024e-05, "loss": 0.0213, "step": 22150 }, { "grad_norm": 0.22117926180362701, "learning_rate": 4.52163989174471e-05, "loss": 0.0205, "step": 22160 }, { "grad_norm": 0.17008715867996216, "learning_rate": 4.51752534267774e-05, "loss": 0.0199, "step": 22170 }, { "grad_norm": 0.0891306921839714, "learning_rate": 4.513411123377649e-05, "loss": 0.0196, "step": 22180 }, { "grad_norm": 0.08897823095321655, "learning_rate": 4.5092972366564675e-05, "loss": 0.019, "step": 22190 }, { "grad_norm": 0.08045018464326859, "learning_rate": 4.505183685325997e-05, "loss": 0.0198, "step": 22200 }, { "grad_norm": 0.06440674513578415, "learning_rate": 4.5010704721978125e-05, "loss": 0.0168, "step": 22210 }, { "grad_norm": 0.08095364272594452, "learning_rate": 4.496957600083255e-05, "loss": 0.0214, "step": 22220 }, { "grad_norm": 0.1790608912706375, "learning_rate": 4.4928450717934343e-05, "loss": 0.0201, "step": 22230 }, { "grad_norm": 0.09979280084371567, "learning_rate": 4.488732890139227e-05, "loss": 0.0204, "step": 22240 }, { "grad_norm": 0.09573759883642197, "learning_rate": 4.4846210579312665e-05, "loss": 0.0204, "step": 22250 }, { "grad_norm": 0.07091377675533295, "learning_rate": 4.480509577979953e-05, "loss": 0.0217, "step": 22260 }, { "grad_norm": 0.08032340556383133, "learning_rate": 4.476398453095445e-05, "loss": 0.0206, "step": 22270 }, { "grad_norm": 0.09079979360103607, "learning_rate": 4.472287686087656e-05, "loss": 0.0203, "step": 22280 }, { "grad_norm": 0.10499966889619827, "learning_rate": 4.468177279766259e-05, "loss": 0.0198, "step": 22290 }, { "grad_norm": 0.07906488329172134, "learning_rate": 4.4640672369406746e-05, "loss": 0.0199, "step": 22300 }, { "grad_norm": 0.07272917032241821, "learning_rate": 4.459957560420082e-05, "loss": 0.0214, "step": 22310 }, { "grad_norm": 0.1092638298869133, "learning_rate": 4.455848253013403e-05, "loss": 0.0229, "step": 22320 }, { "grad_norm": 0.13770103454589844, "learning_rate": 4.4517393175293146e-05, "loss": 0.0229, "step": 22330 }, { "grad_norm": 0.0721767246723175, "learning_rate": 4.447630756776232e-05, "loss": 0.0224, "step": 22340 }, { "grad_norm": 0.08032162487506866, "learning_rate": 4.443522573562318e-05, "loss": 0.0221, "step": 22350 }, { "grad_norm": 0.09771700948476791, "learning_rate": 4.4394147706954776e-05, "loss": 0.0196, "step": 22360 }, { "grad_norm": 0.1630980670452118, "learning_rate": 4.435307350983355e-05, "loss": 0.0223, "step": 22370 }, { "grad_norm": 0.13048921525478363, "learning_rate": 4.4312003172333326e-05, "loss": 0.0231, "step": 22380 }, { "grad_norm": 0.10066667944192886, "learning_rate": 4.427093672252531e-05, "loss": 0.0204, "step": 22390 }, { "grad_norm": 0.1129312664270401, "learning_rate": 4.422987418847802e-05, "loss": 0.0233, "step": 22400 }, { "grad_norm": 0.10812507569789886, "learning_rate": 4.4188815598257325e-05, "loss": 0.0238, "step": 22410 }, { "grad_norm": 0.09606225043535233, "learning_rate": 4.414776097992638e-05, "loss": 0.0219, "step": 22420 }, { "grad_norm": 0.09724973887205124, "learning_rate": 4.4106710361545595e-05, "loss": 0.0217, "step": 22430 }, { "grad_norm": 0.07719162851572037, "learning_rate": 4.406566377117272e-05, "loss": 0.0193, "step": 22440 }, { "grad_norm": 0.10477318614721298, "learning_rate": 4.40246212368627e-05, "loss": 0.0266, "step": 22450 }, { "grad_norm": 0.06930243968963623, "learning_rate": 4.3983582786667715e-05, "loss": 0.0223, "step": 22460 }, { "grad_norm": 0.11056004464626312, "learning_rate": 4.394254844863716e-05, "loss": 0.0194, "step": 22470 }, { "grad_norm": 0.07674813270568848, "learning_rate": 4.390151825081762e-05, "loss": 0.0209, "step": 22480 }, { "grad_norm": 0.10629437863826752, "learning_rate": 4.386049222125286e-05, "loss": 0.0219, "step": 22490 }, { "grad_norm": 0.07922059297561646, "learning_rate": 4.3819470387983774e-05, "loss": 0.0208, "step": 22500 }, { "grad_norm": 0.08589751273393631, "learning_rate": 4.377845277904841e-05, "loss": 0.0186, "step": 22510 }, { "grad_norm": 0.08921796828508377, "learning_rate": 4.37374394224819e-05, "loss": 0.0205, "step": 22520 }, { "grad_norm": 0.09058383852243423, "learning_rate": 4.369643034631648e-05, "loss": 0.0241, "step": 22530 }, { "grad_norm": 0.10338441282510757, "learning_rate": 4.365542557858149e-05, "loss": 0.0195, "step": 22540 }, { "grad_norm": 0.09821773320436478, "learning_rate": 4.361442514730329e-05, "loss": 0.0199, "step": 22550 }, { "grad_norm": 0.07864256203174591, "learning_rate": 4.357342908050528e-05, "loss": 0.0193, "step": 22560 }, { "grad_norm": 0.07554230093955994, "learning_rate": 4.3532437406207895e-05, "loss": 0.0213, "step": 22570 }, { "grad_norm": 0.13846302032470703, "learning_rate": 4.349145015242856e-05, "loss": 0.0201, "step": 22580 }, { "grad_norm": 0.08258354663848877, "learning_rate": 4.345046734718168e-05, "loss": 0.0191, "step": 22590 }, { "grad_norm": 0.11006055027246475, "learning_rate": 4.34094890184786e-05, "loss": 0.0211, "step": 22600 }, { "grad_norm": 0.11582417786121368, "learning_rate": 4.336851519432765e-05, "loss": 0.0203, "step": 22610 }, { "grad_norm": 0.08804883062839508, "learning_rate": 4.332754590273403e-05, "loss": 0.0196, "step": 22620 }, { "grad_norm": 0.15260842442512512, "learning_rate": 4.3286581171699855e-05, "loss": 0.0183, "step": 22630 }, { "grad_norm": 0.07322154194116592, "learning_rate": 4.324562102922416e-05, "loss": 0.0207, "step": 22640 }, { "grad_norm": 0.11413341760635376, "learning_rate": 4.320466550330278e-05, "loss": 0.0205, "step": 22650 }, { "grad_norm": 0.06619322299957275, "learning_rate": 4.3163714621928466e-05, "loss": 0.0207, "step": 22660 }, { "grad_norm": 0.06856369972229004, "learning_rate": 4.312276841309074e-05, "loss": 0.0197, "step": 22670 }, { "grad_norm": 0.11115201562643051, "learning_rate": 4.3081826904775945e-05, "loss": 0.0211, "step": 22680 }, { "grad_norm": 0.11759985238313675, "learning_rate": 4.3040890124967246e-05, "loss": 0.0219, "step": 22690 }, { "grad_norm": 0.1138996034860611, "learning_rate": 4.2999958101644537e-05, "loss": 0.0233, "step": 22700 }, { "grad_norm": 0.09423699229955673, "learning_rate": 4.2959030862784435e-05, "loss": 0.0206, "step": 22710 }, { "grad_norm": 0.06644391268491745, "learning_rate": 4.291810843636036e-05, "loss": 0.0235, "step": 22720 }, { "grad_norm": 0.12035825103521347, "learning_rate": 4.2877190850342375e-05, "loss": 0.0237, "step": 22730 }, { "grad_norm": 0.15319302678108215, "learning_rate": 4.2836278132697294e-05, "loss": 0.0228, "step": 22740 }, { "grad_norm": 0.10413283109664917, "learning_rate": 4.279537031138855e-05, "loss": 0.0211, "step": 22750 }, { "grad_norm": 0.11076223850250244, "learning_rate": 4.275446741437625e-05, "loss": 0.0205, "step": 22760 }, { "grad_norm": 0.09202924370765686, "learning_rate": 4.2713569469617176e-05, "loss": 0.0198, "step": 22770 }, { "grad_norm": 0.10380102694034576, "learning_rate": 4.267267650506465e-05, "loss": 0.021, "step": 22780 }, { "grad_norm": 0.14959321916103363, "learning_rate": 4.263178854866866e-05, "loss": 0.0223, "step": 22790 }, { "grad_norm": 0.092336006462574, "learning_rate": 4.259090562837571e-05, "loss": 0.0237, "step": 22800 }, { "grad_norm": 0.09171196818351746, "learning_rate": 4.255002777212888e-05, "loss": 0.0214, "step": 22810 }, { "grad_norm": 0.0865914598107338, "learning_rate": 4.250915500786783e-05, "loss": 0.0207, "step": 22820 }, { "grad_norm": 0.11806964874267578, "learning_rate": 4.24682873635287e-05, "loss": 0.0201, "step": 22830 }, { "grad_norm": 0.10737127810716629, "learning_rate": 4.242742486704414e-05, "loss": 0.0219, "step": 22840 }, { "grad_norm": 0.09599220007658005, "learning_rate": 4.238656754634327e-05, "loss": 0.0191, "step": 22850 }, { "grad_norm": 0.08292847126722336, "learning_rate": 4.234571542935168e-05, "loss": 0.0222, "step": 22860 }, { "grad_norm": 0.07778360694646835, "learning_rate": 4.230486854399144e-05, "loss": 0.0173, "step": 22870 }, { "grad_norm": 0.08910661190748215, "learning_rate": 4.226402691818098e-05, "loss": 0.022, "step": 22880 }, { "grad_norm": 0.09959922730922699, "learning_rate": 4.2223190579835196e-05, "loss": 0.0234, "step": 22890 }, { "grad_norm": 0.1153210774064064, "learning_rate": 4.218235955686531e-05, "loss": 0.0208, "step": 22900 }, { "grad_norm": 0.08236277103424072, "learning_rate": 4.214153387717894e-05, "loss": 0.022, "step": 22910 }, { "grad_norm": 0.09987059980630875, "learning_rate": 4.210071356868007e-05, "loss": 0.0209, "step": 22920 }, { "grad_norm": 0.09127619117498398, "learning_rate": 4.205989865926898e-05, "loss": 0.0206, "step": 22930 }, { "grad_norm": 0.10399705916643143, "learning_rate": 4.2019089176842294e-05, "loss": 0.0194, "step": 22940 }, { "grad_norm": 0.10061489045619965, "learning_rate": 4.1978285149292894e-05, "loss": 0.0213, "step": 22950 }, { "grad_norm": 0.0780692771077156, "learning_rate": 4.193748660450996e-05, "loss": 0.0209, "step": 22960 }, { "grad_norm": 0.07303300499916077, "learning_rate": 4.189669357037891e-05, "loss": 0.0179, "step": 22970 }, { "grad_norm": 0.1365424543619156, "learning_rate": 4.1855906074781405e-05, "loss": 0.0225, "step": 22980 }, { "grad_norm": 0.06981220841407776, "learning_rate": 4.1815124145595285e-05, "loss": 0.018, "step": 22990 }, { "grad_norm": 0.09299565106630325, "learning_rate": 4.1774347810694644e-05, "loss": 0.0221, "step": 23000 }, { "grad_norm": 0.08541422337293625, "learning_rate": 4.17335770979497e-05, "loss": 0.0214, "step": 23010 }, { "grad_norm": 0.115557461977005, "learning_rate": 4.169281203522687e-05, "loss": 0.02, "step": 23020 }, { "grad_norm": 0.08222471177577972, "learning_rate": 4.1652052650388674e-05, "loss": 0.0202, "step": 23030 }, { "grad_norm": 0.09943108260631561, "learning_rate": 4.1611298971293786e-05, "loss": 0.0196, "step": 23040 }, { "grad_norm": 0.0993277058005333, "learning_rate": 4.1570551025796935e-05, "loss": 0.0239, "step": 23050 }, { "grad_norm": 0.06964816153049469, "learning_rate": 4.152980884174897e-05, "loss": 0.0195, "step": 23060 }, { "grad_norm": 0.11170744150876999, "learning_rate": 4.148907244699682e-05, "loss": 0.0202, "step": 23070 }, { "grad_norm": 0.09803825616836548, "learning_rate": 4.1448341869383395e-05, "loss": 0.0214, "step": 23080 }, { "grad_norm": 0.11413499712944031, "learning_rate": 4.140761713674765e-05, "loss": 0.0189, "step": 23090 }, { "grad_norm": 0.0899387001991272, "learning_rate": 4.1366898276924574e-05, "loss": 0.0215, "step": 23100 }, { "grad_norm": 0.1751697063446045, "learning_rate": 4.132618531774512e-05, "loss": 0.0186, "step": 23110 }, { "grad_norm": 0.0951986238360405, "learning_rate": 4.128547828703622e-05, "loss": 0.0216, "step": 23120 }, { "grad_norm": 0.0994962528347969, "learning_rate": 4.1244777212620725e-05, "loss": 0.02, "step": 23130 }, { "grad_norm": 0.08900479972362518, "learning_rate": 4.120408212231746e-05, "loss": 0.0174, "step": 23140 }, { "grad_norm": 0.055986933410167694, "learning_rate": 4.116339304394111e-05, "loss": 0.0185, "step": 23150 }, { "grad_norm": 0.09350760281085968, "learning_rate": 4.112271000530229e-05, "loss": 0.0204, "step": 23160 }, { "grad_norm": 0.07628501951694489, "learning_rate": 4.10820330342075e-05, "loss": 0.0201, "step": 23170 }, { "grad_norm": 0.09759870916604996, "learning_rate": 4.1041362158459027e-05, "loss": 0.0208, "step": 23180 }, { "grad_norm": 0.11755888909101486, "learning_rate": 4.1000697405855024e-05, "loss": 0.0169, "step": 23190 }, { "grad_norm": 0.16322188079357147, "learning_rate": 4.096003880418951e-05, "loss": 0.0206, "step": 23200 }, { "grad_norm": 0.0977817103266716, "learning_rate": 4.0919386381252215e-05, "loss": 0.0236, "step": 23210 }, { "grad_norm": 0.10414037108421326, "learning_rate": 4.087874016482872e-05, "loss": 0.0211, "step": 23220 }, { "grad_norm": 0.08226557075977325, "learning_rate": 4.0838100182700295e-05, "loss": 0.0251, "step": 23230 }, { "grad_norm": 0.10726364701986313, "learning_rate": 4.079746646264402e-05, "loss": 0.0204, "step": 23240 }, { "grad_norm": 0.08353511989116669, "learning_rate": 4.075683903243262e-05, "loss": 0.0226, "step": 23250 }, { "grad_norm": 0.09877998381853104, "learning_rate": 4.071621791983462e-05, "loss": 0.0232, "step": 23260 }, { "grad_norm": 0.11363904178142548, "learning_rate": 4.06756031526141e-05, "loss": 0.0194, "step": 23270 }, { "grad_norm": 0.13996733725070953, "learning_rate": 4.063499475853092e-05, "loss": 0.0189, "step": 23280 }, { "grad_norm": 0.10069137811660767, "learning_rate": 4.0594392765340506e-05, "loss": 0.0201, "step": 23290 }, { "grad_norm": 0.10526686906814575, "learning_rate": 4.0553797200793954e-05, "loss": 0.0204, "step": 23300 }, { "grad_norm": 0.08315443992614746, "learning_rate": 4.0513208092637926e-05, "loss": 0.0199, "step": 23310 }, { "grad_norm": 0.10611880570650101, "learning_rate": 4.0472625468614735e-05, "loss": 0.0193, "step": 23320 }, { "grad_norm": 0.06002854183316231, "learning_rate": 4.043204935646218e-05, "loss": 0.0198, "step": 23330 }, { "grad_norm": 0.08813697844743729, "learning_rate": 4.0391479783913675e-05, "loss": 0.0203, "step": 23340 }, { "grad_norm": 0.07574756443500519, "learning_rate": 4.0350916778698155e-05, "loss": 0.0211, "step": 23350 }, { "grad_norm": 0.066121406853199, "learning_rate": 4.031036036854001e-05, "loss": 0.0148, "step": 23360 }, { "grad_norm": 0.07908301800489426, "learning_rate": 4.026981058115918e-05, "loss": 0.0196, "step": 23370 }, { "grad_norm": 0.09448972344398499, "learning_rate": 4.022926744427108e-05, "loss": 0.0214, "step": 23380 }, { "grad_norm": 0.08435007929801941, "learning_rate": 4.018873098558654e-05, "loss": 0.0189, "step": 23390 }, { "grad_norm": 0.07248484343290329, "learning_rate": 4.014820123281186e-05, "loss": 0.0192, "step": 23400 }, { "grad_norm": 0.0796123519539833, "learning_rate": 4.0107678213648735e-05, "loss": 0.0186, "step": 23410 }, { "grad_norm": 0.08175020664930344, "learning_rate": 4.006716195579428e-05, "loss": 0.0201, "step": 23420 }, { "grad_norm": 0.06937901675701141, "learning_rate": 4.002665248694096e-05, "loss": 0.019, "step": 23430 }, { "grad_norm": 0.07832269370555878, "learning_rate": 3.998614983477664e-05, "loss": 0.0178, "step": 23440 }, { "grad_norm": 0.0703519657254219, "learning_rate": 3.994565402698448e-05, "loss": 0.0198, "step": 23450 }, { "grad_norm": 0.08064650744199753, "learning_rate": 3.9905165091242975e-05, "loss": 0.0203, "step": 23460 }, { "grad_norm": 0.09264140576124191, "learning_rate": 3.9864683055225936e-05, "loss": 0.0202, "step": 23470 }, { "grad_norm": 0.16012422740459442, "learning_rate": 3.982420794660247e-05, "loss": 0.021, "step": 23480 }, { "grad_norm": 0.10211829841136932, "learning_rate": 3.978373979303691e-05, "loss": 0.0218, "step": 23490 }, { "grad_norm": 0.07851273566484451, "learning_rate": 3.974327862218888e-05, "loss": 0.0211, "step": 23500 }, { "grad_norm": 0.08307259529829025, "learning_rate": 3.970282446171318e-05, "loss": 0.0167, "step": 23510 }, { "grad_norm": 0.07210596650838852, "learning_rate": 3.966237733925988e-05, "loss": 0.0215, "step": 23520 }, { "grad_norm": 0.08991911262273788, "learning_rate": 3.962193728247418e-05, "loss": 0.019, "step": 23530 }, { "grad_norm": 0.13640771806240082, "learning_rate": 3.958150431899651e-05, "loss": 0.0206, "step": 23540 }, { "grad_norm": 0.12057682871818542, "learning_rate": 3.954107847646238e-05, "loss": 0.0157, "step": 23550 }, { "grad_norm": 0.06719955056905746, "learning_rate": 3.950065978250249e-05, "loss": 0.0189, "step": 23560 }, { "grad_norm": 0.07625941187143326, "learning_rate": 3.9460248264742624e-05, "loss": 0.0188, "step": 23570 }, { "grad_norm": 0.09584185481071472, "learning_rate": 3.941984395080371e-05, "loss": 0.0204, "step": 23580 }, { "grad_norm": 0.09383463859558105, "learning_rate": 3.937944686830167e-05, "loss": 0.0201, "step": 23590 }, { "grad_norm": 0.10958480834960938, "learning_rate": 3.933905704484756e-05, "loss": 0.0197, "step": 23600 }, { "grad_norm": 0.09941728413105011, "learning_rate": 3.929867450804743e-05, "loss": 0.0201, "step": 23610 }, { "grad_norm": 0.10093482583761215, "learning_rate": 3.925829928550237e-05, "loss": 0.0195, "step": 23620 }, { "grad_norm": 0.1049128845334053, "learning_rate": 3.921793140480847e-05, "loss": 0.0179, "step": 23630 }, { "grad_norm": 0.12113762646913528, "learning_rate": 3.917757089355677e-05, "loss": 0.0233, "step": 23640 }, { "grad_norm": 0.09762881696224213, "learning_rate": 3.9137217779333326e-05, "loss": 0.0192, "step": 23650 }, { "grad_norm": 0.07480986416339874, "learning_rate": 3.9096872089719083e-05, "loss": 0.0207, "step": 23660 }, { "grad_norm": 0.09690424054861069, "learning_rate": 3.905653385228996e-05, "loss": 0.0204, "step": 23670 }, { "grad_norm": 0.09587876498699188, "learning_rate": 3.901620309461677e-05, "loss": 0.0195, "step": 23680 }, { "grad_norm": 0.1265149861574173, "learning_rate": 3.897587984426518e-05, "loss": 0.0197, "step": 23690 }, { "grad_norm": 0.10944181680679321, "learning_rate": 3.893556412879577e-05, "loss": 0.0196, "step": 23700 }, { "grad_norm": 0.08048505336046219, "learning_rate": 3.889525597576395e-05, "loss": 0.0172, "step": 23710 }, { "grad_norm": 0.08773548901081085, "learning_rate": 3.8854955412719965e-05, "loss": 0.0194, "step": 23720 }, { "grad_norm": 0.07010786980390549, "learning_rate": 3.881466246720887e-05, "loss": 0.0228, "step": 23730 }, { "grad_norm": 0.066439189016819, "learning_rate": 3.8774377166770484e-05, "loss": 0.019, "step": 23740 }, { "grad_norm": 0.13647347688674927, "learning_rate": 3.8734099538939474e-05, "loss": 0.0208, "step": 23750 }, { "grad_norm": 0.08901707828044891, "learning_rate": 3.869382961124518e-05, "loss": 0.0205, "step": 23760 }, { "grad_norm": 0.09309252351522446, "learning_rate": 3.8653567411211736e-05, "loss": 0.0194, "step": 23770 }, { "grad_norm": 0.0844254419207573, "learning_rate": 3.8613312966357987e-05, "loss": 0.0186, "step": 23780 }, { "grad_norm": 0.11835110187530518, "learning_rate": 3.857306630419745e-05, "loss": 0.0213, "step": 23790 }, { "grad_norm": 0.07533097267150879, "learning_rate": 3.853282745223834e-05, "loss": 0.0182, "step": 23800 }, { "grad_norm": 0.08521345257759094, "learning_rate": 3.8492596437983546e-05, "loss": 0.0214, "step": 23810 }, { "grad_norm": 0.12772539258003235, "learning_rate": 3.8452373288930586e-05, "loss": 0.0173, "step": 23820 }, { "grad_norm": 0.13784490525722504, "learning_rate": 3.841215803257159e-05, "loss": 0.0201, "step": 23830 }, { "grad_norm": 0.07759195566177368, "learning_rate": 3.83719506963933e-05, "loss": 0.0182, "step": 23840 }, { "grad_norm": 0.09867724031209946, "learning_rate": 3.8331751307877087e-05, "loss": 0.0233, "step": 23850 }, { "grad_norm": 0.08913695812225342, "learning_rate": 3.82915598944988e-05, "loss": 0.0218, "step": 23860 }, { "grad_norm": 0.08363393694162369, "learning_rate": 3.825137648372893e-05, "loss": 0.0213, "step": 23870 }, { "grad_norm": 0.09445064514875412, "learning_rate": 3.8211201103032465e-05, "loss": 0.0248, "step": 23880 }, { "grad_norm": 0.09330327808856964, "learning_rate": 3.817103377986887e-05, "loss": 0.0199, "step": 23890 }, { "grad_norm": 0.07748174667358398, "learning_rate": 3.813087454169215e-05, "loss": 0.0178, "step": 23900 }, { "grad_norm": 0.09617912024259567, "learning_rate": 3.809072341595078e-05, "loss": 0.02, "step": 23910 }, { "grad_norm": 0.0981249213218689, "learning_rate": 3.8050580430087636e-05, "loss": 0.0207, "step": 23920 }, { "grad_norm": 0.0851464718580246, "learning_rate": 3.8010445611540096e-05, "loss": 0.0193, "step": 23930 }, { "grad_norm": 0.09982088208198547, "learning_rate": 3.797031898773992e-05, "loss": 0.0175, "step": 23940 }, { "grad_norm": 0.1076010912656784, "learning_rate": 3.793020058611329e-05, "loss": 0.0204, "step": 23950 }, { "grad_norm": 0.10608836263418198, "learning_rate": 3.789009043408074e-05, "loss": 0.02, "step": 23960 }, { "grad_norm": 0.0891198143362999, "learning_rate": 3.7849988559057194e-05, "loss": 0.017, "step": 23970 }, { "grad_norm": 0.07710427790880203, "learning_rate": 3.78098949884519e-05, "loss": 0.018, "step": 23980 }, { "grad_norm": 0.12448575347661972, "learning_rate": 3.776980974966843e-05, "loss": 0.0208, "step": 23990 }, { "grad_norm": 0.09658782929182053, "learning_rate": 3.772973287010468e-05, "loss": 0.0207, "step": 24000 }, { "grad_norm": 0.07407540082931519, "learning_rate": 3.768966437715283e-05, "loss": 0.0201, "step": 24010 }, { "grad_norm": 0.08245779573917389, "learning_rate": 3.7649604298199274e-05, "loss": 0.0223, "step": 24020 }, { "grad_norm": 0.13893093168735504, "learning_rate": 3.760955266062473e-05, "loss": 0.0211, "step": 24030 }, { "grad_norm": 0.08436224609613419, "learning_rate": 3.75695094918041e-05, "loss": 0.0198, "step": 24040 }, { "grad_norm": 0.0838548019528389, "learning_rate": 3.752947481910652e-05, "loss": 0.0191, "step": 24050 }, { "grad_norm": 0.0692245364189148, "learning_rate": 3.7489448669895324e-05, "loss": 0.02, "step": 24060 }, { "grad_norm": 0.08750681579113007, "learning_rate": 3.744943107152798e-05, "loss": 0.0196, "step": 24070 }, { "grad_norm": 0.07977908849716187, "learning_rate": 3.7409422051356165e-05, "loss": 0.0176, "step": 24080 }, { "grad_norm": 0.09045680612325668, "learning_rate": 3.736942163672564e-05, "loss": 0.0233, "step": 24090 }, { "grad_norm": 0.09531323611736298, "learning_rate": 3.732942985497636e-05, "loss": 0.0177, "step": 24100 }, { "grad_norm": 0.08215537667274475, "learning_rate": 3.728944673344228e-05, "loss": 0.0197, "step": 24110 }, { "grad_norm": 0.08401689678430557, "learning_rate": 3.72494722994515e-05, "loss": 0.0203, "step": 24120 }, { "grad_norm": 0.07970687747001648, "learning_rate": 3.720950658032617e-05, "loss": 0.0183, "step": 24130 }, { "grad_norm": 0.08127021789550781, "learning_rate": 3.716954960338249e-05, "loss": 0.0196, "step": 24140 }, { "grad_norm": 0.08128118515014648, "learning_rate": 3.712960139593066e-05, "loss": 0.0188, "step": 24150 }, { "grad_norm": 0.08580763638019562, "learning_rate": 3.708966198527493e-05, "loss": 0.0192, "step": 24160 }, { "grad_norm": 0.10955197364091873, "learning_rate": 3.704973139871349e-05, "loss": 0.0187, "step": 24170 }, { "grad_norm": 0.11161687225103378, "learning_rate": 3.700980966353853e-05, "loss": 0.0202, "step": 24180 }, { "grad_norm": 0.10313070565462112, "learning_rate": 3.696989680703619e-05, "loss": 0.023, "step": 24190 }, { "grad_norm": 0.0722343772649765, "learning_rate": 3.69299928564865e-05, "loss": 0.0181, "step": 24200 }, { "grad_norm": 0.09976458549499512, "learning_rate": 3.689009783916345e-05, "loss": 0.0194, "step": 24210 }, { "grad_norm": 0.09148944914340973, "learning_rate": 3.6850211782334895e-05, "loss": 0.02, "step": 24220 }, { "grad_norm": 0.08613935858011246, "learning_rate": 3.681033471326261e-05, "loss": 0.0204, "step": 24230 }, { "grad_norm": 0.10693272203207016, "learning_rate": 3.677046665920216e-05, "loss": 0.0204, "step": 24240 }, { "grad_norm": 0.06294029951095581, "learning_rate": 3.6730607647403005e-05, "loss": 0.018, "step": 24250 }, { "grad_norm": 0.08525843173265457, "learning_rate": 3.6690757705108416e-05, "loss": 0.0204, "step": 24260 }, { "grad_norm": 0.13301615417003632, "learning_rate": 3.665091685955542e-05, "loss": 0.018, "step": 24270 }, { "grad_norm": 0.10474764555692673, "learning_rate": 3.6611085137974896e-05, "loss": 0.0183, "step": 24280 }, { "grad_norm": 0.09324172139167786, "learning_rate": 3.657126256759143e-05, "loss": 0.0178, "step": 24290 }, { "grad_norm": 0.07675094902515411, "learning_rate": 3.653144917562335e-05, "loss": 0.0201, "step": 24300 }, { "grad_norm": 0.07880448549985886, "learning_rate": 3.649164498928277e-05, "loss": 0.0172, "step": 24310 }, { "grad_norm": 0.0994177758693695, "learning_rate": 3.645185003577546e-05, "loss": 0.0239, "step": 24320 }, { "grad_norm": 0.11011205613613129, "learning_rate": 3.6412064342300906e-05, "loss": 0.0187, "step": 24330 }, { "grad_norm": 0.1428551822900772, "learning_rate": 3.637228793605224e-05, "loss": 0.018, "step": 24340 }, { "grad_norm": 0.08191180229187012, "learning_rate": 3.6332520844216264e-05, "loss": 0.0181, "step": 24350 }, { "grad_norm": 0.093088299036026, "learning_rate": 3.6292763093973425e-05, "loss": 0.0211, "step": 24360 }, { "grad_norm": 0.07771889865398407, "learning_rate": 3.6253014712497754e-05, "loss": 0.0159, "step": 24370 }, { "grad_norm": 0.07899230718612671, "learning_rate": 3.621327572695692e-05, "loss": 0.0176, "step": 24380 }, { "grad_norm": 