{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "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": 1.6415215730667114, "learning_rate": 4.5e-07, "loss": 1.3216, "step": 10 }, { "grad_norm": 1.6009840965270996, "learning_rate": 9.5e-07, "loss": 1.3194, "step": 20 }, { "grad_norm": 1.470821738243103, "learning_rate": 1.45e-06, "loss": 1.3131, "step": 30 }, { "grad_norm": 1.0267130136489868, "learning_rate": 1.95e-06, "loss": 1.2985, "step": 40 }, { "grad_norm": 0.650117814540863, "learning_rate": 2.4500000000000003e-06, "loss": 1.2645, "step": 50 }, { "grad_norm": 0.47639691829681396, "learning_rate": 2.95e-06, "loss": 1.2259, "step": 60 }, { "grad_norm": 0.4135073125362396, "learning_rate": 3.4500000000000004e-06, "loss": 1.2035, "step": 70 }, { "grad_norm": 0.5063672661781311, "learning_rate": 3.95e-06, "loss": 1.2003, "step": 80 }, { "grad_norm": 0.584980309009552, "learning_rate": 4.45e-06, "loss": 1.1783, "step": 90 }, { "grad_norm": 0.7574766278266907, "learning_rate": 4.950000000000001e-06, "loss": 1.1456, "step": 100 }, { "grad_norm": 0.5545220375061035, "learning_rate": 5.45e-06, "loss": 1.1015, "step": 110 }, { "grad_norm": 0.611007809638977, "learning_rate": 5.95e-06, "loss": 1.0834, "step": 120 }, { "grad_norm": 0.569625198841095, "learning_rate": 6.45e-06, "loss": 1.0692, "step": 130 }, { "grad_norm": 0.5910418033599854, "learning_rate": 6.950000000000001e-06, "loss": 1.0728, "step": 140 }, { "grad_norm": 0.5891577005386353, "learning_rate": 7.45e-06, "loss": 1.0627, "step": 150 }, { "grad_norm": 0.49145135283470154, "learning_rate": 7.95e-06, "loss": 1.0487, "step": 160 }, { "grad_norm": 0.5444611310958862, "learning_rate": 8.45e-06, "loss": 1.0461, "step": 170 }, { "grad_norm": 0.4576614499092102, "learning_rate": 8.95e-06, "loss": 1.0504, "step": 180 }, { "grad_norm": 0.4975200593471527, "learning_rate": 9.450000000000001e-06, "loss": 1.0392, "step": 190 }, { "grad_norm": 0.4996066093444824, "learning_rate": 9.950000000000001e-06, "loss": 1.0371, "step": 200 }, { "grad_norm": 0.5163521766662598, "learning_rate": 1.045e-05, "loss": 1.0379, "step": 210 }, { "grad_norm": 0.6469461917877197, "learning_rate": 1.095e-05, "loss": 1.0389, "step": 220 }, { "grad_norm": 0.46692055463790894, "learning_rate": 1.145e-05, "loss": 1.0315, "step": 230 }, { "grad_norm": 0.5221322774887085, "learning_rate": 1.195e-05, "loss": 1.0279, "step": 240 }, { "grad_norm": 0.590556263923645, "learning_rate": 1.2450000000000001e-05, "loss": 1.0168, "step": 250 }, { "grad_norm": 1.1641744375228882, "learning_rate": 1.2950000000000001e-05, "loss": 1.0121, "step": 260 }, { "grad_norm": 0.6304851174354553, "learning_rate": 1.3450000000000002e-05, "loss": 0.9971, "step": 270 }, { "grad_norm": 0.7960355877876282, "learning_rate": 1.3950000000000002e-05, "loss": 0.9845, "step": 280 }, { "grad_norm": 0.7681980729103088, "learning_rate": 1.4449999999999999e-05, "loss": 0.9546, "step": 290 }, { "grad_norm": 1.7160232067108154, "learning_rate": 1.4950000000000001e-05, "loss": 0.9153, "step": 300 }, { "grad_norm": 1.1196322441101074, "learning_rate": 1.545e-05, "loss": 0.88, "step": 310 }, { "grad_norm": 1.6232943534851074, "learning_rate": 1.595e-05, "loss": 0.8343, "step": 320 }, { "grad_norm": 1.4979565143585205, "learning_rate": 1.645e-05, "loss": 0.7725, "step": 330 }, { "grad_norm": 1.485659122467041, "learning_rate": 1.6950000000000002e-05, "loss": 0.7215, "step": 340 }, { "grad_norm": 1.5367035865783691, "learning_rate": 1.745e-05, "loss": 0.6825, "step": 350 }, { "grad_norm": 1.3799759149551392, "learning_rate": 1.795e-05, "loss": 0.6418, "step": 360 }, { "grad_norm": 1.5134873390197754, "learning_rate": 1.845e-05, "loss": 0.5917, "step": 370 }, { "grad_norm": 1.5334382057189941, "learning_rate": 1.895e-05, "loss": 0.5604, "step": 380 }, { "grad_norm": 1.4374092817306519, "learning_rate": 1.9450000000000002e-05, "loss": 0.5468, "step": 390 }, { "grad_norm": 1.5520939826965332, "learning_rate": 1.995e-05, "loss": 0.5274, "step": 400 }, { "grad_norm": 1.2317813634872437, "learning_rate": 2.045e-05, "loss": 0.517, "step": 410 }, { "grad_norm": 1.4352760314941406, "learning_rate": 2.095e-05, "loss": 0.4788, "step": 420 }, { "grad_norm": 2.086040496826172, "learning_rate": 2.145e-05, "loss": 0.4635, "step": 430 }, { "grad_norm": 1.8872679471969604, "learning_rate": 2.195e-05, "loss": 0.4381, "step": 440 }, { "grad_norm": 1.6151196956634521, "learning_rate": 2.245e-05, "loss": 0.4234, "step": 450 }, { "grad_norm": 1.8495408296585083, "learning_rate": 2.2950000000000002e-05, "loss": 0.3971, "step": 460 }, { "grad_norm": 1.4908121824264526, "learning_rate": 2.345e-05, "loss": 0.3835, "step": 470 }, { "grad_norm": 1.8049460649490356, "learning_rate": 2.395e-05, "loss": 0.3591, "step": 480 }, { "grad_norm": 1.6240066289901733, "learning_rate": 2.445e-05, "loss": 0.3318, "step": 490 }, { "grad_norm": 1.6680376529693604, "learning_rate": 2.495e-05, "loss": 0.3192, "step": 500 }, { "grad_norm": 1.7371342182159424, "learning_rate": 2.5450000000000002e-05, "loss": 0.3016, "step": 510 }, { "grad_norm": 1.6735833883285522, "learning_rate": 2.595e-05, "loss": 0.2831, "step": 520 }, { "grad_norm": 2.490090847015381, "learning_rate": 2.6450000000000003e-05, "loss": 0.2534, "step": 530 }, { "grad_norm": 1.4055254459381104, "learning_rate": 2.6950000000000005e-05, "loss": 0.2238, "step": 540 }, { "grad_norm": 1.4610975980758667, "learning_rate": 2.7450000000000003e-05, "loss": 0.193, "step": 550 }, { "grad_norm": 1.4327552318572998, "learning_rate": 2.7950000000000005e-05, "loss": 0.1683, "step": 560 }, { "grad_norm": 1.877420425415039, "learning_rate": 2.845e-05, "loss": 0.1631, "step": 570 }, { "grad_norm": 1.8909727334976196, "learning_rate": 2.895e-05, "loss": 0.145, "step": 580 }, { "grad_norm": 1.5793254375457764, "learning_rate": 2.945e-05, "loss": 0.1288, "step": 590 }, { "grad_norm": 1.6723344326019287, "learning_rate": 2.995e-05, "loss": 0.1339, "step": 600 }, { "grad_norm": 1.4947915077209473, "learning_rate": 3.045e-05, "loss": 0.1207, "step": 610 }, { "grad_norm": 1.7394483089447021, "learning_rate": 3.095e-05, "loss": 0.1257, "step": 620 }, { "grad_norm": 1.872894048690796, "learning_rate": 3.145e-05, "loss": 0.1164, "step": 630 }, { "grad_norm": 1.547054409980774, "learning_rate": 3.1950000000000004e-05, "loss": 0.1115, "step": 640 }, { "grad_norm": 1.2903761863708496, "learning_rate": 3.245e-05, "loss": 0.1129, "step": 650 }, { "grad_norm": 1.5901747941970825, "learning_rate": 3.295e-05, "loss": 0.1144, "step": 660 }, { "grad_norm": 1.292582631111145, "learning_rate": 3.345000000000001e-05, "loss": 0.1075, "step": 670 }, { "grad_norm": 1.1780225038528442, "learning_rate": 3.3950000000000005e-05, "loss": 0.1019, "step": 680 }, { "grad_norm": 1.7617524862289429, "learning_rate": 3.445e-05, "loss": 0.1087, "step": 690 }, { "grad_norm": 1.6554834842681885, "learning_rate": 3.495e-05, "loss": 0.1026, "step": 700 }, { "grad_norm": 1.2230702638626099, "learning_rate": 3.545e-05, "loss": 0.1125, "step": 710 }, { "grad_norm": 1.3345097303390503, "learning_rate": 3.595e-05, "loss": 0.1065, "step": 720 }, { "grad_norm": 1.3718043565750122, "learning_rate": 3.645e-05, "loss": 0.0896, "step": 730 }, { "grad_norm": 1.4053155183792114, "learning_rate": 3.6950000000000004e-05, "loss": 0.1007, "step": 740 }, { "grad_norm": 1.7103720903396606, "learning_rate": 3.745e-05, "loss": 0.1046, "step": 750 }, { "grad_norm": 1.5401411056518555, "learning_rate": 3.795e-05, "loss": 0.108, "step": 760 }, { "grad_norm": 1.0559817552566528, "learning_rate": 3.845e-05, "loss": 0.0989, "step": 770 }, { "grad_norm": 1.4230653047561646, "learning_rate": 3.8950000000000005e-05, "loss": 0.0973, "step": 780 }, { "grad_norm": 1.3030253648757935, "learning_rate": 3.9450000000000003e-05, "loss": 0.0986, "step": 790 }, { "grad_norm": 1.2218207120895386, "learning_rate": 3.995e-05, "loss": 0.1023, "step": 800 }, { "grad_norm": 1.3930660486221313, "learning_rate": 4.045000000000001e-05, "loss": 0.0961, "step": 810 }, { "grad_norm": 1.18538236618042, "learning_rate": 4.095e-05, "loss": 0.0951, "step": 820 }, { "grad_norm": 1.191536784172058, "learning_rate": 4.145e-05, "loss": 0.0969, "step": 830 }, { "grad_norm": 1.3301697969436646, "learning_rate": 4.195e-05, "loss": 0.0936, "step": 840 }, { "grad_norm": 1.2634788751602173, "learning_rate": 4.245e-05, "loss": 0.0944, "step": 850 }, { "grad_norm": 1.2995100021362305, "learning_rate": 4.295e-05, "loss": 0.1015, "step": 860 }, { "grad_norm": 1.0792349576950073, "learning_rate": 4.345e-05, "loss": 0.0899, "step": 870 }, { "grad_norm": 1.097067952156067, "learning_rate": 4.3950000000000004e-05, "loss": 0.0967, "step": 880 }, { "grad_norm": 1.1697520017623901, "learning_rate": 4.445e-05, "loss": 0.0863, "step": 890 }, { "grad_norm": 1.3377461433410645, "learning_rate": 4.495e-05, "loss": 0.0807, "step": 900 }, { "grad_norm": 1.3604642152786255, "learning_rate": 4.545000000000001e-05, "loss": 0.0845, "step": 910 }, { "grad_norm": 0.8679919242858887, "learning_rate": 4.5950000000000006e-05, "loss": 0.0814, "step": 920 }, { "grad_norm": 1.1033315658569336, "learning_rate": 4.6450000000000004e-05, "loss": 0.0833, "step": 930 }, { "grad_norm": 1.2989084720611572, "learning_rate": 4.695e-05, "loss": 0.0805, "step": 940 }, { "grad_norm": 0.9687386155128479, "learning_rate": 4.745e-05, "loss": 0.0857, "step": 950 }, { "grad_norm": 1.2134044170379639, "learning_rate": 4.795e-05, "loss": 0.0857, "step": 960 }, { "grad_norm": 1.349519968032837, "learning_rate": 4.845e-05, "loss": 0.0875, "step": 970 }, { "grad_norm": 1.254042625427246, "learning_rate": 4.8950000000000004e-05, "loss": 0.0856, "step": 980 }, { "grad_norm": 0.938538670539856, "learning_rate": 4.945e-05, "loss": 0.0844, "step": 990 }, { "grad_norm": 1.322269320487976, "learning_rate": 4.995e-05, "loss": 0.0891, "step": 1000 }, { "grad_norm": 0.9776185154914856, "learning_rate": 5.045e-05, "loss": 0.0828, "step": 1010 }, { "grad_norm": 1.02939772605896, "learning_rate": 5.095e-05, "loss": 0.0823, "step": 1020 }, { "grad_norm": 0.9924076795578003, "learning_rate": 5.145e-05, "loss": 0.0851, "step": 1030 }, { "grad_norm": 0.9349141716957092, "learning_rate": 5.1949999999999996e-05, "loss": 0.0815, "step": 1040 }, { "grad_norm": 1.2007641792297363, "learning_rate": 5.245e-05, "loss": 0.075, "step": 1050 }, { "grad_norm": 1.0690460205078125, "learning_rate": 5.295e-05, "loss": 0.0764, "step": 1060 }, { "grad_norm": 1.04742431640625, "learning_rate": 5.345e-05, "loss": 0.0776, "step": 1070 }, { "grad_norm": 1.0899754762649536, "learning_rate": 5.3950000000000004e-05, "loss": 0.0716, "step": 1080 }, { "grad_norm": 1.238154649734497, "learning_rate": 5.445e-05, "loss": 0.0738, "step": 1090 }, { "grad_norm": 1.2086185216903687, "learning_rate": 5.495e-05, "loss": 0.0811, "step": 1100 }, { "grad_norm": 0.9972516894340515, "learning_rate": 5.545e-05, "loss": 0.072, "step": 1110 }, { "grad_norm": 1.0104039907455444, "learning_rate": 5.5950000000000005e-05, "loss": 0.0789, "step": 1120 }, { "grad_norm": 1.044769287109375, "learning_rate": 5.645e-05, "loss": 0.0778, "step": 1130 }, { "grad_norm": 0.9649087190628052, "learning_rate": 5.695e-05, "loss": 0.0769, "step": 1140 }, { "grad_norm": 0.9566044211387634, "learning_rate": 5.745e-05, "loss": 0.0796, "step": 1150 }, { "grad_norm": 1.0384883880615234, "learning_rate": 5.7950000000000006e-05, "loss": 0.0706, "step": 1160 }, { "grad_norm": 1.0224559307098389, "learning_rate": 5.8450000000000005e-05, "loss": 0.0757, "step": 1170 }, { "grad_norm": 1.1565721035003662, "learning_rate": 5.895e-05, "loss": 0.0765, "step": 1180 }, { "grad_norm": 0.8071246147155762, "learning_rate": 5.945000000000001e-05, "loss": 0.0722, "step": 1190 }, { "grad_norm": 0.8365389108657837, "learning_rate": 5.995000000000001e-05, "loss": 0.0776, "step": 1200 }, { "grad_norm": 0.973045289516449, "learning_rate": 6.0450000000000006e-05, "loss": 0.0727, "step": 1210 }, { "grad_norm": 0.8420519232749939, "learning_rate": 6.0950000000000004e-05, "loss": 0.0657, "step": 1220 }, { "grad_norm": 0.8094637989997864, "learning_rate": 6.145e-05, "loss": 0.0695, "step": 1230 }, { "grad_norm": 0.8192200660705566, "learning_rate": 6.195e-05, "loss": 0.0716, "step": 1240 }, { "grad_norm": 0.8153567910194397, "learning_rate": 6.245000000000001e-05, "loss": 0.0702, "step": 1250 }, { "grad_norm": 1.0208910703659058, "learning_rate": 6.295e-05, "loss": 0.0631, "step": 1260 }, { "grad_norm": 0.7788609266281128, "learning_rate": 6.345e-05, "loss": 0.0764, "step": 1270 }, { "grad_norm": 0.9219652414321899, "learning_rate": 6.395e-05, "loss": 0.069, "step": 1280 }, { "grad_norm": 0.8976600170135498, "learning_rate": 6.445e-05, "loss": 0.0707, "step": 1290 }, { "grad_norm": 0.8334199786186218, "learning_rate": 6.494999999999999e-05, "loss": 0.0684, "step": 1300 }, { "grad_norm": 0.8179826140403748, "learning_rate": 6.545e-05, "loss": 0.0628, "step": 1310 }, { "grad_norm": 0.921071469783783, "learning_rate": 6.595e-05, "loss": 0.0802, "step": 1320 }, { "grad_norm": 0.8027539253234863, "learning_rate": 6.645e-05, "loss": 0.0659, "step": 1330 }, { "grad_norm": 0.6647353172302246, "learning_rate": 6.695e-05, "loss": 0.0572, "step": 1340 }, { "grad_norm": 1.1116269826889038, "learning_rate": 6.745e-05, "loss": 0.0677, "step": 1350 }, { "grad_norm": 1.0260670185089111, "learning_rate": 6.795e-05, "loss": 0.0721, "step": 1360 }, { "grad_norm": 0.6854798197746277, "learning_rate": 6.845e-05, "loss": 0.0637, "step": 1370 }, { "grad_norm": 0.723114013671875, "learning_rate": 6.895000000000001e-05, "loss": 0.061, "step": 1380 }, { "grad_norm": 0.8218165636062622, "learning_rate": 6.945000000000001e-05, "loss": 0.0615, "step": 1390 }, { "grad_norm": 0.7412631511688232, "learning_rate": 6.995e-05, "loss": 0.0601, "step": 1400 }, { "grad_norm": 1.0243332386016846, "learning_rate": 7.045e-05, "loss": 0.0589, "step": 1410 }, { "grad_norm": 0.6480699181556702, "learning_rate": 7.095e-05, "loss": 0.0618, "step": 1420 }, { "grad_norm": 0.7620859146118164, "learning_rate": 7.145e-05, "loss": 0.0717, "step": 1430 }, { "grad_norm": 0.9057062268257141, "learning_rate": 7.195e-05, "loss": 0.0611, "step": 1440 }, { "grad_norm": 0.7770098447799683, "learning_rate": 7.245000000000001e-05, "loss": 0.0602, "step": 1450 }, { "grad_norm": 0.8656067848205566, "learning_rate": 7.295000000000001e-05, "loss": 0.0665, "step": 1460 }, { "grad_norm": 0.8000733256340027, "learning_rate": 7.345000000000001e-05, "loss": 0.0569, "step": 1470 }, { "grad_norm": 0.6571788787841797, "learning_rate": 7.395000000000001e-05, "loss": 0.0545, "step": 1480 }, { "grad_norm": 0.7306585907936096, "learning_rate": 7.445000000000001e-05, "loss": 0.0621, "step": 1490 }, { "grad_norm": 0.6538251042366028, "learning_rate": 7.495e-05, "loss": 0.0577, "step": 1500 }, { "grad_norm": 0.7312050461769104, "learning_rate": 7.545e-05, "loss": 0.0609, "step": 1510 }, { "grad_norm": 0.746128261089325, "learning_rate": 7.595e-05, "loss": 0.0599, "step": 1520 }, { "grad_norm": 0.5966893434524536, "learning_rate": 7.645e-05, "loss": 0.0602, "step": 1530 }, { "grad_norm": 0.7012491822242737, "learning_rate": 7.695e-05, "loss": 0.0614, "step": 1540 }, { "grad_norm": 0.5652503967285156, "learning_rate": 7.745e-05, "loss": 0.0572, "step": 1550 }, { "grad_norm": 0.6395028233528137, "learning_rate": 7.795e-05, "loss": 0.0606, "step": 1560 }, { "grad_norm": 0.8299902677536011, "learning_rate": 7.845e-05, "loss": 0.0625, "step": 1570 }, { "grad_norm": 0.7250336408615112, "learning_rate": 7.895000000000001e-05, "loss": 0.0554, "step": 1580 }, { "grad_norm": 0.8201026916503906, "learning_rate": 7.945e-05, "loss": 0.056, "step": 1590 }, { "grad_norm": 0.636671781539917, "learning_rate": 7.995e-05, "loss": 0.0632, "step": 1600 }, { "grad_norm": 0.6508738398551941, "learning_rate": 8.045e-05, "loss": 0.0568, "step": 1610 }, { "grad_norm": 0.8877677321434021, "learning_rate": 8.095e-05, "loss": 0.0552, "step": 1620 }, { "grad_norm": 0.6917802095413208, "learning_rate": 8.145e-05, "loss": 0.0592, "step": 1630 }, { "grad_norm": 0.7174184918403625, "learning_rate": 8.195e-05, "loss": 0.0584, "step": 1640 }, { "grad_norm": 0.6945115327835083, "learning_rate": 8.245e-05, "loss": 0.0576, "step": 1650 }, { "grad_norm": 0.5988478660583496, "learning_rate": 8.295000000000001e-05, "loss": 0.056, "step": 1660 }, { "grad_norm": 0.7594149708747864, "learning_rate": 8.345000000000001e-05, "loss": 0.0611, "step": 1670 }, { "grad_norm": 0.6970542073249817, "learning_rate": 8.395000000000001e-05, "loss": 0.057, "step": 1680 }, { "grad_norm": 0.6646000742912292, "learning_rate": 8.445e-05, "loss": 0.0548, "step": 1690 }, { "grad_norm": 0.8323544263839722, "learning_rate": 8.495e-05, "loss": 0.0647, "step": 1700 }, { "grad_norm": 0.5609958171844482, "learning_rate": 8.545e-05, "loss": 0.0578, "step": 1710 }, { "grad_norm": 0.7728407382965088, "learning_rate": 8.595e-05, "loss": 0.0581, "step": 1720 }, { "grad_norm": 0.7541513442993164, "learning_rate": 8.645000000000001e-05, "loss": 0.0583, "step": 1730 }, { "grad_norm": 0.5601022243499756, "learning_rate": 8.695000000000001e-05, "loss": 0.0556, "step": 1740 }, { "grad_norm": 0.7432585954666138, "learning_rate": 8.745000000000001e-05, "loss": 0.0524, "step": 1750 }, { "grad_norm": 0.6351825594902039, "learning_rate": 8.795e-05, "loss": 0.0531, "step": 1760 }, { "grad_norm": 0.5829357504844666, "learning_rate": 8.845e-05, "loss": 0.0679, "step": 1770 }, { "grad_norm": 0.5286015272140503, "learning_rate": 8.895e-05, "loss": 0.0583, "step": 1780 }, { "grad_norm": 0.6164659857749939, "learning_rate": 8.945e-05, "loss": 0.0567, "step": 1790 }, { "grad_norm": 0.7461672425270081, "learning_rate": 8.995e-05, "loss": 0.0629, "step": 1800 }, { "grad_norm": 0.43249624967575073, "learning_rate": 9.045e-05, "loss": 0.056, "step": 1810 }, { "grad_norm": 0.7155604362487793, "learning_rate": 9.095e-05, "loss": 0.0551, "step": 1820 }, { "grad_norm": 0.5769962668418884, "learning_rate": 9.145e-05, "loss": 0.0533, "step": 1830 }, { "grad_norm": 0.6373982429504395, "learning_rate": 9.195e-05, "loss": 0.0525, "step": 1840 }, { "grad_norm": 0.747323215007782, "learning_rate": 9.245e-05, "loss": 0.0532, "step": 1850 }, { "grad_norm": 0.6578043699264526, "learning_rate": 9.295000000000001e-05, "loss": 0.0529, "step": 1860 }, { "grad_norm": 0.5416834950447083, "learning_rate": 9.345000000000001e-05, "loss": 0.0534, "step": 1870 }, { "grad_norm": 0.5876487493515015, "learning_rate": 9.395000000000001e-05, "loss": 0.0518, "step": 1880 }, { "grad_norm": 0.4869372546672821, "learning_rate": 9.445e-05, "loss": 0.0488, "step": 1890 }, { "grad_norm": 0.6166200041770935, "learning_rate": 9.495e-05, "loss": 0.0535, "step": 1900 }, { "grad_norm": 0.6396713256835938, "learning_rate": 9.545e-05, "loss": 0.0553, "step": 1910 }, { "grad_norm": 0.5057575702667236, "learning_rate": 9.595e-05, "loss": 0.0464, "step": 1920 }, { "grad_norm": 0.6108220815658569, "learning_rate": 9.645000000000001e-05, "loss": 0.0503, "step": 1930 }, { "grad_norm": 0.6525489091873169, "learning_rate": 9.695000000000001e-05, "loss": 0.062, "step": 1940 }, { "grad_norm": 0.5705341696739197, "learning_rate": 9.745000000000001e-05, "loss": 0.0514, "step": 1950 }, { "grad_norm": 0.6801689267158508, "learning_rate": 9.795000000000001e-05, "loss": 0.0515, "step": 1960 }, { "grad_norm": 0.5143327713012695, "learning_rate": 9.845000000000001e-05, "loss": 0.053, "step": 1970 }, { "grad_norm": 0.5779087543487549, "learning_rate": 9.895e-05, "loss": 0.0588, "step": 1980 }, { "grad_norm": 0.5929870009422302, "learning_rate": 9.945e-05, "loss": 0.0507, "step": 1990 }, { "grad_norm": 0.6456423997879028, "learning_rate": 9.995e-05, "loss": 0.0512, "step": 2000 }, { "grad_norm": 0.7070780396461487, "learning_rate": 9.99999861593158e-05, "loss": 0.0518, "step": 2010 }, { "grad_norm": 0.66335529088974, "learning_rate": 9.999993831498517e-05, "loss": 0.0543, "step": 2020 }, { "grad_norm": 0.465688019990921, "learning_rate": 9.99998562961682e-05, "loss": 0.0487, "step": 2030 }, { "grad_norm": 0.5122871398925781, "learning_rate": 9.99997401029209e-05, "loss": 0.0543, "step": 2040 }, { "grad_norm": 0.7813759446144104, "learning_rate": 9.999958973532271e-05, "loss": 0.0566, "step": 2050 }, { "grad_norm": 0.6312514543533325, "learning_rate": 9.999940519347642e-05, "loss": 0.0547, "step": 2060 }, { "grad_norm": 0.7885079383850098, "learning_rate": 9.999918647750812e-05, "loss": 0.0524, "step": 2070 }, { "grad_norm": 0.5789466500282288, "learning_rate": 9.999893358756736e-05, "loss": 0.0547, "step": 2080 }, { "grad_norm": 0.5862072706222534, "learning_rate": 9.999864652382692e-05, "loss": 0.0547, "step": 2090 }, { "grad_norm": 0.7843810319900513, "learning_rate": 9.999832528648307e-05, "loss": 0.052, "step": 2100 }, { "grad_norm": 0.5146076679229736, "learning_rate": 9.999796987575532e-05, "loss": 0.0523, "step": 2110 }, { "grad_norm": 0.5503524541854858, "learning_rate": 9.999758029188662e-05, "loss": 0.0505, "step": 2120 }, { "grad_norm": 0.5573360919952393, "learning_rate": 9.999715653514324e-05, "loss": 0.0532, "step": 2130 }, { "grad_norm": 0.5507333874702454, "learning_rate": 9.999669860581482e-05, "loss": 0.0494, "step": 2140 }, { "grad_norm": 0.6182554364204407, "learning_rate": 9.999620650421433e-05, "loss": 0.0522, "step": 2150 }, { "grad_norm": 0.4295720160007477, "learning_rate": 9.999568023067813e-05, "loss": 0.0421, "step": 2160 }, { "grad_norm": 0.5351302027702332, "learning_rate": 9.999511978556595e-05, "loss": 0.0504, "step": 2170 }, { "grad_norm": 0.5350063443183899, "learning_rate": 9.99945251692608e-05, "loss": 0.0491, "step": 2180 }, { "grad_norm": 0.4812272787094116, "learning_rate": 9.999389638216912e-05, "loss": 0.0479, "step": 2190 }, { "grad_norm": 0.4307490587234497, "learning_rate": 9.999323342472068e-05, "loss": 0.0452, "step": 2200 }, { "grad_norm": 0.5751038789749146, "learning_rate": 9.99925362973686e-05, "loss": 0.0465, "step": 2210 }, { "grad_norm": 0.5207296013832092, "learning_rate": 9.999180500058935e-05, "loss": 0.0477, "step": 2220 }, { "grad_norm": 0.412018746137619, "learning_rate": 9.99910395348828e-05, "loss": 0.0454, "step": 2230 }, { "grad_norm": 0.45195141434669495, "learning_rate": 9.99902399007721e-05, "loss": 0.0496, "step": 2240 }, { "grad_norm": 0.5704726576805115, "learning_rate": 9.99894060988038e-05, "loss": 0.0572, "step": 2250 }, { "grad_norm": 0.4141075909137726, "learning_rate": 9.998853812954783e-05, "loss": 0.0564, "step": 2260 }, { "grad_norm": 0.5578204393386841, "learning_rate": 9.998763599359739e-05, "loss": 0.0571, "step": 2270 }, { "grad_norm": 0.6391753554344177, "learning_rate": 9.998669969156912e-05, "loss": 0.049, "step": 2280 }, { "grad_norm": 0.4201097786426544, "learning_rate": 9.998572922410294e-05, "loss": 0.0439, "step": 2290 }, { "grad_norm": 0.5018947720527649, "learning_rate": 9.998472459186219e-05, "loss": 0.0517, "step": 2300 }, { "grad_norm": 0.46052926778793335, "learning_rate": 9.99836857955335e-05, "loss": 0.0491, "step": 2310 }, { "grad_norm": 0.5522165894508362, "learning_rate": 9.998261283582688e-05, "loss": 0.0477, "step": 2320 }, { "grad_norm": 0.5278041362762451, "learning_rate": 9.998150571347571e-05, "loss": 0.0536, "step": 2330 }, { "grad_norm": 0.49618780612945557, "learning_rate": 9.998036442923667e-05, "loss": 0.0495, "step": 2340 }, { "grad_norm": 0.706361711025238, "learning_rate": 9.997918898388983e-05, "loss": 0.0476, "step": 2350 }, { "grad_norm": 0.5459227561950684, "learning_rate": 9.997797937823859e-05, "loss": 0.0528, "step": 2360 }, { "grad_norm": 0.41452929377555847, "learning_rate": 9.997673561310971e-05, "loss": 0.0509, "step": 2370 }, { "grad_norm": 0.6120489835739136, "learning_rate": 9.997545768935331e-05, "loss": 0.0494, "step": 2380 }, { "grad_norm": 0.5120590329170227, "learning_rate": 9.997414560784282e-05, "loss": 0.0498, "step": 2390 }, { "grad_norm": 0.5297312140464783, "learning_rate": 9.997279936947502e-05, "loss": 0.048, "step": 2400 }, { "grad_norm": 0.4149268865585327, "learning_rate": 9.997141897517006e-05, "loss": 0.0438, "step": 2410 }, { "grad_norm": 0.48027360439300537, "learning_rate": 9.997000442587145e-05, "loss": 0.0477, "step": 2420 }, { "grad_norm": 0.5994620323181152, "learning_rate": 9.9968555722546e-05, "loss": 0.0489, "step": 2430 }, { "grad_norm": 0.4365735948085785, "learning_rate": 9.996707286618387e-05, "loss": 0.0443, "step": 2440 }, { "grad_norm": 0.49626806378364563, "learning_rate": 9.996555585779861e-05, "loss": 0.0467, "step": 2450 }, { "grad_norm": 0.5099051594734192, "learning_rate": 9.996400469842708e-05, "loss": 0.0507, "step": 2460 }, { "grad_norm": 0.5789559483528137, "learning_rate": 9.996241938912945e-05, "loss": 0.0438, "step": 2470 }, { "grad_norm": 0.48177093267440796, "learning_rate": 9.99607999309893e-05, "loss": 0.0516, "step": 2480 }, { "grad_norm": 0.5346314311027527, "learning_rate": 9.99591463251135e-05, "loss": 0.0442, "step": 2490 }, { "grad_norm": 0.46124735474586487, "learning_rate": 9.995745857263226e-05, "loss": 0.0486, "step": 2500 }, { "grad_norm": 0.6765794157981873, "learning_rate": 9.995573667469918e-05, "loss": 0.0487, "step": 2510 }, { "grad_norm": 0.4950980246067047, "learning_rate": 9.995398063249111e-05, "loss": 0.0468, "step": 2520 }, { "grad_norm": 0.5047560930252075, "learning_rate": 9.995219044720833e-05, "loss": 0.0445, "step": 2530 }, { "grad_norm": 0.4477350413799286, "learning_rate": 9.99503661200744e-05, "loss": 0.0445, "step": 2540 }, { "grad_norm": 0.4268946945667267, "learning_rate": 9.994850765233624e-05, "loss": 0.0509, "step": 2550 }, { "grad_norm": 0.4856688678264618, "learning_rate": 9.994661504526406e-05, "loss": 0.0434, "step": 2560 }, { "grad_norm": 0.570576548576355, "learning_rate": 9.994468830015148e-05, "loss": 0.0437, "step": 2570 }, { "grad_norm": 0.41368409991264343, "learning_rate": 9.99427274183154e-05, "loss": 0.0478, "step": 2580 }, { "grad_norm": 0.5697697401046753, "learning_rate": 9.994073240109606e-05, "loss": 0.041, "step": 2590 }, { "grad_norm": 0.522297739982605, "learning_rate": 9.993870324985703e-05, "loss": 0.0489, "step": 2600 }, { "grad_norm": 0.5285522937774658, "learning_rate": 9.993663996598523e-05, "loss": 0.0476, "step": 2610 }, { "grad_norm": 0.4956802427768707, "learning_rate": 9.993454255089089e-05, "loss": 0.0454, "step": 2620 }, { "grad_norm": 0.47710078954696655, "learning_rate": 9.993241100600757e-05, "loss": 0.0507, "step": 2630 }, { "grad_norm": 0.3498680293560028, "learning_rate": 9.993024533279215e-05, "loss": 0.046, "step": 2640 }, { "grad_norm": 0.4536656439304352, "learning_rate": 9.992804553272487e-05, "loss": 0.0467, "step": 2650 }, { "grad_norm": 0.49920976161956787, "learning_rate": 9.992581160730926e-05, "loss": 0.0459, "step": 2660 }, { "grad_norm": 0.3306896686553955, "learning_rate": 9.992354355807218e-05, "loss": 0.0478, "step": 2670 }, { "grad_norm": 0.4578983783721924, "learning_rate": 9.992124138656385e-05, "loss": 0.0422, "step": 2680 }, { "grad_norm": 0.4781835675239563, "learning_rate": 9.991890509435774e-05, "loss": 0.0471, "step": 2690 }, { "grad_norm": 0.6572957038879395, "learning_rate": 9.991653468305071e-05, "loss": 0.0454, "step": 2700 }, { "grad_norm": 0.5793191194534302, "learning_rate": 9.991413015426292e-05, "loss": 0.0501, "step": 2710 }, { "grad_norm": 0.5006295442581177, "learning_rate": 9.991169150963781e-05, "loss": 0.0435, "step": 2720 }, { "grad_norm": 0.3858761191368103, "learning_rate": 9.990921875084221e-05, "loss": 0.0474, "step": 2730 }, { "grad_norm": 0.4165707230567932, "learning_rate": 9.990671187956622e-05, "loss": 0.0472, "step": 2740 }, { "grad_norm": 0.4132266640663147, "learning_rate": 9.990417089752324e-05, "loss": 0.0432, "step": 2750 }, { "grad_norm": 0.5843473076820374, "learning_rate": 9.990159580645001e-05, "loss": 0.0461, "step": 2760 }, { "grad_norm": 0.4210019111633301, "learning_rate": 9.98989866081066e-05, "loss": 0.0399, "step": 2770 }, { "grad_norm": 0.4775783121585846, "learning_rate": 9.989634330427636e-05, "loss": 0.0455, "step": 2780 }, { "grad_norm": 0.3904673755168915, "learning_rate": 9.989366589676596e-05, "loss": 0.045, "step": 2790 }, { "grad_norm": 0.44643792510032654, "learning_rate": 9.98909543874054e-05, "loss": 0.0399, "step": 2800 }, { "grad_norm": 0.55136638879776, "learning_rate": 9.988820877804795e-05, "loss": 0.0413, "step": 2810 }, { "grad_norm": 0.5130012035369873, "learning_rate": 9.988542907057022e-05, "loss": 0.0391, "step": 2820 }, { "grad_norm": 0.40746212005615234, "learning_rate": 9.988261526687212e-05, "loss": 0.0467, "step": 2830 }, { "grad_norm": 0.4178461730480194, "learning_rate": 9.987976736887685e-05, "loss": 0.0442, "step": 2840 }, { "grad_norm": 0.4563087224960327, "learning_rate": 9.987688537853091e-05, "loss": 0.0396, "step": 2850 }, { "grad_norm": 0.4131450951099396, "learning_rate": 9.987396929780413e-05, "loss": 0.0409, "step": 2860 }, { "grad_norm": 0.4392145872116089, "learning_rate": 9.987101912868962e-05, "loss": 0.0431, "step": 2870 }, { "grad_norm": 0.3376072943210602, "learning_rate": 9.98680348732038e-05, "loss": 0.0409, "step": 2880 }, { "grad_norm": 0.3986402153968811, "learning_rate": 9.986501653338636e-05, "loss": 0.0385, "step": 2890 }, { "grad_norm": 0.6861705780029297, "learning_rate": 9.986196411130031e-05, "loss": 0.0416, "step": 2900 }, { "grad_norm": 0.40738287568092346, "learning_rate": 9.985887760903197e-05, "loss": 0.042, "step": 2910 }, { "grad_norm": 0.4791441559791565, "learning_rate": 9.985575702869093e-05, "loss": 0.0386, "step": 2920 }, { "grad_norm": 0.44263607263565063, "learning_rate": 9.985260237241008e-05, "loss": 0.0456, "step": 2930 }, { "grad_norm": 0.4556688368320465, "learning_rate": 9.984941364234557e-05, "loss": 0.0486, "step": 2940 }, { "grad_norm": 0.3512797951698303, "learning_rate": 9.98461908406769e-05, "loss": 0.0465, "step": 2950 }, { "grad_norm": 0.45595303177833557, "learning_rate": 9.98429339696068e-05, "loss": 0.0501, "step": 2960 }, { "grad_norm": 0.5537407994270325, "learning_rate": 9.983964303136133e-05, "loss": 0.0437, "step": 2970 }, { "grad_norm": 0.37345680594444275, "learning_rate": 9.983631802818981e-05, "loss": 0.0435, "step": 2980 }, { "grad_norm": 0.35033726692199707, "learning_rate": 9.983295896236484e-05, "loss": 0.0401, "step": 2990 }, { "grad_norm": 0.35160499811172485, "learning_rate": 9.982956583618232e-05, "loss": 0.0422, "step": 3000 }, { "grad_norm": 0.42489761114120483, "learning_rate": 9.982613865196142e-05, "loss": 0.0454, "step": 3010 }, { "grad_norm": 0.5230557322502136, "learning_rate": 9.982267741204458e-05, "loss": 0.042, "step": 3020 }, { "grad_norm": 0.45287561416625977, "learning_rate": 9.981918211879753e-05, "loss": 0.0387, "step": 3030 }, { "grad_norm": 0.46841591596603394, "learning_rate": 9.981565277460927e-05, "loss": 0.0371, "step": 3040 }, { "grad_norm": 0.4899894893169403, "learning_rate": 9.981208938189206e-05, "loss": 0.0402, "step": 3050 }, { "grad_norm": 0.42603999376296997, "learning_rate": 9.980849194308149e-05, "loss": 0.0413, "step": 3060 }, { "grad_norm": 0.468386709690094, "learning_rate": 9.98048604606363e-05, "loss": 0.0426, "step": 3070 }, { "grad_norm": 0.45652076601982117, "learning_rate": 9.980119493703864e-05, "loss": 0.0422, "step": 3080 }, { "grad_norm": 0.5272055864334106, "learning_rate": 9.979749537479383e-05, "loss": 0.0412, "step": 3090 }, { "grad_norm": 0.42237406969070435, "learning_rate": 9.979376177643051e-05, "loss": 0.0441, "step": 3100 }, { "grad_norm": 0.43355509638786316, "learning_rate": 9.978999414450052e-05, "loss": 0.0469, "step": 3110 }, { "grad_norm": 0.38083332777023315, "learning_rate": 9.978619248157904e-05, "loss": 0.0372, "step": 3120 }, { "grad_norm": 0.4129858613014221, "learning_rate": 9.978235679026445e-05, "loss": 0.0484, "step": 3130 }, { "grad_norm": 0.47952157258987427, "learning_rate": 9.97784870731784e-05, "loss": 0.0391, "step": 3140 }, { "grad_norm": 0.3756651282310486, "learning_rate": 9.977458333296584e-05, "loss": 0.042, "step": 3150 }, { "grad_norm": 0.3136999011039734, "learning_rate": 9.977064557229492e-05, "loss": 0.0346, "step": 3160 }, { "grad_norm": 0.3214828372001648, "learning_rate": 9.976667379385705e-05, "loss": 0.0394, "step": 3170 }, { "grad_norm": 0.41712409257888794, "learning_rate": 9.97626680003669e-05, "loss": 0.0412, "step": 3180 }, { "grad_norm": 0.4360157251358032, "learning_rate": 9.975862819456242e-05, "loss": 0.0433, "step": 3190 }, { "grad_norm": 0.4498880207538605, "learning_rate": 9.975455437920477e-05, "loss": 0.0397, "step": 3200 }, { "grad_norm": 0.40135815739631653, "learning_rate": 9.975044655707834e-05, "loss": 0.0385, "step": 3210 }, { "grad_norm": 0.4928305149078369, "learning_rate": 9.974630473099082e-05, "loss": 0.0428, "step": 3220 }, { "grad_norm": 0.38201719522476196, "learning_rate": 9.974212890377311e-05, "loss": 0.0475, "step": 3230 }, { "grad_norm": 0.41888877749443054, "learning_rate": 9.973791907827931e-05, "loss": 0.0405, "step": 3240 }, { "grad_norm": 0.490866482257843, "learning_rate": 9.973367525738683e-05, "loss": 0.0405, "step": 3250 }, { "grad_norm": 0.47912517189979553, "learning_rate": 9.972939744399627e-05, "loss": 0.0448, "step": 3260 }, { "grad_norm": 0.45648565888404846, "learning_rate": 9.97250856410315e-05, "loss": 0.0426, "step": 3270 }, { "grad_norm": 0.3920406699180603, "learning_rate": 9.972073985143955e-05, "loss": 0.0387, "step": 3280 }, { "grad_norm": 0.4068121016025543, "learning_rate": 9.971636007819074e-05, "loss": 0.0412, "step": 3290 }, { "grad_norm": 0.3598281145095825, "learning_rate": 9.971194632427863e-05, "loss": 0.042, "step": 3300 }, { "grad_norm": 0.48422661423683167, "learning_rate": 9.970749859271997e-05, "loss": 0.0416, "step": 3310 }, { "grad_norm": 0.42204749584198, "learning_rate": 9.970301688655473e-05, "loss": 0.0453, "step": 3320 }, { "grad_norm": 0.3968379497528076, "learning_rate": 9.969850120884612e-05, "loss": 0.0399, "step": 3330 }, { "grad_norm": 0.3978451192378998, "learning_rate": 9.969395156268055e-05, "loss": 0.0406, "step": 3340 }, { "grad_norm": 0.34641149640083313, "learning_rate": 9.968936795116768e-05, "loss": 0.0404, "step": 3350 }, { "grad_norm": 0.37458378076553345, "learning_rate": 9.968475037744036e-05, "loss": 0.0367, "step": 3360 }, { "grad_norm": 0.37582969665527344, "learning_rate": 9.968009884465465e-05, "loss": 0.0393, "step": 3370 }, { "grad_norm": 0.40783798694610596, "learning_rate": 9.967541335598984e-05, "loss": 0.0365, "step": 3380 }, { "grad_norm": 0.2936830520629883, "learning_rate": 9.967069391464841e-05, "loss": 0.0406, "step": 3390 }, { "grad_norm": 0.34152132272720337, "learning_rate": 9.966594052385608e-05, "loss": 0.0415, "step": 3400 }, { "grad_norm": 0.4632757604122162, "learning_rate": 9.96611531868617e-05, "loss": 0.0454, "step": 3410 }, { "grad_norm": 0.4160197377204895, "learning_rate": 9.96563319069374e-05, "loss": 0.0396, "step": 3420 }, { "grad_norm": 0.3490140438079834, "learning_rate": 9.965147668737847e-05, "loss": 0.0365, "step": 3430 }, { "grad_norm": 0.5444722175598145, "learning_rate": 9.96465875315034e-05, "loss": 0.0436, "step": 3440 }, { "grad_norm": 0.3627833127975464, "learning_rate": 9.96416644426539e-05, "loss": 0.0354, "step": 3450 }, { "grad_norm": 0.38494232296943665, "learning_rate": 9.963670742419485e-05, "loss": 0.0375, "step": 3460 }, { "grad_norm": 0.3267291486263275, "learning_rate": 9.96317164795143e-05, "loss": 0.0427, "step": 3470 }, { "grad_norm": 0.39965057373046875, "learning_rate": 9.962669161202356e-05, "loss": 0.0384, "step": 3480 }, { "grad_norm": 0.3360757827758789, "learning_rate": 9.962163282515705e-05, "loss": 0.0383, "step": 3490 }, { "grad_norm": 0.36845192313194275, "learning_rate": 9.961654012237241e-05, "loss": 0.035, "step": 3500 }, { "grad_norm": 0.40991878509521484, "learning_rate": 9.961141350715044e-05, "loss": 0.0424, "step": 3510 }, { "grad_norm": 0.4481619894504547, "learning_rate": 9.960625298299516e-05, "loss": 0.0382, "step": 3520 }, { "grad_norm": 0.4100955128669739, "learning_rate": 9.960105855343372e-05, "loss": 0.0399, "step": 3530 }, { "grad_norm": 0.4238847494125366, "learning_rate": 9.959583022201647e-05, "loss": 0.038, "step": 3540 }, { "grad_norm": 0.3574022054672241, "learning_rate": 9.959056799231692e-05, "loss": 0.0392, "step": 3550 }, { "grad_norm": 0.4979078769683838, "learning_rate": 9.958527186793176e-05, "loss": 0.0402, "step": 3560 }, { "grad_norm": 0.3765380382537842, "learning_rate": 9.957994185248086e-05, "loss": 0.0406, "step": 3570 }, { "grad_norm": 0.35520312190055847, "learning_rate": 9.957457794960718e-05, "loss": 0.0394, "step": 3580 }, { "grad_norm": 0.42091912031173706, "learning_rate": 9.956918016297694e-05, "loss": 0.0358, "step": 3590 }, { "grad_norm": 0.4820012152194977, "learning_rate": 9.956374849627948e-05, "loss": 0.0388, "step": 3600 }, { "grad_norm": 0.3668622672557831, "learning_rate": 9.955828295322728e-05, "loss": 0.0369, "step": 3610 }, { "grad_norm": 0.48550260066986084, "learning_rate": 9.955278353755598e-05, "loss": 0.0369, "step": 3620 }, { "grad_norm": 0.3599294424057007, "learning_rate": 9.954725025302439e-05, "loss": 0.0411, "step": 3630 }, { "grad_norm": 0.409374862909317, "learning_rate": 9.954168310341445e-05, "loss": 0.0358, "step": 3640 }, { "grad_norm": 0.35385993123054504, "learning_rate": 9.953608209253126e-05, "loss": 0.0346, "step": 3650 }, { "grad_norm": 0.3876805603504181, "learning_rate": 9.953044722420307e-05, "loss": 0.0352, "step": 3660 }, { "grad_norm": 0.46755698323249817, "learning_rate": 9.952477850228124e-05, "loss": 0.0351, "step": 3670 }, { "grad_norm": 0.33069273829460144, "learning_rate": 9.95190759306403e-05, "loss": 0.0352, "step": 3680 }, { "grad_norm": 0.35191816091537476, "learning_rate": 9.951333951317789e-05, "loss": 0.0432, "step": 3690 }, { "grad_norm": 0.4785519242286682, "learning_rate": 9.950756925381479e-05, "loss": 0.0342, "step": 3700 }, { "grad_norm": 0.3359984755516052, "learning_rate": 9.950176515649496e-05, "loss": 0.0355, "step": 3710 }, { "grad_norm": 0.3901040554046631, "learning_rate": 9.949592722518542e-05, "loss": 0.0368, "step": 3720 }, { "grad_norm": 0.3346503674983978, "learning_rate": 9.949005546387631e-05, "loss": 0.0387, "step": 3730 }, { "grad_norm": 0.31976813077926636, "learning_rate": 9.948414987658098e-05, "loss": 0.0419, "step": 3740 }, { "grad_norm": 0.44033119082450867, "learning_rate": 9.94782104673358e-05, "loss": 0.038, "step": 3750 }, { "grad_norm": 0.3878341019153595, "learning_rate": 9.947223724020034e-05, "loss": 0.0393, "step": 3760 }, { "grad_norm": 0.37310558557510376, "learning_rate": 9.94662301992572e-05, "loss": 0.0316, "step": 3770 }, { "grad_norm": 0.40719032287597656, "learning_rate": 9.946018934861216e-05, "loss": 0.0395, "step": 3780 }, { "grad_norm": 0.34950563311576843, "learning_rate": 9.945411469239409e-05, "loss": 0.0353, "step": 3790 }, { "grad_norm": 0.30794206261634827, "learning_rate": 9.944800623475497e-05, "loss": 0.0387, "step": 3800 }, { "grad_norm": 0.4232628345489502, "learning_rate": 9.944186397986984e-05, "loss": 0.042, "step": 3810 }, { "grad_norm": 0.4690187871456146, "learning_rate": 9.943568793193687e-05, "loss": 0.0435, "step": 3820 }, { "grad_norm": 0.3654971420764923, "learning_rate": 9.942947809517737e-05, "loss": 0.0366, "step": 3830 }, { "grad_norm": 0.2735729515552521, "learning_rate": 9.942323447383568e-05, "loss": 0.0378, "step": 3840 }, { "grad_norm": 0.38253942131996155, "learning_rate": 9.941695707217925e-05, "loss": 0.0392, "step": 3850 }, { "grad_norm": 0.33201614022254944, "learning_rate": 9.941064589449865e-05, "loss": 0.0387, "step": 3860 }, { "grad_norm": 0.3712804317474365, "learning_rate": 9.940430094510748e-05, "loss": 0.0341, "step": 3870 }, { "grad_norm": 0.46236592531204224, "learning_rate": 9.939792222834248e-05, "loss": 0.0353, "step": 3880 }, { "grad_norm": 0.5072771906852722, "learning_rate": 9.939150974856343e-05, "loss": 0.0404, "step": 3890 }, { "grad_norm": 0.420412540435791, "learning_rate": 9.93850635101532e-05, "loss": 0.0436, "step": 3900 }, { "grad_norm": 0.4533492624759674, "learning_rate": 9.937858351751775e-05, "loss": 0.0391, "step": 3910 }, { "grad_norm": 0.4692460894584656, "learning_rate": 9.937206977508604e-05, "loss": 0.044, "step": 3920 }, { "grad_norm": 0.33271098136901855, "learning_rate": 9.936552228731022e-05, "loss": 0.0344, "step": 3930 }, { "grad_norm": 0.3592855632305145, "learning_rate": 9.93589410586654e-05, "loss": 0.0391, "step": 3940 }, { "grad_norm": 0.4104155898094177, "learning_rate": 9.93523260936498e-05, "loss": 0.0365, "step": 3950 }, { "grad_norm": 0.2788482904434204, "learning_rate": 9.934567739678467e-05, "loss": 0.0363, "step": 3960 }, { "grad_norm": 0.3489533066749573, "learning_rate": 9.933899497261433e-05, "loss": 0.0327, "step": 3970 }, { "grad_norm": 0.3011980354785919, "learning_rate": 9.933227882570617e-05, "loss": 0.0336, "step": 3980 }, { "grad_norm": 0.3700128495693207, "learning_rate": 9.932552896065062e-05, "loss": 0.0364, "step": 3990 }, { "grad_norm": 0.3067794442176819, "learning_rate": 9.931874538206114e-05, "loss": 0.0399, "step": 4000 }, { "grad_norm": 0.3326687216758728, "learning_rate": 9.931192809457423e-05, "loss": 0.0337, "step": 4010 }, { "grad_norm": 0.37433281540870667, "learning_rate": 9.930507710284944e-05, "loss": 0.0338, "step": 4020 }, { "grad_norm": 0.40123194456100464, "learning_rate": 9.92981924115694e-05, "loss": 0.0384, "step": 4030 }, { "grad_norm": 0.3712434470653534, "learning_rate": 9.929127402543968e-05, "loss": 0.0346, "step": 4040 }, { "grad_norm": 0.5006332397460938, "learning_rate": 9.928432194918895e-05, "loss": 0.0358, "step": 4050 }, { "grad_norm": 0.32509252429008484, "learning_rate": 9.927733618756889e-05, "loss": 0.0364, "step": 4060 }, { "grad_norm": 0.42753684520721436, "learning_rate": 9.92703167453542e-05, "loss": 0.0309, "step": 4070 }, { "grad_norm": 0.46692103147506714, "learning_rate": 9.926326362734263e-05, "loss": 0.0355, "step": 4080 }, { "grad_norm": 0.3932330310344696, "learning_rate": 9.925617683835489e-05, "loss": 0.0343, "step": 4090 }, { "grad_norm": 0.3205368220806122, "learning_rate": 9.924905638323472e-05, "loss": 0.0325, "step": 4100 }, { "grad_norm": 0.35560742020606995, "learning_rate": 9.92419022668489e-05, "loss": 0.0389, "step": 4110 }, { "grad_norm": 0.38999518752098083, "learning_rate": 9.923471449408723e-05, "loss": 0.0383, "step": 4120 }, { "grad_norm": 0.41645893454551697, "learning_rate": 9.922749306986244e-05, "loss": 0.0393, "step": 4130 }, { "grad_norm": 0.3385359048843384, "learning_rate": 9.922023799911033e-05, "loss": 0.0343, "step": 4140 }, { "grad_norm": 0.4018062353134155, "learning_rate": 9.921294928678968e-05, "loss": 0.0369, "step": 4150 }, { "grad_norm": 0.3111570179462433, "learning_rate": 9.920562693788222e-05, "loss": 0.0342, "step": 4160 }, { "grad_norm": 0.37304162979125977, "learning_rate": 9.919827095739273e-05, "loss": 0.0372, "step": 4170 }, { "grad_norm": 0.34119850397109985, "learning_rate": 9.919088135034898e-05, "loss": 0.0349, "step": 4180 }, { "grad_norm": 0.458831787109375, "learning_rate": 9.918345812180165e-05, "loss": 0.039, "step": 4190 }, { "grad_norm": 0.3612849712371826, "learning_rate": 9.917600127682446e-05, "loss": 0.0339, "step": 4200 }, { "grad_norm": 0.288106769323349, "learning_rate": 9.91685108205141e-05, "loss": 0.033, "step": 4210 }, { "grad_norm": 0.4623330533504486, "learning_rate": 9.916098675799024e-05, "loss": 0.0346, "step": 4220 }, { "grad_norm": 0.2795619070529938, "learning_rate": 9.915342909439548e-05, "loss": 0.0349, "step": 4230 }, { "grad_norm": 0.3959304690361023, "learning_rate": 9.914583783489543e-05, "loss": 0.0338, "step": 4240 }, { "grad_norm": 0.3435482084751129, "learning_rate": 9.913821298467863e-05, "loss": 0.0356, "step": 4250 }, { "grad_norm": 0.3831171691417694, "learning_rate": 9.913055454895661e-05, "loss": 0.0353, "step": 4260 }, { "grad_norm": 0.3277098536491394, "learning_rate": 9.912286253296381e-05, "loss": 0.0344, "step": 4270 }, { "grad_norm": 0.5218454599380493, "learning_rate": 9.911513694195768e-05, "loss": 0.0371, "step": 4280 }, { "grad_norm": 0.33750292658805847, "learning_rate": 9.910737778121859e-05, "loss": 0.0339, "step": 4290 }, { "grad_norm": 0.33034130930900574, "learning_rate": 9.909958505604984e-05, "loss": 0.031, "step": 4300 }, { "grad_norm": 0.3467344045639038, "learning_rate": 9.909175877177767e-05, "loss": 0.0318, "step": 4310 }, { "grad_norm": 0.3563389480113983, "learning_rate": 9.908389893375129e-05, "loss": 0.0365, "step": 4320 }, { "grad_norm": 0.3423028886318207, "learning_rate": 9.907600554734283e-05, "loss": 0.0318, "step": 4330 }, { "grad_norm": 0.38793498277664185, "learning_rate": 9.906807861794734e-05, "loss": 0.035, "step": 4340 }, { "grad_norm": 0.3575578033924103, "learning_rate": 9.906011815098279e-05, "loss": 0.0414, "step": 4350 }, { "grad_norm": 0.3415902554988861, "learning_rate": 9.90521241518901e-05, "loss": 0.0333, "step": 4360 }, { "grad_norm": 0.29345574975013733, "learning_rate": 9.904409662613308e-05, "loss": 0.0353, "step": 4370 }, { "grad_norm": 0.3438984155654907, "learning_rate": 9.903603557919849e-05, "loss": 0.0367, "step": 4380 }, { "grad_norm": 0.3632912039756775, "learning_rate": 9.902794101659594e-05, "loss": 0.0421, "step": 4390 }, { "grad_norm": 0.4533582627773285, "learning_rate": 9.901981294385803e-05, "loss": 0.0335, "step": 4400 }, { "grad_norm": 0.34154587984085083, "learning_rate": 9.901165136654018e-05, "loss": 0.0355, "step": 4410 }, { "grad_norm": 0.41886937618255615, "learning_rate": 9.900345629022079e-05, "loss": 0.0356, "step": 4420 }, { "grad_norm": 0.37298259139060974, "learning_rate": 9.89952277205011e-05, "loss": 0.0334, "step": 4430 }, { "grad_norm": 0.49569690227508545, "learning_rate": 9.898696566300527e-05, "loss": 0.0352, "step": 4440 }, { "grad_norm": 0.3265545964241028, "learning_rate": 9.897867012338032e-05, "loss": 0.0369, "step": 4450 }, { "grad_norm": 0.342781126499176, "learning_rate": 9.897034110729617e-05, "loss": 0.0346, "step": 4460 }, { "grad_norm": 0.3557250499725342, "learning_rate": 9.896197862044564e-05, "loss": 0.0316, "step": 4470 }, { "grad_norm": 0.4999900162220001, "learning_rate": 9.895358266854442e-05, "loss": 0.0349, "step": 4480 }, { "grad_norm": 0.39477622509002686, "learning_rate": 9.894515325733103e-05, "loss": 0.0413, "step": 4490 }, { "grad_norm": 0.2813314199447632, "learning_rate": 9.893669039256693e-05, "loss": 0.0337, "step": 4500 }, { "grad_norm": 0.2879772484302521, "learning_rate": 9.89281940800364e-05, "loss": 0.0312, "step": 4510 }, { "grad_norm": 0.3302120268344879, "learning_rate": 9.891966432554655e-05, "loss": 0.0299, "step": 4520 }, { "grad_norm": 0.41323429346084595, "learning_rate": 9.891110113492745e-05, "loss": 0.0364, "step": 4530 }, { "grad_norm": 0.3877699077129364, "learning_rate": 9.89025045140319e-05, "loss": 0.0369, "step": 4540 }, { "grad_norm": 0.35489898920059204, "learning_rate": 9.889387446873567e-05, "loss": 0.0315, "step": 4550 }, { "grad_norm": 0.41171130537986755, "learning_rate": 9.888521100493726e-05, "loss": 0.0327, "step": 4560 }, { "grad_norm": 0.35178640484809875, "learning_rate": 9.887651412855809e-05, "loss": 0.0314, "step": 4570 }, { "grad_norm": 0.41266337037086487, "learning_rate": 9.88677838455424e-05, "loss": 0.0383, "step": 4580 }, { "grad_norm": 0.2867991030216217, "learning_rate": 9.885902016185725e-05, "loss": 0.038, "step": 4590 }, { "grad_norm": 0.40495339035987854, "learning_rate": 9.885022308349252e-05, "loss": 0.0341, "step": 4600 }, { "grad_norm": 0.4876304566860199, "learning_rate": 9.884139261646095e-05, "loss": 0.0394, "step": 4610 }, { "grad_norm": 0.36860987544059753, "learning_rate": 9.883252876679807e-05, "loss": 0.0331, "step": 4620 }, { "grad_norm": 0.3314003646373749, "learning_rate": 9.882363154056225e-05, "loss": 0.0327, "step": 4630 }, { "grad_norm": 0.36605289578437805, "learning_rate": 9.881470094383465e-05, "loss": 0.0351, "step": 4640 }, { "grad_norm": 0.33016258478164673, "learning_rate": 9.880573698271924e-05, "loss": 0.0322, "step": 4650 }, { "grad_norm": 0.3478606045246124, "learning_rate": 9.879673966334282e-05, "loss": 0.0361, "step": 4660 }, { "grad_norm": 0.357839971780777, "learning_rate": 9.878770899185496e-05, "loss": 0.0336, "step": 4670 }, { "grad_norm": 0.3115417957305908, "learning_rate": 9.877864497442804e-05, "loss": 0.0313, "step": 4680 }, { "grad_norm": 0.348108172416687, "learning_rate": 9.876954761725723e-05, "loss": 0.0327, "step": 4690 }, { "grad_norm": 0.4120684266090393, "learning_rate": 9.876041692656052e-05, "loss": 0.0411, "step": 4700 }, { "grad_norm": 0.34973207116127014, "learning_rate": 9.87512529085786e-05, "loss": 0.0339, "step": 4710 }, { "grad_norm": 0.26118120551109314, "learning_rate": 9.874205556957503e-05, "loss": 0.03, "step": 4720 }, { "grad_norm": 0.28285935521125793, "learning_rate": 9.873282491583608e-05, "loss": 0.0315, "step": 4730 }, { "grad_norm": 0.4501439332962036, "learning_rate": 9.872356095367084e-05, "loss": 0.0331, "step": 4740 }, { "grad_norm": 0.335245817899704, "learning_rate": 9.87142636894111e-05, "loss": 0.0332, "step": 4750 }, { "grad_norm": 0.29537418484687805, "learning_rate": 9.870493312941148e-05, "loss": 0.0366, "step": 4760 }, { "grad_norm": 0.3636355698108673, "learning_rate": 9.869556928004933e-05, "loss": 0.039, "step": 4770 }, { "grad_norm": 0.34819138050079346, "learning_rate": 9.868617214772476e-05, "loss": 0.0326, "step": 4780 }, { "grad_norm": 0.319810152053833, "learning_rate": 9.867674173886059e-05, "loss": 0.0363, "step": 4790 }, { "grad_norm": 0.3196150064468384, "learning_rate": 9.866727805990245e-05, "loss": 0.0342, "step": 4800 }, { "grad_norm": 0.35500988364219666, "learning_rate": 9.865778111731865e-05, "loss": 0.0294, "step": 4810 }, { "grad_norm": 0.3203672766685486, "learning_rate": 9.864825091760028e-05, "loss": 0.0334, "step": 4820 }, { "grad_norm": 0.4217623174190521, "learning_rate": 9.863868746726112e-05, "loss": 0.0326, "step": 4830 }, { "grad_norm": 0.33032843470573425, "learning_rate": 9.862909077283773e-05, "loss": 0.0329, "step": 4840 }, { "grad_norm": 0.3001137673854828, "learning_rate": 9.861946084088933e-05, "loss": 0.0316, "step": 4850 }, { "grad_norm": 0.3654986619949341, "learning_rate": 9.860979767799792e-05, "loss": 0.0296, "step": 4860 }, { "grad_norm": 0.36920028924942017, "learning_rate": 9.860010129076813e-05, "loss": 0.0363, "step": 4870 }, { "grad_norm": 0.28526654839515686, "learning_rate": 9.859037168582741e-05, "loss": 0.0326, "step": 4880 }, { "grad_norm": 0.28547218441963196, "learning_rate": 9.85806088698258e-05, "loss": 0.0323, "step": 4890 }, { "grad_norm": 0.3253670632839203, "learning_rate": 9.857081284943612e-05, "loss": 0.0312, "step": 4900 }, { "grad_norm": 0.26004913449287415, "learning_rate": 9.856098363135385e-05, "loss": 0.0302, "step": 4910 }, { "grad_norm": 0.3161565363407135, "learning_rate": 9.855112122229717e-05, "loss": 0.0332, "step": 4920 }, { "grad_norm": 0.2994840443134308, "learning_rate": 9.854122562900696e-05, "loss": 0.0337, "step": 4930 }, { "grad_norm": 0.3288623094558716, "learning_rate": 9.853129685824673e-05, "loss": 0.0281, "step": 4940 }, { "grad_norm": 0.31222736835479736, "learning_rate": 9.852133491680273e-05, "loss": 0.0337, "step": 4950 }, { "grad_norm": 0.27230173349380493, "learning_rate": 9.851133981148385e-05, "loss": 0.0344, "step": 4960 }, { "grad_norm": 0.33739835023880005, "learning_rate": 9.850131154912164e-05, "loss": 0.0277, "step": 4970 }, { "grad_norm": 0.38315171003341675, "learning_rate": 9.849125013657031e-05, "loss": 0.0354, "step": 4980 }, { "grad_norm": 0.39031025767326355, "learning_rate": 9.848115558070676e-05, "loss": 0.0341, "step": 4990 }, { "grad_norm": 0.39023199677467346, "learning_rate": 9.847102788843055e-05, "loss": 0.0305, "step": 5000 }, { "grad_norm": 0.3125436305999756, "learning_rate": 9.84608670666638e-05, "loss": 0.0329, "step": 5010 }, { "grad_norm": 0.5150122046470642, "learning_rate": 9.845067312235138e-05, "loss": 0.0349, "step": 5020 }, { "grad_norm": 0.33318158984184265, "learning_rate": 9.844044606246074e-05, "loss": 0.0309, "step": 5030 }, { "grad_norm": 0.44595572352409363, "learning_rate": 9.843018589398199e-05, "loss": 0.0301, "step": 5040 }, { "grad_norm": 0.39419084787368774, "learning_rate": 9.841989262392785e-05, "loss": 0.0337, "step": 5050 }, { "grad_norm": 0.24182040989398956, "learning_rate": 9.840956625933367e-05, "loss": 0.0347, "step": 5060 }, { "grad_norm": 0.2685829699039459, "learning_rate": 9.839920680725743e-05, "loss": 0.0322, "step": 5070 }, { "grad_norm": 0.41008442640304565, "learning_rate": 9.83888142747797e-05, "loss": 0.0306, "step": 5080 }, { "grad_norm": 0.3646048307418823, "learning_rate": 9.83783886690037e-05, "loss": 0.0336, "step": 5090 }, { "grad_norm": 0.30716782808303833, "learning_rate": 9.836792999705524e-05, "loss": 0.0311, "step": 5100 }, { "grad_norm": 0.24125367403030396, "learning_rate": 9.835743826608271e-05, "loss": 0.0329, "step": 5110 }, { "grad_norm": 0.3135361969470978, "learning_rate": 9.834691348325709e-05, "loss": 0.0302, "step": 5120 }, { "grad_norm": 0.37114807963371277, "learning_rate": 9.833635565577199e-05, "loss": 0.0312, "step": 5130 }, { "grad_norm": 0.42956632375717163, "learning_rate": 9.832576479084359e-05, "loss": 0.033, "step": 5140 }, { "grad_norm": 0.2811664640903473, "learning_rate": 9.831514089571064e-05, "loss": 0.0327, "step": 5150 }, { "grad_norm": 0.3321332037448883, "learning_rate": 9.830448397763447e-05, "loss": 0.0309, "step": 5160 }, { "grad_norm": 0.3406131863594055, "learning_rate": 9.8293794043899e-05, "loss": 0.0287, "step": 5170 }, { "grad_norm": 0.2640426456928253, "learning_rate": 9.828307110181066e-05, "loss": 0.0277, "step": 5180 }, { "grad_norm": 0.35774198174476624, "learning_rate": 9.827231515869852e-05, "loss": 0.0336, "step": 5190 }, { "grad_norm": 0.3609336316585541, "learning_rate": 9.826152622191413e-05, "loss": 0.0319, "step": 5200 }, { "grad_norm": 