0.06767059117555618, "learning_rate": 3.617354616451211e-05, "loss": 0.0221, "step": 24390 }, { "grad_norm": 0.08593886345624924, "learning_rate": 3.6133826052318116e-05, "loss": 0.0197, "step": 24400 }, { "grad_norm": 0.11010554432868958, "learning_rate": 3.609411541752327e-05, "loss": 0.0184, "step": 24410 }, { "grad_norm": 0.13181664049625397, "learning_rate": 3.6054414287269405e-05, "loss": 0.023, "step": 24420 }, { "grad_norm": 0.07835280895233154, "learning_rate": 3.601472268869188e-05, "loss": 0.0167, "step": 24430 }, { "grad_norm": 0.08684578537940979, "learning_rate": 3.597504064891952e-05, "loss": 0.0174, "step": 24440 }, { "grad_norm": 0.07767417281866074, "learning_rate": 3.5935368195074636e-05, "loss": 0.021, "step": 24450 }, { "grad_norm": 0.10689087957143784, "learning_rate": 3.589570535427297e-05, "loss": 0.0171, "step": 24460 }, { "grad_norm": 0.07853467017412186, "learning_rate": 3.585605215362371e-05, "loss": 0.0219, "step": 24470 }, { "grad_norm": 0.06480186432600021, "learning_rate": 3.581640862022941e-05, "loss": 0.0177, "step": 24480 }, { "grad_norm": 0.1025080531835556, "learning_rate": 3.57767747811861e-05, "loss": 0.0174, "step": 24490 }, { "grad_norm": 0.18343138694763184, "learning_rate": 3.573715066358308e-05, "loss": 0.0165, "step": 24500 }, { "grad_norm": 0.22134597599506378, "learning_rate": 3.569753629450311e-05, "loss": 0.0164, "step": 24510 }, { "grad_norm": 0.0588548369705677, "learning_rate": 3.565793170102221e-05, "loss": 0.0195, "step": 24520 }, { "grad_norm": 0.10840880870819092, "learning_rate": 3.561833691020976e-05, "loss": 0.0219, "step": 24530 }, { "grad_norm": 0.1448626071214676, "learning_rate": 3.5578751949128415e-05, "loss": 0.02, "step": 24540 }, { "grad_norm": 0.08343111723661423, "learning_rate": 3.5539176844834125e-05, "loss": 0.0185, "step": 24550 }, { "grad_norm": 0.08713726699352264, "learning_rate": 3.5499611624376125e-05, "loss": 0.018, "step": 24560 }, { "grad_norm": 0.13140860199928284, "learning_rate": 3.546005631479684e-05, "loss": 0.0177, "step": 24570 }, { "grad_norm": 0.09998907148838043, "learning_rate": 3.542051094313196e-05, "loss": 0.0156, "step": 24580 }, { "grad_norm": 0.1314394474029541, "learning_rate": 3.5380975536410364e-05, "loss": 0.0229, "step": 24590 }, { "grad_norm": 0.11000087112188339, "learning_rate": 3.534145012165415e-05, "loss": 0.0214, "step": 24600 }, { "grad_norm": 0.09657745808362961, "learning_rate": 3.5301934725878546e-05, "loss": 0.0202, "step": 24610 }, { "grad_norm": 0.08834274113178253, "learning_rate": 3.526242937609197e-05, "loss": 0.0211, "step": 24620 }, { "grad_norm": 0.12093129009008408, "learning_rate": 3.522293409929595e-05, "loss": 0.0185, "step": 24630 }, { "grad_norm": 0.08745972812175751, "learning_rate": 3.518344892248513e-05, "loss": 0.0177, "step": 24640 }, { "grad_norm": 0.07902029901742935, "learning_rate": 3.514397387264725e-05, "loss": 0.0187, "step": 24650 }, { "grad_norm": 0.09135068953037262, "learning_rate": 3.5104508976763176e-05, "loss": 0.0177, "step": 24660 }, { "grad_norm": 0.08504068106412888, "learning_rate": 3.506505426180674e-05, "loss": 0.0199, "step": 24670 }, { "grad_norm": 0.11806412041187286, "learning_rate": 3.502560975474488e-05, "loss": 0.0202, "step": 24680 }, { "grad_norm": 0.10134505480527878, "learning_rate": 3.4986175482537566e-05, "loss": 0.0223, "step": 24690 }, { "grad_norm": 0.08558586239814758, "learning_rate": 3.4946751472137725e-05, "loss": 0.0188, "step": 24700 }, { "grad_norm": 0.07484663277864456, "learning_rate": 3.490733775049132e-05, "loss": 0.0163, "step": 24710 }, { "grad_norm": 0.07882849872112274, "learning_rate": 3.4867934344537236e-05, "loss": 0.0206, "step": 24720 }, { "grad_norm": 0.08141504228115082, "learning_rate": 3.482854128120735e-05, "loss": 0.0192, "step": 24730 }, { "grad_norm": 0.11385118961334229, "learning_rate": 3.478915858742643e-05, "loss": 0.0173, "step": 24740 }, { "grad_norm": 0.07783298939466476, "learning_rate": 3.4749786290112205e-05, "loss": 0.0184, "step": 24750 }, { "grad_norm": 0.08666034042835236, "learning_rate": 3.471042441617524e-05, "loss": 0.0145, "step": 24760 }, { "grad_norm": 0.08581425249576569, "learning_rate": 3.467107299251902e-05, "loss": 0.0153, "step": 24770 }, { "grad_norm": 0.1459905058145523, "learning_rate": 3.463173204603984e-05, "loss": 0.0178, "step": 24780 }, { "grad_norm": 0.06473703682422638, "learning_rate": 3.4592401603626924e-05, "loss": 0.017, "step": 24790 }, { "grad_norm": 0.06692042201757431, "learning_rate": 3.45530816921622e-05, "loss": 0.017, "step": 24800 }, { "grad_norm": 0.06885439902544022, "learning_rate": 3.451377233852051e-05, "loss": 0.0162, "step": 24810 }, { "grad_norm": 0.08680752664804459, "learning_rate": 3.4474473569569385e-05, "loss": 0.0148, "step": 24820 }, { "grad_norm": 0.0838310644030571, "learning_rate": 3.443518541216918e-05, "loss": 0.0176, "step": 24830 }, { "grad_norm": 0.07396416366100311, "learning_rate": 3.439590789317299e-05, "loss": 0.0184, "step": 24840 }, { "grad_norm": 0.11085182428359985, "learning_rate": 3.4356641039426607e-05, "loss": 0.02, "step": 24850 }, { "grad_norm": 0.06817880272865295, "learning_rate": 3.431738487776857e-05, "loss": 0.0189, "step": 24860 }, { "grad_norm": 0.06503988802433014, "learning_rate": 3.4278139435030084e-05, "loss": 0.0132, "step": 24870 }, { "grad_norm": 0.06695204973220825, "learning_rate": 3.423890473803504e-05, "loss": 0.0186, "step": 24880 }, { "grad_norm": 0.08354298025369644, "learning_rate": 3.41996808136e-05, "loss": 0.0165, "step": 24890 }, { "grad_norm": 0.07850180566310883, "learning_rate": 3.416046768853413e-05, "loss": 0.0169, "step": 24900 }, { "grad_norm": 0.04981241375207901, "learning_rate": 3.412126538963925e-05, "loss": 0.0156, "step": 24910 }, { "grad_norm": 0.08713600784540176, "learning_rate": 3.4082073943709727e-05, "loss": 0.0165, "step": 24920 }, { "grad_norm": 0.16520896553993225, "learning_rate": 3.404289337753258e-05, "loss": 0.0184, "step": 24930 }, { "grad_norm": 0.08473596721887589, "learning_rate": 3.400372371788736e-05, "loss": 0.0201, "step": 24940 }, { "grad_norm": 0.06861752271652222, "learning_rate": 3.3964564991546124e-05, "loss": 0.0175, "step": 24950 }, { "grad_norm": 0.07494933903217316, "learning_rate": 3.392541722527351e-05, "loss": 0.017, "step": 24960 }, { "grad_norm": 0.07471694052219391, "learning_rate": 3.3886280445826644e-05, "loss": 0.0215, "step": 24970 }, { "grad_norm": 0.096929170191288, "learning_rate": 3.3847154679955154e-05, "loss": 0.0165, "step": 24980 }, { "grad_norm": 0.06403648853302002, "learning_rate": 3.380803995440113e-05, "loss": 0.017, "step": 24990 }, { "grad_norm": 0.13787563145160675, "learning_rate": 3.3768936295899115e-05, "loss": 0.0155, "step": 25000 }, { "grad_norm": 0.0733976885676384, "learning_rate": 3.3729843731176094e-05, "loss": 0.0202, "step": 25010 }, { "grad_norm": 0.1057635173201561, "learning_rate": 3.369076228695146e-05, "loss": 0.0184, "step": 25020 }, { "grad_norm": 0.08558806777000427, "learning_rate": 3.365169198993703e-05, "loss": 0.0152, "step": 25030 }, { "grad_norm": 0.08317510038614273, "learning_rate": 3.361263286683697e-05, "loss": 0.0156, "step": 25040 }, { "grad_norm": 0.051837146282196045, "learning_rate": 3.35735849443478e-05, "loss": 0.0166, "step": 25050 }, { "grad_norm": 0.10843618959188461, "learning_rate": 3.3534548249158435e-05, "loss": 0.0186, "step": 25060 }, { "grad_norm": 0.06195659562945366, "learning_rate": 3.3495522807950086e-05, "loss": 0.0163, "step": 25070 }, { "grad_norm": 0.09501487761735916, "learning_rate": 3.345650864739627e-05, "loss": 0.0158, "step": 25080 }, { "grad_norm": 0.07866220921278, "learning_rate": 3.3417505794162794e-05, "loss": 0.0146, "step": 25090 }, { "grad_norm": 0.08360650390386581, "learning_rate": 3.3378514274907745e-05, "loss": 0.0171, "step": 25100 }, { "grad_norm": 0.10840065777301788, "learning_rate": 3.333953411628147e-05, "loss": 0.0209, "step": 25110 }, { "grad_norm": 0.09571092575788498, "learning_rate": 3.330056534492653e-05, "loss": 0.0179, "step": 25120 }, { "grad_norm": 0.08956930041313171, "learning_rate": 3.32616079874777e-05, "loss": 0.0159, "step": 25130 }, { "grad_norm": 0.11294308304786682, "learning_rate": 3.322266207056197e-05, "loss": 0.0188, "step": 25140 }, { "grad_norm": 0.09040067344903946, "learning_rate": 3.318372762079852e-05, "loss": 0.019, "step": 25150 }, { "grad_norm": 0.10112575441598892, "learning_rate": 3.3144804664798666e-05, "loss": 0.0184, "step": 25160 }, { "grad_norm": 0.08969080448150635, "learning_rate": 3.3105893229165894e-05, "loss": 0.0165, "step": 25170 }, { "grad_norm": 0.07250281423330307, "learning_rate": 3.30669933404958e-05, "loss": 0.015, "step": 25180 }, { "grad_norm": 0.06926926970481873, "learning_rate": 3.302810502537609e-05, "loss": 0.0146, "step": 25190 }, { "grad_norm": 0.06439176201820374, "learning_rate": 3.298922831038655e-05, "loss": 0.0176, "step": 25200 }, { "grad_norm": 0.09872317314147949, "learning_rate": 3.2950363222099073e-05, "loss": 0.014, "step": 25210 }, { "grad_norm": 0.07972279191017151, "learning_rate": 3.291150978707758e-05, "loss": 0.0154, "step": 25220 }, { "grad_norm": 0.07575762271881104, "learning_rate": 3.287266803187798e-05, "loss": 0.0167, "step": 25230 }, { "grad_norm": 0.061481330543756485, "learning_rate": 3.283383798304829e-05, "loss": 0.0164, "step": 25240 }, { "grad_norm": 0.06842108070850372, "learning_rate": 3.279501966712847e-05, "loss": 0.019, "step": 25250 }, { "grad_norm": 0.17146649956703186, "learning_rate": 3.275621311065047e-05, "loss": 0.0169, "step": 25260 }, { "grad_norm": 0.07963988184928894, "learning_rate": 3.271741834013822e-05, "loss": 0.02, "step": 25270 }, { "grad_norm": 0.07976312190294266, "learning_rate": 3.267863538210756e-05, "loss": 0.0185, "step": 25280 }, { "grad_norm": 0.10731546580791473, "learning_rate": 3.2639864263066296e-05, "loss": 0.0181, "step": 25290 }, { "grad_norm": 0.07786326110363007, "learning_rate": 3.26011050095141e-05, "loss": 0.0195, "step": 25300 }, { "grad_norm": 0.0822734534740448, "learning_rate": 3.256235764794259e-05, "loss": 0.0193, "step": 25310 }, { "grad_norm": 0.07049858570098877, "learning_rate": 3.2523622204835194e-05, "loss": 0.0192, "step": 25320 }, { "grad_norm": 0.10534194856882095, "learning_rate": 3.2484898706667214e-05, "loss": 0.0187, "step": 25330 }, { "grad_norm": 0.10204516351222992, "learning_rate": 3.2446187179905806e-05, "loss": 0.0195, "step": 25340 }, { "grad_norm": 0.11832211911678314, "learning_rate": 3.240748765100995e-05, "loss": 0.0194, "step": 25350 }, { "grad_norm": 0.1737392693758011, "learning_rate": 3.236880014643039e-05, "loss": 0.0181, "step": 25360 }, { "grad_norm": 0.15993279218673706, "learning_rate": 3.233012469260969e-05, "loss": 0.0224, "step": 25370 }, { "grad_norm": 0.09162325412034988, "learning_rate": 3.229146131598213e-05, "loss": 0.0198, "step": 25380 }, { "grad_norm": 0.12357461452484131, "learning_rate": 3.2252810042973794e-05, "loss": 0.0172, "step": 25390 }, { "grad_norm": 0.1366214156150818, "learning_rate": 3.2214170900002456e-05, "loss": 0.0154, "step": 25400 }, { "grad_norm": 0.09465321898460388, "learning_rate": 3.217554391347758e-05, "loss": 0.0163, "step": 25410 }, { "grad_norm": 0.09905539453029633, "learning_rate": 3.213692910980037e-05, "loss": 0.02, "step": 25420 }, { "grad_norm": 0.1127220168709755, "learning_rate": 3.2098326515363666e-05, "loss": 0.0197, "step": 25430 }, { "grad_norm": 0.08512213081121445, "learning_rate": 3.205973615655199e-05, "loss": 0.0209, "step": 25440 }, { "grad_norm": 0.09364980459213257, "learning_rate": 3.202115805974149e-05, "loss": 0.0198, "step": 25450 }, { "grad_norm": 0.0695565789937973, "learning_rate": 3.1982592251299916e-05, "loss": 0.022, "step": 25460 }, { "grad_norm": 0.0803309753537178, "learning_rate": 3.1944038757586656e-05, "loss": 0.0169, "step": 25470 }, { "grad_norm": 0.20304669439792633, "learning_rate": 3.190549760495263e-05, "loss": 0.0173, "step": 25480 }, { "grad_norm": 0.11784657090902328, "learning_rate": 3.186696881974039e-05, "loss": 0.019, "step": 25490 }, { "grad_norm": 0.0704912468791008, "learning_rate": 3.1828452428283986e-05, "loss": 0.0182, "step": 25500 }, { "grad_norm": 0.1278069019317627, "learning_rate": 3.178994845690898e-05, "loss": 0.022, "step": 25510 }, { "grad_norm": 0.05963754653930664, "learning_rate": 3.17514569319325e-05, "loss": 0.0187, "step": 25520 }, { "grad_norm": 0.10642091929912567, "learning_rate": 3.171297787966312e-05, "loss": 0.0209, "step": 25530 }, { "grad_norm": 0.08533904701471329, "learning_rate": 3.167451132640093e-05, "loss": 0.0184, "step": 25540 }, { "grad_norm": 0.06169929355382919, "learning_rate": 3.163605729843746e-05, "loss": 0.0171, "step": 25550 }, { "grad_norm": 0.07194297760725021, "learning_rate": 3.159761582205565e-05, "loss": 0.0192, "step": 25560 }, { "grad_norm": 0.08248525857925415, "learning_rate": 3.155918692352992e-05, "loss": 0.0208, "step": 25570 }, { "grad_norm": 0.08410104364156723, "learning_rate": 3.152077062912602e-05, "loss": 0.0186, "step": 25580 }, { "grad_norm": 0.0976022407412529, "learning_rate": 3.148236696510117e-05, "loss": 0.0189, "step": 25590 }, { "grad_norm": 0.0827711820602417, "learning_rate": 3.144397595770388e-05, "loss": 0.0175, "step": 25600 }, { "grad_norm": 0.10284208506345749, "learning_rate": 3.1405597633174036e-05, "loss": 0.0173, "step": 25610 }, { "grad_norm": 0.08150866627693176, "learning_rate": 3.136723201774289e-05, "loss": 0.0169, "step": 25620 }, { "grad_norm": 0.08494960516691208, "learning_rate": 3.132887913763295e-05, "loss": 0.0185, "step": 25630 }, { "grad_norm": 0.07192284613847733, "learning_rate": 3.129053901905806e-05, "loss": 0.0166, "step": 25640 }, { "grad_norm": 0.09620898962020874, "learning_rate": 3.125221168822335e-05, "loss": 0.0164, "step": 25650 }, { "grad_norm": 0.062093690037727356, "learning_rate": 3.1213897171325154e-05, "loss": 0.0168, "step": 25660 }, { "grad_norm": 0.10129398107528687, "learning_rate": 3.1175595494551116e-05, "loss": 0.0163, "step": 25670 }, { "grad_norm": 0.09131728112697601, "learning_rate": 3.1137306684080045e-05, "loss": 0.0169, "step": 25680 }, { "grad_norm": 0.10164066404104233, "learning_rate": 3.1099030766081985e-05, "loss": 0.0184, "step": 25690 }, { "grad_norm": 0.11153025180101395, "learning_rate": 3.106076776671818e-05, "loss": 0.019, "step": 25700 }, { "grad_norm": 0.08351155370473862, "learning_rate": 3.102251771214101e-05, "loss": 0.0179, "step": 25710 }, { "grad_norm": 0.10950012505054474, "learning_rate": 3.098428062849404e-05, "loss": 0.0186, "step": 25720 }, { "grad_norm": 0.1020461916923523, "learning_rate": 3.094605654191195e-05, "loss": 0.0204, "step": 25730 }, { "grad_norm": 0.07710781693458557, "learning_rate": 3.090784547852055e-05, "loss": 0.0174, "step": 25740 }, { "grad_norm": 0.09819675981998444, "learning_rate": 3.0869647464436746e-05, "loss": 0.0153, "step": 25750 }, { "grad_norm": 0.08136934787034988, "learning_rate": 3.0831462525768496e-05, "loss": 0.0175, "step": 25760 }, { "grad_norm": 0.06706683337688446, "learning_rate": 3.079329068861488e-05, "loss": 0.017, "step": 25770 }, { "grad_norm": 0.08675328642129898, "learning_rate": 3.075513197906597e-05, "loss": 0.0222, "step": 25780 }, { "grad_norm": 0.09700522571802139, "learning_rate": 3.071698642320286e-05, "loss": 0.018, "step": 25790 }, { "grad_norm": 0.13029420375823975, "learning_rate": 3.067885404709772e-05, "loss": 0.0185, "step": 25800 }, { "grad_norm": 0.09786514937877655, "learning_rate": 3.0640734876813636e-05, "loss": 0.0165, "step": 25810 }, { "grad_norm": 0.07667297124862671, "learning_rate": 3.060262893840473e-05, "loss": 0.017, "step": 25820 }, { "grad_norm": 0.08109483122825623, "learning_rate": 3.056453625791603e-05, "loss": 0.0184, "step": 25830 }, { "grad_norm": 0.16887038946151733, "learning_rate": 3.052645686138353e-05, "loss": 0.0222, "step": 25840 }, { "grad_norm": 0.07717393338680267, "learning_rate": 3.0488390774834153e-05, "loss": 0.0214, "step": 25850 }, { "grad_norm": 0.07568430155515671, "learning_rate": 3.0450338024285684e-05, "loss": 0.0186, "step": 25860 }, { "grad_norm": 0.07976433634757996, "learning_rate": 3.0412298635746855e-05, "loss": 0.0184, "step": 25870 }, { "grad_norm": 0.10633580386638641, "learning_rate": 3.03742726352172e-05, "loss": 0.0203, "step": 25880 }, { "grad_norm": 0.18544696271419525, "learning_rate": 3.0336260048687125e-05, "loss": 0.0185, "step": 25890 }, { "grad_norm": 0.08322147279977798, "learning_rate": 3.0298260902137897e-05, "loss": 0.0181, "step": 25900 }, { "grad_norm": 0.08655161410570145, "learning_rate": 3.0260275221541566e-05, "loss": 0.02, "step": 25910 }, { "grad_norm": 0.1587589830160141, "learning_rate": 3.0222303032860987e-05, "loss": 0.0166, "step": 25920 }, { "grad_norm": 0.05309627205133438, "learning_rate": 3.018434436204979e-05, "loss": 0.0166, "step": 25930 }, { "grad_norm": 0.0758572667837143, "learning_rate": 3.014639923505237e-05, "loss": 0.0179, "step": 25940 }, { "grad_norm": 0.09847808629274368, "learning_rate": 3.0108467677803863e-05, "loss": 0.0198, "step": 25950 }, { "grad_norm": 0.08572404831647873, "learning_rate": 3.0070549716230156e-05, "loss": 0.0181, "step": 25960 }, { "grad_norm": 0.09122435003519058, "learning_rate": 3.003264537624777e-05, "loss": 0.0202, "step": 25970 }, { "grad_norm": 0.08870457112789154, "learning_rate": 2.9994754683764e-05, "loss": 0.0173, "step": 25980 }, { "grad_norm": 0.13555939495563507, "learning_rate": 2.9956877664676754e-05, "loss": 0.0217, "step": 25990 }, { "grad_norm": 0.0894797071814537, "learning_rate": 2.9919014344874636e-05, "loss": 0.0212, "step": 26000 }, { "grad_norm": 0.0954200029373169, "learning_rate": 2.9881164750236857e-05, "loss": 0.0182, "step": 26010 }, { "grad_norm": 0.08524684607982635, "learning_rate": 2.984332890663326e-05, "loss": 0.0201, "step": 26020 }, { "grad_norm": 0.10879406332969666, "learning_rate": 2.9805506839924292e-05, "loss": 0.0216, "step": 26030 }, { "grad_norm": 0.28024008870124817, "learning_rate": 2.9767698575960968e-05, "loss": 0.0192, "step": 26040 }, { "grad_norm": 0.19011911749839783, "learning_rate": 2.9729904140584913e-05, "loss": 0.019, "step": 26050 }, { "grad_norm": 0.11396462470293045, "learning_rate": 2.9692123559628234e-05, "loss": 0.0232, "step": 26060 }, { "grad_norm": 0.07529015839099884, "learning_rate": 2.9654356858913596e-05, "loss": 0.0168, "step": 26070 }, { "grad_norm": 0.12010021507740021, "learning_rate": 2.9616604064254206e-05, "loss": 0.0218, "step": 26080 }, { "grad_norm": 0.07557950168848038, "learning_rate": 2.9578865201453732e-05, "loss": 0.0196, "step": 26090 }, { "grad_norm": 0.08662091940641403, "learning_rate": 2.9541140296306335e-05, "loss": 0.023, "step": 26100 }, { "grad_norm": 0.09028445184230804, "learning_rate": 2.9503429374596627e-05, "loss": 0.0209, "step": 26110 }, { "grad_norm": 0.06481568515300751, "learning_rate": 2.946573246209967e-05, "loss": 0.0201, "step": 26120 }, { "grad_norm": 0.18823812901973724, "learning_rate": 2.942804958458094e-05, "loss": 0.0204, "step": 26130 }, { "grad_norm": 0.1071404293179512, "learning_rate": 2.9390380767796343e-05, "loss": 0.0185, "step": 26140 }, { "grad_norm": 0.06842785328626633, "learning_rate": 2.9352726037492174e-05, "loss": 0.0227, "step": 26150 }, { "grad_norm": 0.09582655876874924, "learning_rate": 2.9315085419405052e-05, "loss": 0.0215, "step": 26160 }, { "grad_norm": 0.10294234752655029, "learning_rate": 2.927745893926199e-05, "loss": 0.0209, "step": 26170 }, { "grad_norm": 0.19208484888076782, "learning_rate": 2.9239846622780358e-05, "loss": 0.0227, "step": 26180 }, { "grad_norm": 0.06925927847623825, "learning_rate": 2.9202248495667788e-05, "loss": 0.0171, "step": 26190 }, { "grad_norm": 0.17176106572151184, "learning_rate": 2.916466458362227e-05, "loss": 0.0225, "step": 26200 }, { "grad_norm": 0.07771653681993484, "learning_rate": 2.9127094912332033e-05, "loss": 0.0203, "step": 26210 }, { "grad_norm": 0.11785552650690079, "learning_rate": 2.9089539507475606e-05, "loss": 0.0214, "step": 26220 }, { "grad_norm": 0.08118528872728348, "learning_rate": 2.9051998394721748e-05, "loss": 0.0211, "step": 26230 }, { "grad_norm": 0.0661507174372673, "learning_rate": 2.901447159972948e-05, "loss": 0.0181, "step": 26240 }, { "grad_norm": 0.08945668488740921, "learning_rate": 2.8976959148148e-05, "loss": 0.0201, "step": 26250 }, { "grad_norm": 0.06413882970809937, "learning_rate": 2.8939461065616674e-05, "loss": 0.0203, "step": 26260 }, { "grad_norm": 0.10114417225122452, "learning_rate": 2.8901977377765127e-05, "loss": 0.0182, "step": 26270 }, { "grad_norm": 0.09792707860469818, "learning_rate": 2.8864508110213094e-05, "loss": 0.0173, "step": 26280 }, { "grad_norm": 0.06637674570083618, "learning_rate": 2.8827053288570503e-05, "loss": 0.0201, "step": 26290 }, { "grad_norm": 0.10650058835744858, "learning_rate": 2.8789612938437315e-05, "loss": 0.019, "step": 26300 }, { "grad_norm": 0.07611332833766937, "learning_rate": 2.8752187085403683e-05, "loss": 0.0206, "step": 26310 }, { "grad_norm": 0.07953467965126038, "learning_rate": 2.8714775755049818e-05, "loss": 0.0172, "step": 26320 }, { "grad_norm": 0.16190741956233978, "learning_rate": 2.867737897294604e-05, "loss": 0.0227, "step": 26330 }, { "grad_norm": 0.08843854814767838, "learning_rate": 2.8639996764652653e-05, "loss": 0.0187, "step": 26340 }, { "grad_norm": 0.07070910185575485, "learning_rate": 2.8602629155720084e-05, "loss": 0.0208, "step": 26350 }, { "grad_norm": 0.12116185575723648, "learning_rate": 2.8565276171688703e-05, "loss": 0.0173, "step": 26360 }, { "grad_norm": 0.10094000399112701, "learning_rate": 2.8527937838088943e-05, "loss": 0.0198, "step": 26370 }, { "grad_norm": 0.10346783697605133, "learning_rate": 2.84906141804412e-05, "loss": 0.0212, "step": 26380 }, { "grad_norm": 0.24735507369041443, "learning_rate": 2.8453305224255867e-05, "loss": 0.0166, "step": 26390 }, { "grad_norm": 0.07441047579050064, "learning_rate": 2.8416010995033216e-05, "loss": 0.018, "step": 26400 }, { "grad_norm": 0.13448470830917358, "learning_rate": 2.8378731518263524e-05, "loss": 0.0214, "step": 26410 }, { "grad_norm": 0.0791536197066307, "learning_rate": 2.834146681942696e-05, "loss": 0.0183, "step": 26420 }, { "grad_norm": 0.06920647621154785, "learning_rate": 2.8304216923993622e-05, "loss": 0.0203, "step": 26430 }, { "grad_norm": 0.08312463760375977, "learning_rate": 2.8266981857423413e-05, "loss": 0.0181, "step": 26440 }, { "grad_norm": 0.12488197535276413, "learning_rate": 2.8229761645166197e-05, "loss": 0.0204, "step": 26450 }, { "grad_norm": 0.09029130637645721, "learning_rate": 2.81925563126616e-05, "loss": 0.0203, "step": 26460 }, { "grad_norm": 0.07169969379901886, "learning_rate": 2.8155365885339124e-05, "loss": 0.0161, "step": 26470 }, { "grad_norm": 0.17072203755378723, "learning_rate": 2.8118190388618093e-05, "loss": 0.0163, "step": 26480 }, { "grad_norm": 0.12199023365974426, "learning_rate": 2.8081029847907607e-05, "loss": 0.0155, "step": 26490 }, { "grad_norm": 0.07416917383670807, "learning_rate": 2.8043884288606525e-05, "loss": 0.0195, "step": 26500 }, { "grad_norm": 0.11202110350131989, "learning_rate": 2.8006753736103496e-05, "loss": 0.0183, "step": 26510 }, { "grad_norm": 0.06047070026397705, "learning_rate": 2.7969638215776918e-05, "loss": 0.0193, "step": 26520 }, { "grad_norm": 0.06862282752990723, "learning_rate": 2.793253775299487e-05, "loss": 0.0149, "step": 26530 }, { "grad_norm": 0.060388851910829544, "learning_rate": 2.7895452373115184e-05, "loss": 0.0197, "step": 26540 }, { "grad_norm": 0.0773601233959198, "learning_rate": 2.785838210148539e-05, "loss": 0.0175, "step": 26550 }, { "grad_norm": 0.13067573308944702, "learning_rate": 2.782132696344263e-05, "loss": 0.016, "step": 26560 }, { "grad_norm": 0.1229168176651001, "learning_rate": 2.7784286984313745e-05, "loss": 0.0168, "step": 26570 }, { "grad_norm": 0.07871651649475098, "learning_rate": 2.7747262189415236e-05, "loss": 0.0172, "step": 26580 }, { "grad_norm": 0.0655105859041214, "learning_rate": 2.7710252604053205e-05, "loss": 0.0158, "step": 26590 }, { "grad_norm": 0.09347431361675262, "learning_rate": 2.767325825352332e-05, "loss": 0.0171, "step": 26600 }, { "grad_norm": 0.05662967637181282, "learning_rate": 2.7636279163110913e-05, "loss": 0.0171, "step": 26610 }, { "grad_norm": 0.0937347412109375, "learning_rate": 2.7599315358090795e-05, "loss": 