0.3651580810546875, "learning_rate": 9.825070429883167e-05, "loss": 0.039, "step": 5210 }, { "grad_norm": 0.38263654708862305, "learning_rate": 9.82398493968478e-05, "loss": 0.0373, "step": 5220 }, { "grad_norm": 0.30465322732925415, "learning_rate": 9.822896152338173e-05, "loss": 0.0327, "step": 5230 }, { "grad_norm": 0.3118366003036499, "learning_rate": 9.821804068587523e-05, "loss": 0.0281, "step": 5240 }, { "grad_norm": 0.31932297348976135, "learning_rate": 9.820708689179259e-05, "loss": 0.031, "step": 5250 }, { "grad_norm": 0.341619610786438, "learning_rate": 9.819610014862063e-05, "loss": 0.0329, "step": 5260 }, { "grad_norm": 0.34035906195640564, "learning_rate": 9.818508046386868e-05, "loss": 0.0326, "step": 5270 }, { "grad_norm": 0.3582216203212738, "learning_rate": 9.817402784506859e-05, "loss": 0.0322, "step": 5280 }, { "grad_norm": 0.29313188791275024, "learning_rate": 9.816294229977472e-05, "loss": 0.0315, "step": 5290 }, { "grad_norm": 0.27332746982574463, "learning_rate": 9.815182383556394e-05, "loss": 0.0368, "step": 5300 }, { "grad_norm": 0.3789196312427521, "learning_rate": 9.81406724600356e-05, "loss": 0.0366, "step": 5310 }, { "grad_norm": 0.28148892521858215, "learning_rate": 9.812948818081161e-05, "loss": 0.0386, "step": 5320 }, { "grad_norm": 0.29915207624435425, "learning_rate": 9.811827100553623e-05, "loss": 0.0307, "step": 5330 }, { "grad_norm": 0.34457269310951233, "learning_rate": 9.810702094187638e-05, "loss": 0.036, "step": 5340 }, { "grad_norm": 0.3634885847568512, "learning_rate": 9.809573799752135e-05, "loss": 0.0342, "step": 5350 }, { "grad_norm": 0.2654358744621277, "learning_rate": 9.808442218018288e-05, "loss": 0.0292, "step": 5360 }, { "grad_norm": 0.4494386911392212, "learning_rate": 9.807307349759527e-05, "loss": 0.0309, "step": 5370 }, { "grad_norm": 0.2880295515060425, "learning_rate": 9.806169195751525e-05, "loss": 0.0315, "step": 5380 }, { "grad_norm": 0.3165290355682373, "learning_rate": 9.805027756772194e-05, "loss": 0.0307, "step": 5390 }, { "grad_norm": 0.3860031068325043, "learning_rate": 9.803883033601702e-05, "loss": 0.0285, "step": 5400 }, { "grad_norm": 0.3801299035549164, "learning_rate": 9.802735027022453e-05, "loss": 0.031, "step": 5410 }, { "grad_norm": 0.26388493180274963, "learning_rate": 9.8015837378191e-05, "loss": 0.0293, "step": 5420 }, { "grad_norm": 0.30469250679016113, "learning_rate": 9.80042916677854e-05, "loss": 0.0321, "step": 5430 }, { "grad_norm": 0.2883319854736328, "learning_rate": 9.799271314689908e-05, "loss": 0.0294, "step": 5440 }, { "grad_norm": 0.32400673627853394, "learning_rate": 9.798110182344588e-05, "loss": 0.0302, "step": 5450 }, { "grad_norm": 0.2633635103702545, "learning_rate": 9.796945770536204e-05, "loss": 0.0298, "step": 5460 }, { "grad_norm": 0.2450091540813446, "learning_rate": 9.795778080060615e-05, "loss": 0.0275, "step": 5470 }, { "grad_norm": 0.28583553433418274, "learning_rate": 9.79460711171593e-05, "loss": 0.0306, "step": 5480 }, { "grad_norm": 0.3184750974178314, "learning_rate": 9.793432866302497e-05, "loss": 0.0306, "step": 5490 }, { "grad_norm": 0.31813332438468933, "learning_rate": 9.792255344622897e-05, "loss": 0.0289, "step": 5500 }, { "grad_norm": 0.31222784519195557, "learning_rate": 9.791074547481957e-05, "loss": 0.0307, "step": 5510 }, { "grad_norm": 0.36400488018989563, "learning_rate": 9.789890475686742e-05, "loss": 0.033, "step": 5520 }, { "grad_norm": 0.31290507316589355, "learning_rate": 9.788703130046552e-05, "loss": 0.0305, "step": 5530 }, { "grad_norm": 0.34798282384872437, "learning_rate": 9.78751251137293e-05, "loss": 0.0295, "step": 5540 }, { "grad_norm": 0.33720481395721436, "learning_rate": 9.786318620479646e-05, "loss": 0.0306, "step": 5550 }, { "grad_norm": 0.39661023020744324, "learning_rate": 9.78512145818272e-05, "loss": 0.0292, "step": 5560 }, { "grad_norm": 0.3624129295349121, "learning_rate": 9.783921025300398e-05, "loss": 0.0309, "step": 5570 }, { "grad_norm": 0.3051620423793793, "learning_rate": 9.782717322653164e-05, "loss": 0.0274, "step": 5580 }, { "grad_norm": 0.2485840618610382, "learning_rate": 9.781510351063738e-05, "loss": 0.0308, "step": 5590 }, { "grad_norm": 0.3297702372074127, "learning_rate": 9.780300111357075e-05, "loss": 0.0283, "step": 5600 }, { "grad_norm": 0.25240108370780945, "learning_rate": 9.779086604360361e-05, "loss": 0.0287, "step": 5610 }, { "grad_norm": 0.382181853055954, "learning_rate": 9.777869830903019e-05, "loss": 0.0297, "step": 5620 }, { "grad_norm": 0.32797718048095703, "learning_rate": 9.776649791816698e-05, "loss": 0.0273, "step": 5630 }, { "grad_norm": 0.28949353098869324, "learning_rate": 9.775426487935285e-05, "loss": 0.0259, "step": 5640 }, { "grad_norm": 0.34470638632774353, "learning_rate": 9.774199920094898e-05, "loss": 0.0281, "step": 5650 }, { "grad_norm": 0.3074248135089874, "learning_rate": 9.772970089133884e-05, "loss": 0.0362, "step": 5660 }, { "grad_norm": 0.3195187449455261, "learning_rate": 9.77173699589282e-05, "loss": 0.0291, "step": 5670 }, { "grad_norm": 0.3274196684360504, "learning_rate": 9.770500641214513e-05, "loss": 0.031, "step": 5680 }, { "grad_norm": 0.30357179045677185, "learning_rate": 9.769261025944003e-05, "loss": 0.0283, "step": 5690 }, { "grad_norm": 0.43558621406555176, "learning_rate": 9.768018150928552e-05, "loss": 0.0329, "step": 5700 }, { "grad_norm": 0.42299884557724, "learning_rate": 9.766772017017654e-05, "loss": 0.0299, "step": 5710 }, { "grad_norm": 0.37602531909942627, "learning_rate": 9.765522625063032e-05, "loss": 0.0301, "step": 5720 }, { "grad_norm": 0.34564143419265747, "learning_rate": 9.76426997591863e-05, "loss": 0.0296, "step": 5730 }, { "grad_norm": 0.2891349494457245, "learning_rate": 9.763014070440625e-05, "loss": 0.0317, "step": 5740 }, { "grad_norm": 0.3640691339969635, "learning_rate": 9.761754909487415e-05, "loss": 0.0316, "step": 5750 }, { "grad_norm": 0.2734403610229492, "learning_rate": 9.760492493919626e-05, "loss": 0.0288, "step": 5760 }, { "grad_norm": 0.2685346007347107, "learning_rate": 9.759226824600105e-05, "loss": 0.0323, "step": 5770 }, { "grad_norm": 0.33157962560653687, "learning_rate": 9.757957902393928e-05, "loss": 0.029, "step": 5780 }, { "grad_norm": 0.39502382278442383, "learning_rate": 9.756685728168387e-05, "loss": 0.0322, "step": 5790 }, { "grad_norm": 0.31490010023117065, "learning_rate": 9.755410302793004e-05, "loss": 0.0308, "step": 5800 }, { "grad_norm": 0.2915184497833252, "learning_rate": 9.754131627139522e-05, "loss": 0.0282, "step": 5810 }, { "grad_norm": 0.3559376001358032, "learning_rate": 9.752849702081901e-05, "loss": 0.0289, "step": 5820 }, { "grad_norm": 0.31205087900161743, "learning_rate": 9.751564528496324e-05, "loss": 0.0322, "step": 5830 }, { "grad_norm": 0.36154893040657043, "learning_rate": 9.750276107261197e-05, "loss": 0.0307, "step": 5840 }, { "grad_norm": 0.3046317398548126, "learning_rate": 9.748984439257142e-05, "loss": 0.0316, "step": 5850 }, { "grad_norm": 0.2799105942249298, "learning_rate": 9.747689525367005e-05, "loss": 0.0318, "step": 5860 }, { "grad_norm": 0.38410767912864685, "learning_rate": 9.746391366475845e-05, "loss": 0.0341, "step": 5870 }, { "grad_norm": 0.3103953003883362, "learning_rate": 9.745089963470942e-05, "loss": 0.0307, "step": 5880 }, { "grad_norm": 0.30312037467956543, "learning_rate": 9.743785317241791e-05, "loss": 0.028, "step": 5890 }, { "grad_norm": 0.3944193422794342, "learning_rate": 9.742477428680108e-05, "loss": 0.0292, "step": 5900 }, { "grad_norm": 0.3061687648296356, "learning_rate": 9.741166298679821e-05, "loss": 0.0313, "step": 5910 }, { "grad_norm": 0.33588001132011414, "learning_rate": 9.739851928137076e-05, "loss": 0.0278, "step": 5920 }, { "grad_norm": 0.2653016746044159, "learning_rate": 9.738534317950232e-05, "loss": 0.0264, "step": 5930 }, { "grad_norm": 0.41321468353271484, "learning_rate": 9.737213469019864e-05, "loss": 0.0303, "step": 5940 }, { "grad_norm": 0.2523449659347534, "learning_rate": 9.735889382248757e-05, "loss": 0.0281, "step": 5950 }, { "grad_norm": 0.33342188596725464, "learning_rate": 9.734562058541916e-05, "loss": 0.0284, "step": 5960 }, { "grad_norm": 0.2935464680194855, "learning_rate": 9.733231498806552e-05, "loss": 0.0316, "step": 5970 }, { "grad_norm": 0.2822422385215759, "learning_rate": 9.731897703952088e-05, "loss": 0.0285, "step": 5980 }, { "grad_norm": 0.33067262172698975, "learning_rate": 9.730560674890165e-05, "loss": 0.031, "step": 5990 }, { "grad_norm": 0.36957424879074097, "learning_rate": 9.729220412534628e-05, "loss": 0.0318, "step": 6000 }, { "grad_norm": 0.3036111891269684, "learning_rate": 9.72787691780153e-05, "loss": 0.0262, "step": 6010 }, { "grad_norm": 0.2765824496746063, "learning_rate": 9.726530191609142e-05, "loss": 0.0274, "step": 6020 }, { "grad_norm": 0.25716811418533325, "learning_rate": 9.725180234877937e-05, "loss": 0.0268, "step": 6030 }, { "grad_norm": 0.2504397928714752, "learning_rate": 9.723827048530595e-05, "loss": 0.0291, "step": 6040 }, { "grad_norm": 0.3367936611175537, "learning_rate": 9.722470633492011e-05, "loss": 0.0283, "step": 6050 }, { "grad_norm": 0.32778099179267883, "learning_rate": 9.721110990689278e-05, "loss": 0.0255, "step": 6060 }, { "grad_norm": 0.37481582164764404, "learning_rate": 9.719748121051699e-05, "loss": 0.0305, "step": 6070 }, { "grad_norm": 0.2788096070289612, "learning_rate": 9.718382025510785e-05, "loss": 0.0317, "step": 6080 }, { "grad_norm": 0.31610965728759766, "learning_rate": 9.717012705000248e-05, "loss": 0.0257, "step": 6090 }, { "grad_norm": 0.39060840010643005, "learning_rate": 9.715640160456006e-05, "loss": 0.0335, "step": 6100 }, { "grad_norm": 0.3105662763118744, "learning_rate": 9.714264392816181e-05, "loss": 0.0298, "step": 6110 }, { "grad_norm": 0.31575316190719604, "learning_rate": 9.712885403021095e-05, "loss": 0.0338, "step": 6120 }, { "grad_norm": 0.3042646646499634, "learning_rate": 9.711503192013276e-05, "loss": 0.0289, "step": 6130 }, { "grad_norm": 0.36164453625679016, "learning_rate": 9.710117760737454e-05, "loss": 0.027, "step": 6140 }, { "grad_norm": 0.36730605363845825, "learning_rate": 9.708729110140554e-05, "loss": 0.0313, "step": 6150 }, { "grad_norm": 0.3418625593185425, "learning_rate": 9.707337241171709e-05, "loss": 0.0275, "step": 6160 }, { "grad_norm": 0.3052928149700165, "learning_rate": 9.705942154782245e-05, "loss": 0.029, "step": 6170 }, { "grad_norm": 0.3145742118358612, "learning_rate": 9.704543851925696e-05, "loss": 0.0282, "step": 6180 }, { "grad_norm": 0.37798911333084106, "learning_rate": 9.703142333557784e-05, "loss": 0.0276, "step": 6190 }, { "grad_norm": 0.2816234529018402, "learning_rate": 9.701737600636436e-05, "loss": 0.0284, "step": 6200 }, { "grad_norm": 0.2902202606201172, "learning_rate": 9.700329654121771e-05, "loss": 0.0277, "step": 6210 }, { "grad_norm": 0.3782004714012146, "learning_rate": 9.698918494976109e-05, "loss": 0.032, "step": 6220 }, { "grad_norm": 0.3172137439250946, "learning_rate": 9.697504124163965e-05, "loss": 0.03, "step": 6230 }, { "grad_norm": 0.45543521642684937, "learning_rate": 9.696086542652045e-05, "loss": 0.0288, "step": 6240 }, { "grad_norm": 0.3464176654815674, "learning_rate": 9.694665751409256e-05, "loss": 0.0332, "step": 6250 }, { "grad_norm": 0.35600098967552185, "learning_rate": 9.693241751406694e-05, "loss": 0.0325, "step": 6260 }, { "grad_norm": 0.2543141841888428, "learning_rate": 9.69181454361765e-05, "loss": 0.0268, "step": 6270 }, { "grad_norm": 0.3075425922870636, "learning_rate": 9.690384129017605e-05, "loss": 0.0286, "step": 6280 }, { "grad_norm": 0.24728351831436157, "learning_rate": 9.688950508584237e-05, "loss": 0.0272, "step": 6290 }, { "grad_norm": 0.2919873595237732, "learning_rate": 9.68751368329741e-05, "loss": 0.0346, "step": 6300 }, { "grad_norm": 0.37009719014167786, "learning_rate": 9.68607365413918e-05, "loss": 0.032, "step": 6310 }, { "grad_norm": 0.25945308804512024, "learning_rate": 9.684630422093797e-05, "loss": 0.0287, "step": 6320 }, { "grad_norm": 0.3161633014678955, "learning_rate": 9.683183988147693e-05, "loss": 0.0297, "step": 6330 }, { "grad_norm": 0.28605517745018005, "learning_rate": 9.681734353289491e-05, "loss": 0.0316, "step": 6340 }, { "grad_norm": 0.29365938901901245, "learning_rate": 9.680281518510006e-05, "loss": 0.0298, "step": 6350 }, { "grad_norm": 0.3146636188030243, "learning_rate": 9.678825484802236e-05, "loss": 0.0325, "step": 6360 }, { "grad_norm": 0.24865354597568512, "learning_rate": 9.677366253161365e-05, "loss": 0.0298, "step": 6370 }, { "grad_norm": 0.37620100378990173, "learning_rate": 9.675903824584765e-05, "loss": 0.0321, "step": 6380 }, { "grad_norm": 0.26188820600509644, "learning_rate": 9.674438200071991e-05, "loss": 0.0307, "step": 6390 }, { "grad_norm": 0.3363809585571289, "learning_rate": 9.672969380624786e-05, "loss": 0.0267, "step": 6400 }, { "grad_norm": 0.3134056031703949, "learning_rate": 9.671497367247073e-05, "loss": 0.0335, "step": 6410 }, { "grad_norm": 0.3259485960006714, "learning_rate": 9.670022160944959e-05, "loss": 0.0315, "step": 6420 }, { "grad_norm": 0.31705039739608765, "learning_rate": 9.668543762726734e-05, "loss": 0.0292, "step": 6430 }, { "grad_norm": 0.32255417108535767, "learning_rate": 9.667062173602868e-05, "loss": 0.0299, "step": 6440 }, { "grad_norm": 0.4242064952850342, "learning_rate": 9.665577394586013e-05, "loss": 0.0297, "step": 6450 }, { "grad_norm": 0.3888915181159973, "learning_rate": 9.664089426691006e-05, "loss": 0.0294, "step": 6460 }, { "grad_norm": 0.2758578956127167, "learning_rate": 9.662598270934852e-05, "loss": 0.0284, "step": 6470 }, { "grad_norm": 0.27625772356987, "learning_rate": 9.661103928336748e-05, "loss": 0.0301, "step": 6480 }, { "grad_norm": 0.3577733337879181, "learning_rate": 9.659606399918058e-05, "loss": 0.0296, "step": 6490 }, { "grad_norm": 0.29295310378074646, "learning_rate": 9.658105686702334e-05, "loss": 0.0269, "step": 6500 }, { "grad_norm": 0.37261971831321716, "learning_rate": 9.656601789715294e-05, "loss": 0.027, "step": 6510 }, { "grad_norm": 0.34186238050460815, "learning_rate": 9.655094709984842e-05, "loss": 0.0267, "step": 6520 }, { "grad_norm": 0.2675359547138214, "learning_rate": 9.653584448541048e-05, "loss": 0.0291, "step": 6530 }, { "grad_norm": 0.2751590609550476, "learning_rate": 9.652071006416166e-05, "loss": 0.0298, "step": 6540 }, { "grad_norm": 0.3096335232257843, "learning_rate": 9.650554384644617e-05, "loss": 0.0293, "step": 6550 }, { "grad_norm": 0.2830754220485687, "learning_rate": 9.649034584263e-05, "loss": 0.0295, "step": 6560 }, { "grad_norm": 0.30793070793151855, "learning_rate": 9.64751160631008e-05, "loss": 0.032, "step": 6570 }, { "grad_norm": 0.37758633494377136, "learning_rate": 9.645985451826803e-05, "loss": 0.0281, "step": 6580 }, { "grad_norm": 0.39615362882614136, "learning_rate": 9.644456121856275e-05, "loss": 0.0308, "step": 6590 }, { "grad_norm": 0.25282523036003113, "learning_rate": 9.642923617443784e-05, "loss": 0.0323, "step": 6600 }, { "grad_norm": 0.2071130871772766, "learning_rate": 9.641387939636782e-05, "loss": 0.0317, "step": 6610 }, { "grad_norm": 0.24074393510818481, "learning_rate": 9.639849089484888e-05, "loss": 0.0237, "step": 6620 }, { "grad_norm": 0.2605157196521759, "learning_rate": 9.638307068039893e-05, "loss": 0.032, "step": 6630 }, { "grad_norm": 0.3187861442565918, "learning_rate": 9.636761876355753e-05, "loss": 0.0286, "step": 6640 }, { "grad_norm": 0.36089518666267395, "learning_rate": 9.635213515488593e-05, "loss": 0.0294, "step": 6650 }, { "grad_norm": 0.23764535784721375, "learning_rate": 9.633661986496702e-05, "loss": 0.0307, "step": 6660 }, { "grad_norm": 0.27004799246788025, "learning_rate": 9.632107290440535e-05, "loss": 0.0306, "step": 6670 }, { "grad_norm": 0.3744039237499237, "learning_rate": 9.630549428382715e-05, "loss": 0.0293, "step": 6680 }, { "grad_norm": 0.38505762815475464, "learning_rate": 9.628988401388024e-05, "loss": 0.0294, "step": 6690 }, { "grad_norm": 0.30218666791915894, "learning_rate": 9.627424210523407e-05, "loss": 0.0304, "step": 6700 }, { "grad_norm": 0.29727500677108765, "learning_rate": 9.62585685685798e-05, "loss": 0.0286, "step": 6710 }, { "grad_norm": 0.23505505919456482, "learning_rate": 9.624286341463008e-05, "loss": 0.0288, "step": 6720 }, { "grad_norm": 0.26152169704437256, "learning_rate": 9.622712665411927e-05, "loss": 0.0246, "step": 6730 }, { "grad_norm": 0.25982221961021423, "learning_rate": 9.621135829780328e-05, "loss": 0.0278, "step": 6740 }, { "grad_norm": 0.27999165654182434, "learning_rate": 9.619555835645964e-05, "loss": 0.0259, "step": 6750 }, { "grad_norm": 0.32304906845092773, "learning_rate": 9.617972684088747e-05, "loss": 0.0255, "step": 6760 }, { "grad_norm": 0.33710771799087524, "learning_rate": 9.616386376190745e-05, "loss": 0.0264, "step": 6770 }, { "grad_norm": 0.29827550053596497, "learning_rate": 9.614796913036184e-05, "loss": 0.0273, "step": 6780 }, { "grad_norm": 0.34231624007225037, "learning_rate": 9.61320429571145e-05, "loss": 0.0296, "step": 6790 }, { "grad_norm": 0.31028497219085693, "learning_rate": 9.611608525305078e-05, "loss": 0.0272, "step": 6800 }, { "grad_norm": 0.3769231140613556, "learning_rate": 9.610009602907764e-05, "loss": 0.0281, "step": 6810 }, { "grad_norm": 0.28071632981300354, "learning_rate": 9.608407529612357e-05, "loss": 0.0296, "step": 6820 }, { "grad_norm": 0.37568673491477966, "learning_rate": 9.606802306513857e-05, "loss": 0.0272, "step": 6830 }, { "grad_norm": 0.28827348351478577, "learning_rate": 9.605193934709421e-05, "loss": 0.0281, "step": 6840 }, { "grad_norm": 0.3283229470252991, "learning_rate": 9.603582415298354e-05, "loss": 0.0301, "step": 6850 }, { "grad_norm": 0.39044874906539917, "learning_rate": 9.601967749382117e-05, "loss": 0.032, "step": 6860 }, { "grad_norm": 0.25380900502204895, "learning_rate": 9.600349938064316e-05, "loss": 0.0252, "step": 6870 }, { "grad_norm": 0.4052935838699341, "learning_rate": 9.598728982450711e-05, "loss": 0.0311, "step": 6880 }, { "grad_norm": 0.21532359719276428, "learning_rate": 9.597104883649211e-05, "loss": 0.0242, "step": 6890 }, { "grad_norm": 0.27873075008392334, "learning_rate": 9.59547764276987e-05, "loss": 0.0255, "step": 6900 }, { "grad_norm": 0.284502238035202, "learning_rate": 9.593847260924892e-05, "loss": 0.0251, "step": 6910 }, { "grad_norm": 0.2764848470687866, "learning_rate": 9.59221373922863e-05, "loss": 0.027, "step": 6920 }, { "grad_norm": 0.2587053179740906, "learning_rate": 9.59057707879758e-05, "loss": 0.0278, "step": 6930 }, { "grad_norm": 0.247299462556839, "learning_rate": 9.588937280750382e-05, "loss": 0.0237, "step": 6940 }, { "grad_norm": 0.26448091864585876, "learning_rate": 9.587294346207824e-05, "loss": 0.0246, "step": 6950 }, { "grad_norm": 0.35987526178359985, "learning_rate": 9.585648276292836e-05, "loss": 0.0265, "step": 6960 }, { "grad_norm": 0.30306318402290344, "learning_rate": 9.583999072130492e-05, "loss": 0.0249, "step": 6970 }, { "grad_norm": 0.29073023796081543, "learning_rate": 9.582346734848005e-05, "loss": 0.0273, "step": 6980 }, { "grad_norm": 0.3066357970237732, "learning_rate": 9.580691265574735e-05, "loss": 0.0253, "step": 6990 }, { "grad_norm": 0.20487327873706818, "learning_rate": 9.579032665442178e-05, "loss": 0.0252, "step": 7000 }, { "grad_norm": 0.24541904032230377, "learning_rate": 9.577370935583971e-05, "loss": 0.0289, "step": 7010 }, { "grad_norm": 0.3572259843349457, "learning_rate": 9.575706077135893e-05, "loss": 0.0299, "step": 7020 }, { "grad_norm": 0.31388917565345764, "learning_rate": 9.574038091235857e-05, "loss": 0.0247, "step": 7030 }, { "grad_norm": 0.2614099383354187, "learning_rate": 9.572366979023916e-05, "loss": 0.0266, "step": 7040 }, { "grad_norm": 0.2650270462036133, "learning_rate": 9.57069274164226e-05, "loss": 0.0253, "step": 7050 }, { "grad_norm": 0.2947990298271179, "learning_rate": 9.569015380235213e-05, "loss": 0.0257, "step": 7060 }, { "grad_norm": 0.24803505837917328, "learning_rate": 9.567334895949238e-05, "loss": 0.0233, "step": 7070 }, { "grad_norm": 0.2638857364654541, "learning_rate": 9.565651289932928e-05, "loss": 0.0258, "step": 7080 }, { "grad_norm": 0.21081192791461945, "learning_rate": 9.563964563337015e-05, "loss": 0.026, "step": 7090 }, { "grad_norm": 0.2823010981082916, "learning_rate": 9.562274717314357e-05, "loss": 0.023, "step": 7100 }, { "grad_norm": 0.21881507337093353, "learning_rate": 9.56058175301995e-05, "loss": 0.0271, "step": 7110 }, { "grad_norm": 0.29322150349617004, "learning_rate": 9.558885671610918e-05, "loss": 0.0248, "step": 7120 }, { "grad_norm": 0.2910972535610199, "learning_rate": 9.557186474246517e-05, "loss": 0.0216, "step": 7130 }, { "grad_norm": 0.28369542956352234, "learning_rate": 9.555484162088133e-05, "loss": 0.0272, "step": 7140 }, { "grad_norm": 0.2825823724269867, "learning_rate": 9.553778736299279e-05, "loss": 0.0232, "step": 7150 }, { "grad_norm": 0.3774452209472656, "learning_rate": 9.552070198045599e-05, "loss": 0.0251, "step": 7160 }, { "grad_norm": 0.27587994933128357, "learning_rate": 9.550358548494863e-05, "loss": 0.0244, "step": 7170 }, { "grad_norm": 0.35375046730041504, "learning_rate": 9.548643788816965e-05, "loss": 0.0283, "step": 7180 }, { "grad_norm": 0.3801659643650055, "learning_rate": 9.546925920183929e-05, "loss": 0.0252, "step": 7190 }, { "grad_norm": 0.2974262535572052, "learning_rate": 9.545204943769902e-05, "loss": 0.0271, "step": 7200 }, { "grad_norm": 0.26105019450187683, "learning_rate": 9.543480860751155e-05, "loss": 0.0288, "step": 7210 }, { "grad_norm": 0.2898331582546234, "learning_rate": 9.541753672306081e-05, "loss": 0.0262, "step": 7220 }, { "grad_norm": 0.2897789776325226, "learning_rate": 9.540023379615198e-05, "loss": 0.0277, "step": 7230 }, { "grad_norm": 0.3030625879764557, "learning_rate": 9.538289983861146e-05, "loss": 0.0245, "step": 7240 }, { "grad_norm": 0.3437998294830322, "learning_rate": 9.536553486228683e-05, "loss": 0.0251, "step": 7250 }, { "grad_norm": 0.3415971100330353, "learning_rate": 9.534813887904689e-05, "loss": 0.0259, "step": 7260 }, { "grad_norm": 0.27824199199676514, "learning_rate": 9.533071190078163e-05, "loss": 0.0248, "step": 7270 }, { "grad_norm": 0.2864810526371002, "learning_rate": 9.531325393940225e-05, "loss": 0.0219, "step": 7280 }, { "grad_norm": 0.29063108563423157, "learning_rate": 9.529576500684105e-05, "loss": 0.0269, "step": 7290 }, { "grad_norm": 0.30641764402389526, "learning_rate": 9.52782451150516e-05, "loss": 0.0242, "step": 7300 }, { "grad_norm": 0.3504621386528015, "learning_rate": 9.526069427600857e-05, "loss": 0.0242, "step": 7310 }, { "grad_norm": 0.2614533603191376, "learning_rate": 9.524311250170776e-05, "loss": 0.0256, "step": 7320 }, { "grad_norm": 0.2577875256538391, "learning_rate": 9.522549980416619e-05, "loss": 0.0282, "step": 7330 }, { "grad_norm": 0.26209720969200134, "learning_rate": 9.520785619542196e-05, "loss": 0.025, "step": 7340 }, { "grad_norm": 0.3182014226913452, "learning_rate": 9.519018168753428e-05, "loss": 0.027, "step": 7350 }, { "grad_norm": 0.21648049354553223, "learning_rate": 9.517247629258357e-05, "loss": 0.0239, "step": 7360 }, { "grad_norm": 0.30229315161705017, "learning_rate": 9.515474002267124e-05, "loss": 0.0252, "step": 7370 }, { "grad_norm": 0.3451880216598511, "learning_rate": 9.513697288991989e-05, "loss": 0.0253, "step": 7380 }, { "grad_norm": 0.3225553333759308, "learning_rate": 9.51191749064732e-05, "loss": 0.028, "step": 7390 }, { "grad_norm": 0.3308776021003723, "learning_rate": 9.51013460844959e-05, "loss": 0.0249, "step": 7400 }, { "grad_norm": 0.2789253890514374, "learning_rate": 9.508348643617382e-05, "loss": 0.0301, "step": 7410 }, { "grad_norm": 0.360237181186676, "learning_rate": 9.50655959737139e-05, "loss": 0.0266, "step": 7420 }, { "grad_norm": 0.3297739028930664, "learning_rate": 9.504767470934405e-05, "loss": 0.0341, "step": 7430 }, { "grad_norm": 0.3183579742908478, "learning_rate": 9.502972265531332e-05, "loss": 0.0289, "step": 7440 }, { "grad_norm": 0.30840590596199036, "learning_rate": 9.501173982389175e-05, "loss": 0.0291, "step": 7450 }, { "grad_norm": 0.23771487176418304, "learning_rate": 9.499372622737047e-05, "loss": 0.0282, "step": 7460 }, { "grad_norm": 0.3025233745574951, "learning_rate": 9.497568187806156e-05, "loss": 0.033, "step": 7470 }, { "grad_norm": 0.31968528032302856, "learning_rate": 9.495760678829819e-05, "loss": 0.0264, "step": 7480 }, { "grad_norm": 0.3283470571041107, "learning_rate": 9.49395009704345e-05, "loss": 0.0281, "step": 7490 }, { "grad_norm": 0.2870698571205139, "learning_rate": 9.492136443684565e-05, "loss": 0.0295, "step": 7500 }, { "grad_norm": 0.378156453371048, "learning_rate": 9.49031971999278e-05, "loss": 0.0242, "step": 7510 }, { "grad_norm": 0.22175240516662598, "learning_rate": 9.488499927209806e-05, "loss": 0.024, "step": 7520 }, { "grad_norm": 0.31336113810539246, "learning_rate": 9.486677066579456e-05, "loss": 0.0258, "step": 7530 }, { "grad_norm": 0.21571701765060425, "learning_rate": 9.484851139347639e-05, "loss": 0.0237, "step": 7540 }, { "grad_norm": 0.294091135263443, "learning_rate": 9.483022146762358e-05, "loss": 0.0247, "step": 7550 }, { "grad_norm": 0.27946123480796814, "learning_rate": 9.48119009007371e-05, "loss": 0.0266, "step": 7560 }, { "grad_norm": 0.29503756761550903, "learning_rate": 9.479354970533892e-05, "loss": 0.0229, "step": 7570 }, { "grad_norm": 0.33667823672294617, "learning_rate": 9.477516789397189e-05, "loss": 0.0274, "step": 7580 }, { "grad_norm": 0.22726906836032867, "learning_rate": 9.475675547919983e-05, "loss": 0.0228, "step": 7590 }, { "grad_norm": 0.2725505530834198, "learning_rate": 9.47383124736074e-05, "loss": 0.0247, "step": 7600 }, { "grad_norm": 0.28790798783302307, "learning_rate": 9.471983888980028e-05, "loss": 0.0242, "step": 7610 }, { "grad_norm": 0.38724035024642944, "learning_rate": 9.470133474040494e-05, "loss": 0.0258, "step": 7620 }, { "grad_norm": 0.27602747082710266, "learning_rate": 9.468280003806882e-05, "loss": 0.023, "step": 7630 }, { "grad_norm": 0.32050269842147827, "learning_rate": 9.46642347954602e-05, "loss": 0.0289, "step": 7640 }, { "grad_norm": 0.2575589716434479, "learning_rate": 9.464563902526826e-05, "loss": 0.0278, "step": 7650 }, { "grad_norm": 0.24406415224075317, "learning_rate": 9.462701274020303e-05, "loss": 0.0206, "step": 7660 }, { "grad_norm": 0.2367894947528839, "learning_rate": 9.46083559529954e-05, "loss": 0.024, "step": 7670 }, { "grad_norm": 0.25282377004623413, "learning_rate": 9.458966867639712e-05, "loss": 0.0257, "step": 7680 }, { "grad_norm": 0.2910465896129608, "learning_rate": 9.457095092318074e-05, "loss": 0.0264, "step": 7690 }, { "grad_norm": 0.2966388761997223, "learning_rate": 9.455220270613969e-05, "loss": 0.0243, "step": 7700 }, { "grad_norm": 0.24430932104587555, "learning_rate": 9.45334240380882e-05, "loss": 0.0262, "step": 7710 }, { "grad_norm": 0.2881566286087036, "learning_rate": 9.451461493186129e-05, "loss": 0.0205, "step": 7720 }, { "grad_norm": 0.2359827160835266, "learning_rate": 9.449577540031482e-05, "loss": 0.0241, "step": 7730 }, { "grad_norm": 0.31472310423851013, "learning_rate": 9.447690545632544e-05, "loss": 0.0231, "step": 7740 }, { "grad_norm": 0.26023396849632263, "learning_rate": 9.445800511279058e-05, "loss": 0.0242, "step": 7750 }, { "grad_norm": 0.2452656328678131, "learning_rate": 9.443907438262844e-05, "loss": 0.0246, "step": 7760 }, { "grad_norm": 0.22662240266799927, "learning_rate": 9.442011327877798e-05, "loss": 0.0221, "step": 7770 }, { "grad_norm": 0.2209608256816864, "learning_rate": 9.440112181419895e-05, "loss": 0.0241, "step": 7780 }, { "grad_norm": 0.2801649570465088, "learning_rate": 9.438210000187186e-05, "loss": 0.0215, "step": 7790 }, { "grad_norm": 0.26483920216560364, "learning_rate": 9.43630478547979e-05, "loss": 0.0255, "step": 7800 }, { "grad_norm": 0.29883718490600586, "learning_rate": 9.434396538599906e-05, "loss": 0.0248, "step": 7810 }, { "grad_norm": 0.33484065532684326, "learning_rate": 9.4324852608518e-05, "loss": 0.0238, "step": 7820 }, { "grad_norm": 0.2684856951236725, "learning_rate": 9.430570953541816e-05, "loss": 0.0228, "step": 7830 }, { "grad_norm": 0.2833767533302307, "learning_rate": 9.428653617978362e-05, "loss": 0.0257, "step": 7840 }, { "grad_norm": 0.2889496982097626, "learning_rate": 9.42673325547192e-05, "loss": 0.0211, "step": 7850 }, { "grad_norm": 0.2226199060678482, "learning_rate": 9.42480986733504e-05, "loss": 0.0223, "step": 7860 }, { "grad_norm": 0.2534520626068115, "learning_rate": 9.422883454882338e-05, "loss": 0.0224, "step": 7870 }, { "grad_norm": 0.2519126236438751, "learning_rate": 9.4209540194305e-05, "loss": 0.0268, "step": 7880 }, { "grad_norm": 0.27779632806777954, "learning_rate": 9.419021562298278e-05, "loss": 0.022, "step": 7890 }, { "grad_norm": 0.2894861698150635, "learning_rate": 9.417086084806486e-05, "loss": 0.0252, "step": 7900 }, { "grad_norm": 0.29331162571907043, "learning_rate": 9.415147588278005e-05, "loss": 0.0283, "step": 7910 }, { "grad_norm": 0.30507853627204895, "learning_rate": 9.413206074037781e-05, "loss": 0.0241, "step": 7920 }, { "grad_norm": 0.29521265625953674, "learning_rate": 9.411261543412819e-05, "loss": 0.0264, "step": 7930 }, { "grad_norm": 0.1968209147453308, "learning_rate": 9.409313997732188e-05, "loss": 0.026, "step": 7940 }, { "grad_norm": 0.18849462270736694, "learning_rate": 9.407363438327018e-05, "loss": 0.0216, "step": 7950 }, { "grad_norm": 0.28120845556259155, "learning_rate": 9.405409866530496e-05, "loss": 0.0235, "step": 7960 }, { "grad_norm": 0.2509375214576721, "learning_rate": 9.40345328367787e-05, "loss": 0.0233, "step": 7970 }, { "grad_norm": 0.3123689293861389, "learning_rate": 9.401493691106446e-05, "loss": 0.0252, "step": 7980 }, { "grad_norm": 0.2797274589538574, "learning_rate": 9.39953109015559e-05, "loss": 0.0231, "step": 7990 }, { "grad_norm": 0.2274988889694214, "learning_rate": 9.397565482166718e-05, "loss": 0.0219, "step": 8000 }, { "grad_norm": 0.2937757074832916, "learning_rate": 9.395596868483305e-05, "loss": 0.0241, "step": 8010 }, { "grad_norm": 0.26023828983306885, "learning_rate": 9.393625250450881e-05, "loss": 0.0261, "step": 8020 }, { "grad_norm": 0.21730677783489227, "learning_rate": 9.391650629417028e-05, "loss": 0.0228, "step": 8030 }, { "grad_norm": 0.2897912561893463, "learning_rate": 9.38967300673138e-05, "loss": 0.0249, "step": 8040 }, { "grad_norm": 0.2897968292236328, "learning_rate": 9.387692383745624e-05, "loss": 0.0231, "step": 8050 }, { "grad_norm": 0.22514259815216064, "learning_rate": 9.385708761813496e-05, "loss": 0.0219, "step": 8060 }, { "grad_norm": 0.3045806884765625, "learning_rate": 9.383722142290787e-05, "loss": 0.024, "step": 8070 }, { "grad_norm": 0.25448301434516907, "learning_rate": 9.381732526535328e-05, "loss": 0.0243, "step": 8080 }, { "grad_norm": 0.2911117970943451, "learning_rate": 9.379739915907004e-05, "loss": 0.022, "step": 8090 }, { "grad_norm": 0.27273061871528625, "learning_rate": 9.377744311767746e-05, "loss": 0.0227, "step": 8100 }, { "grad_norm": 0.25758904218673706, "learning_rate": 9.375745715481532e-05, "loss": 0.0278, "step": 8110 }, { "grad_norm": 0.23681242763996124, "learning_rate": 9.373744128414382e-05, "loss": 0.0249, "step": 8120 }, { "grad_norm": 0.2425873577594757, "learning_rate": 9.371739551934359e-05, "loss": 0.024, "step": 8130 }, { "grad_norm": 0.3173688054084778, "learning_rate": 9.36973198741158e-05, "loss": 0.0253, "step": 8140 }, { "grad_norm": 0.2119097262620926, "learning_rate": 9.367721436218189e-05, "loss": 0.0235, "step": 8150 }, { "grad_norm": 0.2754736542701721, "learning_rate": 9.365707899728383e-05, "loss": 0.0253, "step": 8160 }, { "grad_norm": 0.17018213868141174, "learning_rate": 9.363691379318391e-05, "loss": 0.0239, "step": 8170 }, { "grad_norm": 0.28796809911727905, "learning_rate": 9.361671876366491e-05, "loss": 0.0242, "step": 8180 }, { "grad_norm": 0.2745251953601837, "learning_rate": 9.35964939225299e-05, "loss": 0.0237, "step": 8190 }, { "grad_norm": 0.23560822010040283, "learning_rate": 9.35762392836024e-05, "loss": 0.0206, "step": 8200 }, { "grad_norm": 0.30954065918922424, "learning_rate": 9.355595486072624e-05, "loss": 0.0235, "step": 8210 }, { "grad_norm": 0.3458588421344757, "learning_rate": 9.353564066776563e-05, "loss": 0.0217, "step": 8220 }, { "grad_norm": 0.3304402530193329, "learning_rate": 9.351529671860516e-05, "loss": 0.0236, "step": 8230 }, { "grad_norm": 0.28776782751083374, "learning_rate": 9.349492302714969e-05, "loss": 0.0221, "step": 8240 }, { "grad_norm": 0.24840810894966125, "learning_rate": 9.347451960732447e-05, "loss": 0.0225, "step": 8250 }, { "grad_norm": 0.2147844433784485, "learning_rate": 9.345408647307506e-05, "loss": 0.0228, "step": 8260 }, { "grad_norm": 0.2100747972726822, "learning_rate": 9.343362363836726e-05, "loss": 0.0264, "step": 8270 }, { "grad_norm": 0.2799692153930664, "learning_rate": 9.341313111718728e-05, "loss": 0.0249, "step": 8280 }, { "grad_norm": 0.26766932010650635, "learning_rate": 9.339260892354153e-05, "loss": 0.0229, "step": 8290 }, { "grad_norm": 0.29418978095054626, "learning_rate": 9.337205707145676e-05, "loss": 0.0202, "step": 8300 }, { "grad_norm": 0.2611863613128662, "learning_rate": 9.335147557497994e-05, "loss": 0.024, "step": 8310 }, { "grad_norm": 0.2928304970264435, "learning_rate": 9.333086444817835e-05, "loss": 0.0245, "step": 8320 }, { "grad_norm": 0.26097697019577026, "learning_rate": 9.331022370513951e-05, "loss": 0.0214, "step": 8330 }, { "grad_norm": 0.273175984621048, "learning_rate": 9.328955335997113e-05, "loss": 0.0218, "step": 8340 }, { "grad_norm": 0.23317550122737885, "learning_rate": 9.326885342680125e-05, "loss": 0.021, "step": 8350 }, { "grad_norm": 0.2957383394241333, "learning_rate": 9.324812391977806e-05, "loss": 0.0204, "step": 8360 }, { "grad_norm": 0.2707246243953705, "learning_rate": 9.322736485306995e-05, "loss": 0.0225, "step": 8370 }, { "grad_norm": 0.28538450598716736, "learning_rate": 9.32065762408656e-05, "loss": 0.0227, "step": 8380 }, { "grad_norm": 0.287155419588089, "learning_rate": 9.318575809737377e-05, "loss": 0.0221, "step": 8390 }, { "grad_norm": 0.25202471017837524, "learning_rate": 9.316491043682351e-05, "loss": 0.0226, "step": 8400 }, { "grad_norm": 0.23376983404159546, "learning_rate": 9.3144033273464e-05, "loss": 0.0222, "step": 8410 }, { "grad_norm": 0.28510352969169617, "learning_rate": 9.312312662156457e-05, "loss": 0.0221, "step": 8420 }, { "grad_norm": 0.26274317502975464, "learning_rate": 9.310219049541471e-05, "loss": 0.0203, "step": 8430 }, { "grad_norm": 0.21665531396865845, "learning_rate": 9.308122490932409e-05, "loss": 0.0227, "step": 8440 }, { "grad_norm": 0.3266764283180237, "learning_rate": 9.306022987762246e-05, "loss": 0.0228, "step": 8450 }, { "grad_norm": 0.28097593784332275, "learning_rate": 9.303920541465974e-05, "loss": 0.0245, "step": 8460 }, { "grad_norm": 0.23681211471557617, "learning_rate": 9.301815153480598e-05, "loss": 0.0263, "step": 8470 }, { "grad_norm": 0.23262999951839447, "learning_rate": 9.299706825245126e-05, "loss": 0.0221, "step": 8480 }, { "grad_norm": 0.24108771979808807, "learning_rate": 9.297595558200583e-05, "loss": 0.0211, "step": 8490 }, { "grad_norm": 0.2679217755794525, "learning_rate": 9.295481353790001e-05, "loss": 0.0221, "step": 8500 }, { "grad_norm": 0.3355872333049774, "learning_rate": 9.293364213458417e-05, "loss": 0.0227, "step": 8510 }, { "grad_norm": 0.317582368850708, "learning_rate": 9.291244138652877e-05, "loss": 0.0237, "step": 8520 }, { "grad_norm": 0.26138558983802795, "learning_rate": 9.289121130822431e-05, "loss": 0.0267, "step": 8530 }, { "grad_norm": 0.2985326647758484, "learning_rate": 9.286995191418137e-05, "loss": 0.0259, "step": 8540 }, { "grad_norm": 0.38802725076675415, "learning_rate": 9.284866321893056e-05, "loss": 0.0238, "step": 8550 }, { "grad_norm": 0.3575741946697235, "learning_rate": 9.282734523702244e-05, "loss": 0.0265, "step": 8560 }, { "grad_norm": 0.27356043457984924, "learning_rate": 9.280599798302771e-05, "loss": 0.021, "step": 8570 }, { "grad_norm": 0.25932496786117554, "learning_rate": 9.278462147153699e-05, "loss": 0.0242, "step": 8580 }, { "grad_norm": 0.29263320565223694, "learning_rate": 9.276321571716094e-05, "loss": 0.0224, "step": 8590 }, { "grad_norm": 0.24929998815059662, "learning_rate": 9.274178073453017e-05, "loss": 0.0213, "step": 8600 }, { "grad_norm": 0.2784077227115631, "learning_rate": 9.27203165382953e-05, "loss": 0.022, "step": 8610 }, { "grad_norm": 0.29332679510116577, "learning_rate": 9.26988231431269e-05, "loss": 0.023, "step": 8620 }, { "grad_norm": 0.24715951085090637, "learning_rate": 9.267730056371551e-05, "loss": 0.018, "step": 8630 }, { "grad_norm": 0.2295784056186676, "learning_rate": 9.265574881477161e-05, "loss": 0.0198, "step": 8640 }, { "grad_norm": 0.2383338063955307, "learning_rate": 9.263416791102561e-05, "loss": 0.0233, "step": 8650 }, { "grad_norm": 0.22256098687648773, "learning_rate": 9.261255786722786e-05, "loss": 0.0243, "step": 8660 }, { "grad_norm": 0.25416797399520874, "learning_rate": 9.259091869814864e-05, "loss": 0.0251, "step": 8670 }, { "grad_norm": 0.23348189890384674, "learning_rate": 9.256925041857807e-05, "loss": 0.0243, "step": 8680 }, { "grad_norm": 0.20493480563163757, "learning_rate": 9.254755304332626e-05, "loss": 0.0226, "step": 8690 }, { "grad_norm": 0.29327622056007385, "learning_rate": 9.252582658722313e-05, "loss": 0.0199, "step": 8700 }, { "grad_norm": 0.19131869077682495, "learning_rate": 9.250407106511853e-05, "loss": 0.0209, "step": 8710 }, { "grad_norm": 0.37497028708457947, "learning_rate": 9.248228649188215e-05, "loss": 0.019, "step": 8720 }, { "grad_norm": 0.36988815665245056, "learning_rate": 9.246047288240354e-05, "loss": 0.023, "step": 8730 }, { "grad_norm": 0.21144229173660278, "learning_rate": 9.243863025159208e-05, "loss": 0.0195, "step": 8740 }, { "grad_norm": 0.273329496383667, "learning_rate": 9.241675861437704e-05, "loss": 0.0234, "step": 8750 }, { "grad_norm": 0.20204149186611176, "learning_rate": 9.239485798570742e-05, "loss": 0.0211, "step": 8760 }, { "grad_norm": 0.2508716583251953, "learning_rate": 9.237292838055215e-05, "loss": 0.0193, "step": 8770 }, { "grad_norm": 0.2460131049156189, "learning_rate": 9.235096981389988e-05, "loss": 0.0198, "step": 8780 }, { "grad_norm": 0.23820099234580994, "learning_rate": 9.232898230075909e-05, "loss": 0.02, "step": 8790 }, { "grad_norm": 0.22252042591571808, "learning_rate": 9.230696585615801e-05, "loss": 0.0235, "step": 8800 }, { "grad_norm": 0.29326286911964417, "learning_rate": 9.228492049514473e-05, "loss": 0.0217, "step": 8810 }, { "grad_norm": 0.24505552649497986, "learning_rate": 9.2262846232787e-05, "loss": 0.0232, "step": 8820 }, { "grad_norm": 0.26486778259277344, "learning_rate": 9.224074308417238e-05, "loss": 0.0264, "step": 8830 }, { "grad_norm": 0.2054905742406845, "learning_rate": 9.221861106440818e-05, "loss": 0.0243, "step": 8840 }, { "grad_norm": 0.2122398316860199, "learning_rate": 9.219645018862142e-05, "loss": 0.0255, "step": 8850 }, { "grad_norm": 0.20386286079883575, "learning_rate": 9.217426047195882e-05, "loss": 0.0236, "step": 8860 }, { "grad_norm": 0.28464439511299133, "learning_rate": 9.21520419295869e-05, "loss": 0.0234, "step": 8870 }, { "grad_norm": 0.2798413634300232, "learning_rate": 9.212979457669178e-05, "loss": 0.0199, "step": 8880 }, { "grad_norm": 0.27767929434776306, "learning_rate": 9.210751842847931e-05, "loss": 0.0238, "step": 8890 }, { "grad_norm": 0.2628227472305298, "learning_rate": 9.208521350017506e-05, "loss": 0.0181, "step": 8900 }, { "grad_norm": 0.16931019723415375, "learning_rate": 9.206287980702422e-05, "loss": 0.0213, "step": 8910 }, { "grad_norm": 0.19995468854904175, "learning_rate": 9.204051736429166e-05, "loss": 0.0203, "step": 8920 }, { "grad_norm": 0.22849637269973755, "learning_rate": 9.20181261872619e-05, "loss": 0.0203, "step": 8930 }, { "grad_norm": 0.2586766481399536, "learning_rate": 9.19957062912391e-05, "loss": 0.0206, "step": 8940 }, { "grad_norm": 0.3252488672733307, "learning_rate": 9.197325769154704e-05, "loss": 0.0233, "step": 8950 }, { "grad_norm": 0.2750413417816162, "learning_rate": 9.195078040352914e-05, "loss": 0.027, "step": 8960 }, { "grad_norm": 0.2571771442890167, "learning_rate": 9.192827444254839e-05, "loss": 0.0213, "step": 8970 }, { "grad_norm": 0.19798806309700012, "learning_rate": 9.190573982398744e-05, "loss": 0.0236, "step": 8980 }, { "grad_norm": 0.31670165061950684, "learning_rate": 9.188317656324845e-05, "loss": 0.0223, "step": 8990 }, { "grad_norm": 0.21561455726623535, "learning_rate": 9.186058467575322e-05, "loss": 0.0262, "step": 9000 }, { "grad_norm": 0.34534984827041626, "learning_rate": 9.183796417694309e-05, "loss": 0.0245, "step": 9010 }, { "grad_norm": 0.2673737704753876, "learning_rate": 9.181531508227897e-05, "loss": 0.0179, "step": 9020 }, { "grad_norm": 0.22644372284412384, "learning_rate": 9.179263740724125e-05, "loss": 0.0238, "step": 9030 }, { "grad_norm": 0.28067800402641296, "learning_rate": 9.176993116732996e-05, "loss": 0.0237, "step": 9040 }, { "grad_norm": 0.20712649822235107, "learning_rate": 9.17471963780646e-05, "loss": 0.0214, "step": 9050 }, { "grad_norm": 0.21479865908622742, "learning_rate": 9.172443305498414e-05, "loss": 0.0207, "step": 9060 }, { "grad_norm": 0.29085516929626465, "learning_rate": 9.170164121364714e-05, "loss": 0.022, "step": 9070 }, { "grad_norm": 0.28230151534080505, "learning_rate": 9.167882086963158e-05, "loss": 0.0234, "step": 9080 }, { "grad_norm": 0.26531803607940674, "learning_rate": 9.165597203853499e-05, "loss": 0.0232, "step": 9090 }, { "grad_norm": 0.28021949529647827, "learning_rate": 9.163309473597428e-05, "loss": 0.0231, "step": 9100 }, { "grad_norm": 0.2633902132511139, "learning_rate": 9.161018897758591e-05, "loss": 0.0226, "step": 9110 }, { "grad_norm": 0.28415557742118835, "learning_rate": 9.158725477902572e-05, "loss": 0.0233, "step": 9120 }, { "grad_norm": 0.20668506622314453, "learning_rate": 9.156429215596905e-05, "loss": 0.0202, "step": 9130 }, { "grad_norm": 0.2438492625951767, "learning_rate": 9.154130112411061e-05, "loss": 0.0222, "step": 9140 }, { "grad_norm": 0.18606123328208923, "learning_rate": 9.151828169916455e-05, "loss": 0.02, "step": 9150 }, { "grad_norm": 0.2390253245830536, "learning_rate": 9.149523389686445e-05, "loss": 0.0207, "step": 9160 }, { "grad_norm": 0.27786266803741455, "learning_rate": 9.147215773296327e-05, "loss": 0.0228, "step": 9170 }, { "grad_norm": 0.30480721592903137, "learning_rate": 9.144905322323334e-05, "loss": 0.0206, "step": 9180 }, { "grad_norm": 0.26637908816337585, "learning_rate": 9.142592038346635e-05, "loss": 0.0221, "step": 9190 }, { "grad_norm": 0.22106043994426727, "learning_rate": 9.140275922947343e-05, "loss": 0.0225, "step": 9200 }, { "grad_norm": 0.2961888909339905, "learning_rate": 9.137956977708496e-05, "loss": 0.0229, "step": 9210 }, { "grad_norm": 0.2997179925441742, "learning_rate": 9.135635204215075e-05, "loss": 0.0235, "step": 9220 }, { "grad_norm": 0.2141832709312439, "learning_rate": 9.133310604053986e-05, "loss": 0.0217, "step": 9230 }, { "grad_norm": 0.28751102089881897, "learning_rate": 9.130983178814077e-05, "loss": 0.0241, "step": 9240 }, { "grad_norm": 0.22307386994361877, "learning_rate": 9.128652930086116e-05, "loss": 0.0203, "step": 9250 }, { "grad_norm": 0.3014931082725525, "learning_rate": 9.126319859462808e-05, "loss": 0.0208, "step": 9260 }, { "grad_norm": 0.1992233842611313, "learning_rate": 9.123983968538787e-05, "loss": 0.0214, "step": 9270 }, { "grad_norm": 0.268445760011673, "learning_rate": 9.121645258910609e-05, "loss": 0.0219, "step": 9280 }, { "grad_norm": 0.2437484860420227, "learning_rate": 9.119303732176762e-05, "loss": 0.0207, "step": 9290 }, { "grad_norm": 0.23573829233646393, "learning_rate": 9.116959389937656e-05, "loss": 0.0228, "step": 9300 }, { "grad_norm": 0.35171520709991455, "learning_rate": 9.114612233795628e-05, "loss": 0.0194, "step": 9310 }, { "grad_norm": 0.2682882845401764, "learning_rate": 9.112262265354937e-05, "loss": 0.0193, "step": 9320 }, { "grad_norm": 0.2331685870885849, "learning_rate": 9.109909486221763e-05, "loss": 0.0217, "step": 9330 }, { "grad_norm": 0.2812448740005493, "learning_rate": 9.107553898004208e-05, "loss": 0.0198, "step": 9340 }, { "grad_norm": 0.2757926881313324, "learning_rate": 9.105195502312298e-05, "loss": 0.0221, "step": 9350 }, { "grad_norm": 0.24807968735694885, "learning_rate": 9.102834300757972e-05, "loss": 0.021, "step": 9360 }, { "grad_norm": 0.24877645075321198, "learning_rate": 9.100470294955087e-05, "loss": 0.0179, "step": 9370 }, { "grad_norm": 0.23852942883968353, "learning_rate": 9.098103486519423e-05, "loss": 0.021, "step": 9380 }, { "grad_norm": 0.28451213240623474, "learning_rate": 9.095733877068666e-05, "loss": 0.0191, "step": 9390 }, { "grad_norm": 0.24236975610256195, "learning_rate": 9.093361468222426e-05, "loss": 0.022, "step": 9400 }, { "grad_norm": 0.30400004982948303, "learning_rate": 9.090986261602221e-05, "loss": 0.0186, "step": 9410 }, { "grad_norm": 0.28756698966026306, "learning_rate": 9.088608258831482e-05, "loss": 0.0216, "step": 9420 }, { "grad_norm": 0.24260251224040985, "learning_rate": 9.08622746153555e-05, "loss": 0.0236, "step": 9430 }, { "grad_norm": 0.23353083431720734, "learning_rate": 9.083843871341679e-05, "loss": 0.0188, "step": 9440 }, { "grad_norm": 0.26354339718818665, "learning_rate": 9.08145748987903e-05, "loss": 0.022, "step": 9450 }, { "grad_norm": 0.2628061771392822, "learning_rate": 9.079068318778674e-05, "loss": 0.0178, "step": 9460 }, { "grad_norm": 0.21343088150024414, "learning_rate": 9.076676359673586e-05, "loss": 0.0197, "step": 9470 }, { "grad_norm": 0.19352935254573822, "learning_rate": 9.074281614198645e-05, "loss": 0.0209, "step": 9480 }, { "grad_norm": 0.25158315896987915, "learning_rate": 9.071884083990638e-05, "loss": 0.0207, "step": 9490 }, { "grad_norm": 0.293828547000885, "learning_rate": 9.069483770688259e-05, "loss": 0.0194, "step": 9500 }, { "grad_norm": 0.2099379450082779, "learning_rate": 9.067080675932093e-05, "loss": 0.0209, "step": 9510 }, { "grad_norm": 0.20176348090171814, "learning_rate": 9.064674801364634e-05, "loss": 0.0184, "step": 9520 }, { "grad_norm": 0.2351505309343338, "learning_rate": 9.062266148630279e-05, "loss": 0.0151, "step": 9530 }, { "grad_norm": 0.208987295627594, "learning_rate": 9.059854719375313e-05, "loss": 0.0185, "step": 9540 }, { "grad_norm": 0.23470504581928253, "learning_rate": 9.057440515247932e-05, "loss": 0.0238, "step": 9550 }, { "grad_norm": 0.2120210826396942, "learning_rate": 9.055023537898215e-05, "loss": 0.0216, "step": 9560 }, { "grad_norm": 0.24216072261333466, "learning_rate": 9.052603788978149e-05, "loss": 0.0193, "step": 9570 }, { "grad_norm": 0.32790547609329224, "learning_rate": 9.050181270141606e-05, "loss": 0.0222, "step": 9580 }, { "grad_norm": 0.3319876790046692, "learning_rate": 9.047755983044355e-05, "loss": 0.0242, "step": 9590 }, { "grad_norm": 0.28189903497695923, "learning_rate": 9.045327929344058e-05, "loss": 0.0185, "step": 9600 }, { "grad_norm": 0.2554161548614502, "learning_rate": 9.042897110700268e-05, "loss": 0.0218, "step": 9610 }, { "grad_norm": 0.24834933876991272, "learning_rate": 9.040463528774423e-05, "loss": 0.0185, "step": 9620 }, { "grad_norm": 0.2864587903022766, "learning_rate": 9.038027185229856e-05, "loss": 0.0193, "step": 9630 }, { "grad_norm": 0.28729113936424255, "learning_rate": 9.035588081731784e-05, "loss": 0.0238, "step": 9640 }, { "grad_norm": 0.2366987019777298, "learning_rate": 9.033146219947313e-05, "loss": 0.0247, "step": 9650 }, { "grad_norm": 0.2232433706521988, "learning_rate": 9.030701601545429e-05, "loss": 0.0224, "step": 9660 }, { "grad_norm": 0.24947834014892578, "learning_rate": 9.028254228197009e-05, "loss": 0.0179, "step": 9670 }, { "grad_norm": 0.2361399382352829, "learning_rate": 9.025804101574808e-05, "loss": 0.0177, "step": 9680 }, { "grad_norm": 0.31840530037879944, "learning_rate": 9.023351223353463e-05, "loss": 0.0228, "step": 9690 }, { "grad_norm": 0.3714565634727478, "learning_rate": 9.020895595209498e-05, "loss": 0.0201, "step": 9700 }, { "grad_norm": 0.16689251363277435, "learning_rate": 9.018437218821305e-05, "loss": 0.0199, "step": 9710 }, { "grad_norm": 0.19709087908267975, "learning_rate": 9.015976095869166e-05, "loss": 0.0189, "step": 9720 }, { "grad_norm": 0.19766882061958313, "learning_rate": 9.013512228035235e-05, "loss": 0.018, "step": 9730 }, { "grad_norm": 0.21910995244979858, "learning_rate": 9.01104561700354e-05, "loss": 0.0198, "step": 9740 }, { "grad_norm": 0.35880985856056213, "learning_rate": 9.008576264459987e-05, "loss": 0.0214, "step": 9750 }, { "grad_norm": 0.2841029465198517, "learning_rate": 9.006104172092357e-05, "loss": 0.0205, "step": 9760 }, { "grad_norm": 0.24283187091350555, "learning_rate": 9.0036293415903e-05, "loss": 0.0184, "step": 9770 }, { "grad_norm": 0.2993418574333191, "learning_rate": 9.00115177464534e-05, "loss": 0.0167, "step": 9780 }, { "grad_norm": 0.20758523046970367, "learning_rate": 8.998671472950872e-05, "loss": 0.0178, "step": 9790 }, { "grad_norm": 0.25678426027297974, "learning_rate": 8.996188438202156e-05, "loss": 0.0193, "step": 9800 }, { "grad_norm": 0.2785191833972931, "learning_rate": 8.993702672096324e-05, "loss": 0.0284, "step": 9810 }, { "grad_norm": 0.2501772940158844, "learning_rate": 8.991214176332377e-05, "loss": 0.0229, "step": 9820 }, { "grad_norm": 0.1744813770055771, "learning_rate": 8.988722952611171e-05, "loss": 0.0242, "step": 9830 }, { "grad_norm": 0.22522394359111786, "learning_rate": 8.986229002635441e-05, "loss": 0.0225, "step": 9840 }, { "grad_norm": 0.3037087321281433, "learning_rate": 8.983732328109775e-05, "loss": 0.0231, "step": 9850 }, { "grad_norm": 0.2677419185638428, "learning_rate": 8.981232930740627e-05, "loss": 0.0195, "step": 9860 }, { "grad_norm": 0.1849060207605362, "learning_rate": 8.978730812236313e-05, "loss": 0.0163, "step": 9870 }, { "grad_norm": 0.24875833094120026, "learning_rate": 8.976225974307004e-05, "loss": 0.0176, "step": 9880 }, { "grad_norm": 0.2132263481616974, "learning_rate": 8.973718418664736e-05, "loss": 0.0179, "step": 9890 }, { "grad_norm": 0.2541581988334656, "learning_rate": 8.971208147023397e-05, "loss": 0.0186, "step": 9900 }, { "grad_norm": 0.3128814399242401, "learning_rate": 8.968695161098739e-05, "loss": 0.02, "step": 9910 }, { "grad_norm": 0.2846480906009674, "learning_rate": 8.96617946260836e-05, "loss": 0.0182, "step": 9920 }, { "grad_norm": 0.23419274389743805, "learning_rate": 8.963661053271716e-05, "loss": 0.0197, "step": 9930 }, { "grad_norm": 0.3069364130496979, "learning_rate": 8.961139934810117e-05, "loss": 0.0166, "step": 9940 }, { "grad_norm": 0.20356646180152893, "learning_rate": 8.958616108946725e-05, "loss": 0.0239, "step": 9950 }, { "grad_norm": 0.27593129873275757, "learning_rate": 8.95608957740655e-05, "loss": 0.0186, "step": 9960 }, { "grad_norm": 0.25585153698921204, "learning_rate": 8.953560341916452e-05, "loss": 0.0187, "step": 9970 }, { "grad_norm": 0.21519580483436584, "learning_rate": 8.95102840420514e-05, "loss": 0.0161, "step": 9980 }, { "grad_norm": 0.23993811011314392, "learning_rate": 8.94849376600317e-05, "loss": 0.018, "step": 9990 }, { "grad_norm": 0.3369923233985901, "learning_rate": 8.945956429042943e-05, "loss": 0.0196, "step": 10000 }, { "grad_norm": 0.24292726814746857, "learning_rate": 8.943416395058705e-05, "loss": 0.017, "step": 10010 }, { "grad_norm": 0.238583505153656, "learning_rate": 8.940873665786544e-05, "loss": 0.0205, "step": 10020 }, { "grad_norm": 0.28148210048675537, "learning_rate": 8.938328242964394e-05, "loss": 0.021, "step": 10030 }, { "grad_norm": 0.260987251996994, "learning_rate": 8.935780128332026e-05, "loss": 0.0185, "step": 10040 }, { "grad_norm": 0.28648218512535095, "learning_rate": 8.933229323631052e-05, "loss": 0.0178, "step": 10050 }, { "grad_norm": 0.1821424663066864, "learning_rate": 8.930675830604925e-05, "loss": 0.0192, "step": 10060 }, { "grad_norm": 0.20678821206092834, "learning_rate": 8.92811965099893e-05, "loss": 0.0185, "step": 10070 }, { "grad_norm": 0.22816571593284607, "learning_rate": 8.925560786560194e-05, "loss": 0.0253, "step": 10080 }, { "grad_norm": 0.2514851987361908, "learning_rate": 8.922999239037677e-05, "loss": 0.0188, "step": 10090 }, { "grad_norm": 0.2326006293296814, "learning_rate": 8.920435010182171e-05, "loss": 0.0186, "step": 10100 }, { "grad_norm": 0.19311098754405975, "learning_rate": 8.917868101746302e-05, "loss": 0.015, "step": 10110 }, { "grad_norm": 0.2547858655452728, "learning_rate": 8.91529851548453e-05, "loss": 0.0184, "step": 10120 }, { "grad_norm": 0.22439683973789215, "learning_rate": 8.912726253153142e-05, "loss": 0.0208, "step": 10130 }, { "grad_norm": 0.2569746971130371, "learning_rate": 8.910151316510255e-05, "loss": 0.0179, "step": 10140 }, { "grad_norm": 0.22633986175060272, "learning_rate": 8.907573707315813e-05, "loss": 0.0172, "step": 10150 }, { "grad_norm": 0.3043188452720642, "learning_rate": 8.904993427331588e-05, "loss": 0.0203, "step": 10160 }, { "grad_norm": 0.28739047050476074, "learning_rate": 8.902410478321176e-05, "loss": 0.0196, "step": 10170 }, { "grad_norm": 0.2847227156162262, "learning_rate": 8.899824862050002e-05, "loss": 0.0204, "step": 10180 }, { "grad_norm": 0.2433919906616211, "learning_rate": 8.897236580285308e-05, "loss": 0.0196, "step": 10190 }, { "grad_norm": 0.2799326777458191, "learning_rate": 8.894645634796159e-05, "loss": 0.0194, "step": 10200 }, { "grad_norm": 0.2597646713256836, "learning_rate": 8.892052027353444e-05, "loss": 0.0228, "step": 10210 }, { "grad_norm": 0.28474023938179016, "learning_rate": 8.889455759729866e-05, "loss": 0.0232, "step": 10220 }, { "grad_norm": 0.2596752345561981, "learning_rate": 8.886856833699955e-05, "loss": 0.0209, "step": 10230 }, { "grad_norm": 0.1954433023929596, "learning_rate": 8.884255251040046e-05, "loss": 