0.0162, "step": 26620 }, { "grad_norm": 0.06525544822216034, "learning_rate": 2.7562366863727407e-05, "loss": 0.0182, "step": 26630 }, { "grad_norm": 0.08016493171453476, "learning_rate": 2.7525433705274695e-05, "loss": 0.02, "step": 26640 }, { "grad_norm": 0.09243352711200714, "learning_rate": 2.748851590797614e-05, "loss": 0.019, "step": 26650 }, { "grad_norm": 0.07199094444513321, "learning_rate": 2.7451613497064675e-05, "loss": 0.0161, "step": 26660 }, { "grad_norm": 0.09009566158056259, "learning_rate": 2.7414726497762765e-05, "loss": 0.0162, "step": 26670 }, { "grad_norm": 0.10383716225624084, "learning_rate": 2.737785493528232e-05, "loss": 0.0153, "step": 26680 }, { "grad_norm": 0.07803336530923843, "learning_rate": 2.7340998834824745e-05, "loss": 0.0157, "step": 26690 }, { "grad_norm": 0.06848856806755066, "learning_rate": 2.7304158221580777e-05, "loss": 0.0178, "step": 26700 }, { "grad_norm": 0.11148717999458313, "learning_rate": 2.7267333120730675e-05, "loss": 0.0209, "step": 26710 }, { "grad_norm": 0.0936250165104866, "learning_rate": 2.7230523557444017e-05, "loss": 0.0191, "step": 26720 }, { "grad_norm": 0.1007392480969429, "learning_rate": 2.7193729556879798e-05, "loss": 0.016, "step": 26730 }, { "grad_norm": 0.08597110211849213, "learning_rate": 2.715695114418637e-05, "loss": 0.0165, "step": 26740 }, { "grad_norm": 0.09572935104370117, "learning_rate": 2.7120188344501475e-05, "loss": 0.018, "step": 26750 }, { "grad_norm": 0.0755123645067215, "learning_rate": 2.7083441182952067e-05, "loss": 0.0195, "step": 26760 }, { "grad_norm": 0.058039844036102295, "learning_rate": 2.7046709684654527e-05, "loss": 0.0174, "step": 26770 }, { "grad_norm": 0.11478573828935623, "learning_rate": 2.700999387471448e-05, "loss": 0.0171, "step": 26780 }, { "grad_norm": 0.11108720302581787, "learning_rate": 2.6973293778226854e-05, "loss": 0.0189, "step": 26790 }, { "grad_norm": 0.06910032033920288, "learning_rate": 2.6936609420275804e-05, "loss": 0.0185, "step": 26800 }, { "grad_norm": 0.07389044016599655, "learning_rate": 2.689994082593472e-05, "loss": 0.0187, "step": 26810 }, { "grad_norm": 0.09344189614057541, "learning_rate": 2.6863288020266264e-05, "loss": 0.0179, "step": 26820 }, { "grad_norm": 0.14125002920627594, "learning_rate": 2.682665102832228e-05, "loss": 0.0182, "step": 26830 }, { "grad_norm": 0.08154676109552383, "learning_rate": 2.67900298751438e-05, "loss": 0.0205, "step": 26840 }, { "grad_norm": 0.10718286037445068, "learning_rate": 2.6753424585761067e-05, "loss": 0.0185, "step": 26850 }, { "grad_norm": 0.10678769648075104, "learning_rate": 2.671683518519341e-05, "loss": 0.0205, "step": 26860 }, { "grad_norm": 0.09141047298908234, "learning_rate": 2.668026169844936e-05, "loss": 0.0189, "step": 26870 }, { "grad_norm": 0.0843677967786789, "learning_rate": 2.6643704150526538e-05, "loss": 0.0208, "step": 26880 }, { "grad_norm": 0.0993688628077507, "learning_rate": 2.6607162566411716e-05, "loss": 0.016, "step": 26890 }, { "grad_norm": 0.0754043310880661, "learning_rate": 2.6570636971080697e-05, "loss": 0.0183, "step": 26900 }, { "grad_norm": 0.2765786945819855, "learning_rate": 2.6534127389498364e-05, "loss": 0.0187, "step": 26910 }, { "grad_norm": 0.076982781291008, "learning_rate": 2.6497633846618696e-05, "loss": 0.0162, "step": 26920 }, { "grad_norm": 0.09517190605401993, "learning_rate": 2.6461156367384677e-05, "loss": 0.0166, "step": 26930 }, { "grad_norm": 0.11929436773061752, "learning_rate": 2.6424694976728316e-05, "loss": 0.022, "step": 26940 }, { "grad_norm": 0.0716400071978569, "learning_rate": 2.6388249699570667e-05, "loss": 0.0182, "step": 26950 }, { "grad_norm": 0.07173044234514236, "learning_rate": 2.6351820560821672e-05, "loss": 0.0189, "step": 26960 }, { "grad_norm": 0.08867035061120987, "learning_rate": 2.631540758538034e-05, "loss": 0.018, "step": 26970 }, { "grad_norm": 0.06721360236406326, "learning_rate": 2.6279010798134597e-05, "loss": 0.017, "step": 26980 }, { "grad_norm": 0.06634267419576645, "learning_rate": 2.6242630223961305e-05, "loss": 0.0214, "step": 26990 }, { "grad_norm": 0.08615075051784515, "learning_rate": 2.6206265887726244e-05, "loss": 0.0173, "step": 27000 }, { "grad_norm": 0.09100867807865143, "learning_rate": 2.6169917814284066e-05, "loss": 0.0214, "step": 27010 }, { "grad_norm": 0.09246454387903214, "learning_rate": 2.6133586028478364e-05, "loss": 0.0176, "step": 27020 }, { "grad_norm": 0.1078685075044632, "learning_rate": 2.609727055514155e-05, "loss": 0.0171, "step": 27030 }, { "grad_norm": 0.0846615880727768, "learning_rate": 2.606097141909494e-05, "loss": 0.0167, "step": 27040 }, { "grad_norm": 0.0930689349770546, "learning_rate": 2.6024688645148644e-05, "loss": 0.0179, "step": 27050 }, { "grad_norm": 0.07850867509841919, "learning_rate": 2.5988422258101564e-05, "loss": 0.0181, "step": 27060 }, { "grad_norm": 0.10224615037441254, "learning_rate": 2.5952172282741453e-05, "loss": 0.0189, "step": 27070 }, { "grad_norm": 0.10244739800691605, "learning_rate": 2.5915938743844853e-05, "loss": 0.0193, "step": 27080 }, { "grad_norm": 0.07763911783695221, "learning_rate": 2.5879721666177003e-05, "loss": 0.0208, "step": 27090 }, { "grad_norm": 0.11373458057641983, "learning_rate": 2.5843521074491972e-05, "loss": 0.0196, "step": 27100 }, { "grad_norm": 0.070803701877594, "learning_rate": 2.5807336993532487e-05, "loss": 0.0204, "step": 27110 }, { "grad_norm": 0.05405993387103081, "learning_rate": 2.577116944803004e-05, "loss": 0.0184, "step": 27120 }, { "grad_norm": 0.08730056881904602, "learning_rate": 2.5735018462704818e-05, "loss": 0.0194, "step": 27130 }, { "grad_norm": 0.09823845326900482, "learning_rate": 2.5698884062265665e-05, "loss": 0.0194, "step": 27140 }, { "grad_norm": 0.07231563329696655, "learning_rate": 2.5662766271410134e-05, "loss": 0.0197, "step": 27150 }, { "grad_norm": 0.10000582039356232, "learning_rate": 2.5626665114824343e-05, "loss": 0.0186, "step": 27160 }, { "grad_norm": 0.08308842033147812, "learning_rate": 2.5590580617183148e-05, "loss": 0.0179, "step": 27170 }, { "grad_norm": 0.11853183805942535, "learning_rate": 2.5554512803149912e-05, "loss": 0.0233, "step": 27180 }, { "grad_norm": 0.32801511883735657, "learning_rate": 2.5518461697376662e-05, "loss": 0.0188, "step": 27190 }, { "grad_norm": 0.08901181071996689, "learning_rate": 2.548242732450402e-05, "loss": 0.0184, "step": 27200 }, { "grad_norm": 0.09557512402534485, "learning_rate": 2.5446409709161095e-05, "loss": 0.0169, "step": 27210 }, { "grad_norm": 0.0990920439362526, "learning_rate": 2.541040887596561e-05, "loss": 0.0177, "step": 27220 }, { "grad_norm": 0.09395725280046463, "learning_rate": 2.537442484952378e-05, "loss": 0.0173, "step": 27230 }, { "grad_norm": 0.09041307866573334, "learning_rate": 2.533845765443037e-05, "loss": 0.018, "step": 27240 }, { "grad_norm": 0.1872943639755249, "learning_rate": 2.530250731526863e-05, "loss": 0.0171, "step": 27250 }, { "grad_norm": 0.08584508299827576, "learning_rate": 2.5266573856610253e-05, "loss": 0.0176, "step": 27260 }, { "grad_norm": 0.07871998101472855, "learning_rate": 2.5230657303015403e-05, "loss": 0.0173, "step": 27270 }, { "grad_norm": 0.08716096729040146, "learning_rate": 2.5194757679032728e-05, "loss": 0.0188, "step": 27280 }, { "grad_norm": 0.09742093831300735, "learning_rate": 2.5158875009199278e-05, "loss": 0.018, "step": 27290 }, { "grad_norm": 0.11704129725694656, "learning_rate": 2.5123009318040537e-05, "loss": 0.0201, "step": 27300 }, { "grad_norm": 0.06969014555215836, "learning_rate": 2.508716063007034e-05, "loss": 0.0159, "step": 27310 }, { "grad_norm": 0.1330014020204544, "learning_rate": 2.5051328969790934e-05, "loss": 0.0169, "step": 27320 }, { "grad_norm": 0.10358835756778717, "learning_rate": 2.501551436169292e-05, "loss": 0.0182, "step": 27330 }, { "grad_norm": 0.06841596215963364, "learning_rate": 2.4979716830255255e-05, "loss": 0.0186, "step": 27340 }, { "grad_norm": 0.06640549749135971, "learning_rate": 2.4943936399945233e-05, "loss": 0.0194, "step": 27350 }, { "grad_norm": 0.1454964578151703, "learning_rate": 2.4908173095218412e-05, "loss": 0.02, "step": 27360 }, { "grad_norm": 0.1542341113090515, "learning_rate": 2.4872426940518663e-05, "loss": 0.0179, "step": 27370 }, { "grad_norm": 0.13682739436626434, "learning_rate": 2.4836697960278156e-05, "loss": 0.0172, "step": 27380 }, { "grad_norm": 0.06485991925001144, "learning_rate": 2.480098617891732e-05, "loss": 0.0175, "step": 27390 }, { "grad_norm": 0.10211925208568573, "learning_rate": 2.4765291620844837e-05, "loss": 0.019, "step": 27400 }, { "grad_norm": 0.21437600255012512, "learning_rate": 2.472961431045756e-05, "loss": 0.0175, "step": 27410 }, { "grad_norm": 0.07842282950878143, "learning_rate": 2.4693954272140622e-05, "loss": 0.019, "step": 27420 }, { "grad_norm": 0.0776517316699028, "learning_rate": 2.4658311530267315e-05, "loss": 0.0157, "step": 27430 }, { "grad_norm": 0.06283346563577652, "learning_rate": 2.4622686109199124e-05, "loss": 0.0155, "step": 27440 }, { "grad_norm": 0.08333742618560791, "learning_rate": 2.4587078033285695e-05, "loss": 0.0167, "step": 27450 }, { "grad_norm": 0.09873288124799728, "learning_rate": 2.45514873268648e-05, "loss": 0.0215, "step": 27460 }, { "grad_norm": 0.08015424758195877, "learning_rate": 2.4515914014262336e-05, "loss": 0.017, "step": 27470 }, { "grad_norm": 0.1402180790901184, "learning_rate": 2.4480358119792345e-05, "loss": 0.0167, "step": 27480 }, { "grad_norm": 0.1023123487830162, "learning_rate": 2.4444819667756942e-05, "loss": 0.0192, "step": 27490 }, { "grad_norm": 0.127866730093956, "learning_rate": 2.4409298682446346e-05, "loss": 0.0201, "step": 27500 }, { "grad_norm": 0.08179797232151031, "learning_rate": 2.437379518813877e-05, "loss": 0.0182, "step": 27510 }, { "grad_norm": 0.05652911961078644, "learning_rate": 2.4338309209100547e-05, "loss": 0.0173, "step": 27520 }, { "grad_norm": 0.07004822045564651, "learning_rate": 2.4302840769586004e-05, "loss": 0.0145, "step": 27530 }, { "grad_norm": 0.055072154849767685, "learning_rate": 2.42673898938375e-05, "loss": 0.0178, "step": 27540 }, { "grad_norm": 0.08376631140708923, "learning_rate": 2.4231956606085343e-05, "loss": 0.0163, "step": 27550 }, { "grad_norm": 0.06210372596979141, "learning_rate": 2.419654093054789e-05, "loss": 0.0172, "step": 27560 }, { "grad_norm": 0.11201035231351852, "learning_rate": 2.4161142891431375e-05, "loss": 0.0199, "step": 27570 }, { "grad_norm": 0.07391005754470825, "learning_rate": 2.412576251293005e-05, "loss": 0.0175, "step": 27580 }, { "grad_norm": 0.08256510645151138, "learning_rate": 2.4090399819226068e-05, "loss": 0.0157, "step": 27590 }, { "grad_norm": 0.07814958691596985, "learning_rate": 2.4055054834489514e-05, "loss": 0.0183, "step": 27600 }, { "grad_norm": 0.07931673526763916, "learning_rate": 2.401972758287832e-05, "loss": 0.0189, "step": 27610 }, { "grad_norm": 0.06821669638156891, "learning_rate": 2.398441808853834e-05, "loss": 0.018, "step": 27620 }, { "grad_norm": 0.06839270889759064, "learning_rate": 2.3949126375603288e-05, "loss": 0.0186, "step": 27630 }, { "grad_norm": 0.08940883725881577, "learning_rate": 2.3913852468194724e-05, "loss": 0.0186, "step": 27640 }, { "grad_norm": 0.0999814122915268, "learning_rate": 2.387859639042201e-05, "loss": 0.0227, "step": 27650 }, { "grad_norm": 0.09131372720003128, "learning_rate": 2.3843358166382368e-05, "loss": 0.0177, "step": 27660 }, { "grad_norm": 0.08513206243515015, "learning_rate": 2.3808137820160757e-05, "loss": 0.0198, "step": 27670 }, { "grad_norm": 0.10739981383085251, "learning_rate": 2.3772935375829975e-05, "loss": 0.0198, "step": 27680 }, { "grad_norm": 0.09025842696428299, "learning_rate": 2.3737750857450553e-05, "loss": 0.0175, "step": 27690 }, { "grad_norm": 0.08644499629735947, "learning_rate": 2.3702584289070805e-05, "loss": 0.019, "step": 27700 }, { "grad_norm": 0.12995128333568573, "learning_rate": 2.36674356947267e-05, "loss": 0.0159, "step": 27710 }, { "grad_norm": 0.0620269849896431, "learning_rate": 2.3632305098442004e-05, "loss": 0.0174, "step": 27720 }, { "grad_norm": 0.1178075522184372, "learning_rate": 2.3597192524228156e-05, "loss": 0.0224, "step": 27730 }, { "grad_norm": 0.06562872976064682, "learning_rate": 2.356209799608424e-05, "loss": 0.0198, "step": 27740 }, { "grad_norm": 0.08420301973819733, "learning_rate": 2.352702153799704e-05, "loss": 0.0191, "step": 27750 }, { "grad_norm": 0.16062228381633759, "learning_rate": 2.3491963173941018e-05, "loss": 0.0194, "step": 27760 }, { "grad_norm": 0.0750262662768364, "learning_rate": 2.3456922927878196e-05, "loss": 0.0195, "step": 27770 }, { "grad_norm": 0.1979132741689682, "learning_rate": 2.3421900823758257e-05, "loss": 0.0214, "step": 27780 }, { "grad_norm": 0.0967133417725563, "learning_rate": 2.3386896885518496e-05, "loss": 0.0196, "step": 27790 }, { "grad_norm": 0.07513219863176346, "learning_rate": 2.335191113708378e-05, "loss": 0.0177, "step": 27800 }, { "grad_norm": 0.09430678188800812, "learning_rate": 2.331694360236651e-05, "loss": 0.0206, "step": 27810 }, { "grad_norm": 0.09853056818246841, "learning_rate": 2.3281994305266702e-05, "loss": 0.0199, "step": 27820 }, { "grad_norm": 0.1083282008767128, "learning_rate": 2.3247063269671826e-05, "loss": 0.017, "step": 27830 }, { "grad_norm": 0.07745910435914993, "learning_rate": 2.321215051945695e-05, "loss": 0.0193, "step": 27840 }, { "grad_norm": 0.06837240606546402, "learning_rate": 2.3177256078484588e-05, "loss": 0.0172, "step": 27850 }, { "grad_norm": 0.06349069625139236, "learning_rate": 2.3142379970604798e-05, "loss": 0.0139, "step": 27860 }, { "grad_norm": 0.10800862312316895, "learning_rate": 2.3107522219655025e-05, "loss": 0.0183, "step": 27870 }, { "grad_norm": 0.09083639085292816, "learning_rate": 2.3072682849460236e-05, "loss": 0.0163, "step": 27880 }, { "grad_norm": 0.06723746657371521, "learning_rate": 2.303786188383281e-05, "loss": 0.0199, "step": 27890 }, { "grad_norm": 0.08210616558790207, "learning_rate": 2.300305934657257e-05, "loss": 0.0181, "step": 27900 }, { "grad_norm": 0.170663520693779, "learning_rate": 2.2968275261466677e-05, "loss": 0.0195, "step": 27910 }, { "grad_norm": 0.11725948005914688, "learning_rate": 2.293350965228977e-05, "loss": 0.0188, "step": 27920 }, { "grad_norm": 0.07987035065889359, "learning_rate": 2.2898762542803776e-05, "loss": 0.0214, "step": 27930 }, { "grad_norm": 0.06570592522621155, "learning_rate": 2.286403395675803e-05, "loss": 0.0207, "step": 27940 }, { "grad_norm": 0.08284555375576019, "learning_rate": 2.28293239178892e-05, "loss": 0.0187, "step": 27950 }, { "grad_norm": 0.08703675121068954, "learning_rate": 2.2794632449921287e-05, "loss": 0.0163, "step": 27960 }, { "grad_norm": 0.05784162878990173, "learning_rate": 2.275995957656555e-05, "loss": 0.0191, "step": 27970 }, { "grad_norm": 0.09599372744560242, "learning_rate": 2.272530532152058e-05, "loss": 0.0182, "step": 27980 }, { "grad_norm": 0.08099710196256638, "learning_rate": 2.2690669708472233e-05, "loss": 0.0147, "step": 27990 }, { "grad_norm": 0.08752192556858063, "learning_rate": 2.2656052761093655e-05, "loss": 0.0164, "step": 28000 }, { "grad_norm": 0.07266896963119507, "learning_rate": 2.262145450304517e-05, "loss": 0.0159, "step": 28010 }, { "grad_norm": 0.09443369507789612, "learning_rate": 2.2586874957974352e-05, "loss": 0.0207, "step": 28020 }, { "grad_norm": 0.053318679332733154, "learning_rate": 2.2552314149516012e-05, "loss": 0.0175, "step": 28030 }, { "grad_norm": 0.06967370957136154, "learning_rate": 2.2517772101292133e-05, "loss": 0.0189, "step": 28040 }, { "grad_norm": 0.0656699687242508, "learning_rate": 2.248324883691188e-05, "loss": 0.0171, "step": 28050 }, { "grad_norm": 0.13160130381584167, "learning_rate": 2.24487443799716e-05, "loss": 0.0184, "step": 28060 }, { "grad_norm": 0.10215429961681366, "learning_rate": 2.241425875405472e-05, "loss": 0.0157, "step": 28070 }, { "grad_norm": 0.07284529507160187, "learning_rate": 2.2379791982731868e-05, "loss": 0.0181, "step": 28080 }, { "grad_norm": 0.4122708737850189, "learning_rate": 2.2345344089560756e-05, "loss": 0.0175, "step": 28090 }, { "grad_norm": 0.08954097330570221, "learning_rate": 2.2310915098086206e-05, "loss": 0.0201, "step": 28100 }, { "grad_norm": 0.10922405123710632, "learning_rate": 2.227650503184009e-05, "loss": 0.0167, "step": 28110 }, { "grad_norm": 0.11993145942687988, "learning_rate": 2.2242113914341357e-05, "loss": 0.0192, "step": 28120 }, { "grad_norm": 0.06291310489177704, "learning_rate": 2.220774176909602e-05, "loss": 0.015, "step": 28130 }, { "grad_norm": 0.172145277261734, "learning_rate": 2.2173388619597114e-05, "loss": 0.014, "step": 28140 }, { "grad_norm": 0.06601810455322266, "learning_rate": 2.21390544893247e-05, "loss": 0.0165, "step": 28150 }, { "grad_norm": 0.09596738964319229, "learning_rate": 2.210473940174585e-05, "loss": 0.0178, "step": 28160 }, { "grad_norm": 0.15070590376853943, "learning_rate": 2.207044338031456e-05, "loss": 0.0163, "step": 28170 }, { "grad_norm": 0.10059724748134613, "learning_rate": 2.203616644847186e-05, "loss": 0.016, "step": 28180 }, { "grad_norm": 0.06811702996492386, "learning_rate": 2.200190862964571e-05, "loss": 0.0149, "step": 28190 }, { "grad_norm": 0.07920338958501816, "learning_rate": 2.1967669947251024e-05, "loss": 0.0203, "step": 28200 }, { "grad_norm": 0.0749339759349823, "learning_rate": 2.1933450424689583e-05, "loss": 0.018, "step": 28210 }, { "grad_norm": 0.0720069631934166, "learning_rate": 2.1899250085350142e-05, "loss": 0.0187, "step": 28220 }, { "grad_norm": 0.10789040476083755, "learning_rate": 2.1865068952608277e-05, "loss": 0.0197, "step": 28230 }, { "grad_norm": 0.06664618104696274, "learning_rate": 2.1830907049826487e-05, "loss": 0.0178, "step": 28240 }, { "grad_norm": 0.05567034333944321, "learning_rate": 2.179676440035411e-05, "loss": 0.0178, "step": 28250 }, { "grad_norm": 0.09864348918199539, "learning_rate": 2.1762641027527337e-05, "loss": 0.0197, "step": 28260 }, { "grad_norm": 0.11861371248960495, "learning_rate": 2.1728536954669143e-05, "loss": 0.019, "step": 28270 }, { "grad_norm": 0.08008891344070435, "learning_rate": 2.169445220508936e-05, "loss": 0.0177, "step": 28280 }, { "grad_norm": 0.06604959070682526, "learning_rate": 2.166038680208461e-05, "loss": 0.0159, "step": 28290 }, { "grad_norm": 0.1004752591252327, "learning_rate": 2.162634076893823e-05, "loss": 0.016, "step": 28300 }, { "grad_norm": 0.09568290412425995, "learning_rate": 2.1592314128920388e-05, "loss": 0.0155, "step": 28310 }, { "grad_norm": 0.06914616376161575, "learning_rate": 2.155830690528799e-05, "loss": 0.0205, "step": 28320 }, { "grad_norm": 0.08096065372228622, "learning_rate": 2.1524319121284613e-05, "loss": 0.0201, "step": 28330 }, { "grad_norm": 0.07676758617162704, "learning_rate": 2.1490350800140607e-05, "loss": 0.0173, "step": 28340 }, { "grad_norm": 0.06297039985656738, "learning_rate": 2.1456401965073002e-05, "loss": 0.0194, "step": 28350 }, { "grad_norm": 0.08131728321313858, "learning_rate": 2.1422472639285524e-05, "loss": 0.0173, "step": 28360 }, { "grad_norm": 0.09783311188220978, "learning_rate": 2.13885628459685e-05, "loss": 0.018, "step": 28370 }, { "grad_norm": 0.09014569222927094, "learning_rate": 2.135467260829901e-05, "loss": 0.0188, "step": 28380 }, { "grad_norm": 0.06971389800310135, "learning_rate": 2.1320801949440654e-05, "loss": 0.0169, "step": 28390 }, { "grad_norm": 0.12277638167142868, "learning_rate": 2.1286950892543744e-05, "loss": 0.0199, "step": 28400 }, { "grad_norm": 0.08199486881494522, "learning_rate": 2.125311946074515e-05, "loss": 0.0185, "step": 28410 }, { "grad_norm": 0.1120416596531868, "learning_rate": 2.1219307677168355e-05, "loss": 0.0194, "step": 28420 }, { "grad_norm": 0.09133369475603104, "learning_rate": 2.118551556492336e-05, "loss": 0.0181, "step": 28430 }, { "grad_norm": 0.07002566009759903, "learning_rate": 2.1151743147106774e-05, "loss": 0.0196, "step": 28440 }, { "grad_norm": 0.15924811363220215, "learning_rate": 2.111799044680172e-05, "loss": 0.018, "step": 28450 }, { "grad_norm": 0.09099172800779343, "learning_rate": 2.1084257487077873e-05, "loss": 0.0206, "step": 28460 }, { "grad_norm": 0.09223033487796783, "learning_rate": 2.1050544290991357e-05, "loss": 0.0173, "step": 28470 }, { "grad_norm": 0.07928154617547989, "learning_rate": 2.101685088158486e-05, "loss": 0.0192, "step": 28480 }, { "grad_norm": 0.08048178255558014, "learning_rate": 2.0983177281887472e-05, "loss": 0.0186, "step": 28490 }, { "grad_norm": 0.08869882673025131, "learning_rate": 2.0949523514914798e-05, "loss": 0.0171, "step": 28500 }, { "grad_norm": 0.06659136712551117, "learning_rate": 2.0915889603668876e-05, "loss": 0.0159, "step": 28510 }, { "grad_norm": 0.06177138164639473, "learning_rate": 2.0882275571138175e-05, "loss": 0.0184, "step": 28520 }, { "grad_norm": 0.07100123167037964, "learning_rate": 2.0848681440297545e-05, "loss": 0.0156, "step": 28530 }, { "grad_norm": 0.11097676306962967, "learning_rate": 2.081510723410827e-05, "loss": 0.0187, "step": 28540 }, { "grad_norm": 0.12224022299051285, "learning_rate": 2.0781552975518003e-05, "loss": 0.016, "step": 28550 }, { "grad_norm": 0.06041055545210838, "learning_rate": 2.074801868746078e-05, "loss": 0.0137, "step": 28560 }, { "grad_norm": 0.0690428763628006, "learning_rate": 2.0714504392856955e-05, "loss": 0.0156, "step": 28570 }, { "grad_norm": 0.08122513443231583, "learning_rate": 2.0681010114613215e-05, "loss": 0.016, "step": 28580 }, { "grad_norm": 0.04773931950330734, "learning_rate": 2.0647535875622597e-05, "loss": 0.0159, "step": 28590 }, { "grad_norm": 0.06074430048465729, "learning_rate": 2.0614081698764432e-05, "loss": 0.0182, "step": 28600 }, { "grad_norm": 0.08468641340732574, "learning_rate": 2.0580647606904334e-05, "loss": 0.0162, "step": 28610 }, { "grad_norm": 0.10399452596902847, "learning_rate": 2.0547233622894208e-05, "loss": 0.0215, "step": 28620 }, { "grad_norm": 0.08157650381326675, "learning_rate": 2.0513839769572157e-05, "loss": 0.0178, "step": 28630 }, { "grad_norm": 0.08136937767267227, "learning_rate": 2.0480466069762584e-05, "loss": 0.0176, "step": 28640 }, { "grad_norm": 0.12639233469963074, "learning_rate": 2.0447112546276104e-05, "loss": 0.0162, "step": 28650 }, { "grad_norm": 0.0910225585103035, "learning_rate": 2.0413779221909547e-05, "loss": 0.0141, "step": 28660 }, { "grad_norm": 0.3761611580848694, "learning_rate": 2.0380466119445912e-05, "loss": 0.0181, "step": 28670 }, { "grad_norm": 0.05687123164534569, "learning_rate": 2.0347173261654373e-05, "loss": 0.0172, "step": 28680 }, { "grad_norm": 0.08772983402013779, "learning_rate": 2.03139006712903e-05, "loss": 0.0159, "step": 28690 }, { "grad_norm": 0.06817562878131866, "learning_rate": 2.028064837109519e-05, "loss": 0.0168, "step": 28700 }, { "grad_norm": 0.07312539964914322, "learning_rate": 2.0247416383796685e-05, "loss": 0.0163, "step": 28710 }, { "grad_norm": 0.07655881345272064, "learning_rate": 2.0214204732108548e-05, "loss": 0.0162, "step": 28720 }, { "grad_norm": 0.08170782029628754, "learning_rate": 2.0181013438730596e-05, "loss": 0.0167, "step": 28730 }, { "grad_norm": 0.04793785512447357, "learning_rate": 2.0147842526348783e-05, "loss": 0.0144, "step": 28740 }, { "grad_norm": 0.07952474057674408, "learning_rate": 2.011469201763511e-05, "loss": 0.0151, "step": 28750 }, { "grad_norm": 0.07231157273054123, "learning_rate": 2.0081561935247665e-05, "loss": 0.0183, "step": 28760 }, { "grad_norm": 0.07715283334255219, "learning_rate": 2.0048452301830523e-05, "loss": 0.0174, "step": 28770 }, { "grad_norm": 0.07760003954172134, "learning_rate": 2.0015363140013788e-05, "loss": 0.0185, "step": 28780 }, { "grad_norm": 0.07402795553207397, "learning_rate": 1.9982294472413606e-05, "loss": 0.0152, "step": 28790 }, { "grad_norm": 0.063426673412323, "learning_rate": 1.9949246321632103e-05, "loss": 0.0166, "step": 28800 }, { "grad_norm": 0.06376315653324127, "learning_rate": 1.9916218710257377e-05, "loss": 0.0158, "step": 28810 }, { "grad_norm": 0.07550597190856934, "learning_rate": 1.988321166086351e-05, "loss": 0.0157, "step": 28820 }, { "grad_norm": 0.07623062282800674, "learning_rate": 1.9850225196010468e-05, "loss": 0.0129, "step": 28830 }, { "grad_norm": 0.06800027936697006, "learning_rate": 1.981725933824421e-05, "loss": 0.0156, "step": 28840 }, { "grad_norm": 0.09057305008172989, "learning_rate": 1.978431411009661e-05, "loss": 0.0178, "step": 28850 }, { "grad_norm": 0.048561181873083115, "learning_rate": 1.9751389534085375e-05, "loss": 0.0156, "step": 28860 }, { "grad_norm": 0.06893578916788101, "learning_rate": 1.9718485632714184e-05, "loss": 0.0203, "step": 28870 }, { "grad_norm": 0.08799061179161072, "learning_rate": 1.968560242847251e-05, "loss": 0.017, "step": 28880 }, { "grad_norm": 0.06360359489917755, "learning_rate": 1.965273994383573e-05, "loss": 0.0156, "step": 28890 }, { "grad_norm": 0.10244636237621307, "learning_rate": 1.961989820126504e-05, "loss": 0.0179, "step": 28900 }, { "grad_norm": 0.07783586531877518, "learning_rate": 1.958707722320746e-05, "loss": 0.0154, "step": 28910 }, { "grad_norm": 0.08424399048089981, "learning_rate": 1.955427703209584e-05, "loss": 0.0169, "step": 28920 }, { "grad_norm": 0.0916755273938179, "learning_rate": 1.9521497650348764e-05, "loss": 0.0187, "step": 28930 }, { "grad_norm": 0.08292637020349503, "learning_rate": 1.948873910037067e-05, "loss": 0.014, "step": 28940 }, { "grad_norm": 0.08396320044994354, "learning_rate": 1.9456001404551678e-05, "loss": 0.0146, "step": 28950 }, { "grad_norm": 0.07205771654844284, "learning_rate": 1.942328458526771e-05, "loss": 0.0154, "step": 28960 }, { "grad_norm": 0.07673470675945282, "learning_rate": 1.9390588664880427e-05, "loss": 0.0167, "step": 28970 }, { "grad_norm": 0.07673560082912445, "learning_rate": 1.9357913665737145e-05, "loss": 0.0158, "step": 28980 }, { "grad_norm": 0.13256670534610748, "learning_rate": 1.932525961017093e-05, "loss": 0.0191, "step": 28990 }, { "grad_norm": 0.072689488530159, "learning_rate": 1.9292626520500533e-05, "loss": 0.015, "step": 29000 }, { "grad_norm": 0.22309279441833496, "learning_rate": 1.9260014419030354e-05, "loss": 0.0183, "step": 29010 }, { "grad_norm": 0.11266376823186874, "learning_rate": 1.9227423328050475e-05, "loss": 0.0166, "step": 29020 }, { "grad_norm": 0.29358863830566406, "learning_rate": 1.9194853269836582e-05, "loss": 0.0142, "step": 29030 }, { "grad_norm": 0.1310613602399826, "learning_rate": 1.916230426664999e-05, "loss": 0.02, "step": 29040 }, { "grad_norm": 0.18993358314037323, "learning_rate": 1.912977634073765e-05, "loss": 0.0157, "step": 29050 }, { "grad_norm": 0.07064948230981827, "learning_rate": 1.9097269514332083e-05, "loss": 0.018, "step": 29060 }, { "grad_norm": 0.06640850007534027, "learning_rate": 1.9064783809651433e-05, "loss": 0.0161, "step": 29070 }, { "grad_norm": 0.11338693648576736, "learning_rate": 1.9032319248899333e-05, "loss": 0.0165, "step": 29080 }, { "grad_norm": 0.0687950924038887, "learning_rate": 1.8999875854265015e-05, "loss": 0.0207, "step": 29090 }, { "grad_norm": 0.06994554400444031, "learning_rate": 1.8967453647923232e-05, "loss": 0.0166, "step": 29100 }, { "grad_norm": 0.09694486856460571, "learning_rate": 1.893505265203427e-05, "loss": 0.0192, "step": 29110 }, { "grad_norm": 0.07302246242761612, "learning_rate": 1.8902672888743907e-05, "loss": 0.0172, "step": 29120 }, { "grad_norm": 0.08020859956741333, "learning_rate": 1.8870314380183396e-05, "loss": 0.0156, "step": 29130 }, { "grad_norm": 0.09155192226171494, "learning_rate": 1.8837977148469448e-05, "loss": 0.0169, "step": 29140 }, { "grad_norm": 0.18424157798290253, "learning_rate": 1.880566121570429e-05, "loss": 0.0152, "step": 29150 }, { "grad_norm": 0.0859399288892746, "learning_rate": 1.877336660397554e-05, "loss": 0.0159, "step": 29160 }, { "grad_norm": 0.08835405111312866, "learning_rate": 1.874109333535628e-05, "loss": 0.0164, "step": 29170 }, { "grad_norm": 0.0682586282491684, "learning_rate": 1.870884143190496e-05, "loss": 0.0171, "step": 29180 }, { "grad_norm": 0.10857293009757996, "learning_rate": 1.867661091566546e-05, "loss": 0.015, "step": 29190 }, { "grad_norm": 0.06685224920511246, "learning_rate": 1.864440180866704e-05, "loss": 0.0181, "step": 29200 }, { "grad_norm": 0.07783001661300659, "learning_rate": 1.8612214132924317e-05, "loss": 0.0162, "step": 29210 }, { "grad_norm": 0.08203861862421036, "learning_rate": 1.858004791043728e-05, "loss": 0.0189, "step": 29220 }, { "grad_norm": 0.06644943356513977, "learning_rate": 1.854790316319123e-05, "loss": 0.0165, "step": 29230 }, { "grad_norm": 0.08235234022140503, "learning_rate": 1.8515779913156766e-05, "loss": 0.0165, "step": 29240 }, { "grad_norm": 0.0634709894657135, "learning_rate": 1.848367818228986e-05, "loss": 0.0148, "step": 29250 }, { "grad_norm": 0.07556535303592682, "learning_rate": 1.8451597992531733e-05, "loss": 0.0192, "step": 29260 }, { "grad_norm": 0.09592963010072708, "learning_rate": 1.8419539365808914e-05, "loss": 0.0186, "step": 29270 }, { "grad_norm": 0.13975532352924347, "learning_rate": 1.838750232403313e-05, "loss": 0.0144, "step": 29280 }, { "grad_norm": 0.07229122519493103, "learning_rate": 1.835548688910142e-05, "loss": 0.0171, "step": 29290 }, { "grad_norm": 0.08428029716014862, "learning_rate": 1.8323493082896037e-05, "loss": 0.0151, "step": 29300 }, { "grad_norm": 0.058495283126831055, "learning_rate": 1.8291520927284454e-05, "loss": 0.016, "step": 29310 }, { "grad_norm": 0.11895062774419785, "learning_rate": 1.8259570444119305e-05, "loss": 0.0179, "step": 29320 }, { "grad_norm": 0.08897474408149719, "learning_rate": 1.8227641655238488e-05, "loss": 0.0195, "step": 29330 }, { "grad_norm": 0.07867228984832764, "learning_rate": 1.819573458246498e-05, "loss": 0.0163, "step": 29340 }, { "grad_norm": 0.07264307141304016, "learning_rate": 1.816384924760699e-05, "loss": 0.0167, "step": 29350 }, { "grad_norm": 0.09409717470407486, "learning_rate": 1.813198567245784e-05, "loss": 0.0146, "step": 29360 }, { "grad_norm": 0.06556922942399979, "learning_rate": 1.8100143878796006e-05, "loss": 0.0133, "step": 29370 }, { "grad_norm": 0.0999930128455162, "learning_rate": 1.8068323888385015e-05, "loss": 0.016, "step": 29380 }, { "grad_norm": 0.06874487549066544, "learning_rate": 1.803652572297355e-05, "loss": 0.0163, "step": 29390 }, { "grad_norm": 0.07122944295406342, "learning_rate": 1.8004749404295353e-05, "loss": 0.0189, "step": 29400 }, { "grad_norm": 0.08265940099954605, "learning_rate": 1.797299495406926e-05, "loss": 0.0157, "step": 29410 }, { "grad_norm": 0.10547750443220139, "learning_rate": 1.7941262393999103e-05, "loss": 0.0159, "step": 29420 }, { "grad_norm": 0.12280306965112686, "learning_rate": 1.7909551745773816e-05, "loss": 0.0162, "step": 29430 }, { "grad_norm": 0.05721360445022583, "learning_rate": 1.7877863031067304e-05, "loss": 0.0152, "step": 29440 }, { "grad_norm": 0.07730048894882202, "learning_rate": 1.7846196271538516e-05, "loss": 0.0165, "step": 29450 }, { "grad_norm": 0.08944372832775116, "learning_rate": 1.7814551488831384e-05, "loss": 0.0174, "step": 29460 }, { "grad_norm": 0.05692332610487938, "learning_rate": 1.7782928704574835e-05, "loss": 0.0145, "step": 29470 }, { "grad_norm": 0.22096964716911316, "learning_rate": 1.775132794038271e-05, "loss": 0.0163, "step": 29480 }, { "grad_norm": 0.12893417477607727, "learning_rate": 1.7719749217853855e-05, "loss": 0.0144, "step": 29490 }, { "grad_norm": 0.10026847571134567, "learning_rate": 1.7688192558572038e-05, "loss": 0.014, "step": 29500 }, { "grad_norm": 0.06271573901176453, "learning_rate": 1.7656657984105906e-05, "loss": 0.018, "step": 29510 }, { "grad_norm": 0.13335904479026794, "learning_rate": 1.7625145516009068e-05, "loss": 0.0181, "step": 29520 }, { "grad_norm": 0.10464290529489517, "learning_rate": 1.7593655175820005e-05, "loss": 0.0156, "step": 29530 }, { "grad_norm": 0.048705145716667175, "learning_rate": 1.7562186985062046e-05, "loss": 0.015, "step": 29540 }, { "grad_norm": 0.08445973694324493, "learning_rate": 1.7530740965243403e-05, "loss": 0.0146, "step": 29550 }, { "grad_norm": 0.06914617866277695, "learning_rate": 1.7499317137857153e-05, "loss": 0.0188, "step": 29560 }, { "grad_norm": 0.08405019342899323, "learning_rate": 1.7467915524381184e-05, "loss": 0.014, "step": 29570 }, { "grad_norm": 0.07931676506996155, "learning_rate": 1.7436536146278182e-05, "loss": 0.0151, "step": 29580 }, { "grad_norm": 0.06816226989030838, "learning_rate": 1.7405179024995688e-05, "loss": 0.0158, "step": 29590 }, { "grad_norm": 0.07851288467645645, "learning_rate": 1.737384418196596e-05, "loss": 0.0163, "step": 29600 }, { "grad_norm": 0.13261465728282928, "learning_rate": 1.734253163860609e-05, "loss": 0.0172, "step": 29610 }, { "grad_norm": 0.06970828771591187, "learning_rate": 1.7311241416317896e-05, "loss": 0.0144, "step": 29620 }, { "grad_norm": 0.08834677189588547, "learning_rate": 1.7279973536487982e-05, "loss": 0.0151, "step": 29630 }, { "grad_norm": 0.06543765217065811, "learning_rate": 1.724872802048761e-05, "loss": 0.0146, "step": 29640 }, { "grad_norm": 0.10835070163011551, "learning_rate": 1.7217504889672803e-05, "loss": 0.0167, "step": 29650 }, { "grad_norm": 0.14792367815971375, "learning_rate": 1.7186304165384287e-05, "loss": 0.016, "step": 29660 }, { "grad_norm": 0.08259417116641998, "learning_rate": 1.7155125868947475e-05, "loss": 0.0166, "step": 29670 }, { "grad_norm": 0.11108500510454178, "learning_rate": 1.7123970021672404e-05, "loss": 0.0163, "step": 29680 }, { "grad_norm": 0.10158473253250122, "learning_rate": 1.709283664485384e-05, "loss": 0.0143, "step": 29690 }, { "grad_norm": 0.05960426107048988, "learning_rate": 1.7061725759771113e-05, "loss": 0.0136, "step": 29700 }, { "grad_norm": 0.10990497469902039, "learning_rate": 1.7030637387688248e-05, "loss": 0.0165, "step": 29710 }, { "grad_norm": 0.06567469984292984, "learning_rate": 1.6999571549853836e-05, "loss": 0.0151, "step": 29720 }, { "grad_norm": 0.10523282736539841, "learning_rate": 1.696852826750112e-05, "loss": 0.0158, "step": 29730 }, { "grad_norm": 0.5418563485145569, "learning_rate": 1.6937507561847844e-05, "loss": 0.0186, "step": 29740 }, { "grad_norm": 0.0941271185874939, "learning_rate": 1.6906509454096385e-05, "loss": 0.0163, "step": 29750 }, { "grad_norm": 0.08398736268281937, "learning_rate": 1.687553396543367e-05, "loss": 0.0178, "step": 29760 }, { "grad_norm": 0.11243259161710739, "learning_rate": 1.6844581117031154e-05, "loss": 0.0161, "step": 29770 }, { "grad_norm": 0.09983205795288086, "learning_rate": 1.681365093004481e-05, "loss": 0.0159, "step": 29780 }, { "grad_norm": 0.10658272355794907, "learning_rate": 1.678274342561511e-05, "loss": 0.0174, "step": 29790 }, { "grad_norm": 0.07865481078624725, "learning_rate": 1.675185862486706e-05, "loss": 0.0197, "step": 29800 }, { "grad_norm": 0.056257475167512894, "learning_rate": 1.6720996548910127e-05, "loss": 0.0184, "step": 29810 }, { "grad_norm": 0.06337353587150574, "learning_rate": 1.6690157218838247e-05, "loss": 0.0175, "step": 29820 }, { "grad_norm": 0.11754006892442703, "learning_rate": 1.665934065572984e-05, "loss": 0.0156, "step": 29830 }, { "grad_norm": 0.06928753852844238, "learning_rate": 1.6628546880647688e-05, "loss": 0.019, "step": 29840 }, { "grad_norm": 0.07797639071941376, "learning_rate": 1.6597775914639076e-05, "loss": 0.0185, "step": 29850 }, { "grad_norm": 0.09780628234148026, "learning_rate": 1.6567027778735654e-05, "loss": 0.0178, "step": 29860 }, { "grad_norm": 0.08905419707298279, "learning_rate": 1.653630249395351e-05, "loss": 0.0204, "step": 29870 }, { "grad_norm": 0.16102439165115356, "learning_rate": 1.6505600081293072e-05, "loss": 0.0148, "step": 29880 }, { "grad_norm": 0.074723020195961, "learning_rate": 1.647492056173912e-05, "loss": 0.0152, "step": 29890 }, { "grad_norm": 0.09310749173164368, "learning_rate": 1.644426395626085e-05, "loss": 0.0191, "step": 29900 }, { "grad_norm": 0.053928814828395844, "learning_rate": 1.641363028581175e-05, "loss": 0.0161, "step": 29910 }, { "grad_norm": 0.05526569113135338, "learning_rate": 1.638301957132965e-05, "loss": 0.0152, "step": 29920 }, { "grad_norm": 0.081576868891716, "learning_rate": 1.6352431833736703e-05, "loss": 0.0171, "step": 29930 }, { "grad_norm": 0.11447159945964813, "learning_rate": 1.6321867093939298e-05, "loss": 0.0171, "step": 29940 }, { "grad_norm": 0.08973095566034317, "learning_rate": 1.629132537282817e-05, "loss": 0.0163, "step": 29950 }, { "grad_norm": 0.06546973437070847, "learning_rate": 1.62608066912783e-05, "loss": 0.0184, "step": 29960 }, { "grad_norm": 0.08490284532308578, "learning_rate": 1.623031107014893e-05, "loss": 0.0155, "step": 29970 }, { "grad_norm": 0.07095350325107574, "learning_rate": 1.619983853028351e-05, "loss": 0.0162, "step": 29980 }, { "grad_norm": 0.11744535714387894, "learning_rate": 1.6169389092509724e-05, "loss": 0.0168, "step": 29990 }, { "grad_norm": 0.10393309593200684, "learning_rate": 1.6138962777639494e-05, "loss": 0.0173, "step": 30000 }, { "grad_norm": 0.10025988519191742, "learning_rate": 1.610855960646891e-05, "loss": 0.0203, "step": 30010 }, { "grad_norm": 0.17401708662509918, "learning_rate": 1.607817959977826e-05, "loss": 0.0181, "step": 30020 }, { "grad_norm": 0.09727147966623306, "learning_rate": 1.6047822778332005e-05, "loss": 0.0202, "step": 30030 }, { "grad_norm": 0.10186654329299927, "learning_rate": 1.6017489162878713e-05, "loss": 0.0202, "step": 30040 }, { "grad_norm": 0.08231864124536514, "learning_rate": 1.5987178774151147e-05, "loss": 0.0169, "step": 30050 }, { "grad_norm": 0.18657517433166504, "learning_rate": 1.5956891632866184e-05, "loss": 0.0176, "step": 30060 }, { "grad_norm": 0.07518764585256577, "learning_rate": 1.592662775972476e-05, "loss": 0.0168, "step": 30070 }, { "grad_norm": 0.07731018960475922, "learning_rate": 1.5896387175411986e-05, "loss": 0.0171, "step": 30080 }, { "grad_norm": 0.06352604925632477, "learning_rate": 1.586616990059699e-05, "loss": 0.0191, "step": 30090 }, { "grad_norm": 0.07714260369539261, "learning_rate": 1.5835975955933002e-05, "loss": 0.0188, "step": 30100 }, { "grad_norm": 0.07688265293836594, "learning_rate": 1.5805805362057295e-05, "loss": 0.0185, "step": 30110 }, { "grad_norm": 0.09432581067085266, "learning_rate": 1.57756581395912e-05, "loss": 0.017, "step": 30120 }, { "grad_norm": 0.09485168755054474, "learning_rate": 1.574553430914006e-05, "loss": 0.0199, "step": 30130 }, { "grad_norm": 0.0959511250257492, "learning_rate": 1.5715433891293206e-05, "loss": 0.0187, "step": 30140 }, { "grad_norm": 0.09481579065322876, "learning_rate": 1.5685356906624016e-05, "loss": 0.0204, "step": 30150 }, { "grad_norm": 0.08606783300638199, "learning_rate": 1.56553033756898e-05, "loss": 0.0203, "step": 30160 }, { "grad_norm": 0.0826239287853241, "learning_rate": 1.5625273319031873e-05, "loss": 0.0195, "step": 30170 }, { "grad_norm": 0.1027667298913002, "learning_rate": 1.5595266757175518e-05, "loss": 0.0209, "step": 30180 }, { "grad_norm": 0.09475474804639816, "learning_rate": 1.55652837106299e-05, "loss": 0.0195, "step": 30190 }, { "grad_norm": 0.08784739673137665, "learning_rate": 1.5535324199888164e-05, "loss": 0.0185, "step": 30200 }, { "grad_norm": 0.09317483752965927, "learning_rate": 1.5505388245427355e-05, "loss": 0.018, "step": 30210 }, { "grad_norm": 0.0775645300745964, "learning_rate": 1.5475475867708418e-05, "loss": 0.0173, "step": 30220 }, { "grad_norm": 0.07626856863498688, "learning_rate": 1.5445587087176195e-05, "loss": 0.0171, "step": 30230 }, { "grad_norm": 0.07953917980194092, "learning_rate": 1.5415721924259346e-05, "loss": 0.0172, "step": 30240 }, { "grad_norm": 0.06952299922704697, "learning_rate": 1.538588039937047e-05, "loss": 0.0142, "step": 30250 }, { "grad_norm": 0.07187306135892868, "learning_rate": 1.535606253290594e-05, "loss": 0.0158, "step": 30260 }, { "grad_norm": 0.07876410335302353, "learning_rate": 1.5326268345245993e-05, "loss": 0.0165, "step": 30270 }, { "grad_norm": 0.0779549852013588, "learning_rate": 1.5296497856754693e-05, "loss": 0.0168, "step": 30280 }, { "grad_norm": 0.06468462198972702, "learning_rate": 1.5266751087779858e-05, "loss": 0.015, "step": 30290 }, { "grad_norm": 0.06683225184679031, "learning_rate": 1.5237028058653142e-05, "loss": 0.016, "step": 30300 }, { "grad_norm": 0.07087993621826172, "learning_rate": 1.5207328789689946e-05, "loss": 0.0164, "step": 30310 }, { "grad_norm": 0.07233627140522003, "learning_rate": 1.5177653301189454e-05, "loss": 0.0152, "step": 30320 }, { "grad_norm": 0.11075931042432785, "learning_rate": 1.5148001613434587e-05, "loss": 0.0198, "step": 30330 }, { "grad_norm": 0.09335537999868393, "learning_rate": 1.5118373746691983e-05, "loss": 0.0143, "step": 30340 }, { "grad_norm": 0.09066010266542435, "learning_rate": 1.5088769721211992e-05, "loss": 0.0148, "step": 30350 }, { "grad_norm": 0.09229134768247604, "learning_rate": 1.505918955722871e-05, "loss": 0.0179, "step": 30360 }, { "grad_norm": 0.15512797236442566, "learning_rate": 1.5029633274959886e-05, "loss": 0.0173, "step": 30370 }, { "grad_norm": 0.14444805681705475, "learning_rate": 1.5000100894606989e-05, "loss": 0.0173, "step": 30380 }, { "grad_norm": 0.08049265295267105, "learning_rate": 1.4970592436355079e-05, "loss": 0.0161, "step": 30390 }, { "grad_norm": 0.07600360363721848, "learning_rate": 1.4941107920372938e-05, "loss": 0.0167, "step": 30400 }, { "grad_norm": 0.11188232153654099, "learning_rate": 1.4911647366812947e-05, "loss": 0.0153, "step": 30410 }, { "grad_norm": 0.07865136116743088, "learning_rate": 1.4882210795811114e-05, "loss": 0.0171, "step": 30420 }, { "grad_norm": 0.08732388913631439, "learning_rate": 1.485279822748708e-05, "loss": 0.0163, "step": 30430 }, { "grad_norm": 0.08783799409866333, "learning_rate": 1.4823409681944034e-05, "loss": 0.0182, "step": 30440 }, { "grad_norm": 0.0742197334766388, "learning_rate": 1.4794045179268763e-05, "loss": 0.0151, "step": 30450 }, { "grad_norm": 0.10789337009191513, "learning_rate": 1.4764704739531643e-05, "loss": 0.0179, "step": 30460 }, { "grad_norm": 0.08505062013864517, "learning_rate": 1.473538838278658e-05, "loss": 0.016, "step": 30470 }, { "grad_norm": 0.0820886641740799, "learning_rate": 1.4706096129071045e-05, "loss": 0.0163, "step": 30480 }, { "grad_norm": 0.05408959463238716, "learning_rate": 1.4676827998405996e-05, "loss": 0.0173, "step": 30490 }, { "grad_norm": 0.09170646965503693, "learning_rate": 1.4647584010795928e-05, "loss": 0.0158, "step": 30500 }, { "grad_norm": 0.09752674400806427, "learning_rate": 1.4618364186228834e-05, "loss": 0.0161, "step": 30510 }, { "grad_norm": 0.07365855574607849, "learning_rate": 1.4589168544676207e-05, "loss": 0.0157, "step": 30520 }, { "grad_norm": 0.07878319919109344, "learning_rate": 1.455999710609296e-05, "loss": 0.0197, "step": 30530 }, { "grad_norm": 0.07059366255998611, "learning_rate": 1.453084989041753e-05, "loss": 0.0161, "step": 30540 }, { "grad_norm": 0.07981716841459274, "learning_rate": 1.4501726917571728e-05, "loss": 0.0182, "step": 30550 }, { "grad_norm": 0.1306271255016327, "learning_rate": 1.4472628207460848e-05, "loss": 0.0167, "step": 30560 }, { "grad_norm": 0.07739271223545074, "learning_rate": 1.4443553779973584e-05, "loss": 0.0171, "step": 30570 }, { "grad_norm": 0.11068470776081085, "learning_rate": 1.4414503654982037e-05, "loss": 0.0171, "step": 30580 }, { "grad_norm": 0.09258244931697845, "learning_rate": 1.4385477852341706e-05, "loss": 0.0171, "step": 30590 }, { "grad_norm": 0.08227650821208954, "learning_rate": 1.4356476391891421e-05, "loss": 0.0179, "step": 30600 }, { "grad_norm": 0.07777561992406845, "learning_rate": 1.4327499293453423e-05, "loss": 0.0167, "step": 30610 }, { "grad_norm": 0.09912478923797607, "learning_rate": 1.4298546576833304e-05, "loss": 0.0143, "step": 30620 }, { "grad_norm": 0.10292749851942062, "learning_rate": 1.426961826181994e-05, "loss": 0.0151, "step": 30630 }, { "grad_norm": 0.09037256985902786, "learning_rate": 1.4240714368185592e-05, "loss": 0.0122, "step": 30640 }, { "grad_norm": 0.06442704796791077, "learning_rate": 1.4211834915685767e-05, "loss": 0.014, "step": 30650 }, { "grad_norm": 0.11192291975021362, "learning_rate": 1.4182979924059313e-05, "loss": 0.0177, "step": 30660 }, { "grad_norm": 0.12888097763061523, "learning_rate": 1.415414941302835e-05, "loss": 0.0158, "step": 30670 }, { "grad_norm": 0.07329816371202469, "learning_rate": 1.4125343402298257e-05, "loss": 0.0173, "step": 30680 }, { "grad_norm": 0.10014515370130539, "learning_rate": 1.4096561911557688e-05, "loss": 0.0168, "step": 30690 }, { "grad_norm": 0.1084328144788742, "learning_rate": 1.4067804960478481e-05, "loss": 0.0173, "step": 30700 }, { "grad_norm": 0.07246338576078415, "learning_rate": 1.4039072568715784e-05, "loss": 0.0176, "step": 30710 }, { "grad_norm": 0.06907422095537186, "learning_rate": 1.401036475590788e-05, "loss": 0.0167, "step": 30720 }, { "grad_norm": 0.0980636477470398, "learning_rate": 1.3981681541676301e-05, "loss": 0.0193, "step": 30730 }, { "grad_norm": 0.0726781114935875, "learning_rate": 1.3953022945625772e-05, "loss": 0.0166, "step": 30740 }, { "grad_norm": 0.05972006544470787, "learning_rate": 1.3924388987344138e-05, "loss": 0.0165, "step": 30750 }, { "grad_norm": 0.06821546703577042, "learning_rate": 1.3895779686402465e-05, "loss": 0.0146, "step": 30760 }, { "grad_norm": 0.16169309616088867, "learning_rate": 1.3867195062354931e-05, "loss": 0.0197, "step": 30770 }, { "grad_norm": 0.07966200262308121, "learning_rate": 1.3838635134738858e-05, "loss": 0.014, "step": 30780 }, { "grad_norm": 0.06047111749649048, "learning_rate": 1.3810099923074704e-05, "loss": 0.0199, "step": 30790 }, { "grad_norm": 0.11754646897315979, "learning_rate": 1.3781589446866006e-05, "loss": 0.0155, "step": 30800 }, { "grad_norm": 0.16101990640163422, "learning_rate": 1.3753103725599387e-05, "loss": 0.0141, "step": 30810 }, { "grad_norm": 0.07343640178442001, "learning_rate": 1.3724642778744579e-05, "loss": 0.0146, "step": 30820 }, { "grad_norm": 0.09292872995138168, "learning_rate": 1.3696206625754376e-05, "loss": 0.0184, "step": 30830 }, { "grad_norm": 0.07737970352172852, "learning_rate": 1.3667795286064639e-05, "loss": 0.0187, "step": 30840 }, { "grad_norm": 0.06359007209539413, "learning_rate": 1.3639408779094215e-05, "loss": 0.0172, "step": 30850 }, { "grad_norm": 0.07349870353937149, "learning_rate": 1.361104712424503e-05, "loss": 0.0146, "step": 30860 }, { "grad_norm": 0.06335665285587311, "learning_rate": 1.358271034090201e-05, "loss": 0.0181, "step": 30870 }, { "grad_norm": 0.06096653640270233, "learning_rate": 1.3554398448433076e-05, "loss": 0.0173, "step": 30880 }, { "grad_norm": 0.055986277759075165, "learning_rate": 1.3526111466189156e-05, "loss": 0.0147, "step": 30890 }, { "grad_norm": 0.0735747441649437, "learning_rate": 1.3497849413504121e-05, "loss": 0.0157, "step": 30900 }, { "grad_norm": 0.06058596074581146, "learning_rate": 1.3469612309694802e-05, "loss": 0.0157, "step": 30910 }, { "grad_norm": 0.08772020787000656, "learning_rate": 1.3441400174061008e-05, "loss": 0.0139, "step": 30920 }, { "grad_norm": 0.09689807146787643, "learning_rate": 1.3413213025885469e-05, "loss": 0.0152, "step": 30930 }, { "grad_norm": 0.08747471123933792, "learning_rate": 1.3385050884433853e-05, "loss": 0.0141, "step": 30940 }, { "grad_norm": 0.11016636341810226, "learning_rate": 1.3356913768954676e-05, "loss": 0.0197, "step": 30950 }, { "grad_norm": 0.07962475717067719, "learning_rate": 1.3328801698679415e-05, "loss": 0.0175, "step": 30960 }, { "grad_norm": 0.08969565480947495, "learning_rate": 1.3300714692822413e-05, "loss": 0.0174, "step": 30970 }, { "grad_norm": 0.08282685279846191, "learning_rate": 1.3272652770580857e-05, "loss": 0.0171, "step": 30980 }, { "grad_norm": 0.08365872502326965, "learning_rate": 1.3244615951134831e-05, "loss": 0.017, "step": 30990 }, { "grad_norm": 0.25245046615600586, "learning_rate": 1.3216604253647219e-05, "loss": 0.0189, "step": 31000 }, { "grad_norm": 0.06882333010435104, "learning_rate": 1.3188617697263738e-05, "loss": 0.0174, "step": 31010 }, { "grad_norm": 0.09779812395572662, "learning_rate": 1.3160656301112956e-05, "loss": 0.0181, "step": 31020 }, { "grad_norm": 0.11200263351202011, "learning_rate": 1.3132720084306211e-05, "loss": 0.0181, "step": 31030 }, { "grad_norm": 0.09881580621004105, "learning_rate": 1.3104809065937668e-05, "loss": 0.0194, "step": 31040 }, { "grad_norm": 0.0867447629570961, "learning_rate": 1.3076923265084212e-05, "loss": 0.0158, "step": 31050 }, { "grad_norm": 0.07054132968187332, "learning_rate": 1.3049062700805543e-05, "loss": 0.0165, "step": 31060 }, { "grad_norm": 0.07974711805582047, "learning_rate": 1.3021227392144092e-05, "loss": 0.0169, "step": 31070 }, { "grad_norm": 