0.0176, "step": 10240 }, { "grad_norm": 0.1979115903377533, "learning_rate": 8.8816510135283e-05, "loss": 0.0163, "step": 10250 }, { "grad_norm": 0.2627243399620056, "learning_rate": 8.879044122944688e-05, "loss": 0.0237, "step": 10260 }, { "grad_norm": 0.2071765959262848, "learning_rate": 8.876434581070996e-05, "loss": 0.0212, "step": 10270 }, { "grad_norm": 0.28134214878082275, "learning_rate": 8.87382238969082e-05, "loss": 0.0211, "step": 10280 }, { "grad_norm": 0.2962311804294586, "learning_rate": 8.871207550589568e-05, "loss": 0.0177, "step": 10290 }, { "grad_norm": 0.21433213353157043, "learning_rate": 8.868590065554458e-05, "loss": 0.0203, "step": 10300 }, { "grad_norm": 0.269072026014328, "learning_rate": 8.865969936374519e-05, "loss": 0.0185, "step": 10310 }, { "grad_norm": 0.28234946727752686, "learning_rate": 8.863347164840581e-05, "loss": 0.0183, "step": 10320 }, { "grad_norm": 0.268440842628479, "learning_rate": 8.860721752745285e-05, "loss": 0.0202, "step": 10330 }, { "grad_norm": 0.22771452367305756, "learning_rate": 8.858093701883077e-05, "loss": 0.0176, "step": 10340 }, { "grad_norm": 0.28804251551628113, "learning_rate": 8.8554630140502e-05, "loss": 0.0222, "step": 10350 }, { "grad_norm": 0.29729944467544556, "learning_rate": 8.85282969104471e-05, "loss": 0.021, "step": 10360 }, { "grad_norm": 0.23023201525211334, "learning_rate": 8.850193734666456e-05, "loss": 0.0191, "step": 10370 }, { "grad_norm": 0.32791566848754883, "learning_rate": 8.84755514671709e-05, "loss": 0.0173, "step": 10380 }, { "grad_norm": 0.194224551320076, "learning_rate": 8.84491392900006e-05, "loss": 0.0167, "step": 10390 }, { "grad_norm": 0.1706506311893463, "learning_rate": 8.842270083320617e-05, "loss": 0.0173, "step": 10400 }, { "grad_norm": 0.24446482956409454, "learning_rate": 8.839623611485801e-05, "loss": 0.0178, "step": 10410 }, { "grad_norm": 0.2398073375225067, "learning_rate": 8.836974515304453e-05, "loss": 0.0157, "step": 10420 }, { "grad_norm": 0.2215758115053177, "learning_rate": 8.834322796587204e-05, "loss": 0.0184, "step": 10430 }, { "grad_norm": 0.1658633053302765, "learning_rate": 8.831668457146478e-05, "loss": 0.017, "step": 10440 }, { "grad_norm": 0.25491926074028015, "learning_rate": 8.829011498796493e-05, "loss": 0.0163, "step": 10450 }, { "grad_norm": 0.21816888451576233, "learning_rate": 8.826351923353253e-05, "loss": 0.0175, "step": 10460 }, { "grad_norm": 0.20212550461292267, "learning_rate": 8.823689732634555e-05, "loss": 0.0184, "step": 10470 }, { "grad_norm": 0.1855219602584839, "learning_rate": 8.82102492845998e-05, "loss": 0.0164, "step": 10480 }, { "grad_norm": 0.22499176859855652, "learning_rate": 8.818357512650896e-05, "loss": 0.015, "step": 10490 }, { "grad_norm": 0.2750430107116699, "learning_rate": 8.815687487030458e-05, "loss": 0.0171, "step": 10500 }, { "grad_norm": 0.23293711245059967, "learning_rate": 8.8130148534236e-05, "loss": 0.0187, "step": 10510 }, { "grad_norm": 0.30794674158096313, "learning_rate": 8.810339613657047e-05, "loss": 0.0207, "step": 10520 }, { "grad_norm": 0.33584827184677124, "learning_rate": 8.807661769559295e-05, "loss": 0.0212, "step": 10530 }, { "grad_norm": 0.2485899031162262, "learning_rate": 8.804981322960628e-05, "loss": 0.0202, "step": 10540 }, { "grad_norm": 0.23962199687957764, "learning_rate": 8.802298275693106e-05, "loss": 0.0211, "step": 10550 }, { "grad_norm": 0.16858799755573273, "learning_rate": 8.799612629590568e-05, "loss": 0.0179, "step": 10560 }, { "grad_norm": 0.236466646194458, "learning_rate": 8.796924386488624e-05, "loss": 0.0194, "step": 10570 }, { "grad_norm": 0.2634967565536499, "learning_rate": 8.794233548224666e-05, "loss": 0.0169, "step": 10580 }, { "grad_norm": 0.2704890966415405, "learning_rate": 8.791540116637853e-05, "loss": 0.0211, "step": 10590 }, { "grad_norm": 0.3307843506336212, "learning_rate": 8.788844093569124e-05, "loss": 0.0202, "step": 10600 }, { "grad_norm": 0.16973918676376343, "learning_rate": 8.786145480861184e-05, "loss": 0.0215, "step": 10610 }, { "grad_norm": 0.27503058314323425, "learning_rate": 8.783444280358507e-05, "loss": 0.0238, "step": 10620 }, { "grad_norm": 0.17043958604335785, "learning_rate": 8.780740493907342e-05, "loss": 0.0177, "step": 10630 }, { "grad_norm": 0.2578222155570984, "learning_rate": 8.778034123355698e-05, "loss": 0.021, "step": 10640 }, { "grad_norm": 0.20597337186336517, "learning_rate": 8.775325170553357e-05, "loss": 0.0184, "step": 10650 }, { "grad_norm": 0.2437574565410614, "learning_rate": 8.77261363735186e-05, "loss": 0.0201, "step": 10660 }, { "grad_norm": 0.2230488806962967, "learning_rate": 8.769899525604517e-05, "loss": 0.0175, "step": 10670 }, { "grad_norm": 0.2237183153629303, "learning_rate": 8.767182837166397e-05, "loss": 0.0154, "step": 10680 }, { "grad_norm": 0.2459309846162796, "learning_rate": 8.764463573894328e-05, "loss": 0.0167, "step": 10690 }, { "grad_norm": 0.18342165648937225, "learning_rate": 8.761741737646902e-05, "loss": 0.018, "step": 10700 }, { "grad_norm": 0.22207790613174438, "learning_rate": 8.759017330284471e-05, "loss": 0.0153, "step": 10710 }, { "grad_norm": 0.2800830602645874, "learning_rate": 8.756290353669142e-05, "loss": 0.0195, "step": 10720 }, { "grad_norm": 0.22740083932876587, "learning_rate": 8.753560809664774e-05, "loss": 0.0192, "step": 10730 }, { "grad_norm": 0.2839958965778351, "learning_rate": 8.750828700136986e-05, "loss": 0.0174, "step": 10740 }, { "grad_norm": 0.2245020717382431, "learning_rate": 8.74809402695315e-05, "loss": 0.0166, "step": 10750 }, { "grad_norm": 0.21464531123638153, "learning_rate": 8.745356791982391e-05, "loss": 0.0183, "step": 10760 }, { "grad_norm": 0.19358249008655548, "learning_rate": 8.742616997095578e-05, "loss": 0.0191, "step": 10770 }, { "grad_norm": 0.26851004362106323, "learning_rate": 8.739874644165341e-05, "loss": 0.0169, "step": 10780 }, { "grad_norm": 0.26004698872566223, "learning_rate": 8.737129735066048e-05, "loss": 0.0208, "step": 10790 }, { "grad_norm": 0.2639443874359131, "learning_rate": 8.734382271673821e-05, "loss": 0.0186, "step": 10800 }, { "grad_norm": 0.2537703812122345, "learning_rate": 8.731632255866525e-05, "loss": 0.0144, "step": 10810 }, { "grad_norm": 0.22399237751960754, "learning_rate": 8.728879689523767e-05, "loss": 0.0182, "step": 10820 }, { "grad_norm": 0.19250230491161346, "learning_rate": 8.726124574526905e-05, "loss": 0.0197, "step": 10830 }, { "grad_norm": 0.22725102305412292, "learning_rate": 8.72336691275903e-05, "loss": 0.0177, "step": 10840 }, { "grad_norm": 0.2699919641017914, "learning_rate": 8.720606706104979e-05, "loss": 0.0168, "step": 10850 }, { "grad_norm": 0.20184136927127838, "learning_rate": 8.71784395645133e-05, "loss": 0.0184, "step": 10860 }, { "grad_norm": 0.20289894938468933, "learning_rate": 8.715078665686392e-05, "loss": 0.0181, "step": 10870 }, { "grad_norm": 0.22391708195209503, "learning_rate": 8.712310835700218e-05, "loss": 0.0167, "step": 10880 }, { "grad_norm": 0.29346707463264465, "learning_rate": 8.709540468384591e-05, "loss": 0.0195, "step": 10890 }, { "grad_norm": 0.21001645922660828, "learning_rate": 8.706767565633033e-05, "loss": 0.017, "step": 10900 }, { "grad_norm": 0.23137769103050232, "learning_rate": 8.7039921293408e-05, "loss": 0.0179, "step": 10910 }, { "grad_norm": 0.2131565660238266, "learning_rate": 8.70121416140487e-05, "loss": 0.0193, "step": 10920 }, { "grad_norm": 0.2943631410598755, "learning_rate": 8.698433663723962e-05, "loss": 0.0178, "step": 10930 }, { "grad_norm": 0.2603611946105957, "learning_rate": 8.695650638198518e-05, "loss": 0.0184, "step": 10940 }, { "grad_norm": 0.2487758845090866, "learning_rate": 8.692865086730713e-05, "loss": 0.0153, "step": 10950 }, { "grad_norm": 0.397062212228775, "learning_rate": 8.69007701122444e-05, "loss": 0.0158, "step": 10960 }, { "grad_norm": 0.18474280834197998, "learning_rate": 8.687286413585328e-05, "loss": 0.016, "step": 10970 }, { "grad_norm": 0.2143416553735733, "learning_rate": 8.684493295720719e-05, "loss": 0.0185, "step": 10980 }, { "grad_norm": 0.22584691643714905, "learning_rate": 8.681697659539685e-05, "loss": 0.0201, "step": 10990 }, { "grad_norm": 0.20893818140029907, "learning_rate": 8.678899506953019e-05, "loss": 0.0172, "step": 11000 }, { "grad_norm": 0.16990144550800323, "learning_rate": 8.676098839873227e-05, "loss": 0.0152, "step": 11010 }, { "grad_norm": 0.1813032627105713, "learning_rate": 8.673295660214545e-05, "loss": 0.0173, "step": 11020 }, { "grad_norm": 0.20561788976192474, "learning_rate": 8.670489969892914e-05, "loss": 0.0189, "step": 11030 }, { "grad_norm": 0.20943671464920044, "learning_rate": 8.667681770826e-05, "loss": 0.0167, "step": 11040 }, { "grad_norm": 0.2430121898651123, "learning_rate": 8.66487106493318e-05, "loss": 0.0183, "step": 11050 }, { "grad_norm": 0.20991162955760956, "learning_rate": 8.662057854135544e-05, "loss": 0.0172, "step": 11060 }, { "grad_norm": 0.23417285084724426, "learning_rate": 8.659242140355897e-05, "loss": 0.018, "step": 11070 }, { "grad_norm": 0.25205478072166443, "learning_rate": 8.65642392551875e-05, "loss": 0.0188, "step": 11080 }, { "grad_norm": 0.24377375841140747, "learning_rate": 8.65360321155033e-05, "loss": 0.0159, "step": 11090 }, { "grad_norm": 0.21934157609939575, "learning_rate": 8.650780000378566e-05, "loss": 0.0177, "step": 11100 }, { "grad_norm": 0.2642960846424103, "learning_rate": 8.647954293933096e-05, "loss": 0.0173, "step": 11110 }, { "grad_norm": 0.20923936367034912, "learning_rate": 8.645126094145264e-05, "loss": 0.0165, "step": 11120 }, { "grad_norm": 0.22267261147499084, "learning_rate": 8.642295402948117e-05, "loss": 0.0174, "step": 11130 }, { "grad_norm": 0.16724027693271637, "learning_rate": 8.639462222276409e-05, "loss": 0.0179, "step": 11140 }, { "grad_norm": 0.1703285127878189, "learning_rate": 8.636626554066589e-05, "loss": 0.0194, "step": 11150 }, { "grad_norm": 0.25311410427093506, "learning_rate": 8.633788400256811e-05, "loss": 0.0142, "step": 11160 }, { "grad_norm": 0.22370311617851257, "learning_rate": 8.630947762786927e-05, "loss": 0.0155, "step": 11170 }, { "grad_norm": 0.22217294573783875, "learning_rate": 8.628104643598483e-05, "loss": 0.0193, "step": 11180 }, { "grad_norm": 0.24603894352912903, "learning_rate": 8.625259044634726e-05, "loss": 0.0165, "step": 11190 }, { "grad_norm": 0.162809357047081, "learning_rate": 8.622410967840597e-05, "loss": 0.0171, "step": 11200 }, { "grad_norm": 0.20372973382472992, "learning_rate": 8.619560415162731e-05, "loss": 0.0202, "step": 11210 }, { "grad_norm": 0.26284265518188477, "learning_rate": 8.616707388549447e-05, "loss": 0.0168, "step": 11220 }, { "grad_norm": 0.23133176565170288, "learning_rate": 8.613851889950771e-05, "loss": 0.0167, "step": 11230 }, { "grad_norm": 0.2587195932865143, "learning_rate": 8.610993921318402e-05, "loss": 0.016, "step": 11240 }, { "grad_norm": 0.3157714009284973, "learning_rate": 8.608133484605738e-05, "loss": 0.0183, "step": 11250 }, { "grad_norm": 0.1984710991382599, "learning_rate": 8.605270581767859e-05, "loss": 0.0201, "step": 11260 }, { "grad_norm": 0.20762470364570618, "learning_rate": 8.602405214761536e-05, "loss": 0.0219, "step": 11270 }, { "grad_norm": 0.1641400307416916, "learning_rate": 8.599537385545215e-05, "loss": 0.016, "step": 11280 }, { "grad_norm": 0.16980759799480438, "learning_rate": 8.596667096079032e-05, "loss": 0.0145, "step": 11290 }, { "grad_norm": 0.2361089140176773, "learning_rate": 8.593794348324806e-05, "loss": 0.0162, "step": 11300 }, { "grad_norm": 0.3061775267124176, "learning_rate": 8.590919144246028e-05, "loss": 0.021, "step": 11310 }, { "grad_norm": 0.1976073980331421, "learning_rate": 8.588041485807876e-05, "loss": 0.0156, "step": 11320 }, { "grad_norm": 0.20734791457653046, "learning_rate": 8.585161374977202e-05, "loss": 0.0168, "step": 11330 }, { "grad_norm": 0.26678434014320374, "learning_rate": 8.582278813722533e-05, "loss": 0.019, "step": 11340 }, { "grad_norm": 0.24402450025081635, "learning_rate": 8.579393804014076e-05, "loss": 0.0201, "step": 11350 }, { "grad_norm": 0.2961506247520447, "learning_rate": 8.576506347823703e-05, "loss": 0.0156, "step": 11360 }, { "grad_norm": 0.2733178436756134, "learning_rate": 8.573616447124968e-05, "loss": 0.018, "step": 11370 }, { "grad_norm": 0.24802494049072266, "learning_rate": 8.570724103893086e-05, "loss": 0.0193, "step": 11380 }, { "grad_norm": 0.17027230560779572, "learning_rate": 8.567829320104951e-05, "loss": 0.0177, "step": 11390 }, { "grad_norm": 0.22346360981464386, "learning_rate": 8.564932097739118e-05, "loss": 0.0166, "step": 11400 }, { "grad_norm": 0.30938804149627686, "learning_rate": 8.562032438775811e-05, "loss": 0.0213, "step": 11410 }, { "grad_norm": 0.2043396383523941, "learning_rate": 8.559130345196921e-05, "loss": 0.0222, "step": 11420 }, { "grad_norm": 0.19620200991630554, "learning_rate": 8.556225818986e-05, "loss": 0.0211, "step": 11430 }, { "grad_norm": 0.3068983554840088, "learning_rate": 8.553318862128265e-05, "loss": 0.0192, "step": 11440 }, { "grad_norm": 0.25059545040130615, "learning_rate": 8.550409476610593e-05, "loss": 0.0164, "step": 11450 }, { "grad_norm": 0.1617857664823532, "learning_rate": 8.547497664421522e-05, "loss": 0.0174, "step": 11460 }, { "grad_norm": 0.22919046878814697, "learning_rate": 8.544583427551249e-05, "loss": 0.018, "step": 11470 }, { "grad_norm": 0.28143110871315, "learning_rate": 8.541666767991628e-05, "loss": 0.0182, "step": 11480 }, { "grad_norm": 0.18151907622814178, "learning_rate": 8.538747687736166e-05, "loss": 0.0169, "step": 11490 }, { "grad_norm": 0.2363259494304657, "learning_rate": 8.535826188780026e-05, "loss": 0.0198, "step": 11500 }, { "grad_norm": 0.19042979180812836, "learning_rate": 8.532902273120029e-05, "loss": 0.0171, "step": 11510 }, { "grad_norm": 0.21291981637477875, "learning_rate": 8.52997594275464e-05, "loss": 0.0177, "step": 11520 }, { "grad_norm": 0.2557770609855652, "learning_rate": 8.52704719968398e-05, "loss": 0.0172, "step": 11530 }, { "grad_norm": 0.2567228376865387, "learning_rate": 8.524116045909818e-05, "loss": 0.019, "step": 11540 }, { "grad_norm": 0.21545298397541046, "learning_rate": 8.521182483435569e-05, "loss": 0.0142, "step": 11550 }, { "grad_norm": 0.2192620486021042, "learning_rate": 8.518246514266295e-05, "loss": 0.0193, "step": 11560 }, { "grad_norm": 0.3180059790611267, "learning_rate": 8.515308140408703e-05, "loss": 0.021, "step": 11570 }, { "grad_norm": 0.19352670013904572, "learning_rate": 8.512367363871145e-05, "loss": 0.0156, "step": 11580 }, { "grad_norm": 0.28733739256858826, "learning_rate": 8.509424186663614e-05, "loss": 0.0152, "step": 11590 }, { "grad_norm": 0.2525884807109833, "learning_rate": 8.506478610797743e-05, "loss": 0.0142, "step": 11600 }, { "grad_norm": 0.25093188881874084, "learning_rate": 8.503530638286805e-05, "loss": 0.0161, "step": 11610 }, { "grad_norm": 0.30053433775901794, "learning_rate": 8.500580271145712e-05, "loss": 0.0203, "step": 11620 }, { "grad_norm": 0.23738454282283783, "learning_rate": 8.497627511391014e-05, "loss": 0.016, "step": 11630 }, { "grad_norm": 0.17790959775447845, "learning_rate": 8.494672361040891e-05, "loss": 0.0162, "step": 11640 }, { "grad_norm": 0.23480264842510223, "learning_rate": 8.491714822115162e-05, "loss": 0.0171, "step": 11650 }, { "grad_norm": 0.2443733513355255, "learning_rate": 8.488754896635277e-05, "loss": 0.0172, "step": 11660 }, { "grad_norm": 0.25445374846458435, "learning_rate": 8.485792586624317e-05, "loss": 0.0152, "step": 11670 }, { "grad_norm": 0.20005185902118683, "learning_rate": 8.482827894106993e-05, "loss": 0.0158, "step": 11680 }, { "grad_norm": 0.24546605348587036, "learning_rate": 8.479860821109646e-05, "loss": 0.0203, "step": 11690 }, { "grad_norm": 0.2099747210741043, "learning_rate": 8.476891369660239e-05, "loss": 0.0204, "step": 11700 }, { "grad_norm": 0.26853132247924805, "learning_rate": 8.473919541788366e-05, "loss": 0.0167, "step": 11710 }, { "grad_norm": 0.24925141036510468, "learning_rate": 8.470945339525245e-05, "loss": 0.0198, "step": 11720 }, { "grad_norm": 0.2118704468011856, "learning_rate": 8.467968764903713e-05, "loss": 0.0137, "step": 11730 }, { "grad_norm": 0.22732135653495789, "learning_rate": 8.46498981995823e-05, "loss": 0.0153, "step": 11740 }, { "grad_norm": 0.3187790811061859, "learning_rate": 8.462008506724879e-05, "loss": 0.016, "step": 11750 }, { "grad_norm": 0.2472190260887146, "learning_rate": 8.459024827241359e-05, "loss": 0.0176, "step": 11760 }, { "grad_norm": 0.286607027053833, "learning_rate": 8.456038783546985e-05, "loss": 0.0169, "step": 11770 }, { "grad_norm": 0.14528051018714905, "learning_rate": 8.453050377682691e-05, "loss": 0.0207, "step": 11780 }, { "grad_norm": 0.2881283462047577, "learning_rate": 8.450059611691026e-05, "loss": 0.0166, "step": 11790 }, { "grad_norm": 0.28947120904922485, "learning_rate": 8.447066487616146e-05, "loss": 0.0212, "step": 11800 }, { "grad_norm": 0.2122950553894043, "learning_rate": 8.444071007503826e-05, "loss": 0.0168, "step": 11810 }, { "grad_norm": 0.22447514533996582, "learning_rate": 8.441073173401449e-05, "loss": 0.0202, "step": 11820 }, { "grad_norm": 0.17744508385658264, "learning_rate": 8.438072987358006e-05, "loss": 0.0172, "step": 11830 }, { "grad_norm": 0.24819692969322205, "learning_rate": 8.435070451424094e-05, "loss": 0.0163, "step": 11840 }, { "grad_norm": 0.17840507626533508, "learning_rate": 8.432065567651919e-05, "loss": 0.0192, "step": 11850 }, { "grad_norm": 0.25329193472862244, "learning_rate": 8.429058338095291e-05, "loss": 0.0195, "step": 11860 }, { "grad_norm": 0.19052043557167053, "learning_rate": 8.426048764809624e-05, "loss": 0.0207, "step": 11870 }, { "grad_norm": 0.240098237991333, "learning_rate": 8.423036849851932e-05, "loss": 0.0172, "step": 11880 }, { "grad_norm": 0.1974698007106781, "learning_rate": 8.42002259528083e-05, "loss": 0.0183, "step": 11890 }, { "grad_norm": 0.23387902975082397, "learning_rate": 8.417006003156532e-05, "loss": 0.025, "step": 11900 }, { "grad_norm": 0.2164762318134308, "learning_rate": 8.413987075540852e-05, "loss": 0.0153, "step": 11910 }, { "grad_norm": 0.35051724314689636, "learning_rate": 8.4109658144972e-05, "loss": 0.0175, "step": 11920 }, { "grad_norm": 0.30202800035476685, "learning_rate": 8.407942222090573e-05, "loss": 0.0188, "step": 11930 }, { "grad_norm": 0.2101394236087799, "learning_rate": 8.404916300387576e-05, "loss": 0.0156, "step": 11940 }, { "grad_norm": 0.2207673192024231, "learning_rate": 8.401888051456391e-05, "loss": 0.0206, "step": 11950 }, { "grad_norm": 0.16465787589550018, "learning_rate": 8.398857477366803e-05, "loss": 0.0187, "step": 11960 }, { "grad_norm": 0.19596551358699799, "learning_rate": 8.395824580190178e-05, "loss": 0.0166, "step": 11970 }, { "grad_norm": 0.22551023960113525, "learning_rate": 8.392789361999473e-05, "loss": 0.0201, "step": 11980 }, { "grad_norm": 0.2649100422859192, "learning_rate": 8.38975182486923e-05, "loss": 0.0181, "step": 11990 }, { "grad_norm": 0.19009381532669067, "learning_rate": 8.386711970875581e-05, "loss": 0.0176, "step": 12000 }, { "grad_norm": 0.27448245882987976, "learning_rate": 8.383669802096232e-05, "loss": 0.022, "step": 12010 }, { "grad_norm": 0.2876286804676056, "learning_rate": 8.38062532061048e-05, "loss": 0.0193, "step": 12020 }, { "grad_norm": 0.22156919538974762, "learning_rate": 8.3775785284992e-05, "loss": 0.019, "step": 12030 }, { "grad_norm": 0.2172042727470398, "learning_rate": 8.374529427844843e-05, "loss": 0.0202, "step": 12040 }, { "grad_norm": 0.19533245265483856, "learning_rate": 8.371478020731442e-05, "loss": 0.0159, "step": 12050 }, { "grad_norm": 0.3024807572364807, "learning_rate": 8.368424309244607e-05, "loss": 0.0156, "step": 12060 }, { "grad_norm": 0.2364368885755539, "learning_rate": 8.365368295471517e-05, "loss": 0.0171, "step": 12070 }, { "grad_norm": 0.32288405299186707, "learning_rate": 8.362309981500931e-05, "loss": 0.0168, "step": 12080 }, { "grad_norm": 0.18788859248161316, "learning_rate": 8.359249369423177e-05, "loss": 0.0158, "step": 12090 }, { "grad_norm": 0.2032059282064438, "learning_rate": 8.356186461330155e-05, "loss": 0.0175, "step": 12100 }, { "grad_norm": 0.18046404421329498, "learning_rate": 8.353121259315334e-05, "loss": 0.0155, "step": 12110 }, { "grad_norm": 0.16062216460704803, "learning_rate": 8.350053765473751e-05, "loss": 0.0184, "step": 12120 }, { "grad_norm": 0.18520262837409973, "learning_rate": 8.346983981902005e-05, "loss": 0.0162, "step": 12130 }, { "grad_norm": 0.2384960800409317, "learning_rate": 8.343911910698271e-05, "loss": 0.0134, "step": 12140 }, { "grad_norm": 0.2092033475637436, "learning_rate": 8.340837553962278e-05, "loss": 0.0155, "step": 12150 }, { "grad_norm": 0.2541310489177704, "learning_rate": 8.337760913795316e-05, "loss": 0.0194, "step": 12160 }, { "grad_norm": 0.22502322494983673, "learning_rate": 8.334681992300244e-05, "loss": 0.0195, "step": 12170 }, { "grad_norm": 0.18429136276245117, "learning_rate": 8.331600791581475e-05, "loss": 0.0146, "step": 12180 }, { "grad_norm": 0.24344132840633392, "learning_rate": 8.328517313744978e-05, "loss": 0.015, "step": 12190 }, { "grad_norm": 0.26929402351379395, "learning_rate": 8.325431560898286e-05, "loss": 0.019, "step": 12200 }, { "grad_norm": 0.22950400412082672, "learning_rate": 8.322343535150478e-05, "loss": 0.0158, "step": 12210 }, { "grad_norm": 0.21809828281402588, "learning_rate": 8.319253238612191e-05, "loss": 0.0167, "step": 12220 }, { "grad_norm": 0.2522887885570526, "learning_rate": 8.316160673395614e-05, "loss": 0.015, "step": 12230 }, { "grad_norm": 0.22579337656497955, "learning_rate": 8.313065841614487e-05, "loss": 0.0161, "step": 12240 }, { "grad_norm": 0.1962343007326126, "learning_rate": 8.309968745384096e-05, "loss": 0.0174, "step": 12250 }, { "grad_norm": 0.23641446232795715, "learning_rate": 8.306869386821282e-05, "loss": 0.016, "step": 12260 }, { "grad_norm": 0.2875693440437317, "learning_rate": 8.30376776804442e-05, "loss": 0.0194, "step": 12270 }, { "grad_norm": 0.22400115430355072, "learning_rate": 8.300663891173443e-05, "loss": 0.0171, "step": 12280 }, { "grad_norm": 0.19768598675727844, "learning_rate": 8.297557758329822e-05, "loss": 0.0168, "step": 12290 }, { "grad_norm": 0.1840212345123291, "learning_rate": 8.294449371636564e-05, "loss": 0.0178, "step": 12300 }, { "grad_norm": 0.23551127314567566, "learning_rate": 8.291338733218226e-05, "loss": 0.0184, "step": 12310 }, { "grad_norm": 0.21753227710723877, "learning_rate": 8.2882258452009e-05, "loss": 0.0138, "step": 12320 }, { "grad_norm": 0.21922315657138824, "learning_rate": 8.285110709712214e-05, "loss": 0.018, "step": 12330 }, { "grad_norm": 0.2387281358242035, "learning_rate": 8.281993328881337e-05, "loss": 0.0205, "step": 12340 }, { "grad_norm": 0.19247713685035706, "learning_rate": 8.278873704838964e-05, "loss": 0.0173, "step": 12350 }, { "grad_norm": 0.2610059678554535, "learning_rate": 8.275751839717334e-05, "loss": 0.0214, "step": 12360 }, { "grad_norm": 0.21669019758701324, "learning_rate": 8.272627735650208e-05, "loss": 0.0176, "step": 12370 }, { "grad_norm": 0.21433617174625397, "learning_rate": 8.269501394772884e-05, "loss": 0.017, "step": 12380 }, { "grad_norm": 0.24689456820487976, "learning_rate": 8.266372819222189e-05, "loss": 0.016, "step": 12390 }, { "grad_norm": 0.26369670033454895, "learning_rate": 8.26324201113647e-05, "loss": 0.0188, "step": 12400 }, { "grad_norm": 0.19424651563167572, "learning_rate": 8.260108972655606e-05, "loss": 0.0156, "step": 12410 }, { "grad_norm": 0.28712067008018494, "learning_rate": 8.256973705921e-05, "loss": 0.0164, "step": 12420 }, { "grad_norm": 0.13783171772956848, "learning_rate": 8.25383621307558e-05, "loss": 0.0138, "step": 12430 }, { "grad_norm": 0.2528391182422638, "learning_rate": 8.25069649626379e-05, "loss": 0.015, "step": 12440 }, { "grad_norm": 0.2284558117389679, "learning_rate": 8.247554557631596e-05, "loss": 0.0239, "step": 12450 }, { "grad_norm": 0.21876786649227142, "learning_rate": 8.244410399326483e-05, "loss": 0.0165, "step": 12460 }, { "grad_norm": 0.20239873230457306, "learning_rate": 8.241264023497457e-05, "loss": 0.0203, "step": 12470 }, { "grad_norm": 0.25739961862564087, "learning_rate": 8.238115432295034e-05, "loss": 0.0159, "step": 12480 }, { "grad_norm": 0.23907962441444397, "learning_rate": 8.234964627871247e-05, "loss": 0.0153, "step": 12490 }, { "grad_norm": 0.19462871551513672, "learning_rate": 8.231811612379639e-05, "loss": 0.0137, "step": 12500 }, { "grad_norm": 0.2074728012084961, "learning_rate": 8.228656387975268e-05, "loss": 0.0154, "step": 12510 }, { "grad_norm": 0.22061991691589355, "learning_rate": 8.225498956814702e-05, "loss": 0.0162, "step": 12520 }, { "grad_norm": 0.24021898210048676, "learning_rate": 8.222339321056014e-05, "loss": 0.0177, "step": 12530 }, { "grad_norm": 0.2286812961101532, "learning_rate": 8.219177482858785e-05, "loss": 0.0168, "step": 12540 }, { "grad_norm": 0.20762425661087036, "learning_rate": 8.216013444384099e-05, "loss": 0.0163, "step": 12550 }, { "grad_norm": 0.20838916301727295, "learning_rate": 8.21284720779455e-05, "loss": 0.0148, "step": 12560 }, { "grad_norm": 0.17453067004680634, "learning_rate": 8.209678775254231e-05, "loss": 0.0152, "step": 12570 }, { "grad_norm": 0.19089099764823914, "learning_rate": 8.206508148928733e-05, "loss": 0.0148, "step": 12580 }, { "grad_norm": 0.21460378170013428, "learning_rate": 8.203335330985151e-05, "loss": 0.0153, "step": 12590 }, { "grad_norm": 0.248976930975914, "learning_rate": 8.200160323592076e-05, "loss": 0.0167, "step": 12600 }, { "grad_norm": 0.18852567672729492, "learning_rate": 8.196983128919598e-05, "loss": 0.0159, "step": 12610 }, { "grad_norm": 0.21633735299110413, "learning_rate": 8.193803749139295e-05, "loss": 0.0188, "step": 12620 }, { "grad_norm": 0.24848385155200958, "learning_rate": 8.190622186424244e-05, "loss": 0.0165, "step": 12630 }, { "grad_norm": 0.3166060447692871, "learning_rate": 8.187438442949016e-05, "loss": 0.0169, "step": 12640 }, { "grad_norm": 0.29790422320365906, "learning_rate": 8.184252520889668e-05, "loss": 0.0174, "step": 12650 }, { "grad_norm": 0.23300330340862274, "learning_rate": 8.181064422423748e-05, "loss": 0.0131, "step": 12660 }, { "grad_norm": 0.23252488672733307, "learning_rate": 8.177874149730289e-05, "loss": 0.0144, "step": 12670 }, { "grad_norm": 0.2059391736984253, "learning_rate": 8.174681704989816e-05, "loss": 0.0162, "step": 12680 }, { "grad_norm": 0.19714505970478058, "learning_rate": 8.171487090384333e-05, "loss": 0.0173, "step": 12690 }, { "grad_norm": 0.21539516746997833, "learning_rate": 8.168290308097328e-05, "loss": 0.0136, "step": 12700 }, { "grad_norm": 0.16343377530574799, "learning_rate": 8.165091360313774e-05, "loss": 0.0131, "step": 12710 }, { "grad_norm": 0.19965815544128418, "learning_rate": 8.161890249220119e-05, "loss": 0.015, "step": 12720 }, { "grad_norm": 0.23422792553901672, "learning_rate": 8.158686977004295e-05, "loss": 0.0137, "step": 12730 }, { "grad_norm": 0.1944032907485962, "learning_rate": 8.155481545855706e-05, "loss": 0.0137, "step": 12740 }, { "grad_norm": 0.16840232908725739, "learning_rate": 8.152273957965233e-05, "loss": 0.0137, "step": 12750 }, { "grad_norm": 0.21794182062149048, "learning_rate": 8.149064215525237e-05, "loss": 0.0157, "step": 12760 }, { "grad_norm": 0.30277231335639954, "learning_rate": 8.14585232072954e-05, "loss": 0.0164, "step": 12770 }, { "grad_norm": 0.2569328844547272, "learning_rate": 8.142638275773449e-05, "loss": 0.0187, "step": 12780 }, { "grad_norm": 0.22498908638954163, "learning_rate": 8.139422082853729e-05, "loss": 0.0136, "step": 12790 }, { "grad_norm": 0.22042568027973175, "learning_rate": 8.136203744168618e-05, "loss": 0.0152, "step": 12800 }, { "grad_norm": 0.22383084893226624, "learning_rate": 8.132983261917819e-05, "loss": 0.0142, "step": 12810 }, { "grad_norm": 0.2473316639661789, "learning_rate": 8.129760638302504e-05, "loss": 0.0169, "step": 12820 }, { "grad_norm": 0.2341151088476181, "learning_rate": 8.126535875525305e-05, "loss": 0.0158, "step": 12830 }, { "grad_norm": 0.22613975405693054, "learning_rate": 8.123308975790316e-05, "loss": 0.0185, "step": 12840 }, { "grad_norm": 0.24011293053627014, "learning_rate": 8.120079941303094e-05, "loss": 0.0167, "step": 12850 }, { "grad_norm": 0.20219309628009796, "learning_rate": 8.116848774270651e-05, "loss": 0.0179, "step": 12860 }, { "grad_norm": 0.2534615099430084, "learning_rate": 8.113615476901461e-05, "loss": 0.0158, "step": 12870 }, { "grad_norm": 0.19790571928024292, "learning_rate": 8.110380051405454e-05, "loss": 0.0145, "step": 12880 }, { "grad_norm": 0.23862721025943756, "learning_rate": 8.107142499994009e-05, "loss": 0.0163, "step": 12890 }, { "grad_norm": 0.18184104561805725, "learning_rate": 8.103902824879966e-05, "loss": 0.0166, "step": 12900 }, { "grad_norm": 0.26385995745658875, "learning_rate": 8.10066102827761e-05, "loss": 0.0148, "step": 12910 }, { "grad_norm": 0.20102955400943756, "learning_rate": 8.097417112402676e-05, "loss": 0.0154, "step": 12920 }, { "grad_norm": 0.16969600319862366, "learning_rate": 8.094171079472355e-05, "loss": 0.0156, "step": 12930 }, { "grad_norm": 0.2270621508359909, "learning_rate": 8.090922931705277e-05, "loss": 0.0165, "step": 12940 }, { "grad_norm": 0.1695328801870346, "learning_rate": 8.08767267132152e-05, "loss": 0.0123, "step": 12950 }, { "grad_norm": 0.2504426836967468, "learning_rate": 8.084420300542608e-05, "loss": 0.015, "step": 12960 }, { "grad_norm": 0.22081902623176575, "learning_rate": 8.081165821591505e-05, "loss": 0.0133, "step": 12970 }, { "grad_norm": 0.2575734257698059, "learning_rate": 8.077909236692615e-05, "loss": 0.0162, "step": 12980 }, { "grad_norm": 0.30112403631210327, "learning_rate": 8.074650548071787e-05, "loss": 0.0226, "step": 12990 }, { "grad_norm": 0.19585277140140533, "learning_rate": 8.071389757956301e-05, "loss": 0.0146, "step": 13000 }, { "grad_norm": 0.2293144166469574, "learning_rate": 8.068126868574876e-05, "loss": 0.017, "step": 13010 }, { "grad_norm": 0.2770582139492035, "learning_rate": 8.064861882157668e-05, "loss": 0.0161, "step": 13020 }, { "grad_norm": 0.16200146079063416, "learning_rate": 8.061594800936263e-05, "loss": 0.0148, "step": 13030 }, { "grad_norm": 0.2417653650045395, "learning_rate": 8.058325627143681e-05, "loss": 0.0212, "step": 13040 }, { "grad_norm": 0.25398463010787964, "learning_rate": 8.055054363014372e-05, "loss": 0.0181, "step": 13050 }, { "grad_norm": 0.1901843100786209, "learning_rate": 8.051781010784211e-05, "loss": 0.0143, "step": 13060 }, { "grad_norm": 0.21282416582107544, "learning_rate": 8.048505572690506e-05, "loss": 0.013, "step": 13070 }, { "grad_norm": 0.23239676654338837, "learning_rate": 8.045228050971988e-05, "loss": 0.0128, "step": 13080 }, { "grad_norm": 0.25679001212120056, "learning_rate": 8.041948447868814e-05, "loss": 0.0143, "step": 13090 }, { "grad_norm": 0.226441890001297, "learning_rate": 8.038666765622558e-05, "loss": 0.0126, "step": 13100 }, { "grad_norm": 0.23862886428833008, "learning_rate": 8.03538300647622e-05, "loss": 0.0161, "step": 13110 }, { "grad_norm": 0.20606917142868042, "learning_rate": 8.03209717267422e-05, "loss": 0.0143, "step": 13120 }, { "grad_norm": 0.19791308045387268, "learning_rate": 8.028809266462395e-05, "loss": 0.0158, "step": 13130 }, { "grad_norm": 0.2334667295217514, "learning_rate": 8.025519290087994e-05, "loss": 0.0163, "step": 13140 }, { "grad_norm": 0.2516237497329712, "learning_rate": 8.022227245799688e-05, "loss": 0.0177, "step": 13150 }, { "grad_norm": 0.1960330307483673, "learning_rate": 8.018933135847557e-05, "loss": 0.0148, "step": 13160 }, { "grad_norm": 0.1997160166501999, "learning_rate": 8.015636962483096e-05, "loss": 0.0182, "step": 13170 }, { "grad_norm": 0.25103020668029785, "learning_rate": 8.012338727959205e-05, "loss": 0.0163, "step": 13180 }, { "grad_norm": 0.2165667563676834, "learning_rate": 8.009038434530198e-05, "loss": 0.0175, "step": 13190 }, { "grad_norm": 0.18595504760742188, "learning_rate": 8.005736084451796e-05, "loss": 0.0166, "step": 13200 }, { "grad_norm": 0.22005362808704376, "learning_rate": 8.002431679981122e-05, "loss": 0.0145, "step": 13210 }, { "grad_norm": 0.20132414996623993, "learning_rate": 7.999125223376706e-05, "loss": 0.0174, "step": 13220 }, { "grad_norm": 0.20195187628269196, "learning_rate": 7.99581671689848e-05, "loss": 0.0173, "step": 13230 }, { "grad_norm": 0.2409990280866623, "learning_rate": 7.992506162807775e-05, "loss": 0.0184, "step": 13240 }, { "grad_norm": 0.1937837153673172, "learning_rate": 7.989193563367328e-05, "loss": 0.0158, "step": 13250 }, { "grad_norm": 0.23448792099952698, "learning_rate": 7.985878920841266e-05, "loss": 0.0151, "step": 13260 }, { "grad_norm": 0.17841683328151703, "learning_rate": 7.982562237495117e-05, "loss": 0.0137, "step": 13270 }, { "grad_norm": 0.1822482943534851, "learning_rate": 7.979243515595802e-05, "loss": 0.0146, "step": 13280 }, { "grad_norm": 0.18212221562862396, "learning_rate": 7.975922757411636e-05, "loss": 0.0116, "step": 13290 }, { "grad_norm": 0.2389056384563446, "learning_rate": 7.972599965212329e-05, "loss": 0.0159, "step": 13300 }, { "grad_norm": 0.2592625617980957, "learning_rate": 7.969275141268973e-05, "loss": 0.0148, "step": 13310 }, { "grad_norm": 0.2461392730474472, "learning_rate": 7.96594828785406e-05, "loss": 0.0152, "step": 13320 }, { "grad_norm": 0.15440548956394196, "learning_rate": 7.962619407241456e-05, "loss": 0.015, "step": 13330 }, { "grad_norm": 0.2526731789112091, "learning_rate": 7.959288501706424e-05, "loss": 0.0142, "step": 13340 }, { "grad_norm": 0.1938915252685547, "learning_rate": 7.955955573525605e-05, "loss": 0.0152, "step": 13350 }, { "grad_norm": 0.1778085082769394, "learning_rate": 7.952620624977026e-05, "loss": 0.0157, "step": 13360 }, { "grad_norm": 0.24643193185329437, "learning_rate": 7.949283658340089e-05, "loss": 0.0162, "step": 13370 }, { "grad_norm": 0.17042769491672516, "learning_rate": 7.945944675895585e-05, "loss": 0.0135, "step": 13380 }, { "grad_norm": 0.20310942828655243, "learning_rate": 7.942603679925671e-05, "loss": 0.0131, "step": 13390 }, { "grad_norm": 0.17970015108585358, "learning_rate": 7.93926067271389e-05, "loss": 0.0151, "step": 13400 }, { "grad_norm": 0.2084377557039261, "learning_rate": 7.935915656545155e-05, "loss": 0.0132, "step": 13410 }, { "grad_norm": 0.21621963381767273, "learning_rate": 7.932568633705752e-05, "loss": 0.013, "step": 13420 }, { "grad_norm": 0.16308289766311646, "learning_rate": 7.929219606483341e-05, "loss": 0.0161, "step": 13430 }, { "grad_norm": 0.22773639857769012, "learning_rate": 7.925868577166948e-05, "loss": 0.0169, "step": 13440 }, { "grad_norm": 0.22066883742809296, "learning_rate": 7.922515548046974e-05, "loss": 0.0153, "step": 13450 }, { "grad_norm": 0.21871130168437958, "learning_rate": 7.919160521415179e-05, "loss": 0.0174, "step": 13460 }, { "grad_norm": 0.20689034461975098, "learning_rate": 7.915803499564694e-05, "loss": 0.0186, "step": 13470 }, { "grad_norm": 0.26997652649879456, "learning_rate": 7.912444484790013e-05, "loss": 0.0158, "step": 13480 }, { "grad_norm": 0.18438412249088287, "learning_rate": 7.909083479386987e-05, "loss": 0.0137, "step": 13490 }, { "grad_norm": 0.2005113661289215, "learning_rate": 7.905720485652836e-05, "loss": 0.0142, "step": 13500 }, { "grad_norm": 0.22723375260829926, "learning_rate": 7.902355505886132e-05, "loss": 0.0161, "step": 13510 }, { "grad_norm": 0.22111943364143372, "learning_rate": 7.898988542386805e-05, "loss": 0.0121, "step": 13520 }, { "grad_norm": 0.23269201815128326, "learning_rate": 7.895619597456147e-05, "loss": 0.0126, "step": 13530 }, { "grad_norm": 0.20017966628074646, "learning_rate": 7.892248673396798e-05, "loss": 0.0122, "step": 13540 }, { "grad_norm": 0.19939561188220978, "learning_rate": 7.888875772512754e-05, "loss": 0.0133, "step": 13550 }, { "grad_norm": 0.19058066606521606, "learning_rate": 7.885500897109359e-05, "loss": 0.018, "step": 13560 }, { "grad_norm": 0.22339694201946259, "learning_rate": 7.882124049493309e-05, "loss": 0.015, "step": 13570 }, { "grad_norm": 0.230096697807312, "learning_rate": 7.878745231972649e-05, "loss": 0.0136, "step": 13580 }, { "grad_norm": 0.20560744404792786, "learning_rate": 7.875364446856766e-05, "loss": 0.0155, "step": 13590 }, { "grad_norm": 0.28056764602661133, "learning_rate": 7.871981696456398e-05, "loss": 0.0129, "step": 13600 }, { "grad_norm": 0.2935844659805298, "learning_rate": 7.868596983083623e-05, "loss": 0.0159, "step": 13610 }, { "grad_norm": 0.2542419731616974, "learning_rate": 7.865210309051858e-05, "loss": 0.014, "step": 13620 }, { "grad_norm": 0.16813772916793823, "learning_rate": 7.861821676675863e-05, "loss": 0.0146, "step": 13630 }, { "grad_norm": 0.2003648579120636, "learning_rate": 7.858431088271739e-05, "loss": 0.0152, "step": 13640 }, { "grad_norm": 0.20491527020931244, "learning_rate": 7.855038546156918e-05, "loss": 0.0152, "step": 13650 }, { "grad_norm": 0.21193893253803253, "learning_rate": 7.851644052650173e-05, "loss": 0.0127, "step": 13660 }, { "grad_norm": 0.23122936487197876, "learning_rate": 7.848247610071609e-05, "loss": 0.0154, "step": 13670 }, { "grad_norm": 0.3063417077064514, "learning_rate": 7.844849220742658e-05, "loss": 0.0155, "step": 13680 }, { "grad_norm": 0.23822136223316193, "learning_rate": 7.841448886986092e-05, "loss": 0.016, "step": 13690 }, { "grad_norm": 0.21274591982364655, "learning_rate": 7.838046611126004e-05, "loss": 0.0143, "step": 13700 }, { "grad_norm": 0.2024126946926117, "learning_rate": 7.834642395487819e-05, "loss": 0.0164, "step": 13710 }, { "grad_norm": 0.21326279640197754, "learning_rate": 7.831236242398285e-05, "loss": 0.0174, "step": 13720 }, { "grad_norm": 0.27649152278900146, "learning_rate": 7.827828154185477e-05, "loss": 0.0149, "step": 13730 }, { "grad_norm": 0.27162566781044006, "learning_rate": 7.82441813317879e-05, "loss": 0.0171, "step": 13740 }, { "grad_norm": 0.2244277000427246, "learning_rate": 7.821006181708944e-05, "loss": 0.0139, "step": 13750 }, { "grad_norm": 0.27550002932548523, "learning_rate": 7.81759230210797e-05, "loss": 0.0136, "step": 13760 }, { "grad_norm": 0.1914265900850296, "learning_rate": 7.814176496709227e-05, "loss": 0.016, "step": 13770 }, { "grad_norm": 0.18624737858772278, "learning_rate": 7.810758767847385e-05, "loss": 0.0174, "step": 13780 }, { "grad_norm": 0.24825862050056458, "learning_rate": 7.807339117858427e-05, "loss": 0.0139, "step": 13790 }, { "grad_norm": 0.22932690382003784, "learning_rate": 7.803917549079655e-05, "loss": 0.015, "step": 13800 }, { "grad_norm": 0.14860840141773224, "learning_rate": 7.800494063849679e-05, "loss": 0.0132, "step": 13810 }, { "grad_norm": 0.23330315947532654, "learning_rate": 7.797068664508416e-05, "loss": 0.0148, "step": 13820 }, { "grad_norm": 0.20748861134052277, "learning_rate": 7.793641353397096e-05, "loss": 0.0124, "step": 13830 }, { "grad_norm": 0.20250585675239563, "learning_rate": 7.790212132858253e-05, "loss": 0.0135, "step": 13840 }, { "grad_norm": 0.209207221865654, "learning_rate": 7.786781005235728e-05, "loss": 0.0151, "step": 13850 }, { "grad_norm": 0.18320150673389435, "learning_rate": 7.783347972874662e-05, "loss": 0.0164, "step": 13860 }, { "grad_norm": 0.23593679070472717, "learning_rate": 7.779913038121504e-05, "loss": 0.0146, "step": 13870 }, { "grad_norm": 0.240645170211792, "learning_rate": 7.776476203323997e-05, "loss": 0.0147, "step": 13880 }, { "grad_norm": 0.15381722152233124, "learning_rate": 7.773037470831185e-05, "loss": 0.0138, "step": 13890 }, { "grad_norm": 0.2189253866672516, "learning_rate": 7.76959684299341e-05, "loss": 0.0161, "step": 13900 }, { "grad_norm": 0.21748559176921844, "learning_rate": 7.76615432216231e-05, "loss": 0.012, "step": 13910 }, { "grad_norm": 0.22933423519134521, "learning_rate": 7.762709910690811e-05, "loss": 0.0127, "step": 13920 }, { "grad_norm": 0.20984798669815063, "learning_rate": 7.759263610933141e-05, "loss": 0.0132, "step": 13930 }, { "grad_norm": 0.16750794649124146, "learning_rate": 7.755815425244811e-05, "loss": 0.0129, "step": 13940 }, { "grad_norm": 0.2410721480846405, "learning_rate": 7.752365355982624e-05, "loss": 0.0145, "step": 13950 }, { "grad_norm": 0.22969438135623932, "learning_rate": 7.748913405504668e-05, "loss": 0.014, "step": 13960 }, { "grad_norm": 0.2862030565738678, "learning_rate": 7.745459576170322e-05, "loss": 0.0145, "step": 13970 }, { "grad_norm": 0.21458926796913147, "learning_rate": 7.742003870340242e-05, "loss": 0.0152, "step": 13980 }, { "grad_norm": 0.30864301323890686, "learning_rate": 7.738546290376373e-05, "loss": 0.0164, "step": 13990 }, { "grad_norm": 0.21145440638065338, "learning_rate": 7.735086838641937e-05, "loss": 0.0204, "step": 14000 }, { "grad_norm": 0.17594081163406372, "learning_rate": 7.731625517501437e-05, "loss": 0.012, "step": 14010 }, { "grad_norm": 0.1991441398859024, "learning_rate": 7.728162329320655e-05, "loss": 0.0141, "step": 14020 }, { "grad_norm": 0.23093664646148682, "learning_rate": 7.724697276466645e-05, "loss": 0.013, "step": 14030 }, { "grad_norm": 0.19602474570274353, "learning_rate": 7.721230361307738e-05, "loss": 0.0163, "step": 14040 }, { "grad_norm": 0.20275533199310303, "learning_rate": 7.71776158621354e-05, "loss": 0.0164, "step": 14050 }, { "grad_norm": 0.24430714547634125, "learning_rate": 7.714290953554925e-05, "loss": 0.0151, "step": 14060 }, { "grad_norm": 0.20464542508125305, "learning_rate": 7.710818465704037e-05, "loss": 0.0147, "step": 14070 }, { "grad_norm": 0.24801324307918549, "learning_rate": 7.707344125034288e-05, "loss": 0.0139, "step": 14080 }, { "grad_norm": 0.2292875200510025, "learning_rate": 7.703867933920359e-05, "loss": 0.0153, "step": 14090 }, { "grad_norm": 0.1830536276102066, "learning_rate": 7.700389894738194e-05, "loss": 0.0142, "step": 14100 }, { "grad_norm": 0.19027842581272125, "learning_rate": 7.696910009864999e-05, "loss": 0.0131, "step": 14110 }, { "grad_norm": 0.19808807969093323, "learning_rate": 7.693428281679241e-05, "loss": 0.0146, "step": 14120 }, { "grad_norm": 0.24026983976364136, "learning_rate": 7.689944712560652e-05, "loss": 0.014, "step": 14130 }, { "grad_norm": 0.22731636464595795, "learning_rate": 7.686459304890214e-05, "loss": 0.0132, "step": 14140 }, { "grad_norm": 0.23374275863170624, "learning_rate": 7.682972061050175e-05, "loss": 0.0151, "step": 14150 }, { "grad_norm": 0.2590828537940979, "learning_rate": 7.679482983424032e-05, "loss": 0.0165, "step": 14160 }, { "grad_norm": 0.2637365460395813, "learning_rate": 7.675992074396534e-05, "loss": 0.0141, "step": 14170 }, { "grad_norm": 0.2404571771621704, "learning_rate": 7.672499336353687e-05, "loss": 0.0185, "step": 14180 }, { "grad_norm": 0.19881968200206757, "learning_rate": 7.669004771682744e-05, "loss": 0.0147, "step": 14190 }, { "grad_norm": 0.21700164675712585, "learning_rate": 7.665508382772206e-05, "loss": 0.0172, "step": 14200 }, { "grad_norm": 0.17308397591114044, "learning_rate": 7.662010172011824e-05, "loss": 0.0118, "step": 14210 }, { "grad_norm": 0.19997023046016693, "learning_rate": 7.658510141792588e-05, "loss": 0.014, "step": 14220 }, { "grad_norm": 0.2513800263404846, "learning_rate": 7.65500829450674e-05, "loss": 0.0151, "step": 14230 }, { "grad_norm": 0.21157430112361908, "learning_rate": 7.651504632547759e-05, "loss": 0.0174, "step": 14240 }, { "grad_norm": 0.1981220692396164, "learning_rate": 7.647999158310364e-05, "loss": 0.0163, "step": 14250 }, { "grad_norm": 0.19800713658332825, "learning_rate": 7.644491874190512e-05, "loss": 0.0147, "step": 14260 }, { "grad_norm": 0.22546017169952393, "learning_rate": 7.6409827825854e-05, "loss": 0.0151, "step": 14270 }, { "grad_norm": 0.18834318220615387, "learning_rate": 7.637471885893459e-05, "loss": 0.0146, "step": 14280 }, { "grad_norm": 0.20382994413375854, "learning_rate": 7.633959186514354e-05, "loss": 0.0146, "step": 14290 }, { "grad_norm": 0.285525918006897, "learning_rate": 7.630444686848984e-05, "loss": 0.0169, "step": 14300 }, { "grad_norm": 0.26423659920692444, "learning_rate": 7.626928389299471e-05, "loss": 0.0145, "step": 14310 }, { "grad_norm": 0.23797716200351715, "learning_rate": 7.623410296269175e-05, "loss": 0.016, "step": 14320 }, { "grad_norm": 0.2137494683265686, "learning_rate": 7.61989041016268e-05, "loss": 0.0164, "step": 14330 }, { "grad_norm": 0.1847822219133377, "learning_rate": 7.616368733385793e-05, "loss": 0.0126, "step": 14340 }, { "grad_norm": 0.19659796357154846, "learning_rate": 7.612845268345547e-05, "loss": 0.0148, "step": 14350 }, { "grad_norm": 0.14658315479755402, "learning_rate": 7.609320017450199e-05, "loss": 0.0129, "step": 14360 }, { "grad_norm": 0.15584145486354828, "learning_rate": 7.605792983109222e-05, "loss": 0.0109, "step": 14370 }, { "grad_norm": 0.24578715860843658, "learning_rate": 7.602264167733313e-05, "loss": 0.0127, "step": 14380 }, { "grad_norm": 0.2046518325805664, "learning_rate": 7.598733573734384e-05, "loss": 0.0164, "step": 14390 }, { "grad_norm": 0.2313671112060547, "learning_rate": 7.595201203525561e-05, "loss": 0.0157, "step": 14400 }, { "grad_norm": 0.17536744475364685, "learning_rate": 7.591667059521187e-05, "loss": 0.0162, "step": 14410 }, { "grad_norm": 0.198593407869339, "learning_rate": 7.588131144136815e-05, "loss": 0.0151, "step": 14420 }, { "grad_norm": 0.20818327367305756, "learning_rate": 7.584593459789212e-05, "loss": 0.0136, "step": 14430 }, { "grad_norm": 0.1928614377975464, "learning_rate": 7.58105400889635e-05, "loss": 0.0133, "step": 14440 }, { "grad_norm": 0.2064107060432434, "learning_rate": 7.57751279387741e-05, "loss": 0.015, "step": 14450 }, { "grad_norm": 0.18215227127075195, "learning_rate": 7.573969817152782e-05, "loss": 0.0142, "step": 14460 }, { "grad_norm": 0.1847037523984909, "learning_rate": 7.570425081144052e-05, "loss": 0.0154, "step": 14470 }, { "grad_norm": 0.19525524973869324, "learning_rate": 7.56687858827402e-05, "loss": 0.0129, "step": 14480 }, { "grad_norm": 0.16676677763462067, "learning_rate": 7.563330340966675e-05, "loss": 0.0157, "step": 14490 }, { "grad_norm": 0.22655132412910461, "learning_rate": 7.559780341647212e-05, "loss": 0.0137, "step": 14500 }, { "grad_norm": 0.17722177505493164, "learning_rate": 7.556228592742026e-05, "loss": 0.0155, "step": 14510 }, { "grad_norm": 0.20414656400680542, "learning_rate": 7.552675096678696e-05, "loss": 0.0131, "step": 14520 }, { "grad_norm": 0.18890370428562164, "learning_rate": 7.549119855886012e-05, "loss": 0.0136, "step": 14530 }, { "grad_norm": 0.21458584070205688, "learning_rate": 7.545562872793941e-05, "loss": 0.0162, "step": 14540 }, { "grad_norm": 0.19247253239154816, "learning_rate": 7.542004149833648e-05, "loss": 0.0125, "step": 14550 }, { "grad_norm": 0.1668834537267685, "learning_rate": 7.538443689437492e-05, "loss": 0.0134, "step": 14560 }, { "grad_norm": 0.22504843771457672, "learning_rate": 7.53488149403901e-05, "loss": 0.0155, "step": 14570 }, { "grad_norm": 0.2331830859184265, "learning_rate": 7.531317566072929e-05, "loss": 0.0148, "step": 14580 }, { "grad_norm": 0.2085515409708023, "learning_rate": 7.527751907975158e-05, "loss": 0.0138, "step": 14590 }, { "grad_norm": 0.16348083317279816, "learning_rate": 7.524184522182793e-05, "loss": 0.0132, "step": 14600 }, { "grad_norm": 0.20645838975906372, "learning_rate": 7.520615411134112e-05, "loss": 0.0134, "step": 14610 }, { "grad_norm": 0.1703270971775055, "learning_rate": 7.517044577268564e-05, "loss": 0.0154, "step": 14620 }, { "grad_norm": 0.21221433579921722, "learning_rate": 7.513472023026782e-05, "loss": 0.015, "step": 14630 }, { "grad_norm": 0.21166104078292847, "learning_rate": 7.509897750850572e-05, "loss": 0.0137, "step": 14640 }, { "grad_norm": 0.25022852420806885, "learning_rate": 7.506321763182918e-05, "loss": 0.0119, "step": 14650 }, { "grad_norm": 0.19630014896392822, "learning_rate": 7.50274406246797e-05, "loss": 0.0181, "step": 14660 }, { "grad_norm": 0.18185678124427795, "learning_rate": 7.499164651151056e-05, "loss": 0.0131, "step": 14670 }, { "grad_norm": 0.1680784523487091, "learning_rate": 7.495583531678669e-05, "loss": 0.0138, "step": 14680 }, { "grad_norm": 0.2087288796901703, "learning_rate": 7.492000706498469e-05, "loss": 0.0134, "step": 14690 }, { "grad_norm": 0.23795852065086365, "learning_rate": 7.488416178059284e-05, "loss": 0.0138, "step": 14700 }, { "grad_norm": 0.17482201755046844, "learning_rate": 7.484829948811107e-05, "loss": 0.0124, "step": 14710 }, { "grad_norm": 0.2314142882823944, "learning_rate": 7.48124202120509e-05, "loss": 0.0145, "step": 14720 }, { "grad_norm": 0.18848326802253723, "learning_rate": 7.477652397693549e-05, "loss": 0.0133, "step": 14730 }, { "grad_norm": 0.1845630258321762, "learning_rate": 7.474061080729955e-05, "loss": 0.0171, "step": 14740 }, { "grad_norm": 0.17894284427165985, "learning_rate": 7.470468072768941e-05, "loss": 0.0133, "step": 14750 }, { "grad_norm": 0.20717774331569672, "learning_rate": 7.466873376266297e-05, "loss": 0.0127, "step": 14760 }, { "grad_norm": 0.1792818307876587, "learning_rate": 7.46327699367896e-05, "loss": 0.0138, "step": 14770 }, { "grad_norm": 0.18773500621318817, "learning_rate": 7.459678927465026e-05, "loss": 0.0122, "step": 14780 }, { "grad_norm": 0.1678897738456726, "learning_rate": 7.456079180083737e-05, "loss": 0.0132, "step": 14790 }, { "grad_norm": 0.2153242826461792, "learning_rate": 7.452477753995489e-05, "loss": 0.0125, "step": 14800 }, { "grad_norm": 0.21711304783821106, "learning_rate": 7.448874651661823e-05, "loss": 0.0117, "step": 14810 }, { "grad_norm": 0.23085622489452362, "learning_rate": 7.445269875545423e-05, "loss": 0.0141, "step": 14820 }, { "grad_norm": 0.1893828809261322, "learning_rate": 7.44166342811012e-05, "loss": 0.0139, "step": 14830 }, { "grad_norm": 0.17103269696235657, "learning_rate": 7.438055311820886e-05, "loss": 0.0155, "step": 14840 }, { "grad_norm": 0.2909722924232483, "learning_rate": 7.434445529143837e-05, "loss": 0.014, "step": 14850 }, { "grad_norm": 0.1612793207168579, "learning_rate": 7.430834082546225e-05, "loss": 0.0116, "step": 14860 }, { "grad_norm": 0.1990976333618164, "learning_rate": 7.427220974496438e-05, "loss": 0.0125, "step": 14870 }, { "grad_norm": 0.16523180902004242, "learning_rate": 7.423606207464005e-05, "loss": 0.0151, "step": 14880 }, { "grad_norm": 0.2934200167655945, "learning_rate": 7.419989783919578e-05, "loss": 0.0154, "step": 14890 }, { "grad_norm": 0.2150823473930359, "learning_rate": 7.416371706334956e-05, "loss": 0.0137, "step": 14900 }, { "grad_norm": 0.16800040006637573, "learning_rate": 7.412751977183056e-05, "loss": 0.0127, "step": 14910 }, { "grad_norm": 0.190083310008049, "learning_rate": 7.409130598937932e-05, "loss": 0.0155, "step": 14920 }, { "grad_norm": 0.17586246132850647, "learning_rate": 7.40550757407476e-05, "loss": 0.0144, "step": 14930 }, { "grad_norm": 0.22874070703983307, "learning_rate": 7.401882905069843e-05, "loss": 0.016, "step": 14940 }, { "grad_norm": 0.22014883160591125, "learning_rate": 7.39825659440061e-05, "loss": 0.0122, "step": 14950 }, { "grad_norm": 0.21213938295841217, "learning_rate": 7.394628644545609e-05, "loss": 0.0126, "step": 14960 }, { "grad_norm": 0.1822558045387268, "learning_rate": 7.390999057984507e-05, "loss": 0.0189, "step": 14970 }, { "grad_norm": 0.1730339080095291, "learning_rate": 7.387367837198097e-05, "loss": 0.0136, "step": 14980 }, { "grad_norm": 0.16640335321426392, "learning_rate": 7.383734984668281e-05, "loss": 0.0131, "step": 14990 }, { "grad_norm": 0.23079530894756317, "learning_rate": 7.38010050287808e-05, "loss": 0.0148, "step": 15000 }, { "grad_norm": 0.195553719997406, "learning_rate": 7.376464394311628e-05, "loss": 0.0127, "step": 15010 }, { "grad_norm": 0.22883008420467377, "learning_rate": 7.372826661454172e-05, "loss": 0.0138, "step": 15020 }, { "grad_norm": 0.22127346694469452, "learning_rate": 7.369187306792068e-05, "loss": 0.0117, "step": 15030 }, { "grad_norm": 0.22120708227157593, "learning_rate": 7.365546332812779e-05, "loss": 0.0112, "step": 15040 }, { "grad_norm": 0.22375716269016266, "learning_rate": 7.361903742004876e-05, "loss": 0.0123, "step": 15050 }, { "grad_norm": 0.20092467963695526, "learning_rate": 7.358259536858039e-05, "loss": 0.0107, "step": 15060 }, { "grad_norm": 0.1906672716140747, "learning_rate": 7.354613719863044e-05, "loss": 0.0114, "step": 15070 }, { "grad_norm": 0.1525750309228897, "learning_rate": 7.350966293511776e-05, "loss": 0.0131, "step": 15080 }, { "grad_norm": 0.2100876271724701, "learning_rate": 7.347317260297212e-05, "loss": 0.0125, "step": 15090 }, { "grad_norm": 0.17300279438495636, "learning_rate": 7.343666622713437e-05, "loss": 0.0135, "step": 15100 }, { "grad_norm": 0.2094549834728241, "learning_rate": 7.340014383255624e-05, "loss": 0.0137, "step": 15110 }, { "grad_norm": 0.18845538794994354, "learning_rate": 7.336360544420044e-05, "loss": 0.0124, "step": 15120 }, { "grad_norm": 0.25401628017425537, "learning_rate": 7.332705108704064e-05, "loss": 0.0139, "step": 15130 }, { "grad_norm": 0.2257518619298935, "learning_rate": 7.329048078606138e-05, "loss": 0.017, "step": 15140 }, { "grad_norm": 0.20038171112537384, "learning_rate": 7.32538945662581e-05, "loss": 0.0165, "step": 15150 }, { "grad_norm": 0.17932122945785522, "learning_rate": 7.321729245263718e-05, "loss": 0.016, "step": 15160 }, { "grad_norm": 0.20804426074028015, "learning_rate": 7.318067447021578e-05, "loss": 0.0146, "step": 15170 }, { "grad_norm": 0.18695075809955597, "learning_rate": 7.314404064402198e-05, "loss": 0.0145, "step": 15180 }, { "grad_norm": 0.21627949178218842, "learning_rate": 7.310739099909461e-05, "loss": 0.0164, "step": 15190 }, { "grad_norm": 0.2315109819173813, "learning_rate": 7.307072556048339e-05, "loss": 0.0146, "step": 15200 }, { "grad_norm": 0.28492677211761475, "learning_rate": 7.30340443532488e-05, "loss": 0.0159, "step": 15210 }, { "grad_norm": 0.23442929983139038, "learning_rate": 7.299734740246208e-05, "loss": 0.014, "step": 15220 }, { "grad_norm": 0.1998727172613144, "learning_rate": 