0.09414006024599075, "learning_rate": 1.2993417358125037e-05, "loss": 0.016, "step": 31080 }, { "grad_norm": 0.06523743271827698, "learning_rate": 1.2965632617756257e-05, "loss": 0.0194, "step": 31090 }, { "grad_norm": 0.05773181468248367, "learning_rate": 1.2937873190028387e-05, "loss": 0.0151, "step": 31100 }, { "grad_norm": 0.07707567512989044, "learning_rate": 1.2910139093914697e-05, "loss": 0.0203, "step": 31110 }, { "grad_norm": 0.05155380815267563, "learning_rate": 1.2882430348371205e-05, "loss": 0.0166, "step": 31120 }, { "grad_norm": 0.09480983018875122, "learning_rate": 1.2854746972336567e-05, "loss": 0.0187, "step": 31130 }, { "grad_norm": 0.07868692278862, "learning_rate": 1.2827088984732132e-05, "loss": 0.0181, "step": 31140 }, { "grad_norm": 0.19121669232845306, "learning_rate": 1.279945640446184e-05, "loss": 0.0203, "step": 31150 }, { "grad_norm": 0.07779959589242935, "learning_rate": 1.2771849250412315e-05, "loss": 0.0166, "step": 31160 }, { "grad_norm": 0.06486549973487854, "learning_rate": 1.2744267541452792e-05, "loss": 0.015, "step": 31170 }, { "grad_norm": 0.07027366757392883, "learning_rate": 1.2716711296435112e-05, "loss": 0.0154, "step": 31180 }, { "grad_norm": 0.05738828331232071, "learning_rate": 1.268918053419369e-05, "loss": 0.0156, "step": 31190 }, { "grad_norm": 0.13401968777179718, "learning_rate": 1.2661675273545564e-05, "loss": 0.0188, "step": 31200 }, { "grad_norm": 0.33627426624298096, "learning_rate": 1.26341955332903e-05, "loss": 0.016, "step": 31210 }, { "grad_norm": 0.07812836766242981, "learning_rate": 1.2606741332210048e-05, "loss": 0.02, "step": 31220 }, { "grad_norm": 0.0706629827618599, "learning_rate": 1.2579312689069495e-05, "loss": 0.0135, "step": 31230 }, { "grad_norm": 0.12550494074821472, "learning_rate": 1.2551909622615881e-05, "loss": 0.0176, "step": 31240 }, { "grad_norm": 0.0775209590792656, "learning_rate": 1.252453215157891e-05, "loss": 0.0159, "step": 31250 }, { "grad_norm": 0.10646721720695496, "learning_rate": 1.2497180294670841e-05, "loss": 0.0139, "step": 31260 }, { "grad_norm": 0.07019829750061035, "learning_rate": 1.2469854070586423e-05, "loss": 0.016, "step": 31270 }, { "grad_norm": 0.07308575510978699, "learning_rate": 1.2442553498002845e-05, "loss": 0.0165, "step": 31280 }, { "grad_norm": 0.08490448445081711, "learning_rate": 1.241527859557981e-05, "loss": 0.0164, "step": 31290 }, { "grad_norm": 0.09593791514635086, "learning_rate": 1.238802938195946e-05, "loss": 0.0159, "step": 31300 }, { "grad_norm": 0.08584902435541153, "learning_rate": 1.2360805875766357e-05, "loss": 0.0141, "step": 31310 }, { "grad_norm": 0.09708937257528305, "learning_rate": 1.2333608095607524e-05, "loss": 0.0188, "step": 31320 }, { "grad_norm": 0.06888283044099808, "learning_rate": 1.230643606007238e-05, "loss": 0.0156, "step": 31330 }, { "grad_norm": 0.06150779128074646, "learning_rate": 1.2279289787732778e-05, "loss": 0.015, "step": 31340 }, { "grad_norm": 0.07680978626012802, "learning_rate": 1.2252169297142906e-05, "loss": 0.0135, "step": 31350 }, { "grad_norm": 0.06484723836183548, "learning_rate": 1.2225074606839392e-05, "loss": 0.0172, "step": 31360 }, { "grad_norm": 0.07008729875087738, "learning_rate": 1.2198005735341173e-05, "loss": 0.0155, "step": 31370 }, { "grad_norm": 0.12674300372600555, "learning_rate": 1.2170962701149584e-05, "loss": 0.0155, "step": 31380 }, { "grad_norm": 0.07783414423465729, "learning_rate": 1.2143945522748285e-05, "loss": 0.0154, "step": 31390 }, { "grad_norm": 0.057101428508758545, "learning_rate": 1.2116954218603266e-05, "loss": 0.0153, "step": 31400 }, { "grad_norm": 0.06376366317272186, "learning_rate": 1.2089988807162817e-05, "loss": 0.0184, "step": 31410 }, { "grad_norm": 0.0714743360877037, "learning_rate": 1.2063049306857543e-05, "loss": 0.0148, "step": 31420 }, { "grad_norm": 0.0768466591835022, "learning_rate": 1.2036135736100346e-05, "loss": 0.0144, "step": 31430 }, { "grad_norm": 0.08964217454195023, "learning_rate": 1.200924811328641e-05, "loss": 0.0158, "step": 31440 }, { "grad_norm": 0.10200809687376022, "learning_rate": 1.1982386456793143e-05, "loss": 0.0182, "step": 31450 }, { "grad_norm": 0.07832469046115875, "learning_rate": 1.1955550784980257e-05, "loss": 0.015, "step": 31460 }, { "grad_norm": 0.06772113591432571, "learning_rate": 1.1928741116189663e-05, "loss": 0.0166, "step": 31470 }, { "grad_norm": 0.0773315504193306, "learning_rate": 1.1901957468745529e-05, "loss": 0.0138, "step": 31480 }, { "grad_norm": 0.0829879641532898, "learning_rate": 1.1875199860954217e-05, "loss": 0.0174, "step": 31490 }, { "grad_norm": 0.07544033974409103, "learning_rate": 1.1848468311104316e-05, "loss": 0.0177, "step": 31500 }, { "grad_norm": 0.06766717880964279, "learning_rate": 1.1821762837466566e-05, "loss": 0.0165, "step": 31510 }, { "grad_norm": 0.10441303998231888, "learning_rate": 1.179508345829392e-05, "loss": 0.0146, "step": 31520 }, { "grad_norm": 0.07636545598506927, "learning_rate": 1.1768430191821478e-05, "loss": 0.0183, "step": 31530 }, { "grad_norm": 0.08262842893600464, "learning_rate": 1.174180305626651e-05, "loss": 0.0164, "step": 31540 }, { "grad_norm": 0.08679072558879852, "learning_rate": 1.1715202069828401e-05, "loss": 0.0148, "step": 31550 }, { "grad_norm": 0.08249740302562714, "learning_rate": 1.1688627250688661e-05, "loss": 0.0186, "step": 31560 }, { "grad_norm": 0.0963330864906311, "learning_rate": 1.1662078617010935e-05, "loss": 0.0212, "step": 31570 }, { "grad_norm": 0.04941009730100632, "learning_rate": 1.1635556186940972e-05, "loss": 0.0167, "step": 31580 }, { "grad_norm": 0.07122586667537689, "learning_rate": 1.1609059978606602e-05, "loss": 0.0157, "step": 31590 }, { "grad_norm": 0.1438826471567154, "learning_rate": 1.1582590010117745e-05, "loss": 0.0218, "step": 31600 }, { "grad_norm": 0.07467549294233322, "learning_rate": 1.1556146299566345e-05, "loss": 0.0187, "step": 31610 }, { "grad_norm": 0.06364589184522629, "learning_rate": 1.1529728865026452e-05, "loss": 0.015, "step": 31620 }, { "grad_norm": 0.06380472332239151, "learning_rate": 1.150333772455412e-05, "loss": 0.0163, "step": 31630 }, { "grad_norm": 0.09053540974855423, "learning_rate": 1.1476972896187466e-05, "loss": 0.0181, "step": 31640 }, { "grad_norm": 0.07481657713651657, "learning_rate": 1.1450634397946592e-05, "loss": 0.0171, "step": 31650 }, { "grad_norm": 0.07612114399671555, "learning_rate": 1.142432224783359e-05, "loss": 0.0173, "step": 31660 }, { "grad_norm": 0.09006745368242264, "learning_rate": 1.1398036463832584e-05, "loss": 0.0188, "step": 31670 }, { "grad_norm": 0.09034028649330139, "learning_rate": 1.1371777063909661e-05, "loss": 0.0144, "step": 31680 }, { "grad_norm": 0.08868905901908875, "learning_rate": 1.1345544066012874e-05, "loss": 0.0228, "step": 31690 }, { "grad_norm": 0.10333284735679626, "learning_rate": 1.131933748807224e-05, "loss": 0.0162, "step": 31700 }, { "grad_norm": 0.08590671420097351, "learning_rate": 1.1293157347999683e-05, "loss": 0.0191, "step": 31710 }, { "grad_norm": 0.09203045815229416, "learning_rate": 1.1267003663689097e-05, "loss": 0.0204, "step": 31720 }, { "grad_norm": 0.09996543824672699, "learning_rate": 1.1240876453016275e-05, "loss": 0.0181, "step": 31730 }, { "grad_norm": 0.16120705008506775, "learning_rate": 1.1214775733838933e-05, "loss": 0.0178, "step": 31740 }, { "grad_norm": 0.08486174046993256, "learning_rate": 1.118870152399666e-05, "loss": 0.0159, "step": 31750 }, { "grad_norm": 0.08543882519006729, "learning_rate": 1.1162653841310905e-05, "loss": 0.0178, "step": 31760 }, { "grad_norm": 0.06459548324346542, "learning_rate": 1.1136632703585043e-05, "loss": 0.017, "step": 31770 }, { "grad_norm": 0.08120886981487274, "learning_rate": 1.111063812860426e-05, "loss": 0.0212, "step": 31780 }, { "grad_norm": 0.15060579776763916, "learning_rate": 1.1084670134135616e-05, "loss": 0.0153, "step": 31790 }, { "grad_norm": 0.08711239695549011, "learning_rate": 1.1058728737927998e-05, "loss": 0.0171, "step": 31800 }, { "grad_norm": 0.06773420423269272, "learning_rate": 1.1032813957712086e-05, "loss": 0.0139, "step": 31810 }, { "grad_norm": 0.07469040900468826, "learning_rate": 1.1006925811200385e-05, "loss": 0.0192, "step": 31820 }, { "grad_norm": 0.15489600598812103, "learning_rate": 1.0981064316087231e-05, "loss": 0.0169, "step": 31830 }, { "grad_norm": 0.11618471890687943, "learning_rate": 1.0955229490048675e-05, "loss": 0.0167, "step": 31840 }, { "grad_norm": 0.060923751443624496, "learning_rate": 1.0929421350742603e-05, "loss": 0.017, "step": 31850 }, { "grad_norm": 0.104413241147995, "learning_rate": 1.090363991580861e-05, "loss": 0.0172, "step": 31860 }, { "grad_norm": 0.17276309430599213, "learning_rate": 1.0877885202868083e-05, "loss": 0.0164, "step": 31870 }, { "grad_norm": 0.10437876731157303, "learning_rate": 1.0852157229524118e-05, "loss": 0.0158, "step": 31880 }, { "grad_norm": 0.06835158914327621, "learning_rate": 1.0826456013361542e-05, "loss": 0.0155, "step": 31890 }, { "grad_norm": 0.10064814239740372, "learning_rate": 1.080078157194691e-05, "loss": 0.0195, "step": 31900 }, { "grad_norm": 0.06389200687408447, "learning_rate": 1.077513392282844e-05, "loss": 0.015, "step": 31910 }, { "grad_norm": 0.09597163647413254, "learning_rate": 1.0749513083536073e-05, "loss": 0.016, "step": 31920 }, { "grad_norm": 0.07733482867479324, "learning_rate": 1.072391907158139e-05, "loss": 0.0158, "step": 31930 }, { "grad_norm": 0.06783926486968994, "learning_rate": 1.0698351904457676e-05, "loss": 0.0192, "step": 31940 }, { "grad_norm": 0.07543597370386124, "learning_rate": 1.0672811599639849e-05, "loss": 0.0148, "step": 31950 }, { "grad_norm": 0.11921126395463943, "learning_rate": 1.0647298174584453e-05, "loss": 0.0215, "step": 31960 }, { "grad_norm": 0.06371691823005676, "learning_rate": 1.0621811646729673e-05, "loss": 0.0144, "step": 31970 }, { "grad_norm": 0.07542401552200317, "learning_rate": 1.0596352033495328e-05, "loss": 0.0182, "step": 31980 }, { "grad_norm": 0.3758288621902466, "learning_rate": 1.057091935228281e-05, "loss": 0.0148, "step": 31990 }, { "grad_norm": 0.12041878700256348, "learning_rate": 1.0545513620475134e-05, "loss": 0.0167, "step": 32000 }, { "grad_norm": 0.0868888720870018, "learning_rate": 1.0520134855436863e-05, "loss": 0.0145, "step": 32010 }, { "grad_norm": 0.0900222435593605, "learning_rate": 1.049478307451413e-05, "loss": 0.0148, "step": 32020 }, { "grad_norm": 0.14782634377479553, "learning_rate": 1.0469458295034656e-05, "loss": 0.0169, "step": 32030 }, { "grad_norm": 0.0749107226729393, "learning_rate": 1.044416053430769e-05, "loss": 0.0149, "step": 32040 }, { "grad_norm": 0.10968632996082306, "learning_rate": 1.0418889809624021e-05, "loss": 0.0127, "step": 32050 }, { "grad_norm": 0.0985405445098877, "learning_rate": 1.0393646138255925e-05, "loss": 0.0156, "step": 32060 }, { "grad_norm": 0.0948045626282692, "learning_rate": 1.036842953745723e-05, "loss": 0.0133, "step": 32070 }, { "grad_norm": 0.10577858239412308, "learning_rate": 1.034324002446324e-05, "loss": 0.0172, "step": 32080 }, { "grad_norm": 0.053909264504909515, "learning_rate": 1.0318077616490751e-05, "loss": 0.0158, "step": 32090 }, { "grad_norm": 0.0737570971250534, "learning_rate": 1.0292942330738054e-05, "loss": 0.0143, "step": 32100 }, { "grad_norm": 0.1264968067407608, "learning_rate": 1.026783418438485e-05, "loss": 0.0151, "step": 32110 }, { "grad_norm": 0.07564615458250046, "learning_rate": 1.0242753194592319e-05, "loss": 0.0171, "step": 32120 }, { "grad_norm": 0.04428642615675926, "learning_rate": 1.0217699378503092e-05, "loss": 0.0128, "step": 32130 }, { "grad_norm": 0.0996691957116127, "learning_rate": 1.0192672753241218e-05, "loss": 0.0172, "step": 32140 }, { "grad_norm": 0.07969129830598831, "learning_rate": 1.0167673335912164e-05, "loss": 0.0143, "step": 32150 }, { "grad_norm": 0.1157047301530838, "learning_rate": 1.0142701143602779e-05, "loss": 0.0168, "step": 32160 }, { "grad_norm": 0.09652918577194214, "learning_rate": 1.0117756193381328e-05, "loss": 0.0171, "step": 32170 }, { "grad_norm": 0.07725165039300919, "learning_rate": 1.0092838502297453e-05, "loss": 0.0171, "step": 32180 }, { "grad_norm": 0.059604350477457047, "learning_rate": 1.0067948087382162e-05, "loss": 0.0121, "step": 32190 }, { "grad_norm": 0.09027843177318573, "learning_rate": 1.0043084965647826e-05, "loss": 0.0134, "step": 32200 }, { "grad_norm": 0.08302181214094162, "learning_rate": 1.001824915408815e-05, "loss": 0.0145, "step": 32210 }, { "grad_norm": 0.08028966188430786, "learning_rate": 9.99344066967815e-06, "loss": 0.0148, "step": 32220 }, { "grad_norm": 0.1723155528306961, "learning_rate": 9.968659529374219e-06, "loss": 0.0157, "step": 32230 }, { "grad_norm": 0.09955475479364395, "learning_rate": 9.943905750114025e-06, "loss": 0.0142, "step": 32240 }, { "grad_norm": 0.15078873932361603, "learning_rate": 9.919179348816559e-06, "loss": 0.0163, "step": 32250 }, { "grad_norm": 0.08396682888269424, "learning_rate": 9.894480342382056e-06, "loss": 0.0138, "step": 32260 }, { "grad_norm": 0.18009990453720093, "learning_rate": 9.869808747692061e-06, "loss": 0.0153, "step": 32270 }, { "grad_norm": 0.10029143840074539, "learning_rate": 9.84516458160939e-06, "loss": 0.0169, "step": 32280 }, { "grad_norm": 0.11044485867023468, "learning_rate": 9.820547860978097e-06, "loss": 0.0183, "step": 32290 }, { "grad_norm": 0.10344855487346649, "learning_rate": 9.795958602623461e-06, "loss": 0.0141, "step": 32300 }, { "grad_norm": 0.14304028451442719, "learning_rate": 9.77139682335203e-06, "loss": 0.0173, "step": 32310 }, { "grad_norm": 0.07676825672388077, "learning_rate": 9.74686253995153e-06, "loss": 0.0156, "step": 32320 }, { "grad_norm": 0.0954967811703682, "learning_rate": 9.722355769190911e-06, "loss": 0.0151, "step": 32330 }, { "grad_norm": 0.0750376507639885, "learning_rate": 9.697876527820333e-06, "loss": 0.0165, "step": 32340 }, { "grad_norm": 0.21070659160614014, "learning_rate": 9.673424832571132e-06, "loss": 0.0159, "step": 32350 }, { "grad_norm": 0.09503341466188431, "learning_rate": 9.649000700155791e-06, "loss": 0.017, "step": 32360 }, { "grad_norm": 0.13319508731365204, "learning_rate": 9.624604147267984e-06, "loss": 0.0189, "step": 32370 }, { "grad_norm": 0.08402855694293976, "learning_rate": 9.600235190582524e-06, "loss": 0.016, "step": 32380 }, { "grad_norm": 0.08041312545537949, "learning_rate": 9.57589384675538e-06, "loss": 0.0181, "step": 32390 }, { "grad_norm": 0.07107436656951904, "learning_rate": 9.551580132423598e-06, "loss": 0.0166, "step": 32400 }, { "grad_norm": 0.14663009345531464, "learning_rate": 9.527294064205405e-06, "loss": 0.0163, "step": 32410 }, { "grad_norm": 0.09864160418510437, "learning_rate": 9.503035658700072e-06, "loss": 0.0167, "step": 32420 }, { "grad_norm": 0.14163042604923248, "learning_rate": 9.478804932488e-06, "loss": 0.0165, "step": 32430 }, { "grad_norm": 0.07094606757164001, "learning_rate": 9.454601902130667e-06, "loss": 0.0185, "step": 32440 }, { "grad_norm": 0.08886317163705826, "learning_rate": 9.430426584170626e-06, "loss": 0.0141, "step": 32450 }, { "grad_norm": 0.08865290880203247, "learning_rate": 9.40627899513145e-06, "loss": 0.0162, "step": 32460 }, { "grad_norm": 0.10393785685300827, "learning_rate": 9.38215915151781e-06, "loss": 0.0159, "step": 32470 }, { "grad_norm": 0.13281621038913727, "learning_rate": 9.358067069815396e-06, "loss": 0.0124, "step": 32480 }, { "grad_norm": 0.11304271966218948, "learning_rate": 9.334002766490901e-06, "loss": 0.0175, "step": 32490 }, { "grad_norm": 0.10865657776594162, "learning_rate": 9.309966257992054e-06, "loss": 0.0163, "step": 32500 }, { "grad_norm": 0.08490326255559921, "learning_rate": 9.2859575607476e-06, "loss": 0.0179, "step": 32510 }, { "grad_norm": 0.1302524209022522, "learning_rate": 9.26197669116723e-06, "loss": 0.0162, "step": 32520 }, { "grad_norm": 0.07771990448236465, "learning_rate": 9.238023665641655e-06, "loss": 0.015, "step": 32530 }, { "grad_norm": 0.08756228536367416, "learning_rate": 9.21409850054254e-06, "loss": 0.0185, "step": 32540 }, { "grad_norm": 0.12142754346132278, "learning_rate": 9.190201212222526e-06, "loss": 0.0153, "step": 32550 }, { "grad_norm": 0.23567336797714233, "learning_rate": 9.166331817015155e-06, "loss": 0.0162, "step": 32560 }, { "grad_norm": 0.08590187132358551, "learning_rate": 9.142490331234965e-06, "loss": 0.018, "step": 32570 }, { "grad_norm": 0.07382259517908096, "learning_rate": 9.11867677117736e-06, "loss": 0.0207, "step": 32580 }, { "grad_norm": 0.5816985368728638, "learning_rate": 9.09489115311869e-06, "loss": 0.0163, "step": 32590 }, { "grad_norm": 0.16971006989479065, "learning_rate": 9.071133493316202e-06, "loss": 0.0184, "step": 32600 }, { "grad_norm": 0.07402152568101883, "learning_rate": 9.047403808008053e-06, "loss": 0.0145, "step": 32610 }, { "grad_norm": 0.11283605545759201, "learning_rate": 9.023702113413224e-06, "loss": 0.0149, "step": 32620 }, { "grad_norm": 0.09208182245492935, "learning_rate": 9.00002842573162e-06, "loss": 0.0145, "step": 32630 }, { "grad_norm": 0.5920184254646301, "learning_rate": 8.976382761143981e-06, "loss": 0.0156, "step": 32640 }, { "grad_norm": 0.06997538357973099, "learning_rate": 8.952765135811909e-06, "loss": 0.0161, "step": 32650 }, { "grad_norm": 0.06067655608057976, "learning_rate": 8.929175565877795e-06, "loss": 0.0134, "step": 32660 }, { "grad_norm": 0.09188032150268555, "learning_rate": 8.905614067464924e-06, "loss": 0.0178, "step": 32670 }, { "grad_norm": 0.07650045305490494, "learning_rate": 8.882080656677333e-06, "loss": 0.0162, "step": 32680 }, { "grad_norm": 0.203887939453125, "learning_rate": 8.85857534959989e-06, "loss": 0.0136, "step": 32690 }, { "grad_norm": 0.12372589856386185, "learning_rate": 8.835098162298256e-06, "loss": 0.0158, "step": 32700 }, { "grad_norm": 0.17514240741729736, "learning_rate": 8.811649110818875e-06, "loss": 0.0179, "step": 32710 }, { "grad_norm": 0.12667877972126007, "learning_rate": 8.78822821118893e-06, "loss": 0.0157, "step": 32720 }, { "grad_norm": 0.10102947056293488, "learning_rate": 8.764835479416405e-06, "loss": 0.0134, "step": 32730 }, { "grad_norm": 0.066282257437706, "learning_rate": 8.741470931490003e-06, "loss": 0.0151, "step": 32740 }, { "grad_norm": 0.09285694360733032, "learning_rate": 8.718134583379189e-06, "loss": 0.0154, "step": 32750 }, { "grad_norm": 0.08258606493473053, "learning_rate": 8.69482645103412e-06, "loss": 0.0163, "step": 32760 }, { "grad_norm": 0.21296454966068268, "learning_rate": 8.671546550385679e-06, "loss": 0.0183, "step": 32770 }, { "grad_norm": 0.10403003543615341, "learning_rate": 8.648294897345466e-06, "loss": 0.0169, "step": 32780 }, { "grad_norm": 0.05699194222688675, "learning_rate": 8.62507150780576e-06, "loss": 0.014, "step": 32790 }, { "grad_norm": 0.06847365200519562, "learning_rate": 8.601876397639542e-06, "loss": 0.0165, "step": 32800 }, { "grad_norm": 0.30557113885879517, "learning_rate": 8.578709582700456e-06, "loss": 0.0192, "step": 32810 }, { "grad_norm": 0.078569196164608, "learning_rate": 8.555571078822783e-06, "loss": 0.0146, "step": 32820 }, { "grad_norm": 0.07130930572748184, "learning_rate": 8.532460901821477e-06, "loss": 0.016, "step": 32830 }, { "grad_norm": 0.0669824555516243, "learning_rate": 8.509379067492128e-06, "loss": 0.0173, "step": 32840 }, { "grad_norm": 0.0640517845749855, "learning_rate": 8.486325591610967e-06, "loss": 0.0184, "step": 32850 }, { "grad_norm": 0.17321401834487915, "learning_rate": 8.463300489934811e-06, "loss": 0.0167, "step": 32860 }, { "grad_norm": 0.0881185382604599, "learning_rate": 8.440303778201093e-06, "loss": 0.0152, "step": 32870 }, { "grad_norm": 0.05230974778532982, "learning_rate": 8.417335472127858e-06, "loss": 0.0155, "step": 32880 }, { "grad_norm": 0.14017577469348907, "learning_rate": 8.394395587413722e-06, "loss": 0.0166, "step": 32890 }, { "grad_norm": 0.10869187861680984, "learning_rate": 8.371484139737889e-06, "loss": 0.0172, "step": 32900 }, { "grad_norm": 0.063339464366436, "learning_rate": 8.348601144760127e-06, "loss": 0.0151, "step": 32910 }, { "grad_norm": 0.1065795049071312, "learning_rate": 8.325746618120716e-06, "loss": 0.015, "step": 32920 }, { "grad_norm": 0.209811270236969, "learning_rate": 8.302920575440531e-06, "loss": 0.0146, "step": 32930 }, { "grad_norm": 0.08055320382118225, "learning_rate": 8.280123032320953e-06, "loss": 0.0173, "step": 32940 }, { "grad_norm": 0.05187869444489479, "learning_rate": 8.257354004343892e-06, "loss": 0.0189, "step": 32950 }, { "grad_norm": 0.0835813507437706, "learning_rate": 8.234613507071754e-06, "loss": 0.0168, "step": 32960 }, { "grad_norm": 0.09621310979127884, "learning_rate": 8.211901556047475e-06, "loss": 0.0188, "step": 32970 }, { "grad_norm": 0.09627998620271683, "learning_rate": 8.189218166794427e-06, "loss": 0.0183, "step": 32980 }, { "grad_norm": 0.11168186366558075, "learning_rate": 8.166563354816509e-06, "loss": 0.0132, "step": 32990 }, { "grad_norm": 0.08829145133495331, "learning_rate": 8.143937135598074e-06, "loss": 0.0156, "step": 33000 }, { "grad_norm": 0.15026435256004333, "learning_rate": 8.121339524603943e-06, "loss": 0.0163, "step": 33010 }, { "grad_norm": 0.06982214003801346, "learning_rate": 8.098770537279338e-06, "loss": 0.0145, "step": 33020 }, { "grad_norm": 0.0818316861987114, "learning_rate": 8.076230189049966e-06, "loss": 0.0158, "step": 33030 }, { "grad_norm": 0.053828801959753036, "learning_rate": 8.05371849532196e-06, "loss": 0.0153, "step": 33040 }, { "grad_norm": 0.07820260524749756, "learning_rate": 8.031235471481813e-06, "loss": 0.0148, "step": 33050 }, { "grad_norm": 0.07903379201889038, "learning_rate": 8.008781132896486e-06, "loss": 0.015, "step": 33060 }, { "grad_norm": 0.07854393869638443, "learning_rate": 7.986355494913312e-06, "loss": 0.0137, "step": 33070 }, { "grad_norm": 0.05082054063677788, "learning_rate": 7.963958572859975e-06, "loss": 0.0136, "step": 33080 }, { "grad_norm": 0.10329494625329971, "learning_rate": 7.941590382044583e-06, "loss": 0.0146, "step": 33090 }, { "grad_norm": 0.05325564369559288, "learning_rate": 7.919250937755574e-06, "loss": 0.0165, "step": 33100 }, { "grad_norm": 0.09366662800312042, "learning_rate": 7.896940255261764e-06, "loss": 0.0152, "step": 33110 }, { "grad_norm": 0.07013720273971558, "learning_rate": 7.874658349812269e-06, "loss": 0.0135, "step": 33120 }, { "grad_norm": 0.09655213356018066, "learning_rate": 7.85240523663659e-06, "loss": 0.0136, "step": 33130 }, { "grad_norm": 0.060472335666418076, "learning_rate": 7.830180930944487e-06, "loss": 0.0164, "step": 33140 }, { "grad_norm": 0.07384414225816727, "learning_rate": 7.80798544792608e-06, "loss": 0.0199, "step": 33150 }, { "grad_norm": 0.07782459259033203, "learning_rate": 7.785818802751776e-06, "loss": 0.0149, "step": 33160 }, { "grad_norm": 0.15355323255062103, "learning_rate": 7.763681010572277e-06, "loss": 0.0171, "step": 33170 }, { "grad_norm": 0.06930548697710037, "learning_rate": 7.741572086518528e-06, "loss": 0.0144, "step": 33180 }, { "grad_norm": 0.132571741938591, "learning_rate": 7.719492045701782e-06, "loss": 0.0174, "step": 33190 }, { "grad_norm": 0.07092025876045227, "learning_rate": 7.697440903213549e-06, "loss": 0.0181, "step": 33200 }, { "grad_norm": 0.0869368314743042, "learning_rate": 7.675418674125573e-06, "loss": 0.0153, "step": 33210 }, { "grad_norm": 0.09438933432102203, "learning_rate": 7.65342537348983e-06, "loss": 0.0139, "step": 33220 }, { "grad_norm": 0.09667079895734787, "learning_rate": 7.631461016338549e-06, "loss": 0.0159, "step": 33230 }, { "grad_norm": 0.07309287786483765, "learning_rate": 7.609525617684138e-06, "loss": 0.0147, "step": 33240 }, { "grad_norm": 0.09034457057714462, "learning_rate": 7.587619192519252e-06, "loss": 0.0148, "step": 33250 }, { "grad_norm": 0.051386184990406036, "learning_rate": 7.565741755816725e-06, "loss": 0.0133, "step": 33260 }, { "grad_norm": 0.0726565271615982, "learning_rate": 7.543893322529588e-06, "loss": 0.0181, "step": 33270 }, { "grad_norm": 0.08254052698612213, "learning_rate": 7.522073907591015e-06, "loss": 0.0167, "step": 33280 }, { "grad_norm": 0.06924402713775635, "learning_rate": 7.500283525914381e-06, "loss": 0.0116, "step": 33290 }, { "grad_norm": 0.07282949984073639, "learning_rate": 7.478522192393211e-06, "loss": 0.0148, "step": 33300 }, { "grad_norm": 0.09355496615171432, "learning_rate": 7.456789921901181e-06, "loss": 0.0131, "step": 33310 }, { "grad_norm": 0.09306053817272186, "learning_rate": 7.435086729292079e-06, "loss": 0.0157, "step": 33320 }, { "grad_norm": 0.08975569903850555, "learning_rate": 7.413412629399824e-06, "loss": 0.0162, "step": 33330 }, { "grad_norm": 0.06385814398527145, "learning_rate": 7.391767637038466e-06, "loss": 0.0145, "step": 33340 }, { "grad_norm": 0.061363574117422104, "learning_rate": 7.370151767002159e-06, "loss": 0.0135, "step": 33350 }, { "grad_norm": 0.058333996683359146, "learning_rate": 7.348565034065147e-06, "loss": 0.0149, "step": 33360 }, { "grad_norm": 0.11023051291704178, "learning_rate": 7.327007452981765e-06, "loss": 0.0188, "step": 33370 }, { "grad_norm": 0.06789059937000275, "learning_rate": 7.305479038486396e-06, "loss": 0.0141, "step": 33380 }, { "grad_norm": 0.13067293167114258, "learning_rate": 7.283979805293517e-06, "loss": 0.0152, "step": 33390 }, { "grad_norm": 0.11768531799316406, "learning_rate": 7.262509768097653e-06, "loss": 0.0152, "step": 33400 }, { "grad_norm": 0.07938476651906967, "learning_rate": 7.241068941573381e-06, "loss": 0.0132, "step": 33410 }, { "grad_norm": 0.09373810142278671, "learning_rate": 7.219657340375296e-06, "loss": 0.0137, "step": 33420 }, { "grad_norm": 0.07017093151807785, "learning_rate": 7.198274979138014e-06, "loss": 0.0154, "step": 33430 }, { "grad_norm": 0.07360205799341202, "learning_rate": 7.176921872476189e-06, "loss": 0.0136, "step": 33440 }, { "grad_norm": 0.08510113507509232, "learning_rate": 7.1555980349844674e-06, "loss": 0.0117, "step": 33450 }, { "grad_norm": 0.0861692801117897, "learning_rate": 7.134303481237492e-06, "loss": 0.0162, "step": 33460 }, { "grad_norm": 0.08449837565422058, "learning_rate": 7.113038225789892e-06, "loss": 0.0138, "step": 33470 }, { "grad_norm": 0.10415560007095337, "learning_rate": 7.091802283176258e-06, "loss": 0.0155, "step": 33480 }, { "grad_norm": 0.05953088030219078, "learning_rate": 7.070595667911162e-06, "loss": 0.0157, "step": 33490 }, { "grad_norm": 0.0906105563044548, "learning_rate": 7.049418394489127e-06, "loss": 0.0179, "step": 33500 }, { "grad_norm": 0.11518794298171997, "learning_rate": 7.028270477384636e-06, "loss": 0.0176, "step": 33510 }, { "grad_norm": 0.10015770047903061, "learning_rate": 7.0071519310520695e-06, "loss": 0.0155, "step": 33520 }, { "grad_norm": 0.08410811424255371, "learning_rate": 6.98606276992575e-06, "loss": 0.0161, "step": 33530 }, { "grad_norm": 0.08006531000137329, "learning_rate": 6.965003008419929e-06, "loss": 0.0163, "step": 33540 }, { "grad_norm": 0.15182246267795563, "learning_rate": 6.943972660928755e-06, "loss": 0.0149, "step": 33550 }, { "grad_norm": 0.05402062460780144, "learning_rate": 6.922971741826273e-06, "loss": 0.0131, "step": 33560 }, { "grad_norm": 0.18986530601978302, "learning_rate": 6.902000265466424e-06, "loss": 0.0128, "step": 33570 }, { "grad_norm": 0.06650877743959427, "learning_rate": 6.881058246182998e-06, "loss": 0.0155, "step": 33580 }, { "grad_norm": 0.08447739481925964, "learning_rate": 6.860145698289666e-06, "loss": 0.0162, "step": 33590 }, { "grad_norm": 0.08322697132825851, "learning_rate": 6.839262636079985e-06, "loss": 0.0154, "step": 33600 }, { "grad_norm": 0.07042117416858673, "learning_rate": 6.818409073827292e-06, "loss": 0.0143, "step": 33610 }, { "grad_norm": 0.08221742510795593, "learning_rate": 6.79758502578483e-06, "loss": 0.0173, "step": 33620 }, { "grad_norm": 0.14910171926021576, "learning_rate": 6.776790506185621e-06, "loss": 0.0143, "step": 33630 }, { "grad_norm": 0.19269083440303802, "learning_rate": 6.756025529242533e-06, "loss": 0.0165, "step": 33640 }, { "grad_norm": 0.09828942269086838, "learning_rate": 6.735290109148223e-06, "loss": 0.0159, "step": 33650 }, { "grad_norm": 0.09496984630823135, "learning_rate": 6.714584260075163e-06, "loss": 0.0144, "step": 33660 }, { "grad_norm": 0.0896792858839035, "learning_rate": 6.693907996175619e-06, "loss": 0.0148, "step": 33670 }, { "grad_norm": 0.10401037335395813, "learning_rate": 6.673261331581593e-06, "loss": 0.0133, "step": 33680 }, { "grad_norm": 0.07708410173654556, "learning_rate": 6.652644280404912e-06, "loss": 0.0152, "step": 33690 }, { "grad_norm": 0.09095858782529831, "learning_rate": 6.6320568567371175e-06, "loss": 0.014, "step": 33700 }, { "grad_norm": 0.0612591952085495, "learning_rate": 6.611499074649518e-06, "loss": 0.0155, "step": 33710 }, { "grad_norm": 0.048915471881628036, "learning_rate": 6.590970948193187e-06, "loss": 0.0134, "step": 33720 }, { "grad_norm": 0.06701546907424927, "learning_rate": 6.5704724913988825e-06, "loss": 0.0121, "step": 33730 }, { "grad_norm": 0.14496572315692902, "learning_rate": 6.550003718277109e-06, "loss": 0.0182, "step": 33740 }, { "grad_norm": 0.08336472511291504, "learning_rate": 6.529564642818087e-06, "loss": 0.0143, "step": 33750 }, { "grad_norm": 0.09867934137582779, "learning_rate": 6.50915527899173e-06, "loss": 0.0151, "step": 33760 }, { "grad_norm": 0.24515050649642944, "learning_rate": 6.4887756407476575e-06, "loss": 0.014, "step": 33770 }, { "grad_norm": 0.052812669426202774, "learning_rate": 6.468425742015149e-06, "loss": 0.0134, "step": 33780 }, { "grad_norm": 0.07745988667011261, "learning_rate": 6.448105596703152e-06, "loss": 0.0121, "step": 33790 }, { "grad_norm": 0.15470531582832336, "learning_rate": 6.427815218700317e-06, "loss": 0.0196, "step": 33800 }, { "grad_norm": 0.0608149953186512, "learning_rate": 6.407554621874911e-06, "loss": 0.0141, "step": 33810 }, { "grad_norm": 0.10252820700407028, "learning_rate": 6.3873238200748864e-06, "loss": 0.015, "step": 33820 }, { "grad_norm": 0.07058625668287277, "learning_rate": 6.367122827127775e-06, "loss": 0.0134, "step": 33830 }, { "grad_norm": 0.05860619246959686, "learning_rate": 6.346951656840777e-06, "loss": 0.0164, "step": 33840 }, { "grad_norm": 0.14384585618972778, "learning_rate": 6.326810323000698e-06, "loss": 0.0147, "step": 33850 }, { "grad_norm": 0.0966719463467598, "learning_rate": 6.306698839373948e-06, "loss": 0.0139, "step": 33860 }, { "grad_norm": 0.08403157442808151, "learning_rate": 6.286617219706547e-06, "loss": 0.0123, "step": 33870 }, { "grad_norm": 0.14680394530296326, "learning_rate": 6.26656547772409e-06, "loss": 0.0147, "step": 33880 }, { "grad_norm": 0.11805108189582825, "learning_rate": 6.246543627131734e-06, "loss": 0.0133, "step": 33890 }, { "grad_norm": 0.11271669715642929, "learning_rate": 6.22655168161424e-06, "loss": 0.0163, "step": 33900 }, { "grad_norm": 0.1283307522535324, "learning_rate": 6.20658965483591e-06, "loss": 0.0136, "step": 33910 }, { "grad_norm": 0.08644664287567139, "learning_rate": 6.186657560440612e-06, "loss": 0.014, "step": 33920 }, { "grad_norm": 0.20954641699790955, "learning_rate": 6.166755412051734e-06, "loss": 0.0127, "step": 33930 }, { "grad_norm": 0.12600791454315186, "learning_rate": 6.1468832232721985e-06, "loss": 0.0135, "step": 33940 }, { "grad_norm": 0.08244306594133377, "learning_rate": 6.127041007684476e-06, "loss": 0.0169, "step": 33950 }, { "grad_norm": 0.0833250880241394, "learning_rate": 6.107228778850527e-06, "loss": 0.0139, "step": 33960 }, { "grad_norm": 0.19294244050979614, "learning_rate": 6.087446550311837e-06, "loss": 0.0187, "step": 33970 }, { "grad_norm": 0.08648063987493515, "learning_rate": 6.067694335589353e-06, "loss": 0.0155, "step": 33980 }, { "grad_norm": 0.09735210984945297, "learning_rate": 6.047972148183528e-06, "loss": 0.0133, "step": 33990 }, { "grad_norm": 0.23860976099967957, "learning_rate": 6.0282800015742936e-06, "loss": 0.0182, "step": 34000 }, { "grad_norm": 0.08890065550804138, "learning_rate": 6.008617909221054e-06, "loss": 0.0157, "step": 34010 }, { "grad_norm": 0.08399511873722076, "learning_rate": 5.988985884562664e-06, "loss": 0.0164, "step": 34020 }, { "grad_norm": 0.1870039850473404, "learning_rate": 5.969383941017414e-06, "loss": 0.0177, "step": 34030 }, { "grad_norm": 0.09295075386762619, "learning_rate": 5.949812091983059e-06, "loss": 0.0153, "step": 34040 }, { "grad_norm": 0.22869336605072021, "learning_rate": 5.9302703508367695e-06, "loss": 0.0134, "step": 34050 }, { "grad_norm": 0.10520032048225403, "learning_rate": 5.910758730935157e-06, "loss": 0.014, "step": 34060 }, { "grad_norm": 0.22250871360301971, "learning_rate": 5.8912772456142016e-06, "loss": 0.0169, "step": 34070 }, { "grad_norm": 0.1047673374414444, "learning_rate": 5.871825908189338e-06, "loss": 0.014, "step": 34080 }, { "grad_norm": 0.06538059562444687, "learning_rate": 5.852404731955352e-06, "loss": 0.0143, "step": 34090 }, { "grad_norm": 0.06616228073835373, "learning_rate": 5.833013730186443e-06, "loss": 0.0136, "step": 34100 }, { "grad_norm": 0.15465764701366425, "learning_rate": 5.813652916136181e-06, "loss": 0.0165, "step": 34110 }, { "grad_norm": 0.07127337902784348, "learning_rate": 5.794322303037508e-06, "loss": 0.0174, "step": 34120 }, { "grad_norm": 0.08368097990751266, "learning_rate": 5.775021904102701e-06, "loss": 0.0123, "step": 34130 }, { "grad_norm": 0.09667235612869263, "learning_rate": 5.755751732523401e-06, "loss": 0.0133, "step": 34140 }, { "grad_norm": 0.24320699274539948, "learning_rate": 5.736511801470595e-06, "loss": 0.0179, "step": 34150 }, { "grad_norm": 0.12066797912120819, "learning_rate": 5.717302124094603e-06, "loss": 0.0149, "step": 34160 }, { "grad_norm": 0.13135237991809845, "learning_rate": 5.6981227135250396e-06, "loss": 0.0148, "step": 34170 }, { "grad_norm": 0.10232607275247574, "learning_rate": 5.678973582870867e-06, "loss": 0.0141, "step": 34180 }, { "grad_norm": 0.10065541416406631, "learning_rate": 5.659854745220317e-06, "loss": 0.0152, "step": 34190 }, { "grad_norm": 0.08123557269573212, "learning_rate": 5.640766213640947e-06, "loss": 0.0159, "step": 34200 }, { "grad_norm": 0.06718101352453232, "learning_rate": 5.6217080011795934e-06, "loss": 0.0184, "step": 34210 }, { "grad_norm": 0.0822458565235138, "learning_rate": 5.602680120862364e-06, "loss": 0.0164, "step": 34220 }, { "grad_norm": 0.1547560840845108, "learning_rate": 5.583682585694616e-06, "loss": 0.0133, "step": 34230 }, { "grad_norm": 0.15630704164505005, "learning_rate": 5.564715408661009e-06, "loss": 0.0143, "step": 34240 }, { "grad_norm": 0.06883077323436737, "learning_rate": 5.5457786027254265e-06, "loss": 0.0158, "step": 34250 }, { "grad_norm": 0.1762944906949997, "learning_rate": 5.526872180830977e-06, "loss": 0.0144, "step": 34260 }, { "grad_norm": 0.10180289298295975, "learning_rate": 5.507996155900041e-06, "loss": 0.0169, "step": 34270 }, { "grad_norm": 0.09738782048225403, "learning_rate": 5.489150540834203e-06, "loss": 0.0148, "step": 34280 }, { "grad_norm": 0.0688006654381752, "learning_rate": 5.47033534851425e-06, "loss": 0.0165, "step": 34290 }, { "grad_norm": 0.0890662744641304, "learning_rate": 5.451550591800198e-06, "loss": 0.0142, "step": 34300 }, { "grad_norm": 0.06964480131864548, "learning_rate": 5.43279628353125e-06, "loss": 0.0153, "step": 34310 }, { "grad_norm": 0.1259564310312271, "learning_rate": 5.4140724365258075e-06, "loss": 0.0142, "step": 34320 }, { "grad_norm": 0.11378549784421921, "learning_rate": 5.395379063581424e-06, "loss": 0.0152, "step": 34330 }, { "grad_norm": 0.10953027009963989, "learning_rate": 5.376716177474866e-06, "loss": 0.0149, "step": 34340 }, { "grad_norm": 0.10107026249170303, "learning_rate": 5.358083790962015e-06, "loss": 0.0157, "step": 34350 }, { "grad_norm": 0.05632511526346207, "learning_rate": 5.33948191677795e-06, "loss": 0.0167, "step": 34360 }, { "grad_norm": 0.08996647596359253, "learning_rate": 5.320910567636866e-06, "loss": 0.0141, "step": 34370 }, { "grad_norm": 0.08821073919534683, "learning_rate": 5.302369756232123e-06, "loss": 0.016, "step": 34380 }, { "grad_norm": 0.06129324063658714, "learning_rate": 5.283859495236165e-06, "loss": 0.0131, "step": 34390 }, { "grad_norm": 0.07884082943201065, "learning_rate": 5.265379797300602e-06, "loss": 0.0162, "step": 34400 }, { "grad_norm": 0.08999728411436081, "learning_rate": 5.246930675056117e-06, "loss": 0.0132, "step": 34410 }, { "grad_norm": 0.05883781239390373, "learning_rate": 5.228512141112535e-06, "loss": 0.0141, "step": 34420 }, { "grad_norm": 0.10614261031150818, "learning_rate": 5.21012420805872e-06, "loss": 0.0172, "step": 34430 }, { "grad_norm": 0.1619044989347458, "learning_rate": 5.191766888462679e-06, "loss": 0.0156, "step": 34440 }, { "grad_norm": 0.11667341738939285, "learning_rate": 5.173440194871437e-06, "loss": 0.0155, "step": 34450 }, { "grad_norm": 0.08380714058876038, "learning_rate": 5.155144139811141e-06, "loss": 0.0163, "step": 34460 }, { "grad_norm": 0.050978515297174454, "learning_rate": 5.13687873578696e-06, "loss": 0.0145, "step": 34470 }, { "grad_norm": 0.06300010532140732, "learning_rate": 5.11864399528314e-06, "loss": 0.0143, "step": 34480 }, { "grad_norm": 0.056702837347984314, "learning_rate": 5.1004399307629346e-06, "loss": 0.0169, "step": 34490 }, { "grad_norm": 0.08254005014896393, "learning_rate": 5.082266554668652e-06, "loss": 0.0132, "step": 34500 }, { "grad_norm": 0.056470971554517746, "learning_rate": 5.06412387942164e-06, "loss": 0.0151, "step": 34510 }, { "grad_norm": 0.09321156144142151, "learning_rate": 5.046011917422239e-06, "loss": 0.0159, "step": 34520 }, { "grad_norm": 0.06057898327708244, "learning_rate": 5.0279306810498026e-06, "loss": 0.014, "step": 34530 }, { "grad_norm": 0.08530881255865097, "learning_rate": 5.009880182662679e-06, "loss": 0.0137, "step": 34540 }, { "grad_norm": 0.08125997334718704, "learning_rate": 4.991860434598217e-06, "loss": 0.0156, "step": 34550 }, { "grad_norm": 0.10222714394330978, "learning_rate": 4.973871449172746e-06, "loss": 0.0178, "step": 34560 }, { "grad_norm": 0.08582849055528641, "learning_rate": 4.9559132386815695e-06, "loss": 0.0148, "step": 34570 }, { "grad_norm": 0.0625634640455246, "learning_rate": 4.93798581539896e-06, "loss": 0.0152, "step": 34580 }, { "grad_norm": 0.103554368019104, "learning_rate": 4.920089191578131e-06, "loss": 0.018, "step": 34590 }, { "grad_norm": 0.06357431411743164, "learning_rate": 4.90222337945126e-06, "loss": 0.0158, "step": 34600 }, { "grad_norm": 0.08813686668872833, "learning_rate": 4.8843883912294605e-06, "loss": 0.0172, "step": 34610 }, { "grad_norm": 0.05970250815153122, "learning_rate": 4.866584239102795e-06, "loss": 0.0187, "step": 34620 }, { "grad_norm": 0.06221269816160202, "learning_rate": 4.8488109352402164e-06, "loss": 0.0175, "step": 34630 }, { "grad_norm": 0.08995941281318665, "learning_rate": 4.831068491789603e-06, "loss": 0.016, "step": 34640 }, { "grad_norm": 0.07874691486358643, "learning_rate": 4.81335692087776e-06, "loss": 0.017, "step": 34650 }, { "grad_norm": 0.07973282039165497, "learning_rate": 4.795676234610374e-06, "loss": 0.0165, "step": 34660 }, { "grad_norm": 0.09476683288812637, "learning_rate": 4.778026445072037e-06, "loss": 0.0126, "step": 34670 }, { "grad_norm": 0.18612784147262573, "learning_rate": 4.760407564326208e-06, "loss": 0.0152, "step": 34680 }, { "grad_norm": 0.05548190325498581, "learning_rate": 4.742819604415222e-06, "loss": 0.0157, "step": 34690 }, { "grad_norm": 0.07658743858337402, "learning_rate": 4.725262577360284e-06, "loss": 0.0151, "step": 34700 }, { "grad_norm": 0.10342096537351608, "learning_rate": 4.707736495161458e-06, "loss": 0.0134, "step": 34710 }, { "grad_norm": 0.09359657019376755, "learning_rate": 4.690241369797671e-06, "loss": 0.0153, "step": 34720 }, { "grad_norm": 0.10351771116256714, "learning_rate": 4.672777213226659e-06, "loss": 0.0157, "step": 34730 }, { "grad_norm": 0.11163835972547531, "learning_rate": 4.655344037385001e-06, "loss": 0.0167, "step": 34740 }, { "grad_norm": 0.07210009545087814, "learning_rate": 4.6379418541881215e-06, "loss": 0.0144, "step": 34750 }, { "grad_norm": 0.10398749262094498, "learning_rate": 4.620570675530245e-06, "loss": 0.0159, "step": 34760 }, { "grad_norm": 0.07321063429117203, "learning_rate": 4.603230513284407e-06, "loss": 0.0143, "step": 34770 }, { "grad_norm": 0.21727025508880615, "learning_rate": 4.58592137930246e-06, "loss": 0.0176, "step": 34780 }, { "grad_norm": 0.08405710011720657, "learning_rate": 4.568643285415009e-06, "loss": 0.0161, "step": 34790 }, { "grad_norm": 0.06914382427930832, "learning_rate": 4.551396243431483e-06, "loss": 0.019, "step": 34800 }, { "grad_norm": 0.06954853981733322, "learning_rate": 4.534180265140075e-06, "loss": 0.013, "step": 34810 }, { "grad_norm": 0.07925461232662201, "learning_rate": 4.5169953623077356e-06, "loss": 0.0143, "step": 34820 }, { "grad_norm": 0.1457682102918625, "learning_rate": 4.499841546680189e-06, "loss": 0.0118, "step": 34830 }, { "grad_norm": 0.08987866342067719, "learning_rate": 4.4827188299819025e-06, "loss": 0.0163, "step": 34840 }, { "grad_norm": 0.11715580523014069, "learning_rate": 4.465627223916086e-06, "loss": 0.0147, "step": 34850 }, { "grad_norm": 0.10966192930936813, "learning_rate": 4.448566740164706e-06, "loss": 0.0145, "step": 34860 }, { "grad_norm": 0.07116763293743134, "learning_rate": 4.431537390388429e-06, "loss": 0.0125, "step": 34870 }, { "grad_norm": 0.0671481266617775, "learning_rate": 4.414539186226674e-06, "loss": 0.0152, "step": 34880 }, { "grad_norm": 0.07868187874555588, "learning_rate": 4.397572139297529e-06, "loss": 0.0148, "step": 34890 }, { "grad_norm": 0.07704394310712814, "learning_rate": 4.380636261197829e-06, "loss": 0.0128, "step": 34900 }, { "grad_norm": 0.166878342628479, "learning_rate": 4.36373156350307e-06, "loss": 0.0153, "step": 34910 }, { "grad_norm": 0.1431778371334076, "learning_rate": 4.3468580577674665e-06, "loss": 0.0173, "step": 34920 }, { "grad_norm": 0.07539903372526169, "learning_rate": 4.330015755523908e-06, "loss": 0.0163, "step": 34930 }, { "grad_norm": 0.3379894196987152, "learning_rate": 4.313204668283926e-06, "loss": 0.0145, "step": 34940 }, { "grad_norm": 0.4637930989265442, "learning_rate": 4.296424807537752e-06, "loss": 0.0149, "step": 34950 }, { "grad_norm": 0.18078993260860443, "learning_rate": 4.279676184754267e-06, "loss": 0.0166, "step": 34960 }, { "grad_norm": 0.10800597816705704, "learning_rate": 4.262958811380985e-06, "loss": 0.0154, "step": 34970 }, { "grad_norm": 0.16901542246341705, "learning_rate": 4.246272698844095e-06, "loss": 0.0181, "step": 34980 }, { "grad_norm": 0.10414145141839981, "learning_rate": 4.229617858548374e-06, "loss": 0.0123, "step": 34990 }, { "grad_norm": 0.04795187711715698, "learning_rate": 4.212994301877271e-06, "loss": 0.0149, "step": 35000 }, { "grad_norm": 0.09151126444339752, "learning_rate": 4.196402040192804e-06, "loss": 0.0128, "step": 35010 }, { "grad_norm": 0.1280326396226883, "learning_rate": 4.179841084835639e-06, "loss": 0.0157, "step": 35020 }, { "grad_norm": 0.15221327543258667, "learning_rate": 4.163311447125046e-06, "loss": 0.0153, "step": 35030 }, { "grad_norm": 0.17066359519958496, "learning_rate": 4.146813138358851e-06, "loss": 0.0166, "step": 35040 }, { "grad_norm": 0.1364479660987854, "learning_rate": 4.130346169813509e-06, "loss": 0.017, "step": 35050 }, { "grad_norm": 0.1032281145453453, "learning_rate": 4.113910552744038e-06, "loss": 0.0166, "step": 35060 }, { "grad_norm": 0.10537926852703094, "learning_rate": 4.097506298384024e-06, "loss": 0.0169, "step": 35070 }, { "grad_norm": 0.12798532843589783, "learning_rate": 4.081133417945626e-06, "loss": 0.017, "step": 35080 }, { "grad_norm": 0.07507933676242828, "learning_rate": 4.0647919226195484e-06, "loss": 0.0149, "step": 35090 }, { "grad_norm": 0.17881716787815094, "learning_rate": 4.04848182357504e-06, "loss": 0.0167, "step": 35100 }, { "grad_norm": 0.06130284070968628, "learning_rate": 4.032203131959911e-06, "loss": 0.0172, "step": 35110 }, { "grad_norm": 0.0745777115225792, "learning_rate": 4.015955858900483e-06, "loss": 0.0156, "step": 35120 }, { "grad_norm": 0.16182108223438263, "learning_rate": 3.999740015501641e-06, "loss": 0.0172, "step": 35130 }, { "grad_norm": 0.11485456675291061, "learning_rate": 3.9835556128467236e-06, "loss": 0.0163, "step": 35140 }, { "grad_norm": 0.05889859423041344, "learning_rate": 3.967402661997644e-06, "loss": 0.0167, "step": 35150 }, { "grad_norm": 0.11381931602954865, "learning_rate": 3.951281173994781e-06, "loss": 0.015, "step": 35160 }, { "grad_norm": 0.1429622620344162, "learning_rate": 3.935191159857016e-06, "loss": 0.0179, "step": 35170 }, { "grad_norm": 0.10252377390861511, "learning_rate": 3.919132630581746e-06, "loss": 0.014, "step": 35180 }, { "grad_norm": 0.07594145834445953, "learning_rate": 3.903105597144802e-06, "loss": 0.0157, "step": 35190 }, { "grad_norm": 0.0923718810081482, "learning_rate": 3.887110070500505e-06, "loss": 0.0164, "step": 35200 }, { "grad_norm": 0.06434468924999237, "learning_rate": 3.871146061581654e-06, "loss": 0.0173, "step": 35210 }, { "grad_norm": 0.13536077737808228, "learning_rate": 3.855213581299505e-06, "loss": 0.0145, "step": 35220 }, { "grad_norm": 0.08787071704864502, "learning_rate": 3.839312640543763e-06, "loss": 0.013, "step": 35230 }, { "grad_norm": 0.10474944114685059, "learning_rate": 3.823443250182551e-06, "loss": 0.0135, "step": 35240 }, { "grad_norm": 0.0631655603647232, "learning_rate": 3.80760542106246e-06, "loss": 0.0183, "step": 35250 }, { "grad_norm": 0.0801677480340004, "learning_rate": 3.7917991640084972e-06, "loss": 0.0161, "step": 35260 }, { "grad_norm": 0.11832869052886963, "learning_rate": 3.7760244898240925e-06, "loss": 0.0164, "step": 35270 }, { "grad_norm": 0.2174445390701294, "learning_rate": 3.760281409291078e-06, "loss": 0.0164, "step": 35280 }, { "grad_norm": 0.06873873621225357, "learning_rate": 3.7445699331697113e-06, "loss": 0.0162, "step": 35290 }, { "grad_norm": 0.07310695946216583, "learning_rate": 3.7288900721986262e-06, "loss": 0.0176, "step": 35300 }, { "grad_norm": 0.1274339258670807, "learning_rate": 3.7132418370948663e-06, "loss": 0.0183, "step": 35310 }, { "grad_norm": 0.10963230580091476, "learning_rate": 3.6976252385538535e-06, "loss": 0.0168, "step": 35320 }, { "grad_norm": 0.07041899114847183, "learning_rate": 3.6820402872493832e-06, "loss": 0.0138, "step": 35330 }, { "grad_norm": 0.05012526363134384, "learning_rate": 3.6664869938336343e-06, "loss": 0.0145, "step": 35340 }, { "grad_norm": 0.08396593481302261, "learning_rate": 3.650965368937115e-06, "loss": 0.0159, "step": 35350 }, { "grad_norm": 0.10144577920436859, "learning_rate": 3.6354754231687214e-06, "loss": 0.0166, "step": 35360 }, { "grad_norm": 0.07432730495929718, "learning_rate": 3.620017167115697e-06, "loss": 0.0192, "step": 35370 }, { "grad_norm": 0.10247863084077835, "learning_rate": 3.604590611343589e-06, "loss": 0.0159, "step": 35380 }, { "grad_norm": 0.10253391414880753, "learning_rate": 3.589195766396325e-06, "loss": 0.0147, "step": 35390 }, { "grad_norm": 0.07279621064662933, "learning_rate": 3.573832642796121e-06, "loss": 0.0137, "step": 35400 }, { "grad_norm": 0.0727882981300354, "learning_rate": 3.5585012510435336e-06, "loss": 0.0151, "step": 35410 }, { "grad_norm": 0.0851818397641182, "learning_rate": 3.5432016016174184e-06, "loss": 0.0143, "step": 35420 }, { "grad_norm": 0.06981167942285538, "learning_rate": 3.5279337049749506e-06, "loss": 0.0146, "step": 35430 }, { "grad_norm": 0.0809643417596817, "learning_rate": 3.5126975715515985e-06, "loss": 0.016, "step": 35440 }, { "grad_norm": 0.07472000271081924, "learning_rate": 3.4974932117610994e-06, "loss": 0.0164, "step": 35450 }, { "grad_norm": 0.08759187906980515, "learning_rate": 3.4823206359955065e-06, "loss": 0.0154, "step": 35460 }, { "grad_norm": 0.05397273600101471, "learning_rate": 3.4671798546251133e-06, "loss": 0.0149, "step": 35470 }, { "grad_norm": 0.09660078585147858, "learning_rate": 3.4520708779985088e-06, "loss": 0.0138, "step": 35480 }, { "grad_norm": 0.06215536966919899, "learning_rate": 3.4369937164425493e-06, "loss": 0.0138, "step": 35490 }, { "grad_norm": 0.17053501307964325, "learning_rate": 3.421948380262313e-06, "loss": 0.0161, "step": 35500 }, { "grad_norm": 0.06268353015184402, "learning_rate": 3.406934879741158e-06, "loss": 0.0136, "step": 35510 }, { "grad_norm": 0.3467918634414673, "learning_rate": 3.391953225140665e-06, "loss": 0.0129, "step": 35520 }, { "grad_norm": 0.07921838760375977, "learning_rate": 3.3770034267006533e-06, "loss": 0.0144, "step": 35530 }, { "grad_norm": 0.05369197949767113, "learning_rate": 3.362085494639178e-06, "loss": 0.0149, "step": 35540 }, { "grad_norm": 0.16645070910453796, "learning_rate": 3.3471994391524986e-06, "loss": 