7.296063473320528e-05, "loss": 0.0128, "step": 15230 }, { "grad_norm": 0.1873689889907837, "learning_rate": 7.292390637057113e-05, "loss": 0.0142, "step": 15240 }, { "grad_norm": 0.17843392491340637, "learning_rate": 7.288716233966314e-05, "loss": 0.0134, "step": 15250 }, { "grad_norm": 0.2235615998506546, "learning_rate": 7.285040266559551e-05, "loss": 0.0149, "step": 15260 }, { "grad_norm": 0.20769090950489044, "learning_rate": 7.281362737349312e-05, "loss": 0.0126, "step": 15270 }, { "grad_norm": 0.27400484681129456, "learning_rate": 7.277683648849153e-05, "loss": 0.0147, "step": 15280 }, { "grad_norm": 0.2078503668308258, "learning_rate": 7.2740030035737e-05, "loss": 0.0132, "step": 15290 }, { "grad_norm": 0.16756433248519897, "learning_rate": 7.270320804038634e-05, "loss": 0.0124, "step": 15300 }, { "grad_norm": 0.18765461444854736, "learning_rate": 7.266637052760708e-05, "loss": 0.0119, "step": 15310 }, { "grad_norm": 0.2619507908821106, "learning_rate": 7.262951752257728e-05, "loss": 0.0138, "step": 15320 }, { "grad_norm": 0.14576967060565948, "learning_rate": 7.259264905048564e-05, "loss": 0.0124, "step": 15330 }, { "grad_norm": 0.22171777486801147, "learning_rate": 7.255576513653142e-05, "loss": 0.0139, "step": 15340 }, { "grad_norm": 0.23593313992023468, "learning_rate": 7.251886580592439e-05, "loss": 0.0147, "step": 15350 }, { "grad_norm": 0.1614341288805008, "learning_rate": 7.248195108388496e-05, "loss": 0.0139, "step": 15360 }, { "grad_norm": 0.18240545690059662, "learning_rate": 7.244502099564395e-05, "loss": 0.0163, "step": 15370 }, { "grad_norm": 0.2552392780780792, "learning_rate": 7.240807556644271e-05, "loss": 0.0156, "step": 15380 }, { "grad_norm": 0.17230820655822754, "learning_rate": 7.237111482153314e-05, "loss": 0.0142, "step": 15390 }, { "grad_norm": 0.196395605802536, "learning_rate": 7.233413878617751e-05, "loss": 0.0142, "step": 15400 }, { "grad_norm": 0.23924414813518524, "learning_rate": 7.229714748564864e-05, "loss": 0.0142, "step": 15410 }, { "grad_norm": 0.1811290830373764, "learning_rate": 7.22601409452297e-05, "loss": 0.0132, "step": 15420 }, { "grad_norm": 0.22539308667182922, "learning_rate": 7.222311919021433e-05, "loss": 0.0148, "step": 15430 }, { "grad_norm": 0.1644640862941742, "learning_rate": 7.218608224590655e-05, "loss": 0.0161, "step": 15440 }, { "grad_norm": 0.21822653710842133, "learning_rate": 7.214903013762074e-05, "loss": 0.0128, "step": 15450 }, { "grad_norm": 0.1642625778913498, "learning_rate": 7.21119628906817e-05, "loss": 0.014, "step": 15460 }, { "grad_norm": 0.21434438228607178, "learning_rate": 7.207488053042454e-05, "loss": 0.0158, "step": 15470 }, { "grad_norm": 0.24862536787986755, "learning_rate": 7.203778308219467e-05, "loss": 0.0164, "step": 15480 }, { "grad_norm": 0.3047561049461365, "learning_rate": 7.200067057134787e-05, "loss": 0.0211, "step": 15490 }, { "grad_norm": 0.27686813473701477, "learning_rate": 7.196354302325019e-05, "loss": 0.0174, "step": 15500 }, { "grad_norm": 0.21062923967838287, "learning_rate": 7.192640046327795e-05, "loss": 0.0155, "step": 15510 }, { "grad_norm": 0.21044205129146576, "learning_rate": 7.188924291681777e-05, "loss": 0.0149, "step": 15520 }, { "grad_norm": 0.1823122799396515, "learning_rate": 7.185207040926643e-05, "loss": 0.0147, "step": 15530 }, { "grad_norm": 0.19005687534809113, "learning_rate": 7.181488296603103e-05, "loss": 0.0125, "step": 15540 }, { "grad_norm": 0.21500226855278015, "learning_rate": 7.177768061252885e-05, "loss": 0.0145, "step": 15550 }, { "grad_norm": 0.2189425379037857, "learning_rate": 7.174046337418729e-05, "loss": 0.0131, "step": 15560 }, { "grad_norm": 0.16784660518169403, "learning_rate": 7.170323127644403e-05, "loss": 0.0124, "step": 15570 }, { "grad_norm": 0.23555439710617065, "learning_rate": 7.166598434474683e-05, "loss": 0.0114, "step": 15580 }, { "grad_norm": 0.1869751363992691, "learning_rate": 7.162872260455364e-05, "loss": 0.0143, "step": 15590 }, { "grad_norm": 0.20053590834140778, "learning_rate": 7.159144608133248e-05, "loss": 0.015, "step": 15600 }, { "grad_norm": 0.20073874294757843, "learning_rate": 7.155415480056153e-05, "loss": 0.0138, "step": 15610 }, { "grad_norm": 0.20738346874713898, "learning_rate": 7.151684878772902e-05, "loss": 0.0124, "step": 15620 }, { "grad_norm": 0.1707349419593811, "learning_rate": 7.147952806833324e-05, "loss": 0.013, "step": 15630 }, { "grad_norm": 0.16149358451366425, "learning_rate": 7.14421926678826e-05, "loss": 0.0137, "step": 15640 }, { "grad_norm": 0.21355143189430237, "learning_rate": 7.140484261189543e-05, "loss": 0.0147, "step": 15650 }, { "grad_norm": 0.25564366579055786, "learning_rate": 7.136747792590017e-05, "loss": 0.016, "step": 15660 }, { "grad_norm": 0.20612789690494537, "learning_rate": 7.133009863543524e-05, "loss": 0.0135, "step": 15670 }, { "grad_norm": 0.22040824592113495, "learning_rate": 7.129270476604901e-05, "loss": 0.0141, "step": 15680 }, { "grad_norm": 0.1734144240617752, "learning_rate": 7.125529634329988e-05, "loss": 0.0126, "step": 15690 }, { "grad_norm": 0.21928580105304718, "learning_rate": 7.12178733927561e-05, "loss": 0.0112, "step": 15700 }, { "grad_norm": 0.22273288667201996, "learning_rate": 7.118043593999593e-05, "loss": 0.0132, "step": 15710 }, { "grad_norm": 0.2186920940876007, "learning_rate": 7.114298401060752e-05, "loss": 0.0128, "step": 15720 }, { "grad_norm": 0.2509160041809082, "learning_rate": 7.11055176301889e-05, "loss": 0.0147, "step": 15730 }, { "grad_norm": 0.22144587337970734, "learning_rate": 7.1068036824348e-05, "loss": 0.0134, "step": 15740 }, { "grad_norm": 0.20081378519535065, "learning_rate": 7.10305416187026e-05, "loss": 0.0146, "step": 15750 }, { "grad_norm": 0.2671731412410736, "learning_rate": 7.099303203888029e-05, "loss": 0.0153, "step": 15760 }, { "grad_norm": 0.2024042308330536, "learning_rate": 7.095550811051855e-05, "loss": 0.0142, "step": 15770 }, { "grad_norm": 0.2093038409948349, "learning_rate": 7.09179698592646e-05, "loss": 0.0138, "step": 15780 }, { "grad_norm": 0.21970857679843903, "learning_rate": 7.088041731077551e-05, "loss": 0.0132, "step": 15790 }, { "grad_norm": 0.23714672029018402, "learning_rate": 7.084285049071806e-05, "loss": 0.0126, "step": 15800 }, { "grad_norm": 0.21697290241718292, "learning_rate": 7.080526942476886e-05, "loss": 0.0118, "step": 15810 }, { "grad_norm": 0.19337470829486847, "learning_rate": 7.076767413861418e-05, "loss": 0.0159, "step": 15820 }, { "grad_norm": 0.23979127407073975, "learning_rate": 7.073006465795005e-05, "loss": 0.0116, "step": 15830 }, { "grad_norm": 0.18791146576404572, "learning_rate": 7.06924410084822e-05, "loss": 0.0159, "step": 15840 }, { "grad_norm": 0.14511416852474213, "learning_rate": 7.065480321592604e-05, "loss": 0.0139, "step": 15850 }, { "grad_norm": 0.2203586995601654, "learning_rate": 7.061715130600663e-05, "loss": 0.014, "step": 15860 }, { "grad_norm": 0.18041642010211945, "learning_rate": 7.057948530445873e-05, "loss": 0.012, "step": 15870 }, { "grad_norm": 0.16929127275943756, "learning_rate": 7.054180523702668e-05, "loss": 0.0153, "step": 15880 }, { "grad_norm": 0.1761600226163864, "learning_rate": 7.050411112946442e-05, "loss": 0.0143, "step": 15890 }, { "grad_norm": 0.30532336235046387, "learning_rate": 7.046640300753557e-05, "loss": 0.0163, "step": 15900 }, { "grad_norm": 0.2156222015619278, "learning_rate": 7.042868089701325e-05, "loss": 0.0151, "step": 15910 }, { "grad_norm": 0.2052801549434662, "learning_rate": 7.039094482368016e-05, "loss": 0.0174, "step": 15920 }, { "grad_norm": 0.2568667531013489, "learning_rate": 7.035319481332858e-05, "loss": 0.0134, "step": 15930 }, { "grad_norm": 0.17159533500671387, "learning_rate": 7.031543089176023e-05, "loss": 0.0126, "step": 15940 }, { "grad_norm": 0.1726926565170288, "learning_rate": 7.027765308478644e-05, "loss": 0.0155, "step": 15950 }, { "grad_norm": 0.1917310208082199, "learning_rate": 7.023986141822798e-05, "loss": 0.0124, "step": 15960 }, { "grad_norm": 0.17304325103759766, "learning_rate": 7.02020559179151e-05, "loss": 0.0123, "step": 15970 }, { "grad_norm": 0.21636155247688293, "learning_rate": 7.016423660968748e-05, "loss": 0.012, "step": 15980 }, { "grad_norm": 0.17498339712619781, "learning_rate": 7.012640351939428e-05, "loss": 0.0128, "step": 15990 }, { "grad_norm": 0.15309765934944153, "learning_rate": 7.008855667289404e-05, "loss": 0.0165, "step": 16000 }, { "grad_norm": 0.23016251623630524, "learning_rate": 7.005069609605476e-05, "loss": 0.0131, "step": 16010 }, { "grad_norm": 0.17324469983577728, "learning_rate": 7.001282181475377e-05, "loss": 0.0154, "step": 16020 }, { "grad_norm": 0.14664284884929657, "learning_rate": 6.997493385487775e-05, "loss": 0.0131, "step": 16030 }, { "grad_norm": 0.16713006794452667, "learning_rate": 6.99370322423228e-05, "loss": 0.012, "step": 16040 }, { "grad_norm": 0.2046881467103958, "learning_rate": 6.989911700299433e-05, "loss": 0.0134, "step": 16050 }, { "grad_norm": 0.2616203725337982, "learning_rate": 6.9861188162807e-05, "loss": 0.0143, "step": 16060 }, { "grad_norm": 0.19361984729766846, "learning_rate": 6.982324574768487e-05, "loss": 0.0115, "step": 16070 }, { "grad_norm": 0.19225288927555084, "learning_rate": 6.978528978356117e-05, "loss": 0.0144, "step": 16080 }, { "grad_norm": 0.21811658143997192, "learning_rate": 6.974732029637846e-05, "loss": 0.0122, "step": 16090 }, { "grad_norm": 0.17327697575092316, "learning_rate": 6.970933731208855e-05, "loss": 0.0121, "step": 16100 }, { "grad_norm": 0.22081270813941956, "learning_rate": 6.967134085665244e-05, "loss": 0.0123, "step": 16110 }, { "grad_norm": 0.2023918330669403, "learning_rate": 6.963333095604034e-05, "loss": 0.0125, "step": 16120 }, { "grad_norm": 0.2391110509634018, "learning_rate": 6.959530763623166e-05, "loss": 0.013, "step": 16130 }, { "grad_norm": 0.1563355028629303, "learning_rate": 6.955727092321497e-05, "loss": 0.0128, "step": 16140 }, { "grad_norm": 0.23777998983860016, "learning_rate": 6.951922084298803e-05, "loss": 0.0129, "step": 16150 }, { "grad_norm": 0.21209730207920074, "learning_rate": 6.948115742155769e-05, "loss": 0.0139, "step": 16160 }, { "grad_norm": 0.17944036424160004, "learning_rate": 6.944308068493996e-05, "loss": 0.0119, "step": 16170 }, { "grad_norm": 0.1788674145936966, "learning_rate": 6.940499065915992e-05, "loss": 0.0131, "step": 16180 }, { "grad_norm": 0.13610397279262543, "learning_rate": 6.936688737025173e-05, "loss": 0.0127, "step": 16190 }, { "grad_norm": 0.22552938759326935, "learning_rate": 6.932877084425867e-05, "loss": 0.0125, "step": 16200 }, { "grad_norm": 0.22872839868068695, "learning_rate": 6.929064110723297e-05, "loss": 0.0125, "step": 16210 }, { "grad_norm": 0.1548631489276886, "learning_rate": 6.925249818523598e-05, "loss": 0.014, "step": 16220 }, { "grad_norm": 0.15049795806407928, "learning_rate": 6.921434210433801e-05, "loss": 0.0126, "step": 16230 }, { "grad_norm": 0.25977087020874023, "learning_rate": 6.917617289061841e-05, "loss": 0.0146, "step": 16240 }, { "grad_norm": 0.20121604204177856, "learning_rate": 6.913799057016547e-05, "loss": 0.0101, "step": 16250 }, { "grad_norm": 0.19498664140701294, "learning_rate": 6.909979516907641e-05, "loss": 0.0149, "step": 16260 }, { "grad_norm": 0.16682720184326172, "learning_rate": 6.906158671345746e-05, "loss": 0.0126, "step": 16270 }, { "grad_norm": 0.25250115990638733, "learning_rate": 6.902336522942374e-05, "loss": 0.0154, "step": 16280 }, { "grad_norm": 0.21951323747634888, "learning_rate": 6.898513074309924e-05, "loss": 0.0118, "step": 16290 }, { "grad_norm": 0.2150856852531433, "learning_rate": 6.894688328061693e-05, "loss": 0.0129, "step": 16300 }, { "grad_norm": 0.19527357816696167, "learning_rate": 6.890862286811853e-05, "loss": 0.0136, "step": 16310 }, { "grad_norm": 0.21654702723026276, "learning_rate": 6.88703495317547e-05, "loss": 0.0128, "step": 16320 }, { "grad_norm": 0.1715468466281891, "learning_rate": 6.883206329768492e-05, "loss": 0.0112, "step": 16330 }, { "grad_norm": 0.1913924366235733, "learning_rate": 6.879376419207743e-05, "loss": 0.0108, "step": 16340 }, { "grad_norm": 0.18757453560829163, "learning_rate": 6.875545224110935e-05, "loss": 0.0142, "step": 16350 }, { "grad_norm": 0.15702694654464722, "learning_rate": 6.871712747096651e-05, "loss": 0.0149, "step": 16360 }, { "grad_norm": 0.2035713642835617, "learning_rate": 6.867878990784353e-05, "loss": 0.0101, "step": 16370 }, { "grad_norm": 0.16480673849582672, "learning_rate": 6.864043957794377e-05, "loss": 0.0099, "step": 16380 }, { "grad_norm": 0.2709570825099945, "learning_rate": 6.860207650747934e-05, "loss": 0.0125, "step": 16390 }, { "grad_norm": 0.16192464530467987, "learning_rate": 6.856370072267104e-05, "loss": 0.0151, "step": 16400 }, { "grad_norm": 0.21209171414375305, "learning_rate": 6.852531224974831e-05, "loss": 0.0145, "step": 16410 }, { "grad_norm": 0.18680693209171295, "learning_rate": 6.848691111494936e-05, "loss": 0.0146, "step": 16420 }, { "grad_norm": 0.210549995303154, "learning_rate": 6.844849734452097e-05, "loss": 0.0112, "step": 16430 }, { "grad_norm": 0.16632266342639923, "learning_rate": 6.841007096471862e-05, "loss": 0.0109, "step": 16440 }, { "grad_norm": 0.15905670821666718, "learning_rate": 6.837163200180636e-05, "loss": 0.0123, "step": 16450 }, { "grad_norm": 0.21504250168800354, "learning_rate": 6.833318048205684e-05, "loss": 0.0143, "step": 16460 }, { "grad_norm": 0.14715366065502167, "learning_rate": 6.829471643175136e-05, "loss": 0.0114, "step": 16470 }, { "grad_norm": 0.18304364383220673, "learning_rate": 6.825623987717969e-05, "loss": 0.0128, "step": 16480 }, { "grad_norm": 0.13682237267494202, "learning_rate": 6.821775084464022e-05, "loss": 0.011, "step": 16490 }, { "grad_norm": 0.12541836500167847, "learning_rate": 6.817924936043982e-05, "loss": 0.0117, "step": 16500 }, { "grad_norm": 0.1786738634109497, "learning_rate": 6.81407354508939e-05, "loss": 0.0098, "step": 16510 }, { "grad_norm": 0.22356818616390228, "learning_rate": 6.810220914232636e-05, "loss": 0.0116, "step": 16520 }, { "grad_norm": 0.15445014834403992, "learning_rate": 6.806367046106959e-05, "loss": 0.0128, "step": 16530 }, { "grad_norm": 0.17288091778755188, "learning_rate": 6.802511943346435e-05, "loss": 0.0112, "step": 16540 }, { "grad_norm": 0.1711214780807495, "learning_rate": 6.798655608585997e-05, "loss": 0.0099, "step": 16550 }, { "grad_norm": 0.20919623970985413, "learning_rate": 6.79479804446141e-05, "loss": 0.0127, "step": 16560 }, { "grad_norm": 0.17527872323989868, "learning_rate": 6.790939253609284e-05, "loss": 0.0135, "step": 16570 }, { "grad_norm": 0.17299884557724, "learning_rate": 6.787079238667065e-05, "loss": 0.0107, "step": 16580 }, { "grad_norm": 0.1479242742061615, "learning_rate": 6.783218002273039e-05, "loss": 0.0114, "step": 16590 }, { "grad_norm": 0.17470099031925201, "learning_rate": 6.779355547066322e-05, "loss": 0.0116, "step": 16600 }, { "grad_norm": 0.1438913643360138, "learning_rate": 6.775491875686865e-05, "loss": 0.0104, "step": 16610 }, { "grad_norm": 0.25050368905067444, "learning_rate": 6.771626990775457e-05, "loss": 0.0125, "step": 16620 }, { "grad_norm": 0.19620919227600098, "learning_rate": 6.767760894973704e-05, "loss": 0.0143, "step": 16630 }, { "grad_norm": 0.18764100968837738, "learning_rate": 6.763893590924048e-05, "loss": 0.0121, "step": 16640 }, { "grad_norm": 0.1991969496011734, "learning_rate": 6.760025081269756e-05, "loss": 0.017, "step": 16650 }, { "grad_norm": 0.2328876256942749, "learning_rate": 6.756155368654915e-05, "loss": 0.0113, "step": 16660 }, { "grad_norm": 0.21162749826908112, "learning_rate": 6.752284455724442e-05, "loss": 0.015, "step": 16670 }, { "grad_norm": 0.16933129727840424, "learning_rate": 6.748412345124065e-05, "loss": 0.0118, "step": 16680 }, { "grad_norm": 0.16464757919311523, "learning_rate": 6.744539039500335e-05, "loss": 0.0109, "step": 16690 }, { "grad_norm": 0.15710483491420746, "learning_rate": 6.740664541500625e-05, "loss": 0.0118, "step": 16700 }, { "grad_norm": 0.17266719043254852, "learning_rate": 6.736788853773112e-05, "loss": 0.0126, "step": 16710 }, { "grad_norm": 0.17738789319992065, "learning_rate": 6.732911978966796e-05, "loss": 0.0115, "step": 16720 }, { "grad_norm": 0.2968637943267822, "learning_rate": 6.729033919731482e-05, "loss": 0.0124, "step": 16730 }, { "grad_norm": 0.2564302086830139, "learning_rate": 6.725154678717787e-05, "loss": 0.0171, "step": 16740 }, { "grad_norm": 0.24289926886558533, "learning_rate": 6.721274258577138e-05, "loss": 0.014, "step": 16750 }, { "grad_norm": 0.22708813846111298, "learning_rate": 6.717392661961763e-05, "loss": 0.0137, "step": 16760 }, { "grad_norm": 0.2037508189678192, "learning_rate": 6.713509891524697e-05, "loss": 0.0148, "step": 16770 }, { "grad_norm": 0.20059825479984283, "learning_rate": 6.709625949919777e-05, "loss": 0.0133, "step": 16780 }, { "grad_norm": 0.20171183347702026, "learning_rate": 6.705740839801642e-05, "loss": 0.0111, "step": 16790 }, { "grad_norm": 0.18458503484725952, "learning_rate": 6.701854563825727e-05, "loss": 0.0127, "step": 16800 }, { "grad_norm": 0.17456521093845367, "learning_rate": 6.697967124648266e-05, "loss": 0.0095, "step": 16810 }, { "grad_norm": 0.17271097004413605, "learning_rate": 6.694078524926285e-05, "loss": 0.0116, "step": 16820 }, { "grad_norm": 0.21357113122940063, "learning_rate": 6.690188767317607e-05, "loss": 0.0128, "step": 16830 }, { "grad_norm": 0.18304695188999176, "learning_rate": 6.686297854480843e-05, "loss": 0.0119, "step": 16840 }, { "grad_norm": 0.193607896566391, "learning_rate": 6.682405789075398e-05, "loss": 0.0142, "step": 16850 }, { "grad_norm": 0.17595820128917694, "learning_rate": 6.67851257376146e-05, "loss": 0.0115, "step": 16860 }, { "grad_norm": 0.17803433537483215, "learning_rate": 6.674618211200004e-05, "loss": 0.0118, "step": 16870 }, { "grad_norm": 0.1642601042985916, "learning_rate": 6.670722704052792e-05, "loss": 0.0131, "step": 16880 }, { "grad_norm": 0.1798069328069687, "learning_rate": 6.666826054982365e-05, "loss": 0.0114, "step": 16890 }, { "grad_norm": 0.17660661041736603, "learning_rate": 6.662928266652048e-05, "loss": 0.0131, "step": 16900 }, { "grad_norm": 0.23995427787303925, "learning_rate": 6.659029341725941e-05, "loss": 0.0126, "step": 16910 }, { "grad_norm": 0.25247427821159363, "learning_rate": 6.655129282868923e-05, "loss": 0.0164, "step": 16920 }, { "grad_norm": 0.17478935420513153, "learning_rate": 6.651228092746646e-05, "loss": 0.0111, "step": 16930 }, { "grad_norm": 0.17510239779949188, "learning_rate": 6.647325774025539e-05, "loss": 0.0119, "step": 16940 }, { "grad_norm": 0.20688548684120178, "learning_rate": 6.643422329372798e-05, "loss": 0.0139, "step": 16950 }, { "grad_norm": 0.18811146914958954, "learning_rate": 6.639517761456392e-05, "loss": 0.0152, "step": 16960 }, { "grad_norm": 0.21283963322639465, "learning_rate": 6.635612072945054e-05, "loss": 0.0135, "step": 16970 }, { "grad_norm": 0.20902179181575775, "learning_rate": 6.631705266508289e-05, "loss": 0.0126, "step": 16980 }, { "grad_norm": 0.18937839567661285, "learning_rate": 6.62779734481636e-05, "loss": 0.0112, "step": 16990 }, { "grad_norm": 0.18740883469581604, "learning_rate": 6.623888310540294e-05, "loss": 0.0139, "step": 17000 }, { "grad_norm": 0.25331729650497437, "learning_rate": 6.619978166351882e-05, "loss": 0.0161, "step": 17010 }, { "grad_norm": 0.16693159937858582, "learning_rate": 6.616066914923666e-05, "loss": 0.0146, "step": 17020 }, { "grad_norm": 0.21039333939552307, "learning_rate": 6.612154558928955e-05, "loss": 0.0115, "step": 17030 }, { "grad_norm": 0.21447831392288208, "learning_rate": 6.608241101041804e-05, "loss": 0.0137, "step": 17040 }, { "grad_norm": 0.25312530994415283, "learning_rate": 6.604326543937025e-05, "loss": 0.0142, "step": 17050 }, { "grad_norm": 0.20596736669540405, "learning_rate": 6.60041089029018e-05, "loss": 0.0137, "step": 17060 }, { "grad_norm": 0.21248844265937805, "learning_rate": 6.596494142777583e-05, "loss": 0.0133, "step": 17070 }, { "grad_norm": 0.16340607404708862, "learning_rate": 6.592576304076294e-05, "loss": 0.0099, "step": 17080 }, { "grad_norm": 0.21101166307926178, "learning_rate": 6.588657376864119e-05, "loss": 0.014, "step": 17090 }, { "grad_norm": 0.16713811457157135, "learning_rate": 6.584737363819605e-05, "loss": 0.0107, "step": 17100 }, { "grad_norm": 0.25736886262893677, "learning_rate": 6.580816267622048e-05, "loss": 0.0108, "step": 17110 }, { "grad_norm": 0.19597190618515015, "learning_rate": 6.576894090951478e-05, "loss": 0.0123, "step": 17120 }, { "grad_norm": 0.17261570692062378, "learning_rate": 6.572970836488665e-05, "loss": 0.0116, "step": 17130 }, { "grad_norm": 0.2134222686290741, "learning_rate": 6.569046506915119e-05, "loss": 0.0128, "step": 17140 }, { "grad_norm": 0.15698380768299103, "learning_rate": 6.56512110491308e-05, "loss": 0.0121, "step": 17150 }, { "grad_norm": 0.17754176259040833, "learning_rate": 6.561194633165523e-05, "loss": 0.0124, "step": 17160 }, { "grad_norm": 0.1699327975511551, "learning_rate": 6.557267094356155e-05, "loss": 0.0107, "step": 17170 }, { "grad_norm": 0.1676630675792694, "learning_rate": 6.553338491169414e-05, "loss": 0.0105, "step": 17180 }, { "grad_norm": 0.20552654564380646, "learning_rate": 6.54940882629046e-05, "loss": 0.0143, "step": 17190 }, { "grad_norm": 0.2588731348514557, "learning_rate": 6.545478102405184e-05, "loss": 0.0105, "step": 17200 }, { "grad_norm": 0.15482494235038757, "learning_rate": 6.541546322200199e-05, "loss": 0.0094, "step": 17210 }, { "grad_norm": 0.13819435238838196, "learning_rate": 6.537613488362837e-05, "loss": 0.0112, "step": 17220 }, { "grad_norm": 0.20664070546627045, "learning_rate": 6.533679603581155e-05, "loss": 0.0149, "step": 17230 }, { "grad_norm": 0.2045435756444931, "learning_rate": 6.529744670543926e-05, "loss": 0.0136, "step": 17240 }, { "grad_norm": 0.22695471346378326, "learning_rate": 6.52580869194064e-05, "loss": 0.0107, "step": 17250 }, { "grad_norm": 0.13303326070308685, "learning_rate": 6.521871670461499e-05, "loss": 0.0134, "step": 17260 }, { "grad_norm": 0.1895119547843933, "learning_rate": 6.517933608797422e-05, "loss": 0.0113, "step": 17270 }, { "grad_norm": 0.1422095149755478, "learning_rate": 6.513994509640038e-05, "loss": 0.0116, "step": 17280 }, { "grad_norm": 0.2485627382993698, "learning_rate": 6.510054375681682e-05, "loss": 0.0103, "step": 17290 }, { "grad_norm": 0.22645290195941925, "learning_rate": 6.506113209615398e-05, "loss": 0.0119, "step": 17300 }, { "grad_norm": 0.17901358008384705, "learning_rate": 6.502171014134938e-05, "loss": 0.0114, "step": 17310 }, { "grad_norm": 0.14344483613967896, "learning_rate": 6.498227791934755e-05, "loss": 0.0118, "step": 17320 }, { "grad_norm": 0.1904594600200653, "learning_rate": 6.494283545710003e-05, "loss": 0.0149, "step": 17330 }, { "grad_norm": 0.14739374816417694, "learning_rate": 6.490338278156538e-05, "loss": 0.0112, "step": 17340 }, { "grad_norm": 0.18312238156795502, "learning_rate": 6.486391991970913e-05, "loss": 0.0136, "step": 17350 }, { "grad_norm": 0.17270250618457794, "learning_rate": 6.482444689850377e-05, "loss": 0.0128, "step": 17360 }, { "grad_norm": 0.20093265175819397, "learning_rate": 6.478496374492875e-05, "loss": 0.0115, "step": 17370 }, { "grad_norm": 0.1512458622455597, "learning_rate": 6.474547048597042e-05, "loss": 0.0119, "step": 17380 }, { "grad_norm": 0.1535632461309433, "learning_rate": 6.470596714862205e-05, "loss": 0.0118, "step": 17390 }, { "grad_norm": 0.18372896313667297, "learning_rate": 6.46664537598838e-05, "loss": 0.0136, "step": 17400 }, { "grad_norm": 0.15520493686199188, "learning_rate": 6.462693034676271e-05, "loss": 0.0128, "step": 17410 }, { "grad_norm": 0.12560561299324036, "learning_rate": 6.458739693627265e-05, "loss": 0.011, "step": 17420 }, { "grad_norm": 0.17283397912979126, "learning_rate": 6.454785355543432e-05, "loss": 0.0115, "step": 17430 }, { "grad_norm": 0.21149086952209473, "learning_rate": 6.450830023127528e-05, "loss": 0.0117, "step": 17440 }, { "grad_norm": 0.24468104541301727, "learning_rate": 6.446873699082982e-05, "loss": 0.0139, "step": 17450 }, { "grad_norm": 0.1289641261100769, "learning_rate": 6.44291638611391e-05, "loss": 0.01, "step": 17460 }, { "grad_norm": 0.21438941359519958, "learning_rate": 6.43895808692509e-05, "loss": 0.0123, "step": 17470 }, { "grad_norm": 0.20527398586273193, "learning_rate": 6.434998804221986e-05, "loss": 0.0123, "step": 17480 }, { "grad_norm": 0.185546413064003, "learning_rate": 6.431038540710732e-05, "loss": 0.0119, "step": 17490 }, { "grad_norm": 0.1968475729227066, "learning_rate": 6.427077299098129e-05, "loss": 0.0121, "step": 17500 }, { "grad_norm": 0.2155316174030304, "learning_rate": 6.423115082091651e-05, "loss": 0.0123, "step": 17510 }, { "grad_norm": 0.19458886981010437, "learning_rate": 6.419151892399429e-05, "loss": 0.0093, "step": 17520 }, { "grad_norm": 0.16246825456619263, "learning_rate": 6.415187732730273e-05, "loss": 0.0119, "step": 17530 }, { "grad_norm": 0.20532077550888062, "learning_rate": 6.411222605793645e-05, "loss": 0.0113, "step": 17540 }, { "grad_norm": 0.16466063261032104, "learning_rate": 6.407256514299674e-05, "loss": 0.0102, "step": 17550 }, { "grad_norm": 0.16486495733261108, "learning_rate": 6.403289460959147e-05, "loss": 0.012, "step": 17560 }, { "grad_norm": 0.17727941274642944, "learning_rate": 6.399321448483501e-05, "loss": 0.0126, "step": 17570 }, { "grad_norm": 0.15085871517658234, "learning_rate": 6.395352479584844e-05, "loss": 0.0125, "step": 17580 }, { "grad_norm": 0.1294821947813034, "learning_rate": 6.391382556975923e-05, "loss": 0.012, "step": 17590 }, { "grad_norm": 0.23016437888145447, "learning_rate": 6.387411683370144e-05, "loss": 0.0105, "step": 17600 }, { "grad_norm": 0.19415654242038727, "learning_rate": 6.383439861481562e-05, "loss": 0.0116, "step": 17610 }, { "grad_norm": 0.17330412566661835, "learning_rate": 6.379467094024879e-05, "loss": 0.0108, "step": 17620 }, { "grad_norm": 0.18749502301216125, "learning_rate": 6.375493383715445e-05, "loss": 0.0132, "step": 17630 }, { "grad_norm": 0.23724499344825745, "learning_rate": 6.371518733269254e-05, "loss": 0.0122, "step": 17640 }, { "grad_norm": 0.20826347172260284, "learning_rate": 6.367543145402942e-05, "loss": 0.0097, "step": 17650 }, { "grad_norm": 0.16613823175430298, "learning_rate": 6.363566622833785e-05, "loss": 0.0098, "step": 17660 }, { "grad_norm": 0.1581864058971405, "learning_rate": 6.359589168279698e-05, "loss": 0.0111, "step": 17670 }, { "grad_norm": 0.1459369957447052, "learning_rate": 6.355610784459235e-05, "loss": 0.0149, "step": 17680 }, { "grad_norm": 0.207834392786026, "learning_rate": 6.351631474091585e-05, "loss": 0.0109, "step": 17690 }, { "grad_norm": 0.23361380398273468, "learning_rate": 6.347651239896566e-05, "loss": 0.0118, "step": 17700 }, { "grad_norm": 0.20223908126354218, "learning_rate": 6.343670084594633e-05, "loss": 0.0108, "step": 17710 }, { "grad_norm": 0.19518092274665833, "learning_rate": 6.339688010906866e-05, "loss": 0.0129, "step": 17720 }, { "grad_norm": 0.16672860085964203, "learning_rate": 6.335705021554975e-05, "loss": 0.0117, "step": 17730 }, { "grad_norm": 0.14828521013259888, "learning_rate": 6.3317211192613e-05, "loss": 0.0147, "step": 17740 }, { "grad_norm": 0.2540571391582489, "learning_rate": 6.327736306748795e-05, "loss": 0.0101, "step": 17750 }, { "grad_norm": 0.2677644193172455, "learning_rate": 6.323750586741047e-05, "loss": 0.0164, "step": 17760 }, { "grad_norm": 0.19795310497283936, "learning_rate": 6.319763961962252e-05, "loss": 0.0119, "step": 17770 }, { "grad_norm": 0.12763312458992004, "learning_rate": 6.315776435137233e-05, "loss": 0.012, "step": 17780 }, { "grad_norm": 0.16893605887889862, "learning_rate": 6.311788008991432e-05, "loss": 0.013, "step": 17790 }, { "grad_norm": 0.19781309366226196, "learning_rate": 6.307798686250891e-05, "loss": 0.0105, "step": 17800 }, { "grad_norm": 0.1577647626399994, "learning_rate": 6.303808469642284e-05, "loss": 0.015, "step": 17810 }, { "grad_norm": 0.21063017845153809, "learning_rate": 6.29981736189288e-05, "loss": 0.0121, "step": 17820 }, { "grad_norm": 0.1704251915216446, "learning_rate": 6.295825365730567e-05, "loss": 0.0099, "step": 17830 }, { "grad_norm": 0.14565099775791168, "learning_rate": 6.291832483883835e-05, "loss": 0.0108, "step": 17840 }, { "grad_norm": 0.21278028190135956, "learning_rate": 6.28783871908178e-05, "loss": 0.0106, "step": 17850 }, { "grad_norm": 0.2122192531824112, "learning_rate": 6.283844074054107e-05, "loss": 0.0122, "step": 17860 }, { "grad_norm": 0.22311478853225708, "learning_rate": 6.279848551531112e-05, "loss": 0.0143, "step": 17870 }, { "grad_norm": 0.294739305973053, "learning_rate": 6.275852154243702e-05, "loss": 0.0115, "step": 17880 }, { "grad_norm": 0.3008247911930084, "learning_rate": 6.271854884923377e-05, "loss": 0.0145, "step": 17890 }, { "grad_norm": 0.2129630446434021, "learning_rate": 6.267856746302228e-05, "loss": 0.0121, "step": 17900 }, { "grad_norm": 0.16989493370056152, "learning_rate": 6.263857741112948e-05, "loss": 0.0133, "step": 17910 }, { "grad_norm": 0.17656710743904114, "learning_rate": 6.259857872088821e-05, "loss": 0.0141, "step": 17920 }, { "grad_norm": 0.17534944415092468, "learning_rate": 6.255857141963719e-05, "loss": 0.0116, "step": 17930 }, { "grad_norm": 0.154433012008667, "learning_rate": 6.251855553472101e-05, "loss": 0.0127, "step": 17940 }, { "grad_norm": 0.19803553819656372, "learning_rate": 6.247853109349016e-05, "loss": 0.0101, "step": 17950 }, { "grad_norm": 0.1830909252166748, "learning_rate": 6.243849812330098e-05, "loss": 0.0162, "step": 17960 }, { "grad_norm": 0.24604752659797668, "learning_rate": 6.239845665151563e-05, "loss": 0.0128, "step": 17970 }, { "grad_norm": 0.1954261064529419, "learning_rate": 6.235840670550204e-05, "loss": 0.0108, "step": 17980 }, { "grad_norm": 0.1593979448080063, "learning_rate": 6.231834831263403e-05, "loss": 0.0132, "step": 17990 }, { "grad_norm": 0.1711994707584381, "learning_rate": 6.22782815002911e-05, "loss": 0.0113, "step": 18000 }, { "grad_norm": 0.20040611922740936, "learning_rate": 6.223820629585852e-05, "loss": 0.0129, "step": 18010 }, { "grad_norm": 0.25090646743774414, "learning_rate": 6.219812272672737e-05, "loss": 0.0118, "step": 18020 }, { "grad_norm": 0.1442956030368805, "learning_rate": 6.215803082029434e-05, "loss": 0.0108, "step": 18030 }, { "grad_norm": 0.19707712531089783, "learning_rate": 6.211793060396188e-05, "loss": 0.0115, "step": 18040 }, { "grad_norm": 0.12304724752902985, "learning_rate": 6.207782210513811e-05, "loss": 0.013, "step": 18050 }, { "grad_norm": 0.1893119066953659, "learning_rate": 6.203770535123683e-05, "loss": 0.0142, "step": 18060 }, { "grad_norm": 0.19330863654613495, "learning_rate": 6.199758036967747e-05, "loss": 0.0143, "step": 18070 }, { "grad_norm": 0.20136208832263947, "learning_rate": 6.195744718788503e-05, "loss": 0.0105, "step": 18080 }, { "grad_norm": 0.2159934639930725, "learning_rate": 6.191730583329021e-05, "loss": 0.0131, "step": 18090 }, { "grad_norm": 0.21069687604904175, "learning_rate": 6.187715633332921e-05, "loss": 0.0131, "step": 18100 }, { "grad_norm": 0.18376757204532623, "learning_rate": 6.183699871544386e-05, "loss": 0.0109, "step": 18110 }, { "grad_norm": 0.22966180741786957, "learning_rate": 6.179683300708152e-05, "loss": 0.0162, "step": 18120 }, { "grad_norm": 0.22798649966716766, "learning_rate": 6.175665923569503e-05, "loss": 0.0127, "step": 18130 }, { "grad_norm": 0.17848022282123566, "learning_rate": 6.171647742874281e-05, "loss": 0.0102, "step": 18140 }, { "grad_norm": 0.20084713399410248, "learning_rate": 6.167628761368875e-05, "loss": 0.0094, "step": 18150 }, { "grad_norm": 0.21924132108688354, "learning_rate": 6.163608981800222e-05, "loss": 0.0113, "step": 18160 }, { "grad_norm": 0.1638786941766739, "learning_rate": 6.159588406915803e-05, "loss": 0.0112, "step": 18170 }, { "grad_norm": 0.1594478338956833, "learning_rate": 6.155567039463639e-05, "loss": 0.0117, "step": 18180 }, { "grad_norm": 0.10721606016159058, "learning_rate": 6.151544882192302e-05, "loss": 0.0121, "step": 18190 }, { "grad_norm": 0.253061443567276, "learning_rate": 6.147521937850895e-05, "loss": 0.0129, "step": 18200 }, { "grad_norm": 0.16330011188983917, "learning_rate": 6.143498209189066e-05, "loss": 0.0103, "step": 18210 }, { "grad_norm": 0.16380460560321808, "learning_rate": 6.139473698956993e-05, "loss": 0.0107, "step": 18220 }, { "grad_norm": 0.225546196103096, "learning_rate": 6.13544840990539e-05, "loss": 0.0139, "step": 18230 }, { "grad_norm": 0.17959408462047577, "learning_rate": 6.131422344785507e-05, "loss": 0.0147, "step": 18240 }, { "grad_norm": 0.1390865445137024, "learning_rate": 6.127395506349119e-05, "loss": 0.0128, "step": 18250 }, { "grad_norm": 0.13866685330867767, "learning_rate": 6.123367897348533e-05, "loss": 0.0126, "step": 18260 }, { "grad_norm": 0.16285304725170135, "learning_rate": 6.119339520536584e-05, "loss": 0.0137, "step": 18270 }, { "grad_norm": 0.2929856479167938, "learning_rate": 6.115310378666625e-05, "loss": 0.0118, "step": 18280 }, { "grad_norm": 0.2634677290916443, "learning_rate": 6.11128047449254e-05, "loss": 0.0105, "step": 18290 }, { "grad_norm": 0.17749059200286865, "learning_rate": 6.107249810768729e-05, "loss": 0.0109, "step": 18300 }, { "grad_norm": 0.23247526586055756, "learning_rate": 6.1032183902501125e-05, "loss": 0.0159, "step": 18310 }, { "grad_norm": 0.1655990332365036, "learning_rate": 6.099186215692131e-05, "loss": 0.0101, "step": 18320 }, { "grad_norm": 0.23088665306568146, "learning_rate": 6.095153289850734e-05, "loss": 0.0155, "step": 18330 }, { "grad_norm": 0.2412264496088028, "learning_rate": 6.0911196154823904e-05, "loss": 0.0123, "step": 18340 }, { "grad_norm": 0.19608210027217865, "learning_rate": 6.087085195344079e-05, "loss": 0.0119, "step": 18350 }, { "grad_norm": 0.20541472733020782, "learning_rate": 6.083050032193286e-05, "loss": 0.0107, "step": 18360 }, { "grad_norm": 0.12654614448547363, "learning_rate": 6.0790141287880097e-05, "loss": 0.0099, "step": 18370 }, { "grad_norm": 0.20298174023628235, "learning_rate": 6.0749774878867496e-05, "loss": 0.0123, "step": 18380 }, { "grad_norm": 0.12288478761911392, "learning_rate": 6.0709401122485146e-05, "loss": 0.0088, "step": 18390 }, { "grad_norm": 0.18290309607982635, "learning_rate": 6.066902004632811e-05, "loss": 0.0105, "step": 18400 }, { "grad_norm": 0.2429719716310501, "learning_rate": 6.062863167799646e-05, "loss": 0.0108, "step": 18410 }, { "grad_norm": 0.13302136957645416, "learning_rate": 6.058823604509529e-05, "loss": 0.0111, "step": 18420 }, { "grad_norm": 0.16686822474002838, "learning_rate": 6.054783317523462e-05, "loss": 0.0134, "step": 18430 }, { "grad_norm": 0.22065620124340057, "learning_rate": 6.050742309602944e-05, "loss": 0.0127, "step": 18440 }, { "grad_norm": 0.19283829629421234, "learning_rate": 6.046700583509965e-05, "loss": 0.0129, "step": 18450 }, { "grad_norm": 0.23267686367034912, "learning_rate": 6.042658142007007e-05, "loss": 0.0095, "step": 18460 }, { "grad_norm": 0.17693354189395905, "learning_rate": 6.038614987857041e-05, "loss": 0.0114, "step": 18470 }, { "grad_norm": 0.1644120216369629, "learning_rate": 6.0345711238235224e-05, "loss": 0.0103, "step": 18480 }, { "grad_norm": 0.17835617065429688, "learning_rate": 6.030526552670399e-05, "loss": 0.014, "step": 18490 }, { "grad_norm": 0.17295216023921967, "learning_rate": 6.0264812771620925e-05, "loss": 0.0118, "step": 18500 }, { "grad_norm": 0.17012985050678253, "learning_rate": 6.022435300063512e-05, "loss": 0.0113, "step": 18510 }, { "grad_norm": 0.22091884911060333, "learning_rate": 6.0183886241400466e-05, "loss": 0.0116, "step": 18520 }, { "grad_norm": 0.18115690350532532, "learning_rate": 6.0143412521575584e-05, "loss": 0.0103, "step": 18530 }, { "grad_norm": 0.17684204876422882, "learning_rate": 6.010293186882389e-05, "loss": 0.0123, "step": 18540 }, { "grad_norm": 0.20416569709777832, "learning_rate": 6.0062444310813525e-05, "loss": 0.0129, "step": 18550 }, { "grad_norm": 0.14916561543941498, "learning_rate": 6.0021949875217355e-05, "loss": 0.0105, "step": 18560 }, { "grad_norm": 0.15632431209087372, "learning_rate": 5.998144858971295e-05, "loss": 0.0098, "step": 18570 }, { "grad_norm": 0.14448459446430206, "learning_rate": 5.994094048198257e-05, "loss": 0.0094, "step": 18580 }, { "grad_norm": 0.2088163048028946, "learning_rate": 5.990042557971307e-05, "loss": 0.0121, "step": 18590 }, { "grad_norm": 0.19549661874771118, "learning_rate": 5.985990391059607e-05, "loss": 0.01, "step": 18600 }, { "grad_norm": 0.1636260598897934, "learning_rate": 5.981937550232771e-05, "loss": 0.0128, "step": 18610 }, { "grad_norm": 0.20587463676929474, "learning_rate": 5.9778840382608794e-05, "loss": 0.0129, "step": 18620 }, { "grad_norm": 0.22980909049510956, "learning_rate": 5.9738298579144695e-05, "loss": 0.013, "step": 18630 }, { "grad_norm": 0.15688377618789673, "learning_rate": 5.9697750119645314e-05, "loss": 0.0105, "step": 18640 }, { "grad_norm": 0.1592477411031723, "learning_rate": 5.96571950318252e-05, "loss": 0.0104, "step": 18650 }, { "grad_norm": 0.17292410135269165, "learning_rate": 5.9616633343403316e-05, "loss": 0.0107, "step": 18660 }, { "grad_norm": 0.1760510951280594, "learning_rate": 5.957606508210324e-05, "loss": 0.0099, "step": 18670 }, { "grad_norm": 0.2527870237827301, "learning_rate": 5.953549027565297e-05, "loss": 0.0132, "step": 18680 }, { "grad_norm": 0.18233086168766022, "learning_rate": 5.949490895178501e-05, "loss": 0.0119, "step": 18690 }, { "grad_norm": 0.15785139799118042, "learning_rate": 5.945432113823632e-05, "loss": 0.0113, "step": 18700 }, { "grad_norm": 0.15468233823776245, "learning_rate": 5.9413726862748276e-05, "loss": 0.0097, "step": 18710 }, { "grad_norm": 0.12792924046516418, "learning_rate": 5.9373126153066694e-05, "loss": 0.0098, "step": 18720 }, { "grad_norm": 0.23507103323936462, "learning_rate": 5.933251903694177e-05, "loss": 0.0121, "step": 18730 }, { "grad_norm": 0.15603339672088623, "learning_rate": 5.929190554212807e-05, "loss": 0.0102, "step": 18740 }, { "grad_norm": 0.2416888177394867, "learning_rate": 5.9251285696384565e-05, "loss": 0.0101, "step": 18750 }, { "grad_norm": 0.12027798593044281, "learning_rate": 5.921065952747451e-05, "loss": 0.0078, "step": 18760 }, { "grad_norm": 0.1729857325553894, "learning_rate": 5.917002706316552e-05, "loss": 0.0099, "step": 18770 }, { "grad_norm": 0.19667159020900726, "learning_rate": 5.912938833122952e-05, "loss": 0.0107, "step": 18780 }, { "grad_norm": 0.14239494502544403, "learning_rate": 5.908874335944265e-05, "loss": 0.0087, "step": 18790 }, { "grad_norm": 0.25532954931259155, "learning_rate": 5.904809217558542e-05, "loss": 0.0113, "step": 18800 }, { "grad_norm": 0.12284036725759506, "learning_rate": 5.90074348074425e-05, "loss": 0.0099, "step": 18810 }, { "grad_norm": 0.20078302919864655, "learning_rate": 5.8966771282802814e-05, "loss": 0.0142, "step": 18820 }, { "grad_norm": 0.23168456554412842, "learning_rate": 5.892610162945952e-05, "loss": 0.0109, "step": 18830 }, { "grad_norm": 0.17115157842636108, "learning_rate": 5.8885425875209924e-05, "loss": 0.0108, "step": 18840 }, { "grad_norm": 0.15967683494091034, "learning_rate": 5.884474404785553e-05, "loss": 0.0103, "step": 18850 }, { "grad_norm": 0.21698129177093506, "learning_rate": 5.8804056175201983e-05, "loss": 0.0103, "step": 18860 }, { "grad_norm": 0.15489612519741058, "learning_rate": 5.876336228505904e-05, "loss": 0.0092, "step": 18870 }, { "grad_norm": 0.1857990026473999, "learning_rate": 5.872266240524062e-05, "loss": 0.0117, "step": 18880 }, { "grad_norm": 0.2154826670885086, "learning_rate": 5.86819565635647e-05, "loss": 0.0131, "step": 18890 }, { "grad_norm": 0.1533348113298416, "learning_rate": 5.8641244787853334e-05, "loss": 0.0109, "step": 18900 }, { "grad_norm": 0.2406163364648819, "learning_rate": 5.860052710593265e-05, "loss": 0.0119, "step": 18910 }, { "grad_norm": 0.23578913509845734, "learning_rate": 5.855980354563276e-05, "loss": 0.0112, "step": 18920 }, { "grad_norm": 0.18160945177078247, "learning_rate": 5.8519074134787874e-05, "loss": 0.0119, "step": 18930 }, { "grad_norm": 0.1525498330593109, "learning_rate": 5.847833890123614e-05, "loss": 0.0114, "step": 18940 }, { "grad_norm": 0.19919167459011078, "learning_rate": 5.8437597872819737e-05, "loss": 0.0122, "step": 18950 }, { "grad_norm": 0.20340481400489807, "learning_rate": 5.839685107738473e-05, "loss": 0.0127, "step": 18960 }, { "grad_norm": 0.21752406656742096, "learning_rate": 5.835609854278118e-05, "loss": 0.0111, "step": 18970 }, { "grad_norm": 0.16528844833374023, "learning_rate": 5.831534029686308e-05, "loss": 0.011, "step": 18980 }, { "grad_norm": 0.19341592490673065, "learning_rate": 5.82745763674883e-05, "loss": 0.0115, "step": 18990 }, { "grad_norm": 0.16319723427295685, "learning_rate": 5.823380678251861e-05, "loss": 0.0114, "step": 19000 }, { "grad_norm": 0.1666461080312729, "learning_rate": 5.81930315698196e-05, "loss": 0.0107, "step": 19010 }, { "grad_norm": 0.2114814817905426, "learning_rate": 5.815225075726076e-05, "loss": 0.0113, "step": 19020 }, { "grad_norm": 0.2424546182155609, "learning_rate": 5.811146437271543e-05, "loss": 0.0126, "step": 19030 }, { "grad_norm": 0.16187028586864471, "learning_rate": 5.807067244406066e-05, "loss": 0.0095, "step": 19040 }, { "grad_norm": 0.18938584625720978, "learning_rate": 5.8029874999177405e-05, "loss": 0.0097, "step": 19050 }, { "grad_norm": 0.18250781297683716, "learning_rate": 5.798907206595029e-05, "loss": 0.0091, "step": 19060 }, { "grad_norm": 0.20773102343082428, "learning_rate": 5.794826367226773e-05, "loss": 0.0116, "step": 19070 }, { "grad_norm": 0.17175565659999847, "learning_rate": 5.790744984602193e-05, "loss": 0.0136, "step": 19080 }, { "grad_norm": 0.15875191986560822, "learning_rate": 5.786663061510872e-05, "loss": 0.0115, "step": 19090 }, { "grad_norm": 0.1778509020805359, "learning_rate": 5.782580600742765e-05, "loss": 0.01, "step": 19100 }, { "grad_norm": 0.13944341242313385, "learning_rate": 5.7784976050881965e-05, "loss": 0.0101, "step": 19110 }, { "grad_norm": 0.156545951962471, "learning_rate": 5.774414077337855e-05, "loss": 0.0122, "step": 19120 }, { "grad_norm": 0.16350023448467255, "learning_rate": 5.770330020282796e-05, "loss": 0.0096, "step": 19130 }, { "grad_norm": 0.13645535707473755, "learning_rate": 5.7662454367144317e-05, "loss": 0.01, "step": 19140 }, { "grad_norm": 0.1758413314819336, "learning_rate": 5.762160329424536e-05, "loss": 0.0104, "step": 19150 }, { "grad_norm": 0.11712971329689026, "learning_rate": 5.7580747012052416e-05, "loss": 0.0089, "step": 19160 }, { "grad_norm": 0.20563821494579315, "learning_rate": 5.753988554849037e-05, "loss": 0.0096, "step": 19170 }, { "grad_norm": 0.2107222080230713, "learning_rate": 5.749901893148766e-05, "loss": 0.0126, "step": 19180 }, { "grad_norm": 0.15058927237987518, "learning_rate": 5.745814718897621e-05, "loss": 0.0115, "step": 19190 }, { "grad_norm": 0.2263890504837036, "learning_rate": 5.74172703488915e-05, "loss": 0.0114, "step": 19200 }, { "grad_norm": 0.18013018369674683, "learning_rate": 5.737638843917242e-05, "loss": 0.01, "step": 19210 }, { "grad_norm": 0.15816931426525116, "learning_rate": 5.73355014877614e-05, "loss": 0.0088, "step": 19220 }, { "grad_norm": 0.19758100807666779, "learning_rate": 5.7294609522604316e-05, "loss": 0.0111, "step": 19230 }, { "grad_norm": 0.17493964731693268, "learning_rate": 5.7253712571650376e-05, "loss": 0.0103, "step": 19240 }, { "grad_norm": 0.19114360213279724, "learning_rate": 5.721281066285229e-05, "loss": 0.013, "step": 19250 }, { "grad_norm": 0.16378401219844818, "learning_rate": 5.717190382416615e-05, "loss": 0.0095, "step": 19260 }, { "grad_norm": 0.18763716518878937, "learning_rate": 5.713099208355135e-05, "loss": 0.0103, "step": 19270 }, { "grad_norm": 0.21215319633483887, "learning_rate": 5.709007546897074e-05, "loss": 0.0115, "step": 19280 }, { "grad_norm": 0.1468556970357895, "learning_rate": 5.704915400839037e-05, "loss": 0.0108, "step": 19290 }, { "grad_norm": 0.18272003531455994, "learning_rate": 5.700822772977971e-05, "loss": 0.0088, "step": 19300 }, { "grad_norm": 0.155790314078331, "learning_rate": 5.696729666111148e-05, "loss": 0.0084, "step": 19310 }, { "grad_norm": 0.18160168826580048, "learning_rate": 5.692636083036168e-05, "loss": 0.0103, "step": 19320 }, { "grad_norm": 0.23969252407550812, "learning_rate": 5.688542026550958e-05, "loss": 0.0128, "step": 19330 }, { "grad_norm": 0.1943143755197525, "learning_rate": 5.684447499453763e-05, "loss": 0.0124, "step": 19340 }, { "grad_norm": 0.13108666241168976, "learning_rate": 5.680352504543156e-05, "loss": 0.0101, "step": 19350 }, { "grad_norm": 0.17718032002449036, "learning_rate": 5.67625704461803e-05, "loss": 0.0103, "step": 19360 }, { "grad_norm": 0.21999873220920563, "learning_rate": 5.672161122477589e-05, "loss": 0.0099, "step": 19370 }, { "grad_norm": 0.21122248470783234, "learning_rate": 5.668064740921359e-05, "loss": 0.0116, "step": 19380 }, { "grad_norm": 0.18564988672733307, "learning_rate": 5.663967902749179e-05, "loss": 0.0124, "step": 19390 }, { "grad_norm": 0.2320733517408371, "learning_rate": 5.6598706107611965e-05, "loss": 0.0109, "step": 19400 }, { "grad_norm": 0.21916463971138, "learning_rate": 5.655772867757876e-05, "loss": 0.0127, "step": 19410 }, { "grad_norm": 0.21790042519569397, "learning_rate": 5.651674676539982e-05, "loss": 0.0093, "step": 19420 }, { "grad_norm": 0.2126656323671341, "learning_rate": 5.647576039908593e-05, "loss": 0.0109, "step": 19430 }, { "grad_norm": 0.15996992588043213, "learning_rate": 5.6434769606650864e-05, "loss": 0.0101, "step": 19440 }, { "grad_norm": 0.23525521159172058, "learning_rate": 5.639377441611143e-05, "loss": 0.0116, "step": 19450 }, { "grad_norm": 0.1840997040271759, "learning_rate": 5.635277485548751e-05, "loss": 0.009, "step": 19460 }, { "grad_norm": 0.20370618999004364, "learning_rate": 5.631177095280186e-05, "loss": 0.0117, "step": 19470 }, { "grad_norm": 0.14914971590042114, "learning_rate": 5.627076273608027e-05, "loss": 0.0126, "step": 19480 }, { "grad_norm": 0.14675380289554596, "learning_rate": 5.622975023335148e-05, "loss": 0.0094, "step": 19490 }, { "grad_norm": 0.15997780859470367, "learning_rate": 5.618873347264716e-05, "loss": 0.0083, "step": 19500 }, { "grad_norm": 0.1783151477575302, "learning_rate": 5.614771248200188e-05, "loss": 0.0129, "step": 19510 }, { "grad_norm": 0.1663934290409088, "learning_rate": 5.6106687289453066e-05, "loss": 0.0093, "step": 19520 }, { "grad_norm": 0.18541212379932404, "learning_rate": 5.606565792304108e-05, "loss": 0.0104, "step": 19530 }, { "grad_norm": 0.18584875762462616, "learning_rate": 5.602462441080909e-05, "loss": 0.0098, "step": 19540 }, { "grad_norm": 0.244888573884964, "learning_rate": 5.5983586780803135e-05, "loss": 0.0119, "step": 19550 }, { "grad_norm": 0.19530506432056427, "learning_rate": 5.594254506107205e-05, "loss": 0.01, "step": 19560 }, { "grad_norm": 0.1422291100025177, "learning_rate": 5.590149927966743e-05, "loss": 0.0129, "step": 19570 }, { "grad_norm": 0.1442227065563202, "learning_rate": 5.58604494646437e-05, "loss": 0.0081, "step": 19580 }, { "grad_norm": 0.20947405695915222, "learning_rate": 5.5819395644058025e-05, "loss": 0.011, "step": 19590 }, { "grad_norm": 0.2942711114883423, "learning_rate": 5.577833784597031e-05, "loss": 0.0123, "step": 19600 }, { "grad_norm": 0.11002390831708908, "learning_rate": 5.573727609844316e-05, "loss": 0.0094, "step": 19610 }, { "grad_norm": 0.16405245661735535, "learning_rate": 5.5696210429541884e-05, "loss": 0.0103, "step": 19620 }, { "grad_norm": 0.17300032079219818, "learning_rate": 5.565514086733451e-05, "loss": 0.0084, "step": 19630 }, { "grad_norm": 0.16733013093471527, "learning_rate": 5.5614067439891657e-05, "loss": 0.0124, "step": 19640 }, { "grad_norm": 0.11676310747861862, "learning_rate": 5.557299017528666e-05, "loss": 0.0092, "step": 19650 }, { "grad_norm": 0.15724556148052216, "learning_rate": 5.5531909101595436e-05, "loss": 0.0089, "step": 19660 }, { "grad_norm": 0.16378900408744812, "learning_rate": 5.549082424689649e-05, "loss": 0.0111, "step": 19670 }, { "grad_norm": 0.19446758925914764, "learning_rate": 5.544973563927095e-05, "loss": 0.0118, "step": 19680 }, { "grad_norm": 0.16659249365329742, "learning_rate": 5.540864330680249e-05, "loss": 0.0097, "step": 19690 }, { "grad_norm": 0.1392250955104828, "learning_rate": 5.536754727757733e-05, "loss": 0.0087, "step": 19700 }, { "grad_norm": 0.15242059528827667, "learning_rate": 5.532644757968422e-05, "loss": 0.011, "step": 19710 }, { "grad_norm": 0.1876497119665146, "learning_rate": 5.528534424121441e-05, "loss": 0.0099, "step": 19720 }, { "grad_norm": 0.1452251821756363, "learning_rate": 5.524423729026165e-05, "loss": 0.0126, "step": 19730 }, { "grad_norm": 0.18570837378501892, "learning_rate": 5.5203126754922164e-05, "loss": 0.0095, "step": 19740 }, { "grad_norm": 0.12852148711681366, "learning_rate": 5.5162012663294585e-05, "loss": 0.0092, "step": 19750 }, { "grad_norm": 0.14368732273578644, "learning_rate": 5.512089504348003e-05, "loss": 0.008, "step": 19760 }, { "grad_norm": 0.2419804483652115, "learning_rate": 5.5079773923582e-05, "loss": 0.0117, "step": 19770 }, { "grad_norm": 0.15602976083755493, "learning_rate": 5.50386493317064e-05, "loss": 0.0095, "step": 19780 }, { "grad_norm": 0.18634602427482605, "learning_rate": 5.49975212959615e-05, "loss": 0.0104, "step": 19790 }, { "grad_norm": 0.14215458929538727, "learning_rate": 5.4956389844457904e-05, "loss": 0.0086, "step": 19800 }, { "grad_norm": 0.17820538580417633, "learning_rate": 5.491525500530859e-05, "loss": 0.0099, "step": 19810 }, { "grad_norm": 0.2595156729221344, "learning_rate": 5.487411680662882e-05, "loss": 0.0109, "step": 19820 }, { "grad_norm": 0.18005190789699554, "learning_rate": 5.483297527653618e-05, "loss": 0.0118, "step": 19830 }, { "grad_norm": 0.1714392602443695, "learning_rate": 5.4791830443150516e-05, "loss": 0.0102, "step": 19840 }, { "grad_norm": 0.16832739114761353, "learning_rate": 5.475068233459392e-05, "loss": 0.0089, "step": 19850 }, { "grad_norm": 0.14892463386058807, "learning_rate": 5.470953097899075e-05, "loss": 0.0123, "step": 19860 }, { "grad_norm": 0.13181041181087494, "learning_rate": 5.466837640446756e-05, "loss": 0.0126, "step": 19870 }, { "grad_norm": 0.19003096222877502, "learning_rate": 5.462721863915312e-05, "loss": 0.0117, "step": 19880 }, { "grad_norm": 0.21867692470550537, "learning_rate": 5.4586057711178374e-05, "loss": 0.0108, "step": 19890 }, { "grad_norm": 0.22292470932006836, "learning_rate": 5.454489364867642e-05, "loss": 0.0113, "step": 19900 }, { "grad_norm": 0.19985413551330566, "learning_rate": 5.4503726479782523e-05, "loss": 0.0133, "step": 19910 }, { "grad_norm": 0.19733060896396637, "learning_rate": 5.446255623263403e-05, "loss": 0.0114, "step": 19920 }, { "grad_norm": 0.24446991086006165, "learning_rate": 5.4421382935370445e-05, "loss": 0.0099, "step": 19930 }, { "grad_norm": 0.16191977262496948, "learning_rate": 5.438020661613331e-05, "loss": 0.0107, "step": 19940 }, { "grad_norm": 0.11695453524589539, "learning_rate": 5.433902730306625e-05, "loss": 0.008, "step": 19950 }, { "grad_norm": 0.17674188315868378, "learning_rate": 5.429784502431495e-05, "loss": 0.0109, "step": 19960 }, { "grad_norm": 0.22174127399921417, "learning_rate": 5.42566598080271e-05, "loss": 0.0123, "step": 19970 }, { "grad_norm": 0.23859024047851562, "learning_rate": 5.421547168235241e-05, "loss": 0.0094, "step": 19980 }, { "grad_norm": 0.11793214082717896, "learning_rate": 5.417428067544258e-05, "loss": 0.0093, "step": 19990 }, { "grad_norm": 0.14707548916339874, "learning_rate": 5.413308681545126e-05, "loss": 0.01, "step": 20000 }, { "grad_norm": 0.19218401610851288, "learning_rate": 5.409189013053408e-05, "loss": 0.0143, "step": 20010 }, { "grad_norm": 0.14852648973464966, "learning_rate": 5.4050690648848576e-05, "loss": 0.0102, "step": 20020 }, { "grad_norm": 0.2007443904876709, "learning_rate": 5.400948839855421e-05, "loss": 0.01, "step": 20030 }, { "grad_norm": 0.15986040234565735, "learning_rate": 5.396828340781234e-05, "loss": 0.0095, "step": 20040 }, { "grad_norm": 0.198152557015419, "learning_rate": 5.392707570478617e-05, "loss": 0.0093, "step": 20050 }, { "grad_norm": 0.16483210027217865, "learning_rate": 5.388586531764078e-05, "loss": 0.0108, "step": 20060 }, { "grad_norm": 0.21011587977409363, "learning_rate": 5.384465227454311e-05, "loss": 0.0112, "step": 20070 }, { "grad_norm": 0.1796422302722931, "learning_rate": 5.380343660366184e-05, "loss": 0.0097, "step": 20080 }, { "grad_norm": 0.13423745334148407, "learning_rate": 5.376221833316752e-05, "loss": 0.012, "step": 20090 }, { "grad_norm": 0.2316216677427292, "learning_rate": 5.3720997491232436e-05, "loss": 0.0142, "step": 20100 }, { "grad_norm": 0.20026564598083496, "learning_rate": 5.367977410603068e-05, "loss": 0.0131, "step": 20110 }, { "grad_norm": 0.2273377776145935, "learning_rate": 5.3638548205738004e-05, "loss": 0.0135, "step": 20120 }, { "grad_norm": 0.1917782425880432, "learning_rate": 5.359731981853194e-05, "loss": 0.0119, "step": 20130 }, { "grad_norm": 0.20582503080368042, "learning_rate": 5.35560889725917e-05, "loss": 0.0102, "step": 20140 }, { "grad_norm": 0.17976713180541992, "learning_rate": 5.3514855696098176e-05, "loss": 0.013, "step": 20150 }, { "grad_norm": 0.19070473313331604, "learning_rate": 5.347362001723394e-05, "loss": 0.0096, "step": 20160 }, { "grad_norm": 0.16021256148815155, "learning_rate": 5.3432381964183176e-05, "loss": 0.0095, "step": 20170 }, { "grad_norm": 0.1686970591545105, "learning_rate": 5.3391141565131685e-05, "loss": 0.0091, "step": 20180 }, { "grad_norm": 0.14528436958789825, "learning_rate": 5.3349898848266935e-05, "loss": 0.0099, "step": 20190 }, { "grad_norm": 0.1483396589756012, "learning_rate": 5.330865384177789e-05, "loss": 0.0095, "step": 20200 }, { "grad_norm": 0.139942467212677, "learning_rate": 5.326740657385515e-05, "loss": 0.0096, "step": 20210 }, { "grad_norm": 0.17186284065246582, "learning_rate": 5.322615707269083e-05, "loss": 0.0116, "step": 20220 }, { "grad_norm": 0.11839289218187332, "learning_rate": 5.318490536647856e-05, "loss": 0.0078, "step": 20230 }, { "grad_norm": 0.21482403576374054, "learning_rate": 5.3143651483413524e-05, "loss": 0.0093, "step": 20240 }, { "grad_norm": 0.16619282960891724, "learning_rate": 5.310239545169232e-05, "loss": 0.0112, "step": 20250 }, { "grad_norm": 0.20561708509922028, "learning_rate": 5.30611372995131e-05, "loss": 0.0117, "step": 20260 }, { "grad_norm": 0.20493929088115692, "learning_rate": 5.30198770550754e-05, "loss": 0.0094, "step": 20270 }, { "grad_norm": 0.19817513227462769, "learning_rate": 5.297861474658019e-05, "loss": 0.0097, "step": 20280 }, { "grad_norm": 0.1776421070098877, "learning_rate": 5.29373504022299e-05, "loss": 0.0101, "step": 20290 }, { "grad_norm": 0.1504308581352234, "learning_rate": 5.28960840502283e-05, "loss": 0.0088, "step": 20300 }, { "grad_norm": 0.1576734185218811, "learning_rate": 5.285481571878056e-05, "loss": 0.0146, "step": 20310 }, { "grad_norm": 0.2245406210422516, "learning_rate": 5.281354543609321e-05, "loss": 0.0096, "step": 20320 }, { "grad_norm": 0.19511690735816956, "learning_rate": 5.277227323037406e-05, "loss": 0.0105, "step": 20330 }, { "grad_norm": 0.14418284595012665, "learning_rate": 5.273099912983233e-05, "loss": 0.0101, "step": 20340 }, { "grad_norm": 0.15494099259376526, "learning_rate": 5.268972316267843e-05, "loss": 0.0115, "step": 20350 }, { "grad_norm": 0.1642446219921112, "learning_rate": 5.26484453571241e-05, "loss": 0.0108, "step": 20360 }, { "grad_norm": 0.12848328053951263, "learning_rate": 5.260716574138235e-05, "loss": 0.0107, "step": 20370 }, { "grad_norm": 0.13435272872447968, "learning_rate": 5.256588434366739e-05, "loss": 0.0109, "step": 20380 }, { "grad_norm": 0.11088842898607254, "learning_rate": 5.25246011921947e-05, "loss": 0.0124, "step": 20390 }, { "grad_norm": 0.1587885171175003, "learning_rate": 5.248331631518089e-05, "loss": 0.0114, "step": 20400 }, { "grad_norm": 0.15072987973690033, "learning_rate": 5.244202974084379e-05, "loss": 0.0101, "step": 20410 }, { "grad_norm": 0.1783527284860611, "learning_rate": 5.240074149740239e-05, "loss": 0.0071, "step": 20420 }, { "grad_norm": 0.14056755602359772, "learning_rate": 5.2359451613076814e-05, "loss": 0.0078, "step": 20430 }, { "grad_norm": 0.18160699307918549, "learning_rate": 5.231816011608832e-05, "loss": 0.0099, "step": 20440 }, { "grad_norm": 0.21187055110931396, "learning_rate": 5.227686703465924e-05, "loss": 0.0079, "step": 20450 }, { "grad_norm": 0.1519923061132431, "learning_rate": 5.2235572397013e-05, "loss": 0.0105, "step": 20460 }, { "grad_norm": 0.1735655963420868, "learning_rate": 5.2194276231374114e-05, "loss": 0.0097, "step": 20470 }, { "grad_norm": 0.18974429368972778, "learning_rate": 5.21529785659681e-05, "loss": 0.0116, "step": 20480 }, { "grad_norm": 0.22063840925693512, "learning_rate": 5.2111679429021565e-05, "loss": 0.0123, "step": 20490 }, { "grad_norm": 0.20016400516033173, "learning_rate": 5.207037884876205e-05, "loss": 0.0109, "step": 20500 }, { "grad_norm": 0.19901202619075775, "learning_rate": 5.202907685341809e-05, "loss": 0.0092, "step": 20510 }, { "grad_norm": 0.1856459230184555, "learning_rate": 5.198777347121926e-05, "loss": 0.0137, "step": 20520 }, { "grad_norm": 0.13889014720916748, "learning_rate": 5.194646873039598e-05, "loss": 0.0127, "step": 20530 }, { "grad_norm": 0.10956840217113495, "learning_rate": 5.1905162659179696e-05, "loss": 0.0082, "step": 20540 }, { "grad_norm": 0.21422898769378662, "learning_rate": 5.18638552858027e-05, "loss": 0.0117, "step": 20550 }, { "grad_norm": 0.17448019981384277, "learning_rate": 5.182254663849818e-05, "loss": 0.0126, "step": 20560 }, { "grad_norm": 0.15047144889831543, "learning_rate": 5.178123674550023e-05, "loss": 0.0111, "step": 20570 }, { "grad_norm": 0.18853096663951874, "learning_rate": 5.173992563504375e-05, "loss": 0.0119, "step": 20580 }, { "grad_norm": 0.12601545453071594, "learning_rate": 5.169861333536451e-05, "loss": 0.0092, "step": 20590 }, { "grad_norm": 0.1831161230802536, "learning_rate": 5.165729987469907e-05, "loss": 0.0095, "step": 20600 }, { "grad_norm": 0.15223097801208496, "learning_rate": 5.161598528128478e-05, "loss": 0.0097, "step": 20610 }, { "grad_norm": 0.21472303569316864, "learning_rate": 5.157466958335981e-05, "loss": 0.0084, "step": 20620 }, { "grad_norm": 0.17257115244865417, "learning_rate": 5.1533352809163025e-05, "loss": 0.0123, "step": 20630 }, { "grad_norm": 0.18351468443870544, "learning_rate": 5.1492034986934046e-05, "loss": 0.0101, "step": 20640 }, { "grad_norm": 0.19325654208660126, "learning_rate": 5.1450716144913225e-05, "loss": 0.012, "step": 20650 }, { "grad_norm": 0.14511889219284058, "learning_rate": 5.1409396311341595e-05, "loss": 0.0125, "step": 20660 }, { "grad_norm": 0.1698617935180664, "learning_rate": 5.136807551446089e-05, "loss": 0.0119, "step": 20670 }, { "grad_norm": 0.2380286157131195, "learning_rate": 5.132675378251346e-05, "loss": 0.01, "step": 20680 }, { "grad_norm": 0.19212239980697632, "learning_rate": 5.1285431143742325e-05, "loss": 0.0096, "step": 20690 }, { "grad_norm": 0.15524843335151672, "learning_rate": 5.1244107626391136e-05, "loss": 0.0111, "step": 20700 }, { "grad_norm": 0.16969504952430725, "learning_rate": 5.12027832587041e-05, "loss": 0.0107, "step": 20710 }, { "grad_norm": 0.22336475551128387, "learning_rate": 5.116145806892607e-05, "loss": 0.0102, "step": 20720 }, { "grad_norm": 0.20108655095100403, "learning_rate": 5.1120132085302384e-05, "loss": 0.0115, "step": 20730 }, { "grad_norm": 0.2764921486377716, "learning_rate": 5.107880533607898e-05, "loss": 0.0125, "step": 20740 }, { "grad_norm": 0.18169629573822021, "learning_rate": 5.103747784950231e-05, "loss": 0.01, "step": 20750 }, { "grad_norm": 0.21673887968063354, "learning_rate": 5.09961496538193e-05, "loss": 0.0113, "step": 20760 }, { "grad_norm": 0.18885833024978638, "learning_rate": 5.095482077727742e-05, "loss": 0.0129, "step": 20770 }, { "grad_norm": 0.13495148718357086, "learning_rate": 5.091349124812452e-05, "loss": 0.0086, "step": 20780 }, { "grad_norm": 0.1699829250574112, "learning_rate": 5.087216109460897e-05, "loss": 0.0092, "step": 20790 }, { "grad_norm": 0.21483220160007477, "learning_rate": 5.083083034497954e-05, "loss": 0.0092, "step": 20800 }, { "grad_norm": 0.2148907631635666, "learning_rate": 5.07894990274854e-05, "loss": 0.0083, "step": 20810 }, { "grad_norm": 0.2973144054412842, "learning_rate": 5.074816717037614e-05, "loss": 0.0091, "step": 20820 }, { "grad_norm": 0.16701437532901764, "learning_rate": 5.070683480190165e-05, "loss": 0.01, "step": 20830 }, { "grad_norm": 0.1720982789993286, "learning_rate": 5.066550195031223e-05, "loss": 0.0091, "step": 20840 }, { "grad_norm": 0.19352880120277405, "learning_rate": 5.062416864385852e-05, "loss": 0.009, "step": 20850 }, { "grad_norm": 0.1822950541973114, "learning_rate": 5.058283491079142e-05, "loss": 0.0108, "step": 20860 }, { "grad_norm": 0.13041743636131287, "learning_rate": 5.054150077936216e-05, "loss": 0.0091, "step": 20870 }, { "grad_norm": 0.18575985729694366, "learning_rate": 5.0500166277822214e-05, "loss": 0.0107, "step": 20880 }, { "grad_norm": 0.1572202444076538, "learning_rate": 5.0458831434423334e-05, "loss": 0.0098, "step": 20890 }, { "grad_norm": 0.13561148941516876, "learning_rate": 5.0417496277417506e-05, "loss": 0.0079, "step": 20900 }, { "grad_norm": 0.1602659970521927, "learning_rate": 5.037616083505691e-05, "loss": 0.0099, "step": 20910 }, { "grad_norm": 0.09795871376991272, "learning_rate": 5.0334825135593935e-05, "loss": 0.0092, "step": 20920 }, { "grad_norm": 0.1538141816854477, "learning_rate": 5.029348920728111e-05, "loss": 0.009, "step": 20930 }, { "grad_norm": 0.14994008839130402, "learning_rate": 5.0252153078371186e-05, "loss": 0.0123, "step": 20940 }, { "grad_norm": 0.1490968018770218, "learning_rate": 5.021081677711704e-05, "loss": 0.0083, "step": 20950 }, { "grad_norm": 0.20658724009990692, "learning_rate": 5.016948033177159e-05, "loss": 0.0096, "step": 20960 }, { "grad_norm": 0.20378297567367554, "learning_rate": 5.012814377058793e-05, "loss": 0.0099, "step": 20970 }, { "grad_norm": 0.14254041016101837, "learning_rate": 5.008680712181921e-05, "loss": 0.0105, "step": 20980 }, { "grad_norm": 0.14637422561645508, "learning_rate": 5.0045470413718645e-05, "loss": 0.0079, "step": 20990 }, { "grad_norm": 0.1280822455883026, "learning_rate": 5.00041336745395e-05, "loss": 0.0122, "step": 21000 }, { "grad_norm": 0.15485943853855133, "learning_rate": 4.996279693253499e-05, "loss": 0.0092, "step": 21010 }, { "grad_norm": 0.216073676943779, "learning_rate": 4.992146021595847e-05, "loss": 0.0111, "step": 21020 }, { "grad_norm": 0.14522597193717957, "learning_rate": 4.988012355306313e-05, "loss": 0.0098, "step": 21030 }, { "grad_norm": 0.20835141837596893, "learning_rate": 4.98387869721022e-05, "loss": 0.0103, "step": 21040 }, { "grad_norm": 0.2412874549627304, "learning_rate": 4.9797450501328866e-05, "loss": 0.0125, "step": 21050 }, { "grad_norm": 0.15745939314365387, "learning_rate": 4.97561141689962e-05, "loss": 0.0087, "step": 21060 }, { "grad_norm": 0.16182297468185425, "learning_rate": 4.971477800335721e-05, "loss": 0.0132, "step": 21070 }, { "grad_norm": 0.15224532783031464, "learning_rate": 4.967344203266475e-05, "loss": 0.0103, "step": 21080 }, { "grad_norm": 0.20321892201900482, "learning_rate": 4.9632106285171584e-05, "loss": 0.0095, "step": 21090 }, { "grad_norm": 0.1375395953655243, "learning_rate": 4.959077078913031e-05, "loss": 0.01, "step": 21100 }, { "grad_norm": 0.19705088436603546, "learning_rate": 4.954943557279333e-05, "loss": 0.0129, "step": 21110 }, { "grad_norm": 0.15188318490982056, "learning_rate": 4.9508100664412916e-05, "loss": 0.0088, "step": 21120 }, { "grad_norm": 0.18120349943637848, "learning_rate": 4.946676609224105e-05, "loss": 0.0105, "step": 21130 }, { "grad_norm": 0.20678679645061493, "learning_rate": 4.942543188452952e-05, "loss": 0.0111, "step": 21140 }, { "grad_norm": 0.1690378338098526, "learning_rate": 4.938409806952988e-05, "loss": 0.0113, "step": 21150 }, { "grad_norm": 0.18460261821746826, "learning_rate": 4.93427646754934e-05, "loss": 0.0097, "step": 21160 }, { "grad_norm": 0.12360084056854248, "learning_rate": 4.930143173067108e-05, "loss": 0.0095, "step": 21170 }, { "grad_norm": 0.16178223490715027, "learning_rate": 4.9260099263313565e-05, "loss": 0.0098, "step": 21180 }, { "grad_norm": 0.17360815405845642, "learning_rate": 4.921876730167123e-05, "loss": 0.0107, "step": 21190 }, { "grad_norm": 0.15102417767047882, "learning_rate": 4.917743587399409e-05, "loss": 0.0099, "step": 21200 }, { "grad_norm": 0.17307902872562408, "learning_rate": 4.913610500853178e-05, "loss": 0.0088, "step": 21210 }, { "grad_norm": 0.17866092920303345, "learning_rate": 4.909477473353354e-05, "loss": 0.0095, "step": 21220 }, { "grad_norm": 0.13881123065948486, "learning_rate": 4.9053445077248236e-05, "loss": 0.0105, "step": 21230 }, { "grad_norm": 0.13407909870147705, "learning_rate": 4.901211606792429e-05, "loss": 0.0093, "step": 21240 }, { "grad_norm": 0.1405525505542755, "learning_rate": 4.89707877338097e-05, "loss": 0.009, "step": 21250 }, { "grad_norm": 0.15884876251220703, "learning_rate": 4.892946010315199e-05, "loss": 0.009, "step": 21260 }, { "grad_norm": 0.18559172749519348, "learning_rate": 4.8888133204198204e-05, "loss": 0.0115, "step": 21270 }, { "grad_norm": 0.17541788518428802, "learning_rate": 4.8846807065194886e-05, "loss": 0.0105, "step": 21280 }, { "grad_norm": 0.15451651811599731, "learning_rate": 4.880548171438806e-05, "loss": 0.0097, "step": 21290 }, { "grad_norm": 0.1759735643863678, "learning_rate": 4.8764157180023245e-05, "loss": 0.0083, "step": 21300 }, { "grad_norm": 0.15284259617328644, "learning_rate": 4.872283349034533e-05, "loss": 0.012, "step": 21310 }, { "grad_norm": 0.13582564890384674, "learning_rate": 4.8681510673598674e-05, "loss": 0.0096, "step": 21320 }, { "grad_norm": 0.18203262984752655, "learning_rate": 4.8640188758027046e-05, "loss": 0.0073, "step": 21330 }, { "grad_norm": 0.11013861000537872, "learning_rate": 4.859886777187357e-05, "loss": 0.0108, "step": 21340 }, { "grad_norm": 0.21045057475566864, "learning_rate": 4.855754774338077e-05, "loss": 0.011, "step": 21350 }, { "grad_norm": 0.17074747383594513, "learning_rate": 4.851622870079048e-05, "loss": 0.0092, "step": 21360 }, { "grad_norm": 0.17731010913848877, "learning_rate": 4.847491067234389e-05, "loss": 0.0098, "step": 21370 }, { "grad_norm": 0.16429683566093445, "learning_rate": 4.843359368628146e-05, "loss": 0.0082, "step": 21380 }, { "grad_norm": 0.19893762469291687, "learning_rate": 4.8392277770842975e-05, "loss": 0.0117, "step": 21390 }, { "grad_norm": 0.16713592410087585, "learning_rate": 4.83509629542675e-05, "loss": 0.0101, "step": 21400 }, { "grad_norm": 0.13601337373256683, "learning_rate": 4.830964926479329e-05, "loss": 0.0102, "step": 21410 }, { "grad_norm": 0.17397554218769073, "learning_rate": 4.826833673065785e-05, "loss": 0.0106, "step": 21420 }, { "grad_norm": 0.1983485221862793, "learning_rate": 4.822702538009794e-05, "loss": 0.0095, "step": 21430 }, { "grad_norm": 0.14726927876472473, "learning_rate": 4.818571524134945e-05, "loss": 0.0112, "step": 21440 }, { "grad_norm": 0.16635772585868835, "learning_rate": 4.8144406342647496e-05, "loss": 0.0122, "step": 21450 }, { "grad_norm": 0.14058612287044525, "learning_rate": 4.81030987122263e-05, "loss": 0.0086, "step": 21460 }, { "grad_norm": 0.15117084980010986, "learning_rate": 4.806179237831926e-05, "loss": 0.0079, "step": 21470 }, { "grad_norm": 0.11527121812105179, "learning_rate": 4.802048736915884e-05, "loss": 0.008, "step": 21480 }, { "grad_norm": 0.17178112268447876, "learning_rate": 4.797918371297666e-05, "loss": 0.0096, "step": 21490 }, { "grad_norm": 0.13796019554138184, "learning_rate": 4.793788143800334e-05, "loss": 0.0117, "step": 21500 }, { "grad_norm": 0.1591343730688095, "learning_rate": 4.789658057246862e-05, "loss": 0.0077, "step": 21510 }, { "grad_norm": 0.16274389624595642, "learning_rate": 4.7855281144601227e-05, "loss": 0.0094, "step": 21520 }, { "grad_norm": 0.13272596895694733, "learning_rate": 4.781398318262897e-05, "loss": 0.0099, "step": 21530 }, { "grad_norm": 0.1976625919342041, "learning_rate": 4.777268671477858e-05, "loss": 0.0105, "step": 21540 }, { "grad_norm": 0.17483191192150116, "learning_rate": 4.773139176927582e-05, "loss": 0.0084, "step": 21550 }, { "grad_norm": 0.20885606110095978, "learning_rate": 4.769009837434539e-05, "loss": 0.0097, "step": 21560 }, { "grad_norm": 0.15231774747371674, "learning_rate": 4.764880655821095e-05, "loss": 0.0081, "step": 21570 }, { "grad_norm": 0.15307556092739105, "learning_rate": 4.760751634909508e-05, "loss": 0.0114, "step": 21580 }, { "grad_norm": 0.21584713459014893, "learning_rate": 4.756622777521919e-05, "loss": 0.0099, "step": 21590 }, { "grad_norm": 0.15144118666648865, "learning_rate": 4.752494086480368e-05, "loss": 0.0116, "step": 21600 }, { "grad_norm": 0.20636078715324402, "learning_rate": 4.7483655646067744e-05, "loss": 0.0091, "step": 21610 }, { "grad_norm": 0.14778591692447662, "learning_rate": 4.744237214722944e-05, "loss": 0.0105, "step": 21620 }, { "grad_norm": 0.18479369580745697, "learning_rate": 4.740109039650567e-05, "loss": 0.0107, "step": 21630 }, { "grad_norm": 0.19186697900295258, "learning_rate": 4.73598104221121e-05, "loss": 0.0108, "step": 21640 }, { "grad_norm": 0.17472074925899506, "learning_rate": 4.731853225226322e-05, "loss": 0.0079, "step": 21650 }, { "grad_norm": 0.27743983268737793, "learning_rate": 4.727725591517225e-05, "loss": 0.0106, "step": 21660 }, { "grad_norm": 0.16567134857177734, "learning_rate": 4.723598143905119e-05, "loss": 0.0103, "step": 21670 }, { "grad_norm": 0.1490890383720398, "learning_rate": 4.719470885211077e-05, "loss": 0.0099, "step": 21680 }, { "grad_norm": 0.19859552383422852, "learning_rate": 4.7153438182560387e-05, "loss": 0.011, "step": 21690 }, { "grad_norm": 0.24610866606235504, "learning_rate": 4.711216945860815e-05, "loss": 0.0106, "step": 21700 }, { "grad_norm": 0.1421017050743103, "learning_rate": 4.707090270846088e-05, "loss": 0.0104, "step": 21710 }, { "grad_norm": 0.15958398580551147, "learning_rate": 4.702963796032397e-05, "loss": 0.0143, "step": 21720 }, { "grad_norm": 0.16089777648448944, "learning_rate": 4.6988375242401514e-05, "loss": 0.0117, "step": 21730 }, { "grad_norm": 0.12640534341335297, "learning_rate": 4.694711458289618e-05, "loss": 0.0083, "step": 21740 }, { "grad_norm": 0.14844389259815216, "learning_rate": 4.690585601000925e-05, "loss": 0.0088, "step": 21750 }, { "grad_norm": 0.3079046607017517, "learning_rate": 4.686459955194055e-05, "loss": 0.0104, "step": 21760 }, { "grad_norm": 0.13693170249462128, "learning_rate": 4.6823345236888504e-05, "loss": 0.0115, "step": 21770 }, { "grad_norm": 0.23592813313007355, "learning_rate": 4.678209309305002e-05, "loss": 0.0098, "step": 21780 }, { "grad_norm": 0.1501479595899582, "learning_rate": 4.674084314862057e-05, "loss": 0.008, "step": 21790 }, { "grad_norm": 0.23382174968719482, "learning_rate": 4.669959543179409e-05, "loss": 0.0082, "step": 21800 }, { "grad_norm": 0.13573576509952545, "learning_rate": 4.665834997076303e-05, "loss": 0.01, "step": 21810 }, { "grad_norm": 0.19503094255924225, "learning_rate": 4.661710679371823e-05, "loss": 0.0076, "step": 21820 }, { "grad_norm": 0.18774037063121796, "learning_rate": 4.657586592884905e-05, "loss": 0.0094, "step": 21830 }, { "grad_norm": 0.2163110375404358, "learning_rate": 4.653462740434322e-05, "loss": 0.0086, "step": 21840 }, { "grad_norm": 0.17695583403110504, "learning_rate": 4.649339124838689e-05, "loss": 0.0109, "step": 21850 }, { "grad_norm": 0.13103888928890228, "learning_rate": 4.6452157489164574e-05, "loss": 0.0088, "step": 21860 }, { "grad_norm": 0.17191457748413086, "learning_rate": 4.6410926154859155e-05, "loss": 0.0105, "step": 21870 }, { "grad_norm": 0.1975998729467392, "learning_rate": 4.636969727365186e-05, "loss": 0.0088, "step": 21880 }, { "grad_norm": 0.1828552484512329, "learning_rate": 4.632847087372226e-05, "loss": 0.0095, "step": 21890 }, { "grad_norm": 0.18255199491977692, "learning_rate": 4.628724698324818e-05, "loss": 0.0099, "step": 21900 }, { "grad_norm": 0.18992950022220612, "learning_rate": 4.6246025630405795e-05, "loss": 0.0091, "step": 21910 }, { "grad_norm": 0.16371794044971466, "learning_rate": 4.6204806843369474e-05, "loss": 0.0083, "step": 21920 }, { "grad_norm": 0.1320754736661911, "learning_rate": 4.616359065031191e-05, "loss": 0.0086, "step": 21930 }, { "grad_norm": 0.1296696811914444, "learning_rate": 4.6122377079403946e-05, "loss": 0.0096, "step": 21940 }, { "grad_norm": 0.1600361317396164, "learning_rate": 4.6081166158814695e-05, "loss": 0.0075, "step": 21950 }, { "grad_norm": 0.12568065524101257, "learning_rate": 4.603995791671144e-05, "loss": 0.0087, "step": 21960 }, { "grad_norm": 0.13129852712154388, "learning_rate": 4.599875238125957e-05, "loss": 0.0117, "step": 21970 }, { "grad_norm": 0.15781764686107635, "learning_rate": 4.595754958062273e-05, "loss": 0.0106, "step": 21980 }, { "grad_norm": 0.11107010394334793, "learning_rate": 4.591634954296265e-05, "loss": 0.0093, "step": 21990 }, { "grad_norm": 0.1817159652709961, "learning_rate": 4.587515229643913e-05, "loss": 0.0071, "step": 22000 }, { "grad_norm": 0.49801334738731384, "learning_rate": 4.583395786921013e-05, "loss": 0.0094, "step": 22010 }, { "grad_norm": 0.14203199744224548, "learning_rate": 4.579276628943164e-05, "loss": 0.0087, "step": 22020 }, { "grad_norm": 0.161062553524971, "learning_rate": 4.575157758525772e-05, "loss": 0.0088, "step": 22030 }, { "grad_norm": 0.16448312997817993, "learning_rate": 4.571039178484046e-05, "loss": 0.0095, "step": 22040 }, { "grad_norm": 0.15264831483364105, "learning_rate": 4.566920891632998e-05, "loss": 0.0078, "step": 22050 }, { "grad_norm": 0.1655844748020172, "learning_rate": 4.562802900787436e-05, "loss": 0.0082, "step": 22060 }, { "grad_norm": 0.19021935760974884, "learning_rate": 4.558685208761968e-05, "loss": 0.0092, "step": 22070 }, { "grad_norm": 0.15308038890361786, "learning_rate": 4.554567818370998e-05, "loss": 0.008, "step": 22080 }, { "grad_norm": 0.19119708240032196, "learning_rate": 4.550450732428726e-05, "loss": 0.0095, "step": 22090 }, { "grad_norm": 0.1329852044582367, "learning_rate": 4.546333953749137e-05, "loss": 0.0116, "step": 22100 }, { "grad_norm": 0.18684661388397217, "learning_rate": 4.5422174851460154e-05, "loss": 0.0081, "step": 22110 }, { "grad_norm": 0.1996850222349167, "learning_rate": 4.538101329432924e-05, "loss": 0.0092, "step": 22120 }, { "grad_norm": 0.19784417748451233, "learning_rate": 4.5339854894232195e-05, "loss": 0.0119, "step": 22130 }, { "grad_norm": 0.13785500824451447, "learning_rate": 4.52986996793004e-05, "loss": 0.0098, "step": 22140 }, { "grad_norm": 0.15687792003154755, "learning_rate": 4.5257547677663024e-05, "loss": 0.0093, "step": 22150 }, { "grad_norm": 0.16868650913238525, "learning_rate": 4.52163989174471e-05, "loss": 0.0086, "step": 22160 }, { "grad_norm": 0.11781078577041626, "learning_rate": 4.51752534267774e-05, "loss": 0.0095, "step": 22170 }, { "grad_norm": 0.13698574900627136, "learning_rate": 4.513411123377649e-05, "loss": 0.0098, "step": 22180 }, { "grad_norm": 0.16692788898944855, "learning_rate": 4.5092972366564675e-05, "loss": 0.0082, "step": 22190 }, { "grad_norm": 0.23256874084472656, "learning_rate": 4.505183685325997e-05, "loss": 0.0092, "step": 22200 }, { "grad_norm": 0.15656298398971558, "learning_rate": 4.5010704721978125e-05, "loss": 0.0088, "step": 22210 }, { "grad_norm": 0.18951044976711273, "learning_rate": 4.496957600083255e-05, "loss": 0.0119, "step": 22220 }, { "grad_norm": 0.24899084866046906, "learning_rate": 4.4928450717934343e-05, "loss": 0.0094, "step": 22230 }, { "grad_norm": 0.16123567521572113, "learning_rate": 4.488732890139227e-05, "loss": 0.0097, "step": 22240 }, { "grad_norm": 0.15257540345191956, "learning_rate": 4.4846210579312665e-05, "loss": 0.0147, "step": 22250 }, { "grad_norm": 0.20026490092277527, "learning_rate": 4.480509577979953e-05, "loss": 0.0112, "step": 22260 }, { "grad_norm": 0.12938980758190155, "learning_rate": 4.476398453095445e-05, "loss": 0.0089, "step": 22270 }, { "grad_norm": 0.15767347812652588, "learning_rate": 4.472287686087656e-05, "loss": 0.0092, "step": 22280 }, { "grad_norm": 0.14391085505485535, "learning_rate": 4.468177279766259e-05, "loss": 0.009, "step": 22290 }, { "grad_norm": 0.1687774509191513, "learning_rate": 4.4640672369406746e-05, "loss": 0.0086, "step": 22300 }, { "grad_norm": 0.1470276415348053, "learning_rate": 4.459957560420082e-05, "loss": 0.0086, "step": 22310 }, { "grad_norm": 0.18076777458190918, "learning_rate": 4.455848253013403e-05, "loss": 0.0088, "step": 22320 }, { "grad_norm": 0.1564243882894516, "learning_rate": 4.4517393175293146e-05, "loss": 0.0115, "step": 22330 }, { "grad_norm": 0.16907314956188202, "learning_rate": 4.447630756776232e-05, "loss": 0.0088, "step": 22340 }, { "grad_norm": 0.14601297676563263, "learning_rate": 4.443522573562318e-05, "loss": 0.0081, "step": 22350 }, { "grad_norm": 0.15225690603256226, "learning_rate": 4.4394147706954776e-05, "loss": 0.0093, "step": 22360 }, { "grad_norm": 0.16601470112800598, "learning_rate": 4.435307350983355e-05, "loss": 0.0092, "step": 22370 }, { "grad_norm": 0.14531563222408295, "learning_rate": 4.4312003172333326e-05, "loss": 0.0082, "step": 22380 }, { "grad_norm": 0.23597106337547302, "learning_rate": 4.427093672252531e-05, "loss": 0.0107, "step": 22390 }, { "grad_norm": 0.17309577763080597, "learning_rate": 4.422987418847802e-05, "loss": 0.0102, "step": 22400 }, { "grad_norm": 0.16573943197727203, "learning_rate": 4.4188815598257325e-05, "loss": 0.0088, "step": 22410 }, { "grad_norm": 0.1147768497467041, "learning_rate": 4.414776097992638e-05, "loss": 0.0114, "step": 22420 }, { "grad_norm": 0.22173896431922913, "learning_rate": 4.4106710361545595e-05, "loss": 0.0107, "step": 22430 }, { "grad_norm": 0.1916019767522812, "learning_rate": 4.406566377117272e-05, "loss": 0.0112, "step": 22440 }, { "grad_norm": 0.22433191537857056, "learning_rate": 4.40246212368627e-05, "loss": 0.0086, "step": 22450 }, { "grad_norm": 0.1461784690618515, "learning_rate": 4.3983582786667715e-05, "loss": 0.008, "step": 22460 }, { "grad_norm": 0.12175039201974869, "learning_rate": 4.394254844863716e-05, "loss": 0.0085, "step": 22470 }, { "grad_norm": 0.1566801369190216, "learning_rate": 4.390151825081762e-05, "loss": 0.0081, "step": 22480 }, { "grad_norm": 0.1426217406988144, "learning_rate": 4.386049222125286e-05, "loss": 0.0078, "step": 22490 }, { "grad_norm": 0.17295722663402557, "learning_rate": 4.3819470387983774e-05, "loss": 0.0089, "step": 22500 }, { "grad_norm": 0.16864672303199768, "learning_rate": 4.377845277904841e-05, "loss": 0.0086, "step": 22510 }, { "grad_norm": 0.123735211789608, "learning_rate": 4.37374394224819e-05, "loss": 0.0086, "step": 22520 }, { "grad_norm": 0.11421903222799301, "learning_rate": 4.369643034631648e-05, "loss": 0.0094, "step": 22530 }, { "grad_norm": 0.1643579602241516, "learning_rate": 4.365542557858149e-05, "loss": 0.0102, "step": 22540 }, { "grad_norm": 0.13534316420555115, "learning_rate": 4.361442514730329e-05, "loss": 0.0081, "step": 22550 }, { "grad_norm": 0.1571989357471466, "learning_rate": 4.357342908050528e-05, "loss": 0.0091, "step": 22560 }, { "grad_norm": 0.1383044421672821, "learning_rate": 4.3532437406207895e-05, "loss": 0.0085, "step": 22570 }, { "grad_norm": 0.10885154455900192, "learning_rate": 4.349145015242856e-05, "loss": 0.0071, "step": 22580 }, { "grad_norm": 0.13458888232707977, "learning_rate": 4.345046734718168e-05, "loss": 0.0077, "step": 22590 }, { "grad_norm": 0.13075116276741028, "learning_rate": 4.34094890184786e-05, "loss": 0.0095, "step": 22600 }, { "grad_norm": 0.23433074355125427, "learning_rate": 4.336851519432765e-05, "loss": 0.0082, "step": 22610 }, { "grad_norm": 0.175811305642128, "learning_rate": 4.332754590273403e-05, "loss": 0.0073, "step": 22620 }, { "grad_norm": 0.1636539250612259, "learning_rate": 4.3286581171699855e-05, "loss": 0.0076, "step": 22630 }, { "grad_norm": 0.2885567843914032, "learning_rate": 4.324562102922416e-05, "loss": 0.0082, "step": 22640 }, { "grad_norm": 0.16347546875476837, "learning_rate": 4.320466550330278e-05, "loss": 0.0111, "step": 22650 }, { "grad_norm": 0.09006626904010773, "learning_rate": 4.3163714621928466e-05, "loss": 0.0084, "step": 22660 }, { "grad_norm": 0.21672572195529938, "learning_rate": 4.312276841309074e-05, "loss": 0.0133, "step": 22670 }, { "grad_norm": 0.1688414216041565, "learning_rate": 4.3081826904775945e-05, "loss": 0.0092, "step": 22680 }, { "grad_norm": 0.1708773970603943, "learning_rate": 4.3040890124967246e-05, "loss": 0.0095, "step": 22690 }, { "grad_norm": 0.1189764067530632, "learning_rate": 4.2999958101644537e-05, "loss": 0.0093, "step": 22700 }, { "grad_norm": 0.13810311257839203, "learning_rate": 4.2959030862784435e-05, "loss": 0.0094, "step": 22710 }, { "grad_norm": 0.1684558391571045, "learning_rate": 4.291810843636036e-05, "loss": 0.008, "step": 22720 }, { "grad_norm": 0.13123774528503418, "learning_rate": 4.2877190850342375e-05, "loss": 0.008, "step": 22730 }, { "grad_norm": 0.18786272406578064, "learning_rate": 4.2836278132697294e-05, "loss": 0.0088, "step": 22740 }, { "grad_norm": 0.17799048125743866, "learning_rate": 4.279537031138855e-05, "loss": 0.0091, "step": 22750 }, { "grad_norm": 0.11346162855625153, "learning_rate": 4.275446741437625e-05, "loss": 0.0082, "step": 22760 }, { "grad_norm": 0.13824167847633362, "learning_rate": 4.2713569469617176e-05, "loss": 0.0094, "step": 22770 }, { "grad_norm": 0.17265506088733673, "learning_rate": 4.267267650506465e-05, "loss": 0.0088, "step": 22780 }, { "grad_norm": 0.15415988862514496, "learning_rate": 4.263178854866866e-05, "loss": 0.0089, "step": 22790 }, { "grad_norm": 0.1714065670967102, "learning_rate": 4.259090562837571e-05, "loss": 0.0089, "step": 22800 }, { "grad_norm": 0.20310041308403015, "learning_rate": 4.255002777212888e-05, "loss": 0.0114, "step": 22810 }, { "grad_norm": 0.17763139307498932, "learning_rate": 4.250915500786783e-05, "loss": 0.0071, "step": 22820 }, { "grad_norm": 0.19055987894535065, "learning_rate": 4.24682873635287e-05, "loss": 0.01, "step": 22830 }, { "grad_norm": 0.14725296199321747, "learning_rate": 4.242742486704414e-05, "loss": 0.0078, "step": 22840 }, { "grad_norm": 0.21124479174613953, "learning_rate": 4.238656754634327e-05, "loss": 0.0099, "step": 22850 }, { "grad_norm": 0.1386217474937439, "learning_rate": 4.234571542935168e-05, "loss": 0.0091, "step": 22860 }, { "grad_norm": 0.2597929537296295, "learning_rate": 4.230486854399144e-05, "loss": 0.0096, "step": 22870 }, { "grad_norm": 0.12002526968717575, "learning_rate": 4.226402691818098e-05, "loss": 0.0085, "step": 22880 }, { "grad_norm": 0.1700126975774765, "learning_rate": 4.2223190579835196e-05, "loss": 0.0085, "step": 22890 }, { "grad_norm": 0.24111826717853546, "learning_rate": 4.218235955686531e-05, "loss": 0.0107, "step": 22900 }, { "grad_norm": 0.1739354133605957, "learning_rate": 4.214153387717894e-05, "loss": 0.0103, "step": 22910 }, { "grad_norm": 0.17050059139728546, "learning_rate": 4.210071356868007e-05, "loss": 0.0097, "step": 22920 }, { "grad_norm": 0.20963041484355927, "learning_rate": 4.205989865926898e-05, "loss": 0.0085, "step": 22930 }, { "grad_norm": 0.15687301754951477, "learning_rate": 4.2019089176842294e-05, "loss": 0.0094, "step": 22940 }, { "grad_norm": 0.15760250389575958, "learning_rate": 4.1978285149292894e-05, "loss": 0.0102, "step": 22950 }, { "grad_norm": 0.18713267147541046, "learning_rate": 4.193748660450996e-05, "loss": 0.0097, "step": 22960 }, { "grad_norm": 0.20177888870239258, "learning_rate": 4.189669357037891e-05, "loss": 0.0095, "step": 22970 }, { "grad_norm": 0.12646222114562988, "learning_rate": 4.1855906074781405e-05, "loss": 0.0079, "step": 22980 }, { "grad_norm": 0.16672171652317047, "learning_rate": 4.1815124145595285e-05, "loss": 0.0074, "step": 22990 }, { "grad_norm": 0.15966327488422394, "learning_rate": 4.1774347810694644e-05, "loss": 0.0093, "step": 23000 }, { "grad_norm": 0.16173486411571503, "learning_rate": 4.17335770979497e-05, "loss": 0.0083, "step": 23010 }, { "grad_norm": 0.12503378093242645, "learning_rate": 4.169281203522687e-05, "loss": 0.0077, "step": 23020 }, { "grad_norm": 0.18300968408584595, "learning_rate": 4.1652052650388674e-05, "loss": 0.0079, "step": 23030 }, { "grad_norm": 0.15811452269554138, "learning_rate": 4.1611298971293786e-05, "loss": 0.0086, "step": 23040 }, { "grad_norm": 0.20368769764900208, "learning_rate": 4.1570551025796935e-05, "loss": 0.0076, "step": 23050 }, { "grad_norm": 0.22139635682106018, "learning_rate": 4.152980884174897e-05, "loss": 0.01, "step": 23060 }, { "grad_norm": 0.1645168662071228, "learning_rate": 4.148907244699682e-05, "loss": 0.0101, "step": 23070 }, { "grad_norm": 0.20446233451366425, "learning_rate": 4.1448341869383395e-05, "loss": 0.011, "step": 23080 }, { "grad_norm": 0.15781505405902863, "learning_rate": 4.140761713674765e-05, "loss": 0.0086, "step": 23090 }, { "grad_norm": 0.20597721636295319, "learning_rate": 4.1366898276924574e-05, "loss": 0.0099, "step": 23100 }, { "grad_norm": 0.21658384799957275, "learning_rate": 4.132618531774512e-05, "loss": 0.008, "step": 23110 }, { "grad_norm": 0.1538633108139038, "learning_rate": 4.128547828703622e-05, "loss": 0.0075, "step": 23120 }, { "grad_norm": 0.11489703506231308, "learning_rate": 4.1244777212620725e-05, "loss": 0.0085, "step": 23130 }, { "grad_norm": 0.13507422804832458, "learning_rate": 4.120408212231746e-05, "loss": 0.0072, "step": 23140 }, { "grad_norm": 0.1545163094997406, "learning_rate": 4.116339304394111e-05, "loss": 0.0093, "step": 23150 }, { "grad_norm": 0.14658571779727936, "learning_rate": 4.112271000530229e-05, "loss": 0.0086, "step": 23160 }, { "grad_norm": 0.09225703775882721, "learning_rate": 4.10820330342075e-05, "loss": 0.0079, "step": 23170 }, { "grad_norm": 0.1449969857931137, "learning_rate": 4.1041362158459027e-05, "loss": 0.0082, "step": 23180 }, { "grad_norm": 0.16345062851905823, "learning_rate": 4.1000697405855024e-05, "loss": 0.0102, "step": 23190 }, { "grad_norm": 0.21196973323822021, "learning_rate": 4.096003880418951e-05, "loss": 0.0126, "step": 23200 }, { "grad_norm": 0.1534031331539154, "learning_rate": 4.0919386381252215e-05, "loss": 0.0073, "step": 23210 }, { "grad_norm": 0.1506410539150238, "learning_rate": 4.087874016482872e-05, "loss": 0.007, "step": 23220 }, { "grad_norm": 0.19430266320705414, "learning_rate": 4.0838100182700295e-05, "loss": 0.0081, "step": 23230 }, { "grad_norm": 0.14946280419826508, "learning_rate": 4.079746646264402e-05, "loss": 0.0086, "step": 23240 }, { "grad_norm": 0.14922939240932465, "learning_rate": 4.075683903243262e-05, "loss": 0.0062, "step": 23250 }, { "grad_norm": 0.14144590497016907, "learning_rate": 4.071621791983462e-05, "loss": 0.0071, "step": 23260 }, { "grad_norm": 0.17051586508750916, "learning_rate": 4.06756031526141e-05, "loss": 0.0083, "step": 23270 }, { "grad_norm": 0.17764799296855927, "learning_rate": 4.063499475853092e-05, "loss": 0.0096, "step": 23280 }, { "grad_norm": 0.14898580312728882, "learning_rate": 4.0594392765340506e-05, "loss": 0.0085, "step": 23290 }, { "grad_norm": 0.14183489978313446, "learning_rate": 4.0553797200793954e-05, "loss": 0.008, "step": 23300 }, { "grad_norm": 0.181218683719635, "learning_rate": 4.0513208092637926e-05, "loss": 0.0088, "step": 23310 }, { "grad_norm": 0.1509699821472168, "learning_rate": 4.0472625468614735e-05, "loss": 0.0108, "step": 23320 }, { "grad_norm": 0.13012640178203583, "learning_rate": 4.043204935646218e-05, "loss": 0.0088, "step": 23330 }, { "grad_norm": 0.19520790874958038, "learning_rate": 4.0391479783913675e-05, "loss": 0.0078, "step": 23340 }, { "grad_norm": 0.17447778582572937, "learning_rate": 4.0350916778698155e-05, "loss": 0.0095, "step": 23350 }, { "grad_norm": 0.15067771077156067, "learning_rate": 4.031036036854001e-05, "loss": 0.0093, "step": 23360 }, { "grad_norm": 0.12965214252471924, "learning_rate": 4.026981058115918e-05, "loss": 0.0087, "step": 23370 }, { "grad_norm": 0.17848654091358185, "learning_rate": 4.022926744427108e-05, "loss": 0.0116, "step": 23380 }, { "grad_norm": 0.1364479959011078, "learning_rate": 4.018873098558654e-05, "loss": 0.0086, "step": 23390 }, { "grad_norm": 0.1695675551891327, "learning_rate": 4.014820123281186e-05, "loss": 0.0085, "step": 23400 }, { "grad_norm": 0.1503438651561737, "learning_rate": 4.0107678213648735e-05, "loss": 0.0085, "step": 23410 }, { "grad_norm": 0.1487322747707367, "learning_rate": 4.006716195579428e-05, "loss": 0.0108, "step": 23420 }, { "grad_norm": 0.19472025334835052, "learning_rate": 4.002665248694096e-05, "loss": 0.0108, "step": 23430 }, { "grad_norm": 0.18238352239131927, "learning_rate": 3.998614983477664e-05, "loss": 0.0078, "step": 23440 }, { "grad_norm": 0.13501794636249542, "learning_rate": 3.994565402698448e-05, "loss": 0.0094, "step": 23450 }, { "grad_norm": 0.16723477840423584, "learning_rate": 3.9905165091242975e-05, "loss": 0.0094, "step": 23460 }, { "grad_norm": 0.17331324517726898, "learning_rate": 3.9864683055225936e-05, "loss": 0.0089, "step": 23470 }, { "grad_norm": 0.16006670892238617, "learning_rate": 3.982420794660247e-05, "loss": 0.0074, "step": 23480 }, { "grad_norm": 0.19046545028686523, "learning_rate": 3.978373979303691e-05, "loss": 0.0055, "step": 23490 }, { "grad_norm": 0.1879490166902542, "learning_rate": 3.974327862218888e-05, "loss": 0.0083, "step": 23500 }, { "grad_norm": 0.15673476457595825, "learning_rate": 3.970282446171318e-05, "loss": 0.0106, "step": 23510 }, { "grad_norm": 0.2034757435321808, "learning_rate": 3.966237733925988e-05, "loss": 0.01, "step": 23520 }, { "grad_norm": 0.198495015501976, "learning_rate": 3.962193728247418e-05, "loss": 0.0093, "step": 23530 }, { "grad_norm": 0.176111102104187, "learning_rate": 3.958150431899651e-05, "loss": 0.0101, "step": 23540 }, { "grad_norm": 0.16606231033802032, "learning_rate": 3.954107847646238e-05, "loss": 0.008, "step": 23550 }, { "grad_norm": 0.1542590856552124, "learning_rate": 3.950065978250249e-05, "loss": 0.0069, "step": 23560 }, { "grad_norm": 0.15086489915847778, "learning_rate": 3.9460248264742624e-05, "loss": 0.0083, "step": 23570 }, { "grad_norm": 0.18309669196605682, "learning_rate": 3.941984395080371e-05, "loss": 0.0119, "step": 23580 }, { "grad_norm": 0.16028225421905518, "learning_rate": 3.937944686830167e-05, "loss": 0.0083, "step": 23590 }, { "grad_norm": 0.15866830945014954, "learning_rate": 3.933905704484756e-05, "loss": 0.008, "step": 23600 }, { "grad_norm": 0.18073751032352448, "learning_rate": 3.929867450804743e-05, "loss": 0.0072, "step": 23610 }, { "grad_norm": 0.12902824580669403, "learning_rate": 3.925829928550237e-05, "loss": 0.0073, "step": 23620 }, { "grad_norm": 0.16490912437438965, "learning_rate": 3.921793140480847e-05, "loss": 0.0081, "step": 23630 }, { "grad_norm": 0.17436659336090088, "learning_rate": 3.917757089355677e-05, "loss": 0.0072, "step": 23640 }, { "grad_norm": 0.16375045478343964, "learning_rate": 3.9137217779333326e-05, "loss": 0.0073, "step": 23650 }, { "grad_norm": 0.14989304542541504, "learning_rate": 3.9096872089719083e-05, "loss": 0.0073, "step": 23660 }, { "grad_norm": 0.22852040827274323, "learning_rate": 3.905653385228996e-05, "loss": 0.0087, "step": 23670 }, { "grad_norm": 0.1337893158197403, "learning_rate": 3.901620309461677e-05, "loss": 0.0066, "step": 23680 }, { "grad_norm": 0.22191014885902405, "learning_rate": 3.897587984426518e-05, "loss": 0.0075, "step": 23690 }, { "grad_norm": 0.12031374126672745, "learning_rate": 3.893556412879577e-05, "loss": 0.0073, "step": 23700 }, { "grad_norm": 0.16402360796928406, "learning_rate": 3.889525597576395e-05, "loss": 0.0067, "step": 23710 }, { "grad_norm": 0.19532358646392822, "learning_rate": 3.8854955412719965e-05, "loss": 0.0102, "step": 23720 }, { "grad_norm": 0.15397261083126068, "learning_rate": 3.881466246720887e-05, "loss": 0.0077, "step": 23730 }, { "grad_norm": 0.13853824138641357, "learning_rate": 3.8774377166770484e-05, "loss": 0.0095, "step": 23740 }, { "grad_norm": 0.22396612167358398, "learning_rate": 3.8734099538939474e-05, "loss": 0.0075, "step": 23750 }, { "grad_norm": 0.123240627348423, "learning_rate": 3.869382961124518e-05, "loss": 0.009, "step": 23760 }, { "grad_norm": 0.19599762558937073, "learning_rate": 3.8653567411211736e-05, "loss": 0.0077, "step": 23770 }, { "grad_norm": 0.1768662929534912, "learning_rate": 3.8613312966357987e-05, "loss": 0.0084, "step": 23780 }, { "grad_norm": 0.18946142494678497, "learning_rate": 3.857306630419745e-05, "loss": 0.0091, "step": 23790 }, { "grad_norm": 0.12922170758247375, "learning_rate": 3.853282745223834e-05, "loss": 0.0094, "step": 23800 }, { "grad_norm": 0.15243369340896606, "learning_rate": 3.8492596437983546e-05, "loss": 0.0084, "step": 23810 }, { "grad_norm": 0.19902177155017853, "learning_rate": 3.8452373288930586e-05, "loss": 0.011, "step": 23820 }, { "grad_norm": 0.14146891236305237, "learning_rate": 3.841215803257159e-05, "loss": 0.0065, "step": 23830 }, { "grad_norm": 0.1333857774734497, "learning_rate": 3.83719506963933e-05, "loss": 0.0094, "step": 23840 }, { "grad_norm": 0.11339139938354492, "learning_rate": 3.8331751307877087e-05, "loss": 0.0078, "step": 23850 }, { "grad_norm": 0.16688261926174164, "learning_rate": 3.82915598944988e-05, "loss": 0.0073, "step": 23860 }, { "grad_norm": 0.13645026087760925, "learning_rate": 3.825137648372893e-05, "loss": 0.0083, "step": 23870 }, { "grad_norm": 0.14580705761909485, "learning_rate": 3.8211201103032465e-05, "loss": 0.0076, "step": 23880 }, { "grad_norm": 0.11923098564147949, "learning_rate": 3.817103377986887e-05, "loss": 0.0089, "step": 23890 }, { "grad_norm": 0.1394934505224228, "learning_rate": 3.813087454169215e-05, "loss": 0.0085, "step": 23900 }, { "grad_norm": 0.19539019465446472, "learning_rate": 3.809072341595078e-05, "loss": 0.0078, "step": 23910 }, { "grad_norm": 0.11205853521823883, "learning_rate": 3.8050580430087636e-05, "loss": 0.0078, "step": 23920 }, { "grad_norm": 0.14864791929721832, "learning_rate": 3.8010445611540096e-05, "loss": 0.0067, "step": 23930 }, { "grad_norm": 0.19110019505023956, "learning_rate": 3.797031898773992e-05, "loss": 0.0082, "step": 23940 }, { "grad_norm": 0.19897352159023285, "learning_rate": 3.793020058611329e-05, "loss": 0.0089, "step": 23950 }, { "grad_norm": 0.13029567897319794, "learning_rate": 3.789009043408074e-05, "loss": 0.0095, "step": 23960 }, { "grad_norm": 0.14397679269313812, "learning_rate": 3.7849988559057194e-05, "loss": 0.0065, "step": 23970 }, { "grad_norm": 0.1977427899837494, "learning_rate": 3.78098949884519e-05, "loss": 0.0063, "step": 23980 }, { "grad_norm": 0.15137562155723572, "learning_rate": 3.776980974966843e-05, "loss": 0.0081, "step": 23990 }, { "grad_norm": 0.17947152256965637, "learning_rate": 3.772973287010468e-05, "loss": 0.008, "step": 24000 }, { "grad_norm": 0.11673620343208313, "learning_rate": 3.768966437715283e-05, "loss": 0.008, "step": 24010 }, { "grad_norm": 0.1539871096611023, "learning_rate": 3.7649604298199274e-05, "loss": 0.0086, "step": 24020 }, { "grad_norm": 0.16469880938529968, "learning_rate": 3.760955266062473e-05, "loss": 0.0076, "step": 24030 }, { "grad_norm": 0.16704225540161133, "learning_rate": 3.75695094918041e-05, "loss": 0.0084, "step": 24040 }, { "grad_norm": 0.12788213789463043, "learning_rate": 3.752947481910652e-05, "loss": 0.0086, "step": 24050 }, { "grad_norm": 0.16932734847068787, "learning_rate": 3.7489448669895324e-05, "loss": 0.0097, "step": 24060 }, { "grad_norm": 0.13932354748249054, "learning_rate": 3.744943107152798e-05, "loss": 0.0077, "step": 24070 }, { "grad_norm": 0.15991701185703278, "learning_rate": 3.7409422051356165e-05, "loss": 0.0061, "step": 24080 }, { "grad_norm": 0.17441889643669128, "learning_rate": 3.736942163672564e-05, "loss": 0.0094, "step": 24090 }, { "grad_norm": 0.3126179277896881, "learning_rate": 3.732942985497636e-05, "loss": 0.0074, "step": 24100 }, { "grad_norm": 0.13072718679904938, "learning_rate": 3.728944673344228e-05, "loss": 0.0077, "step": 24110 }, { "grad_norm": 0.18476502597332, "learning_rate": 3.72494722994515e-05, "loss": 0.0086, "step": 24120 }, { "grad_norm": 0.14454035460948944, "learning_rate": 3.720950658032617e-05, "loss": 0.008, "step": 24130 }, { "grad_norm": 0.15941934287548065, "learning_rate": 3.716954960338249e-05, "loss": 0.0078, "step": 24140 }, { "grad_norm": 0.07481976598501205, "learning_rate": 3.712960139593066e-05, "loss": 0.0074, "step": 24150 }, { "grad_norm": 0.14394988119602203, "learning_rate": 3.708966198527493e-05, "loss": 0.0085, "step": 24160 }, { "grad_norm": 0.17499324679374695, "learning_rate": 3.704973139871349e-05, "loss": 0.0069, "step": 24170 }, { "grad_norm": 0.1678711324930191, "learning_rate": 3.700980966353853e-05, "loss": 0.0107, "step": 24180 }, { "grad_norm": 0.16327610611915588, "learning_rate": 3.696989680703619e-05, "loss": 0.0097, "step": 24190 }, { "grad_norm": 0.18219530582427979, "learning_rate": 3.69299928564865e-05, "loss": 0.0068, "step": 24200 }, { "grad_norm": 0.15564405918121338, "learning_rate": 3.689009783916345e-05, "loss": 0.0068, "step": 24210 }, { "grad_norm": 0.15930718183517456, "learning_rate": 3.6850211782334895e-05, "loss": 0.0107, "step": 24220 }, { "grad_norm": 0.13595156371593475, "learning_rate": 3.681033471326261e-05, "loss": 0.0097, "step": 24230 }, { "grad_norm": 0.1386430263519287, "learning_rate": 3.677046665920216e-05, "loss": 0.0073, "step": 24240 }, { "grad_norm": 0.18713878095149994, "learning_rate": 3.6730607647403005e-05, "loss": 0.0081, "step": 24250 }, { "grad_norm": 0.13632437586784363, "learning_rate": 3.6690757705108416e-05, "loss": 0.0077, "step": 24260 }, { "grad_norm": 0.15752005577087402, "learning_rate": 3.665091685955542e-05, "loss": 0.0072, "step": 24270 }, { "grad_norm": 0.13475722074508667, "learning_rate": 3.6611085137974896e-05, "loss": 0.008, "step": 24280 }, { "grad_norm": 0.11252006143331528, "learning_rate": 3.657126256759143e-05, "loss": 0.0108, "step": 24290 }, { "grad_norm": 0.11059316247701645, "learning_rate": 3.653144917562335e-05, "loss": 0.0082, "step": 24300 }, { "grad_norm": 0.16120773553848267, "learning_rate": 3.649164498928277e-05, "loss": 0.0085, "step": 24310 }, { "grad_norm": 0.1500127613544464, "learning_rate": 3.645185003577546e-05, "loss": 0.0073, "step": 24320 }, { "grad_norm": 0.12706314027309418, "learning_rate": 3.6412064342300906e-05, "loss": 0.0073, "step": 24330 }, { "grad_norm": 0.1535467505455017, "learning_rate": 3.637228793605224e-05, "loss": 0.0091, "step": 24340 }, { "grad_norm": 0.1635219305753708, "learning_rate": 3.6332520844216264e-05, "loss": 0.0073, "step": 24350 }, { "grad_norm": 0.17854264378547668, "learning_rate": 3.6292763093973425e-05, "loss": 0.0091, "step": 24360 }, { "grad_norm": 0.1641182005405426, "learning_rate": 3.6253014712497754e-05, "loss": 0.0096, "step": 24370 }, { "grad_norm": 0.15096613764762878, "learning_rate": 3.621327572695692e-05, "loss": 0.0076, "step": 24380 }, { "grad_norm": 0.14580675959587097, "learning_rate": 3.617354616451211e-05, "loss": 0.0081, "step": 24390 }, { "grad_norm": 0.10694447159767151, "learning_rate": 3.6133826052318116e-05, "loss": 0.0071, "step": 24400 }, { "grad_norm": 0.20513644814491272, "learning_rate": 3.609411541752327e-05, "loss": 0.0089, "step": 24410 }, { "grad_norm": 0.1731848418712616, "learning_rate": 3.6054414287269405e-05, "loss": 0.0068, "step": 24420 }, { "grad_norm": 0.1438877433538437, "learning_rate": 3.601472268869188e-05, "loss": 0.0087, "step": 24430 }, { "grad_norm": 0.19746215641498566, "learning_rate": 3.597504064891952e-05, "loss": 0.0103, "step": 24440 }, { "grad_norm": 0.16547168791294098, "learning_rate": 3.5935368195074636e-05, "loss": 0.0088, "step": 24450 }, { "grad_norm": 0.1682264506816864, "learning_rate": 3.589570535427297e-05, "loss": 0.0071, "step": 24460 }, { "grad_norm": 0.1809171885251999, "learning_rate": 3.585605215362371e-05, "loss": 0.0076, "step": 24470 }, { "grad_norm": 0.13854192197322845, "learning_rate": 3.581640862022941e-05, "loss": 0.0077, "step": 24480 }, { "grad_norm": 0.21182401478290558, "learning_rate": 3.57767747811861e-05, "loss": 0.0101, "step": 24490 }, { "grad_norm": 0.1638992428779602, "learning_rate": 3.573715066358308e-05, "loss": 0.011, "step": 24500 }, { "grad_norm": 0.196949765086174, "learning_rate": 3.569753629450311e-05, "loss": 0.0081, "step": 24510 }, { "grad_norm": 0.1459006667137146, "learning_rate": 3.565793170102221e-05, "loss": 0.0088, "step": 24520 }, { "grad_norm": 0.10597516596317291, "learning_rate": 3.561833691020976e-05, "loss": 0.0097, "step": 24530 }, { "grad_norm": 0.13819803297519684, "learning_rate": 3.5578751949128415e-05, "loss": 0.0104, "step": 24540 }, { "grad_norm": 0.10244859009981155, "learning_rate": 3.5539176844834125e-05, "loss": 0.009, "step": 24550 }, { "grad_norm": 0.16009418666362762, "learning_rate": 3.5499611624376125e-05, "loss": 0.0079, "step": 24560 }, { "grad_norm": 0.15063585340976715, "learning_rate": 3.546005631479684e-05, "loss": 0.0105, "step": 24570 }, { "grad_norm": 0.1990552693605423, "learning_rate": 3.542051094313196e-05, "loss": 0.0077, "step": 24580 }, { "grad_norm": 0.19878070056438446, "learning_rate": 3.5380975536410364e-05, "loss": 0.0075, "step": 24590 }, { "grad_norm": 0.16253550350666046, "learning_rate": 3.534145012165415e-05, "loss": 0.0093, "step": 24600 }, { "grad_norm": 0.21085651218891144, "learning_rate": 3.5301934725878546e-05, "loss": 0.0083, "step": 24610 }, { "grad_norm": 0.18506868183612823, "learning_rate": 3.526242937609197e-05, "loss": 0.009, "step": 24620 }, { "grad_norm": 0.22256743907928467, "learning_rate": 3.522293409929595e-05, "loss": 0.0072, "step": 24630 }, { "grad_norm": 0.16752125322818756, "learning_rate": 3.518344892248513e-05, "loss": 0.0063, "step": 24640 }, { "grad_norm": 0.1291397511959076, "learning_rate": 3.514397387264725e-05, "loss": 0.0073, "step": 24650 }, { "grad_norm": 0.1438145488500595, "learning_rate": 3.5104508976763176e-05, "loss": 0.0084, "step": 24660 }, { "grad_norm": 0.13368865847587585, "learning_rate": 3.506505426180674e-05, "loss": 0.0084, "step": 24670 }, { "grad_norm": 0.16011470556259155, "learning_rate": 3.502560975474488e-05, "loss": 0.0094, "step": 24680 }, { "grad_norm": 0.13446350395679474, "learning_rate": 3.4986175482537566e-05, "loss": 0.009, "step": 24690 }, { "grad_norm": 0.11768905073404312, "learning_rate": 3.4946751472137725e-05, "loss": 0.0077, "step": 24700 }, { "grad_norm": 0.1580425500869751, "learning_rate": 3.490733775049132e-05, "loss": 0.0071, "step": 24710 }, { "grad_norm": 0.17357319593429565, "learning_rate": 3.4867934344537236e-05, "loss": 0.0101, "step": 24720 }, { "grad_norm": 0.15468107163906097, "learning_rate": 3.482854128120735e-05, "loss": 0.0077, "step": 24730 }, { "grad_norm": 0.1766677349805832, "learning_rate": 3.478915858742643e-05, "loss": 0.0077, "step": 24740 }, { "grad_norm": 0.13128530979156494, "learning_rate": 3.4749786290112205e-05, "loss": 0.0073, "step": 24750 }, { "grad_norm": 0.15641796588897705, "learning_rate": 3.471042441617524e-05, "loss": 0.0097, "step": 24760 }, { "grad_norm": 0.14969705045223236, "learning_rate": 3.467107299251902e-05, "loss": 0.0068, "step": 24770 }, { "grad_norm": 0.4082764983177185, "learning_rate": 3.463173204603984e-05, "loss": 0.007, "step": 24780 }, { "grad_norm": 0.17633521556854248, "learning_rate": 3.4592401603626924e-05, "loss": 0.0087, "step": 24790 }, { "grad_norm": 0.1785987913608551, "learning_rate": 3.45530816921622e-05, "loss": 0.0077, "step": 24800 }, { "grad_norm": 0.16904151439666748, "learning_rate": 3.451377233852051e-05, "loss": 0.0093, "step": 24810 }, { "grad_norm": 0.15357579290866852, "learning_rate": 3.4474473569569385e-05, "loss": 0.0068, "step": 24820 }, { "grad_norm": 0.21429137885570526, "learning_rate": 3.443518541216918e-05, "loss": 0.0077, "step": 24830 }, { "grad_norm": 0.1400027573108673, "learning_rate": 3.439590789317299e-05, "loss": 0.0074, "step": 24840 }, { "grad_norm": 0.15837453305721283, "learning_rate": 3.4356641039426607e-05, "loss": 0.0082, "step": 24850 }, { "grad_norm": 0.08771836757659912, "learning_rate": 3.431738487776857e-05, "loss": 0.0066, "step": 24860 }, { "grad_norm": 0.12160355597734451, "learning_rate": 3.4278139435030084e-05, "loss": 0.0071, "step": 24870 }, { "grad_norm": 0.12005526572465897, "learning_rate": 3.423890473803504e-05, "loss": 0.0081, "step": 24880 }, { "grad_norm": 0.18000318109989166, "learning_rate": 3.41996808136e-05, "loss": 0.006, "step": 24890 }, { "grad_norm": 0.1382383406162262, "learning_rate": 3.416046768853413e-05, "loss": 0.0076, "step": 24900 }, { "grad_norm": 0.11196350306272507, "learning_rate": 3.412126538963925e-05, "loss": 0.0071, "step": 24910 }, { "grad_norm": 0.14402484893798828, "learning_rate": 3.4082073943709727e-05, "loss": 0.01, "step": 24920 }, { "grad_norm": 0.20175747573375702, "learning_rate": 3.404289337753258e-05, "loss": 0.0078, "step": 24930 }, { "grad_norm": 0.18990498781204224, "learning_rate": 3.400372371788736e-05, "loss": 0.0071, "step": 24940 }, { "grad_norm": 0.13470380008220673, "learning_rate": 3.3964564991546124e-05, "loss": 0.0077, "step": 24950 }, { "grad_norm": 0.19008556008338928, "learning_rate": 3.392541722527351e-05, "loss": 0.0083, "step": 24960 }, { "grad_norm": 0.10956967622041702, "learning_rate": 3.3886280445826644e-05, "loss": 0.0066, "step": 24970 }, { "grad_norm": 0.199712336063385, "learning_rate": 3.3847154679955154e-05, "loss": 0.0094, "step": 24980 }, { "grad_norm": 0.10868700593709946, "learning_rate": 3.380803995440113e-05, "loss": 0.0079, "step": 24990 }, { "grad_norm": 0.14568662643432617, "learning_rate": 3.3768936295899115e-05, "loss": 0.0082, "step": 25000 }, { "grad_norm": 0.09372022747993469, "learning_rate": 3.3729843731176094e-05, "loss": 0.0073, "step": 25010 }, { "grad_norm": 0.1791258007287979, "learning_rate": 3.369076228695146e-05, "loss": 0.0076, "step": 25020 }, { "grad_norm": 0.16399726271629333, "learning_rate": 3.365169198993703e-05, "loss": 0.0084, "step": 25030 }, { "grad_norm": 0.15457507967948914, "learning_rate": 3.361263286683697e-05, "loss": 0.0066, "step": 25040 }, { "grad_norm": 0.09638374298810959, "learning_rate": 3.35735849443478e-05, "loss": 0.0087, "step": 25050 }, { "grad_norm": 0.11854623258113861, "learning_rate": 3.3534548249158435e-05, "loss": 0.0075, "step": 25060 }, { "grad_norm": 0.13422705233097076, "learning_rate": 3.3495522807950086e-05, "loss": 0.0097, "step": 25070 }, { "grad_norm": 0.174208864569664, "learning_rate": 3.345650864739627e-05, "loss": 0.0073, "step": 25080 }, { "grad_norm": 0.15896868705749512, "learning_rate": 3.3417505794162794e-05, "loss": 0.0069, "step": 25090 }, { "grad_norm": 0.12968681752681732, "learning_rate": 3.3378514274907745e-05, "loss": 0.0083, "step": 25100 }, { "grad_norm": 0.15773753821849823, "learning_rate": 3.333953411628147e-05, "loss": 0.007, "step": 25110 }, { "grad_norm": 0.18110370635986328, "learning_rate": 3.330056534492653e-05, "loss": 0.0095, "step": 25120 }, { "grad_norm": 0.17556613683700562, "learning_rate": 3.32616079874777e-05, "loss": 0.0078, "step": 25130 }, { "grad_norm": 0.19437731802463531, "learning_rate": 3.322266207056197e-05, "loss": 0.0094, "step": 25140 }, { "grad_norm": 0.17792992293834686, "learning_rate": 3.318372762079852e-05, "loss": 0.0073, "step": 25150 }, { "grad_norm": 0.20708651840686798, "learning_rate": 3.3144804664798666e-05, "loss": 0.0068, "step": 25160 }, { "grad_norm": 0.14624513685703278, "learning_rate": 3.3105893229165894e-05, "loss": 0.0063, "step": 25170 }, { "grad_norm": 0.14108099043369293, "learning_rate": 3.30669933404958e-05, "loss": 0.0084, "step": 25180 }, { "grad_norm": 0.17060768604278564, "learning_rate": 3.302810502537609e-05, "loss": 0.008, "step": 25190 }, { "grad_norm": 0.09576835483312607, "learning_rate": 3.298922831038655e-05, "loss": 0.0056, "step": 25200 }, { "grad_norm": 0.14283387362957, "learning_rate": 3.2950363222099073e-05, "loss": 0.0062, "step": 25210 }, { "grad_norm": 0.17296771705150604, "learning_rate": 3.291150978707758e-05, "loss": 0.0091, "step": 25220 }, { "grad_norm": 0.19543424248695374, "learning_rate": 3.287266803187798e-05, "loss": 0.0085, "step": 25230 }, { "grad_norm": 0.17219136655330658, "learning_rate": 3.283383798304829e-05, "loss": 0.008, "step": 25240 }, { "grad_norm": 0.1310252547264099, "learning_rate": 3.279501966712847e-05, "loss": 0.0063, "step": 25250 }, { "grad_norm": 0.12776915729045868, "learning_rate": 3.275621311065047e-05, "loss": 0.007, "step": 25260 }, { "grad_norm": 0.10843731462955475, "learning_rate": 3.271741834013822e-05, "loss": 0.0078, "step": 25270 }, { "grad_norm": 0.1719367504119873, "learning_rate": 3.267863538210756e-05, "loss": 0.0074, "step": 25280 }, { "grad_norm": 0.13112284243106842, "learning_rate": 3.2639864263066296e-05, "loss": 0.0098, "step": 25290 }, { "grad_norm": 0.1631331741809845, "learning_rate": 3.26011050095141e-05, "loss": 0.0082, "step": 25300 }, { "grad_norm": 0.17984090745449066, "learning_rate": 3.256235764794259e-05, "loss": 0.0093, "step": 25310 }, { "grad_norm": 0.13072384893894196, "learning_rate": 3.2523622204835194e-05, "loss": 0.0082, "step": 25320 }, { "grad_norm": 0.1137094646692276, "learning_rate": 3.2484898706667214e-05, "loss": 0.0071, "step": 25330 }, { "grad_norm": 0.11347147822380066, "learning_rate": 3.2446187179905806e-05, "loss": 0.007, "step": 25340 }, { "grad_norm": 0.09274634718894958, "learning_rate": 3.240748765100995e-05, "loss": 0.0061, "step": 25350 }, { "grad_norm": 0.13624130189418793, "learning_rate": 3.236880014643039e-05, "loss": 0.0065, "step": 25360 }, { "grad_norm": 0.19910277426242828, "learning_rate": 3.233012469260969e-05, "loss": 0.0085, "step": 25370 }, { "grad_norm": 0.16156314313411713, "learning_rate": 3.229146131598213e-05, "loss": 0.0085, "step": 25380 }, { "grad_norm": 0.12920299172401428, "learning_rate": 3.2252810042973794e-05, "loss": 0.0091, "step": 25390 }, { "grad_norm": 0.1784958690404892, "learning_rate": 3.2214170900002456e-05, "loss": 0.0092, "step": 25400 }, { "grad_norm": 0.12197820097208023, "learning_rate": 3.217554391347758e-05, "loss": 0.0109, "step": 25410 }, { "grad_norm": 0.09983208775520325, "learning_rate": 3.213692910980037e-05, "loss": 0.0093, "step": 25420 }, { "grad_norm": 0.16612248122692108, "learning_rate": 3.2098326515363666e-05, "loss": 0.0076, "step": 25430 }, { "grad_norm": 0.1673526167869568, "learning_rate": 3.205973615655199e-05, "loss": 0.0093, "step": 25440 }, { "grad_norm": 0.15440049767494202, "learning_rate": 3.202115805974149e-05, "loss": 0.0067, "step": 25450 }, { "grad_norm": 0.18256409466266632, "learning_rate": 3.1982592251299916e-05, "loss": 0.0096, "step": 25460 }, { "grad_norm": 0.18049393594264984, "learning_rate": 3.1944038757586656e-05, "loss": 0.0118, "step": 25470 }, { "grad_norm": 0.15795087814331055, "learning_rate": 3.190549760495263e-05, "loss": 0.011, "step": 25480 }, { "grad_norm": 0.1518244445323944, "learning_rate": 3.186696881974039e-05, "loss": 0.0087, "step": 25490 }, { "grad_norm": 0.20663301646709442, "learning_rate": 3.1828452428283986e-05, "loss": 0.0078, "step": 25500 }, { "grad_norm": 0.11698968708515167, "learning_rate": 3.178994845690898e-05, "loss": 0.0081, "step": 25510 }, { "grad_norm": 0.1572457104921341, "learning_rate": 3.17514569319325e-05, "loss": 0.0099, "step": 25520 }, { "grad_norm": 0.1492062509059906, "learning_rate": 3.171297787966312e-05, "loss": 0.0096, "step": 25530 }, { "grad_norm": 0.13637743890285492, "learning_rate": 3.167451132640093e-05, "loss": 0.0088, "step": 25540 }, { "grad_norm": 0.1656186282634735, "learning_rate": 3.163605729843746e-05, "loss": 0.0053, "step": 25550 }, { "grad_norm": 0.12074429541826248, "learning_rate": 3.159761582205565e-05, "loss": 0.0066, "step": 25560 }, { "grad_norm": 0.1417388916015625, "learning_rate": 3.155918692352992e-05, "loss": 0.0082, "step": 25570 }, { "grad_norm": 0.17018698155879974, "learning_rate": 3.152077062912602e-05, "loss": 0.0086, "step": 25580 }, { "grad_norm": 0.19251440465450287, "learning_rate": 3.148236696510117e-05, "loss": 0.0095, "step": 25590 }, { "grad_norm": 0.14897240698337555, "learning_rate": 3.144397595770388e-05, "loss": 0.0094, "step": 25600 }, { "grad_norm": 0.10426434129476547, "learning_rate": 3.1405597633174036e-05, "loss": 0.0068, "step": 25610 }, { "grad_norm": 0.13515706360340118, "learning_rate": 3.136723201774289e-05, "loss": 0.0077, "step": 25620 }, { "grad_norm": 0.14807753264904022, "learning_rate": 3.132887913763295e-05, "loss": 0.0081, "step": 25630 }, { "grad_norm": 0.22371618449687958, "learning_rate": 3.129053901905806e-05, "loss": 0.0064, "step": 25640 }, { "grad_norm": 0.13449916243553162, "learning_rate": 3.125221168822335e-05, "loss": 0.0073, "step": 25650 }, { "grad_norm": 0.18245762586593628, "learning_rate": 3.1213897171325154e-05, "loss": 0.007, "step": 25660 }, { "grad_norm": 0.13983219861984253, "learning_rate": 3.1175595494551116e-05, "loss": 0.0112, "step": 25670 }, { "grad_norm": 0.13128964602947235, "learning_rate": 3.1137306684080045e-05, "loss": 0.0083, "step": 25680 }, { "grad_norm": 0.14939244091510773, "learning_rate": 3.1099030766081985e-05, "loss": 0.0053, "step": 25690 }, { "grad_norm": 0.18631906807422638, "learning_rate": 3.106076776671818e-05, "loss": 0.0089, "step": 25700 }, { "grad_norm": 0.19762320816516876, "learning_rate": 3.102251771214101e-05, "loss": 0.0089, "step": 25710 }, { "grad_norm": 0.12866324186325073, "learning_rate": 3.098428062849404e-05, "loss": 0.006, "step": 25720 }, { "grad_norm": 0.13313360512256622, "learning_rate": 3.094605654191195e-05, "loss": 0.0083, "step": 25730 }, { "grad_norm": 0.16001516580581665, "learning_rate": 3.090784547852055e-05, "loss": 0.0089, "step": 25740 }, { "grad_norm": 0.16606426239013672, "learning_rate": 3.0869647464436746e-05, "loss": 0.0081, "step": 25750 }, { "grad_norm": 0.22317761182785034, "learning_rate": 3.0831462525768496e-05, "loss": 0.0083, "step": 25760 }, { "grad_norm": 0.1071072444319725, "learning_rate": 3.079329068861488e-05, "loss": 0.007, "step": 25770 }, { "grad_norm": 0.16590730845928192, "learning_rate": 3.075513197906597e-05, "loss": 0.0074, "step": 25780 }, { "grad_norm": 0.17802214622497559, "learning_rate": 3.071698642320286e-05, "loss": 0.0077, "step": 25790 }, { "grad_norm": 0.12919239699840546, "learning_rate": 3.067885404709772e-05, "loss": 0.0064, "step": 25800 }, { "grad_norm": 0.13973009586334229, "learning_rate": 3.0640734876813636e-05, "loss": 0.007, "step": 25810 }, { "grad_norm": 0.1617988646030426, "learning_rate": 3.060262893840473e-05, "loss": 0.008, "step": 25820 }, { "grad_norm": 0.1415114551782608, "learning_rate": 3.056453625791603e-05, "loss": 0.0093, "step": 25830 }, { "grad_norm": 0.1093520075082779, "learning_rate": 3.052645686138353e-05, "loss": 0.0078, "step": 25840 }, { "grad_norm": 0.09051970392465591, "learning_rate": 3.0488390774834153e-05, "loss": 0.0078, "step": 25850 }, { "grad_norm": 0.15875619649887085, "learning_rate": 3.0450338024285684e-05, "loss": 0.0107, "step": 25860 }, { "grad_norm": 0.10913573950529099, "learning_rate": 3.0412298635746855e-05, "loss": 0.006, "step": 25870 }, { "grad_norm": 0.14418752491474152, "learning_rate": 3.03742726352172e-05, "loss": 0.0101, "step": 25880 }, { "grad_norm": 0.16836325824260712, "learning_rate": 3.0336260048687125e-05, "loss": 0.0081, "step": 25890 }, { "grad_norm": 0.33571940660476685, "learning_rate": 3.0298260902137897e-05, "loss": 0.0072, "step": 25900 }, { "grad_norm": 0.20975206792354584, "learning_rate": 3.0260275221541566e-05, "loss": 0.0088, "step": 25910 }, { "grad_norm": 0.18612483143806458, "learning_rate": 3.0222303032860987e-05, "loss": 0.0093, "step": 25920 }, { "grad_norm": 0.08571882545948029, "learning_rate": 3.018434436204979e-05, "loss": 0.006, "step": 25930 }, { "grad_norm": 0.13446979224681854, "learning_rate": 3.014639923505237e-05, "loss": 0.0059, "step": 25940 }, { "grad_norm": 0.1486126333475113, "learning_rate": 3.0108467677803863e-05, "loss": 0.0066, "step": 25950 }, { "grad_norm": 0.10150545090436935, "learning_rate": 3.0070549716230156e-05, "loss": 0.006, "step": 25960 }, { "grad_norm": 0.1589099019765854, "learning_rate": 3.003264537624777e-05, "loss": 0.0053, "step": 25970 }, { "grad_norm": 0.15557131171226501, "learning_rate": 2.9994754683764e-05, "loss": 0.0065, "step": 25980 }, { "grad_norm": 0.14369945228099823, "learning_rate": 2.9956877664676754e-05, "loss": 0.0067, "step": 25990 }, { "grad_norm": 0.16825537383556366, "learning_rate": 2.9919014344874636e-05, "loss": 0.0066, "step": 26000 }, { "grad_norm": 0.22749370336532593, "learning_rate": 2.9881164750236857e-05, "loss": 0.0109, "step": 26010 }, { "grad_norm": 0.16315551102161407, "learning_rate": 2.984332890663326e-05, "loss": 0.0061, "step": 26020 }, { "grad_norm": 0.1434134989976883, "learning_rate": 2.9805506839924292e-05, "loss": 0.0091, "step": 26030 }, { "grad_norm": 0.14315757155418396, "learning_rate": 2.9767698575960968e-05, "loss": 0.0063, "step": 26040 }, { "grad_norm": 0.1448092758655548, "learning_rate": 2.9729904140584913e-05, "loss": 0.006, "step": 26050 }, { "grad_norm": 0.15720626711845398, "learning_rate": 2.9692123559628234e-05, "loss": 0.0071, "step": 26060 }, { "grad_norm": 0.1447758674621582, "learning_rate": 2.9654356858913596e-05, "loss": 0.008, "step": 26070 }, { "grad_norm": 0.14103275537490845, "learning_rate": 2.9616604064254206e-05, "loss": 0.0081, "step": 26080 }, { "grad_norm": 0.16991212964057922, "learning_rate": 2.9578865201453732e-05, "loss": 0.0083, "step": 26090 }, { "grad_norm": 0.14139512181282043, "learning_rate": 2.9541140296306335e-05, "loss": 0.0075, "step": 26100 }, { "grad_norm": 0.12161282449960709, "learning_rate": 2.9503429374596627e-05, "loss": 0.0062, "step": 26110 }, { "grad_norm": 0.13157734274864197, "learning_rate": 2.946573246209967e-05, "loss": 0.0063, "step": 26120 }, { "grad_norm": 0.2168293595314026, "learning_rate": 2.942804958458094e-05, "loss": 0.0084, "step": 26130 }, { "grad_norm": 0.11482065171003342, "learning_rate": 2.9390380767796343e-05, "loss": 0.0068, "step": 26140 }, { "grad_norm": 0.15135586261749268, "learning_rate": 2.9352726037492174e-05, "loss": 0.0085, "step": 26150 }, { "grad_norm": 0.10693573951721191, "learning_rate": 2.9315085419405052e-05, "loss": 0.0068, "step": 26160 }, { "grad_norm": 0.1405724138021469, "learning_rate": 2.927745893926199e-05, "loss": 0.0087, "step": 26170 }, { "grad_norm": 0.12093085050582886, "learning_rate": 2.9239846622780358e-05, "loss": 0.0055, "step": 26180 }, { "grad_norm": 0.1572103649377823, "learning_rate": 2.9202248495667788e-05, "loss": 0.006, "step": 26190 }, { "grad_norm": 0.1567637175321579, "learning_rate": 2.916466458362227e-05, "loss": 0.0083, "step": 26200 }, { "grad_norm": 0.18340471386909485, "learning_rate": 2.9127094912332033e-05, "loss": 0.0084, "step": 26210 }, { "grad_norm": 0.1746019423007965, "learning_rate": 2.9089539507475606e-05, "loss": 0.0071, "step": 26220 }, { "grad_norm": 0.16922958195209503, "learning_rate": 2.9051998394721748e-05, "loss": 0.0093, "step": 26230 }, { "grad_norm": 0.1289130300283432, "learning_rate": 2.901447159972948e-05, "loss": 0.0064, "step": 26240 }, { "grad_norm": 0.18304474651813507, "learning_rate": 2.8976959148148e-05, "loss": 0.0056, "step": 26250 }, { "grad_norm": 0.09473846107721329, "learning_rate": 2.8939461065616674e-05, "loss": 0.0076, "step": 26260 }, { "grad_norm": 0.198639914393425, "learning_rate": 2.8901977377765127e-05, "loss": 0.0068, "step": 26270 }, { "grad_norm": 0.08833685517311096, "learning_rate": 2.8864508110213094e-05, "loss": 0.0049, "step": 26280 }, { "grad_norm": 0.12744861841201782, "learning_rate": 2.8827053288570503e-05, "loss": 0.0058, "step": 26290 }, { "grad_norm": 0.10931090265512466, "learning_rate": 2.8789612938437315e-05, "loss": 0.0077, "step": 26300 }, { "grad_norm": 0.17164646089076996, "learning_rate": 2.8752187085403683e-05, "loss": 0.0067, "step": 26310 }, { "grad_norm": 0.1402716189622879, "learning_rate": 2.8714775755049818e-05, "loss": 0.008, "step": 26320 }, { "grad_norm": 0.13458628952503204, "learning_rate": 2.867737897294604e-05, "loss": 0.0089, "step": 26330 }, { "grad_norm": 0.14755605161190033, "learning_rate": 2.8639996764652653e-05, "loss": 0.0081, "step": 26340 }, { "grad_norm": 0.14871661365032196, "learning_rate": 2.8602629155720084e-05, "loss": 0.0101, "step": 26350 }, { "grad_norm": 0.1597025841474533, "learning_rate": 2.8565276171688703e-05, "loss": 0.0066, "step": 26360 }, { "grad_norm": 0.11614828556776047, "learning_rate": 2.8527937838088943e-05, "loss": 0.0063, "step": 26370 }, { "grad_norm": 0.18733203411102295, "learning_rate": 2.84906141804412e-05, "loss": 0.0079, "step": 26380 }, { "grad_norm": 0.14878809452056885, "learning_rate": 2.8453305224255867e-05, "loss": 0.0082, "step": 26390 }, { "grad_norm": 0.09468419849872589, "learning_rate": 2.8416010995033216e-05, "loss": 0.0065, "step": 26400 }, { "grad_norm": 0.17668770253658295, "learning_rate": 2.8378731518263524e-05, "loss": 0.0072, "step": 26410 }, { "grad_norm": 0.11528423428535461, "learning_rate": 2.834146681942696e-05, "loss": 0.0075, "step": 26420 }, { "grad_norm": 0.1446715146303177, "learning_rate": 2.8304216923993622e-05, "loss": 0.007, "step": 26430 }, { "grad_norm": 0.13740858435630798, "learning_rate": 2.8266981857423413e-05, "loss": 0.0089, "step": 26440 }, { "grad_norm": 0.14379021525382996, "learning_rate": 2.8229761645166197e-05, "loss": 0.007, "step": 26450 }, { "grad_norm": 0.19750069081783295, "learning_rate": 2.81925563126616e-05, "loss": 0.0084, "step": 26460 }, { "grad_norm": 0.16917815804481506, "learning_rate": 2.8155365885339124e-05, "loss": 0.007, "step": 26470 }, { "grad_norm": 0.168304443359375, "learning_rate": 2.8118190388618093e-05, "loss": 0.0076, "step": 26480 }, { "grad_norm": 0.13830360770225525, "learning_rate": 2.8081029847907607e-05, "loss": 0.0086, "step": 26490 }, { "grad_norm": 0.14247706532478333, "learning_rate": 2.8043884288606525e-05, "loss": 0.0066, "step": 26500 }, { "grad_norm": 0.13556070625782013, "learning_rate": 2.8006753736103496e-05, "loss": 0.0076, "step": 26510 }, { "grad_norm": 0.11990997195243835, "learning_rate": 2.7969638215776918e-05, "loss": 0.0071, "step": 26520 }, { "grad_norm": 0.15720023214817047, "learning_rate": 2.793253775299487e-05, "loss": 0.0073, "step": 26530 }, { "grad_norm": 0.19641806185245514, "learning_rate": 2.7895452373115184e-05, "loss": 0.0098, "step": 26540 }, { "grad_norm": 0.17398390173912048, "learning_rate": 2.785838210148539e-05, "loss": 0.0076, "step": 26550 }, { "grad_norm": 0.15320712327957153, "learning_rate": 2.782132696344263e-05, "loss": 0.0076, "step": 26560 }, { "grad_norm": 0.12763258814811707, "learning_rate": 2.7784286984313745e-05, "loss": 0.0088, "step": 26570 }, { "grad_norm": 0.137028768658638, "learning_rate": 2.7747262189415236e-05, "loss": 0.0091, "step": 26580 }, { "grad_norm": 0.1461276262998581, "learning_rate": 2.7710252604053205e-05, "loss": 0.0066, "step": 26590 }, { "grad_norm": 0.09235314279794693, "learning_rate": 2.767325825352332e-05, "loss": 0.0063, "step": 26600 }, { "grad_norm": 0.11778637766838074, "learning_rate": 2.7636279163110913e-05, "loss": 0.0098, "step": 26610 }, { "grad_norm": 0.11217886209487915, "learning_rate": 2.7599315358090795e-05, "loss": 0.0095, "step": 26620 }, { "grad_norm": 0.18453432619571686, "learning_rate": 2.7562366863727407e-05, "loss": 0.0072, "step": 26630 }, { "grad_norm": 0.10458539426326752, "learning_rate": 2.7525433705274695e-05, "loss": 0.0068, "step": 26640 }, { "grad_norm": 0.11669723689556122, "learning_rate": 2.748851590797614e-05, "loss": 0.0059, "step": 26650 }, { "grad_norm": 0.128043532371521, "learning_rate": 2.7451613497064675e-05, "loss": 0.0078, "step": 26660 }, { "grad_norm": 0.1572285294532776, "learning_rate": 2.7414726497762765e-05, "loss": 0.0074, "step": 26670 }, { "grad_norm": 0.12604421377182007, "learning_rate": 2.737785493528232e-05, "loss": 0.0062, "step": 26680 }, { "grad_norm": 0.14792181551456451, "learning_rate": 2.7340998834824745e-05, "loss": 0.0062, "step": 26690 }, { "grad_norm": 0.19823665916919708, "learning_rate": 2.7304158221580777e-05, "loss": 0.0068, "step": 26700 }, { "grad_norm": 0.15777398645877838, "learning_rate": 2.7267333120730675e-05, "loss": 0.0085, "step": 26710 }, { "grad_norm": 0.2026006430387497, "learning_rate": 2.7230523557444017e-05, "loss": 0.0068, "step": 26720 }, { "grad_norm": 0.1327875256538391, "learning_rate": 2.7193729556879798e-05, "loss": 0.0063, "step": 26730 }, { "grad_norm": 0.1439688354730606, "learning_rate": 2.715695114418637e-05, "loss": 0.0049, "step": 26740 }, { "grad_norm": 0.15009596943855286, "learning_rate": 2.7120188344501475e-05, "loss": 0.0061, "step": 26750 }, { "grad_norm": 0.1012357547879219, "learning_rate": 2.7083441182952067e-05, "loss": 0.0074, "step": 26760 }, { "grad_norm": 0.16789594292640686, "learning_rate": 2.7046709684654527e-05, "loss": 0.0059, "step": 26770 }, { "grad_norm": 0.15876422822475433, "learning_rate": 2.700999387471448e-05, "loss": 0.0091, "step": 26780 }, { "grad_norm": 0.08962107449769974, "learning_rate": 2.6973293778226854e-05, "loss": 0.0046, "step": 26790 }, { "grad_norm": 0.14767222106456757, "learning_rate": 2.6936609420275804e-05, "loss": 0.007, "step": 26800 }, { "grad_norm": 0.1050594300031662, "learning_rate": 2.689994082593472e-05, "loss": 0.0069, "step": 26810 }, { "grad_norm": 0.1225721538066864, "learning_rate": 2.6863288020266264e-05, "loss": 0.0071, "step": 26820 }, { "grad_norm": 0.29070985317230225, "learning_rate": 2.682665102832228e-05, "loss": 0.008, "step": 26830 }, { "grad_norm": 0.14667218923568726, "learning_rate": 2.67900298751438e-05, "loss": 0.0066, "step": 26840 }, { "grad_norm": 0.104786217212677, "learning_rate": 2.6753424585761067e-05, "loss": 0.0065, "step": 26850 }, { "grad_norm": 0.15320415794849396, "learning_rate": 2.671683518519341e-05, "loss": 0.0064, "step": 26860 }, { "grad_norm": 0.20236627757549286, "learning_rate": 2.668026169844936e-05, "loss": 0.0083, "step": 26870 }, { "grad_norm": 0.18686991930007935, "learning_rate": 2.6643704150526538e-05, "loss": 0.0075, "step": 26880 }, { "grad_norm": 0.1996002495288849, "learning_rate": 2.6607162566411716e-05, "loss": 0.0095, "step": 26890 }, { "grad_norm": 0.11067535728216171, "learning_rate": 2.6570636971080697e-05, "loss": 0.0072, "step": 26900 }, { "grad_norm": 0.16065098345279694, "learning_rate": 2.6534127389498364e-05, "loss": 0.0068, "step": 26910 }, { "grad_norm": 0.1430581957101822, "learning_rate": 2.6497633846618696e-05, "loss": 0.0067, "step": 26920 }, { "grad_norm": 0.20466536283493042, "learning_rate": 2.6461156367384677e-05, "loss": 0.0067, "step": 26930 }, { "grad_norm": 0.12641486525535583, "learning_rate": 2.6424694976728316e-05, "loss": 0.0071, "step": 26940 }, { "grad_norm": 0.1652490496635437, "learning_rate": 2.6388249699570667e-05, "loss": 0.0099, "step": 26950 }, { "grad_norm": 0.14111272990703583, "learning_rate": 2.6351820560821672e-05, "loss": 0.0077, "step": 26960 }, { "grad_norm": 0.14761875569820404, "learning_rate": 2.631540758538034e-05, "loss": 0.0066, "step": 26970 }, { "grad_norm": 0.1773013174533844, "learning_rate": 2.6279010798134597e-05, "loss": 0.0063, "step": 26980 }, { "grad_norm": 0.15327906608581543, "learning_rate": 2.6242630223961305e-05, "loss": 0.0075, "step": 26990 }, { "grad_norm": 0.15380048751831055, "learning_rate": 2.6206265887726244e-05, "loss": 0.0064, "step": 27000 }, { "grad_norm": 0.13561950623989105, "learning_rate": 2.6169917814284066e-05, "loss": 0.0071, "step": 27010 }, { "grad_norm": 0.1199551671743393, "learning_rate": 2.6133586028478364e-05, "loss": 0.0063, "step": 27020 }, { "grad_norm": 0.13037465512752533, "learning_rate": 2.609727055514155e-05, "loss": 0.0067, "step": 27030 }, { "grad_norm": 0.14359724521636963, "learning_rate": 2.606097141909494e-05, "loss": 0.0059, "step": 27040 }, { "grad_norm": 0.12354855239391327, "learning_rate": 2.6024688645148644e-05, "loss": 0.006, "step": 27050 }, { "grad_norm": 0.14373862743377686, "learning_rate": 2.5988422258101564e-05, "loss": 0.0083, "step": 27060 }, { "grad_norm": 0.11047320067882538, "learning_rate": 2.5952172282741453e-05, "loss": 0.0072, "step": 27070 }, { "grad_norm": 0.1683477759361267, "learning_rate": 2.5915938743844853e-05, "loss": 0.008, "step": 27080 }, { "grad_norm": 0.08568648248910904, "learning_rate": 2.5879721666177003e-05, "loss": 0.0054, "step": 27090 }, { "grad_norm": 0.12492317706346512, "learning_rate": 2.5843521074491972e-05, "loss": 0.0061, "step": 27100 }, { "grad_norm": 0.1174103394150734, "learning_rate": 2.5807336993532487e-05, "loss": 0.0055, "step": 27110 }, { "grad_norm": 0.13149665296077728, "learning_rate": 2.577116944803004e-05, "loss": 0.0072, "step": 27120 }, { "grad_norm": 0.181332066655159, "learning_rate": 2.5735018462704818e-05, "loss": 0.0064, "step": 27130 }, { "grad_norm": 0.07384335994720459, "learning_rate": 2.5698884062265665e-05, "loss": 0.0059, "step": 27140 }, { "grad_norm": 0.18804487586021423, "learning_rate": 2.5662766271410134e-05, "loss": 0.0078, "step": 27150 }, { "grad_norm": 0.15050609409809113, "learning_rate": 2.5626665114824343e-05, "loss": 0.008, "step": 27160 }, { "grad_norm": 0.18112565577030182, "learning_rate": 2.5590580617183148e-05, "loss": 0.0089, "step": 27170 }, { "grad_norm": 0.11045557260513306, "learning_rate": 2.5554512803149912e-05, "loss": 0.0061, "step": 27180 }, { "grad_norm": 0.1922568529844284, "learning_rate": 2.5518461697376662e-05, "loss": 0.0086, "step": 27190 }, { "grad_norm": 0.18531367182731628, "learning_rate": 2.548242732450402e-05, "loss": 0.0063, "step": 27200 }, { "grad_norm": 0.1107790395617485, "learning_rate": 2.5446409709161095e-05, "loss": 0.0067, "step": 27210 }, { "grad_norm": 0.11976032704114914, "learning_rate": 2.541040887596561e-05, "loss": 0.0062, "step": 27220 }, { "grad_norm": 0.11133287101984024, "learning_rate": 2.537442484952378e-05, "loss": 0.0058, "step": 27230 }, { "grad_norm": 0.19558189809322357, "learning_rate": 2.533845765443037e-05, "loss": 0.0061, "step": 27240 }, { "grad_norm": 0.136002779006958, "learning_rate": 2.530250731526863e-05, "loss": 0.0073, "step": 27250 }, { "grad_norm": 0.14057812094688416, "learning_rate": 2.5266573856610253e-05, "loss": 0.007, "step": 27260 }, { "grad_norm": 0.1159990131855011, "learning_rate": 2.5230657303015403e-05, "loss": 0.0063, "step": 27270 }, { "grad_norm": 0.09135857969522476, "learning_rate": 2.5194757679032728e-05, "loss": 0.0063, "step": 27280 }, { "grad_norm": 0.10946536809206009, "learning_rate": 2.5158875009199278e-05, "loss": 0.0072, "step": 27290 }, { "grad_norm": 0.1442740559577942, "learning_rate": 2.5123009318040537e-05, "loss": 0.0087, "step": 27300 }, { "grad_norm": 0.1207728162407875, "learning_rate": 2.508716063007034e-05, "loss": 0.0075, "step": 27310 }, { "grad_norm": 0.16442663967609406, "learning_rate": 2.5051328969790934e-05, "loss": 0.0068, "step": 27320 }, { "grad_norm": 0.11674011498689651, "learning_rate": 2.501551436169292e-05, "loss": 0.0054, "step": 27330 }, { "grad_norm": 0.09351354837417603, "learning_rate": 2.4979716830255255e-05, "loss": 0.006, "step": 27340 }, { "grad_norm": 0.10488874465227127, "learning_rate": 2.4943936399945233e-05, "loss": 0.0056, "step": 27350 }, { "grad_norm": 0.10465189814567566, "learning_rate": 2.4908173095218412e-05, "loss": 0.0054, "step": 27360 }, { "grad_norm": 0.13993042707443237, "learning_rate": 2.4872426940518663e-05, "loss": 0.0078, "step": 27370 }, { "grad_norm": 0.15641196072101593, "learning_rate": 2.4836697960278156e-05, "loss": 0.0071, "step": 27380 }, { "grad_norm": 0.11871989816427231, "learning_rate": 2.480098617891732e-05, "loss": 0.0055, "step": 27390 }, { "grad_norm": 0.14676538109779358, "learning_rate": 2.4765291620844837e-05, "loss": 0.0061, "step": 27400 }, { "grad_norm": 0.17533309757709503, "learning_rate": 2.472961431045756e-05, "loss": 0.0064, "step": 27410 }, { "grad_norm": 0.17883041501045227, "learning_rate": 2.4693954272140622e-05, "loss": 0.0071, "step": 27420 }, { "grad_norm": 0.14305458962917328, "learning_rate": 2.4658311530267315e-05, "loss": 0.0082, "step": 27430 }, { "grad_norm": 0.16444824635982513, "learning_rate": 2.4622686109199124e-05, "loss": 0.0068, "step": 27440 }, { "grad_norm": 0.15223005414009094, "learning_rate": 2.4587078033285695e-05, "loss": 0.0092, "step": 27450 }, { "grad_norm": 0.13009341061115265, "learning_rate": 2.45514873268648e-05, "loss": 0.0064, "step": 27460 }, { "grad_norm": 0.12910203635692596, "learning_rate": 2.4515914014262336e-05, "loss": 0.0079, "step": 27470 }, { "grad_norm": 0.14144155383110046, "learning_rate": 2.4480358119792345e-05, "loss": 0.0068, "step": 27480 }, { "grad_norm": 0.12789224088191986, "learning_rate": 2.4444819667756942e-05, "loss": 0.0055, "step": 27490 }, { "grad_norm": 0.08218580484390259, "learning_rate": 2.4409298682446346e-05, "loss": 0.0067, "step": 27500 }, { "grad_norm": 0.11858776956796646, "learning_rate": 2.437379518813877e-05, "loss": 0.0055, "step": 27510 }, { "grad_norm": 0.14291489124298096, "learning_rate": 2.4338309209100547e-05, "loss": 0.0063, "step": 27520 }, { "grad_norm": 0.13565152883529663, "learning_rate": 2.4302840769586004e-05, "loss": 0.0051, "step": 27530 }, { "grad_norm": 0.19016306102275848, "learning_rate": 2.42673898938375e-05, "loss": 0.0075, "step": 27540 }, { "grad_norm": 0.15721531212329865, "learning_rate": 2.4231956606085343e-05, "loss": 0.0063, "step": 27550 }, { "grad_norm": 0.12558703124523163, "learning_rate": 2.419654093054789e-05, "loss": 0.0057, "step": 27560 }, { "grad_norm": 0.14139050245285034, "learning_rate": 2.4161142891431375e-05, "loss": 0.0075, "step": 27570 }, { "grad_norm": 0.17669184505939484, "learning_rate": 2.412576251293005e-05, "loss": 0.0067, "step": 27580 }, { "grad_norm": 0.13525575399398804, "learning_rate": 2.4090399819226068e-05, "loss": 0.0084, "step": 27590 }, { "grad_norm": 0.12340106815099716, "learning_rate": 2.4055054834489514e-05, "loss": 0.0067, "step": 27600 }, { "grad_norm": 0.13419505953788757, "learning_rate": 2.401972758287832e-05, "loss": 0.0083, "step": 27610 }, { "grad_norm": 0.15593582391738892, "learning_rate": 2.398441808853834e-05, "loss": 0.0098, "step": 27620 }, { "grad_norm": 0.12948988378047943, "learning_rate": 2.3949126375603288e-05, "loss": 0.0081, "step": 27630 }, { "grad_norm": 0.16512109339237213, "learning_rate": 2.3913852468194724e-05, "loss": 0.0063, "step": 27640 }, { "grad_norm": 0.09248364716768265, "learning_rate": 2.387859639042201e-05, "loss": 0.0067, "step": 27650 }, { "grad_norm": 0.12751780450344086, "learning_rate": 2.3843358166382368e-05, "loss": 0.0058, "step": 27660 }, { "grad_norm": 0.1387290209531784, "learning_rate": 2.3808137820160757e-05, "loss": 0.0052, "step": 27670 }, { "grad_norm": 0.19807712733745575, "learning_rate": 2.3772935375829975e-05, "loss": 0.0053, "step": 27680 }, { "grad_norm": 0.1611603945493698, "learning_rate": 2.3737750857450553e-05, "loss": 0.0065, "step": 27690 }, { "grad_norm": 0.1616864800453186, "learning_rate": 2.3702584289070805e-05, "loss": 0.0071, "step": 27700 }, { "grad_norm": 0.17608794569969177, "learning_rate": 2.36674356947267e-05, "loss": 0.0061, "step": 27710 }, { "grad_norm": 0.09717947244644165, "learning_rate": 2.3632305098442004e-05, "loss": 0.0049, "step": 27720 }, { "grad_norm": 0.14569348096847534, "learning_rate": 2.3597192524228156e-05, "loss": 0.0055, "step": 27730 }, { "grad_norm": 0.12210720777511597, "learning_rate": 2.356209799608424e-05, "loss": 0.0064, "step": 27740 }, { "grad_norm": 0.1403607726097107, "learning_rate": 2.352702153799704e-05, "loss": 0.0074, "step": 27750 }, { "grad_norm": 0.11857055127620697, "learning_rate": 2.3491963173941018e-05, "loss": 0.0069, "step": 27760 }, { "grad_norm": 0.1309659630060196, "learning_rate": 2.3456922927878196e-05, "loss": 0.0074, "step": 27770 }, { "grad_norm": 0.1624280959367752, "learning_rate": 2.3421900823758257e-05, "loss": 0.0052, "step": 27780 }, { "grad_norm": 0.09850896149873734, "learning_rate": 2.3386896885518496e-05, "loss": 0.0068, "step": 27790 }, { "grad_norm": 0.1740904450416565, "learning_rate": 2.335191113708378e-05, "loss": 0.0068, "step": 27800 }, { "grad_norm": 0.14196395874023438, "learning_rate": 2.331694360236651e-05, "loss": 0.0067, "step": 27810 }, { "grad_norm": 0.1405695080757141, "learning_rate": 2.3281994305266702e-05, "loss": 0.0078, "step": 27820 }, { "grad_norm": 0.10827218741178513, "learning_rate": 2.3247063269671826e-05, "loss": 0.0057, "step": 27830 }, { "grad_norm": 0.1490347981452942, "learning_rate": 2.321215051945695e-05, "loss": 0.006, "step": 27840 }, { "grad_norm": 0.16826367378234863, "learning_rate": 2.3177256078484588e-05, "loss": 0.0062, "step": 27850 }, { "grad_norm": 0.12869463860988617, "learning_rate": 2.3142379970604798e-05, "loss": 0.0067, "step": 27860 }, { "grad_norm": 0.18721209466457367, "learning_rate": 2.3107522219655025e-05, "loss": 0.0075, "step": 27870 }, { "grad_norm": 0.12990720570087433, "learning_rate": 2.3072682849460236e-05, "loss": 0.0069, "step": 27880 }, { "grad_norm": 0.13474705815315247, "learning_rate": 2.303786188383281e-05, "loss": 0.008, "step": 27890 }, { "grad_norm": 0.12063317745923996, "learning_rate": 2.300305934657257e-05, "loss": 0.0054, "step": 27900 }, { "grad_norm": 0.14066319167613983, "learning_rate": 2.2968275261466677e-05, "loss": 0.0083, "step": 27910 }, { "grad_norm": 0.1627531200647354, "learning_rate": 2.293350965228977e-05, "loss": 0.0052, "step": 27920 }, { "grad_norm": 0.12412192672491074, "learning_rate": 2.2898762542803776e-05, "loss": 0.0085, "step": 27930 }, { "grad_norm": 0.11584415286779404, "learning_rate": 2.286403395675803e-05, "loss": 0.0059, "step": 27940 }, { "grad_norm": 0.15606330335140228, "learning_rate": 2.28293239178892e-05, "loss": 0.009, "step": 27950 }, { "grad_norm": 0.17535457015037537, "learning_rate": 2.2794632449921287e-05, "loss": 0.0068, "step": 27960 }, { "grad_norm": 0.14024819433689117, "learning_rate": 2.275995957656555e-05, "loss": 0.0074, "step": 27970 }, { "grad_norm": 0.15494433045387268, "learning_rate": 2.272530532152058e-05, "loss": 0.0064, "step": 27980 }, { "grad_norm": 0.13115137815475464, "learning_rate": 2.2690669708472233e-05, "loss": 0.0078, "step": 27990 }, { "grad_norm": 0.13469837605953217, "learning_rate": 2.2656052761093655e-05, "loss": 0.0061, "step": 28000 }, { "grad_norm": 0.12620800733566284, "learning_rate": 2.262145450304517e-05, "loss": 0.0074, "step": 28010 }, { "grad_norm": 0.13469962775707245, "learning_rate": 2.2586874957974352e-05, "loss": 0.0059, "step": 28020 }, { "grad_norm": 0.1882493644952774, "learning_rate": 2.2552314149516012e-05, "loss": 0.0065, "step": 28030 }, { "grad_norm": 0.11320345103740692, "learning_rate": 2.2517772101292133e-05, "loss": 0.0061, "step": 28040 }, { "grad_norm": 0.16895204782485962, "learning_rate": 2.248324883691188e-05, "loss": 0.0093, "step": 28050 }, { "grad_norm": 0.1850971132516861, "learning_rate": 2.24487443799716e-05, "loss": 0.0065, "step": 28060 }, { "grad_norm": 0.14286985993385315, "learning_rate": 2.241425875405472e-05, "loss": 0.0053, "step": 28070 }, { "grad_norm": 0.19782587885856628, "learning_rate": 2.2379791982731868e-05, "loss": 0.0061, "step": 28080 }, { "grad_norm": 0.10775984078645706, "learning_rate": 2.2345344089560756e-05, "loss": 0.0043, "step": 28090 }, { "grad_norm": 0.1418152153491974, "learning_rate": 2.2310915098086206e-05, "loss": 0.0087, "step": 28100 }, { "grad_norm": 0.13941629230976105, "learning_rate": 2.227650503184009e-05, "loss": 0.0052, "step": 28110 }, { "grad_norm": 0.11720606684684753, "learning_rate": 2.2242113914341357e-05, "loss": 0.005, "step": 28120 }, { "grad_norm": 0.12179052084684372, "learning_rate": 2.220774176909602e-05, "loss": 0.006, "step": 28130 }, { "grad_norm": 0.10966113954782486, "learning_rate": 2.2173388619597114e-05, "loss": 0.0065, "step": 28140 }, { "grad_norm": 0.12315461039543152, "learning_rate": 2.21390544893247e-05, "loss": 0.0047, "step": 28150 }, { "grad_norm": 0.08811647444963455, "learning_rate": 2.210473940174585e-05, "loss": 0.0062, "step": 28160 }, { "grad_norm": 0.1571943461894989, "learning_rate": 2.207044338031456e-05, "loss": 0.0077, "step": 28170 }, { "grad_norm": 0.06968982517719269, "learning_rate": 2.203616644847186e-05, "loss": 0.0051, "step": 28180 }, { "grad_norm": 0.16321155428886414, "learning_rate": 2.200190862964571e-05, "loss": 0.0077, "step": 28190 }, { "grad_norm": 0.13087525963783264, "learning_rate": 2.1967669947251024e-05, "loss": 0.0057, "step": 28200 }, { "grad_norm": 0.07810088992118835, "learning_rate": 2.1933450424689583e-05, "loss": 0.0064, "step": 28210 }, { "grad_norm": 0.14314931631088257, "learning_rate": 2.1899250085350142e-05, "loss": 0.0061, "step": 28220 }, { "grad_norm": 0.11833231896162033, "learning_rate": 2.1865068952608277e-05, "loss": 0.0055, "step": 28230 }, { "grad_norm": 0.11534276604652405, "learning_rate": 2.1830907049826487e-05, "loss": 0.0065, "step": 28240 }, { "grad_norm": 0.14759768545627594, "learning_rate": 2.179676440035411e-05, "loss": 0.0063, "step": 28250 }, { "grad_norm": 0.11934524774551392, "learning_rate": 2.1762641027527337e-05, "loss": 0.0066, "step": 28260 }, { "grad_norm": 0.1615467667579651, "learning_rate": 2.1728536954669143e-05, "loss": 0.0049, "step": 28270 }, { "grad_norm": 0.137218177318573, "learning_rate": 2.169445220508936e-05, "loss": 0.0065, "step": 28280 }, { "grad_norm": 0.1358463019132614, "learning_rate": 2.166038680208461e-05, "loss": 0.0052, "step": 28290 }, { "grad_norm": 0.13409213721752167, "learning_rate": 2.162634076893823e-05, "loss": 0.0068, "step": 28300 }, { "grad_norm": 0.15883095562458038, "learning_rate": 2.1592314128920388e-05, "loss": 0.0069, "step": 28310 }, { "grad_norm": 0.12331993132829666, "learning_rate": 2.155830690528799e-05, "loss": 0.0073, "step": 28320 }, { "grad_norm": 0.11132751405239105, "learning_rate": 2.1524319121284613e-05, "loss": 0.006, "step": 28330 }, { "grad_norm": 0.1259261518716812, "learning_rate": 2.1490350800140607e-05, "loss": 0.0056, "step": 28340 }, { "grad_norm": 0.15265712141990662, "learning_rate": 2.1456401965073002e-05, "loss": 0.0052, "step": 28350 }, { "grad_norm": 0.15366388857364655, "learning_rate": 2.1422472639285524e-05, "loss": 0.0053, "step": 28360 }, { "grad_norm": 0.13140733540058136, "learning_rate": 2.13885628459685e-05, "loss": 0.0055, "step": 28370 }, { "grad_norm": 0.12969054281711578, "learning_rate": 2.135467260829901e-05, "loss": 0.0066, "step": 28380 }, { "grad_norm": 0.1491304188966751, "learning_rate": 2.1320801949440654e-05, "loss": 0.0071, "step": 28390 }, { "grad_norm": 0.10746683925390244, "learning_rate": 2.1286950892543744e-05, "loss": 0.0048, "step": 28400 }, { "grad_norm": 0.12664039433002472, "learning_rate": 2.125311946074515e-05, "loss": 0.0086, "step": 28410 }, { "grad_norm": 0.1186598613858223, "learning_rate": 2.1219307677168355e-05, "loss": 0.0053, "step": 28420 }, { "grad_norm": 0.13914239406585693, "learning_rate": 2.118551556492336e-05, "loss": 0.0065, "step": 28430 }, { "grad_norm": 0.11408048868179321, "learning_rate": 2.1151743147106774e-05, "loss": 0.0054, "step": 28440 }, { "grad_norm": 0.16323459148406982, "learning_rate": 2.111799044680172e-05, "loss": 0.0066, "step": 28450 }, { "grad_norm": 0.26691246032714844, "learning_rate": 2.1084257487077873e-05, "loss": 0.0062, "step": 28460 }, { "grad_norm": 0.1445283442735672, "learning_rate": 2.1050544290991357e-05, "loss": 0.0078, "step": 28470 }, { "grad_norm": 0.20528005063533783, "learning_rate": 2.101685088158486e-05, "loss": 0.0074, "step": 28480 }, { "grad_norm": 0.1193489357829094, "learning_rate": 2.0983177281887472e-05, "loss": 0.0059, "step": 28490 }, { "grad_norm": 0.11887570470571518, "learning_rate": 2.0949523514914798e-05, "loss": 0.0064, "step": 28500 }, { "grad_norm": 0.10665610432624817, "learning_rate": 2.0915889603668876e-05, "loss": 0.0044, "step": 28510 }, { "grad_norm": 0.12794984877109528, "learning_rate": 2.0882275571138175e-05, "loss": 0.0049, "step": 28520 }, { "grad_norm": 0.1091630682349205, "learning_rate": 2.0848681440297545e-05, "loss": 0.0071, "step": 28530 }, { "grad_norm": 0.1814165711402893, "learning_rate": 2.081510723410827e-05, "loss": 0.0061, "step": 28540 }, { "grad_norm": 0.18409353494644165, "learning_rate": 2.0781552975518003e-05, "loss": 0.0064, "step": 28550 }, { "grad_norm": 0.10584204643964767, "learning_rate": 2.074801868746078e-05, "loss": 0.0066, "step": 28560 }, { "grad_norm": 0.11454281210899353, "learning_rate": 2.0714504392856955e-05, "loss": 0.0073, "step": 28570 }, { "grad_norm": 0.08510760217905045, "learning_rate": 2.0681010114613215e-05, "loss": 0.0063, "step": 28580 }, { "grad_norm": 0.11780007183551788, "learning_rate": 2.0647535875622597e-05, "loss": 0.0058, "step": 28590 }, { "grad_norm": 0.10383728891611099, "learning_rate": 2.0614081698764432e-05, "loss": 0.0058, "step": 28600 }, { "grad_norm": 0.1713823527097702, "learning_rate": 2.0580647606904334e-05, "loss": 0.0064, "step": 28610 }, { "grad_norm": 0.13236626982688904, "learning_rate": 2.0547233622894208e-05, "loss": 0.0061, "step": 28620 }, { "grad_norm": 0.16259834170341492, "learning_rate": 2.0513839769572157e-05, "loss": 0.0078, "step": 28630 }, { "grad_norm": 0.11888734996318817, "learning_rate": 2.0480466069762584e-05, "loss": 0.0049, "step": 28640 }, { "grad_norm": 0.14402924478054047, "learning_rate": 2.0447112546276104e-05, "loss": 0.0071, "step": 28650 }, { "grad_norm": 0.11946439743041992, "learning_rate": 2.0413779221909547e-05, "loss": 0.0053, "step": 28660 }, { "grad_norm": 0.11668211966753006, "learning_rate": 2.0380466119445912e-05, "loss": 0.006, "step": 28670 }, { "grad_norm": 0.18390747904777527, "learning_rate": 2.0347173261654373e-05, "loss": 0.0069, "step": 28680 }, { "grad_norm": 0.1696925163269043, "learning_rate": 2.03139006712903e-05, "loss": 0.0065, "step": 28690 }, { "grad_norm": 0.14103500545024872, "learning_rate": 2.028064837109519e-05, "loss": 0.0062, "step": 28700 }, { "grad_norm": 0.1229960098862648, "learning_rate": 2.0247416383796685e-05, "loss": 0.007, "step": 28710 }, { "grad_norm": 0.10216871649026871, "learning_rate": 2.0214204732108548e-05, "loss": 0.005, "step": 28720 }, { "grad_norm": 0.145614892244339, "learning_rate": 2.0181013438730596e-05, "loss": 0.0063, "step": 28730 }, { "grad_norm": 0.10077041387557983, "learning_rate": 2.0147842526348783e-05, "loss": 0.0053, "step": 28740 }, { "grad_norm": 0.16495083272457123, "learning_rate": 2.011469201763511e-05, "loss": 0.0066, "step": 28750 }, { "grad_norm": 0.1431622952222824, "learning_rate": 2.0081561935247665e-05, "loss": 0.0064, "step": 28760 }, { "grad_norm": 0.15815503895282745, "learning_rate": 2.0048452301830523e-05, "loss": 0.0067, "step": 28770 }, { "grad_norm": 0.1058272197842598, "learning_rate": 2.0015363140013788e-05, "loss": 0.007, "step": 28780 }, { "grad_norm": 0.10527917742729187, "learning_rate": 1.9982294472413606e-05, "loss": 0.0051, "step": 28790 }, { "grad_norm": 0.10168062150478363, "learning_rate": 1.9949246321632103e-05, "loss": 0.0052, "step": 28800 }, { "grad_norm": 0.10574644058942795, "learning_rate": 1.9916218710257377e-05, "loss": 0.0049, "step": 28810 }, { "grad_norm": 0.14228218793869019, "learning_rate": 1.988321166086351e-05, "loss": 0.0081, "step": 28820 }, { "grad_norm": 0.0973120853304863, "learning_rate": 1.9850225196010468e-05, "loss": 0.0051, "step": 28830 }, { "grad_norm": 0.12101136893033981, "learning_rate": 1.981725933824421e-05, "loss": 0.0064, "step": 28840 }, { "grad_norm": 0.10478976368904114, "learning_rate": 1.978431411009661e-05, "loss": 0.0053, "step": 28850 }, { "grad_norm": 0.10653668642044067, "learning_rate": 1.9751389534085375e-05, "loss": 0.0054, "step": 28860 }, { "grad_norm": 0.13857401907444, "learning_rate": 1.9718485632714184e-05, "loss": 0.006, "step": 28870 }, { "grad_norm": 0.09771370142698288, "learning_rate": 1.968560242847251e-05, "loss": 0.0062, "step": 28880 }, { "grad_norm": 0.10453565418720245, "learning_rate": 1.965273994383573e-05, "loss": 0.0054, "step": 28890 }, { "grad_norm": 0.12171143293380737, "learning_rate": 1.961989820126504e-05, "loss": 0.0066, "step": 28900 }, { "grad_norm": 0.15017622709274292, "learning_rate": 1.958707722320746e-05, "loss": 0.0052, "step": 28910 }, { "grad_norm": 0.12244899570941925, "learning_rate": 1.955427703209584e-05, "loss": 0.0051, "step": 28920 }, { "grad_norm": 0.1184072345495224, "learning_rate": 1.9521497650348764e-05, "loss": 0.0076, "step": 28930 }, { "grad_norm": 0.09321122616529465, "learning_rate": 1.948873910037067e-05, "loss": 0.0055, "step": 28940 }, { "grad_norm": 0.1774321347475052, "learning_rate": 1.9456001404551678e-05, "loss": 0.0056, "step": 28950 }, { "grad_norm": 0.13103167712688446, "learning_rate": 1.942328458526771e-05, "loss": 0.0089, "step": 28960 }, { "grad_norm": 0.15450406074523926, "learning_rate": 1.9390588664880427e-05, "loss": 0.0062, "step": 28970 }, { "grad_norm": 0.14117863774299622, "learning_rate": 1.9357913665737145e-05, "loss": 0.0056, "step": 28980 }, { "grad_norm": 0.16721564531326294, "learning_rate": 1.932525961017093e-05, "loss": 0.0069, "step": 28990 }, { "grad_norm": 0.18123139441013336, "learning_rate": 1.9292626520500533e-05, "loss": 0.0062, "step": 29000 }, { "grad_norm": 0.11578291654586792, "learning_rate": 1.9260014419030354e-05, "loss": 0.0062, "step": 29010 }, { "grad_norm": 0.13156931102275848, "learning_rate": 1.9227423328050475e-05, "loss": 0.0065, "step": 29020 }, { "grad_norm": 0.11758281290531158, "learning_rate": 1.9194853269836582e-05, "loss": 0.005, "step": 29030 }, { "grad_norm": 0.13876351714134216, "learning_rate": 1.916230426664999e-05, "loss": 0.0089, "step": 29040 }, { "grad_norm": 0.11975447833538055, "learning_rate": 1.912977634073765e-05, "loss": 0.0046, "step": 29050 }, { "grad_norm": 0.09462428092956543, "learning_rate": 1.9097269514332083e-05, "loss": 0.0052, "step": 29060 }, { "grad_norm": 0.16680018603801727, "learning_rate": 1.9064783809651433e-05, "loss": 0.0067, "step": 29070 }, { "grad_norm": 0.16939598321914673, "learning_rate": 1.9032319248899333e-05, "loss": 0.006, "step": 29080 }, { "grad_norm": 0.12409448623657227, "learning_rate": 1.8999875854265015e-05, "loss": 0.0066, "step": 29090 }, { "grad_norm": 0.16976520419120789, "learning_rate": 1.8967453647923232e-05, "loss": 0.006, "step": 29100 }, { "grad_norm": 0.1559596061706543, "learning_rate": 1.893505265203427e-05, "loss": 0.0054, "step": 29110 }, { "grad_norm": 0.17795461416244507, "learning_rate": 1.8902672888743907e-05, "loss": 0.0059, "step": 29120 }, { "grad_norm": 0.13229025900363922, "learning_rate": 1.8870314380183396e-05, "loss": 0.0063, "step": 29130 }, { "grad_norm": 0.1547555923461914, "learning_rate": 1.8837977148469448e-05, "loss": 0.0083, "step": 29140 }, { "grad_norm": 0.16452474892139435, "learning_rate": 1.880566121570429e-05, "loss": 0.0064, "step": 29150 }, { "grad_norm": 0.22565117478370667, "learning_rate": 1.877336660397554e-05, "loss": 0.0069, "step": 29160 }, { "grad_norm": 0.0913868322968483, "learning_rate": 1.874109333535628e-05, "loss": 0.0062, "step": 29170 }, { "grad_norm": 0.1167052835226059, "learning_rate": 1.870884143190496e-05, "loss": 0.0058, "step": 29180 }, { "grad_norm": 0.12078125029802322, "learning_rate": 1.867661091566546e-05, "loss": 0.0065, "step": 29190 }, { "grad_norm": 0.06370024383068085, "learning_rate": 1.864440180866704e-05, "loss": 0.0045, "step": 29200 }, { "grad_norm": 0.18804723024368286, "learning_rate": 1.8612214132924317e-05, "loss": 0.0052, "step": 29210 }, { "grad_norm": 0.15059787034988403, "learning_rate": 1.858004791043728e-05, "loss": 0.0064, "step": 29220 }, { "grad_norm": 0.19297988712787628, "learning_rate": 1.854790316319123e-05, "loss": 0.0064, "step": 29230 }, { "grad_norm": 0.1798355132341385, "learning_rate": 1.8515779913156766e-05, "loss": 0.0076, "step": 29240 }, { "grad_norm": 0.12346334755420685, "learning_rate": 1.848367818228986e-05, "loss": 0.0066, "step": 29250 }, { "grad_norm": 0.10277598351240158, "learning_rate": 1.8451597992531733e-05, "loss": 0.0052, "step": 29260 }, { "grad_norm": 0.0763036459684372, "learning_rate": 1.8419539365808914e-05, "loss": 0.0063, "step": 29270 }, { "grad_norm": 0.11036748439073563, "learning_rate": 1.838750232403313e-05, "loss": 0.0051, "step": 29280 }, { "grad_norm": 0.1481596827507019, "learning_rate": 1.835548688910142e-05, "loss": 0.0063, "step": 29290 }, { "grad_norm": 0.21268847584724426, "learning_rate": 1.8323493082896037e-05, "loss": 0.0068, "step": 29300 }, { "grad_norm": 0.1149691492319107, "learning_rate": 1.8291520927284454e-05, "loss": 0.0054, "step": 29310 }, { "grad_norm": 0.11709994822740555, "learning_rate": 1.8259570444119305e-05, "loss": 0.0057, "step": 29320 }, { "grad_norm": 0.12079418450593948, "learning_rate": 1.8227641655238488e-05, "loss": 0.0071, "step": 29330 }, { "grad_norm": 0.08343882113695145, "learning_rate": 1.819573458246498e-05, "loss": 0.0048, "step": 29340 }, { "grad_norm": 0.1170654147863388, "learning_rate": 1.816384924760699e-05, "loss": 0.006, "step": 29350 }, { "grad_norm": 0.11326844245195389, "learning_rate": 1.813198567245784e-05, "loss": 0.006, "step": 29360 }, { "grad_norm": 0.12143893539905548, "learning_rate": 1.8100143878796006e-05, "loss": 0.0064, "step": 29370 }, { "grad_norm": 0.14295394718647003, "learning_rate": 1.8068323888385015e-05, "loss": 0.0052, "step": 29380 }, { "grad_norm": 0.1421777904033661, "learning_rate": 1.803652572297355e-05, "loss": 0.0053, "step": 29390 }, { "grad_norm": 0.12959903478622437, "learning_rate": 1.8004749404295353e-05, "loss": 0.0055, "step": 29400 }, { "grad_norm": 0.1271144449710846, "learning_rate": 1.797299495406926e-05, "loss": 0.0047, "step": 29410 }, { "grad_norm": 0.1283779889345169, "learning_rate": 1.7941262393999103e-05, "loss": 0.0047, "step": 29420 }, { "grad_norm": 0.16300131380558014, "learning_rate": 1.7909551745773816e-05, "loss": 0.0061, "step": 29430 }, { "grad_norm": 0.12625417113304138, "learning_rate": 1.7877863031067304e-05, "loss": 0.0058, "step": 29440 }, { "grad_norm": 0.10956133157014847, "learning_rate": 1.7846196271538516e-05, "loss": 0.0054, "step": 29450 }, { "grad_norm": 0.13353444635868073, "learning_rate": 1.7814551488831384e-05, "loss": 0.0061, "step": 29460 }, { "grad_norm": 0.12282971292734146, "learning_rate": 1.7782928704574835e-05, "loss": 0.0041, "step": 29470 }, { "grad_norm": 0.16876423358917236, "learning_rate": 1.775132794038271e-05, "loss": 0.0055, "step": 29480 }, { "grad_norm": 0.11432288587093353, "learning_rate": 1.7719749217853855e-05, "loss": 0.0049, "step": 29490 }, { "grad_norm": 0.14458800852298737, "learning_rate": 1.7688192558572038e-05, "loss": 0.0061, "step": 29500 }, { "grad_norm": 0.11755567044019699, "learning_rate": 1.7656657984105906e-05, "loss": 0.0069, "step": 29510 }, { "grad_norm": 0.1920827478170395, "learning_rate": 1.7625145516009068e-05, "loss": 0.0065, "step": 29520 }, { "grad_norm": 0.1326322853565216, "learning_rate": 1.7593655175820005e-05, "loss": 0.0046, "step": 29530 }, { "grad_norm": 0.12316484749317169, "learning_rate": 1.7562186985062046e-05, "loss": 0.008, "step": 29540 }, { "grad_norm": 0.07319024950265884, "learning_rate": 1.7530740965243403e-05, "loss": 0.005, "step": 29550 }, { "grad_norm": 0.14340348541736603, "learning_rate": 1.7499317137857153e-05, "loss": 0.006, "step": 29560 }, { "grad_norm": 0.10233306884765625, "learning_rate": 1.7467915524381184e-05, "loss": 0.005, "step": 29570 }, { "grad_norm": 0.08017901331186295, "learning_rate": 1.7436536146278182e-05, "loss": 0.0048, "step": 29580 }, { "grad_norm": 0.09702812135219574, "learning_rate": 1.7405179024995688e-05, "loss": 0.0048, "step": 29590 }, { "grad_norm": 0.18944452702999115, "learning_rate": 1.737384418196596e-05, "loss": 0.0073, "step": 29600 }, { "grad_norm": 0.19397638738155365, "learning_rate": 1.734253163860609e-05, "loss": 0.0082, "step": 29610 }, { "grad_norm": 0.15713097155094147, "learning_rate": 1.7311241416317896e-05, "loss": 0.0071, "step": 29620 }, { "grad_norm": 0.14754126965999603, "learning_rate": 1.7279973536487982e-05, "loss": 0.0052, "step": 29630 }, { "grad_norm": 0.107975535094738, "learning_rate": 1.724872802048761e-05, "loss": 0.0051, "step": 29640 }, { "grad_norm": 0.09816084057092667, "learning_rate": 1.7217504889672803e-05, "loss": 0.0055, "step": 29650 }, { "grad_norm": 0.23629701137542725, "learning_rate": 1.7186304165384287e-05, "loss": 0.005, "step": 29660 }, { "grad_norm": 0.10367613285779953, "learning_rate": 1.7155125868947475e-05, "loss": 0.0053, "step": 29670 }, { "grad_norm": 0.10835756361484528, "learning_rate": 1.7123970021672404e-05, "loss": 0.0052, "step": 29680 }, { "grad_norm": 0.1337396502494812, "learning_rate": 1.709283664485384e-05, "loss": 0.0052, "step": 29690 }, { "grad_norm": 0.17120641469955444, "learning_rate": 1.7061725759771113e-05, "loss": 0.0066, "step": 29700 }, { "grad_norm": 0.15485481917858124, "learning_rate": 1.7030637387688248e-05, "loss": 0.0058, "step": 29710 }, { "grad_norm": 0.10069240629673004, "learning_rate": 1.6999571549853836e-05, "loss": 0.0055, "step": 29720 }, { "grad_norm": 0.10719725489616394, "learning_rate": 1.696852826750112e-05, "loss": 0.0087, "step": 29730 }, { "grad_norm": 0.10301770269870758, "learning_rate": 1.6937507561847844e-05, "loss": 0.0043, "step": 29740 }, { "grad_norm": 0.1410502791404724, "learning_rate": 1.6906509454096385e-05, "loss": 0.007, "step": 29750 }, { "grad_norm": 0.11980791389942169, "learning_rate": 1.687553396543367e-05, "loss": 0.0061, "step": 29760 }, { "grad_norm": 0.12975217401981354, "learning_rate": 1.6844581117031154e-05, "loss": 0.0066, "step": 29770 }, { "grad_norm": 0.11102506518363953, "learning_rate": 1.681365093004481e-05, "loss": 0.0064, "step": 29780 }, { "grad_norm": 0.1634426712989807, "learning_rate": 1.678274342561511e-05, "loss": 0.0056, "step": 29790 }, { "grad_norm": 0.16181807219982147, "learning_rate": 1.675185862486706e-05, "loss": 0.0094, "step": 29800 }, { "grad_norm": 0.1600041687488556, "learning_rate": 1.6720996548910127e-05, "loss": 0.004, "step": 29810 }, { "grad_norm": 0.10400348156690598, "learning_rate": 1.6690157218838247e-05, "loss": 0.006, "step": 29820 }, { "grad_norm": 0.09524272382259369, "learning_rate": 1.665934065572984e-05, "loss": 0.0061, "step": 29830 }, { "grad_norm": 0.14951974153518677, "learning_rate": 1.6628546880647688e-05, "loss": 0.0055, "step": 29840 }, { "grad_norm": 0.1697375476360321, "learning_rate": 1.6597775914639076e-05, "loss": 0.0063, "step": 29850 }, { "grad_norm": 0.15202240645885468, "learning_rate": 1.6567027778735654e-05, "loss": 0.0077, "step": 29860 }, { "grad_norm": 0.17178009450435638, "learning_rate": 1.653630249395351e-05, "loss": 0.0062, "step": 29870 }, { "grad_norm": 0.1572396457195282, "learning_rate": 1.6505600081293072e-05, "loss": 0.0077, "step": 29880 }, { "grad_norm": 0.1283363401889801, "learning_rate": 1.647492056173912e-05, "loss": 0.0078, "step": 29890 }, { "grad_norm": 0.14474748075008392, "learning_rate": 1.644426395626085e-05, "loss": 0.0054, "step": 29900 }, { "grad_norm": 0.20124311745166779, "learning_rate": 1.641363028581175e-05, "loss": 0.007, "step": 29910 }, { "grad_norm": 0.09107106178998947, "learning_rate": 1.638301957132965e-05, "loss": 0.0056, "step": 29920 }, { "grad_norm": 0.11459647119045258, "learning_rate": 1.6352431833736703e-05, "loss": 0.0082, "step": 29930 }, { "grad_norm": 0.10006976127624512, "learning_rate": 1.6321867093939298e-05, "loss": 0.007, "step": 29940 }, { "grad_norm": 0.1619967371225357, "learning_rate": 1.629132537282817e-05, "loss": 0.0066, "step": 29950 }, { "grad_norm": 0.10565415024757385, "learning_rate": 1.62608066912783e-05, "loss": 0.0048, "step": 29960 }, { "grad_norm": 0.13850298523902893, "learning_rate": 1.623031107014893e-05, "loss": 0.0053, "step": 29970 }, { "grad_norm": 0.11840171366930008, "learning_rate": 1.619983853028351e-05, "loss": 0.0041, "step": 29980 }, { "grad_norm": 0.14376935362815857, "learning_rate": 1.6169389092509724e-05, "loss": 0.0051, "step": 29990 }, { "grad_norm": 0.12784157693386078, "learning_rate": 1.6138962777639494e-05, "loss": 0.0053, "step": 30000 }, { "grad_norm": 0.11800028383731842, "learning_rate": 1.610855960646891e-05, "loss": 0.0046, "step": 30010 }, { "grad_norm": 0.10807386785745621, "learning_rate": 1.607817959977826e-05, "loss": 0.0084, "step": 30020 }, { "grad_norm": 0.12346983700990677, "learning_rate": 1.6047822778332005e-05, "loss": 0.0059, "step": 30030 }, { "grad_norm": 0.09645577520132065, "learning_rate": 1.6017489162878713e-05, "loss": 0.0046, "step": 30040 }, { "grad_norm": 0.09389755129814148, "learning_rate": 1.5987178774151147e-05, "loss": 0.0071, "step": 30050 }, { "grad_norm": 0.16287505626678467, "learning_rate": 1.5956891632866184e-05, "loss": 0.0072, "step": 30060 }, { "grad_norm": 0.12135220319032669, "learning_rate": 1.592662775972476e-05, "loss": 0.0051, "step": 30070 }, { "grad_norm": 0.09171595424413681, "learning_rate": 1.5896387175411986e-05, "loss": 0.0049, "step": 30080 }, { "grad_norm": 0.09102007746696472, "learning_rate": 1.586616990059699e-05, "loss": 0.0065, "step": 30090 }, { "grad_norm": 0.12205309420824051, "learning_rate": 1.5835975955933002e-05, "loss": 0.0064, "step": 30100 }, { "grad_norm": 0.13660471141338348, "learning_rate": 1.5805805362057295e-05, "loss": 0.0068, "step": 30110 }, { "grad_norm": 