0.0141, "step": 35550 }, { "grad_norm": 0.07835856080055237, "learning_rate": 3.3323452704150825e-06, "loss": 0.0145, "step": 35560 }, { "grad_norm": 0.09311626106500626, "learning_rate": 3.3175229985796253e-06, "loss": 0.0144, "step": 35570 }, { "grad_norm": 0.06610327214002609, "learning_rate": 3.3027326337770115e-06, "loss": 0.0146, "step": 35580 }, { "grad_norm": 0.07628348469734192, "learning_rate": 3.2879741861163216e-06, "loss": 0.0133, "step": 35590 }, { "grad_norm": 0.08116751909255981, "learning_rate": 3.2732476656847976e-06, "loss": 0.0147, "step": 35600 }, { "grad_norm": 0.07424712926149368, "learning_rate": 3.2585530825478883e-06, "loss": 0.0177, "step": 35610 }, { "grad_norm": 0.0925835594534874, "learning_rate": 3.243890446749204e-06, "loss": 0.0141, "step": 35620 }, { "grad_norm": 0.1592162400484085, "learning_rate": 3.2292597683105107e-06, "loss": 0.0146, "step": 35630 }, { "grad_norm": 0.07804395258426666, "learning_rate": 3.214661057231755e-06, "loss": 0.0172, "step": 35640 }, { "grad_norm": 0.06732676178216934, "learning_rate": 3.2000943234910097e-06, "loss": 0.0149, "step": 35650 }, { "grad_norm": 0.06830788403749466, "learning_rate": 3.1855595770444835e-06, "loss": 0.0133, "step": 35660 }, { "grad_norm": 0.078715018928051, "learning_rate": 3.1710568278265574e-06, "loss": 0.0173, "step": 35670 }, { "grad_norm": 0.07232034206390381, "learning_rate": 3.156586085749719e-06, "loss": 0.0148, "step": 35680 }, { "grad_norm": 0.13785597681999207, "learning_rate": 3.1421473607045893e-06, "loss": 0.0151, "step": 35690 }, { "grad_norm": 0.08483551442623138, "learning_rate": 3.127740662559886e-06, "loss": 0.0155, "step": 35700 }, { "grad_norm": 0.15069332718849182, "learning_rate": 3.1133660011624655e-06, "loss": 0.0148, "step": 35710 }, { "grad_norm": 0.11380244791507721, "learning_rate": 3.099023386337263e-06, "loss": 0.0154, "step": 35720 }, { "grad_norm": 0.06341075152158737, "learning_rate": 3.0847128278873317e-06, "loss": 0.0189, "step": 35730 }, { "grad_norm": 0.09595402330160141, "learning_rate": 3.0704343355938027e-06, "loss": 0.0131, "step": 35740 }, { "grad_norm": 0.0657360628247261, "learning_rate": 3.056187919215886e-06, "loss": 0.014, "step": 35750 }, { "grad_norm": 0.09368480741977692, "learning_rate": 3.041973588490865e-06, "loss": 0.0159, "step": 35760 }, { "grad_norm": 0.13505925238132477, "learning_rate": 3.0277913531341128e-06, "loss": 0.0152, "step": 35770 }, { "grad_norm": 0.10092353075742722, "learning_rate": 3.0136412228390476e-06, "loss": 0.0132, "step": 35780 }, { "grad_norm": 0.10807076096534729, "learning_rate": 2.999523207277155e-06, "loss": 0.0146, "step": 35790 }, { "grad_norm": 0.08964905887842178, "learning_rate": 2.9854373160979552e-06, "loss": 0.014, "step": 35800 }, { "grad_norm": 0.11089273542165756, "learning_rate": 2.9713835589290362e-06, "loss": 0.0151, "step": 35810 }, { "grad_norm": 0.05960863456130028, "learning_rate": 2.9573619453759913e-06, "loss": 0.0141, "step": 35820 }, { "grad_norm": 0.07477321475744247, "learning_rate": 2.943372485022489e-06, "loss": 0.0155, "step": 35830 }, { "grad_norm": 0.10290581732988358, "learning_rate": 2.9294151874301623e-06, "loss": 0.0142, "step": 35840 }, { "grad_norm": 0.07682594656944275, "learning_rate": 2.9154900621387093e-06, "loss": 0.0152, "step": 35850 }, { "grad_norm": 0.05539606139063835, "learning_rate": 2.9015971186658164e-06, "loss": 0.0136, "step": 35860 }, { "grad_norm": 0.07863938063383102, "learning_rate": 2.887736366507182e-06, "loss": 0.0145, "step": 35870 }, { "grad_norm": 0.05982555076479912, "learning_rate": 2.873907815136501e-06, "loss": 0.0136, "step": 35880 }, { "grad_norm": 0.12596294283866882, "learning_rate": 2.8601114740054556e-06, "loss": 0.0134, "step": 35890 }, { "grad_norm": 0.06168777123093605, "learning_rate": 2.8463473525437155e-06, "loss": 0.0144, "step": 35900 }, { "grad_norm": 0.11147040128707886, "learning_rate": 2.8326154601589216e-06, "loss": 0.0147, "step": 35910 }, { "grad_norm": 0.06880584359169006, "learning_rate": 2.818915806236705e-06, "loss": 0.0138, "step": 35920 }, { "grad_norm": 0.1899726837873459, "learning_rate": 2.8052484001406478e-06, "loss": 0.0177, "step": 35930 }, { "grad_norm": 0.06503883004188538, "learning_rate": 2.791613251212277e-06, "loss": 0.0142, "step": 35940 }, { "grad_norm": 0.05381951853632927, "learning_rate": 2.7780103687711166e-06, "loss": 0.0128, "step": 35950 }, { "grad_norm": 0.04292993247509003, "learning_rate": 2.764439762114579e-06, "loss": 0.0115, "step": 35960 }, { "grad_norm": 0.051444873213768005, "learning_rate": 2.750901440518061e-06, "loss": 0.0124, "step": 35970 }, { "grad_norm": 0.11725978553295135, "learning_rate": 2.7373954132348776e-06, "loss": 0.0154, "step": 35980 }, { "grad_norm": 0.13747233152389526, "learning_rate": 2.7239216894962784e-06, "loss": 0.0142, "step": 35990 }, { "grad_norm": 0.29799604415893555, "learning_rate": 2.710480278511407e-06, "loss": 0.0142, "step": 36000 }, { "grad_norm": 0.07093815505504608, "learning_rate": 2.6970711894673483e-06, "loss": 0.0145, "step": 36010 }, { "grad_norm": 0.13041901588439941, "learning_rate": 2.6836944315291047e-06, "loss": 0.0139, "step": 36020 }, { "grad_norm": 0.07167463004589081, "learning_rate": 2.67035001383954e-06, "loss": 0.0154, "step": 36030 }, { "grad_norm": 0.07404111325740814, "learning_rate": 2.6570379455194527e-06, "loss": 0.0135, "step": 36040 }, { "grad_norm": 0.04906477406620979, "learning_rate": 2.6437582356675206e-06, "loss": 0.014, "step": 36050 }, { "grad_norm": 0.10161199420690536, "learning_rate": 2.6305108933602775e-06, "loss": 0.0158, "step": 36060 }, { "grad_norm": 0.07474133372306824, "learning_rate": 2.6172959276521803e-06, "loss": 0.0158, "step": 36070 }, { "grad_norm": 0.07357864081859589, "learning_rate": 2.6041133475755153e-06, "loss": 0.0152, "step": 36080 }, { "grad_norm": 0.1518239676952362, "learning_rate": 2.590963162140475e-06, "loss": 0.0182, "step": 36090 }, { "grad_norm": 0.07009082287549973, "learning_rate": 2.5778453803350587e-06, "loss": 0.014, "step": 36100 }, { "grad_norm": 0.13663436472415924, "learning_rate": 2.5647600111251723e-06, "loss": 0.0155, "step": 36110 }, { "grad_norm": 0.0684775784611702, "learning_rate": 2.551707063454528e-06, "loss": 0.0123, "step": 36120 }, { "grad_norm": 0.05841892585158348, "learning_rate": 2.5386865462446897e-06, "loss": 0.0162, "step": 36130 }, { "grad_norm": 0.056359488517045975, "learning_rate": 2.52569846839506e-06, "loss": 0.0155, "step": 36140 }, { "grad_norm": 0.052819930016994476, "learning_rate": 2.512742838782883e-06, "loss": 0.0158, "step": 36150 }, { "grad_norm": 0.10013320297002792, "learning_rate": 2.4998196662631924e-06, "loss": 0.0148, "step": 36160 }, { "grad_norm": 0.08092306554317474, "learning_rate": 2.4869289596688604e-06, "loss": 0.0153, "step": 36170 }, { "grad_norm": 0.09093661606311798, "learning_rate": 2.4740707278105567e-06, "loss": 0.0182, "step": 36180 }, { "grad_norm": 0.07272279262542725, "learning_rate": 2.461244979476779e-06, "loss": 0.0137, "step": 36190 }, { "grad_norm": 0.05954035371541977, "learning_rate": 2.4484517234337866e-06, "loss": 0.0126, "step": 36200 }, { "grad_norm": 0.0685371682047844, "learning_rate": 2.4356909684256634e-06, "loss": 0.0155, "step": 36210 }, { "grad_norm": 0.10341599583625793, "learning_rate": 2.4229627231742545e-06, "loss": 0.0149, "step": 36220 }, { "grad_norm": 0.07356594502925873, "learning_rate": 2.4102669963791956e-06, "loss": 0.0136, "step": 36230 }, { "grad_norm": 0.14521673321723938, "learning_rate": 2.3976037967179065e-06, "loss": 0.0159, "step": 36240 }, { "grad_norm": 0.08692482113838196, "learning_rate": 2.3849731328455582e-06, "loss": 0.012, "step": 36250 }, { "grad_norm": 0.12100858241319656, "learning_rate": 2.3723750133950894e-06, "loss": 0.0171, "step": 36260 }, { "grad_norm": 0.09907107055187225, "learning_rate": 2.3598094469771957e-06, "loss": 0.0176, "step": 36270 }, { "grad_norm": 0.05511576309800148, "learning_rate": 2.347276442180324e-06, "loss": 0.0203, "step": 36280 }, { "grad_norm": 0.08520962297916412, "learning_rate": 2.3347760075706715e-06, "loss": 0.0143, "step": 36290 }, { "grad_norm": 0.07146132737398148, "learning_rate": 2.3223081516921595e-06, "loss": 0.0148, "step": 36300 }, { "grad_norm": 0.09201943874359131, "learning_rate": 2.3098728830664494e-06, "loss": 0.0166, "step": 36310 }, { "grad_norm": 0.08928460627794266, "learning_rate": 2.2974702101929314e-06, "loss": 0.0157, "step": 36320 }, { "grad_norm": 0.10168711841106415, "learning_rate": 2.2851001415487185e-06, "loss": 0.0182, "step": 36330 }, { "grad_norm": 0.06409987062215805, "learning_rate": 2.272762685588631e-06, "loss": 0.0153, "step": 36340 }, { "grad_norm": 0.08670008927583694, "learning_rate": 2.260457850745218e-06, "loss": 0.0159, "step": 36350 }, { "grad_norm": 0.053884245455265045, "learning_rate": 2.2481856454286963e-06, "loss": 0.0157, "step": 36360 }, { "grad_norm": 0.107914038002491, "learning_rate": 2.235946078027018e-06, "loss": 0.0169, "step": 36370 }, { "grad_norm": 0.07730334997177124, "learning_rate": 2.223739156905802e-06, "loss": 0.0145, "step": 36380 }, { "grad_norm": 0.07008013874292374, "learning_rate": 2.2115648904083796e-06, "loss": 0.0166, "step": 36390 }, { "grad_norm": 0.09105318039655685, "learning_rate": 2.199423286855734e-06, "loss": 0.0141, "step": 36400 }, { "grad_norm": 0.21505846083164215, "learning_rate": 2.1873143545465323e-06, "loss": 0.0151, "step": 36410 }, { "grad_norm": 0.0924505814909935, "learning_rate": 2.1752381017571267e-06, "loss": 0.0124, "step": 36420 }, { "grad_norm": 0.08159168809652328, "learning_rate": 2.1631945367415142e-06, "loss": 0.0193, "step": 36430 }, { "grad_norm": 0.08255303651094437, "learning_rate": 2.15118366773136e-06, "loss": 0.0161, "step": 36440 }, { "grad_norm": 0.059480514377355576, "learning_rate": 2.1392055029359926e-06, "loss": 0.0138, "step": 36450 }, { "grad_norm": 0.05250483378767967, "learning_rate": 2.127260050542357e-06, "loss": 0.0141, "step": 36460 }, { "grad_norm": 0.13168562948703766, "learning_rate": 2.1153473187150608e-06, "loss": 0.0162, "step": 36470 }, { "grad_norm": 0.06782876700162888, "learning_rate": 2.1034673155963467e-06, "loss": 0.015, "step": 36480 }, { "grad_norm": 0.12262195348739624, "learning_rate": 2.0916200493060976e-06, "loss": 0.0147, "step": 36490 }, { "grad_norm": 0.08122136443853378, "learning_rate": 2.0798055279417916e-06, "loss": 0.0144, "step": 36500 }, { "grad_norm": 0.07326089590787888, "learning_rate": 2.0680237595785356e-06, "loss": 0.016, "step": 36510 }, { "grad_norm": 0.08887138217687607, "learning_rate": 2.0562747522690774e-06, "loss": 0.0138, "step": 36520 }, { "grad_norm": 0.08565720915794373, "learning_rate": 2.0445585140437373e-06, "loss": 0.0142, "step": 36530 }, { "grad_norm": 0.08265312016010284, "learning_rate": 2.0328750529104602e-06, "loss": 0.0156, "step": 36540 }, { "grad_norm": 0.09844355285167694, "learning_rate": 2.021224376854791e-06, "loss": 0.0162, "step": 36550 }, { "grad_norm": 0.07304804772138596, "learning_rate": 2.0096064938398386e-06, "loss": 0.0137, "step": 36560 }, { "grad_norm": 0.10087087005376816, "learning_rate": 1.998021411806328e-06, "loss": 0.0143, "step": 36570 }, { "grad_norm": 0.07276404649019241, "learning_rate": 1.9864691386725586e-06, "loss": 0.0143, "step": 36580 }, { "grad_norm": 0.04830918461084366, "learning_rate": 1.9749496823343814e-06, "loss": 0.0166, "step": 36590 }, { "grad_norm": 0.10843813419342041, "learning_rate": 1.963463050665254e-06, "loss": 0.0139, "step": 36600 }, { "grad_norm": 0.3352086544036865, "learning_rate": 1.9520092515161682e-06, "loss": 0.0147, "step": 36610 }, { "grad_norm": 0.07525701075792313, "learning_rate": 1.940588292715695e-06, "loss": 0.0183, "step": 36620 }, { "grad_norm": 0.0612567737698555, "learning_rate": 1.929200182069951e-06, "loss": 0.0132, "step": 36630 }, { "grad_norm": 0.17009352147579193, "learning_rate": 1.9178449273625997e-06, "loss": 0.0154, "step": 36640 }, { "grad_norm": 0.06679387390613556, "learning_rate": 1.90652253635486e-06, "loss": 0.0159, "step": 36650 }, { "grad_norm": 0.1094704419374466, "learning_rate": 1.8952330167854693e-06, "loss": 0.0158, "step": 36660 }, { "grad_norm": 0.1043052226305008, "learning_rate": 1.8839763763707107e-06, "loss": 0.0138, "step": 36670 }, { "grad_norm": 0.09831273555755615, "learning_rate": 1.8727526228043912e-06, "loss": 0.0173, "step": 36680 }, { "grad_norm": 0.0877072662115097, "learning_rate": 1.8615617637578408e-06, "loss": 0.0147, "step": 36690 }, { "grad_norm": 0.11482412368059158, "learning_rate": 1.850403806879908e-06, "loss": 0.0136, "step": 36700 }, { "grad_norm": 0.07204224914312363, "learning_rate": 1.8392787597969475e-06, "loss": 0.0168, "step": 36710 }, { "grad_norm": 0.07797136157751083, "learning_rate": 1.828186630112827e-06, "loss": 0.0129, "step": 36720 }, { "grad_norm": 0.1175796240568161, "learning_rate": 1.8171274254089098e-06, "loss": 0.0177, "step": 36730 }, { "grad_norm": 0.06012725830078125, "learning_rate": 1.806101153244061e-06, "loss": 0.0155, "step": 36740 }, { "grad_norm": 0.06668232381343842, "learning_rate": 1.7951078211546401e-06, "loss": 0.0194, "step": 36750 }, { "grad_norm": 0.10888425260782242, "learning_rate": 1.7841474366544764e-06, "loss": 0.0153, "step": 36760 }, { "grad_norm": 0.07287228107452393, "learning_rate": 1.7732200072348825e-06, "loss": 0.0149, "step": 36770 }, { "grad_norm": 0.06843118369579315, "learning_rate": 1.7623255403646677e-06, "loss": 0.0152, "step": 36780 }, { "grad_norm": 0.07485683262348175, "learning_rate": 1.7514640434900865e-06, "loss": 0.0147, "step": 36790 }, { "grad_norm": 0.10465840250253677, "learning_rate": 1.7406355240348838e-06, "loss": 0.0167, "step": 36800 }, { "grad_norm": 0.13808292150497437, "learning_rate": 1.7298399894002392e-06, "loss": 0.0186, "step": 36810 }, { "grad_norm": 0.0608515627682209, "learning_rate": 1.7190774469647952e-06, "loss": 0.016, "step": 36820 }, { "grad_norm": 0.096861831843853, "learning_rate": 1.7083479040846617e-06, "loss": 0.0135, "step": 36830 }, { "grad_norm": 0.19904930889606476, "learning_rate": 1.697651368093367e-06, "loss": 0.0159, "step": 36840 }, { "grad_norm": 0.0946023166179657, "learning_rate": 1.6869878463019017e-06, "loss": 0.0167, "step": 36850 }, { "grad_norm": 0.10708705335855484, "learning_rate": 1.6763573459986858e-06, "loss": 0.0155, "step": 36860 }, { "grad_norm": 0.1176966205239296, "learning_rate": 1.6657598744495516e-06, "loss": 0.0127, "step": 36870 }, { "grad_norm": 0.11630651354789734, "learning_rate": 1.6551954388977774e-06, "loss": 0.0159, "step": 36880 }, { "grad_norm": 0.11779125034809113, "learning_rate": 1.644664046564054e-06, "loss": 0.0161, "step": 36890 }, { "grad_norm": 0.0744086429476738, "learning_rate": 1.634165704646501e-06, "loss": 0.0129, "step": 36900 }, { "grad_norm": 0.1221378743648529, "learning_rate": 1.6237004203206175e-06, "loss": 0.0148, "step": 36910 }, { "grad_norm": 0.07441767305135727, "learning_rate": 1.6132682007393373e-06, "loss": 0.0174, "step": 36920 }, { "grad_norm": 0.13842175900936127, "learning_rate": 1.6028690530329847e-06, "loss": 0.0159, "step": 36930 }, { "grad_norm": 0.05729187652468681, "learning_rate": 1.5925029843092742e-06, "loss": 0.0124, "step": 36940 }, { "grad_norm": 0.07973341643810272, "learning_rate": 1.5821700016533214e-06, "loss": 0.0151, "step": 36950 }, { "grad_norm": 0.07437010854482651, "learning_rate": 1.5718701121276268e-06, "loss": 0.014, "step": 36960 }, { "grad_norm": 0.08299902081489563, "learning_rate": 1.5616033227720483e-06, "loss": 0.0146, "step": 36970 }, { "grad_norm": 0.052661996334791183, "learning_rate": 1.5513696406038558e-06, "loss": 0.0148, "step": 36980 }, { "grad_norm": 0.08388043940067291, "learning_rate": 1.5411690726176653e-06, "loss": 0.0152, "step": 36990 }, { "grad_norm": 0.08217655122280121, "learning_rate": 1.5310016257854832e-06, "loss": 0.0145, "step": 37000 }, { "grad_norm": 0.08201023191213608, "learning_rate": 1.5208673070566504e-06, "loss": 0.0174, "step": 37010 }, { "grad_norm": 0.0518101342022419, "learning_rate": 1.5107661233578818e-06, "loss": 0.0162, "step": 37020 }, { "grad_norm": 0.048677898943424225, "learning_rate": 1.5006980815932437e-06, "loss": 0.017, "step": 37030 }, { "grad_norm": 0.06654418259859085, "learning_rate": 1.490663188644148e-06, "loss": 0.0148, "step": 37040 }, { "grad_norm": 0.08507633954286575, "learning_rate": 1.4806614513693418e-06, "loss": 0.0146, "step": 37050 }, { "grad_norm": 0.09306958317756653, "learning_rate": 1.4706928766049289e-06, "loss": 0.0141, "step": 37060 }, { "grad_norm": 0.10205383598804474, "learning_rate": 1.4607574711643257e-06, "loss": 0.016, "step": 37070 }, { "grad_norm": 0.07888300716876984, "learning_rate": 1.4508552418382949e-06, "loss": 0.0152, "step": 37080 }, { "grad_norm": 0.07060149312019348, "learning_rate": 1.4409861953949112e-06, "loss": 0.0152, "step": 37090 }, { "grad_norm": 0.0578920878469944, "learning_rate": 1.43115033857959e-06, "loss": 0.0146, "step": 37100 }, { "grad_norm": 0.05727788433432579, "learning_rate": 1.4213476781150258e-06, "loss": 0.0138, "step": 37110 }, { "grad_norm": 0.11201096326112747, "learning_rate": 1.411578220701254e-06, "loss": 0.013, "step": 37120 }, { "grad_norm": 0.061802271753549576, "learning_rate": 1.4018419730156108e-06, "loss": 0.0153, "step": 37130 }, { "grad_norm": 0.11616197973489761, "learning_rate": 1.3921389417127283e-06, "loss": 0.0166, "step": 37140 }, { "grad_norm": 0.060386329889297485, "learning_rate": 1.3824691334245298e-06, "loss": 0.0133, "step": 37150 }, { "grad_norm": 0.08265932649374008, "learning_rate": 1.3728325547602505e-06, "loss": 0.0114, "step": 37160 }, { "grad_norm": 0.10217718780040741, "learning_rate": 1.3632292123063828e-06, "loss": 0.0144, "step": 37170 }, { "grad_norm": 0.04280533269047737, "learning_rate": 1.3536591126267317e-06, "loss": 0.0161, "step": 37180 }, { "grad_norm": 0.06009168177843094, "learning_rate": 1.3441222622623761e-06, "loss": 0.0143, "step": 37190 }, { "grad_norm": 0.07607030868530273, "learning_rate": 1.3346186677316575e-06, "loss": 0.0137, "step": 37200 }, { "grad_norm": 0.1066872775554657, "learning_rate": 1.3251483355301852e-06, "loss": 0.0163, "step": 37210 }, { "grad_norm": 0.047500062733888626, "learning_rate": 1.315711272130854e-06, "loss": 0.0133, "step": 37220 }, { "grad_norm": 0.07187804579734802, "learning_rate": 1.306307483983804e-06, "loss": 0.0135, "step": 37230 }, { "grad_norm": 0.036092184484004974, "learning_rate": 1.2969369775164275e-06, "loss": 0.0125, "step": 37240 }, { "grad_norm": 0.04848887026309967, "learning_rate": 1.28759975913339e-06, "loss": 0.0134, "step": 37250 }, { "grad_norm": 0.12101773917675018, "learning_rate": 1.2782958352165863e-06, "loss": 0.014, "step": 37260 }, { "grad_norm": 0.08192415535449982, "learning_rate": 1.2690252121251633e-06, "loss": 0.0144, "step": 37270 }, { "grad_norm": 0.08147703856229782, "learning_rate": 1.2597878961954968e-06, "loss": 0.0149, "step": 37280 }, { "grad_norm": 0.24960540235042572, "learning_rate": 1.250583893741214e-06, "loss": 0.0127, "step": 37290 }, { "grad_norm": 0.13037365674972534, "learning_rate": 1.2414132110531607e-06, "loss": 0.0144, "step": 37300 }, { "grad_norm": 0.12072500586509705, "learning_rate": 1.2322758543994062e-06, "loss": 0.0148, "step": 37310 }, { "grad_norm": 0.04193107411265373, "learning_rate": 1.2231718300252603e-06, "loss": 0.0123, "step": 37320 }, { "grad_norm": 0.12175855785608292, "learning_rate": 1.214101144153229e-06, "loss": 0.0119, "step": 37330 }, { "grad_norm": 0.12181798368692398, "learning_rate": 1.205063802983042e-06, "loss": 0.0128, "step": 37340 }, { "grad_norm": 0.2173980325460434, "learning_rate": 1.1960598126916301e-06, "loss": 0.0162, "step": 37350 }, { "grad_norm": 0.05841341242194176, "learning_rate": 1.187089179433154e-06, "loss": 0.0154, "step": 37360 }, { "grad_norm": 0.2240191251039505, "learning_rate": 1.178151909338937e-06, "loss": 0.0145, "step": 37370 }, { "grad_norm": 0.18850913643836975, "learning_rate": 1.1692480085175317e-06, "loss": 0.0167, "step": 37380 }, { "grad_norm": 0.10416349768638611, "learning_rate": 1.1603774830546587e-06, "loss": 0.0132, "step": 37390 }, { "grad_norm": 0.06407332420349121, "learning_rate": 1.1515403390132573e-06, "loss": 0.0168, "step": 37400 }, { "grad_norm": 0.19843870401382446, "learning_rate": 1.1427365824334124e-06, "loss": 0.0137, "step": 37410 }, { "grad_norm": 0.11481281369924545, "learning_rate": 1.133966219332422e-06, "loss": 0.0173, "step": 37420 }, { "grad_norm": 0.06906480342149734, "learning_rate": 1.1252292557047405e-06, "loss": 0.0155, "step": 37430 }, { "grad_norm": 0.04739775508642197, "learning_rate": 1.1165256975219974e-06, "loss": 0.0153, "step": 37440 }, { "grad_norm": 0.07347685843706131, "learning_rate": 1.1078555507330002e-06, "loss": 0.0133, "step": 37450 }, { "grad_norm": 0.0773715004324913, "learning_rate": 1.0992188212637089e-06, "loss": 0.0139, "step": 37460 }, { "grad_norm": 0.14491184055805206, "learning_rate": 1.0906155150172458e-06, "loss": 0.0183, "step": 37470 }, { "grad_norm": 0.06537749618291855, "learning_rate": 1.0820456378738852e-06, "loss": 0.0145, "step": 37480 }, { "grad_norm": 0.1512938290834427, "learning_rate": 1.0735091956910692e-06, "loss": 0.0171, "step": 37490 }, { "grad_norm": 0.10926222801208496, "learning_rate": 1.0650061943033695e-06, "loss": 0.0159, "step": 37500 }, { "grad_norm": 0.07115606963634491, "learning_rate": 1.0565366395225041e-06, "loss": 0.0155, "step": 37510 }, { "grad_norm": 0.09358591586351395, "learning_rate": 1.0481005371373366e-06, "loss": 0.0134, "step": 37520 }, { "grad_norm": 0.06304822117090225, "learning_rate": 1.0396978929138602e-06, "loss": 0.0144, "step": 37530 }, { "grad_norm": 0.13104940950870514, "learning_rate": 1.0313287125952032e-06, "loss": 0.0166, "step": 37540 }, { "grad_norm": 0.06315246224403381, "learning_rate": 1.0229930019016288e-06, "loss": 0.0154, "step": 37550 }, { "grad_norm": 0.05361852049827576, "learning_rate": 1.0146907665305183e-06, "loss": 0.0134, "step": 37560 }, { "grad_norm": 0.13381406664848328, "learning_rate": 1.0064220121563551e-06, "loss": 0.0175, "step": 37570 }, { "grad_norm": 0.0713755339384079, "learning_rate": 9.98186744430768e-07, "loss": 0.0128, "step": 37580 }, { "grad_norm": 0.08834373205900192, "learning_rate": 9.899849689824826e-07, "loss": 0.0158, "step": 37590 }, { "grad_norm": 0.11225444823503494, "learning_rate": 9.818166914173365e-07, "loss": 0.0157, "step": 37600 }, { "grad_norm": 0.07400171458721161, "learning_rate": 9.736819173182642e-07, "loss": 0.0139, "step": 37610 }, { "grad_norm": 0.17411494255065918, "learning_rate": 9.655806522453126e-07, "loss": 0.0145, "step": 37620 }, { "grad_norm": 0.08709681034088135, "learning_rate": 9.575129017356187e-07, "loss": 0.0138, "step": 37630 }, { "grad_norm": 0.08450818806886673, "learning_rate": 9.494786713034109e-07, "loss": 0.0176, "step": 37640 }, { "grad_norm": 0.07434330135583878, "learning_rate": 9.414779664400075e-07, "loss": 0.015, "step": 37650 }, { "grad_norm": 0.08003165572881699, "learning_rate": 9.335107926138342e-07, "loss": 0.015, "step": 37660 }, { "grad_norm": 0.06711646914482117, "learning_rate": 9.255771552703519e-07, "loss": 0.0166, "step": 37670 }, { "grad_norm": 0.07330671697854996, "learning_rate": 9.176770598321449e-07, "loss": 0.0152, "step": 37680 }, { "grad_norm": 0.0491466298699379, "learning_rate": 9.098105116988498e-07, "loss": 0.0141, "step": 37690 }, { "grad_norm": 0.08464260399341583, "learning_rate": 9.019775162471767e-07, "loss": 0.0165, "step": 37700 }, { "grad_norm": 0.10538582503795624, "learning_rate": 8.94178078830904e-07, "loss": 0.0144, "step": 37710 }, { "grad_norm": 0.08925462514162064, "learning_rate": 8.864122047808731e-07, "loss": 0.0149, "step": 37720 }, { "grad_norm": 0.08783804625272751, "learning_rate": 8.786798994049882e-07, "loss": 0.0144, "step": 37730 }, { "grad_norm": 0.07772394269704819, "learning_rate": 8.709811679882052e-07, "loss": 0.0136, "step": 37740 }, { "grad_norm": 0.07570521533489227, "learning_rate": 8.633160157925313e-07, "loss": 0.0146, "step": 37750 }, { "grad_norm": 0.08455765247344971, "learning_rate": 8.556844480570369e-07, "loss": 0.0146, "step": 37760 }, { "grad_norm": 0.03706727921962738, "learning_rate": 8.480864699978164e-07, "loss": 0.0159, "step": 37770 }, { "grad_norm": 