0.09289707988500595, "learning_rate": 1.57756581395912e-05, "loss": 0.0053, "step": 30120 }, { "grad_norm": 0.13337768614292145, "learning_rate": 1.574553430914006e-05, "loss": 0.0055, "step": 30130 }, { "grad_norm": 0.16399650275707245, "learning_rate": 1.5715433891293206e-05, "loss": 0.0095, "step": 30140 }, { "grad_norm": 0.1442788988351822, "learning_rate": 1.5685356906624016e-05, "loss": 0.0071, "step": 30150 }, { "grad_norm": 0.11243351548910141, "learning_rate": 1.56553033756898e-05, "loss": 0.0056, "step": 30160 }, { "grad_norm": 0.09043826907873154, "learning_rate": 1.5625273319031873e-05, "loss": 0.007, "step": 30170 }, { "grad_norm": 0.09412682801485062, "learning_rate": 1.5595266757175518e-05, "loss": 0.0056, "step": 30180 }, { "grad_norm": 0.10605674982070923, "learning_rate": 1.55652837106299e-05, "loss": 0.0062, "step": 30190 }, { "grad_norm": 0.07926937937736511, "learning_rate": 1.5535324199888164e-05, "loss": 0.0073, "step": 30200 }, { "grad_norm": 0.1374395191669464, "learning_rate": 1.5505388245427355e-05, "loss": 0.0064, "step": 30210 }, { "grad_norm": 0.18960073590278625, "learning_rate": 1.5475475867708418e-05, "loss": 0.008, "step": 30220 }, { "grad_norm": 0.10392557084560394, "learning_rate": 1.5445587087176195e-05, "loss": 0.005, "step": 30230 }, { "grad_norm": 0.09911428391933441, "learning_rate": 1.5415721924259346e-05, "loss": 0.0054, "step": 30240 }, { "grad_norm": 0.16546973586082458, "learning_rate": 1.538588039937047e-05, "loss": 0.008, "step": 30250 }, { "grad_norm": 0.11664903908967972, "learning_rate": 1.535606253290594e-05, "loss": 0.0047, "step": 30260 }, { "grad_norm": 0.11320020258426666, "learning_rate": 1.5326268345245993e-05, "loss": 0.0062, "step": 30270 }, { "grad_norm": 0.15377750992774963, "learning_rate": 1.5296497856754693e-05, "loss": 0.0078, "step": 30280 }, { "grad_norm": 0.14288783073425293, "learning_rate": 1.5266751087779858e-05, "loss": 0.0063, "step": 30290 }, { "grad_norm": 0.16140545904636383, "learning_rate": 1.5237028058653142e-05, "loss": 0.0053, "step": 30300 }, { "grad_norm": 0.08228013664484024, "learning_rate": 1.5207328789689946e-05, "loss": 0.005, "step": 30310 }, { "grad_norm": 0.1346340924501419, "learning_rate": 1.5177653301189454e-05, "loss": 0.0068, "step": 30320 }, { "grad_norm": 0.10760292410850525, "learning_rate": 1.5148001613434587e-05, "loss": 0.0061, "step": 30330 }, { "grad_norm": 0.16008447110652924, "learning_rate": 1.5118373746691983e-05, "loss": 0.0067, "step": 30340 }, { "grad_norm": 0.12271366268396378, "learning_rate": 1.5088769721211992e-05, "loss": 0.0059, "step": 30350 }, { "grad_norm": 0.14413033425807953, "learning_rate": 1.505918955722871e-05, "loss": 0.0061, "step": 30360 }, { "grad_norm": 0.11594375967979431, "learning_rate": 1.5029633274959886e-05, "loss": 0.0071, "step": 30370 }, { "grad_norm": 0.13890960812568665, "learning_rate": 1.5000100894606989e-05, "loss": 0.0054, "step": 30380 }, { "grad_norm": 0.12178099900484085, "learning_rate": 1.4970592436355079e-05, "loss": 0.0057, "step": 30390 }, { "grad_norm": 0.10682504624128342, "learning_rate": 1.4941107920372938e-05, "loss": 0.0047, "step": 30400 }, { "grad_norm": 0.10532570630311966, "learning_rate": 1.4911647366812947e-05, "loss": 0.0059, "step": 30410 }, { "grad_norm": 0.1218567043542862, "learning_rate": 1.4882210795811114e-05, "loss": 0.0058, "step": 30420 }, { "grad_norm": 0.1238861158490181, "learning_rate": 1.485279822748708e-05, "loss": 0.0036, "step": 30430 }, { "grad_norm": 0.0763607770204544, "learning_rate": 1.4823409681944034e-05, "loss": 0.0038, "step": 30440 }, { "grad_norm": 0.18494197726249695, "learning_rate": 1.4794045179268763e-05, "loss": 0.0052, "step": 30450 }, { "grad_norm": 0.10338404774665833, "learning_rate": 1.4764704739531643e-05, "loss": 0.0046, "step": 30460 }, { "grad_norm": 0.13795632123947144, "learning_rate": 1.473538838278658e-05, "loss": 0.0051, "step": 30470 }, { "grad_norm": 0.097343809902668, "learning_rate": 1.4706096129071045e-05, "loss": 0.0044, "step": 30480 }, { "grad_norm": 0.1593656837940216, "learning_rate": 1.4676827998405996e-05, "loss": 0.005, "step": 30490 }, { "grad_norm": 0.1399906724691391, "learning_rate": 1.4647584010795928e-05, "loss": 0.0064, "step": 30500 }, { "grad_norm": 0.13979575037956238, "learning_rate": 1.4618364186228834e-05, "loss": 0.0049, "step": 30510 }, { "grad_norm": 0.11846279352903366, "learning_rate": 1.4589168544676207e-05, "loss": 0.0058, "step": 30520 }, { "grad_norm": 0.10265812277793884, "learning_rate": 1.455999710609296e-05, "loss": 0.0067, "step": 30530 }, { "grad_norm": 0.08181165158748627, "learning_rate": 1.453084989041753e-05, "loss": 0.0051, "step": 30540 }, { "grad_norm": 0.15693101286888123, "learning_rate": 1.4501726917571728e-05, "loss": 0.0045, "step": 30550 }, { "grad_norm": 0.1262279748916626, "learning_rate": 1.4472628207460848e-05, "loss": 0.0069, "step": 30560 }, { "grad_norm": 0.15670989453792572, "learning_rate": 1.4443553779973584e-05, "loss": 0.0086, "step": 30570 }, { "grad_norm": 0.14260876178741455, "learning_rate": 1.4414503654982037e-05, "loss": 0.0066, "step": 30580 }, { "grad_norm": 0.10580242425203323, "learning_rate": 1.4385477852341706e-05, "loss": 0.0045, "step": 30590 }, { "grad_norm": 0.1387321650981903, "learning_rate": 1.4356476391891421e-05, "loss": 0.005, "step": 30600 }, { "grad_norm": 0.29809680581092834, "learning_rate": 1.4327499293453423e-05, "loss": 0.0067, "step": 30610 }, { "grad_norm": 0.13273519277572632, "learning_rate": 1.4298546576833304e-05, "loss": 0.0043, "step": 30620 }, { "grad_norm": 0.11408399045467377, "learning_rate": 1.426961826181994e-05, "loss": 0.006, "step": 30630 }, { "grad_norm": 0.12704163789749146, "learning_rate": 1.4240714368185592e-05, "loss": 0.0059, "step": 30640 }, { "grad_norm": 0.0751732587814331, "learning_rate": 1.4211834915685767e-05, "loss": 0.0046, "step": 30650 }, { "grad_norm": 0.16202788054943085, "learning_rate": 1.4182979924059313e-05, "loss": 0.0065, "step": 30660 }, { "grad_norm": 0.10149211436510086, "learning_rate": 1.415414941302835e-05, "loss": 0.0046, "step": 30670 }, { "grad_norm": 0.16428042948246002, "learning_rate": 1.4125343402298257e-05, "loss": 0.0049, "step": 30680 }, { "grad_norm": 0.09551575779914856, "learning_rate": 1.4096561911557688e-05, "loss": 0.0054, "step": 30690 }, { "grad_norm": 0.15602990984916687, "learning_rate": 1.4067804960478481e-05, "loss": 0.0055, "step": 30700 }, { "grad_norm": 0.0998486578464508, "learning_rate": 1.4039072568715784e-05, "loss": 0.0057, "step": 30710 }, { "grad_norm": 0.10501308739185333, "learning_rate": 1.401036475590788e-05, "loss": 0.006, "step": 30720 }, { "grad_norm": 0.10947232693433762, "learning_rate": 1.3981681541676301e-05, "loss": 0.0063, "step": 30730 }, { "grad_norm": 0.19015459716320038, "learning_rate": 1.3953022945625772e-05, "loss": 0.0061, "step": 30740 }, { "grad_norm": 0.11612392216920853, "learning_rate": 1.3924388987344138e-05, "loss": 0.0042, "step": 30750 }, { "grad_norm": 0.09843410551548004, "learning_rate": 1.3895779686402465e-05, "loss": 0.0056, "step": 30760 }, { "grad_norm": 0.131010502576828, "learning_rate": 1.3867195062354931e-05, "loss": 0.0054, "step": 30770 }, { "grad_norm": 0.09718844294548035, "learning_rate": 1.3838635134738858e-05, "loss": 0.0065, "step": 30780 }, { "grad_norm": 0.1828058660030365, "learning_rate": 1.3810099923074704e-05, "loss": 0.0043, "step": 30790 }, { "grad_norm": 0.18324223160743713, "learning_rate": 1.3781589446866006e-05, "loss": 0.007, "step": 30800 }, { "grad_norm": 0.12295352667570114, "learning_rate": 1.3753103725599387e-05, "loss": 0.0062, "step": 30810 }, { "grad_norm": 0.1771109402179718, "learning_rate": 1.3724642778744579e-05, "loss": 0.0069, "step": 30820 }, { "grad_norm": 0.0718613713979721, "learning_rate": 1.3696206625754376e-05, "loss": 0.0067, "step": 30830 }, { "grad_norm": 0.14964383840560913, "learning_rate": 1.3667795286064639e-05, "loss": 0.0045, "step": 30840 }, { "grad_norm": 0.11179175972938538, "learning_rate": 1.3639408779094215e-05, "loss": 0.0043, "step": 30850 }, { "grad_norm": 0.11428510397672653, "learning_rate": 1.361104712424503e-05, "loss": 0.0073, "step": 30860 }, { "grad_norm": 0.1481141597032547, "learning_rate": 1.358271034090201e-05, "loss": 0.008, "step": 30870 }, { "grad_norm": 0.1705075204372406, "learning_rate": 1.3554398448433076e-05, "loss": 0.0061, "step": 30880 }, { "grad_norm": 0.15245184302330017, "learning_rate": 1.3526111466189156e-05, "loss": 0.0047, "step": 30890 }, { "grad_norm": 0.173567995429039, "learning_rate": 1.3497849413504121e-05, "loss": 0.0063, "step": 30900 }, { "grad_norm": 0.19111678004264832, "learning_rate": 1.3469612309694802e-05, "loss": 0.0051, "step": 30910 }, { "grad_norm": 0.10782533138990402, "learning_rate": 1.3441400174061008e-05, "loss": 0.0056, "step": 30920 }, { "grad_norm": 0.18425829708576202, "learning_rate": 1.3413213025885469e-05, "loss": 0.0057, "step": 30930 }, { "grad_norm": 0.15212729573249817, "learning_rate": 1.3385050884433853e-05, "loss": 0.0042, "step": 30940 }, { "grad_norm": 0.12670892477035522, "learning_rate": 1.3356913768954676e-05, "loss": 0.0053, "step": 30950 }, { "grad_norm": 0.08508129417896271, "learning_rate": 1.3328801698679415e-05, "loss": 0.0061, "step": 30960 }, { "grad_norm": 0.08339177072048187, "learning_rate": 1.3300714692822413e-05, "loss": 0.0049, "step": 30970 }, { "grad_norm": 0.0790681391954422, "learning_rate": 1.3272652770580857e-05, "loss": 0.0049, "step": 30980 }, { "grad_norm": 0.11728303879499435, "learning_rate": 1.3244615951134831e-05, "loss": 0.007, "step": 30990 }, { "grad_norm": 0.12257551401853561, "learning_rate": 1.3216604253647219e-05, "loss": 0.0048, "step": 31000 }, { "grad_norm": 0.11988640576601028, "learning_rate": 1.3188617697263738e-05, "loss": 0.0065, "step": 31010 }, { "grad_norm": 0.14765819907188416, "learning_rate": 1.3160656301112956e-05, "loss": 0.0068, "step": 31020 }, { "grad_norm": 0.12915444374084473, "learning_rate": 1.3132720084306211e-05, "loss": 0.0044, "step": 31030 }, { "grad_norm": 0.1604388803243637, "learning_rate": 1.3104809065937668e-05, "loss": 0.0086, "step": 31040 }, { "grad_norm": 0.09912508726119995, "learning_rate": 1.3076923265084212e-05, "loss": 0.0051, "step": 31050 }, { "grad_norm": 0.11206257343292236, "learning_rate": 1.3049062700805543e-05, "loss": 0.0051, "step": 31060 }, { "grad_norm": 0.13269564509391785, "learning_rate": 1.3021227392144092e-05, "loss": 0.0056, "step": 31070 }, { "grad_norm": 0.09017789363861084, "learning_rate": 1.2993417358125037e-05, "loss": 0.0058, "step": 31080 }, { "grad_norm": 0.12643511593341827, "learning_rate": 1.2965632617756257e-05, "loss": 0.0043, "step": 31090 }, { "grad_norm": 0.13446730375289917, "learning_rate": 1.2937873190028387e-05, "loss": 0.0061, "step": 31100 }, { "grad_norm": 0.12908729910850525, "learning_rate": 1.2910139093914697e-05, "loss": 0.0055, "step": 31110 }, { "grad_norm": 0.11212699115276337, "learning_rate": 1.2882430348371205e-05, "loss": 0.0061, "step": 31120 }, { "grad_norm": 0.1368834525346756, "learning_rate": 1.2854746972336567e-05, "loss": 0.0048, "step": 31130 }, { "grad_norm": 0.1248813346028328, "learning_rate": 1.2827088984732132e-05, "loss": 0.0049, "step": 31140 }, { "grad_norm": 0.22855018079280853, "learning_rate": 1.279945640446184e-05, "loss": 0.0051, "step": 31150 }, { "grad_norm": 0.13207076489925385, "learning_rate": 1.2771849250412315e-05, "loss": 0.0067, "step": 31160 }, { "grad_norm": 0.12421707063913345, "learning_rate": 1.2744267541452792e-05, "loss": 0.0054, "step": 31170 }, { "grad_norm": 0.1295749992132187, "learning_rate": 1.2716711296435112e-05, "loss": 0.0054, "step": 31180 }, { "grad_norm": 0.1382538378238678, "learning_rate": 1.268918053419369e-05, "loss": 0.0048, "step": 31190 }, { "grad_norm": 0.12648507952690125, "learning_rate": 1.2661675273545564e-05, "loss": 0.005, "step": 31200 }, { "grad_norm": 0.15239274501800537, "learning_rate": 1.26341955332903e-05, "loss": 0.0054, "step": 31210 }, { "grad_norm": 0.1543363481760025, "learning_rate": 1.2606741332210048e-05, "loss": 0.007, "step": 31220 }, { "grad_norm": 0.10007019340991974, "learning_rate": 1.2579312689069495e-05, "loss": 0.0052, "step": 31230 }, { "grad_norm": 0.13134139776229858, "learning_rate": 1.2551909622615881e-05, "loss": 0.0082, "step": 31240 }, { "grad_norm": 0.09975650906562805, "learning_rate": 1.252453215157891e-05, "loss": 0.0054, "step": 31250 }, { "grad_norm": 0.14129692316055298, "learning_rate": 1.2497180294670841e-05, "loss": 0.0046, "step": 31260 }, { "grad_norm": 0.10956517606973648, "learning_rate": 1.2469854070586423e-05, "loss": 0.0063, "step": 31270 }, { "grad_norm": 0.10877707600593567, "learning_rate": 1.2442553498002845e-05, "loss": 0.0083, "step": 31280 }, { "grad_norm": 0.14241750538349152, "learning_rate": 1.241527859557981e-05, "loss": 0.0068, "step": 31290 }, { "grad_norm": 0.13222330808639526, "learning_rate": 1.238802938195946e-05, "loss": 0.0078, "step": 31300 }, { "grad_norm": 0.13087226450443268, "learning_rate": 1.2360805875766357e-05, "loss": 0.0041, "step": 31310 }, { "grad_norm": 0.11237813532352448, "learning_rate": 1.2333608095607524e-05, "loss": 0.0047, "step": 31320 }, { "grad_norm": 0.11211016774177551, "learning_rate": 1.230643606007238e-05, "loss": 0.0051, "step": 31330 }, { "grad_norm": 0.14170387387275696, "learning_rate": 1.2279289787732778e-05, "loss": 0.0058, "step": 31340 }, { "grad_norm": 0.09350341558456421, "learning_rate": 1.2252169297142906e-05, "loss": 0.0061, "step": 31350 }, { "grad_norm": 0.10840052366256714, "learning_rate": 1.2225074606839392e-05, "loss": 0.0056, "step": 31360 }, { "grad_norm": 0.12087666988372803, "learning_rate": 1.2198005735341173e-05, "loss": 0.0057, "step": 31370 }, { "grad_norm": 0.16086970269680023, "learning_rate": 1.2170962701149584e-05, "loss": 0.0063, "step": 31380 }, { "grad_norm": 0.11036689579486847, "learning_rate": 1.2143945522748285e-05, "loss": 0.0082, "step": 31390 }, { "grad_norm": 0.07316151261329651, "learning_rate": 1.2116954218603266e-05, "loss": 0.0059, "step": 31400 }, { "grad_norm": 0.1337687075138092, "learning_rate": 1.2089988807162817e-05, "loss": 0.006, "step": 31410 }, { "grad_norm": 0.15312905609607697, "learning_rate": 1.2063049306857543e-05, "loss": 0.0078, "step": 31420 }, { "grad_norm": 0.0958588644862175, "learning_rate": 1.2036135736100346e-05, "loss": 0.005, "step": 31430 }, { "grad_norm": 0.1263469159603119, "learning_rate": 1.200924811328641e-05, "loss": 0.0048, "step": 31440 }, { "grad_norm": 0.1643855720758438, "learning_rate": 1.1982386456793143e-05, "loss": 0.005, "step": 31450 }, { "grad_norm": 0.13466140627861023, "learning_rate": 1.1955550784980257e-05, "loss": 0.0059, "step": 31460 }, { "grad_norm": 0.11934806406497955, "learning_rate": 1.1928741116189663e-05, "loss": 0.0055, "step": 31470 }, { "grad_norm": 0.13310131430625916, "learning_rate": 1.1901957468745529e-05, "loss": 0.005, "step": 31480 }, { "grad_norm": 0.09613321721553802, "learning_rate": 1.1875199860954217e-05, "loss": 0.0054, "step": 31490 }, { "grad_norm": 0.14352735877037048, "learning_rate": 1.1848468311104316e-05, "loss": 0.0058, "step": 31500 }, { "grad_norm": 0.09949253499507904, "learning_rate": 1.1821762837466566e-05, "loss": 0.0045, "step": 31510 }, { "grad_norm": 0.0888509452342987, "learning_rate": 1.179508345829392e-05, "loss": 0.0042, "step": 31520 }, { "grad_norm": 0.11577223986387253, "learning_rate": 1.1768430191821478e-05, "loss": 0.0044, "step": 31530 }, { "grad_norm": 0.09853500127792358, "learning_rate": 1.174180305626651e-05, "loss": 0.0065, "step": 31540 }, { "grad_norm": 0.08999276161193848, "learning_rate": 1.1715202069828401e-05, "loss": 0.0079, "step": 31550 }, { "grad_norm": 0.1447075456380844, "learning_rate": 1.1688627250688661e-05, "loss": 0.0051, "step": 31560 }, { "grad_norm": 0.10752806067466736, "learning_rate": 1.1662078617010935e-05, "loss": 0.0039, "step": 31570 }, { "grad_norm": 0.11594665050506592, "learning_rate": 1.1635556186940972e-05, "loss": 0.0043, "step": 31580 }, { "grad_norm": 0.11333058774471283, "learning_rate": 1.1609059978606602e-05, "loss": 0.0045, "step": 31590 }, { "grad_norm": 0.12573981285095215, "learning_rate": 1.1582590010117745e-05, "loss": 0.0054, "step": 31600 }, { "grad_norm": 0.1072617620229721, "learning_rate": 1.1556146299566345e-05, "loss": 0.0046, "step": 31610 }, { "grad_norm": 0.1836840659379959, "learning_rate": 1.1529728865026452e-05, "loss": 0.0073, "step": 31620 }, { "grad_norm": 0.1606292724609375, "learning_rate": 1.150333772455412e-05, "loss": 0.0035, "step": 31630 }, { "grad_norm": 0.16421450674533844, "learning_rate": 1.1476972896187466e-05, "loss": 0.0079, "step": 31640 }, { "grad_norm": 0.14136114716529846, "learning_rate": 1.1450634397946592e-05, "loss": 0.0051, "step": 31650 }, { "grad_norm": 0.17482806742191315, "learning_rate": 1.142432224783359e-05, "loss": 0.0071, "step": 31660 }, { "grad_norm": 0.1785348653793335, "learning_rate": 1.1398036463832584e-05, "loss": 0.0073, "step": 31670 }, { "grad_norm": 0.10303580015897751, "learning_rate": 1.1371777063909661e-05, "loss": 0.0055, "step": 31680 }, { "grad_norm": 0.09065265953540802, "learning_rate": 1.1345544066012874e-05, "loss": 0.0044, "step": 31690 }, { "grad_norm": 0.18073056638240814, "learning_rate": 1.131933748807224e-05, "loss": 0.0061, "step": 31700 }, { "grad_norm": 0.14137771725654602, "learning_rate": 1.1293157347999683e-05, "loss": 0.0048, "step": 31710 }, { "grad_norm": 0.1313239485025406, "learning_rate": 1.1267003663689097e-05, "loss": 0.0044, "step": 31720 }, { "grad_norm": 0.1880386620759964, "learning_rate": 1.1240876453016275e-05, "loss": 0.0051, "step": 31730 }, { "grad_norm": 0.08091960847377777, "learning_rate": 1.1214775733838933e-05, "loss": 0.0055, "step": 31740 }, { "grad_norm": 0.15887881815433502, "learning_rate": 1.118870152399666e-05, "loss": 0.0059, "step": 31750 }, { "grad_norm": 0.10599994659423828, "learning_rate": 1.1162653841310905e-05, "loss": 0.005, "step": 31760 }, { "grad_norm": 0.18042989075183868, "learning_rate": 1.1136632703585043e-05, "loss": 0.0056, "step": 31770 }, { "grad_norm": 0.090924933552742, "learning_rate": 1.111063812860426e-05, "loss": 0.005, "step": 31780 }, { "grad_norm": 0.195891335606575, "learning_rate": 1.1084670134135616e-05, "loss": 0.006, "step": 31790 }, { "grad_norm": 0.2063804566860199, "learning_rate": 1.1058728737927998e-05, "loss": 0.0058, "step": 31800 }, { "grad_norm": 0.07469817250967026, "learning_rate": 1.1032813957712086e-05, "loss": 0.0055, "step": 31810 }, { "grad_norm": 0.10875401645898819, "learning_rate": 1.1006925811200385e-05, "loss": 0.005, "step": 31820 }, { "grad_norm": 0.055063601583242416, "learning_rate": 1.0981064316087231e-05, "loss": 0.0055, "step": 31830 }, { "grad_norm": 0.11239868402481079, "learning_rate": 1.0955229490048675e-05, "loss": 0.0051, "step": 31840 }, { "grad_norm": 0.10568016022443771, "learning_rate": 1.0929421350742603e-05, "loss": 0.005, "step": 31850 }, { "grad_norm": 0.1050238311290741, "learning_rate": 1.090363991580861e-05, "loss": 0.005, "step": 31860 }, { "grad_norm": 0.13022996485233307, "learning_rate": 1.0877885202868083e-05, "loss": 0.0045, "step": 31870 }, { "grad_norm": 0.07090496271848679, "learning_rate": 1.0852157229524118e-05, "loss": 0.0045, "step": 31880 }, { "grad_norm": 0.06340966373682022, "learning_rate": 1.0826456013361542e-05, "loss": 0.0038, "step": 31890 }, { "grad_norm": 0.14553028345108032, "learning_rate": 1.080078157194691e-05, "loss": 0.0048, "step": 31900 }, { "grad_norm": 0.09328698366880417, "learning_rate": 1.077513392282844e-05, "loss": 0.0054, "step": 31910 }, { "grad_norm": 0.12014104425907135, "learning_rate": 1.0749513083536073e-05, "loss": 0.0044, "step": 31920 }, { "grad_norm": 0.16355247795581818, "learning_rate": 1.072391907158139e-05, "loss": 0.0048, "step": 31930 }, { "grad_norm": 0.13339413702487946, "learning_rate": 1.0698351904457676e-05, "loss": 0.0074, "step": 31940 }, { "grad_norm": 0.1652868539094925, "learning_rate": 1.0672811599639849e-05, "loss": 0.005, "step": 31950 }, { "grad_norm": 0.15850265324115753, "learning_rate": 1.0647298174584453e-05, "loss": 0.0075, "step": 31960 }, { "grad_norm": 0.12402410060167313, "learning_rate": 1.0621811646729673e-05, "loss": 0.0045, "step": 31970 }, { "grad_norm": 0.14623311161994934, "learning_rate": 1.0596352033495328e-05, "loss": 0.0066, "step": 31980 }, { "grad_norm": 0.18222825229167938, "learning_rate": 1.057091935228281e-05, "loss": 0.0043, "step": 31990 }, { "grad_norm": 0.12138678878545761, "learning_rate": 1.0545513620475134e-05, "loss": 0.0052, "step": 32000 }, { "grad_norm": 0.12662525475025177, "learning_rate": 1.0520134855436863e-05, "loss": 0.0068, "step": 32010 }, { "grad_norm": 0.14526483416557312, "learning_rate": 1.049478307451413e-05, "loss": 0.0049, "step": 32020 }, { "grad_norm": 0.13335928320884705, "learning_rate": 1.0469458295034656e-05, "loss": 0.0045, "step": 32030 }, { "grad_norm": 0.1048998162150383, "learning_rate": 1.044416053430769e-05, "loss": 0.0042, "step": 32040 }, { "grad_norm": 0.1478457897901535, "learning_rate": 1.0418889809624021e-05, "loss": 0.0039, "step": 32050 }, { "grad_norm": 0.09549279510974884, "learning_rate": 1.0393646138255925e-05, "loss": 0.0057, "step": 32060 }, { "grad_norm": 0.11158299446105957, "learning_rate": 1.036842953745723e-05, "loss": 0.0041, "step": 32070 }, { "grad_norm": 0.17665588855743408, "learning_rate": 1.034324002446324e-05, "loss": 0.0063, "step": 32080 }, { "grad_norm": 0.07096729427576065, "learning_rate": 1.0318077616490751e-05, "loss": 0.0052, "step": 32090 }, { "grad_norm": 0.14049983024597168, "learning_rate": 1.0292942330738054e-05, "loss": 0.004, "step": 32100 }, { "grad_norm": 0.04825165495276451, "learning_rate": 1.026783418438485e-05, "loss": 0.0042, "step": 32110 }, { "grad_norm": 0.12670931220054626, "learning_rate": 1.0242753194592319e-05, "loss": 0.005, "step": 32120 }, { "grad_norm": 0.14080588519573212, "learning_rate": 1.0217699378503092e-05, "loss": 0.0047, "step": 32130 }, { "grad_norm": 0.08964796364307404, "learning_rate": 1.0192672753241218e-05, "loss": 0.0054, "step": 32140 }, { "grad_norm": 0.09152176231145859, "learning_rate": 1.0167673335912164e-05, "loss": 0.0069, "step": 32150 }, { "grad_norm": 0.09290209412574768, "learning_rate": 1.0142701143602779e-05, "loss": 0.0043, "step": 32160 }, { "grad_norm": 0.20657579600811005, "learning_rate": 1.0117756193381328e-05, "loss": 0.0045, "step": 32170 }, { "grad_norm": 0.15948635339736938, "learning_rate": 1.0092838502297453e-05, "loss": 0.005, "step": 32180 }, { "grad_norm": 0.10030163079500198, "learning_rate": 1.0067948087382162e-05, "loss": 0.0035, "step": 32190 }, { "grad_norm": 0.10568553954362869, "learning_rate": 1.0043084965647826e-05, "loss": 0.0042, "step": 32200 }, { "grad_norm": 0.11338215321302414, "learning_rate": 1.001824915408815e-05, "loss": 0.0085, "step": 32210 }, { "grad_norm": 0.16098810732364655, "learning_rate": 9.99344066967815e-06, "loss": 0.0045, "step": 32220 }, { "grad_norm": 0.13260214030742645, "learning_rate": 9.968659529374219e-06, "loss": 0.0055, "step": 32230 }, { "grad_norm": 0.16576944291591644, "learning_rate": 9.943905750114025e-06, "loss": 0.0055, "step": 32240 }, { "grad_norm": 0.13184745609760284, "learning_rate": 9.919179348816559e-06, "loss": 0.0039, "step": 32250 }, { "grad_norm": 0.08803502470254898, "learning_rate": 9.894480342382056e-06, "loss": 0.0068, "step": 32260 }, { "grad_norm": 0.15657739341259003, "learning_rate": 9.869808747692061e-06, "loss": 0.0054, "step": 32270 }, { "grad_norm": 0.17848068475723267, "learning_rate": 9.84516458160939e-06, "loss": 0.0072, "step": 32280 }, { "grad_norm": 0.15464334189891815, "learning_rate": 9.820547860978097e-06, "loss": 0.0062, "step": 32290 }, { "grad_norm": 0.11767511069774628, "learning_rate": 9.795958602623461e-06, "loss": 0.0055, "step": 32300 }, { "grad_norm": 0.16101683676242828, "learning_rate": 9.77139682335203e-06, "loss": 0.0045, "step": 32310 }, { "grad_norm": 0.11118463426828384, "learning_rate": 9.74686253995153e-06, "loss": 0.004, "step": 32320 }, { "grad_norm": 0.09555570036172867, "learning_rate": 9.722355769190911e-06, "loss": 0.0051, "step": 32330 }, { "grad_norm": 0.10516241192817688, "learning_rate": 9.697876527820333e-06, "loss": 0.0043, "step": 32340 }, { "grad_norm": 0.04958585649728775, "learning_rate": 9.673424832571132e-06, "loss": 0.0055, "step": 32350 }, { "grad_norm": 0.10750148445367813, "learning_rate": 9.649000700155791e-06, "loss": 0.0049, "step": 32360 }, { "grad_norm": 0.08006107062101364, "learning_rate": 9.624604147267984e-06, "loss": 0.0051, "step": 32370 }, { "grad_norm": 0.17025591433048248, "learning_rate": 9.600235190582524e-06, "loss": 0.0045, "step": 32380 }, { "grad_norm": 0.15787535905838013, "learning_rate": 9.57589384675538e-06, "loss": 0.007, "step": 32390 }, { "grad_norm": 0.13895903527736664, "learning_rate": 9.551580132423598e-06, "loss": 0.0061, "step": 32400 }, { "grad_norm": 0.14529480040073395, "learning_rate": 9.527294064205405e-06, "loss": 0.0073, "step": 32410 }, { "grad_norm": 0.16590222716331482, "learning_rate": 9.503035658700072e-06, "loss": 0.0052, "step": 32420 }, { "grad_norm": 0.17759554088115692, "learning_rate": 9.478804932488e-06, "loss": 0.0051, "step": 32430 }, { "grad_norm": 0.09911449998617172, "learning_rate": 9.454601902130667e-06, "loss": 0.0045, "step": 32440 }, { "grad_norm": 0.10666514188051224, "learning_rate": 9.430426584170626e-06, "loss": 0.0044, "step": 32450 }, { "grad_norm": 0.12114718556404114, "learning_rate": 9.40627899513145e-06, "loss": 0.0057, "step": 32460 }, { "grad_norm": 0.09793233126401901, "learning_rate": 9.38215915151781e-06, "loss": 0.0065, "step": 32470 }, { "grad_norm": 0.1833406537771225, "learning_rate": 9.358067069815396e-06, "loss": 0.0061, "step": 32480 }, { "grad_norm": 0.14711925387382507, "learning_rate": 9.334002766490901e-06, "loss": 0.0045, "step": 32490 }, { "grad_norm": 0.09480039030313492, "learning_rate": 9.309966257992054e-06, "loss": 0.0036, "step": 32500 }, { "grad_norm": 0.08221329003572464, "learning_rate": 9.2859575607476e-06, "loss": 0.0046, "step": 32510 }, { "grad_norm": 0.17053641378879547, "learning_rate": 9.26197669116723e-06, "loss": 0.0055, "step": 32520 }, { "grad_norm": 0.1307002454996109, "learning_rate": 9.238023665641655e-06, "loss": 0.0051, "step": 32530 }, { "grad_norm": 0.1469285637140274, "learning_rate": 9.21409850054254e-06, "loss": 0.0093, "step": 32540 }, { "grad_norm": 0.13424040377140045, "learning_rate": 9.190201212222526e-06, "loss": 0.0047, "step": 32550 }, { "grad_norm": 0.13593736290931702, "learning_rate": 9.166331817015155e-06, "loss": 0.0044, "step": 32560 }, { "grad_norm": 0.15826277434825897, "learning_rate": 9.142490331234965e-06, "loss": 0.0052, "step": 32570 }, { "grad_norm": 0.1734384298324585, "learning_rate": 9.11867677117736e-06, "loss": 0.0071, "step": 32580 }, { "grad_norm": 0.10311893373727798, "learning_rate": 9.09489115311869e-06, "loss": 0.0084, "step": 32590 }, { "grad_norm": 0.084555983543396, "learning_rate": 9.071133493316202e-06, "loss": 0.0075, "step": 32600 }, { "grad_norm": 0.0978347510099411, "learning_rate": 9.047403808008053e-06, "loss": 0.004, "step": 32610 }, { "grad_norm": 0.1232111006975174, "learning_rate": 9.023702113413224e-06, "loss": 0.0037, "step": 32620 }, { "grad_norm": 0.14481356739997864, "learning_rate": 9.00002842573162e-06, "loss": 0.0059, "step": 32630 }, { "grad_norm": 0.14614881575107574, "learning_rate": 8.976382761143981e-06, "loss": 0.0046, "step": 32640 }, { "grad_norm": 0.13096579909324646, "learning_rate": 8.952765135811909e-06, "loss": 0.0037, "step": 32650 }, { "grad_norm": 0.1585482507944107, "learning_rate": 8.929175565877795e-06, "loss": 0.0054, "step": 32660 }, { "grad_norm": 0.0957646444439888, "learning_rate": 8.905614067464924e-06, "loss": 0.005, "step": 32670 }, { "grad_norm": 0.07693032920360565, "learning_rate": 8.882080656677333e-06, "loss": 0.0067, "step": 32680 }, { "grad_norm": 0.1005522608757019, "learning_rate": 8.85857534959989e-06, "loss": 0.0051, "step": 32690 }, { "grad_norm": 0.1512012630701065, "learning_rate": 8.835098162298256e-06, "loss": 0.0059, "step": 32700 }, { "grad_norm": 0.09608282893896103, "learning_rate": 8.811649110818875e-06, "loss": 0.0034, "step": 32710 }, { "grad_norm": 0.10148421674966812, "learning_rate": 8.78822821118893e-06, "loss": 0.0045, "step": 32720 }, { "grad_norm": 0.13164813816547394, "learning_rate": 8.764835479416405e-06, "loss": 0.0061, "step": 32730 }, { "grad_norm": 0.10969623923301697, "learning_rate": 8.741470931490003e-06, "loss": 0.0036, "step": 32740 }, { "grad_norm": 0.17406441271305084, "learning_rate": 8.718134583379189e-06, "loss": 0.0057, "step": 32750 }, { "grad_norm": 0.09100556373596191, "learning_rate": 8.69482645103412e-06, "loss": 0.0045, "step": 32760 }, { "grad_norm": 0.1620873063802719, "learning_rate": 8.671546550385679e-06, "loss": 0.0058, "step": 32770 }, { "grad_norm": 0.15910902619361877, "learning_rate": 8.648294897345466e-06, "loss": 0.0047, "step": 32780 }, { "grad_norm": 0.14453473687171936, "learning_rate": 8.62507150780576e-06, "loss": 0.0041, "step": 32790 }, { "grad_norm": 0.10824941843748093, "learning_rate": 8.601876397639542e-06, "loss": 0.0043, "step": 32800 }, { "grad_norm": 0.17719988524913788, "learning_rate": 8.578709582700456e-06, "loss": 0.0047, "step": 32810 }, { "grad_norm": 0.12903079390525818, "learning_rate": 8.555571078822783e-06, "loss": 0.0042, "step": 32820 }, { "grad_norm": 0.17760266363620758, "learning_rate": 8.532460901821477e-06, "loss": 0.0071, "step": 32830 }, { "grad_norm": 0.13551394641399384, "learning_rate": 8.509379067492128e-06, "loss": 0.0044, "step": 32840 }, { "grad_norm": 0.10387018322944641, "learning_rate": 8.486325591610967e-06, "loss": 0.004, "step": 32850 }, { "grad_norm": 0.1363234668970108, "learning_rate": 8.463300489934811e-06, "loss": 0.0044, "step": 32860 }, { "grad_norm": 0.07056408375501633, "learning_rate": 8.440303778201093e-06, "loss": 0.0043, "step": 32870 }, { "grad_norm": 0.08637937158346176, "learning_rate": 8.417335472127858e-06, "loss": 0.0065, "step": 32880 }, { "grad_norm": 0.0877026915550232, "learning_rate": 8.394395587413722e-06, "loss": 0.0037, "step": 32890 }, { "grad_norm": 0.08452928811311722, "learning_rate": 8.371484139737889e-06, "loss": 0.0048, "step": 32900 }, { "grad_norm": 0.10324373096227646, "learning_rate": 8.348601144760127e-06, "loss": 0.0034, "step": 32910 }, { "grad_norm": 0.12956708669662476, "learning_rate": 8.325746618120716e-06, "loss": 0.0048, "step": 32920 }, { "grad_norm": 0.11994627863168716, "learning_rate": 8.302920575440531e-06, "loss": 0.0049, "step": 32930 }, { "grad_norm": 0.12125612050294876, "learning_rate": 8.280123032320953e-06, "loss": 0.0066, "step": 32940 }, { "grad_norm": 0.09358123689889908, "learning_rate": 8.257354004343892e-06, "loss": 0.0051, "step": 32950 }, { "grad_norm": 0.10897181183099747, "learning_rate": 8.234613507071754e-06, "loss": 0.0048, "step": 32960 }, { "grad_norm": 0.12014123797416687, "learning_rate": 8.211901556047475e-06, "loss": 0.0042, "step": 32970 }, { "grad_norm": 0.12165801972150803, "learning_rate": 8.189218166794427e-06, "loss": 0.0048, "step": 32980 }, { "grad_norm": 0.08668100088834763, "learning_rate": 8.166563354816509e-06, "loss": 0.0057, "step": 32990 }, { "grad_norm": 0.059870049357414246, "learning_rate": 8.143937135598074e-06, "loss": 0.0054, "step": 33000 }, { "grad_norm": 0.1538211703300476, "learning_rate": 8.121339524603943e-06, "loss": 0.0041, "step": 33010 }, { "grad_norm": 0.08409732580184937, "learning_rate": 8.098770537279338e-06, "loss": 0.0048, "step": 33020 }, { "grad_norm": 0.0889618843793869, "learning_rate": 8.076230189049966e-06, "loss": 0.0066, "step": 33030 }, { "grad_norm": 0.06910098344087601, "learning_rate": 8.05371849532196e-06, "loss": 0.0054, "step": 33040 }, { "grad_norm": 0.1258387714624405, "learning_rate": 8.031235471481813e-06, "loss": 0.0055, "step": 33050 }, { "grad_norm": 0.07846518605947495, "learning_rate": 8.008781132896486e-06, "loss": 0.0054, "step": 33060 }, { "grad_norm": 0.0943121537566185, "learning_rate": 7.986355494913312e-06, "loss": 0.0043, "step": 33070 }, { "grad_norm": 0.08501707762479782, "learning_rate": 7.963958572859975e-06, "loss": 0.0068, "step": 33080 }, { "grad_norm": 0.15984497964382172, "learning_rate": 7.941590382044583e-06, "loss": 0.0066, "step": 33090 }, { "grad_norm": 0.1188863143324852, "learning_rate": 7.919250937755574e-06, "loss": 0.0049, "step": 33100 }, { "grad_norm": 0.08642945438623428, "learning_rate": 7.896940255261764e-06, "loss": 0.0048, "step": 33110 }, { "grad_norm": 0.08854768425226212, "learning_rate": 7.874658349812269e-06, "loss": 0.0047, "step": 33120 }, { "grad_norm": 0.12905184924602509, "learning_rate": 7.85240523663659e-06, "loss": 0.0049, "step": 33130 }, { "grad_norm": 0.23844318091869354, "learning_rate": 7.830180930944487e-06, "loss": 0.0076, "step": 33140 }, { "grad_norm": 0.11044133454561234, "learning_rate": 7.80798544792608e-06, "loss": 0.0058, "step": 33150 }, { "grad_norm": 0.12549830973148346, "learning_rate": 7.785818802751776e-06, "loss": 0.0049, "step": 33160 }, { "grad_norm": 0.1257677674293518, "learning_rate": 7.763681010572277e-06, "loss": 0.0061, "step": 33170 }, { "grad_norm": 0.13521018624305725, "learning_rate": 7.741572086518528e-06, "loss": 0.006, "step": 33180 }, { "grad_norm": 0.16314275562763214, "learning_rate": 7.719492045701782e-06, "loss": 0.0038, "step": 33190 }, { "grad_norm": 0.13149304687976837, "learning_rate": 7.697440903213549e-06, "loss": 0.005, "step": 33200 }, { "grad_norm": 0.12612763047218323, "learning_rate": 7.675418674125573e-06, "loss": 0.0068, "step": 33210 }, { "grad_norm": 0.06345174461603165, "learning_rate": 7.65342537348983e-06, "loss": 0.0048, "step": 33220 }, { "grad_norm": 0.17518404126167297, "learning_rate": 7.631461016338549e-06, "loss": 0.0053, "step": 33230 }, { "grad_norm": 0.13949444890022278, "learning_rate": 7.609525617684138e-06, "loss": 0.0045, "step": 33240 }, { "grad_norm": 0.0793524831533432, "learning_rate": 7.587619192519252e-06, "loss": 0.0047, "step": 33250 }, { "grad_norm": 0.11676632612943649, "learning_rate": 7.565741755816725e-06, "loss": 0.0074, "step": 33260 }, { "grad_norm": 0.11490501463413239, "learning_rate": 7.543893322529588e-06, "loss": 0.0056, "step": 33270 }, { "grad_norm": 0.10609713941812515, "learning_rate": 7.522073907591015e-06, "loss": 0.0054, "step": 33280 }, { "grad_norm": 0.08005775511264801, "learning_rate": 7.500283525914381e-06, "loss": 0.0049, "step": 33290 }, { "grad_norm": 0.13392208516597748, "learning_rate": 7.478522192393211e-06, "loss": 0.0044, "step": 33300 }, { "grad_norm": 0.12900963425636292, "learning_rate": 7.456789921901181e-06, "loss": 0.0049, "step": 33310 }, { "grad_norm": 0.0912073478102684, "learning_rate": 7.435086729292079e-06, "loss": 0.0034, "step": 33320 }, { "grad_norm": 0.1588309109210968, "learning_rate": 7.413412629399824e-06, "loss": 0.0074, "step": 33330 }, { "grad_norm": 0.11755749583244324, "learning_rate": 7.391767637038466e-06, "loss": 0.0047, "step": 33340 }, { "grad_norm": 0.10597342252731323, "learning_rate": 7.370151767002159e-06, "loss": 0.0042, "step": 33350 }, { "grad_norm": 0.15359748899936676, "learning_rate": 7.348565034065147e-06, "loss": 0.0046, "step": 33360 }, { "grad_norm": 0.09716379642486572, "learning_rate": 7.327007452981765e-06, "loss": 0.0051, "step": 33370 }, { "grad_norm": 0.17210014164447784, "learning_rate": 7.305479038486396e-06, "loss": 0.0059, "step": 33380 }, { "grad_norm": 0.11805742233991623, "learning_rate": 7.283979805293517e-06, "loss": 0.0054, "step": 33390 }, { "grad_norm": 0.12324059754610062, "learning_rate": 7.262509768097653e-06, "loss": 0.0055, "step": 33400 }, { "grad_norm": 0.15502367913722992, "learning_rate": 7.241068941573381e-06, "loss": 0.0056, "step": 33410 }, { "grad_norm": 0.14992782473564148, "learning_rate": 7.219657340375296e-06, "loss": 0.004, "step": 33420 }, { "grad_norm": 0.13391658663749695, "learning_rate": 7.198274979138014e-06, "loss": 0.0044, "step": 33430 }, { "grad_norm": 0.1191791221499443, "learning_rate": 7.176921872476189e-06, "loss": 0.009, "step": 33440 }, { "grad_norm": 0.11050005257129669, "learning_rate": 7.1555980349844674e-06, "loss": 0.0039, "step": 33450 }, { "grad_norm": 0.09730647504329681, "learning_rate": 7.134303481237492e-06, "loss": 0.0035, "step": 33460 }, { "grad_norm": 0.1412803679704666, "learning_rate": 7.113038225789892e-06, "loss": 0.0056, "step": 33470 }, { "grad_norm": 0.12271174788475037, "learning_rate": 7.091802283176258e-06, "loss": 0.0069, "step": 33480 }, { "grad_norm": 0.06315620243549347, "learning_rate": 7.070595667911162e-06, "loss": 0.0046, "step": 33490 }, { "grad_norm": 0.06299101561307907, "learning_rate": 7.049418394489127e-06, "loss": 0.004, "step": 33500 }, { "grad_norm": 0.16739913821220398, "learning_rate": 7.028270477384636e-06, "loss": 0.0075, "step": 33510 }, { "grad_norm": 0.09256424009799957, "learning_rate": 7.0071519310520695e-06, "loss": 0.0035, "step": 33520 }, { "grad_norm": 0.11096489429473877, "learning_rate": 6.98606276992575e-06, "loss": 0.0067, "step": 33530 }, { "grad_norm": 0.10250645875930786, "learning_rate": 6.965003008419929e-06, "loss": 0.004, "step": 33540 }, { "grad_norm": 0.10387054830789566, "learning_rate": 6.943972660928755e-06, "loss": 0.0048, "step": 33550 }, { "grad_norm": 0.08915679156780243, "learning_rate": 6.922971741826273e-06, "loss": 0.0047, "step": 33560 }, { "grad_norm": 0.16510076820850372, "learning_rate": 6.902000265466424e-06, "loss": 0.0044, "step": 33570 }, { "grad_norm": 0.1642686128616333, "learning_rate": 6.881058246182998e-06, "loss": 0.0052, "step": 33580 }, { "grad_norm": 0.11209522932767868, "learning_rate": 6.860145698289666e-06, "loss": 0.005, "step": 33590 }, { "grad_norm": 0.0825229361653328, "learning_rate": 6.839262636079985e-06, "loss": 0.004, "step": 33600 }, { "grad_norm": 0.1331264078617096, "learning_rate": 6.818409073827292e-06, "loss": 0.0057, "step": 33610 }, { "grad_norm": 0.06450708210468292, "learning_rate": 6.79758502578483e-06, "loss": 0.0047, "step": 33620 }, { "grad_norm": 0.11397846043109894, "learning_rate": 6.776790506185621e-06, "loss": 0.0054, "step": 33630 }, { "grad_norm": 0.09200803935527802, "learning_rate": 6.756025529242533e-06, "loss": 0.0039, "step": 33640 }, { "grad_norm": 0.19086885452270508, "learning_rate": 6.735290109148223e-06, "loss": 0.0055, "step": 33650 }, { "grad_norm": 0.10476039350032806, "learning_rate": 6.714584260075163e-06, "loss": 0.0047, "step": 33660 }, { "grad_norm": 0.1098833978176117, "learning_rate": 6.693907996175619e-06, "loss": 0.0043, "step": 33670 }, { "grad_norm": 0.11161475628614426, "learning_rate": 6.673261331581593e-06, "loss": 0.0041, "step": 33680 }, { "grad_norm": 0.13376878201961517, "learning_rate": 6.652644280404912e-06, "loss": 0.0062, "step": 33690 }, { "grad_norm": 0.16067646443843842, "learning_rate": 6.6320568567371175e-06, "loss": 0.0036, "step": 33700 }, { "grad_norm": 0.13483066856861115, "learning_rate": 6.611499074649518e-06, "loss": 0.005, "step": 33710 }, { "grad_norm": 0.07381264120340347, "learning_rate": 6.590970948193187e-06, "loss": 0.0047, "step": 33720 }, { "grad_norm": 0.145905539393425, "learning_rate": 6.5704724913988825e-06, "loss": 0.0046, "step": 33730 }, { "grad_norm": 0.11226441711187363, "learning_rate": 6.550003718277109e-06, "loss": 0.0032, "step": 33740 }, { "grad_norm": 0.12114682793617249, "learning_rate": 6.529564642818087e-06, "loss": 0.0056, "step": 33750 }, { "grad_norm": 0.11183162033557892, "learning_rate": 6.50915527899173e-06, "loss": 0.0044, "step": 33760 }, { "grad_norm": 0.084511399269104, "learning_rate": 6.4887756407476575e-06, "loss": 0.0047, "step": 33770 }, { "grad_norm": 0.11510977894067764, "learning_rate": 6.468425742015149e-06, "loss": 0.0052, "step": 33780 }, { "grad_norm": 0.09318853169679642, "learning_rate": 6.448105596703152e-06, "loss": 0.005, "step": 33790 }, { "grad_norm": 0.08561722189188004, "learning_rate": 6.427815218700317e-06, "loss": 0.0054, "step": 33800 }, { "grad_norm": 0.12982109189033508, "learning_rate": 6.407554621874911e-06, "loss": 0.0051, "step": 33810 }, { "grad_norm": 0.0828063115477562, "learning_rate": 6.3873238200748864e-06, "loss": 0.0031, "step": 33820 }, { "grad_norm": 0.09700774401426315, "learning_rate": 6.367122827127775e-06, "loss": 0.0055, "step": 33830 }, { "grad_norm": 0.11540450155735016, "learning_rate": 6.346951656840777e-06, "loss": 0.0038, "step": 33840 }, { "grad_norm": 0.18149146437644958, "learning_rate": 6.326810323000698e-06, "loss": 0.0054, "step": 33850 }, { "grad_norm": 0.11943185329437256, "learning_rate": 6.306698839373948e-06, "loss": 0.0035, "step": 33860 }, { "grad_norm": 0.08964837342500687, "learning_rate": 6.286617219706547e-06, "loss": 0.0039, "step": 33870 }, { "grad_norm": 0.13462571799755096, "learning_rate": 6.26656547772409e-06, "loss": 0.0051, "step": 33880 }, { "grad_norm": 0.17268551886081696, "learning_rate": 6.246543627131734e-06, "loss": 0.0049, "step": 33890 }, { "grad_norm": 0.13628382980823517, "learning_rate": 6.22655168161424e-06, "loss": 0.0061, "step": 33900 }, { "grad_norm": 0.07346891611814499, "learning_rate": 6.20658965483591e-06, "loss": 0.003, "step": 33910 }, { "grad_norm": 0.10774172842502594, "learning_rate": 6.186657560440612e-06, "loss": 0.0049, "step": 33920 }, { "grad_norm": 0.14240187406539917, "learning_rate": 6.166755412051734e-06, "loss": 0.0075, "step": 33930 }, { "grad_norm": 0.08941400051116943, "learning_rate": 6.1468832232721985e-06, "loss": 0.0039, "step": 33940 }, { "grad_norm": 0.09828276187181473, "learning_rate": 6.127041007684476e-06, "loss": 0.0035, "step": 33950 }, { "grad_norm": 0.07621365040540695, "learning_rate": 6.107228778850527e-06, "loss": 0.005, "step": 33960 }, { "grad_norm": 0.1201208084821701, "learning_rate": 6.087446550311837e-06, "loss": 0.006, "step": 33970 }, { "grad_norm": 0.20689044892787933, "learning_rate": 6.067694335589353e-06, "loss": 0.0061, "step": 33980 }, { "grad_norm": 0.17012935876846313, "learning_rate": 6.047972148183528e-06, "loss": 0.0048, "step": 33990 }, { "grad_norm": 0.1065211296081543, "learning_rate": 6.0282800015742936e-06, "loss": 0.0058, "step": 34000 }, { "grad_norm": 0.06807030737400055, "learning_rate": 6.008617909221054e-06, "loss": 0.0039, "step": 34010 }, { "grad_norm": 0.11443083733320236, "learning_rate": 5.988985884562664e-06, "loss": 0.0044, "step": 34020 }, { "grad_norm": 0.07919599860906601, "learning_rate": 5.969383941017414e-06, "loss": 0.0067, "step": 34030 }, { "grad_norm": 0.07791068404912949, "learning_rate": 5.949812091983059e-06, "loss": 0.0043, "step": 34040 }, { "grad_norm": 0.0851743221282959, "learning_rate": 5.9302703508367695e-06, "loss": 0.0048, "step": 34050 }, { "grad_norm": 0.09936099499464035, "learning_rate": 5.910758730935157e-06, "loss": 0.0062, "step": 34060 }, { "grad_norm": 0.1433669626712799, "learning_rate": 5.8912772456142016e-06, "loss": 0.0045, "step": 34070 }, { "grad_norm": 0.10426842421293259, "learning_rate": 5.871825908189338e-06, "loss": 0.0051, "step": 34080 }, { "grad_norm": 0.11822552978992462, "learning_rate": 5.852404731955352e-06, "loss": 0.0053, "step": 34090 }, { "grad_norm": 0.08644860237836838, "learning_rate": 5.833013730186443e-06, "loss": 0.0042, "step": 34100 }, { "grad_norm": 0.12732964754104614, "learning_rate": 5.813652916136181e-06, "loss": 0.0065, "step": 34110 }, { "grad_norm": 0.08037684112787247, "learning_rate": 5.794322303037508e-06, "loss": 0.0051, "step": 34120 }, { "grad_norm": 0.16758009791374207, "learning_rate": 5.775021904102701e-06, "loss": 0.0056, "step": 34130 }, { "grad_norm": 0.15849696099758148, "learning_rate": 5.755751732523401e-06, "loss": 0.0037, "step": 34140 }, { "grad_norm": 0.10926928371191025, "learning_rate": 5.736511801470595e-06, "loss": 0.005, "step": 34150 }, { "grad_norm": 0.09849990904331207, "learning_rate": 5.717302124094603e-06, "loss": 0.004, "step": 34160 }, { "grad_norm": 0.09173113852739334, "learning_rate": 5.6981227135250396e-06, "loss": 0.0043, "step": 34170 }, { "grad_norm": 0.11118234694004059, "learning_rate": 5.678973582870867e-06, "loss": 0.0055, "step": 34180 }, { "grad_norm": 0.1324654370546341, "learning_rate": 5.659854745220317e-06, "loss": 0.0044, "step": 34190 }, { "grad_norm": 0.13308650255203247, "learning_rate": 5.640766213640947e-06, "loss": 0.0036, "step": 34200 }, { "grad_norm": 0.13928331434726715, "learning_rate": 5.6217080011795934e-06, "loss": 0.0037, "step": 34210 }, { "grad_norm": 0.11528155952692032, "learning_rate": 5.602680120862364e-06, "loss": 0.0043, "step": 34220 }, { "grad_norm": 0.19230277836322784, "learning_rate": 5.583682585694616e-06, "loss": 0.0049, "step": 34230 }, { "grad_norm": 0.18856890499591827, "learning_rate": 5.564715408661009e-06, "loss": 0.0053, "step": 34240 }, { "grad_norm": 0.0789102166891098, "learning_rate": 5.5457786027254265e-06, "loss": 0.0043, "step": 34250 }, { "grad_norm": 0.15234464406967163, "learning_rate": 5.526872180830977e-06, "loss": 0.0063, "step": 34260 }, { "grad_norm": 0.1608467400074005, "learning_rate": 5.507996155900041e-06, "loss": 0.0053, "step": 34270 }, { "grad_norm": 0.15052346885204315, "learning_rate": 5.489150540834203e-06, "loss": 0.0052, "step": 34280 }, { "grad_norm": 0.2707231938838959, "learning_rate": 5.47033534851425e-06, "loss": 0.0054, "step": 34290 }, { "grad_norm": 0.14046698808670044, "learning_rate": 5.451550591800198e-06, "loss": 0.0036, "step": 34300 }, { "grad_norm": 0.14264942705631256, "learning_rate": 5.43279628353125e-06, "loss": 0.0034, "step": 34310 }, { "grad_norm": 0.132435604929924, "learning_rate": 5.4140724365258075e-06, "loss": 0.0059, "step": 34320 }, { "grad_norm": 0.07035424560308456, "learning_rate": 5.395379063581424e-06, "loss": 0.0044, "step": 34330 }, { "grad_norm": 0.12009648978710175, "learning_rate": 5.376716177474866e-06, "loss": 0.0041, "step": 34340 }, { "grad_norm": 0.11709900200366974, "learning_rate": 5.358083790962015e-06, "loss": 0.0055, "step": 34350 }, { "grad_norm": 0.14151741564273834, "learning_rate": 5.33948191677795e-06, "loss": 0.0051, "step": 34360 }, { "grad_norm": 0.1415831446647644, "learning_rate": 5.320910567636866e-06, "loss": 0.0039, "step": 34370 }, { "grad_norm": 0.11112584173679352, "learning_rate": 5.302369756232123e-06, "loss": 0.0048, "step": 34380 }, { "grad_norm": 0.2599598169326782, "learning_rate": 5.283859495236165e-06, "loss": 0.0073, "step": 34390 }, { "grad_norm": 0.1185302585363388, "learning_rate": 5.265379797300602e-06, "loss": 0.0035, "step": 34400 }, { "grad_norm": 0.08974748849868774, "learning_rate": 5.246930675056117e-06, "loss": 0.0039, "step": 34410 }, { "grad_norm": 0.09778138250112534, "learning_rate": 5.228512141112535e-06, "loss": 0.0036, "step": 34420 }, { "grad_norm": 0.12982523441314697, "learning_rate": 5.21012420805872e-06, "loss": 0.0046, "step": 34430 }, { "grad_norm": 0.0911635309457779, "learning_rate": 5.191766888462679e-06, "loss": 0.0034, "step": 34440 }, { "grad_norm": 0.11775994300842285, "learning_rate": 5.173440194871437e-06, "loss": 0.005, "step": 34450 }, { "grad_norm": 0.09342534095048904, "learning_rate": 5.155144139811141e-06, "loss": 0.0042, "step": 34460 }, { "grad_norm": 0.0542573407292366, "learning_rate": 5.13687873578696e-06, "loss": 0.0037, "step": 34470 }, { "grad_norm": 0.0923205316066742, "learning_rate": 5.11864399528314e-06, "loss": 0.0048, "step": 34480 }, { "grad_norm": 0.10080615431070328, "learning_rate": 5.1004399307629346e-06, "loss": 0.0043, "step": 34490 }, { "grad_norm": 0.07391030341386795, "learning_rate": 5.082266554668652e-06, "loss": 0.0043, "step": 34500 }, { "grad_norm": 0.11632867157459259, "learning_rate": 5.06412387942164e-06, "loss": 0.0051, "step": 34510 }, { "grad_norm": 0.16400280594825745, "learning_rate": 5.046011917422239e-06, "loss": 0.0054, "step": 34520 }, { "grad_norm": 0.12580889463424683, "learning_rate": 5.0279306810498026e-06, "loss": 0.0046, "step": 34530 }, { "grad_norm": 0.08084551990032196, "learning_rate": 5.009880182662679e-06, "loss": 0.0041, "step": 34540 }, { "grad_norm": 0.08568085730075836, "learning_rate": 4.991860434598217e-06, "loss": 0.0045, "step": 34550 }, { "grad_norm": 0.11348602920770645, "learning_rate": 4.973871449172746e-06, "loss": 0.0043, "step": 34560 }, { "grad_norm": 0.10379137098789215, "learning_rate": 4.9559132386815695e-06, "loss": 0.0045, "step": 34570 }, { "grad_norm": 0.11038627475500107, "learning_rate": 4.93798581539896e-06, "loss": 0.0061, "step": 34580 }, { "grad_norm": 0.1230790987610817, "learning_rate": 4.920089191578131e-06, "loss": 0.0045, "step": 34590 }, { "grad_norm": 0.184021458029747, "learning_rate": 4.90222337945126e-06, "loss": 0.0038, "step": 34600 }, { "grad_norm": 0.04903575778007507, "learning_rate": 4.8843883912294605e-06, "loss": 0.0053, "step": 34610 }, { "grad_norm": 0.12203627079725266, "learning_rate": 4.866584239102795e-06, "loss": 0.0056, "step": 34620 }, { "grad_norm": 0.09511405974626541, "learning_rate": 4.8488109352402164e-06, "loss": 0.0062, "step": 34630 }, { "grad_norm": 0.08801677823066711, "learning_rate": 4.831068491789603e-06, "loss": 0.0055, "step": 34640 }, { "grad_norm": 0.17952102422714233, "learning_rate": 4.81335692087776e-06, "loss": 0.0069, "step": 34650 }, { "grad_norm": 0.11433185636997223, "learning_rate": 4.795676234610374e-06, "loss": 0.0035, "step": 34660 }, { "grad_norm": 0.11232054978609085, "learning_rate": 4.778026445072037e-06, "loss": 0.0042, "step": 34670 }, { "grad_norm": 0.0905328169465065, "learning_rate": 4.760407564326208e-06, "loss": 0.0054, "step": 34680 }, { "grad_norm": 0.11653953790664673, "learning_rate": 4.742819604415222e-06, "loss": 0.0041, "step": 34690 }, { "grad_norm": 0.13263165950775146, "learning_rate": 4.725262577360284e-06, "loss": 0.0054, "step": 34700 }, { "grad_norm": 0.120051808655262, "learning_rate": 4.707736495161458e-06, "loss": 0.0047, "step": 34710 }, { "grad_norm": 0.1660643070936203, "learning_rate": 4.690241369797671e-06, "loss": 0.0049, "step": 34720 }, { "grad_norm": 0.1267959326505661, "learning_rate": 4.672777213226659e-06, "loss": 0.0044, "step": 34730 }, { "grad_norm": 0.10288127511739731, "learning_rate": 4.655344037385001e-06, "loss": 0.006, "step": 34740 }, { "grad_norm": 0.09816370904445648, "learning_rate": 4.6379418541881215e-06, "loss": 0.0051, "step": 34750 }, { "grad_norm": 0.07653448730707169, "learning_rate": 4.620570675530245e-06, "loss": 0.0037, "step": 34760 }, { "grad_norm": 0.1329886019229889, "learning_rate": 4.603230513284407e-06, "loss": 0.0044, "step": 34770 }, { "grad_norm": 0.07173751294612885, "learning_rate": 4.58592137930246e-06, "loss": 0.0038, "step": 34780 }, { "grad_norm": 0.14739100635051727, "learning_rate": 4.568643285415009e-06, "loss": 0.004, "step": 34790 }, { "grad_norm": 0.09701795130968094, "learning_rate": 4.551396243431483e-06, "loss": 0.0036, "step": 34800 }, { "grad_norm": 0.15098589658737183, "learning_rate": 4.534180265140075e-06, "loss": 0.0052, "step": 34810 }, { "grad_norm": 0.26877135038375854, "learning_rate": 4.5169953623077356e-06, "loss": 0.004, "step": 34820 }, { "grad_norm": 0.09490196406841278, "learning_rate": 4.499841546680189e-06, "loss": 0.0058, "step": 34830 }, { "grad_norm": 0.12705445289611816, "learning_rate": 4.4827188299819025e-06, "loss": 0.0039, "step": 34840 }, { "grad_norm": 0.07817655801773071, "learning_rate": 4.465627223916086e-06, "loss": 0.0037, "step": 34850 }, { "grad_norm": 0.08476079255342484, "learning_rate": 4.448566740164706e-06, "loss": 0.0036, "step": 34860 }, { "grad_norm": 0.08462618291378021, "learning_rate": 4.431537390388429e-06, "loss": 0.0066, "step": 34870 }, { "grad_norm": 0.07071801275014877, "learning_rate": 4.414539186226674e-06, "loss": 0.004, "step": 34880 }, { "grad_norm": 0.07938460260629654, "learning_rate": 4.397572139297529e-06, "loss": 0.0052, "step": 34890 }, { "grad_norm": 0.08012641966342926, "learning_rate": 4.380636261197829e-06, "loss": 0.0033, "step": 34900 }, { "grad_norm": 0.13619233667850494, "learning_rate": 4.36373156350307e-06, "loss": 0.0041, "step": 34910 }, { "grad_norm": 0.12560714781284332, "learning_rate": 4.3468580577674665e-06, "loss": 0.0038, "step": 34920 }, { "grad_norm": 0.10630500316619873, "learning_rate": 4.330015755523908e-06, "loss": 0.0044, "step": 34930 }, { "grad_norm": 0.09850698709487915, "learning_rate": 4.313204668283926e-06, "loss": 0.0058, "step": 34940 }, { "grad_norm": 0.1325218379497528, "learning_rate": 4.296424807537752e-06, "loss": 0.0042, "step": 34950 }, { "grad_norm": 0.14138007164001465, "learning_rate": 4.279676184754267e-06, "loss": 0.0043, "step": 34960 }, { "grad_norm": 0.08204315602779388, "learning_rate": 4.262958811380985e-06, "loss": 0.004, "step": 34970 }, { "grad_norm": 0.1504564881324768, "learning_rate": 4.246272698844095e-06, "loss": 0.0069, "step": 34980 }, { "grad_norm": 0.10977467149496078, "learning_rate": 4.229617858548374e-06, "loss": 0.0045, "step": 34990 }, { "grad_norm": 0.10450221598148346, "learning_rate": 4.212994301877271e-06, "loss": 0.0045, "step": 35000 }, { "grad_norm": 0.13981597125530243, "learning_rate": 4.196402040192804e-06, "loss": 0.0083, "step": 35010 }, { "grad_norm": 0.15631428360939026, "learning_rate": 4.179841084835639e-06, "loss": 0.0047, "step": 35020 }, { "grad_norm": 0.1279870867729187, "learning_rate": 4.163311447125046e-06, "loss": 0.0053, "step": 35030 }, { "grad_norm": 0.09445207566022873, "learning_rate": 4.146813138358851e-06, "loss": 0.0047, "step": 35040 }, { "grad_norm": 0.12059291452169418, "learning_rate": 4.130346169813509e-06, "loss": 0.0041, "step": 35050 }, { "grad_norm": 0.09990250319242477, "learning_rate": 4.113910552744038e-06, "loss": 0.0037, "step": 35060 }, { "grad_norm": 0.207473024725914, "learning_rate": 4.097506298384024e-06, "loss": 0.0053, "step": 35070 }, { "grad_norm": 0.11854177713394165, "learning_rate": 4.081133417945626e-06, "loss": 0.0084, "step": 35080 }, { "grad_norm": 0.18300406634807587, "learning_rate": 4.0647919226195484e-06, "loss": 0.0048, "step": 35090 }, { "grad_norm": 0.19320467114448547, "learning_rate": 4.04848182357504e-06, "loss": 0.0044, "step": 35100 }, { "grad_norm": 0.11782951653003693, "learning_rate": 4.032203131959911e-06, "loss": 0.0078, "step": 35110 }, { "grad_norm": 0.18443754315376282, "learning_rate": 4.015955858900483e-06, "loss": 0.0049, "step": 35120 }, { "grad_norm": 0.0764031857252121, "learning_rate": 3.999740015501641e-06, "loss": 0.0042, "step": 35130 }, { "grad_norm": 0.07626740634441376, "learning_rate": 3.9835556128467236e-06, "loss": 0.0035, "step": 35140 }, { "grad_norm": 0.0969700813293457, "learning_rate": 3.967402661997644e-06, "loss": 0.0036, "step": 35150 }, { "grad_norm": 0.13844192028045654, "learning_rate": 3.951281173994781e-06, "loss": 0.0038, "step": 35160 }, { "grad_norm": 0.11326711624860764, "learning_rate": 3.935191159857016e-06, "loss": 0.0039, "step": 35170 }, { "grad_norm": 0.06856437027454376, "learning_rate": 3.919132630581746e-06, "loss": 0.0061, "step": 35180 }, { "grad_norm": 0.10690998286008835, "learning_rate": 3.903105597144802e-06, "loss": 0.0059, "step": 35190 }, { "grad_norm": 0.11153792589902878, "learning_rate": 3.887110070500505e-06, "loss": 0.0077, "step": 35200 }, { "grad_norm": 0.08482632040977478, "learning_rate": 3.871146061581654e-06, "loss": 0.004, "step": 35210 }, { "grad_norm": 0.16460488736629486, "learning_rate": 3.855213581299505e-06, "loss": 0.0038, "step": 35220 }, { "grad_norm": 0.11069052666425705, "learning_rate": 3.839312640543763e-06, "loss": 0.0047, "step": 35230 }, { "grad_norm": 0.14676107466220856, "learning_rate": 3.823443250182551e-06, "loss": 0.0046, "step": 35240 }, { "grad_norm": 0.06383409351110458, "learning_rate": 3.80760542106246e-06, "loss": 0.0048, "step": 35250 }, { "grad_norm": 0.06977903097867966, "learning_rate": 3.7917991640084972e-06, "loss": 0.0044, "step": 35260 }, { "grad_norm": 0.12870782613754272, "learning_rate": 3.7760244898240925e-06, "loss": 0.0041, "step": 35270 }, { "grad_norm": 0.08651324361562729, "learning_rate": 3.760281409291078e-06, "loss": 0.0033, "step": 35280 }, { "grad_norm": 0.07982809096574783, "learning_rate": 3.7445699331697113e-06, "loss": 0.0047, "step": 35290 }, { "grad_norm": 0.09411156922578812, "learning_rate": 3.7288900721986262e-06, "loss": 0.005, "step": 35300 }, { "grad_norm": 0.16679877042770386, "learning_rate": 3.7132418370948663e-06, "loss": 0.0043, "step": 35310 }, { "grad_norm": 0.1624399572610855, "learning_rate": 3.6976252385538535e-06, "loss": 0.005, "step": 35320 }, { "grad_norm": 0.15947985649108887, "learning_rate": 3.6820402872493832e-06, "loss": 0.0055, "step": 35330 }, { "grad_norm": 0.07556191086769104, "learning_rate": 3.6664869938336343e-06, "loss": 0.005, "step": 35340 }, { "grad_norm": 0.14018021523952484, "learning_rate": 3.650965368937115e-06, "loss": 0.005, "step": 35350 }, { "grad_norm": 0.09416328370571136, "learning_rate": 3.6354754231687214e-06, "loss": 0.0035, "step": 35360 }, { "grad_norm": 0.07682691514492035, "learning_rate": 3.620017167115697e-06, "loss": 0.0077, "step": 35370 }, { "grad_norm": 0.10946854203939438, "learning_rate": 3.604590611343589e-06, "loss": 0.0034, "step": 35380 }, { "grad_norm": 0.10065441578626633, "learning_rate": 3.589195766396325e-06, "loss": 0.005, "step": 35390 }, { "grad_norm": 0.21337327361106873, "learning_rate": 3.573832642796121e-06, "loss": 0.005, "step": 35400 }, { "grad_norm": 0.08886122703552246, "learning_rate": 3.5585012510435336e-06, "loss": 0.0041, "step": 35410 }, { "grad_norm": 0.20000170171260834, "learning_rate": 3.5432016016174184e-06, "loss": 0.0056, "step": 35420 }, { "grad_norm": 0.05506390333175659, "learning_rate": 3.5279337049749506e-06, "loss": 0.0045, "step": 35430 }, { "grad_norm": 0.13723047077655792, "learning_rate": 3.5126975715515985e-06, "loss": 0.0076, "step": 35440 }, { "grad_norm": 0.11354485899209976, "learning_rate": 3.4974932117610994e-06, "loss": 0.005, "step": 35450 }, { "grad_norm": 0.07871881872415543, "learning_rate": 3.4823206359955065e-06, "loss": 0.0061, "step": 35460 }, { "grad_norm": 0.11929038912057877, "learning_rate": 3.4671798546251133e-06, "loss": 0.006, "step": 35470 }, { "grad_norm": 0.09699244052171707, "learning_rate": 3.4520708779985088e-06, "loss": 0.005, "step": 35480 }, { "grad_norm": 0.08749765157699585, "learning_rate": 3.4369937164425493e-06, "loss": 0.0032, "step": 35490 }, { "grad_norm": 0.06486394256353378, "learning_rate": 3.421948380262313e-06, "loss": 0.0061, "step": 35500 }, { "grad_norm": 0.10376526415348053, "learning_rate": 3.406934879741158e-06, "loss": 0.0075, "step": 35510 }, { "grad_norm": 0.07596731185913086, "learning_rate": 3.391953225140665e-06, "loss": 0.0052, "step": 35520 }, { "grad_norm": 0.1283978521823883, "learning_rate": 3.3770034267006533e-06, "loss": 0.0055, "step": 35530 }, { "grad_norm": 0.1273457556962967, "learning_rate": 3.362085494639178e-06, "loss": 0.0053, "step": 35540 }, { "grad_norm": 0.1618768721818924, "learning_rate": 3.3471994391524986e-06, "loss": 0.0037, "step": 35550 }, { "grad_norm": 0.05062861740589142, "learning_rate": 3.3323452704150825e-06, "loss": 0.0047, "step": 35560 }, { "grad_norm": 0.11782445758581161, "learning_rate": 3.3175229985796253e-06, "loss": 0.0044, "step": 35570 }, { "grad_norm": 0.11645174026489258, "learning_rate": 3.3027326337770115e-06, "loss": 0.0039, "step": 35580 }, { "grad_norm": 0.21009889245033264, "learning_rate": 3.2879741861163216e-06, "loss": 0.0044, "step": 35590 }, { "grad_norm": 0.076136015355587, "learning_rate": 3.2732476656847976e-06, "loss": 0.004, "step": 35600 }, { "grad_norm": 0.12457643449306488, "learning_rate": 3.2585530825478883e-06, "loss": 0.0046, "step": 35610 }, { "grad_norm": 0.12375102937221527, "learning_rate": 3.243890446749204e-06, "loss": 0.0056, "step": 35620 }, { "grad_norm": 0.15315528213977814, "learning_rate": 3.2292597683105107e-06, "loss": 0.0065, "step": 35630 }, { "grad_norm": 0.20842120051383972, "learning_rate": 3.214661057231755e-06, "loss": 0.0068, "step": 35640 }, { "grad_norm": 0.1746738851070404, "learning_rate": 3.2000943234910097e-06, "loss": 0.0046, "step": 35650 }, { "grad_norm": 0.08623704314231873, "learning_rate": 3.1855595770444835e-06, "loss": 0.0046, "step": 35660 }, { "grad_norm": 0.06780292093753815, "learning_rate": 3.1710568278265574e-06, "loss": 0.0048, "step": 35670 }, { "grad_norm": 0.13154341280460358, "learning_rate": 3.156586085749719e-06, "loss": 0.005, "step": 35680 }, { "grad_norm": 0.1886025071144104, "learning_rate": 3.1421473607045893e-06, "loss": 0.0063, "step": 35690 }, { "grad_norm": 0.09757710248231888, "learning_rate": 3.127740662559886e-06, "loss": 0.005, "step": 35700 }, { "grad_norm": 0.09783927351236343, "learning_rate": 3.1133660011624655e-06, "loss": 0.0039, "step": 35710 }, { "grad_norm": 0.09166249632835388, "learning_rate": 3.099023386337263e-06, "loss": 0.007, "step": 35720 }, { "grad_norm": 0.1335921436548233, "learning_rate": 3.0847128278873317e-06, "loss": 0.0052, "step": 35730 }, { "grad_norm": 0.12222278118133545, "learning_rate": 3.0704343355938027e-06, "loss": 0.0042, "step": 35740 }, { "grad_norm": 0.14979994297027588, "learning_rate": 3.056187919215886e-06, "loss": 0.0044, "step": 35750 }, { "grad_norm": 0.12402854114770889, "learning_rate": 3.041973588490865e-06, "loss": 0.0039, "step": 35760 }, { "grad_norm": 0.09379731863737106, "learning_rate": 3.0277913531341128e-06, "loss": 0.0053, "step": 35770 }, { "grad_norm": 0.08624675124883652, "learning_rate": 3.0136412228390476e-06, "loss": 0.0031, "step": 35780 }, { "grad_norm": 0.0827358067035675, "learning_rate": 2.999523207277155e-06, "loss": 0.0038, "step": 35790 }, { "grad_norm": 0.13197356462478638, "learning_rate": 2.9854373160979552e-06, "loss": 0.0052, "step": 35800 }, { "grad_norm": 0.10067587345838547, "learning_rate": 2.9713835589290362e-06, "loss": 0.0059, "step": 35810 }, { "grad_norm": 0.13812105357646942, "learning_rate": 2.9573619453759913e-06, "loss": 0.0037, "step": 35820 }, { "grad_norm": 0.08482318371534348, "learning_rate": 2.943372485022489e-06, "loss": 0.0041, "step": 35830 }, { "grad_norm": 0.17205646634101868, "learning_rate": 2.9294151874301623e-06, "loss": 0.0049, "step": 35840 }, { "grad_norm": 0.07378888875246048, "learning_rate": 2.9154900621387093e-06, "loss": 0.0037, "step": 35850 }, { "grad_norm": 0.11135890334844589, "learning_rate": 2.9015971186658164e-06, "loss": 0.003, "step": 35860 }, { "grad_norm": 0.10897106677293777, "learning_rate": 2.887736366507182e-06, "loss": 0.0068, "step": 35870 }, { "grad_norm": 0.13499580323696136, "learning_rate": 2.873907815136501e-06, "loss": 0.0032, "step": 35880 }, { "grad_norm": 0.113094761967659, "learning_rate": 2.8601114740054556e-06, "loss": 0.0039, "step": 35890 }, { "grad_norm": 0.10833685845136642, "learning_rate": 2.8463473525437155e-06, "loss": 0.0043, "step": 35900 }, { "grad_norm": 0.14855515956878662, "learning_rate": 2.8326154601589216e-06, "loss": 0.0072, "step": 35910 }, { "grad_norm": 0.10961218923330307, "learning_rate": 2.818915806236705e-06, "loss": 0.003, "step": 35920 }, { "grad_norm": 0.10878510028123856, "learning_rate": 2.8052484001406478e-06, "loss": 0.0039, "step": 35930 }, { "grad_norm": 0.08109720051288605, "learning_rate": 2.791613251212277e-06, "loss": 0.0038, "step": 35940 }, { "grad_norm": 0.06443909555673599, "learning_rate": 2.7780103687711166e-06, "loss": 0.0048, "step": 35950 }, { "grad_norm": 0.09467297047376633, "learning_rate": 2.764439762114579e-06, "loss": 0.0035, "step": 35960 }, { "grad_norm": 0.1649504452943802, "learning_rate": 2.750901440518061e-06, "loss": 0.0049, "step": 35970 }, { "grad_norm": 0.0937161073088646, "learning_rate": 2.7373954132348776e-06, "loss": 0.0051, "step": 35980 }, { "grad_norm": 0.1243826150894165, "learning_rate": 2.7239216894962784e-06, "loss": 0.005, "step": 35990 }, { "grad_norm": 0.0754496306180954, "learning_rate": 2.710480278511407e-06, "loss": 0.006, "step": 36000 }, { "grad_norm": 0.09357879310846329, "learning_rate": 2.6970711894673483e-06, "loss": 0.004, "step": 36010 }, { "grad_norm": 0.1334192007780075, "learning_rate": 2.6836944315291047e-06, "loss": 0.0037, "step": 36020 }, { "grad_norm": 0.11008691787719727, "learning_rate": 2.67035001383954e-06, "loss": 0.0071, "step": 36030 }, { "grad_norm": 0.05868982896208763, "learning_rate": 2.6570379455194527e-06, "loss": 0.0055, "step": 36040 }, { "grad_norm": 0.07468260824680328, "learning_rate": 2.6437582356675206e-06, "loss": 0.004, "step": 36050 }, { "grad_norm": 0.05711318925023079, "learning_rate": 2.6305108933602775e-06, "loss": 0.0051, "step": 36060 }, { "grad_norm": 0.1148805096745491, "learning_rate": 2.6172959276521803e-06, "loss": 0.0048, "step": 36070 }, { "grad_norm": 0.13096219301223755, "learning_rate": 2.6041133475755153e-06, "loss": 0.0051, "step": 36080 }, { "grad_norm": 0.12265970557928085, "learning_rate": 2.590963162140475e-06, "loss": 0.0052, "step": 36090 }, { "grad_norm": 0.08351477235555649, "learning_rate": 2.5778453803350587e-06, "loss": 0.0066, "step": 36100 }, { "grad_norm": 0.05298015847802162, "learning_rate": 2.5647600111251723e-06, "loss": 0.0044, "step": 36110 }, { "grad_norm": 0.07731904089450836, "learning_rate": 2.551707063454528e-06, "loss": 0.0047, "step": 36120 }, { "grad_norm": 0.11311899125576019, "learning_rate": 2.5386865462446897e-06, "loss": 0.0043, "step": 36130 }, { "grad_norm": 0.04252247139811516, "learning_rate": 2.52569846839506e-06, "loss": 0.0033, "step": 36140 }, { "grad_norm": 0.12981542944908142, "learning_rate": 2.512742838782883e-06, "loss": 0.0044, "step": 36150 }, { "grad_norm": 0.0844641849398613, "learning_rate": 2.4998196662631924e-06, "loss": 0.0037, "step": 36160 }, { "grad_norm": 0.056174300611019135, "learning_rate": 2.4869289596688604e-06, "loss": 0.0057, "step": 36170 }, { "grad_norm": 0.06325317174196243, "learning_rate": 2.4740707278105567e-06, "loss": 0.004, "step": 36180 }, { "grad_norm": 0.11231549084186554, "learning_rate": 2.461244979476779e-06, "loss": 0.0038, "step": 36190 }, { "grad_norm": 0.08330222964286804, "learning_rate": 2.4484517234337866e-06, "loss": 0.0037, "step": 36200 }, { "grad_norm": 0.09436263889074326, "learning_rate": 2.4356909684256634e-06, "loss": 0.0035, "step": 36210 }, { "grad_norm": 0.06044640764594078, "learning_rate": 2.4229627231742545e-06, "loss": 0.0037, "step": 36220 }, { "grad_norm": 0.09760399162769318, "learning_rate": 2.4102669963791956e-06, "loss": 0.0059, "step": 36230 }, { "grad_norm": 0.18275633454322815, "learning_rate": 2.3976037967179065e-06, "loss": 0.0043, "step": 36240 }, { "grad_norm": 0.04297090321779251, "learning_rate": 2.3849731328455582e-06, "loss": 0.0041, "step": 36250 }, { "grad_norm": 0.15459564328193665, "learning_rate": 2.3723750133950894e-06, "loss": 0.0045, "step": 36260 }, { "grad_norm": 0.059206247329711914, "learning_rate": 2.3598094469771957e-06, "loss": 0.0045, "step": 36270 }, { "grad_norm": 0.1203082799911499, "learning_rate": 2.347276442180324e-06, "loss": 0.0047, "step": 36280 }, { "grad_norm": 0.1120951920747757, "learning_rate": 2.3347760075706715e-06, "loss": 0.0043, "step": 36290 }, { "grad_norm": 0.12357941269874573, "learning_rate": 2.3223081516921595e-06, "loss": 0.0036, "step": 36300 }, { "grad_norm": 0.14096878468990326, "learning_rate": 2.3098728830664494e-06, "loss": 0.0039, "step": 36310 }, { "grad_norm": 0.05705545097589493, "learning_rate": 2.2974702101929314e-06, "loss": 0.0037, "step": 36320 }, { "grad_norm": 0.11060120910406113, "learning_rate": 2.2851001415487185e-06, "loss": 0.0063, "step": 36330 }, { "grad_norm": 0.197111114859581, "learning_rate": 2.272762685588631e-06, "loss": 0.0054, "step": 36340 }, { "grad_norm": 0.08376791328191757, "learning_rate": 2.260457850745218e-06, "loss": 0.0042, "step": 36350 }, { "grad_norm": 0.06770630925893784, "learning_rate": 2.2481856454286963e-06, "loss": 0.0034, "step": 36360 }, { "grad_norm": 0.14049479365348816, "learning_rate": 2.235946078027018e-06, "loss": 0.0047, "step": 36370 }, { "grad_norm": 0.13744154572486877, "learning_rate": 2.223739156905802e-06, "loss": 0.0048, "step": 36380 }, { "grad_norm": 0.15911760926246643, "learning_rate": 2.2115648904083796e-06, "loss": 0.0069, "step": 36390 }, { "grad_norm": 0.11309318989515305, "learning_rate": 2.199423286855734e-06, "loss": 0.0036, "step": 36400 }, { "grad_norm": 0.1529247760772705, "learning_rate": 2.1873143545465323e-06, "loss": 0.0043, "step": 36410 }, { "grad_norm": 0.17883995175361633, "learning_rate": 2.1752381017571267e-06, "loss": 0.0039, "step": 36420 }, { "grad_norm": 0.0830119252204895, "learning_rate": 2.1631945367415142e-06, "loss": 0.0048, "step": 36430 }, { "grad_norm": 0.08594987541437149, "learning_rate": 2.15118366773136e-06, "loss": 0.0041, "step": 36440 }, { "grad_norm": 0.172734797000885, "learning_rate": 2.1392055029359926e-06, "loss": 0.0041, "step": 36450 }, { "grad_norm": 0.13854128122329712, "learning_rate": 2.127260050542357e-06, "loss": 0.0039, "step": 36460 }, { "grad_norm": 0.13064046204090118, "learning_rate": 2.1153473187150608e-06, "loss": 0.0042, "step": 36470 }, { "grad_norm": 0.09800492227077484, "learning_rate": 2.1034673155963467e-06, "loss": 0.004, "step": 36480 }, { "grad_norm": 0.11974498629570007, "learning_rate": 2.0916200493060976e-06, "loss": 0.0041, "step": 36490 }, { "grad_norm": 0.14650046825408936, "learning_rate": 2.0798055279417916e-06, "loss": 0.0041, "step": 36500 }, { "grad_norm": 0.0953458920121193, "learning_rate": 2.0680237595785356e-06, "loss": 0.0029, "step": 36510 }, { "grad_norm": 0.04210381954908371, "learning_rate": 2.0562747522690774e-06, "loss": 0.003, "step": 36520 }, { "grad_norm": 0.05211317539215088, "learning_rate": 2.0445585140437373e-06, "loss": 0.0033, "step": 36530 }, { "grad_norm": 0.13119874894618988, "learning_rate": 2.0328750529104602e-06, "loss": 0.0051, "step": 36540 }, { "grad_norm": 0.1469733566045761, "learning_rate": 2.021224376854791e-06, "loss": 0.004, "step": 36550 }, { "grad_norm": 0.05667876452207565, "learning_rate": 2.0096064938398386e-06, "loss": 0.0032, "step": 36560 }, { "grad_norm": 0.10322444140911102, "learning_rate": 1.998021411806328e-06, "loss": 0.0037, "step": 36570 }, { "grad_norm": 0.11567311733961105, "learning_rate": 1.9864691386725586e-06, "loss": 0.0063, "step": 36580 }, { "grad_norm": 0.13503201305866241, "learning_rate": 1.9749496823343814e-06, "loss": 0.0055, "step": 36590 }, { "grad_norm": 0.11576809734106064, "learning_rate": 1.963463050665254e-06, "loss": 0.0046, "step": 36600 }, { "grad_norm": 0.12321760505437851, "learning_rate": 1.9520092515161682e-06, "loss": 0.0055, "step": 36610 }, { "grad_norm": 0.049010682851076126, "learning_rate": 1.940588292715695e-06, "loss": 0.0049, "step": 36620 }, { "grad_norm": 0.0918067917227745, "learning_rate": 1.929200182069951e-06, "loss": 0.0033, "step": 36630 }, { "grad_norm": 0.11673735082149506, "learning_rate": 1.9178449273625997e-06, "loss": 0.0041, "step": 36640 }, { "grad_norm": 0.09901941567659378, "learning_rate": 1.90652253635486e-06, "loss": 0.0062, "step": 36650 }, { "grad_norm": 0.11020391434431076, "learning_rate": 1.8952330167854693e-06, "loss": 0.003, "step": 36660 }, { "grad_norm": 0.11400393396615982, "learning_rate": 1.8839763763707107e-06, "loss": 0.0039, "step": 36670 }, { "grad_norm": 0.08733084052801132, "learning_rate": 1.8727526228043912e-06, "loss": 0.0042, "step": 36680 }, { "grad_norm": 0.08484763652086258, "learning_rate": 1.8615617637578408e-06, "loss": 0.005, "step": 36690 }, { "grad_norm": 0.06250357627868652, "learning_rate": 1.850403806879908e-06, "loss": 0.0043, "step": 36700 }, { "grad_norm": 0.07423382252454758, "learning_rate": 1.8392787597969475e-06, "loss": 0.0042, "step": 36710 }, { "grad_norm": 0.11211419850587845, "learning_rate": 1.828186630112827e-06, "loss": 0.0051, "step": 36720 }, { "grad_norm": 0.0551023930311203, "learning_rate": 1.8171274254089098e-06, "loss": 0.0035, "step": 36730 }, { "grad_norm": 0.10951977968215942, "learning_rate": 1.806101153244061e-06, "loss": 0.0037, "step": 36740 }, { "grad_norm": 0.05320459604263306, "learning_rate": 1.7951078211546401e-06, "loss": 0.0062, "step": 36750 }, { "grad_norm": 0.08427160233259201, "learning_rate": 1.7841474366544764e-06, "loss": 0.0056, "step": 36760 }, { "grad_norm": 0.05615941062569618, "learning_rate": 1.7732200072348825e-06, "loss": 0.0025, "step": 36770 }, { "grad_norm": 0.08564957231283188, "learning_rate": 1.7623255403646677e-06, "loss": 0.004, "step": 36780 }, { "grad_norm": 0.10736817866563797, "learning_rate": 1.7514640434900865e-06, "loss": 0.0045, "step": 36790 }, { "grad_norm": 0.14349091053009033, "learning_rate": 1.7406355240348838e-06, "loss": 0.0052, "step": 36800 }, { "grad_norm": 0.05427663400769234, "learning_rate": 1.7298399894002392e-06, "loss": 0.0034, "step": 36810 }, { "grad_norm": 0.13388285040855408, "learning_rate": 1.7190774469647952e-06, "loss": 0.0051, "step": 36820 }, { "grad_norm": 0.193474680185318, "learning_rate": 1.7083479040846617e-06, "loss": 0.0063, "step": 36830 }, { "grad_norm": 0.14598165452480316, "learning_rate": 1.697651368093367e-06, "loss": 0.004, "step": 36840 }, { "grad_norm": 0.10731293261051178, "learning_rate": 1.6869878463019017e-06, "loss": 0.0057, "step": 36850 }, { "grad_norm": 0.10935071855783463, "learning_rate": 1.6763573459986858e-06, "loss": 0.0054, "step": 36860 }, { "grad_norm": 0.07871192693710327, "learning_rate": 1.6657598744495516e-06, "loss": 0.0044, "step": 36870 }, { "grad_norm": 0.033566564321517944, "learning_rate": 1.6551954388977774e-06, "loss": 0.0044, "step": 36880 }, { "grad_norm": 0.14016328752040863, "learning_rate": 1.644664046564054e-06, "loss": 0.0034, "step": 36890 }, { "grad_norm": 0.1344791054725647, "learning_rate": 1.634165704646501e-06, "loss": 0.0038, "step": 36900 }, { "grad_norm": 0.15467724204063416, "learning_rate": 1.6237004203206175e-06, "loss": 0.0077, "step": 36910 }, { "grad_norm": 0.06716378033161163, "learning_rate": 1.6132682007393373e-06, "loss": 0.004, "step": 36920 }, { "grad_norm": 0.05275420472025871, "learning_rate": 1.6028690530329847e-06, "loss": 0.0033, "step": 36930 }, { "grad_norm": 0.14317867159843445, "learning_rate": 1.5925029843092742e-06, "loss": 0.0056, "step": 36940 }, { "grad_norm": 0.16556763648986816, "learning_rate": 1.5821700016533214e-06, "loss": 0.007, "step": 36950 }, { "grad_norm": 0.050935257226228714, "learning_rate": 1.5718701121276268e-06, "loss": 0.0076, "step": 36960 }, { "grad_norm": 0.09845215827226639, "learning_rate": 1.5616033227720483e-06, "loss": 0.0047, "step": 36970 }, { "grad_norm": 0.06096182391047478, "learning_rate": 1.5513696406038558e-06, "loss": 0.0037, "step": 36980 }, { "grad_norm": 0.16769573092460632, "learning_rate": 1.5411690726176653e-06, "loss": 0.0035, "step": 36990 }, { "grad_norm": 0.06821444630622864, "learning_rate": 1.5310016257854832e-06, "loss": 0.0057, "step": 37000 }, { "grad_norm": 0.12987563014030457, "learning_rate": 1.5208673070566504e-06, "loss": 0.0044, "step": 37010 }, { "grad_norm": 0.08534862846136093, "learning_rate": 1.5107661233578818e-06, "loss": 0.0033, "step": 37020 }, { "grad_norm": 0.14834456145763397, "learning_rate": 1.5006980815932437e-06, "loss": 0.0042, "step": 37030 }, { "grad_norm": 0.07516218721866608, "learning_rate": 1.490663188644148e-06, "loss": 0.0061, "step": 37040 }, { "grad_norm": 0.12365204095840454, "learning_rate": 1.4806614513693418e-06, "loss": 0.0044, "step": 37050 }, { "grad_norm": 0.12049157917499542, "learning_rate": 1.4706928766049289e-06, "loss": 0.0055, "step": 37060 }, { "grad_norm": 0.06882105767726898, "learning_rate": 1.4607574711643257e-06, "loss": 0.0041, "step": 37070 }, { "grad_norm": 0.13495254516601562, "learning_rate": 1.4508552418382949e-06, "loss": 0.0038, "step": 37080 }, { "grad_norm": 0.1203780546784401, "learning_rate": 1.4409861953949112e-06, "loss": 0.0052, "step": 37090 }, { "grad_norm": 0.08026383072137833, "learning_rate": 1.43115033857959e-06, "loss": 0.0041, "step": 37100 }, { "grad_norm": 0.1502707451581955, "learning_rate": 1.4213476781150258e-06, "loss": 0.0035, "step": 37110 }, { "grad_norm": 0.15550421178340912, "learning_rate": 1.411578220701254e-06, "loss": 0.003, "step": 37120 }, { "grad_norm": 0.06099868193268776, "learning_rate": 1.4018419730156108e-06, "loss": 0.0034, "step": 37130 }, { "grad_norm": 0.06466396898031235, "learning_rate": 1.3921389417127283e-06, "loss": 0.0033, "step": 37140 }, { "grad_norm": 0.11801125854253769, "learning_rate": 1.3824691334245298e-06, "loss": 0.0045, "step": 37150 }, { "grad_norm": 0.10980676859617233, "learning_rate": 1.3728325547602505e-06, "loss": 0.0037, "step": 37160 }, { "grad_norm": 0.059253063052892685, "learning_rate": 1.3632292123063828e-06, "loss": 0.0047, "step": 37170 }, { "grad_norm": 0.04841199144721031, "learning_rate": 1.3536591126267317e-06, "loss": 0.0053, "step": 37180 }, { "grad_norm": 0.16246099770069122, "learning_rate": 1.3441222622623761e-06, "loss": 0.0069, "step": 37190 }, { "grad_norm": 0.07149355858564377, "learning_rate": 1.3346186677316575e-06, "loss": 0.0049, "step": 37200 }, { "grad_norm": 0.10179715603590012, "learning_rate": 1.3251483355301852e-06, "loss": 0.0063, "step": 37210 }, { "grad_norm": 0.061882857233285904, "learning_rate": 1.315711272130854e-06, "loss": 0.0045, "step": 37220 }, { "grad_norm": 0.07067196071147919, "learning_rate": 1.306307483983804e-06, "loss": 0.0061, "step": 37230 }, { "grad_norm": 0.12607939541339874, "learning_rate": 1.2969369775164275e-06, "loss": 0.0034, "step": 37240 }, { "grad_norm": 0.12566575407981873, "learning_rate": 1.28759975913339e-06, "loss": 0.0047, "step": 37250 }, { "grad_norm": 0.07420851290225983, "learning_rate": 1.2782958352165863e-06, "loss": 0.0041, "step": 37260 }, { "grad_norm": 0.15541905164718628, "learning_rate": 1.2690252121251633e-06, "loss": 0.0036, "step": 37270 }, { "grad_norm": 0.138152077794075, "learning_rate": 1.2597878961954968e-06, "loss": 0.0054, "step": 37280 }, { "grad_norm": 0.0800509974360466, "learning_rate": 1.250583893741214e-06, "loss": 0.0029, "step": 37290 }, { "grad_norm": 0.14294755458831787, "learning_rate": 1.2414132110531607e-06, "loss": 0.0053, "step": 37300 }, { "grad_norm": 0.09310872107744217, "learning_rate": 1.2322758543994062e-06, "loss": 0.0036, "step": 37310 }, { "grad_norm": 0.08327431976795197, "learning_rate": 1.2231718300252603e-06, "loss": 0.007, "step": 37320 }, { "grad_norm": 0.10100974142551422, "learning_rate": 1.214101144153229e-06, "loss": 0.005, "step": 37330 }, { "grad_norm": 0.1005554050207138, "learning_rate": 1.205063802983042e-06, "loss": 0.0052, "step": 37340 }, { "grad_norm": 0.05461946874856949, "learning_rate": 1.1960598126916301e-06, "loss": 0.0049, "step": 37350 }, { "grad_norm": 0.08563163131475449, "learning_rate": 1.187089179433154e-06, "loss": 0.005, "step": 37360 }, { "grad_norm": 0.06347856670618057, "learning_rate": 1.178151909338937e-06, "loss": 0.0037, "step": 37370 }, { "grad_norm": 0.14462734758853912, "learning_rate": 1.1692480085175317e-06, "loss": 0.0041, "step": 37380 }, { "grad_norm": 0.06010375916957855, "learning_rate": 1.1603774830546587e-06, "loss": 0.0059, "step": 37390 }, { "grad_norm": 0.10535765439271927, "learning_rate": 1.1515403390132573e-06, "loss": 0.0045, "step": 37400 }, { "grad_norm": 0.0465153232216835, "learning_rate": 1.1427365824334124e-06, "loss": 0.0067, "step": 37410 }, { "grad_norm": 0.06910575181245804, "learning_rate": 1.133966219332422e-06, "loss": 0.0037, "step": 37420 }, { "grad_norm": 0.08821554481983185, "learning_rate": 1.1252292557047405e-06, "loss": 0.0034, "step": 37430 }, { "grad_norm": 0.09474597126245499, "learning_rate": 1.1165256975219974e-06, "loss": 0.0073, "step": 37440 }, { "grad_norm": 0.10442539304494858, "learning_rate": 1.1078555507330002e-06, "loss": 0.0045, "step": 37450 }, { "grad_norm": 0.10226097702980042, "learning_rate": 1.0992188212637089e-06, "loss": 0.0066, "step": 37460 }, { "grad_norm": 0.09822197258472443, "learning_rate": 1.0906155150172458e-06, "loss": 0.0052, "step": 37470 }, { "grad_norm": 0.16420325636863708, "learning_rate": 1.0820456378738852e-06, "loss": 0.0047, "step": 37480 }, { "grad_norm": 0.0989886149764061, "learning_rate": 1.0735091956910692e-06, "loss": 0.0052, "step": 37490 }, { "grad_norm": 0.12012454867362976, "learning_rate": 1.0650061943033695e-06, "loss": 0.0056, "step": 37500 }, { "grad_norm": 0.09384386241436005, "learning_rate": 1.0565366395225041e-06, "loss": 0.0038, "step": 37510 }, { "grad_norm": 0.11919771879911423, "learning_rate": 1.0481005371373366e-06, "loss": 0.0069, "step": 37520 }, { "grad_norm": 0.14669275283813477, "learning_rate": 1.0396978929138602e-06, "loss": 0.0055, "step": 37530 }, { "grad_norm": 0.0841396227478981, "learning_rate": 1.0313287125952032e-06, "loss": 0.0043, "step": 37540 }, { "grad_norm": 0.06388873606920242, "learning_rate": 1.0229930019016288e-06, "loss": 0.0046, "step": 37550 }, { "grad_norm": 0.05802198871970177, "learning_rate": 1.0146907665305183e-06, "loss": 0.0041, "step": 37560 }, { "grad_norm": 0.06294536590576172, "learning_rate": 1.0064220121563551e-06, "loss": 0.0066, "step": 37570 }, { "grad_norm": 0.10056362301111221, "learning_rate": 9.98186744430768e-07, "loss": 0.0039, "step": 37580 }, { "grad_norm": 0.14177250862121582, "learning_rate": 9.899849689824826e-07, "loss": 0.0041, "step": 37590 }, { "grad_norm": 0.061599038541316986, "learning_rate": 9.818166914173365e-07, "loss": 0.0025, "step": 37600 }, { "grad_norm": 0.10386981815099716, "learning_rate": 9.736819173182642e-07, "loss": 0.0037, "step": 37610 }, { "grad_norm": 0.12679524719715118, "learning_rate": 9.655806522453126e-07, "loss": 0.0034, "step": 37620 }, { "grad_norm": 0.08579432964324951, "learning_rate": 9.575129017356187e-07, "loss": 0.0029, "step": 37630 }, { "grad_norm": 0.16281160712242126, "learning_rate": 9.494786713034109e-07, "loss": 0.0033, "step": 37640 }, { "grad_norm": 0.14133389294147491, "learning_rate": 9.414779664400075e-07, "loss": 0.0051, "step": 37650 }, { "grad_norm": 0.07282757014036179, "learning_rate": 9.335107926138342e-07, "loss": 0.0047, "step": 37660 }, { "grad_norm": 0.1643914431333542, "learning_rate": 9.255771552703519e-07, "loss": 0.0057, "step": 37670 }, { "grad_norm": 0.11990325897932053, "learning_rate": 9.176770598321449e-07, "loss": 0.007, "step": 37680 }, { "grad_norm": 0.08579728752374649, "learning_rate": 9.098105116988498e-07, "loss": 0.0059, "step": 37690 }, { "grad_norm": 0.05950339883565903, "learning_rate": 9.019775162471767e-07, "loss": 0.0036, "step": 37700 }, { "grad_norm": 0.09631351381540298, "learning_rate": 8.94178078830904e-07, "loss": 0.0045, "step": 37710 }, { "grad_norm": 0.08225615322589874, "learning_rate": 8.864122047808731e-07, "loss": 0.004, "step": 37720 }, { "grad_norm": 0.06338236480951309, "learning_rate": 8.786798994049882e-07, "loss": 0.0033, "step": 37730 }, { "grad_norm": 0.13052614033222198, "learning_rate": 8.709811679882052e-07, "loss": 0.0038, "step": 37740 }, { "grad_norm": 0.05356431007385254, "learning_rate": 8.633160157925313e-07, "loss": 0.0043, "step": 37750 }, { "grad_norm": 0.05861632525920868, "learning_rate": 8.556844480570369e-07, "loss": 0.0037, "step": 37760 }, { "grad_norm": 0.09285953640937805, "learning_rate": 8.480864699978164e-07, "loss": 0.0028, "step": 37770 }, { "grad_norm": 0.05119621753692627, "learning_rate": 8.405220868080154e-07, "loss": 0.0056, "step": 37780 }, { "grad_norm": 0.09192308783531189, "learning_rate": 8.329913036578263e-07, "loss": 0.0029, "step": 37790 }, { "grad_norm": 0.150011345744133, "learning_rate": 8.254941256944593e-07, "loss": 0.0038, "step": 37800 }, { "grad_norm": 0.06083458289504051, "learning_rate": 8.18030558042171e-07, "loss": 0.0043, "step": 37810 }, { "grad_norm": 0.1100180596113205, "learning_rate": 8.106006058022419e-07, "loss": 0.005, "step": 37820 }, { "grad_norm": 0.1534986048936844, "learning_rate": 8.03204274052971e-07, "loss": 0.0032, "step": 37830 }, { "grad_norm": 0.1391812562942505, "learning_rate": 7.958415678496756e-07, "loss": 0.004, "step": 37840 }, { "grad_norm": 0.06554927676916122, "learning_rate": 7.885124922247022e-07, "loss": 0.0056, "step": 37850 }, { "grad_norm": 0.05173212289810181, "learning_rate": 7.812170521874051e-07, "loss": 0.0037, "step": 37860 }, { "grad_norm": 0.07653755694627762, "learning_rate": 7.739552527241456e-07, "loss": 0.0033, "step": 37870 }, { "grad_norm": 0.08721320331096649, "learning_rate": 7.667270987983033e-07, "loss": 0.0038, "step": 37880 }, { "grad_norm": 0.07579607516527176, "learning_rate": 7.595325953502375e-07, "loss": 0.0037, "step": 37890 }, { "grad_norm": 0.11376312375068665, "learning_rate": 7.523717472973368e-07, "loss": 0.0034, "step": 37900 }, { "grad_norm": 0.061757542192935944, "learning_rate": 7.452445595339696e-07, "loss": 0.0037, "step": 37910 }, { "grad_norm": 0.09684731811285019, "learning_rate": 7.381510369315059e-07, "loss": 0.0035, "step": 37920 }, { "grad_norm": 0.11430849879980087, "learning_rate": 7.310911843382895e-07, "loss": 0.0037, "step": 37930 }, { "grad_norm": 0.10919754207134247, "learning_rate": 7.240650065796716e-07, "loss": 0.0037, "step": 37940 }, { "grad_norm": 0.10594075918197632, "learning_rate": 7.170725084579777e-07, "loss": 0.0037, "step": 37950 }, { "grad_norm": 0.06725946813821793, "learning_rate": 7.101136947525233e-07, "loss": 0.0043, "step": 37960 }, { "grad_norm": 0.07469627261161804, "learning_rate": 7.031885702195707e-07, "loss": 0.0046, "step": 37970 }, { "grad_norm": 0.09987401217222214, "learning_rate": 6.962971395924e-07, "loss": 0.0043, "step": 37980 }, { "grad_norm": 0.07113064080476761, "learning_rate": 6.894394075812217e-07, "loss": 0.0031, "step": 37990 }, { "grad_norm": 0.13405995070934296, "learning_rate": 6.826153788732359e-07, "loss": 0.0058, "step": 38000 }, { "grad_norm": 0.0956515520811081, "learning_rate": 6.758250581326009e-07, "loss": 0.0038, "step": 38010 }, { "grad_norm": 0.11938880383968353, "learning_rate": 6.690684500004485e-07, "loss": 0.0039, "step": 38020 }, { "grad_norm": 0.09364815056324005, "learning_rate": 6.623455590948458e-07, "loss": 0.0029, "step": 38030 }, { "grad_norm": 0.046164195984601974, "learning_rate": 6.556563900108226e-07, "loss": 0.0034, "step": 38040 }, { "grad_norm": 0.16069309413433075, "learning_rate": 6.490009473203662e-07, "loss": 0.0048, "step": 38050 }, { "grad_norm": 0.09410692751407623, "learning_rate": 6.423792355724156e-07, "loss": 0.0042, "step": 38060 }, { "grad_norm": 0.07561647146940231, "learning_rate": 6.35791259292845e-07, "loss": 0.0029, "step": 38070 }, { "grad_norm": 0.15516580641269684, "learning_rate": 6.292370229844691e-07, "loss": 0.0047, "step": 38080 }, { "grad_norm": 0.11173947155475616, "learning_rate": 6.227165311270488e-07, "loss": 0.0065, "step": 38090 }, { "grad_norm": 0.1954604983329773, "learning_rate": 6.16229788177275e-07, "loss": 0.0061, "step": 38100 }, { "grad_norm": 0.03809060901403427, "learning_rate": 6.097767985687841e-07, "loss": 0.0038, "step": 38110 }, { "grad_norm": 0.08047151565551758, "learning_rate": 6.033575667121317e-07, "loss": 0.0044, "step": 38120 }, { "grad_norm": 0.1333116590976715, "learning_rate": 5.969720969947967e-07, "loss": 0.0065, "step": 38130 }, { "grad_norm": 0.12273000925779343, "learning_rate": 5.906203937811883e-07, "loss": 0.003, "step": 38140 }, { "grad_norm": 0.14484736323356628, "learning_rate": 5.843024614126447e-07, "loss": 0.0032, "step": 38150 }, { "grad_norm": 0.1069556176662445, "learning_rate": 5.780183042074061e-07, "loss": 0.0045, "step": 38160 }, { "grad_norm": 0.1497611403465271, "learning_rate": 5.717679264606312e-07, "loss": 0.0055, "step": 38170 }, { "grad_norm": 0.05641280114650726, "learning_rate": 5.655513324444028e-07, "loss": 0.0032, "step": 38180 }, { "grad_norm": 0.1251291185617447, "learning_rate": 5.593685264076942e-07, "loss": 0.0043, "step": 38190 }, { "grad_norm": 0.13232654333114624, "learning_rate": 5.532195125764029e-07, "loss": 0.004, "step": 38200 }, { "grad_norm": 0.09163253009319305, "learning_rate": 5.471042951533167e-07, "loss": 0.0041, "step": 38210 }, { "grad_norm": 0.1090649962425232, "learning_rate": 5.410228783181371e-07, "loss": 0.0042, "step": 38220 }, { "grad_norm": 0.08001293241977692, "learning_rate": 5.349752662274444e-07, "loss": 0.0057, "step": 38230 }, { "grad_norm": 0.07622145861387253, "learning_rate": 5.289614630147321e-07, "loss": 0.0041, "step": 38240 }, { "grad_norm": 0.11646436899900436, "learning_rate": 5.229814727903737e-07, "loss": 0.0047, "step": 38250 }, { "grad_norm": 0.06895823031663895, "learning_rate": 5.170352996416439e-07, "loss": 0.0045, "step": 38260 }, { "grad_norm": 0.07749106734991074, "learning_rate": 5.111229476326918e-07, "loss": 0.0037, "step": 38270 }, { "grad_norm": 0.16457562148571014, "learning_rate": 5.052444208045459e-07, "loss": 0.0041, "step": 38280 }, { "grad_norm": 0.1220884844660759, "learning_rate": 4.993997231751312e-07, "loss": 0.0047, "step": 38290 }, { "grad_norm": 0.06816622614860535, "learning_rate": 4.935888587392468e-07, "loss": 0.0037, "step": 38300 }, { "grad_norm": 0.1253966987133026, "learning_rate": 4.878118314685598e-07, "loss": 0.0049, "step": 38310 }, { "grad_norm": 0.0584893599152565, "learning_rate": 4.820686453116119e-07, "loss": 0.0037, "step": 38320 }, { "grad_norm": 0.08117762953042984, "learning_rate": 4.763593041938186e-07, "loss": 0.0053, "step": 38330 }, { "grad_norm": 0.12317889928817749, "learning_rate": 4.706838120174584e-07, "loss": 0.0071, "step": 38340 }, { "grad_norm": 0.14347277581691742, "learning_rate": 4.650421726616838e-07, "loss": 0.0041, "step": 38350 }, { "grad_norm": 0.11848006397485733, "learning_rate": 4.594343899824993e-07, "loss": 0.0041, "step": 38360 }, { "grad_norm": 0.22837737202644348, "learning_rate": 4.538604678127667e-07, "loss": 0.0051, "step": 38370 }, { "grad_norm": 0.17004840075969696, "learning_rate": 4.483204099622107e-07, "loss": 0.0049, "step": 38380 }, { "grad_norm": 0.1668739914894104, "learning_rate": 4.4281422021740793e-07, "loss": 0.0039, "step": 38390 }, { "grad_norm": 0.09055084735155106, "learning_rate": 4.373419023417924e-07, "loss": 0.0036, "step": 38400 }, { "grad_norm": 0.15021556615829468, "learning_rate": 4.319034600756389e-07, "loss": 0.0065, "step": 38410 }, { "grad_norm": 0.06868670880794525, "learning_rate": 4.264988971360739e-07, "loss": 0.0046, "step": 38420 }, { "grad_norm": 0.12016773223876953, "learning_rate": 4.211282172170594e-07, "loss": 0.0049, "step": 38430 }, { "grad_norm": 0.19455979764461517, "learning_rate": 4.1579142398940894e-07, "loss": 0.0035, "step": 38440 }, { "grad_norm": 0.1679660677909851, "learning_rate": 4.1048852110076585e-07, "loss": 0.0041, "step": 38450 }, { "grad_norm": 0.0795559287071228, "learning_rate": 4.052195121756197e-07, "loss": 0.0054, "step": 38460 }, { "grad_norm": 0.08538047969341278, "learning_rate": 3.9998440081528974e-07, "loss": 0.0047, "step": 38470 }, { "grad_norm": 0.07063426077365875, "learning_rate": 3.9478319059791936e-07, "loss": 0.0042, "step": 38480 }, { "grad_norm": 0.0966687723994255, "learning_rate": 3.8961588507848143e-07, "loss": 0.0048, "step": 38490 }, { "grad_norm": 0.0454234853386879, "learning_rate": 3.844824877887954e-07, "loss": 0.006, "step": 38500 }, { "grad_norm": 0.06983405351638794, "learning_rate": 3.7938300223748224e-07, "loss": 0.0054, "step": 38510 }, { "grad_norm": 0.07858243584632874, "learning_rate": 3.743174319099929e-07, "loss": 0.0032, "step": 38520 }, { "grad_norm": 0.08717841655015945, "learning_rate": 3.6928578026859673e-07, "loss": 0.004, "step": 38530 }, { "grad_norm": 0.0759018287062645, "learning_rate": 3.642880507523816e-07, "loss": 0.005, "step": 38540 }, { "grad_norm": 0.08264084160327911, "learning_rate": 3.5932424677724285e-07, "loss": 0.0069, "step": 38550 }, { "grad_norm": 0.08507038652896881, "learning_rate": 3.5439437173589996e-07, "loss": 0.0034, "step": 38560 }, { "grad_norm": 0.10220137238502502, "learning_rate": 3.4949842899787423e-07, "loss": 0.004, "step": 38570 }, { "grad_norm": 0.11609067022800446, "learning_rate": 3.446364219094944e-07, "loss": 0.0036, "step": 38580 }, { "grad_norm": 0.15489405393600464, "learning_rate": 3.3980835379389675e-07, "loss": 0.0061, "step": 38590 }, { "grad_norm": 0.10559788346290588, "learning_rate": 3.350142279510249e-07, "loss": 0.004, "step": 38600 }, { "grad_norm": 0.12703129649162292, "learning_rate": 3.3025404765761324e-07, "loss": 0.0046, "step": 38610 }, { "grad_norm": 0.12162133306264877, "learning_rate": 3.255278161672037e-07, "loss": 0.0061, "step": 38620 }, { "grad_norm": 0.04289935529232025, "learning_rate": 3.20835536710129e-07, "loss": 0.0035, "step": 38630 }, { "grad_norm": 0.1219756081700325, "learning_rate": 3.1617721249351804e-07, "loss": 0.004, "step": 38640 }, { "grad_norm": 0.11614067852497101, "learning_rate": 3.115528467012963e-07, "loss": 0.005, "step": 38650 }, { "grad_norm": 0.1447312980890274, "learning_rate": 3.0696244249416305e-07, "loss": 0.0045, "step": 38660 }, { "grad_norm": 0.14914990961551666, "learning_rate": 3.024060030096254e-07, "loss": 0.0036, "step": 38670 }, { "grad_norm": 0.08447648584842682, "learning_rate": 2.9788353136196435e-07, "loss": 0.0049, "step": 38680 }, { "grad_norm": 0.13365095853805542, "learning_rate": 2.9339503064224615e-07, "loss": 0.004, "step": 38690 }, { "grad_norm": 0.09139462560415268, "learning_rate": 2.889405039183224e-07, "loss": 0.0037, "step": 38700 }, { "grad_norm": 0.11969394236803055, "learning_rate": 2.8451995423481315e-07, "loss": 0.0043, "step": 38710 }, { "grad_norm": 0.07189628481864929, "learning_rate": 2.801333846131293e-07, "loss": 0.0041, "step": 38720 }, { "grad_norm": 0.20835795998573303, "learning_rate": 2.757807980514504e-07, "loss": 0.0046, "step": 38730 }, { "grad_norm": 0.11905965954065323, "learning_rate": 2.714621975247189e-07, "loss": 0.0048, "step": 38740 }, { "grad_norm": 0.15201571583747864, "learning_rate": 2.671775859846626e-07, "loss": 0.0061, "step": 38750 }, { "grad_norm": 0.09741232544183731, "learning_rate": 2.629269663597722e-07, "loss": 0.0036, "step": 38760 }, { "grad_norm": 0.1472184658050537, "learning_rate": 2.587103415553127e-07, "loss": 0.0058, "step": 38770 }, { "grad_norm": 0.061498694121837616, "learning_rate": 2.545277144533009e-07, "loss": 0.0037, "step": 38780 }, { "grad_norm": 0.13907574117183685, "learning_rate": 2.5037908791252227e-07, "loss": 0.0041, "step": 38790 }, { "grad_norm": 0.07469802349805832, "learning_rate": 2.462644647685308e-07, "loss": 0.0045, "step": 38800 }, { "grad_norm": 0.13287514448165894, "learning_rate": 2.421838478336269e-07, "loss": 0.0047, "step": 38810 }, { "grad_norm": 0.06206384301185608, "learning_rate": 2.3813723989687398e-07, "loss": 0.0055, "step": 38820 }, { "grad_norm": 0.1178857758641243, "learning_rate": 2.3412464372409293e-07, "loss": 0.0044, "step": 38830 }, { "grad_norm": 0.06664584577083588, "learning_rate": 2.301460620578455e-07, "loss": 0.0044, "step": 38840 }, { "grad_norm": 0.07452195137739182, "learning_rate": 2.262014976174731e-07, "loss": 0.0043, "step": 38850 }, { "grad_norm": 0.052395693957805634, "learning_rate": 2.2229095309903025e-07, "loss": 0.007, "step": 38860 }, { "grad_norm": 0.17344343662261963, "learning_rate": 2.184144311753511e-07, "loss": 0.0042, "step": 38870 }, { "grad_norm": 0.2057366967201233, "learning_rate": 2.14571934495994e-07, "loss": 0.0042, "step": 38880 }, { "grad_norm": 0.11611562222242355, "learning_rate": 2.1076346568726923e-07, "loss": 0.0038, "step": 38890 }, { "grad_norm": 0.05279689282178879, "learning_rate": 2.0698902735223346e-07, "loss": 0.0054, "step": 38900 }, { "grad_norm": 0.09295405447483063, "learning_rate": 2.0324862207067863e-07, "loss": 0.0049, "step": 38910 }, { "grad_norm": 0.09800247102975845, "learning_rate": 1.9954225239913195e-07, "loss": 0.0038, "step": 38920 }, { "grad_norm": 0.0646526962518692, "learning_rate": 1.9586992087087252e-07, "loss": 0.0036, "step": 38930 }, { "grad_norm": 0.04511716961860657, "learning_rate": 1.9223162999588705e-07, "loss": 0.0041, "step": 38940 }, { "grad_norm": 0.1474868357181549, "learning_rate": 1.886273822609308e-07, "loss": 0.0056, "step": 38950 }, { "grad_norm": 0.13326996564865112, "learning_rate": 1.8505718012946094e-07, "loss": 0.0042, "step": 38960 }, { "grad_norm": 0.17217156291007996, "learning_rate": 1.8152102604168664e-07, "loss": 0.0035, "step": 38970 }, { "grad_norm": 0.1308603584766388, "learning_rate": 1.7801892241452457e-07, "loss": 0.0042, "step": 38980 }, { "grad_norm": 0.07337290048599243, "learning_rate": 1.745508716416433e-07, "loss": 0.005, "step": 38990 }, { "grad_norm": 0.11944092065095901, "learning_rate": 1.7111687609341342e-07, "loss": 0.004, "step": 39000 }, { "grad_norm": 0.14084549248218536, "learning_rate": 1.6771693811693523e-07, "loss": 0.0054, "step": 39010 }, { "grad_norm": 0.06611872464418411, "learning_rate": 1.643510600360443e-07, "loss": 0.0032, "step": 39020 }, { "grad_norm": 0.05366404727101326, "learning_rate": 1.610192441512781e-07, "loss": 0.0052, "step": 39030 }, { "grad_norm": 0.15739084780216217, "learning_rate": 1.5772149273990956e-07, "loss": 0.0056, "step": 39040 }, { "grad_norm": 0.128244087100029, "learning_rate": 1.5445780805591337e-07, "loss": 0.0028, "step": 39050 }, { "grad_norm": 0.1127924919128418, "learning_rate": 1.5122819232998852e-07, "loss": 0.0064, "step": 39060 }, { "grad_norm": 0.13815225660800934, "learning_rate": 1.4803264776955262e-07, "loss": 0.005, "step": 39070 }, { "grad_norm": 0.15272153913974762, "learning_rate": 1.448711765587196e-07, "loss": 0.0029, "step": 39080 }, { "grad_norm": 0.09595058858394623, "learning_rate": 1.4174378085833885e-07, "loss": 0.005, "step": 39090 }, { "grad_norm": 0.07226362079381943, "learning_rate": 1.3865046280594486e-07, "loss": 0.0052, "step": 39100 }, { "grad_norm": 0.14897842705249786, "learning_rate": 1.3559122451579643e-07, "loss": 0.0033, "step": 39110 }, { "grad_norm": 0.10114357620477676, "learning_rate": 1.3256606807884874e-07, "loss": 0.0065, "step": 39120 }, { "grad_norm": 0.05709192156791687, "learning_rate": 1.2957499556277563e-07, "loss": 0.006, "step": 39130 }, { "grad_norm": 0.11146177351474762, "learning_rate": 1.266180090119473e-07, "loss": 0.0045, "step": 39140 }, { "grad_norm": 0.07799983024597168, "learning_rate": 1.2369511044742488e-07, "loss": 0.0031, "step": 39150 }, { "grad_norm": 0.15136972069740295, "learning_rate": 1.2080630186698805e-07, "loss": 0.0041, "step": 39160 }, { "grad_norm": 0.03185140714049339, "learning_rate": 1.1795158524511851e-07, "loss": 0.0033, "step": 39170 }, { "grad_norm": 0.049844495952129364, "learning_rate": 1.1513096253297773e-07, "loss": 0.0038, "step": 39180 }, { "grad_norm": 0.09777430444955826, "learning_rate": 1.1234443565843466e-07, "loss": 0.0047, "step": 39190 }, { "grad_norm": 0.09101809561252594, "learning_rate": 1.0959200652606028e-07, "loss": 0.0041, "step": 39200 }, { "grad_norm": 0.13238760828971863, "learning_rate": 1.0687367701711082e-07, "loss": 0.005, "step": 39210 }, { "grad_norm": 0.10942031443119049, "learning_rate": 1.0418944898953898e-07, "loss": 0.0044, "step": 39220 }, { "grad_norm": 0.09186723828315735, "learning_rate": 1.0153932427798829e-07, "loss": 0.004, "step": 39230 }, { "grad_norm": 0.22304384410381317, "learning_rate": 9.892330469379874e-08, "loss": 0.0069, "step": 39240 }, { "grad_norm": 0.06406592577695847, "learning_rate": 9.634139202498449e-08, "loss": 0.0026, "step": 39250 }, { "grad_norm": 0.10063044726848602, "learning_rate": 9.379358803627281e-08, "loss": 0.0032, "step": 39260 }, { "grad_norm": 0.10850845277309418, "learning_rate": 9.12798944690485e-08, "loss": 0.0031, "step": 39270 }, { "grad_norm": 0.16494768857955933, "learning_rate": 8.880031304141501e-08, "loss": 0.0069, "step": 39280 }, { "grad_norm": 0.1327551156282425, "learning_rate": 8.635484544812777e-08, "loss": 0.0067, "step": 39290 }, { "grad_norm": 0.10503501445055008, "learning_rate": 8.394349336064422e-08, "loss": 0.004, "step": 39300 }, { "grad_norm": 0.09040160477161407, "learning_rate": 8.156625842710153e-08, "loss": 0.0047, "step": 39310 }, { "grad_norm": 0.13396571576595306, "learning_rate": 7.92231422723222e-08, "loss": 0.0045, "step": 39320 }, { "grad_norm": 0.044754549860954285, "learning_rate": 7.691414649780293e-08, "loss": 0.0051, "step": 39330 }, { "grad_norm": 0.060839373618364334, "learning_rate": 7.463927268171467e-08, "loss": 0.003, "step": 39340 }, { "grad_norm": 0.12448371946811676, "learning_rate": 7.239852237891365e-08, "loss": 0.0055, "step": 39350 }, { "grad_norm": 0.11499923467636108, "learning_rate": 7.019189712093588e-08, "loss": 0.0044, "step": 39360 }, { "grad_norm": 0.09477554261684418, "learning_rate": 6.801939841598603e-08, "loss": 0.0035, "step": 39370 }, { "grad_norm": 0.11521543562412262, "learning_rate": 6.588102774894856e-08, "loss": 0.0037, "step": 39380 }, { "grad_norm": 0.0992189571261406, "learning_rate": 6.377678658137653e-08, "loss": 0.0039, "step": 39390 }, { "grad_norm": 0.13576383888721466, "learning_rate": 6.17066763514973e-08, "loss": 0.0036, "step": 39400 }, { "grad_norm": 0.07371159642934799, "learning_rate": 5.967069847421792e-08, "loss": 0.0036, "step": 39410 }, { "grad_norm": 0.06413822621107101, "learning_rate": 5.7668854341108583e-08, "loss": 0.0042, "step": 39420 }, { "grad_norm": 0.14740988612174988, "learning_rate": 5.570114532040815e-08, "loss": 0.0049, "step": 39430 }, { "grad_norm": 0.08708333969116211, "learning_rate": 5.3767572757029706e-08, "loss": 0.0062, "step": 39440 }, { "grad_norm": 0.03366042673587799, "learning_rate": 5.186813797254386e-08, "loss": 0.005, "step": 39450 }, { "grad_norm": 0.16206304728984833, "learning_rate": 5.000284226521213e-08, "loss": 0.004, "step": 39460 }, { "grad_norm": 0.07614514976739883, "learning_rate": 4.8171686909931346e-08, "loss": 0.004, "step": 39470 }, { "grad_norm": 0.07195644825696945, "learning_rate": 4.637467315828925e-08, "loss": 0.004, "step": 39480 }, { "grad_norm": 0.12345919758081436, "learning_rate": 4.461180223852002e-08, "loss": 0.0042, "step": 39490 }, { "grad_norm": 0.1355839967727661, "learning_rate": 4.288307535553204e-08, "loss": 0.0037, "step": 39500 }, { "grad_norm": 0.21991275250911713, "learning_rate": 4.1188493690896836e-08, "loss": 0.0078, "step": 39510 }, { "grad_norm": 0.1275297999382019, "learning_rate": 3.952805840284346e-08, "loss": 0.005, "step": 39520 }, { "grad_norm": 0.06809234619140625, "learning_rate": 3.790177062625855e-08, "loss": 0.0031, "step": 39530 }, { "grad_norm": 0.06853342801332474, "learning_rate": 3.6309631472702946e-08, "loss": 0.004, "step": 39540 }, { "grad_norm": 0.09344737231731415, "learning_rate": 3.475164203038395e-08, "loss": 0.0036, "step": 39550 }, { "grad_norm": 0.035461924970149994, "learning_rate": 3.3227803364171974e-08, "loss": 0.0034, "step": 39560 }, { "grad_norm": 0.07495958358049393, "learning_rate": 3.173811651560055e-08, "loss": 0.0033, "step": 39570 }, { "grad_norm": 0.12882933020591736, "learning_rate": 3.0282582502855204e-08, "loss": 0.0054, "step": 39580 }, { "grad_norm": 0.06949090212583542, "learning_rate": 2.8861202320773496e-08, "loss": 0.0051, "step": 39590 }, { "grad_norm": 0.14310134947299957, "learning_rate": 2.7473976940867175e-08, "loss": 0.0038, "step": 39600 }, { "grad_norm": 0.07929772138595581, "learning_rate": 2.6120907311277808e-08, "loss": 0.0045, "step": 39610 }, { "grad_norm": 0.09018973261117935, "learning_rate": 2.480199435682673e-08, "loss": 0.0029, "step": 39620 }, { "grad_norm": 0.10084343701601028, "learning_rate": 2.3517238978976174e-08, "loss": 0.0037, "step": 39630 }, { "grad_norm": 0.08126329630613327, "learning_rate": 2.2266642055834842e-08, "loss": 0.0037, "step": 39640 }, { "grad_norm": 0.0732973963022232, "learning_rate": 2.105020444219119e-08, "loss": 0.0032, "step": 39650 }, { "grad_norm": 0.10291371494531631, "learning_rate": 1.9867926969452388e-08, "loss": 0.0043, "step": 39660 }, { "grad_norm": 0.10089748352766037, "learning_rate": 1.8719810445705366e-08, "loss": 0.0048, "step": 39670 }, { "grad_norm": 0.10334035009145737, "learning_rate": 1.760585565567241e-08, "loss": 0.0043, "step": 39680 }, { "grad_norm": 0.04421366751194, "learning_rate": 1.6526063360727816e-08, "loss": 0.0042, "step": 39690 }, { "grad_norm": 0.05793644115328789, "learning_rate": 1.5480434298908997e-08, "loss": 0.0042, "step": 39700 }, { "grad_norm": 0.074440598487854, "learning_rate": 1.4468969184883164e-08, "loss": 0.005, "step": 39710 }, { "grad_norm": 0.0839373841881752, "learning_rate": 1.3491668709980643e-08, "loss": 0.0039, "step": 39720 }, { "grad_norm": 0.1319737285375595, "learning_rate": 1.254853354217822e-08, "loss": 0.005, "step": 39730 }, { "grad_norm": 0.06543730199337006, "learning_rate": 1.1639564326099139e-08, "loss": 0.0069, "step": 39740 }, { "grad_norm": 0.0734311193227768, "learning_rate": 1.0764761683018653e-08, "loss": 0.003, "step": 39750 }, { "grad_norm": 0.14188313484191895, "learning_rate": 9.924126210852925e-09, "loss": 0.0044, "step": 39760 }, { "grad_norm": 0.10478965193033218, "learning_rate": 9.117658484170122e-09, "loss": 0.0036, "step": 39770 }, { "grad_norm": 0.04910323768854141, "learning_rate": 8.345359054179324e-09, "loss": 0.0037, "step": 39780 }, { "grad_norm": 0.07561340183019638, "learning_rate": 7.60722844874162e-09, "loss": 0.0056, "step": 39790 }, { "grad_norm": 0.11997492611408234, "learning_rate": 6.903267172364558e-09, "loss": 0.0042, "step": 39800 }, { "grad_norm": 0.07368051260709763, "learning_rate": 6.23347570619659e-09, "loss": 0.0042, "step": 39810 }, { "grad_norm": 0.11801760643720627, "learning_rate": 5.59785450803818e-09, "loss": 0.005, "step": 39820 }, { "grad_norm": 0.09146593511104584, "learning_rate": 4.996404012325151e-09, "loss": 0.0046, "step": 39830 }, { "grad_norm": 0.14287440478801727, "learning_rate": 4.429124630145331e-09, "loss": 0.0047, "step": 39840 }, { "grad_norm": 0.12751300632953644, "learning_rate": 3.89601674922746e-09, "loss": 0.0058, "step": 39850 }, { "grad_norm": 0.08644873648881912, "learning_rate": 3.3970807339467336e-09, "loss": 0.004, "step": 39860 }, { "grad_norm": 0.13520771265029907, "learning_rate": 2.9323169253248073e-09, "loss": 0.0042, "step": 39870 }, { "grad_norm": 0.0767550989985466, "learning_rate": 2.501725641018693e-09, "loss": 0.0037, "step": 39880 }, { "grad_norm": 0.14261342585086823, "learning_rate": 2.105307175331861e-09, "loss": 0.0065, "step": 39890 }, { "grad_norm": 0.05744582787156105, "learning_rate": 1.7430617992197918e-09, "loss": 0.006, "step": 39900 }, { "grad_norm": 0.16145870089530945, "learning_rate": 1.4149897602677708e-09, "loss": 0.0044, "step": 39910 }, { "grad_norm": 0.06828194856643677, "learning_rate": 1.1210912827130937e-09, "loss": 0.0053, "step": 39920 }, { "grad_norm": 0.10761803388595581, "learning_rate": 8.613665674284121e-10, "loss": 0.0062, "step": 39930 }, { "grad_norm": 0.06959618628025055, "learning_rate": 6.358157919383878e-10, "loss": 0.0038, "step": 39940 }, { "grad_norm": 0.05851806700229645, "learning_rate": 4.4443911040303877e-10, "loss": 0.0026, "step": 39950 }, { "grad_norm": 0.11949149519205093, "learning_rate": 2.8723665362329066e-10, "loss": 0.0046, "step": 39960 }, { "grad_norm": 0.11629597097635269, "learning_rate": 1.6420852905207895e-10, "loss": 0.0033, "step": 39970 }, { "grad_norm": 0.055131927132606506, "learning_rate": 7.535482076659328e-11, "loss": 0.0046, "step": 39980 }, { "grad_norm": 0.15388669073581696, "learning_rate": 2.0675589512686443e-11, "loss": 0.0032, "step": 39990 }, { "grad_norm": 0.09898225963115692, "learning_rate": 1.7087265491433358e-13, "loss": 0.0033, "step": 40000 } ], "logging_steps": 10, "max_steps": 40000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "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": 48, "trial_name": null, "trial_params": null }