0.07178615778684616, "learning_rate": 8.405220868080154e-07, "loss": 0.0126, "step": 37780 }, { "grad_norm": 0.07766041159629822, "learning_rate": 8.329913036578263e-07, "loss": 0.0143, "step": 37790 }, { "grad_norm": 0.09354519844055176, "learning_rate": 8.254941256944593e-07, "loss": 0.0136, "step": 37800 }, { "grad_norm": 0.049269743263721466, "learning_rate": 8.18030558042171e-07, "loss": 0.0134, "step": 37810 }, { "grad_norm": 0.08349484205245972, "learning_rate": 8.106006058022419e-07, "loss": 0.0164, "step": 37820 }, { "grad_norm": 0.059483788907527924, "learning_rate": 8.03204274052971e-07, "loss": 0.0168, "step": 37830 }, { "grad_norm": 0.18729177117347717, "learning_rate": 7.958415678496756e-07, "loss": 0.0171, "step": 37840 }, { "grad_norm": 0.06414800882339478, "learning_rate": 7.885124922247022e-07, "loss": 0.0142, "step": 37850 }, { "grad_norm": 0.06789879500865936, "learning_rate": 7.812170521874051e-07, "loss": 0.0152, "step": 37860 }, { "grad_norm": 0.07205114513635635, "learning_rate": 7.739552527241456e-07, "loss": 0.0149, "step": 37870 }, { "grad_norm": 0.0782490000128746, "learning_rate": 7.667270987983033e-07, "loss": 0.0117, "step": 37880 }, { "grad_norm": 0.10720033943653107, "learning_rate": 7.595325953502375e-07, "loss": 0.015, "step": 37890 }, { "grad_norm": 0.09725473076105118, "learning_rate": 7.523717472973368e-07, "loss": 0.0173, "step": 37900 }, { "grad_norm": 0.07242182642221451, "learning_rate": 7.452445595339696e-07, "loss": 0.0137, "step": 37910 }, { "grad_norm": 0.0981801226735115, "learning_rate": 7.381510369315059e-07, "loss": 0.0156, "step": 37920 }, { "grad_norm": 0.08167260140180588, "learning_rate": 7.310911843382895e-07, "loss": 0.0172, "step": 37930 }, { "grad_norm": 0.08242647349834442, "learning_rate": 7.240650065796716e-07, "loss": 0.0147, "step": 37940 }, { "grad_norm": 0.050070833414793015, "learning_rate": 7.170725084579777e-07, "loss": 0.0143, "step": 37950 }, { "grad_norm": 0.10235315561294556, "learning_rate": 7.101136947525233e-07, "loss": 0.0183, "step": 37960 }, { "grad_norm": 0.17226669192314148, "learning_rate": 7.031885702195707e-07, "loss": 0.016, "step": 37970 }, { "grad_norm": 0.2646254003047943, "learning_rate": 6.962971395924e-07, "loss": 0.0162, "step": 37980 }, { "grad_norm": 0.1113836020231247, "learning_rate": 6.894394075812217e-07, "loss": 0.0142, "step": 37990 }, { "grad_norm": 0.13465094566345215, "learning_rate": 6.826153788732359e-07, "loss": 0.0153, "step": 38000 }, { "grad_norm": 0.14871619641780853, "learning_rate": 6.758250581326009e-07, "loss": 0.0144, "step": 38010 }, { "grad_norm": 0.0908045619726181, "learning_rate": 6.690684500004485e-07, "loss": 0.0133, "step": 38020 }, { "grad_norm": 1.5369114875793457, "learning_rate": 6.623455590948458e-07, "loss": 0.0129, "step": 38030 }, { "grad_norm": 0.0680365189909935, "learning_rate": 6.556563900108226e-07, "loss": 0.0155, "step": 38040 }, { "grad_norm": 0.0569254495203495, "learning_rate": 6.490009473203662e-07, "loss": 0.0141, "step": 38050 }, { "grad_norm": 0.09543342888355255, "learning_rate": 6.423792355724156e-07, "loss": 0.0159, "step": 38060 }, { "grad_norm": 0.13226300477981567, "learning_rate": 6.35791259292845e-07, "loss": 0.0148, "step": 38070 }, { "grad_norm": 0.06083917245268822, "learning_rate": 6.292370229844691e-07, "loss": 0.0168, "step": 38080 }, { "grad_norm": 0.07970287650823593, "learning_rate": 6.227165311270488e-07, "loss": 0.0144, "step": 38090 }, { "grad_norm": 0.07350444793701172, "learning_rate": 6.16229788177275e-07, "loss": 0.0148, "step": 38100 }, { "grad_norm": 0.5557743906974792, "learning_rate": 6.097767985687841e-07, "loss": 0.0203, "step": 38110 }, { "grad_norm": 0.09656988084316254, "learning_rate": 6.033575667121317e-07, "loss": 0.0155, "step": 38120 }, { "grad_norm": 0.13167938590049744, "learning_rate": 5.969720969947967e-07, "loss": 0.0158, "step": 38130 }, { "grad_norm": 0.1429625004529953, "learning_rate": 5.906203937811883e-07, "loss": 0.0146, "step": 38140 }, { "grad_norm": 0.29539960622787476, "learning_rate": 5.843024614126447e-07, "loss": 0.017, "step": 38150 }, { "grad_norm": 0.057415541261434555, "learning_rate": 5.780183042074061e-07, "loss": 0.0146, "step": 38160 }, { "grad_norm": 0.061070989817380905, "learning_rate": 5.717679264606312e-07, "loss": 0.014, "step": 38170 }, { "grad_norm": 0.13852694630622864, "learning_rate": 5.655513324444028e-07, "loss": 0.0152, "step": 38180 }, { "grad_norm": 0.07931506633758545, "learning_rate": 5.593685264076942e-07, "loss": 0.015, "step": 38190 }, { "grad_norm": 0.07673317193984985, "learning_rate": 5.532195125764029e-07, "loss": 0.0178, "step": 38200 }, { "grad_norm": 0.11609195172786713, "learning_rate": 5.471042951533167e-07, "loss": 0.0178, "step": 38210 }, { "grad_norm": 0.08838611841201782, "learning_rate": 5.410228783181371e-07, "loss": 0.0169, "step": 38220 }, { "grad_norm": 0.13369299471378326, "learning_rate": 5.349752662274444e-07, "loss": 0.0162, "step": 38230 }, { "grad_norm": 0.06098444387316704, "learning_rate": 5.289614630147321e-07, "loss": 0.0154, "step": 38240 }, { "grad_norm": 0.07973547279834747, "learning_rate": 5.229814727903737e-07, "loss": 0.0165, "step": 38250 }, { "grad_norm": 0.3889649510383606, "learning_rate": 5.170352996416439e-07, "loss": 0.0178, "step": 38260 }, { "grad_norm": 0.1188381090760231, "learning_rate": 5.111229476326918e-07, "loss": 0.0144, "step": 38270 }, { "grad_norm": 0.08118944615125656, "learning_rate": 5.052444208045459e-07, "loss": 0.0176, "step": 38280 }, { "grad_norm": 0.07285989075899124, "learning_rate": 4.993997231751312e-07, "loss": 0.0157, "step": 38290 }, { "grad_norm": 0.1365593522787094, "learning_rate": 4.935888587392468e-07, "loss": 0.0148, "step": 38300 }, { "grad_norm": 0.08947479724884033, "learning_rate": 4.878118314685598e-07, "loss": 0.0165, "step": 38310 }, { "grad_norm": 0.06969830393791199, "learning_rate": 4.820686453116119e-07, "loss": 0.0123, "step": 38320 }, { "grad_norm": 0.19963669776916504, "learning_rate": 4.763593041938186e-07, "loss": 0.0118, "step": 38330 }, { "grad_norm": 0.10503017902374268, "learning_rate": 4.706838120174584e-07, "loss": 0.0133, "step": 38340 }, { "grad_norm": 0.08987042307853699, "learning_rate": 4.650421726616838e-07, "loss": 0.0172, "step": 38350 }, { "grad_norm": 0.07450859248638153, "learning_rate": 4.594343899824993e-07, "loss": 0.0141, "step": 38360 }, { "grad_norm": 0.19649052619934082, "learning_rate": 4.538604678127667e-07, "loss": 0.0159, "step": 38370 }, { "grad_norm": 0.07490156590938568, "learning_rate": 4.483204099622107e-07, "loss": 0.0156, "step": 38380 }, { "grad_norm": 0.08044340461492538, "learning_rate": 4.4281422021740793e-07, "loss": 0.0175, "step": 38390 }, { "grad_norm": 0.06945043802261353, "learning_rate": 4.373419023417924e-07, "loss": 0.0165, "step": 38400 }, { "grad_norm": 0.15665487945079803, "learning_rate": 4.319034600756389e-07, "loss": 0.0137, "step": 38410 }, { "grad_norm": 0.14141279458999634, "learning_rate": 4.264988971360739e-07, "loss": 0.0145, "step": 38420 }, { "grad_norm": 0.09170199185609818, "learning_rate": 4.211282172170594e-07, "loss": 0.0156, "step": 38430 }, { "grad_norm": 0.06971639394760132, "learning_rate": 4.1579142398940894e-07, "loss": 0.0153, "step": 38440 }, { "grad_norm": 0.09167597442865372, "learning_rate": 4.1048852110076585e-07, "loss": 0.0152, "step": 38450 }, { "grad_norm": 0.07892017066478729, "learning_rate": 4.052195121756197e-07, "loss": 0.0132, "step": 38460 }, { "grad_norm": 0.06295889616012573, "learning_rate": 3.9998440081528974e-07, "loss": 0.0125, "step": 38470 }, { "grad_norm": 0.06941143423318863, "learning_rate": 3.9478319059791936e-07, "loss": 0.0179, "step": 38480 }, { "grad_norm": 0.06827767938375473, "learning_rate": 3.8961588507848143e-07, "loss": 0.0145, "step": 38490 }, { "grad_norm": 0.1723705381155014, "learning_rate": 3.844824877887954e-07, "loss": 0.0207, "step": 38500 }, { "grad_norm": 0.07531668245792389, "learning_rate": 3.7938300223748224e-07, "loss": 0.0149, "step": 38510 }, { "grad_norm": 0.06443745642900467, "learning_rate": 3.743174319099929e-07, "loss": 0.0151, "step": 38520 }, { "grad_norm": 0.10573366284370422, "learning_rate": 3.6928578026859673e-07, "loss": 0.0128, "step": 38530 }, { "grad_norm": 0.10766985267400742, "learning_rate": 3.642880507523816e-07, "loss": 0.0121, "step": 38540 }, { "grad_norm": 0.08644318580627441, "learning_rate": 3.5932424677724285e-07, "loss": 0.0158, "step": 38550 }, { "grad_norm": 0.07941470295190811, "learning_rate": 3.5439437173589996e-07, "loss": 0.0189, "step": 38560 }, { "grad_norm": 0.0956042930483818, "learning_rate": 3.4949842899787423e-07, "loss": 0.0156, "step": 38570 }, { "grad_norm": 0.13261766731739044, "learning_rate": 3.446364219094944e-07, "loss": 0.0144, "step": 38580 }, { "grad_norm": 0.05613025277853012, "learning_rate": 3.3980835379389675e-07, "loss": 0.0161, "step": 38590 }, { "grad_norm": 0.12218651175498962, "learning_rate": 3.350142279510249e-07, "loss": 0.0155, "step": 38600 }, { "grad_norm": 0.07963351160287857, "learning_rate": 3.3025404765761324e-07, "loss": 0.0165, "step": 38610 }, { "grad_norm": 0.12509658932685852, "learning_rate": 3.255278161672037e-07, "loss": 0.0139, "step": 38620 }, { "grad_norm": 0.13403378427028656, "learning_rate": 3.20835536710129e-07, "loss": 0.0165, "step": 38630 }, { "grad_norm": 0.08441600948572159, "learning_rate": 3.1617721249351804e-07, "loss": 0.016, "step": 38640 }, { "grad_norm": 0.11743763834238052, "learning_rate": 3.115528467012963e-07, "loss": 0.0153, "step": 38650 }, { "grad_norm": 0.14868460595607758, "learning_rate": 3.0696244249416305e-07, "loss": 0.0134, "step": 38660 }, { "grad_norm": 0.31800398230552673, "learning_rate": 3.024060030096254e-07, "loss": 0.0157, "step": 38670 }, { "grad_norm": 0.06297490000724792, "learning_rate": 2.9788353136196435e-07, "loss": 0.0137, "step": 38680 }, { "grad_norm": 0.08786781132221222, "learning_rate": 2.9339503064224615e-07, "loss": 0.014, "step": 38690 }, { "grad_norm": 0.05260872468352318, "learning_rate": 2.889405039183224e-07, "loss": 0.0135, "step": 38700 }, { "grad_norm": 0.1688620001077652, "learning_rate": 2.8451995423481315e-07, "loss": 0.0154, "step": 38710 }, { "grad_norm": 0.14455996453762054, "learning_rate": 2.801333846131293e-07, "loss": 0.0127, "step": 38720 }, { "grad_norm": 0.1215178519487381, "learning_rate": 2.757807980514504e-07, "loss": 0.0154, "step": 38730 }, { "grad_norm": 0.07977542281150818, "learning_rate": 2.714621975247189e-07, "loss": 0.0181, "step": 38740 }, { "grad_norm": 0.12887723743915558, "learning_rate": 2.671775859846626e-07, "loss": 0.0147, "step": 38750 }, { "grad_norm": 0.13053327798843384, "learning_rate": 2.629269663597722e-07, "loss": 0.0146, "step": 38760 }, { "grad_norm": 0.12467323988676071, "learning_rate": 2.587103415553127e-07, "loss": 0.0168, "step": 38770 }, { "grad_norm": 0.08114802837371826, "learning_rate": 2.545277144533009e-07, "loss": 0.0153, "step": 38780 }, { "grad_norm": 0.07939638197422028, "learning_rate": 2.5037908791252227e-07, "loss": 0.017, "step": 38790 }, { "grad_norm": 0.07396992295980453, "learning_rate": 2.462644647685308e-07, "loss": 0.014, "step": 38800 }, { "grad_norm": 0.08121822774410248, "learning_rate": 2.421838478336269e-07, "loss": 0.0173, "step": 38810 }, { "grad_norm": 0.06387663632631302, "learning_rate": 2.3813723989687398e-07, "loss": 0.0147, "step": 38820 }, { "grad_norm": 0.2404939830303192, "learning_rate": 2.3412464372409293e-07, "loss": 0.0132, "step": 38830 }, { "grad_norm": 0.08056836575269699, "learning_rate": 2.301460620578455e-07, "loss": 0.0162, "step": 38840 }, { "grad_norm": 0.22636783123016357, "learning_rate": 2.262014976174731e-07, "loss": 0.0145, "step": 38850 }, { "grad_norm": 0.0908656194806099, "learning_rate": 2.2229095309903025e-07, "loss": 0.0145, "step": 38860 }, { "grad_norm": 0.10758741945028305, "learning_rate": 2.184144311753511e-07, "loss": 0.0174, "step": 38870 }, { "grad_norm": 0.06797367334365845, "learning_rate": 2.14571934495994e-07, "loss": 0.0159, "step": 38880 }, { "grad_norm": 0.07519116997718811, "learning_rate": 2.1076346568726923e-07, "loss": 0.0132, "step": 38890 }, { "grad_norm": 0.12487878650426865, "learning_rate": 2.0698902735223346e-07, "loss": 0.0119, "step": 38900 }, { "grad_norm": 0.08601144701242447, "learning_rate": 2.0324862207067863e-07, "loss": 0.0145, "step": 38910 }, { "grad_norm": 0.07077411562204361, "learning_rate": 1.9954225239913195e-07, "loss": 0.0143, "step": 38920 }, { "grad_norm": 0.0978090763092041, "learning_rate": 1.9586992087087252e-07, "loss": 0.0136, "step": 38930 }, { "grad_norm": 0.07082546502351761, "learning_rate": 1.9223162999588705e-07, "loss": 0.0132, "step": 38940 }, { "grad_norm": 0.06451688706874847, "learning_rate": 1.886273822609308e-07, "loss": 0.0126, "step": 38950 }, { "grad_norm": 0.18216544389724731, "learning_rate": 1.8505718012946094e-07, "loss": 0.0145, "step": 38960 }, { "grad_norm": 0.16132916510105133, "learning_rate": 1.8152102604168664e-07, "loss": 0.0131, "step": 38970 }, { "grad_norm": 0.05615728348493576, "learning_rate": 1.7801892241452457e-07, "loss": 0.0125, "step": 38980 }, { "grad_norm": 0.0679105669260025, "learning_rate": 1.745508716416433e-07, "loss": 0.0118, "step": 38990 }, { "grad_norm": 0.16153861582279205, "learning_rate": 1.7111687609341342e-07, "loss": 0.0137, "step": 39000 }, { "grad_norm": 0.07785637676715851, "learning_rate": 1.6771693811693523e-07, "loss": 0.0167, "step": 39010 }, { "grad_norm": 0.058240946382284164, "learning_rate": 1.643510600360443e-07, "loss": 0.0154, "step": 39020 }, { "grad_norm": 0.10190368443727493, "learning_rate": 1.610192441512781e-07, "loss": 0.0135, "step": 39030 }, { "grad_norm": 0.08182048797607422, "learning_rate": 1.5772149273990956e-07, "loss": 0.0135, "step": 39040 }, { "grad_norm": 0.07722129672765732, "learning_rate": 1.5445780805591337e-07, "loss": 0.0159, "step": 39050 }, { "grad_norm": 0.059665389358997345, "learning_rate": 1.5122819232998852e-07, "loss": 0.0132, "step": 39060 }, { "grad_norm": 0.14074084162712097, "learning_rate": 1.4803264776955262e-07, "loss": 0.0145, "step": 39070 }, { "grad_norm": 0.1485411673784256, "learning_rate": 1.448711765587196e-07, "loss": 0.0111, "step": 39080 }, { "grad_norm": 0.23770898580551147, "learning_rate": 1.4174378085833885e-07, "loss": 0.0145, "step": 39090 }, { "grad_norm": 0.09066405892372131, "learning_rate": 1.3865046280594486e-07, "loss": 0.0136, "step": 39100 }, { "grad_norm": 0.0871458426117897, "learning_rate": 1.3559122451579643e-07, "loss": 0.0162, "step": 39110 }, { "grad_norm": 0.06258244812488556, "learning_rate": 1.3256606807884874e-07, "loss": 0.0127, "step": 39120 }, { "grad_norm": 0.20055264234542847, "learning_rate": 1.2957499556277563e-07, "loss": 0.014, "step": 39130 }, { "grad_norm": 0.05935979261994362, "learning_rate": 1.266180090119473e-07, "loss": 0.0181, "step": 39140 }, { "grad_norm": 0.11770831048488617, "learning_rate": 1.2369511044742488e-07, "loss": 0.0134, "step": 39150 }, { "grad_norm": 0.06745099276304245, "learning_rate": 1.2080630186698805e-07, "loss": 0.0135, "step": 39160 }, { "grad_norm": 0.09624669700860977, "learning_rate": 1.1795158524511851e-07, "loss": 0.013, "step": 39170 }, { "grad_norm": 0.04467061907052994, "learning_rate": 1.1513096253297773e-07, "loss": 0.0115, "step": 39180 }, { "grad_norm": 0.0951341837644577, "learning_rate": 1.1234443565843466e-07, "loss": 0.0141, "step": 39190 }, { "grad_norm": 0.05784020572900772, "learning_rate": 1.0959200652606028e-07, "loss": 0.0142, "step": 39200 }, { "grad_norm": 0.0526997372508049, "learning_rate": 1.0687367701711082e-07, "loss": 0.0127, "step": 39210 }, { "grad_norm": 0.057354118674993515, "learning_rate": 1.0418944898953898e-07, "loss": 0.0149, "step": 39220 }, { "grad_norm": 0.042050525546073914, "learning_rate": 1.0153932427798829e-07, "loss": 0.0127, "step": 39230 }, { "grad_norm": 0.0818597674369812, "learning_rate": 9.892330469379874e-08, "loss": 0.0146, "step": 39240 }, { "grad_norm": 0.0867827758193016, "learning_rate": 9.634139202498449e-08, "loss": 0.016, "step": 39250 }, { "grad_norm": 0.28608787059783936, "learning_rate": 9.379358803627281e-08, "loss": 0.0146, "step": 39260 }, { "grad_norm": 0.0772475004196167, "learning_rate": 9.12798944690485e-08, "loss": 0.013, "step": 39270 }, { "grad_norm": 0.07961549609899521, "learning_rate": 8.880031304141501e-08, "loss": 0.0145, "step": 39280 }, { "grad_norm": 0.14272280037403107, "learning_rate": 8.635484544812777e-08, "loss": 0.0159, "step": 39290 }, { "grad_norm": 0.1320405900478363, "learning_rate": 8.394349336064422e-08, "loss": 0.0148, "step": 39300 }, { "grad_norm": 0.057322874665260315, "learning_rate": 8.156625842710153e-08, "loss": 0.0167, "step": 39310 }, { "grad_norm": 0.0690556988120079, "learning_rate": 7.92231422723222e-08, "loss": 0.0174, "step": 39320 }, { "grad_norm": 0.06633639335632324, "learning_rate": 7.691414649780293e-08, "loss": 0.0137, "step": 39330 }, { "grad_norm": 0.656382143497467, "learning_rate": 7.463927268171467e-08, "loss": 0.0219, "step": 39340 }, { "grad_norm": 0.07443644851446152, "learning_rate": 7.239852237891365e-08, "loss": 0.0139, "step": 39350 }, { "grad_norm": 0.0992092415690422, "learning_rate": 7.019189712093588e-08, "loss": 0.0152, "step": 39360 }, { "grad_norm": 0.10425129532814026, "learning_rate": 6.801939841598603e-08, "loss": 0.0127, "step": 39370 }, { "grad_norm": 0.08999470621347427, "learning_rate": 6.588102774894856e-08, "loss": 0.0152, "step": 39380 }, { "grad_norm": 0.06829449534416199, "learning_rate": 6.377678658137653e-08, "loss": 0.014, "step": 39390 }, { "grad_norm": 0.14680273830890656, "learning_rate": 6.17066763514973e-08, "loss": 0.0157, "step": 39400 }, { "grad_norm": 0.0815383717417717, "learning_rate": 5.967069847421792e-08, "loss": 0.0134, "step": 39410 }, { "grad_norm": 0.1853141486644745, "learning_rate": 5.7668854341108583e-08, "loss": 0.0129, "step": 39420 }, { "grad_norm": 0.07555830478668213, "learning_rate": 5.570114532040815e-08, "loss": 0.0152, "step": 39430 }, { "grad_norm": 0.07587763667106628, "learning_rate": 5.3767572757029706e-08, "loss": 0.0168, "step": 39440 }, { "grad_norm": 0.14569541811943054, "learning_rate": 5.186813797254386e-08, "loss": 0.0161, "step": 39450 }, { "grad_norm": 0.4301559031009674, "learning_rate": 5.000284226521213e-08, "loss": 0.0149, "step": 39460 }, { "grad_norm": 0.09269678592681885, "learning_rate": 4.8171686909931346e-08, "loss": 0.0153, "step": 39470 }, { "grad_norm": 0.12098932266235352, "learning_rate": 4.637467315828925e-08, "loss": 0.0156, "step": 39480 }, { "grad_norm": 0.05951690673828125, "learning_rate": 4.461180223852002e-08, "loss": 0.0158, "step": 39490 }, { "grad_norm": 0.07740315794944763, "learning_rate": 4.288307535553204e-08, "loss": 0.0149, "step": 39500 }, { "grad_norm": 0.04563309624791145, "learning_rate": 4.1188493690896836e-08, "loss": 0.0159, "step": 39510 }, { "grad_norm": 0.08389141410589218, "learning_rate": 3.952805840284346e-08, "loss": 0.0181, "step": 39520 }, { "grad_norm": 0.2111765295267105, "learning_rate": 3.790177062625855e-08, "loss": 0.0145, "step": 39530 }, { "grad_norm": 0.05785960704088211, "learning_rate": 3.6309631472702946e-08, "loss": 0.0163, "step": 39540 }, { "grad_norm": 0.09903743863105774, "learning_rate": 3.475164203038395e-08, "loss": 0.0155, "step": 39550 }, { "grad_norm": 0.12281688302755356, "learning_rate": 3.3227803364171974e-08, "loss": 0.0128, "step": 39560 }, { "grad_norm": 0.07555437833070755, "learning_rate": 3.173811651560055e-08, "loss": 0.0173, "step": 39570 }, { "grad_norm": 0.09130040556192398, "learning_rate": 3.0282582502855204e-08, "loss": 0.0154, "step": 39580 }, { "grad_norm": 0.07236438244581223, "learning_rate": 2.8861202320773496e-08, "loss": 0.0162, "step": 39590 }, { "grad_norm": 0.0910479798913002, "learning_rate": 2.7473976940867175e-08, "loss": 0.014, "step": 39600 }, { "grad_norm": 0.13184477388858795, "learning_rate": 2.6120907311277808e-08, "loss": 0.0169, "step": 39610 }, { "grad_norm": 0.06744330376386642, "learning_rate": 2.480199435682673e-08, "loss": 0.0146, "step": 39620 }, { "grad_norm": 0.0687052458524704, "learning_rate": 2.3517238978976174e-08, "loss": 0.0142, "step": 39630 }, { "grad_norm": 0.05135873705148697, "learning_rate": 2.2266642055834842e-08, "loss": 0.0136, "step": 39640 }, { "grad_norm": 0.06908178329467773, "learning_rate": 2.105020444219119e-08, "loss": 0.0156, "step": 39650 }, { "grad_norm": 0.09742460399866104, "learning_rate": 1.9867926969452388e-08, "loss": 0.0166, "step": 39660 }, { "grad_norm": 0.08620912581682205, "learning_rate": 1.8719810445705366e-08, "loss": 0.0149, "step": 39670 }, { "grad_norm": 0.09050711244344711, "learning_rate": 1.760585565567241e-08, "loss": 0.0144, "step": 39680 }, { "grad_norm": 0.15585745871067047, "learning_rate": 1.6526063360727816e-08, "loss": 0.0151, "step": 39690 }, { "grad_norm": 0.05776698514819145, "learning_rate": 1.5480434298908997e-08, "loss": 0.0144, "step": 39700 }, { "grad_norm": 0.090537428855896, "learning_rate": 1.4468969184883164e-08, "loss": 0.0166, "step": 39710 }, { "grad_norm": 0.07864557206630707, "learning_rate": 1.3491668709980643e-08, "loss": 0.0154, "step": 39720 }, { "grad_norm": 0.1347825527191162, "learning_rate": 1.254853354217822e-08, "loss": 0.0157, "step": 39730 }, { "grad_norm": 0.07158537954092026, "learning_rate": 1.1639564326099139e-08, "loss": 0.0152, "step": 39740 }, { "grad_norm": 0.09416475892066956, "learning_rate": 1.0764761683018653e-08, "loss": 0.0157, "step": 39750 }, { "grad_norm": 0.07146769762039185, "learning_rate": 9.924126210852925e-09, "loss": 0.016, "step": 39760 }, { "grad_norm": 0.10990043729543686, "learning_rate": 9.117658484170122e-09, "loss": 0.016, "step": 39770 }, { "grad_norm": 0.07795156538486481, "learning_rate": 8.345359054179324e-09, "loss": 0.0163, "step": 39780 }, { "grad_norm": 0.1200723797082901, "learning_rate": 7.60722844874162e-09, "loss": 0.0173, "step": 39790 }, { "grad_norm": 0.06299757212400436, "learning_rate": 6.903267172364558e-09, "loss": 0.0137, "step": 39800 }, { "grad_norm": 0.10748946666717529, "learning_rate": 6.23347570619659e-09, "loss": 0.0206, "step": 39810 }, { "grad_norm": 0.06452438980340958, "learning_rate": 5.59785450803818e-09, "loss": 0.0148, "step": 39820 }, { "grad_norm": 0.09297434985637665, "learning_rate": 4.996404012325151e-09, "loss": 0.0156, "step": 39830 }, { "grad_norm": 0.11044002324342728, "learning_rate": 4.429124630145331e-09, "loss": 0.0168, "step": 39840 }, { "grad_norm": 0.0921916738152504, "learning_rate": 3.89601674922746e-09, "loss": 0.0176, "step": 39850 }, { "grad_norm": 0.07360287755727768, "learning_rate": 3.3970807339467336e-09, "loss": 0.019, "step": 39860 }, { "grad_norm": 0.22157584130764008, "learning_rate": 2.9323169253248073e-09, "loss": 0.0158, "step": 39870 }, { "grad_norm": 0.18157655000686646, "learning_rate": 2.501725641018693e-09, "loss": 0.0141, "step": 39880 }, { "grad_norm": 0.09164756536483765, "learning_rate": 2.105307175331861e-09, "loss": 0.0135, "step": 39890 }, { "grad_norm": 0.2393665462732315, "learning_rate": 1.7430617992197918e-09, "loss": 0.0158, "step": 39900 }, { "grad_norm": 0.06962273269891739, "learning_rate": 1.4149897602677708e-09, "loss": 0.0188, "step": 39910 }, { "grad_norm": 0.11387611925601959, "learning_rate": 1.1210912827130937e-09, "loss": 0.0185, "step": 39920 }, { "grad_norm": 0.10982028394937515, "learning_rate": 8.613665674284121e-10, "loss": 0.0129, "step": 39930 }, { "grad_norm": 0.10932990163564682, "learning_rate": 6.358157919383878e-10, "loss": 0.015, "step": 39940 }, { "grad_norm": 0.07189736515283585, "learning_rate": 4.4443911040303877e-10, "loss": 0.0171, "step": 39950 }, { "grad_norm": 0.08147009462118149, "learning_rate": 2.8723665362329066e-10, "loss": 0.0157, "step": 39960 }, { "grad_norm": 0.12532201409339905, "learning_rate": 1.6420852905207895e-10, "loss": 0.0131, "step": 39970 }, { "grad_norm": 0.06756307929754257, "learning_rate": 7.535482076659328e-11, "loss": 0.0149, "step": 39980 }, { "grad_norm": 0.0817355066537857, "learning_rate": 2.0675589512686443e-11, "loss": 0.0167, "step": 39990 }, { "grad_norm": 0.10262326151132584, "learning_rate": 1.7087265491433358e-13, "loss": 0.0156, "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": 32, "trial_name": null, "trial_params": null }